DEVELOPMENT OF FLUORESCENT PROTEIN TAGS FOR LIVE-CELL IMAGING By Elizabeth Marie Santos A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Chemistry – Doctor of Philosophy 2017 ABSTRACT DEVELOPMENT OF FLUORESCENT PROTEIN TAGS FOR LIVE-CELL IMAGING By Elizabeth Marie Santos Our primary goal is to develop fluorescent proteins that span the entire visible spectra, to be used when conventional fluorescent proteins are inadequate. In our system, fluorescence is activated upon coupling of the protein and ligand, such that temporal control can be achieved, whereas intrinsically fluorescent proteins are constitutively on. Additionally, our system does not require oxygen and can therefore find potential uses in obligate anaerobes. Our lab has demonstrated the ability to effectively control the absorption profile of conjugated polyenes. The initial aim of this project was to regulate the emissive properties of bound fluorophores with the same degree of control. This was achieved by the coupling of the solvatochromic fluorophore ThioFluor to hCRBPII mutants. ThioFluor yielded mutants with absorption maxima varying from 501 nm to 705 nm and emission maxima from 613 nm to 744 nm. This is equivalent to regulation over 204 nm in absorption and 131 nm in emission, covering both the red and far-red fluorescence wavelength regimes. Furthermore, we have shown its utility in live-cell imaging in whole cells, and with targeting to the nucleus and extranuclear space; fortuitously, negligible background fluorescence is apparent. In the course of optimizing binding of ThioFluor to hCRBPII, we discovered that it was observed that not only is the protonated Schiff base (PSB) fluorescent, but the Schiff base (SB) is as well. However, while PSB emission wavelength could be altered over 204 nm, SB emission remains nearly constant at 500 nm. Interestingly, select mutants displayed a far-red emission upon SB irradiation, similar to that obtained upon irradiation of the PSB, presumably through protonation in the excited state. This serendipitous discovery leads to more than a 200 nm Stokes shift with high quantum yield (> 60%). One such hCRBPII/ThioFluor complex displays ideal spectroscopic properties including fast iminium formation with a half-life of 1.7 min, low pKa of 5.3 (rendering almost complete SB formation at physiological pH), high quantum yield (0.51) and large Stokes shift (208 nm). This fluorescent protein was successfully used to visualize whole cell fluorescence, as well as targeting to the nucleus. The last major endeavor was to develop a protein-based pH sensor, with the ability to report pH values with high accuracy. We have previously reported an absorptive system that was capable of ratiometric sensing of pH. However, we have now developed a single protein fluorescent ratiometric pH sensor, based on the titration of an acidic residue near the iminium. Standard curves were generated based on the ratio of emissions at two excitation wavelengths, allowing for concentration independent sensing of pH. We have been able to demonstrate its applicability as an in vivo fluorescent pH probe, obtaining a pH value of 6.7 when hCRBPII is targeted to the nucleus of HeLa cells. Dedicated to my family and friends for their encouragement and support. iv ACKNOWLEDGMENTS Firstly, I am thankful to my undergraduate advisor and mentor Dr. Montserrat Rabago Smith. Without her guidance I would not have pursued an education and career in chemistry. I am very grateful to my PhD advisor, Professor Babak Borhan for instilling the knowledge needed to become a successful scientist, particularly in providing a new perspective of thinking. I also appreciate the freedom he allows in choosing and starting new projects, as well as the assistance provided in learning new techniques, and most importantly fixing instruments! Upon joining the lab, senior graduate students were pivotal in teaching me experimental techniques, how to design projects and analyze results. For this, I would also like to thank Dr. Tetyana Berbasova, Dr. Ipek Yapici and Dr. Arvind Jaganathan. For the aforementioned, I would also like to thank Dr. Chrysoula Vasileiou. Anytime I have needed assistance, she has been more than helpful. Many of these projects would not have been successful without collaborators, to which I would also like to express my gratitude. Special thanks to Alireza Ghanbarpour and Zahra Assar for the crystal structures, Hadi Gholami for the synthesis of chromophores and Wei Sheng for the measurement of quantum yields. Additionally, confocal imaging would not have been possible without the guidance of Dr. Melinda Frame. I want to extend my appreciation to my committee members Professors Jim Geiger, Xuefei Huang and Ned Jackson. v Lastly, I would like to thank my parents and siblings for their love and support. I have unfound gratitude to my late mom who always instilled upon me the importance of a good education. vi TABLE OF CONTENTS LIST OF TABLES ................................................................................................... x LIST OF FIGURES .............................................................................................. xiv LIST OF SCHEMES ......................................................................................... xxviii KEY TO ABBREVIATIONS ............................................................................... xxix CHAPTER I: METHODS EMPLOYED TO FLUORESCENTLY LABEL PROTEINS OF INTEREST ................................................................................... 1 I.1 Intrinsically fluorescent proteins ................................................................... 1 I.2 Extrinsically fluorescent proteins .................................................................. 8 I.2.1 Site-specific chemical labeling ......................................................... 11 I.2.2 Bioorthogonal reactions.................................................................... 20 I.2.3 Small protein tags............................................................................. 22 I.2.3.1 Fluorogen activating proteins ............................................. 23 I.2.3.2 Photoactive yellow protein .................................................. 25 I.2.3.3 In situ formation of a cyanine dye ....................................... 30 REFERENCES ................................................................................................ 33 CHAPTER II: MODULATION OF ABSORPTION AND EMISSION VIA A PROTEIN EMBEDDED SOLVATOCHROMIC FLUOROPHORE ...................... 56 II.1 Application of merocyanine aldehyde with hCRBPII mutants for in vivo imaging ............................................................................................................ 58 II.2 Principles of solvatochromicity .................................................................. 70 II.3 Choice of fluorophore ................................................................................ 74 II.4 Spectroscopic properties of ThioFluor ....................................................... 77 II.5 Removal of polarity induces a red-shift in absorption and emission.......... 81 II.6 Attempts to encapsulate the binding cavity ............................................... 89 II.7 Removal of water and extensive packing of the chromophore leads to an enhanced red-shift in absorption and emission ............................................... 97 II.7.1 Investigation of combinatorial residue effects on the wavelength of Q108K:K40L:T51V:T53S:R58W:Y19W .................................................. 105 II.8 Localization of charge leads to a blue-shift in absorption and emission.. 107 II.8.1 Design of blue-shifted hCRBPII/ThioFluor complexes that promote iminium formation ................................................................................... 111 II.8.2 Exploring the additive effects of acidic residues on the absorption and emission wavelength of hCRBPII/ThioFluor .................................... 117 II.9 Emission is linearly correlated to absorbance ......................................... 121 II.10 Visualization of hCRBPII/ThioFluor in bacteria ...................................... 124 vii II.11 Visualization of hCRBPII/ThioFluor in mammalian cells ........................ 126 II.11.1 Labeling of hCRBPII variants that have a propensity to oligomerize is unproductive ....................................................................................... 126 II.11.2 ThioFluor successfully labels hCRBPII in mammalian cells ........ 131 II.12 Conclusions ........................................................................................... 133 REFERENCES .............................................................................................. 135 CHAPTER III: STRUCTURE-PROPERTY RELATIONSHIPS OF THIOFLUOR ANALOGUES BOUND TO hCRBPII MUTANTS ............................................. 143 III.1 Prior work exploring substituent effects on absorption wavelength in fluorescent proteins ....................................................................................... 145 III.2 Structure of ThioFluor analogues to be explored.................................... 148 III.3 Evaluation of ThioFluor analogues with n-butyl amine in ethanol .......... 151 III.4 Coupling of ThioFluor analogues to Q108K:K40L:T51V:T53S:R58W:Y19W:L117E.............................................. 156 III.5 Coupling of ThioFluor analogues to Q108K:K40L:T51V:T53S:R58W:Y19W:A33W .............................................. 168 III.6 Coupling of ThioFluor analogues to Q108K:K40L:R58F ........................ 172 III.7 Coupling of ThioFluor analogues to Q108K:K40L:T51V:T53S:R58W .... 174 III.8 Conclusions and proposed projects ....................................................... 176 III.8.1 Proposed design of a multi-input fluorogenic probe ..................... 178 III.8.2 Development of near infrared and large Stokes shift fluorescent probes .................................................................................................... 184 APPENDIX .................................................................................................... 186 REFERENCES .............................................................................................. 193 CHAPTER IV: ENGINEERING OF hCRBPII/THIOFLUOR INTO A LARGE STOKES SHIFT FLUORESCENT PROTEIN ................................................... 202 IV.1 Preliminary observation of dual fluorescence in hCRBPII/ThioFluor...... 203 IV.2 Exploration of amino acid substitution at residue 40 on fluorescence ... 210 IV.3 Modification of active site residues to glutamic acid .............................. 222 IV.4 Effect of point mutations in Q108K:K40E:T53A:R58L:Q38F:Q4F on ESPT fluorescent properties .................................................................................... 227 IV.5 Energy transfer from tryptophan fluorescence to the bound ligand ....... 230 IV.6 Live cell imaging of large Stokes shift hCRBPII/ThioFluor complexes ... 235 IV.7 Excited state proton transfer is observed in ThioFluor analogues ......... 238 IV.8 Conclusions............................................................................................ 247 REFERENCES .............................................................................................. 248 CHAPTER V: DEVELOPMENT OF A SINGLE PROTEIN FLUORESCENT RATIOMETRIC PH SENSOR ........................................................................... 255 V.1 Initial observation of pH induced wavelength shifts in a ThioFluor/hCRBPII complex ......................................................................................................... 258 V.2 Optimization of hCRBPII for a ratiometric pH probe ............................... 265 viii V.3 L117D mutation leads to a much larger change in the iminium maxima . 272 V.4 Imaging of hCRBPII mutants displaying the pH induced wavelength shift with ThioFluor ................................................................................................ 282 V.5 ThioFluor analogues also show a pH induced wavelength shift ............. 288 V.6 Conclusions............................................................................................. 289 REFERENCES .............................................................................................. 291 CHAPTER VI: PROMOTING MONOMERIZATION OF hCRBPII MUTANTS .. 293 VI.1 Effect of T51 mutations on the monomer/dimer ratio during bacterial overexpression in hCRBPII mutants.............................................................. 295 VI.2 Spectroscopic properties of Q108K:K40L:T51X mutants....................... 297 VI.3 Imine isomerization of Q108K:K40L:T51X mutants ............................... 299 VI.4 Thermal imine isomerization that leads to a different absorption wavelength of the iminium ............................................................................. 302 VI.5 Different rates of iminium formation between monomer and dimer with retinal ............................................................................................................. 304 VI.6 Mutation of Y60 in the triple mutant Q108K:K40L:T51F ........................ 309 VI.7 Rational mutation to promote monomerization in hCRBPII mutants ...... 312 VI.8 Dimer formation in CRABPII, another iLBP ............................................ 313 VI.9 Conclusions............................................................................................ 315 REFERENCES .............................................................................................. 318 CHAPTER VII: MATERIALS AND METHODS ................................................ 321 VII.1 Site-directed mutagenesis of hCRBPII and CRABPII ........................... 321 VII.2 Protein expression and purification of hCRBPII and CRABPII in pET-17b expression plasmid ........................................................................................ 328 VII.3 Protein extinction coefficient determination ........................................... 331 VII.4 UV-Vis measurements of hCRBPII/chromophore complexes ............... 337 VII.4.1 pKa measurements of hCRBPII/chromophore complexes .......... 337 VII.4.2 Kinetic measurements of hCRBPII/chromophore PSB formation 338 VII.5 Fluorescence measurements ................................................................ 339 VII.6 Quantum yield measurements .............................................................. 340 VII.7 Live cell imaging in bacteria .................................................................. 340 VII.8 Cloning for mammalian expression vectors .......................................... 341 VII.8.1 General cloning protocol ............................................................. 341 VII.8.2 Preparation of plasmids for mammalian expression ................... 343 VII.8.3 Sequences of plasmids described in this thesis ......................... 349 VII.9 Mammalian cell culture ......................................................................... 355 VII.10 General confocal imaging methods..................................................... 356 VII.11 Synthesis of chromophores................................................................. 357 VII.11.1 Synthesis of ThioFluor .............................................................. 357 VII.11.2 Synthesis of ThioFluor-2 ........................................................... 360 REFERENCES .............................................................................................. 364 ix LIST OF TABLES Table II-1. Spectroscopic properties of hCRBPII mutants coupled with merocyanine retinal analog (MCRA) ................................................................... 60 Table II-2. Spectral properties of KSD-1-3 in solvents of different polarities. ..... 76 Table II-3. Spectroscopic characterization of ThioFluor in various solvents ..... 78 Table II-4. Spectroscopic properties of ThioFluor-PSB with n-butyl amine ....... 80 Table II-5. Spectroscopic change as a result of Q4F mutation ........................... 84 Table II-6. Spectroscopic change as a result of T51V mutation ......................... 85 Table II-7. Spectroscopic change as a result of mutating T53............................ 86 Table II-8. Spectroscopic change as a result of Q38F mutation ......................... 87 Table II-9. Spectroscopic change as a result of mutating R58 ........................... 88 Table II-10. Additive effects of T51 and T53 mutations ...................................... 89 Table II-11. Spectroscopic changes as a result of mutating R58 in Q108K:K40L:T51V:T53S to aromatic residues. .................................................. 89 Table II-12. Spectroscopic changes as a result of mutating R58 in Q108K:K40L:T51V:T53S .................................................................................... 94 Table II-13. Spectroscopic changes as a result of Y19W mutation .................... 98 Table II-14. Spectroscopic changes as a result of introducing bulky residues at the entrance of the binding cavity ..................................................................... 100 Table II-15. Spectroscopic changes a result of mutating active site residues to tryptophan ......................................................................................................... 102 Table II-16. Additive effect of F16Y, leading to a final bathochromic shift in wavelength ........................................................................................................ 105 Table II-17. Deducing the origins of the bathochromic shift in Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor ........................................... 106 x Table II-18. Spectroscopic changes observed upon the introduction of an acidic residue at L117 ................................................................................................. 110 Table II-19. Spectroscopic property of protein mutants designed in order to increase iminium formation ............................................................................... 116 Table II-20. Spectroscopic changes as a result of introducing one acidic residue near the iminium................................................................................................ 117 Table II-21. Spectroscopic changes as a result of introducing two acidic residues near the iminium................................................................................................ 120 Table II-22. Spectroscopic properties of Q108K:K40L:T53S:V62N/merocyanine aldehyde............................................................................................................ 128 Table III-1. Effect of Tyr66 replacement on spectroscopic properties of GFP .. 148 Table III-2. Spectroscopic properties of ThioFluor analogues (free aldehyde) in ethanol .............................................................................................................. 152 Table III-3. Spectroscopic properties of ThioFluor analogues (SB and PSB) in ethanol .............................................................................................................. 154 Table III-4. Spectroscopic properties of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W:Y19W:L117E ................................................. 159 Table III-5. Rates and half-lives of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W:Y19W:L117E ................................................. 166 Table III-6. Spectroscopic properties of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W:Y19W:A33W .................................................. 169 Table III-7. Rates and half-lives of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W:Y19W:A33W .................................................. 172 Table III-8. Spectroscopic properties of ThioFluor-PSB analogues coupled to Q108K:K40L:R58F ............................................................................................ 173 Table III-9. Spectroscopic properties of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W ....................................................................... 174 Table IV-1. Spectroscopic properties of large Stokes shift fluorescent proteins due to ESPT ...................................................................................................... 208 xi Table IV-2. Spectroscopic properties of the SB as a result of substitution at residue 40 in the hCRBPII protein Q108K:K40L:T53A:R58L:Q38F:Q4F .......... 213 Table IV-3. Spectroscopic properties of the PSB as a result of substitution at residue 40 in the hCRBPII protein Q108K:K40L:T53A:R58L:Q38F:Q4F .......... 217 Table IV-4. Protein expression yields of hCRBPII mutants with glutamic acid introduced at varying locations in Q108K:K40L:T53A:R58L:Q38F:Q4F ........... 223 Table IV-5. Spectroscopic properties of the SB as a result of glutamic acid introduction in the hCRBPII protein Q108K:K40L:T53A:R58L:Q38F:Q4F ........ 224 Table IV-6. Fluorescence lifetimes of hCRBPII/ThioFluor complexes upon 1060 nm excitation (three photon excitation) ............................................................. 226 Table IV-7. Protein expression yields as a result of point mutations in Q108K:K40E:T53A:R58L:Q38F:Q4F ................................................................ 228 Table IV-8. Spectroscopic properties of the SB as a result of point mutations in Q108K:K40E:T53A:R58L:Q38F:Q4F ................................................................ 229 Table IV-9. Spectroscopic properties of the PSB as a result of point mutations in Q108K:K40E:T53A:R58L:Q38F:Q4F ................................................................ 229 Table IV-10. Protein expression yields as a result of point mutations in Q108K:K40D:T53A:R58L:Q38F:Q4F ................................................................ 230 Table IV-11. Spectroscopic properties of the SB with tryptophan to phenylalanine mutations in Q108K:K40E:R58L:Q38F:Q4F ..................................................... 232 Table IV-12. Spectroscopic properties of ThioFluor analogues coupled with Q108K:K40E:T53A:R58L:Q38F:Q4F ................................................................ 245 Table V-1. Emission ratios as a function of pH for Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor................................ 262 Table V-2. Initial mutagenesis of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E for development toward a single fluorescent protein ratiometric sensor ........... 265 Table V-3. Effect of L117D versus L117E in the development toward a single fluorescent protein ratiometric sensor ............................................................... 272 xii Table V-4. Ratio of emissions obtained upon excitation at 514 nm and 633 nm for EGFP-Q108K:K40L:T51V:T53S:R58W:Y19W:L117E-3NLS/ThioFluor ..... 284 Table V-5. pH induced wavelength shift observed in Q108K:K40L:T51V:T53S:R58W:Y19W:L117E with ThioFluor analogues....... 289 Table VI-1. Effect of T51 mutation in KL:T51X:T53S:R58W:Y19W:F16Y ........ 293 Table VI-2. hCRBPII dimerization promoted by T53S, V62N, T51A and T51N mutations........................................................................................................... 294 Table VI-3. Effect of T51X mutations on monomer/dimer formation in Q108K:K40L ..................................................................................................... 296 Table VI-4. Spectroscopic properties of Q108K:K40L:T51X mutants bound to retinal ................................................................................................................ 298 Table VI-5. Thermal isomerization half-lives with substitution of small aliphatic residues at T51 ................................................................................................. 302 Table VI-6. Effect of Y60 mutation in Q108K:K40L:T51F on protein expression yields and rate of iminium formation ................................................................. 310 Table VI-7. Effect of T51 and T53 mutations on monomer/dimer ratio Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E ...................................................... 313 Table VII-1. PCR cycling conditions for site-directed mutagenesis .................. 321 Table VII-2. FPLC Source 15Q method ............................................................ 330 Table VII-2. FPLC SEC method........................................................................ 330 Table VII-3. Extinction coefficients of hCRBPII and CRABPII mutants ............ 332 Table VII-4. PCR cycling conditions for cloning ................................................ 342 xiii LIST OF FIGURES Figure I-1. a. Crystal structure of GFP. The chromophore is shown in green and is located in the center of the β-barrel. Coordinates were obtained from PDB 1EMA. b. Proposed scheme for the maturation of the GFP chromophore. Proper protein conformation and subsequent cyclization, oxidation and dehydration are required to form the fluorescent molecule ............................................................. 2 Figure I-2. Select fluorescent proteins and chromophore structures from each spectral class; the conjugated system of each chromophore is colored according to its emission ....................................................................................................... 4 Figure I-3. Equilibrium between the two ionization states of GFP – the neutral phenol species and the anionic phenolate species (top). UV-Vis of GFP at physiological conditions (solid line); excitation with light at each species yields essentially identical emission spectra (dotted line) ............................................... 5 Figure I-4. Water network responsible for excited state proton transfer in GFP .. 6 Figure I-5. The chemical structure of bilirubin and the crystal structure of bilirubin-bound wild-type UnaG. Coordinates were obtained from PDB 4I3B ....... 8 Figure I-6. a. Structure of biliverdin IXα (BV). b. Schematic of the bacterial phytochrome photoreceptor (BphP) PAS and GAF domains, showing the BV binding site and reactive cysteine residue. The PHY and effector domains are omitted for clarity. c. Crystal structure of BV-bound BphP with BV and the reactive cysteine (Cys12) shown in green. Coordinates were obtained from PDB 3C2W .................................................................................................................. 10 Figure I-7. Self-labeling protein tags: a. SNAP-tag. b. CLIP-tag. c. HaloTag. POI = protein of interest. The conjugated fluorophore is represented by a red star... 12 Figure I-8. Engineering a coumarin ligase. a. Two step probe targeting by ligation of an alkyl azide followed by ligation of a fluorophore-conjugated cyclooctyne. The fluorophore is represented by a red star. b. Direct coumarin ligation by an LpIA mutant. c. Cell-permeable coumarin analogue used for in vivo live-cell imaging. d. Coumarin labeling in mammalian cells. NES = nuclear export signal, CAAX = prenylation tag, NLS = nuclear localization signal. MAP2 = microtubule-associated protein 2. All scale bars are 10 μm ............................... 14 Figure I-9. Coumarin based fluorogenic probes for use in no-wash protein labeling. POI = protein of interest ........................................................................ 17 xiv Figure I-10. A fluorogenic probe for cell surface proteins. A Nile Red BG derivative targeting SNAP-tagged membrane receptor. The free probe is nonfluorescent, but upon reaction with SNAP-tag fluorescence is activated. The probe (2 μM) was incubated with Chinese hamster ovary (CHO) cells stably expressing SNAP-tagged HIR for 30 min at 37 °C and imaged without and with washing. All scale bars are 10 μm ...................................................................... 18 Figure I-11. Methods of site-specific chemical labeling. The site for fluorophore incorporation is represented by a red star........................................................... 22 Figure I-12. a. Chemical structure of the fluorogen malachite green. b. Live cell imaging of the expressed FAP-mCer3 localized to the cytosol, nucleus, mitochondria, peroxisome and endoplasmic reticulum (ER) of HEK cells. All scale bars are 10 μm .................................................................................................... 25 Figure I-13. Crystal structure of photoactive yellow protein (PYP) bound to 4hydroxycinnamic acid through Cys69. Chemical structures of 4-hydoxycinnamic acid and 7-hydroxycoumarin-3-carboxylic acid, which also bind PYP, are also shown. Coordinates were obtained from PDB 2PHY .......................................... 26 Figure I-14. a. Chemical structures of the developed cell-permeable probes for labeling PYP. b. Live cell imaging of MBP-PYP, PYP-NLS and PYP-EGFR with TMBDMA. The top row shows fluorescence obtained upon excitation at 473 nm, using a 490-590 nm emission filter. The controls for nontransfected cells are shown in the bottom panel. MBP = maltose binding protein, NLS = nuclear localization signal, EGFR = epidermal growth factor receptor. Scale bar = 10 μm. c. Time-lapse live cell imaging of PYP-NLS with TMBDMA (left). Scale bar = 5 μm and plot of average fluorescence intensity of TMBDMA against incubation time (right) ........................................................................................................... 28 Figure I-15. Scheme of Yellow Fluorescence-Activating and absorption-Shifting Tag (Y-FAST). Ligand HBR and HMBR are non-fluorescent, but become fluorescent and absorption red-shifts upon interaction with Y-FAST .................. 29 Figure I-16. Chemical structure of merocyanine retinal aldehyde (MCRA) and MCRA bound to an engineered CRABPII variant. Coordinates were obtained from PDB 3FEP. Fluorescence imaging of E. coli cells expressing CRABPII variant (left; overlay with brightfield on the right). Scale bar is 10 μm ................. 31 Figure II-1. Reaction and spectroscopic properties of merocyanine aldehyde (MCRA) with hCRBPII. Coordinates were obtained from PDB 3FEP ................. 58 Figure II-2. UV-Vis (blue curve) and emission spectra (red curve) of Q108K:K40L:T51V:R58F/MCRA ........................................................................ 61 xv Figure II-3. Labeling of EGFP-hCRBPIItetra in HeLa cells. All samples were incubated with 250 nM MCRA for 1 h at 37 °C, washed with PBS buffer two times and then incubated at 37 °C for the indicated amount of time (left to right: 0 min, 2 h, 4 h and 8 h). The right most panel is labeling of EGFP-hCRBPIItetraQ108L after incubating with 250 nM MCRA for 1 h at 37 °C, washed with PBS buffer two times and imaged immediately. All scale bars are 20 μM ................................... 62 Figure II-4. Overlay of CRABPII (R111K:R132L:Y134F:T54V:R59W, shown in magenta) and hCRBPII (Q108K:K40L:T51V:T53C:R58W:T29L:Y19W:Q4R, shown in green). Coordinates were obtained from PDB 4I9S and 4EEJ ............ 65 Figure II-5. Optimization of MCRA concentration for live cell imaging in HeLa cells expressing EGFP-hCRBPIInona; various concentrations (125 nM, 250 nM and 500 nM) were tested. All scale bars are 20 μM ........................................... 66 Figure II-6. Labeling of EGFP-hCRBPIInona in Hela cells. All samples were incubated with 250 nM MCRA for 30 sec at room temperature, washed with PBS buffer two times and imaged immediately. Localization signals 3NLS and NES targeted hCRBPII to the nucleus and extranuclear space, respectively. NLS = nuclear localization sequence, NES = nuclear export sequence. All scale bars are 20 μM .................................................................................................................. 67 Figure II-7. a. Simplified Jablonski diagram. b. Diagram of absorption and emission spectra; the difference in the band maxima is the Stokes shift ............ 70 Figure II-8. Examples of fluorescent solvatochromic dyes. The wavy lines indicate the common point for dye conjugation ................................................... 72 Figure II-9. Origins of solvatochromicity in fluorescence .................................... 73 Figure II-10. a. Structure of Betaine 30. b. Solutions of Betaine 30 in (1) methanol, (2) ethanol, (3) 1-octanol, (4) N,N-dimethylacetamide and (5) dichloromethane.................................................................................................. 74 Figure II-11. Fluorescence response of A-41S to HSA. a. 5 μM A41-S with HSA at 0, 1, 2, 3, 4, 5, 7.5, 10, 15, 20 and 30 μM. b. Structure of A41-S. c. Corresponding intensity changes are the ratio of intensity of A-41-S with/without HSA ..................................................................................................................... 75 Figure II-12. Structures of Dapoxyl dyes used to investigate the effect of the pilinker.................................................................................................................... 75 Figure II-13. Spectroscopic properties of ThioFluor in different solvents. a. UVVis and b. Fluorescence spectra of ThioFluor. ThioFluor under c. white light xvi and d. UV irradiation at 365 nm (with TLC handlamp). e. Stokes shift in different solvents versus the ET(30) value indicates that ThioFluor is solvatochromic .... 78 Figure II-14. Formation of the protonated Schiff base (PSB) of ThioFluor with nbutyl amine in ethanol ......................................................................................... 79 Figure II-15. Stokes shift in different solvents versus the ET(30) value indicates that ThioFluor-PSB is not solvatochromic. Different correlations are observed in protic and aprotic solvents .................................................................................. 81 Figure II-16. Polar residues mutated in order to remove polarity from the binding pocket. Coordinates obtained from PDB 4EXZ (hCRBPII-Q108K:K40L/retinal) . 82 Figure II-17. Water mediated hydrogen bonding between Q4 and the hCRBPII/retinal iminium. Coordinates obtained from PDB 4EFG (hCRBPIIQ108K:K40L:T51V:T53C:R58W:T29L:Y19W/retinal) ......................................... 83 Figure II-18. Water mediated hydrogen bonding between T51 and T53. Coordinates obtained from PDB 4EXZ (hCRBPII-Q108K:K40L/retinal) ............. 85 Figure II-19. Water mediated hydrogen bonding between Q38 and Q128. Coordinates obtained from PDB 4EXZ (hCRBPII-Q108K:K40L/retinal) ............. 87 Figure II-20. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor and space-filling representation. Electron density is shown at 1σ ............................. 90 Figure II-21. a. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor (cyan) overlaid with Q108K:K40L:T51V:T53S:R58Y/ThioFluor (green). b. Water mediated hydrogen bonding between R58Y and T29 ......................................... 92 Figure II-22. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor (cyan) overlaid with Q108K:K40L:T51V:T53S:R58Y/ThioFluor (green) and Q108K:K40L:T53A:R58F/ThioFluor (pink) ........................................................ 93 Figure II-23. a. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor, highlighting the water mediated network from Y19 to ThioFluor. b. Crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor shows this water network is abolished............................................................................................ 97 Figure II-24. Residues mutated (shown in green) in an attempt to encapsulate the binding cavity. Crystal structure is of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor ............................................. 99 xvii Figure II-25. Residues where tryptophan was introduced in an attempt to redshift wavelength are shown in green. Crystal structure is of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor ........................................... 101 Figure II-26. Crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E/ThioFluor. Mutation of I25 to tryptophan would sterically clash with A33W ................................................ 103 Figure II-27. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor. The water mediated hydrogen bonding between T53S and R58W is highlighted .... 107 Figure II-28. Residues chosen for introduction of an acidic residue, in order to interact with the iminium, based on the crystal structure of Q108K:K40L/retinal. Coordinates obtained from PDB 4EXZ ............................................................. 108 Figure II-29. Iminium formation of Q108K:K40L:T53A and Q108K:K40L:T53S monomers and dimers. From left to right are the UV-Vis of the thermodynamic protein/ThioFluor complex, iminium formation as a function of time for the monomer and dimer. Data is shown in black, while curve fitting is a dotted red line (fit to pseudo first order kinetics, measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7) ......................................................................... 111 Figure II-30. Iminium formation of KLA:R58L(R58F) and KLA:Q38A(M) monomers and dimers. From left to right are the UV-Vis of the thermodynamic protein/ThioFluor complex, iminium formation as a function of time for the monomer and dimer. Data is shown in black, while curve fitting is a dotted red line (fit to pseudo first order kinetics, measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7). KLA = Q108K:K40L:T53A .............................. 112 Figure II-31. Iminium formation of KLA:R58L:Q38F and KLA:R58L:Q38M monomer. From left to right are the UV-Vis of the thermodynamic protein/ThioFluor complex, iminium formation as a function of time. Data is shown in black, while curve fitting is a dotted red line (fit to pseudo first order kinetics, measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7). KLA = Q108K:K40L:T53A ............................................................................ 115 Figure II-32. Positions for introducing an acidic reside to blue-shift wavelength. Based on the crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor . 118 Figure II-33. a. Crystal structure of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor, highlighting the acidic residues tested in this series of protein mutants. b. – c. The residues L117 and K40 (shown in grey) were mutated to glutamic acid in Pymol........................... 119 xviii Figure II-34. Range of absorption and emission achieved in this study by coupling hCRBPII monomers with ThioFluor. Mutants from left to right are (1) Q108K:K40L:T51D:T53A:R58Y:Q38F:Q4F:L117E (note: data was smoothed by averaging because of low absorbance intensity), (2) Q108K:K40D:T53A:R58Y:Q38F:Q4F, (3) Q108K:K40L:T53A:R58L, (4) Q108K:K40L:T51V:T53S:R58G, (5) Q108K:K40L:T51V:T53S:R58W:Y19W, (6) Q108K:K40L:T51V:T53S:R58W:Y19W:F16Y, (7) Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Q4W ........................................ 122 Figure II-35. Plot of emission versus absorbance for the approximately 80 hCRBPII/ThioFluor complexes examined in this study yields a linear fit, with R2 = 0.93 ................................................................................................................... 122 Figure II-36. Spectroscopic properties of Q108K:K40L:T53S:V62N with ThioFluor including UV-Vis and fluorescence spectra, pKa, binding affinity, and rate plot. Note: It was later found that the hCRBPII species data shown here is for the dimer ........................................................................................................... 124 Figure II-37. Bacterial imaging of Q108K:K40L:T53S:V62N, labeled with ThioFluor.......................................................................................................... 125 Figure II-38. Labeling of HeLa cells expressing Q108K:K40L:T53S:V62N-ECFP with ThioFluor at various incubation times and concentrations of ThioFluor.. 126 Figure II-39. Drastically different spectroscopic properties are observed upon binding ThioFluor to Q108K:K40L:T53S:V62N monomer and dimer (for both identical conditions are used: 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7 and 23 °C). The spectra depict the course of iminium formation (in minutes) .. 127 Figure II-40. Labeling of HeLa cells expressing Q108K:K40L:T53S:V62N-ECFP with merocyanine aldehyde............................................................................... 129 Figure II-41. Labeling of U2OS cells expressing Q108K:K40L:T53S:V62N-ECFP with ThioFluor and merocyanine aldehyde ...................................................... 130 Figure II-42. Spectroscopic properties of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E with ThioFluor including UV-Vis and fluorescence spectra, pKa, and rate plot .................................................... 132 Figure II-43. Labeling of HeLa cells expressing hCRBPII-EGFP, hCRBPII-EGFP3NLS and hCRBPII-EGFP-NES with 10 μM ThioFluor (incubated at 37°C for 1 hour). NLS = nuclear localization sequence, NES = nuclear export signal ....... 133 xix Figure III-1. Proposed scheme for the maturation of the GFP and mCherry chromophores. Proper protein conformation and subsequent cyclization, oxidation and dehydration are required to form the fluorescent state ............... 147 Figure III-2. ThioFluor analogues investigated to study structure-property relationships in absorption and fluorescence. Extinction coefficients are for λmax of each chromophore ........................................................................................ 149 Figure III-3. Plot of Hammett values versus absorption wavelength for ThioFluor-5, ThioFluor-6, ThioFluor-4 and ThioFluor .................................. 153 Figure III-4. In the hCRBPII heptamutant Q108K:K40L:T51V:T35S:R58W:Y19W:L117E, the distance between the iminium nitrogen and L117E is 2.9 Å; it is this ionic interaction that presumably leads to the high pKa and holds the iminium as a trans isomer. Density is shown at 1σ 157 Figure III-5. Schematic representation of the geometric arrangement of the donor and acceptor groups upon excitation in planarized intramolecular charge transfer (PLICT) and twisted intramolecular charge transfer (TICT) .............................. 163 Figure III-6. Emission maxima versus absorption maxima for ThioFluor analogues bound to hCRBPII heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E ................................................. 165 Figure III-7. Double iminium pKas observed upon coupling of hCRBPII pentamutant Q108K:K40L:T51V:T53S:R58W with a. ThioFluor, b. ThioFluor-3 and c. ThioFluor-4. Absorption intensity of the iminium is plotted as a function of pH. Two curve fittings provide the two different pKa values .............................. 175 Figure III-8. Plots of Hammett value versus absorption wavelength, rate of iminium formation and pKa for ThioFluor, ThioFluor-4 and ThioFluor-6 coupled with the hCRBPII mutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E ......... 178 Figure III-9. Proposed fluorophore and reaction scheme for an improved fluorogenic probe .............................................................................................. 180 Figure III-10. Proposed chemical steps of the Pseudomonas sp. strain 4chlorobenzoyl-CoA dehalogenase-catalyzed conversion of 4-chlorobenzoyl-CoA to 4-hydroxybenzoyl-CoA .................................................................................. 182 Figure III-11. Crystal structure of 4-chlorobenzoyl coenzyme A dehalogenase from Pseudonomas sp. Catalytic residues are shown in cyan and aromatic residues responsible for hydrophobic packing are shown in green. Coordinates obtained from PDB 1NZY.................................................................................. 183 xx Figure III-12. Formation of ThioFluor-PSB with n-butyl amine in ethanol ....... 187 Figure III-13. Formation of ThioFluor-2-PSB with n-butyl amine in ethanol .... 187 Figure III-14. Formation of ThioFluor-3-PSB with n-butyl amine in ethanol .... 188 Figure III-15. Formation of ThioFluor-4-PSB with n-butyl amine in ethanol .... 188 Figure III-16. Formation of ThioFluor-5-PSB with n-butyl amine in ethanol .... 189 Figure III-17. Formation of ThioFluor-6-PSB with n-butyl amine in ethanol .... 189 Figure III-18. Formation of ThioFluor-7-PSB with n-butyl amine in ethanol .... 190 Figure III-19. Formation of ThioFluor-8-PSB with n-butyl amine in ethanol .... 190 Figure III-20. Formation of ThioFluor-9-PSB with n-butyl amine in ethanol .... 191 Figure III-21. Formation of ThioFluor-10-PSB with n-butyl amine in ethanol .. 191 Figure III-22. Formation of ThioFluor-11-PSB with n-butyl amine in ethanol .. 192 Figure IV-1. Dual fluorescence observed upon irradiation of the SB of Q108K:K40D:T53A:R58L:Q38F:Q4F bound with ThioFluor ............................ 203 Figure IV-2. Comparison of fluorescence spectra upon irradiation of the SB and PSB of Q108K:K40D:T53A:R58L:Q38F:Q4F/ThioFluor .................................. 204 Figure IV-3. UV-Vis and fluorescence spectra obtained as a result of changing pH in Q108K:K40D:T53A:R58L:Q38F:Q4F/ThioFluor ..................................... 205 Figure IV-4. UV-Vis (blue) and emission spectra (red) of Q108K:K40D:T53A:R58L:Q38F:Q4F-dimer/ThioFluor at pH 7.1 and pH 5.2 .. 206 Figure IV-5. Water network responsible for the excited state proton transfer in GFP ................................................................................................................... 207 Figure IV-6. UV-Vis, emission and excitation spectra of Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E/ThioFluor .................................... 209 Figure IV-7. Location of residue 40 in the crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor. L117D (shown in gray) was modeled in Pymol ............................................................................................................ 211 xxi Figure IV-8. Characterization of SB fluorescent properties of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor. a. UV-Vis and emission spectra after 397 nm excitation. b. Excitation spectrum at 605 nm. c. Emission spectra upon 280 nm excitation. d. Overlay of 280 nm emission spectrum and UV-Vis of the hCRBPII-SB showing spectral overlap. All data was collected at pH 7.2 ... 214 Figure IV-9. a. pKa titration of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor. b. SB emission as a function of pH ....................................................................... 215 Figure IV-10. Plot of fluorescence intensity at 605 nm versus time for binding of ThioFluor to Q108K:K40E:T53A:R58L:Q38F:Q4F at 23 °C fit to second order rate kinetics ....................................................................................................... 216 Figure IV-11. a. Comparison of fluorescence spectra upon irradiation of the SB and PSB of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor. b. PSB emission spectra as a function of pH ............................................................................... 218 Figure IV-12. Chains A, B and C of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor. Density is shown at 1σ in Pymol ............................................................................................................ 219 Figure IV-13. Hydrogen bonding network centered around K40E and T51, based on chain A of the crystal structure Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor. Some side chains are not shown for clarity .......................................................................................... 220 Figure IV-14. UV-Vis and emission spectra of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor SB (pH 7.1) and PSB (pH 4.5) ............................................................................................................. 221 Figure IV-15. Residues mutated to glutamic acid to test whether its location allows ESPT to occur; two views are shown. Based on the crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor. Mutation to glutamic acid of residues shown in grey did not lead to soluble protein expression.................................. 222 Figure IV-16. Crystal structure of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor (Chain A), showing the location of residues 40, 42, and 51 ................................................................... 225 Figure IV-17. Four tryptophan residues are present in wild type hCRBPII. Based on the crystal structure of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ ThioFluor.......................................................................................................... 231 xxii Figure IV-18. Emission spectra of Q108K:K40E:R58L:Q38F:Q4F and subsequent mutation of the four tryptophan residues in hCRBPII .................... 233 Figure IV-19. Live cell imaging of HeLa cells expressing EGFPQ108K:K40E:T53A:R58L:Q38F:Q4F and EGFPQ108K:K40E:T53A:R58L:Q38F:Q4F-3NLS, labeled with 10 μM ThioFluor for one minute at 37 °C. NLS = nuclear localization sequence .............................. 236 Figure IV-20. Non-transfected HeLa cells incubated with ThioFluor show background fluorescence when excited at 405 nm (LP 615 nm emission filter) 237 Figure IV-21. UV-Vis and emission spectra of ThioFluor, ThioFluor-SB and ThioFluor-PSB when bound to n-butyl amine .................................................. 238 Figure IV-22. UV-Vis and emission spectra of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-3 at pH 7.1 and pH 4.7 ......... 239 Figure IV-23. UV-Vis and emission spectra of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-4 at pH 7.0 and pH 4.5 ......... 240 Figure IV-24. UV-Vis and emission spectra of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-7 at pH 7.1............................ 241 Figure IV-25. UV-Vis, emission spectra and pH titration of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-8 ........................................... 242 Figure IV-26. UV-Vis, emission spectra and pH titration of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-9 ........................................... 243 Figure IV-27. UV-Vis, emission spectra and pH titration of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-10 ......................................... 244 Figure V-1. Two-protein-based ratiometric probe design. a. The selected proteins have distinct wavelength (UV-Vis) and pKa values. b. pH titration of the fused protein. (c) Ratiometric analysis of the fused protein system at two different concentrations ................................................................................................... 256 Figure V-2. Single protein ratiometric probe design. a. Titration of L117E leads to a shift in wavelength. b. Removal of L117E does not lead to the pH induced wavelength change. c. A blue-shift in wavelength results due to localizing charge on the iminium through interaction with the carboxylate, while protonation of the L117E side chain leads to less localized charge............................................... 257 xxiii Figure V-3. Acid-base titration of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor................................ 259 Figure V-4. Crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor highlighting the location of L117E. Density is shown at 1σ...................................................................... 260 Figure V-5. a. Crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor. Density is shown at 1σ. b. Overlay of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor (magenta) and Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (cyan)..................... 261 Figure V-6. Select standard curves obtained upon excitation of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor at 514/633 nm, 543/633 nm and 594/633 nm ............................................................................ 263 Figure V-7. In vitro characterization of Q108K:K40L:T51V:T53S:R58H:Y19W:L117E/ThioFluor as a single protein ratiometric pH sensor. a. Formation of PSB, monitored at 539 nm via UV-Vis, fit to second order rate kinetics. b. UV-Vis titration of the hCRBPII/ThioFluor complex. c. Double pKa curve obtained by plotting the absorbance intensity at 598 nm as a function of pH. d. Standard curve generated by plotting the emission ratio of 514/633 nm versus pH .......................................................................... 266 Figure V-8. Overlay of Q108K:K40L:T51V:T53:R58W:Y19W:L117E/ThioFluor (green) and Q108K:K40L:T51V:T53:R58H:Y19W:L117E/ThioFluor (magenta), highlighting R58 and L117E .............................................................................. 267 Figure V-9. Overlay of Q108K:K40L:T51V:T53:R58W:Y19W:L117E/ThioFluor (green) and Q108K:K40L:T51V:T53:R58W:Y19W:L117E:A33W/ThioFluor (magenta) .......................................................................................................... 268 Figure V-10. In vitro characterization of Q108K:K40L:T51V:T53S:R58W:Y19W:117E:A33W/ThioFluor as a single protein ratiometric pH sensor a. UV-Vis titration of the hCRBPII/ThioFluor complex. b. pKa curve obtained by plotting the absorbance intensity at 628 nm as a function of pH. c. pKa curve obtained by plotting the absorbance intensity at 539 nm as a function of pH. d. Formation of PSB, monitored at 628 nm via UVVis, fit to second order rate kinetics .................................................................. 269 Figure V-11. In vitro characterization of Q108K:K40L:T51V:T53S:R58H:Y19W:L117E:A33W/ThioFluor as a single protein ratiometric pH sensor. a. Formation of PSB, monitored at 585 nm via UVVis, fit to second order rate kinetics. b. UV-Vis titration of the hCRBPII/ThioFluor xxiv complex. c. pKa curve obtained by plotting the absorbance intensity at 614 nm as a function of pH. d. pKa curve obtained by plotting the absorbance intensity at 547 nm as a function of pH ............................................................................... 271 Figure V-12. In vitro characterization of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor as a single protein ratiometric pH sensor. a. UV-Vis titration of the hCRBPII/ThioFluor complex. b. pKa curve obtained by plotting the absorbance intensity at 618 nm as a function of pH. c. pKa curve obtained by plotting the absorbance intensity at 525 nm as a function of pH. d. Standard curve generated by plotting the emission ratio at 514/633 nm versus pH ...................................................................................... 273 Figure V-13. Overlay of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (green) and Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor (magenta) .............. 274 Figure V-14. In vitro characterization of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F/ThioFluor as a single protein ratiometric pH sensor. a. UV-Vis titration of the hCRBPII/ThioFluor complex. b. pKa curve obtained by plotting the absorbance intensity at 623 nm as a function of pH. c. pKa curve obtained by plotting the absorbance intensity at 539 nm as a function of pH. d. Standard curves generated by plotting the emission ratio of 514/633 nm, 543/633 nm and 594/633 nm versus pH ...................................... 276 Figure V-15. Overlay of the crystal structures Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor (magenta) and Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F/ThioFluor (cyan) ............ 277 Figure V-16. Overlay of crystal structures of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (green) and Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F/ThioFluor (cyan) ............ 278 Figure V-17. a. Acidification of Q108K:K40L:T51V:T53S:R58H:Y19W:L117D:Q4F/ThioFluor complex. b. Basification of Q108K:K40L:T51V:T53S:R58H:Y19W:L117D:Q4F/ThioFluor complex ............................................................................................................. 280 Figure V-18. Titration of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:A33W:Q4F/ThioFluor complex and pKa obtained by plotting absorbance intensity versus pH at 649 nm ......... 281 Figure V-19. Imaging of HeLa cells expressing EGFPQ108K:K40L:T51V:T53S:R58W:Y19W:L117E-3NLS, labeled with ThioFluor for one hour at 37 °C .............................................................................................. 283 xxv Figure V-20. Optimized imaging of HeLa cells expressing EGFPQ108K:K40L:T51V:T53S:R58W:Y19W:L117E-3NLS upon excitation at 514 nm and 633 nm ....................................................................................................... 283 Figure V-21. Linear fit to the working range of the in vitro generated standard curve for Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor................. 285 Figure V-22. Imaging of HeLa cells expressing Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F3NLS, labeled with ThioFluor for ten minutes at 37 °C ................................................................... 287 Figure VI-1. Formation of domain-swapped dimer in hCRBPIIQ108K:K40L:Y60W. Coordinates obtained from PDB 4ZR2 (dimer) and 4ZJ0 (monomer)......................................................................................................... 293 Figure VI-2. Location of T51 in Q108K:K40L/retinal. Coordinates were obtained from PDB 4EXZ................................................................................................. 295 Figure VI-3. Cycle for imine isomerization in retinal bound hCRBPII and CRABPII variants ............................................................................................................. 300 Figure VI-4. a. UV-Vis of Q108K:K40L-monomer/retinal. b. Absorbance intensity at SB and PSB as a function of time. c. UV irradiation and subsequent isomerization of the imine. d. Rate (pseudo-first order) of imine isomerization after UV irradiation ............................................................................................ 301 Figure VI-5. a. UV-Vis of Q108K:K40L:T51N-monomer/retinal. b. Normalized absorption intensity versus time indicates that protein unfolding occurs after iminium formation .............................................................................................. 303 Figure VI-6. Crystal structure of Q108K:K40L:T51N-dimer/retinal ................... 304 Figure VI-7. UV-Vis of Q108K:K40L:T51F-monomer/retinal and Q108K:K40L:T51F-dimer/retinal ....................................................................... 305 Figure VI-8. Overlay of two chains in the crystal structure of Q108K:K40L:T51Fdimer/retinal ...................................................................................................... 306 Figure VI-9. Overlay of three chains in the crystal structure of Q108K:K40L:T51F-monomer/retinal ................................................................. 307 Figure VI-10. Overlay of the crystal structures of Q108K:K40L:T51Fmonomer/retinal (green) and Q108K:K40L:T51F-dimer/retinal (purple) ........... 307 xxvi Figure VI-11. Overlay of the crystal structures of Q108K:K40L:T51F-dimer/retinal (purple) and apo Q108K:K40L:T51F-dimer (green) .......................................... 308 Figure VI-12. UV-Vis scanning of Q108K:K40L:T51F dimer and monomer over time with retinal ................................................................................................. 309 Figure VI-13. Absorbance intensity of PSB and SB as a function of time, followed by UV-Vis .......................................................................................................... 311 Figure VI-14. Crystal structure of Q108K:K40L:T51F:Y60A-dimer/retinal ....... 312 Figure VI-15. Size exclusion chromatography trace and denaturing protein gel for each fraction obtained from expression of I52A ................................................ 315 Figure VI-16. Residues shown to affect the formation of domain swapped dimer in hCRBPII. Coordinates obtained from 4ZH9 .................................................. 316 xxvii LIST OF SCHEMES Scheme II-1. Synthesis of ThioFluor ................................................................. 77 Scheme VII-1. Detailed synthesis of ThioFluor ............................................... 358 Scheme VII-2. Detailed synthesis of ThioFluor-2............................................ 361 xxviii KEY TO ABBREVIATIONS A alanine, Ala C cysteine, Cys D aspartic acid, Asp E glutamic acid, Glu F phenylalanine, Phe G glycine, Gly H histidine, His I isoleucine, Ile K lysine, Lys L leucine, Leu M methionine, Met N asparagine, Asn P proline, Pro Q glutamine, Gln R arginine, Arg S serine, Ser T threonine, Thr V valine, Val W tryptophan, Trp Y tyrosine, Tyr xxix Å angstrom Da dalton nm nanometer μm micrometer cm centimeter M molar μM micromolar nM nanomolar mmol millimole ε extinction coefficient h hour min minute sec second t time t1/2 maturation half-life Arch Archaerhodopsin-3 BC O2-benzylcytosine Betaine 30 4-(2,4,6-triphenylpyridinium)-2,6-diphenyloxide BFP blue fluorescent protein BG O6-benzylguanine BP band pass BphP Bacterial phytochromes photoreceptor xxx BV biliverdin IXα CFP cyan fluorescent protein CHO Chinese hamster ovary CRABPII cellular retinoic acid binding protein II CuAAC CuI-catalyzed alkyne-azide cycloaddition DMEM Dulbecco’s modified eagle medium DMSO dimethyl sulfoxide DPBS Dulbecco’s phosphate buffered saline eDHFR Escherichia coli dihydrofolate reductase EGFP enhanced GFP EGFR epidermal growth factor receptor equiv equivalent ER emission ratio ESPT excited state proton transfer FABP fatty acid binding protein FAP fluorogen-activating proteins FBS fetal bovine serum FlAsH fluorescein arsenical hairpin binder FP fluorescent protein FPLC fast protein liquid chromatography FRET Förster resonance energy transfer FRFP far-red fluorescent protein xxxi GAF cGMP phosphodiesterase/adenylyl cyclase/FhlA GFP green fluorescent protein hAGT human O6-alkylguanine-DNA alkyltransferase HBR 4-hydroxybenzylidene-rhodanine hCRBPII human Cellular Retinol Binding Protein II hCRBPIInona Q108K:K40L:T51V:T53C:R58W:T29L:A33W:Q4F:L117E hCRBPIItetra Q108K:K40L:T51V:R58F HIR human insulin receptor HMBR 4-hydroxy-3-methylbenzylidene-rhodanine HOMO highest occupied molecular orbital HSA human serum albumin HWE Horner-Wadsworth-Emmons ICT intramolecular charge transfer IDT Integrated DNA Technologies iLBP intracellular lipid binding protein IPTG isopropyl-1-thio-D-galactopyranoside JRA julolidine retinal analog KL Q108K:K40L KLA Q108K:K40L:T53A KLF Q108K:K40L:R58F LAP LplA acceptor peptide LB Luria broth xxxii LP long pass LpIA lipoic acid ligase LSS large Stokes shift LUMO lowest unoccupied molecular orbital MAP2 microtubule-associated protein 2 mBeRFP monomer blue light-excited RFP MBP maltose binding protein MCRA merocyanine retinal aldehyde NBD nitrobenzoxadiazole nd not determined NES nuclear export sequence NIR near-infrared NLS nuclear localization sequence OD optical density OFP orange fluorescent protein p-HBI 4-(p-hydroxy-benzylidene)-5-imidazolinone PAFP photoactivatable FP PALM photoactivated localization microscopy PAS period/ARNT/single-minded PBS phosphate buffered saline PCR polymerase chain reaction PDB protein data bank xxxiii PeT photoinduced electron transfer PHY phytochrome PLICT planarized intramolecular charge transfer PSB protonated Schiff base PSFP photoswitchable FP PSG penicillin-streptomycin-glutamine PYP photoactive yellow protein QY quantum yield (Φ) ReAsH resofurin arsenical hairpin binder RESOLFT reversibly saturatable optical fluorescence transitions RFP red fluorescent protein rsFP reversibly photoswitchable FP SB Schiff base scFv human single-chain antibodies SEC size-exclusion chromatography SNAr nucleophilic aromatic substitution SPAAC strain-promoted alkyne-azide cycloaddition SQ Source 15Q TB terrific broth TBET through-bond energy transfer THF tetrahydrofuran TICT twisted intramolecular charge transfer xxxiv TMP trimethoprim UV ultraviolet Vis visible wtGFP wild-type green fluorescent protein Y-FAST Yellow Fluorescence-Activating and absorption-Shifting Tag YFP yellow fluorescent protein xxxv CHAPTER I: METHODS EMPLOYED TO FLUORESCENTLY LABEL PROTEINS OF INTEREST Fluorescence imaging is a powerful tool, typically used to analyze subcellular localization and protein dynamics.1-12 This chapter serves to review the most commonly employed tools for fluorescence imaging to date. I.1 Intrinsically fluorescent proteins The most prevalent imaging tools are fluorescent proteins (FPs), which can be genetically encoded to allow for visualization of a protein of interest; fluorescent proteins have revolutionized the field of cell biology due to their ease of use and the wide array of colors that have been developed.13-23 Green fluorescent protein (GFP) is perhaps the canonical example of fluorescent proteins, with the Nobel Prize in Chemistry in 2008 being awarded to Osamu Shimomura, Martin Chalfie and Roger Tsien for the discovery and development of the Aequorea victoria GFP. GFP (27 kDa, containing 238 amino acids) was originally isolated from the jellyfish Aequorea victoria in the early 1960s by Shimomura.24, 25 It was not until 1992 that Douglas Prasher was successful in cloning the gene encoding GFP.26, 27 Soon after, Chalfie was able to express recombinant GFP in C. elegans.28, 29 Subsequently, Tsien led the effort of protein engineering for the development of innumerable FP technologies. The results of many studies have led to optimized folding time, protein stability and photostability. Nonetheless, all conventional fluorescent proteins share the same fold and have similar mechanisms of fluorophore formation. 1 Subsequent to protein folding, the fluorophore is formed auto-catalytically from three amino acids contained in the beta barrel of the protein.18, 21, 23, 30-33 Crystal structures of GFP show an 11-stranded beta barrel structure with a single α-helix at the center of the barrel; it is along this central helix, in the protein’s hydrophobic core, that the 4-(p-hydroxy-benzylidene)-5-imidazolinone (p-HBI) chromophore forms (Figure I-1a).34-36 In GFP, the p-HBI chromophore is formed via the cyclization of Ser65-Tyr66-Gly67. The most widely accepted mechanism (Figure I-1b) begins with attack of the Gly67 amide nitrogen to the carbonyl carbon of Ser65.37-39 Subsequently, dehydration yields the imidazolin-5one intermediate. Lastly, oxygen catalyzes the dehydrogenation of the Cα-Cβ bond of Tyr66, yielding the fully conjugated ring structure. It is important to note that upon denaturation, GFP becomes nonfluorescent, but upon refolding it regains its fluorescence.19, 40 a. b. Tyr66 HO N H O H N N H O N H O Folding HN HO O N H OH Gly67 O O Ser65 OH Cyclization O O N N HO N H O OH N Oxidation (O2) -H2O HN HO N H OH O OH Figure I-1. a. Crystal structure of GFP. The chromophore is shown in green and is located in the center of the β-barrel. Coordinates were obtained from PDB 1EMA. b. Proposed scheme for the maturation of the GFP chromophore. Proper protein conformation and subsequent cyclization, oxidation and dehydration are required to form the fluorescent molecule. 2 Since its discovery nearly 60 years ago, GFP has been one of the most widely employed proteins in cell biological applications. In the past half a century, GFP has led to the development of enhanced genetic variants of different colors, spanning nearly the entire visible spectrum (Figure I-2).13, 15, 16, 33 Interestingly, GFP-like proteins have been found in numerous animal species, including jellyfishes and coral polyps,41-44 comb jellies,45 crustaceans, and lancelets.46, 47 FPs are typically categorized based upon their emission maxima: blue (BFPs: 440-470 nm), cyan (CFPs: 471-500 nm), green (GFPs: 501-520 nm), yellow (YFPs: 521-550 nm), orange (OFPs: 551-575 nm); red (RFPs: 576-610 nm) and far-red (FRFPs: 611-660 nm).13, 15, 16, 18, 33 The optical properties of the mature chromophore depend on its specific molecular structure, which results from chemical transformations of the tripeptide maturation (including oxidation, dehydration, cyclization or protonation) and also on interactions of surrounding residues on the polypeptide moiety, such as π-stacking or hydrogen bonding. In RFPs, the conjugation of the π -system in p-HBI is extended;18, 21, 30 for example in mCherry, the chromophore’s conjugation is extended with an acylimine substituent as a result of backbone oxidation (Figure I-2).33, 48 Red FPs are of particular interest because the red emission channel has lower signal from cellular autofluorescence.30, 33, 48-53 Furthermore, longer wavelengths allow for great tissue penetration and cause less damage to proteins and DNA. The majority of RFPs were isolated from Anthozoa species.4144 It should be noted that in their natural form, anthozoan FPs often form tightly 3 associated tetramers,41 which has led many researchers to undertake the feat of designing monomeric variants. O N NH N N EBFP λex = 383 nm λem = 445 nm N OH O EGFP λex = 484 nm λem = 507 nm O N mOrange λex = 548 nm λem = 562 nm O O O N N N N rry spbe mRa mGr ape1 erry mCh rry mStr awbe e gerin mTan ato tdTom mOr ange nana mBa ew neyd mHo Citrin e ECFP λex = 433 nm λem = 475 nm EGF P EC F P EBF P N H Citrine λex = 516 nm λem = 529 nm O O mCherry λex = 587 nm λem = 610 nm O N N R mGr ape2 N O N m N N O N mPlu O Figure I-2. Select fluorescent proteins and chromophore structures from each spectral class; the conjugated system of each chromophore is colored according to its emission. In general, the protein microenvironment dictates the fluorescence characteristics of the mature chromophore. In wild-type GFP, the p-HBI species experiences two ionization states of Tyr66 – an anionic phenolate species and a neutral, protonated species. The neutral species predominates at physiological conditions, maximally absorbing at 396 nm, while the phenolate is red-shifted to 475 nm (Figure I-3).32, 34 While the absorbances are drastically different, the two species emit at virtually identical wavelengths, 508 and 503 nm, respectively. This is due to an excited state proton transfer (ESPT), occurring upon excitation 4 at 400 nm, 54-57 which decreases the pKa of the tyrosine moiety of the chromophore by several units.58 The probable mechanism is that proton transfer occurs via the hydrogen bonds of a buried water and Ser205 to Glu222. Meanwhile the side chain of Thr203 rotates to solvate and stabilize the phenolate oxyanion (Figure I-4).56, 59 O O HO N H O N O OH R-OH Normalized absorption N H+ R-O N H O OH Normalized fluorescence N N Wavelength (nm) Figure I-3. Equilibrium between the two ionization states of GFP – the neutral phenol species and the anionic phenolate species (top). UV-Vis of GFP at physiological conditions (solid line); excitation with light at each species yields essentially identical emission spectra (dotted line). The most commonly used mutation to cause ionization of the phenol of the chromophore is replacement of Ser 65 by Thr.60, 61 In S65T, the wild-type 396 nm excitation peak due to the neutral phenol is suppressed and the 475 nm peak due to the anion is enhanced five- to six-fold in amplitude. As previously stated, S65 donates a hydrogen bond to the buried side chain of Glu222 to allow 5 ionization of the carboxylic acid. Threonine is too large to adopt the correct conformation in the crowded interior of the protein, forcing the carboxyl group of Glu222 to remain neutral. Consequently, the S65T mutation, along with a F64L substitution designed to improve protein folding efficiency at 37 °C, has been incorporated into an enhanced version of GFP (EGFP) and most of its color variants. Thr203 O OH N R O N N O H His148 H O H O O λabs = 396 nm λem = 508 nm N R O H NH O Thr203 H R O N Ser65 N R N H His148 H O H O H O H H O Glu222 Ser205 λabs = 475 nm λem = 503 nm O H Ser65 O Glu222 Ser205 Figure I-4. Water network responsible for excited state proton transfer in GFP. It was later shown that ESPT is also responsible for numerous large Stokes shifts (LSS) in red FPs, such as mKeima,62-64 LssmOrange,65-69 LSSmKates,70-75 and mBeRFP.76 A large Stokes shift includes a difference between the excitation and emission maxima larger than 100 nm. As an example, LSSmKate-1 and LSSmKate-2 (evolved via rational mutagenesis of monomeric mKate) possess large Stokes shifts with excitation and emission at 463/624 nm and 460/605 nm, respectively.71, 72 The only disadvantage compared to EGFP is that their quantum yields are smaller, 0.08 and 0.17, respectively. However, these probes have been utilized simultaneously with ECFP for multicolor imaging of tumor cell migration in living mice.69 6 The broad array of variants that have been developed makes fluorescent proteins useful in a variety of applications to study the organization and function of living systems. Fluorescent proteins have become well established as in vivo imaging tools to study subcellular localization, visualize protein-protein interactions, and monitor gene expression, among other applications.4, 6, 16, 22, 32, 51 Additionally, as microscopy has become more sophisticated, new FPs that are not constitutively fluorescent (as opposed to conventional FPs which are permanently fluorescent) continue to be developed. These include irreversibly photoactivatable and photoswitchable FPs which require induction by light.6, 77, 78 Members of the first group exhibit an irreversible photoconversion from the nonfluorescent or green fluorescent state to the red fluorescent state; these are termed photoactivatable FPs (PAFPs) and photoswitchable FPs (PSFPs), respectively. Reversibly photoswitchable FPs (rsFPs) can be repeatedly photoswitched, assuming they are fatigue resistant, between fluorescent and nonfluorescent states. These systems have found utility in super-resolution techniques (techniques developed to overcome the inherent diffraction limit of light), such as reversibly saturatable optical fluorescence transition (RESOLFT) imaging79-81 and photoactivated localization microscopy (PALM).82, 83 While fluorescent proteins have been employed in innumerable applications, they still have their limitations. GFP is advantageously monomeric, which allows for it to be fused to a wide variety of partners, without experiencing pitfalls such as aggregation. However, the tendency of FPs to form dimers, 7 tetramers or other oligomers is problematic when the FP is to be fused to another protein.16, 43, 84 Additionally, oxygen is required to produce the mature chromophore, which limits the use of fluorescent proteins in obligate anaerobes. Thus, even after decades of research, there is still no fluorescent protein that can satisfy the necessity of all applications. I.2 Extrinsically fluorescent proteins Fluorescent proteins are intrinsically fluorescent; they become fluorescent after folding, rather than upon the addition of exogenous ligand. As the limitations of intrinsically fluorescent proteins become apparent, focus has been turned toward the development of fluorescent proteins that bind endogenous biomolecules or require the addition of exogenous ligand. Extrinsically fluorescent proteins that utilize endogenous biomolecules as fluorophores include UnaG as a green fluorescent protein and phytochromes as near infrared proteins. O NH HO O HN O OH O NH NH Bilirubin Figure I-5. The chemical structure of bilirubin and the crystal structure of bilirubin-bound wild-type UnaG. Coordinates were obtained from PDB 4I3B. UnaG is a protein with no intrinsic fluorescence, first isolated from muscle of the Japenese eel Unagi.85 Studies showed the isolated protein to be a 8 monomer with a molecular mass of approximately 17 kDa. UnaG belongs to the fatty acid binding protein (FABP) family. Its fluorescence is triggered by bilirubin (Figure I-5),86 a membrane-permeant heme metabolite, which unlike GFP occurs in an oxygen independent manner. Upon binding (non-covalently), green fluorescence at 527 is observed, with a quantum yield of 0.51. Bacterial phytochromes photoreceptors (BphPs), which bind biliverdin IXα (BV) (Figure I-6a) and fluoresce weakly in the near-infrared, have certain advantages over phytochromes from plants and cyanobacteria, particularly because they bind biliverdin, which is abundant in eukaryotic cells.87, 88 This and the fact that BphPs exhibit red-shifted NIR absorbance and fluorescence relative to other phytochromoes, renders them more practical toward development of probes for imaging. BphPs bind BV (Figure I-6c) in the GAF domain and subsequently a thioether bond is formed with a conserved cysteine in the PAS domain (Figure I-6b). Deletion of the PHY and effector domains from Deinococcus radiodurans and Rhodopseudomonas palustris BphPs, as well as extensive mutagenesis of the PAS and GAF domains yield near-infrared FPs, including IFP1.4, Wi-Phy and RFP, all of which have emission at approximately 700-720 nm, with quantum yields less than 0.10.89, 90 They have been used for whole-body imaging, and imaging of livers and tumors in mice.91 These engineered BphPs have also served as templates for the development of other novel fluorescent proteins. 9 a. b. site of Cys attack O O O O PAS NH O N H GAF HN N H Biliverdin IXα (BV) SH O binding site of BV c. Figure I-6. a. Structure of biliverdin IXα (BV). b. Schematic of the bacterial phytochrome photoreceptor (BphP) PAS and GAF domains, showing the BV binding site and reactive cysteine residue. The PHY and effector domains are omitted for clarity. c. Crystal structure of BV-bound BphP with BV and the reactive cysteine (Cys12) shown in green. Coordinates were obtained from PDB 3C2W. Recently, it was also shown that BV binds to sandercyanin; the blue colored protein ligand complex was isolated from the skin mucus of walleye.92 However, sandercyanin differs from previously reported fluorescent BV-binding BphPs in that the chromophore does not appear to be covalently bound via a cysteine residue. BV-bound sandercyanin shows absorbance maxima at 280, 375 and 630 nm at physiological pH, and displays 675 nm red fluorescence when excited at either 375 or 630 nm, with a quantum yield of 0.016. 10 In terms of utility, these extrinsically fluorescent proteins that bind endogenous biomolecules are (in experimental terms) used similarly to intrinsically fluorescent proteins; however, the endogenous concentration of fluorophore has to be considered, requiring the need to supplement the cells with additional ligand. I.2.1 Site-specific chemical labeling Because of the aforementioned limitations of fluorescent proteins, in addition to their large size (~30 kDa), a large effort has been devoted to developing alternatives to fluorescent proteins. Perhaps the largest area of research has been focused on chemical-based methods, particularly in the development of extrinsically fluorescent proteins that become fluorescent upon the addition of an exogenous ligand to an apoprotein. The most developed alternative to fluorescent proteins is site-specific chemical labeling systems. These methods incorporate a tag fused with a protein of interest and a fluorescent small-molecule; the tags covalently react with a small-molecule substrate containing a fluorophore. One advantage of this chemical recognition mediated labeling approach is that a wide variety of molecules, including NIR fluorescent dyes, can be incorporated into a probe by replacement of the fluorophore moiety of the probe. The most widely used tags are SNAP-tag, CLIP-tag, HaloTag, and tetracysteine tags, to name a few. The SNAP-tag is a self-labeling protein tag that is mutated from the human suicide protein O6-alkylguanine-DNA alkyl- 11 transferase (hAGT), engineered to specifically bind O6-benzylguanine (BG) derivatives (Figure I-7a).93-97 A second generation CLIP-tag, reacts irreversibly with O2-benzylcytosine (BC) derivatives.93, 98 Together SNAP-tag and CLIP-tag systems have been exploited for simultaneous labeling with different probes.99 a. SNAP-Tag S POI POI S O N N N H b. N O NH2 N NH N H eDHFR N NH2 O S POI POI NH S OR O NH O N O NH2 N O NH2 c. HaloTag O POI O O H N O O POI O O Cl Cl O NH Figure I-7. Self-labeling protein tags: a. SNAP-tag. b. CLIP-tag. c. HaloTag. POI = protein of interest. The conjugated fluorophore is represented by a red star. 12 The Cornish lab introduced a similar self-labeling tag. 2,4-Diamino-5(3,4,5-trimethoxybenzyl)pyrimidine (trimethoprim or TMP) tag uses an engineered 18 kDa, monomeric Escherichia coli dihydrofolate reductase (eDHFR) to react with trimethoprim-fused fluorophores (Figure I-7b).100, 101 TMPfluorophore conjugates were shown to label eDHFR fusion proteins (with nM binding affinity) in mammalian cells with low background and fast kinetics.100, 102104 HaloTag (33 kDa) is derived from the bacterial haloalkane dehalogenase enzyme from Rhodococcus rhodochrous. A covalent bond is formed between the HaloTag protein and fluorophore-conjugated alkyl halide ligand (Figure I-7c).105 Interaction of the enzyme and ligand generates an alkyl-enzyme intermediate via nucleophilic displacement of the alkyl chloride with Asp106 (it should be noted that a H272P mutation is included to prevent hydrolysis and release of the enzyme).106 As opposed to the self-labeling proteins discussed above, short peptide sequences have also been developed which prove advantageous because of their small targeting motif. Fluorescein arsenical hairpin binder (FlAsH)107 and the more red-shifted analog resofurin arsenical hairpin binder (ReAsH)108 are small biarsenical dyes that covalently react with a short peptide sequence containing four cysteine residues (via metal-ligand recognition). In these, the peptide sequence is Cys-Cys-Xaa-Xaa-Cys-Cys, where Xaa is any amino acid other than cysteine. As in the other systems, the sequence is genetically introduced into the 13 sequence of the target protein. Tsien and coworkers first described the use of biarsenical reagents for site-specific protein labeling in live cells.107, 109 However, similar tags have also been developed based on recognition for histidine,110-113 aspartate,114, 115 and serine116 among other recognition moieties. a. O LpIA, ATP NH3 HN HN 7 LAP O O engineered LpIA, ATP NH3 HN O O 4 c. O O O OH O O OH H N 4 LAP O H N HO O 7 LAP 7 b. LAP N N N N3 HO O O N3 O O O O O H N 4 cellular esterases O O HO O O O OH H N 4 O coumarin-AM2 LAP2-NES LAP2-CAAX LAP2-NLS LAP2-β-actin Vimentin-LAP2 LAP2-MAP2 Coumarin + DIC Coumarin d. Figure I-8. Engineering a coumarin ligase. a. Two step probe targeting by ligation of an alkyl azide followed by ligation of a fluorophore-conjugated cyclooctyne. The fluorophore is represented by a red star. b. Direct coumarin ligation by an LpIA mutant. c. Cell-permeable coumarin analogue used for in vivo live-cell imaging. d. Coumarin labeling in mammalian cells. NES = nuclear export signal, CAAX = prenylation tag, NLS = nuclear localization signal. MAP2 = microtubule-associated protein 2. All scale bars are 10 μm. 14 Several enzymes have also been developed to label a protein of interest with a short peptide sequence, including farnesyl transferase,117, 118 biotin ligase,119-122 myristyl transferase,123-125 formylglycine-generating enzyme,126-128 sortase,129-132 lipoic acid ligase (LplA)133-135 and tissue transglutaminase. As an example, LplA is based on an Escherichia coli enzyme whose natural function is to ligate lipoic acid onto three E. coli proteins involved in oxidative metabolism, essentially catalyzing the acylation of lysine residues of proteins with lipoic acid.133 The Ting lab showed that LplA could be used for labeling of cell-surface proteins; they demonstrated that the wild-type enzyme could ligate an azidoalkanoic acid instead of lipoic acid.134 Ligated azide could then be chemoselectively derivatized using cyclooctyne-fluorophore conjugates (Figure I8a). More noteworthy, through LplA active site mutagenesis, they created a ligase capable of directly attaching the fluorophore 7-hydroxycoumarin to proteins fused to a 13-amino acid LplA acceptor peptide (LAP) (Figure I-8b).135 A final analogue of the probe coumarin-AM2, in which both the carboxylic acid and the 7hydroxy group are protected with acetoxymethyl groups in order to allow for cell permeability was used for imaging; the groups can be cleaved to reveal reactive probe with intracellular endogenous esterases (Figure I-8c). The system was evaluated in mammalian cell lines, including HEK, HeLa and COS-7. Labeling of various LAP2 fusion proteins was also undertaken, where signaling peptides for localization to the nucleus (LAP2-NLS), cytosol (LAP2-NES) or plasma membrane (LAP2-CAAX) were included. Additionally, LAP2 fusions to the 15 cytoskeletal proteins β-actin and vimentin or microtubule-associated protein 2 (MAP2) could be specifically labeled with coumarin (Figure I-8d). In all of these site-specific chemical labeling systems, the reactive group that covalently binds to the tag is independent of the attached fluorophore, allowing a wide variety of fluorophores to be attached. The capability to explicitly attach chemical probes to individual proteins represents a powerful approach to study processes in living cells. The major drawback in some of these systems is the complexity associated with adding an external substrate that can be fluorescent itself when unbound or fluorescent due to non-specific binding, which reduces the signal-to-noise ratio and detection of protein labeling. Hence, these systems typically require stringent washing to reduce background. Ultimately, this delay in acquisition of data prevents real-time monitoring of fluorescence and it is not always possible to completely remove the unbound probe. It is therefore desirable to build in both spatial and temporal control. To address the issue of background fluorescence from unreacted probe, much effort has focused on the development of fluorogenic probes to be used in imaging.136-139 The underlying principle in the design of a fluorogenic probe is that its intrinsic fluorescence is quenched or suppressed until a reaction abolishes the quenching effect to restore the latent fluorescence. Numerous types of quenching mechanisms exist, but they can be generally categorized as follows: environmental quenching, energy transfer, or electron transfer.140 Environmental quenching arises from the fluorophore’s inherent sensitivity to its environment 16 such as solvent polarity, temperature, and pH, among others.136, 141-144 Solvatochromic fluorophores are most often used in environment-sensitive fluorogenic protein labeling, because these fluorophores typically are more fluorescent in a less polar medium or environment, such as the hydrophobic environment of a protein active site. More typically, researchers have employed probes reliant on reaction based quenching mechanisms. These include energy transfer, such as Forster resonance energy transfer (FRET),145-147 through-bond energy transfer (TBET),148-151 and electron transfer, as observed in photoinduced electron transfer (PeT).152-155 O POI R POI N POI O R O HS N O S N O N O R O HS O Intramolecular Michael addition R1 O Dimaleimide fluorogen O N HS O N S R1 O NONFLUORESCENT S R1 FLUORESCENT Figure I-9. Coumarin based fluorogenic probes for use in no-wash protein labeling. POI = protein of interest. As an example, the Keillor lab has developed coumarin-based fluorogenic probes for no-wash protein labeling. In this system, a protein of interest is genetically tagged with a short peptide containing two cysteine residues that are separated by two turns of an α-helix (approximately 10 Å apart).156-159 The probe becomes fluorescent upon selective reaction with synthetic fluorogenic ligands (Figure I-9). These ligands are comprised of a fluorophore and a dimaleimide 17 moiety, which exists to quench the fluorescence of the fluorophore via PeT until both maleimide groups undergo thiol additions. It should be noted that few native proteins contain two free cysteine residues on their surfaces at the proper distance to react with both maleimide groups. After tuning the reactivity of the probes, the developed labeling method enabled specific labeling of a protein of interest in live cells with no washing and low toxicity.156 Brightfield No Wash O O N H N O N H 2N Three Washes N Nile Red N O SNAP-tag S S O N target receptor target receptor O O N quenched probe fluorescent probe Figure I-10. A fluorogenic probe for cell surface proteins. A Nile Red BG derivative targeting SNAP-tagged membrane receptor. The free probe is nonfluorescent, but upon reaction with SNAP-tag fluorescence is activated. The probe (2 μM) was incubated with Chinese hamster ovary (CHO) cells stably expressing SNAP-tagged HIR for 30 min at 37 °C and imaged without and with washing. All scale bars are 10 μm. 18 Several groups have also developed FRET-based covalent fluorogenic probes. One such example is a fluorogenic probe for SNAP-tagged plasma membrane proteins based on the dye Nile Red (Figure I-10).97, 160 It takes advantage of Nile Red, a solvatochromic molecule highly fluorescent in an apolar environment, such as cellular membranes, but almost dark in a polar aqueous environment. The low fluorescence properties of the dye in a polar environment are believed to be due to the formation of aggregates that causes self-quenching, and to hydrogen bonding that has been reported to incite radiationless deactivation in anthraquinone and fluorenone derivatives.161-164 Similar to most solvatochromic probes, the absorption and emission of Nile Red-shifts strongly toward the red part of the spectrum as the solvent becomes more polar.163 As expected, Nile Red can spontaneously insert into membranes to become highly fluorescent due to the low polarity of these structures, and is thus a popular probe for staining membranes.165, 166 However, it is this propensity to localize in the membrane that prevents Nile Red from being used to visualize only a subset of membranes or proteins. Johnsson and coworkers hypothesized that tuning the membrane affinity of Nile Red through chemical derivatization would prevent this spontaneous insertion.97, 160 The probe possesses a tuned affinity for membranes allowing its Nile Red moiety to insert into the lipid bilayer of the plasma membrane, becoming fluorescent, only after its conjugation to a SNAP-tagged plasma membrane protein. CHO cells stably expressing a human insulin receptor (HIR)-SNAP-tag fusion protein were treated with the probe for 30 19 min at room temperature. Live cells were imaged for fluorescence after the labeling with and without wash steps. This probe proved to efficiently label SNAP-tagged plasma membrane proteins with minimal background in no-wash experiments (excitation at 572 nm and emission at 632 ± 30 nm). I.2.2 Bioorthogonal reactions Similar to the previously described site-specific chemical labeling approaches, other bioorthogonal reactions have been developed. Bioorthogonal reactions are between functional groups that do not react with biological entities under physiological conditions (ambient temperature, pressure, neutral pH, aqueous conditions) but selectively react with each other.167-171 In these systems, the protein is labeled with one of the functional groups and the fluorophore is labeled with the second moiety (Figure I-11). A group that is essentially absent from biological functionalities is the azide group. Azide-bearing unnatural amino acids have been incorporated into proteins and used in a variety of chemical reactions,172-174 including reaction with phosphines in Staudinger ligations175, [3+2] cycloadditions.177, 178 176 and reaction with terminal alkynes in Notably, Wang and coworkers developed the first “smart azide probes”.179 An azide in the 3-position of coumarin leads to significant quenching of its fluorescence via intramolecular charge transfer (ICT). Upon reaction with an alkyne, the lone pair of the azide is delocalized into the triazole, leading to fluorescence enhancement. 3-Azido-7-hydroxycoumarin showed the most impressive fluorescence enhancement upon click reaction, over 20 a hundred- fold in pH 7 phosphate buffered saline (PBS). The robust enhancement in fluorescence of this probe and its moderate water solubility and cell permeability have made this probe a popular choice for the detection of alkynes on intracellular and extracellular targets in biological systems. The CuI-catalyzed alkyne-azide cycloaddition (CuAAC) suffers from the necessity for Cu+, a species toxic to many living systems.180, 181 Unfortunately decreasing the Cu+ concentration severely hampers the reaction rate.182-184 The alkyne-azide cycloaddition is accelerated with respect to the uncatalyzed reaction by introducing ring strain into the alkyne (rather than using metal catalysis), creating a reaction dubbed the “strain-promoted alkyne-azide cycloaddition” (SPAAC).185, 186 These reactions have shown usefulness in labeling biomolecules within complex biological systems, including live mammalian cells and animals.187 A light induced 1,3-dipolar cycloaddition reaction between tetrazoles and terminal alkenes has been developed for use in biological settings.188, 189 This photoclick chemistry has been used to modify purified proteins in vitro and also to visualize proteins in living cells.190-193 Because the system is inducible by UVlight, it has the advantage of spatiotemporal control. Inverse-electron demand Diels-Alder reactions between tetrazines and strained alkenes or alkynes (including norbornenes and cyclopropenes, among others) yield dihydropyridazines or pyridiazines with nitrogen gas as the byproduct. These reactions have been explored as chemoselective reactions for 21 labeling in their native settings.194, 195 Advantageously, they are extraordinarily fast and can be made fluorogenic by conjugation of the tetrazine to a fluorophore; the fluorescence is quenched by the tetrazine, but upon cycloaddition fluorescence is restored.196 Staundinger Ligation: O R N N N -N2 O + H 2O Ph2P HO O R Ph + NH Ph P O Copper(I)-Catalyzed Azide-Alkyne Cycloaddition (CuAAC): R N N N Cu(I) + ligand R N N N Strain-Promoted Copper-Free Azide-Alkyne Cycloaddition (SPAAC): N N R N N N + R N Photoclick Cycloadditions: N N N N R hυ + -N2 R N N Diels-Alder Cycloadditions: + R N N N N -N2 R N NH Figure I-11. Methods of site-specific chemical labeling. The site for fluorophore incorporation is represented by a red star. I.2.3 Small protein tags The last major class of fluorogenic proteins consists of small protein tags, which are preferable to the aforementioned intrinsically fluorescent proteins due to their smaller size, as larger protein tags might cause unfavorable steric effects on the function or intracellular trafficking of proteins of interest. 22 Fluorogens are fluorescence generating dyes that are dark until constrained, bound or activated, ensuring that unreacted dye produces a minimal background signal. Fluorescence is produced only when the dye is bound or modified by the target, providing temporal control of the signal based on addition of the dye. Spatial control of the signal is achieved by fusing the activating protein to a specific protein or peptide that targets an organelle. This section will describe three varieties of small protein tags and fluorogen protein/small molecule pairs developed toward no-wash live-cell imaging. I.2.3.1 Fluorogen activating proteins Fluorogen-activating proteins (FAPs) are single-chain antibodies that bind a nonfluorescent molecule and stabilize it in a fluorescent state. Library screening of human single-chain antibodies (scFvs) was conducted by yeast surface display to isolate proteins that noncovalently bind to fluorogens such as thiazole orange or malachite green.197 Human scFvs were chosen because of their relatively small size of less than 30 kDa and the fact that they are easy to use as recombinant tags in fusion proteins. Thiazole orange and malachite green (Figure I-12a) are essentially nonfluorescent in solution, because the free rotation of aromatic rings in the fluorogen is thought to cause the promotion of nonradiative decay and the loss of fluorescence. Rotation around the single bond of the fluorogen is constrained upon binding to the FAP, and thus, fluorescence is “activated.” 23 In pioneering studies conducted by Waggoner and coworkers, the developed yeast FAPs were displayed on yeast and mammalian cell surfaces. The FAPs were shown to bind fluorogens with nanomolar affinity and brightness levels typical of fluorescent proteins.197 FAP domains provided a report of protein location and abundance with both temporal and spatial control upon addition of the fluorogen, rendering them more useful than the site-specific chemical labeling systems. The FAP-fluorogen system also allows for multicolor imaging of cells expressing different FAPs. Fluorogens derived from the malachite green chromophore (Figure I-12a) are highly fluorogenic, emit in the far-red spectral region (as compared to thiazole orange that emits in the yellow region), and display low nonspecific labeling in living cells.197 Bruchez and coworkers evaluated genetically encoded malachite green FAPs for targeting, brightness, and photostability in the cytosol, nucleus, mitochondria, peroxisomes, and endoplasmic reticulum (Figure I-12b).198 Labeling was realized within 20-30 minutes for each protein upon addition of nM concentrations of dye, producing a signal that colocalized with a linked mCerulean3 fluorescent protein. At a higher dye concentration of 1 μM, in the nuclear-targeted FAP, the malachite green dye achieved its labeling plateau in 7 minutes. The FAPs and malachite green dye can be used as specific, rapid and wash-free labels for intracellular organelles in live cells with far-red excitation and emission properties. Notably, labeling was proficiently achieved even in the reducing environment of the cytosol and nucleus, in which FAPs have a 24 propensity to misfold due to the presence of disulfide bonds within each variable domain. a. b. cytosol nucleus mitochondria peroxisome ER mCer3 FAP + MG N N malachite green (MG) merge O Figure I-12. a. Chemical structure of the fluorogen malachite green. b. Live cell imaging of the expressed FAP-mCer3 localized to the cytosol, nucleus, mitochondria, peroxisome and endoplasmic reticulum (ER) of HEK cells. All scale bars are 10 μm. I.2.3.2 Photoactive yellow protein Another small protein tag is photoactive yellow protein tag (PYP-tag), which consists of 125 amino acids, nearly half the size of GFP. The PYP-tag is derived from photoactive yellow protein, a small cytosolic photoreceptor originally isolated found in the purple photosynthetic bacteria Halorhodospira halophile.199 PYP binds to a natural cofactor, CoA thioester of 4-hydroxycinnamic acid, through transthioesterification with Cys69, the only cysteine residue in the protein (Figure I-13).200 It is known that PYP-tag also binds to the thioester derivative of 7hydroxycoumarin-3-carboxylic acid (Figure I-13);201 this coumarin ligand has been utilized to design a fluorogenic probe based on the mechanism of static quenching. 7-Hydroxycoumarin-3-carboxylic acid thioester is linked to fluorescein 25 through an ethylene glycol linker.202 In the absence of PYP, the probe is not fluorescent. However, upon binding to PYP, dissociation of the coumarin from fluorescein allows for fluorescence turn-on. This first generation probe required a wash step in live-cell imaging due to the extremely long incubation time required (24 h). O OH HO 4-hydroxycinnamic acid O OH HO O O 7-hydroxycoumarin-3-carboxylic acid Figure I-13. Crystal structure of photoactive yellow protein (PYP) bound to 4hydroxycinnamic acid through Cys69. Chemical structures of 4-hydoxycinnamic acid and 7-hydroxycoumarin-3-carboxylic acid, which also bind PYP, are also shown. Coordinates were obtained from PDB 2PHY. Later, the same group replaced the coumarin ligand with cinnamic acid thioester, in order to reduce the intramolecular stacking interaction between the ligand and fluorophore moieties.203 As compared to the probe containing the coumarin ligand, these probes displayed significantly improved kinetic properties, with up to 110 times acceleration in binding of the probe to PYP (t1/2 ~ 15 min). Unfortunately, the probe is not cell-permeable, so direct imaging of cell-surface proteins was undertaken. Epidermal growth factor receptor (EGFR) fused with 26 the PYP tag allowed for clear fluorescence imaging along the plasma membrane in cells treated with the probe, even without a wash step. Second generation probes were designed to be fluorogenic, cellpermeable and require a short incubation time. Designed probes were based on 7-dimethylaminocoumarin thioester derivatives.204 Notably, 7- dialkylaminocoumarin derivatives are environment-sensitive, in that they are scarcely fluorescent in polar solvents, but become fluorescent in low polarity solvents. It was hypothesized that the designed probe would be nonfluorescent in aqueous buffer, but emit when in the protein interior. It should be noted that the trimethylammonium and carboxylic acid moieties were included to increase water solubility (Figure I-14a). TMBDMA labeled 50% of PYP-tag for 1.1 min. As previously accomplished, no-wash live-cell imaging of cell surface proteins on the cell membrane allowed for fluorescence observation without washing the cells (Figure I-14b). In contrast, non-transfected cells proved to be essentially nonfluorescent. More impressively, genes encoding PYP-tag fused to maltosebinding protein (MBP-PYP) and nuclear localization signals (PYP-NLS) were created and expressed in HEK293T cells. Live cell images were obtained with a no-wash protocol immediately after TMBDMA was incubated with cells expressing either construct for 30 min (Figure I-14b). Time-lapse experiments showed the amount of incubation time necessary for fluorescence detection of PYP-tag inside cells. Fluorescence appeared in as little as 2 min after addition of the probe, with the reaction being nearly completely within 6 min (Figure I-14c). 27 PYP-EGFR a. b. O O CMBDMA TMBDMA O N H c. N No wash COOH Nontransfected S N Wash Transfected O PYP-NLS MBP-PYP 0 min 2 min 3 min 4 min 5 min 6 min 7 min 8 min 9 min 10 min Figure I-14. a. Chemical structures of the developed cell-permeable probes for labeling PYP. b. Live cell imaging of MBP-PYP, PYP-NLS and PYP-EGFR with TMBDMA. The top row shows fluorescence obtained upon excitation at 473 nm, using a 490-590 nm emission filter. The controls for nontransfected cells are shown in the bottom panel. MBP = maltose binding protein, NLS = nuclear localization signal, EGFR = epidermal growth factor receptor. Scale bar = 10 μm. c. Time-lapse live cell imaging of PYP-NLS with TMBDMA (left). Scale bar = 5 μm and plot of average fluorescence intensity of TMBDMA against incubation time (right). While TMBDMA showed better kinetic properties than CMBDMA, it was apparent that CMBDMA was brighter than TMBDMA. Thus, latest efforts focused on improving the kinetics of CMBDMA with PYP.205, 206 The CMBDMA2/PYP3RNLS pair enabled rapid and clear detection of fluorescence in nuclei only 1 min after the addition of the probe, showing saturation after 3 min.205 Most recently, Yellow Fluorescence-Activating and absorption-Shifting Tag (Y-FAST), a small protein tag enabling fluorescent labeling of proteins in living cells and multicellular organisms, was developed (Figure I-15).207 Y-FAST is an 28 engineered variant of PYP that was evolved through yeast surface display to reversibly bind 4-hydroxybenzylidene-rhodanine (HBR) or 4-hydroxy-3- methylbenzylidene-rhodanine (HMBR). λabs = 470-480 nm λabs = 400 nm λem = 530-540 nm HBR or HMBR Y-FAST Y-FAST POI POI HBR R = H HMBR R = Me O = S HO R O = NH S O S R NH S Figure I-15. Scheme of Yellow Fluorescence-Activating and absorption-Shifting Tag (Y-FAST). Ligands HBR and HMBR are non-fluorescent, but become fluorescent and their absorption red-shifts upon interaction with Y-FAST. HBR and HMBR are themselves nonfluorescent, but fortuitously they fluoresce specifically and instantaneously when bound to Y-FAST. Its high selectivity is reliant on two spectroscopic changes for fluorogen activation: 1) binding of H(M)BR to Y-FAST results in a significant increase of fluorescence quantum yield and 2) binding induces a large absorption red-shift. Hence, unbound fluorogen does not contribute to the fluorescence signal, ensuring high signal-to-noise ratio. HeLa cells expressing various Y-FAST fusions were successfully labeled with 5 μM HMBR (in vitro Y-FAST/HMBR complex has a 29 quantum yield of 0.33), targeted to the nucleus, cell membrane, mitochondria, golgi and microtubules. I.2.3.3 In situ formation of a cyanine dye Work by our lab in using engineered variants of human cellular retinol binding protein II (hCRBPII) bound with retinal through an active site lysine to modulate absorption wavelength,208, 209 inspired the design of a system in which the union of a protein and a nonfluorescent ligand produces a fluorescent protein. Protein engineering of cellular retinoic acid binding protein II (CRABPII), a member of the intracellular lipid binding protein family (iLBPs), led to a genetically encoded red fluorescent protein tag. Advantageously, CRABPII is a small protein (15.6 kDa), with a relatively large binding cavity and is remarkably tolerant to mutations. CRABPII was engineered to bind a merocyanine retinal aldehyde (MCRA), to generate a cyanine dye in situ.210 Reaction of the merocyanine aldehyde with the protein’s active site lysine renders a fluorescent iminium (Figure I-16). Advantageously, the free aldehyde is nonfluorescent and blueshifted relative to the iminium, rendering negligible background fluorescence from unbound merocyanine aldehyde. Formation of the iminium yields a resonating cation, leading to a bathochromic shift in absorption. Fortuitously, due to the extremely high degree of conjugation between the two nitrogen atoms, CRABPII/merocyanine complexes display high pKa, allowing for measurements at physiological pH, and red-shifted emission up to 619 nm (upon excitation at 594 nm). Additionally, high quantum efficiencies were obtained (up to 39%). In a 30 proof-of-principle experiment, merocyanine aldehyde was able to instantaneously label E. coli cells transformed with CRABPII mutants. Unfortunately, translation to mammalian cells was not fruitful. λem > 615 nm O λex = 594 nm N MCRA Figure I-16. Chemical structure of merocyanine retinal aldehyde (MCRA) and MCRA bound to an engineered CRABPII variant. Coordinates were obtained from PDB 3FEP. Fluorescence imaging of E. coli cells expressing CRABPII variant (left; overlay with brightfield on the right). Scale bar is 10 μm. Drawing inspiration from the idea of using synthetic chromophores, the Arnold lab was able to generate a near-infrared FP based on the combination of Archaerhodopsin-3 (Arch), a proton-pumping microbial rhodopsin, and a longer merocyanine aldehyde (one double bond longer than MCRA).211 Arch had also previously been engineered to bind its natural all-trans-retinal ligand, but brightness limits its usefulness in live cell imaging.212-214 Because the previously described merocyanine retinal analogue was capable of binding to CRABPII, it was anticipated that the longer merocyanine aldehyde would be able to bind to 31 Arch; Additionally, in Arch, the natural ligand all-trans-retinal is covalently bound to a conserved lysine residue via a Schiff base (SB). Directed evolution of Arch around the merocyanine retinal allowed for the enhancement of chromophore affinity, as well as the augmentation of fluorescent brightness in the near-infrared. The best variant displayed excitation and emission maxima of 759 and 770 nm, with a quantum yield of 0.13. The potential of the merocyanine aldehyde bound Arch variants were shown in bacterial cells, where fluorescence is readily detected even under low power light excitation at 727 nm. Merocyanine dependent fluorescence could be detected in eukaryotic HEK cells, but the authors noted that the expression and stability of Arch should be optimized for eukaryotic imaging. 32 REFERENCES 33 REFERENCES 1. Dean, K.M. & Palmer, A.E. Advances in fluorescence labeling strategies for dynamic cellular imaging. Nature Chemical Biology 10, 512-523 (2014). 2. Schaferling, M. The Art of Fluorescence Imaging with Chemical Sensors. Angewandte Chemie-International Edition 51, 3532-3554 (2012). 3. Kowada, T., Maeda, H. & Kikuchi, K. BODIPY-based probes for the fluorescence imaging of biomolecules in living cells. Chemical Society Reviews 44, 4953-4972 (2015). 4. Specht, E.A., Braselmann, E. & Palmer, A.E. A Critical and Comparative Review of Fluorescent Tools for Live-Cell Imaging. Annual Review of Physiology 79, 93-117 (2017). 5. Yuan, L., Lin, W.Y., Zheng, K.B., He, L.W. & Huang, W.M. Far-red to near infrared analyte-responsive fluorescent probes based on organic fluorophore platforms for fluorescence imaging. Chemical Society Reviews 42, 622-661 (2013). 6. Nienhaus, K. & Nienhaus, G.U. Fluorescent proteins for live-cell imaging with super-resolution. Chemical Society Reviews 43, 1088-1106 (2014). 7. Yang, Z.G., Cao, J.F., He, Y.X., Yang, J.H., Kim, T., Peng, X.J. & Kim, J.S. Macro-/micro-environment-sensitive chemosensing and biological imaging. Chemical Society Reviews 43, 4563-4601 (2014). 8. Hilderbrand, S.A. & Weissleder, R. Near-infrared fluorescence: application to in vivo molecular imaging. Current Opinion in Chemical Biology 14, 7179 (2010). 9. Guo, Z.Q., Park, S., Yoon, J. & Shin, I. Recent progress in the development of near-infrared fluorescent probes for bioimaging applications. Chemical Society Reviews 43, 16-29 (2014). 10. Luo, S.L., Zhang, E.L., Su, Y.P., Cheng, T.M. & Shi, C.M. A review of NIR dyes in cancer targeting and imaging. Biomaterials 32, 7127-7138 (2011). 11. Lee, M.H., Kim, J.S. & Sessler, J.L. Small molecule-based ratiometric fluorescence probes for cations, anions, and biomolecules. Chemical Society Reviews 44, 4185-4191 (2015). 34 12. Sevick-Muraca, E.M. Translation of Near-Infrared Fluorescence Imaging Technologies: Emerging Clinical Applications. Annual Review of Medicine 63, 217-231 (2012). 13. Shaner, N.C., Steinbach, P.A. & Tsien, R.Y. A guide to choosing fluorescent proteins. Nature Methods 2, 905-909 (2005). 14. Shu, X.K., Shaner, N.C., Yarbrough, C.A., Tsien, R.Y. & Remington, S.J. Novel chromophores and buried charges control color in mFruits. Biochemistry 45, 9639-9647 (2006). 15. Olenych, S.G., Claxton, N.S., Ottenberg, G.K. & Davidson, M.W. The fluorescent protein color palette. Current Protocols in Cell Biology 21.5, 134 (2007). 16. Day, R.N. & Davidson, M.W. The fluorescent protein palette: tools for cellular imaging. Chemical Society Reviews 38, 2887-2921 (2009). 17. Rizzo, M.A., Davidson, M.W. & Piston, D.W. Fluorescent protein tracking and detection: fluorescent protein structure and color variants. Cold Spring Harbor Protocols 2009, 1-21 (2009). 18. Sample, V., Newman, R.H. & Zhang, J. The structure and function of fluorescent proteins. Chemical Society Reviews 38, 2852-2864 (2009). 19. Remington, S.J. Green fluorescent protein: a perspective. Protein Science 20, 1509-1519 (2011). 20. Wu, B., Piatkevich, K.D., Lionnet, T., Singer, R.H. & Verkhusha, V.V. Modern fluorescent proteins and imaging technologies to study gene expression, nuclear localization, and dynamics. Current Opinion in Cell Biology 23, 310-317 (2011). 21. Stepanenko, O.V., Stepanenko, O.V., Kuznetsova, I.M., Verkhusha, V.V. & Turoverov, K.K. Beta-Barrel Scaffold of Fluorescent Proteins: Folding, Stability and Role in Chromophore Formation. International Review of Cell and Molecular Biology 302, 221-278 (2013). 22. Mishin, A.S., Belousov, V.V., Solntsev, K.M. & Lukyanov, K.A. Novel uses of fluorescent proteins. Current Opinion in Chemical Biology 27, 1-9 (2015). 35 23. Nienhaus, K. & Nienhaus, G.U. Chromophore photophysics and dynamics in fluorescent proteins of the GFP family. Journal of Physics: Condensed Matter 28, 443001 (2016). 24. Shimomura, O. & Johnson, F.H. Properties of the bioluminescent protein aequorin. Biochemistry 8, 3991-3997 (1969). 25. Shimomura, O., Johnson, F.H. & Saiga, Y. Extraction, Purification and Properties of Aequorin, a Bioluminescent Protein from Luminous Hydromedusan, Aequorea. Journal of Cellular and Comparative Physiology 59, 223-& (1962). 26. Prasher, D., Mccann, R.O. & Cormier, M.J. Cloning and Expression of the Cdna Coding for Aequorin, a Bioluminescent Calcium-Binding Protein. Biochemical and Biophysical Research Communications 126, 1259-1268 (1985). 27. Prasher, D.C., Eckenrode, V.K., Ward, W.W., Prendergast, F.G. & Cormier, M.J. Primary Structure of the Aequorea-Victoria GreenFluorescent Protein. Gene 111, 229-233 (1992). 28. Wang, S. & Hazelrigg, T. Implications for bcd mRNA localization from spatial distribution of exu protein in Drosophila oogenesis. Nature 369, 400-403 (1994). 29. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W.W. & Prasher, D.C. Green Fluorescent Protein as a Marker for Gene-Expression. Science 263, 802805 (1994). 30. Subach, F.V. & Verkhusha, V.V. Chromophore Transformations in Red Fluorescent Proteins. Chemical Reviews 112, 4308-4327 (2012). 31. Remington, S.J. Fluorescent proteins: maturation, photochemistry and photophysics. Current Opinion in Structural Biology 16, 714-721 (2006). 32. Zimmer, M. GFP: from jellyfish to the Nobel prize and beyond. Chemical Society Reviews 38, 2823-2832 (2009). 33. Piatkevich, K.D., Efremenko, E.N., Verkhusha, V.V. & Varfolomeev, D.D. Red fluorescent proteins and their properties. Russian Chemical Reviews 79, 243-258 (2010). 34. Tsien, R.Y. The green fluorescent protein. Annual Review of Biochemistry 67, 509-544 (1998). 36 35. Ormo, M., Cubitt, A.B., Kallio, K., Gross, L.A., Tsien, R.Y. & Remington, S.J. Crystal structure of the Aequorea victoria green fluorescent protein. Science 273, 1392-1395 (1996). 36. Yang, F., Moss, L.G. & Phillips, G.N. The molecular structure of green fluorescent protein. Nature Biotechnology 14, 1246-1251 (1996). 37. Heim, R., Prasher, D.C. & Tsien, R.Y. Wavelength Mutations and Posttranslational Autoxidation of Green Fluorescent Protein. Proceedings of the National Academy of Sciences of the United States of America 91, 12501-12504 (1994). 38. Cubitt, A.B., Heim, R., Adams, S.R., Boyd, A.E., Gross, L.A. & Tsien, R.Y. Understanding, Improving and Using Green Fluorescent Proteins. Trends in Biochemical Sciences 20, 448-455 (1995). 39. Reid, B.G. & Flynn, G.C. Chromophore formation in green fluorescent protein. Biochemistry 36, 6786-6791 (1997). 40. Bokman, S.H. & Ward, W.W. Renaturation of Aequorea GreenFluorescent Protein. Biochemical and Biophysical Research Communications 101, 1372-1380 (1981). 41. Baird, G.S., Zacharias, D.A. & Tsien, R.Y. Biochemistry, mutagenesis, and oligomerization of DsRed, a red fluorescent protein from coral. Proceedings of the National Academy of Sciences of the United States of America 97, 11984-11989 (2000). 42. Matz, M.V., Fradkov, A.F., Labas, Y.A., Savitsky, A.P., Zaraisky, A.G., Markelov, M.L. & Lukyanov, S.A. Fluorescent proteins from nonbioluminescent Anthozoa species. Nature Biotechnology 17, 969-973 (1999). 43. Verkhusha, V.V. & Lukyanov, K.A. The molecular properties and applications of Anthozoa fluorescent proteins and chromoproteins. Nature Biotechnology 22, 289-296 (2004). 44. Gross, L.A., Baird, G.S., Hoffman, R.C., Baldridge, K.K. & Tsien, R.Y. The structure of the chromophore within DsRed, a red fluorescent protein from coral. Proceedings of the National Academy of Sciences of the United States of America 97, 11990-11995 (2000). 37 45. Haddock, S.H., Mastroianni, N. & Christianson, L.M. A photoactivatable green-fluorescent protein from the phylum Ctenophora. Proceedings of the Royal Society B: Biological Sciences 277, 1155-1160 (2010). 46. Deheyn, D.D., Kubokawa, K., McCarthy, J.K., Murakami, A., Porrachia, M., Rouse, G.W. & Holland, N.D. Endogenous green fluorescent protein (GFP) in amphioxus. The Biological Bulletin 213, 95-100 (2007). 47. Pletnev, V.Z., Pletneva, N.V., Lukyanov, K.A., Souslova, E.A., Fradkov, A.F., Chudakov, D.M., Chepurnykh, T., Yampolsky, I.V., Wlodawer, A., Dauter, Z. & Pletnev, S. Structure of the red fluorescent protein from a lancelet (Branchiostoma lanceolatum): a novel GYG chromophore covalently bound to a nearby tyrosine. Acta Crystallographica Section D 69, 1850-1860 (2013). 48. Shcherbakova, D.M., Subach, O.M. & Verkhusha, V.V. Red fluorescent proteins: advanced imaging applications and future design. Angewandte Chemie-International Edition 51, 10724-10738 (2012). 49. Campbell, R.E., Tour, O., Palmer, A.E., Steinbach, P.A., Baird, G.S., Zacharias, D.A. & Tsien, R.Y. A monomeric red fluorescent protein. Proceedings of the National Academy of Sciences of the United States of America 99, 7877-7882 (2002). 50. Shaner, N.C., Campbell, R.E., Steinbach, P.A., Giepmans, B.N.G., Palmer, A.E. & Tsien, R.Y. Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp red fluorescent protein. Nature Biotechnology 22, 1567-1572 (2004). 51. Piatkevich, K.D. & Verkhusha, V.V. Advances in engineering of fluorescent proteins and photoactivatable proteins with red emission. Current Opinion in Chemical Biology 14, 23-29 (2010). 52. Subach, F.V., Piatkevich, K.D. & Verkhusha, V.V. Directed molecular evolution to design advanced red fluorescent proteins. Nature Methods 8, 1019-1026 (2011). 53. Shcherbakova, D.M., Baloban, M., Emelyanov, A.V., Brenowitz, M., Guo, P. & Verkhusha, V.V. Bright monomeric near-infrared fluorescent proteins as tags and biosensors for multiscale imaging. Nature Communications 7 (2016). 54. Chattoraj, M., King, B.A., Bublitz, G.U. & Boxer, S.G. Ultra-fast excited state dynamics in green fluorescent protein: multiple states and proton 38 transfer. Proceedings of the National Academy of Sciences of the United States of America 93, 8362-8367 (1996). 55. Lossau, H., Kummer, A., Heinecke, R., PollingerDammer, F., Kompa, C., Bieser, G., Jonsson, T., Silva, C.M., Yang, M.M., Youvan, D.C. & MichelBeyerle, M.E. Time-resolved spectroscopy of wild-type and mutant Green Fluorescent Proteins reveals excited state deprotonation consistent with fluorophore-protein interactions. Chemical Physics 213, 1-16 (1996). 56. Jung, G., Wiehler, J. & Zumbusch, A. The photophysics of green fluorescent protein: Influence of the key amino acids at positions 65, 203, and 222. Biophysical Journal 88, 1932-1947 (2005). 57. Oltrogge, L.M., Wang, Q. & Boxer, S.G. Ground-state proton transfer kinetics in green fluorescent protein. Biochemistry 53, 5947-5957 (2014). 58. Scharnagl, C. & Raupp-Kossmann, R.A. Solution pK(a) values of the green fluorescent protein chromophore from hybrid quantum-classical calculations. Journal of Physical Chemistry B 108, 477-489 (2004). 59. Brejc, K., Sixma, T.K., Kitts, P.A., Kain, S.R., Tsien, R.Y., Ormo, M. & Remington, S.J. Structural basis for dual excitation and photoisomerization of the Aequorea victoria green fluorescent protein. Proceedings of the National Academy of Sciences of the United States of America 94, 23062311 (1997). 60. Heim, R., Cubitt, A.B. & Tsien, R.Y. Improved Green Fluorescence. Nature 373, 663-664 (1995). 61. Hein, R. & Tsien, R.Y. Engineering green fluorescent protein for improved brightness, longer wavelengths and fluorescence resonance energy transfer. Current Biology 6, 178-182 (1996). 62. Kogure, T., Karasawa, S., Araki, T., Saito, K., Kinjo, M. & Miyawaki, A. A fluorescent variant of a protein from the stony coral Montipora facilitates dual-color single-laser fluorescence cross-correlation spectroscopy. Nature Biotechnology 24, 577-581 (2006). 63. Henderson, J.N., Osborn, M.F., Koon, N., Gepshtein, R., Huppert, D. & Remington, S.J. Excited State Proton Transfer in the Red Fluorescent Protein mKeima. Journal of the American Chemical Society 131, 1321213213 (2009). 39 64. Violot, S., Carpentier, P., Blanchoin, L. & Bourgeois, D. Reverse pHDependence of Chromophore Protonation Explains the Large Stokes Shift of the Red Fluorescent Protein mKeima. Journal of the American Chemical Society 131, 10356-10357 (2009). 65. Shcherbakova, D.M., Hink, M.A., Joosen, L., Gadella, T.W.J. & Verkhusha, V.V. An Orange Fluorescent Protein with a Large Stokes Shift for Single-Excitation Multicolor FCCS and FRET Imaging. Journal of the American Chemical Society 134, 7913-7923 (2012). 66. Pletnev, S., Shcherbakova, D.M., Subach, O.M., Pletneva, N.V., Malashkevich, V.N., Almo, S.C., Dauter, Z. & Verkhusha, V.V. Orange Fluorescent Proteins: Structural Studies of LSSmOrange, PSmOrange and PSmOrange2. Plos One 9 (2014). 67. Fron, E., De Keersmaecker, H., Rocha, S., Baeten, Y., Lu, G., Uji-I, H., Van der Auweraer, M., Hofkens, J. & Mizuno, H. Mechanism Behind the Apparent Large Stokes Shift in LSSmOrange Investigated by TimeResolved Spectroscopy. Journal of Physical Chemistry B 119, 1488014891 (2015). 68. Bergeler, M., Mizuno, H., Fron, E. & Harvey, J.N. QM/MM-Based Calculations of Absorption and Emission Spectra of LSSmOrange Variants. Journal of Physical Chemistry B 120, 12454-12465 (2016). 69. Chu, J., Oh, Y., Sens, A., Ataie, N., Dana, H., Macklin, J.J., Laviv, T., Welf, E.S., Dean, K.M., Zhang, F.J., Kim, B.B., Tang, C.T., Hu, M., Baird, M.A., Davidson, M.W., Kay, M.A., Fiolka, R., Yasuda, R., Kim, D.S., Ng, H.L. & Lin, M.Z. A bright cyan-excitable orange fluorescent protein facilitates dual-emission microscopy and enhances bioluminescence imaging in vivo. Nature Biotechnology 34, 760-767 (2016). 70. Shcherbo, D., Merzlyak, E.M., Chepurnykh, T.V., Fradkov, A.F., Ermakova, G.V., Solovieva, E.A., Lukyanov, K.A., Bogdanova, E.A., Zaraisky, A.G., Lukyanov, S. & Chudakov, D.M. Bright far-red fluorescent protein for whole-body imaging. Nature Methods 4, 741-746 (2007). 71. Piatkevich, K.D., Hulit, J., Subach, O.M., Wu, B., Abdulla, A., Segall, J.E. & Verkhusha, V.V. Monomeric red fluorescent proteins with a large Stokes shift. Proceedings of the National Academy of Sciences of the United States of America 107, 5369-5374 (2010). 72. Piatkevich, K.D., Malashkevich, V.N., Almo, S.C. & Verkhusha, V.V. Engineering ESPT Pathways Based on Structural Analysis of LSSmKate 40 Red Fluorescent Proteins with Large Stokes Shift. Journal of the American Chemical Society 132, 10762-10770 (2010). 73. Randino, C., Moreno, M., Gelabert, R. & Lluch, J.M. Peek at the Potential Energy Surfaces of the LSSmKate1 and LSSmKate2 Proteins. Journal of Physical Chemistry B 116, 14302-14310 (2012). 74. Chen, F.S., Zeng, Q., Zhuang, W. & Liang, W.Z. Characterizing the Structures, Spectra, and Energy Landscapes Involved in the Excited-State Proton Transfer Process of Red Fluorescent Protein LSSmKate1. Journal of Physical Chemistry B 120, 9833-9842 (2016). 75. Piatkevich, K.D., English, B.P., Malashkevich, V.N., Xiao, H., Almo, S.C., Singer, R.H. & Verkhusha, V.V. Photoswitchable Red Fluorescent Protein with a Large Stokes Shift. Chemistry & Biology 21, 1402-1414 (2014). 76. Yang, J., Wang, L., Yang, F., Luo, H.M., Xu, L.L., Lu, J.L., Zeng, S.Q. & Zhang, Z.H. mBeRFP, an Improved Large Stokes Shift Red Fluorescent Protein. Plos One 8 (2013). 77. Shcherbakova, D.M., Sengupta, P., Lippincott-Schwartz, J. & Verkhusha, V.V. Photocontrollable Fluorescent Proteins for Superresolution Imaging. Annual Review of Biophysics 43, 303-329 (2014). 78. Hell, S.W., Sahl, S.J., Bates, M., Zhuang, X.W., Heintzmann, R., Booth, M.J., Bewersdorf, J., Shtengel, G., Hess, H., Tinnefeld, P., Honigmann, A., Jakobs, S., Testa, I., Cognet, L., Lounis, B., Ewers, H., Davis, S.J., Eggeling, C., Klenerman, D., Willig, K.I., Vicidomini, G., Castello, M., Diaspro, A. & Cordes, T. The 2015 super-resolution microscopy roadmap. Journal of Physics D: Applied Physics 48 (2015). 79. Hell, S.W. & Wichmann, J. Breaking the Diffraction Resolution Limit by Stimulated-Emission - Stimulated-Emission-Depletion Fluorescence Microscopy. Optics Letters 19, 780-782 (1994). 80. Gustafsson, M.G.L. Nonlinear structured-illumination microscopy: Widefield fluorescence imaging with theoretically unlimited resolution. Proceedings of the National Academy of Sciences of the United States of America 102, 13081-13086 (2005). 81. Rust, M.J., Bates, M. & Zhuang, X.W. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nature Methods 3, 793-795 (2006). 41 82. Betzig, E., Patterson, G.H., Sougrat, R., Lindwasser, O.W., Olenych, S., Bonifacino, J.S., Davidson, M.W., Lippincott-Schwartz, J. & Hess, H.F. Imaging intracellular fluorescent proteins at nanometer resolution. Science 313, 1642-1645 (2006). 83. Hess, S.T., Girirajan, T.P.K. & Mason, M.D. Ultra-high resolution imaging by fluorescence photoactivation localization microscopy. Biophysical Journal 91, 4258-4272 (2006). 84. Shaner, N.C., Patterson, G.H. & Davidson, M.W. Advances in fluorescent protein technology. Journal of Cell Science 120, 4247-4260 (2007). 85. Hayashi, S. & Toda, Y. A novel fluorescent protein purified from eel muscle. Fisheries Science 75, 1461-1469 (2009). 86. Kumagai, A., Ando, R., Miyatake, H., Greimel, P., Kobayashi, T., Hirabayashi, Y., Shimogori, T. & Miyawaki, A. A Bilirubin-Inducible Fluorescent Protein from Eel Muscle. Cell 153, 1602-1611 (2013). 87. Auldridge, M.E. & Forest, K.T. Bacterial phytochromes: More than meets the light. Critical Reviews in Biochemistry and Molecular Biology 46, 67-88 (2011). 88. Piatkevich, K.D., Subach, F.V. & Verkhusha, V.V. Engineering of bacterial phytochromes for near-infrared imaging, sensing, and light-control in mammals. Chemical Society Reviews 42, 3441-3452 (2013). 89. Shcherbakova, D.M., Baloban, M. & Verkhusha, V.V. Near-infrared fluorescent proteins engineered from bacterial phytochromes. Current Opinion in Chemical Biology 27, 52-63 (2015). 90. Shcherbakova, D.M. & Verkhusha, V.V. Near-infrared fluorescent proteins for multicolor in vivo imaging. Nature Methods 10, 751-754 (2013). 91. Kaberniuk, A.A., Shemetov, A.A. & Verkhusha, V.V. A bacterial phytochrome-based optogenetic system controllable with near-infrared light. Nature Methods 13, 591-597 (2016). 92. Ghosh, S., Yu, C.L., Ferraro, D.J., Sudha, S., Pal, S.K., Schaefer, W.F., Gibson, D.T. & Ramaswamy, S. Blue protein with red fluorescence. Proceedings of the National Academy of Sciences of the United States of America 113, 11513-11518 (2016). 42 93. Gautier, A., Juillerat, A., Heinis, C., Correa, I.R., Jr., Kindermann, M., Beaufils, F. & Johnsson, K. An engineered protein tag for multiprotein labeling in living cells. Chemistry & Biology 15, 128-136 (2008). 94. Keppler, A., Gendreizig, S., Gronemeyer, T., Pick, H., Vogel, H. & Johnsson, K. A general method for the covalent labeling of fusion proteins with small molecules in vivo. Nature Biotechnology 21, 86-89 (2003). 95. Keppler, A., Pick, H., Arrivoli, C., Vogel, H. & Johnsson, K. Labeling of fusion proteins with synthetic fluorophores in live cells. Proceedings of the National Academy of Sciences of the United States of America 101, 99559959 (2004). 96. Lukinavicius, G., Umezawa, K., Olivier, N., Honigmann, A., Yang, G., Plass, T., Mueller, V., Reymond, L., Correa, I.R., Jr., Luo, Z.G., Schultz, C., Lemke, E.A., Heppenstall, P., Eggeling, C., Manley, S. & Johnsson, K. A near-infrared fluorophore for live-cell super-resolution microscopy of cellular proteins. Nature Chemistry 5, 132-139 (2013). 97. Sun, X., Zhang, A., Baker, B., Sun, L., Howard, A., Buswell, J., Maurel, D., Masharina, A., Johnsson, K., Noren, C.J., Xu, M.Q. & Correa, I.R., Jr. Development of SNAP-tag fluorogenic probes for wash-free fluorescence imaging. ChemBioChem 12, 2217-2226 (2011). 98. Hoehnel, S. & Lutolf, M.P. Capturing Cell-Cell Interactions via SNAP-tag and CLIP-tag Technology. Bioconjugate Chemistry 26, 1678-1686 (2015). 99. Provost, C.R. & Sun, L. Fluorescent labeling of COS-7 expressing SNAPtag fusion proteins for live cell imaging. Journal of Visualized Experiments (2010). 100. Calloway, N.T., Choob, M., Sanz, A., Sheetz, M.P., Miller, L.W. & Cornish, V.W. Optimized fluorescent trimethoprim derivatives for in vivo protein labeling. ChemBioChem 8, 767-774 (2007). 101. Miller, L.W., Cai, Y.F., Sheetz, M.P. & Cornish, V.W. In vivo protein labeling with trimethoprim conjugates: a flexible chemical tag. Nature Methods 2, 255-257 (2005). 102. Liu, W., Li, F., Chen, X., Hou, J., Yi, L. & Wu, Y.W. A rapid and fluorogenic TMP-AcBOPDIPY probe for covalent labeling of proteins in live cells. Journal of the American Chemical Society 136, 4468-4471 (2014). 43 103. Miller, L.W., Cai, Y., Sheetz, M.P. & Cornish, V.W. In vivo protein labeling with trimethoprim conjugates: a flexible chemical tag. Nature Methods 2, 255-257 (2005). 104. Jing, C. & Cornish, V.W. A fluorogenic TMP-tag for high signal-tobackground intracellular live cell imaging. ACS Chemical Biology 8, 17041712 (2013). 105. Los, G.V., Encell, L.P., McDougall, M.G., Hartzell, D.D., Karassina, N., Zimprich, C., Wood, M.G., Learish, R., Ohane, R.F., Urh, M., Simpson, D., Mendez, J., Zimmerman, K., Otto, P., Vidugiris, G., Zhu, J., Darzins, A., Klaubert, D.H., Bulleit, R.F. & Wood, K.V. HaloTag: A novel protein labeling technology for cell imaging and protein analysis. ACS Chemical Biology 3, 373-382 (2008). 106. Pries, F., Kingma, J., Krooshof, G.H., Jeronimusstratingh, C.M., Bruins, A.P. & Janssen, D.B. Histidine-289 Is Essential for Hydrolysis of the AlkylEnzyme Intermediate of Haloalkane Dehalogenase. Journal of Biological Chemistry 270, 10405-10411 (1995). 107. Griffin, B.A., Adams, S.R. & Tsien, R.Y. Specific covalent labeling of recombinant protein molecules inside live cells. Science 281, 269-272 (1998). 108. Adams, S.R., Campbell, R.E., Gross, L.A., Martin, B.R., Walkup, G.K., Yao, Y., Llopis, J. & Tsien, R.Y. New biarsenical Ligands and tetracysteine motifs for protein labeling in vitro and in vivo: Synthesis and biological applications. Journal of the American Chemical Society 124, 6063-6076 (2002). 109. Griffin, B.A., Adams, S.R., Jones, J. & Tsien, R.Y. Fluorescent labeling of recombinant proteins in living cells with FlAsH. Method in Enzymology 327, 565-578 (2000). 110. Hochuli, E., Bannwarth, W., Dobeli, H., Gentz, R. & Stuber, D. Genetic Approach to Facilitate Purification of Recombinant Proteins with a Novel Metal Chelate Adsorbent. Nature Biotechnology 6, 1321-1325 (1988). 111. Lata, S., Reichel, A., Brock, R., Tampe, R. & Piehler, J. High-affinity adaptors for switchable recognition of histidine-tagged proteins. Journal of the American Chemical Society 127, 10205-10215 (2005). 112. Tinazli, A., Tang, J.L., Valiokas, R., Picuric, S., Lata, S., Piehler, J., Liedberg, B. & Tampe, R. High-affinity chelator thiols for switchable and 44 oriented immobilization of histidine-tagged proteins: A generic platform for protein chip technologies. Chemistry-A European Journal 11, 5249-5259 (2005). 113. Wieneke, R., Raulf, A., Kollmannsperger, A., Heilemann, M. & Tampe, R. SLAP: Small Labeling Pair for Single-Molecule Super-Resolution Imaging. Angewandte Chemie-International Edition 54, 10216-10219 (2015). 114. Nonaka, H., Tsukiji, S., Ojida, A. & Hamachi, I. Non-enzymatic covalent protein labeling using a reactive tag. Journal of the American Chemical Society 129, 15777-15779 (2007). 115. Ojida, A., Honda, K., Shinmi, D., Kiyonaka, S., Mori, Y. & Hamachi, I. Oligo-Asp Tag/Zn(II) complex probe as a new pair for labeling and fluorescence imaging of proteins. Journal of the American Chemical Society 128, 10452-10459 (2006). 116. Halo, T.L., Appelbaum, J., Hobert, E.M., Balkin, D.M. & Schepartz, A. Selective Recognition of Protein Tetraserine Motifs with a Cell-Permeable, Pro-fluorescent Bis-boronic Acid. Journal of the American Chemical Society 131, 438-439 (2009). 117. Rashidian, M., Dozier, J.K., Lenevich, S. & Distefano, M.D. Selective labeling of polypeptides using protein farnesyltransferase via rapid oxime ligation. Chemical Communications 46, 8998-9000 (2010). 118. Duckworth, B.P., Zhang, Z., Hosokawa, A. & Distefano, M.D. Selective labeling of proteins by using protein farnesyltransferase. ChemBioChem 8, 98-105 (2007). 119. Chen, I., Howarth, M., Lin, W. & Ting, A.Y. Site-specific labeling of cell surface proteins with biophysical probes using biotin ligase. Nature Methods 2, 99-104 (2005). 120. Slavoff, S.A., Chen, I., Choi, Y.A. & Ting, A.Y. Expanding the substrate tolerance of biotin ligase through exploration of enzymes from diverse species. Journal of the American Chemical Society 130, 1160-1162 (2008). 121. Sueda, S., Yoneda, S. & Hayashi, H. Site-specific labeling of proteins by using biotin protein ligase conjugated with fluorophores. ChemBioChem 12, 1367-1375 (2011). 45 122. Roux, K.J., Kim, D.I., Raida, M. & Burke, B. A promiscuous biotin ligase fusion protein identifies proximal and interacting proteins in mammalian cells. The Journal of Cell Biology 196, 801-810 (2012). 123. Farazi, T.A., Waksman, G. & Gordon, J.I. The biology and enzymology of protein N-myristoylation. The Journal of Biological Chemistry 276, 3950139504 (2001). 124. Heal, W.P., Wickramasinghe, S.R., Bowyer, P.W., Holder, A.A., Smith, D.F., Leatherbarrow, R.J. & Tate, E.W. Site-specific N-terminal labelling of proteins in vitro and in vivo using N-myristoyl transferase and bioorthogonal ligation chemistry. Chemical Communications, 480-482 (2008). 125. Heal, W.P., Wickramasinghe, S.R., Leatherbarrow, R.J. & Tate, E.W. NMyristoyl transferase-mediated protein labelling in vivo. Organic & Biomolecular Chemistry 6, 2308-2315 (2008). 126. Roeser, D., Preusser-Kunze, A., Schmidt, B., Gasow, K., Wittmann, J.G., Dierks, T., von Figura, K. & Rudolph, M.G. A general binding mechanism for all human sulfatases by the formylglycine-generating enzyme. Proceedings of the National Academy of Sciences of the United States of America 103, 81-86 (2006). 127. Wu, P., Shui, W., Carlson, B.L., Hu, N., Rabuka, D., Lee, J. & Bertozzi, C.R. Site-specific chemical modification of recombinant proteins produced in mammalian cells by using the genetically encoded aldehyde tag. Proceedings of the National Academy of Sciences of the United States of America 106, 3000-3005 (2009). 128. Rabuka, D., Rush, J.S., deHart, G.W., Wu, P. & Bertozzi, C.R. Sitespecific chemical protein conjugation using genetically encoded aldehyde tags. Nature Protocols 7, 1052-1067 (2012). 129. Mao, H., Hart, S.A., Schink, A. & Pollok, B.A. Sortase-mediated protein ligation: a new method for protein engineering. Journal of the American Chemical Society 126, 2670-2671 (2004). 130. Popp, M.W., Antos, J.M., Grotenbreg, G.M., Spooner, E. & Ploegh, H.L. Sortagging: a versatile method for protein labeling. Nature Chemical Biology 3, 707-708 (2007). 46 131. Proft, T. Sortase-mediated protein ligation: an emerging biotechnology tool for protein modification and immobilisation. Biotechnology Letters 32, 1-10 (2010). 132. Ritzefeld, M. Sortagging: a robust and efficient chemoenzymatic ligation strategy. Chemistry-A European Journal 20, 8516-8529 (2014). 133. Cronan, J.E., Zhao, X. & Jiang, Y.F. Function, attachment and synthesis of lipoic acid in Escherichia coli. Advances in Microbial Physiology 50, 103146 (2005). 134. Fernandez-Suarez, M., Baruah, H., Martinez-Hernandez, L., Xie, K.T., Baskin, J.M., Bertozzi, C.R. & Ting, A.Y. Redirecting lipoic acid ligase for cell surface protein labeling with small-molecule probes. Nature Biotechnology 25, 1483-1487 (2007). 135. Uttamapinant, C., White, K.A., Baruah, H., Thompson, S., FernandezSuarez, M., Puthenveetil, S. & Ting, A.Y. A fluorophore ligase for sitespecific protein labeling inside living cells. Proceedings of the National Academy of Sciences of the United States of America 107, 10914-10919 (2010). 136. Klymchenko, A.S. Solvatochromic and Fluorogenic Dyes as EnvironmentSensitive Probes: Design and Biological Applications. Accounts of Chemical Research 50, 366-375 (2017). 137. Hori, Y. & Kikuchi, K. Protein labeling with fluorogenic probes for no-wash live-cell imaging of proteins. Current Opinion in Chemical Biology 17, 644650 (2013). 138. Baranczak, A., Connelly, S., Liu, Y., Choi, S., Grimster, N.P., Powers, E.T., Wilson, I.A. & Kelly, J.W. Fluorogenic small molecules requiring reaction with a specific protein to create a fluorescent conjugate for biological imaging-what we know and what we need to learn. Biopolymers 101, 484-495 (2014). 139. Bruchez, M.P. Dark dyes-bright complexes: fluorogenic protein labeling. Current Opinion in Chemical Biology 27, 18-23 (2015). 140. Chen, Y.C., Tsao, K. & Keillor, J.W. Fluorogenic protein labelling: a review of photophysical quench mechanisms and principles of fluorogen design. Canadian Journal of Chemistry 93, 389-398 (2015). 47 141. Machado, V.G., Stock, R.I. & Reichardt, C. Pyridinium N-Phenolate Betaine Dyes. Chemical Reviews 114, 10429-10475 (2014). 142. Klymchenko, A.S. & Mely, Y. Fluorescent Environment-Sensitive Dyes as Reporters of Biomolecular Interactions. Progress in Molecular Biology and Translational Science 113, 35-58 (2013). 143. Marini, A., Munoz-Losa, A., Biancardi, A. & Mennucci, B. What is Solvatochromism? Journal of Physical Chemistry B 114, 17128-17135 (2010). 144. Loving, G.S., Sainlos, M. & Imperiali, B. Monitoring protein interactions and dynamics with solvatochromic fluorophores. Trends in Biotechnology 28, 73-83 (2010). 145. Kikuchi, K., Takakusa, H. & Nagano, T. Recent advances in the design of small molecule-based FRET sensors for cell biology. TrAC Trends in Analytical Chemistry 23, 407-415 (2004). 146. Yuan, L., Lin, W.Y., Zheng, K.B. & Zhu, S.S. FRET-Based Small-Molecule Fluorescent Probes: Rational Design and Bioimaging Applications. Accounts of Chemical Research 46, 1462-1473 (2013). 147. Zhang, J., Campbell, R.E., Ting, A.Y. & Tsien, R.Y. Creating new fluorescent probes for cell biology. Nature Reviews Molecular Cell Biology 3, 906-918 (2002). 148. Jiao, G.S., Thoresen, L.H. & Burgess, K. Fluorescent, through-bond energy transfer cassettes for Labeling multiple biological molecules in one experiment. Journal of the American Chemical Society 125, 14668-14669 (2003). 149. Jiao, G.S., Thoresen, L.H., Kim, T.G., Haaland, W.C., Gao, F., Topp, M.R., Hochstrasser, R.M., Metzker, M.L. & Burgess, K. Syntheses, photophysical properties, and application of through-bond energy-transfer cassettes for biotechnology. Chemistry-A European Journal 12, 78167826 (2006). 150. Lin, W.Y., Yuan, L., Cao, Z.M., Feng, Y.M. & Song, J.Z. Through-Bond Energy Transfer Cassettes with Minimal Spectral Overlap between the Donor Emission and Acceptor Absorption: Coumarin-Rhodamine Dyads with Large Pseudo-Stokes Shifts and Emission Shifts. Angewandte Chemie-International Edition 49, 375-379 (2010). 48 151. Speiser, S. Photophysics and mechanisms of intramolecular electronic energy transfer in bichromophoric molecular systems: Solution and supersonic jet studies. Chemical Reviews 96, 1953-1976 (1996). 152. Fox, M.A. Photoinduced Electron-Transfer. Photobiology 52, 617-627 (1990). 153. Shieh, P., Hangauer, M.J. & Bertozzi, C.R. Fluorogenic azidofluoresceins for biological imaging. Journal of the American Chemical Society 134, 17428-17431 (2012). 154. Shieh, P., Siegrist, M.S., Cullen, A.J. & Bertozzi, C.R. Imaging bacterial peptidoglycan with near-infrared fluorogenic azide probes. Proceedings of the National Academy of Sciences of the United States of America 111, 5456-5461 (2014). 155. Ueno, T., Urano, Y., Setsukinai, K., Takakusa, H., Kojima, H., Kikuchi, K., Ohkubo, K., Fukuzumi, S. & Nagano, T. Rational principles for modulating fluorescence properties of fluorescein. Journal of the American Chemical Society 126, 14079-14085 (2004). 156. Chen, Y., Clouthier, C.M., Tsao, K., Strmiskova, M., Lachance, H. & Keillor, J.W. Coumarin-based fluorogenic probes for no-wash protein labeling. Angewandte Chemie-International Edition 53, 13785-13788 (2014). 157. Guy, J., Castonguay, R., Campos-Reales Pineda, N.B., Jacquier, V., Caron, K., Michnick, S.W. & Keillor, J.W. De novo helical peptides as target sequences for a specific, fluorogenic protein labelling strategy. Molecular Biosystems 6, 976-987 (2010). 158. Caron, K., Lachapelle, V. & Keillor, J.W. Dramatic increase of quench efficiency in "spacerless" dimaleimide fluorogens. Organic & Biomolecular Chemistry 9, 185-197 (2011). 159. Girouard, S., Houle, M.H., Grandbois, A., Keillor, J.W. & Michnick, S.W. Synthesis and characterization of dimaleimide fluorogens designed for specific labeling of proteins. Journal of the American Chemical Society 127, 559-566 (2005). 160. Prifti, E., Reymond, L., Umebayashi, M., Hovius, R., Riezman, H. & Johnsson, K. A fluorogenic probe for SNAP-tagged plasma membrane proteins based on the solvatochromic molecule Nile Red. ACS Chemical Biology 9, 606-612 (2014). 49 Photochemistry and 161. Biczok, L., Berces, T. & Inoue, H. Effects of molecular structure and hydrogen bonding on the radiationless deactivation of singlet excited fluorenone derivatives. Journal of Physical Chemistry A 103, 3837-3842 (1999). 162. Cser, A., Nagy, K. & Biczok, L. Fluorescence lifetime of Nile Red as a probe for the hydrogen bonding strength with its microenvironment. Chemical Physics Letters 360, 473-478 (2002). 163. Jose, J. & Burgess, K. Benzophenoxazine-based fluorescent dyes for labeling biomolecules. Tetrahedron 62, 11021-11037 (2006). 164. Yatsuhashi, T. & Inoue, H. Molecular mechanism of radiationless deactivation of aminoanthraquinones through intermolecular hydrogenbonding interaction with alcohols and hydroperoxides. Journal of Physical Chemistry A 101, 8166-8173 (1997). 165. Diaz, G., Melis, M., Batetta, B., Angius, F. & Falchi, A.M. Hydrophobic characterization of intracellular lipids in situ by Nile Red red/yellow emission ratio. Micron 39, 819-824 (2008). 166. Greenspan, P., Mayer, E.P. & Fowler, S.D. Nile red: a selective fluorescent stain for intracellular lipid droplets. The Journal of Cell Biology 100, 965-973 (1985). 167. Gong, Y.K. & Pan, L.F. Recent advances in bioorthogonal reactions for site-specific protein labeling and engineering. Tetrahedron Letters 56, 2123-2132 (2015). 168. Lang, K. & Chin, J.W. Bioorthogonal Reactions for Labeling Proteins. ACS Chemical Biology 9, 16-20 (2014). 169. Li, J. & Chen, P.R. Development and application of bond cleavage reactions in bioorthogonal chemistry. Nature Chemical Biology 12, 129137 (2016). 170. Lim, R.K.V. & Lin, Q. Photoinducible Bioorthogonal Chemistry: A Spatiotemporally Controllable Tool to Visualize and Perturb Proteins in Live Cells. Accounts of Chemical Research 44, 828-839 (2011). 171. Shieh, P. & Bertozzi, C.R. Design strategies for bioorthogonal smart probes. Organic & Biomolecular Chemistry 12, 9307-9320 (2014). 50 172. Davis, L. & Chin, J.W. Designer proteins: applications of genetic code expansion in cell biology. Nature Reviews Molecular Cell Biology 13, 168182 (2012). 173. Lang, K. & Chin, J.W. Cellular Incorporation of Unnatural Amino Acids and Bioorthogonal Labeling of Proteins. Chemical Reviews 114, 4764-4806 (2014). 174. Liu, C.C. & Schultz, P.G. Adding New Chemistries to the Genetic Code. Annual Review of Biochemistry 79, 413-444 (2010). 175. Tsao, M.L., Tian, F. & Schultz, P.G. Selective staudinger modification of proteins containing p-azidophenylalanine. ChemBioChem 6, 2147-2149 (2005). 176. Yanagisawa, T., Ishii, R., Fukunaga, R., Kobayashi, T., Sakamoto, K. & Yokoyama, S. Multistep Engineering of Pyrrolysyl-tRNA Synthetase to Genetically Encode N(epsilon)-(o-Azidobenzyloxycarbonyl) lysine for SiteSpecific Protein Modification. Chemistry & Biology 15, 1187-1197 (2008). 177. Rostovtsev, V.V., Green, L.G., Fokin, V.V. & Sharpless, K.B. A stepwise Huisgen cycloaddition process: Copper(I)-catalyzed regioselective "ligation" of azides and terminal alkynes. Angewandte ChemieInternational Edition 41, 2596-2599 (2002). 178. Tornoe, C.W., Christensen, C. & Meldal, M. Peptidotriazoles on solid phase: [1,2,3]-triazoles by regiospecific copper(I)-catalyzed 1,3-dipolar cycloadditions of terminal alkynes to azides. Journal of Organic Chemistry 67, 3057-3064 (2002). 179. Sivakumar, K., Xie, F., Cash, B.M., Long, S., Barnhill, H.N. & Wang, Q. A fluorogenic 1,3-dipolar cycloaddition reaction of 3-azidocoumarins and acetylenes. Organic Letters 6, 4603-4606 (2004). 180. Hong, V., Steinmetz, N.F., Manchester, M. & Finn, M.G. Labeling live cells by copper-catalyzed alkyne--azide click chemistry. Bioconjugate Chemistry 21, 1912-1916 (2010). 181. Wolbers, F., ter Braak, P., Le Gac, S., Luttge, R., Andersson, H., Vermes, I. & van den Berg, A. Viability study of HL60 cells in contact with commonly used microchip materials. Electrophoresis 27, 5073-5080 (2006). 51 182. Besanceney-Webler, C., Jiang, H., Zheng, T., Feng, L., Soriano del Amo, D., Wang, W., Klivansky, L.M., Marlow, F.L., Liu, Y. & Wu, P. Increasing the efficacy of bioorthogonal click reactions for bioconjugation: a comparative study. Angewandte Chemie-International Edition 50, 80518056 (2011). 183. Presolski, S.I., Hong, V., Cho, S.H. & Finn, M.G. Tailored ligand acceleration of the Cu-catalyzed azide-alkyne cycloaddition reaction: practical and mechanistic implications. Journal of the American Chemical Society 132, 14570-14576 (2010). 184. Rodionov, V.O., Presolski, S.I., Diaz, D.D., Fokin, V.V. & Finn, M.G. Ligand-accelerated Cu-catalyzed azide-alkyne cycloaddition: a mechanistic report. Journal of the American Chemical Society 129, 1270512712 (2007). 185. Agard, N.J., Baskin, J.M., Prescher, J.A., Lo, A. & Bertozzi, C.R. A comparative study of bioorthogonal reactions with azides. ACS Chemical Biology 1, 644-648 (2006). 186. Agard, N.J., Prescher, J.A. & Bertozzi, C.R. A strain-promoted [3 + 2] azide-alkyne cycloaddition for covalent modification of biomolecules in living systems. Journal of the American Chemical Society 126, 1504615047 (2004). 187. Sletten, E.M. & Bertozzi, C.R. From mechanism to mouse: a tale of two bioorthogonal reactions. Accounts of Chemical Research 44, 666-676 (2011). 188. Johnson, J.A., Lu, Y.Y., Van Deventer, J.A. & Tirrell, D.A. Residue-specific incorporation of non-canonical amino acids into proteins: recent developments and applications. Current Opinion in Chemical Biology 14, 774-780 (2010). 189. Ngo, J.T. & Tirrell, D.A. Noncanonical amino acids in the interrogation of cellular protein synthesis. Accounts of Chemical Research 44, 677-685 (2011). 190. Song, W., Wang, Y., Qu, J. & Lin, Q. Selective functionalization of a genetically encoded alkene-containing protein via "photoclick chemistry" in bacterial cells. Journal of the American Chemical Society 130, 9654-9655 (2008). 52 191. Song, W., Wang, Y., Qu, J., Madden, M.M. & Lin, Q. A photoinducible 1,3dipolar cycloaddition reaction for rapid, selective modification of tetrazolecontaining proteins. Angewandte Chemie-International Edition 47, 28322835 (2008). 192. Yu, Z., Ohulchanskyy, T.Y., An, P., Prasad, P.N. & Lin, Q. Fluorogenic, two-photon-triggered photoclick chemistry in live mammalian cells. Journal of the American Chemical Society 135, 16766-16769 (2013). 193. Yu, Z., Pan, Y., Wang, Z., Wang, J. & Lin, Q. Genetically encoded cyclopropene directs rapid, photoclick-chemistry-mediated protein labeling in mammalian cells. Angewandte Chemie-International Edition 51, 1060010604 (2012). 194. Blackman, M.L., Royzen, M. & Fox, J.M. Tetrazine ligation: fast bioconjugation based on inverse-electron-demand Diels-Alder reactivity. Journal of the American Chemical Society 130, 13518-13519 (2008). 195. Devaraj, N.K., Weissleder, R. & Hilderbrand, S.A. Tetrazine-based cycloadditions: application to pretargeted live cell imaging. Bioconjugate Chemistry 19, 2297-2299 (2008). 196. Lang, K., Davis, L., Wallace, S., Mahesh, M., Cox, D.J., Blackman, M.L., Fox, J.M. & Chin, J.W. Genetic Encoding of bicyclononynes and transcyclooctenes for site-specific protein labeling in vitro and in live mammalian cells via rapid fluorogenic Diels-Alder reactions. Journal of the American Chemical Society 134, 10317-10320 (2012). 197. Szent-Gyorgyi, C., Schmidt, B.A., Creeger, Y., Fisher, G.W., Zakel, K.L., Adler, S., Fitzpatrick, J.A.J., Woolford, C.A., Yan, Q., Vasilev, K.V., Berget, P.B., Bruchez, M.P., Jarvik, J.W. & Waggoner, A. Fluorogen-activating single-chain antibodies for imaging cell surface proteins. Nature Biotechnology 26, 235-240 (2008). 198. Telmer, C.A., Verma, R., Teng, H.B., Andreko, S., Law, L. & Bruchez, M.P. Rapid, Specific, No-wash, Far-red Fluorogen Activation in Subcellular Compartments by Targeted Fluorogen Activating Proteins. ACS Chemical Biology 10, 1239-1246 (2015). 199. Kumauchi, M., Hara, M.T., Stalcup, P., Xie, A.H. & Hoff, W.D. Identification of six new photoactive yellow proteins - Diversity and structure-function relationships in a bacterial blue light photoreceptor. Photochemistry and Photobiology 84, 956-969 (2008). 53 200. Kyndt, J.A., Meyer, T.E., Cusanovich, M.A. & Van Beeumen, J.J. Characterization of a bacterial tyrosine ammonia lyase, a biosynthetic enzyme for the photoactive yellow protein. FEBS Letters 512, 240-244 (2002). 201. van der Horst, M.A., Arents, J.C., Kort, R. & Hellingwerf, K.J. Binding, tuning and mechanical function of the 4-hydroxy-cinnamic acid chromophore in photoactive yellow protein. Photochemical & Photobiological Sciences 6, 571-579 (2007). 202. Hori, Y., Ueno, H., Mizukami, S. & Kikuchi, K. Photoactive Yellow ProteinBased Protein Labeling System with Turn-On Fluorescence Intensity. Journal of the American Chemical Society 131, 16610-16611 (2009). 203. Hori, Y., Nakaki, K., Sato, M., Mizukami, S. & Kikuchi, K. Development of Protein-Labeling Probes with a Redesigned Fluorogenic Switch Based on Intramolecular Association for No-Wash Live-Cell Imaging. Angewandte Chemie-International Edition 51, 5611-5614 (2012). 204. Hori, Y., Norinobu, T., Sato, M., Arita, K., Shirakawa, M. & Kikuchi, K. Development of Fluorogenic Probes for Quick No-Wash Live-Cell Imaging of Intracellular Proteins. Journal of the American Chemical Society 135, 12360-12365 (2013). 205. Hori, Y., Hirayama, S., Sato, M. & Kikuchi, K. Redesign of a Fluorogenic Labeling System To Improve Surface Charge, Brightness, and Binding Kinetics for Imaging the Functional Localization of Bromodomains. Angewandte Chemie-International Edition 54, 14368-14371 (2015). 206. Kamikawa, Y., Hori, Y., Yamashita, K., Jin, L., Hirayama, S., Standley, D.M. & Kikuchi, K. Design of a protein tag and fluorogenic probe with modular structure for live-cell imaging of intracellular proteins. Chemical Science 7, 308-314 (2016). 207. Plamont, M.A., Billon-Denis, E., Maurin, S., Gauron, C., Pimenta, F.M., Specht, C.G., Shi, J., Querard, J., Pan, B.Y., Rossignol, J., Morellet, N., Volovitch, M., Lescop, E., Chen, Y., Triller, A., Vriz, S., Le Saux, T., Jullien, L. & Gautier, A. Small fluorescence-activating and absorptionshifting tag for tunable protein imaging in vivo. Proceedings of the National Academy of Sciences of the United States of America 113, 497-502 (2016). 208. Wang, W.J., Geiger, J.H. & Borhan, B. The photochemical determinants of color vision. Bioessays 36, 65-74 (2014). 54 209. Wang, W.J., Nossoni, Z., Berbasova, T., Watson, C.T., Yapici, I., Lee, K.S.S., Vasileiou, C., Geiger, J.H. & Borhan, B. Tuning the Electronic Absorption of Protein-Embedded All-trans-Retinal. Science 338, 13401343 (2012). 210. Yapici, I., Lee, K.S.S., Berbasova, T., Nosrati, M., Jia, X.F., Vasileiou, C., Wang, W.J., Santos, E.M., Geiger, J.H. & Borhan, B. "Turn-On" Protein Fluorescence: In Situ Formation of Cyanine Dyes. Journal of the American Chemical Society 137, 1073-1080 (2015). 211. Herwig, L., Rice, A.J., Bedbrook, C.N., Zhang, R.J.K., Lignell, A., Cahn, J.K.B., Renata, H., Dodani, S.C., Cho, I., Cai, L., Gradinaru, V. & Arnold, F.H. Directed Evolution of a Bright Near-Infrared Fluorescent Rhodopsin Using a Synthetic Chromophore. Cell Chemical Biology 24, 415-425 (2017). 212. Engqvist, M.K.M., McIsaac, R.S., Dollinger, P., Flytzanis, N.C., Abrams, M., Schor, S. & Arnold, F.H. Directed Evolution of Gloeobacter violaceus Rhodopsin Spectral Properties. Journal of Molecular Biology 427, 205-220 (2015). 213. Flytzanis, N.C., Bedbrook, C.N., Chiu, H., Engqvist, M.K.M., Xiao, C., Chan, K.Y., Sternberg, P.W., Arnold, F.H. & Gradinaru, V. Archaerhodopsin variants with enhanced voltage-sensitive fluorescence in mammalian and Caenorhabditis elegans neurons. Nature Communications 5 (2014). 214. McIsaac, R.S., Engqvist, M.K.M., Wannier, T., Rosenthal, A.Z., Herwig, L., Flytzanis, N.C., Imasheva, E.S., Lanyi, J.K., Balashov, S.P., Gradinaru, V. & Arnold, F.H. Directed evolution of a far-red fluorescent rhodopsin. Proceedings of the National Academy of Sciences of the United States of America 111, 13034-13039 (2014). 55 CHAPTER II: MODULATION OF ABSORPTION AND EMISSION VIA A PROTEIN EMBEDDED SOLVATOCHROMIC FLUOROPHORE Our overarching goal is to develop fluorescent proteins that span the entire visible spectra, to be used when conventional fluorescent proteins are inadequate. In our system, fluorescence is activated upon coupling of the protein and ligand, such that temporal control can be achieved, whereas intrinsically fluorescent proteins are constitutively on; it is only upon ligand addition and irradiation that our complex will become fluorescent. Additionally, our system does not require oxygen and can therefore find potential in obligate anaerobes. Our lab has demonstrated the ability to effectively control the absorption profile of conjugated polyenes;1-3 however, in this case we envision using solvatochromic fluorophores, in hopes that emission wavelength can be regulated with the same level of control. The term solvatochromism is used to describe the marked change in wavelength, and sometimes intensity, of an absorption and/or emission band that results from changing the polarity of the media.4, 5 It is proposed that this effect can be mimicked inside of a protein cavity by the introduction of specific interactions between the bound fluorophore and amino acid side chains. As seen in Chapter I, a variety of methods can be employed in developing fluorogenic (or “turn-on”) proteins. Our lab has been successful in labeling expressed variants of human cellular retinol binding protein II (hCRBPII) in living cells with merocyanine aldehyde in multiple cell lines, localized to a number of 56 intracellular targets (allowing for spatial control). Because fluorescence is activated upon coupling of the protein and ligand, temporal control can also be achieved. Advantageously, this system does not require oxygen and can therefore find potential in obligate anaerobes. Additionally, because the protonated Schiff base (PSB), formed as a result of protein/ligand condensation, is inherently sensitive to pH, innumerable applications can be envisioned such as multi-color imaging and in vivo pH sensing, among others. Specificity is imperative in engineering a practical and efficient fluorogenic system. As will first be described, the most deleterious drawback of merocyanine aldehyde in imaging live cells is the background fluorescence that arises from non-specific labeling (i.e. from iminium formation with nonspecific proteins), due to the high pKa and unchanging excitation and emission profiles of merocyanine aldehyde bound hCRBPII complexes. To exclude the observation of background fluorescence due to nonspecific binding, we focused on the coupling of aldehydic derivatives of known solvatochromic fluorophores with engineered variants of hCRBPII. This offers two advantages: 1) Emission of solvatochromic fluorophores is variable based on the polarity of their environment. Thus, it should be possible to develop probes of many different colors. 2) We also should be able to alter the hCRBPII/fluorophore emission away from that of nonspecific labeling. This would result in a probe that yields little background fluorescence. 57 II.1 Application of merocyanine aldehyde with hCRBPII mutants for in vivo imaging CHO N MCRA HN NH2 hCRBPII Lys108 Abs λmax 603 nm MCRA Abs λmax 492 nm Em λmax 623 nm 1 Emission (normalized) Absorbance (normalized) 1 hCRBPII/MCRA Lys108 0 0 400 500 600 Wavelength (nm) 700 Figure II-1. Reaction and spectroscopic properties of merocyanine aldehyde (MCRA) with hCRBPII. Coordinates were obtained from PDB 3FEP. Our lab has recently described the spectroscopic characterization of merocyanine retinal analog (MCRA), a weakly fluorescent aldehyde, which upon imine formation becomes fluorophoric (turn-on fluorescence).6 The complex also exhibits a substantial bathochromic shift in absorption wavelength, rendering any background fluorescence from unbound reagent obsolete (Figure II-1). Dr. Ipek Yapici was successful in conducting preliminary studies with MCRA and cellular retinoic acid binding protein II (CRABPII). These studies highlighted the 58 advantage of using the ‘turn-on’ approach in bacteria. Nonetheless, expression of CRABPII in mammalian cell cultures was not fruitful, most probably as a result of protein misfolding. On the other hand, hCRBPII variants have shown no deleterious problem in expression or folding, producing highly fluorescent pigments upon complexation with MCRA. Dr. Wenjing Wang and Dr. Tetyana Berbasova pioneered the efforts to couple hCRBPII with MCRA, to produce fast binding mutants that are complete within one minute (it should be noted that all in vitro data detailed in this section was completed by Dr. Wang and Dr. Berbasova). With promising in vitro data for hCRBPII/MCRA complexes, we sought to prove the system’s expediency in live cells. Preliminary screening of hCRBPII mutants coupled with MCRA yielded six potential candidates for further optimization (Table II-1, entries 3-8). Q108K:K40L (KL) and Q108K:K40L:R58F bind MCRA relatively fast with half-life (t1/2) maturation times of 5.7 min and 10.0 min, respectively (Table II-1, entries 3 and 4). With respectable half-lives, we sought to increase the complex’s quantum yield. In previous wavelength regulation studies with retinal, it was shown that R58, located at the entrance of the protein cavity, was important for encapsulating the binding cavity; upon substitution of R58 with phenylalanine, quantum efficiency did not change (18% for KL and 16% for KL:R58F). Fortuitously though, T51V doubled the quantum efficiency of the hCRBPII/MCRA complex (Table II-1, entries 5 and 6), yielding a two-fold increase in brightness 59 relative to KL:R58F. Unfortunately, the T51V addition led to substantially slower formation of the iminium. Table II-1. Spectroscopic properties of hCRBPII mutants coupled with merocyanine retinal analog (MCRA). Entry Mutant λmax (abs/em) Φ a ε (10-3 M-1cm-1) t1/2b (min) 1 mRFP 584/607 0.25 50 - 2 mRaspberry 598/625 0.15 86 - 3 Q108K:K40L (KL) 600/619 0.18 162 5.7 4 KL:R58F 600/619 0.16 165 10.0 5 KL:T51V (KLV) 598/616 0.37 166 140.3 6 KLV:R58F 603/623 0.36 161 76.6 7 KLV:T53C:R58W:T29L:Y19W:A33W:Q4F 600/619 0.19 190 10.4 8 KLV:T53C:R58W:T29L:A33W:Q4F 605/625 0.19 170 2.7 9 KLV:R58F:L117E 593/615 0.36 157 16.1 10 KLV:R58F:L117E:Q4F 594/615 0.32 161 14.4 11 KLV:T53C:R58W:T29L:A33W:Q4F:L117E 590/611 0.37 166 0.15 12 KLV:T53C:R58W:T29L:A33W:Q4F:L117D 590/613 0.30 197 0.16 a Quantum yields were determined based on two fluorescent standards (Oxazine1 and Oxazine-170). bHalf-lives are based on rate constants obtained from fitting the data to second order kinetics. We then screened hCRBPII mutants previously, produced in our lab, for faster kinetics of iminium formation with MCRA, retaining the T51V mutation. Iminium formation was faster in a few mutants (Table II-1, entries 7 and 8). Unfortunately, the quantum yield was significantly decreased, as compared to KLV:R58F. Regardless of the low quantum yield, the hCRBPII/MCRA complexes display brightness (the product of quantum yield and extinction coefficient) higher than FPs emitting in the same wavelength regime (Table II-1, entries 1 and 2). 60 603 nm 623 nm 2 108 0.2 1 108 0.1 0 500 3 108 550 600 650 Wavelength (nm) Emission (count/sec) Abs. (a. u.) 0.3 0 700 Figure II-2. UV-Vis (blue curve) and emission spectra (red curve) of Q108K:K40L:T51V:R58F/MCRA. Because KL:T51V:R58F/MCRA is an acceptably bright complex, we chose to test its utility as a fluorescent tag in mammalian cells. Additionally, its absorption and emission spectra (Figure II-2) are well aligned with available lasers on the confocal microscope (594 nm excitation and 615 nm emission long pass filter). Labeling of KL:T51V:R58F (hCRBPIItetra) with MCRA was evaluated in cervical carcinoma (HeLa) cells. hCRBPII was cloned into a pFlag-CMV2 vector containing EGFP, which serves as a positive control of transfection. Cells expressing EGFP-hCRBPIItetra, were incubated with 250 nM MCRA for 1 hour at 37 °C. Whole cell fluorescence is observed upon excitation of EGFP at 488 nm (BP 505-530 emission filter). To our delight, upon excitation at 594 nm (LP 615 nm emission filter), whole cell fluorescence can be detected. While it was pleasing to see that MCRA can efficiently penetrate the cell membrane, significant background fluorescence was observed (Figure II-3, EGFP- hCRBPIItetra at 0 min) when cells were imaged immediately after MCRA 61 incubation. As hCRBPII/MCRA complexes do not show wavelength regulation, background fluorescence is presumably due to off-target iminium formation that absorbs and emits in the same wavelength regime as our hCRBPII/MCRA complex. EGFP-CRBPtetra 0 min 2h EGFP-CRBPtetraQ108L 4h 8h EGFP ex: 488 nm em: BP 505-530 nm MCRA/hCRBPII ex: 594 nm em: LP 615 nm MCRA/hCRBPII + DIC Figure II-3. Labeling of EGFP-hCRBPIItetra in HeLa cells. All samples were incubated with 250 nM MCRA for 1 h at 37 °C, washed with PBS buffer two times and then incubated at 37 °C for the indicated amount of time (left to right: 0 min, 2 h, 4 h and 8 h). The right most panel is labeling of EGFP-hCRBPIItetraQ108L after incubating with 250 nM MCRA for 1 h at 37 °C, washed with PBS buffer two times and imaged immediately. All scale bars are 20 μM. To this end, we sought to test whether an incubation period after MCRA staining would allow for nonspecific staining to be degraded. HeLa cells expressing the protein fusion were labeled with MCRA for 1 hour at 37 °C and imaged after various post MCRA incubation times. Cells were imaged 0 min, 2 hours, 4 hours and 8 hours after the initial 1 hour MCRA incubation period (Figure II-3, EGFP-hCRBPIItetra). As expected, the EGFP channel is essentially the same in each experiment. However, in the red channel indicative of 62 hCRBPII/MCRA, more background fluorescence is observed with no post MCRA incubation period (cells were imaged directly after MCRA incubation and washout), while background fluorescence is dimmest 8 hours post MCRA staining. This indicates that a ‘degradation’ period is necessary after the 1 hour MCRA incubation, to allow for the abatement of nonspecific fluorescence. To irrefutably show that MCRA is labeling hCRBPII in the expressed EGFP-hCRBPIItetra fusion, we made the active site lysine free variant, EGFPhCRBPIItetraQ108L, in which Q108 is mutated to leucine. It should not form an iminium with MCRA, and thus will not lead to a turn-on response by the addition of MCRA. HeLa cells expressing EGFP-hCRBPIItetraQ108L were incubated with 250 nM MCRA for 1 hour at 37 °C, washed two times with PBS buffer and imaged immediately (Figure II-3, EGFP-hCRBPIItetraQ108L). Whole cell fluorescence is observed upon excitation of EGFP at 488 nm (BP 505-530 emission filter). However, upon excitation at 594 nm, no whole cell fluorescence is detected; only significant background fluorescence is observed. Colocalization analyses of MCRA in non-transfected cells indicate that nonspecific iminium formation proceeds in the lysosomes (see Dr. Tetyana Berbasova’s thesis, Figure IV-26). Because the fluorescence from this nonspecific iminium formation is in the same channel as specific hCRBPII/MCRA fluorescence, imaging will be encumbered due to low signal to noise ratio. It was hypothesized that with a lower incubation time (requiring a faster reaction between hCRBPII and MCRA), hCRBPII/MCRA iminium formation would be able to outcompete nonspecific binding. 63 Having shown that MCRA can label hCRBPIItetra in live cells, we focused our efforts on the generation of a fast reacting hCRBPII mutant. This was desirable for a number of reasons, but most notably, to avoid the reaction of MCRA with off-target lysines that produce complexes fluorescing at the same wavelength as hCRBPII/MCRA. As aforementioned, the long incubation time of 1 hour required for labeling hCRBPIItetra, due to the slow rate of iminium formation, leads to substantial nonspecific iminium formation. Additionally, a faster reacting hCRBPII mutant would enable quick imaging and visualization; the 8 h incubation post-addition of MCRA, necessary for high contrast imaging, would hamper visualization of processes within this time frame. To engineer a fast binding hCRBPII mutant for MCRA, we took inspiration from the PLP-dependent enzymes and aldolases.7 In these enzymes, iminium formation is optimized as a necessary part of their function. To aid in the first step of iminium formation, these enzymes employ an acidic residue to activate the substrate’s carbonyl for nucleophilic attack by an active-site lysine. We have also observed similar acid catalyzed iminium formation in our lab. Previous studies in CRABPII showed that L121E was successful at enhancing the rate of iminium formation.6 An overlay of CRABPII and hCRBPII identified hCRBPII’s L117 as an analogous residue to L121 in CRABPII (Figure II-4). Gratifyingly, when L117E was introduced into hCRBPIItetra, the rate of iminium formation increased by nearly four-fold (Table II-1, entry 9). Fortuitously, the quantum yield and brightness of the fluorescent complex were not affected. 64 Thus, we chose to retain the L117E mutation, which led to a number of fast binding hCRBPII mutants. A Q4F mutation was also included as it has been shown to lead to higher protein expression yields without negatively affecting the hCRBPII/MCRA complex’s quantum yield or rate of iminium formation (Table II1, entry 10). L117 (hCRBPII) L121 (CRABPII) Figure II-4. Overlay of CRABPII (R111K:R132L:Y134F:T54V:R59W, shown in magenta) and hCRBPII (Q108K:K40L:T51V:T53C:R58W:T29L:Y19W:Q4R, shown in green). Coordinates were obtained from PDB 4I9S and 4EEJ. Iminium formation was fastest when L117E was incorporated into the Q4F containing octa mutant shown in entry 8 (Table II-1). The resulting Q108K:K40L:T51V:T53C:R58W:T29L:A33W:Q4F:L117E-hCRPBII nona mutant (hCRBPIInona) shows extraordinarily fast reaction kinetics of iminium formation, with a half-life for maturation of 10 seconds (based on fitting to second order kinetics; measured at 23 °C with 20 μM protein and 0.5 equiv MCRA at pH ~ 7; the rate of iminium formation was too fast at 37 °C). The introduction of the shorter aspartic acid residue (L117D) led to similar results (Table II-1, entry 12). Fortuitously, the quantum efficiency of the hCRBPIInona/MCRA complex was equivalent to that of hCRBPIItetra/MCRA, presumably due to the strong electrostatic interaction between the negatively charged carboxylate and the 65 positively charged iminium, confining the conformation of the generated cyanine dye. 125 nM 250 nM 500 nM EGFP ex: 488 nm em: BP 505 - 530 nm MCRA/hCRBPII ex: 594 nm em: LP 615 nm MCRA/hCRBPII + DIC Figure II-5. Optimization of MCRA concentration for live cell imaging in HeLa cells expressing EGFP-hCRBPIInona; various concentrations (125 nM, 250 nM and 500 nM) were tested. All scale bars are 20 μM. With a fast binding hCRBPII variant (hCRBPIInona) in hand, we sought to test its applicability in live cell imaging. First, we chose to optimize the MCRA concentration necessary for effective labeling of hCRBPII; in these experiments, we incubated the cells with MCRA for 30 sec at room temperature following by washing with PBS buffer two times before imaging. Three concentrations of MCRA of 125 nM, 250 nM and 500 nM were tested in HeLa cells for hCRBPIInona labeling (Figure II-5). The best results were obtained when using a MCRA 66 concentration of 250 nM. A MCRA concentration of 500 nM led to substantially more background fluorescence, as evidenced by the red fluorescence from nontransfected cells (transfected cells are indicated by the EGFP channel). On the other hand, the intensity of red fluorescence from hCRBPII/MCRA, when using a MCRA concentration of 125 nM was more dim than when 250 nM MCRA was used. Therefore, we chose to proceed with a MCRA concentration of 250 nM, as the best signal-to-noise ratio is obtained after 30 sec incubation. EGFP-CRBPnona3NLS EGFP-CRBPnona COS-7 U2OS EGFP-CRBPnonaNES HeLa EGFP ex: 488 nm em: BP 505-530 nm MCRA/hCRBPII ex: 594 nm em: LP 615 nm MCRA/hCRBPII + DIC Figure II-6. Labeling of EGFP-hCRBPIInona in Hela cells. All samples were incubated with 250 nM MCRA for 30 sec at room temperature, washed with PBS buffer two times and imaged immediately. Localization signals 3NLS and NES targeted hCRBPII to the nucleus and extranuclear space, respectively. NLS = nuclear localization sequence, NES = nuclear export sequence. All scale bars are 20 μM. For subsequent experiments, routine imaging was performed by incubating the cells with 250 nM MCRA for 30 seconds at room temperature. After washing with buffer two times, confocal images were acquired immediately, 67 demonstrating that the hCRBPIInona/MCRA complex is formed almost instantaneously in vivo and can be used for imaging soon after addition of ligand. EGFP-hCRBPIInona was successfully expressed and labeled with MCRA in multiple cell lines including COS7, U2OS and HeLa cells (Figure II-6). As before, excitation at 488 nm (BP 505-530 emission filter) leads to whole cell fluorescence from EGFP. Similarly, excitation at 594 nm (LP 615 nm emission filter) shows whole cell fluorescence as a result of hCRBPII/MCRA complex. Minimal background in the red channel is observed in both COS7 and HeLa cells; moderate background fluorescence is apparent in U2OS cells. Subsequently, targeting of EGFP-hCRBPIInona was investigated by the inclusion of a nuclear localization sequence (3NLS) and nuclear export sequence (NES).8 Fortuitously, the nucleus fluoresces as efficiently as the cytosol localization under identical conditions, even though the chromophore needs to penetrate not only the cellular membrane, but also enter the nucleus before binding to hCRBPIInona (Figure II-6). EGFP-hCRBPIInona-NES showed fluorescence in the extranuclear space in both the EGFP and hCRBPII/MCRA (red) channels. Gratifyingly, nonspecific fluorescence is not observable in any of the hCRBPIInona fusion constructs in HeLa cells. It appears that labeling of hCRBPIInona with MCRA occurs within 30 seconds and can outcompete nonspecific iminium formation in the lysosomes. Hence, long incubation times 68 following the buffer exchange necessary for hCRBPIItetra were not needed for hCRBPIInona, due to its fast iminium formation with MCRA. As has been shown, hCRBPII/MCRA can successfully label mammalian cells, targeting to intracellular organelles, with an unprecedented half-life of 10 seconds. Advantageously, hCRBPII/MCRA complexes also exhibit high quantum yields and extinction coefficients. The only pitfall with the system is that background fluorescence is observable from nonspecific iminium formation. Presumably, this is due to the high pKa of the in situ generated cyanine dye. Additionally, the absorption wavelength is essentially constant in different hCRBPII variants when coupled with MCRA due to the high degree of conjugation between the iminium and terminal nitrogen. Thus, we can imagine developing fluorogenic probes in two ways. Firstly, we can prevent nonspecific iminium formation. It is envisioned that this can be done by using an initially nonfluorescent probe that becomes fluorescent upon iminium formation with hCRBPII. However, it would require a less reactive fluorophore than merocyanine aldehyde (namely the methyl ketone variant), such that the probe does not react nonspecifically, rendering it only fluorescent upon reaction inside an activated hCRBPII variant. A second option would involve utilizing environmentally sensitive probes to tailor the specific hCRBPII/fluorophore emission, such that the fluorescence from the desired reaction is red-shifted as compared to that of nonspecific iminium formation. Presumably, this can be achieved by changing the electrostatic environment 69 around the fluorophore inside of the protein cavity. This chapter will focus on the latter. II.2 Principles of solvatochromicity a. Stokes shift b. Excitation 10-15 sec Fluorescence 10-9 sec 1 absorption 0 440 460 emission 480 500 520 540 Wavelength (nm) Normalized emission Internal conversion 10-12 sec Normalized abosorbance 1 0 560 Figure II-7. a. Simplified Jablonski diagram. b. Diagram of absorption and emission spectra; the difference in the band maxima is the Stokes shift. Fluorescence generally involves three processes; these are excitation, nonradiative transition, and fluorescence emission as is shown by the simplified Jablonski diagram in Figure II-7a.9-11 First, a photon is supplied and the fluorophore is excited to a vibrational level within the first excited state (S1). This process occurs on a 10-15 seconds time scale. The excited state exists for a finite time (typically nanoseconds), after which the fluorophore undergoes conformational changes such as internal conversion (10-12 sec) or intersystem crossing (this produces the triplet state which leads to phosphorescence). Upon returning to the ground state, a photon is emitted (10-9 seconds). 70 Emission from fluorophores occurs at wavelengths bathochromically shifted from the absorption due to the dissipation of energy in the excited state. This difference in energy of the emission maxima and absorption maxima is referred to as the Stokes shift (Figure II-7b). Importantly, a larger Stokes shift allows for emission photons to be isolated from excitation photons. Because the time required to electronically excite a fluorophore is much shorter than the time required to undergo vibrational motion, the position of nuclei of the absorbing molecule do not change between the ground state and excited state.12 However, if the lifetime of the excited state is long enough, then reorientation of the surrounding solvent molecules can take place, after which fluorescence occurs from this state. Thus, solvent and environment can affect a fluorophore’s emission spectra and quantum yields.13 Examples include solvent polarity, viscosity, rate of solvent relaxation, fluorophore conformational changes, rigidity, charge transfer, excited state reactions, and changes in radiative and non-radiative decay rates.13 These factors have all led to the development of a variety of environment-sensitive dyes.14-16 One such class of dyes is solvatochromic fluorophores.14-16 Solvatochromic fluorophores display emission properties, including emission wavelengths, quantum yields and fluorescence lifetimes, that are highly sensitive to their environment. The structures of some of the most common solvatochromic fluorophores are shown in Figure II-8. Typically, only fluorophores that are themselves polar exhibit substantial sensitivity to solvent polarity. These include 71 fluorophores derived from a variety of molecules, including dansyl,17, 18 prodan and derivatives,19-23 anthradan,24 fluorene,25-28 dapoxyl,29-31 nitrobenzoxadiazole (NBD),32-35 dimethylaminophthalimide,36-39 coumarin,40-42 phenoxazine,43-47 among others. The similarity amongst these dyes is that they bear a conjugated system with an electron-donor and electron-acceptor group, generating a pushpull system. Most solvatochromic dyes show relatively low fluorescence quantum yields in polar solvents and high fluorescence quantum yields in apolar solvents.13-16 Dansyl Prodan Phenoxazine NBD O O Cl S O O O N N O N O O N N N HN N Dimethylaminophthalimide O N N Dapoxyl Coumarin N RO O O O (or NR2) N O O S N O R (or the amide variant) Figure II-8. Examples of fluorescent solvatochromic dyes. The wavy lines indicate the common point for dye conjugation. Typically, these fluorophores have a larger dipole moment in the excited state than in the ground state. If the dipole moment of the molecule increases upon excitation, then a more polar solvent will serve to stabilize the excited state. As a result, the separation between the ground state and excited state energies is decreased. Hence, the emission is red-shifted; this effect is called positive 72 solvatochromism. Conversely, a hypsochromic shift of the emission band with increasing solvent polarity is termed negative solvatochromism. The magnitude of the shift is dependent on the change in the dipole moment of the probe. The Jablonski diagram depicting the energies of the different electronic states of the system is shown in Figure II-9. Internal conversion Solvent relaxation Excitation Fluorophore Fluorescence Solvent Figure II-9. Origins of solvatochromicity in fluorescence. Numerous scales of solvent polarity have been developed; many of these scales are based on the spectral data of a single standard probe.4, 48-57 Reichardt developed one such scale, the ET(30) scale, which is based on the spectroscopic behavior of the negatively solvatochromic betaine indicator dye 2,6-diphenyl-4(2,4,6-triphenyl-1-pyridinio)phenolate (Betaine 30).57-59 Betaine 30 shows solvent dependent spectral shifts (Figure II-10), in which absorption wavelength is higher (i.e. lower in energy) in nonpolar solvents.59 Since its introduction, the ET(30) parameter has been used for empirically measuring the polarity of many systems. 73 a. b. Ph Ph N Ph 1 2 3 4 5 Ph Ph O Betaine 30 λabs~ 510 550 700 600 675 Figure II-10. a. Structure of Betaine 30. b. Solutions of Betaine 30 in (1) methanol, (2) ethanol, (3) 1-octanol, (4) N,N-dimethylacetamide and (5) dichloromethane. II.3 Choice of fluorophore In the late 1990s, Diwu and coworkers developed the Dapoxyl dyes (Figure II-8); these strongly solvatochromic compounds exhibit distinctive variations in quantum yields, fluorescence wavelength maxima (i.e. larger Stokes shifts) and extinction coefficients with changes in polarity.29 Fortuitously, they also display a much lower quantum yield in polar solvents than nonpolar solvents (Φhexane=0.91 and Φ1:4 acetonitrile:water=0.04). Their characteristic of being almost nonfluorescent in aqueous solutions should translate to almost no fluorescence being observed from the unbound probe in an aqueous solution, a feature beneficial in live cell imaging. Since their discovery, many derivatives of the Dapoxyl dyes have been synthesized. Min and coworkers were successful in generating a Dapoxyl dye library of 80 unique structures.30 Through a fluorescence-based high throughput screening of their dapoxyl library, one Dapoxyl derivative was found to be highly sensitive toward human serum albumin (HSA). Under physiological conditions 74 (50 mM HEPES, pH 7.4), A-41S displayed a fluorescence increase of 55-fold and a blue-shift of the fluorescence emission (520 to 473 nm) (Figure II-11). a. b. N Relative fluorescence intensity HO3S HSA HO O F A41-S I/I0 c. Wavelength (nm) [HSA] (µM) Figure II-11. Fluorescence response of A-41S to HSA. a. 5 μM A41-S with HSA at 0, 1, 2, 3, 4, 5, 7.5, 10, 15, 20 and 30 μM. b. Structure of A41-S. c. Corresponding intensity changes are the ratio of intensity of A-41-S with/without HSA. As aforementioned, Dapoxyl dyes exhibit a typical push pull system, conjugated through a pi-system, where the effect of donor and acceptor substitutions has already been evaluated. To this end, Koji Suzuki and coworkers investigated the effect of the central heteroaromatic ring system (Figure I-12), which serves as a pi-linker between the donor and acceptor moieties of the fluorophore, replacing the oxazole with thiophene (KSD-3), furan (KSD-2) and pyrrole (KSD-1).31 X O N X = NH : KSD-1 X = O : KSD-2 X = S : KSD-3 Figure II-12. Structures of Dapoxyl dyes used to investigate the effect of the pilinker. 75 Some spectroscopic properties of the chromophores are detailed in Table II-2.31 The absorption maxima do not change significantly upon changing polarity of the solvent, indicating that the solvent does not stabilize the ground state conformations of these molecules. Conversely, strong positive solvatochromic shifts are observed when the solvent polarity is increased for all three dyes. As described previously, this indicates a solvent stabilized excited state, presumably due to the difference in dipole moments of the ground state and excited state. Table II-2. Spectral properties of KSD-1-3 in solvents of different polarities. Solvent KSD-1 KSD-2 KSD-3 λabs λem Φ λabs λem Φ λabs λem Φ Toluene 358 413 0.32 368 428 0.60 386 455 0.67 1,4-Dioxane 352 415 0.59 366 434 0.82 385 466 0.76 Ethyl acetate 352 423 0.64 364 446 0.72 384 479 0.61 Chloroform 365 443 0.33 378 465 0.62 395 490 0.54 Acetone 355 449 0.77 369 474 0.78 389 506 0.60 DMSO 363 462 0.82 380 499 0.74 400 531 0.49 Acetonitrile 355 458 0.77 371 491 0.80 390 519 0.57 i-Propanol 366 479 0.34 379 512 0.60 394 535 0.46 Ethanol 365 495 0.20 380 528 0.40 395 548 0.18 Methanol 363 509 0.11 381 541 0.13 396 559 0.02 The shifts in emission spectra for the dyes from toluene to methanol are 96 nm, 113 nm and 104 nm (indicating a polarity dependent intramolecular charge transfer from the electron donating N,N-dimethyl amino phenyl moiety to the electron withdrawing acetyl unit), for KSD-1, KSD-2 and KSD-3, respectively. It is also observed that when thiophene is used as the pi-linker, the longest 76 emission wavelengths are obtained. This is likely due to better effective conjugation of the thiophene moiety. 2.0 equiv EtO 1. 2.0 equiv NaH O P CN Br 2. EtO Br II-1 0.84 M Na2CO3 in H2O, 0.05 equiv Pd(PPh3)4 2.0 equiv N,N-dimethylphenylboronic acid O yield = 90% CN S N CN S II-2 S II-3 0.28 M PhCH3, 0.56 M MeOH, 1. 1.5 equiv DIBAL, CH2Cl2 H 2. HPLC separation of cis and trans isomer S yield (over two steps) = 4% N O ThioFluor Scheme II-1. Synthesis of ThioFluor. Based on the previous data, we chose to start from the aldehydic analog of KSD-3. ThioFluor was synthesized in three steps from the commercially available 2-acetyl-3-bromothiophene (Scheme II-1). Horner-Wadsworth-Emmons (HWE) reaction with ylide II-1 yields intermediate II-2. II-2 was coupled with N,Ndimethylphenylboronic acid, resulting in II-3. Final reduction of the nitrile resulted in the desired fluorophore ThioFlour. II.4 Spectroscopic properties of ThioFluor Spectroscopic characterization of the free aldehyde was first undertaken. Results are summarized in Table II-3. As is the case for the known solvatochromic fluorophores, ThioFluor’s absorbance does not change significantly in different solvents (Figure II-13a), but its emission varies over 140 nm from nonpolar toluene to the more polar ethanol, with a larger Stokes shift in polar solvents (Figure II-13b). Thus, under white light, all solutions appear colorless to faint yellow (Figure II-3c), while under UV irradiation (with a TLC 77 headlamp, 365 nm), the solutions show radically different colors (Figure II-13d). As expected the free aldehyde’s absorbance is nearly constant while its emission changes as a function of the solvent’s ET(30) value (Figure II-13e).4 Table II-3. Spectroscopic characterization of ThioFluor in various solvents. ε Solvent λabs λem Stokes shift Φa -1 (M cm-1) 422 516 94 32,149 0.02 Ethyl acetate 414 558 144 28,939 0.11 Dimethyl sulfoxide 437 623 186 26,891 0.53 Ethanol 430 654 224 27,021 0.06 PBS buffer 393 - - 8,842 0.00 Absolute quantum yield was measured on a Quantaurus-QY. Abs. (a. u.) a. 0.3 0.2 Toluene Ethyl acetate Ethanol DMSO PBS Buffer b. 0.1 0 300 350 400 450 500 550 600 650 Wavelength (nm) c. 1 Toluene Ethyl acetate Ethanol DMSO Normalized Fluorescence Intensity a Toluene 0 450 500 550 600 650 700 750 800 Wavelength (nm) e. 240 d. Toluene Ethyl acetate DMSO Ethanol Stokes Shift [nm] 220 y = -131.9 + 6.96x R= 0.99 200 180 160 140 120 100 80 30 35 40 45 ET(30) [kcal/mol] 50 55 Figure II-13. Spectroscopic properties of ThioFluor in different solvents. a. UVVis and b. Fluorescence spectra of ThioFluor. ThioFluor under c. white light and d. UV irradiation at 365 nm (with TLC handlamp). e. Stokes shift in different solvents versus the ET(30) value indicates that ThioFluor is solvatochromic. 78 The extinction coefficients of ThioFluor in toluene, ethyl acetate, dimethyl sulfoxide and ethanol do not drastically differ; they are all approximately 30,000 M-1cm-1. However, the formation of aggregates is quite apparent from the UV-Vis spectrum of ThioFluor in PBS buffer.60 Fortuitously, this leads to essentially a non-fluorescent species in PBS buffer, which may indicate that when applied to in vivo applications the free aldehyde should not be fluorescent, acting as a fluorogenic probe. Most solvatochromic fluorophores are highly fluorescent in nonpolar solvents; however, ThioFluor is only slightly emissive in toluene. Interestingly, ThioFluor is only highly fluorescent in dimethyl sulfoxide, with a quantum yield 0f 0.53. H O S n-BuNH2 N R Bu HCl R H N Bu N ThioFluor λabs = 430 nm ThioFluor-SB λabs = 400 nm ThioFluor-PSB λabs = 521 nm 0.4 521 nm 400 nm Abs. (a. u.) 0.3 430 nm 0.2 0.1 0 300 350 400 450 500 550 600 650 Wavelength (nm) Figure II-14. Formation of the protonated Schiff base (PSB) of ThioFluor with nbutyl amine in ethanol. 79 The aldehyde was then coupled with n-butyl amine in ethanol, in order to mimic the product that will be obtained upon condensation with an active site lysine in the protein pocket. Reaction of the aldehyde with n-butyl amine produces a blue-shifted imine, or Schiff base (SB), with absorption maxima of 400 nm (Figure II-14). Subsequent protonation of the ThioFluor-SB yields a resonating cation, resulting in a large red-shift in UV-Vis, with an absorbance maximum of 521 nm. ThioFluor-PSB was formed in a variety of solvents. Results are summarized in Table II-4. Table II-4. Spectroscopic properties of ThioFluor-PSB with n-butyl amine. λabs λem Stokes Shift Φa PBS buffer 464 - - - Ethanol 521 689 168 0.02 Tetrahydrofuran 494 674 180 0.17 Formamide 525 707 182 0.01 Acetone 493 687 194 0.03 Dimethyl sulfoxide 512 716 204 0.02 Solvent a Absolute quantum yield was measured on a Quantaurus-QY. Neither the absorption maxima nor the emission maxima of the ThioFluor- PSB vary significantly in different solvents. It should be noted that THF was the least polar solvent employed, because acidification was carried out using HCl, which requires the solvents to be miscible. Nonetheless, it is seen that the ThioFluor-PSB is much less solvatochromic than the ThioFluor aldehyde. Additionally, the emission wavelength of the PSB is different in aprotic and protic solvents, indicating that PSB emission does not exactly follow a purely solvatochromic trend (as a result of a purely polar effect) based on just solvent 80 polarity (Figure II-15).4 This suggests that some other factor, such as hydrogen bonding, is affecting the emission wavelength of ThioFluor-PSB; this is not surprising given that ThioFluor possess a positively charged nitrogen, which can hydrogen bond with polar solvents. ThioFluor-PSB was much less solvatochromic than ThioFluor; however, we chose to start coupling the chromophore to hCRBPII protein mutants, in order to deduce whether ThioFluor’s emission could be altered by the introduction of point mutations in hCRBPII. aprotic solvents 210 DMSO Stokes Shift [nm] 200 180 protic solvents Acetone 190 Formamide THF 170 160 35 Ethanol 40 45 50 ET(30) [kcal/mol] 55 60 Figure II-15. Stokes shift in different solvents versus the ET(30) value indicates that ThioFluor-PSB is not solvatochromic. Different correlations are observed in protic and aprotic solvents. II.5 Removal of polarity induces a red-shift in absorption and emission It was quickly realized that global polarity changes of the binding pocket led to little to no soluble protein expression, and the application of solvatochromicity purely in terms of its definition would not be possible. In other 81 words, it is not possible to introduce enough polar residues into the binding pocket to induce a global polarity change, while still obtaining soluble protein after expression. It is hypothesized that the introduction of hydrophilic residues in the protein active site leads to protein destabilization, causing protein unfolding so that the polar residues are exposed to the solvent exterior. Hence, it was decided to introduce interactions to the chromophore via point mutations, which could alter its absorption. Q38 R58 T53 T51 Q4 Q108K Figure II-16. Polar residues mutated in order to remove polarity from the binding pocket. Coordinates obtained from PDB 4EXZ (hCRBPII-Q108K:K40L/retinal). Our group has previously shown, utilizing retinal as a ligand, that an even distribution of electrostatic potential across the entire chromophore is essential for maximal bathochromic shifting.1-3 To this end, we tried to remove polarity in 82 two ways: 1) removal of the more polar residues in the binding pocket nearest the chromophore and 2) removal of the water molecule directly interacting with the iminium. Removal of these negative dipoles in the PSB region should encourage positive charge delocalization. Generally, hCRBPII has a hydrophobic binding pocket, with minimal polar residues. To this end, mutagenesis was carried out on T51, T53, Q4, Q38 and R58 residues (Figure II-16), mutating them to similarly sized hydrophobic residues in the parent double mutant Q108K:K40L (KL). KL was used, as PSB could not be observed for the Q108K mutant, as acidification of the complex only led to precipitation. Presumably K40L increases the pKa slightly, such that some PSB is formed. 3.4 Å 3.5 Å Q4 Q108K Figure II-17. Water mediated hydrogen bonding between Q4 and the hCRBPII/retinal iminium. Coordinates obtained from PDB 4EFG (hCRBPIIQ108K:K40L:T51V:T53C:R58W:T29L:Y19W/retinal). We chose to start with mutation of Q4 because in crystal structures of hCRBPII/retinal with Q4, there is a water mediated hydrogen bond from the iminium to Q4, capable of stabilizing the protonated state of the PSB (Figure II17).3 To disrupt this presumed water network, Q4 was mutated to hydrophobic 83 phenylalanine in KL (Table II-5, entry 1). As expected, the removal of polarity leads to a red-shift in both absorption and fortuitously emission, a change of 27 nm and 17 nm for absorption and emission, respectively from KL (Table II-5, entry 2). From this initial hit, it was suspected that a bathochromic shift in absorption should result in a bathochromic shift in emission; hence, studies were continued by attempting to red-shift the absorption (and in effect emission). Table II-5. Spectroscopic change as a result of Q4F mutation. Entry a Mutant λabs λem Φa nd 1 Q108K:K40L 580 674 2 Q108K:K40L:Q4F 607 691 0.12 Absolute quantum yield was measured on a Quantaurus-QY. Hydrophobic valine was substituted for the more polar T51 as an isosteric replacement. Gratifyingly, the Q108K:K40L:T51V hCRBPII triple mutant led to a 28 nm red-shift in absorption and 16 nm red-shift in emission from the parent Q108K:K40L (Table II-6). This red-shift is in agreement with the prior conclusion that removing negative polarity in the Schiff base region leads to more delocalization along the polyene, in the case of hCRBPII/retinal. Beyond the T51V mutation, other residues were not tested at position 51. The T51V mutation served an even larger purpose than red-shifting the wavelength; fortuitously, T51V is also capable of monomerizing hCRBPII variants (see Chapter VI for a full discussion) and it was thus retained in most mutants tested. As will be discussed later (Chapter II.8.1), complexes of hCRBPII with ThioFluor displayed significantly different spectroscopic properties of the monomer and domain- 84 swapped dimer. Only spectroscopic properties of the monomer will be discussed here. Table II-6. Spectroscopic change as a result of T51V mutation. Entry a Mutant λabs λem Φa nd 1 Q108K:K40L 580 674 2 Q108K:K40L:T51V 608 690 0.12 Absolute quantum yield was measured on a Quantaurus-QY. As shown in Figure II-18, T53 is situated near the middle of the polyene in the crystal structure of KL/retinal, providing a hydrogen bond to T51 through a water molecule. In order to alter its polarity, T53 was replaced with various residues including alanine, serine and valine in the parent template KL. Interestingly, all three substitutions led to at least a slight red-shift in absorption. From KL the change in absorption was 5 nm, 7 nm, and 17 nm for T53A, T53S and T53V, respectively (Table II-7). T51 T53 2.5 Å 3.1 Å Q108K Figure II-18. Water mediated hydrogen bonding between T51 and T53. Coordinates obtained from PDB 4EXZ (hCRBPII-Q108K:K40L/retinal). 85 There seems to be no trend when looking at the absorption wavelength versus the polarity. Mutations to nonpolar alanine and valine red-shifted absorption, but did not significantly affect the emission wavelength. T53S moderately affected the absorption wavelength, yet had the largest effect on emission (674 nm in KL to 682 nm). Threonine and serine are of approximately the same polarity; it is the volume of the side chain that is different. Presumably, in these cases, polarity is not the only determining factor since no correlation is found between the electrostatic substitution and wavelength. It is possible that chromophore orientation is not preserved through the course of these mutations, or the conformation of residues in close contact with T53 have changed. Table II-7. Spectroscopic change as a result of mutating T53. Entry a Mutant λabs λem Φa nd 1 Q108K:K40L 580 674 2 Q108K:K40L:T53A 585 677 0.11 3 Q108K:K40L:T53S 587 682 0.08 4 Q108K:K40L:T53V 597 672 0.13 Absolute quantum yield was measured on a Quantaurus-QY. As seen in the crystal structure of KL/retinal (Figure II-18), T51 and T53 interact via a water-mediated hydrogen bond. Mutation of T51 to valine would abolish this interaction, leading to a red-shift in wavelength, as discussed above. T53S could potentially incite a similar effect. From the KL/retinal structure, a water-mediated hydrogen bonding interaction between Q38 and Q128 is observed (Figure II-19). If T53S could form a tight hydrogen bond with Q38 (meaning that T53S has a higher propensity to interact with Q38 rather than 86 T51), it would disrupt the conserved water-mediated interaction between Q38 and Q128 that is presumably apparent in the KL/ThioFluor structure. The overall effect is the ejection of water, and thus reduced polarity, near the chromophore. Q128 3.0 Å 3.1 Å Q108K 2.9 Å 2.8 Å Q38 Figure II-19. Water mediated hydrogen bonding between Q38 and Q128. Coordinates obtained from PDB 4EXZ (hCRBPII-Q108K:K40L/retinal). If interrupting the water mediated hydrogen bonding between Q38 and Q128 leads to a red-shift in wavelength, one would hypothesize that a Q38 to nonpolar mutation should also result in a red-shift in wavelength. To this end, Q38 was mutated to phenylalanine in KL. Gratifyingly, a red-shift of 8 nm was observed in absorption (Table II-8). This further supports the theory that T53S mutation must be removing polarity in the pocket by way of interfering with the water mediated interaction between Q38 and Q128. Table II-8. Spectroscopic change as a result of Q38F mutation. Entry a Mutant λabs λem Φa nd 1 Q108K:K40L 580 674 2 Q108K:K40L:Q38F 588 680 0.13 Absolute quantum yield was measured on a Quantaurus-QY. 87 Although located at the entrance of the binding pocket, R58 is the last of the more polar residues that is within interacting distance of the chromophore. Replacement of polar charged arginine with aromatic tyrosine and phenylalanine in KL both led to a bathochromic shift in absorption and emission with ThioFluor (Table II-9). This is in line with previous studies that showed that encapsulation of the retinal chromophore in hCRBPII resulted in a red-shift in wavelength due to more effective shielding of the binding pocket.3 A wider screening of R58 mutations is detailed in Section II.6. Table II-9. Spectroscopic change as a result of mutating R58. Entry a Mutant λabs λem Φa nd 1 Q108K:K40L 580 674 2 Q108K:K40L:R58Y 591 681 0.09 3 Q108K:K40L:R58F 597 678 0.18 Absolute quantum yield was measured on a Quantaurus-QY. In summary, it should be noted that residues closer to the iminium, Q4 and T51, led to a larger red-shift than the T53 and Q38 residues farther away. As aforementioned and reiterated in Table II-10, both T51V and T53S red-shift the absorption and emission. As previously stated, T51V is also capable of monomerizing hCRBPII variants (see Chapter VI for a full discussion). To look at the combinatorial effect of T51V and T53S, the tetramutant Q108K:K40L:T51V:T53S was made. The additive effect from the addition of T51V and T53S to KL was quite apparent. Gratifyingly, red-shifts in absorption and emission from KL were substantial, at 47 nm and 23 nm, respectively (Table II-10, entry 4). Thus, the template Q108K:K40L:T51V:T53S was retained for 88 further protein engineering. Q4 mutations were not included, due to prior studies showing that the pKa of hCRBPII/retinal complexes are severely depressed when Q4 was substituted with phenylalanine.61, 62 Table II-10. Additive effects of T51 and T53 mutations. Entry a Mutant λabs λem Φa nd 1 Q108K:K40L 580 674 2 Q108K:K40L:T53S 587 682 0.08 3 Q108K:K40L:T51V 608 690 0.12 4 Q108K:K40L:T51V:T53S 627 697 0.15 Absolute quantum yield was measured on a Quantaurus-QY. II.6 Attempts to encapsulate the binding cavity Studies conducted by Dr. Wenjing Wang previously showed that an aromatic residue at position 58 leads to a bathochromic shift in absorption when retinal is used as a ligand,3 presumably due to more effective shielding of the binding pocket. To this end, R58F, R58Y, R58W and R58H mutations were introduced into the parent mutant Q108K:K40L:T51V:T53S (Table II-11). Table II-11. Spectroscopic changes as a result of mutating R58 in Q108K:K40L:T51V:T53S to aromatic residues. t1/2b Entry Mutant λabs λem Φa (min) 1 Q108K:K40L:T51V:T53S:R58F 617 691 0.12 184 2 Q108K:K40L:T51V:T53S:R58Y 622 691 0.11 226 3 Q108K:K40L:T51V:T53S:R58W 623 697 0.08 320 4 Q108K:K40L:T51V:T53S 627 697 0.15 146 5 Q108K:K40L:T51V:T53S:R58H 643 705 0.14 176 a Absolute quantum yield was measured on a Quantaurus-QY. bHalf-life based on the rate constant obtained from pseudo first order fitting; measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7. 89 Unexpectedly, of the four aromatic residues, only R58H led to a bathochromic shift (Table II-11, entry 5), while the other three aromatic mutations actually led to a minor blue-shift in absorption. However, unlike retinal, ThioFluor contains heteroatoms that can interact with surrounding residues through hydrogen bonding, hydrophobic interactions, pi-pi stacking, pi-cation interactions, ionic interactions, etc. Q108K 3.7 Å R58W Figure II-20. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor and space-filling representation. Electron density is shown at 1σ. From the crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor, it is quite obvious that the tryptophan at position 58 does not serve to sequester the binding pocket, in contrast to when retinal is used as a ligand. R58W sits directly on top of the N,N-dimethyl amino moiety of the bound ThioFluor. As shown in Figure II-20, the closest distance between the tryptophan and nitrogen is 3.7 Å, 90 well within van der Waals’ contact. This could suggest that ThioFluor’s amino group possesses a partial positive charge that can interact with R58W through an amino-aromatic stacking interaction, in which R58W serves as the hydrogen bonding acceptor. Similar interactions have been observed between the side chains of asparagine and glutamine with aromatic residues, where the amino group hydrogen atoms favor polar interactions with the pi-electrons of the aromatic. However, this interaction does not significantly affect wavelength, only a 6 nm hypsochromic shift from Q108K:K40L:T51V:T53S is obtained. It was hypothesized that substitution of R58 for histidine, could result in a red-shift if a stronger hydrogen bonding interaction between the positively charged histidine side chain and N,N-dimethyl amino moiety of the ligand was introduced. Fortuitously, Q108K:K40L:T51V:T53S:R58H shows a 16 nm shift in absorption and 8 nm shift in emission from Q108K:K40L:T51V:T53S. Unfortunately, crystallization of Q108K:K40L:T51V:T53S:R58H was not fruitful. Based on in silico modeling of R58H in the crystal structure of Q108K:K40L:T51V:T53S:R58W with ThioFluor, it is expected that histidine adopts a similar conformation as for tryptophan. Histidine could potentially act as a hydrogen bond donor to the N,N-dimethyl amino moiety. Localization of charge on the tail end of the bound chromophore would promote positive charge delocalization from the iminium, resulting in a red-shift in wavelength, as was observed upon the introduction of R58H to Q108K:K40L:T51V:T53S. 91 a. b. R58Y R58W R58Y 2.7 Å 3.2 Å T29 Figure II-21. a. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor (cyan) overlaid with Q108K:K40L:T51V:T53S:R58Y/ThioFluor (green). b. Water mediated hydrogen bonding between R58Y and T29. Next, R58F and R58Y were introduced into the same parent template Q108K:K40L:T51V:T53S. Both mutations actually led to a blue-shift in the absorption profile of the hCRBPII/ThioFluor complex (Table II-11). The crystal structure of Q108K:K40L:T51V:T53S:R58Y/ThioFluor shows a vastly different trajectory of residue 58 and chromophore orientation than that observed for Q108K:K40L:T51V:T53S:R58W/ThioFluor (Figure II-21a). As aforementioned, R58W flips in to stack with the N,N-dimethyl moiety of the ligand. However, tyrosine does not participate in aromatic interactions as often as tryptophan, and that is the case here. Tyrosine is flipped out of the pocket, providing no stacking interaction with the ligand. Alternatively, the phenolic group of R58Y participates in a water-mediated hydrogen bond with the hydroxyl side chain of T29 (Figure II-21b). As the ligand is not ‘held’ in place by the R58 mutation, this presumably causes the translocation and rotation of the chromophore. These changes, and 92 the fact that the binding pocket is no more sequestered, may be responsible for the negligible effect on wavelength. Flexible loop R58W R58F R58Y Figure II-22. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor (cyan) overlaid with Q108K:K40L:T51V:T53S:R58Y/ThioFluor (green) and Q108K:K40L:T53A:R58F/ThioFluor (pink). The crystal structure of Q108K:K40L:T51V:T53S:R58F/ThioFluor was not obtained, but Ms. Zahra Assar provided the structure of the similar variant Q108K:K40L:T53A:R58F with ThioFluor. Most notably, the flexibility of the loop at the entrance of the binding pocket was revealed, upon overlay of the structure with both Q108K:K40L:T51V:T53S:R58W/ThioFluor and Q108K:K40L:T51V: T53S:R58Y/ThioFluor (Figure II-22). In the R58F mutation, the loop moves such that the phenylalanine is within van der Waals contact of the ligand. A slight bathochromic shift (4 nm) in absorption is observed for Q108K:K40L:T53A:R58F from Q108K:K40L:T53A, suggesting again that R58F does not sequester the binding cavity. This suggestion is further supported by looking at the surface of the protein, which again shows a large hole at the binding cavity’s entrance. 93 After exhaustive screening of R58 for aromatic residues, we chose to explore the effect of other residues including mutation to glycine, leucine, proline, glutamine, and glutamic acid in the same parent template Q108K:K40L:T51V:T53S. As shown in Table II-12, mutation of R58 to the nonpolar aliphatic glycine and leucine has essentially no effect on the wavelength of absorption (Table II-12, entries 1 and 3). Because arginine is polar and charged, it is likely that R58 is exposed to the aqueous exterior of the protein, and projects minimal electrostatic potential onto the chromophore. It would then be expected that the removal of the arginine would not significantly affect wavelength, as is observed for mutation to leucine and glycine. Table II-12. Spectroscopic changes as a result of mutating R58 in Q108K:K40L:T51V:T53S. t1/2b Entry Mutant λabs λem Φa (min) 1 Q108K:K40L:T51V:T53S:R58G 621 699 0.09 249 2 Q108K:K40L:T51V:T53S 627 697 0.15 146 3 Q108K:K40L:T51V:T53S:R58L 629 697 0.14 205 4 Q108K:K40L:T51V:T53S:R58Q 632 698 0.13 225 5 Q108K:K40L:T51V:T53S:R58P 636 703 0.14 168 6 Q108K:K40L:T51V:T53S:R58E 636 700 0.14 240 a b Absolute quantum yield was measured on a Quantaurus-QY. Half-lives are based on the rate constant obtained from pseudo first order fitting; measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7. On the other hand, in the case of mutating R58 to the shorter polar uncharged glutamine and polar charged glutamic acid residues, bathochromic shifts in wavelength are observed from the parent Q108K:K40L:T51V:T53S. 94 R58Q and R58E mutations result in 5 nm and 9 nm, respectively (Table II-12, entries 4 and 6). It is possible that because these residues are shorter than arginine, and thus can flip into the pocket, they may have a higher propensity to interact with the ligand. In this case, polar residues could result in a similar orientation of the chromophore as that in Q108K:K40L:T51V:T53S:R58W, in which residue 58 is set up to interact with the N,N-dimethyl amino group of ThioFluor. The hydrogen bonding interaction of glutamine’s amino hydrogen atoms or glutamic acid’s carboxylic acid hydrogen atom (assuming the side chain is protonated) with the N,N-dimethyl amino group of ThioFluor would again promote positive charge delocalization along the chromophore. Nonetheless, further crystallographic analyses are required in order to test these hypotheses. Similar to R58E, mutation of R58 to proline results in a 9 nm red-shift in wavelength (Table II-12, entries 5), even though glutamic acid is much more polar than proline. It was previously shown in the crystal structure of Q108K:K40L:T53A:R58F/ThioFluor that the loop at the entrance of the binding cavity is flexible; it is likely that introduction of the rigid proline alters the loop’s location and potentially results in a different orientation of the chromophore. It is known that the distinctive cyclic structure of proline’s side chain provides conformational rigidity compared to other amino acids. In these studies, this theory may be supported by the rate of iminium formation in the Q108K:K40L:T51V:T53S:R58X series (fitting was based on pseudo first order rate kinetics as the product concentration cannot be accurately determined; 95 measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7). While many factors can influence the rate of iminium formation, it is fastest in the case of the rigidifying R58P mutation and slowest when the bulky tryptophan residue is introduced at the mouth of the binding cavity (Table II-11 and Table II-12). Presumably, in the case of R58P, the loop is permanently out of the way, leading to a more accessible binding cavity, and thus a faster rate of iminium formation. It can be concluded that R58 mutations do not significantly affect wavelength, in contrast to that seen for hCRBPII/retinal complexes; it appears that a residue at this position may have the propensity to interact with the ligand, due to the presence of an N,N-dimethyl amino moiety in its vicinity. As aforementioned, as observed in the crystal structure of Q108K:K40L:T51V:T53S:R58W, the tryptophan does not serve to further sequester the binding pocket, but appears to interact with the nitrogen atom of the chromophore, through a stacking amino-aromatic interaction. In the same context, R58Y again does not sufficiently enclose the binding pocket. However, there is a rotation of the chromophore, which could also affect the complexes absorption and emission profiles. Further crystallographic analysis would need to be conducted in order to determine the effect of other R58 mutations on the spectroscopic properties of the complexes, as they are shown to dramatically alter the chromophore orientation. Nonetheless, it was hypothesized that no mutation at R58 will be capable of sequestering the binding cavity, in order to 96 bathochromically shift the wavelength. Thus, we chose alternate positions for mutagenesis, in hopes to achieve the same effect. II.7 Removal of water and extensive packing of the chromophore leads to an enhanced red-shift in absorption and emission As shown in the crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor, there is a conserved water network from the hydroxyl group of Y19 to the thiophene sulfur atom of ThioFluor (Figure II-23a). As removal of polarity was previously shown to red-shift wavelength it was predicted that mutation of Y19, to a residue less likely to provide this hydrogen bonding would result in removal of this water network. Tryptophan was introduced at Y19 for this purpose. A red-shift of 30 nm is observed with the introduction of Y19W to Q108K:K40L:T51V:T53S:R58W (Table II-13, entry 2). a. R58W b. R58W 3.0 Å Y19 2.2 Å 2.9 Å Y19W 2.5 Å Figure II-23. a. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor, highlighting the water mediated network from Y19 to ThioFluor. b. Crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor shows this water network is abolished. 97 The crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor (obtained by Ms. Zahra Assar) showed that substitution of tryptophan at Y19 did indeed abolish the water Q108K:K40L:T51V:T53S:R58W/ThioFluor network (Figure present II-23b), with no in other significant changes of the chromophore’s orientation or any of the other residues. However, there could be two reasons for this red-shift in wavelength. It could be solely due to the removal of the water network localizing polarity on the thiophene core or to the replacement of this water network with a polarizable residue, which would more efficiently delocalize charge along the ligand. Table II-13. Spectroscopic changes as a result of Y19W mutation. Entry a Mutant λabs λem Φa 1 Q108K:K40L:T51V:T53S:R58W 623 697 0.08 2 Q108K:K40L:T51V:T53S:R58W:Y19W 653 719 0.10 3 Q108K:K40L:T51V:T53S:R58Y 622 691 0.11 4 Q108K:K40L:T51V:T53S:R58Y:Y19W 635 698 0.16 5 Q108K:K40L:T51V:T53S:R58H 643 705 0.14 6 Q108K:K40L:T51V:T53S:R58H:Y19W 642 705 0.15 Absolute quantum yield was measured on a Quantaurus-QY. The effect of the Y19W mutation was also tested in the templates with R58Y and R58H (Table II-13, entries 4 and 6). A significantly smaller red-shift of 13 nm was observed in adding Y19W to Q108K:K40L:T51V:T53S:R58Y, and addition of Y19W to Q108K:K40L:T51V:T53S:R58H unfortunately led to no change. For this reason, Q108K:K40L:T51V:T53S:R58W:Y19W was utilized for further protein engineering in an attempt to red-shift absorption and emission wavelengths. Introduction of four mutations to the parent KL protein 98 (Q108K:K40L:T51V:T53S:R58W:Y19W, λabs = 653 nm, λem = 719 nm) leads to the largest change in absorption (Δabs = 73 nm) and emission (Δem = 45 nm) from Q108K:K40L (λabs = 580 nm, λem = 674 nm) thus far. R58W A33 F16 Y19W L77 Figure II-24. Residues mutated (shown in green) in an attempt to encapsulate the binding cavity. Crystal structure is of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor. With the red-shifted hexamutant in hand, we sought to encapsulate the binding cavity via the introduction of a bulky tryptophan residue at L77, A33 and F16 to Q108K:K40L:T51V:T53S:R58W:Y19W (Figure II-24). Fortuitously, both L77W and A33W mutations red-shift the wavelength, 5 nm and 13 nm, respectively (Table II-14, entries 3 and 4). Unfortunately, crystallization of Q108K:K40L:T51V:T53S:R58W:Y19W:A33W was not fruitful. Presumably though, A33W sufficiently covers the binding cavity (see Figure II-26 for a similar mutant crystallized with ThioFluor). 99 While direct encapsulation of the pocket was presumably achieved through the introduction of tryptophans at L77 and A33, located at the interface between the aqueous media and protein cavity, it was also suspected that introducing a bulky residue near R58W could force this residue to flip out and result in a similar effect. F16 is located at the interior pocket, but it can be envisioned that mutation of F16 to tryptophan would lead to steric clash of F16W with R58W, and may be able to flip R58W outward, in effect sequestering the binding pocket. Unfortunately, room temperature expression of Q108K:K40L:T51V:T53S:R58W:Y19W:F16W did not result in soluble protein. Conversely, Q108K:K40L:T51V:T53S:R58W:F16Y did provide soluble protein, perhaps because tyrosine is more similar to the wild-type phenylalanine than tryptophan. A red-shift from Q108K:K40L:T51V:T53S:R58W:Y19W of 11 nm was observed via the introduction of F16Y (Table II-14, entry 2). The possibilities for this red-shift are two-fold. It is possible that R58W is flipped outward as a result of introducing the longer tyrosine at F16, or the tyrosine phenol may serve as a hydrogen bond acceptor to the N,N-dimethyl amino moiety of ThioFluor. Table II-14. Spectroscopic changes as a result of introducing bulky residues at the entrance of the binding cavity. Entry a Mutant λabs λem Φa 1 Q108K:K40L:T51V:T53S:R58W:Y19W 653 719 0.10 2 Q108K:K40L:T51V:T53S:R58W:Y19W:F16Y 664 726 0.05 3 Q108K:K40L:T51V:T53S:R58W:Y19W:L77W 658 711 0.10 4 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W 666 724 0.14 Absolute quantum yield was measured on a Quantaurus-QY. 100 Because the introduction of polarizable tryptophan residues at R58 and Y19 lead to a red-shift, from Q108K:K40L:T51V:T53S (λabs = 627 nm, λem = 697 nm) to Q108K:K40L:T51V:T53S:R58W:Y19W (λabs = 653 nm, λem = 719 nm), we chose to investigate the effect of tryptophan in other positions. It was also predicted that these residues might provide tight packing of the chromophore, which should lead to chromophore rigidification and restriction. To this end, tryptophan mutations were singly made at each residue in the protein cavity using the template Q108K:K40L:T51V:T53S:R58W:Y19W:A33W, along the entire chromophore’s length. Q38 K40L F57 Q4 T53 T51V T29 L119 S76 Y60 F64 I25 Figure II-25. Residues where tryptophan was introduced in an attempt to redshift wavelength are shown in green. Crystal structure is of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor. Residues mutated in this study (Figure II-25) were T51, K40, Q38, I25, Y60, S76, L119, T53, T29, F57, F64 and Q4. Mutation of residues located at the mouth of the binding cavity, including S76, T29 and F57 to tryptophan led to negligible changes in the wavelength of the iminium (Table II-15, entries 7, 10 101 and 11). Presumably, the binding cavity Q108K:K40L:T51V:T53S:R58W:Y19W:A33W is in the parent completely mutant sequestered, rendering the introduction of new tryptophans at the entrance to the binding cavity as obsolete. Table II-15. Spectroscopic changes a result of mutating active site residues to tryptophan. Entry a Mutant λabs λem Φa nd 1 Q108K:K40L:T51W:T53S:R58W:Y19W:A33W 603 689 2 Q108K:K40H:T51V:T53S:R58W:Y19W:A33W 612 694 0.19 3 Q108K:K40W:T51V:T53S:R58W:Y19W:A33W 627 701 0.17 4 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Q38W 641 712 0.10 5 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:I25W 646 711 0.09 6 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Y60W 661 718 0.10 7 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:S76W 665 721 0.14 8 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L119W 665 729 nd 9 Q108K:K40L:T51V:T53W:R58W:Y19W:A33W 666 724 nd 10 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:T29W 666 726 0.12 11 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:F57W 667 722 0.12 12 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:F64W 669 727 0.12 13 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Q4W 705 744 nd Absolute quantum yield was measured on a Quantaurus-QY. Only I25W has an effect on the wavelength; Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:I25W is 20 nm blue-shifted in comparison to Q108K:K40L:T51V:T53S:R58W:Y19W:A33W (Table II-15, entry 5). While efforts to crystallize Q108K:K40L:T51V:T53S:R58W:Y19W:A33W/ThioFluor were not successful, a crystal structure 102 of Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E/ThioFluor was achieved. Assuming, the two structures are similar, it is apparent that the introduction of a large tryptophan residue at residue 25 would force the conformational change of a neighboring residue to avoid steric clashing (Figure II-26). Indeed, upon in silico mutation of I25 to tryptophan using the crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E/ThioFluor as a template, all rotamers of I25W sterically clash with A33W. This may force A33W to adopt a different conformation, negating the sequestering effect it delivered before. R58W A33W Y19W I25 Figure II-26. Crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E/ThioFluor. Mutation of I25 to tryptophan would sterically clash with A33W. Mutation of Q38, Y60, K40 and T51 to tryptophan each led to a blue-shift in absorption wavelength of 25 nm, 5 nm, 39 nm and 63 nm from the parent mutant Q108K:K40L:T51V:T53S:R58W:Y19W:A33W, respectively (Table II-15, entries 4, 6, 3 and 1). Interestingly, the most hypsochromic shifts were observed via introduction of tryptophan at the residues closest to the iminium, at positions 103 K40 and T51 (the same result was also achieved via introduction of histidine at K40; Table II-15, entry 2). Presumably, the introduction of a polarizable tryptophan in close proximity to the iminium localizes the charge, leading to a blue-shift in wavelength. While some exchanges led to significant hypsochromic shifts in wavelength, numerous substitutions including tryptophan substitution at L119, T53 and F64, had a negligible effect on the wavelength (Table II-15, entries 8, 9 and 12). As these residues are at the interior of the protein pocket and size was significantly altered for each point mutation, without crystal structures it is not possible to explain the null effect on wavelength. As a result of increased steric bulk, it is possible that the chromophore has changed conformation, or the changes have impeded the interaction of other residues with the chromophore as previously described. The only substitution of tryptophan that led to a significant bathochromic shift in wavelength was at Q4 (Table II-15, entry 13). Intuitively, the red-shift would be attributable to the removal of polarity by removing the water mediated network between Q4 and the iminium. However, there is another possibility. From previous studies with hCRBPII/retinal it was observed that the removal of the water mediated hydrogen bond from Q4 to a residue with a non-hydrogen bonding side chain leads to a thermodynamically stable trans isomer of the iminium, rather than the cis iminium. It is quite possible that the addition of Q4W to Q108K:K40L:T51V:T53S:R58W:Y19W:A33W also leads to a trans isomer of 104 the iminium. This is supported by the fact that Q4W dramatically alters the pKa of the iminium. Q108K:K40L:T51V:T53S:R58W:Y19W:A33W/ThioFluor exhibits a high pKa of 9.3. Conversely, at a neutral pH of 7, no iminium species is apparent by UV-Vis for Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Q4W/ThioFluor. After exhaustive tryptophan screening, one last effort was undertaken to red-shift the absorption and emission profiles of the hCRBPII/ThioFluor mutants through combining residues that red-shifted wavelength individually. To this end, a few mutants were made, as detailed in Table II-16. The most red-shifted hCRBPII mutant coupled with ThioFluor was Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Q4W:F16Y with found an to be absorption maxima of 705 and emission maxima of 744 nm. From Q108K:K40L (λabs = 580 nm, λem = 674 nm), a change of 125 nm is obtained in absorption and 70 nm in emission. Table II-16. Additive effect of F16Y, leading to a final bathochromic shift in wavelength. Entry Mutanta λabs λem Φb 1 KL:T51V:T53S:R58W:Y19W:A33W:F16Y 670 735 0.06 2 KL:T51V:T53S:R58W:Y19W:L77W:F16Y 672 723 0.06 3 KL:T51V:T53S:R58W:Y19W:A33W:Q4W:F16Y 705 744 nd a KL = Q108K:K40L. bAbsolute quantum yield was measured on a QuantaurusQY. II.7.1 Investigation of combinatorial residue effects on the wavelength of Q108K:K40L:T51V:T53S:R58W:Y19W As previously discussed, the introduction of four mutations to the parent KL protein (Q108K:K40L:T51V:T53S:R58W:Y19W, λabs = 653 nm, λem = 719 nm) 105 leads to a red-shift in absorption (Δabs = 73 nm) and emission (Δem = 45 nm). By mutating residues back to their wild type, we sought to investigate which residues are the most necessary to red-shift wavelength. Interestingly, when introduced to Q108K:K40L, the addition of T51V and T53S (Table II-17, entry 4) yields a red-shift of 47 nm, while the addition R58W and Y19W (Table II-17, entry 1) leads to only a 25 nm red-shift in absorption. Introduction of R58W or Y19W individually to Q108K:K40L:T51V:T53S do not substantially alter the wavelength of absorption (Table II-17, entries 3 and 5). However, their combination leads to the red-shifted hexamutant Q108K:K40L:T51V:T53S:R58W:Y19W (Table II-17, entry 7). Table II-17. Deducing the origins of the bathochromic shift in Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor. Entry a Mutant λabs λem Φa 1 Q108K: K40L: 2 Q108K: K40L: T51V: 3 Q108K: K40L: T51V: T53S: 4 Q108K: K40L: T51V: T53S 5 Q108K: K40L: T51V: T53S: Y19W 628 699 0.12 6 Q108K: K40L: T53S: R58W: Y19W 635 704 0.16 7 Q108K: K40L: T53S: R58W: Y19W 653 719 0.10 T51V: R58W: Y19W 605 691 0.14 R58W: Y19W 612 691 0.07 R58W 623 697 0.08 627 697 0.15 Absolute quantum yield was measured on a Quantaurus-QY. The crystal structures of Q108K:K40L:T51V:T53S:R58W/ThioFluor and Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor may provide insight into the origin of this synergistic combination that leads to a red-shift in wavelength. In the crystal structure, two conserved hydrogen bonding water mediated networks are 106 apparent. First, the hydroxyl of T53S hydrogen bonds with both the indole side chain of R58W and the amide moiety of Q38 through a water mediated hydrogen bond (Figure II-27). Secondly, Y19 interacts with the sulfur atom of ThioFluor through an extensive water network (Figure II-23a). These interactions presumably dictate the chromophore’s orientation. Q38 Q128 2.7 Å R58W 2.9 Å 2.7 Å T53S Figure II-27. Crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor. The water mediated hydrogen bonding between T53S and R58W is highlighted. As seen in the crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W/ ThioFluor and described previously (Section II.7, Figure II-23), the introduction of Y19W abolishes the water network from Y19 to the sulfur atom of ThioFluor, presumably leading to a red-shift in wavelength. It is predicted that in the absence of the R58W mutation, ThioFluor cannot adopt an orientation in which the Y19W mutation is able to eliminate this water network. II.8 Localization of charge leads to a blue-shift in absorption and emission In studies with hCRBPII/retinal, Dr. Wenjing Wang previously demonstrated that while an even distribution of electrostatic potential across the 107 polyene led to a bathochromic shift in absorption, it was localizing the cation on the iminium nitrogen that led to a blue-shift.3 Similarly, we sought to introduce negatively polarized residues near the iminium. It is widely accepted that hydrophobic interactions can influence protein folding and stability.63-65 While tryptophans were introduced with relative ease into the hCRBPII sequence, it was more of a challenge to substitute acidic residues for otherwise non-charged or less polar residues. The primary challenges observed in introducing acidic residues were the ability to obtain soluble protein during protein expression, protein stability and the formation of domain-swapped dimers. K40L T51 L115 L117 Q108K F64 L93 Figure II-28. Residues chosen for introduction of an acidic residue, in order to interact with the iminium, based on the crystal structure of Q108K:K40L/retinal. Coordinates obtained from PDB 4EXZ. We chose to start again from Q108K:K40L, introducing acidic residues which should reside in close proximity to the iminium, namely at residues K40, 108 T51, L117, L115, F64 and L93 (Figure II-28). Q108K:K40L:L115E, Q108K:K40L:F64E and Q108K:K40L:L93E led to no soluble protein expression, so exploration at these positions were not followed. Incubation of ThioFluor with Q108K:K40L:L117E, Q108K:K40E and Q108K:K40D led to no substantial change when monitored by UV-Vis; only precipitation was apparent. Only Q108K:K40L:T51E yielded an observable iminium species(λabs = 592 nm, λem = 685 nm). Interestingly, a red-shift of 12 nm from Q108K:K40L (λabs = 580 nm, λem = 674 nm) was observed. This proves counterintuitive, as the introduction of polarity near the iminium is expected to result in a blue-shift in wavelength. It was hypothesized that the introduction of an acidic residue near the basic active site lysine may cause the formation of a salt bridge, preventing attack of the lysine to ligand to form an iminium. To test this, Q108K:K40E was incubated with retinal. PSB forms quickly, and further acidification results in a clear iminium species. As retinal is capable of forming the iminium with Q108K:K40E, it was assumed that ThioFluor should be able to also. Therefore, it was hypothesized that the pKa of the protein/ThioFluor complexes are at least two units lower than physiological pH. However, iminium cannot be observed upon lowering the pH of the protein/ThioFluor complexes, due to the sensitivity of the protein to acid (only precipitation is observed in these cases). As discussed above, we were not able to conclude anything based on the introduction of an acidic residue in the Q108K template. Thus, we chose to introduce an acidic residue into one of the red-shifted mutants, in order to test 109 whether a carboxylate can localize charge on the iminium, leading to a blue-shift in wavelength. To this end, L117 was mutated to both aspartic acid and glutamic acid in Q108K:K40L:T51V:T53S:R58W:Y19W and Q108K:K40L:T51V:T53S:R58W:Y19W:A33W, both of which when bound with ThioFluor produced a clearly observable red-shifted iminium. Fortuitously, the complexes were stable, with the complex of Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E surviving to a pH as low as 1. Noteworthy, large hypsochromic shifts were observed under circumstances when the carboxylate side chain is either protonated or deprotonated (these mutants will be discussed in more detail in Chapter IV for their application toward the development of a ratiometric pH sensor). This means that the introduction of a counter anion blue-shifts the wavelength more than simply just introducing polarity (consider L117E-COO– versus L117E-COOH). Table II-18. Spectroscopic changes observed upon the introduction of an acidic residue at L117. Entry Mutant λabs1 λabs2a Φb 653 0.10 1 Q108K:K40L:T51V:T53S:R58W:Y19W 2 Q108K:K40L:T51V:T53S:R58W:Y19W:L117E 558 614 0.31 3 Q108K:K40L:T51V:T53S:R58W:Y19W:L117D 525 618 0.30 4 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W 5 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E 539 628 0.28 6 Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117D 510 645 0.16 666 0.14 a λabs1 is absorbance maxima at basic pH and λabs2 is absorbance maxima at acidic pH. bAbsolute quantum yield was measured on a Quantaurus-QY. With the knowledge that a blue-shift in wavelength can be achieved via the localization of charge on the iminium, we sought to design a protein template that 110 is similar in wavelength (or at least not significantly red-shifted) to Q108K:K40L and produces an iminium at physiological pH or is acid stable enough to observe the iminium by lowering the pH of the solution. II.8.1 Design of blue-shifted hCRBPII/ThioFluor complexes that promote iminium formation From the UV-vis of Q108K:K40L/ThioFluor it is quite obvious that the complex suffers from either a severely suppressed pKa or extinction coefficient of the iminium. In preliminary studies, introduction of T53A or T53S led to a mix of hCRBPII monomer and dimer. Interestingly, the spectroscopic properties of these UV-Vis of Q108K:K40L:T53A/ThioFluor 607 nm 0.2 585 nm 0 300 500 600 Wavelength (nm) 0.5 Monomer Dimer k = 0.0128 min-1 t1/2 = 54 min R2 = 0.997 0.03 Abs. at 587 nm (a. u.) 0.3 0.2 587 nm 0.1 500 600 Wavelength (nm) 700 k = 0.0206 min-1 t1/2 = 34 min R2 = 0.990 0.1 0 PSB Formation of Q108K:K40L:T53SMonomer/ThioFluor 50 100 150 200 250 300 350 400 Time (min) PSB Formation of Q108K:K40L:T53SDimer/ThioFluor 0.5 0.09 0.06 k = 0.00845 min-1 t1/2 = 82 min R2 = 0.995 0.03 0 400 0.2 50 100 150 200 250 300 350 400 Time (min) 0.12 617 nm 0.3 0 0 700 0.4 Abs. (a. u.) 0.06 0 400 UV-Vis of Q108K:K40L:T53S/ThioFluor 0 300 0.09 0.4 Abs. at 617 nm (a. u.) Abs. (a. u.) 0.3 0.1 PSB Formation of Q108K:K40L:T53ADimer/ThioFluor Abs. at 607 nm (a. u.) Monomer Dimer PSB Formation of Q108K:K40L:T53AMonomer/ThioFluor 0.12 Abs. at 585 nm (a. u.) 0.4 0.4 0.3 k = 0.0206 min-1 t1/2 = 34 min R2 = 0.983 0.2 0.1 0 0 200 400 600 Time (min) 800 1000 0 100 200 300 400 500 600 700 800 Time (min) Figure II-29. Iminium formation of Q108K:K40L:T53A and Q108K:K40L:T53S monomers and dimers. From left to right are the UV-Vis of the thermodynamic protein/ThioFluor complex, iminium formation as a function of time for the monomer and dimer. Data is shown in black, while curve fitting is a dotted red line (fit to pseudo first order kinetics, measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7). 111 UV-Vis of KLA:R58L/ThioFluor PSB Formation of KLA:R58LMonomer/ThioFluor 0.5 Abs. at 579 nm (a. u.) Abs. (a. u.) 0.4 0.3 579 nm 0.2 0.1 0 300 0.25 0.5 0.2 0.4 609 nm 0.15 Abs. at 609 nm (a. u.) Monomer Dimer PSB Formation of KLA:R58LDimer/ThioFluor k = 0.0041 min-1 t1/2 = 169 min R2 = 0.994 0.1 0.05 0 400 500 600 Wavelength (nm) 200 UV-Vis of KLA:R58F/ThioFluor 0.3 k = 0.0201 min-1 t1/2 = 35 min R2 = 0.988 0.2 0.1 0 0 700 0.3 400 600 Time (min) 800 1000 0 200 400 600 Time (min) 800 1000 PSB Formation of KLA:R58FMonomer/ThioFluor 0.3 Monomer Abs. at 589 nm (a. u.) 0.2 0.1 0 300 0.2 k = 0.0047 min-1 t1/2 = 147 min R2 = 0.992 0.1 0 400 500 600 Wavelength (nm) 0 700 200 UV-Vis of KLA:Q38A/ThioFluor 0.3 800 1000 0.2 571 nm 0.1 0.25 0.15 min-1 k = 0.0119 t1/2 = 58 min R2 = 0.977 0.1 0.05 0 400 500 600 Wavelength (nm) PSB Formation of KLA:Q38ADimer/ThioFluor 0.3 600 nm Abs. at 571 nm (a. u.) Abs. (a. u.) Monomer Dimer 400 600 Time (min) PSB Formation of KLA:Q38AMonomer/ThioFluor 0.2 0 300 no dimer was obtained during KLA:R58F protein expression at room temperature Abs. at 600 nm (a. u.) Abs. (a. u.) 589 nm k = 0.0831 min-1 t1/2 = 8 min R2 = 0.984 0.15 0.1 0.05 0 0 700 0.2 UV-Vis of KLA:Q38M/ThioFluor 50 100 150 200 250 300 350 400 Time (min) 0 20 40 60 80 Time (min) 100 120 PSB Formation of KLA:Q38MMonomer/ThioFluor 0.3 0.25 Monomer Abs. at 584 nm (a. u.) Abs. (a. u.) 584 nm 0.2 0.1 0 300 0.2 0.15 no dimer was obtained during KLA:Q38M protein expression at room temperature k = 0.0132 min-1 t1/2 = 53 min R2 = 0.926 0.1 0.05 0 400 500 600 Wavelength (nm) 700 0 200 400 600 Time (min) 800 1000 Figure II-30. Iminium formation of KLA:R58L(R58F) and KLA:Q38A(M) monomers and dimers. From left to right are the UV-Vis of the thermodynamic protein/ThioFluor complex, iminium formation as a function of time for the monomer and dimer. Data is shown in black, while curve fitting is a dotted red line (fit to pseudo first order kinetics, measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7). KLA = Q108K:K40L:T53A. 112 are vastly different. Under identical conditions, the dimers of Q108K:K40L:T53A and Q108K:K40L:T53S both led to a significantly larger amount of iminium formation than their associated monomers (Figure II-29); additionally, the absorption wavelength of the dimer is more red than that of the monomer for both proteins. Interestingly, a difference in the rates of iminium formation is apparent (Figure II-29), in which the dimer forms iminium substantially faster with ThioFluor for both Q108K:K40L:T53A and Q108K:K40L:T53S. While these structures have not been obtained, it is likely that the chromophore’s iminium experiences different local environments, leading to either a difference in extinction coefficient or pKa. It should be noted that we did not measure pKas or extinction coefficients, so the origin of difference between monomer and dimer cannot be determined at this time. This information may prove invaluable in future studies to develop differentially binding probes to hCRBPII monomers and dimers in vivo. It is hypothesized that the extent of iminium formation may be due to different conformations of the iminium (cis or trans), which has previously been shown to alter the iminium pKa in both CRABPII and hCRBPII when bound to retinal. However, further crystallographic studies should be undertaken to understand the true difference between the holohCRBPII monomers and dimers. From the Q108K:K40L:T53A (KLA) mutant, it was observed that further removal of polarity at R58 and Q38 led to a further increase in the extent of iminium formation, although the reason is not apparent at this time. We chose to 113 mutate R58 to the nonpolar residues leucine and phenylalanine. At the same time, Q38 was replaced with alanine, methionine and phenylalanine (unfortunately, introduction of Q38F led to low soluble protein expression). As seen in Figure II-30, both Q108K:K40L:T53A:R58L-monomer and Q108K:K40L:T53A:R58F-monomer lead to a more intense iminium species, as did Q108K:K40L:T53A:Q38M-monomer and Q108K:K40L:T53A:Q38A-monomer. It should be noted that the mutation of Q38 to both alanine and methionine led to a less stable protein, as the complex begins to precipitate after formation of the iminium with ThioFluor. Dimer/ThioFluor is also presented for comparison purposes. In all cases, the dimer/ThioFluor iminium is more red-shifted than the monomer/ThioFluor iminium. Additionally, the rate of iminium rate formation for the dimer is faster than the monomer. Final combination of iminium promoting mutations (at residues 53, 58, and Q38) were introduced to yield Q108K:K40L:T53A:R58L:Q38M and Q108K:K40L:T53A:R58L:Q38F. Only monomer was obtained during protein expression at room temperature. R58L was retained in lieu of R58F because the complex of Q108K:K40L:T53A:R58L was more blue-shifted than Q108K:K40L:T53A:R58F by 10 nm. As seen in Figure II-31, both pentamutants produced a clearly observable iminium species for the monomer-bound ThioFluor. Fortuitously, Q108K:K40L:T53A:R58L:Q38F (λabs = 577 nm, λem = 671 nm). is of approximately the same wavelength as Q108K:K40L (λabs = 580 114 nm, λem = 674 nm), rendering the effect of the installed substitutions on the wavelength negligible as was desired (Table II-19). UV-Vis of KLA:R58L:Q38F/ThioFluor 0.4 PSB Formation of KLA:R58L:Q38FMonomer/ThioFluor 0.4 Abs. at 577 nm (a. u.) 577 nm Abs. (a. u.) 0.3 0.2 0.1 0 300 0.3 k = 0.0082 min-1 t1/2 = 85 min R2 = 0.992 0.2 0.1 0 400 500 600 Wavelength (nm) 0 700 UV-Vis of KLA:R58L:Q38M/ThioFluor 200 400 600 Time (min) 800 1000 PSB Formation of KLA:R58L:Q38MMonomer/ThioFluor 0.4 0.4 Abs. at 582 nm (a. u.) 582 nm Abs. (a. u.) 0.3 0.2 0.1 0 300 0.3 k = 0.0056 min-1 t1/2 = 124 min R2 = 0.995 0.2 0.1 0 400 500 600 Wavelength (nm) 0 700 200 400 600 Time (min) 800 1000 Figure II-31. Iminium formation of KLA:R58L:Q38F and KLA:R58L:Q38M monomer. From left to right are the UV-Vis of the thermodynamic protein/ThioFluor complex, iminium formation as a function of time. Data is shown in black, while curve fitting is a dotted red line (fit to pseudo first order kinetics, measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7). KLA = Q108K:K40L:T53A. It should be noted that the introduction of Q128L, yielding Q108K:K40L:T53A:R58L:Q128L, provided the same effect, however, the protein was obtained with a lower expression yield. Presumably mutation of either Q38 or Q128 to a nonpolar residue removes the conserved water mediated network between the two residues. A large set of other mutations were tested, however, 115 for simplicity and with the goal of blue-shifting wavelength, we continued with the template Q108K:K40L:T53A:R58L:Q38F for further protein engineering. The spectroscopic properties of these mutants are reported in Table II-19. It should be reiterated that in order to understand the effect of these mutations, it would be essential to collect pKa and extinction data on these mutants. Table II-19. Spectroscopic property of protein mutants designed in order to increase iminium formation. Entry a Mutant λabs λem Φa 1 Q108K:K40L:T53A:Q38A 571 667 0.14 2 Q108K:K40L:T53A:R58L:Q38A 575 669 0.13 3 Q108K:K40L:T53A:R58L:Q38F 577 671 0.08 4 Q108K:K40L:T53A:R58L:Q38F:Y19W 577 667 0.08 5 Q108K:K40L:T53A:R58L:Y19F 578 672 0.10 6 Q108K:K40L:T53A:R58L:Q38F:S76A 578 670 7 Q108K:K40L:T53A:R58L:Q128L 579 670 0.10 8 Q108K:K40L:T53A:R58L 579 672 0.09 9 Q108K:K40L 580 674 10 Q108K:K40L:T53A:R58L:Q38M 582 674 0.11 11 Q108K:K40L:T53A:R58L:A33W 584 672 0.06 12 Q108K:K40L:T53A:Q38M 584 677 13 Q108K:K40L:T53A 585 677 0.11 14 Q108K:K40L:T53A:R58W 586 678 0.11 15 Q108K:K40L:T53V:R58L 587 678 0.16 16 Q108K:K40L:T53A:R58F 589 681 0.10 17 Q108K:K40L:T53A:R58L:Q38F:Y60F 590 675 0.19 18 Q108K:K40L:T53A:R58L:Q38F:F16Y 594 677 0.13 19 Q108K:K40L:T51C:T53A:R58L:Q38F 594 677 0.12 20 Q108K:K40L:T53A:R58L:Q38F:L77W 597 678 0.15 Absolute quantum yield was measured on a Quantaurus-QY. 116 nd nd nd II.8.2 Exploring the additive effects of acidic residues on the absorption and emission wavelength of hCRBPII/ThioFluor Having in hand a blue-shifted hCRBPII mutant capable of forming an iminium with ThioFluor at physiological pH, we chose to introduce acidic residues in an attempt to blue-shift its wavelength. It was quickly realized that the introduction of an acidic residue at residue 40 to Q108K:K40L:T53A:R58L:Q38F would not prove successful, as soluble protein was not obtained from either Q108K:K40D:T53A:R58L:Q38F or Q108K:K40E:T53A:R58L:Q38F. Table II-20. Spectroscopic changes as a result of introducing one acidic residue near the iminium. λabs λem Φa 1 Q108K:K40L:T53A:R58Y:Q38F:Q4F:L117E 510 653 0.20 2 Q108K:K40D:T53A:R58L:Q38F:Q4F 546 649 0.21 3 Q108K:K40D:T53A:R58Y:Q38F:Q4F 551 641 0.26 4 Q108K:K40E:T53A:R58L:Q38F:Q4F 575 671 0.15 5 Q108K:K40L:T51D:T53A:R58Y:Q38F:Q4F 576 665 0.25 6 Q108K:K40L:T53A:R58L:Q38F:Q4F:V62E 584 656 0.15 7 Q108K:K40L:T53A:R58L:Q38F:Q4F 616 691 0.10 Entry a Mutant Absolute quantum yield was measured on a Quantaurus-QY. Fortuitously though, our lab has observed the general trend that the Q4F mutation is capable of increasing protein expression yield while also resulting in a higher level of soluble protein. Thus, acidic residues were introduced to the Q108K:K40L:T53A:R58L:Q38F:Q4F hCRBPII hexamutant. To our delight soluble protein was obtained in most of the desired proteins. It should be noted that in some cases R58Y was included instead of R58L, as R58Y led to better soluble 117 protein expression. However, in comparing Q108K:K40D:T53A:R58L:Q38F:Q4F (Table II-20, entry 2) and Q108K:K40D:T53A:R58Y:Q38F:Q4F (Table II-20, entry 3), it is obvious that the substitution at R58 does not drastically alter the wavelength in these mutants, rendering them comparable in terms of spectroscopic properties. K40L T51V L117 V62 Figure II-32. Positions for introducing an acidic reside to blue-shift wavelength. Based on the crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor. Acidic residues were introduced at K40, T51, V62 and L117 (Figure II-32). Results are detailed Table II-20. The introduction of L117E provides the largest hypsochromic shift from Q108K:K40L:T53A:R58L:Q38F:Q4F, an astounding 106 nm in absorption maxima and 38 nm in emission maxima. The crystal structure of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor (R58W and Y19W seem to rigidify the chromophore, such that ligand density is well-ordered; a structure was not successfully obtained without these two tryptophans) was obtained and solved by Mr. Alireza Ghanbarpour. If the chromophore adopts the same 118 orientation in each mutant (and exists as the same isomer of the iminium), L117 would be closest to the iminium (Figure II-33); and V62 would be furthest from the iminium when mutated to glutamic acid (mutagenesis suggests the glutamic acid must be oriented away from the iminium to avoid steric clash with the methyl group). It would appear that the distance from the glutamic acid residue to the iminium is roughly correlated with how much of a hypsochromic shift is observed with its introduction. While crystallographic data would provide concrete evidence, it is hypothesized the larger shift is observed as the residue resides closer to the iminium or is involved in a tighter hydrogen bonding network with the iminium. a. K40E c. b. T51E 4.7 Å L117E 3.4 Å 3.8 Å Figure II-33. a. Crystal structure of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor, highlighting the acidic residues tested in this series of protein mutants. b. – c. The residues L117 and K40 (shown in grey) were mutated to glutamic acid in pymol. Additionally, higher quantum yields are obtained with the introduction of an acidic residue; every mutation has a higher quantum yield than that of the parent template. It could be that the quantum yield is higher solely based on the fact that these mutants are more blue-shifted than their parent protein. From the crystal 119 structure of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor, it is also apparent that the residues are within distance from the iminium to interact via a water mediated network. It is possible that the presence of an acidic residue rigidifies ThioFluor, which could lead to an enhancement of quantum yield. Table II-21. Spectroscopic changes as a result of introducing two acidic residues near the iminium. Entry a Mutant λabs λem Φa 1 Q108K:K40L:T51D:T53A:R58Y:Q38F:Q4F:L117E 501 613 0.49 2 Q108K:K40D:T53A:R58Y:Q38F:Q4F:L117E 505 653 0.13 3 Q108K:K40E:T53A:R58L:Q38F:Q4F:V62E 521 656 0.24 4 Q108K:K40D:T51V:R58L:Q38F:Q4F:V62E 528 621 0.65 5 Q108K:K40D:T51V:T53A:R58L:Q38F:Q4F:V62E 571 670 0.21 6 Q108K:K40L:T53A:R58L:Q38F:Q4F 616 691 0.10 Absolute quantum yield was measured on a Quantaurus-QY. We next sought to investigate whether the introduction of two acidic residues concurrently could further shift the wavelength or increase the quantum yield. Results are shown in Table II-21. It should be noted that T51V was included in some mutants in order to obtain soluble monomer during protein expression (see Chapter VI for a more detailed description); Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E produced only dimer at both room temperature expression, as well as at 16 °C (we know from previous studies that dimer formation is temperature dependent).66 Similarly, the combination of K40D and T51D also led to mainly dimer formation during expression of Q108K:K40D:T51D:T53A:R58Y:Q38F:Q4F. 120 Again, inclusion of L117E led to the most blue-shifted mutants (Table II21, entries 1 and 2). Interestingly, while absorption was not affected substantially, Q108K:K40L:T51D:T53A:R58Y:Q38F:Q4F:L117E was 40 nm blue-shifted in emission as compared to the mutant without T51D (Table II-20, entry 1). It is possible that we have reached the threshold at which the absorption can be varied. Fortuitously, Q108K:K40L:T51D:T53A:R58Y:Q38F:Q4F:L117E (Table II21, entry 1) and Q108K:K40D:T51V:R58L:Q38F:Q4F:V62E (Table II-21, entry 4) both show high quantum yields of 0.49 and 0.65, respectively, when coupled with ThioFluor. Interestingly, these mutants do not contain acidic residues in common. If it is the presence of the acidic residue that is responsible for increasing the quantum yield of the complex, one would venture to hypothesize that the chromophore binding in the two proteins is dissimilar, such that the iminium experiences different microenvironments. It is also possible that there is a purely synergistic combination (i.e. orientation and distance to the iminium). However, the origin of the large quantum yields cannot be deduced at this time; crystallization of these hCRBPII mutants with ThioFluor may provide this insight. II.9 Emission is linearly correlated to absorbance As was first stated at the opening of this chapter, our overall goal is to develop fluorescent proteins that span the entire visible spectra. Figure II-34 depicts select hCRBPII mutants coupled with ThioFluor, clearly shows that we have achieved this goal. hCRBPII mutants bound with ThioFluor yielded mutants 121 Normalized Emission Normalized Absorbance 450 500 550 600 650 700 Wavelength (nm) 750 550 λabs,max = 501 - 705 nm Δabs = 204 nm 600 650 700 750 Wavelength (nm) 800 850 λem,max = 613 - 744 nm Δem = 131 nm Figure II-34. Range of absorption and emission achieved in this study by coupling hCRBPII monomers with ThioFluor. Mutants from left to right are (1) Q108K:K40L:T51D:T53A:R58Y:Q38F:Q4F:L117E (note: data was smoothed by averaging because of low absorbance intensity), (2) Q108K:K40D:T53A:R58Y:Q38F:Q4F, (3) Q108K:K40L:T53A:R58L, (4) Q108K:K40L:T51V:T53S:R58G, (5) Q108K:K40L:T51V:T53S:R58W:Y19W, (6) Q108K:K40L:T51V:T53S:R58W:Y19W:F16Y, (7) Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Q4W. 760 y = 341.8 + 0.5702x R2= 0.9312 Emission (nm) 720 680 640 600 500 550 600 650 Absorbance (nm) 700 Figure II-35. Plot of emission versus absorbance for the approximately 80 hCRBPII/ThioFluor complexes examined in this study yields a linear fit, with R2 = 0.93. 122 with absorption maxima varying from 501 nm to 705 nm and emission maxima from 613 nm to 744 nm. This is equivalent to regulation over 204 nm in absorption and 131 nm in emission, covering both the red and far-red fluorescence wavelength regimes. A plot of emission versus absorbance shows a linear response (R2 = 0.93), with a slope of 0.57 (Figure II-35) indicating that approximately a 1 nm red-shift is gained in emission for a 2 nm change in absorption. The ultimate goal would now be to demonstrate success in multi-color imaging by using two different hCRBPII variants targeted to different intracellular organelles (i.e. one variant targeted to the nucleus and the second targeted to the extra nuclear space), labeled with the same ThioFluor dye, to allow for multiorganelle imaging simultaneously. However, the debilitating factor here is that the developed blue-shifted hCRBPII/ThioFluor complexes exhibit little iminium at physiological pH due to suppressed pKa. While this study has shown that emission wavelength can be regulated with absorption wavelength, we will require a new fluorophore that can shift wavelength in the same manner, while also possessing a high pKa in these mutants at physiological pH. Nonetheless, the rest of this chapter will prove the probe’s usefulness in live cell imaging, most prominently showing that ThioFluor is cell permeable and displays no background fluorescence when excited in the red, even when long incubation periods with ThioFluor are required, unlike merocyanine aldehyde. 123 II.10 Visualization of hCRBPII/ThioFluor in bacteria We next sought to demonstrate the engineered hCRBPII/ThioFluor system in cells. The mutant Q108K:K40L:T53S:V62N was chosen for these preliminary studies, for imaging in E. coli because of it’s moderate quantum yield (0.23), sufficient pKa of 8.3 and nanomolar binding affinity with the Q108L variant (Kd = 3.2 nM) with ThioFluor (Figure II-36). Additionally, binding was relatively fast; the pseudo first order half-life of iminium formation is 30 min at 23 °C (20 μM protein and 0.5 equiv ThioFluor at pH ~ 7). pKa Titration UV-Vis and Fluorescence Spectra 5 107 4 107 3 107 0.08 2 107 0.04 Emission Absorbance 0.12 1 107 0 400 500 600 700 800 ΔAbs at 613 nm (a. u.) 6 10 613 nm 688 nm 0.16 0.4 7 0 900 0.3 pKa = 8.3 R2 = 0.99 0.2 0.1 0 5 6 Wavelength (nm) 8 pH 9 10 11 PSB Formation 0.4 1 Abs. at 613 nm (a. u.) Relative Fluorescence Binding Affinity 7 0.8 Kd = 3.2 nM R2 = 0.93 0.6 0 0.3 0.2 k = 0.0232 min-1 t1/2 = 30 min R2 = 0.99 0.1 0 2 10-7 4 10-7 6 10-7 8 10-7 1 10-6 Chromophore Concentration, (M) 0 50 100 150 200 Time (min) 250 300 Figure II-36. Spectroscopic properties of Q108K:K40L:T53S:V62N with ThioFluor including UV-Vis and fluorescence spectra, pKa, binding affinity, and rate plot. Note: It was later found that the hCRBPII species data shown here is for the dimer. 124 BL21 cells, expressing Q108K:K40L:T53S:V62N, were incubated with 10 μM ThioFluor for 1 hour at 37°C and washed with PBS buffer. As evidenced from confocal imaging (Figure II-37), the bacterial cells readily uptake the chromophore. Fluorescence is observed upon excitation at 594 nm, and collecting emission with a long pass (LP) 615 nm filter (fluorescence greater than 615 nm is collected). The control cells, which were not transformed with the hCRBPII plasmid, lack fluorescence, demonstrating that background fluorescence due to unbound chromophore is minimal. Transfected cells Non-transfected cells ThioFluor/hCRBPII ex: 594 nm em: LP 615 nm ThioFluor/hCRBPII + Brightfield Figure II-37. Bacterial imaging of Q108K:K40L:T53S:V62N, labeled with ThioFluor. 125 II.11 Visualization of hCRBPII/ThioFluor in mammalian cells II.11.1 Labeling of hCRBPII variants that have a propensity to oligomerize is unproductive To further highlight its application as a fluorescent protein tag, Q108K:K40L:T53S:V62N was cloned into a pECFP-N1 vector containing ECFP to serve as a positive control for transfection. The final product of cloning was XhoI-hCRBPII-BamHI-AgeI-ECFP-Stop-NotI. No localization signal is included, so fluorescence is expected in the entire cell. 5 µM ThioFluor, 10 µM ThioFluor, 5 µM ThioFluor, 10 µM ThioFluor, incubated 2 h @ 37°C incubated 2 h @ 37°C incubated 4 h @ 37°C incubated 4 h @ 37°C ECFP ex: 458 nm em: BP 475 - 525 nm ThioFluor/hCRBPII ex: 594 nm em: LP 615 nm Figure II-38. Labeling of HeLa cells expressing Q108K:K40L:T53S:V62N-ECFP with ThioFluor at various incubation times and concentrations of ThioFluor. HeLa cells were incubated with ThioFluor at 37 °C with various concentrations (1 μM, 5 μM, 10 μM, 50 μM, 100 μM, 200 μM) and incubation times (1 h, 2 h, 4 h). Fluorescence from ECFP is observable upon excitation at 458 nm (band pass (BP) 475-535 emission filter). Excitation at 594 nm (LP 615 nm emission filter) did not provide images corresponding to that of ECFP (Figure II-38). Unfortunately, in all cases (all data not shown), no whole cell fluorescence 126 from ThioFluor is visible when hCRBPII-ECFP is expressed and incubated with ThioFluor. Faint fluorescence is observable in the red channel, presumably due to nonspecific labeling, which indicates ThioFluor is cell permeable, but unfortunately is unable to bind to the expressed hCRBPII-ECFP fusion protein. Q108K:K40L:T53S:V62N-dimer/ThioFluor 614 nm Abs. (a. u.) 0.4 0.3 0.2 0 10 20 30 40 50 60 70 80 90 100 110 0 10 20 30 40 50 60 70 80 90 100 110 0.5 0.4 Abs. (a. u.) 0.5 Q108K:K40L:T53S:V62N-monomer/ThioFluor 0.1 0.3 0.2 597 nm 0.1 0 300 400 500 600 700 Wavelength (nm) 0 300 800 400 500 600 700 Wavelength (nm) 800 Figure II-39. Drastically different spectroscopic properties are observed upon binding ThioFluor to Q108K:K40L:T53S:V62N monomer and dimer (for both identical conditions are used: 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7 and 23 °C). The spectra depict the course of iminium formation (in minutes). These preliminary in vivo studies were disheartening. However, it was soon realized that this protein makes both monomer and dimer during protein expression at room temperature, 8.0 mg/L and 7.1 mg/L, respectively. After separation of the monomer and dimer via size exclusion chromatography, we were able to look at the spectroscopic properties of the monomer and dimer independently. Data presented in the previous section (Chapter II.10) is equivalent to that obtained for the dimer. While the spectroscopic properties of Q108K:K40L:T53S:V62N-dimer/ThioFluor should be suitable for imaging, there appears to be little iminium formation of ThioFluor with the monomer of 127 Q108K:K40L:T53S:V62N (Figure II-39) under identical conditions. Disadvantageously, the monomer is also not stable to acid. Because ThioFluor has not been previously used in live cell imaging, we chose to test merocyanine aldehyde with the same hCRBPII mutant, as merocyanine aldehyde has shown success in labeling live cells (although with other hCRBPII mutants). Binding to the monomer and dimer independently was verified in vitro under identical conditions. Spectroscopic properties, including the wavelength of absorption, extinction coefficient and rate of iminium formation are essentially the same for both monomer and dimer (Table II-22). It should be noted that brightness is almost twelve times higher for the hCRBPII monomer complexed with merocyanine aldehyde than for ThioFluor, and its brightness is comparable to the previously reported hCRBPII/merocyanine aldehyde complexes. Table II-22. Spectroscopic properties of Q108K:K40L:T53S:V62N/merocyanine aldehyde. ε k t1/2b a hCRBPII λabs λem Φ (M-1cm-1) (min-1) (min) monomer 598 nd nd 212,000 0.3148 2.2 dimer 595 613 0.37 207,000 0.3475 2.0 a Absolute quantum yield was measured on a Quantaurus-QY. Half-lives based on the rate constant obtained from pseudo first order fitting of the data; measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7. b Next, we attempted to label the same hCRBPII-ECFP fusion in vivo with merocyanine aldehyde (Figure II-40). The same hCRBPII-ECFP construct was expressed in HeLa cells and subsequently incubated with 250 nM merocyanine 128 aldehyde for 1 h at 37 °C. Again, fluorescence from the ECFP channel is visible, denoting that transfection of the DNA was successful. However, as was the case for ThioFluor, no whole cell fluorescence is visible upon excitation at 594 nm (LP 615 nm emission filter). The fluorescence from the red channel seems to be consistent with background fluorescence due to nonspecific labeling. ECFP ex: 458 nm em: BP 475 - 525 nm ThioFluor/hCRBPII ex: 594 nm em: LP 615 nm ThioFluor/hCRBPII + Brightfield Non-transfected cells Transfected cells ThioFluor/hCRBPII ex: 594 nm em: LP 615 nm Figure II-40. Labeling of HeLa cells expressing Q108K:K40L:T53S:V62N-ECFP with merocyanine aldehyde. Interestingly labeling in U2OS cells shows a bit more hope. The same hCRBPII-ECFP fusion was expressed in U2OS cells and subsequently labeled with ThioFluor (incubated with 50 μM for 1 h at 37 °C) or merocyanine aldehyde (incubated with 250 nM for 5 min at 37 °C) (Figure II-41). As expected, 129 fluorescence from the ECFP channel is visible, showing that transfection of the DNA was successful. It should be noted that the ECFP channel of cells labeled with ThioFluor shows background, which is presumably due to the excitation of free chromophore. Upon excitation at 594 nm (LP 615 nm emission filter), both experiments show whole cell labeling, though more dim than expected with a substantial amount of background fluorescence. Perhaps, different cell lines lead to differences in monomer/dimer ratios or protein folding. However, further experiments were not conducted in order to optimize this system. ECFP + Brightfield Red Channel ex: 594 nm em: LP 615 nm ThioFluor merocyanine aldehyde ECFP ex: 458 nm em: BP 475 - 525 nm Figure II-41. Labeling of U2OS cells expressing Q108K:K40L:T53S:V62N-ECFP with ThioFluor and merocyanine aldehyde. From the data presented in this section, it was hypothesized that in vivo labeling was unsuccessful due to the propensity of Q108K:K40L:T53S:V62N to 130 oligomerize (which is due to both the presence of T53S and V62N). Hence, mutants that produced a majority of monomer during protein expression were utilized for future imaging of mammalian cells. II.11.2 ThioFluor successfully labels hCRBPII in mammalian cells In the previous section, it was observed that many mutations make hCRBPII prone to domain-swapped dimerization (see Chapter VI for a full discussion), and that one such mutant (Q108K:K40L:T53S:V62N) displays vastly different spectroscopic properties when coupled with ThioFluor. Additionally, ThioFluor was not able to label the fusion protein when expressed in mammalian cells. It was predicted that it is the propensity to form oligomers that renders this mutant unsuccessful when moving from labeling in vitro to in vivo. The mutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E was chosen for application toward in vivo imaging, due to its high quantum yield (0.31), high pKa (11.2) and reasonable half-life of 82 min (based on fitting to second order kinetics; measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7) with ThioFluor (Figure II-42). This absorption and emission wavelengths are also well aligned with lasers available on the confocal microscope (in vitro this mutant displayed an absorption maxima of 563 nm and emission maxima of 673 nm at ~ pH 7). Additionally, it is solely expressed in the monomeric form. hCRBPII was cloned into the pFlag-CMV2 vector containing EGFP, both with and without localization peptides, for labeling the whole cell (hCRBPII-EGFP), nucleus (hCRBPII-EGFP-3NLS) and the extranuclear space (hCRBPII-EGFP- 131 NES). Imaging was performed by incubating HeLa cells with 10 μM ThioFluor for 1 h at 37 °C. The cells were then washed with PBS buffer two times before imaging. UV-Vis and Fluorescence Spectra 563 nm 0.15 673 nm 2.5 10 6 0.1 1.5 10 6 1 106 0.05 Emission Absorbance 2 106 5 105 0 0 400 500 600 700 800 Wavelength (nm) pKa Titration PSB Formation 1.2 10 -5 Concentration of Complex (M) ΔAbs at 563 nm (a. u.) 0.15 0.1 pKa = 11.2 R2 = 0.95 0.05 0 1 10-5 8 10-6 k = 1224 M-1 min-1 t1/2 = 82 min R2 = 0.99 6 10-6 4 10-6 2 10-6 0 8 9 10 11 12 13 0 pH 200 400 600 Time (min) 800 1000 Figure II-42. Spectroscopic properties of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E with ThioFluor including UV-Vis and fluorescence spectra, pKa, and rate plot. In all cases, green fluorescence was observed when excited at 488 nm, indicating that transfection was successful and the fusion protein is being expressed (Figure II-43, green channel). Excitation at 594 nm of the hCRBPII/ThioFluor complex provided almost identical images (Figure II-43, red 132 channel). Noteworthy, background fluorescence is not apparent in the red channel, indicating that ThioFluor does not label off target lysines. This proves the utility of labeling Q108K:K40L:T51V:T53S:R58W:Y19W:L117E with ThioFluor in mammalian cells. EGFP ex: 488 nm em: BP 505 - 530 nm ThioFluor/hCRBPII ex: 594 nm em: LP 615 nm ThioFluor/hCRBPII + DIC EGFP hCRBPII 5864 bp pFLAG-CMV2 EGFP hCRBPII 3NLS 5864 bp pFLAG-CMV2 EGFP hCRBPII NES 5864 bp pFLAG-CMV2 Figure II-43. Labeling of HeLa cells expressing hCRBPII-EGFP, hCRBPII-EGFP3NLS and hCRBPII-EGFP-NES with 10 μM ThioFluor (incubated at 37°C for 1 hour). NLS = nuclear localization sequence, NES = nuclear export signal. II.12 Conclusions As stated in the introduction of this chapter, our overarching goal is to develop fluorescent proteins that span the entire visible spectra. This was achieved by the coupling of the solvatochromic fluorophore ThioFluor to various 133 hCRBPII mutants. ThioFluor yielded mutants with absorption maxima varying from 501 nm to 705 nm and emission maxima from 613 nm to 744 nm. This is equivalent to regulation over 204 nm in absorption and 131 nm in emission, covering both the red and far-red fluorescence wavelength regimes. A plot of emission versus absorbance shows a linear response (R2 = 0.93), with a slope of 0.57 (Figure II-35) indicating that approximately a 1 nm red-shift is gained in emission for a 2 nm change in absorption. Furthermore, we have shown its utility in live-cell imaging in whole cells, and with targeting to the nucleus and extranuclear space; fortuitously, negligible background fluorescence is apparent. The ultimate goal would now be to demonstrate success in multi-color imaging by using two different hCRBPII variants targeted to different intracellular organelles (i.e. one variant targeted to the nucleus and the second targeted to the extra nuclear space), labeled with the same ThioFluor dye, to allow for multiorganelle imaging simultaneously. However, blue-shifted hCRBPII/ThioFluor complexes exhibit little iminium at physiological pH due to suppressed pKa. We will require a new fluorophore that can shift wavelength in the same manner, while also possessing a high pKa in these mutants at physiological pH in order to show success in multi-color imaging. 134 REFERENCES 135 REFERENCES 1. Wang, W., Geiger, J.H. & Borhan, B. The photochemical determinants of color vision: revealing how opsins tune their chromophore's absorption wavelength. Bioessays 36, 65-74 (2014). 2. Berbasova, T., Nosrati, M., Vasileiou, C., Wang, W., Lee, K.S., Yapici, I., Geiger, J.H. & Borhan, B. Rational design of a colorimetric pH sensor from a soluble retinoic acid chaperone. Journal of the American Chemical Society 135, 16111-16119 (2013). 3. Wang, W., Nossoni, Z., Berbasova, T., Watson, C.T., Yapici, I., Lee, K.S., Vasileiou, C., Geiger, J.H. & Borhan, B. Tuning the electronic absorption of protein-embedded all-trans-retinal. Science 338, 1340-1343 (2012). 4. Reichardt, C. Solvatochromic Dyes as Solvent Polarity Indicators. Chemical Reviews 94, 2319-2358 (1994). 5. Marini, A., Munoz-Losa, A., Biancardi, A. & Mennucci, B. What is Solvatochromism? Journal of Physical Chemistry B 114, 17128-17135 (2010). 6. Yapici, I., Lee, K.S.S., Berbasova, T., Nosrati, M., Jia, X.F., Vasileiou, C., Wang, W.J., Santos, E.M., Geiger, J.H. & Borhan, B. "Turn-On" Protein Fluorescence: In Situ Formation of Cyanine Dyes. Journal of the American Chemical Society 137, 1073-1080 (2015). 7. Richard, J.P., Amyes, T.L., Crugeiras, J. & Rios, A. Pyridoxal 5 'phosphate: electrophilic catalyst extraordinaire. Current Opinion in Chemical Biology 13, 475-483 (2009). 8. Uttamapinant, C., White, K.A., Baruah, H., Thompson, S., FernandezSuarez, M., Puthenveetil, S. & Ting, A.Y. A fluorophore ligase for sitespecific protein labeling inside living cells. Proceedings of the National Academy of Sciences of the United States of America 107, 10914-10919 (2010). 9. Lakowicz, J.R. Principles of Fluorescence Spectroscopy, Edn. 3rd. (Springer, 2006). 10. Engelborghs, Y. & Visser, A.J.W.G. Fluorescence Spectroscopy and Microscopy: Methods and Protocols. (Springer, 2014). 136 11. Klan, P. & Wirz, J. in Photochemistry of Organic Compounds: From Concepts to Practice 25-72 (Wiley-Blackwell, 2009). 12. Ahmed, F. A Good Example of the Franck-Condon Principle. Journal of Chemical Education 64, 427-428 (1987). 13. Lakowicz, J.R. in Principles of Fluorescence Spectroscopy, Edn. 3rd 205235 (Springer, 2006). 14. Klymchenko, A.S. & Mely, Y. Fluorescent Environment-Sensitive Dyes as Reporters of Biomolecular Interactions. Progress in Molecular Biology and Translational Science 113, 35-58 (2013). 15. Loving, G.S., Sainlos, M. & Imperiali, B. Monitoring protein interactions and dynamics with solvatochromic fluorophores. Trends in Biotechnology 28, 73-83 (2010). 16. Klymchenko, A.S. Solvatochromic and Fluorogenic Dyes as EnvironmentSensitive Probes: Design and Biological Applications. Accounts of Chemical Research 50, 366-375 (2017). 17. Holmesfarley, S.R. & Whitesides, G.M. Fluorescence Properties of Dansyl Groups Covalently Bonded to the Surface of Oxidatively Functionalized Low-Density Polyethylene Film. Langmuir 2, 266-281 (1986). 18. Sameiro, M. & Goncalves, T. Fluorescent Labeling of Biomolecules with Organic Probes. Chemical Reviews 109, 190-212 (2009). 19. Weber, G. & Farris, F.J. Synthesis and Spectral Properties of a Hydrophobic Fluorescent-Probe 6-Propionyl-2(Dimethylamino)Naphthalene. Biochemistry 18, 3075-3078 (1979). 20. Lehrer, S.S. & Ishii, Y. Fluorescence Properties of Acrylodan-Labeled Tropomyosin and Tropomyosin Actin - Evidence for Myosin Subfragment1 Induced Changes in Geometry between Tropomyosin and Actin. Biochemistry 27, 5899-5906 (1988). 21. Hiratsuka, T. ATP-induced opposite changes in the local environments around Cys(697) (SH2) and Cys(707) (SH1) of the myosin motor domain revealed by the prodan fluorescence. Journal of Biological Chemistry 274, 29156-29163 (1999). 22. Shi, J.X., Boyd, A.E., Radic, Z. & Taylor, P. Reversibly bound and covalently attached ligands induce conformational changes in the omega 137 loop, Cys(69)-Cys(96), of mouse acetylcholinesterase. Journal of Biological Chemistry 276, 42196-42204 (2001). 23. Harikumar, K.G., Pinon, D.L., Wessels, W.S., Prendergast, F.G. & Miller, L.J. Environment and mobility of a series of fluorescent reporters at the amino terminus of structurally related peptide agonists and antagonists bound to the cholecystokinin receptor. Journal of Biological Chemistry 277, 18552-18560 (2002). 24. Lu, Z.K., Lord, S.J., Wang, H., Moerner, W.E. & Twieg, R.J. Longwavelength analogue of PRODAN: Synthesis and properties of Anthradan, a fluorophore with a 2,6-donor-acceptor anthracene structure. Journal of Organic Chemistry 71, 9651-9657 (2006). 25. Kucherak, O.A., Didier, P., Mely, Y. & Klymchenko, A.S. Fluorene Analogues of Prodan with Superior Fluorescence Brightness and Solvatochromism. Journal of Physical Chemistry Letters 1, 616-620 (2010). 26. Carlotti, B., Flamini, R., Kikas, I., Mazzucato, U. & Spalletti, A. Intramolecular charge transfer, solvatochromism and hyperpolarizability of compounds bearing ethenylene or ethynylene bridges. Chemical Physics 407, 9-19 (2012). 27. Singh, P., Baheti, A., Thomas, K.R.J., Lee, C.P. & Ho, K.C. Fluorenebased organic dyes containing acetylene linkage for dye-sensitized solar cells. Dyes and Pigments 95, 523-533 (2012). 28. Tigreros, A., Ortiz, A. & Insuasty, B. Effect of pi-conjugated linkage on photophysical properties: Acetylene linker as the better connection group for highly solvatochromic probes. Dyes and Pigments 111, 45-51 (2014). 29. Diwu, Z., Lu, Y.X., Zhang, C.L., Klaubert, D.H. & Haugland, R.P. Fluorescent molecular probes II. The synthesis, spectral properties and use of fluorescent solvatochromic Dapoxyl(TM) dyes. Photochemistry and Photobiology 66, 424-431 (1997). 30. Min, J., Lee, J.W., Ahn, Y.H. & Chang, Y.T. Combinatorial dapoxyl dye library and its application to site selective probe for human serum albumin. Journal of Combinatorial Chemistry 9, 1079-1083 (2007). 31. Ando, Y., Homma, Y., Hiruta, Y., Citterio, D. & Suzuki, K. Structural characteristics and optical properties of a series of solvatochromic 138 fluorescent dyes displaying Pigments 83, 198-206 (2009). long-wavelength emission. Dyes and 32. Dattelbaum, J.D., Looger, L.L., Benson, D.E., Sali, K.M., Thompson, R.B. & Hellinga, H.W. Analysis of allosteric signal transduction mechanisms in an engineered fluorescent maltose biosensor. Protein Science 14, 284291 (2005). 33. Shi, Z.D., Karki, R.G., Oishi, S., Worthy, K.M., Bindu, L.K., Dharmawardana, P.G., Nicklaus, M.C., Bottaro, D.P., Fisher, R.J. & Burke, T.R. Utilization of a nitrobenzoxadiazole (NBD) fluorophore in the design of a Grb2 SH2 domain-binding peptide mimetic. Bioorganic & Medicinal Chemistry Letters 15, 1385-1388 (2005). 34. Jiang, W., Fu, Q.Q., Fan, H.Y. & Wang, W. An NBD fluorophore-based sensitive and selective fluorescent probe for zinc ion. Chemical Communications, 259-261 (2008). 35. Zhou, G.D., Wang, H.L., Ma, Y. & Chen, X.Q. An NBD fluorophore-based colorimetric and fluorescent chemosensor for hydrogen sulfide and its application for bioimaging. Tetrahedron 69, 867-870 (2013). 36. Soujanya, T., Fessenden, R.W. & Samanta, A. Role of nonfluorescent twisted intramolecular charge transfer state on the photophysical behavior of aminophthalimide dyes. Journal of Physical Chemistry 100, 3507-3512 (1996). 37. Vazquez, M.E., Rothman, D.M. & Imperiali, B. A new environmentsensitive fluorescent amino acid for Fmoc-based solid phase peptide synthesis. Organic & Biomolecular Chemistry 2, 1965-1966 (2004). 38. Vazquez, M.E., Blanco, J.B. & Imperiali, B. Photophysics and biological applications of the environment-sensitive fluorophore 6-N,NDimethylamino-2,3-naphthalimide. Journal of the American Chemical Society 127, 1300-1306 (2005). 39. Loving, G. & Imperiali, B. A versatile amino acid analogue of the solvatochromic fluorophore 4-N,N-dimethylamino-1,8-naphthalimide: A powerful tool for the study of dynamic protein interactions. Journal of the American Chemical Society 130, 13630-13638 (2008). 40. Liu, X., Cole, J.M. & Low, K.S. Solvent Effects on the UV-vis Absorption and Emission of Optoelectronic Coumarins: a Comparison of Three 139 Empirical Solvatochromic Models. Journal of Physical Chemistry C 117, 14731-14741 (2013). 41. Maity, B., Chatterjee, A. & Seth, D. Photophysics of a Coumarin in Different Solvents: Use of Different Solvatochromic Models. Photochemistry and Photobiology 90, 734-746 (2014). 42. Lanke, S.K. & Sekar, N. Coumarin Push-Pull NLOphores with Red Emission: Solvatochromic and Theoretical Approach. Journal of Fluorescence 26, 949-962 (2016). 43. Greenspan, P., Mayer, E.P. & Fowler, S.D. Nile Red - a Selective Fluorescent Stain for Intracellular Lipid Droplets. Journal of Cell Biology 100, 965-973 (1985). 44. Jose, J. & Burgess, K. Benzophenoxazine-based fluorescent dyes for labeling biomolecules. Tetrahedron 62, 11021-11037 (2006). 45. Sherman, D.B., Pitner, J.B., Ambroise, A. & Thomas, K.J. Synthesis of thiol-reactive, long-wavelength fluorescent phenoxazine derivatives for biosensor applications. Bioconjugate Chemistry 17, 387-392 (2006). 46. Thomas, K.J., Sherman, D.B., Amiss, T.J., Andaluz, S.A. & Pitner, J.B. A long-wavelength fluorescent glucose biosensor based on bioconjugates of galactose/glucose binding protein and Nile Red derivatives. Diabetes Technology & Therapeutics 8, 261-268 (2006). 47. Kawski, A., Bojarski, P. & Kuklinski, B. Estimation of ground- and excitedstate dipole moments of Nile Red dye from solvatochromic effect on absorption and fluorescence spectra. Chemical Physics Letters 463, 410412 (2008). 48. Kamlet, M.J. & Taft, R.W. Solvatochromic Comparison Method 1. BetaScale of Solvent Hydrogen-Bond Acceptor (Hba) Basicities. Journal of the American Chemical Society 98, 377-383 (1976). 49. Taft, R.W. & Kamlet, M.J. Solvatochromic Comparison Method 2. AlphaScale of Solvent Hydrogen-Bond Donor (Hbd) Acidities. Journal of the American Chemical Society 98, 2886-2894 (1976). 50. Kamlet, M.J., Abboud, J.L.M., Abraham, M.H. & Taft, R.W. Linear Solvation Energy Relationships 23. A Comprehensive Collection of the Solvatochromic Parameters, Pi-Star, Alpha and Beta, and Some Methods 140 for Simplifying the Generalized Solvatochromic Equation. Journal of Organic Chemistry 48, 2877-2887 (1983). 51. Abraham, M.H., Grellier, P.L., Abboud, J.L.M., Doherty, R.M. & Taft, R.W. Solvent Effects in Organic-Chemistry - Recent Developments. Canadian Journal of Chemistry-Revue Canadienne De Chimie 66, 2673-2686 (1988). 52. Catalan, J., Lopez, V., Perez, P., Martinvillamil, R. & Rodriguez, J.G. Progress Towards a Generalized Solvent Polarity Scale - the Solvatochromism of 2-(Dimethylamino)-7-Nitrofluorene and Its Homomorph 2-Fluoro-7-Nitrofluorene. Liebigs Annalen, 241-252 (1995). 53. Catalan, J., Diaz, C., Lopez, V., Perez, P., dePaz, J.L.G. & Rodriguez, J.G. A generalized solvent basicity scale: The solvatochromism of 5nitroindoline and its homomorph 1-methyl-5-nitroindoline. Liebigs Annalen, 1785-1794 (1996). 54. Catalan, J. & Diaz, C. A generalized solvent acidity scale: The solvatochromism of o-tert-butylstilbazolium betaine dye and its homomorph o,o'-di-tert-butylstilbazolium betaine dye. Liebigs Annalen, 1941-1949 (1997). 55. Zou, J.W., Yu, Q.S. & Shang, Z.C. Correlation between empirical solvent polarity scales and computed quantities derived from molecular surface electrostatic potentials. Journal of the Chemical Society-Perkin Transactions 2, 1439-1443 (2001). 56. Catalan, J. Toward a Generalized Treatment of the Solvent Effect Based on Four Empirical Scales: Dipolarity (SdP, a New Scale), Polarizability (SP), Acidity (SA), and Basicity (SB) of the Medium. Journal of Physical Chemistry B 113, 5951-5960 (2009). 57. Ceron-Carrasco, J.P., Jacquemin, D., Laurence, C., Planchat, A., Reichardt, C. & Sraidi, K. Solvent polarity scales: determination of new ET(30) values for 84 organic solvents. Journal of Physical Organic Chemistry 27, 512-518 (2014). 58. Reichardt, C. Pyridinium N-phenolate betaine dyes as empirical indicators of solvent polarity: Some new findings. Pure and Applied Chemistry 76, 1903-1919 (2004). 59. Machado, V.G., Stock, R.I. & Reichardt, C. Pyridinium N-Phenolate Betaine Dyes. Chemical Reviews 114, 10429-10475 (2014). 141 60. Eisfeld, A. & Briggs, J.S. The J- and H-bands of organic dye aggregates. Chemical Physics 324, 376-384 (2006). 61. Wang, W. in PhD Thesis, Michigan State University (2012). 62. Yapici, I. in PhD Thesis, Michigan State University (2015). 63. Yutani, K., Ogasahara, K., Tsujita, T. & Sugino, Y. Dependence of Conformational Stability on Hydrophobicity of the Amino-Acid Residue in a Series of Variant Proteins Substituted at a Unique Position of Tryptophan Synthase Alpha-Subunit. Proceedings of the National Academy of Sciences of the United States of America 84, 4441-4444 (1987). 64. Pace, C.N., Trevino, S., Prabhakaran, E. & Scholtz, J.M. Protein structure, stability and solubility in water and other solvents. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 359, 1225-1234 (2004). 65. Pace, C.N., Grimsley, G.R. & Scholtz, J.M. Protein Ionizable Groups: pK Values and Their Contribution to Protein Stability and Solubility. Journal of Biological Chemistry 284, 13285-13289 (2009). 66. Assar, Z., Nossoni, Z., Wang, W., Santos, E.M., Kramer, K., McCornack, C., Vasileiou, C., Borhan, B. & Geiger, J.H. Domain-Swapped Dimers of Intracellular Lipid-Binding Proteins: Evidence for Ordered Folding Intermediates. Structure 24, 1590-1598 (2016). 142 CHAPTER III: STRUCTURE-PROPERTY RELATIONSHIPS OF THIOFLUOR ANALOGUES BOUND TO hCRBPII MUTANTS In developing novel fluorescent proteins and small molecule tags, there are numerous desirable qualities of the probe that should be achieved. It is these requirements that present a challenge in the design of new synthetic small molecule probes. Perhaps the most imperative characteristics are that the probe is fluorogenic (ideally in the red or near-infrared wavelength regime), cell permeable and rapid (i.e. short incubation time). Fluorogenic probes should allow for temporal control, in which fluorescence is produced only when the dye is bound to the target.1-4 Nonspecific labeling of a dye may yield background fluorescence that decreases the detection limit or signal to noise ratio of the dye. Undesired fluorescence may also originate from the autofluroescence of endogenous compounds in mammalian cells, such as riboflavin and flavin coenzymes, among others.5-8 When using green or yellow fluorescent proteins (FPs), whose emission overlays cellular autofluorescence, detection of the FP may be hampered. Choosing red or far red emitting fluorescent dyes can help to increase the signal to noise ratio. Red-shifted emission can also provide the added benefit of reduced scattering and deeper tissue penetration, due to the longer wavelength of light.9, 10 In line with fluorogenicity, it is advantageous for the system to require a short incubation time with the dye. Not only does this decrease the amount of background fluorescence in some cases (as was seen for merocyanine aldehyde), but a faster reaction time also allows for performing 143 real time measurements and the monitoring of molecular events directly after labeling.11-13 Cell permeable probes again extend the applicability of the dye, as they allow for the imaging of intracellular organelles such as the nucleus, while impermeable dyes limit imaging to the cell membrane,14, 15 which limits their usefulness. Other desirable properties of dyes in live cell imaging are brightness and photostability, and their ability to be targeted to allow for spatial control. Brightness is the product of the extinction coefficient (at the relevant excitation wavelength) and quantum yield. Extinction coefficient is a measure of the ability of a dye to absorb light, while quantum yield is the number of photons emitted per the number of photons absorbed. Thus, both a higher extinction coefficient and higher quantum yield can lead to increased brightness.16-19 Along the same lines, it is desirable that the dye be photostable. During excitation of a fluorophore, the dye can be photobleached, which can lead to its permanent loss of fluorescence due to photon-induced damage.20-22 Some fluorophores bleach quickly, while others are robust and can undergo cycling between excitation and emission many times before bleaching. One of the ways to minimize photobleaching of the fluorophore is to limit the amount of light it is exposed to, which can be done by reducing light intensity (which also reduces signal intensity); it is in such a case that brighter fluorophores are desirable. Chapter II detailed the design of fluorescent proteins based on hCRBPII coupled with the solvatochromic fluorophore ThioFluor. Significant achievements 144 include the wavelength and emission regulation, via site directed mutagenesis, with absorptions spanning 501 nm to 705 nm and emissions 613 nm to 744 nm. The system displayed many desirable properties, such as being fluorogenic, which allowed for minimal background staining and spatial control (as the tags can be localized to organelles via sequencing peptides), and decent quantum yields. However, properties to be improved are reaction kinetics and pKa of the in situ formed iminium. To this end, we sought to generate a small library of ThioFluor analogues, in order to understand how the spectroscopic properties of hCRBPII and an aldehydic ligand pair can be manipulated, such as the absorption wavelength, emission wavelength, the rate of iminium formation, pKa and quantum yield. Herein, the goal is to make rational, meticulous modifications on ThioFluor, and individually look at the effect of the substitution. Better understanding of the structure-property relationships should establish design parameters for improved dyes. We used ThioFluor as a starting point for modification to develop analogues, generally based on substitution of the electron donor and acceptor substituent’s and the π linker between the electron donor and electron acceptor moieties. III.1 Prior work exploring substituent effects on absorption wavelength in fluorescent proteins Excitation wavelength tuning in fluorescent proteins is dictated by the covalent structure of the chromophore, while also being influenced by hydrogen 145 bonding and electrostatic interactions with surrounding residues.20, 23-26 The chromophore of green fluorescent protein (GFP) is comprised of phenol and imidazolinone rings derived from the inherent tyrosine 66 residue and backbone atoms of adjacent residues (serine 65 and glycine 67).27-29 An olefin group bridges the two rings to create an extended conjugated π system (Figure III-1). Additionally, in most red fluorescent proteins (RFPs), oxidation of the N-Cα bond immediately N-terminal to the chromophore extends conjugation through the acylimine group.26, 30 The chromophore of mCherry is shown in Figure III-1 as an example. In this case, the extra conjugation from the acylimine group is partially responsible for the 100 nm red-shift in absorption wavelength.26 31 The approach of systematically exploring a chromophore’s donor and acceptor abilities is not an unprecedented one. The Schultz lab used unnatural amino acid mutagenesis to selectively substitute tyrosine 66 of GFP with tyrosine analogues bearing different substituents at the para position of the phenyl ring including amino, methoxy, iodo, and bromo groups (Table III-1, entries 1 – 5).32 When the unnatural amino acids were introduced, single maxima were observed for both absorbance and emission spectra, unlike wild-type GFP which displays two absorbance maxima due to the presence of both the protonated and unprotonated forms of the tyrosine’s hydroxyl group (Table III-1, entry 2). The wavelengths for both absorbance and emission peaks increase in the order of electron donating ability: p-bromo, p-iodo, p-methoxy, p-hydroxyl, p-amino and phydroxide (Table III-1). The quantum yields of the modified GFPs with p-amino 146 and p-methoxy are lower than wild-type GFP, but their extinction coefficients are slightly higher. Tyr66 HO H N HN O O N H O Folding N H HN HO O Res65 OH Gly67 O O R O O N H Cyclization O O N HO Oxidation (O2) N O R O O -H2O N H N HO HN R OH O HN +- H+ O N R O O O O Oxidation (O2) N N N O N O R N H O EGFP mCherry λex = 484 nm λex = 587 nm λem = 507 nm λem = 610 nm Figure III-1. Proposed scheme for the maturation of the GFP and mCherry chromophores. Proper protein conformation and subsequent cyclization, oxidation and dehydration are required to form the fluorescent state. The same group later introduced numerous other electron substituents via the site-specific incorporation of unnatural amino acids at Tyr66. While absorption wavelength data was not provided for the other substituents (including nitrile, nitro, acetyl, azido, o-allyl and phenyl) it is observed that the emission maxima are all blue-shifted as compared to the wild-type tyrosine (Table III-1, 147 entries 6 – 11).33, 34 The blue-shift in these wavelengths is because the electron donating ability of these functional groups is lower than the hydroxyl group of tyrosine. Additionally, the replacement of the hydroxyl group of tyrosine reduces the fluorescence quantum yields substantially in all cases, presumably because the electron-withdrawing group prevents the formation of a fluorescent push-pull system. Alternatively, it has been shown that the incorporation of halogens in porphyrins leads to a higher formation of the triplet state upon excitation,35 which leads to a decrease in fluorescence quantum yield. Table III-1. Effect of Tyr66 replacement on spectroscopic properties of GFP. ε Entry Substituent λabs λem Φ -1 (M cm-1) 1 p-bromo 375 428 20,000 0.01 2 p-hydroxy (wildtype) 397,475 506 25,000 0.76 3 p-iodo 381 438 16,000 0.01 4 p-methoxy 394 460 27,000 0.37 5 p-amino 435 498 31,000 0.43 6 p-nitro nd - - - 7 p-acetyl nd 435 nd 0.04 8 p-cyano nd 442 nd 0.04 9 p-phenyl nd 459 nd 0.07 10 p-vinyloxy nd 461 nd 0.04 11 p-azido nd 501 nd 0.11 III.2 Structure of ThioFluor analogues to be explored As aforementioned, the goal is to develop a small library of ThioFluor analogues, in order to gain a better understanding of the structure-property relationships apparent in this system. It is hoped this would establish design 148 parameters for the rational design of dyes with improved spectroscopic properties. We chose to start with ThioFluor, as it has already shown success in live cell imaging. Mr. Hadi Gholami was successful in synthesizing a variety of analogues, which differ based on their electron donating or acceptor groups and the aromatic linker between the two groups (Figure III-2). H H H O O O S S N S F3C N ThioFluor ε430nm (EtOH) = 27,021 M-1cm-1 ThioFluor-5 ε360nm (EtOH) = 23,412 M-1cm-1 H H O O N ThioFluor-9 ε416nm (EtOH) = 15,688 M-1cm-1 N O S S H ThioFluor-2 ε445nm (EtOH) = 29,763 M-1cm-1 H H ThioFluor-6 ε364nm (EtOH) = 17,708 M-1cm-1 O H ThioFluor-10 ε369nm (EtOH) = 22,071 M-1cm-1 F O O S S S CF3 N N ThioFluor-3 ε418nm (EtOH) = 26,592 M-1cm-1 H N ThioFluor-7 ε340nm (EtOH/Et3N) = 20,682 M-1cm-1 H ThioFluor-11 ε429nm (EtOH) = 26,900 M-1cm-1 H O O S S S N O ThioFluor-4 ε384nm (EtOH) = 21,509 M-1cm-1 ThioFluor-8 ε472nm (EtOH) = 14,821 M-1cm-1 Figure III-2. ThioFluor analogues investigated to study structure-property relationships in absorption and fluorescence. Extinction coefficients are for λmax of each chromophore. The first set of analogues was made in order to explore the role of substituent effects on ThioFluor. For this reason, we tested the substitution of 149 electron donating groups on ThioFluor, replacing its para N,N-dimethyl aniline group with azetidine (ThioFluor-3), methoxy (ThioFluor-4), trifluoromethyl (ThioFluor-5), and hydrogen (ThioFluor-6). To alter the electron acceptor moiety, while keeping the same ThioFluor scaffold, we are more limited as iminium formation requires an aldehydic moiety. To this end we made ThioFluor7 and ThioFluor-11, in which a trifluoromethyl group is introduced beta to the aldehyde and a fluoro group is introduced alpha to the aldehyde, respectively. The second set of analogues was synthesized in order to test the effect of conjugation across the pi system. For this reason, we kept the electron donating and electron acceptor groups equivalent to ThioFluor, and chose to alter the aromatic rings between the two. Both the bithiophene and biphenyl variants, ThioFluor-8 and ThioFluor-10, were synthesized as well as ThioFluor-9, in which the order of the two aromatic rings is reversed. It should be noted that the shortest chromophore ThioFluor-2 was made in order to create a dye that is more cyanine like, in that it should remain fully conjugated. Cyanine dyes have been a focus in recent research in the field of fluorescence,36-42 as well as in our lab.43 They satisfy many of the aforementioned requirements for use in imaging such as exhibiting fluorogenic behavior, high molar extinction coefficient and high quantum yields. Additionally, as shown in Chapter II.1, merocyanine aldehyde can bind optimized hCRBPII mutants, with half-lives as fast as ten seconds. With the synthesized analogues in hand, we first chose to test the spectroscopic properties of the free aldehyde, and the model Schiff bases (SB) 150 and protonated Schiff bases (PSB). This data will provide insight into the relative wavelengths of both absorption and emission for the ThioFluor analogues. III.3 Evaluation of ThioFluor analogues with n-butyl amine in ethanol Spectroscopic properties of the ligands were first measured. Results are summarized in Table III-2, shown in increasing order of absorption wavelength. All spectroscopic data was recorded in ethanol. It should be noted that ThioFluor-7, with the trifluoromethyl group at the beta position of the aldehyde, exists in equilibrium with its hemiacetal form in ethanol. The extinction coefficient was measured by adding triethylamine to the solution to force the equilibrium almost solely to the hemiacetal form (λabs = 340 nm). ThioFluor-10 also shows two bands in absorption; presumably the blue-shifted absorption band is due to a twisted ground state, which does not allow the chromophore to be coplanar. A plot of Hammett value versus absorption wavelength is shown in Figure III-3.44 An increase in electron donating ability of the group at the para position of the phenyl ring leads to a red-shift in absorption, increasing in order for trifluoromethyl, hydrogen, methoxy and N,N-dimethyl. Changing the aromatic rings also results in regulation of wavelength; interestingly, reversal of the thiophene and phenyl rings in ThioFluor-9 leads to no significant change in absorption wavelength from the parent ThioFluor. However, the biphenyl ThioFluor-10 is more blue-shifted than ThioFluor, while the bithiophene ThioFluor-8 is more red-shifted. It stands to reason that ThioFluor-8 is more conjugated. This is supported by the resonance energies of benzene and 151 thiophene. The resonance energy of benzene is 36 kcal/mol, while the resonance energy of thiophene is 29 kcal/mol;45 thus it should be easier to break aromaticity and resonate through a thiophene pi linker, as opposed to a phenyl linker. The overall effect is better delocalization over the whole molecule, which leads to a smaller energy gap between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO). Table III-2. Spectroscopic properties of ThioFluor analogues (free aldehyde) in ethanol. ε Ligand λabs λema -1 (M cm-1) ThioFluor-5 360 402 23,412 ThioFluor-6 364 396 17,708 ThioFluor-10 271, 369 556 22,071 ThioFluor-4 384 447 21,509 ThioFluor-9 416 592 15,688 ThioFluor-3 418 562, 626, 660 26,592 ThioFluor-11 429 ~ 556, 630 653 26,900 ThioFluor 430 634 27,021 ThioFluor-7 320, 438 601 20,682 ThioFluor-2 445 498 29,763 ThioFluor-8 472 669 14,821 a Emission maxima is recorded as a result of exciting the absorption peak that is expected to correspond to the fully conjugated aldehyde. Unless otherwise noted, the emission peaks are singular, implying that emission occurs from only one excited state. The outliers are seen in ThioFluor3 and ThioFluor-11. Azetidine has been hypothesized to minimize formation of the twisted intramolecular charge transfer (TICT) in numerous fluorophore 152 families,46, 47 and perhaps that is the reason for the multiple emission bands observed upon excitation of the aldehyde in solution. Presumably emission occurs from both the locally excited state and the intramolecular charge transfer state, which should correspond to the blue-shifted and red-shifted bands, respectively. ThioFluor, possessing a para N,N-dimethyl substituent displays only a red-shifted emission band (which is close in wavelength to the red-shifted emission band in ThioFluor-3). ThioFluor-5 and ThioFluor-6, which do not have electron donating groups in the chromophore, are not capable of charge transfer, and thus display small Stokes shift upon excitation of the aldehyde. y = 6.187 - 0.01645x R2= 0.8583 0.6 -CF3 p Hammett Value (σ ) 0.4 0.2 0 -H -0.2 -0.4 -OCH3 -0.6 -0.8 -1 340 -N(CH3)2 360 380 400 420 Absorbance (nm) 440 Figure III-3. Plot of Hammett values versus absorption wavelength for ThioFluor-5, ThioFluor-6, ThioFluor-4 and ThioFluor. Model Schiff bases were then prepared by reacting each ThioFluor analogue with n-butyl amine in ethanol. In all cases, the SB displays an absorption maxima blue-shifted than the free aldehyde. Subsequently the protonated Schiff base (PSB) was formed by acidification of the SB. 153 Spectroscopic properties are detailed in Table III-3. Not unexpectedly, the most red-shifted PSB belongs to the most conjugated aldehyde (ThioFluor-8). Interestingly though, the iminiums formed with ThioFluor-11 and ThioFluor-7 are more red-shifted than with ThioFluor, even though the absorption wavelengths of the free aldehyde were approximately the same. Upon excitation of the PSB, singular peaks are observed. With the exception of ThioFluor-5-PSB, ThioFluor-6-PSB and ThioFluor-4-PSB (with a trifluoromethyl group, hydrogen, and methoxy group at the para position of the phenyl, respectively), the emissions fall in the red emission range. Interestingly the largest Stokes shift (238 nm) is observed from ThioFluor-10-PSB. Similarly reported biphenyl push-pull probes also display large Stokes shifts (LSS).48 Table III-3. Spectroscopic properties of ThioFluor analogues (SB and PSB) in ethanol. Ligand a SB PSB λabs λema λabs λem Stokes Shift ThioFluor-5 352 399 403 489 86 ThioFluor-6 347 395 418 503 85 ThioFluor-10 349 488 443 681 238 ThioFluor-4 358 440 440 556 116 ThioFluor-9 390 535 506 690 184 ThioFluor-3 392 550 508 690 182 ThioFluor-11 397 552 543 698 155 ThioFluor 400 540 521 688 167 ThioFluor-7 317, 397 ~ 487, 585 336, 568 - - ThioFluor-2 400 485 510 531 21 ThioFluor-8 432 584 600 709 109 Emission recorded as a result of exciting the most red-shifted absorption peak. 154 After testing the ThioFluor analogues with n-butyl amine in solution, we chose to test them with four hCRBPII protein mutants. As stated previously, the problem with the hCRBPII/ThioFluor protein series (i.e. generally when acidic residues are present) is that many of the blue-shifted mutants suffer from a severely suppressed pKa. For this reason, we chose to start with mutants that were expected to exhibit a pKa higher than physiological pH, as evidenced by the formation of iminium at a pH of approximately 7. We chose to test the following mutants – Q108K:K40L:T51V:T53S:R58W:Y19W:L117E, Q108KK40L:T51V:T53S:R58W:Y19W:A33W, Q108K:K40L:R58F, and Q108K:K40L:T51V:T53S:R58W. Firstly, Q108K:K40L:T51V:T53S:R58W:Y19W:L117E is an ideal mutant when coupled with ThioFluor because it has a high quantum yield (0.31), high pKa (11.2), and this hCRBPII/ThioFluor complex was successfully utilized in live cell imaging. Additionally, we predicted that ThioFluor analogues would bind in the same conformation as ThioFluor, due to the stabilization of the iminium by L117E. Thus, we predicted that we could make a direct correlation between the electron donating ability of the para-substituent or overall conjugation and the spectroscopic properties of the resultant complex, such as absorption wavelength, emission wavelength, quantum yield and the rate of iminium formation. The other three hCRBPII mutants chosen also showed substantial iminium at physiological pH when 155 coupled to ThioFluor. Q108K:K40L:T51V:T53S:R58W:Y19W:L117E was the most blue-shifted mutant (λabs = 558 nm and λem = 663 nm with ThioFluor) tested with ThioFluor analogues. It should be noted that we did attempt to test the more blue-shifted mutant Q108K:K40D:T51V:R58L:Q38F:Q4F:V62E with ThioFluor analogues, but the protein was not acid stable enough to form an iminium with many of the analogues, due to their extremely low pKas. spanning to the red wavelength regime we tested Q108K:K40L:R58F (λabs = 597 nm and λem = 678 nm with ThioFluor), Q108K:K40L:T51V:T53S:R58W (λabs = 623 nm and λem = 697 nm with ThioFluor) and Q108K:K40L:T51V:T53S:R58W:Y19W:A33W (λabs = 666 nm and λem = 724 nm with ThioFluor). III.4 Coupling of ThioFluor analogues to Q108K:K40L:T51V:T53S:R58W:Y19W:L117E In Chapter II.11.2, it was shown that ThioFluor could successfully label expressed Q108K:K40L:T51V:T53S:R58W:Y19W:L117E in mammalian cells. It is an ideal mutant when coupled with ThioFluor for live cell imaging due to its high quantum yield (0.31), high pKa (11.2) and reasonable half-life of 82 min (based on fitting to second order kinetics; measured at 23 °C with 20 μM protein and 0.5 equiv ThioFluor at pH ~ 7). With the use of ThioFluor analogues, it would be ideal to retain the high iminium pKa and quantum yield, while decreasing the rate of iminium formation. The longer half-life, requiring an hour of ThioFluor incubation, hampers visualization of fluorescence within this time domain and prohibits the collection of real time measurements. 156 L117E 2.9 Å Q108K Figure III-4. In the hCRBPII heptamutant Q108K:K40L:T51V:T35S:R58W:Y19W:L117E, the distance between the iminium nitrogen and L117E is 2.9 Å; it is this ionic interaction that presumably leads to the high pKa and holds the iminium as a trans isomer. Density is shown at 1σ. From the crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor, a trans iminium of ThioFluor is observed, which puts the iminium nitrogen just 2.9 Å away from the neighboring L117E residue (Figure III-4). Presumably, the ionic interaction between the positively charged iminium and the negatively charged carboxylate side chain of L117E leads to the high observed pKa of 11.2. Indeed, upon acid titration of the Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor complex from a neutral pH, a red-shift in absorption is observed, presumably due to protonation of L117E. The pKa of L117E was found to be 6.6, rendering the 157 carboxylate species predominate at pH 7. This result supports the theory that the elevated iminium pKa is due to an ionic interaction with the carboxylate side chain of the nearby glutamate residue. As aforementioned, we predicted that the pKa of the iminium should be correlated with the degree of conjugation along the chromophore. To this end, the electron donating ability of the para substituent should affect iminium pKa, in which a stronger electron-donating group can stabilize the positive charge of the iminium nitrogen better. However, it was hypothesized that ThioFluor analogues, if bound to Q108K:K40L:T51V:T53S:R58W:Y19W:L117E in the same manner, could all adopt a similar trans iminium and be stabilized by the carboxylate present at residue 117, leading to the iminium possessing a high pKa, rendering the effect of chromophore conjugation insignificant. Assuming binding of the chromophore to be comparable (in terms of chromophore trajectory and isomer of the resultant iminium), we should be able to make a direct correlation between the electron donating ability of the parasubstituent or overall conjugation and the spectroscopic properties of the resultant complex, such as absorption wavelength, emission wavelength, quantum yield and the rate of iminium formation. Of the eleven chromophores tested with the hCRBPII heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E, formation of iminium was observed for ten; it appears that iminium formation does not occur for ThioFluor11. On steric grounds, fluorine is not significantly larger than hydrogen, so this 158 result is unexpected. A summary of the spectroscopic properties of the ThioFluor analogues, coupled with Q108K:K40L:T51V:T53S:R58W:Y19W:L117E is provided in Table III-4. Table III-4. Spectroscopic properties of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. ε Ligand λabs λem Φ pKa (M-1cm-1) ThioFluor-5 423 - - 23,530 10.9 ThioFluor-6 424 502 - 13,820 10.6 ThioFluor-4 466 543 - 29,878 10.9 ThioFluor-10 489 675 0.06 7,050 11.3 ThioFluor-2 515 536 0.13 50,578 11.3 ThioFluor-3 548 679 0.27 21,223 10.9 ThioFluor 558 673 0.31 15,091 11.2 ThioFluor-9 562 700 0.02 15,406 11.1 ThioFluor-8 641 715 0.09 22,948 11.2 ThioFluor-7 714 780 - 22,517 4.3 Generally the absorption wavelength seems to be correlated to the degree of conjugation between the electron donor and electron acceptor groups. The most blue-shifted analogues when coupled with Q108K:K40L:T51V:T53S:R58W:Y19W:L117E are ThioFluor-5 (λabs = 423 nm) and ThioFluor-6 (λabs = 424 nm), which possess a trifluoromethyl group and hydrogen at the para-phenyl position, respectively. The pigments formed are essentially nonfluorescent, but their iminiums do exhibit high pKas. ThioFluor-4 contains the electron donating methoxy group at the para position, and as expected, is more red-shifted than with hydrogen or trifluoromethyl groups (λabs = 159 466 nm). Again, the iminium is essentially nonfluorescent, suggesting that methoxy is not electron donating enough to generate a strong enough push-pull system. Further, N,N-dimethylamino and azetidine groups provide the largest redshifts amongst those tested at the para position (λabs = 558 nm and 548 nm, respectively). Notably, a bathochromic shift of 134 nm is achieved in absorption wavelength via the introduction of a para N,N-dimethyl amino group, as compared to ThioFluor-6 which simply contains an electronically neutral hydrogen atom. Interestingly, the quantum yields of the para substituted N,Ndimethylamino and azetidine groups are similar. Previous studies by the Lavis lab showed that replacement of the N,N-dimethyl group with azetidine (and later aziridine) led to a large increase in quantum yield relative to that of the parent fluorophore in numerous fluorophores such as coumarin, napthalimide, acridine, rhodol, carborhodamine and oxazine.46, 47 It was predicted that azetidine and aziridine might inhibit TICT in these cases. With this knowledge, we are compelled to hypothesize that with ThioFluor, fluorescence is not lost due to nonradiative decay from the excited state as a result of TICT in the case of ThioFluor. Interestingly, the crystal structures of ThioFluor and ThioFluor-3 show that both the N,N-dimethyl and azetidine groups are planar with the rest of the chromophore, indicating that they may exhibit high TICT resistance. Because ThioFluor-11 did not form an iminium with Q108K:K40L:T51V:T53S:R58W:Y19W:L117E, only one modification of the 160 electron acceptor moiety was tested. ThioFluor-7, which contains a trifluoromethyl group in the beta position to the aldehyde led to the most redshifted absorption wavelength obtained in this hCRBPII mutant. From the parent ThioFluor, a remarkable red-shift of 156 nm is obtained from the structurally small modification (λabs = 714 nm). Just as a stronger electron donating group in the para position leads to a bathochromic shift, a red-shift can also be obtained by decreasing the aromaticity of the central rings, such that it is easier for electrons to delocalize across the entire chromophore. As previously described, the resonance energy of benzene (36 kcal/mol) is higher than that of thiophene (29 kcal/mol). For this reason, it should be easier to break aromaticity and resonate through a thiophene π linker, as opposed to a phenyl linker. The overall effect is better delocalization over the whole molecule, leading to a smaller gap between the HOMO and LUMO. To our satisfaction, substitution of the phenyl ring in ThioFluor, to yield the bithiophene scaffold in ThioFluor-8, results in a red-shift of almost 100 nm from ThioFluor (λabs = 641 nm), which is indicative of a more conjugated π system. It is hypothesized that the two aromatic rings in ThioFluor-8 are coplanar when bound to the protein, because ThioFluor is essentially planar in hCRBPII crystal structures. Interestingly, replacement of the thiophene ring in ThioFluor with a second phenyl ring yields an approximately 70 nm hypsochromic shift (λabs = 489 nm), yet with a remarkable Stokes shift of 186 nm. Previously it has been shown 161 that the torsion on push-pull biphenyl analogs enhances their fluorescence solvatochromism. The observed hypsochromic shift in absorption wavelength could occur for various reasons, but we will make two suggestions here. It is possible that steric hindrance forces the biphenyl ThioFluor-10 to remain in a twisted ground state, such that only the two aromatic phenyl rings are coplanar, and the donating N,N-dimethyl is rotated out of the plane (Figure III-5a). This would account for the large blue-shift observed in absorption wavelength. Subsequently, a planarized intramolecular charge transfer (PLICT) could explain the large Stokes shift.49, 50 In this scenario, upon excitation of the molecule, a charge transfer first occurs from the donor to acceptor. Then the single bond between the N,N-dimethyl group and phenyl ring can rotate around 90° in order to stabilize the positive charge on the donor group. Due to the geometric relaxation in the excited state, a large Stokes shift and quantum yield should be observed. An alternative explanation would be almost the exact opposite. If ThioFluor-10 is planar in the ground state when bound to the protein (Figure III5b), it would stand to reason that the large hypsochromic shift in absorption wavelength is solely due to the decrease in delocalization along the molecule, due to the increasing aromaticity of the second ring (from thiophene to phenyl). Assuming, the donor and acceptor moieties are in the same plane in the ground state, excitation could result in an intramolecular charge transfer (ICT) from the donor to acceptor.49-51 Subsequent relaxation in the excited state is achieved by 162 rotation around the single bond connecting the donor and acceptor, yielding a perpendicular conformation. This twisted intramolecular charge transfer (TICT) state typically emits dual fluorescence in emission spectra and results in large Stokes shift. a. planarized ICT (PLICT) b. twisted ICT (TICT) D D’ A D A D’ hυ hυ D D D’ A A D’ rotation D’ D rotation A D D’ A Figure III-5. Schematic representation of the geometric arrangement of the donor and acceptor groups upon excitation in planarized intramolecular charge transfer (PLICT) and twisted intramolecular charge transfer (TICT). The emission of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor10 appears to follow a normal distribution, and clearly no conclusion can be drawn as to the mechanism which leads to the LSS. One subtle factor that may provide information is the rate of iminium formation. Iminium formation of ThioFluor-10 with Q108K:K40L:T51V:T53S:R58W:Y19W:L117E is not significantly different than with ThioFluor. This suggests that the aldehyde is of comparable electrophilicity. As seen 163 from the crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E with ThioFluor, ThioFluor is planar in the ground state; this would lead one to hypothesize that ThioFluor-10 possesses a completely planar configuration of the chromophore in the ground state. Nonetheless, crystal structures may provide more information, as they could provide details of the ground state chromophore configuration. It would also be interesting to investigate the spectroscopic properties of azetidine substituted ThioFluor-10. If fluorescence does involve a TICT mechanism, the inclusion of azetidine may decrease or eliminate TICT and improve fluorescence properties. Changing the order of the phenyl and thiophene rings does not seem to have a significant effect on the absorption wavelength. However, the quantum yield is drastically decreased in comparison to ThioFluor. They are 0.31 and 0.02 for ThioFluor and ThioFluor-9, respectively. Interestingly, removal of the phenyl ring altogether leads to an absorption maxima of 515 nm, which will be shown to remain constant in all proteins tested. ThioFluor-2 was evaluated because it was expected to lead to a more Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor-2 cyanine-like dye. exhibits high a iminium pKa, high extinction coefficient and extremely small Stokes shift. Both of these characteristics are reminiscent of typical cyanine dyes. A plot of absorption wavelength versus emission wavelength is given in Figure III-6a. In general a linear trend is observed. However, two outliers are clearly observed. These belong to Q108K:K40L:T51V:T53S:R58W:Y19W:L117E coupled with ThioFluor-2 and ThioFluor-10. As ThioFluor-2 is cyanine-like and 164 the wavelength is not regulatable (to be shown later) due to its more permanent resonance, it is acceptable to remove this data from the series. On the other hand, the emission of ThioFluor-10 is much more red-shifted than would have been predicted; Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor-10 exhibits a Stokes shift of 186 nm. This is likely due to an intramolecular charge transfer from the electron donating N,N-dimethyl amino group to the electron accepting iminium. A plot of absorption wavelength versus emission wavelength shows a linear correlation (Figure III-6b). y = 143.9 + 0.9168x R2= 0.7399 b. y = 121.5 + 0.9561x R2= 0.9088 800 800 750 750 700 700 Emission (nm) Emission (nm) a. 650 600 550 650 600 550 500 500 450 350 400 450 500 550 600 650 700 750 Absorbance (nm) 450 350 400 450 500 550 600 650 700 750 Absorbance (nm) Figure III-6. Emission maxima versus absorption maxima for ThioFluor analogues bound to hCRBPII heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. Rate constants and half-lives are given in Table III-5. All rates were plotted according to second order kinetics (measured by UV-Vis at 23 °C with 20 μM protein and 0.5 equiv ligand at pH ~ 7). Drastically different rates of iminium formation are observed. The fastest reactions are those that possess the weakest push-pull system. Presumably the trifluoromethyl group and hydrogen in the para position, in ThioFluor-5 and ThioFluor-6, leads to a more electrophilic 165 aldehyde. Following the same trend, a methoxy substituent results in almost a five-fold enhancement of rate. Interestingly, reversal of the thiophene and phenyl rings leads to a six-fold increase in reaction rate, suggesting that donation of the N,N-dimethylamino group to the aldehyde is much less when aromaticity has to be broken further from the electron donating site. In other words, the aldehyde is more electrophilic in ThioFluor-9 than ThioFluor. This is further supported by the extremely low quantum yield Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor-9 as compared of to Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor. Unfortunately, the rate of iminium formation cannot be measured with ThioFluor-7 under analogous conditions, as the pKa is several units lower than necessary to observe iminium formation. Table III-5. Rates and half-lives of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. k (min-1M-1) t1/2 (min) ThioFluor-6 31,151 3 ThioFluor-5 10,699 9 ThioFluor-9 7,624 13 ThioFluor-4 6,001 17 ThioFluor-3 3,659 27 ThioFluor 1,224 82 ThioFluor-10 1,082 92 ThioFluor-8 284 352 ThioFluor-2 89 1124 Ligand 166 Conversely, the most conjugated systems lead to much slower iminium formation. Both the cyanine like ThioFluor-2 and bithiophene analog ThioFluor8 are substantially slower than ThioFluor. These slow reaction rates would find essentially no use in live cell imaging experiments. Upon analysis of ThioFluor analogues with Q108K:K40L:T51V:T53S:R58W:Y19W:L117E, the following general trends are concluded, which will be discussed in greater detail below. • As expected, absorption wavelength seems to be correlated to the degree of conjugation between the electron donor and electron acceptor groups. • Emission wavelength is generally linearly correlated with absorption wavelength, albeit no significant environmental factors are in play. • The reaction rate is faster for those chromophores containing electronwithdrawing substituents in the para position, which is not surprising, as less electron donation should result in a more electrophilic aldehyde center. • The iminium pKa of all the hCRBPII/ThioFluor analog complexes, with the exception of ThioFluor-7, are high. This indicates that the ligand experiences the same iminium environment, presumably due to an ionic interaction between the protonated nitrogen of the iminium and the negatively charge carboxylate of the neighboring L117E residue. It is predicted that the pKa would be variable were this interaction was not present. 167 • Quantum yield is hard to assess. However, it is quite apparent that a pushpull system needs to be present in order to show fluorescence. Three other hCRBPII mutants were chosen for binding with the ThioFluor analogues to verify the trends observed. III.5 Coupling of ThioFluor analogues to Q108K:K40L:T51V:T53S:R58W:Y19W:A33W As previously mentioned, the iminium pKa of ThioFluor analogs with the heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E were high, presumably due to stabilization via the carboxylate installed at residue 117. We now turned our attention to a protein of similar sequence, but with no stabilizing ionic interaction at the iminium. It is hypothesized that without the interaction, the pKa of the iminium will be variable, dependent on chromophore structure. The heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:A33W was chosen because it is one of the more red-shifted mutants when coupled with ThioFluor (λabs = 666 nm), and it was desirable to see if the analogues of ThioFluor could also regulate wavelength, as was shown in Chapter II. Spectroscopic properties of the iminium formed between ThioFluor analogs and Q108K:K40L:T51V:T53S:R58W:Y19W:A33W are given in Table III-6. It is apparent that the absorption wavelength is more red-shifted as conjugation is extended across the molecule. Para substitution of trifluoromethyl (ThioFluor-5, λabs = 484 nm), hydrogen (ThioFluor-6, λabs = 495 nm), methoxy (ThioFluor-4, λabs = 530 nm) and N,N-dimethyl (ThioFluor, λabs = 666 nm) 168 becomes increasingly more red-shifted with the increased electron donating ability of the para substituent. It appears that the simple introduction of N,Ndimethyl at the para position is responsible for a 171 nm red-shift in absorption wavelength, five unit increase in iminium pKa and quantum yield of 0.14. Table III-6. Spectroscopic properties of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W:Y19W:A33W. ε Ligand λabs λem Φ pKa -1 (M cm-1) ThioFluor-5 484 - - 39,228 3.7 ThioFluor-6 495 - - 36,566 4.0 ThioFluor-2 519 536 - 36,594 11.1 ThioFluor-4 530 577 - 24,875 9.1 ThioFluor-10 603 733 - 15,547 8.7 ThioFluor-3 658 732 0.10 40,965 9.2 ThioFluor 666 724 0.14 57,316 9.3 ThioFluor-9 693 769 - 31,531 9.5 ThioFluor-11 694 744 - - - ThioFluor-8 722 751 0.02 62,382 9.1 ThioFluor-7 797 - - - - It should be noted that the iminium pKa in both complexes of ThioFluor-5 and ThioFluor-6 with Q108K:K40L:T51V:T53S:R58W:Y19W:A33W are approximately five units lower than ThioFluor. This can be due to the fact that there is no stabilization of the iminium via an electron donating para substituent; it is also possible that the opposite isomer of the iminium is formed, which exhibits a decreased pKa. As opposed Q108K:K40L:T51V:T53S:R58W:Y19W:L117E analogues, the quantum 169 to coupled the heptamutant with ThioFluor yields of Q108K:K40L:T51V:T53S:R58W:Y19W:A33W/ThioFluor analogues are much smaller or nonexistent. For example, the quantum yield of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (excited at 558 nm) is 0.31, while Q108K:K40L:T51V:T53S:R58W:Y19W:A33W/ThioFluor is 0.14 (excited at 666 nm). The quantum yield can be decreased as a result of the bathochromic shift of over 100 nm, or due to the removal of L117E, which may confer restriction on the chromophore. Modification of the electron acceptor moiety with alpha fluoro (ThioFluor11) and beta trifluoromethyl (ThioFluor-7) groups leads to red-shifts of 28 nm and 131 nm from the parent ThioFluor, Q108K:K40L:T51V:T53S:R58W:Y19W:A33W/ThioFluor-7 respectively. absorbs in the infrared at an astonishing 797 nm, but does suffer from a severely depressed pKa (it cannot be measured because it is too low) and fluorescence is not observed upon excitation of the iminium at 797 nm. It should be noted that our fluorometer does restrict fluorescence observation to below 900 nm; although we do not think the complex will emit higher than this cutoff, it has to be considered as a possibility. As previously suggested, the aromaticity of the central hetero rings can significantly alter the spectroscopic properties of the hCRBPII bound ligand. Better delocalization across the molecule is observed for the bithiophene ThioFluor-8 (λabs = 722 nm) as compared to ThioFluor (λabs = 666 nm). The biphenyl ThioFluor-10 (λabs = 603 nm) is again more blue-shifted. Removal of 170 the phenyl ring results in an iminium with absorption maxima of 519 nm, which is consistent with other hCRBPII mutants tested. Rate constants and half-lives are given in Table III-7. All rates were plotted according to second order kinetics (measured by UV-Vis at 23 °C with 20 μM protein and 0.5 equiv ligand at pH ~ 7). Varying rates of iminium formation are observed with the different ThioFluor analogues. Rate constants for many of the ligands could not be measured by UV-Vis because of their extremely low iminium pKas; additionally, the ligands were not fluorescent when bound to Q108K:K40L:T51V:T53S:R58W:Y19W:A33W, negating the measurement of rates by fluorescence. Generally, the rate of iminium formation is slower than those observed with Q108K:K40L:T51V:T53S:R58W:Y19W:L117E (Table III-5). Presumably, the carboxylate at residue 117 provides some means of acid catalyzed iminium formation. It is also possible that a different isomer of the iminium is formed, yet there is not direct evidence to support this. As similarly reported for the previous hCRBPII heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E, the rate of iminium formation is faster with the less electron donating methoxy group at the para position (ThioFluor-4) as compared to the N,N-dimethyl group. The fastest rate is observed for ThioFluor-9, which is presumably due to its more electrophilic aldehyde. The only pitfall is that the hCRBPII/ThioFluor-9 complex is not fluorescent. 171 Table III-7. Rates and half-lives of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W:Y19W:A33W. k (min-1M-1) t1/2 (min) ThioFluor-9 6252 16 ThioFluor-4 2903 34 ThioFluor-10 503 199 ThioFluor-3 452 221 ThioFluor-8 284 352 ThioFluor 238 420 Ligand III.6 Coupling of ThioFluor analogues to Q108K:K40L:R58F Similar trends in absorption wavelength were obtained for the hCRBPII triple mutant Q108K:K40L:R58F as for the heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:A33W. Spectroscopic properties for the iminium formed upon condensation of Q108K:K40L:R58F and the ThioFluor analogues are detailed in Table III-8. Para substitution of trifluoromethyl (ThioFluor-5, λabs = 442 nm), hydrogen (ThioFluor-6, λabs = 458 nm), methoxy (ThioFluor-4, λabs = 496 nm) and N,Ndimethyl (ThioFluor, λabs = 597 nm) becomes increasingly more red-shifted with the increased electron donating ability of the para substituent. As observed in the previous two hCRBPII mutants, only the iminium with ThioFluor is fluorescent, exhibiting a quantum yield of 0.18. It should be noted that the iminium pKas for the ligands are much more similar to each other. This is in Q108K:K40L:T51V:T53S:R58W:Y19W:A33W, contrast in to which complexes with ThioFluor-5 and ThioFluor-6 had iminium pKa values five units below that of ThioFluor. It is 172 predicted that the large change in pKa is due to isomerization of the iminium (this will be further supported by coupling of the ligands to Q108K:K40L:T51V:T53S:R58W in Section III.7). Our lab has previously observed the trend that CRABPII/retinal cis imine showed a higher pKa than the trans imine;52, 53 presumably, a similar phenomena is occurring here. Table III-8. Spectroscopic properties of ThioFluor-PSB analogues coupled to Q108K:K40L:R58F. ε Ligand λabs λem Φ pKa -1 (M cm-1) ThioFluor-5 442 - - 13,721 8.0 ThioFluor-6 458 509 - 12,634 8.4 ThioFluor-10 487 675 0.02 4,632 8.2 ThioFluor-4 496 554 - 31,580 8.8 ThioFluor-2 518 534 - 34,188 10.9 ThioFluor-3 581 683 0.13 24,105 8.7 ThioFluor 597 678 0.18 18,136 9.3 ThioFluor-9 602 709 - 14,235 8.2 ThioFluor-11 619 695 - - - ThioFluor-8 673 712 0.06 20,147 9.1 Absorption wavelength is altered via substitution of the central aromatic rings. In the case of the bithiophene containing ThioFluor-8, the absorption maxima for the iminium is 673 nm, while the biphenyl ThioFluor-10 shows a hypsochromic shift, with absorption maxima of 487 nm. Removal of the phenyl ring (ThioFluor-2) results in an iminium with absorption maxima of 518 nm, which is shown to be static in all hCRBPII mutants tested. 173 III.7 Coupling of ThioFluor analogues to Q108K:K40L:T51V:T53S:R58W The general trends already discussed for the hCRBPII/ThioFluor analogues complexes are applicable here, including absorption regulation and the necessity of a push-pull system to generate fluorescence. Spectroscopic data for the iminium complexes are presented in Table III-9. Table III-9. Spectroscopic properties of ThioFluor-PSB analogues coupled to Q108K:K40L:T51V:T53S:R58W. λabs λem Φ pKaa ThioFluor-5 446 503 - - ThioFluor-6 469 517 - nd, 7.1 ThioFluor-4 502 556 - 4.8, 8.4 ThioFluor-2 516 531 - 10.9 ThioFluor-10 517 684 - 7.7 ThioFluor-3 610 699 0.08 6.4, 8.7 ThioFluor 623 697 0.08 6.4, 8.4 ThioFluor-9 639 745 - nd, 8.4 ThioFluor-11 661 723 - 4.3 ThioFluor-8 706 736 0.02 8.1 Ligand a Presumably the two pKas correspond to the trans and cis iminiums. Only one data appears to be an outlier from the trends observed thus far. In the cases of hCRBPII mutants Q108K:K40L:T51V:T53S:R58W:Y19W:A33W and Q108K:K40L:R58F, the iminium formed with ThioFluor-8 (λabs = 722 nm and λabs = 673 nm, respectively) is more red-shifted than for ThioFluor-11 (λabs = 694 nm and λabs = 619 nm, respectively). In this case, the bithiophene ThioFluor-8 exhibits a more red-shifted iminium than ThioFluor-11. 174 0.3 0.2 4.3 4.7 5.2 5.7 6.2 6.6 0.3 7.1 7.9 9.3 10.4 10.9 ΔAbs at 623 nm (a. u.) Abs. (a. u.) a. 0.1 0.1 4 350 400 450 500 550 600 650 700 750 Wavelength (nm) 4.2 4.7 5.2 5.6 6.1 6.6 7.2 7.7 8.6 9.9 10.5 11.2 0.2 9 10 11 pKa = 8.7 R2 = 0.99 0.1 4 2.0 2.9 3.9 4.4 4.8 5.3 5.8 6.3 6.7 7.1 7.9 8.5 5 6 7 8 pH 9 10 11 12 0.4 9.5 10.3 10.7 11.2 ΔAbs at 502 nm (a. u.) Abs. (a. u.) 8 pKa = 6.4 R2 = 0.99 0.2 350 400 450 500 550 600 650 700 750 Wavelength (nm) 0.4 7 0 0 0.5 6 0.3 0.1 c. 5 pH ΔAbs at 610 nm (a. u.) Abs. (a. u.) 0.3 pKa = 8.4 R2 = 0.98 0 0 b. pKa = 6.4 R2 = 0.99 0.2 0.3 0.2 0.1 0 300 350 400 450 500 550 600 650 Wavelength (nm) pKa = 4.8 R2 = 0.99 0.3 0.2 pKa = 8.4 R2 = 0.99 0.1 0 2 4 6 8 10 12 pH Figure III-7. Double iminium pKas observed upon coupling of hCRBPII pentamutant Q108K:K40L:T51V:T53S:R58W with a. ThioFluor, b. ThioFluor-3 and c. ThioFluor-4. Absorption intensity of the iminium is plotted as a function of pH. Two curve fittings provide the two different pKa values. 175 While absorption wavelength regulation supports the general trends already observed in the other hCRBPII mutants, one interesting new observation was made with regard to the iminium pKa. The striking difference in this hCRBPII mutant is the observation of two independent pKas for the iminium. For this reason, extinction coefficients are not presented, as the concentration of the iminium complex cannot be accurately determined. The pKa curves are presented for those hCRBPII/ligand complexes which clearly show two pKa curves (Figure III-7). These two pKas presumably correspond to the cis and trans isomers of the iminium; Dr. Meisam Nosrati has previously shown that these two species to be present crystallographically when retinal was used as a ligand in CRABPII mutants.53 It should be noted that the iminium with the lower pKa is also observed with ThioFluor-9 and ThioFluor-6, but the pKa is too low to measure in this hCRBPII mutant, due to protein stability in acidic conditions. With this knowledge in hand, all pKa data should be considered cautiously, as the values are not only dependent on the chromophore electronics, but also the isomer of the iminium, which can lead to drastically different pKa values. Presumably, the chromophore structure dictates the isomer of the iminium formed. However, at this time, no conclusions can be drawn as to the exact origin; crystal structures may provide further insight. III.8 Conclusions and proposed projects While none of the analogues investigated in this chapter will be deemed suitable for use for fluorescent live cell imaging (due to their low quantum yields), 176 the data presented in this chapter is no less pertinent. The structure-property relationships revealed should establish design parameters for improved dyes. To conclude, the following relationships have been determined: 1. Extended conjugation leads to a bathochromic shift in absorption wavelength. Emission wavelength is linearly correlated to absorption wavelength, when no other environmental factors (i.e. intramolecular charge transfer) are present. 2. A more electrophilic aldehyde (through either less donation from an electron donating group or increased aromaticity of the hetero rings) is correlated with an increased rate of iminium formation. 3. A push-pull system is necessary for fluorescence. When electron withdrawing or weakly electron donating groups are substituted at the para position of the phenyl ring of the chromophore, a small peak in fluorescence is detectable, which equates to a less than 0.01 quantum yield. 4. The iminium pKa can be manipulated by protein environment (such as the introduction of a counter anion near the iminium), but in general a more conjugated system provides a higher pKa. This is most apparent by the cyanine-like dye ThioFluor-2, which in all proteins tested retained a high pKa. As an example of the previous four conclusions, plots of Hammett value versus absorption wavelength, rate of iminium formation and pKa for select ThioFluor 177 analogues (ThioFluor, ThioFluor-4, ThioFluor-6) coupled with the hCRBPII mutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E are provided in Figure III8. y = 2.615 - 0.006177x R2= 0.9998 0.2 -H p Hammett Value (σ ) 0 S CHO R ThioFluor-6; R = H ThioFuor-4; R = OCH3 ThioFluor; R = N(CH3)2 -0.2 -OCH3 -0.4 -0.6 -0.8 -N(CH3)2 -1 420 440 460 480 500 520 540 560 Absorbance (nm) y = 0.00445 - 0.01031x R2= 0.9505 0 0 p Hammett Value (σ ) 0.2 p Hammett Value (σ ) y = 14.71 - 1.383x R2= 0.9609 0.2 -0.2 -0.4 -0.6 -0.2 -0.4 -0.6 -0.8 -0.8 -1 10.5 10.6 10.7 10.8 10.9 11 11.1 11.2 11.3 pK a -1 0 20 40 60 t (min) 80 100 1/2 Figure III-8. Plots of Hammett value versus absorption wavelength, rate of iminium formation and pKa for ThioFluor, ThioFluor-4 and ThioFluor-6 coupled with the hCRBPII mutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. III.8.1 Proposed design of a multi-input fluorogenic probe The structure-property relationships described in this chapter can be exploited to generate a truly fluorogenic (turn-on fluorescence) system, with fast iminium formation. As was shown in Chapter II.4, ThioFluor is fluorescent in numerous solvents, yet it can be used in live cell imaging. This is because the 178 hCRBPII/ThioFluor complex is much more red-shifted than the free aldehyde in solution, and can thus be specifically excited as long as iminium formation does not occur with off target lysines. However, it would be ideal to use a probe that is always nonfluorescent. Free aldehyde will then provide no background fluorescence, whether in aqueous media or embedded in the hydrophobic cellular membranes (recall that most solvatochromic fluorophores are fluorescent in hydrophobic environments). Additionally, even if the aldehyde reacts with off target lysines, fluorescence from these nonspecific reactions will not be apparent, because both the imine and iminium are still not fluorescent. This should hold true for fluorophores that contain an electron withdrawing or weakly electron donating group at the para position of the phenyl substituent. The proposed fluorophore and reaction scheme when bound to hCRBPII is shown in Figure III-9. Hammett values are also provided as a reference. Turn on fluorescence would require four inputs including: 1) iminium formation with hCRBPII, 2) nucleophilic substitution reaction with an active site glutamic acid or aspartic acid residue, 3) hydrolysis of the resultant ester, and 4) deprotonation of the hydroxyl group. The concept of multi-input fluorescent probes, which require two or more parameters to turn on fluorescence, is not a new one, though it does appear to be a severely underdeveloped field. Probes have been developed that require chromophore binding to a target, followed by a change in pH, chelation to a metal or enzymatic reaction to produce a fluorescent probe.54-60 179 Iminium formation with hCRBPII SNAr with active site Glu/Asp Fluorescence Hydrolysis of ester Deprotonation of hydroxyl H N Lys108 hCRBPII-Lys108 O S S Cl Cl H+ H N Lys108 H N Lys108 SNAr S O Glu/Asp S O Cl H 2O H N Lys108 H N Lys108 H+ S HO S σp (Cl) = 0.23 σp (OCOMe) = 0.31 σp (OH) = - 0.37 σp (O-) = -0.81 O Figure III-9. Proposed fluorophore and reaction scheme for an improved fluorogenic probe. It is suggested that a para-chloro substituent be used in the same scaffold as ThioFluor, as the system has now been well developed and the probe shown to be cell permeable. In the aldehydic, imine and iminium states, quantum yield is expected to be less than 0.01. Additionally, iminium formation should proceed swiftly due to the electron withdrawing nature of the chlorine group in conjugation with the aldehyde. Upon iminium formation, the aromatic chlorine is set up for nucleophilic aromatic substitution (a similar mechanism is suggested for the dehalogenation in the enzyme 4-chlorobenzoyl-CoA dehalogenase), being 180 activated by the electron withdrawing iminium. This is envisioned by attack of a nearby aspartic acid or glutamic acid residue. This intermediate still would not be fluorescent, predicted by its high Hammett value, even if off target reactions occur. Further hydrolysis of the ester should result in the unmasking of a hydroxyl group, which can be deprotonated by a nearby basic residue such as histidine to turn on fluorescence. It is predicted that fluorescent properties will be similar to ThioFluor, as the Hammett values of N,N-dimethyl and phenoxide are similar. The proposal outlined is similar to reactions catalyzed by hydrolytic aromatic dehalogenases, chlorothalonil dehalogenase, including 4-chlorobenzoyl-CoA atrazine dehalogenase,61 as dehalogenase, well as the nonhydrolytic dehalogenase glutathione-S transferase.62 The best studied hydrolytic aromatic dehalogenase is 4-chlorobenzoyl-CoA dehalogenase.61, 63-71 This enzyme is found in several 4-chlorobenzoate degrading bacteria, and is responsible for catalyzing the conversion of 4-chlorobenzoyl-CoA to 4hydroxybenzoyl-CoA (Figure III-10). The steady state kcat and Km values measured for wild-type enzyme are 0.60 s-1 and 3.7 μM, respectively.67 The first step is the ligation of CoA to the substrate (4-chlorobenzoate) via 4-chlorobenzoate CoA ligase. The CoA ligated product is then bound by 4chlorobenzoyl-CoA dehalogenase (ES). The mechanism proceeds by nucleophilic aromatic substitution (SNAr) through an enzyme-Meisenheimer (EMc) intermediate, via Asp145.68 The collapse of this intermediate, with the expulsion of a chloride ion, leads to the arylated enzyme (EAr) complex. This 181 product is attacked by a water molecule, which is activated by His90, to form the 4-hydroxybenzoyl-CoA product (EP).70 Phe64 Phe64 Glu114 N H O O Asp145 Glu114 H N N N H H O SCoA Cl H O O Cl Asp145 O H SCoA His90 H N NH NH Trp137 ES EMc Phe64 Glu114 N H N N H O His90 N Phe64 N O H H N N Trp137 Glu114 H O H H O SCoA SCoA Cl O H N Trp137 H OH O Asp145 O H O Asp145 O His90 N H N NH ES Trp137 H His90 N NH EMc Figure III-10. Proposed chemical steps of the Pseudomonas sp. strain 4chlorobenzoyl-CoA dehalogenase-catalyzed conversion of 4-chlorobenzoyl-CoA to 4-hydroxybenzoyl-CoA. The crystal structure provides insight into the favorable interactions between the enzyme and product ligand (Figure III-11).72 4-Chlorobenzoyl-CoA is bound in a primarily hydrophobic environment, except for the catalytic Asp145 and His90 residues. The aromatic groups (Phe64, Phe82, Trp89 and Trp137) 182 around the benzoyl ring assist in the polarization of the π-electrons in the benzoyl ring during the reaction.65, 73 It was shown that Trp137 (which forms a hydrogen bond with Asp145) has a significant effect on the formation of the Meisenheimer intermediate; the mutation W137F causes a 20,000-fold reduction in the rate of formation of the Meisenheimer intermediate,66 but does not affect binding of the substrate;65 presumably Trp137 positions Asp145 for catalysis. 4-chlorobenzoyl coenzyme A Trp137 3.1 Å Asp145 3.1 Å 3.2 Å Phe82 Phe64 4.9 Å His90 Trp89 Figure III-11. Crystal structure of 4-chlorobenzoyl coenzyme A dehalogenase from Pseudonomas sp. Catalytic residues are shown in cyan and aromatic residues responsible for hydrophobic packing are shown in green. Coordinates obtained from PDB 1NZY. From the crystal structure, it is also evident that the thioester carbonyl of the ligand is within hydrogen-bonding distance of the backbone amide hydrogen atoms of Gly114 and Phe64 (Figure III-10).72 Catalysis in F64P, F64A and G114A dehalogenase mutants were shown to be less active than the wild-type enzyme, while the binding affinities of the ligands are unchanged.67 183 The proposed system should meet most of the requirements outlined at the beginning of this chapter. Fluorescence should only be turned on when coupled to the target hCRBPII (with multiple inputs). This would allow for high specificity, in which fluorescence should never be observed from free probe or if it does bind to off target lysines. Additionally, upon reaction of the probe with hCRBPII, the fluorophore should be red-shifted (due to the presence of the electron donating phenoxide). Lastly the probe should prove to be cell permeable, as the analogue ThioFluor has already shown utility in live cell imaging. It should be noted that this scheme requires three reactions – iminium formation, SNAr and hydrolysis of the ester; it is hard predict the reaction kinetics under such rigorous requirements. III.8.2 Development of near infrared and large Stokes shift fluorescent probes Other noteworthy observations include the LSS observed with ThioFluor10 when coupled with select hCRBPII mutants. Preliminary results show a Stokes shift of 186 nm and a quantum yield of 0.06 in the hCRBPII heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. It is suggested that hCRBPII mutants be optimized for binding to ThioFluor-10, in order to improve brightness and be applicable as a LSS fluorescent probe. ThioFluor-7, or similar analogues, with a beta trifluoromethyl group (or perhaps other electron withdrawing groups) to the aldehyde can be developed into near infrared probes. The current pitfall with ThioFluor-7 is the severely 184 depressed pKa and the lack of fluorescence upon excitation of the PSB. It is suggested that other fluorophore scaffolds that are inherently brighter than ThioFluor, be screened for electron withdrawing groups near the aldehyde. It is possible that halogens such as the trifluoromethyl group in ThioFluor-7 and the fluorine in ThioFluor-11 are responsible for the fluorescence quenching observed. 185 APPENDIX 186 APPENDIX H O S n-BuNH2, N R HCl Bu H N R Ethanol Bu N ThioFluor ϕ = 0.063 ThioFluor-SB ThioFluor ThioFluor-SB ThioFluor-PSB 2.5 10 6 Fluorescence Intensity (cps) 0.4 400 nm Abs. (a. u.) 0.3 430 nm ThioFluor-PSB 521 nm 0.2 0.1 540 nm 2 106 654 nm 1.5 10 6 1 106 5 105 689 nm 0 0 400 300 350 400 450 500 550 600 650 Wavelength (nm) 500 600 700 Wavelength (nm) 800 Figure III-12. Formation of ThioFluor-PSB with n-butyl amine in ethanol. H S n-BuNH2, Bu HCl Ethanol ThioFluor-2 ϕ < 0.001 ThioFluor-2-SB ThioFluor-2 ThioFluor-2-SB ThioFluor-2-PSB 1 510 nm 0.8 Abs. (a. u.) N R 0.6 0.4 400 nm 445 nm 0.2 H N R Bu ThioFluor-2-PSB 4 105 Fluorescence Intensity (cps) N O 531 nm 3.5 10 5 3 105 2.5 10 5 498 nm 2 105 1.5 10 5 0 300 350 400 450 500 550 600 650 Wavelength (nm) 1 105 5 10 485 nm 4 0 400 500 Wavelength (nm) Figure III-13. Formation of ThioFluor-2-PSB with n-butyl amine in ethanol. 187 600 H n-BuNH2, O S N R HCl Bu Ethanol R H N Bu N ThioFluor-3 ϕ = 0.060 ThioFluor-3-SB 392 nm Abs. (a. u.) 0.3 418 nm 508 nm 0.2 0.1 3.5 10 6 Fluorescence Intensity (cps) ThioFluor-3 ThioFluor-3-SB ThioFluor-3-PSB 0.4 ThioFluor-3-PSB 550 nm 3 106 2.5 10 6 2 106 1.5 10 6 562, 626, 660 nm 1 106 5 105 0 690 nm 0 400 300 350 400 450 500 550 600 650 Wavelength (nm) 500 600 700 Wavelength (nm) 800 Figure III-14. Formation of ThioFluor-3-PSB with n-butyl amine in ethanol. H S O n-BuNH2, N R Bu HCl Ethanol ThioFluor-4 ϕ = 0.005 ThioFluor-4 ThioFluor-4-SB ThioFluor-4-PSB 0.4 0.3 ThioFluor-4-SB 358 nm 440 nm Abs. (a. u.) 384 nm 0.2 0.1 2.5 10 6 Fluorescence Intensity (cps) H3CO 0 300 R H N ThioFluor-4-PSB 440 nm 2 106 1.5 10 6 556 nm 1 10 6 5 10 5 447 nm 0 350 400 450 500 Wavelength (nm) 400 550 500 600 Wavelength (nm) 700 Figure III-15. Formation of ThioFluor-4-PSB with n-butyl amine in ethanol. Bu 188 H O S n-BuNH2, N R HCl Bu Ethanol ThioFluor-5 ϕ < 0.001 ThioFluor-5-SB ThioFluor-5 ThioFluor-5-SB ThioFluor-5-PSB 0.3 352 nm 360 nm 403 nm Abs. (a. u.) 0.2 0.1 0 300 Bu ThioFluor-5-PSB 5 105 Fluorescence Intensity (cps) F3C R H N 399 nm 4 10 5 3 105 402 nm 2 105 1 105 489 nm 0 350 400 450 Wavelength (nm) 400 500 500 Wavelength (nm) 600 Figure III-16. Formation of ThioFluor-5-PSB with n-butyl amine in ethanol. H O S n-BuNH2, N R HCl Bu R H N Ethanol Bu H ThioFluor-6 ϕ < 0.001 ThioFluor-6-SB 347 nm 364 nm 418 nm Abs. (a. u.) 0.2 0.1 6 105 Fluorescence Intensity (cps) ThioFluor-6 ThioFluor-6-SB ThioFluor-6-PSB 0.3 0 300 5 10 ThioFluor-6-PSB 395 nm 5 4 105 3 105 2 105 396 nm 503 nm 1 105 0 350 400 450 Wavelength (nm) 400 500 500 Wavelength (nm) 600 Figure III-17. Formation of ThioFluor-6-PSB with n-butyl amine in ethanol. 189 O Bu N H S N n-BuNH2, CF3 R Ethanol ThioFluor-7 Bu N H HCl R CF3 CF3 ThioFluor-7-SB ThioFluor-7-PSB HO R ϕ = 0.562 ThioFluor-7 ThioFluor-7-SB ThioFluor-7-PSB ThioFluor-7+Et N 3 340 nm Abs. (a. u.) 0.2 397 nm 568 nm 438 nm 0.1 Fluorescence Intensity (cps) 0.3 F3C OEt 0 300 400 500 600 Wavelength (nm) 5 105 4 105 3 105 2 105 1 105 0 400 1.2 10 7 1 107 8 106 6 106 4 106 2 106 0 487, 585 nm Ex @ 438 nm Ex @ 397 nm 601 nm 500 600 Ex @ 320 nm Ex @ 317 nm Ex @ 336 nm 484 nm 401, 478 nm 400 700 700 401,483 nm 500 Wavelength (nm) 600 Figure III-18. Formation of ThioFluor-7-PSB with n-butyl amine in ethanol. H n-BuNH2, O S S N R Bu HCl Ethanol R H N Bu N ThioFluor-8 ϕ = 0.017 ThioFluor-8-SB Abs. (a. u.) 600 nm 0.2 432 nm 472 nm 0.1 0 300 1.2 10 6 Fluorescence Intensity (cps) ThioFluor-8 ThioFluor-8-SB ThioFluor-8-PSB 0.3 ThioFluor-8-PSB 584 nm 1 106 8 105 6 105 4 105 669 nm 2 105 709 nm 0 400 500 600 Wavelength (nm) 500 700 600 700 Wavelength (nm) 800 Figure III-19. Formation of ThioFluor-8-PSB with n-butyl amine in ethanol. 190 O H S N n-BuNH2, HCl R Ethanol ThioFluor-9 ϕ < 0.007 ThioFluor-9-SB ThioFluor-9 ThioFluor-9-SB ThioFluor-9-PSB 390 nm Abs. (a. u.) 506 nm 416 nm 0.1 0 6 10 Bu N H 535 nm 5 5 105 4 105 3 105 2 105 592 nm 1 105 690 nm 0 400 300 350 400 450 500 550 600 650 Wavelength (nm) R ThioFluor-9-PSB 7 105 Fluorescence Intensity (cps) 0.3 0.2 N Bu 500 600 700 Wavelength (nm) 800 Figure III-20. Formation of ThioFluor-9-PSB with n-butyl amine in ethanol. O H n-BuNH2, HCl R Ethanol ThioFluor-10 ϕ = 0.011 0.3 ThioFluor-10 ThioFluor-10-SB ThioFluor-10-PSB 349 nm 369 nm 0.2 Abs. (a. u.) Bu ThioFluor-10-SB 443 nm 0.1 6 105 Fluorescence Intensity (cps) N N 0 R N H Bu ThioFluor-10-PSB 488 nm 5 105 4 105 3 105 2 105 556 nm 1 105 681 nm 0 400 250 300 350 400 450 500 550 600 Wavelength (nm) 500 600 700 Wavelength (nm) 800 Figure III-21. Formation of ThioFluor-10-PSB with n-butyl amine in ethanol. 191 F H O S F n-BuNH2, N R HCl Bu Ethanol F R H N Bu N ThioFluor-11 ϕ = 0.023 397 nm Abs. (a. u.) 543 nm 429 nm 0.2 0.1 2.5 10 6 Fluorescence Intensity (cps) ThioFluor-11 ThioFluor-11-SB ThioFluor-11-PSB 0.4 0.3 ThioFluor-11-SB 0 552 nm 2 106 1.5 10 6 1 106 556, 630, 653 nm 5 105 698 nm 0 400 300 350 400 450 500 550 600 650 700 Wavelength (nm) ThioFluor-11-PSB 500 600 700 Wavelength (nm) 800 Figure III-22. Formation of ThioFluor-11-PSB with n-butyl amine in ethanol. 192 REFERENCES 193 REFERENCES 1. Hori, Y. & Kikuchi, K. Protein labeling with fluorogenic probes for no-wash live-cell imaging of proteins. Current Opinion in Chemical Biology 17, 644650 (2013). 2. Baranczak, A., Connelly, S., Liu, Y., Choi, S., Grimster, N.P., Powers, E.T., Wilson, I.A. & Kelly, J.W. Fluorogenic small molecules requiring reaction with a specific protein to create a fluorescent conjugate for biological imaging-what we know and what we need to learn. Biopolymers 101, 484-495 (2014). 3. Bruchez, M.P. Dark dyes-bright complexes: fluorogenic protein labeling. Current Opinion in Chemical Biology 27, 18-23 (2015). 4. Chen, Y.C., Tsao, K. & Keillor, J.W. Fluorogenic protein labelling: a review of photophysical quench mechanisms and principles of fluorogen design. Canadian Journal of Chemistry 93, 389-398 (2015). 5. Andersson, H., Baechi, T., Hoechl, M. & Richter, C. Autofluorescence of living cells. Journal of Microscopy-Oxford 191, 1-7 (1998). 6. Aubin, J.E. Autofluorescence of Viable Cultured Mammalian-Cells. Journal of Histochemistry & Cytochemistry 27, 36-43 (1979). 7. Benson, R.C., Meyer, R.A., Zaruba, M.E. & Mckhann, G.M. Cellular Autofluorescence - Is It Due to Flavins. Journal of Histochemistry & Cytochemistry 27, 44-48 (1979). 8. Muller, M.G., Georgakoudi, I., Zhang, Q.G., Wu, J. & Feld, M.S. Intrinsic fluorescence spectroscopy in turbid media: disentangling effects of scattering and absorption. Applied Optics 40, 4633-4646 (2001). 9. Weissleder, R. A clearer vision for in vivo imaging. Nature Biotechnology 19, 316-317 (2001). 10. Weissleder, R. & Ntziachristos, V. Shedding light onto live molecular targets. Nature Medicine 9, 123-128 (2003). 11. Venkatraman, P., Nguyen, T.T., Sainlos, M., Bilsel, O., Chitta, S., Imperiali, B. & Stern, L.J. Fluorogenic probes for monitoring peptide 194 binding to class II MHC proteins in living cells. Nature Chemical Biology 3, 222-228 (2007). 12. Fujikawa, Y., Urano, Y., Komatsu, T., Hanaoka, K., Kojima, H., Terai, T., Inoue, H. & Nagano, T. Design and Synthesis of Highly Sensitive Fluorogenic Substrates for Glutathione S-Transferase and Application for Activity Imaging in Living Cells. Journal of the American Chemical Society 130, 14533-14543 (2008). 13. Barnett, E.M., Zhang, X., Maxwell, D., Chang, Q. & Piwnica-Worms, D. Single-cell imaging of retinal ganglion cell apoptosis with a cellpenetrating, activatable peptide probe in an in vivo glaucoma model. Proceedings of the National Academy of Sciences of the United States of America 106, 9391-9396 (2009). 14. Hori, Y., Nakaki, K., Sato, M., Mizukami, S. & Kikuchi, K. Development of Protein-Labeling Probes with a Redesigned Fluorogenic Switch Based on Intramolecular Association for No-Wash Live-Cell Imaging. Angewandte Chemie-International Edition 51, 5611-5614 (2012). 15. Hori, Y., Norinobu, T., Sato, M., Arita, K., Shirakawa, M. & Kikuchi, K. Development of Fluorogenic Probes for Quick No-Wash Live-Cell Imaging of Intracellular Proteins. Journal of the American Chemical Society 135, 12360-12365 (2013). 16. Davidson, M.W. & Campbell, R.E. Engineered fluorescent proteins: innovations and applications. Nature Methods 6, 713-717 (2009). 17. Marx, V. Probes: paths to photostability. Nature Methods 12, 187-190 (2015). 18. Cranfill, P.J., Sell, B.R., Baird, M.A., Allen, J.R., Lavagnino, Z., de Gruiter, H.M., Kremers, G.J., Davidson, M.W., Ustione, A. & Piston, D.W. Quantitative assessment of fluorescent proteins. Nature Methods 13, 557562 (2016). 19. Chudakov, D.M., Lukyanov, S. & Lukyanov, K.A. Fluorescent proteins as a toolkit for in vivo imaging. Trends in Biotechnology 23, 605-613 (2005). 20. Shaner, N.C., Steinbach, P.A. & Tsien, R.Y. A guide to choosing fluorescent proteins. Nature Methods 2, 905-909 (2005). 21. Johnson, I.D. in Handbook Of Biological Confocal Microscopy, Edn. 3rd. (ed. J.B. Pawley) 353-367 (Springer, 2006). 195 22. Remington, S.J. Fluorescent proteins: maturation, photochemistry and photophysics. Current Opinion in Structural Biology 16, 714-721 (2006). 23. Olenych, S.G., Claxton, N.S., Ottenberg, G.K. & Davidson, M.W. The fluorescent protein color palette. Current Protocols in Cell Biology 21.5, 134 (2007). 24. Day, R.N. & Davidson, M.W. The fluorescent protein palette: tools for cellular imaging. Chemical Society Reviews 38, 2887-2921 (2009). 25. Sample, V., Newman, R.H. & Zhang, J. The structure and function of fluorescent proteins. Chemical Society Reviews 38, 2852-2864 (2009). 26. Piatkevich, K.D., Efremenko, E.N., Verkhusha, V.V. & Varfolomeev, D.D. Red fluorescent proteins and their properties. Russian Chemical Reviews 79, 243-258 (2010). 27. Heim, R., Prasher, D.C. & Tsien, R.Y. Wavelength Mutations and Posttranslational Autoxidation of Green Fluorescent Protein. Proceedings of the National Academy of Sciences of the United States of America 91, 12501-12504 (1994). 28. Cubitt, A.B., Heim, R., Adams, S.R., Boyd, A.E., Gross, L.A. & Tsien, R.Y. Understanding, Improving and Using Green Fluorescent Proteins. Trends in Biochemical Sciences 20, 448-455 (1995). 29. Reid, B.G. & Flynn, G.C. Chromophore formation in green fluorescent protein. Biochemistry 36, 6786-6791 (1997). 30. Shcherbakova, D.M., Subach, O.M. & Verkhusha, V.V. Red fluorescent proteins: advanced imaging applications and future design. Angewandte Chemie-International Edition 51, 10724-10738 (2012). 31. Shu, X., Shaner, N.C., Yarbrough, C.A., Tsien, R.Y. & Remington, S.J. Novel chromophores and buried charges control color in mFruits. Biochemistry 45, 9639-9647 (2006). 32. Wang, L., Xie, J.M., Deniz, A.A. & Schultz, P.G. Unnatural amino acid mutagenesis of green fluorescent protein. Journal of Organic Chemistry 68, 174-176 (2003). 33. Wang, F., Niu, W., Guo, J. & Schultz, P.G. Unnatural amino acid mutagenesis of fluorescent proteins. Angewandte Chemie-International Edition 51, 10132-10135 (2012). 196 34. Young, D.D., Jockush, S., Turro, N.J. & Schultz, P.G. Synthetase polyspecificity as a tool to modulate protein function. Bioorganic & Medicinal Chemistry Letters 21, 7502-7504 (2011). 35. Bonnett, R., Harriman, A. & Kozyrev, A.N. Photophysics of Halogenated Porphyrins. Journal of the Chemical Society-Faraday Transactions 88, 763-769 (1992). 36. Ozhalici-Unal, H., Pow, C.L., Marks, S.A., Jesper, L.D., Silva, G.L., Shank, N.I., Jones, E.W., Burnette, J.M., Berget, P.B. & Armitage, B.A. A rainbow of fluoromodules: A promiscuous scFv protein binds to and activates a diverse set of fluorogenic cyanine dyes. Journal of the American Chemical Society 130, 12620-12621 (2008). 37. Ptaszek, M. Rational Design of Fluorophores for In Vivo Applications. Progress in Molecular Biology and Translational Science 113, 59-108 (2013). 38. Yuan, L., Lin, W.Y., Zheng, K.B., He, L.W. & Huang, W.M. Far-red to near infrared analyte-responsive fluorescent probes based on organic fluorophore platforms for fluorescence imaging. Chemical Society Reviews 42, 622-661 (2013). 39. Guo, Z.Q., Park, S., Yoon, J. & Shin, I. Recent progress in the development of near-infrared fluorescent probes for bioimaging applications. Chemical Society Reviews 43, 16-29 (2014). 40. Juette, M.F., Terry, D.S., Wasserman, M.R., Zhou, Z., Altman, R.B., Zheng, Q. & Blanchard, S.C. The bright future of single-molecule fluorescence imaging. Current Opinion in Chemical Biology 20, 103-111 (2014). 41. Gorka, A.P., Nani, R.R. & Schnermann, M.J. Cyanine polyene reactivity: scope and biomedical applications. Organic & Biomolecular Chemistry 13, 7584-7598 (2015). 42. Hague, A., Faizi, M.S.H., Rather, J.A. & Khan, M.S. Next generation NIR fluorophores for tumor imaging and fluorescence-guided surgery: A review. Bioorganic & Medicinal Chemistry 25, 2017-2034 (2017). 43. Yapici, I., Lee, K.S.S., Berbasova, T., Nosrati, M., Jia, X.F., Vasileiou, C., Wang, W.J., Santos, E.M., Geiger, J.H. & Borhan, B. "Turn-On" Protein Fluorescence: In Situ Formation of Cyanine Dyes. Journal of the American Chemical Society 137, 1073-1080 (2015). 197 44. Hansch, C., Leo, A. & Taft, R.W. A Survey of Hammett Substituent Constants and Resonance and Field Parameters. Chemical Reviews 91, 165-195 (1991). 45. Chen, W., Li, H., Widawsky, J.R., Appayee, C., Venkataraman, L. & Breslow, R. Aromaticity decreases single-molecule junction conductance. Journal of the American Chemical Society 136, 918-920 (2014). 46. Grimm, J.B., English, B.P., Chen, J., Slaughter, J.P., Zhang, Z., Revyakin, A., Patel, R., Macklin, J.J., Normanno, D., Singer, R.H., Lionnet, T. & Lavis, L.D. A general method to improve fluorophores for live-cell and single-molecule microscopy. Nature Methods 12, 244-250 (2015). 47. Liu, X., Qiao, Q., Tian, W., Liu, W., Chen, J., Lang, M.J. & Xu, Z. Aziridinyl Fluorophores Demonstrate Bright Fluorescence and Superior Photostability by Effectively Inhibiting Twisted Intramolecular Charge Transfer. Journal of the American Chemical Society 138, 6960-6963 (2016). 48. Sasaki, S., Niko, Y., Klymchenko, A.S. & Konishi, G.I. Design of donoracceptor geometry for tuning excited-state polarization: fluorescence solvatochromism of push-pull biphenyls with various torsional restrictions on their aryl-aryl bonds. Tetrahedron 70, 7551-7559 (2014). 49. Haberhauer, G., Gleiter, R. & Burkhart, C. Planarized Intramolecular Charge Transfer: A Concept for Fluorophores with both Large Stokes Shifts and High Fluorescence Quantum Yields. Chemistry-A European Journal 22, 971-978 (2016). 50. Zhang, Z.Y., Zhang, G.Q., Wang, J.X., Sun, S.S. & Zhang, Z.Z. The mechanisms of Large Stokes Shift and Fluorescence Quantum Yields in anilino substituted Rhodamine analogue: TICT and PICT. Computational and Theoretical Chemistry 1095, 44-53 (2016). 51. Sasaki, S., Drummen, G.P.C. & Konishi, G. Recent advances in twisted intramolecular charge transfer (TICT) fluorescence and related phenomena in materials chemistry. Journal of Materials Chemistry C 4, 2731-2743 (2016). 52. Berbasova, T., Santos, E.M., Nosrati, M., Vasileiou, C., Geiger, J.H. & Borhan, B. Light-Activated Reversible Imine Isomerization: Towards a Photochromic Protein Switch. ChemBioChem 17, 407-414 (2016). 198 53. Nosrati, M., Berbasova, T., Vasileiou, C., Borhan, B. & Geiger, J.H. A Photoisomerizing Rhodopsin Mimic Observed at Atomic Resolution. Journal of the American Chemical Society 138, 8802-8808 (2016). 54. Hettie, K.S., Klockow, J.L. & Glass, T.E. Three-input logic gates with potential applications for neuronal imaging. Journal of the American Chemical Society 136, 4877-4880 (2014). 55. Debieu, S. & Romieu, A. Dual enzyme-responsive "turn-on" fluorescence sensing systems based on in situ formation of 7-hydroxy-2-iminocoumarin scaffolds. Organic & Biomolecular Chemistry 13, 10348-10361 (2015). 56. Klockow, J.L., Hettie, K.S., Secor, K.E., Barman, D.N. & Glass, T.E. Tunable Molecular Logic Gates Designed for Imaging Released Neurotransmitters. Chemistry-A European Journal 21, 11446-11451 (2015). 57. Romieu, A. "AND" luminescent "reactive" molecular logic gates: a gateway to multi-analyte bioimaging and biosensing. Organic & Biomolecular Chemistry 13, 1294-1306 (2015). 58. Dai, C.G., Liu, X.L., Du, X.J., Zhang, Y. & Song, Q.H. Two-Input Fluorescent Probe for Thiols and Hydrogen Sulfide Chemosensing and Live Cell Imaging. ACS Sensors 1, 888-895 (2016). 59. Grimm, J.B., Gruber, T.D., Ortiz, G., Brown, T.A. & Lavis, L.D. Virginia Orange: A Versatile, Red-Shifted Fluorescein Scaffold for Single- and Dual-Input Fluorogenic Probes. Bioconjugate Chemistry 27, 474-480 (2016). 60. Debieu, S. & Romieu, A. In situ formation of pyronin dyes for fluorescence protease sensing. Organic & Biomolecular Chemistry 15, 2575-2584 (2017). 61. Arora, P.K. & Bae, H. Role of Dehalogenases in Aerobic Bacterial Degradation of Chlorinated Aromatic Compounds. Journal of Chemistry (2014). 62. Graminski, G.F., Zhang, P.H., Sesay, M.A., Ammon, H.L. & Armstrong, R.N. Formation of the 1-(S-Glutathionyl)-2,4,6-Trinitrocyclohexadienate Anion at the Active-Site of Glutathione S-Transferase - Evidence for Enzymic Stabilization of Sigma-Complex Intermediates in Nucleophilic Aromatic-Substitution Reactions. Biochemistry 28, 6252-6258 (1989). 199 63. Crooks, G.P. & Copley, S.D. A Surprising Effect of Leaving Group on the Nucleophilic Aromatic-Substitution Reaction Catalyzed by 4Chlorobenzoyl-Coa Dehalogenase. Journal of the American Chemical Society 115, 6422-6423 (1993). 64. Taylor, K.L., Liu, R.Q., Liang, P.H., Price, J., Dunawaymariano, D., Tonge, P.J., Clarkson, J. & Carey, P.R. Evidence for Electrophilic Catalysis in the 4-Chlorobenzoyl-Coa Dehalogenase Reaction - Uv, Raman, and C-13Nmr Spectral Studies of Dehalogenase Complexes of Benzoyl-Coa Adducts. Biochemistry 34, 13881-13888 (1995). 65. Yang, G.A., Liu, R.Q., Taylor, K.L., Xiang, H., Price, J. & DunawayMariano, D. Identification of active site residues essential to 4chlorobenzoyl-coenzyme A dehalogenase catalysis by chemical modification and site directed mutagenesis. Biochemistry 35, 10879-10885 (1996). 66. Lau, E.Y. & Bruice, T.C. The active site dynamics of 4-chlorobenzoyl-CoA dehalogenase. Proceedings of the National Academy of Sciences of the United States of America 98, 9527-9532 (2001). 67. Luo, L.S., Taylor, K.L., Xiang, H., Wei, Y.S., Zhang, W.H. & DunawayMariano, D. Role of active site binding interactions in 4-chlorobenzoylcoenzyme A dehalogenase catalysis. Biochemistry 40, 15684-15692 (2001). 68. Dong, J., Carey, P.R., Wei, Y.S., Luo, L.S., Lu, X.F., Liu, R.Q. & DunawayMariano, D. Raman evidence for Meisenheimer complex formation in the hydrolysis reactions of 4-fluorobenzoyl- and 4-nitrobenzoyl-coenzyme a catalyzed by 4-chlorobenzoyl-coenzyme A dehalogenase. Biochemistry 41, 7453-7463 (2002). 69. de Jong, R.M. & Dijkstra, B.W. Structure and mechanism of bacterial dehalogenases: different ways to cleave a carbon-halogen bond. Current Opinion in Structural Biology 13, 722-730 (2003). 70. Xu, D.G. & Guo, H. Electrostatic influence of active-site waters on the nucleophilic aromatic substitution catalyzed by 4-chlorobenzoyl-CoA dehalogenase. FEBS Letters 579, 4249-4253 (2005). 71. Wu, R., Cao, J., Lu, X.F., Reger, A.S., Gulick, A.M. & Dunaway-Mariano, D. Mechanism of 4-chlorobenzoate: Coenzyme a ligase catalysis. Biochemistry 47, 8026-8039 (2008). 200 72. Benning, M.M., Taylor, K.L., Liu, R.Q., Yang, G., Xiang, H., Wesenberg, G., DunawayMariano, D. & Holden, H.M. Structure of 4-chlorobenzoyl coenzyme A dehalogenase determined to 1.8 angstrom resolution: An enzyme catalyst generated via adaptive mutation. Biochemistry 35, 81038109 (1996). 73. Clarkson, J., Tonge, P.J., Taylor, K.L., DunawayMariano, D. & Carey, P.R. Raman study of the polarizing forces promoting catalysis in 4chlorobenzoate-CoA dehalogenase. Biochemistry 36, 10192-10199 (1997). 201 CHAPTER IV: ENGINEERING OF hCRBPII/THIOFLUOR INTO A LARGE STOKES SHIFT FLUORESCENT PROTEIN Our efforts have been focused on developing protein fusion tags based on the reengineering of human Cellular Retinol Binding Protein II (hCRBPII) to bind aldehydes, which become fluorescent upon iminium formation. During the course of reengineering hCRBPII to bind the solvatochromic fluorophore ThioFluor, it was observed that not only the protonated Schiff base (PSB), but also the Schiff base (SB) is fluorescent. However, while the PSB emission wavelength can be regulated over a range of 130 nm (613 nm to 744 nm), the emission of the SB remains nearly consistent at approximately 500 nm. Interestingly though, we observed that when specific mutants were irradiated at the Schiff base (SB), they displayed dual fluorescence. These two bands correspond to the expected SB emission and to a large Stokes shift (LSS) species, emitting in the far-red wavelength regime, more reminiscent of PSB emission. An increased Stokes shift, and hence better spectral separation, has numerous advantages, namely to diminish self-absorption and light scattering.1, 2 However, in most cases, high fluorescence efficiency and large Stokes shift are not compatible, particularly in synthetic probes.3-8 Fluorophores with large Stokes shift can be achieved by a variety of mechanisms: intramolecular charge transfer (ICT),9-12 solvatochromism,12-14 excimer/exciplex formation,15-17 or excited state intramolecular proton transfer.18-21 Unfortunately, the precise structure-property 202 relationships are difficult to predict, and in many cases these phenomena lead to fluorescence quenching. IV.1 Preliminary observation of dual fluorescence in hCRBPII/ThioFluor Q108K:K40L:T53A:R58L:Q38F:Q4F/ThioFluor (pH 7.2) irradiated at 379 nm 5 106 Abs (a. u.) 3 106 0.1 2 10 6 1 106 0 639 nm 7 106 6 106 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) 5 106 481 nm 0.1 4 106 3 106 2 106 1 106 0 Fluorescence Intensity (cps) 0.2 Fluorescence Intensity (cps) 4 10 6 393 nm 0.2 Abs (a. u.) 379 nm 474 nm Q108K:K40D:T53A:R58L:Q38F:Q4F/ThioFluor (pH 7.1) irradiated at 393 nm 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) Figure IV-1. Dual fluorescence observed upon irradiation of the SB of Q108K:K40D:T53A:R58L:Q38F:Q4F bound with ThioFluor. Preliminary analysis of sequences comparing the initial hit of the hCRBPII/ThioFluor LSS protein versus those displaying normal SB emission, revealed that an acidic residue had been introduced in close proximity to the iminium. As shown in Figure IV-1, upon irradiation of the SB at 379 nm of the hCRBPII hexamutant Q108K:K40L:T53A:R58L:Q38F:Q4F bound with ThioFluor, an emission maxima of 474 nm is observed. Under identical conditions, irradiation of the SB (at 393 nm) of Q108K:K40D:T53A:R58L:Q38F:Q4F bound with ThioFluor results in fluorescence with emission maxima of 481 nm and 639 nm. The first band corresponds to the expected SB emission, while the second band displays an emission maximum more indicative of PSB emission. The quantum yields of these two mutants are quite different; they are 0.03 and 0.34 203 for Q108K:K40L:T53A:R58L:Q38F:Q4F and Q108K:K40D:T53A:R58L:Q38F:Q4F, respectively. Next, we sought to investigate the spectroscopic properties of the PSB, to deduce whether its emission maxima is in agreement with that obtained from excitation at the SB. Figure IV-2 shows a plot of the emissions obtained upon excitation of the SB and PSB. As observed, the red-shifted band of the emission as a result of irradiating the SB is not normally distributed, possessing a redshifted shoulder. Moreover, from the overlay with normalized emission intensity, it appears that the emission maximum from PSB excitation (649 nm) is equivalent to this red shoulder. The other notable difference is that the quantum yield of SB emission is 0.34, while the quantum yield of the PSB is 0.21. Normalized absorbance 7 10 1 Irr. @ SB Irr. @ PSB 6 6 106 5 106 4 106 3 106 2 106 Irr. @ SB Irr. @ PSB 1 Normalized emission Fluorescence Intensity (cps) 8 106 1 106 0 0 450 500 550 600 650 Wavelength (nm) 700 750 0 440 480 520 560 600 640 680 720 Wavelength (nm) Figure IV-2. Comparison of fluorescence spectra upon irradiation of the SB and PSB of Q108K:K40D:T53A:R58L:Q38F:Q4F/ThioFluor. It should be noted that the emission maxima of both SB and PSB emissions do not change significantly with pH (Figure IV-3). This phenomena will be discuss further in depth later when a larger change in emission maxima is 204 observed upon excitation of the SB and PSB (section IV.2). Additionally, the ratio of the two emission bands observed upon excitation of the SB remains nearly constant at pH values of 7.1, 8.0 and 9.0. Presumably, the excited state pKa is far enough from this pH regime, and thus the protonation state of the imine is unaffected. UV-Vis 7.1 5.2 0.2 0.1 6 106 Fluorescence Intensity (cps) Abs (a. u.) 0.3 0 4 106 3 106 2 106 1 106 0 450 500 550 600 650 700 750 800 Wavelength (nm) UV-Vis 7.1 8.1 9.0 0.2 0.1 1 107 Fluorescence Intensity (cps) Abs (a. u.) 7.1 5.2 5 106 350 400 450 500 550 600 650 700 Wavelength (nm) 0.3 Irradiation @ PSB 0 8 106 Irradiation @ SB 7.1 8.1 9.0 6 106 4 106 2 106 0 350 400 450 500 550 600 650 700 Wavelength (nm) 450 500 550 600 650 700 750 800 Wavelength (nm) Figure IV-3. UV-Vis and fluorescence spectra obtained as a result of changing pH in Q108K:K40D:T53A:R58L:Q38F:Q4F/ThioFluor. During protein expression at room temperature, both monomer and dimer were obtained due to the presence of K40D (see Chapter VI for a full discussion). Dimer was favored during expression, which led to only 35 mol% 205 monomer (8.3 mg/L monomer and 30.8 mg/L dimer). Interestingly, the dimer of the same hexamutant displays essentially the same fluorescent properties. Upon excitation of the SB at 395 nm, dual fluorescence is observed (Figure IV-4), with a quantum yield of 0.35. Similar to the monomer, the red-shifted emission band appears to be composed of two species, of which the major species has an emission maximum of 623 nm. Additionally, excitation of the PSB leads to a more red-shifted emission band at 658 nm with a lower quantum yield (0.17). pH 7.1 Excitation @ 395 nm 1 107 8 106 6 106 0.1 4 106 2 106 0 549 nm 0.3 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) 658 nm 6 106 5 106 4 106 0.2 3 106 2 106 0.1 1 106 0 Fluorescence Intensity (cps) Abs (a. u.) 1.2 10 7 Fluorescence Intensity (cps) 0.2 623 nm Abs (a. u.) 395 nm pH 5.2 Excitation @ 549 nm 0 400 500 600 700 Wavelength (nm) Figure IV-4. UV-Vis (blue) and emission spectra (red) Q108K:K40D:T53A:R58L:Q38F:Q4F-dimer/ThioFluor at pH 7.1 and pH 5.2. of From the aforementioned data, we hypothesized that a proton transfer occurs in the excited state. This would, simply put, produce a PSB in the excited state, which would explain the red-shifted emission band. However, efficiency of the excited state proton transfer (ESPT) dictates that some emission from the SB still occurs. It is further hypothesized that the ESPT emission band is composed of two species, as indicated by the red-shifted shoulder of 623 nm (Figure IV-4). 206 This process is similar to the ESPT observed in wild-type green fluorescent protein (wtGFP), in which the two absorption bands of wtGFP with maxima at 396 and 476 correspond to the neutral and anionic ground-state forms of the chromophore’s tyrosyl group (Figure IV-5).22, 23 Excitation at either absorption leads to green fluorescence. Upon excitation of the neutral chromophore, the pKa is decreased by several units and proton transfer from the hydroxyl group takes place to generate the anionic species that emits green fluorescence.24-27 Excitation of the anionic species directly leads to green fluorescence. Crystal structures show a proton wire connecting the hydroxyl group through a water molecule and Ser205 to the proton acceptor Glu222.25, 28 Thr203 O OH N R O N N O H His148 H O H O O λabs = 396 nm λem = 508 nm N R O H NH O Thr203 R O N Ser65 N R N H H His148 H O H O H O H H O Glu222 Ser205 λabs = 475 nm λem = 503 nm O H Ser65 O Glu222 Ser205 Figure IV-5. Water network responsible for the excited state proton transfer in GFP. It has been shown that excited state proton transfer (ESPT) is also responsible for the large Stokes shift (LSS) in red FPs, including mKeima,29-31 LSSmOrange32-36 and LSSmKates37-42 (Table IV-1). mBeRFP (monomer Blue light-excited RFP) is a monomer that was developed via site-directed mutagenesis from mKate (mKate is majorly monomeric but there is weak dimer 207 formation).43 It displays improved brightness and a faster maturation time than the initially developed mKeima and LSS-mKates. Additionally, even though it requires blue light excitation, it displays higher photostability than its predecessors; it takes 60 minutes to photobleach in E. coli cells. Nonetheless, each of these variants has proven suitable for imaging in cells. Table IV-1. Spectroscopic properties of large Stokes shift fluorescent proteins due to ESPT. ε Maturation t1/2 Protein λabs λem Φ Brightness -1 -1 (M cm ) at 37 °C (h) LSSmOrange 437 572 52,000 0.45 23,400 2.3 mKeima 440 620 13,400 0.24 3,216 4.4 mBeRFP 446 611 65,000 0.45 29,250 1.0 LSSmKate2 460 605 26,000 0.17 4,420 2.5 LSSmKate1 463 624 31,200 0.08 2,496 1.8 In this class of LSS fluorescent proteins, the chromophore environments exhibit high structural similarity; additionally, the interactions between the chromophores and neighboring residues are similar. The active site residues have been optimized such that the chromophore is neutral in the ground state (stabilized by an acidic residue through hydrogen bonding), and the pKa is decreased several units in the excited state, such that the acidic residue now acts as a proton acceptor. Notably, Piatkevich and coworkers showed that several conventional red-shifted FPs, including mNeptune, mCherry, mStrawberry, mOrange and mKO, could be developed in LSS fluorescent proteins by modification of neighboring residues to the chromophore’s hydroxyl group to Asp or Glu.40 208 As shown in Figure IV-1, excitation of the SB of Q108K:K40D:T53A:R58L:Q38F:Q4F/ThioFluor does not lead to complete ESPT. We hypothesized that the presence of two proton acceptors would lead to complete excited state protonation of the SB, leading to only the LSS fluorescent species. To this end, we chose to introduce another acidic residue in the vicinity of the iminium. We added V62E to the same hCRBPII mutant Q108K:K40D:T53A:R58L:Q38F:Q4F. Unfortunately, this mutant is expressed solely in the dimeric form at both room temperature and at 16 °C. UV-Vis and emission spectrum (excited at 400 nm) Abs (a. u.) 5 106 4 106 3 106 0.1 2 106 1 106 0 Fluorescence Intensity (cps) 6 106 0.2 2 107 7 106 Fluorescence Intensity (cps) 0.3 Excitation spectrum at 612 nm 1 107 0 250 300 350 400 450 500 550 600 Wavelength (nm) 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) Figure IV-6. UV-Vis, emission and excitation Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E/ThioFluor. spectra of Fortuitously though, upon excitation of the SB at 400 nm, a major emission peak with maxima at 612 nm is observed (Figure IV-6); only a small blue-shifted shoulder corresponding to normal SB emission is observed. Additionally, a quantum yield of 0.50 is obtained; this is a substantial increase than that observed in Q108K:K40D:T53A:R58L:Q38F:Q4F (0.34 for the monomer and 0.35 for the dimer). Presumably, the ionic interactions between the acidic residues and 209 the iminium leads to chromophore rigidification, leading to an increase in quantum yield. To irrefutably show that the LSS fluorescent species is coming from the SB, an excitation spectrum was collected (Figure IV-6). In this experiment, the emission wavelength is held constant while the excitation wavelength is scanned, indicating which wavelengths produced its fluorescence. This clearly shows that the fluorescence at 612 nm originates from the SB and tryptophans inherent to the protein (to be discussed in Section IV.5). While the dimeric species of hCRBPII is not desirable, these results tell us that in our system, changing the environment around the iminium can alter ESPT. Next we chose to investigate both the introduction of different ionizable residues at position 40 and how the location of a glutamic acid from the iminium affects the ESPT pathway and fluorescent properties. IV.2 Exploration of amino acid substitution at residue 40 on fluorescence The two protein sequences (Q108K:K40L:T53A:R58L:Q38F:Q4F and Q108K:K40D:T53A:R58L:Q38F:Q4F) just discussed only vary by the substitution at residue 40. The position of residue 40 is highlighted in the crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor (Figure IV-7). Presumably, the introduction of aspartic acid is responsible for the ESPT observed (L117 has been mutated to aspartic acid via mutagenesis in Pymol in Figure IV-7). To this end, we introduced other ionizable amino acids at position 40 to see if they also produce the red-shifted emission. We wanted to limit the number of variables that 210 are changed at one time, so we chose to explore modifications retaining Q108K:K40L:T53A:R58L:Q38F:Q4F as the parent hCRBPII mutant. This mutant was chosen as introduction of an acidic residue near the iminium was already shown to lead to ESPT. Additionally, when bound to ThioFluor, it possesses a low pKa (5.1), such that almost no iminium is present at physiological pH. Fortuitously, ThioFluor also displays an acceptable half-life for iminium formation (16.0 minutes; binding of 20 μM protein with 10 μM ThioFluor was measured via fluorescence at 23 °C and plotted to pseudo-first order rate kinetics), and in vitro protein expression yields only monomeric hCRBPII (10.7 mg/L). L117 4.9 Å Figure IV-7. Location of residue 40 in the crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor. L117D (shown in gray) was modeled in Pymol. Numerous amino acids, including cysteine, glutamic acid, histidine, glutamine, arginine, threonine, tyrosine and the wild-type lysine, were introduced at position 40 of Q108K:K40L:T53A:R58L:Q38F:Q4F. Interestingly, the iminium pKa of these mutants coupled with ThioFluor varied over a range of about two pH units (Table IV-2). The most perplexing result is that introduction of the basic 211 residues histidine and lysine at position 40 leads to higher pKas (7.8 and 7.4, respectively) than the introduction of the acidic residues aspartic acid and glutamic acid (6.4 and 5.2, respectively). This is counterintuitive, as it would be expected that a carboxylate would more easily stabilize a positive charge on the nitrogen atom of the iminium, while a basic residue would destabilize this state. Presumably, the imine exists as a different isomer in each of these cases, such that the dissimilar environments around the iminium dictate the pKa. Unfortunately, crystal structures have not been determined for any of these protein/ThioFluor complexes. Fluorescent properties upon irradiation of the SB were analyzed (all data shown is for monomer) at pH 7.0 – 7.2. We defined ΦESPT as the fraction of photons absorbed at the absorption maximum that leads to photochemical protonation of the chromophore’s imine. ΦESPT was derived by integrating the area of the emission spectra of both the deprotonated and protonated forms obtained upon excitation of the SB; ΦESPT was then obtained as a ratio the redshifted emission band area over total fluorescence area. Both cysteine and threonine do not lead to ESPT upon excitation of the SB at physiological pH, as was similarly observed for the introduction of nonpolar leucine at residue 40. However, upon acidification of the solution, a very minor amount of the LSS species can be observed. Basic residues histidine, lysine and arginine all results in partial ESPT. Presumably, they are capable of directly protonating the imine in the excited state, or they polarize a water molecule, 212 which can act as a proton donor in the excited state. However, quantum yields were much lower than that observed with K40D (0.34). Similarly, glutamine and tyrosine showed dual fluorescence upon excitation of the SB, with suppressed quantum yields. Table IV-2. Spectroscopic properties of the SB as a result of substitution at residue 40 in the hCRBPII protein Q108K:K40L:T53A:R58L:Q38F:Q4F. Residue 40 λabs λem (Blue) λem (Red) ΦESPT Φ pKa t1/2 (min)a K40C 383 480 - < 0.02 0.07 6.5 12.8 K40D 393 481 639 0.80 0.34 6.4 13.8 K40E 397 - 605 > 0.98 0.51 5.2 3.6 K40H 393 477 642 0.40 0.09 7.8 15.4 K40K 389 491 661 0.37 0.08 7.4 28.8 K40L 379 474 - < 0.02 0.03 5.1 16.0 K40Q 388 488 663 0.07 0.11 7.5 40.7 K40R 392 492 653 0.69 - - nd K40T 386 487 - < 0.02 0.13 6.0 nd K40Y 393 495 674 0.09 0.04 6.8 12.9 a Half-lives were computed from the rate constant that was obtained from plotting data to pseudo-first order kinetics, to ensure the numbers are comparable (note, they should only be compared in relative terms). Second order rates are not given, as concentration is difficult to determine when both SB and PSB are formed in the course of binding. Mutants with K40E and K40L were measured via fluorescence monitoring SB emission while others were measured via UV-Vis following PSB formation. Fortuitously, the introduction of glutamic acid at position 40 resulted in complete ESPT upon excitation of Q108K:K40E:T53A:R58L:Q38F:Q4F/ ThioFluor at its SB absorption maxima of 397 nm. A single emission band with maximum of 605 nm was obtained (Figure IV-8a), which is more blue-shifted than the other mutants. An excitation spectrum, collected at 605 nm showed that 213 the fluorescence originates from the SB and tryptophans in hCRBPII (Figure IV8b). The emission spectrum obtained upon excitation of tryptophan of the hCRBPII protein at 280 nm suggests the existence of an energy transfer process from tryptophan to the bound ThioFluor-SB (Figure IV-8c). The emission spectrum obtained upon excitation of 280 nm overlays with the emission spectrum obtained upon excitation of the SB at 397 nm, indicating that fluorescence results form the same excited state species. a. 0.2 1 107 0.1 0 5 106 1.5 10 7 d. Ex @ 280 nm Ex @ 397 nm 1 1 Normalized absorbance Ex @ 280 nm UV-Vis 1 107 5 106 0 250 5 106 Normalized emission 7 1 107 0 200 250 300 350 400 450 500 550 600 Wavelength (nm) 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) c. 1.5 10 Fluorescence Intensity (cps) Abs (a. u.) 1.5 10 7 Fluorescence Intensity (cps) 0.3 Fluorescence Intensity (cps) 2 107 b. spectral overlap 350 450 550 650 Wavelength (nm) 750 0 300 0 400 500 600 Wavelength (nm) 700 Figure IV-8. Characterization of SB fluorescent properties of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor. a. UV-Vis and emission spectra after 397 nm excitation. b. Excitation spectrum at 605 nm. c. Emission spectra upon 280 nm excitation. d. Overlay of 280 nm emission spectrum and UV-Vis of the hCRBPII-SB showing spectral overlap. All data was collected at pH 7.2. 214 Figure IV-8d shows the spectral overlap between the normalized protein emission and the ThioFluor-SB absorption spectra. This spectral overlap indicates the possibility of energy transfer from the excited tryptophan(s) in hCRBPII to the bound SB, which then undergoes ESPT. a. 0.4 7.1 6.7 6.2 5.7 5.3 4.9 0.4 4.5 4.0 3.6 ΔAbs at 575 nm (a. u.) 10.1 9.0 7.9 Abs. (a. u.) 0.3 0.2 0.1 Fluorescence Intensity (cps) 1.6 10 7 10.1 9.0 7.9 7.1 6.2 4.9 4.0 1.4 10 7 1.2 10 7 1 10 7 8 106 6 106 4 106 2 10 6 0 pKa = 5.2 R2 = 0.99 0.2 0.1 0 3 Fluorescence Intensity @ 605 nm b. 0 300 350 400 450 500 550 600 650 700 Wavelength (nm) 0.3 4 5 6 7 pH 8 9 10 11 1.6 10 7 1.4 10 7 1.2 10 7 1 107 8 106 working range 6 106 4 106 2 106 0 450 500 550 600 650 Wavelength (nm) 700 750 2 4 6 8 10 pH Figure IV-9. a. pKa titration of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor. b. SB emission as a function of pH. To our delight, a high quantum yield of 0.51 and low pKa of 5.2 (Figure IV9a) were observed. This is desirable since at physiological pH the system will exist mostly as the SB form, ready to undergo ESPT. This should lead to background free fluorescence live cell imaging, as fluorescence will originate 215 from the ESPT emitting state, which is not expected to occur with nonspecific entities. Additionally, upon excitation of the SB, ESPT is observed over a wide range of pH values. Select emission curves obtained upon SB excitation at varying pH values are shown in Figure IV-9b. Typical SB emission only becomes apparent at pH 10.1. Notably, the reaction between Q108K:K40E:T53A:R58L:Q38F:Q4F and ThioFluor is complete within 10 minutes, with a half-life of 1.7 minutes at 23 °C (measured via fluorescence with 20 μM protein with 10 μM ThioFluor and plotted to second order rate kinetics; concentration should be accurate as the pKa is low enough that only SB is formed at pH 7.2) (Figure IV-10). This fast reaction time should allow for almost instantaneous visualization of hCRBPII/ThioFluor after dye addition. Concentration of Complex (M) 1.2 10 -5 1 10-5 8 10-6 k = 59,684 M-1 min-1 t1/2 = 1.7 min R2 = 0.988 6 10-6 4 10-6 2 10-6 0 0 10 20 30 Time (min) 40 50 Figure IV-10. Plot of fluorescence intensity at 605 nm versus time for binding of ThioFluor to Q108K:K40E:T53A:R58L:Q38F:Q4F at 23 °C fit to second order rate kinetics. 216 Fluorescent properties obtained upon excitation of the PSB were also collected (Table IV-3). In most cases, the emission maxima upon excitation of the PSB was within a few nanometers of the emission maxima of the LSS fluorescent species observed upon irradiation of the SB. The introduction of K40E, yielding Q108K:K40E:T53A:R58L:Q38F:Q4F, was the exception (and to a lesser extent K40D, shown in Figure IV-2). Table IV-3. Spectroscopic properties of the PSB as a result of substitution at residue 40 in the hCRBPII protein Q108K:K40L:T53A:R58L:Q38F:Q4F. Residue 40 λabs λem Φ K40C 558 663 0.17 K40D 546 649 0.21 K40E 575 671 0.15 K40H 544 648 0.23 K40K 557 664 0.20 K40L 613 690 0.10 K40Q 558 664 0.17 K40R - 657 - K40T 550 666 0.12 K40Y 595 677 0.16 Figure VI-11a shows an overlay of the normalized emission spectra obtained upon excitation of the SB and PSB (at pH 7.1 and 4.0, respectively; the UV-Vis is shown in Figure IV-9a) of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor. It is quite apparent that the emission maxima are dramatically different – 605 nm and 671 nm for the SB and PSB, respectively. Emission spectra were collected at different pHs. Similarly to that observed for the SB, the PSB emission maxima does not change 217 significantly with pH (Figure IV-11b). This would suggest that hydrogen bonding does not affect the wavelength of emission (as the carboxylate of K40E would be expected to exist at basic pH and the protonated form at lower pH). It would stand to reason that the observed difference between the red-shifted SB emission and PSB emission is not due to charged state interactions between the imine and surrounding residues. It is hypothesized that the difference is due to the presence of different isomers at the imine (cis versus trans). a. b. pH 7.1, Irr. @ 397 nm pH 4.0, Irr. @ 575 nm 3.2 10 6 Normalized emission Fluorescence Intensity (cps) 1 Normalized emission 1 0 2.8 10 6.2 5.7 5.3 4.9 4.5 4.0 6 2.4 10 6 2 106 1.6 10 6 1.2 10 6 8 105 4 105 0 600 0 450 500 550 600 650 700 750 800 Wavelength (nm) 650 700 Wavelength (nm) 750 Figure IV-11. a. Comparison of fluorescence spectra upon irradiation of the SB and PSB of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor. b. PSB emission spectra as a function of pH. Unfortunately, attempts to obtain the crystal structure of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor were not fruitful. However, Mr. Alireza Ghanbarpour was able to obtain the crystal structure of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor (it should be noted that the crystal was frozen after approximately 2 months at room temperature), in which tryptophans are introduced at positions 58 and 19 in an attempt to rigidify 218 the chromophore. Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor is crystallized with four molecules in the asymmetric unit. Three of the chains show good electron density for the chromophore. Interestingly, two distinct orientations of the active site residues Q38F, R58W and F57 are apparent, while K40E conformation and chromophore trajectory remain the same (Figure IV-12). An overlay of Chains A and B is shown in Figure IV-12. It should be noted that both cis and trans isomers of the imine are observed, indicating that both isomers are present in the thermodynamic ground state. Interestingly, in the trans isomer, the imine nitrogen atom is pointing towards K40E. However, in the cis isomer, the nitrogen is pointing toward tryptophan 106, stabilized through a π-cation interaction (these crystals were obtained at pH 4.5). Chain A Chain B Q38F F57 Q38F F57 K40E R58W K40E R58W Chain C F57 Q38F Overlay of Chains A and B Q38F K40E K40E F57 R58W R58W Figure IV-12. Chains A, B and C of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor. Density is shown at 1σ in Pymol. It is interesting that the K40E rotamer is constant, while the surrounding aromatic residues exist in two conformations. The distances from K40E to the 219 imine nitrogen atom are 4.7 Å, 5.2 Å and 5.2 Å for chains A, B and C, respectively. In Chain A of the Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ ThioFluor crystal structure, an extensive hydrogen-bonding network is observed (Figure IV-13); however, it should be noted that the water molecule between K40E and T51 is not seen in the other two chains. Perhaps, this is the reason that K40E does not change orientation even when Q38F, R58W and F57 do adopt different conformations. In this network, a conserved water molecule lies between the side chains of K40E and T51E and the main chain carbonyl of Y60. Side chain interactions of other neighboring residues presumably hold T51 in the orientation shown. V51 2.9 Å E40 K52 2.9 Å T51 2.8 Å 4.7 Å 2.9 Å 3.0 Å 3.1 Å D61 Y60 2.8 Å 2.5 Å D63 V62 Figure IV-13. Hydrogen bonding network centered around K40E and T51, based on chain A of the crystal structure Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor. Some side chains are not shown for clarity. 220 Spectroscopic properties of this hCRBPII/ThioFluor complex are shown in Figure IV-14. Similar to Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor, the pKa of the iminium is low and upon excitation of the SB at 392 nm at pH 7.1, a single emission band at 621 nm is observed, indicating full ESPT. However, the quantum yield is 0.30, which is Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor much (0.51). lower than Additionally, upon excitation of the ground state PSB at 612 nm (pH 4.5), a much more red-shifted emission band is observed, with a maximum of 687 nm. It should be noted that introduction of the tryptophans at positions 58 and 19 individually did not change the quantum yield substantially upon excitation of the SB, both of which showed complete ESPT (Q108K:K40E:T53A:R58W:Q38F:Q4F was 0.44 and Q108K:K40E:T53A:R58L:Q38F:Q4F:Y19W was 0.43). S pH 7.1 Lys108 7 106 6 10 6 5 106 4 106 0.1 3 106 2 106 1 106 0 500 600 700 Wavelength (nm) S H N Lys108 pH 4.5 8 106 λabs = 392 nm λem = 621 nm 400 N 0.2 λabs = 612 nm λem = 687 nm 1 106 0.1 5 105 0 0 800 1.5 10 6 400 500 600 700 Wavelength (nm) Fluorescence Intensity (cps) Abs (a. u.) pKa Fluorescence Intensity (cps) 0.2 N Abs (a. u.) N 0 800 Figure IV-14. UV-Vis and emission spectra of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor SB (pH 7.1) and PSB (pH 4.5). 221 IV.3 Modification of active site residues to glutamic acid As seen in the previous section, modification of residue 40 to glutamic acid provided the most striking results. Thus, we next sought to investigate how location of the proton donor affects ESPT and fluorescent properties of the SB. To this end we chose to modify other residues along the entire chromophores’ length. Glutamic acid was utilized for this purpose; however, it should be noted that aspartic acid might have produced better results at different positions, as the length of the side chain ostensibly affects the proton wire that is responsible for the ESPT pathway. W8 W8 Q38 K40L I42 F130 T53S F130 L115 Q38 Q4 K40L L115 Q128 T51V I42 Q128 L117 Q4 V62 T51V F49 T53S L117 F49 L93 W106 L119 F64 F64 L119 V62 C95 V86 V86 W106 L93 C95 Figure IV-15. Residues mutated to glutamic acid to test whether its location allows ESPT to occur; two views are shown. Based on the crystal structure of Q108K:K40L:T51V:T53S:R58W/ThioFluor. Mutation to glutamic acid of residues shown in grey did not lead to soluble protein expression. We introduced glutamic acid in the same parent hCRBPII hexamutant (Q108K:K40L:T53A:R58L:Q38F:Q4F) at the following 19 residues: Q4, W8, Q38, K40, I42, F49, T51, T53, V62, F64, V86, L93, C95, W106, L115, L117, L119, Q128 and F130 (Figure IV-15). Residues as far as 10 Å from the imine were 222 tested, as it was predicted that ESPT could occur as a result of hydrogen bonding water networks. Unfortunately, many of the mutants led to insoluble protein expression, as evidenced by the presence of hCRBPII in the pellet of lysed cells via gel electrophoresis. Soluble protein was obtained after substitution of glutamic acid at nine positions, including at residues K40, T51, V62, I42, L117, T53, L119, Q38 and Q128. Interestingly, mutation of the conserved isoleucine at residue 42 to glutamic acid led to soluble protein expression, while previous attempts to insert an I42E mutation in other templates was unsuccessful. Presumably it is the presence of Q4F that allows for soluble protein expression, although the reason is unknown at this time. It should also be noted that the addition of F49E led to soluble protein expression, but only one time, and is thus not included because the results could not be reproduced. Table IV-4. Protein expression yields of hCRBPII mutants with glutamic acid introduced at varying locations in Q108K:K40L:T53A:R58L:Q38F:Q4F. Monomer Dimer Mol% hCRBPII Mutant (mg/L) (mg/L) monomer Q108K:K40L:T53A:R58L:Q4F:Q38E 16.3 - > 95 Q108K:K40E:T53A:R58L:Q38F:Q4F 22.6 - > 95 Q108K:K40L:T53A:R58L:Q38F:Q4F:I42E 26.9 - > 95 Q108K:K40L:T53A:R58L:Q38F:Q4F:T51E 35.4 2.0 > 95 Q108K:K40L:T53E:R58L:Q38F:Q4F 22.7 - > 95 Q108K:K40L:T53A:R58L:Q38F:Q4F:V62E 16.0 - > 95 Q108K:K40L:T53A:R58L:Q38F:Q4F:L117E 32.6 1.5 98 Q108K:K40L:T53A:R58L:Q38F:Q4F:L119E 17.2 - > 95 Q108K:K40L:T53A:R58L:Q38F:Q4F:Q128E 19.6 - > 95 223 Fortuitously, only monomeric protein was isolated during protein expression except for those mutants when T51E and L117E were introduced (Table IV-4). The spectroscopic properties upon excitation of the SB were explored. Results, including absorption and emission maxima, quantum yield and select half-lives are detailed in Table IV-4. Table IV-5. Spectroscopic properties of the SB as a result of glutamic acid introduction in the hCRBPII protein Q108K:K40L:T53A:R58L:Q38F:Q4F. t1/2 hCRBPII Mutant λabs λem Φ ΦESPT (min)a Q108K:K40L:T53A:R58L:Q4F:Q38E 389 485 0.05 < 0.02 nd Q108K:K40E:T53A:R58L:Q38F:Q4F 397 605 0.51 > 0.98 3.6 Q108K:K40L:T53A:R58L:Q38F:Q4F:I42E 406 603 0.41 0.95 11.5 Q108K:K40L:T53A:R58L:Q38F:Q4F:T51E 399 602 0.12 0.94 6.4 Q108K:K40L:T53E:R58L:Q38F:Q4F 387 0.04 < 0.02 nd Q108K:K40L:T53A:R58L:Q38F:Q4F:V62E 386 0.08 0.27 nd Q108K:K40L:T53A:R58L:Q38F:Q4F:L117E 391 - 0.74 nd Q108K:K40L:T53A:R58L:Q38F:Q4F:L119E 384 480 474, 598 486, 653 479 0.05 0.03 nd Q108K:K40L:T53A:R58L:Q38F:Q4F:Q128E 385 473 0.05 < 0.02 nd a Half-lives were computed from the rate constant that was obtained from plotting data to pseudo-first order kinetics, to ensure the numbers are comparable (note, they should only be compared in relative terms). Second order rates are not given, as concentration is difficult to determine when both SB and PSB are formed in course of binding. Rates were measured via fluorescence monitoring SB emission. Fluorescent properties obtained upon excitation of the SB were obtained (Table IV-5). Again, we defined ΦESPT as the fraction of photons absorbed at the absorption maximum that leads to photochemical protonation of the chromophore’s imine. ΦESPT was derived by integrating the area of the emission spectra of both the deprotonated and protonated forms obtained upon excitation 224 of the SB; ΦESPT was then obtained as a ratio the red-shifted emission band area over total fluorescence area. High ΦESPT values were obtained for only K40E, I42E, and T51E mutations. Mutations V62E and L117E significantly reduced ΦESPT, while all other mutations tested led to typical SB emission as the major event. Among those mutations exhibiting a majority of ESPT fluorescence upon excitation of the SB, high quantum yields were observed for K40E and I42E, while a pronounced reduction is observed for T51E. As seen in Figure IV-16, all three residues are in close proximity to the imine. However, crystal structures are not available for these hCRBPII mutant/ThioFluor complexes. T51 K40E I42 Figure IV-16. Crystal structure of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor (Chain A), showing the location of residues 40, 42, and 51. The striking difference in quantum yields is quite interesting. It was first suspected that the hCRBPII/ThioFluor complexes have different excited state lifetimes. With the aid of Dr. Christopher Mancuso (Dantus lab), we were able to 225 obtain lifetimes for the red-shifted fluorescent species that were observed upon SB excitation at pH 7 (1060 nm laser corresponds to three-photon excitation). As seen in Table IV-6, there is no significant difference between the lifetimes associated with these hCRBPII mutants; it is concluded that the differences in quantum yield are not due to lifetime. Further crystallographic evidence may be useful to elucidate the decrease in quantum yield in the case of T51E, as compared to either K40E or I42E. It is possible that T51E adopts a rotamer, in which the side chain is not directly interacting with the imine. Table IV-6. Fluorescence lifetimes of hCRBPII/ThioFluor complexes upon 1060 nm excitation (three photon excitation). hCRBPII Mutant Φ Lifetime (ns) Q108K:K40E:T53A:R58L:Q38F:Q4F 0.51 1.6 Q108K:K40L:T53A:R58L:Q38F:Q4F:I42E 0.41 1.9 Q108K:K40L:T53A:R58L:Q38F:Q4F:T51E 0.12 1.7 In preliminary studies, it was observed that when we added V62E to the hCRBPII mutant Q108K:K40D:T53A:R58L:Q38F:Q4F (ΦESPT = 0.80), ESPT was the major pathway for fluorescence. A major emission peak with maxima at 612 nm was observed (Figure IV-6); only a small blue-shifted shoulder corresponding to normal SB emission is apparent. Additionally, a quantum yield of 0.50 is obtained; this is a substantial increase than that observed in Q108K:K40D:T53A:R58L:Q38F:Q4F (0.34 for the monomer and 0.35 for the dimer). Based on these results, we sought to investigate whether the introduction of a second acidic residue to the optimized Q108K:K40E:T53A:R58L:Q38F:Q4F could increase quantum yield. 226 To this end, we attempted to introduced a second glutamic acid at the following sites: L115, F130, V62, T51, W8 and 42E. Unfortunately, no soluble protein was obtained during protein expression for L115E, F130E, W8E and I42E. hCRBPII was found to be in the pellet of lysed cells (as evidenced by protein gel electrophoresis). Two hCRBPII Q108K:K40E:T53A:R58L:Q38F:Q4F:V62E Q108K:K40E:T53A:R58L:Q38F:Q4F:T51E mutants, and provided soluble protein during expression. Unfortunately, only dimer was isolated from the expression of Q108K:K40E:T53A:R58L:Q38F:Q4F:T51E. On the other hand, Q108K:K40E:T53A:R58L:Q38F:Q4F:V62E led to both monomer and dimer (6.6 mg/L monomer and 13.8 mg/L dimer). Upon excitation of the SB with ThioFluor, for both monomer and dimer, single emission bands are observed at 605 nm and 608 nm, respectively. While ESPT properties did not change, quantum yield did decrease for the monomer (0.34) from the parent hCRBPII mutant Q108K:K40E:T53A:R58L:Q38F:Q4F (0.51). However, an increase in the quantum yield of the dimer upon SB was observed (0.67). IV.4 Effect of point mutations in Q108K:K40E:T53A:R58L:Q38F:Q4F on ESPT fluorescent properties Presumably, substitutions of residues in the putative proton wire, as well as point mutations in the immediate chromophore environment can change the ESPT pathway. Thus, we sought to investigate how the single point mutations in Q108K:K40E:T53A:R58L:Q38F:Q4F affect the fluorescent properties upon 227 excitation of the SB. To this end, we removed each mutation independently from the parent mutant. Unfortunate, it was observed that hCRBPII oligomerization is significantly affected by single point mutations (Table IV-7). Additionally, soluble protein is not obtained during expression in the absence of Q4F. Table IV-7. Protein expression yields as a result of point mutations in Q108K:K40E:T53A:R58L:Q38F:Q4F. Monomer Dimer Mol% hCRBPII Mutant (mg/L) (mg/L) monomer Q108K:K40E:T53A:R58L:Q38F:Q4F 22.6 - > 95 Q108K:T53A:R58L:Q38F:Q4F 16.6 2.5 93 Q108K:K40E:R58L:Q38F:Q4F 66.7 4.3 97 Q108K:K40E:T53A:Q38F:Q4F 3.7 4.0 65 Q108K:K40E:T53A:R58L:Q4F 5.6 36.3 24 Interestingly, the single point mutations did not affect the degree of ESPT, as all showed a single emission band centered at approximately 600 nm (Table IV-8). This was the case for all monomers and dimers. Unfortunately, in the hCRBPII monomers, quantum yield of the SB was markedly decreased, almost in half for each mutant (Table IV-8). Rates for iminium formation, measured via fluorescence at 23 °C (plotted to pseudo-first order rate kinetics using 20 μM protein with 10 μM ThioFluor) did not substantially change between any of the mutants. It was previously observed that the emission maxima observed upon excitation of the SB and PSB were drastically different in Q108K:K40E:T53A:R58L:Q38F:Q4F. We wanted to investigate whether this is a one time phenomena, or if it is true of mutants that contain K40E. To this end, we 228 explored the fluorescent properties of the PSB in the monomeric hCRBPII mutants just discussed. In comparing the SB emission maxima (Table IV-8) to the PSB emission maxima (Table IV-9), it is quite apparent that the emission obtained upon PSB excitation is redder than that obtained upon excitation of the SB, in all cases. Table IV-8. Spectroscopic properties of the SB as a result of point mutations in Q108K:K40E:T53A:R58L:Q38F:Q4F. t1/2 hCRBPII Mutant λabs λem Φ (min) Q108K:K40E:T53A:R58L:Q38F:Q4F 397 605 0.51 3.6 Q108K:K40E:R58L:Q38F:Q4F (monomer) 396 612 0.23 5.1 Q108K:K40E:R58L:Q38F:Q4F (dimer) 397 611 0.28 3.4 Q108K:K40E:T53A:Q38F:Q4F (monomer) 396 607 0.25 3.5 Q108K:K40E:T53A:Q38F:Q4F (dimer) 394 613 0.40 5.1 Q108K:K40E:T53A:R58L:Q4F (monomer) 397 619 0.23 2.3 Q108K:K40E:T53A:R58L:Q4F (dimer) 409 628 0.39 9.0 Table IV-9. Spectroscopic properties of the PSB as a result of point mutations in Q108K:K40E:T53A:R58L:Q38F:Q4F. hCRBPII Mutant λabs λem Q108K:K40E:T53A:R58L:Q38F:Q4F 575 671 Q108K:K40E:R58L:Q38F:Q4F (monomer) 554 651 Q108K:K40E:R58L:Q38F:Q4F (dimer) 557 660 Q108K:K40E:T53A:Q38F:Q4F (monomer) 568 664 Q108K:K40E:T53A:Q38F:Q4F (dimer) 566 674 Q108K:K40E:T53A:R58L:Q4F (monomer) 566 670 Q108K:K40E:T53A:R58L:Q4F (dimer) 583 670 We wanted to see if the same mutations also produced different amounts of monomer in the similar hCRBPII mutant Q108K:K40D:T53A:R58L:Q38F:Q4F. 229 Again, residues were changed individually back to wild type, and proteins were expressed at room temperature. As shown in Table IV-10, similar results were obtained; removal of R58L or Q38F results in substantially less monomer formation. In the case of R58L removal, no monomer was isolated during protein expression. Beneficially though, in this case, the removal of T53A led to a significant increase in the amount of monomer produced. Due to the large amount of dimer formation in these variants, they were not pursued further for development into a LSS fluorescent probe. Table IV-10. Protein expression yields as a result of point mutations in Q108K:K40D:T53A:R58L:Q38F:Q4F. Monomer Dimer Mol% hCRBPII Mutant (mg/L) (mg/L) monomer Q108K:K40D:T53A:R58L:Q38F:Q4F 8.3 30.8 35 Q108K:K40D:R58L:Q38F:Q4F 22.5 3.0 94 Q108K:K40D:T53A:Q38F:Q4F - 22.5 <5 Q108K:K40D:T53A:R58L:Q4F 3.2 45.7 12 IV.5 Energy transfer from tryptophan fluorescence to the bound ligand As briefly discussed in Section IV.2, the emission spectra obtained upon excitation of tryptophan 280 nm suggests the existence of an energy transfer process from tryptophan to the bound ThioFluor-SB. Figure IV-8 shows the spectral overlap between the protein emission and the ThioFluor-SB absorption spectra of Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor. This spectral overlap indicates the possibility of energy transfer from the excited tryptophans to the bound ligand. This leads to over a 300 nm pseudo-Stokes shift (280 nm excitation with ~ 600 nm emission). 230 The energy transfer process between hCRBPII’s tryptophans and bound ThioFluor has been preliminarily studied. Wild-type hCRBPII contains four tryptophan residues (Figure IV-17). In an attempt to understand which tryptophan is responsible for the energy transfer, each were independently mutated to phenylalanine in the parent template Q108K:K40E:R58L:Q38F:Q4F. This protein template was used because protein expression is higher without the T53A mutation (Table IV-7, 66.7 mg/L monomer), yet still shows complete ESPT upon excitation of the SB. W8 W109 W106 W88 Figure IV-17. Four tryptophan residues are present in wild type hCRBPII. Based on the crystal structure of Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W/ThioFluor. To this end, Q108K:K40E:R58L:Q38F:Q4F:W88F, Q108K:K40E:R58L:Q38F:Q4F:W8F, Q108K:K40E:R58L:Q38F:Q4F:W106F, and Q108K:K40E:R58L:Q38F:Q4F:W109F were expressed and bound with 231 ThioFluor. All complexes showed essentially complete ESPT upon excitation of the SB (Table IV-11). Table IV-11. Spectroscopic properties of the SB with tryptophan to phenylalanine mutations in Q108K:K40E:R58L:Q38F:Q4F. Monomer hCRBPII Mutant Expression λabs λem Φ (mg/L) Q108K:K40E:R58L:Q38F:Q4F:W8F 27.0 398 616 0.38 Q108K:K40E:R58L:Q38F:Q4F:W88F 24.2 399 615 0.30 Q108K:K40E:R58L:Q38F:Q4F:W106F 8.9 394 615 0.33 Q108K:K40E:R58L:Q38F:Q4F:W109F 17.0 399 614 0.37 The spectroscopic properties of the parent mutant Q108K:K40E:R58L:Q38F:Q4F coupled with ThioFluor are shown in Figure IV18. From the overlay of the emission spectrum (blue curve) upon excitation at 280 nm and the absorption spectrum of the hCRBPII/ThioFluor complex (green curve), it is obvious that that there is spectral overlap between tryptophan emission (λmax = 355 nm) and SB absorbance (λmax = 396 nm). This suggests that tryptophan should be able to transfer energy to the SB. As shown in Figure IV-18, not only does excitation at 280 nm lead to tryptophan emission, but it also leads to a red-shifted emission band (λmax = 612 nm; blue curve), similar to that observed upon irradiation of the SB at 396 nm (red curve). This indicates that excitation of tryptophan residues at 280 nm leads to energy transfer to the ThioFluor-SB, which subsequently undergoes a proton transfer in the excited state. The final result is a pseudo Stokes shift of 332 nm. 232 1 280 nm 396 nm UV-Vis Normalized emission Q108K:K40E:R58L:Q38F:Q4F spectral overlap 0.8 0.6 0.4 0.2 0 300 Q108K:K40E:R58L:Q38F:Q4F:W8F 400 500 600 Wavelength (nm) Q108K:K40E:R58L:Q38F:Q4F:W88F 1 1 280 nm 399 nm 0.8 Normalized emission Normalized emission 280 nm 398 nm 0.6 0.4 0.2 0 0.8 0.6 0.4 0.2 0 300 400 500 600 Wavelength (nm) 700 Q108K:K40E:R58L:Q38F:Q4F:W106F 300 400 500 600 Wavelength (nm) 700 Q108K:K40E:R58L:Q38F:Q4F:W109F 1 1 280 nm 394 nm 280 nm 399 nm 0.8 Normalized emission Normalized emission 700 0.6 0.4 0.2 0 0.8 0.6 0.4 0.2 0 300 400 500 600 Wavelength (nm) 700 300 400 500 600 Wavelength (nm) 700 Figure IV-18. Emission spectra of Q108K:K40E:R58L:Q38F:Q4F subsequent mutation of the four tryptophan residues in hCRBPII. and Similar results were obtained after each tryptophan was individually mutated to phenylalanine. Overlays of the emission spectra upon excitation at 233 280 nm (blue curve) and the SB at approximately 390 nm (red curve) of the hCRBPII/ThioFluor complexes are shown in Figure IV-18. In all cases the spectra are dominated by tryptophan emission at approximately 350 nm. However, the ratio of emissions from tryptophan (~350 nm) and the ESPT pathway (~615 nm) do vary. The most obvious difference is seen in Q108K:K40E:R58L:Q38F:Q4F:W109F. When the emission spectrum was obtained at 280 nm excitation, the quenched fluorescence emission of the tryptophan residue around 350 nm is more, as compared to the other mutants. It is interesting to note that W109 is located on the protein surface, directed toward the solvent exterior. The emission spectra of the mutants containing W8F, W88F and W106F are comparable to that of the parent hCRBPII. Because the removal of W109 increases energy transfer, while the others do not change the fluorescent properties, it can be inferred that W109 impedes the energy transfer process to the SB. However, no other conclusions can be made at this time, except each tryptophan allows for at least partial energy transfer of their fluorescence to the bound SB. subsequent ESPT leads to the red-shifted emission band at approximately 600 nm. When each tryptophan was removed individually, all of the mutants still showed partial energy transfer and complete ESPT upon 280 nm excitation. It can be concluded that more than one tryptophan residue is responsible for the energy transfer to the bound SB. We attempted to express and purify double 234 tryptophan to phenylalanine mutations, but no hCRBPII protein was isolated in any case. IV.6 Live cell imaging of large Stokes shift hCRBPII/ThioFluor complexes The mutant Q108K:K40E:T53A:R58L:Q38F:Q4F was chosen for live-cell imaging with ThioFluor. It displays ideal spectroscopic properties including fast iminium formation with a half-life of 1.7 min (measured via fluorescence with 20 μM protein with 10 μM ThioFluor and plotted to second order rate kinetics; concentration should be accurate as the pKa is low enough that only SB is formed at pH 7.2), low pKa of 5.3 (rendering almost complete SB formation at physiological pH), high quantum yield (0.51) and large Stokes shift (208 nm). Additionally, in vitro protein expression produces solely the monomer form of hCRBPII. We cloned the hCRBPII hexamutant Q108K:K40E:T53A:R58L:Q38F:Q4F into plasmids containing EGFP, to serve as a control of transfection. The same backbone vector pFlag-CMV2 was used, as we have previously demonstrated that this plasmid efficiently expresses our protein in living HeLa, U2O2 and COS7 cell lines. Cells were Q108K:K40E:T53A:R58L:Q38F:Q4F transfected and with EGFPEGFP- Q108K:K40E:T53A:R58L:Q38F:Q4F-3NLS (nuclear localization) and labeled with 10 μM ThioFluor for one minute at 37 °C. Cells were subsequently washed twice and imaged immediately (Figure IV-19). 235 ThioFluor/hCRBPII ex: 405 nm em: LP 615 nm ThioFluor/hCRBPII + DIC EGFP-hCRBPII EGFP-hCRBPII-3NLS EGFP ex: 488 nm em: BP 505 - 530 nm Figure IV-19. Live cell imaging of HeLa cells expressing EGFPQ108K:K40E:T53A:R58L:Q38F:Q4F and EGFPQ108K:K40E:T53A:R58L:Q38F:Q4F-3NLS, labeled with 10 μM ThioFluor for one minute at 37 °C. NLS = nuclear localization sequence. As seen in Figure IV-19, ThioFluor was sufficient in labeling expressed hCRBPII. Excitation at 488 nm led to green fluorescence from EGFP, indicating that transfection was efficient. Fortuitously, excitation at 405 nm of the hCRBPII/ThioFluor complex resulted in red fluorescence, which was collected with a LP 615 nm emission filter. While the system shows utility, the disadvantage of this system should be noted. For images shown in Figure IV-19, background fluorescence is not visible, presumably due to the short incubation time (one minute) required for ThioFluor binding to Q108K:K40E:T53A:R58L:Q38F:Q4F. Longer incubation 236 times and higher ThioFluor concentrations lead to background fluorescence under the same imaging conditions. It is hypothesized that the background fluorescence is due to unbound ThioFluor, due to its solvatochromic properties; ThioFluor absorbs at approximately 430 nm in solution and emits in the red wavelength regime (Figure II-13), which would disadvantageously show fluorescence under the imaging conditions for ESPT. Images of non-transfected cells, incubated with ThioFluor, are shown in Figure IV-20. 20 µM ThioFluor 5 min @ 37 °C 10 µM ThioFluor 5 min @ 37 °C 5 µM ThioFluor 5 min @ 37 °C ThioFluor/hCRBPII + DIC ThioFluor/hCRBPII ex: 405 nm em: LP 615 nm 20 µM ThioFluor 2 min @ 37 °C Figure IV-20. Non-transfected HeLa cells incubated with ThioFluor show background fluorescence when excited at 405 nm (LP 615 nm emission filter). It is suggested that a fluorophore be used that does not display solvatochromic properties, because the far red emission observed from excitation at 405 nm is presumably due to the free chromophore. Spectroscopic properties of ThioFluor bound to n-butyl amine are shown in Figure IV-21. In ethanol, ThioFluor absorbs at 430 nm and emits at 654 nm. This large Stokes shift would allow for any unbound chromophore to be visible in the same LP 615 nm 237 emission channel, at which hCRBPII/ThioFluor fluorescence as a result of ESPT. When bound to n-butyl amine, the absorbance blue-shifts to 400 nm, and emits with a maximum of 540 nm. Only a small amount of ESPT occurs form the SB in ethanol. However, even this fluorescence over 615 nm may lead to background fluorescence upon excitation of non-specifically formed imine. ThioFluor ThioFluor-SB ThioFluor-PSB 400 nm Abs. (a. u.) 0.3 430 nm 521 nm 0.2 0.1 Emission ThioFluor ThioFluor-SB ThioFluor-PSB 2.5 10 6 Fluorescence Intensity (cps) UV-Vis 0.4 540 nm 2 106 654 nm 1.5 10 0 6 1 106 5 105 689 nm 0 400 300 350 400 450 500 550 600 650 Wavelength (nm) 500 600 700 800 Wavelength (nm) 900 Figure IV-21. UV-Vis and emission spectra of ThioFluor, ThioFluor-SB and ThioFluor-PSB when bound to n-butyl amine. To this end, it is suggested that a more hydrophilic fluorophore be synthesized and tested, as it is expected to stay in the cell’s aqueous environment and remain nonfluorescent (recall ThioFluor is not fluorescent due to aggregation in PBS buffer). The addition of a sulfonic acid group to ThioFluor should fulfill this need, assuming other spectroscopic properties are not detrimentally altered. IV.7 Excited state proton transfer is observed in ThioFluor analogues Gratifyingly, the phenomenon of ESPT was also observed in ThioFluor analogues. ThioFluor-3, ThioFluor-4, ThioFluor-7, ThioFluor-8, ThioFluor-9, 238 and ThioFluor-10 were analyzed with the hCRBPII mutant (Q108K:K40E:T53A:R58L:Q38F:Q4F) that was optimized for complete ESPT at physiological pH with ThioFluor. N N S pKa N Lys108 pH 7.1 Excitation @ 395 nm 395 nm 611 nm 2 106 Abs (a. u.) 1 106 0 0.2 8 105 6 105 0.1 4 105 2 105 0 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) 1 106 Fluorescence Intensity (cps) 4 10 6 674 nm 1.2 10 6 Fluorescence Intensity (cps) 5 10 6 3 106 0.1 554 nm 7 106 6 106 0.2 Lys108 pH 4.7 Excitation @ 554 nm Abs (a. u.) 0.3 S H N 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) Figure IV-22. UV-Vis and emission spectra Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-3 at pH 7.1 and pH 4.7. of Spectroscopic properties of Q108K:K40E:T53A:R58L:Q38F:Q4F bound with ThioFluor-3 are shown in Figure IV-22. At neutral pH, SB is present, which upon excitation yields a red-shifted emission band at 611 nm. This is indicative of complete ESPT, as was observed for ThioFluor. The quantum yield is 0.58, a small increase over that observed for ThioFluor (0.51). Upon acidification and formation of PSB in the ground state, an absorption maximum of 554 nm is observed. Excitation of this peak results in an emission maximum of 674 nm. Spectroscopic properties of Q108K:K40E:T53A:R58L:Q38F:Q4F bound with ThioFluor-4 are shown in Figure IV-23. At neutral pH, only SB is present, which upon excitation yields emission with a maximum of band of 479 nm, with a 239 quantum yield of 0.02. Upon acidification and formation of PSB in the ground state, an absorption maximum of 470 nm is observed. Excitation of this peak results in an emission maximum of 550 nm. From this data it is obvious that substitution of the N,N-dimethyl donating group in ThioFluor for the less donating methoxy group in ThioFluor-4 leads to no ESPT upon SB excitation. O S N O Lys108 pH 7.0 Excitation @ 369 nm 369 nm 479 nm 6 105 Abs (a. u.) 2 105 1 105 0 1 105 470 nm 8 104 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) 6 104 0.1 4 104 2 104 0 Fluorescence Intensity (cps) 4 10 5 3 105 0.1 Lys108 550 nm 0.2 Fluorescence Intensity (cps) 5 105 0.2 S pH 4.5 Excitation @ 470 nm Abs (a. u.) 0.3 H N pKa 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) Figure IV-23. UV-Vis and emission spectra Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-4 at pH 7.0 and pH 4.5. of Spectroscopic properties of Q108K:K40E:T53A:R58L:Q38F:Q4F bound with ThioFluor-7 are shown in Figure IV-24. At neutral pH, only SB is present, which upon excitation yields emission at 516 nm, with a quantum yield of 0.11. Upon acidification no PSB was observed, even at pH 2.8. As the emission of the PSB cannot be determined, no definite conclusion can be drawn. However, an emission maximum of 516 nm is more indicative of SB emission (ThioFluor-7-SB 240 with n-butyl amine, Table III-3). It would stand to reason that the pKa of the imine is too low and thus is incapable of abstracting a proton in the excited state. Lys108 N 387 nm 516 nm 7 106 6 106 CF3 pH 7.1 Excitation @ 387 nm Abs (a. u.) S 5 106 0.1 4 106 3 106 0.05 2 106 1 106 0 Fluorescence Intensity (cps) N 0.15 0 350 400 450 500 550 600 650 700 750 Wavelength (nm) Figure IV-24. UV-Vis and emission Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-7 at pH 7.1. spectra of Spectroscopic properties of Q108K:K40E:T53A:R58L:Q38F:Q4F bound with ThioFluor-8 are shown in Figure IV-25. At neutral pH, SB is present, which upon excitation yields a red-shifted emission band at 642 nm. This is indicative of complete ESPT, as was observed for ThioFluor. The quantum yield is 0.45, a small decrease to that observed for ThioFluor (0.51), but about 30 nm more redshifted in emission. Upon acidification and formation of the PSB in the ground state, an absorption maximum of 650 nm is observed. Excitation of this peak results in an emission maximum of 701 nm, which is 50 nm more red-shifted than the emission maximum obtained upon excitation of the SB. The iminium pKa was determined to be 5.5, which is comparable to the pKa determined for ThioFluor with the same hCRBPII mutant (5.3). 241 N pKa S N S N S Lys108 UV-Vis and Emission at pH 7.1 Excitation @ 426 nm 426 nm 650 nm 642 nm 3 10 6 Abs (a. u.) 1.5 10 6 0.05 1 10 6 5 105 0 400 500 600 700 Wavelength (nm) 4 105 0.2 3 105 0.15 2 105 0.1 1 105 0.05 0 800 0 400 pKa curve 7.1 7.8 8.3 9.0 10.0 0.3 ΔAbs at 650 nm (a. u.) Abs. (a. u.) 0.3 5.7 6.0 6.6 0.2 0.1 0 300 0 800 500 600 700 Wavelength (nm) pH Titration 4.3 4.7 5.3 Fluorescence Intensity (cps) 2 106 701 nm 0.25 Fluorescence Intensity (cps) 2.5 10 6 0.1 Lys108 UV-Vis and Emission at pH 4.3 Excitation @ 650 nm Abs (a. u.) 0.15 H N S 0.2 pKa = 5.5 R2 = 0.99 0.1 0 400 500 600 700 Wavelength (nm) 4 800 5 6 7 8 9 10 11 pH Figure IV-25. UV-Vis, emission spectra Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-8. and pH titration of Spectroscopic properties of Q108K:K40E:T53A:R58L:Q38F:Q4F bound with ThioFluor-9 are shown in Figure IV-26. At neutral pH, SB is present, which upon excitation yields dual fluorescence with emission maxima of 470 nm and 635 nm. This is indicative of only partial ESPT, as some of the typical SB emission occurs at 470 nm. The quantum yield is 0.10, a substantial decrease than that observed for ThioFluor (0.51). Upon acidification and formation of the 242 PSB in the ground state, an absorption maximum of 550 nm is observed. Excitation of this peak results in an emission maximum of 677 nm, which is approximately 40 nm more red-shifted than the red-shifted emission maximum obtained upon excitation of the SB. The iminium pKa was determined to be 4.6, which is less than one unit different from the pKa determined for ThioFluor (5.3) with the same hCRBPII mutant. N N Lys108 pKa S S N UV-Vis and Emission at pH 7.2 Excitation @ 385 nm 385 nm Abs (a. u.) 635 nm 0.1 1 106 5 105 0.05 0 400 3.8 4.1 4.4 4.8 500 600 700 Wavelength (nm) pH Titration 5.4 6.9 5.8 7.2 6.1 7.5 6.6 8.2 0.1 3 104 2 104 0.05 1 104 0 400 9.0 10.0 10.5 0.1 0 300 4 104 0 800 0.2 5 104 0.15 0 800 500 600 700 Wavelength (nm) pKa curve 0.2 ΔAbs at 550 nm (a. u.) Abs. (a. u.) 0.3 1.5 10 6 677 nm Fluorescence Intensity (cps) 2 10 6 Fluorescence Intensity (cps) 0.2 0.15 550 nm 2.5 10 6 470 nm Lys108 UV-Vis and Emission at pH 3.8 Excitation @ 550 nm Abs (a. u.) 0.25 N H 0.15 pKa = 4.6 R2 = 0.99 0.1 0.05 0 400 500 600 700 Wavelength (nm) 2 800 4 6 8 10 pH Figure IV-26. UV-Vis, emission spectra Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-9. 243 and pH titration of N Lys108 N H pKa N N UV-Vis and Emission at pH 7.2 Excitation @ 345 nm 1 106 0.2 8 105 0.15 6 105 0.1 4 105 0.05 2 105 400 Fluorescence Intensity (cps) 1.2 10 6 2.5 10 4 0.08 2 104 0.06 1.5 10 4 0.04 1 104 0.02 5000 0 300 400 5.3 5.6 6.1 6.8 7.2 0.2 0.1 0 300 0 800 0.15 ΔAbs at 477 nm (a. u.) Abs. (a. u.) 0.3 4.6 4.9 500 600 700 Wavelength (nm) pKa curve pH Titration 3.5 4.4 3.5 10 4 3 104 0.1 0 800 500 600 700 Wavelength (nm) 666 nm Fluorescence Intensity (cps) 451 nm 0 300 477 nm 1.4 10 6 608 nm 0.25 Abs (a. u.) UV-Vis and Emission at pH 4.4 Excitation @ 477 nm Abs (a. u.) 345 nm 0.3 Lys108 0.1 pKa = 5.2 R2 = 0.99 0.05 0 400 500 600 700 Wavelength (nm) 3 800 4 5 6 7 8 pH Figure IV-27. UV-Vis, emission spectra Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor-10. and pH titration of Spectroscopic properties of Q108K:K40E:T53A:R58L:Q38F:Q4F bound with ThioFluor-10 are shown in Figure IV-27. At neutral pH, SB is present, which upon excitation yields dual fluorescence with emission maxima of 451 nm and 608 nm. This is indicative of only partial ESPT, as some of the typical SB emission occurs at 451 nm. The quantum yield is 0.17, a substantial decrease than that observed for ThioFluor (0.51). Upon acidification and formation of the 244 PSB in the ground state, an absorption maximum of 477 nm is observed. Excitation of this peak results in an emission maximum of 666 nm, which is approximately 60 nm more red-shifted than the red-shifted emission maximum obtained upon excitation of the SB. The iminium pKa was determined to be 5.2, which is almost the same as the pKa determined for ThioFluor (5.3) with the same hCRBPII mutant. Table IV-12. Spectroscopic properties of ThioFluor analogues coupled with Q108K:K40E:T53A:R58L:Q38F:Q4F. Ligand SB PSB pKa λabs λem Φ ΦESPT λabs λem ThioFluor 397 605 0.51 > 0.98 575 671 5.3 ThioFluor-3 395 611 0.58 > 0.98 554 674 nd ThioFluor-4 369 479 0.02 < 0.02 470 550 nd ThioFluor-8 426 642 0.45 > 0.98 650 701 5.5 ThioFluor-9 385 470, 635 0.10 0.45 550 677 4.6 ThioFluor-10 345 451, 608 0.17 0.66 477 666 5.2 The data for ThioFluor analogues bound with Q108K:K40E:T53A:R58L:Q38F:Q4F excited at SB and PSB is summarized in Table IV-12. In general, it seems that the highly conjugated systems (ThioFluor, ThioFluor-3 and ThioFluor-8) result in only the red-shifted emission band upon SB excitation, indicating that the fluorescence occurs via a pronated excited state. These three complexes have the most red-shift SB and PSB, indicating the higher degree of conjugation. On the other hand, the SB of ThioFluor-4 (with the less donating methoxy group) is almost 30 nm blue-shifted than that of ThioFluor; in addition, the PSB is 100 nm more blue-shifted. It can be inferred 245 that the less electron donating group leads to less overall conjugation to the iminium. As no ESPT emission is observed, it can be reasoned that the methoxy group cannot sufficiently stabilize a protonated excited state. ThioFluor-9 and ThioFluor-10 show partial ESPT when the SBs of the Q108K:K40E:T53A:R58L:Q38F:Q4F complexes are irradiated. At first, it was counterintuitive for ThioFluor-10 to show more ESPT than ThioFluor-9. It was expected that it would be harder to break aromaticity in the biphenyl ThioFluor10, leading to less conjugation than ThioFluor-9, which possess a phenyl ring and the less aromatic thiophene ring as the second aromatic linker.44 The hypothesis that ThioFluor-10 is less conjugated than ThioFluor-9 is supported by the UV-Vis absorption of ThioFluor-10-SB (λmax = 345 nm) and ThioFluor-9SB (λmax = 385 nm); ThioFluor-10-SB displays an absorption maxima 40 nm hypsochromically shifted of ThioFluor-9-SB. From this data, it is apparent that another process is partially responsible for the dual fluorescence observed upon irradiation of ThioFluor-10-SB. Large Stokes shifts of hCRBPII/ThioFluor-10 complexes were previously observed (Chapter III); we hypothesized that this was due to the formation of a twisted TICT state upon excitation.10, 45, 46 In this case, it is hypothesized that excitation of ThioFluor-10-SB leads to a proton transfer in the excited state, assisted via TICT. These two processes can take place cooperatively, and lead to the emission profile observed. TICT facilitated ESPT was previously shown to be 246 energetically preferable (through computational studies) in hydrogen bonding donor-acceptor models.47-49 IV.8 Conclusions Typically, large Stokes shift resulting from proton transfer do not lead to probes that are able to meet practical demands, as there are many parameters that are difficult to control, such as pH, solvent, etc. In contrast, our developed protein/probe complex exhibits excited state proton transfer only when an acidic residue is placed in the appropriate orientation to the protein-bound iminium, rendering the probe ideally suited for live-cell imaging and a turn on fluorescent response over a broad pH range. To our delight, the Q108K:K40E:T53A:R58L:Q38F:Q4F/ThioFluor probe can be used to image live cells with a Stokes shift of over 200 nm (and quantum yields up to 0.51). The limitations of this system stem from the solvatochromic nature of the fluorophores, as the free aldehyde appears to emit background fluorescence above 600 nm. Efforts should be focused on fluorophore design applicable toward live-cell imaging without the need for a wash step, to allow real time measurements. Additionally, crystallographic studies and ultrafast spectroscopy may prove invaluable in shedding a detailed light on the mechanism of the excited state proton transfer process observed here. 247 REFERENCES 248 REFERENCES 1. Weissleder, R. A clearer vision for in vivo imaging. Nature Biotechnology 19, 316-317 (2001). 2. Weissleder, R. & Ntziachristos, V. Shedding light onto live molecular targets. Nature Medicine 9, 123-128 (2003). 3. He, B., Nie, H., Chen, L., Lou, X., Hu, R., Qin, A., Zhao, Z. & Tang, B.Z. High Fluorescence Efficiencies and Large Stokes Shifts of Folded Fluorophores Consisting of a Pair of Alkenyl-Tethered, pi-Stacked Oligo-pphenylenes. Organic Letters 17, 6174-6177 (2015). 4. Miura, T., Urano, Y., Tanaka, K., Nagano, T., Ohkubo, K. & Fukuzumi, S. Rational design principle for modulating fluorescence properties of fluorescein-based probes by photoinduced electron transfer. Journal of the American Chemical Society 125, 8666-8671 (2003). 5. Trenor, S.R., Shultz, A.R., Love, B.J. & Long, T.E. Coumarins in polymers: From light harvesting to photo-cross-linkable tissue scaffolds. Chemical Reviews 104, 3059-3077 (2004). 6. Loudet, A. & Burgess, K. BODIPY dyes and their derivatives: Syntheses and spectroscopic properties. Chemical Reviews 107, 4891-4932 (2007). 7. Beija, M., Afonso, C.A.M. & Martinho, J.M.G. Synthesis and applications of Rhodamine derivatives as fluorescent probes. Chemical Society Reviews 38, 2410-2433 (2009). 8. Lu, H., Mack, J., Yang, Y. & Shen, Z. Structural modification strategies for the rational design of red/NIR region BODIPYs. Chemical Society Reviews 43, 4778-4823 (2014). 9. Grabowski, Z.R., Rotkiewicz, K. & Rettig, W. Structural changes accompanying intramolecular electron transfer: Focus on twisted intramolecular charge-transfer states and structures. Chemical Reviews 103, 3899-4031 (2003). 10. Sasaki, S., Drummen, G.P.C. & Konishi, G. Recent advances in twisted intramolecular charge transfer (TICT) fluorescence and related phenomena in materials chemistry. Journal of Materials Chemistry C 4, 2731-2743 (2016). 249 11. Dereka, B., Koch, M. & Vauthey, E. Looking at Photoinduced Charge Transfer Processes in the IR: Answers to Several Long-Standing Questions. Accounts of Chemical Research 50, 426-434 (2017). 12. Klymchenko, A.S. Solvatochromic and Fluorogenic Dyes as EnvironmentSensitive Probes: Design and Biological Applications. Accounts of Chemical Research 50, 366-375 (2017). 13. Loving, G.S., Sainlos, M. & Imperiali, B. Monitoring protein interactions and dynamics with solvatochromic fluorophores. Trends in Biotechnology 28, 73-83 (2010). 14. Klymchenko, A.S. & Mely, Y. Fluorescent Environment-Sensitive Dyes as Reporters of Biomolecular Interactions. Progress in Molecular Biology and Translational Science 113, 35-58 (2013). 15. Vollmer, F., Rettig, W. & Birckner, E. Photochemical mechanisms producing large fluorescence stokes shifts. Journal of Fluorescence 4, 6569 (1994). 16. Okamoto, A., Ichiba, T. & Saito, I. Pyrene-labeled oligodeoxynucleotide probe for detecting base insertion by excimer fluorescence emission. Journal of the American Chemical Society 126, 8364-8365 (2004). 17. Honda, K., Fujishima, S.H., Ojida, A. & Hamachi, I. Pyrene excimer-based dual-emission detection of a oligoaspartate tag-fused protein by using a Zn(II)-DpaTyr probe. ChemBiochem 8, 1370-1372 (2007). 18. Padalkar, V.S. & Seki, S. Excited-state intramolecular proton-transfer (ESIPT)-inspired solid state emitters. Chemical Society Reviews 45, 169202 (2016). 19. Serdiuk, I.E. & Roshal, A.D. Exploring double proton transfer: A review on photochemical features of compounds with two proton-transfer sites. Dyes and Pigments 138, 223-244 (2017). 20. Song, P. & Ma, F.C. Intermolecular hydrogen-bonding effects on photophysics and photochemistry. International Reviews in Physical Chemistry 32, 589-609 (2013). 21. Klymchenko, A.S., Pivovarenko, V.G., Ozturk, T. & Demchenko, A.P. Modulation of the solvent-dependent dual emission in 3hydroxychromones by substituents. New Journal of Chemistry 27, 13361343 (2003). 250 22. Zimmer, M. GFP: from jellyfish to the Nobel prize and beyond. Chemical Society Reviews 38, 2823-2832 (2009). 23. Tsien, R.Y. The green fluorescent protein. Annual Review of Biochemistry 67, 509-544 (1998). 24. Oltrogge, L.M., Wang, Q. & Boxer, S.G. Ground-state proton transfer kinetics in green fluorescent protein. Biochemistry 53, 5947-5957 (2014). 25. Jung, G., Wiehler, J. & Zumbusch, A. The photophysics of green fluorescent protein: Influence of the key amino acids at positions 65, 203, and 222. Biophysical Journal 88, 1932-1947 (2005). 26. Lossau, H., Kummer, A., Heinecke, R., PollingerDammer, F., Kompa, C., Bieser, G., Jonsson, T., Silva, C.M., Yang, M.M., Youvan, D.C. & MichelBeyerle, M.E. Time-resolved spectroscopy of wild-type and mutant Green Fluorescent Proteins reveals excited state deprotonation consistent with fluorophore-protein interactions. Chemical Physics 213, 1-16 (1996). 27. Chattoraj, M., King, B.A., Bublitz, G.U. & Boxer, S.G. Ultra-fast excited state dynamics in green fluorescent protein: multiple states and proton transfer. Proceedings of the National Academy of Sciences of the United States of America 93, 8362-8367 (1996). 28. Brejc, K., Sixma, T.K., Kitts, P.A., Kain, S.R., Tsien, R.Y., Ormo, M. & Remington, S.J. Structural basis for dual excitation and photoisomerization of the Aequorea victoria green fluorescent protein. Proceedings of the National Academy of Sciences of the United States of America 94, 23062311 (1997). 29. Henderson, J.N., Osborn, M.F., Koon, N., Gepshtein, R., Huppert, D. & Remington, S.J. Excited State Proton Transfer in the Red Fluorescent Protein mKeima. Journal of the American Chemical Society 131, 1321213213 (2009). 30. Kogure, T., Karasawa, S., Araki, T., Saito, K., Kinjo, M. & Miyawaki, A. A fluorescent variant of a protein from the stony coral Montipora facilitates dual-color single-laser fluorescence cross-correlation spectroscopy. Nature Biotechnology 24, 577-581 (2006). 31. Violot, S., Carpentier, P., Blanchoin, L. & Bourgeois, D. Reverse pHDependence of Chromophore Protonation Explains the Large Stokes Shift of the Red Fluorescent Protein mKeima. Journal of the American Chemical Society 131, 10356-10357 (2009). 251 32. Bergeler, M., Mizuno, H., Fron, E. & Harvey, J.N. QM/MM-Based Calculations of Absorption and Emission Spectra of LSSmOrange Variants. Journal of Physical Chemistry B 120, 12454-12465 (2016). 33. Chu, J., Oh, Y., Sens, A., Ataie, N., Dana, H., Macklin, J.J., Laviv, T., Welf, E.S., Dean, K.M., Zhang, F.J., Kim, B.B., Tang, C.T., Hu, M., Baird, M.A., Davidson, M.W., Kay, M.A., Fiolka, R., Yasuda, R., Kim, D.S., Ng, H.L. & Lin, M.Z. A bright cyan-excitable orange fluorescent protein facilitates dual-emission microscopy and enhances bioluminescence imaging in vivo. Nature Biotechnology 34, 760-767 (2016). 34. Fron, E., De Keersmaecker, H., Rocha, S., Baeten, Y., Lu, G., Uji-I, H., Van der Auweraer, M., Hofkens, J. & Mizuno, H. Mechanism Behind the Apparent Large Stokes Shift in LSSmOrange Investigated by TimeResolved Spectroscopy. Journal of Physical Chemistry B 119, 1488014891 (2015). 35. Pletnev, S., Shcherbakova, D.M., Subach, O.M., Pletneva, N.V., Malashkevich, V.N., Almo, S.C., Dauter, Z. & Verkhusha, V.V. Orange Fluorescent Proteins: Structural Studies of LSSmOrange, PSmOrange and PSmOrange2. Plos One 9 (2014). 36. Shcherbakova, D.M., Hink, M.A., Joosen, L., Gadella, T.W.J. & Verkhusha, V.V. An Orange Fluorescent Protein with a Large Stokes Shift for Single-Excitation Multicolor FCCS and FRET Imaging. Journal of the American Chemical Society 134, 7913-7923 (2012). 37. Chen, F.S., Zeng, Q., Zhuang, W. & Liang, W.Z. Characterizing the Structures, Spectra, and Energy Landscapes Involved in the Excited-State Proton Transfer Process of Red Fluorescent Protein LSSmKate1. Journal of Physical Chemistry B 120, 9833-9842 (2016). 38. Piatkevich, K.D., English, B.P., Malashkevich, V.N., Xiao, H., Almo, S.C., Singer, R.H. & Verkhusha, V.V. Photoswitchable Red Fluorescent Protein with a Large Stokes Shift. Chemistry & Biology 21, 1402-1414 (2014). 39. Piatkevich, K.D., Hulit, J., Subach, O.M., Wu, B., Abdulla, A., Segall, J.E. & Verkhusha, V.V. Monomeric red fluorescent proteins with a large Stokes shift. Proceedings of the National Academy of Sciences of the United States of America 107, 5369-5374 (2010). 40. Piatkevich, K.D., Malashkevich, V.N., Almo, S.C. & Verkhusha, V.V. Engineering ESPT Pathways Based on Structural Analysis of LSSmKate 252 Red Fluorescent Proteins with Large Stokes Shift. Journal of the American Chemical Society 132, 10762-10770 (2010). 41. Randino, C., Moreno, M., Gelabert, R. & Lluch, J.M. Peek at the Potential Energy Surfaces of the LSSmKate1 and LSSmKate2 Proteins. Journal of Physical Chemistry B 116, 14302-14310 (2012). 42. Shcherbo, D., Merzlyak, E.M., Chepurnykh, T.V., Fradkov, A.F., Ermakova, G.V., Solovieva, E.A., Lukyanov, K.A., Bogdanova, E.A., Zaraisky, A.G., Lukyanov, S. & Chudakov, D.M. Bright far-red fluorescent protein for whole-body imaging. Nature Methods 4, 741-746 (2007). 43. Yang, J., Wang, L., Yang, F., Luo, H.M., Xu, L.L., Lu, J.L., Zeng, S.Q. & Zhang, Z.H. mBeRFP, an Improved Large Stokes Shift Red Fluorescent Protein. Plos One 8 (2013). 44. Chen, W., Li, H., Widawsky, J.R., Appayee, C., Venkataraman, L. & Breslow, R. Aromaticity decreases single-molecule junction conductance. Journal of the American Chemical Society 136, 918-920 (2014). 45. Zhang, Z.Y., Zhang, G.Q., Wang, J.X., Sun, S.S. & Zhang, Z.Z. The mechanisms of Large Stokes Shift and Fluorescence Quantum Yields in anilino substituted Rhodamine analogue: TICT and PICT. Computational and Theoretical Chemistry 1095, 44-53 (2016). 46. Haberhauer, G., Gleiter, R. & Burkhart, C. Planarized Intramolecular Charge Transfer: A Concept for Fluorophores with both Large Stokes Shifts and High Fluorescence Quantum Yields. Chemistry-A European Journal 22, 971-978 (2016). 47. Kim, C.H., Park, J., Seo, J., Park, S.Y. & Joo, T. Excited state intramolecular proton transfer and charge transfer dynamics of a 2-(2'hydroxyphenyl)benzoxazole derivative in solution. The Journal of Physical Chemistry A 114, 5618-5629 (2010). 48. Wang, D.D., Lu, R., Yuan, M.H., Chen, J.S., Feng, L.Q., Fu, A.P., Tian, F.H., Varandas, A.J.C. & Chu, T.S. Photoinduced coupled twisted intramolecular charge transfer and excited-state proton transfer via intermolecular hydrogen bonding: A DFT/TD-DFT study. Chemical Physics Letters 610, 179-185 (2014). 49. Wang, Y., Shi, Y., Cong, L. & Li, H. TDDFT study of twisted intramolecular charge transfer and intermolecular double proton transfer in the excited 253 state of 4 '-dimethylaminoflavonol in ethanol solvent. Spectrochimica Acta Part A-Molecular and Biomolecular Spectroscopy 137, 913-918 (2015). 254 CHAPTER V: DEVELOPMENT OF A SINGLE PROTEIN FLUORESCENT RATIOMETRIC PH SENSOR Our lab previously showed success in reengineering cellular retinoic acid binding protein II (CRABPII) to bind retinal as a protonated Schiff base (PSB), demonstrating wavelength tuning over a 166 nm range (474 nm to 640 nm) and pKa modulation from 2.4 to 8.1.1 This led to the development of a colorimetric pH sensor, in which a CRABPII-CRABPII protein fusion was used to measure pH ratiometrically. The two CRABPII R111K:R132L:Y134F:T54V:L121Q:R59W (pKa mutants chosen 4.4, 505 λmax nm) were and R111K:R132L:Y134F:T54V:R59W:A32W:M93L:E73A (pKa 2.6, λmax 607 nm). Figure V-1a depicts the absorption spectra of the individually expressed CRABPII mutants when bound to retinal. The two proteins were cloned into the expression vector pET-17b with a three amino acid linker. The expressed protein fusion was incubated with 2 equiv of retinal in citrate buffer, and subsequently titrated (Figure V-1b). As expected, upon acidification, an iminium corresponding to the high pKa mutant first appears; further acidification leads to the iminium of the low pKa mutant. To build a standard curve, the ratio of absorbance intensity at 480 nm to 650 nm was plotted against pH (Figure V-1c); standard curves at concentration of 58 μM and 19 μM overlapped, purporting the concentration independent response of the system. As a proof of principle, the protein fusion was used to measure the pH of known buffer solutions with accuracy within 0.2 pH units. 255 a. 1 b. λmax1, pKa1 pH 5.68 pH 4.92 pH 4.52 pH 3.90 pH 3.35 Abs (a. u.) 0.8 0.6 pH 2.79 pH 2.29 pH 1.84 pH 1.33 0.4 0.2 λmax2, pKa2 0 320 2 480 560 640 Wavelength (nm) 720 58 mM 19 mM 4 /A 480 a 607 nm pK = 2.6 a Ratio of A 505 nm pK = 4.4 650 1.5 Abs (a. u.) 5 c. 400 1 0.5 3 2 1 0 400 500 600 Wavelength (nm) 1 700 2 3 4 5 6 pH Figure V-1. Two-protein-based ratiometric probe design. a. The selected proteins have distinct wavelength (UV-Vis) and pKa values. b. pH titration of the fused protein. c. Ratiometric analysis of the fused protein system at two different concentrations. The system just discussed relied on the ratio of absorbances. However, fluorescence based pH probes are suitable for more applications. Fluorescence detection is greater than absorption spectrophotometric methods, which leads to greater limits of detection. Thus, Dr. Tetyana Berbasova pioneered efforts to develop a fluorescent ratiometric pH sensor. The fortuitous discovery that human Cellular Retinol Binding Protein (hCRBPII) mutants containing a L117E mutation showed a pH dependent wavelength shift of the protonated Schiff based (PSB) when bound to julolidine retinal analog (JRA) allowed for the generation of a single protein ratiometric pH sensor.2 As seen in Figure V-2a, the pentamutant 256 Q108K:K40L:T51V:R58F:L117E showed two distinct absorption maxima at pH 5.0 and pH 8.0. The tetramutant lacking the L117E mutation shows the necessity of the titratable carboxylate side chain near the iminium (Figure V-2b). It was hypothesized that the carboxylate form of L117E localizes charge on the PSB, which leads to the observed blue-shift (Figure V-2c). Acidification results in protonation of the side chain, leading to better delocalization of charge and a redshift in absorption wavelength. Q108K:K40L:T51V:R58F:L117E Q108K:K40L:T51V:R58F:L117E 1.2 pH 8.0 pH 7.3 pH 7.0 pH 6.4 pH 5.0 Abs (a. u.) 0.9 1.5 1.2 0.3 644 nm pH 10.24 pH 10.00 pH 9.63 pH 8.96 pH 8.50 pH 7.84 pH 7.02 pH 6.42 pH 5.65 615 nm 581 nm 0.6 Q108K:K40L:T51V:R58F Q108K:K40L:T51V:R58F b. Abs (a. u.) a. 0.9 0.6 0.3 0 500 0 550 600 650 Wavelength (nm) 700 400 500 600 Wavelength (nm) 700 c. N H H+ N O λmax = 581 nm O N H N Lys108 O Glu117 λmax = 615 nm more localized charge OH Lys108 Glu117 less localized charge Figure V-2. Single protein ratiometric probe design. a. Titration of L117E leads to a shift in wavelength. b. Removal of L117E does not lead to the pH induced wavelength change. c. A blue-shift in wavelength results due to localizing charge on the iminium through interaction with the carboxylate, while protonation of the L117E side chain leads to less localized charge. 257 Through mutagenesis of active site residues, the largest shift between the two PSB maxima achieved was 95 nm (565 nm and 660 nm) in the nonamutant Q108K:K40L:T51V:T53C:R58W:T29L:A33W:Q4F:L117D, with a quantum yield of 0.18.2 The large shift in absorption wavelength allows for excitation at two distinct wavelengths, from which the emission ratios can be plotted as a function of pH to develop a ratiometric standard curve. It was envisioned that the ratio of emission intensities at the two excitation wavelengths in live cells expressing hCRBPIIL117E could be obtained. From the standard curve, the in vivo pH could then be determined. Unfortunately live cell imaging was not fruitful; JRA leads to significant background fluorescence when incubated with both HeLa and U2OS cells expressing hCRBPII, while specific labeling was not achieved. V.1 Initial observation of pH induced wavelength shifts in a ThioFluor/hCRBPII complex In the course of optimizing binding of ThioFluor to hCRBPII mutants in wavelength regulation studies (Chapter II), it was observed that the inclusion of L117E produced the same pH induced wavelength shift as was observed for JRA. ThioFluor was coupled with Q108K:K40L:T51V:T53S:R58W:Y19W:L117E (Figure the heptamutant V-3). At pH 7, an absorption maximum of 558 nm is apparent; acidification leads to a bathochromic shift in absorption with a maximum of 614 nm, a 56 nm change. When the absorption intensity at 614 nm was plotted as a function of pH, a double pKa 258 curve is obtained. Fitting two pKas to the curve yields values of 6.6 and 11.2, which presumably correspond to the pKas of L117E and iminium, respectively. 5.0 5.4 5.9 0.25 6.5 7.0 7.6 8.4 9.8 10.7 11.2 11.6 11.9 12.3 0.25 Abs. (a. u.) 0.2 ΔAbs at 614 nm (a. u.) 614 nm 558 nm 0.15 0.1 0.05 0.2 pka = 6.6 R2 = 0.99 0.15 0.1 pka = 11.2 R2 = 0.97 0.05 0 0 4 350 400 450 500 550 600 650 700 Wavelength (nm) 6 8 10 12 pH Figure V-3. Acid-base titration Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor. To our delight, the crystal of structure of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E (crystallized at 1.2 Å resolution by Ms. Zahra Assar) shows a distance of 2.9 Å from the glutamic acid side chain to the iminium (Figure V-4). Unexpectedly, this high pKa mutant exists as a trans iminium when crystallized at pH 4 – 4.5; in crystal structures of retinal/hCRBPII, it was previously shown that high pKa mutants typically exhibited a cis iminium.3, 4 As previously Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor stated, has an absorption maximum of 558 nm; acidification leads to a bathochromic shift in absorption with a maximum of 614 nm, a 56 nm change. This bathochromic shift in wavelength with protonation of the L117E side chain is in agreement with the idea that an even distribution of electrostatic potential along the chromophore leads to red- 259 shifting. Moreover, the mutant without L117E (Q108K:K40L:T51V:T53S:R58W:Y19W) is more red-shifted, with an absorption maximum of 653 nm and extinction coefficient of approximately 40,000 M-1cm-1. L117E 2.9 Å Figure V-4. Crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor highlighting the location of L117E. Density is shown at 1σ. Not only does the addition of L117E to this mutant produce the pH induced wavelength change, but it also has the added benefit of increasing the rate of iminium formation by 4.5 fold Q108K:K40L:T51V:T53S:R58W:Y19W (t1/2 and = 370 82 minutes minutes for for Q108K:K40L:T51V:T53S:R58W:Y19W:L117E; iminium formation was measured via UV-Vis at 23°C with 20 μM protein and 0.5 equiv ligand, and plotted to second order rate kinetics). A disadvantage of the latter system is that the extinction coefficient of hCRBPII/ThioFluor is significantly decreased with the addition of L117E (approximately two-fold). Ms. Zahra Assar was also successful in obtaining the crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor (at 1.2 Å 260 Figure V-5. a. Crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor. Density is shown at 1σ. b. Overlay of Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor (magenta) and Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (cyan). resolution) which clearly shows a cis iminium stabilized by a water mediated hydrogen bond to Q4 (Figure V-5a). An overlay Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor of and shows that the chromophore is translocated toward the protein exterior when the trans iminium is present (Figure V-5b). However, there are no significant differences in any of the nearby residues. It is hypothesized that the decrease in extinction coefficient upon L117E introduction is due to the isomerization at the iminium, the 261 translocation of the chromophore, or because of simply introducing polarity near the iminium. Fortuitously this Q108K:K40L:T51V:T53S:R58W:Y19W:L117E complex has shown success in labeling live HeLa cells with whole cell labeling, and both nuclear and cytosolic targeting (Chapter II.11.2). We first chose to explore how the excitation wavelengths affect the emission ratio. To this end, we collected emission spectra at four excitation wavelengths (at each pH) that correspond to lasers available on the confocal microscope: 514 nm, 543 nm, 594 nm and 633 nm. Each emission spectrum was then integrated from either 615 nm to 900 nm or 650 nm to 900 nm, in order to mimic the long pass (LP) 615 nm and 650 nm filters available on the confocal microscope. All data is shown in Table V-1 for Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor. Table V-1. Emission ratios (in vitro) as a function of pH for Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor. λem1/λem2 (LP 615 nm) λem1/λem2 (LP 650 nm) pH 514/ 543 514/ 594 543/ 594 514/ 543 514/ 594 514/ 633 543/ 594 543/ 633 594/ 633 5.0 0.55 0.37 0.67 0.55 0.36 0.50 0.66 0.91 1.39 5.4 0.57 0.40 0.70 0.57 0.39 0.56 0.69 0.99 1.44 5.9 0.62 0.50 0.79 0.61 0.47 0.71 0.77 1.17 1.52 6.5 0.71 0.72 1.00 0.70 0.66 1.15 0.95 1.65 1.74 7.0 0.79 1.05 1.32 0.78 0.98 2.35 1.25 3.01 2.40 7.6 0.82 1.27 1.54 0.81 1.20 4.29 1.48 5.29 3.57 8.4 0.83 1.38 1.67 0.82 1.33 6.52 1.62 7.94 4.90 9.8 0.83 1.42 1.71 0.82 1.37 7.59 1.67 9.21 5.53 10.7 0.84 1.49 1.77 0.83 1.44 8.32 1.72 10.00 5.80 262 As seen in Table V-1, cutting the emission off at 615 nm or 650 nm makes essentially no difference in the ratio of emission intensities (i.e. 0.84 and 0.83 for excitation wavelengths of 514 nm and 543 nm, respectively). However, the emission ratio significantly varies depending on the two excitation wavelengths used. The smallest changes between pH 5.0 and pH 10.7 are observed when the excitation wavelengths are closer (594 and 633 nm). On the other hand, when the excitation wavelengths are further apart, and close to the maxima of the hCRBPII/ThioFluor complex, the emission ratio is higher. 10 594 nm / 633 nm 543 nm / 633 nm 514 nm / 633 nm 11x increase 17x increase 6 4x increase λ em1 /λ em2 8 4 2 0 5 6 7 8 9 10 pH Figure V-6. Select standard curves obtained upon excitation of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor at 514/633 nm, 543/633 nm and 594/633 nm. In the case of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor, the highest emission ratio is obtained upon using excitation wavelengths of 543 nm and 633 nm. Standard curves can be generated by plotting emission ratio versus pH. An overlay of the three best standard curves is shown in Figure V-6. The largest increase in fluorescence emission ratio is observed when 514 nm 263 and 633 nm excitation wavelengths are used, leading to a 17-fold dynamic range. Presumably this is because the two different iminiums can be selectively excited at the tail end of their absorbances. In doing so the emission ratio starts more toward zero, indicating that with an excitation wavelength of 514 nm, little fluorescence is observed from the iminium when L117E is protonated. Based on this curve, the dynamic range is 6.5 to 8.5, which should prove useful for measuring cells at physiological pH. With this initial hit, we sought to optimize its use for live cell imaging. To this end, we made point Q108K:K40L:T51V:T53S:R58W:Y19W:L117E mutations heptamutant. in The the following requirements were kept in mind when engineering hCRBPII for applicability toward a single fluorescent protein ratiometric pH sensor. Firstly, we sought to increase the change between the two absorptive states to allow for excitation at two distinct wavelengths. It is also essential to have a high pKa value of the iminium, such that the iminium remains protonated during the titration. This will ensure that a change to the absorption spectra of the iminium is registered as the result of titrating the side chain carboxylic acid. It is important to note that the titratable side chain should possess a pKa that allows a pH working range that is useful at physiological pH. Lastly, as in all imaging methods, fast kinetics and high brightness (the product of extinction coefficient and quantum yield) are ideal in live cell imaging, so that measurements can be taken directly following fluorophore incubation. 264 V.2 Optimization of hCRBPII for a ratiometric pH probe In screening hCRBPII mutants for ThioFluor binding, it was observed that T53D and R58H mutations increased the rate of iminium formation. To this end, they were introduced individually into the parent template Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. In both cases, the absorption wavelengths of the iminium, upon titrating the L117E, are more separated. Advantageously, the quantum yield of the PSB was also increased. Unfortunately though, the introduction of T53D (Table V-2, entry 2) significantly decreased both the rate of iminium formation and the extinction coefficient of the bound complex. Table V-2. Initial mutagenesis of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E for development toward a single fluorescent protein ratiometric sensor. t1/2 Entry hCRBPII mutanta λabsb ΔAbs pKac Φd (min)e 558, 11.2, 1 KLVS:R58W:Y19W:L117E 56 0.31 82 614 6.6 530, 9.7, 2 KLVD:R58W:Y19W:L117E 76 0.41 217 606 6.9 530, 10.9, 3 KLVS:R58H:Y19W:L117E 68 0.37 17 598 6.6 539, 11.3, 4 KLVS:R58W:Y19W:L117E:A33W 89 0.28 241 628 8.0 547, 10.9, 5 KLVS:R58H:Y19W:L117E:A33W 67 0.29 30 614 7.0 a KLVX = Q108K:K40L:T51V:T53X. bλmax of absorbance at basic pH and λmax of absorbance at acidic pH. cpKa corresponding to the iminium and pKa corresponding to Glu117. dQuantum yield measured at pH ~7.2. eKinetic measurements were performed at 23 °C with 20 μM protein and 0.5 equiv ligand. PSB formation was monitored by UV-Vis at λmax for each complex over time at pH ~7.2. Data was then fit to second order kinetics. To our delight the substitution of R58H to yield Q108K:K40L:T51V:T53S:R58H:Y19W:L117E (Table V-2, entry 3) led to fast iminium formation (Figure V-7a), and retained the high iminium pKa (Figure V-7b 265 and Figure V-7c). Fortuitously, the highest emission ratio obtained when excited at 514 nm and 633 nm is also significantly increased to approximately 40 (Figure V-7d), which should allow for more sensitive detection. The disadvantage of this complex is that the working range (7.4 – 9.4) is not useful for most physiological applications, but it may find use in measuring the pH of mitochondria (pH ~ 8).5 a. 5.0 5.5 6.0 6.5 6.9 b. 1 10 0.3 -5 8 10-6 6 10-6 4 10 k = 6112 M-1 min-1 t1/2 = 16.4 min R2 = 0.99 -6 2 10-6 598 nm 530 nm 0.2 0.15 0.1 0.05 0 0 0 c. 50 100 150 200 250 300 350 400 Time (min) d. 0.3 350 400 450 500 550 600 650 700 Wavelength (nm) 50 40 0.15 0.1 pka = 10.9 R2 = 0.97 0.05 Working range: 7.4 - 9.4 30 /λ pka = 6.6 R2 = 0.99 em,514nm 0.2 em,633nm 0.25 20 λ ΔAbs at 598 nm (a. u.) 10.4 10.8 11.2 11.6 12.2 0.25 Abs. (a. u.) Concentration of Complex (M) 1.2 10 -5 7.3 7.7 8.1 8.8 9.9 10 0 0 4 6 8 10 12 5 pH 6 7 8 9 10 pH Figure V-7. In vitro characterization of Q108K:K40L:T51V:T53S:R58H:Y19W:L117E/ThioFluor as a single protein ratiometric pH sensor. a. Formation of PSB, monitored at 539 nm via UV-Vis, fit to second order rate kinetics. b. UV-Vis titration of the hCRBPII/ThioFluor complex. c. Double pKa curve obtained by plotting the absorbance intensity at 598 nm as a function of pH. d. Standard curve generated by plotting the emission ratio of 514/633 nm versus pH. 266 Ms. Zahra Assar obtained the crystal structure of Q108K:K40L:T51V:T53S:R58H:Y19W:L117E with ThioFluor at 1.1 Å resolution. The crystal structures show similar trajectories for the chromophore and the active site lysine, location of L117E and all surrounding residues. Even histidine at residue 58 adopts the same rotamer as R58W. Assuming similar binding affinities of ThioFluor to the hCRBPII mutants, it is only the substitution at R58 that affects the rate of iminium formation. Thus, it is hypothesized that the less sterically encumbering histidine accelerates the reaction. It is also possible that the histidine hydrogen bonds with the N,N-dimethyl moiety of the chromophore leading to a partially positively charged nitrogen. Less electron donation from the N,N-dimethyl moiety would lead to a more electrophilic aldehyde. R58W R58H Figure V-8. Overlay of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (green) and Q108K:K40L:T51V:T53:R58H:Y19W:L117E/ThioFluor (magenta), highlighting R58 and L117E. Besides fast iminium formation, it would also be ideal to develop a bright hCRBPII/ThioFluor complex. The quantum yields obtained with L117E are sufficient, so attempts to increase brightness were undertaken by increasing the complex’s extinction coefficient. In wavelength regulation studies, it was 267 observed that A33W led to an increase in extinction coefficient, presumably by encapsulation of the binding cavity. To this end, A33W was added to the parent hCRBPII heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117E (Table V-2, entry 4). Interestingly the extinction coefficient at the carboxylate state of L117E is ~15,000 M-1cm-1, which is similar to that observed for Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. However the extinction coefficient of the iminium with the protonated L117E residue increases by two-fold to ~40,000 M-1cm-1 (Figure V-10a). At the same time, the quantum yield is not significantly changed; this leads to a two-fold increase in brightness. The crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E:A33W was crystallized with ThioFluor by Mr. Alireza Ghanbarpour with 1.3 Å resolution. It appears that the inclusion of A33W results in a cis iminium (which interacts with Q4 through a water molecule), with the concomitant translation of the chromophore deeper into the binding pocket (Figure V-9). R58W L117E A33 A33W Q4 2.9 Å 2.8 Å Figure V-9. Overlay of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (green) and Q108K:K40L:T51V:T53S:R58W:Y19W:L117E:A33W/ThioFluor (magenta). 268 2.9 4.3 4.6 4.9 5.2 5.9 6.5 6.8 7.3 7.8 0.5 Abs. (a. u.) 0.4 0.3 8.4 9.4 10.1 10.5 11.0 11.6 12.1 12.3 12.7 628 nm b. 0.5 ΔAbs at 628 nm (a. u.) a. 539 nm 0.2 0.1 0.3 pka = 8.0 R2 = 0.99 0.2 0.1 0 0 350 400 450 500 550 600 650 700 Wavelength (nm) c. 4 d. pka = 11.3 R2 = 0.96 0.05 8 10 12 1.2 10 -5 Concentration of Complex (M) 0.1 6 pH 0.15 ΔAbs at 539 nm (a. u.) 0.4 1 10-5 8 10-6 6 10-6 k = 415 M-1 min-1 t1/2 = 241 min R2 = 0.99 4 10-6 2 10-6 0 0 4 6 8 10 0 12 pH 200 400 600 Time (min) 800 Figure V-10. In vitro characterization of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E:A33W/ThioFluor as a single protein ratiometric pH sensor a. UV-Vis titration of the hCRBPII/ThioFluor complex. b. pKa curve obtained by plotting the absorbance intensity at 628 nm as a function of pH. c. pKa curve obtained by plotting the absorbance intensity at 539 nm as a function of pH. d. Formation of PSB, monitored at 628 nm via UVVis, fit to second order rate kinetics. As shown in the overlay of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E (green) and Q108K:K40L:T51V:T53S:R58W:Y19W:L117E:A33W (magenta), L117E does not change position. As discussed previously, L117E interacts with the trans imine nitrogen (2.9 Å) in Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. However, in the case of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E:A33W, it is quite apparent that L117E does not face the imine nitrogen atom when it is in 269 the cis orientation. It is hypothesized that this is why the pKa of L117E is higher when A33W is present. Fortuitously, a much larger change is also observed between the two iminium maxima. Q108K:K40L:T51V:T53S:R58W:Y19W:L117E:A33W displays a 89 nm shift (Figure V-10a), as compared to 56 nm for Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. Additionally, a high pKa of 11.3 is achieved (Figure V-10c). A pKa of 8.0 is observed for the glutamic acid side chain at position 117 (Figure V-10b); this is much higher than would be expected for a carboxylic acid moiety. While brightness is higher and the iminium maxima are more separated, this mutant does experience a much slower rate of iminium formation. For this reason, this hCRBPII mutant was not further explored for used as a single fluorescent protein ratiometric sensor. We next Y19W:L117E:A33W made the mutant (Table V-2, entry Q108K:K40L:T51V:T53S:R58H: 5). The previously discussed Q108K:K40L:T51V:T53S:R58W:Y19W:L117E:A33W mutant was ideal in terms of the large separation between the two iminium maxima. However, the rate of iminium formation was slow. The R58H mutation was added in order to increase the reaction rate. To our delight, the half-life of iminium formation was decreased from 241 minutes to 30 minutes (Figure V-11a), with the R58H substitution, comparable to that observed in Q108K:K40L:T51V:T53S:R58H:Y19W:L117E. However, the separation between the iminium maxima decreased to 67 nm (Figure V-11b). From these 270 data, it is obvious that Figure V-11. In vitro characterization of Q108K:K40L:T51V:T53S:R58H:Y19W:L117E:A33W/ThioFluor as a single protein ratiometric pH sensor. a. Formation of PSB, monitored at 585 nm via UVVis, fit to second order rate kinetics. b. UV-Vis titration of the hCRBPII/ThioFluor complex. c. pKa curve obtained by plotting the absorbance intensity at 614 nm as a function of pH. d. pKa curve obtained by plotting the absorbance intensity at 547 nm as a function of pH. Q108K:K40L:T51V:T53S:R58H:Y19W:L117E:A33W behaves similar to that of Q108K:K40L:T51V:T53S:R58H:Y19W:L117E when bound to ThioFluor and that A33W does not lead to a larger separation between the absorbance maxima in all cases. It should also be noted that the pKa of the L117E side chain is similar to the mutant without A33W (Figure V-11c and Figure V-11d), but the extinction 271 coefficients at both iminium states are advantageously as high as Q108K:K40L:T51V:T53S:R58W:Y19W:L117E:A33W (Figure V-11b). V.3 L117D mutation leads to a much larger change in the iminium maxima We next sought to compare the effect of L117D versus L117E as the titratable acid. The comparison was made in three different hCRBPII series. The rate of iminium formation was significantly increased with aspartate at residue 117, as compared to glutamate (Table V-3). Interestingly, L117D led to a more hypsochromic shift in absorption wavelength of the iminium at basic pH, as compared to L117E. Table V-3. Effect of L117D versus L117E in the development toward a single fluorescent protein ratiometric sensor. t1/2 h hCRBPII mutanta λabsab λabsc ΔAbs pKad pKae Φf g E.R. (min) KLVSWW:L117E 558 614 56 11.2 6.6 0.31 82 7 KLVSWW:L117D 525 618 93 10.9 6.7 0.30 14 21 KLVSWW:L117E:A33W 539 628 89 11.3 8.0 0.28 241 - KLVSWW:L117D:A33W 510 645 135 11.2 7.6 0.16 20 17 KLVSWW:L117E:Q4F 565 619 54 11.5 6.3 0.25 262 - KLVSWW:L117D:Q4F 529 623 94 11.0 5.7 0.31 8 16 a b KLVSWW = Q108K:K40L:T51V:T53S:R58W:Y19W. λmax of absorbance at basic pH. cλmax of absorbance at acidic pH. dpKa corresponding to the iminium. e pKa corresponding to Glu/Asp117. fQuantum yield measured at pH ~7.2. gKinetic measurements were performed at 23 °C with 20 μM protein and 0.5 equiv ligand. PSB formation was monitored by UV-Vis at λmax for each complex over time at pH ~7.2. Data was then fit to second order kinetics. hER = the highest emission ratio collected as a result of excitation at 514 nm and 633 nm. Titration of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:A33W leads to the largest change in the two absorption maxima, 135 nm. However, as was the 272 case with L117E in this mutant, the working pH range of the complex is too high, rendering the probe not suitable for use at physiological pH. 0.32 0.28 7.9 8.3 8.7 9.1 9.4 9.9 10.4 10.8 11.2 11.5 11.9 b. 0.3 Abs. (a. u.) 0.24 618 nm 0.2 525 nm 0.16 0.12 0.08 ΔAbs at 618 nm (a. u.) 5.0 5.3 5.8 6.4 6.9 7.5 a. pka = 6.7 R2 = 0.99 0.2 0.1 0.04 0 0 4 350 400 450 500 550 600 650 700 Wavelength (nm) c. 0.12 6 7 8 9 10 11 pH 25 0.1 em,633nm 20 pka = 10.9 R2 = 0.98 0.08 15 em,514nm /λ 0.06 Working range: pH 7.0 - 9.0 0.04 10 λ ΔAbs at 525 nm (a. u.) 5 d. 5 0.02 0 0 7 8 9 10 pH 11 12 13 5 6 7 8 9 10 pH Figure V-12. In vitro characterization of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor as a single protein ratiometric pH sensor. a. UV-Vis titration of the hCRBPII/ThioFluor complex. b. pKa curve obtained by plotting the absorbance intensity at 618 nm as a function of pH. c. pKa curve obtained by plotting the absorbance intensity at 525 nm as a function of pH. d. Standard curve generated by plotting the emission ratio of 514/633 nm versus pH. The heptamutant Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ ThioFluor iminium displayed a change of 93 nm upon titration of the aspartic acid side chain at residue 117. Characterization of the complex is shown in 273 Figure V-12. Similar to that previously observed for mutants containing L117E, the titration of L117D shows an effect on absorption wavelength of the iminium. The aspartic acid side chain exhibits a pKa of 6.7 (Figure V-12b), rendering an absorption maximum of 525 nm at pH 7.5. Acidification of the complex yields a bathochromic shift in absorption maximum to 618 nm (Figure V-12a). To our delight, generation of a standard curve based on excitation wavelengths of 514 nm and 633 nm yields a high emission ratio of approximately 20 (Figure V-12d), which is more than two-fold higher than for Q108K:K40L:T51V:T53S:R58W:Y19W:L117E (Figure V-6). Unfortunately, the working range (pH 7.0 – 9.0) is outside physiological pH. However, this probe would be useful to measure the pH of more basic organelles, such as the mitochondria (pH ~ 8).5 L117D R58W L117E Figure V-13. Overlay Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (green) Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor (magenta). of and It is interesting to note that the absorption maxima of the iminium at low pH is similar for Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor and Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor. However, a difference 274 of 33 nm is observed at high pH. According to the theory that blue-shifted absorption wavelengths occur from localization of charge, it has to be concluded that aspartate more efficiently localizes charge on the iminium as compared to glutamate. Fortuitously, Mr. Alireza Ghanbarpour provided the crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor at 1.0 Å resolution. An overlay of the crystal Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor structures and Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor is shown in Figure V13. No significant differences are apparent when comparing the crystal structures of Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor and Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor, except for the isomeric state of the iminium. In the case of L117E, the trans iminium is stabilized by the glutamate side chain. Conversely, the shorter L117D mutation does not provide this interaction (it interacts with Q4 through an acetate molecule). However, it is quite important to note that these structures were crystallized at low pH 4.0 – 4.5. From the UV-Vis, we can see that the absorption maxima of these two complexes are not significantly different at low pH. It is hypothesized that an acid induced isomerization occurs upon acidification when L117D is present. In the case of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ ThioFluor, it is proposed that at low pH a trans iminium is present, such that the carboxylate of L117D can interact with the imine to localize charge. However, upon acidification, a cis iminium forms, in which L117D does not interact with the 275 iminium. This would lead to the observed bathochromic shift in absorption wavelength with acidification. This posited mechanism may also explain the difference in extinction coefficients of the iminium at low and high pH. It is possible that the cis or trans iminium leads to a higher extinction. However, evidence for these hypotheses has not been obtained. 2.8 3.9 4.4 4.7 4.9 5.2 5.6 0.3 Abs. (a. u.) 0.25 6.0 6.4 6.7 7.1 7.8 8.9 10.3 10.8 11.1 11.4 11.7 12.0 b. 0.25 0.2 623 nm 529 nm 0.15 0.1 0.05 ΔAbs at 623 nm (a. u.) a. 0.1 0.05 350 400 450 500 550 600 650 700 Wavelength (nm) 0.1 2 d. 0.08 3 20 4 5 6 pH 7 8 9 10 594/633 543/633 514/633 15 /λ em2 0.06 0.04 em1 pka = 10.9 R2 = 0.94 Working range: pH 5.7 - 7.7 10 λ ΔAbs at 529 nm (a. u.) pka = 5.7 R2 = 0.99 0.15 0 0 c. 0.2 5 0.02 0 0 6 7 8 9 10 pH 11 12 5 13 6 7 8 9 10 pH Figure V-14. In vitro characterization of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F/ThioFluor as a single protein ratiometric pH sensor. a. UV-Vis titration of the hCRBPII/ThioFluor complex. b. pKa curve obtained by plotting the absorbance intensity at 623 nm as a function of pH. c. pKa curve obtained by plotting the absorbance intensity at 539 nm as a function of pH. d. Standard curves generated by plotting the emission ratio of 514/633 nm, 543/633 nm and 594/633 nm versus pH. 276 The working range of the Q108K:K40L:T51V:T53S:R58W:Y19W:L117D was too high, so we chose to introduce a Q4F mutation, in hopes that the pKa of the aspartate side chain would decrease. To our delight, the addition of Q4F to Q108K:K40L:T51V:T53S:R58W:Y19W:L117D results in one unit decrease in the pKa of the aspartic acid side chain (from 6.7 to 5.7), while having no effect on any of the other spectroscopic properties (Figure V-14). Standard curves plotted at different wavelengths are shown in Figure V-14d. Excitation wavelengths of 514/633 nm and 543/633 nm show high emission ratios. From the standard curve generated with excitation wavelengths of 514 nm and 633 nm, this hCRBPII/ThioFluor pair is shown to be useful at physiological pH with a working range of 5.7 to 7.7. L117D R58W Figure V-15. Overlay of the crystal structures Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor (magenta) and Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F/ThioFluor (cyan). Mr. Alireza Ghanbarpour was successful in obtaining the crystal structure of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F with ThioFluor at 1.6 Å resolution. An overlay Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F/ThioFluor of and Q108K:K40L:T51V:T53S:R58W:Y19W:L117D/ThioFluor indicates that with the 277 addition of Q4F a trans iminium is observed (Figure V-15). Neighboring residues, including L117D and R58W do not move. It is the closer distance between L117D and the iminium hydrogen atom in the trans isomer that presumably leads to the decreased pKa. Q128 2.6 Å L117D 2.6 Å L117E 2.9 Å Figure V-16. Overlay of crystal structures Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (green) Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F/ThioFluor (cyan). It is most interesting to compare the structures Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F/ThioFluor of and of with Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor (Figure V-16), as the chromophore and active site lysine adopt the same orientation, both displaying a trans iminium. The major difference between the structures is that L117E interacts with both the iminium and Q128, with distances of 2.9 Å and 2.6 Å, respectively. However, L117D, which is one methylene shorter, only forms a hydrogen bond with Q128 (2.6 Å). Upon titration of L117E in Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor, a shift of 56 nm is observed between the two iminium maxima. A larger shift of 94 nm is observed in 278 Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F/ThioFluor. The respective absorbance maxima of the red-shifted iminiums are 614 nm and 623 nm. The insignificant change in absorption maxima is presumably due to the unchanging electrostatic potential along the chromophore when L117D and L117E are protonated; in other words, the polarity is essentially the same. These crystals were obtained at pH 4.0 – 4.5, which yields no information about the state of the iminium with the carboxylate at position 117 (at more basic pH). Crystallization at a more basic pH, in which the imine is protonated and acidic side chain is deprotonated may provide insight into the reason why L117D mutants are more blue-shifted than with L117E, as well as the reason for the significant increase in extinction coefficient observed upon protonation of the acidic side chain. Lastly, final efforts were made to optimize Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F. We chose to test the effect of an R58H substitution, as it was already shown to enhance the rate of iminium formation. It should be noted that we first tried the mutants Q108K:K40L:T51V:T53S:R58H:Y19W:L117D and Q108K:K40L:T51V:T53S:R58H:Y19W:L117D:A33W, but no soluble protein was obtained from the expressions. Fortunately, expression of Q108K:K40L:T51V:T53S:R58H:Y19W:L117D:Q4F was fruitful with an expression yield of 21 mg/mL. In this hCRBPII mutant, the titration of L117D in absorption is hard to observe, due to its extremely low absorbance intensity. Starting from pH 279 7.2, the Q108K:K40L:T51V:T53S:R58H:Y19W:L117D:Q4F/ThioFluor complex was first acidified (Figure V-17a). Upon acidification, the SB at ~400 nm decreases, while a peak at 617 nm forms. The pKa for this species is 5.8. Presumably, this corresponds to the pKa of the aspartic acid. Additionally, it appears that the small peak at ~500 nm decreases with acidification. a. 0.3 4.3 4.8 5.2 0.25 7.2 ΔAbs at 617 nm (a. u.) Abs. (a. u.) 0.25 5.7 6.2 6.6 0.2 0.15 0.1 0.05 0.2 0.1 0.05 0 0 350 400 450 500 550 600 650 700 Wavelength (nm) 0.3 7.2 7.5 7.9 Abs. (a. u.) 0.25 8.7 9.5 10.2 4 0.2 0.15 0.1 0.05 5 6 pH 7 8 0.1 11.1 11.7 ΔAbs at 500 nm (a. u.) b. pKa = 5.8 R2 = 0.99 0.15 0.08 0.06 pKa = 10.9 R2 = 0.94 0.04 0.02 0 0 7 350 400 450 500 550 600 650 700 Wavelength (nm) 8 9 10 11 12 pH Figure V-17. a. Acidification of Q108K:K40L:T51V:T53S:R58H:Y19W:L117D:Q4F/ThioFluor complex. b. Basification of Q108K:K40L:T51V:T53S:R58H:Y19W:L117D:Q4F/ThioFluor complex. On the other hand, basification of the Q108K:K40L:T51V:T53S:R58H:Y19W:L117D:Q4F/ThioFluor complex from pH 280 7.2 (Figure V-17b) leads to a decrease in the peak at ~500 nm and an increase in the SB at ~400 nm. It would stand to reason that this is the peak corresponding to the iminium when the aspartic acid is deprotonated (pKa = 10.9). However, it is predicted that a mixture of imine isomers is present since the SB can be converted to both the 617 nm and ~500 nm peak. It should also be noted that the blue-shifted iminium is almost unobservable due to its low absorbance intensity, presumably due to its extremely low extinction coefficient, which is going to limit its use as a single protein ratiometric pH sensor. 0.35 Abs. (a. u.) 0.3 0.25 7.9 8.5 9.4 10.3 10.9 11.4 12.1 0.4 ΔAbs at 649 nm (a. u.) 3.3 4.1 4.6 5.0 5.6 6.1 6.7 7.3 0.4 0.2 0.15 0.1 0.3 pKa = 5.7 R2 = 0.99 0.2 0.1 0.05 0 0 3 350 400 450 500 550 600 650 700 750 Wavelength (nm) 4 5 6 7 pH 8 9 10 11 Figure V-18. Titration of Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:A33W:Q4F/ThioFluor complex and pKa obtained by plotting absorbance intensity versus pH at 649 nm. The last hCRBPII mutant investigated was Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:A33W:Q4F because the mutant with A33W lead to the largest shift in absorption maxima upon titration of the aspartic acid side chain, but the pKa led to a working range too high for use at physiological pH. Q4F was included to lower the pKa of the side chain. This hCRBPII mutants behaves similar to the previous mutant; a blue-shifted iminium 281 at ~500 nm exists at pH 7.3, while a red-shifted peak with absorption maximum of 649 nm (pKa = 5.7) is formed upon acidification (Figure V-18). The almost unobservable blue-shifted iminium renders this hCRBPII/ThioFluor complex also inadequate for application as a single protein ratiometric sensor. V.4 Imaging of hCRBPII mutants displaying the pH induced wavelength shift with ThioFluor We chose to test two hCRBPII mutants for their applicability as a single fluorescent protein ratiometric sensor. Q108K:K40L:T51V:T53S:R58W: Y19W:L117E was first chosen as ThioFluor has also shown success in labeling. It also shows an absorption wavelength shift with pH (56 nm) and an emission ratio of 7 when excitation wavelengths of 514 nm and 633 nm are used. HeLa cells expressing EGFP-Q108K:K40L:T51V:T53S:R58W:Y19W:L117E-3NLS were incubated with 10 μM ThioFluor for one hour at 37 °C. Cells were then washed twice with PBS buffer and subsequently imaged (Figure V-19). Green fluorescence was observed when the cells excited at 488 nm (BP 505 – 530 nm), indicating that transfection was successful and the fusion protein is being expressed in the nucleus. Imaging wavelengths of 514 nm, 543 nm, 594 nm and 633 nm were tested for their ability to fluoresce the hCRBPII/ThioFluor complex. The gain and offset settings were optimized for the 594 nm excitation with a long pass (LP) 615 nm emission filter. As seen, there is minimal background fluorescence under these confocal settings. However, significant background fluorescence is observed at the lower excitation wavelengths of 514 282 nm and 543 nm, with both LP 615 nm and 650 nm emission filters. This is presumably due to nonspecific iminium formation or free aldehyde. Unfortunately, background fluorescence is also observed at the most red-shifted excitation wavelength of 633 nm. It only makes sense that this comes from two-photon excitation (~315 nm) of nonspecific imine formation. 514 nm LP 615 nm 543 nm LP 615 nm 594 nm LP 615 nm 488 nm BP 505-530 nm + DIC 514 nm LP 650 nm 543 nm LP 650 nm 594 nm LP 650 nm 633 nm LP 650 nm Figure V-19. Imaging of HeLa cells expressing EGFPQ108K:K40L:T51V:T53S:R58W:Y19W:L117E-3NLS, labeled with ThioFluor for one hour at 37 °C. 5 4 3 2 6 1 11 10 514 nm LP 650 nm 7 8 9 633 nm LP 650 nm Figure V-20. Optimized imaging of HeLa cells expressing EGFPQ108K:K40L:T51V:T53S:R58W:Y19W:L117E-3NLS upon excitation at 514 nm and 633 nm. 283 Nonetheless, it should be noted that background fluorescence observed upon excitation with 633 nm could be minimized upon optimization of the gain and offset settings in this channel (Figure V-20). Under the same settings, the background fluorescence with 514 nm is also drastically reduced. In the optimized 633 nm excitation channel, the nucleus of each cell was selected (cells are numbered in Figure V-20). From the confocal software a histogram of fluorescence intensity was then output. The sum of the area under the curve was then computed for each individual cell. In order to ensure the exact same area is being analyzed when comparing the excitation of 514 nm and 633 nm, the same nucleus selections were superimposed on the imaged obtained from 514 nm. Again, the area under the curve of the generated histogram was determined. Lastly, the ratio of fluorescence intensity at 514 nm to 633 nm was determined for each cell. It should be noted that two cells did not provide reliable data, as their fluorescence intensities were outside of the threshold limits (cells 3 and 11 are too bright). This data is summarized in Table V-4. Table V-4. Ratio of emissions obtained upon excitation at 514 nm and 633 nm for EGFP-Q108K:K40L:T51V:T53S:R58W:Y19W:L117E-3NLS/ThioFluor. λem,514nm/ λem,514nm/ Cell Cell λem,633nm λem,633nm 1 1.60 7 1.60 2 1.37 8 1.38 4 1.38 9 1.39 5 1.61 10 1.42 6 1.65 Note: Cells 3 and 11 omitted were removed because their fluorescence intensities were outside of the threshold limits. 284 Based on the data collected from nine cells, the average emission ratio of 514 nm to 633 nm is 1.49 with a standard deviation of 0.12. The in vitro standard curve (Figure V-6) was previously obtained based on plotting the ratio of emissions at 514 nm and 633 nm versus pH. To make this plot useful, we fit a linear line to the linear portion of the standard curve (6.5 – 8.5). This plot is shown in Figure V-21. Substitution of the observed 1.49 ± 0.12 emission ratio average over the nine cells yields an in vivo pH of 6.65 ± 0.043 for the nucleus of these HeLa cells. y = -17.586 + 2.8696x R2= 0.99793 8 λ em,514nm /λ em,633nm 7 6 5 4 3 2 1 0 5 6 7 8 9 10 pH Figure V-21. Linear fit to the working range of the in vitro generated standard curve for Q108K:K40L:T51V:T53S:R58W:Y19W:L117E/ThioFluor. Intracellular pH is a fundamental modulator of cell function. Abnormal pH values are associated with some common diseases such as cancer and Alzheimer’s. Thus, the ability to measure intracellular pH is needed in order to study the physiological processes of organelles. To this end, research has focused on the development of intracellular pH sensors based on small fluorescent organic molecules, nanoparticles and fluorescent proteins.6 More specifically, green fluorescent protein (GFP) variants were used to measure 285 cytosolic pH of HeLa cells (7.4).7, 8 More recently, nano-pH probes were developed to measure the pH levels of individual cells. Pourmand and coworkers measured single-cell intracellular pH measurements using non-cancerous and cancerous cell lines, including human fibroblasts, HeLa, MDA-MB-231 and MCF7. The average intracellular pH levels were 7.4, 6.8, 6.9 and 6.9 for human fibroblast, HeLa, MCF-7 and MDA-MB-231, respectively.9 While most intracellular pH sensors do not differentiate between nuclear and cytoplasmic pH, small molecules have been targeted for this purpose. U2OS cells expressing HaloTag3NLS showed a pH of 7.4, while cytosol targeting indicated a pH of 7.1.10 From these data it is apparent that there is a gap in the field of intracellular pH measurement to which we can contribute. As shown, in vivo pH can be measured in HeLa cells expressing EGFPQ108K:K40L:T51V:T53S:R58W:Y19W:L117E-3NLS, labeled with ThioFluor. The only disadvantage with this mutant is that high background fluorescence is observed, presumably due to unbound chromophore or nonspecific imine formation. Additionally, the pH is on the edge of the working range for this mutant (6.5 – 8.5). To this end, we chose to test the utility of the fastest binding mutant developed in this study. Advantageously, it also has a lower working range of 5.7 – 7.7, which should allow for more accurate estimation of the in vivo pH. Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F was cloned into the pFlag-CMV2 vector, retaining the nuclear localization sequence. It should be noted that EGFP was not retained, in order to eliminate any crosstalk that may 286 arise from undesired EGFP excitation at 514 nm or 543 nm (even though the emission should be filtered with both LP 615 nm and LP 650 nm emission filters). 514 nm LP 650 nm 633 nm LP 650 nm Figure V-22. Imaging of HeLa cells expressing Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F-3NLS, labeled with ThioFluor for ten minutes at 37 °C. HeLa cells expressing Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F3NLS were incubated with 10 μM ThioFluor for ten minutes at 37 °C. Images were collected with excitation wavelengths of 514 nm and 633 nm, with LP 650 nm emission filter. The gain and offset settings were optimized for the 633 nm excitation. Background fluorescence is negligible under excitation with 633 nm (Figure V-22), but unfortunately significant background fluorescence is again observed upon 514 nm excitation. The background fluorescence is even brighter than the fluorescence from hCRBPII/ThioFluor in all cells. As compared to HeLa cells expressing Q108K:K40L:T51V:T53S:R58W:Y19W:L117E, the background fluorescence is brighter at 514 nm; in vitro data shows that the extinction coefficient is decreased by about half (~8,000 M-1cm-1 for Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F and ~15,000 M-1cm-1 for 287 Q108K:K40L:T51V:T53S:R58W:Y19W:L117E) which leads to a reduction in brightness. It should be noted that expression of this hCRBPII mutant leads to fluorescence from the hCRBPII/ThioFluor complex that seems to be ‘leaky’ as it is not fully localized to the nucleus. It is hypothesized that this is due to overexpression of the mutant in vivo (in vitro protein expression yield is 23 mg/mL). If this mutant it to be used for live cell imaging, it will be necessary to manipulate the in vivo protein expression, possibly by imaging earlier after transfection. V.5 ThioFluor analogues also show a pH induced wavelength shift While most of the ThioFluor analogues developed were not fluorescent (Chapter III), we sought to investigate whether they also showed a pH dependent wavelength shift of the iminium due to titration of the glutamic acid side chain in Q108K:K40L:T51V:T53S:R58W:Y19W:L117E. Results are summarized in Table V-5. As expected, all of the chromophores show a bathochromic shift in wavelength upon putative protonation of the glutamic acid side chain, except ThioFluor-2, which was already shown not to be able to regulate wavelength. Small shifts are observed with weakly electron donating groups at the para position of the phenyl ring (ThioFluor-6 = H, ThioFluor-4 = methoxy). These chromophores also show a lower pKa of the glutamic acid side chain. On the other hand, the more conjugated ThioFluor analogues (ThioFluor, ThioFluor-3, 288 ThioFluor-8) showed a larger difference between the iminium maxima and also an increased pKa of the glutamic acid side chain at residue 117. Presumably the donation along the chromophore leads to the high pKa in the other chromophores. Table V-5. pH induced wavelength shift observed in Q108K:K40L:T51V:T53S: R58W:Y19W:L117E with ThioFluor analogues. t1/2 Ligand λabsab λabsb ΔAbs pKad pKae Φf (min)g ThioFluor 558 614 56 11.2 6.6 0.31 82 ThioFluor-2 515 - - 11.3 - 0.13 1124 ThioFluor-3 548 611 63 10.9 6.6 0.27 27 ThioFluor-4 466 493 27 10.9 5.7 - 17 ThioFluor-5 423 nd nd 10.9 nd - 9 ThioFluor-6 424 458 34 10.6 5.4 - 3 ThioFluor-8 641 687 46 11.2 6.8 0.09 352 ThioFluor-9 562 624 62 11.1 6.6 0.02 13 ThioFluor-10 489 545 56 11.3 6.1 0.06 92 a b KLVSWW = Q108K:K40L:T51V:T53S:R58W:Y19W. λmax of absorbance at basic pH. cλmax of absorbance at acidic pH. dpKa corresponding to the iminium. e pKa corresponding to Glu117/Asp117. fQuantum yield measured at pH ~7.2. g Kinetic measurements were performed at 23 °C with 20 μM protein and 0.5 equiv ligand. PSB formation was monitored by UV-Vis at λmax for each complex over time at pH ~7.2. Data was fit to second order kinetics. V.6 Conclusions We have shown that the titration of an L117D or L117E mutation in hCRBPII mutants leads to a pH induced absorption wavelength shift of the iminium. Shifts up to 135 nm were observed, with quantum yields greater than 30%, and half-lives as fast as 8 minutes. Advantageously, the hCRBPII/ThioFluor system can successfully be used to image live cells. To our 289 delight, based on the ratio of emissions at two excitation wavelengths, the average over nine cells yields an in vivo pH of 6.65 ± 0.043 for the nucleus of HeLa cells. The only pitfall is that background fluorescence is observed. To address this correctly, it is suggested that a fluorogenic analogue of ThioFluor be developed, such that it is only fluorescent when bound as the iminium to hCRBPII. It would also be beneficial to test the affect of ThioFluor concentration. A decrease in ThioFluor concentration would be expected to decrease the background fluorescence, if it is indeed due to unbound aldehyde. 290 REFERENCES 291 REFERENCES 1. Berbasova, T., Nosrati, M., Vasileiou, C., Wang, W.J., Sing, K., Lee, S., Yapici, I., Geiger, J.H. & Borhan, B. Rational Design of a Colorimetric pH Sensor from a Soluble Retinoic Acid Chaperone. Journal of the American Chemical Society 135, 16111-16119 (2013). 2. Berbasova, T. in PhD Thesis, Michigan State University (2014). 3. Wang, W. in PhD Thesis, Michigan State University (2012). 4. Wang, W., Nossoni, Z., Berbasova, T., Watson, C.T., Yapici, I., Lee, K.S., Vasileiou, C., Geiger, J.H. & Borhan, B. Tuning the electronic absorption of protein-embedded all-trans-retinal. Science 338, 1340-1343 (2012). 5. Casey, J.R., Grinstein, S. & Orlowski, J. Sensors and regulators of intracellular pH. Nature Reviews - Molecular Cell Biology 11, 50-61 (2010). 6. Han, J.Y. & Burgess, K. Fluorescent Indicators for Intracellular pH. Chemical Reviews 110, 2709-2728 (2010). 7. Llopis, J., McCaffery, J.M., Miyawaki, A., Farquhar, M.G. & Tsien, R.Y. Measurement of cytosolic, mitochondrial, and Golgi pH in single living cells with green fluorescent proteins. Proceedings of the National Academy of Sciences of the United States of America 95, 6803-6808 (1998). 8. Kneen, M., Farinas, J., Li, Y.X. & Verkman, A.S. Green fluorescent protein as a noninvasive intracellular pH indicator. Biophysical Journal 74, 15911599 (1998). 9. Ozel, R.E., Lohith, A., Mak, W.H. & Pourmand, N. Single-cell intracellular nano-pH probes. RSC Advances 5, 52436-52443 (2015). 10. Benink, H.A., McDougall, M.G., Klaubert, D.H. & Los, G.V. Direct pH measurements by using subcellular targeting of 5(and 6-) carboxyseminaphthorhodafluor in mammalian cells. Biotechniques 47, 769-774 (2009). 292 CHAPTER VI: PROMOTING MONOMERIZATION OF hCRBPII MUTANTS Our lab has observed the formation of domain-swapped dimers during protein bacterial overexpression. Characterization of domain-swapped dimers from hCRBPII-Y60 mutants has been provided by both size-exclusion chromatography and crystallography. It was hypothesized that the combination of two open monomer intermediates leads to the formation of the domain-swapped dimer (Figure VI-1).1 monomer open monomer intermediate domain-swapped dimer Figure VI-1. Formation of domain-swapped dimer in hCRBPIIQ108K:K40L:Y60W. Coordinates obtained from PDB 4ZR2 (dimer) and 4ZJ0 (monomer). In the course of engineering hCRBPII to bind ThioFluor, it was observed that certain residues are more prone to foster hCRBPII dimerization. For example, T51 was shown to form a significant amount of dimer when mutated to alanine (Table VI-1), albeit in the presence of other mutations. Table VI-1. Effect of T51 mutation in KL:T51X:T53S:R58W:Y19W:F16Y. hCRBPII mutant Major Fraction Q10K:K40L:T51V:T53S:R58W:Y19W:F16Y Monomer Q10K:K40L:T51C:T53S:R58W:Y19W:F16Y Monomer Q10K:K40L:T51A:T53S:R58W:Y19W:F16Y Dimer Other mutations shown to affect the monomer/dimer ratio formed during bacterial expression were K40, T51, T53, V62, and R58, among others. A few 293 select examples are shown in Table V1-2. It should be noted that Q108K:K40L produces only monomer during expression. Single point mutations on wild type hCRBPII were also shown to promote dimerization, including I25E (27.5 mg/L monomer, 16.9 mg/L dimer; 77 mol% monomer), Y60F (40.6 mg/L monomer, 2.9 mg/L dimer; 97 mol% monomer), L77D (31.3 mg/L monomer, 19.3 mg/L dimer; 76 mol% monomer), L77E (45.7 mg/L monomer, 7.6 mg/L dimer; 92 mol% monomer), and Q128L (6.0 mg/L monomer, 5.6 mg/L dimer; 68 mol% monomer). Table VI-2. hCRBPII dimerization promoted by T53S, V62N, T51A and T51N mutations. mol % mg/L mg/L hCRBPII mutant monomer monomer dimer Q108K:K40E 31 3.3 14.6 Q108K:K40D 33 2.5 10.1 Q108K:K40D:Q4F > 95 7.0 - Q108K:K40L:T53A 63 5.5 6.6 Q108K:K40L:T53S 85 11.4 3.9 Q108K:K40L:T53V 84 10.6 4.0 Q108K:K40L:V62N 76 14.5 9.4 Q108K:K40L:T53S:V62N 69 8.0 7.1 Q108K:K40L:T53S:V62E 34 5.5 21.6 Q108K:K40L:T51A:T53S:V62N <5 - 5.8 Q108K:K40L:T51N:T53S:V62N 22 1.7 12.4 We previously showed that hCRBPII variants prone to oligomerization are not good candidates for fluorescent labeling in live cells (see Chapter II.11.1 for data and discussion). For this reason, we wanted to determine if we could rationally introduce point mutations to a sequence of interest, in order to promote the formation of hCRBPII monomer over dimer. We chose to start by exploring 294 the affect of T51 on the formation of domain-swapped dimers, as this position was shown to drastically alter the ratio of monomer and dimer formed during bacterial protein overexpression (Table VI-1). Because we have also observed a difference between hCRBPII monomer and dimer binding to ThioFluor (Chapter II.8.1), we also sought to investigate the spectroscopic properties of these mutants when bound to retinal. VI.1 Effect of T51 mutations on the monomer/dimer ratio during bacterial overexpression in hCRBPII mutants In order to study the effect of T51 mutations (Figure VI-2) on the formation of dimer-swapped dimers in hCRBPII, we started from the parent template Q108K:K40L. This was chosen as the template for mutation for two reasons. Firstly, the inclusion of K40L was shown by Dr. Wenjing Wang to increase the iminium pKa due to the removal of basic charge.2 Additionally, we wanted to prevent the binding of retinal at the secondary K40 site. Dr. Meisam Nosrati previously showed occupancy of retinal at K40 in the dimer Q108K:T51D.3 4.9 Å Figure VI-2. Location of T51 in Q108K:K40L/retinal. Coordinates were obtained from PDB 4EXZ. 295 To this end, all twenty amino acids were introduced at residue 51 in Q108K:K40L. Fortunately, all substitutions except T51D, T51G, T51P and T51R led to soluble protein during expression at room temperature in terrific broth (TB). Isolated protein yields are shown in Table VI-3, in increasing order of monomer. If only one oligomeric state of hCRBPII was isolated after size exclusion chromatography, mol percent monomer is expressed as 99% upon isolation of monomer, or 1% upon isolation of dimer. Table VI-3. Effect of T51X mutations on monomer/dimer formation in Q108K:K40L. mol % mg/L mg/L hCRBPII mutant monomer monomer dimer Q108K:K40L:T51K 1 - 6.1 Q108K:K40L:T51A 14 1.1 13.0 Q108K:K40L:T51N 17 2.7 25.6 Q108K:K40L:T51W 25 1.0 5.9 Q108K:K40L:T51S 28 1.7 8.7 Q108K:K40L:T51H 58 4.6 6.8 Q108K:K40L:T51Q 65 10.2 11.2 Q108K:K40L:T51Y 86 7.9 2.6 Q108K:K40L:T51F 90 10.2 2.3 Q108K:K40L:T51M 92 15.1 2.5 Q108K:K40L:T51E 96 44.7 3.8 Q108K:K40L:T51L 99 57.6 0.7 Q108K:K40L:T51C 99 9.5 - Q108K:K40L:T51I 99 12.0 - Q108K:K40L:T51V 99 12.5 - Q108K:K40L 99 13.3 - 296 It was quickly realized that the mol percent monomer ratio cannot be correlated to any of the commonly used hydrophobicity scales, including KyteDoolittle,4 Hoops-Woods,5 Cornette,6 Eisenberg,7 Rose,8 Janin9 or Engelman.10 Additionally, there also is no correlation between residue size and the monomer/dimer ratio obtained during expression. Non polar side chains, including leucine, isoleucine and valine lead to monomer formation. However, alanine leads to majorly dimer formation. On the other hand, no general trend is observed for the polar side chains. Asparagine, glutamine and serine lead to substantial dimer formation, while threonine and cysteine lead to solely monomer formation. The aromatic residues of phenylalanine, tyrosine and tryptophan all produced both monomer and dimer, albeit at different monomer/dimer ratios. Charged side chains again presented contrasting results. Mutation to the basic lysine and histidine produced a large population of dimer. In fact T51K yields dimer only. Conversely, glutamic acid produces mostly monomer. VI.2 Spectroscopic properties of Q108K:K40L:T51X mutants Purified monomers and dimers were coupled with retinal, in order to see if any differences in spectroscopic properties, including absorption wavelength and half-life of binding, were different in the monomer and dimer of the same mutation (Table VI-4). All rates were measured via UV-Vis at 23 °C at approximately pH 7.2 (20 μM protein and 10 μM retinal, plotted according to pseudo-first order rate kinetics). It should be noted that half-lives less than two minutes should not be 297 seen as significantly different, as the initial rates may not be accurate due to the fast reaction rate. The most interesting mutants are highlighted, due to the different spectroscopic properties observed between the dimer and monomer. Table VI-4. Spectroscopic properties of Q108K:K40L:T51X mutants bound to retinal. λabs λabs t1/2 (min) t1/2 (min) hCRBPII mutant monomer dimer monomer dimer Q108K:K40L:T51A 519 520 1.9 2.0 Q108K:K40L:T51C 489 - 5.3 - Q108K:K40L:T51E 500 461 nd nd Q108K:K40L:T51F 529 538 108.3 6.3 Q108K:K40L:T51H 509 512 1.5 1.5 Q108K:K40L:T51I 544 - 12.9 - Q108K:K40L:T51K - 494 - nd Q108K:K40L:T51L 535 494/517 60.9 1.7 Q108K:K40L:T51M 536 nd 18.7 nd Q108K:K40L:T51N 482/493 489/494 1.1 0.9 Q108K:K40L:T51Q nd nd nd nd Q108K:K40L:T51S 503 503 1.2 ns Q108K:K40L 505 - 3.3 - Q108K:K40L:T51V 535 - 2.1 - Q108K:K40L:T51W 503 503 78.8 86.6 Q108K:K40L:T51Y nd nd nd nd ns = protein is not stable, and precipitates under the experimental conditions very quickly In general, the rate of iminium formation is fast in the hCRBPII variants. The exceptions are when phenylalanine, tryptophan or leucine are substituted at position 51. These three residues lead to slow iminium formation, presumably due to their larger size. However, in the dimer, only tryptophan leads to slow 298 iminium formation. It is quite interesting to note that T51F and T51L actually lead to substantially faster iminium formation in the dimer. Attempts to understand the difference between Q108K:K40L:T51F monomer and dimer were undertaken as it produced more dimer during bacterial expression than Q108K:K40L:T51L (Table VI-3). This will be discussed in Section VI.5. The other major noticeable difference between retinal bound monomer and dimer is seen in the absorption wavelengths of Q108K:K40L:T51E. The iminium in the monomer is almost 40 nm red-shifted of the dimer. Presumably, the glutamic acid adopts a different orientation in the two structures. However, crystal structures would be necessary to understand the difference. All other monomer and dimer pairs exhibit essentially the same absorption wavelengths and rates of iminium formation. VI.3 Imine isomerization of Q108K:K40L:T51X mutants As discussed above, the substitution at residue 51 can have a significant affect on the spectroscopic properties of bound retinal; additionally differences are apparent between the monomer and dimer. These include the absorption wavelength and rate of iminium formation. Because residue 51 was shown to drastically alter the rate of iminium formation, we sought to test whether it also affects the rate of imine isomerization. We again used Q108K:K40L as the parent mutant. The spectroscopic properties of Q108K:K40L-monomer/retinal are shown in Figure VI-4. Iminium formation is fast, exhibiting a half-life of 3.3 minutes. The 299 highest intensity of iminium is observed at 24 minutes, which then decays into SB (Figure VI-4a and Figure VI-4b). This is indicative of cis-trans isomerization of the imine, which we have previously observed between retinal and both hCRBPII and CRABPII mutants. (Figure VI-3).11, 12 briefly, the low pKa trans imine can be converted to the high pKa cis imine with UV light irradiation, which is subsequently protonated to yield the cis iminium. Thermal relaxation eventually produces the thermodynamic trans imine. H high pKa N Lys N Lys cis-iminium λmax > 500 nm kinetic product cis-imine λmax ~ 360 nm UV light N Lys heat or visible light low pKa trans-imine λmax ~ 360 nm thermodynamic product N H Lys trans-iminium λmax > 500 nm nm Figure VI-3. Cycle for imine isomerization in retinal bound hCRBPII and CRABPII variants. The half-life of the thermal isomerization observed in Q108K:K40Lmonomer/retinal is 110.0 minutes. Interestingly, upon UV irradiation of this thermodynamic state (365 nm TLC handlamp for one minute), only iminium is formed, evidenced by the lack of SB at approximately 360 nm (Figure VI-4c). The rate of isomerization after UV irradiation is almost identical to that observed thermally, with a half-life of 111.6 minutes (Figure VI-4d), indicating that the conversion is to the same species. 300 a. b. 24 min 1298 min 0.5 0.3 0.2 0.1 PSB (λ = 360 nm) max = 505 nm) 0.2 0 0 0.5 0.2 0.1 400 600 800 1000 1200 0.5 0.4 Abs. at 505 nm (a. u.) 0.3 200 Time (min) d. UV 1 min 898 min 0.4 Abs. (a. u.) 0.3 max 0.1 0 300 350 400 450 500 550 600 650 Wavelength (nm) c. SB (λ 0.4 Abs. (a. u.) Abs. (a. u.) 0.4 0.5 0.3 0.2 k = 0.00621 min-1 R2 = 0.997 t1/2 = 111.6 min 0.1 0 300 350 400 450 500 550 600 650 Wavelength (nm) 0 0 200 400 600 Time (min) 800 Figure VI-4. a. UV-Vis of Q108K:K40L-monomer/retinal. b. Absorbance intensity at SB and PSB as a function of time. c. UV irradiation and subsequent isomerization of the imine. d. Rate (pseudo-first order) of imine isomerization after UV irradiation. Substitution of valine, alanine and isoleucine also allow imine isomerization to occur. However, the monomers display drastically different halflives of thermal isomerization (Table VI-5). The isomerization of Q108K:K40L:T51I-monomer/retinal was extremely slow and was not complete within 800 minutes. On the other hand, it does not appear that imine isomerization occurs after iminium formation with substitutions of serine or cysteine (in the monomer or dimer). However, in the case of serine, the protein 301 complex does appear to precipitate after iminium formation. It can be concluded that residue 51 significantly affects the imine isomerization. Table VI-5. Thermal isomerization half-lives with substitution of small aliphatic residues at T51. t1/2 (min) hCRBPII mutant Monomer Dimer Q108K:K40L 110.0 - Q108K:K40L:T51A 36.5 36.5 Q108K:K40L:T51V 356.1 - VI.4 Thermal imine isomerization that leads to a different absorption wavelength of the iminium From the data we have collected, the absorption of the PSB remains consistent, whether formed immediately with retinal or after UV irradiation. However, here we show that with a T51N mutation, two absorption wavelengths of PSB are observed. Spectroscopic properties of Q108K:K40L:T51N- monomer/retinal are shown in Figure VI-5. Iminium formation is fast, exhibiting a half-life of 1.1 minutes and absorption wavelength of 482 nm. The highest intensity of iminium is observed at 10 minutes, which then decays into SB (Figure VI-5a). Again, this is indicative of imine isomerization. However, in this case, a bathochromic shift in the wavelength to 493 nm is also observed. This is the first case, in which a shift in the wavelength is observed, presumably of the cis and trans iminium or different conformations of active site residues. The halflife of this thermal isomerization is 43.5 minutes. From the plot of the intensity at 280 nm (corresponding to the protein absorbance), it is quite obvious that after 302 the thermodynamic product is formed, the complex is not stable and the protein unfolds over time (Figure VI-5b). It does not appear to precipitate, because the baseline at 750 nm does not change significantly. a. Abs. (a. u.) 0.4 482 nm 10 min 244 min 0.3 493 nm 0.2 0.1 0 300 350 400 450 500 550 600 650 Wavelength (nm) Normalized Abs. Intensity b. Protein (280 nm) SB (360 nm) PSB (493 nm) Baseline (750 nm) 0 0 200 400 600 800 1000 1200 Time (min) Figure VI-5. a. UV-Vis of Q108K:K40L:T51N-monomer/retinal. b. Normalized absorption intensity versus time indicates that protein unfolding occurs after iminium formation. Q108K:K40L:T51N-dimer behaves the same as the monomer when bound to retinal. In this case, the absorption wavelength of the kinetically formed iminium is 489 nm. Subsequent thermal decay then yields an iminium with an absorption maximum of 494 nm. We were not able to obtain the crystal structure of Q108K:K40L:T51N-monomer/retinal, however, Mr. Alireza Ghanbarpour was successful in obtaining the crystal structure of Q108K:K40L:T51N-dimer/retinal. There were 12 molecules in the asymmetric unit, of which sufficient retinal density was observed in three chains (Figure IV-6). The sigma level (from Pymol) is given for each chain. An overlay of retinal in the three chains shows that the chromophore is quite variable. This conformational change may account for the shift in absorption wavelength observed over time. 303 0.5σ 1.0σ 1.0σ Figure VI-6. Crystal structure of Q108K:K40L:T51N-dimer/retinal. VI.5 Different rates of iminium formation between monomer and dimer with retinal As previously mentioned, the rate of iminium formation was drastically different in the monomer and dimer when phenylalanine and leucine were substituted at residue 51 (Table VI-4). The monomer of Q108K:K40L:T51L exhibited a half-life of 60.9 minutes, while the dimer was much faster with a halflife of 1.7 minutes. It should also be noted that the initial iminium of Q108K:K40L:T51L-dimer/retinal formed (presumably the kinetic product) had an absorption wavelength of 494 nm, while the thermodynamic iminium showed a maximum at 506 nm. This phenomenon was not observed in the monomer. Similarly, the half-life of iminium formation of Q108K:K40L:T51F-monomer and dimer were 108.3 minutes and 6.3 minutes respectively. UV-Vis data is 304 shown in Figure VI-7. It is interesting to note that not only is the rate of iminium formation faster in the dimer, but so is the thermal isomerization. Additionally, the pKas are drastically different between the monomer and dimer, as evidenced by the absorbance intensity of the iminium. Q108K:K40L:T51F-monomer/retinal 438 min 1498 min 0.4 0.3 Abs. (a. u.) Abs. (a. u.) 0.4 0.2 0.1 PSB (λ max = 529 nm) 0.3 0.2 SB (λ max = 360 nm) 0.1 0 300 350 400 450 500 550 600 650 Wavelength (nm) 0 0 500 1000 1500 Time (min) Q108K:K40L:T51F-dimer/retinal 0.5 42 min 898 min 0.5 SB (λ 0.4 Abs. (a. u.) Abs. (a. u.) 0.4 0.3 0.2 0.1 max = 360 nm) 0.3 0.2 PSB (λ max = 538 nm) 0.1 0 300 350 400 450 500 550 600 650 Wavelength (nm) Figure VI-7. UV-Vis of Q108K:K40L:T51F-dimer/retinal. 0 0 200 400 600 800 Time (min) Q108K:K40L:T51F-monomer/retinal and We wanted to see if structural data could provide insight into these very different rates of iminium formation (though structural data only provides the thermodynamically static picture), as it may yield hints on how to engineer faster 305 binders for fluorescence imaging. Mr. Alireza Ghanbarpour was successful in obtaining the crystal structures of apo-dimer (1.9 Å), holo-dimer (1.9 Å) and holomonomer (1.6 Å) of Q108K:K40L:T51F. In the crystal structure of Q108K:K50L:T51F-dimer/retinal there were four molecules in the asymmetric unit, of which sufficient retinal density was observed in two chains (Figure VI-8); the sigma level shown is 1.0. An overlay of retinal in the two chains shows that both retinal and the residue T51F exist in two conformations. Both cis and trans isomers of the iminium are shown, which would be expected based on the existence of both cis iminium and trans imine at pH 7 (Figure VI-7). Figure VI-8. Overlay of two chains in the crystal structure of Q108K:K40L:T51Fdimer/retinal. The crystal structure of Q108K:K40L:T51F-monomer/retinal showed four molecules in the asymmetric unit, with retinal occupancy in three molecules. In each, retinal adopts essentially the same conformation of the ligand. Again, both cis and trans isomers of the imine are observed (Figure VI-9). However, in this 306 case, T51F does not change conformation. Scrutiny of the crystal structure indicates that there are no significant differences of any active site residues. Figure VI-9. Overlay of three Q108K:K40L:T51F-monomer/retinal. chains in the crystal structure of Both Q108K:K40L:T51F-monomer and dimer bind retinal as both cis and trans imine isomers. An overlay of the cis imine conformations of the monomer and dimer is shown in Figure VI-10, indicating that the trajectory of chromophore binding is quite different in the two cases. The major difference in the active-site residues is in the interaction of Y60 with E72. Y60 T51F E72 Figure VI-10. Overlay of the crystal structures of Q108K:K40L:T51Fmonomer/retinal (green) and Q108K:K40L:T51F-dimer/retinal (purple). 307 In the domain-swapped dimer crystal structure, Y60 hydrogen bonds with E72 (2.2 Å). This hydrogen bond is not seen in the monomer structure because the residues are not close enough; this hydrogen bond presumably stabilizes the domain-swapped dimer when retinal is bound. Interestingly, in the apo Q108K:K40L:T51F-dimer structure, Y60 is flipped out of the binding cavity (Figure V-11), while N59 is flipped in. Careful inspection of the crystal structures does not reveal any other significant changes of the side chains in the active site. Binding of retinal requires N59 to flip out of the active site, or else it would sterically hinder the binding of retinal. As a result of this conformational change, Y60 flips in to hydrogen bond with E72 in the holo domain-swapped dimer. This hydrogen bond may prevent Y60 from impeding on imine formation in the case of the domain-swapped dimer. Conversely, the difference in rates of iminium formation may simply be due to the conformation of Phe51. Figure VI-11. Overlay of the crystal structures of Q108K:K40L:T51F-dimer/retinal (purple) and apo Q108K:K40L:T51F-dimer (green). Careful analysis of the UV-Vis binding data of the monomer and dimer with retinal indicate that it is binding of retinal that is the slow process in iminium 308 formation in the monomer. Free retinal absorbs at 380 nm, and upon binding to hCRBPII exhibits fine vibronic structure. This fine vibronic structure forms within two minutes in the case of Q108K:K40L:T51F-dimer/retinal (Figure VI-12). However, in the case of the monomer, vibronic structure at the SB is not apparent in such a time frame. Vibronic structure is only weakly apparent after 40 minutes (Figure VI-12). This may indicate that it is the step of retinal binding to the monomer that is slow. It would be interesting to measure the binding affinity of retinal to Q108K:K40L:T51F monomer and dimer to see if this is the reason, or if it is just a matter of the dimer binding pocket being more accessible. Q108K:K40L:T51F-monomer/retinal 20 min scans 0.4 0.4 0.3 0.3 Abs. (a. u.) Abs. (a. u.) Q108K:K40L:T51F-dimer/retinal 2 min scans 0.2 0.1 0.2 0.1 0 0 350 400 450 500 550 Wavelength (nm) 600 650 350 400 450 500 550 Wavelength (nm) 600 650 Figure VI-12. UV-Vis scanning of Q108K:K40L:T51F dimer and monomer over time with retinal. VI.6 Mutation of Y60 in the triple mutant Q108K:K40L:T51F As we have predicted that the Y60-E72 hydrogen bond is critical in minimizing the formation of the domain-swapped dimer and may play some role in retinal binding, we wanted to investigate the properties observed upon mutation of Y60. To this end, we mutated Y60 to alanine, phenylalanine, leucine, 309 histidine and tryptophan and E72 to leucine and alanine. Unfortunately, in the context of Q108K:K40L:T51F, E72A and E72L mutations did not provide soluble protein. Expression yields for the Y60 mutants are shown in Table VI-6. Table VI-6. Effect of Y60 mutation in Q108K:K40L:T51F on protein expression yields and rate of iminium formation. mol % mg/L mg/L t1/2 (min) t1/2 (min) hCRBPII mutant monomer monomer dimer monomer dimer Q108K:K40L:T51F 82 10.2 2.3 108.3 6.3 Q108K:K40L:T51F:Y60A <5 - 12.9 76.8 40.8 Q108K:K40L:T51F:Y60F 40 5.3 7.8 73.7 17.8 Q108K:K40L:T51F:Y60L <5 - 5.9 - 24.8 Q108K:K40L:T51F:Y60H <5 - 6.6 - 23.1 Q108K:K40L:T51F:Y60W 71 8.2 3.4 94.9 13.1 As previously shown, Y60 mutation led to the formation of more dimer.1 Expression at room temperature led to only dimer formation in the case of Q108K:K40L:T51F:Y60A, Q108K:K40L:T51F:Y60L and Q108K:K40L:T51F:Y60H. However, expression at lower temperature can yield monomer (Mr. Alireza Ghanbarpour provided Q108K:K40L:T51F:Y60A monomer form an 19 °C expression). Interestingly, only mutation to Y60A, in the context of Q108K:K40L:T51F led to significantly slower iminium formation in the dimer. The plot of absorbance intensity at the PSB and SB (Figure VI-13) is quite similar to that observed for Q108K:K40L:T51F-monomer/retinal (Figure VI-7), including the absorbance intensity at the PSB (indicating a similar pKa to the monomer and higher than the dimer), and a slow rate of thermal isomerization. From this data, it is inferred that 310 the binding of retinal to Q108K:K40L:T51F:Y60A-dimer is similar to Q108K:K40L:T51F-monomer. Q108K:K40L:T51F:Y60A-monomer/retinal Q108K:K40L:T51F:Y60A-dimer/retinal 0.4 0.4 PSB (λ 0.3 max = 505 nm) Abs. (a. u.) Abs. (a. u.) PSB (λ 0.2 SB (λ max = 360 nm) 0.1 max = 528 nm) 0.3 0.2 SB (λ max = 360 nm) 0.1 0 0 0 200 400 600 800 1000 1200 1400 0 200 Time (min) 400 600 800 1000 Time (min) Figure VI-13. Absorbance intensity of PSB and SB as a function of time, followed by UV-Vis. All of the holo domain-swapped dimer structures collected so far show an asymmetric orientation of the two domains, due to the flipped in orientation of N59 and flipping out of Y60. From the crystal structure of Q108K:K40L:T51F:Y60A-dimer/retinal, for the first time in a domain-swapped dimer structure, N59 is flipped into the binding pocket (Figure VI-14). This allows for a symmetric domain-swapped dimer as seen for most of the apo-dimer structures. Additionally, the ligand density is disordered; electron density is apparent for only the first four carbons of the bound retinylidene, suggesting that the ligand is not static inside of the protein cavity. No concrete conclusions can be drawn as to the difference in the rates of iminium formation between the monomer and dimer. However, it is hypothesized that the slow step is in the rate of imine formation, presumably due to a sterically 311 hindered active site. Nonetheless, the most interesting observation made in this study is the mutation of Y60A, which leads to the first holo symmetric domainswapped dimer, in which N59 adopts a flipped in conformation. Y60A T51F N59 Figure VI-14. Crystal structure of Q108K:K40L:T51F:Y60A-dimer/retinal. VI.7 Rational mutation to promote monomerization in hCRBPII mutants As described in Chapter 4, we Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E, designed which a hCRBPII displayed mutant, desirable spectroscopic properties when bound to ThioFluor as a large Stokes shift fluorescent protein (Figure IV-6). However, live cell imaging was not successful with this mutant. We hypothesized that it was due to the propensity of the mutant to oligomerize; this mutant produces only dimer under numerous protein expression conditions. To this end, we wanted to see if we could rationally monomerize the heptamutant to deem it suitable for in vivo experiments. As seen in Table VI-3, Q108K:K40L:T51V does not lead to any dimer expression at room 312 temperature. We hypothesized that T51V should be able to “monomerize” our dimer-forming mutant. Therefore, we introduced Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E hCRBPII T51V mutant to in the hopes of producing monomeric hCRBPII. To our delight, the addition of T51V does allow for a small amount of monomer formation (15 mol percent). Moreover, removal of T53A yields 72 mol percent monomer. Here we have shown that with the rational introduction of two mutations into Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E (which produces only dimer during protein expression), we can obtain a greater ratio of monomer than dimer. Table VI-7. Effect of T51 and T53 mutations on monomer/dimer ratio Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E. mol % mg/L mg/L hCRBPII mutant monomer monomer dimer Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E <5 - 12.0 Q108K:K40D:T51V:T53A:R58L:Q38F:Q4F:V62E 15 1.5 17.4 Q108K:K40D:T51V:R58L:Q38F:Q4F:V62E 72 9.8 7.5 VI.8 Dimer formation in CRABPII, another iLBP Because T51 mutations showed a propensity toward dimerization in hCRBPII, we wanted to see if the equivalent residue in another intracellular lipid binding protein (iLBP) displayed the same inclination. To this end, the equivalent residue I52 in cellular retinoic acid binding protein II (CRABPII) was mutated to alanine, asparagine, tryptophan and serine. In hindsight this was not a good position to test, as protein expression yields were low, presumably due to the fact that I52 is a conserved residue. 313 Nonetheless, only monomer was isolated under a number of different expression conditions for I52W, I52N and I52S. Fortuitously dimer was isolated from the expression of I52A under the following expression conditions: a single colony (from BL21(DE3) pLysS cells transformed with I52A DNA) was inoculated to 10 mL Luria broth (LB) with ampicillin (100 mg/L) and chloramphenicol (27 mg/L) and grown at 37 °C. After 16 hours, the solution was transferred to 1 L terrific broth (TB) with fresh ampicillin and chloramphenicol and grown for an additional 6 hours at 37 °C. Isopropyl-1-thio-D-galactopyranoside (IPTG) was added to the culture (at a final concentration 1 mM) to induce protein expression, and the culture was shaken at 18 °C for 29 hours. The size exclusion chromatography (SEC) trace (run at 1 mL/min with 4 mL fractions) for this purification is shown in Figure VI-15. Based on the size of the fractions eluted, both monomer and dimer, as well as higher-order oligomers are isolated from this expression. To confirm that these proteins are CRABPII, rather than proteins from E. coli, a SDS-page denaturing gel was run (Figure VI15). The order of fractions is in the order of elution from gel filtration. From the denaturing gel, it is quite obvious that this expression does not cleanly produce only CRABPII. Fractions 12 to 17 from gel filtration are not CRABPII, but likely E. coli proteins. However, fractions 17 to 21 (though with some impurity) show the correct molecular weight for CRABPII. Based on this data, it is observed that CRABPII can indeed form the dimer species; whether it is domain swapped dimer or homo dimer, in which structural elements have not 314 been exchanged, is yet to be investigated. It is suggested that mutations that affect dimerization in hCRBPII be carried out on equivalent non-conserved residues in CRABPII. Abs. (mAu) Fraction 17-18 dimer 260 nm 280 nm 360 nm 300 200 Fraction 20-21 monomer 100 0 40 60 80 100 Time (min) 120 140 Fraction from SEC Standards 12 13 14 15 17 18 19 20 21 250 kD 150 kD 100 kD 75 kD 50 kD 37 kD 25 kD 20 kD CRABPII ~ 15 kD 10 kD Figure VI-15. Size exclusion chromatography trace and denaturing protein gel for each fraction obtained from expression of I52A. VI.9 Conclusions The onset of this chapter described the seemingly random mutations that lead to dimer formation in hCRBPII. It was previously hypothesized that the interaction between Y60 and E72 was important for retaining the monomeric fold, 315 by controlling the conformation of the open monomer intermediate. It was further predicted that residues along the same strand (52 – 64) may also contribute to controlling the relative orientation of the two halves.1 The mutations shown to affect monomer/dimer ratio during bacterial protein overexpression are shown in the wild-type domain swapped dimer of hCRBPII (Figure IV-16). Presumably these mutations alter the hinge loop or the open interface, which can change the equilibrium between the monomer and the domain-swapped dimer. It is possible that new interactions (ionic, hydrophobic packing, etc.) at the open interface make the domain swapped dimer more favorable; this would explain why residues distal from the hinge loop (I25, K40, L77, Q128) could significantly alter the formation of domain swapped dimer in hCRBPII. I25 L77 Q128 T53 T51 E72 V62 Y60 R58 K40 Q4 Figure VI-16. Residues shown to affect the formation of domain swapped dimer in hCRBPII. Coordinates obtained from 4ZH9. 316 In the absence of more crystal structures and protein folding assays, we cannot make any concrete conclusions as to why mutations at residue 51 render hCRBPII prone to dimerization. We hypothesize that the ratio of monomer/dimer formation is correlated to the stability of the monomer and dimer. If there exists an equilibrium between the closed monomer and open monomer intermediate, and the domain swapped dimer and the open monomer intermediate, it would stand to reason that the most stable product would prevail. In the same manner, destabilization of the monomer would lead to more domain swapped dimer. To this end, it would be interesting to measure the thermal stability of all the monomers and dimers developed in this study in order to ascertain whether the ratio of monomer/dimer corresponds with stability of the products. 317 REFERENCES 318 REFERENCES 1. Assar, Z., Nossoni, Z., Wang, W., Santos, E.M., Kramer, K., McCornack, C., Vasileiou, C., Borhan, B. & Geiger, J.H. Domain-Swapped Dimers of Intracellular Lipid-Binding Proteins: Evidence for Ordered Folding Intermediates. Structure 24, 1590-1598 (2016). 2. Wang, W. in PhD Thesis, Michigan State University (2012). 3. Nosrati, M. in PhD Thesis, Michigan State University (2016). 4. Kyte, J. & Doolittle, R.F. A simple method for displaying the hydropathic character of a protein. Journal of Molecular Biology 157, 105-132 (1982). 5. Hopp, T.P. & Woods, K.R. A computer program for predicting protein antigenic determinants. Molecular Immunology 20, 483-489 (1983). 6. Cornette, J.L., Cease, K.B., Margalit, H., Spouge, J.L., Berzofsky, J.A. & DeLisi, C. Hydrophobicity scales and computational techniques for detecting amphipathic structures in proteins. Journal of Molecular Biology 195, 659-685 (1987). 7. Eisenberg, D., Schwarz, E., Komaromy, M. & Wall, R. Analysis of membrane and surface protein sequences with the hydrophobic moment plot. Journal of Molecular Biology 179, 125-142 (1984). 8. Rose, G.D., Geselowitz, A.R., Lesser, G.J., Lee, R.H. & Zehfus, M.H. Hydrophobicity of amino acid residues in globular proteins. Science 229, 834-838 (1985). 9. Janin, J. Surface and inside volumes in globular proteins. Nature 277, 491-492 (1979). 10. Engelman, D.M., Steitz, T.A. & Goldman, A. Identifying nonpolar transbilayer helices in amino acid sequences of membrane proteins. Annual Review of Biophysics and Biophysical Chemistry 15, 321-353 (1986). 11. Berbasova, T., Santos, E.M., Nosrati, M., Vasileiou, C., Geiger, J.H. & Borhan, B. Light-Activated Reversible Imine Isomerization: Towards a Photochromic Protein Switch. ChemBioChem 17, 407-414 (2016). 319 12. Nosrati, M., Berbasova, T., Vasileiou, C., Borhan, B. & Geiger, J.H. A Photoisomerizing Rhodopsin Mimic Observed at Atomic Resolution. Journal of the American Chemical Society 138, 8802-8808 (2016). 320 CHAPTER VII: MATERIALS AND METHODS VII.1 Site-directed mutagenesis of hCRBPII and CRABPII The pET-17b plasmid, containing hCRBPII-Q108K:K40L cloned between NdeI and XhoI, was used as a template for mutagenesis of hCRBPII.1, 2 Single point mutations were made on wild-type CRABPII in pET-17b (cloned between NdeI and EcoRI). Site-directed mutagenesis was conducted via polymerase chain reaction (PCR), with the specified cycling conditions shown in Table VII-1. Table VII-1. PCR cycling conditions for site-directed mutagenesis. PCR Program 1x Time (min) 94 °C 3:00 94 °C 0:20 3 – 5 °C below Tm 0:55 72 °C 3:30 1x 72 °C 10:00 1x 4 °C 5:00 20x Reactant Volume DNA template 70 ng (x μL) Forward primer 20 pmol (y μL) Reverse primer 20 pmol (z μL) 10 mM dNTP 1 μL 50 mM MgCl2 1 μL DMSO 5 μL 10 x Cloned Pfu Reaction Buffer 5 μL Pfu Turbo DNA polymerase (2.5 U/μL) 1 μL DI water 50 μL – x – y – z – 7 μL 321 The primers used for mutagenesis were ordered from Integrated DNA Technologies (IDT), with melting temperatures (Tm) from approximately 52 °C to 65 °C. The sequences of the forward primers (5’ to 3’) are listed below. It should be noted that, in all cases, the reverse primer is the reverse complement of the forward primer. All primers correspond to hCRBPII, unless otherwise indicated. Q4E-GACGAGGGACGAGAATGGAACC Q4F-GACGAGGGACTTCAATGGAACC Q4W-GACGAGGGACTGGAATGGAACC W8E-AATGGAACCGAGGAGATGGAGAGT W8F-CAGAATGGAACCTTCGAGATGGAGAGT F16W:Y19W-AGTAATGAAAACTGGGAGGGCTGGATGAAGGCC F16Y-GAGTAATGAAAACTATGAGGGCTACATG F16Y:Y19W-GAGAGTAATGAAAACTATGAGGGCTGGATG Y19F-CTTTGAGGGCTTCATGAAGGC Y19W-CTTTGAGGGCTGGATGAAGGC I25W-AAGGCCCTGGATTGGGATTTTGCCACC T29W:A33W-GATTTTGCCTGGCGCAAGATTTGGGTACGTCTC A33W-CGCAAGATTTGGGTACGTCTCAC Q38A:K40L-GTACGTCTCACTGCGACGCTGGTTATTGATCAA Q38E:K40L-GTACGTCTCACTGAGACGCTGGTTATTGATCAA Q38F:K40C-GTACGTCTCACTTTTACGTGTGTTATTGAT Q38F:K40D-GTACGTCTCACTTTTACGGACGTTATTGATCAA 322 Q38F:K40E-GTACGTCTCACTTTTACGGAAGTTATTGATCAA Q38F:K40H-GTACGTCTCACTTTTACGCATGTTATTGAT Q38F:K40L-GTACGTCTCACTTTTACGCTGGTTATTGATCAA K40L:I42E-ACGCTGGTTGAAGATCAAGATGGT Q38F:K40Q-GTACGTCTCACTTTTACGCAAGTTATTGAT Q38F:K40R-GTACGTCTCACTTTTACGCGAGTTATTGAT Q38F:K40T-GTACGTCTCACTTTTACGACAGTTATTGAT Q38F:K40Y-GTACGTCTCACTTTTACGTACGTTATTGAT Q38M:K40L-CGTCTCACTATGACGCTGGTTATTG Q38W:K40L-GTACGTCTCACTTGGACGCTGGTTATTGATCAA K40D-CTCACTCAGACGGATGTTATTGATCAAGATGG K40E-CTCACTCAGACGGAGGTTATTGATCAA K40H-CTCACTCAGACGCACGTTATTGATCAA K40L-CTCACTCAGACGCTGGTTATTGATCAAGATGG K40W-GTCTCACTCAGACGTGGGTTATTGATCAAGA F49E:T53A-GATGGTGATAACGAAAAGACAAAAGCAACTAGCACATTC T51A-ATAACTTCAAGGCAAAAACCACTAGC T51A:T53S-ATAACTTCAAGGCAAAATCCACTAGCACATT T51C-GGTGATAACTTCAAGTGTAAAACCACTAGCAC T51C:T53A-GATAACTTCAAGTGCAAAGCCACTAGCACATTC T51C:T53S-GATAACTTCAAGTGTAAAAGCACTAGCACA T51D-GGTGATAACTTCAAGGATAAAACCACTAGCAC 323 T51D:T53A-GATAACTTCAAGGACAAAGCCACTAGCACATTC T51E-GATAACTTCAAGGAGAAAACCACTAGCAC T51E:T53A-GATAACTTCAAGGAAAAAGCAACTAGCACATTC T51F-GGTGATAACTTCAAGTTCAAAACCACTAGCAC T51G-GGTGATAACTTCAAGGGCAAAACCACTAGCAC T51H-GGTGATAACTTCAAG CAC AAAACCACTAGCAC T51I-GGTGATAACTTCAAGATTAAAACCACTAGCAC T51K-GATAACTTCAAGAAAAAAACCACTAGC T51L-GGTGATAACTTCAAGCTGAAAACCACTAGCAC T51M-GGTGATAACTTCAAGATGAAAACCACTAGCAC T51N-GGTGATAACTTCAAGCTGAAAACCACTAGCAC T51N:T53S-GTGATAACTTCAAGAATAAAAGCACTAGCAC T51P-GGTGATAACTTCAAGCCAAAAACCACTAGCAC T51Q-GGTGATAACTTCAAGCAAAAAACCACTAGCAC T51R-GGTGATAACTTCAAGCGAAAAACCACTAGCAC T51S-GGTGATAACTTCAAGTCGAAAACCACTAGCAC T51V-GGTGATAACTTCAAGGTAAAAACCACTAGCAC T51V:T53A-GATAACTTCAAGGTAAAAGCCACTAGCACATTC T51V:T53D-CTTCAAGGTAAAAGACACTAGCACATTC T51V:T53S-CAAGGTAAAAAGCACTAGCACATTC T51W-GGTGATAACTTCAAGTGGAAAACCACTAGCAC T51W:T53S-GATAACTTCAAGTGGAAAAGCACTAGCACATTC 324 T51Y-GGTGATAACTTCAAG TAC AAAACCACTAGCAC T53A-TTCAAGACAAAAGCCACTAGCACATTC T53E-TTCAAGACAAAAGAGACTAGCACATTC T53S-CAAGACAAAAAGCACTAGCACATTCCG T53V-CTTCAAGACAAAAGTCACTAGCACATTCCG F57W:R58W-CACTAGCACATGGTGGAACTATGATG R58E-CTAGCACATTCGAGAACTATGATGTG R58F-CTAGCACATTCTTCAACTATGATGTG R58G-CTAGCACATTCGGCAACTATGATGTG R58H-CTAGCACATTCCACAACTATGATGTG R58L-CTAGCACATTCCTGAACTATGATGTG R58L:Y60F-ACTAGCACATTCCTCAACTTTGATGTGGATTTC R58P-CTAGCACATTCCCCAACTATGATGTG R58Q-CTAGCACATTCCAGAACTATGATGTG R58W-CTAGCACATTCTGGAACTATGATGTG R58W:Y60W-GCACATTCTGGAACTGGGATGTGGATTTC R58Y-CTAGCACATTCTACAACTATGATGTG Y60A-TTCCGCAACGCTGATGTGGAT Y60F-CACATTCCGCAACTTTGATGTGGATTTCAC Y60H-CACATTCCGCAACCATGATGTGGATTTCAC Y60L-TTCCGCAACCTGGATGTGGAT Y60W-CACATTCCGCAACTGGGATGTGGATTTCAC 325 V62E-AACTATGATGAGGATTTCACTGTTGGAGTA F64E-AACTATGATGTGGATGAAACTGTTGGAGTAGAG F64W-GATGTGGATTGGACTGTTGGAGTAG V62N-AACTATGATAACGATTTCACTGTTGGAGTAG E72A-GTAGAGTTTGACGCGTACACAAAGAGC E72L-GTAGAGTTTGACCTGTACACAAAGAGC S76A-GAGTACACAAAGGCCCTGGATAACCGG S76W-GAGTACACAAAGTGGCTGGATAACCGG L77W-GTACACAAAGAGCTGGGATAACCGGCATG V86E-TTAAGGCACTGGAGACCTGGGAAGG W88F-GCACTGGTCACCTTCGAAGGTGATGTC L93E-TGGGAAGGTGATGTCGAAGTGTGTGTGCAAAAG C95E-GATGTCCTTGTGGAGGTGCAAAAGGGG W106E:Q108K-GAGAACCGCGGCGAGAAGAAGTGGATTGAGGGG W106F:Q108K-GAGAACCGCGGCTTCAAGAAGTGGATT Q108K:W109F-GGCTGGAAGAAGTTCATTGAGGGGGAC L115E -GGGGACAAGGAGTACCTGGAGC L117D-CAAGCTGTACGAGGAGCTGACC L117E-CAAGCTGTACGACGAGCTGACC L119E -GCTGTACCTGGAGGAGACCTGTGGTGAC L119W-GTACCTGGAGTGGACCTGTGGTG Q128E-CAGGTGTGCCGTGAGGTGTTCAAAAAG 326 Q128L-CAGGTGTGCCGTCTGGTGTTCAAAAAG F130E -TGCCGTCAAGTGGAGAAAAAGAAGTGA I52A (CRABPII)-GACACTTTCTACGCCAAAACCTCCACC I52S (CRABPII)-GACACTTTCTACUCCAAAACCTCCACC I52W (CRABPII)-GACACTTTCTACTGGAAAACCTCCACC I52N (CRABPII)-GACACTTTCTACAACAAAACCTCCACC The crude PCR product was then digested with 20 units DpnI enzyme (New England Biolabs) for one hour at 37 °C. The resulting solution (7 μL) was then added to E. coli XL-1 Blue competent cells (Novagen, 100 μL) on ice for 30 min. Subsequently, the cells were heat shocked for 30 sec at 42 °C and then gently spread on a Luria broth (LB) agar plate supplemented with 100 μg/mL ampicillin and 12.5 μg/mL tetracycline. The plate was incubated at 37 °C for approximately 16 h. A single colony was then inoculated into 10 mL LB media supplemented with 100 μg/mL ampicillin and 12.5 μg/mL tetracycline. LB media was prepared by adding 10 g tryptone, 7 g yeast extract and 5 g NaCl to 1 L DI water. The media was autoclaved and cooled to room temperature before use. The inoculated culture was shaken at 37 °C for 12 h. DNA purification was performed using a Promega Wizard Plus SV miniprep DNA purification system (A1330). The concentration of the isolated plasmid was measured via Nandrop; the average concentration was 100 ng/μL. Every sampled was sequenced by the Research 327 Technology Support Facility at Michigan State University, using a primer corresponding to the T7 promoter for all samples in pET-17b plasmid. VII.2 Protein expression and purification of hCRBPII and CRABPII in pET17b expression plasmid The target gene (100 ng of DNA for 100 μL cell solution) was added to thawed BL21(DE3) pLysS competent cells (Invitrogen) on ice and incubated for 30 min. Subsequently, the cells were heat shocked for 30 sec at 42 °C and then gently spread on a Luria broth (LB) agar plate supplemented with 100 μg/mL ampicillin and 27 μg/mL chloramphenicol. The plate was incubated at 37 °C for approximately 12 h. A single colony was then inoculated into 1 L terrific broth (TB) media supplemented with 100 μg/mL ampicillin and 27 μg/mL chloramphenicol. TB media was prepared by mixing two solutions and autoclaving them separately. The first solution consists of 12 g tryptone, 24 g yeast extract and 4 mL glycerol to 900 mL DI water. The second solution was prepared by mixing 2.31 KH2PO4 and 12.54 g of K2HPO4 in 90 mL DI water. Before inoculation the solutions were mixed. The inoculated culture was shaken at 37 °C until optical density (OD) at 600 nm was approximately 1; this typically takes eight to nine h. Overexpression was induced by the addition of isopropyl-β-D-thiogalactopyranoside (IPTG, Gold Biotechnology) at a final concentration of 1 mM. The culture was then shaken at 23 °C for 20 h. 328 The cells were harvested by centrifugation (5000 rpm, 10 min, 4 °C) and resuspended in Tris-binding buffer (10 mM Tris, pH=8.0, 50 mL). The cells were then lysed by sonication (Biologics, Inc, power 60%, 3 min). The solution was again centrifuged to separate the pellet and supernatant (5000 rpm, 30 min, 4 °C). All further protein purification was also conducted at 4 °C. The supernatant was then loaded onto a FastQ anion exchange column pre equilibrated with Tris buffer (10 mM Tris, pH = 8.0). After binding of the protein to the FastQ anion exchange resin (GE Healthcare), the column was washed twice with Tris buffer (100 mL). Lastly the protein was eluted with Triselution buffer (10 mM Tris, 200 mM sodium chloride, pH = 8.0, 100 mL). The eluent from the FastQ anion exchange column was then desalted with Tris buffer using an ultrafiltration cell under nitrogen pressure (~20 psi) equipped with a 10 kDa molecular weight cutoff membrane (Millipore, Regenerated Cellulose membrane, diameter 63.5 mm, NMWL: 10,000). The protein was first concentrated to ~10 mL and then diluted to 150 mL with Tris buffer. This was again concentrated to less than 30 mL. Purification was continued with Fast Protein Liquid Chromatography (NGC chromatography system, Biorad), equipped with a column loaded with Source 15Q (Q Sepharose Fast Flow, GE Healthcare) anion exchange resin. The method for FPLC source Q is shown in Table VII-2. The pH at all steps was set to 8.1. Percent B corresponds to the percent salt, where 100% is equivalent to 1 mM NaCl. 329 Table VII-2. FPLC Source 15Q method. Step Description %B Volume Flow Rate 1. Isocratic flow 0 12 mL 3 mL/min 2. Load sample n/a Sample volume 2 mL/min 3. Isocratic flow 0 10 mL 3 mL/min 4. Gradient flow 0 to 4 12 mL 3 mL/min 5. Isocratic flow 4 20 mL 3 mL/min 6. Gradient flow 4 to 8 15 mL 3 mL/min 7. Isocratic flow 8 20 mL 3 mL/min 8. Gradient flow 8 to 15 15 mL 3 mL/min 9. Isocratic flow 15 40 mL 3 mL/min 10. Gradient flow 15 to 75 10 mL 3 mL/min 11. Isocratic flow 100 20 mL 3 mL/min 12. Isocratic flow 0 35 mL 3 mL/min Protein was then collected from 40 mM, 80 mM or 150 mM NaCl and concentrated to 1 mL using a 10 kDa Centriprep centrifugal filter (Millipore, Regenerated Cellulose membrane, NMWL: 10,000). The concentrated sample was then loaded to the Fast Protein Liquid Chromatography (NGC chromatography system, Biorad), equipped with a column loaded with size exclusion chromatography (SEC) Superdex 75 Prep Grade resin (GE Healthcare). The method for FPLC source Q is shown in Table VII-3. At all steps, the pH was set to 8.1. Table VII-2. FPLC SEC method. Step Description %B Volume Flow Rate 1. Load sample 0 4 mL 1 mL/min 2. Isocratic flow 20 139 mL 1 mL/min 330 VII.3 Protein extinction coefficient determination The extinction coefficients of the proteins were measured at 280 nm, following the method described by Gill and von Hippel.3 The theoretical extinction coefficient (εtheor) is calculated by the following equation: !!!!"# = ! × !!"# + ! × !!"# + ! × !!"# where a, b and c are the number of tryptophans, tyrosines and cysteine residues, respectively. The extinctions of tryptophan, tyrosine and cysteine are 5690 M-1 cm-1, 1280 M-1 cm-1 and 120 M-1 cm-1, respectively. The absorbance at 280 nm of the protein was measured at the same concentration under native (2 x PBS buffer) and denaturing (final concentration 6 M guanidine HCl) conditions. The ratio of absorbance intensities under native (Anative) and denaturing (Adenaturing) conditions, multiplied by εtheor yielded εexp as shown in the following equation. !!"# = !!"#$%& !!"#$%&'(#) × !!!!"# A stock solution of PBS buffer (10 x) was prepared by dissolving 2.0 g KCl (26 mM), 2.4 g KH2PO4 (17.6 mM), 80 g NaCl (1368 mM), 11.45 g Na2HPO4 (80.7 mM) in 1000 mL DI water. The solution was autoclaved and then diluted to 2 x with autoclaved DI water and filtered before using. The extinction coefficients of all purified hCRBPII and CRABPII monomers and dimers described in this thesis are given in Table VII-3. All proteins are hCRBPII, unless otherwise noted. 331 Table VII-3. Extinction coefficients of hCRBPII and CRABPII mutants. εexp(280 nm) Protein Monomer Dimer Q108K:K40C:T53A:R58L:Q38F:Q4F 27,948 - Q108K:K40D 28,612 58,690 Q108K:K40D:Q4F 28,145 - Q108K:K40D:R58L:Q38F:Q4F 29,070 58,142 Q108K:K40D:T51V:R58L:Q38F:Q4F:V62E 27,532 55,904 Q108K:K40D:T51V:T53A:R58L:Q38F:Q4F:V62E 26,357 53,118 - 55,372 28,682 56,298 - 56,982 Q108K:K40D:T53A:R58L:Q4F 25,887 55,852 Q108K:K40D:T53A:R58Y:Q38F:Q4F 28,690 - Q108K:K40D:T53A:R58Y:Q38F:Q4F:L117E 32,800 59,040 Q108K:K40E 31,611 56,834 Q108K:K40E:R58L:Q38F:Q4F 29,350 52,302 Q108K:K40E:R58L:Q38F:Q4F:W106F 21,632 - Q108K:K40E:R58L:Q38F:Q4F:W109F 22,206 - Q108K:K40E:R58L:Q38F:Q4F:W88F 20,811 - Q108K:K40E:R58L:Q38F:Q4F:W8F 21,353 - Q108K:K40E:T53A:Q38F:Q4F 28,083 53,656 Q108K:K40E:T53A:R58L:Q38F:Q4F 27,681 - Q108K:K40E:T53A:R58L:Q38F:Q4F:V62E 25,267 53,774 Q108K:K40E:T53A:R58L:Q38F:Q4F:Y19W 32,892 - Q108K:K40E:T53A:R58L:Q4F 28,344 54,796 Q108K:K40E:T53A:R58W:Q38F:Q4F 34,382 - Q108K:K40E:T53A:R58W:Q38F:Q4F:Y19W 38,874 - Q108K:K40H:T51V:T53S:R58W:Y19W:A33W 42,995 - Q108K:K40H:T53A:R58L:Q38F:Q4F 27,742 - Q108K:K40D:T53A:Q38F:Q4F Q108K:K40D:T53A:R58L:Q38F:Q4F Q108K:K40D:T53A:R58L:Q38F:Q4F:V62E 332 Table VII-3 (cont’d) Q108K:K40L 27,618 - Q108K:K40L:L117E 28,540 - Q108K:K40L:Q38F 29,409 - Q108K:K40L:Q4F 30,761 - Q108K:K40L:R58F 26,372 - Q108K:K40L:R58W:Y19W 38,360 - Q108K:K40L:R58Y 28,800 - Q108K:K40L:T51A 29,926 59,182 - 61,640 Q108K:K40L:T51C 31,624 - Q108K:K40L:T51C:T53A:R58L:Q38F 28,527 - Q108K:K40L:T51D:T53A:R58Y:Q38F:Q4F 28,763 - Q108K:K40L:T51D:T53A:R58Y:Q38F:Q4F:L117E 30,317 - Q108K:K40L:T51E 27,398 58,828 Q108K:K40L:T51F 30,065 63,540 Q108K:K40L:T51F:Y60A - 49,397 Q108K:K40L:T51F:Y60F 29,077 56,631 Q108K:K40L:T51F:Y60H - 56,329 Q108K:K40L:T51F:Y60L - 59,024 Q108K:K40L:T51F:Y60W 30,651 62,700 Q108K:K40L:T51H 26,956 53,057 Q108K:K40L:T51I 28,736 - Q108K:K40L:T51K - 57,708 Q108K:K40L:T51L 27,987 64,330 Q108K:K40L:T51M 26,772 62,166 Q108K:K40L:T51N 30,281 60,452 Q108K:K40L:T51N:T53S:V62N 32,319 57,556 Q108K:K40L:T51Q 27,663 55,434 Q108K:K40L:T51A:T53S:V62N 333 Table VII-3 (cont’d) Q108K:K40L:T51S 33,186 56,882 Q108K:K40L:T51V 29,609 - Q108K:K40L:T51V:R58W:Y19W 37,342 - Q108K:K40L:T51V:T53D:R58W:Y19W:L117E 37,899 - Q108K:K40L:T51V:T53S 30,888 - Q108K:K40L:T51V:T53S:R58E 28,960 60,986 Q108K:K40L:T51V:T53S:R58F 29,833 - Q108K:K40L:T51V:T53S:R58G 30,371 - Q108K:K40L:T51V:T53S:R58H 28,504 - Q108K:K40L:T51V:T53S:R58H:Y19W 34,571 - Q108K:K40L:T51V:T53S:R58H:Y19W:L117D:Q4F 31,920 - Q108K:K40L:T51V:T53S:R58H:Y19W:L117E 32,192 - Q108K:K40L:T51V:T53S:R58H:Y19W:L117E:A33W 38,408 - Q108K:K40L:T51V:T53S:R58L 29,079 60,938 Q108K:K40L:T51V:T53S:R58P 30,155 - Q108K:K40L:T51V:T53S:R58Q 29,013 60,989 Q108K:K40L:T51V:T53S:R58W 33,984 - Q108K:K40L:T51V:T53S:R58W:Y19W 36,880 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W 43,042 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:F16Y 44,754 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:F57W 47,166 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:F64W 49,830 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:I25W 54,006 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117D 43,373 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E 42,130 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L119W 49,720 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Q38W 48,544 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Q4W 52,151 - 334 Table VII-3 (cont’d) Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Q4W:F16Y 57,192 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:S76W 51,910 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:T29W 52,454 - Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:Y60W 49,545 118,327 Q108K:K40L:T51V:T53S:R58W:Y19W:F16Y 39,990 - Q108K:K40L:T51V:T53S:R58W:Y19W:L117D 38,626 - Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:A33W:Q4F 42,730 - Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F 38,135 - Q108K:K40L:T51V:T53S:R58W:Y19W:L117E 35,721 - Q108K:K40L:T51V:T53S:R58W:Y19W:L117E:Q4F 37,554 - Q108K:K40L:T51V:T53S:R58W:Y19W:L77W 44,867 - Q108K:K40L:T51V:T53S:R58W:Y19W:L77W:F16Y 45,310 - Q108K:K40L:T51V:T53S:R58Y 28,768 - Q108K:K40L:T51V:T53S:R58Y:Y19W 36,971 - Q108K:K40L:T51V:T53S:Y19W 34,288 - Q108K:K40L:T51V:T53W:R58W:Y19W:A33W 57,298 - Q108K:K40L:T51W 38,181 75,400 Q108K:K40L:T51W:T53S:R58W:Y19W:A33W 57,571 - Q108K:K40L:T51Y 28,044 54,498 Q108K:K40L:T53A 29,917 61,852 Q108K:K40L:T53A:Q38A 29,631 57,823 Q108K:K40L:T53A:Q38M 29,587 - Q108K:K40L:T53A:R58F 28,499 - Q108K:K40L:T53A:R58L 28,621 54,716 Q108K:K40L:T53A:R58L:A33W 35,160 - Q108K:K40L:T53A:R58L:Q128L 33,769 65,392 Q108K:K40L:T53A:R58L:Q38A 29,882 58,515 Q108K:K40L:T53A:R58L:Q38F 29,023 - 335 Table VII-3 (cont’d) Q108K:K40L:T53A:R58L:Q38F:F16Y 29,004 - Q108K:K40L:T53A:R58L:Q38F:L77W 34,305 - Q108K:K40L:T53A:R58L:Q38F:Q4F 31,014 - Q108K:K40L:T53A:R58L:Q38F:Q4F:I42E 29,326 - Q108K:K40L:T53A:R58L:Q38F:Q4F:F49E 32,681 - Q108K:K40L:T53A:R58L:Q38F:Q4F:L117E 28,159 57,373 Q108K:K40L:T53A:R58L:Q38F:Q4F:L119E 29,595 - Q108K:K40L:T53A:R58L:Q38F:Q4F:Q128E 30,636 - Q108K:K40L:T53A:R58L:Q4F:Q38E 29,153 - Q108K:K40L:T53A:R58L:Q38F:Q4F:T51E 29,050 63,812 Q108K:K40L:T53A:R58L:Q38F:Q4F:V62E 28,660 - Q108K:K40L:T53A:R58L:Q38F:S76A 30,174 - Q108K:K40L:T53A:R58L:Q38F:Y19W 36,749 - Q108K:K40L:T53A:R58L:Q38F:Y60F 26,367 - Q108K:K40L:T53A:R58L:Q38M 28,828 - Q108K:K40L:T53A:R58L:Y19F 29,082 64,206 Q108K:K40L:T53A:R58W 37,876 71,046 Q108K:K40L:T53A:R58Y:Q38F:Q4F:L117E 28,820 - Q108K:K40L:T53E:R58L:Q38F:Q4F 27,822 - Q108K:K40L:T53S 28,195 64,629 Q108K:K40L:T53S:R58W:Y19W 38,340 - Q108K:K40L:T53S:V62E 29,792 60,016 Q108K:K40L:T53S:V62N 32,430 59,200 Q108K:K40L:T53V 28,710 61,756 Q108K:K40L:T53V:R58L 27,460 - Q108K:K40L:V62N 29,091 58,408 Q108K:K40Q:T53A:R58L:Q38F:Q4F 28,759 - Q108K:K40R:T53A:R58L:Q38F:Q4F 27,776 - 336 Table VII-3 (cont’d) Q108K:K40T:T53A:R58L:Q38F:Q4F 28,048 - Q108K:K40W:T51V:T53S:R58W:Y19W:A33W 53,280 - Q108K:K40Y:T53A:R58L:Q38F:Q4F 28,587 - Q108K:T53A:R58L:Q38F:Q4F 29,370 - I52A (CRABPII) 18,391 - I52S (CRABPII) 16,914 - I52N (CRABPII) 19,582 - I52W (CRABPII) 26,101 - VII.4 UV-Vis measurements of hCRBPII/chromophore complexes UV-Vis spectra were recorded with a Cary 300 Bio WinUV, Varian spectrophotometer. For all experiments, unless otherwise indicated, 20 μM protein was incubated with ligand (0.5 equiv) in 2 x PBS buffer and incubated at room temperature until Schiff base (SB) or protonated Schiff base (PSB) formation was complete. This was verified by UV-Vis. VII.4.1 pKa measurements of hCRBPII/chromophore complexes For pKa determination, protein (20 μM in PBS) was incubated with ligand (0.5 equiv) and incubated at room temperature until Schiff base (SB) or protonated Schiff base (PSB) formation was complete. This was verified by UVVis. The solution was then titrated with acid (with 1 M NaOAc, pH 4) or base (1 M NaOH) in ~ 0.3 pH units, and the absorption spectra were recorded at each point. Absorbance change at λmax was plotted as a function of pH. A curve fit, as previously described for bacteriorhodopsin, was applied for pKa determination: ∆! = ∆!! + !"#$%&#% (1 + 10!"!!!! ) 337 The two parameters are: ΔA0, the total absorbance change of the PSB and pKa, the midpoint of titration. It should be noted that a constant is included to account the deviation from zero absorbance intensity of the deprotonated PSB. VII.4.2 Kinetic measurements of hCRBPII/chromophore PSB formation For kinetic measurements, PSB formation was followed by UV-Vis in 2 x PBS buffer (~ pH 7.2) at 23 °C using a Cary temperature controller. pH was verified each time before recording the spectrum. The experiment was performed with a final protein concentration of 20 μM protein and 0.5 equiv ligand. The spectra were recorded immediately after mixing the protein and chromophore and the absorbance intensity plotted as a function of time. The data were fit to either a pseudo-first order rate equation or second order rate equation. Those samples that exhibited high PSB pKa (pKa > 9) were fit to a second order rate equation, as we can be confident in the chromophore concentration. However, when both SB and PSB are formed under the conditions for the kinetic measurements, we had to fit a pseudo-first order rate equation, as concentration could not be accurately determined. The fit for pseudo-first order rate eis shown below; A is the absorbance value at each recorded time point, A0 is the final absorbance value (at completion of the reaction), k is the rate constant and t is the time after ligand addition. ! = !! × 1 − ! !!" The data was fit in Kaleidagraph using the equation shown below, where m3 is a constant that accounts for the time delay after ligand addition before 338 recording the first time point. The half-life (t1/2) of the reaction can be calculated by ln(2)/k, where m2 is the rate constant. ! = !1 × 1 − ! !!! × !! + !3 For those samples exhibiting high PSB pKa, we chose to fit the second order rate equation (in Kaleidagraph) shown below. The second order rate equation for product growth was previously derived,4, 5 where m3 is the concentration of the limiting reactant (in all cases this should be 10 μM, assuming greater than 50% of the protein is functional), m2 is the rate constant k and m1 is 1/m3. The half-life (t1/2) of the reaction can be calculated by 1/(k*[A]0). In all cases 10 μM was assumed for [A]0. ! = !3 − 1 !2 × !0 + !1 First, the concentration of product (hCRBPII/ligand complex) versus time was plotted. The concentration of product at each time point was calculated by Beer’s Law using the absorbance intensity at each time point and the extinction coefficient of the hCRBPII/ligand complex. For kinetic experiments conducted by fluorescence, product concentration was calculated by dividing the fluorescence intensity at each time point by the apparent extinction (maximum fluorescence intensity divided by the limiting reactant concentration). VII.5 Fluorescence measurements Fluorescence spectra were recorded using a Fluorolog®-3 spectrofluorometer (HORIBA, Ltd.). An entrance slit of 1 nm and exit slit of 12 nm was used for all measurements. 339 VII.6 Quantum yield measurements Absolute fluorescence quantum yields (Φ) were measured on a Quantaurus-QY (model C11347-11, Hamamatsu Photonics) equipped with a xenon light source, monochromator, integration sphere and a multichannel backthinned CCD detector at room temperature. All absolute quantum yields were measured by Mr. Wei Sheng and taken as an average of five recordings, excited at lambda max. VII.7 Live cell imaging in bacteria The target gene (100 ng of DNA for 100 μL cell solution) was added to thawed BL21(DE3) pLysS competent cells (Invitrogen) on ice and incubated for 30 min. Subsequently, the cells were heat shocked for 30 sec at 42 °C and then gently spread on a Luria broth (LB) agar plate supplemented with 100 μg/mL ampicillin and 27 μg/mL chloramphenicol. The plate was incubated at 37 °C for approximately 12 h. A single colony was used for inoculation to 10 mL LB. The culture was shaken at 37°C for 6 h followed by induction with IPTG (final concentration 1 mM). The culture was then shaken at 25°C for 16 h. For the control experiment, 5 μL of BL21 (DE3)pLysS cells was inoculated into 10 mL of LB. Culture growth and induction with IPTG proceeded the same as aforementioned. The cells were harvested by centrifugation and resuspended in 1 mL of PBS buffer (pH 7.2). The cultures were heated to 37°C for 10 min. Subsequently, ThioFluor was added and incubated at 37 °C, at the indicated final concentration 340 and time. After the incubation period, the cells were harvested by centrifugation and the supernatant was discarded. The pellets were washed 3x with warm PBS buffer. After the last wash, the cells were resuspended in PBS buffer. For imaging, the cells were plated on a microslide (VWR, 1 in x 3 in, 1.2 mm thick) and covered with a glass coverslip (Corning, No.1, 22 mm sq.). In order to stop the bacteria from floating in solution, the slide was heated at 42 °C for ~1 min. The sample was placed with the coverslip facing the objective. For bacterial imaging, a 63 x oil objective was used. VII.8 Cloning for mammalian expression vectors VII.8.1 General cloning protocol The DNA fragment was amplified using Phusion High-Fidelity DNA Polymerase (NEB, M0530) with the appropriate primers (see below for details). PCR conditions are specified in Table VII-4 using a Bio-Rad iCycler thermal cycler. Four reactions were set up for each cloning, to ensure that enough amplified product was obtained. The PCR amplified gene was purified by Wizard® SV Gel and PCR CleanUp System (Promega) from 1% agarose gel in amount of 20-50 ng/mL. The product was digested with the proper enzymes and ligated to a similarly prepared plasmid (50 ng/mL). Ligation between the insert fragment and plasmid was performed with 30 ng of plasmid and 90 ng of insert using T4 DNA Ligase (New England BioLabs). The ligated product was transformed into E. coli XL1-blue competent cells and grown on LB-agar plates supplemented with antibiotics (100 341 μg/mL ampicillin, 7.5 μg/mL tetracycline) at 37 ºC for 18 h. Colonies were inoculated in LB medium (15 mL) with proper antibiotics (100 μg/mL ampicillin, 7.5 μg/mL tetracycline) and incubated at 37 ºC while shaking, for 10 h. DNA purification was performed using Promega Wizard® Plus SV Miniprep DNA purification kit (A1330) following the manufacturer’s protocol. Table VII-4. PCR cycling conditions for cloning. PCR Program 1x Time (min) 98 °C 0:30 98 °C 0:10 55 °C 0:30 72 °C Extension time 1x 72 °C 10:00 1x 4 °C 10:00 40x Reactant Volume DNA template 100 ng (x μL) Forward primer 20 pmol (y μL) Reverse primer 20 pmol (z μL) 10 mM dNTP 1 μL 50 mM MgCl2 1 μL DMSO 5 μL 5 x Phusion HF Reaction Buffer 10 μL Phusion HF DNA Polymerase 0.5 μL DI water 50 μL – x – y – z – 17.5 μL The DNA sequence was verified with the corresponding sequencing primers by the MSU gene sequencing facility. Sequencing primers used are shown below. 342 CMV end_Seq: 5’-GGTCTATATAAGCAGAGCTGGTTTAG-3’ midGFP: 5’-CGTGCTGCTGCCCGACAACC-3’ CFP FOR: 5’-CGGTCGCCACCATGGTGAGCAAG-3’ VII.8.2 Preparation of plasmids for mammalian expression Plasmid1 (HindIII-EGFP-NotI-hCRBPIItetra-Stop-BamHI in pFlag-CMV2) and Plasmid2 (HindIII-EGFP-NotI-hCRBPIItetraQ108L-Stop-BamHI in pFlag-CMV2) were constructed by amplifying the hCRBPII gene of interest from pET-17b, Q108K:K40L:T51V:R58F and Q108L:K40L:T51V:R58F, respectively. Forward and reverse primers used were 5’-CGGCGGCCGCATGACGAGGGCC-3’ and 5’GCGGATCCTCACTTCTTTTTGAACAC-3’ (see Table VII-4 for protocol, extension time = 15 sec). The purified PCR product was then inserted between NotI and BamHI of Plasmid3. The products were sequenced with the primers CMVend_Seq and midGFP. Plasmid3 (HindIII-EGFP-NotI-CRBPIInona-Stop-BamHI in pFlag-CMV2) was prepared in multiple steps. The first step (completed by Dr. Tetyana Berbasova) served to obtain the construct HindIII-EGFP-hCRBPIInona-Stop-BamHI in pFlagCVM2.6 The EGFP- hCRBPIInona insert was amplified in two rounds of PCR (the italicized portion of the primer indicates the overlap of the sequences): 1) EGFP was amplified from EGFP-RB (donated from Professor Bill Henry’s Lab at MSU) with the forward primer 5’-CGAAGCTTATGGTGAGCAAGGGCGAG-3’ and reverse 5’-CCGGACTTGTACAGC-3’ primer (see Table S3 for protocol, extension time = 30 sec). 2) hCRBPIInona was amplified from pET-17b with the 343 forward primer 5‘-GCTGTACAAGTCCGGAGCCGCTGCAGGAGGC AGCCAAATGACGAGGTTC-3’ and GCGGATCCTCACTTCTTTTTGAACAC-3’ reverse (see Table primer VI-4 for 5’protocol, extension time = 30 sec). The PCR products were combined and amplified with the abovementioned HindIII-EGFP forward primer and hCRBPII-Stop-BamHI reverse primer (see Table VII-4 for protocol, extension time = 45 sec). The fulllength insert was ligated between HindIII and BamHI of the digested pFLAGCMV2-AML1B vector (Addgene #12504). A NotI cutting site was then introduced by two sequential site-directed mutageneses. The DNA was purified and sequenced after each step. The first PCR was conducted with CTGCAGGAGGCGGCCAAATGACGA-3’ the and forward reverse primer 5’- primer 5’- CTCGTCATTTGGCCGCCTCCTGCAG-3’ to introduce the first half of the NotI cutting site (see Table S1 for protocol). The second PCR was conducted with the forward primer 5’-CAGGAGGCGGCCGCATGACGAGG-3’ and reverse primer 5’CCTCGTCATGCGGCCGCCTCCTG-3’ to finish introducing the NotI cutting site (see Table VII-1 for protocol). After the two PCRs, the product is HindIII-EGFPNotI-hCRBPIInona-Stop-BamHI in pFlag-CVM2. This was done to allow for easy exchange of the hCRBPII variant. The product was sequenced with the primers CMVend_Seq and midGFP. Plasmid4 (HindIII-EGFP-NotI-hCRBPIInona-EcoRI-3NLS-Stop in pFlagCMV2) was prepared in multiple steps. The first step served to obtain the 344 construct HindIII-hCRBPIInona-3NLS-Stop-EcoRI in pFlag-CVM2 (completed by Dr. Tetyana Berbasova).6 Two sequential PCR amplifications were necessary to amplify the full-length insert. In the first step, hCRBPIInona was amplified from the pET-17b vector using a reverse primer containing a partial 3NLS sequence. The forward and reverse primers GCAAGCTTACGAGGGACTTCAATGGAACC-3’ used are and 5’5’- CTTCTTGGGGTCCACCTTCCTCTTCTTCTTGGGGTCAGCCCTGCTCTTCTTTT TGAACAC-3’ (see Table VII-4 for protocol, extension time = 15 sec). The PCR product was purified by gel extraction as described above. A second PCR was performed to complete the full-length 3NLS localization signal, add a stop codon and an EcoRI site. The second PCR was done in two 50 μL reaction volume using 20 ng of the product from the first PCR. Forward and reverse primers used in the second reaction are 5’-GCAAGCTTACGAGGGACTTCAATGGAACC-3’ and 5’-GAATTCACACCTTCCTCTTCTTCTTGGGGTCCACCTTCCTCTTCTTCT TGGGGTCCAC-3’ (see Table VII-4 for protocol, extension time = 15 sec). The purified PCR product was then inserted between HindIII and EcoRI of the digested pFLAG-CMV2-AML1B vector. Subsequently, the existing EcoRI cutting site was removed by site-directed mutagenesis with the forward primer 5’-GGAAGGTGTGAAAACATCGATAGA-3’ and reverse primer 5’-TCTATCGATGTTTTCACACCTTCC-3’ (see Table VII-1 for protocol) and a new EcoRI cutting site was introduced via site-directed mutagenesis before the 3NLS signal by PCR with the forward primer 5’- 345 AAAAAGAAGGAATTCGCTGACCCC-3’ and reverse primer 5’- GGGGTCAGCGAATTCCTTCTTTT-3’ (see Table VII-1 for protocol). After these two PCRs, the product is HindIII-hCRBPIInona-EcoRI-3NLS-Stop in pFlag-CVM2. This was done to allow for easy exchange of the hCRBPII variant. EGFP-NotI-hCRBPIInona was then amplified from Plasmid3 with the forward primer 5’-CGAAGCTTATGGTGAGCAAGGGCGAG-3’ and reverse primer 5’-CGGAATTCGCACTTCTTTTTGAACACTTG-3’ (see Table VII-4 for protocol, extension time = 30 sec). The purified PCR product was then inserted between HindIII and EcoRI of the HindIII-hCRBPIInona-EcoRI-3NLS-Stop construct in pFlag-CVM2. The product was sequenced with the primers CMVend_Seq and midGFP. To construct Plasmid5 (HindIII-EGFP-NotI-hCRBPIInona-EcoRI- NES(ELAEKLAGLDIN)-Stop-BamHI in pFlag-CMV2), a gBlock gene fragment of hCRBPIInona-EcoRI-NES was ordered from IDT. The fragment was amplified with the forward primer 5’-CGGCGGCCGCATGACGAGGTTC-3’ and reverse primer 5’-CGGGATCCTCAATTTATGTCAAGCCCGGCAAG-3’ (see Table VII-4 for protocol). The purified PCR product was then inserted between NotI and BamHI of Plasmid3. The product was sequenced with the primers CMVend_Seq and midGFP. Plasmid6 (XhoI-Q108K:K40L:T53S:V62N-BamHI-AgeI-ECFP-Stop-NotI in pECFP-N1) was constructed by amplifying Q108K:K40L:T53S:V62N from pET17b. Forward and reverse 346 primers used were 5’- GTTTACTTTACTCGAGCTCGCCACCATGACGAGGGACC-3’ CGGTGGATCCTTCGACTTCTTTTTGAACACTTG-3’ (see and Table VII-4 5’for protocol, extension time = 15 sec). The purified PCR product was then inserted between XhoI and BamHI of GalT-ECFP (Addgene #11937). The product was sequenced with the primers CMVend_Seq and CFP FOR. Plasmid7 (HindIII-EGFP-NotI-Q108K:K40L:T51V:T53S:R58W:Y19W: L117E-BamHI-RVASL-Stop in pFlag-CMV2) was constructed by amplifying Q108K:K40L:T51V:T53S:R8W:Y19W:L117E from pET-17b. Forward and reverse primers used were 5’-CGGCGGCCGCATGACGAGGGAC-3’ and 5’- GCCATGGATCCGCACTTCTTTTTGAACA-3’ (see Table VII-4 for protocol, extension time = 15 sec). The purified product was then inserted between NotI and BamHI of Plasmid3. The product was sequenced with the primers CMVend_Seq and midGFP. Plasmid8 (HindIII-EGFP-NotI-Q108K:K40L:T51V:T53S:R58W:Y19W: L117E-EcoRI-3NLS-Stop in pFlag-CMV2) was made by amplifying Q108K:K40L:T51V:T53S:R8W:Y19W:L117E from pET-17b. Forward and reverse primers used were 5’-CGGCGGCCGCATGACGAGGGAC-3’ and 5’- CGGAATTCGCACTTCTTTTTGAACACTTG-3’ (see Table VII-4 for protocol, extension time = 15 sec). The purified product was then inserted between NotI and EcoRI of Plasmid4. The product was sequenced with the primers CMVend_Seq and midGFP. 347 Plasmid9 (HindIII-EGFP-NotI-Q108K:K40L:T51V:T53S:R58W:Y19W: L117E-EcoRI-NES(ELAEKLAGLDIN)-Stop-BamHI in pFlag-CMV2) was constructed by amplifying Q108K:K40L:T51V:T53S:R8W:Y19W:L117E from pET17b. The forward and reverse CGGCGGCCGCATGACGAGGGAC-3’ primers and used were 5’- 5’-CGGAATTCGCACTTCTTTTT GAACACTTG-3’ (see Table VII-4 for protocol, extension time = 15 sec). The purified product was then inserted between NotI and EcoRI of Plasmid5. The product was sequenced with the primers CMVend_Seq and midGFP. Plasmid10 EcoRI-3NLS-Stop (HindIII-EGFP-NotI-Q108K:K40E:T53A:R58L:Q38F:Q4Fin pFlag-CMV2) was made by amplifying Q108K:K40E:T53A:R58L:Q38F:Q4F from pET-17b. The forward and reverse primers used were 5’-CGGCGGCCGCATGACGAGGGACTTCAAT-3’ and 5’CGGAATTCGGACTTCTTTTTGAACACTTG-3’ (see Table VII-4 for protocol, extension time = 15 sec). The purified product was then inserted between NotI and EcoRI of Plasmid8. The product was sequenced with the primers CMVend_Seq and midGFP. Plasmid11 (HindIII-EGFP-NotI-Q108K:K40E:T53A:R58L:Q38F:Q4F-StopBamHI in pFlag-CMV2) was made by amplifying Q108K:K40E:T53A:R58L:Q38F:Q4F from pET-17b. The forward and reverse primers used were 5’-CGGCGGCCGCATGACGAGGGACTTCAAT-3’ and 5’GCGGATCCTCACTTCTTTTTGAACAC-3’ (see Table VII-4 for protocol, extension time = 15 sec). The purified product was then inserted between NotI 348 and BamHI of Plasmid7. The product was sequenced with the primers CMVend_Seq and midGFP. Plasmid12 EcoRI-3NLS-Stop (HindIII-Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4Fin pFlag-CMV2) was constructed by amplifying Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F from pET-17b. The forward and reverse primers used were 5’-GATATAAAGCTTATGACGAGGGACCAG-3’ and 5’-CGGAATTCGCACTTCTTTTTGAACACTTG-3’. The purified product was then inserted between HindIII and EcoRI of HindIII-hCRBPIInona-EcoRI-3NLSStop in pFlag-CVM2 (an intermediate of Plasmid4). The product was sequenced with the primer CMVend_Seq. VII.8.3 Sequences of plasmids described in this thesis Plasmid1: HindIII-EGFP-NotI-Q108K:K40L:T51V:R58F-Stop-BamHI (pFlagCMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA AGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACCAGAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTACATGAAGGCCCTGGATA TTGATTTTGCCACCCGCAAGATTGCAGTACGTCTCACTCAGACGCTGGTTATTGATCAA GATGGTGATAACTTCAAGGTAAAAACCACTAGCACATTCTTCAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC CGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTGTACCTGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGAGGATCC 349 Plasmid2: HindIII-EGFP-NotI-Q108L:K40L:T51V:R58F-Stop-BamHI (pFlagCMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA AGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACCAGAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTACATGAAGGCCCTGGATA TTGATTTTGCCACCCGCAAGATTGCAGTACGTCTCACTCAGACGCTGGTTATTGATCAA GATGGTGATAACTTCAAGGTAAAAACCACTAGCACATTCTTCAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC CGCGGCTGGAAGCTGTGGATTGAGGGGGACAAGCTGTACCTGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGAGGATCC Plasmid3: HindIII-EGFP-NotI-Q108K:K40L:T51V:T53C:R58W:T29L:A33W: Q4F:L117E-Stop-BamHI (pFlag-CMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA AGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACTTCAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTACATGAAGGCCCTGGATA TTGATTTTGCCCTGCGCAAGATTTGGGTACGTCTCACTCAGACGCTGGTTATTGATCAA GATGGTGATAACTTCAAGGTAAAATGCACTAGCACATTCTGGAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC 350 CGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTGTACGAGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGAGGATCC Plasmid4: HindIII-EGFP-NotI-Q108K:K40L:T51V:T53C:R58W:T29L:A33W: Q4F:L117E-EcoRI-3NLS-Stop (pFlag-CMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA AGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACTTCAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTACATGAAGGCCCTGGATA TTGATTTTGCCCTGCGCAAGATTTGGGTACGTCTCACTCAGACGCTGGTTATTGATCAA GATGGTGATAACTTCAAGGTAAAATGCACTAGCACATTCTGGAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC CGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTGTACGAGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGCGAATTCGCTGACCCCAAGAAGAAGA GGAAGGTGGACCCCAAGAAGAAGAGGAAGGTGGACCCCAAGAAGAAGAGGAAGGTGTGA Plasmid5: HindIII-EGFP-NotI-Q108K:K40L:T51V:T53C:R58W:T29L:A33W: Q4F:L117E-EcoRI-NES(ELAEKLAGLDIN)-Stop-BamHI (pFlag-CMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA AGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACTTCAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTACATGAAGGCCCTGGATA TTGATTTTGCCCTGCGCAAGATTTGGGTACGTCTCACTCAGACGCTGGTTATTGATCAA 351 GATGGTGATAACTTCAAGGTAAAATGCACTAGCACATTCTGGAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC CGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTGTACGAGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGCGAATTCGAGCTTGCCGAGAAACTTG CCGGGCTTGACATAAATTGAGGATCC Plasmid6: XhoI-Q108K:K40L:T53S:V62N-BamHI-AgeI-ECFP-Stop-NotI (pECFP-N1) CTCGAGCTCGCCACCATGACGAGGGACCAGAATGGAACCTGGGAGATGGAGAGTAATGA AAACTTTGAGGGCTACATGAAGGCCCTGGATATTGATTTTGCCACCCGCAAGATTGCAG TACGTCTCACTCAGACGCTGGTTATTGATCAAGATGGTGATAACTTCAAGACAAAAAGC ACTAGCACATTCCGCAACTATGATAACGATTTCACTGTTGGAGTAGAGTTTGACGAGTA CACAAAGAGCCTGGATAACCGGCATGTTAAGGCACTGGTCACCTGGGAAGGTGATGTCC TTGTGTGTGTGCAAAAGGGGGAGAAGGAGAACCGCGGCTGGAAGAAGTGGATTGAGGGG GACAAGCTGTACCTGGAGCTGACCTGTGGTGACCAGGTGTGCCGTCAAGTGTTCAAAAA GAAGTCGAAGGATCCACCGGTCGCCACCATGGTGAGCAAGGGCGAGGAGCTGTTCACCG GGGTGGTGCCCATCCTGGTCGAGCTGGACGGCGACGTAAACGGCCACGAGCTGGACGGC GACGTAACGGCCACAGGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATGCCACCTACGGC AAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCCTGGCCCACCCT CGTGACCACCCTGACCTGGGGCGTGCAGTGCTTCAGCCGCTACCCCGACCACATGAAGC AGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGCGTACCATCTTC TTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAGGGCGACACCCT GGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAACATCCTGGGGC ACAAGCTGGAGTACAACTACATCAGCCACAACGTCTATATCACCGCCGACAAGCAGAAG AACGGCATCAAGGCCCACTTCAAGATCCGCCACAACATCGAGGACGGCAGCGTGCAGCT CGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCTGCTGCCCGACA ACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGAAGCGCGATCAC ATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATGGACGAGCTGTA CAAGTAAAGCGGCCGC Plasmid7: HindIII-EGFP-NotI-Q108K:K40L:T51V:T53S:R58W:Y19W:L117E- BamHI-RVASL-Stop (pFlag-CMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA 352 AGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACCAGAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTGGATGAAGGCCCTGGATA TTGATTTTGCCACCCGCAAGATTGCAGTACGTCTCACTCAGACGCTGGTTATTGATCAA GATGGTGATAACTTCAAGGTAAAAAGCACTAGCACATTCTGGAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC CGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTGTACGAGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGCGGATCCCGGGTGGCATCCCTGTGA Plasmid8: HindIII-EGFP-NotI-Q108K:K40L:T51V:T53S:R58W:Y19W:L117E- EcoRI-3NLS-Stop (pFlag-CMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA AGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACCAGAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTGGATGAAGGCCCTGGATA TTGATTTTGCCACCCGCAAGATTGCAGTACGTCTCACTCAGACGCTGGTTATTGATCAA GATGGTGATAACTTCAAGGTAAAAAGCACTAGCACATTCTGGAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC CGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTGTACGAGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGCGAATTCGCTGACCCCAAGAAGAAGA GGAAGGTGGACCCCAAGAAGAAGAGGAAGGTGGACCCCAAGAAGAAGAGGAAGGTGTGA Plasmid9: HindIII-EGFP-NotI-Q108K:K40L:T51V:T53S:R58W:Y19W:L117E- EcoRI-NES(ELAEKLAGLDIN)-Stop-BamHI (pFlag-CMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG 353 ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA AGCCCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACCAGAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTGGATGAAGGCCCTGGATA TTGATTTTGCCACCCGCAAGATTGCAGTACGTCTCACTCAGACGCTGGTTATTGATCAA GATGGTGATAACTTCAAGGTAAAAAGCACTAGCACATTCTGGAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC CGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTGTACGAGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGCGAATTCGAGCTTGCCGAGAAACTTG CCGGGCTTGACATAAATTGAGGATCC Plasmid10: HindIII-EGFP-NotI-Q108K:K40E:T53A:R58L:Q38F:Q4F-EcoRI- 3NLS-Stop (pFlag-CMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA AGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACTTCAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTACATGAAGGCCCTGGATA TTGATTTTGCCACCCGCAAGATTGCAGTACGTCTCACTTTTACGGAAGTTATTGATCAA GATGGTGATAACTTCAAGACAAAAGCCACTAGCACATTCCTGAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC CGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTGTACCTGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGCGAATTCGCTGACCCCAAGAAGAAGA GGAAGGTGGACCCCAAGAAGAAGAGGAAGGTGGACCCCAAGAAGAAGAGGAAGGTGTGA Plasmid11: HindIII-EGFP-NotI-Q108K:K40E:T53A:R58L:Q38F:Q4F-Stop- BamHI (pFlag-CMV2) AAGCTTATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGA GCTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATG CCACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCC TGGCCCACCCTCGTGACCACCCTGACCTACGGCGTGCAGTGCTTCAGCCGCTACCCCGA 354 CCACATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGC GCACCATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAG GGCGACACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAA CATCCTGGGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCG ACAAGCAGAAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGC AGCGTGCAGCTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCT GCTGCCCGACAACCACTACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGA AGCGCGATCACATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATG GACGAGCTGTACAAGTCCGGAGCCGCTGCAGGAGGCGGCCGCATGACGAGGGACTTCAA TGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGAGGGCTACATGAAGGCCCTGGATA TTGATTTTGCCACCCGCAAGATTGCAGTACGTCTCACTTTTACGGAAGTTATTGATCAA GATGGTGATAACTTCAAGACAAAAGCCACTAGCACATTCCTGAACTATGATGTGGATTT CACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAGCCTGGATAACCGGCATGTTAAGG CACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTGTGCAAAAGGGGGAGAAGGAGAAC CGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTGTACCTGGAGCTGACCTGTGGTGA CCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGAGGATCC Plasmid12: HindIII-Q108K:K40L:T51V:T53S:R58W:Y19W:L117D:Q4F-EcoRI3NLS-Stop (pFlag-CMV2) AAGCTTATGACGAGGGACTTCAATGGAACCTGGGAGATGGAGAGTAATGAAAACTTTGA GGGCTGGATGAAGGCCCTGGATATTGATTTTGCCACCCGCAAGATTGCAGTACGTCTCA CTCAGACGCTGGTTATTGATCAAGATGGTGATAACTTCAAGGTAAAAAGCACTAGCACA TTCTGGAACTATGATGTGGATTTCACTGTTGGAGTAGAGTTTGACGAGTACACAAAGAG CCTGGATAACCGGCATGTTAAGGCACTGGTCACCTGGGAAGGTGATGTCCTTGTGTGTG TGCAAAAGGGGGAGAAGGAGAACCGCGGCTGGAAGAAGTGGATTGAGGGGGACAAGCTG TACGACGAGCTGACCTGTGGTGACCAGGTGTGCCGTCAAGTGTTCAAAAAGAAGTGCGA ATTCGCTGACCCCAAGAAGAAGAGGAAGGTGGACCCCAAGAAGAAGAGGAAGGTGGACC CCAAGAAGAAGAGGAAGGTGTGA VII.9 Mammalian cell culture All cell lines (HeLa, U2OS and COS-7) were cultured in Dulbecco’s Modified Eagle medium (DMEM, supplemented with phenol red, 4.5 g/L Dglucose, L-glutamine and 110 mg/L Sodium Pyruvate; purchased from SigmaAldrich) supplemented with 10% (v/v) Fetal Bovine Serum (FBS, BioWest) and 1x Penicillin-Streptomycin-Glutamine (PSG; purchased from GIBCO) at 37 °C within a 5% CO2 and 10% O2 atmosphere. 355 For microscopic imaging, cells were seeded on an ibidi 1 μ-Slide 8 well ibiTreat plate. After approximately 12 h the cells were transiently transfected using Genjet Ver. II (purchased from SignaGen) according to the manufacturer’s protocol. Following transfection (after 48 h), the media was removed and the cells were incubated with a media containing the fluorophore. To prepare the fluorophore solution, a stock solution in DMSO (~0.003 M) was warmed to room temperature. The exact concentration of the fluorophore stock solution was determined using UV-Vis (using the associated extinction coefficient). The stock solution was then diluted to the specified concentration with pre-heated (37 °C) DMEM. After the indicated fluorophore incubation period, the cells were washed two times with Dulbecco’s Phosphate Buffered Saline (DPBS, supplemented with calcium chloride and magnesium chloride; purchased from Sigma Aldrich) and incubated in RPMI-1640 medium (without phenol red) for imaging immediately. VII.10 General confocal imaging methods Before imaging, the slide lid was replaced with a DIC lid for μ-dishes (purchased from Ibidi). Cell microscopy was performed using an inverted laser scanning confocal microscope (LSM510Meta, Carl Zeiss, Jena, Germany) equipped with diode, argon and HeNe lasers. A 40x oil-immersed objective was used. hCRBPII/MCRA was imaged with 594 nm excitation, 594 nm primary dichroic, 545 nm secondary dichroic and LP 615 nm emission. EGFP was imaged using 488 nm excitation, 488 nm primary dichroic, 545 nm secondary 356 dichroic and BP 505-530 nm emission. ECFP was imaged using 458 nm excitation, 458 nm primary dichroic, 515 nm secondary dichroic and BP 475 – 525 nm emission. hCRBPII/ThioFluor was imaged using multiple settings, as indicated in the main text: 1) 594 nm excitation, 594 nm primary dichroic, 545 nm secondary dichroic and LP 615 nm (or LP 650 nm) emission; 2) 633 nm excitation, 633 nm primary dichroic, 545 nm secondary dichroic and LP 650 nm emission; 3) 543 nm excitation, 543 nm primary dichroic, 545 nm secondary dichroic and LP 615 nm (or LP 650 nm) emission; 4) 514 nm excitation, 514 nm primary dichroic, 545 nm secondary dichroic and LP 615 nm (or LP 650 nm) emission; 5) 405 nm excitation, 405 nm primary dichroic, 545 nm secondary dichroic and LP 615 nm emission. DICII images were also collected. Kalman averaging 8 was applied in all confocal images. Fluorescence in each experiment was normalized to the same intensity adjusting the gain and amplifier offset. All images are pseudocolored. VII.11 Synthesis of chromophores VII.11.1 Synthesis of ThioFluor The synthesis of ThioFluor is shown in Scheme VII-1. The synthesis of ethyl phosphonate VII-1 was described previously. 2-Acetyl-5-bromothiophene, (CAS number 5370-25-2) and 4-(dimethylaminophenyl) boronic acid (CAS number 28611-39-4) were purchased from Oakwood Chemical. 357 O P EtO EtO 1. 2.0 equiv NaH, THF, 0 °C, 10 min 2. 2.0 equiv VII-1, 0 °C to RT, 25 min CN Br S 3. 1.0 equiv 2-acetyl-5-bromothiophene, RT, 18 h VII-1 Br CN VII-2 CN 1. 0.28 M PhCH3, 0.56 M MeOH 2. 2.0 equiv 4-(dimethylaminophenyl) boronic acid S CN S 3. 0.84 M Na2CO3 in H2O, 4. 0.05 equiv Pd(PPh3)4, 52 °C, 23 h VII-2 N VII-3 H CN S N VII-3 1. 1.5 equiv DIBAL-H, DCM, RT, 1.5 h 2. HPLC separation of cis and trans isomer O S N ThioFluor Scheme VII-1. Detailed synthesis of ThioFluor. Synthesis of compound VII-2: NaH (60% by weight, 0.38 g suspension containing 0.23 g NaH, 9.8 mmol) was placed in a dry round bottom flask and flushed with nitrogen. Anhydrous THF (40 mL) was added and the solution was cooled to 0 °C via an ice-water bath and stirred for 10 minutes. Subsequently, diethyl(cyanomethyl)phosphonate (1.7 g, 9.8 mmol) was added dropwise and stirred for 25 min, gradually warming to room temperature. 2-Acetyl-5bromothiophene (1.0 g, 4.9 mmol) in anhydrous THF (10 mL) was added dropwise. The reaction was kept under nitrogen and stirred for 18 h. Work up: Brine solution was added to the reaction mixture and the resulting aqueous layer was extracted three times with diethyl ether (3 x 30 mL). The combined organic phase was dried over anhydrous Na2SO4. The solvent was then removed under reduced pressure to yield a light brown solid. The crude product was purified by silica gel column chromatography with 20% ethyl 358 acetate/hexane mixture to yield VII-2 and the starting 2-acetyl-5-bromothiophene. This product was carried to the next step. Synthesis of compound VII-3: Toluene (7 mL) and methanol (3.6 mL) were added to a dry round bottom flask. The mixture was dearated by sonication under a nitrogen balloon. 4-(dimethylaminophenyl) boronic acid (0.33 g, 2.0 mmol) and VII-2 (0.23 g, 1.0 mmol) were then added and the solution was dearated again. Subsequently, sodium carbonate (0.32 g in 3.6 DI water) was added. After dearating the solution, tetrakis(triphenylphosphine) palladium (58 mg, 0.05 mmol) was added. The mixture was then heated at 52 °C under nitrogen for 23 h. Work up: The solution was extracted three times with dichloromethane (3 x 30 mL) and the combined organic phase was dried over anhydrous Na2SO4. The solvent was then removed under reduced pressure. The crude product was purified by silica gel column chromatography with a gradient from 100% hexane to 30% ethyl acetate/hexane. The resulting product was impure and was carried to the next step. Synthesis of ThioFluor: Compound VII-3 (0.2 g) was added to a dry round bottom flask, flushed with nitrogen and dissolved in anhydrous dichloromethane (30 mL). Subsequently, a 1 M solution of DIBAL-H in THF (1.5 mL, 1.5 mmol) was added dropwise. The reaction was stirred at room temperature, under nitrogen, for 1.5 h. Work up: The reaction was quenched by slowly adding five drops of water, followed by five drops of 10% HCl, followed by stirring for 15 min. An 359 additional 20 mL of brine solution was added and the aqueous layer extracted with dichloromethane (3 x 40 mL). The resulting organic layer was dried over anhydrous Na2SO4 and the solvent was removed under reduced pressure. The crude product was initially purified by silica gel column chromatography using 10% ethyl acetate/hexane to 20% ethyl acetate/hexane. The impure product was combined and a second column was run with 50% dichloromethane/hexane to 30% ethyl acetate/hexane. The crude product after the second column was dissolved in 70% ethyl acetate/hexane and the solid was filtered through cotton. HPLC was used to purify the soluble portion. The sample was separated by normal-phase HPLC (silica column, Zorbax Rx-SIL, 9.4 mm x 25 cm) after manual injection. The sample was eluted with 20% ethyl acetate/hexane at 3 mL min-1. The product was detected at 400 nm. The overall yield of the trans isomer of ThioFluor was approximately 4% (red solid). 1 H NMR (500 MHz, CDCl3): δ 10.10 (d, J = 8.0 Hz, 1H), 7.51 (d, J = 8.5 Hz, 2H), 7.38 (d, J = 4.0 Hz, 1H), 7.15 (d, J = 3.5 Hz, 1H), 6.71 (d, J = 9.0 Hz, 2H), 6.43 (d, J = 7.5 Hz, 1H), 3.01 (s, 6H), 2.56 (s, 3H) 13 C NMR (125 MHz, CDCl3): δ 190.51, 150.61, 150.13, 149.62, 140.78, 129.62, 127.02, 123.18, 121.79, 121.46, 112.23, 40.30, 15.63 HRMS (ES+): m/z calcd for C16H17NOS: 272.1109, found: 272.1069 [M+H]+. VII.11.2 Synthesis of ThioFluor-2 The synthesis of ThioFluor-2 is shown in Scheme VII-2. The synthesis of ethyl phosphonate VII-1 was described previously. 2-Acetyl-5-bromothiophene, (CAS number 5370-25-2) was purchased from Oakwood Chemical. 360 Br 3.0 equiv NH(Me)2, H2O, 120 °C, Sealed Tube, 16 h S S N O O VII-4 O O P O 1. 2.0 equiv NaH, THF, 0 °C, 5 min 2. 2.0 equiv VII-4, 0 °C to RT, 18 h N NC VII-1 N CN VII-5 1. 1.5 equiv DIBAL-H, DCM, RT, 3 h S S CN 2. HPLC separation of cis and trans isomer VII-5 N S CHO ThioFluor-2 Scheme VII-2. Detailed synthesis of ThioFluor-2. Synthesis of compound VII-4: 2-Acetyl-5-bromothiophene (1.0 g, 5.0 mmol) was added to a sealed tube. Then N,N-dimethyl amine (40%, 2.0 mL, 15.0 mmol) and DI water (5 mL) were added. Subsequently, the reaction mixture was heated at reflux for 16 h. Work up: Water was added to the reaction mixture and the resulting aqueous layer was extracted three times with dichloromethane (3 x 30 mL). The combined organic phase was dried over anhydrous Na2SO4. The solvent was then removed under reduced pressure and the crude product was carried to the next step. Synthesis of compound VII-5: NaH (60% by weight, 0.40 g suspension containing 0.24 g NaH, 10.0 mmol) was placed in a dry round bottom flask and flushed with nitrogen. Anhydrous THF (25 mL) was added and then the solution was cooled to 0 °C via an ice-water bath and stirred for five min. Subsequently, 361 diethyl(cyanomethyl)phosphonate (1.8 g, 10.0 mmol) was added dropwise and stirred for 10 min, gradually warming to room temperature. Compound VII-4 (0.85 g, 5.0 mmol) in 6 mL anhydrous THF was added dropwise. The reaction was kept under nitrogen and stirred for 18 h. Work up: Water was added to the reaction mixture and the resulting aqueous layer was extracted three times with diethyl ether (3 x 40 mL). The combined organic phase was dried over anhydrous Na2SO4. The solvent was then removed under reduced pressure to yield a yellow-orange solid. The crude product was purified by silica gel column chromatography with 20% ethyl acetate/hexane mixture to yield VII-5 and the starting 2-acetyl-5-bromothiophene. This product was carried to the next step. Synthesis of ThioFluor-2: Compound VII-5 (0.5 g) was added to a dry round bottom flask, flushed with nitrogen and dissolved in anhydrous dichloromethane (30 mL). Subsequently, a 1 M solution of DIBAL-H in THF (5.2 mL, 5.2 mmol) was added dropwise. The reaction was stirred at room temperature, under nitrogen, for 3 h. An additional 2 mL of 1 M DIBAL-H in THF was added and the reaction was stirred for 1.5 h. Work up: The reaction was quenched by slowly adding 1 mL of water. The reaction was stirred vigorously for five min. Next, 3 mL of 10% HCl was added, stirred for five min, and an additional 20 mL of brine solution was added and the aqueous layer was extracted with dichloromethane (3 x 40 mL). The resulting organic layer was dried over anhydrous Na2SO4 and the solvent was 362 removed under reduced pressure. The crude product was initially purified by silica gel column chromatography using 10% ethyl acetate/hexane to 30% ethyl acetate/hexane (using a gradient of 10% ethyl acetate/hexane). The crude product after the second column was dissolved in 80% ethyl acetate/hexane and the solid was filtered through cotton. HPLC was used to purify the soluble portion. The sample was separated by normal-phase HPLC (silica column, Zorbax RxSIL, 9.4 mm x 25 cm) after manual injection. The sample was eluted with 30% ethyl acetate/hexane at 3 mL min-1. The product was detected at 420 nm. The overall yield of the trans isomer of ThioFluor-2 was 6% (brown solid). 1 H NMR (500 MHz, CDCl3): δ 9.95 (d, J = 8.5 Hz, 1H), 7.25 (d, J = 4.5 Hz, 1H), 6.12 (d, J = 7.5 Hz, 1H), 5.85 (d, J = 4.5 Hz, 1H), 3.04 (s, 6H), 2.47 (s, 3H) 13 C NMR (125 MHz, CDCl3): δ 189.72, 163.04, 150.98, 131.62, 126.50, 119.12, 103.3, 42.08, 14.82 HRMS (ES+): m/z calcd for C10H13NOS: 196.0796, found: 196.0749 [M+H]+. 363 REFERENCES 364 REFERENCES 1. Wang, W. in PhD Thesis, Michigan State University (2012). 2. Wang, W., Nossoni, Z., Berbasova, T., Watson, C.T., Yapici, I., Lee, K.S., Vasileiou, C., Geiger, J.H. & Borhan, B. Tuning the electronic absorption of protein-embedded all-trans-retinal. Science 338, 1340-1343 (2012). 3. Gill, S.C. & von Hippel, P.H. Calculation of protein extinction coefficients from amino acid sequence data. Analytical Biochemistry 182, 319-326 (1989). 4. Yapici, I., Lee, K.S.S., Berbasova, T., Nosrati, M., Jia, X.F., Vasileiou, C., Wang, W.J., Santos, E.M., Geiger, J.H. & Borhan, B. "Turn-On" Protein Fluorescence: In Situ Formation of Cyanine Dyes. Journal of the American Chemical Society 137, 1073-1080 (2015). 5. Yapici, I. in PhD Thesis, Michigan State University (2015). 6. Berbasova, T. in PhD Thesis, Michigan State University (2014). 365