DEVELOPMENT OF PHOTOSWITCHABLE RHODOPSIN MIMICS : SPECTROSCOPY , STRUCTURAL STUDIES AND WAVELENGTH REGULATION STUDIES OF THESE SYSTEMS By Meisam Nosrati A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Chemistry Ð Doctor of Philosophy 2015 ABSTRACT DEVELOPMENT OF PHOTOSWITCHABLE RHODOPSIN MIMICS: SPECTROSCOPY, STRUCTURAL STUDIES AND WAVELENGTH REGULATION STUDIES OF THESE SYSTEMS By Meisam Nosrati Retinal photo pigments ha ve been an essential part of life on the planet from bacteria to mammals. Photo isomerization of the retinal is the heart of the function of all of these pigments. Vision and circadian rhythm in hig her organisms and energy production and phototaxis in bacteria are all mediated by these light absorbing proteins. Retinal isomerization has never been recapitulated in an artificial retinal protein pigment before . Reengineering of human Cellular Retinoic Acid Binding Prot einII (hCRABPII) into a pho to switchable retinal pigment provided a light absorbing pigment with unique properties. The reengineered protein can isomerize thermodynamically and photochemically from a 15 -cis retinal to 15 -trans retinal and vice versa. The isomerization is accompanied by a 3 -5 unit of pKa change from above physiological pH to lower similar to other retinal pigments. Isomerization of retinal is quantitative both in solution and more importantly in the crystalline form. High-resolution X-ray crystal structures provide a crystal clear image for the isomerization of the retinal in this system. The quantitative isomerization of the retinal in the crystalline form can be a valuable tool in characterizing the intermediates of retinal isomerizati on. This protein in combination with fluorescent proteins can be used as a fluorescent quencher to develop a Reversibly Swi tchable Fluorescent Protein (RSFP). Wavelength regulation of rhodopsins has been a long standing question with no solid answer that has been debated for five decades . Point charges from charged residues and the dipole moment from the polar residues especially the hydroxyl containing side chains and the conformational changes in the retinal chromophor e have been suggested as the main theories for wavelength regulation. However, t he effect of the polar water molecules in wavelength regulation of retinal pigments has been less investigated. Using a reengineered human Cellular Retinol Binding ProteinII (h CRBPII), we were able to systematically demonstrate a coherent effect of the polar water molecules in the wavelength regula tion of hCRBPII retinal protein complexes , which can be attributed to other known retinal light absorbing pigments as well. Mutating the residues interacting with these water molecules, which are located close to the retinal polyene from the beginning to the end of it , regulate the absorption of the protein. !"#!ACKNOWLEDGEMENTS I am grateful and indebted to all the people that had a positive impact on my life and changed it for good and brought hope into it. I would not have been able to progress and move forward without their help and support and be where I am. They are: My father for always caring and checking on my education and providing a peaceful environment to focus on my studies and being the most unmaterialistic person that I have ever met in my entire life and I am sure I will never find anyone like him being thi s much generous and giving, M y mother for being the simplest woman in the world with least imaginable expectations from life who I always felt comfortable to talk about literally everything going on in my life, M y brothers Emran and Ehsan and my sister Fatemeh, my sister -in-law Nahid, brother -in-law Mohammad Reza and my niece Bahar who has brought a lot of joy to our lives. They have been all always there for me and gave me their unconditional support at all times. I have had good teachers and professors that made me interested in science and I enjoyed their mentorship. My English language teachers, in guidance school Mr.Moalla and in high school, Mr.Memarzadeh and Mr.Khalilnejad, my high school chemistry teacher, Mr.Taheri, and my god blessed b iology teacher, Mr.Hasani, were all the people that I was blessed to meet and be their student in my hometown of Damghan Iran. In University of Tehran during my bachelors and masters Dr.Adib played the role in making me interested in Organic Chemistry which still is the most appealing part of the chemistry for myself afterwards till now. !"!Coming to the United States and adopting a new life probably was the most challenging transition, and life far away from home and family especially in the first two years w as a big challenge. I am thankful to Mehrdad Mahdavi, Faramarz Vafaee, Reza Sheikhi and Roozbeh Yousefi that I have made long lasting friendships with. I have always turned to Mehrdad and Faramarz in hard times in here, and they have been always helpful to me and put time for me. They have not been less than brothers to me, and I am lucky to have them in my life. I should especially thank professor Borhan for an initial belief in me and choosing me over a wide number of applicants to come to Michigan State University for PhD studies and his scienti fic mentorship during my PhD, his meticulous point of views in science and perfectionism. I have enjoyed being close to the same type of scientists in his lab like Dr.Chrysoula Vasileiou, Dr.Tetyana Berbasova, Dr.W enjing Wang and Dr.Roozbeh Yousefi. Our collaborative work especially with Dr.Berbasova and then Dr.Yapici helped us to build an interactive and fruitful research environment, which I hope to find or build in a new environment. Also I feel now that it was unique to that time. I should give a special thank to Dr.Vasileiou for her extensive help to me and everyone else in the lab. She always had time for everyone and the door of her office was always open to the people of the lab. The most influential person however, was my PhD academic advisor prof.James H. Geiger. He gives a lot of freedom in his research lab and accepts ideas generously that is essential for making an independent thinker. Also, we started with a regular student advisor relationship, but ov ertime I found him the best person to get not only advise for my research, but for my !"#!personal matters as well where I felt the lack of experience to approach them. I admire his stability and great deal of experience in life, his humble, social, simple and happy personality with an anger that subsides quickly and follows sympathy. Jim h as been always the person that made me feel confident and inspired me to believe in myself. I also enjoyed being labmates with Dr.Susan Strom, Dr.Remie Fawaz, Dr.Rafida Nosso ni, Dr.Camille Watson, Lindsey, Janice, Zahra, Hadi, Hasan and Alireza in prof.Geiger lab. Also, we were a small lab all the time, but I found it easy to find peace, focus and relax. I am thankful to chemistry department staff at Michigan State University that helped to get my research, the research labs and equipment going. I had the privilege to be one of the very few Iranians with the permission to use the Advance Photon Source of Argonne National Lab at Chicago on -site. I made a lot of trips to there wi th my labmates and collected data with the help of super friendly staff of the LS -CAT section 21 at Advanced Photon Source. I am also grateful to anyone else who has helped me during this time and my mind does not help me to remember them and put their nam e here as person or part of a group. !"##!TABLE OF CONTENTS LIST OF TABLES .............................................................................................. ix LIST OF FIGURES ............................................................................................ xi KEY TO SYMBOL AND ABBREVIATIONS ................................................... xxii Chapter I: Photoswitchable protei ns: Biological Significance and Applications ........................................................................................ 1 I.1 Retinal Based Photoswitchable Proteins ..................................................... 1 I.2 Photo cascade of Rhodopsin ....................................................................... 4 I.3 Mechanism of Rhodopsin Signaling ............................................................. 9 I.4 Photoisomerization in Bacterial Rhodopsins .............................................. 11 I.5 Photocycle of Bacteriorhodopsin ............................................................... 13 I.6 Reversibly Switchable fluorescent Proteins (RSFPs) ................................ 16 I.7 Optoch emical Genetics .............................................................................. 19 I.8 Photoactive Yellow P rotein ........................................................................ 20 I.9 A photo switchable rhodopsin mimic .......................................................... 21 REFERENCES ................................................................................................ 25 Chapter II: Development of a Photoswitchable engineered human Cellular Retinic Acid Binding ProteinII .............................. 33 II.1 Introduction ............................................................................................... 33 II.2 Developing of hCRABPII to a Photoswitchable Rhodopsin mimic ............ 39 II.3 Isomerization of hCRABPII mutants in solution ........................................ 56 II.4 The mystery of colorful crystals of M2 mutant after PSB loss ................... 60 II.5 hCRBPII isomerization solution and structural studies ............................. 64 II.6 Conclusion ................................................................................................ 71 REF ERENCES ................................................................................................ 72 Chapter III : Theories of Wavelength Regulation ........................... 75 III.1 Introduction .............................................................................................. 75 III.2 The Point Charge Theory ......................................................................... 76 III.3 The Conformational Theory ..................................................................... 88 III.4 Water Molecules in Wavelength R egulation ............................................ 92 REFERENCES ................................................................................................ 98 Chapter IV : Water networks in the wavelength regulation of human Cellular Retinol Binding Protein II reengineered Protein ........................................................................................................ 105 IV.1 Important waters in the binding pocket of hCRBPII protein ................... 105 IV.2 Gln4 water molecule .............................................................................. 108 IV.3 Thr51 and Thr53 water molecule ........................................................... 109 IV.4 Gln38 and Gln128 water molecule s ...................................................... 110 IV.5 Arg58 and Ala33 mutations ................................................................... 115 !"###!IV.6 Counter anion in the hCRBPII system ................................................... 118 IV.7 Crystal growth for Neutron Diffraction ................................................... 120 IV.7 -1 Q108K:K40L mutant of hCRBPII ........................................................ 121 IV.7 -2 Q108K:K40L:T51V:R58F mutant of hCRBPII .................................... 126 IV.7 -3 Q108K:K40L:T51V:R58Y:Y19W mutant of hCRBPII ......................... 126 IV.7 -4 hCRBPII Expression in Minimal Media ............................................... 127 IV.8 Wavelength Regulation in second generation hCRABPII mutants ........ 128 IV.8 -1 R59W and A32W mutations ............................................................... 128 IV.8 -2 Y134F and T54V mutations ................................................................ 130 IV.8 -3 P39Q, R132Q and F3Q mutations ..................................................... 133 IV.9 Conclusion ............................................................................................. 135 IV.10 Other Structural Studies of hCRABPII and hCRBPII Reengineering .. 136 REFERENCES .............................................................................................. 139 Chapter V: Stabilizing the interaction between Cellular Retinoic Acid Binding Protein II and Retinoic Acid Receptor 142 V.1 Introduction ............................................................................................. 142 V.2 Synthesis of CD -270 ............................................................................... 146 V.3 Conclusion .............................................................................................. 151 REFERENCES .............................................................................................. 152 Chapter VI: Materials and Methods ................................................. 156 VI.1 Site -directed mutagenesis ..................................................................... 156 VI. 2 Protein Expression and Purification of hCRABPII and hCRBPII/pET -17b Mutants ......................................................................................................... 158 VI. 3 Extinction Coefficient Determination ...................................................... 160 VI. 4 UV-vis Measurements ........................................................................... 160 VI. 5 pKa determinations ................................................................................ 160 VI.8 Protein Crystallization of hCRABPII ...................................................... 171 VI. 8-1 R!111K:P39Y:R132Q:Y134F:T54V:R59Y .......................................... 171 VI. 8-2 R!111K:P39Q:R132Q:Y134F:T54V:R59Y .......................................... 172 VI.8 -3 R111K:P39Q:R132Q:Y134F:T54V:R59Y:A32Y:F3 Q ......................... 172 VI.8 -4 R!111K:P39Q:R132Q:Y134F:T54V:R59Y:A32W:F3Q ....................... 173 VI.9 Protein crystallization of hCRBPII .......................................................... 173 VI. 10 Data Collection and Refinement .......................................................... 174 VI. 11 CD-270 Synthesis ................................................................................ 183 VI.11 -1 C-H activation reaction ..................................................................... 183 VI.11 -2 SNAr reaction .................................................................................... 184 VI.11 -3 Protection Reaction .......................................................................... 185 REFERENCES .............................................................................................. 186 !"#!LIST OF TABLES Table II-1. Summary of the mutations on R59 position of fir st generation of hCRABPII mutant s .......................................................................................... 35 Table II-2. Summary of the mutations on R59 position of first generation of hCRABPII mutants .......................................................................................... 37 Table II-3. Mutants of second generation hCRABPII with pKas below physiological pH .............................................................................................. 38 Table II-4. Different mutant of hCRBPII with high pK as .................................. 40 Table II-5. Scree ning for hCRABPII mutants with fast kinetic PSB formation and slow PSB lo ss .......................................................................................... 45 Table IV-1. Gln4 mutation summary ............................................................. 108 Table IV-2. T51V mutation causes 25 -30 nm red shift ................................. 110 Table IV -3. Gln38 and Gln128 mutation on the shorter hCRBPII mutatnts .. 112 Table IV -4. The effect of the Gln38 and Gln128 mutations on hCRBPII longer series proteins ............................................................................................... 113 Table IV-5. The effect of R58W and A33W on red shifting the hCRBPII mutants ......................................................................................................... 115 Table IV -6. Protein shift with reference to Q108K:K40L, R58W enhancement effect ............................................................................................................. 116 Table IV -7. R58F mutation enhances the effect of the internal mutations by a factor of 1.5 -2 ................................................................................................ 116 Table IV-8. R59W and A32W mutations contribute significantly to the wavelength regulation of hCRABPII .............................................................. 129 Table IV -9. The effect of the R132Q:P39Q mutation on the wavelength absorbed by the hCRABPII mutants ............................................................. 133 Table IV -10. The effect of the F3Q mutation on the wavelength absorbed by the hCRABPII mutants .................................................................................. 135 Table VI -1. PCR protocol for hCRABPII and hCRBPII mutagenesis ............ 156! !"!Table VI -2. Anion Exchange purification protocol for h -CRABPII and h-CRBPII ...................................................................................................................... 160 Table VI -3. hCRABPII mutants ..................................................................... 163 Table VI -4. hCRBPII Proteins Extinction Coefficient .................................... 168 Table VI -5. hCRBPII mutants ....................................................................... 169 Table VI -6. The hCRABPII crystal table ....................................................... 176 Table VI -7. The hCRBPII crystal table .......................................................... 179 !"#!LIST OF FIGURES Figure I-1) a) Eye the gate way to vision b) A fundus photo of the Left eye showing the optic nerve, blood vessels an macula c) Different compartments of the eye and the cellular structure of the macula . 1 Figure I -2) a) The structure of a Rod cell. b) Comparison of the characteristics of Rod and Cone cells .................................................................... 2 Figure I-3) The structure of a rod cell, rhodopsin and 11 -cis retinal ................. 3 Figure I -4) Formation of a retinal iminium PSB is essential to light absorbtion of rhodopsin ................................................................................... 3 Figure I -5) Different rhodopsins absorb a photon of light with the !max highlighted to isomerize from 11 -cis to all-trans retinal .................... 4 Figure I-6) Photo cycle of the Rhodopsin .......................................................... 5 Figure I-7) a) Structural differences between the dark state (Green) and excited state (Cyan) of Rhodopsin. Red = TM3, Black = TM5 and Blue = TM6 in the ground state. b) C -terminus peptide of the G !-subunit bound to MetaII -Rhodopsin. c ) Trp265 positioning on TM6 in the ground state and excited state as retinal isomerizes .............. 7 Figure I -8) The retinal conformation in Rhodopsin: Ground state (Red - PDB ID: 1U19), Batho (Gray Ð PDB ID: 2G87), Lumi (White Ð PDB ID: 2HPY) and MetaII (Yellow Ð PDB ID: 3PXO). The arrows show the rotation of the retinal from Lumi to MetaII .............................................. 9 Figure I-9) Rhodopsin signaling ..................................................................... 10 Figure I -10) a) Bacteriorhodopsin in the ground state of all -trans retinal (Left - PDB ID: 1C3W) and excited state of 13 -cis retinal (Right - PDB ID: 1IXF). b) Chemical diagram of the isomerization of the bacterial rhodopsins with the !max of each indicated ........................ 11 Figure I -11) a) Incorporation of Channel Rhodopsin (ChR) or Halo Rhodopsin (HR) can confer the sensitivity of light to neurons. b) The application of optogenetic tools in mice helps to understand the function of different parts of the brain by spatial and temporal act ivation of neurons .................................................................... 12 Figure I-12) Photo cycle of the Microbial rhodopsins ...................................... 13 Figure I -13) The overlay structure of the bacteriorhodopsin in the: Ground state (Green Ð PDB ID: 1C3W), light adapted O -intermediate (Cyan Ð PDB ID: 3VHZ) and dark adapted O -intermediate !"## !(Magenta Ð PDB ID: 3VI0) shows that the light adapted form can reproduce the grou nd state geometry, but not the dark adapted crystal ........................................................................................... 15 Figure I -14) Conventional immunofluorescence imaging of microtubules (a and c) and STORM images which uses RSFPs (b and d) .......... 16 Figure I -15) The mechanism of the switch in RSFPs is a cis -trans isomerization of the fluorophoric ligand ...................................... 17 Figure I -16) The loss of a water molecule attributes to the excitation decoupled on -off switch of the fluorescent protein ..................... 18 Figure I -17) a) Photochromic ligands (PTLs) in both form s can act as a ligand for pr otein, but just one of the two forms activates the protein. b) Photo switchable Tethered Ligand (PTL) is covalently bound to the protein, and just one form fits into the binding pocket of the protein ......................................................................................... 19 Figure I -18) Photo isomerization of PYP. Overlay str ucture of the Ground State (Cyan, PDB ID: 2ZOH) and Excited state (Yellow, PDB ID: 4BBV) ......................................................................................... 21 Figure I -19) Formation of a Retinal Schiff Base and Protonated Schiff Base. It is the PSB that absorbs in the visible region of light ................... 22 Figure I -20) human Cellular Retinoic Acid Binding ProteinII (hCRABPII) bound to Retinoic Acid (Left -PDB ID: 2FR3 ). Human Cellular Retinol Binding ProteinII (hCRBPII) bound to retinol (Right -PDB ID: 4QYN) ................................................................................... 23 Figure I -21) a) Reengineered hCRBPII protein retinal pigments absorbing light from 425nm to 644nm. Incubated at pH = 7.3. b) Reengineered hCRABPII protein retinal pigments absorbing light from 482nm to 630nm ................................................................. 23 Figure II-1) Chemical structure of d ifferent r etinoids ...................................... 33 Figure II-2) a) Crystal structure of hCRABPII bound to Retinoic Acid. b) Binding pocket of hCRABPII and interacting residues with the Retinoic Acid (PDB ID: 2FR3) ..................................................... 34 Figure II -3) a) Crystal structure of the R132K:R111L:L121E mutant of hCRABPII (PDB ID: 2G7B) shows the retinal covalently attached to t he protein. b) Interacting residues with the i minium Schiff Base in this crystal structure ....................................................... 35 Figure II -4) Crystal structure of R132K:R111L:L121E mutant, shows the ionone ring is positioned out of th e binding pocket (PDB ID: 2G7B) ......................................................................................... 36 !"### ! Figure II-5) a) Chemical representation of retinal C15 -analog compared to retinal. b) Overlay structure of R132K:R111L:L121E -retinal complex with R132K:R111L:L121E:R59W -C15 retinal complex 37 Figure II -6) a) Overlay crystal structures of R111K:R132L:Y134F:T54V:R59W and R132K:R111L:L121E shows the retinal structure deeper inside the binding pocket of hCRA BPII. b) Different mutants of second generation hCRABPII absorb from 474 to 640 ............... 38 Figure II -7) PSB loss of the incubated R111K:R132L:Y134F:T54V:R59W mutant of hCRABP II with retinal at pH=5.2 ................................. 39 Figure II -8) a) Retinol inside the binding pocket of wt -hCRBPII (PDB ID: 4QYN). b) Retinal PSB covalently bound inside the binding pocket of Q108K:K40L mutant of hCRBPII (PDB ID: 4RUU) ..... 39 Figure II-9) The structures of the hCRABPII mutant R111K:R132L:Y134F: T54V:R59W (Yellow, PDB ID: 4I9S) and the hCRBPII mutant Q108K:K40L (Cyan, PDB ID: 4RUU) were overlaid and the region around the retinal binding site is shown. Note the three glutamines in hCRBPII are substituted with hydr ophobic residues in hCRABPII ................................................................................ 40 Figure II -10) pKa t itration of M1 mutant in citrat e buffer at pH=5. Protein concentration is 50µM incubated with 0.5 equiv alent retinal ...... 41 Figure II -11) a) Base titration of kinetic product b) acid titration of thermodynamic product of R111K:R132L:Y134F:T54V:R59W mutant of hCRABPII ................................................................... 42 Figure II -12) a) a chemical representation of 15 -cis retinal (left) and 15 -trans retinal (right) b) Retinal density in cr ystal structure of Q108K:K40L:T51V:R58Y:Y19W mutant of hCRBPII shows a cis -PSB density for retinal c) Retinal density in crystal structure of Q108K:K40L:T51V:T53C:R58W:T29L:Y19W:Q4H mutant of hCRBPII shows a trans -SB density for retinal. d) Retinal density in crystal structure of R111K:R132L:Y134F:T54V:R59W:A32W mutant of hCRABPII shows both 15 -cis retinal (left) and 15 -trans retinal (right) which can be fit into the density of retinal ( all the maps are contoured at 1! ) ........................................................ 43 Figure II -13) Thermal interconversion in M1 a) UV/Vis spectra of the M1/retinal complex 3.5 hours and 24 hours after addition of retinal to the protein. b) Retinal density in the crystal structure of R111K:R134F:T54V:R132Q:P39Q:R59Y at 1.83 † contoured at 1! shows a trans-SB (PDB ID: 4YBP) ........................................ 44 Figure II -14) Thermal isomeri zation of a CRABPII rhodopsin mimic. a) PSB loss over time for mutant M2 . M2 was incubated with retinal, and !"#$ !UV/Vis spectra were taken at various time points after the addition; b) Crystal structure of M2 showing the trans -SB at 1.62 †, with the electron density contoured at 2.0 !, (Yellow, protein was incubated with 4 equiv retinal for 24 hours at room temperature before crystallization, PDB ID: 4YFQ ), overlaid on the cis -PSB of the same mutant crystallized after a 20 minute retinal incubation (Cyan) ; c) Crystal structure of the M2 mutant in cis -PSB form at 1.97 † at 1.5 !, (Cyan, the protein was incubated with 4 equivalents retinal at room temperature for 20 minutes and followed by immediate crystallization , PDB ID: 4YFP ) overlaid on the trans -SB crystal structure of the same mutant (Y ellow) ........................................................................... 46 Figure II -15) Cycling of the hCRABPII proteins in solution with UV light (340 -380 nm) and Green light (520 -600 nm) ...................................... 48 Figure II -16) Photo interconversion in both solut ion and crystal. a) UV irradiation of M2 after complete PSB loss shows the PSB recovery in solution ; b) Cycles of UV and visible light irradiation of M2 crystals . (Each Cycle is 5 min UV light and 5 min room light at pH 7.5); c) The structure of M2 obtained from a crystal that was UV irradiated for 30 minutes at 1.95 † . The electron density was contoured at 1.5 ! and the structure clearly shows the retinal in a cis -iminium conformation (Magenta, PDB ID: 4YFR ) overlaid on the structure of the trans -SB (Yellow, PDB ID: 4YFQ ) and cis -PSB (Cyan, PDB ID: 4YFP ) crystal structure s ................................. 49! Figure II -17) Cycling of the M2 mutant crystals: a) Overlay of the retinal for the first three cycles of UV light irradiated crystals of the M2 mutant for 30 minutes. b) Overlay of the retinal for the first three cycles of visible light irradiated crystals of the M2 mu tant for 30 minutes. c) Overlay of the retinal for the first three cycles of UV light irradiated crystals (cyan) and visible light irradiated crystals (yellow) of the M2 mutant each for 30 minutes ........................... 50 Figure II -18) The environment surrounding the iminium i n M2 protein. The trans -SB (yellow) and cis -PSB (cyan) ......................................... 51 Figure II -19) Thermodynamic and light induced cis , trans -retinal iminium isomerization in hCRABP II protein ............................................. 52 Figure II -20) M2 crystal structure indicates the difference in the two forms of the protein. a) a "-cation interaction between Trp109 and the PSB nitrogen of the cis -PSB contributes to the higher pKa in this form. b) Aliphatic hydrophobic residues surrounding the trans -SB in M2 depress the pKa in this form ............................................. 53 Figure II -21) Crystal structure of Q108K:K40L shows that Gln4 contributes to the stabilization of a cis -PSB in the hCRBPII mutants (PDB ID: 4RUU) ......................................................................................... 54 !"#! Figure II -22) Mutants M2 and M3 are shined with green light till maximum PSB loss after maximum PSB formation. PSB recovery has been tracked in dark. In 1 hour M3 mutant has significant PSB recovery, while M2 mutant spectrum shows much less PSB recovery ...................................................................................... 55 Figure II-23) PSB loss of M3 mutant over time ............................................. 55 Figure II -24) Structures of retinal bound M3 (PDB ID: 4YKM) and cis -PSB retinal bound M2 are overlaid, showing the effect of the Gln 3 on the environment about the cis -PSB, increasing the pK a by 1 unit and altering the nature of the thermodynamic p roduct ............... 56 Figure II -25) Cycling of the mutant M2 after PSB loss completion. a) After the first cycle the yield of the switch drops considerably . b) Cycling with room light for visible light irradiat ion of the solution. c) Cycling in dark with no green light involved ................................ 58 Figure II-26) Cycling of M2 mutant at pH 5.5 ................................................. 59 Figure II -27) Cycling of the mutant M3 in solution. More than 50% of the PSB absorption is recovered in each cycle ......................................... 60 Figure II -28) a) M2 SB -crystal grown at pH 5 b) M2 SB crystal grown at pH 6.5 c) M2 SB -Soaked crystal to pH 7.5 d) M2 -UV irradiated crystal after soak for 30 minutes ................................................. 61 Figure II-29) pKa titration of M2 protein in 50 mM Malonate .......................... 61 Figure II -30) M2 protein was incuba ted with 4 equivalents retinal and then mixed with the cry stallization condition . The protein was allowed to sit in the cold room for 24h ..................................................... 62 Figure II -31) a) The UV -Vis spectrum of the soaked crystal of M2 mutant b) The UV -Vis Spectrum of the UV irradiated crystal for 30 minues after soak .................................................................................... 63 Figure II-32) Crystal UV -Vis spectrum of mutans: a) M3 b) M11 ................... 63 Figure II -33) Retinal incubation of the hCRBPII mutants: Left: Q108K:K40L , Right: Q108K:K40L:T51V:T53C:R58W:T29L:Y19W:Q4H .......... 64 Figure II-34) UV-Green light cycling of hCRBPII mutants for: a) Q108K:K40L: T51V:T53C:R58W :T29L:Y19W:Q4H b) Q108K:K40L:T51V:T53C :R58W: T29L:Y19W:Q4A c) Q108K:K40L:T51V:R58F. With Dr. Tetyana BerbosvaÕs permission .................................................. 66 Figure II -35) Retinal incubation followed by green light irradiation for: a) Q108K:K40L:T51V:R58F mutant of hCRBPII and b) Q108K !"#$ !:K40L:T51V:T53C:R58W:T29L:Y19W mutant of hCRB PII. c) Electron density map of the crystal obtained from Q108K: K40L:T51V:R58F mutant contoured at 1 ! shows a cis -iminium and not a trans -iminium .............................................................. 68 Figure II-36) a) Overlay structure of the mutant Q108K:K40L after 20 minutes incubation and after 24 hours incubation. b) Overlay structure of the mutant Q108K:K40L:T51V:R58F after 20 minute s incubation and after 16 hours incubation and 1 hour green light (Yellow) ... 69 Figure III -1) structure and absorption of 11 -cis retinal, its n -butyl amin e Schiff Base and Protonated Schiff base ............................................... 75 Figure III -2) The position of the putative negative charge or negative dipole moment determines the absorption of the protein ...................... 76 Figure III -3) Overlay structure of the models of the Green opsin (Green - PDB ID: 1KPW) and Red opsin (Magenta - PDB ID: 1KPX). Mutating the highlighted residues in Green to equivalent residues in Red will provide a Red absorbing like pigment .................................. 78 Figure III -4) Overlay structure of the Rhodopsin (Green - PDB ID: 1F88) and the model of B lue opsin (Blue - PDB ID: 1KPN). The mutation of the highlighted residues on Rhodopsin to equivalent residues of Blue provides a pigment absorbing at 438 nm. The residue numbering is that of Rhodopsin .................................................. 79 Figure III -5) The overlay structure of the model of Green opsin (Green -PDB ID: 1KPW) and Rhodopsin (Yellow -PDB ID: 1F88) .................... 80 Figure III -6) The effect of mutating or introducing hydroxyl containing residues or charge residues along the retinal polyene of Rhodopsin ........ 81 Figure III -7) Electrostatic potential of BR, HR and SRII at the van der Waals surface of the retinal. The magnitude of the potential i s color coded. PDB IDs used are : BR ( PDB ID: 1QHJ), HR ( PDB ID: 1E12) and SRII (PDB ID: 1JGJ) ................................................. 83 Figure III -8) Electrostatic potential calculation for Rhodopsi n, blue, green and red opsin ..................................................................................... 83 Figure III -9) Overlay structure of the BR (Green - PDB ID: 1C3W) and SRII (Cyan -PDB ID: 1JGJ). The residue numbering is that of SRII ... 85 Figure III -10) Overlay structure of the BR (Green - PDB ID: 1C3W) and SRII (Cyan -PDB ID: 1JGJ). The residue numbering is that of SRII. The differences in the trajectory of the retinal and the Arg 72 residue are highlighted ............................................................... 86 !"#$$ !Figure III -11) Electrostatic charge distribution on the surface of the retinal projected from the protein for lef t: Q108K:K40L mutant of hCRBPII, !max = 508 nm. Right: Q108K:K40L:T51V:T53C: R58W :T29L:Y19W:Q4R, !max = 622 nm ................................................ 87 Figure III-12) The planes of rotation for 11 -cis retinal .................................... 88 Figure III -13) The gas phase absorption of: a) retinal n -butyl iminium PSB. b) analogs of retinal n -butyl iminium PSB ....................................... 90 Figure III -14) The studies on developing blue shifted bacterial rhodopsin optogenetic tools: a) The UV -Vis spectrum of blue shifted C1C2 Channel Rhodopsin (ChR) proteins. b) The UV -Vis spectrum of blue shifted archaerhodopsin -3 (AR -3) proteins. c) The crystal structure of the wild type C1C2 protein shows 6s -trans retinal (PDB ID: 3UG9). d) The crystal structure of the T198G:G202A mutant of C1C2 shows 6s -cis retinal (PDB ID: 4YZI) ................. 91 Figure III -15) The position of Wat2a and Wat2b in the binding pocket of Rhodopsin and interacting residues (PDB ID: 1L9H) ................. 93 Figure III -16) The switch of the counter anion to Glu181 in the MetaII Rhodopsin -transducin complex. a) Glu181 as counter anion in MetaII Rhodopsin - transduc in complex (Dark blue Ð PDB ID: 3PQR). b) The overlay structure of the ground state Rhodopsin (Yellow Ð PDB ID: 1L9H) and MetaII Rhodopsin Ð transducin complex (Blue - PDB ID: 3PQR) ................................................. 94 Figure III -17) a) water network in Rhodopsin structure (PDB ID: 1U19). b) Propagation of the water network all the way to the cytoplasmic surface of the Rhodopsin in the crystal structure of MetaII Rhodopsin with GaCT2 peptide fragment of G -protein (PDB ID: 3PQR) ......................................................................................... 95 Figure III -18) Overlay structure of the L93A mutant of Bac teriorhodopsin (Gold Ð PDB ID: 3VHZ) with wild type Bacteriorhodopsin (Cyan Ð PDB ID: 1C3W). Waters 701 -704 are inserted in the cavity made by L93A mutation near retinal PSB ............................................. 96 Figure III -19) The water molecules crucial in the w avelength regulation of hCRBPII (PDB ID: 4RUU) ........................................................... 96 Figure IV-1) a) The crystal structure of wt hCRBPII bound to Retinol. b) Binding pocket of this protein with the important residues highlighted (PDB ID: 4QYN) ..................................................... 105 Figure IV -2) Retinal bound covalently inside the binding pocket of the hCRBPII Q108K:K40L mutant (PDB ID: 4RUU) ....................... 106 !"#$$$ !Figure IV -3) Crystal structure of the hCRBPII Q108K:K40L mutant (PDB ID: 4RUU). The important water molecules and the interacting residues with these waters are highlighted ............................... 107 Figure IV -4) Gln4 is fixed in its conform ation through a water molecule that is in interaction with the main c hain carbon yl of Asp91 and Thr1 (PDB ID: 4RUU) ........................................................................ 108 Figure IV -5) The overlaid structure of KL:T51V:T53C:R58W:Y19W:T29L (Cyan, PDB ID: 4EFG, 591 nm) and of KL:T51V:T53C:R58W: Y19W:T29L:Q4R (Green, PDB ID: 4EEJ, 622 nm) .................. 109 Figure IV -6) a) Crystal structure of the mutant Q108K:K40L:T51V shows the loss of W2 which counts for the 25 nm red shift of the protein. b) Crystal structure of the mutant Q108K:K40L:T53C shows the presence of the W2, which counts for the wavelength of the protein not to change ................................................................ 110 Figure IV -7) a) The Orientation of the Gln38 and Gln128 is defined by adjacent residues that fi x the amide side chain of these residues in place. Crystal structure of Q108K:K40L. (PDB ID: 4RUU) b) Chemical representation of the orientation of these two rsidues 111 Figure IV -8) a) Gln38 -Gln128 Water molecules in the two chains of crystal structure of KL:T51V:R58Y:Y19W. b) Gln38 -Gln128 Water molecules in the two chains of crystal structure of KL:T51V:R58Y:Y19W:Q38L ..................................................... 114 Figure IV -9) a) Surface representation of the Q108K:K40L:T51 V:T53C mutant, !max = 539 nm (left) b) Surface representation of Q108K:K40L:T51V:T53C:R58W:T29L:Y19W (Right) mutant, !max = 591 nm. c) R58W mutation forces th e Phe 57 to flip out of the binding pocket ........................................................................... 117 Figure IV -10) a) Surface re presentation of the Q108K:K40L:T51V:T53C: R58W:T29L:Y19W mutant, !max = 591 nm (PDB ID: 4EFG) . b) Q108K:K40L:T51V:T53C:R58W:T29L:Y19W:A33W mutant, !max = 606 nm (PDB ID: 4EDE). c) A33W mutation helps to cover the entrance of the binding pocket more effectively ....................... 118 Figure IV-11) a) The retinal density in chain G of the Q108K:T51D crystal structure, Lys 40 points at C12 of retinal. b) The retinal density in chain L of the Q108K:T51D crystal structure, Lys 40 points at C14 of retinal. Map is contoured at " = 1.0 ............................... 120 Figure I V-12) PSB loss of the mutant Q108K:K40L of hCRBPII in a) H 2O b) D2O c) The kinetic of the PSB loss comparison in H 2O and D2O122 Figure IV -13) Crystallization trials of Q108K:K40L in D 2O. a) crystals from sitting drop at 6 mg/ml and 35 µl size. b,c,d) Crystals from !"#" !hanging drop at 6 -8 mg/ml. e) Crystals from hanging drop at 12 mg/ml by screening into lower PEG4000. f) Crystals from hanging drop at 20 -25 mg/ml by screening into lower PEG 4000. g ) Macro seeding of the crystal in panel e. h and i) crystals from sitting drop at 25 -30 mg/ml and 80 µl size. j and k) Neutron diffraction pattern of the crystal in panel h ................................ 125 Figure IV -14) Crystallization of Q108K:K40L:T51V:R58F mutant of hCRBPI I in D 2O: a) with 20 -25 mg/ml protein in 10 µl hanging drops. b) microseeded drops provides smaller crystals ........................... 126 Figure IV -15) Crystal of Q108K:K40L:T51V:R58Y:Y19W mutant of hCRBPII in H2O ....................................................................................... 127 Figure IV -16) a) Expression of hCRBPII in regular and H 2O minimal media for mutant: Q108K:K40L:T51V:T53C:R58W:T29L:Y19W 1) MW marker 2,3,4) regular media, elutions from FastQ each 30ml 5,6,7) H 2O minimal media, elutions from FastQ, each 30ml 15b) Expression of Q108K:K40L in H 2O minimal media without cell adaptation 1) MW marker 2) Fast Q elution 3 -9) Source Q elutions. Fraction 4 provides crystals ........................................ 128 Figure IV -17) a) Crystal structure of R111K:R132L:Y134F:T54V: R59W at 2.6 † (Green, PDB ID: 4I9S) !max = 556 nm and crystal structure of R111K:R132L:Y134F:T54V: R59W:A32W at 2.6 † (Cyan, PDB ID: 4I9R) !max = 610 nm of hCRABPII overlaid. b) Surface representation of these mutants: Left: R111K:R132L:Y134F :T54V: R59W (magenta) and Right: R111K:R132L:Y134F:T54V :R59W:A32W (cyan) ................................................................. 130 Figure IV -18) Retinal incubation of hCRABPII mutants in citrat e buffer at pH5 for mutants: a) R111K:R132Q:P39Q:R59Y b) R111K:Y134F :R132Q:P39Q:R59Y c) R111K:Y134F:T54V: R132Q:P39Q:R59Y d) R111K:Y134F:T54V:R132Q:P39Q:R59Y:A32W .................. 131 Figure IV -19) The retinal incubation of hCRABPII mutants at physiological pH: a) R111K:R132Q:P39Y:R59Y b) R111K:Y134F: R132Q :P39Y:R59Y c) R111K:Y134F: T54V:R132Q:P39Y:R59Y ........ 132 Figure IV -20) a) The orientation of the P39Q and R132Q in the R111K:Y134F:T54V:R132Q: P39Q: R59Y mutant of hCRABPII (PDB ID: 4YBP) b,c) The orientation of the P39Q and R132Q in the R111K:Y134F: T54V:R132Q:P39Q:R59Y:A32W:F3Q mutant of hCRABPII in the two chains of the protein (PDB ID:4YKM). d) The orientation of the Gln 38 and Gln 128 in the Q108K :K40L mutant of hCRBPII (PDB ID: 4RUU) ......................................... 134 Figure IV -21) a) Crystal structure of the R132K:K111L mutant of hCRABPII with Merocyanine . b) Different modes of the binding of Merocyanine inside the binding pocket of hCRABPII ............... 136 !""! Figure IV -22) The crystal structure of the Q108K:K40L:T51V:R58Y:Y19W mutant of hCRBPII with: a) 1,1 -di demethyl retinal in 6s -trans retinal conformation and b) 5 -de methyl retinal in 6s -cis conformation ............................................................................. 137 Figure IV-23) a) Crystal structure of the Q108K:K40L:T51V:T53 C:R58W:T29L :Y19W:Q4H mutant of hCRBPII with Julolidene b) pH dependent change of the absorption of the protein with the introduction of L117E mutation c) pH dependent fluorescence ratio from excitation at the two absorption peaks of the protein ................ 138 Figure V-1) Structural organization of Retinoic Acid Receptors ................... 142 Figure V -2) a) Crystal structure of the Ligand Binding Domain of holo -human RAR -! shows the Retinoic Acid inside the binding pock et of the protein (PDB ID: 2LBD) ............................................................ 143 Figure V -3) a) The result of the pull down assay for the Retinoic acid and CD 270 M) MW marker 1) CRABPII at 50% percent of total input. 2) RAR "-LBD on the Ni -NTA resin. 3) RAR "-LBD incubated with CRABPII in the absence of ligand 4,5) RAR "-LBD incubated with CRABPII in the presence of 7.5 and 18.75 µM Retinoic acid, respectively. 6,7) RAR "-LBD incubated with CRABPII in the presence of 7.5 and 18.75 µM of CD -270, respectively. b) Chemical structure of CD -270 .................................................. 145 Figure V -4) CD-270 synthesis trial with the original p roposed synthesis was not successful ........................................................................... 146 Figure V -5) a) Reported modified Suzuki cross coupling reaction . b) proposed retrosynthesis for CD -270 ......................................................... 147 Figure V-6) The synthesis of the first fragment of the CD -270 ..................... 147 Figure V-7) The first two steps in the synthesis of the second fragment ..... 148 Figure V-8) The rest of the synthetic plan for the second fragment ............. 148 Figure V -9) a) Overlay of the structure of the Apo -CRABPII -Chain A (PDB ID: 2FS7) with the Holo structure (PDB ID: 2FR3) does not show the difference in position of Ala 35 and Thr 57. b) Overlay of the structure of the Apo -CRABPII -Chain B and Holo -CRABPII shows the t wo residues of Ala35 and Thr57 get close to each other upon Retinoic acid binding. c) Overlay structure of the Apo ( PDB ID: 1CBI) and Holo CRABPI (PDB ID: 1CBR) shows that the A35 and Thr57 get close to each other upon Retinoic acid binding . 150 Figure VI -1) Retinal incubation for crystallization of: a) R111K:P39Q:R132Q: Y134F:T54V:R59Y mutant of hCRABPII b) R111K:P39Y:R132Q !""#!:Y134F:T54V:R59Y mutant of hCRABPII. c) Q108K:K40L mutant of hCRBPII d) Q108K:T51D mutant of hCRBPII. e) A photo of R111K:P39Y:R132Q:Y134F:T54V:R59Y crystal. f) A photo of R111K:P39Q:R132Q:Y134F:T54V:R59Y:A32W:F3Q .............. 175 Figure VI -2) Borylation reaction ................................................................... 183 Figure VI -3) Nucleophili c aromatic substitution reaction .............................. 184 Figure VI -4) CD -270 synthesis: pro tection reaction of the aldehyde ............ 185 !""## !KEY TO SYMBOLS AND ABBREVIATIONS † Angstrom ! Sigma PDB ID Protein Data Bank Identifier GPCR G-Protein coupled receptor RET Retinal TM trans -membrane PDE phosphodiesterase GTP Guanosine triphosphate GDP Guanosine diphosphate GMP Guanosine monophosphate c-GMP cyclic -Guanosine mono phosphate BR Bacteriorhodopsin SRII Sensory RhodopsinII HR Halorhodopsin ChR Channelrhodopsin AR Archaerhodopsin RSFP Reversibly Switchable Fluorescent Protein NMR Nuclear Magnetic Resonance PSB Protonated Schiff Base iLBP intracellular Lipid Binding Proteins hCRBPII human Cellular Retinol Binding ProteinII hCRABPII human Cellular Retinoic Acid Binding ProteinII RAR Retinoic Acid R eceptors SB Schiff Base !""### ! UV Ultra Violet WT Wild T ype IPTG Isopropyl !-D-1-thiogalactopyranoside PCR Polymerase Chain Reaction PEG Polyethylene glycol Vis Visible " Extinction coefficient QY Quantum Yield #max Maximum wavelength #ex Excitation wavelength #em Emission Wavelength PCL Photo chromic ligand PTL Photo switchable tethered ligand GABA Gama amino butyric acid DBD DNA binding domain LBD Ligand binding domain PYP Photoactivable Yellow Protein Kd Dissociation constant SDS -PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis KDa Kilo Dalton E. Coli Escherichia coli s second min minute h hour mm millimeter !""#$ ! cm centimeter M molar mM Milimolar µM Micromolar nM Nanomolar mol mole mmol millimole mg milligram ml milliliter DNA deoxyribonucleic acid dNTP deoxynucleotide triphosphates rpm rotation per minute ppm parts per million ¡C degrees of centigrade K degrees of kelvin pH Logarithmic scale of hydrogen ion activity pKa Logarithmic constant of acid dissociation HCl Hydrochloric acid NaOH Sodium hydroxide Cod Cyclooctadiene NaH Sodium hydride DMF Dimethylformamide dtbpy 4,4!-Di-tert -butyl-2,2!-dipyridyl B2Pin 2 Bis(pinacolato)diboron TMOF Trimethylorthoformate !""#! Sphos 2-Dicyclohexylphosphino -2!,6!-dimethoxybiphenyl TLC Thin Layer Chromatography Equiv equivalent RMSD root mean square deviation Ala, A Alanine Arg, R Arginine Asn, N Asparagine Asp, D Aspartate Cys, C Cysteine Gln, Q Glutamine Glu, E Glutamic acid His, H Histidine Ile, I Isoleucine Leu, L Leucine Lys, K Lysine Met, M Methionine Phe, F Phenylalanine Pro, P Proline Ser, S Serine Thr, T Threonine Trp, W Tryptophan Tyr, Y Tyrosine Val, V Valine !"! Cha pter I: Photoswitchable proteins: Biological Significance and Applications I.1 Retinal Based Photos witchable Proteins Vision is a multi step process, which starts with focusing of the light on a small spot on retina curtain in the back of the eye called macula . Absorption of photons by visual pigment s in macula causes a change in the membrane potential of the cells to develop a neuronal signal to the brain for analyzing the data. These events translate into what we know as the ability for vision (Figure I-1) (1-3). Figure I -1) a) Eye the gate way to vision b ) A fundus photo of t he Left eye showing the optic nerve, blood vessels an macula c) Different compartments of the eye and the cellular structure of the macula. Optical nerve Macula Blood vessels b) c) a) !"! Macula is a sm all circular portion, with a 5mm diameter near the center of retina, where the first step of the visual process happens. Macula is loaded with specific type of neur al visual cells called rod and c ones. Rod cells are more abundant and sensitive than cone cells specialized for dark vision while cone cells are the l ess sensitive pigments that are specialized for color vision (Figure I-2) (4). Both r od and cone cells are subdivided into t wo portions of outer and inner segment based on their shape (Figure I-2). Outer segment is mainly an extensively folded cellular membrane loaded with a trans membrane protein called Rhodopsin . There are four different types of Rhodopsins, which includes rod, blue, green and red rhodopsins. Rods are the pigment in Rod Figure I -2) a) The structure of a Rod cell. b ) Comparison of the characteristics of Rod and Cone cells. Outer Segment Inner Segment Response Sensitivity Dark adaptation Small/fast Low Fast Large/slow High Slow Scotopic vision (twilight) Colour vision (bright light ) Rods Cones !"!cells. Blue, green and red rhodopsins are the pigments in the three different types of the cone cells (Figure I-3) (5, 6). The ability of rhodopsin photo pigments to absorb light comes from the retinal chromophore inside the binding pocket of the pro tein. Retinal aldehyde makes a Protonated Schiff B ase (PSB) with Lys 296 . The delocalization of the charge along the polyene contributes to the absorption of the protein in the visible region ( Figure I-4) (7). NHLys 296NH2Lys 296OHFigure I -4) Formation of a retinal iminium PSB is essential to light absorbtion of rhodopsin. NHLys 296Figure I -3) The structure of a rod cell, rhodopsin and 11 -cis retinal. 11 12 !"!The wavelength of light absorbed by Blue, Rod, Green and Red rhodopsins are respectively 425, 500, 530 and 560nm for humans . As a photon of light hits the rhodopsin pigment , it triggers the first step in the vision process which is the photo isomerization of the 11 -cis retinal chromophore to all -trans inside the protein (Figure I-5) (6). It should be noted that the quantum yield of the retinal isomerization in Rhodopsin is 65 percent (8). I.2 Photo cascade of Rhodopsin Different studies have shown that after the light irradiation, Rhodopsin goes through the formation of multiple short -lived photo intermediates. All of these photo intermediates have different retinal geometries, which are accompanied Figure I -5) Different rhodopsins absorb a photon of light with the !max highlighted to isomerize from 11 -cis to all -trans retinal. !max = 425 nm !max = 500 nm !max = 530 nm !max = 560 nm 11 12 11 12 h" PDBID: 1F88 PDBID: 3PXO !"!by conformational cha nges in the Helixes 3,5 and 6 of opsin protein to facilitate its signaling (7) (Figure I-6). The first photo intermediate is photo rhodopsin, which is red shifted with a distorted all -trans retinal chromophore. The decay of the Photo Rhodopsin produces Batho in nano seconds followed by the formation of Lumi. The photo cycle proceeds with the formation of MetaI and MetaII (9). MetaII is the photo intermediate in which the retinal PSB deprotonates and the Rhodopsin blue Figure I -6) Photocycle of the Rhodopsin. !"!shifts significantly to where the Retinal -SB absorbs (380 nm). MetaII is the photo intermediate that can activate G -protein and start the initial events that leads to the signaling of G -protein (10). MetaII can take two different pathways for decay. One is the generation of free retinal and opsin and the other is the isomerization t o MetaIII. MetaIII has been found to have a 15 -cis retinal conformation. MetaIII is slower than MetaII in its de cay to free retinal and Opsin and it has been found that the rate of the decay can change from hours at basic pHs to minutes in acidic pHs. Meta III is a partial agonist compared to MetaII, which is a fully active agonist. It has been suggested that the formation of MetaIII holds back the release of the free retinal and its accumulation that can be toxic for the cell (11-14). The overlay of these two structures shows that the trans membrane !-helices can translocate significantly as the retinal expands from 11 -cis to all -trans conformation ( Figure I-7a). !"! Rhodopsins belong to a family of seven !-helical trans membrane proteins called G -Protein Coupled Receptors (GPCRs). Trans membrane proteins are hard to crystallize as their hydrophobic portions that help them to incorporate into the membrane makes it prob lematic for protein purification and Intracellular side Extracellular side Extracellular side Intracellular side a) b) c) Trp 265 Lys 296 1.9 † 3.9 † Figure I -7) a) Struc tural differences between the dark state (Green) and excited state (Cyan) of Rhodopsin. Red = TM3, Black = TM5 and Blue = TM6 in the ground state. b) C -terminus peptide of the G !-subunit bound to MetaII -Rhodopsin. c) Trp265 positioning on TM6 in the ground state and excited state as retinal isomerizes. !"!crystallization in aqueous hydrophilic environment. However, the improvement of the protein expression and crystallization techniques in the last decade has facilitated the crystallization of these proteins, but still t hey include a small number of all of the proteins crystallized so far. Rhodopsin was the first GPCR protein to be crystallized in 2000 by Paleczewski. The crystal structure shows the 11 -cis retinal inside the binding pocket of the protein in the dark state (Ground state) (15). It took another 11 years to crystallize this protein with all -trans retinal (Excited state) in the MetaII state, which is one of the final photoproducts in the visual cascade of the rhodopsin (10). These conformational changes provide a groove for the !-fragment of the G -prot ein to interact with the rhodopsin in order to carry on a signal, which in the next steps will translate into a membrane potential change to become a neuronal signal into the brain for analyzing the stimuli received from the outside world as colors ( Figure I-7b). The most noticeable movement of the !-helices is related to TM6. TM6 contains a conserved residue Trp265, which changes position as the chromophore goes from 11 -cis to all -trans (Figure 7c). The importance of TM6 movements in rhodopsin signaling wa s discovered by photo affinity labeling studies in 2000 as well (9). The crystal structures of the Batho (16) and Lumi (17) were also solved in 2006 and the overlay structure of these proteins shows a twisted all -trans retinal in the Batho that is relaxed by the dislocation of the ionone ring in Lumi. Also the conformational changes of the Helixes 3,5 and 6 w ere reported in the Batho and Lumi structure but the large conformational changes, especially that of Helix 6 waited to be more clarified by MetaII structure. Compared to Lumi structure in MetaII both the ionone ring and the polyene tail !"!are translated along the retinal chromophore axis, giving it a quite different geometry compared to the ground state, batho and Lumi ( Figure I-8). It should be noted here that besides vision, circadian rythme, which is the body clock, is also regulated by retinal photo isomerization in ganglion cells of the retina (18-20). I.3 Mechanism of Rhodopsin Signaling Rhodopsin signaling is the cooperation of multiple proteins in a series of biological reactions in a matter of milliseconds. After binding of the G -protein to Rhodopsin and opsin proteins, it exchanges a GDP molecule for a GTP. Binding of GTP to G -protein dissociates the !-subunit of the protein from the " and # subunits. The dissociated !-subunit binds to the inhibitory subunit of the cGMP phosphodiesterase (PDE) and the activated PDE hydrolyzes the cGMP to 5Õ -GMP. This results in the closure of the cGMP -gated calcium ion Figure I -8) The retinal conformation in Rhodopsin: Ground state (Red - PDB ID: 1U19), Batho (Gray Ð PDB ID: 2 G87 ), Lumi (White Ð PDB ID: 2HPY) and MetaII (Yellow Ð PDB ID: 3PXO). The arrows show the rotation of the retinal from Lumi to MetaII. !"#!channels and hyperpolarization of the rod and cone cells that will end up to the neuronal signal. A rhodopsin kinase protein phosphorylates the Theronine and Serine residues in the C -terminus of Rhodopsin. The phosphorylation of the Rhodop sin facilitates the binding of another protein called arrestin that prevents further interaction of the rhodopsin with G -protein. Each Rhodopsin can activate 500 G -proteins and each PDE molecule can hydrolyze 4200 molecules of cGMP per second and the fact that multiple cGMP molecules are required to keep one cGMP gated ion channel open makes the signaling very effective (21, 22) (Figure I-9). !"#!"#GTPFigure I -9) Rhodopsin signaling . a) b) !""!I.4 Photoisomerization in Bacterial Rhodopsins Rhodopsin s in bacteria play a variety of different functions, which includes energy production by pumping H + outside the cells (Bacteriorhodopsin) , phototaxis by transducer binding signaling (Sensory Rhodopsins), photo taxis by Na+, K + and other cations pumping across the membrane (Channel Rhodopsins) and Cl - pumping across the membrane (Halor hodopsin) (7, 23). Bacterial rhodopsins in the ground state bind to all -trans retinal (24) and upon absor ption of light they isomerize to the 13 -cis form (Figure I-10) (25). There have been a lot of studies to elucidate the photo intermediates of the isomerization process (26-30), but a lot of the structures are subject to X -ray damage during data collection on the crystals (31). Using t he photo isomerization of Channel Rhodopsin -2 (ChR -2), which is a Na+ and K+ pump and halo rhodopsin which is a Cl - pump in neurons to trigger neuronal si gnaling in the brain is called O ptogenetics. Optogenetics is a very promising field in neurobiology that helps to understand the particular function of each part of the brain. The advantage of using light in controlling the neural 13 14 13 14 Figure I -10) a) Bacteriorhodopsin in the ground state of all -trans retinal (Left - PDB ID: 1C3W) and excited state of 13 -cis retinal (Right - PDB ID: 1IXF). b) Chemical diagram of the isomerization of the bacter ial rhodopsins with the !max of each indicated. a) b) !"#!activity is in the temporal and spatial firing of neurons. Specific signals will be produced right where the brain is irradiated with light and as the light source goes off the signal will be terminated. Optogenetics has been already used to understand fear learning, Cocaine dependence, Prefrontal cortex responses, restora tion of vision and fibrillation (Figure I-11) (32, 33). a) b) Figure I -11) a) Incorporation of Channel Rhodopsin (ChR) or Halo Rhodopsin (HR) can confer the sensitivity of light to neurons. b) The application of optogenetic tools in mice helps to understand the function of different p arts of the brain by spatial and temporal activation of neurons. !"#!I.5 Photocycle of Bacteriorhodopsin After a photon of light hits the bacterial rhodopsins, the isomerization of all -trans retinal to 13-cis retinal goes through multiple intermediates (7). The iso merization of the bacteriorhodopsin (BR) is the one that has been studied extensively in the last two decades by different techniques including FTIR, resonance Raman, NMR, theoretical calculations and X-ray crystallography (7, 34, 35) (Figure I-12). K and O are the red shifted intermediates and L, M and N are all blue -shifted intermediates. The deprotonation happens in the M st ate and the thermal Figure I -12) Photocycle of the Microbial rhodopsins. !"#!reisomeriza tion back to all -trans retinal happens in the O state. It is clear from figure 9 that the isomeric state of the L and N intermediates is the same while a proton must be released from the intracellular space to the extracellular space. Multiple studies showed t hat the M state is subdivided into two M1 and M2 state s, in which the M1 delivers the proton to the extracellular environment, and M2 faces back again toward the intracellular space. Therefore, the switch happens between M1/M2 states. The protonation and deprotonation of the Retinal PSB happens via Asp 96 in the cytosolic side and Asp85 toward the extracellular side. The deprotonation of the Asp96 in the N state and protonation of the Asp85 in the M, N and O state has been confirmed in different studies (7). Bacteriorhodopsin has seven water molecules toward the extracellular side of the protein and only two water molecules toward the cytoplasmic side. This uneven distribution of water molecules is due to the fast kinetic s of proton release of bacteriorhodopsin to the extracellular environment , while the reprotonation of the protein re quires conformational changes of the retinal chromophore and Helix F of the protein to completely segregate the two sides of the protein fro m each other for efficient proton pumping (24, 36). Crystallographic studies of the bacteriorhodopsin photo intermediates mainly includes the irradiation of t he crystals of this protein by corresponding wavelength of light that accumulates the related intermediates and point mutations that elongates their lifetime to help to trap them. Also, the crystallographic data suggests that it is the M state that the swi tch from cytosolic to extracellular environment happens in (28, 29), but the structural models presented for the L intermediate have been controversial (27, 30) and !"#!X-ray radiation damages have been found as a contributing factor for these discrepancies (25, 31). One of the interesting characteristics of the bacteriorhodopsin photo isomerization found before the crystal structures was the mixture of 13 -cis -15-cis retinal along with 13 -trans -15-trans retinal in its dark isomerization (37-39). In the crystal structure of the O -intermediate it was found that the crystals that are shined with green light followed by adaptation in dark the two isomers can be modeled, but when the crystals are treated with red light of 650 nm and kept irradiated during th e process show a complete 13 -trans -15-trans retinal model (26) (Figure I-13). It should be mentioned at the end of this part that the photo isomerization of the rhodopsin and bacteriorhodo sin with the highlighted conformational changes mentioned drops the pKa of the retinal PSB. In Rhodopsin the pKa of the retinal PSB drops from 16 in the 11 -cis retinal ground state to above 8 in Figure I -13) The ove rlay structure of the bacteriorhodopsin in the: Ground state (Green Ð PDB ID: 1C3W), light adapted O -intermediate (Cyan Ð PDB ID: 3VHZ) and dark adapted O-intermediate (Magenta Ð PDB ID: 3VI0) shows that the light adapted form can repro duce the ground state geometry, but not the dark adapted crystal. Lys 216 !"#!the all -trans retinal of MetaI/ MetaII form s and to 5.1 in the 15 -cis retinal MetaIII form (11, 40, 41). In bacteriorhodopsi n the pKa drops from 13 in the all-trans retinal ground state form to below 7.5 in the 13-cis retinal of M form (42-44). Therefore, the pKa drop is one of the essential characteristics of the rhodopsins for their proper functioning. The idea of combining a retinal light absorbing photo switchable pigment to a fluorescent protein has never been tried. This can make a Reversibly Switchable Fluorescent Protein (RSFP), which has been proven to be an extremely valuable biological tool. I.6 Reversibly Swi tchable f luorescent Proteins (RSFPs) Compared to f luorescent proteins that are all the time in an on -state , Reversibly Switchable Fluorescent P roteins (RSFPs) can be switched between a fluorescent and non -fluorescent state back and forth. The development of RSFPs has pushed the limits of biological mic roscopy from 200 nm to 30-40 nm (Figure I-14) (45-48). RSFPs are the key element to a number of growing advanced imaging techniques, which includes: STORM , PALM and FPALM (49-53), Figure I -14) Conventional immunofluorescence imaging of microtubules (a and c) and STORM images which uses RSFPs ( b and d). a) b) c) d) !"#!photochromic FRET (54), optical lock -in detection (55), frequency domain imaging (56), and bio data storage (57). The 2014 chemistry Nobel prize was awarded to the development of super resolution fluorescent microscopy which requires the use of RSFPs. The mechanism of the switch in the RSFPs has been proven crysta lographically in multiple publications to a cis - trans photo isomerization of the fluorescent chr omophore, which in one form the protein, is fluorescent and in the other form it is not (Figure I-15) (58-61). The loss of a water molecule that extends the conjugation of the chromophore is also another mechanism that is structurally proved for the mechanism of the switch in RSFPs . In the off state the water makes a covalent bond with the fluorophore to break its conjugation and turns the fluorescence of the protein off. In this specific switchable protein the wavelengths used for the switch are decoupled from the excitation wavelength (62) (Figure I-16). Figure I -15) The mechanism of the switch in RSFPs is a cis -trans isomerization of the fluorophoric ligand. !"#! Photo bleaching is a process in which the RSFPs lose their fluorescence due to photo destruction of the chromophore. Photo bleaching limits the reversibility of the RSFPs. There are two mechanisms that are proven structurally to be t he reason for this phen omenon. With regular laser powers required for the wide -field fluorescence microscopy the Met159 and C ys171 residues undergo oxidation to lock the chromophore in a protonated state . However, with stronger laser powers used in nanoscopy, the decarboxylation of the Glu212 residue , changing of the H -bonding network around the chromophore and the change of the hybridization of the methylene bridge Figure I -16) The loss of a wa ter molecule attributes to the excitation decoupled on -off switch of the fluorescent protein. !"#!carbon of the fluorophore from SP 2 to SP 3 counts for the los s of the fluorescence (63). I.7 Optochemical Genetics Optochemical genetics uses t he small synthetic photo switchable molecules in receptor proteins to give them the ability to respond to light. Like Optogenetics in Optochemical genetics the temporal and spatial control of the protein receptors become feasible by the use of these molecu les. Diazo switchable a) b) Figure I -17) a) Photochromic ligands (PTLs) in both form s can act as a ligand for protein, but just one of the two forms activates the protein. b) Photo switchable Tethered Ligand (PTL) is covalently bound to the protein, and just one form fits into the binding pocket of the protein. !"#!molecules are the main molecules that have been used to enable trans membrane proteins and neurotransmitters to undergo photo switch by light irradiation (64, 65). Two main categories of ligands that have been used are Photo Chromic Ligands (PCLs) and Photo switchable Tethered Ligands (PTLs) (Figure I-17). In PCLs from the two cis and trans forms of the ligand, one of them is the one that is a n agonist of the protein or the switch is from the antagonist in one form to agonist in the other form . Upon irradiation and isomerization the ligand goes from one form to another and activates the function of the protein. In PTLs the ligand is tethered to the protein via a covalent bond formation between the protein and the ligand. In one form the tethered ligand does not fit into the binding pocket of the protein, u pon irradiation the ligand isomerizes into the appropriate length and spatial availability a nd locks in the binding pocket of the protein and activates the protein function. Photo switchable azo drugs are a main category of molecules that have been used in ligand-gated neuro transmission. They have been used in GABA receptors, Neuromuscular recep tors, Ni cotinic acetyl choline receptors , glutamate receptors , allosteric regulation of GPCRs (66) and enzymatic enantioselectivity (67). They have been used to control heartbeat (68) and pain sensat ion (69) and to restore vision (70). I.8 Photoactive Yellow P rotein Photoactive Yellow Protein (PYP) is a small 14KDa protein, which absorbs the blue light !max = 446 nm in Halorhodospira halophila. It controls the phototaxis in bacteria to swim away from the harmful blue light and toward the more !"#!photo synthetically suitable lights. PYP is one of the proteins , which forms crystals that respond to light in the crystalline state. PYPs photo intermediates have been extensively studied by time -resolved X -ray crystallography. It should be mentioned that just 10% of the overall population of the molecules respond to light in the crystalline form. The mechanism of the light absorption is a cis -trans isomerization around the chromophore of this protein (Figure I-18) (71). I.9 A photo switchable rhodopsin mimic The result of a reaction of an aldehyde with a primary amine is a well -known reaction that makes a C=N bond known as Schiff base formation. All of the Figure I -18) Photo isomerization of PYP. Overlay structure of the Ground State (Cyan, PDB ID: 2ZOH) and Excited state (Yellow, PDB ID: 4BBV). HNSh!= 446nm"HNSOOOOHGround StateExcited StateOOCys 69 !""!retinal light absorbing pigments use the same strategy to make a Schiff base which upon protonation makes a Pr otonated Schiff Base (PSB). Delocalization of the charge along the retinal chromophore causes the absorption of light in the visible region (Figure I-19). Working with rhodopsins is accompanied by multiple challenges that are typical of trans -membrane proteins. These include: expression, purification, solubility, mutagenesis and crystallization. In the search for mimics that can be good candidates for making a retinal PSB we arrived at Cellular Retinoic Acid Binding ProteinII (CRABPII) and Cell ular Retinol Binding ProteinII (CRBPII). These belong to intr acellular Lipid Binding Protein Family (iLBPs). CRABPII and CRBPII are soluble small cytosolic transport proteins that can be expressed, purified, mutated and crystal lized extensively in our hand s ( Figure I-20). Figure I -19) Formation of a Retinal Schiff Base and Protonated Schiff Base. It is the PSB that absorbs in the visible region of light. !"#! These two proteins have been reengineered to retinal light absorbing pigments successfully, spanning over 200 nm of the visible light spectrum (72, 73) (Figure I-21). retinal 380 425 644 474 482 508 533 570 577 590 613 622 a) b) Figure I -21) a) Reengineered hCRBPII protein retinal pigments absorbing light from 425nm to 644nm. Incubated at pH = 7.3. b) Reengineered hCRABPII protein retinal pigments absorbing light from 482nm to 630nm. Figure I -20) human Cellular Retinoic Acid Binding ProteinII (hCRABPII) bound to Retinoic Acid (Left -PDB ID: 2FR3 ). Human Cellular Retinol Binding ProteinII (hCRBPII) bound to retinol (Right -PDB ID: 4QYN). !"#!We have been reengineering hCRBPII and hCRABPII for a variety of different applications besides wavelength regulation studies , which includes in vitro and in vivo Fluorescent studies and pH sensing (72-82). The common theme for all of these studies lies behind the abili ty of these prot eins to make a Protonated Schiff Base (PSB) upon aldehyde chromophore binding. The significance of a system that can provide fine tuned isomerization of retinal in a mimic protein intrigued us to focus our studies in this area as well . In the next chap ter the details of reengineering of hCRABPII to a protein that is able to do retinal dark isomerization and photo isomerization, the UV -Vis assays to highlight the spectroscopic characteristics of the protein and the crystallographic data with atomic resolution description of the retinal and protein con formational changes are delineated. This proteins have clean and complete dark isomerization from one isomer to another a characteristic that the rhodopsin is devoid of and in bacteriorhodopsin provides mixture of isomers. The photo isomerization also prov ides clean and complete photoproducts . Interestingly, these isomerizations are accompanied by 3 -5 pKa unit changes, not as significant as visual and microbial rhodopsins pKa change s though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``%! C7)@*&!F@$+@.FT/G#H$Ia$.@%!!$$%!M1'1C(B^14!M!L!]<^3C!A!+"TT#/!U'3!1CC IP61B7 I$HI7DBI63P(B1C! )'6<4<>'<63!7(B!```!(7!1!>16P(1C!1^6<>36P(37! 7(B!```G!63)3BP!7>3)P6<7)<>()!7PN2(37!1P':1D7! 7(B%!8%+("$73,$%&'"()*+,$%&'"$73,9%+("$(,D,8008 !,+$F/G$#F-I$#H-%!!$.%!]<^3C!A8!KN23=3!&8!A12N!`8!&(3_36P!S8!L!&'3E37!M!+"TTF/!5'7(B!```%! !"#$%&'" ()*+!F.+.$/G$T"HHI$T"#F%!!$F%!b(443641BB!X8!A(PP36!V8!Z16PC!S[8!*C3P37!43P16'<2<>7(B!```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a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`:57V!aN377(4,(;!+4!S78:5(;6(46(! L()(6)+:4;&! >))-3+$1"5/#6"@#1 !A/.%0@%/?B%>?&!$>&!]V32!T !"#$"%&' !.%#=#0!L3)3! ;):53,(!O3;(V!:4!)% ;#60" -<"()*#+)#1"-<"$/#"?+*$#;"($%$#1"-<">6#4*)% !=#%.">0@="#>#B="#>A&!!$?&!C(4V(5;:4!b_F!]+!C`F!a320!D)586)8537! O3;+;!9:5!5(Q(5;+O7(!)%;#60" -<"()*#+)#1"-<"$/#"?+*$#;" ($%$#1"-<">6#4*)% !=#A.=/0@//>%B//>>&!!/#&!]4V5(;(4!K !"#$"%&' !.%##>0!D)586)8537!O3;+;!9:5!5(Q(5;+O7(!)%;#60"-<"()*#+)#1"-<"$/#"?+*$#;" ($%$#1"-<">6#4*)% !=#A."%0@="##$B="##?&!!/=&!I7()4(Q!DF!D8O36N!STF!L38)(5!ZF!`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`a479 F39`!I73B(-9!E+437(*C(9C(N!-9!*-B4!E37,'B-39!3E! CD'9-9(!GD(*%!>+6'3$*%+:%#"ANO?$"FO?@?%!! !""!Chapter II: Development of a Photoswitchable engineered human Cellular Retinic Acid Binding ProteinII !II.1 Introduction Retinoid Binding Proteins (RBPs) are a large family of binding proteins, which their function is the transfer , metabolism and regulation of the biological effects of retinoids in living organisms (Figure II-1) (1, 2). Cellular Retinoic Acid Binding Protein II (CRABPII) is one of the proteins of this family , which binds to all -trans retinoic acid (3, 4) and solubilizes this hydrophobic molecule in the polar environment of the cells and delivers the retinoic acid through trans -localization to t he nucle us to Retinoic Acid Receptors (RARs) (5, 6), which are transcription factors. Crystal structure of the wild type human CRABPII protein bound to Retinoic Acid shows the interacting residues with the ligand in the binding pocket of the protein (Figure II-2) (7). OHOHOOHRetinolRetinalRetinoic AcidFigure II -1) Chemical structure of different r etinoids. !"#! In collaboration with Professor BorhanÕs lab we have been reengineering this protein in the last decade for a variety of different research studies, which include wavelength regulat ion (7, 8), pH sensing (9) and fluorescent studies (10). The first attempts of wavelength regulation studies was reengineering the binding pocket of the protein in a way that enables the protein to make a Schiff Base (SB , imine ) and upon protonation a Protonated Schiff Base (PSB , iminium).This idea comes from Rhodopsins that make a PSB to absorb light (11). To this end, the Arg 132 was mutated to a lysine to introduce the nucleophilic lysine residue for PSB formation, Arg111 was mutated to hydrophobic Leucine to remove the positive charge close to the retinal PSB and finally the Leu121 was mutated to a glutamic acid residue to intro duce the counter anion for PSB stabilization. R132K:R111L:L121E mutant (KLE) absorbs at 450 nm with quick PSB formation completion in minutes and a pKa of 8.7. Unfortunately, the crystallization trials of different mutants of this series have not provided any retinal bound structures except the KLE mutant . The crystal struc ture of this mutant shows the retinal covalently bound to Lys 132 and the other interacting residues with the PSB (Figure II-3) (8). Y134 R111 Retinoic acid 2.71 † R132 T54 L121 2.57 † 3.01 † 2.73 † 3.02 † Figure II -2) a) Crystal structure of hCRAB PII bound to Retinoic Acid. b) B inding pocket of hCRABPII and interacting residues with the Retinoic Acid (PDB ID: 2FR3) . a) b) !"#! Interestingly, different mutations on the entrance of the binding pocket of this protein show no effect on the wavelength of the mutated proteins. Additionally, different mutant proteins show no absorption above 450 nm (Table II-1). A surface representation of the crystal structure of KLE mutant protein shows that the retinal chromophore is not completely embedded in the binding pocket of the protein, leaving the c hromophore exposed to the aqueous buffer, which have a much bigger dielectric constant compared to the one that Mutant !max (nm) Kd R111K:R111L:L121E 449 1±5 R111K:R111L:L121E:R59K 444 157±24 R111K:R111L:L121E:R59H 440 18±12 R111K:R111L:L121E:R59A 442 92±8 R111K:R111L:L121E:R59L 443 10±9 R111K:R111L:L121E:R59F 440 11±9 R111K:R111L:L121E:R59Y 435 2±8 R111K:R111L:L121E:R59W 442 38±8 R111K:R111L:L121E:R59Q 444 16±8 R111K:R111L:L121E:R59D 440 73±11 R111K:R111L:L121E:R59E 450 37±6 Retinal 2.75 † Y134 3.85 † 2.6 † R111L R132K L121E Figure II -3) a) Crystal structure of the R132K:R111L:L121E mutant of hCRABPII (PDB ID: 2G7B) shows the retinal covalently attached to the protein. b) Interacting residues with the i minium Schiff Base in this crystal structure. a)!b)!!Table II -1. Summary of the mutations on R59 position of fir st generation of hCRABPII mutant s. !"#!is implied to the chromophore through mutations. In other words , the aqueous media that the chromophore has remained exposed to wipes out the ef fect of the mutations on the wavelength of the protein (Figure II-4) (12). This category of proteins will be referred to as the first generation of hCRABPII mutant proteins . To solve this problem there are three strategies that proved to be effective: (1) Shortening the length of the chromophore by removing two double bonds of the polyene to make a C15 -anolog of retinal (12). (2) Reengineering the binding pocket of the protein in a way that it can completely encapsulate the retinal (9). (3) Using another protein from Retinoid Binding Protein Family as surrogate (13). With the first strategy not only the chromophore is completely embedded by the protein (Figure II-5), but also the mutations that originally had no effect on the wavelength of the retinal (Table II-1) have a clear ef fect on the wavelength of the C15-reti nal analog (Table II-2). R59 R132K Figure II -4) Crystal structure of R132K:R111L:L121E mutant, shows the ionone ring is posit ioned out of the binding pocket (PDB ID: 2G7B). !"#! With the second strategy not only the retinal is completely encapsulated in the binding pocket of the protein, but with the right mutations a broad range o f wavelengths has been covered, which pr ovide proteins that are more red shifted than any retinlydene protein pigment known. This category of proteins will be referred to as second generat ion of hCRABPII mutant proteins (9) (Figure II-6). Mutant !max (nm) Ð C15 analog Kd R111K:R111L:L121E 333 157±24 R111K:R111L:L121E:R59K 330 1390±161 R111K:R111L:L121E:R59H 398 719±96 R111K:R111L:L121E:R59A 397 236±29 R111K:R111L:L121E:R59L 391 144±19 R111K:R111L:L121E:R59F 392 279±33 R111K:R111L:L121E:R59Y 396 226±33 R111K:R111L:L121E:R59W 404 259±17 R111K:R111L:L121E:R59Q 413 259±26 R111K:R111L:L121E:R59D 409 262±35 R111K:R111L:L121E:R59E 424 1±4 ORetinalOC15-Retinal analogR59W R59 Retinal C15 -Retinal R132K a) b) Figure II -5) a) Chemical representation of retinal C15 -analog compared to retinal. b) Overlay structure of R132K:R111L:L121E -retinal complex with R132K:R111L:L121E :R59W -C15 retinal complex. Table II -2. Summary of the mutations on R59 position of firs t generation of hCRABPII mutants . !!"#! Two of the observed characteristics of second generation hCRABPII proteins is their l ow pK as, which essentially leaves these proteins with no PSB at physiological pH 7.3 (Table II-3) and their quick PSB loss overtime, sometimes in minutes upon incubation with retinal (Figure II-7) (9). Mutant !max (nm) pKa 1-R111K:R132L:Y134F:T54V:R59W:A32W 610 7.0 2-R111K:R132L:Y134F:T54V:L121Q:R59Y:A32W 538 5.9 3-R111K:R132L:Y134F:T54V:L121Q 492 5.3 4-R111K:R132L:Y134F:T54V:L121Y:R59Y:A32W 529 5.1 5-R111K:R132L:Y134F:T54V:L121Q:R59W 505 4.4 6-R111K:R132L:Y134F:T54V:L121N 495 4.2 7-R111K:R132L:Y134F:T54V:L121Q:R59W:A32W:M93L:E73A 560 4.1 8-R111K:R132L:Y134F:T54V:L121Y:R59W:A32W:M93L:E73A 530 3.0 9-R111K:R132L:Y134F:T54V:L121Y:R59W 504 2.8 10-R111K:R132Q:Y134F:T54V:R59W:A32W:M93L:E73A 630 2.6 11-R111K:R132L:Y134F:T54V:R59W:A32W:E73A 620 2.4 a) b) Figure II-6) a) Overlay crystal structures of R111K:R132L:Y134F:T54V:R59W and R132K:R111L:L121E shows the retinal structure deeper inside the binding pocket of hCRA BPII. b) Different mutants of second generation hCRABPII absorb from 474 to 640 nm. Table II -3. Mutants of second generation hCRABPII with pKas below physiological pH. !!"#! In this chapter I will discuss the underlying reasoning for these two observations and how it took us to a photo swi tchable retinal -based protein. II.2 Developing of hCRABPII to a Photoswitchable Rhodopsin mimic One of the other members of iLBPs that we have been reengineering is human Cellular Retinol Binding Protein (hCRBPII), which also shows a wide range of absorption spectrum for the mu tants of this protein (Figure II-8). Figure II -7) PSB loss of the incubated R111K:R132L:Y134F:T54V:R59W mutant of hCRABPII with retinal at pH=5.2 . Figure II -8) a) Retinol inside the binding pocket of wt -hCRBPII (PDB ID: 4QYN). b) Retinal PSB covalently bound inside the binding pocket of Q108K:K40L mutant of hCRBPII (PDB ID: 4RUU) . a) b) !"#!Unlike the second generation of hCRABPII proteins , hCRBPII proteins have high pKa mutant s with pK as as high as 10.2 (13) (Table 4 ). This led us to compare the binding pocket of these two protein s to figure out the differences, and to see if the problem of the low pK a of hCRABPII second generation mutants can be solved by mimicking the binding pocket of hCRBPII into hCRABPII. Mutant !max (nm) pKa Q108K:K40L 508 8.3 Q108K:K40L:R58Y 535 9.5 Q108K:K40L:T51V:R58Y 555 10.0 Q108K:K40L:T51V:R58Y:Y19W 561 10.2 Q108K:K40L:T51V:T53C:R58W:T29L:Y19W 591 8.2 Q108K:K40L:T51V:T53C:R58W:T51V:Y19W:Q4A 613 7.0 Figure II -9) The structure s of the hCRABPII mutant R111K:R132L:Y134F:T54V:R59W (Yellow, PDB ID 4I9S ) and the hCRBPII mutant Q108K:K40L (Cyan , PDB ID 4RUU ) were overlaid and the region around the retinal binding site is shown. Note the three Gl utamines in hCRBP II are substituted with hydrophobic residues in hCRAB PII. Gln 4 Phe 3 Gln 128 R132L Q108K R111K Gln 38 Pro 39 Arg58 R59W Table II -4. Different mutant of hCRBPII with high pK as. !"#!A comparison of the binding pocket of hCRBPII with hCRABPII shows three glutamine residues in hCRBPII i n the vicinity of the chromophore that are n ot conserved in hCRABPII ( Figure II-9). In each case mutation of these residues to Gln in hCRABPII would result in a more hydrophilic binding cavity, which we suspected might lead to stabilization of the retinal PSB. In addition, mutation of Arg58 (Arg 59 in hCRABPII) to Tyr in hCRBPII also reliably resu lted in PSBÕs with higher pKaÕs. (Table II-4). This led us to the R111K:Y134F:T54V: R132Q:P39Q :R59Y ( M1) hCRABPII mutant. Incubation of M1 with retinal at pH=5.0 s howed complete PSB formation in 3 hours and 30 minutes . However, the acid -base titration of this mutant revealed a broad curve, ranging from pH of 4.3 to 10.5, inconsistent with a single pKa. Other mutants containing P39Q and R132Q mutations gave similar r esults (Figure II-10). A possible explanation for this behavior came from our previous studies of retinal -bound hCRABPII mutants, where the PSB was lost over 1 -2 hours or in some cases minutes (Figure II-7). Subsequent acidification regenerated the 00.10.20.30.40.50.60.70.8300400500600700800M1pH 4.284.685.005.455.816.216.657.157.658.238.9410.0910.48Absorption Wavelength (nm) 00.10.20.30.40.50.64567891011M1Absorption at 554 nm pHFigure II -10) pKa t itration of M1 mutant in citrat e buffer at pH=5. Protein concentration is 50µM incubated with 0.5 equiv alent retinal. !"#!PSB and acid -base titration showed the conversion of the retinal PSB from a higher to a lower pKa form, suggesting a change of environment in the vicinity of the iminium nitrogen. Such a change in environment could be attributed to an isomerization of one of the double bonds in the chromophore polyene, analogous to that seen in the natural rhodopsin systems, where light isomerization of retinal , most commonly of the 11 or 13 double bonds, leads to the change in the position and therefore pKa of the iminium (11). If correct this would suggest a thermal analog of a functional rhodop sin system, where isomerization is occuring thermally and results in a significant cha nge in the pKa o f the iminium (Figure II-11). It should be mentioned that the loss of the PSB can not be attributed to retinal loss of the protein, since after the loss of the PSB it can be recovered by acidification to form a new PSB . Figure II -11) a) Base titration of kinetic product b) acid titration of thermodynamic product of R111K:R132L:Y134F:T54V:R59W mutant of hCRABPII. a) b) !"#!The most possible explanation for this event comes from our already present structural data , which was the retinal iminium double bond isomerization itself. We have shown in our hCRBPII structural studies that different mutants of this protein can adopt a cis -PSB retinal (15 -cis retinal) when the Gln4 residue is preserved and a trans -SB retinal (15 -trans retinal) when this residue is mutated (Figure II-12b, Figure II -12c). Also the structural data from the early mutants of the hCRABPII , which the P39Q and R132Q mutations are not implied the refi nement of both isomers of 15-cis retinal and 15 -trans retinal into the electron density map is feasible (Figure II-12d). Like the other second generation hCRABPII mutants, M1 also converts over time from a higher to lower iminium pKa form. However, the rate of N15-cis retinalN15-trans retinala) b) c) d) Figure II -12) a) a chemical representation of 15 -cis retinal (left) and 15 -trans retinal (right) b ) Retinal density in crystal structure of Q108K:K40L:T51V:R58Y:Y19W mutant of hCRBPII shows a cis -PSB density for retinal c ) Retinal density in crystal structure of Q108K:K40L:T51V:T53C:R58W:T29L:Y19W:Q4H mutant of hCRBPII shows a trans -SB density for retinal. d) Retinal density in crystal structure of R111K:R132L:Y134F:T54V:R59W:A32W mutant of hCR ABPII shows both 15-cis retinal (left) and 15 -trans retinal (right) which can be fit into the density of retinal ( all the maps are contoured at 1! ). !""!interconversion between high and low pKa forms is substantially slower than that seen in previous mutants (Compare Figure II-7 and Figure II-13a), which may allow for the trapping and unequivocal identification of each form. In an attempt to identify the final product, crystallization of M1 was initiated only after all of the PSB had disappeared (24 hours). This resulted in a clear electron density map consistent with an all trans conformation for all the chromophoreÕs double bonds, including the iminium double bond ( Figure II-13b). In this isomer the environment around the iminium nitrogen is very hydropho bic, consistent with its low pKa (2.8 pH units lower than a potential 15-cis isomer). However, attempts at trapping the high pKa form of M1 by initiating crystallization trials immediately after maximum PSB formation resulted in crystals with ambiguous electron density for the chromophore , especially in the vicinity of the iminium region. There are two potential problems with this approach. First and most obvious is that partial conversion may occur before 00.20.40.60.8300400500600700800Retinal Incubation of M1 Protein 3h30min 24h incubation Absorption Wavelength (nm) Figure II -13) Thermal inter conversion in M1 a) UV/Vis spectra of the M1/retinal complex 3.5 hours and 24 hours after addition of retinal to the protein . b ) Retinal density in the crystal structure of R111K:R13 4F:T54V:R132Q:P39Q:R59Y at 1. 83 † contoured at 1 ! shows a trans -SB (PDB ID: 4YBP). a) b)!"#!crystal n ucleation, resulting in an ambiguous mixture of the two isomeric forms. The other problem is the comparatively long incubation time required for the PSB to fully form in the first place ( 3 hours and 30 minutes ). This probably results in a situation where the meta -stable form is never the exclusive form in solution. The only way to overcome this problem is to significantly accelerate the formation of the PSB, so that the high pKa form is fully formed before it has time to significantly degrade to the low pKa form. Table 5 summarizes the mutants made in pursuit of an ideal system for this study. It should be mentioned that R132Y mutation fails to accelerate the PSB formation, P39N mutation drops the pKa of the protein and in some cases is accompanied by p rotein expres sion problem and P39E mutation does not provide stable proteins. Clear from the data is that the presence of both Gln39 and Gln132 leads to both slow PSB formation and slow subsequent PSB loss. However, mutants with Tyr at position 39 by and large form the initial PSB much faster, with PSB formatio n complete in about 10 minutes. Based on Mutant * Low p Ka ( !max nm) High p Ka!$!max nm) PSB rise (hours) PSB fall (hours) M1-KFVQ:P39Q:R59Y 5.4 (530) 8.4 (556) 3 24 M2-KFVQ:P39Y:R59Y 3.4 (564) 8.15 (573) 0.2 10 M3-KFVQ:P39Q:R59Y:A32W:F3Q 9.4 (557) - 5 50 % after 24h M4-KFVQ: P39Q: R59W 5.1 (537) 7.8 (555) 3.5 24 M5-KFVQ:P39Q: R59Y:A32Y 5.6 (554) 8.4 (575) 4 24 M6-KFVQ:P39Q:R59Y:A32W 6.0 (563) 8.3 (584) 2 24 M7-KFVQ:P39Q:R59W:A32W 6.2 (582) 8.6 (598) 2 24 M8-KFVQ:P39Y: R59Y:A32Y 3.2 (587) 8.4 (588) 0.1 16 M9-KFVQ:P39Y:R59Y:A32W 2.5 (600) 8.3 (596) 0.2 12 M10 -KFVQ:P39Y:R59W:A32W 3.1 (601) 8.3 (595) 0.3 10 M11 -KFVQ:P39Q:R59Y:A32Y:F3Q 9.3 (545) - 3 - M12 -KFVQ:P39Q:R59W:A32W:F3Q 9.0 (571) - 3 - Table II.5 Screening for hCRABPII mutants with fast kinetic PSB formation and slow PSB loss. ¥KFVQ is the abbreviation for R111K:Y134F:T54V:R132Q . !"#!these observations mutant R111K:R134F:T54V:R132Q:P39Y:R59Y ( M2) was made. M2 forms a PSB within 20 minutes of retinal addition, and slowly deprotonates ove r a course of 10 hours (Figure II-14a). Crystallization trials of retinal -bound (M2) was initiated both 20 minutes and 24 hours after retinal addition. Gratifyingly, electron density pr oduced from crystals grown after short retinal incubation clearly showed the presence of a cis -PSB retinal , while crystals grown from protein incubated with retinal for 24 hours produced electron density that was unambiguously that of a trans -SB retinal , a result similar to that of the M1 mutant ( Figure II-14b and Figure II-14c). 00.10.20.30.40.50.60.70.8300400500600700800M2 - PSB loss over time 20min 1h2h3h4h5h6h7h8h9h10hAbsorption Wavelength (nm) a) c) b) Figure II -14) Thermal isomeri zation of a CRABPII rhodopsin mimic. a) PSB loss over time for mutant M2 . M2 was incubated with retinal, and UV/Vis spectra were taken at various time points after the addition ; b) Crystal structure of M2 showing the trans -SB at 1.62 † , with the electron density contoured at 2.0 !, (Yellow, protein was incubated with 4 equiv retinal for 24 hours at room temperature before crystallization, PDB ID: 4YFQ ), overlaid on the cis -PSB of the same mutant crystallized after a 20 minute retinal incubation (Cyan) ; c) Crystal structure of the M2 mut ant in cis -PSB form at 1.97 † at 1.5 !, (Cyan, t he protein was incubated with 4 equiv alents retinal at room temperature for 20 minutes and followed by immediate crystallization , PDB ID: 4YFP ) overlaid on the trans -SB crystal structure of the same mutant (Yellow) . !"#!These results suggest that crystal nucleation occurs rapidly, allowing the less stable cis -PSB retinal to nucleate before signific ant amounts of the trans -SB retinal have formed. ! The M2 mutant has all the necessary characteristics of a functional rhodopsin mimic in that it can interconvert via chromophore isomerization between two distinct iminium pKa environments. However, the key characteristic of such a system is that it must b e photoswitchable . A particularly attractive aspect of the M2 protein is the low pKa (pKa = 3.4) for its 15 -trans retinal form. At a pH of 7, essentially none of this form is expected to be protonated, which means that only the high pKa form will absorb visible light. On the other hand, with a pKa of 8.2, almost all of the high pKa form should be protonated and therefore not absorb in the UV region of the spectrum. In principle, this should make it possible to quantitatively convert one form into the ot her depending on the wavelength of light used for irradiation. To demonstrate this light induced interconversion, we incubated M2 with retinal, and after the PSB absorbance reached its maximum, the complex was exposed to visible light. The PSB absorbance at 571 nm was immediately reduced, reaching a minimum in 1 minute of exposure. To test the systemÕs reversibility, the sample was then irradiated with UV light, resulting in an immediate reduction in the SB peak at 370 nm with a concurrent increase in the PSB absorbance at 571 nm, indicating the system to be reversible. This can be repeated for several cycles, suggesting that the interconversion between the two forms can be repeated (Figure II-15). !"#! We next sought to identify the two interconverting species crystallographically . Different experiments including the UV irradiation of the solution before crystallization and irradiati ng for every 12 hours during crystallization process, UV irradiation of the crystals on the X -ray beam before data collection and UV irradiation of the crystals formed from a solution after PSB loss right before freezing them were tried to prove that the later is the most effective way of approaching this experiment. We took the crystals of retinal -bound M2 that had been grown after 24 hours of incubation, resulting in complete loss of the meta -stable high p Ka form, and soaked them in a stabilizer at pH 7. 5. As shown, these crystals were completely colorless, indicating minimal absorption in the visible spectrum. As previously described, the structure obtained from such crystals shows exclusively a trans -SB bound chromophore ( Figure II-14b). We then expo sed these crystals to UV light, resulting in a dramatic color change of the crystals. The crystals were then 00.20.40.60.811.20246810121416M2 - Cycling Abs at 373 nm Abs at 573 nm Absorption Time (min) 00.20.40.60.811.2300400500600700800M2 - Cycling Retinal-20min Green-1min-1 UV-3min-1 Green-1min-2 UV-3min-2 Green-1min-3 UV-3min-3 Green-1min-4 UV-3min-4 Green-1min-5 UV-3min-5 Absorption Wavelength (nm) Figure II -15) Cycling of the hCRABPII proteins in solution with UV light (340 -380 nm) and Green light (520 -600 nm) . !"#!exposed to visible light resulting in the colorless crystals originally obtained. This process could be repeated several times indicating the rever sibility of the photo-process in the crystal ( Figure II-16b). The crystals exposed to UV light were immediately frozen in liquid nitrogen. The structure obtained from these crystals unambiguously shows a cis -PSB retinal , demonstrating that the UV light-induced isomerization involves the trans to cis isomerization of the imine double bond. For these proteins, the same process that was previously observed occurring thermall y also occurs photochemically. 00.20.40.60.81300400500600700800 M2 - UV irradiation 24h incubation UV340-380-1min Absorption Wavelength (nm) Figure II -16) Photo interconversion in both solut ion and crystal. a) UV irradiation of M2 after complete PSB loss shows the PSB recovery in solution ; b) Cycles of UV and visible light irradiation of M2 crystals . ( Each Cycle is 5 min UV light and 5 min room light at pH 7.5); c) The structure of M2 obtained from a crystal that was UV irradiated for 30 min utes at 1.95 † . The electron density was contoured at 1.5 ! and the structure clearly shows the retinal in a cis -iminium conformation (Magenta, PDB ID: 4YFR ) overlai d on the str ucture of the trans -SB (Yellow, PDB ID: 4YFQ ) and cis -PSB (Cyan, PDB ID: 4YFP ) crystal structure s. a b c !"#!An overlay of the meta -stable 15-cis -retinal structure obtained by immediate crystallization after PSB formation and the UV irradiated crystal structure shows the two to be essentially identical, while the low p Ka 15-trans -retinal structure i s distinctly different ( Figure II-16c). Crystal structures were also obtained of both UV and visible light irradiated crystals after two and three cycles of crystal irradiation for M1 and M2 proteins, to confirm structurally that the photo -isomerization reliably gives the same two products ( Figure II-17). The overlay of the two isomeric forms of M2 indicates that the critical residues surrounding the chromophore are unchan ged ( Figure II-18). However, the most drastic difference is in the trajectory of the lysine 111, which forms the iminium. An isomerization from trans to cis is accompanied by an outward movement of the carbons of the lysine, showing that most of the motion required for the interconversion involves the change of the confor mation of the Figure II -17) Cycling of the M2 mutant crystals: a) Overlay of the retinal for the first three cycles of UV light irradiated crystals of the M2 mutant for 30 minutes. b) Overlay of the retinal for the first three cycles of visible light irradiated crystals of the M2 mutant for 30 minutes. c) Overlay of the retinal for the first three cycles of UV light irradiated crystals (cyan) and visible light irradiated crystals (yellow) of the M2 mutant each for 30 minutes. ¥All of these crys tals diffract between 1.6 -2.2. PDB IDs in supporting information. a b c !"#!lysine side chain , not the chromophore. The last carbon of the Lys111 in 15-cis -retinal PSB overlays with the nitrogen of the same residue in 15-trans -retinal SB form, putting the protonated nitrogen of the lysine in a 3 -3.5 † distance from Trp109, close enough to make a strong !-cation interaction. This up and outward movement of the Lys 111 causes the carbons of the polyene chain and angular methyls to move u p and out to accommodate to the new iminium position, being more noticeable in the first few carbons of the polyene. Towards the end of the polyene and the ionone ring, the outward movement of the chromophore becomes less significant. The only noticeable movement of the residues inside the binding pocket is related to Leu121. In the 15-trans -retinal form this residue has enough space to rotate for two different conformations. However, as the lysine moves out in the 15-trans -retinal form, one of the conforma tions becomes unavailable for Leu121 due t o steric hindrance (Figure II-18). Trp 109 Y134F R59 Y P39Y Leu 121 R132Q R111K T54V 2.6 † 3.3 † 2.8 † 3.5 † Figure II -18) The environment surrounding the iminium in M2 protein . The trans -SB (Yellow) and cis -PSB (Cyan). !"#!Figure 19 summarizes the retinal isomerization in hCRABPII proteins. Starting with a cis -PSB as the initial kinetic form of the protein, a thermal isomerization to trans -SB form reduces the p Ka of the protein below physological pH and blue shifts the protein -retina l complex by 200 nm. This process can be switched back and forth to the cis -PSB and trans -SB form by UV light and green light irradiation, respectively. pKa titration of each of the cis -PSB, trans -SB and UV irradiated forms of these proteins provides a single proton p Ka titration curve . In the case of R132Q:P39Q mutants there is a 2 -3 p Ka unit change between the 15 -cis -retinal and 15 -trans retinal forms of the protein. This difference in the p Ka increases to 5 units in the R132Q:P39Y mutants. The signi ficant p Ka change between the 15 -cis and 15-trans retinal forms indicate two different retinal iminium environments. In the 15-cis retinal form the Lys111 nitrogen faces towards Trp109, ma king a !-cation int eraction ( Figure II-20a). However, in the trans -SB form the Lys111 nitrogen faces hydrophobic aliphatic residues Retinal-Kinetic SBRetinal-Thermodynamic SBheat or green lightRetinal-Kinetic PSBNHNpKa = 8.2N!max = 556 nm!max = 360 nm!max = 360 nmUV lightpKa = 5.1Retinal-Thermodynamic PSBN!max = 540 nmHheat or green lightUV lightFigure II -19) Thermodynamic and light induced cis , trans -retinal imin ium isomerization in hCRABPII protein. !"#!including Val41, Il e43, Leu119 and Leu121 ( Figure II-20b), which leads to a pKa drop. So far, we have successfully designed rhodopsin mimics that recapitulate the basic chemistry of a rhodopsin in that the iminium moiety can be shifted between two distinct p Ka environments via photo isomerization of a polyene double bond. However, all of these systems thermally relax to the lower p Ka all trans -retinylidene PSB. We next sought to create a system where the thermal product was the higher p Ka cis -PSB form. This is the case in many rhodopsin systems such as 11 -cis -retinal in the visual rhodopsins, where the all trans -unprotonated imine is meta stable and ultimately exits the active site. To produce s uch a variant it is necessary to stabilize the cis -PSB retinal form of the protein. We again took a cue from our work on hCRBPII, where we have shown that the presence of a glutamine at position 4 (residue 3 for hCRABPII) gives the cis -PSB, while mutation of this residue gives the trans -iminium (Figure II-12a and Figure II-12b). A water mediated interaction Val41 Leu119 Leu121 R111K Ile43 4.9 † 5.1 † 3.7 † 4.0 † Trp 109 R111K 3.3 † a b Figure II -20) M2 crystal structure indicates the difference in the two form s of the protein . a) A !-cation interaction between Trp109 and the PSB nitrogen of the cis -PSB contributes to the higher pKa in this form ; b) Aliphatic hydrophobic residues surrounding the trans -SB in M2 depress the pKa in this form . !"#!between Gln4 and the imin ium nitrogen explains the stabilization of the cis conformation (Figure II-21). Analogously, we added the F3Q m utation (equivalent to the position 4 in hCRBPII) to the R111K:Y134F:T54V:R132Q:P39Q:R59Y:A32W (M6) scaffold resulting in the R111K:Y134F:T54V:R132Q:P39Q:R59Y:A32W:F3Q mutant ( M3). Gratifyingly retinal -bound M3 gave the most stable higher p Ka form so far observed, with almost 50% of the high p Ka form still present 24 hours after formation. The p Ka of this mutant was also increased by 1.1 units to 9.4 (compared to 8.3 for M6), and the absorbance blue shifted about 27 nm (compare entry M6 to entry M3 in Table 5 ). Exposure of M3 to visible light results in loss of the PSB absorbance with a concomitant increase in SB absorbance, but the PSB absorbance recovers in darkness, indicating thermal isomerization to the higher p Ka cis -iminium form ( Figure II-22). Q108K Retinal Gln4 2.9 † 2.9 † Figure II -21) crystal structure of Q108K:K40L shows that Gln4 contributes to the stabilization of a cis -PSB in the hCRBPII mutants (PDB ID: 4RUU) . !""! Crystallization of M3 was initiated at various times after retinal addition. Incubation of the protein for over 24 hours with retinal followed by an additional 20 minutes of green light prior to the initiation of crystallization was also attempted. Howeve r, only the cis -PSB was seen in the structure, consistent with the ext ensive stabilization of the cis form confered by the F3Q mutation. M3 mutant loses 50% of its PSB in the course of 24 hours. We are not sure what this loss of PSB corresponds to, as the crystal structures show the same retinal geometry (Figure II-23). 00.20.40.60.8300400500600700800M3 - PSB loss over time Retinal-3h Retinal-24h Absorption Wavelength (nm) Figure II -23) PSB loss of M3 mutant over time. 00.10.20.30.40.50.60.70.8300400500600700800M3 - Green light - PSB Recovery Retinal-5h Green-3min wait-10min wait-1h wait-12h Absorption Wavelength (nm) 00.20.40.60.8300400500600700800M2- Green light - PSB Recovery Retinal-20min Green-2min wait-1h wait-3h wait-6h wait-10h Absorption Wavelength (nm) Figure II -22) Mutants M2 and M3 are shined with green light till maximum PSB loss after maximum PSB formation. PSB recovery has been tracked in dark. In 1 hour M3 mutant has significant PSB recovery, while M2 mutant spectrum shows much less PSB recovery. !"#!In addition, the structure does not show a direct or even wat er mediated interaction between Gln3 and the iminium, with the distance between the Gln3 side chain and the iminium nitrogen at about 5 †. It seems that the increased hydrophilicity in the cis -iminium region relative to Phe3 containing mutants is sufficient to produce the higher pKa, blue shifted spectrum, and increased stability of the cis -iminium (Figure II-24). Figure II-24 shows the F3Q mutant structure overlaid with one of the Phe3 containing structures to show how this position defines the hydrophobicity/ hydrophobicity of the iminium environment. The introduced glutamine is fixed in its conformation via a water network t o main chain carbonyls of Asp91 and Arg111 , which does not exist when Phe3 is not mutated . Unfortunately, M3 crystallizes in a crystal form that cannot be photoisomerized in the crystalline state. II.3 Isomerization of hCRABPII mutants in solution As show n above the isomerization of the hCRABPII mutants in the solid crystalline form is very effective, and the electron density maps show the Figure II-24) Structures of retinal -bound M3 (PDB ID: 4YKM) and cis -PSB retinal -bound M2 are overlaid, showing the effect of Gln 3 on the environment about the cis -PSB, increasing the p Ka by 1 unit and altering the nature of the thermodynamic product. 2.2 † R111K Phe3 Trp109 F3Q W1 Asp91 3.3 † 3.5 † 3.6 † 2.6 † 5.0 † !"#!complete conversion of the two forms to each other. Part of our efforts has been to get the same efficiency in the sol ution as well . The mutants that show complete conversion in the solid phase sound the reasonable ones to follow in the solution as well. For this purpose mutant M2 was targeted for following the isomerization in the solution. Protein -retinal complex was irradiated with UV light and green light for multiple cycl es. In figure 15 the cycling of the M2 mutant starting at PSB maximum formation is shown, sta rting by green light irradiation of the solution. It can be seen that less than 50% of the original absorption can be recovered. This indicates the low isomerization yield of the mutant M2 in solution. To solve this problem different strategies were taken. Cycl ing of mutant M2 was done after the PSB loss, and UV irradiating the solution, which restores a significant amount of the PSB in the first cycle, initiated isomerization . The common theme in figure 15 and ( Figure II-25a) is that after irradiating the solution the first time with the green light , not much PSB can be recovered by UV irradiating the solution. Therefore, I thought it might be the green light source that is problematic. In the next experiment room light was us ed as the source of the green light. Exposing the solution after the first time UV irradiation to room light causes the solution to go colorless again in 10 minutes. However, the solution was allowed to s tay in the room light for two hours to make sure tha t all of the protein retinal complex relaxes back to the trans -SB form , but it did not help to get more of the PSB back in the next cycles ( Figure II-25b). In the next experiment the protein was allowed to relax in dark, in the absence of the green light, which proved to be effective (Figure II-25c). !"#! Figure II -25) Cycling of the mutant M2 after PSB loss completion. a) After the first cycle the yield of the switch drops considerably. b) Cycling with room light for visible light irradiation of the solution. c) Cycling in dark with no green light involved. 00.10.20.30.40.50.60.70.8300400500600700800M2 - Cycling after PSB loss Retinal-12h UV1min-1 Green1min-1 UV1min-2 Green1min-2 UV1min-3 Green1min-3 UV1min-4 Green1min-4 UV1min-5 Green1min-5 Absorption Wavelength (nm) 00.10.20.30.40.50.60.70.802468101214M2 - Cycling after PSB loss Abs at 371nm Abs at 572nm Absorption Time (min) 00.20.40.60.811.2300400500600700800M2 - Cycling in Room light Retinal-24h UV-2min-1 Roomlight-2h-1 UV-2min-2 Roomlight-2h-2 UV-2min-3 Absorption Wavelength (nm) 00.20.40.60.8050100150200250300M2 - Cycling in room light Abs at 375nm Abs at 575nm Absorption Time (min) 00.10.20.30.40.50.60.7300400500600700800M2 - Cycling in dark Retinal-24h UV-2min-1 Wait-10h UV2min-2 Wait-6h UV2min-3 Absorption Wavelength (nm) 00.10.20.30.40.50.60510152025M2 - Cycling in dark Abs at 375nm Abs at 575nm Absorption Time (min) a b c !"#! Since the trans -SB form of the protein has a pKa of 3.4 and the cis -PSB form has a pKa of 8.2, the cycling was done at pH 5.5 to check if the isomerization process is pH dependent, but the isomerization yield of the protein -retinal complex was found to be independent from the pH ( Figure II-26). The answer to these observation s might be in Figure 22 that indicates mutant M2 cannot recover PSB thermally after Green light irradiation. Since the trans -SB is the thermodynamically stable form of the protein, after green light irradiation, part of the SB does not isomerize because it is not thermodynamically favored to go to the ki netic cis -PSB form. With the same hypothesis, it is expected that mutant M3 will be able to isomerize in the solu tion more efficiently, because F igure 22 shows that after green light irradiation a lot of the PSB can be recovered in short time thermally proving the cis -PSB to be the thermodynamic form of the protein. Therefore, unlike mutant M2, UV irradiation will favor the formation of the 00.20.40.60.811.2300400500600700800M2 - Cycling at PSB max at pH=5.5 Retinal-20min pHdrop=5.5 UV1min-1 Green-1min UV1min-2 Green-3min UV1min-3 Green-5min UV1min-4 Green-10min Absorption Wavelength (nm) 00.10.20.30.40.50.60.70246810121416M2 - Cycling at maximum PSB formation Abs at 375nm Abs at 575nm Absorption Time (min) Figure II -26) Cycling of M2 mutant at pH 5.5. !"#!thermodynamic form of the protein. Incubation of the M3 w ith retinal till maximum PSB formation followed by green light irradiation and then UV irradiation for multip le cycles shows that more than 5 0 % of the absorpt ion of the PSB can be recovered in agreement with the rationalization described (Figure II-27). II.4 The mystery of colorful crystals of M2 mutant after PSB loss As mentioned above in order to trap the thermodyna mic form of the hCRABPII second generation mutants, M1 and M2 mutants were crystallized after incubation with four equivalents of retinal and 24 hour s wait period for complete PSB loss. Interestingly, the crystals that are grown from the pale yellowish sol ution with free retinal color , which comes from the excess retinal in crystallization solution , turned to be colorful with an intense blue color for mutant M2 at pH 5 and 6 and light blue color at pH 6.5. (Figure II-28). 00.10.20.30.40.5300400500600700800M3 - Cycling Retinal-5h Green-2min-1 UV-1min-1 Green-2min-2 UV-1min-2 Green-2min-3 UV-1min-3 Green-2min-4 UV-1min-4 Green-2min-5 UV-2min-5 Absorption Wavelength (nm) Figure II -27) Cycling of the mutant M3 in solution. More than 50% of the PSB absorption is recovered in each cycle. !"#! For mutant M1 the color of the crystals is pale red at pH 6.0. The reason that crystals are soaked before freezing is to make sure the PSB remained in the crystals convert to SB for a fair comparison between the SB of the 15-trans retinal and PSB of the 15-cis retinal . However, with a low pKa of 3.4 it is not expected for the crystals to be colorful at pHs of 5, 6 or 6.5 . To find out the reason the pKa titration o f the protein was done in 50 mM m alonate, which is the buffer in crystallization condition , as a control for condition, but it did not give a diffe rent pKa than that in phosphate buffer (Figure II-29). Figure II -28) a) M2 SB -crystal grown at pH 5 b) M2 SB crystal grown at pH 6.5 c) M2 SB -Soaked crystal to pH 7.5 d) M2 -UV irradiated crystal after soak for 30 minutes. a) c) b) d) Figure II -29) pKa titration of M2 protein in 50 mM Malonate. 00.10.20.30.40.50.60.72468!Abs. at 574 nm pHM2 -Acid titration in 50 mM Malonate pKa=4.0 00.10.20.30.40.50.60.70.8300400500600700800M2 - Titration in 50 mM Malonate - pKa = 4.0 pH 8.097.286.65.975.124.153.592.92.48Absorption Wavelength (nm) !"#!The next experiment was that after the protein was incubated for 24 hours, the pH of the protein solution was dropped with 50 mM malonate at pHs of 5 and 6 and the protein wa s allowed to sit for 24 hours ( the crystalli zation time ). The same experiment was repeated with addition of 12% PEG3350 which is the precipitant reagent in the crys tallization condition (Figure II-30). The result of this experiment shows that at the time that the protein forms crystals , there is considerable PSB at pH =5 and a little at pH =6, which is probably the reason why the crystals have color. At pH =5 and 12% precipitant reagent (PEG 3350) and 24 hours in the cold room a lot of protein precipitation is observed as well along with the blue color of the s olution. The UV -Vis spectrum of the crystals after the soak and after the UV irradiation was also collected. The comparison of the spectrums shows mar ginal amounts of PSB in the soaked crystals and considerable amounts of the PSB in the UV irr adiated crystals . This provides more evidence that our 00.511.522.53300400500600700800M2- Retinal 24h then 50 mM Malonate Retinal-24h Malonate-pH5-24h-4C Malonate-pH6-24h-4C Absorption Wavelength (nm) 00.511.522.53300400500600700800M2- Retinal 24h then 12% PEG3350 + 50mM Malonate Retinal-24h PEG-Malonate-PH5-24h-4C PEG-Malonate-pH6-24h-4C Absorption Wavelength (nm) Figure II -30) M2 protein was incubated with 4 equivalents retinal and then mixed with the cry stallization condition . The protein was allowed to sit in the cold room for 24h. !"#!comparison of the trans -SB retinal in colorless crystals and cis -PSB in colorful crystals is valid and relevant (Figure II-31). The crystal UV -Vis spectrum of mutants M3 and M11 was also measured, which matches with that of solution Table 5 (Figure II-32). !!00.511.522.53300400500600700800M2- Soaked Crystal-UV Absorption Wavelength (nm) 00.511.522.53300400500600700800M2- UV irradiated Crystal for 30 min-UV Absorption Wavelength (nm) Figure II -31) a) The UV -Vis spectrum of the soaked crystal of M2 mutant b) The UV-Vis Spectrum of the UV irradiated crystal for 30 minues after soak. a) b) 00.511.52300400500600700800M3 - Crystal UV - 552 nm Absorption Wavelength (nm) 00.511.5300400500600700800M11 - Crystal UV - 550 nm Absorption Wavelength (nm) Figure II -32) Crystal UV -Vis spectrum of mutans: a) M3 b) M11. !"#!II.5 hCRBPII isomerization solution and structural studies Along our efforts to develop a photoisomerizing p rotein from hCRABPII and providing structural data for the switch, we have been aiming for the same in hCRBPII protein as well . Developing a photoswichable hCRBPII protein is important as this protein is more sutiable for in -vivo imaging than hCRABPII . Developing a p hotoswichable hCRBPII protein in -vitro can pave the way for the development of a photoswitchable fluorescent protein in -vivo as well . In the anaerobic condition that proteins like GFP and its derivatives does not work this can be of super importance (14-16). hCRBPII proteins also lose their PSB over time . Like hCRABPII proteins in the presence of Gln4 this loss is not complete and up to 50% of the protein stays as PSB even after l ong incubations with retinal. However, when the Gln4 is mutated the PSB loss goes to completion (Figure II-33). 00.050.10.150.20.250.30.35300400500600700800KL - Retinal Incubation retinal-10min 20min 1h2h4h12hAbsorption Wavelength (nm) 00.050.10.150.20.250.30.35300400500600700800KLVCWLWQ4H - Retinal Incubation Retinal-10min 30min 1h2h3h4h6hAbsorption Wavelength (nm) Figure II -33) Retinal incubation of the hCRBPII mutants: Left: Q108K:K40L , Right: Q108K:K40L:T51V:T53C:R58W:T29L:Y19W:Q4H. !"#!The mutants of hCRBPII respond to UV light and green light irradiation as well. In the presence of the Gln4 the yield of the switch is much better th an when this residue is mutated (Figure II-34 Ð with Dr.Tetyana Berbasova Õs permission ). When this residue is mutated, during photo switch cycles a lot of the protein -retinal complex stays as SB and does not participate in the SB -PSB switch. This might be because of the same reasoning that discussed for hCRABPII mutants on F3Q mutation that switches the thermodynamic form of the protein to 15 -cis retinal ratherthan 15 -trans retinal in the absence of this mutaiton . With Gln4 the 15-cis retinal PSB is the thermodynamic form of the protein -retinal complex and will be recovered immediately upon UV light irradiation, but when this residue is mutated the thermodynamic form of the retinal -protein complex is the 15-trans retinal SB , which might not respond to the UV irradiation to go to the kinetic form which is 15-cis retinal PSB . !""! a) b) c) Figure II -34) UV -Gree n light cyc ling of hCRBPII mutants for: a) Q108K:K40L:T51V:T53C:R58W:T29L:Y19W:Q4H b) Q108K:K40L:T51V:T53C:R58W:T29L:Y19W:Q4A c) Q108K:K40L:T51V:R58F. With Dr. Tetyana BerbosvaÕs permission !"#!Our extensive structural data on the hCRBPII mutant series has proven that whenever the Gln4 residue is present, the iminium geometry is cis -PSB retinal and whenever this residue is mutated it is trans -SB retinal . Gln4 stabilizes the cis -PSB retinal via a water molecule that stays in between the nitrogen atom of Schiff base and the oxygen of the amide group of Gln4 . Therefore , the most possible structural explanation for this event is the isomerization of the iminium bond (Figure II-12 and Figure II-21). Our first attempts to structurally trap the two forms of the protein in the same mutant started with allowing the mutants with the Gln4 residue to lose their PSB to the maximum amount and then irradiating the solution with green light and then crystallizing the protein. Our hypothesis was to connect the PSB loss of these mutants to the iminium isomerization . The data collected on the muta nts crystallized this way disapprovingly showed a cis -PSB with no isomerization along the polyene ( Figure II-35). Also the green light irradiation indicates the PSB loss in solution, it is possible that for this mutants since the cis -PSB retinal is the thermodynamic form, the SB form that is made from green ligh t irradiation recovers back to the cis -PSB retinal before the crystallization of the protein or during crystallization. !"#! The overlay structure of the same mutants that are crystallized right a fter PSB maximum formation and after maximum PSB loss shows the critical residues inside the binding pocket in the same place. However, the second angular methyl of the retinal polyene tail rotates around 45 degrees. Whether this conformational change of the polyene is related to the PSB loss of the protein or not , it does not change the !max of the protein as there is no shift seen in the UV-Vis spectrums over time (Figure II-33 and Figure II-36). 01234300400500600700800Q108K:K40L:T51V:T53C:R58W:T29L:Y19W - Retinal incubation - Green light Retinal-16h Green-2min Green-6min Green-12min Absorption Wavelength (nm) 00.511.522.533.5300400500600700800Q108K:K40L:T51V:R58F - Retinal incubation - Green light Retinal-16h Greenlight-5min Greenlight15min Greenlight30min Greenlight1h Absorption Wavelength (nm) Figure II -35) Retinal incubation followed by green light irradiation for: a) Q108K:K40L:T51V:R58F mutant of hCRBPII and b) Q108K:K40L:T51V:T53C:R58W:T29L:Y19W mutant of hCRBPII . c) Electron density map of the crystal obtained from Q108K:K40L:T51V:R58F mutant contoured at 1 " shows a cis -iminium and not a trans -iminium. ¥Q108K:K40L:T51V:T53C:R58W:T29L:Y19W mutant crystal structu re also shows a cis -iminium, but it only shows the density for the first five carbons close to the iminium. c) b) a) !"#! Being unable to correlate the PSB loss of the hCRBPII mutants with retinal isomerization, we decided to find a way to prove the conversion of the trans -SB to cis -PSB. As mentioned above mutants with Gln4 mutation show a trans -SB. Therefore, we decided to crystallize the se mutants after complete PSB loss , soak and UV irr adiate the obtained crystals and then freeze them and collect crystallography data . We were hypothesizing that the UV T53 Q38 R58 Q128 T51 Q108K K40L Figure II-36) a) Overlay structure of the mutant Q108K:K40L after 20 minutes incubation (Cyan , PDB ID: 4EXZ ) and after 24 hours incubation (Yellow , PDB ID:4RUU ). b) Overlay structure of the mutant Q108K:K40L :T51V:R58F after 20 minute s incubation (Cyan) and after 16 hours incubation and 1 hour green light (Yellow). Q38 Q4 T51V Q108K Q128 R58F K40L a) b) !"#!irradiation will isomerize the iminium back to cis . This experiment was done for the Q108K:K40L:T51V:T53C:R58W:T29L:Y19W :Q4H mutant. The c rystals have a pale grey color, and after the soak they lose the color, by UV irradiation some of the color is returned back. Although the color change after the UV irradiation was recognizable , the data collection on these crystals shows the trans -SB reti nal again. The structural isomerization experiments done with hCRBPII prove again that the Gln4 in hCRBPII and F3Q mutation in hCRABPII thermally stabilizes the cis -PSB that no iso merization is seen overtime despite the partial PSB loss overtime . In hCRABPII for these mutants the structural data shows no retinal isomerization or any conformational changes. In hCRBPII although so far no isomerization is recognizable for these mutants, but the position of angular methyls change overtime ( Figure II-36). The PSB loss can be seen in hCRBPII mutants with Q4X mutation as well, but so far we have not been able to characterize the initial kinetic PSB form that is made. The fact that all of the hCRBPII mutants with Gln4 conserved show a 15 -cis retinal and all o f the mutants with Q4X mutation show a 15 -trans retinal suggests the isomerization of the imine double bond as a possible explanation for the PSB loss over time, but this isomerization has not been structurally characterized in the same mutant of hCRBPII. The hCRBPII retinal isomerization can be tracked in solution. Different mutants of this protein with and without Gln4 mutations show the SB -PSB conversion upon irradiation with green and UV light. However, so far we have not been able to structurally eluci date the light irradiated forms of the protein . !"#!II.6 Conclusion Using a rational design strategy based on the Cellular Retinoic Acid Binding Protein II scaffold, we have created a series of rhodopsin family mimics that reproduce the basic chemistry in these systems, in that photoisomerization switches the retinal between two distinct p Ka regimes. We have further demonstrated the ability to independently control both p Kas, and alter the thermal product from the low to the high p Ka moiety. Both forms ha ve been unambigously identified from high resolution crystal structures, and photo -isomerization in the crystalline state was also demonstrated, with quantitative conversion between each form shown at atomic resolution. Besides photoswiching, the clean an d complete dark isomeriz ation is unique of this protein . This represents a retinal -based photoswitchable protein that has a variety of potential applications from a rhodopsin photoswitch model to a fluor escent quencher for further development of photoswit chable imaging techniques. The ability to control the basic chemistry of each form will be an exceedingly valuable tool in these applications. It can be used as a valuable tool for time resolved spectroscopy and crystallography techniques to unveil the intermediates of the retinal isomerization in this system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`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hapter III : Theories of Wavelength Regulation III.1 Introduction In chapter 1 the mechanism of vision was explained. Four different type s of Rhodopsin proteins all work with the same chromophore 11 -cis retinal to absorb light . One of the long standing questions that in the last 50 years scientists have been trying to answer is the wavelength regulation question: how these proteins with the same chromophore absorb different wavelengths of light. As mentioned Rhodopsins work via the formation of an iminium Protonated Schiff Base (PSB) . The 11 -cis retinal chromophore by itself absorbs at 380 nm. The 11 -cis retinal Schiff Base (SB) , which is not protonated , absorbs at 365 nm and upon protonation it will absorb at 440 nm in ethanol solution . The absorption of the n-butyl retinal iminium PSB can be tuned up to 500 nm depending on the solvent (Figure III-1) (1, 2). This can span the absorption of the Blue and Rod Rhodopsin, but not the Green and Red Rhodopsin, which is at 530 and 560 nm respectively . The difference of the absorption of retinal PSB in protein compared to the ethano l solution (440 nm) is referred to ÒOpsin s hiftÓ, which is attributed to the interac tion of the retinal PSB with protein environment . NHOH1112N1112111211-cis retinal11-cis retinalSchiff Base11-cis retinalProtonated Schiff Base!max = 380 nm!max = 365 nm!max = 440 nmFigure III -1) structure and absorption of 11 -cis retinal, its n -butyl amine Schiff Base and Protonated Schiff base. !"#!The fact tha t none of the colored opsins is crystallized so far has hampered the development of a solid answer to the wavelength regulation question. However, extensive studies that have been done in the last few decades have elucidated a lot of the governing principles in wavelength regulation. The contributing effects have been attributed to: (1) Point charges or dipole moments from the c harged or polar residues . (2) Conformational geometry of the retinal (3) Polar w ater networks (4) Ground state (S0) -Excited st ate (S1) dynamics of the retinal (5) Excitonic coupling to Aromatic Residues III.2 The Point Char ge Theory In brief , the point charge theory emphasizes the stabilization of the positive charge of the iminium via negative c harge or negative dipole moment directly interacting with the iminium nitrogen or along the retinal polyene . If the negative charge or negative dipole moment is close to the iminium or toward the beginning of the polyene the protein is blue shifted. On the other hand , when it is toward the end of the polyene the positive charge is more delocalized and the protein is red shifted (Figure III-2). Figure III -2) The position of the putative negative charge or negative dipole moment determines the absorption of the protein. !""!The formation of a covalent bond between a lysine residue of rhodopsin and retinal was confirmed by different studies via reductive amination and proteolytic fragmentation decades before the crystal structure of the bovine rhodopsin (3-5). The protonation of the Schiff Base was confirmed by resonance raman spectroscopy (6), and NMR studies by Nakanishi and others proved the presence of a negative charge adjacent to the PSB positive charge to be the reason for the stabilizat ion of the positive charge of the PSB (7-9). The pKa titration studies of retinal analogs postulated the pKa of the bovine Rhodopsin to be around 16. The pKa of 16 is nine units above the pKa of a PSB in solution (10). Oprian mutated all of the Glu and Asp residues buried in Rhodopsin to figure out that Glu113 is the only important residue for the wavelength and pKa of the Rhodopsin. E113Q mutation dropped the pKa of the protein to about 6 and blue shifted the protein to 380 nm where the SB absorbs (11). Mutating th is counter anion to different residues can maximally red shift the protein by 30 nm depending on the solute anion. The red shift of the protein is expected as the removal of the negative charge of the counter anion will increase the delocalization of the c harge on the retinal chromophore (12). Studying the model compounds of retinal showed th at both the angle and the distance of the counter anion is important in the absorption and pKa of the retinal PSB in solution (13). After the crystal structure of the Bovine Rhodopsin , Glu113 was found 3.45 † away from the PSB nitrogen (14). Palczewski modeled the structures of the rest of the colored pigments based on the structure of the bovine rhodopsin (15). Besides the importance of the counter anion effect in the pKa and wavelength of the rhodopsins , other studies attempting to convert the !"#!absorption of the colored opsins to one another shed more light into the point -charge model theory. Mutation of th ree residues in Green Rhodopsin including A180S, F277Y and A285T provides a pigment like Red Rhodopsin (16, 17). These residues are all toward the ionone ring of the chromophore. Introducing these polar groups will increase the negative dipole moment toward the end of the retinal polyene and causes the red shift observed. However, for converting the Red opsin to Green there are more mutations needed which includes S116Y, S180A, I230T, A233S, Y277F, T285A and Y309F , but the majority of the spectral shift is related to the mutations at position 180, 277 and 285 (18). In Rhodopsin F261Y/A269T mutant, equivalent of F277Y/A285T in Green opsin, red sh ifts by 20 nm to absorb at 520 nm, while the triple mutant does not provide a detectable PSB (Figure III-3) (19). 3.3 † 3.2 † A180S F277Y A285T Figure III -3) Overlay structure of the models of the Green opsin (Green - PDB ID: 1KPW) and Red opsin ( Magenta - PDB ID: 1KPX). Mutating the highlighted residues in Green to equivalent residues in Red will provide a Red absorbing like pigment. Lys 312 !"#!In another study the important residues in blue shifting the rhodopsin from 500 nm to 438 nm were characterized. These included nine mutations of M86L/G90S on Helix 2, A117 G/E122L/A124T on Helix 3, W265Y on Helix 6 and A292S/A295S/A299C on Helix 7 (20). These mutations are mainly th e introduction of a polar group toward the PSB and E122L mutation removes a negative charge close to the ionone ring of the chromophore (Figure III-4). In another study the triple mutant of T118A/E122D/A292S was shown to be a functional mutant with 50 % rate of transducing activation compared to the wild type protein absorbing at 453 nm (21). The difference of the absorption maxima between the Green opsin and Rhodopsin has been attributed t o anion binding sites near PSB. Cl - and Br - has been found as the two anions that can cause these spectral shifts. The Figure III -4) Overlay structure of the Rhodopsin (Green - PDB ID: 1F88 ) and the model of Blue opsin ( Blue - PDB ID: 1KPN ). The mutation of the highlighted residues on Rhodopsin to equivalent residues of Blue provides a pigment absorbing at 438 nm. The residue numbering is that of Rhodopsin. Lys 296 Lys293 A292S A299C M86L A124T A117G G90S E122L A295S W265Y !"#!interacting residues are i dentified to be His 197 and Lys 200 . These residues are conserved in G reen and Red o psins and absent in Rhodopsin and Blue Opsin. The increased hydrogen bonding between the PSB and waters in the absence of the chloride causes the blue shift observed in Rhodopsin and Blue opsin. The mutation of these residues in Green opsin can provide pigments that absorb at 500 nm like Rhodopsin. The mutants H197E/K200Q, H197E and K200Q absorb respectively at 500, 500 and 523 nm (Figure III-5) (22-24). From the studies that are done in converting t he absorption maxima of the opsin retinal pigments to one another, it is clear that the pol arized groups along the polyene, especially hydroxyl containing amino acids , play an important role in the wavelength regul ation of these pigments. Positively charge d amino acids , by providing chloride binding sites , and Glutamic acid residues , by introducing negative charge along the polyene or by attracting waters , can also regulate the wavelength absorbed by these proteins. Lys 312 Lys 296 Figure III -5) The overlay structure of the model of Green opsin (Green -PDB ID: 1KPW) and Rhodopsin (Yellow -PDB ID: 1F88 ). His 197 Glu 181 Lys 200 Gln 184 !"#!Quantum mechanical and molecular mechanical calculations on some of the mutants described above was done in 201 1 to rationalize how the electronics of the ground state and excited state of the retinal PSB changes upon the mutation of residues toward the PSB or toward the ionone ring in rhodopsin (Figure III-6) (25). The mutations that were studied are: (1) E113D, which is changing the counter an ion to shorter Aspartic Acid residue that was shown to cause 10 nm red shift by Sakmar (12). (2) A269T, which caus es 15 nm red shift, and is toward the ionone ring (19). (3) T118A, which causes 16 nm red shift which is toward the middle of the retinal polyene (21). (4) A292S, which causes 9 nm blue shift and is toward the iminium nitrogen PSB (20). S0S1WTMutantLys 296 T118A A269T A292T E113D 5.2 † 3.0 † 3.5 † Loss of Excited State stabilization Destabilization of Ground State Stabilization of Ground State Excited State stabilization 3.2 † S1S0WTMutantS0S1WTMutantS0S1WTMutantFigure III -6) The effect of mutating or introducing hydroxyl containing residues or charge residues along the retinal polyene of Rhodopsin . !"#!In addition, Mathies and Sakmar studied resonance Raman vibrational spectra of the rhodopsin visual pigments and concluded that the main factor for the opsin shift in color vision is the interaction of the dipolar amino acids with the ground state and excited state cha rge distributions of the chromophore (26). Electrostatic potential calculation for three bacterial rhodopsins i ncluding sensory rhodopsin II (SRII) 486 nm, bacteriorhodopsin (BR) 552 nm and halorhodopsin (HR) 576 nm was done for the assessment of the distribution of the charge along the retinal polyene and its correlation to the absorption of these proteins. Since the chromophore geometry is the same in these bacterial rhodopsins , the absorption difference has been attributed to changes in the distribution of the charge and dipole moments from the protein. For the sensory rhodopsin its absorption is close to 6 0-90 nm more blue shifted compared to the other two bacterial rhodopsins . These calculations show the most nega tive electrostatic potential near th e PSB region of the retinal while for the other two t his negative electrostatic potential is more localized toward the ionone ring. Bacteriorhodopsin and halorhodopsin differences ar e less compared to the sensory rhodopsin II. These calculations emphasize one more time the importance of where the negative dipole moment is set along the retinal polyene. For these theoretical calculations the electrostatic potential of the retinal has s et to zero and the one projected from the protein onto the surfa ce of the retinal is calculated ( Figure III-7) (27). !"#! Dr.Lee in Borhan lab did the same calculation for the Rhodopsin and three other color opsins. The same ideas rationalized for the wavelength regulation concluded so far can be one more time taken from these calculati ons as well (Figure III-8) (28). Amon g the bacterial rhodopsins sensory rhodopsinII (SRII) absorbs ~ 500 n m, while others absorb 560 -590 nm. SRII is the light pigment that mediates the phototaxis in bacteria to avoid harm ful blue light. In an effort to understand the blue shift of SRII, t he sequence of Bacteriorhodopsin (BR) and SRII where Figure III -7) Electrostatic potential of BR, HR and SRII at the van der Waals surface of the retinal. The magnitude of the potential is color coded. PDB IDs used are : BR ( PDB ID: 1QHJ), HR ( PDB ID: 1E12) and SRII ( PDB ID: 1JGJ). Figure III -8) Electrostatic potential calculation for Rhodopsin, blue, green and red opsin. !"#!compared and three residues where found to be important in the wavelength shift. These three residues are conserved in the other bacterial rhodopsins, while they are not the same in SRII. They include Va l108, Gly130 and Thr 204, which the equivalent residues in BR a re Methionine, Serine and Ala. The single mutants of V108M, G130S and T204A were absorbing respectively at 502, 503 and 508 nm. The triple mutant V108M/G130S/ T204A provided a pigment, which abs orbed at 515 nm. Therefore , these residues just count for 30% of the opsin shift between the BR (560 nm) and SRII. Gly 130 and Val 108 are toward the end of the retinal chromophore, while Thr 204 is close to the PSB (29). A closer comparison between the residues in the binding pocket of the two proteins highlighted seven residues that are different, which includes Ile 43, Ile 83, Asn 105, Val 108, Phe 127, Gly 130 and Phe 134. Mutation of these residues to equivalent positions in BR to provide I43V/I83L/N105D/V108M /F127W/G130S/F134M mutant provides a pigment that absorbs at 509 nm. In the search for more differences in the residues inside the binding pocket three more amino acids w ere highlighte d within 5 † distance of the chromophore. These three residues are Met 109, Ala 131 and Thr 204 . The addition of these mutation to provide the mutant I43V/I83L/N105D/V108M /F127W/G130S/F134M/A131T/ M109I/T204A, provided a pigment which absorbs at 524 nm. Th is still counts for just 44% of the Opsi n shift between the SRII and BR. It was concluded that other structural factors like the angle between the !-ionone ring and polyene chain or the distance between the Schiff Base and its counter anion might be the co ntributing factors (Figure III-9) (30). !"#! It was shown in other studies that L93A or M145A mutations on BR, which are the F134M and I83L mutations above, largely blue shifts the protein , changes the isomer content of the protein and affects the kinetic of the photoisomerization of t he BR during its photocy cle (31, 32). Soon after these findings, the SRII crystal structure was published and the retinal polyene and its ring were found planar as BR. The two Aspartic acid residues, Asp 75 and 201 that act as counter anion for retinal PSB were also in identical positions as BR. However, three structural factors that contribute to the blue shift of the SRII are explained as: (1) The displacement of Arg 72 by 1.1 † and its conformational change with a rotation away from the Schiff base that strengthens the PSB and its counter anion interaction . The repositioning of the Arg 72 is attributed to the helix backbone differences between BR a nd SRII that change s from Phe 208 to Ile 197, Glu 194 to Pro 183 and Glu 204 to Asp 192. Also , the position of other Lys 205 I43V I83L M109I V108M T204A N105D F127W F134M G130S A131T Figure III -9) Overlay structure of the BR (Green - PDB ID: 1C3W) and SRII (Cyan -PDB ID: 1JGJ). The residue numbering is that of SRII. !"#!residues inside the binding pocket of the SRII is altered by interacting with the residues outside of it. (2) Removal of two hydroxyl group s near the !-ionone ring, which are Ser 141 and Thr 142 in BR. (3) Change in the tilt and position of the retinal which alters the interaction of the retinal polyene with the binding pocket, res ulting in the modulation of the Ground state and Excited state energy levels. It was rationalized that although earlier studies highlighted the differences in amino acids in the binding pockets of the two proteins and indicated the 44% of the opsin shift between the BR and SRII, but the aforementioned structural differences were not considered as part of the differences between the two proteins (Figure III-10) (33). The most updated study on the point charge theory model was done in professor Borhan and Geiger lab at Michigan State University. In a 2012 science paper human Cellular Retinol Binding Protein II (hCRBPII) was Arg 72 Asp 201 Asp 75 Lys 205 Figure III -10) Overlay structure of the BR (Green - PDB ID: 1C3W) and SRII (Cyan -PDB ID: 1JGJ). The re sidue numbering is that of SRII. The differences in the trajectory of the retinal and the Arg 72 residue are highlighted. !"#!reengineered to a rhodopsin mimic. The wavelength of the protein was tuned from 425 nm to 646 nm. The introduction of counter anions close to the PSB works well for blue shifting the protein as well a s the elimination of the counter anions and waters near the PSB that red shifts the protein. Counter intuitively, the introduction of negatively charged amino acids at the ionone ring of the retinal does not red shift the protein, instead the introduction of the aromatic bulky residues like Trp, Phe and Tyr not only red shifts the protein, but increases the effect of the internal mutations by a factor of 1 -2. The same electrostatic calculation that was done for the microbial rhodopsins was done by the authors on the two mutant proteins absorbing at 508 nm and 622 nm. The result of these calculations indicates that the key element for red shifting a retinal pigment is not in changing the negative electrostatic potential from the PSB to the ionone ring, but in equally distributing it along the retinal chromophore (34) (Figure III-11). The electronic polarization of the chrom ophore by aromatic residues (35-39) or the overall en vironment of the binding pocket (40-42) has been implicated as contributing factors in the wavelength regulation of the retinal PSB in Rhodopsins as well. S21 As shown in Figure S5a, the KL (M4) mutant exhibits overall localized potentials on the two ends of the chromophore. While a negative potential (in red) is localized on the Schiff base region, a positive potential (in blue) is observed around the !-ionone ring . While previous hypotheses would have predicted that an electros tatic potential polarized opposite to that seen in Figure S 5a would lead to maximally red -shifted pigments, this was not the case. Figure S 5b depicts the electrostatic potential for KL:T51V:T53C:Y19W:R58W:T29L:Q4R (M10). To achieve the maximum bathochromic shift, an evenly distributed electrostatic potential was accomplished through judicious mutagenesis of 8 amino acid residues . The resultant protein red -shifts substantially to 622 nm; the electrostatic potential calculated from its crystal structure and p rojected onto the surface of the chromophore illustrates the uniformly neutral distribution of the electric field, in agreement with our proposed hypothesis. a b Figure S5. Electrostatic potential calculations of a) KL ( M4, 508 nm) and b) KL: T51V:T53C:Y19W:R58W:T29L:Q4R ( M10 , 622 nm) hCRBPII mutants (k bTec-1, APBS program suite) using the cryst al structures of the corresponding mutants. The picture depicts the electrostatic of the protein's binding site, as projected on the Van der Waals surface of retinal (the electrostatic potential of the retinylidene chromophore was set to zero to illustrat e the electrostatic contribution of the protein; both faces of the chromophore are shown). Figure III -11) Electrostatic charge distribution on the surface of the retinal projected from the protein for left: Q108K:K40L mutant of hCRBPII, !max = 508 nm. Right: Q108K:K40L:T51V:T53C:R58W:T29L:Y 19W:Q4R, !max = 622 nm. !""!III.3 The C onformational Theory In the conformational theory the emphasis is on the retinal traj ectory and the planarity or twist of the chromophore. The binding pocket of the protein can force the retinal to keep its double bonds in plane with each other or it can force it to twist. The net result of the planarity of the chromophore is that the double bonds will stay in conjugation , and the positive charge of the PSB can travel all the way to the end of the polyene. However, when it is twisted then the do uble bonds will be less conjugated and the PSB positive charge will not delocalize which translates into blue shift (Figure III-12). The reason for the twist of the retinal chromophore is in the steric repulsion of its methyl groups with its hydrogens. As shown above for 11 -cis retinal the C -13 and C -5 methyl groups are in steric clas h with the hydrogens of C -10 and C-8, respectively. Therefore, the molecule can relieve stress by rotating around any of the planes that are highlighted. Depending on how the residues inside the binding pocket of a retinal pigment protein are arranged, the Figure III -12) The planes of rotation for 11 -cis retinal. !"#!dihedral angle of these planes can c hange, which will directly translate into the wavelength absorbed by the protein (43-48). In 2010 the gas phase absorption of retinal n -butyl iminium PSB was measured by Garavelli , and showed a broad flat top absorption from 530 to 610 nm. This broad peak was attrib uted to conformational changes of the retinal and in particular to the !-ionone ring rotation, and the dihedral angle between the C5 -C6 and C7 -C8 double bonds of retinal. Retinal analogs with locked conformation of 6s -trans and twisted 6s -cis were made and interestingly, the absorption of the 6s -trans analog fell on the 610 nm edge and the absorption of the 6s -cis fell on the 530 nm edge of the spectrum. The difference in the S0 energy level between these forms of the retinal was found to be 0 .6 kcal/mol and up to 4 kcal/mol for the analogs . The absorption of all of the bacterial rhodopsins that have 6s -trans retinal PSB is blue shifted compared to the gas phase absorption of 6s-trans retinal PSB . The absorption of Rhodopsin and color opsins, w hich have twisted 6s -cis retinal PSB is also blue shifted compared to the gas phase absorption of 6s -cis retinal PSB in the gas phase . From these it was concluded that retinal proteins mask and smoothen the blue shifting effects of the counter anion. Excep tions to these were the SRII and blue opsin, which absorb at significantly blue shifted wavelengths. Therefore, the 610 nm absorption of the all -trans retinal PSB in gas phase was made as the upper limit and most red shifted that a retinal PSB can absorb (45) (Figure III -13). Twisted 6s -cis retinal conformation in Rhodopsin and color opsins and 6s -trans retinal in bacterial rhodopsins is one of the contributing factors to the fact that the first absorb all below 560 nm while the later can absorb up to 590 nm. !"#! However, hCRBPII -reengineered proteins absorbing as red as 646 nm hit this barrier, and to date are the most red shifted a retinal PSB can absorb (34). The most recent study in taking advantage of the 6s -trans to 6s -cis retinal isomerization is related to the mutagenesis studies in making blue shifted absorbing pigments of Channelrhodopsin s (ChRs) and Archaerhodopsin -3 (AR-3) as optogenetic tools. It was shown that by two mutations on C1C2 ChR and four mutations on AR -3 the wavelength of the protein can shift by NH56786s-cis-retinal PSB!max = 530 nm5678NH6s-trans-retinal PSB!max = 610 nm5678NH1,1-di demethyl retinal!max = 618 nmNH56785,6 - dihydro-retinal!max = 525 nma) b) Figure III -13) The gas phase absorption of: a) retinal n -butyl iminium PSB. b) analogs of retinal n-butyl iminium PSB . !"#!100 nm. The crystal structure of the C1C2 was reported in 2012 showing a 6s-trans retinal inside the binding pocket of the protein like the other bacterial rhodopsins (49). However, the crystal structure of the double mutant of this protein showed a 6s -cis retinal structure. The double mutant absorbs at 455 nm, while the wild type protein absorbs at 476 nm. AR -3 absorbs at 550 nm and mutants of this protein by three or four mutations were blue shifted all the way to 450 nm (Figure III-14) (50). a) c) d) b) Figure III -14) The studies on developing blue shifted bacterial rhodopsin optogenetic tools: a) The UV -Vis spectrum of blue shifted C1C2 Channel Rhodopsin (ChR) proteins. b) The UV -Vis spectrum of blue shifted archaerhodopsin -3 ( AR-3) proteins. c) The crystal structure of the wild type C1C2 protein shows 6s -trans retinal (PDB ID: 3UG9) . d) The crystal structure of the T198G:G202A mutant of C1C2 shows 6s -cis retinal (PDB ID: 4YZI) . WT T198G/G202A !"#!III.4 Water M olecules in Waveleng th R egulation Waters play an important role in protein functions, especially when interacting with conserved residues. Most of the studies on the effect of the water molecules on retinal signaling proteins have been accomplished in the last decade, and over time the effect of these waters have been more and more elucidated. It has been shown that waters play a critical role in wavelength regulation of Rhodopsin and visual pigments (51, 52) as well as their signaling (51, 53-58), wavelength regulation of bacteriorhodopsin (59) and its proton pumping activities (60, 61), bulk water in retinal -protein complex hydrolysis (62) and signaling of squid rhodopsin (63). The first crystal structure of bovine rhodopsin at 2.8 † resolution failed to highlight some of the water m olecules important in the signaling and wavelength regulation of Rhodopsin (14). In the next few years , two more structures with 2.6 † (51) and 2.2 † resolution (53) showed two water molecules inside the binding pocket of the Rhodopsin near the PSB important for the wavelength regulation of this protein and theoretical calculations showed that each one can contribute up to 32 nm to wavelength shift of the protein (52). These two waters were called Wat2a and Wat2b (Figure III-15). !"#! Wat2b is hydrogen bonding with Glu113, the counter anion of the Rhodopsin. It possibly contributes to further stabilization of the negative charge of the counter anion. The further stabilization of the negative charge of the counter anion might directly translate into the wavelength shift of the protein as well as the high pKa of 16 . Wat2a is hydro gen bonding with Glu181 and Ser 186 and the hydrogen bonding network is extended to Glu113 indirectly via the main chain carbonyl of Cys187. The position of Wat2a is criti cal in terms of the color tuning of opsins, bec ause in Green and Red opsin Glu 181 is replaced by a h istidine residue that facilitates the binding of Chloride anion in the binding pocket of the protein. Wat2a in rhodopsin most probably sits in this position and contributes to the blue shift of the Rhodopsin and blue opsin compared to the green and red opsin. It has been shown that an E181Q mutation can cause 10nm red shift, which might be because of the change of the polarity of the water due to the mutation (64). Glu 113 Cys187 Ser186 Glu181 Lys 296 3.1 † 3.5 † 2.8 † 3.4 † 2.9 † 2.6 † 2.5 † Figure III -15) The position of Wat2a and Wat2b in the binding pocket of Rhodopsin and interacting residues (PDB ID: 1L9H ). Wat2b Wat2a !"#!It has been also suggested that Wat2a is the water molecule that facilitates the switch of the counter anion from Glu113 to Glu181 in the photocycle of Rhodoopsin. An overlay structure of the 2.6 † resolution structure of the dark state with MetaII proves this hypothesis (53, 58) (Figure III-16). These structures also highlighted three water molecules Wat1a, wat1b and Wat1c toward the cytoplasmic side of the rho dopsin that are important in it s signaling. From these structures it was known that these waters inter act with the highly conserved motif of N302P303XXY306X5,6 F on helix 7. When the crystal structure of the Meta II rhodopsin was solved in 2012 by Ernst and co -workers , it showed a clear picture that the signaling of the Rhodopsin relies on a propagation of a water netwo rk that is made by structural rearr angement of the conserved motif on Helix7 and another conserved motif of E 134(D)R 135Y136 on Helix3. The water network spans all the way from the binding pocket of the protein to the cytoplasmic surface of the rhodopsin wh ere the G -protein binds (Figure III-17) (58). Glu113 Lys296 Glu181 Ser186 Tyr268 3.3 † 3.4 † 2.7 † 3.3 † 2.7 † Wat2a Figure III -16) The switch of the counter anion to Glu181 in the MetaII Rhodopsin -transducin complex . a) Glu181 as counter anion in MetaII Rhodopsin - transducin complex (Dark blue Ð PDB ID: 3PQR). b) The overlay structure of the ground state Rhodopsin (Yellow Ð PDB ID: 1L9H) and MetaII Rhodopsin Ð transducin complex (Blue - PDB ID: 3PQR). Glu181 Ser186 Lys296 Glu113 2.0 † Wat2a !"#! In bacterio rhodopsin it was also show n that an L93A mutation causes 30 nm blue shift and the crystal structure of the protein showed that the cavity, which is made by this mutation, is occupied by four water molecules that contribute to the blue shift of the protein (Figure III-18). L93A mutation also increases the lifetime of the O intermediate of the bacteriorhodopsin photocycle by a factor of ~250 (59). a) b) Figure III -17) a) water network in Rhodopsin structure (PDB ID: 1U19). b) Propagation of the water network all the way to the cytoplasmic surface of the Rhodopsin in the crystal structure of MetaII Rhodopsin with Ga CT2 peptide fragment of G -protein (PDB ID: 3PQR). !"#! The importance of the water molecules in the wavelength regulation of retinal proteins was highlighted one more time in the studies of reengineered hCRBPII p rotein done in Professor Borhan and Geiger lab at Department of T53 T51 Q4 Q108K Q38 Q128 Wat1 Wat2 Wat4 Wat3 25-30 nm 20-25 nm Next Chapter Figure III -19) The water molecules crucial in the w avelength regulation of hCRBPII (PDB ID: 4RUU). Figure III -18) Overlay structure of the L93A mutant of Bacteriorhodopsin (Gold Ð PDB ID: 3VHZ) with wild type Bacteriorhodopsin ( Cyan Ð PDB ID: 1C3W). Waters 701 -704 are inserted in the cavity made by L93A mutation near retinal PSB. !"#!Chemistry of Michigan State University (Figure III-19) (34, 65, 66). At the end of this ch apter it should be highlighted that the polar amino acids , specially the hydroxyl containing side chains, and the charged amino acids along the retinal chromophore , the twist of the retinal along its double bonds specially the ionone ring and the crucial r ole of water molecules are the main contributing factors to the wavelength regulation of retinal light absorbing pigments. The next chapter will be the continuation of the reengineering of the hCRBPII protein and a more coherent story of the role of waters that puts all of the pieces of the puzzle together to prove the unanimous effect of the conserved waters in the binding po cket of hCRBPII that govern the wavelength regulation of this protein like the other retinal protein pigments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`%^%:RMP!O%R\^WM[.:![:!`[C\W^!5[RL%:MC$! -$.'%)/&$ 0&4+(& ;3('"*)%&=+(3"*)/&@$*"(45 !#H#6@I9BJH8T;JH87$!!"$!L+11&(4&)/? !6#"8J9!P[RP ;O%C.^\M[.:!C.^[V ;CMWM%!Z ;#I;:LO! CM\Va!._!ZWO0.:C!Z ;K!W:V!Z ;#@!._!MP%!ZPO.L.5P.O%!._!0.`[:%! OP.V.5C[:! ;!%`[V%:Z%!_.O!W!7 ;C;Z[C!Z.:_.OLWM[.:!Y[MP! :%RWM[`% ;ZPWOR%!5%OM\O0 WM[.:!:%WO!Z ;#@$!A.'$<()%&-$.'%)/&$0& 7"$*+(3"#4'5 !#7I6#9B";#T$!!#H$!C3<',F<&G!RN!.33+1<,GD'!LN!/!CD! C(D'EE;02*+A*',! ;!2!C3S>=!U'3D!W&3'E'('21!5'G?<,3*$! 7"$<+5#"*)/&8$.'%)/ !7T6K9B#T"";#KH@$!!##$!bDS)+X*)=!%W!/!.A&'2,!VV!6#"8"9!%EE<(3!+E!(2&F+c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`L!RaW!VFU!]W!Y!-PHF>Q!''!5"66=7!,L?B@K?>H!OBJBHUFQ>QJD!LM!NKU>Q! HBOb[HBBQ!@L?LH!OFD@HFUFQ>JFLQ3! =,#$"% !";5478""9"<""9T3!!"63!0N>Q!%W!/BB!,W!Y!$>EU>H!%1!5"66;7!*QJHLOK@JFL Q!LM!NIOHL_I? <^B>HFQ[! >UFQL!>@FOD!@>KDBD!^>JNL@NHLUF@!DPB@JH>?!DNFMJD!FQ!HNLOLPDFQ3!RUFQL!>@FO! DK^DJFJKJFLQD!HBDPLQDF^?B!MLH!HBO <[HBBQ!@L?LH!PF[UBQJ!DPB@JH>?!JKQFQ[3! ?+,( !"#$%&'(")(7/"'"QFD UD!LM!DPB@JH>?!JKQFQ[!FQ!^?KB!@LQB!ZFDK>?! PF[UBQJD3!cFDF^?B!>QO!H>U>Q!DPB@JHLD@LPI!LM!^?KB QJD3! ?+,(!"#$%&'(")(7/"'"QA!],!Y!X>HHBQD!'/!5;##"7!.Q[FQBBHFQ[!>!MKQ@JFLQ>?!^?KB ZB?BQ[JN QJ3! 6/"0+,./9*$3 !=#5;=78:;"6<:;;:3!!;;3!G?BFQD@NUFOJ!]!Y!+>HLDF!X*!5"66;7!RQFLQ!DBQDFJFZFJI!>QO!DPB@JH>?!JKQFQ[!LM! @LQB!ZFDK>?!PF[UBQJD!FQ!DFJK3! :$"0,,;/%<9(")(*+,(=&*/"%&'(-0&;,.3(")( 20/,%0,9(")(*+,(>%/*,;(2*&*,9(")(-.,$/0& !T65"6786"T"<6"T43!!;93!)>Q[!dW!RDBQ`L!RaW!Y!-PHF>Q!''!5"6697!*OBQJFMF@>JFLQ!LM!JNB!0?5 <7<^FQOFQ[! DFJB!FQ!JNB!NKU>Q!HBO!>QO![HBBQ!@L?LH!ZFDFLQ!PF[UBQJD3! 6/"0+,./9*$3 !9;5678;";4<;"9#3!!;=3!GL@NBQOLBHMBH!SSW!)>Q[!dW!-PHF>Q!''W!Y!,>JNFBD!VR!5"66:7!VBDLQ>Q@B! V>U>Q!B_>UFQ>JFLQ! LM!JNB!C>ZB?BQ[JN!HB[K?>JFLQ!UB@N>QFDU!FQ!NKU>Q! ZFDK>?!PF[UBQJD3! 6/"0+,./9*$3 !9\5;;78\4::<\4T:3!!;43!V>`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`6=3(,&! 0"*+##%($4&'*6'1,#'7.1(*$./'8+.%#-2'*6' 9+(#$+#&'*6'1,#':$(1#%'91.1#&'*6'8-#"(+. !LLA"PCM%LQNO%LQQ&!!N$&!0.++,*6FJ*!Y@1!a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`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`I!EQ/+)3O)1X!4I!- !:ST23)Q!G!5$##N8! :WW)R2!PW!*PS/Q1]/21P3!P3!2O)!P*T13!TO1W2!13!QOPUP*T13T'!$'!:X*1Q1R/S! *PS/Q1]/21P3!XPU)ST!WPQ!*QP2)13T'! :"&,;#5*.-2,#6,0"+)(%-2,%"&'()$*+<,= !""$5%&8M""7LNA""7&N'!!7$'!4PQ11!9!5$##$8!a1[Q/21P3/S!*PS/Q1]/21P3!/3U!P*T13!TO1W2!PW!Q)213/S!TRO1WW! [/T)TM!2O)PQ)21R/S!T2+UV'! 4#5*.-2,#6,$"&,8'&*(%-.,9"&'(%-2,7#%(&$+ !"$75%"8M6$&$A6$&&'!!7%'!./012/31!9I!./012/31!4I!)X[Q/3!? >,&$,-2< !5$##K8!B2Q+R2+Q)I!T*)R2QPTRP*VI!/3U!T*)R2Q/S!2+313Z!PW! 2O)!Z/T A*O/T)!Q)213/S!ROQPXP*OPQ)M!2O)![)2/ A1P3P3)!cO/3US)c!/3U!/S0VS! ZQP+*!)WW)R2'! :"&,;#5*.-2,#6,0"+)(%-2,% "&'()$*+<,8 !"#65$68MLK6&ALL#K'!!7K'!,&$,-2< !5$#"#8!HQP[13Z!/3U!XPU)S13Z!2O)!/[TPQ*21P3!PW!Q)213/S! *QP2)13!ROQPXP*OPQ)T!13!\/R+P'! 8.3&?-./$&,9"&'(& !765"#8M"&6#A"&6%'!!7L'!@+Z2)3[+QZ!` >,&$,-2< !5"6NL8!G:>9?F=BG!E;4:<=;<9;D;HB=F'! 4#5*.-2,#6,$"&, 8'&*(%-.,9"&'(%-2,7#%(&$+ !"#N5""8M%"#7A%"#K'!!7&'!a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``'! 6&(70$!%J";J$%#=KPL"Ea"$J'!!L<'!b9854!^ 4)$()&25 !;D#"D=!C2O*08/!*02)G0)21!.@!091!Q!B5012H1F B801!.@!091!X1)<$ EEc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`>! +;>2)C*-2,!A+]9@9,8*=&! !"#>..'-.(?(3)@.('39(2)A")@.("3()$14'2B($)4414'2( '39(9)A)4#,&)3*'4(+"#4#6- !H%4"6K%J:I$&!! !"#$!Chapter IV: Water networks in the wavelength regulation of human Cellular Retinol Binding Protein II reengineered Protein IV.1 Important waters in the binding pocket of hCRBPII protein As our wavelength regulation studies showed little success with hCRABPII in the first generation series with the proteins absorbing maximally at 450 nm (1), one of the solutions was to use another protein from the Retinoid Binding Protein family. Since we knew that we can successfully make a Protonated Schiff Base (PS B) with hCRABPII, we decided to use a protein with the same fold and close homology to hCRABPII. human Cellular Retinol Binding Protein II (hCRBPII) was the target that closely met the requirements. hCRBPII binds, transfers and helps for the metabolism of retinol in the cellular environment (2, 3). It is a small cytosolic protein with a !-barrel fold and a smaller binding pocket than hCRABPII . The crystal structure of the wild type protein in complex with retinol shows the fold of this protein a nd the importan t residues inside the binding pocket of this protein (4) (Figure IV-1). F16 Q38 Q128 K40 Q108 2.7 † 3.0 † Figure IV-1) a) The crystal structure of wt hCRBPII bound to Retinol. b) Binding pocket of this protein with the important residues highlighted (PDB ID: 4QYN) . a) b) !"#$!Residue Gln 108 deep inside the binding pocket of the protein was chosen for introducing the nucleophilic lysine residue . Mutant Q108K upon incuba tion with Retinal does not show any PSB in the physiological pH, upon acidification to pHs lower than 6 there is a slight absorption at 506 nm, but the protein can not be titrated to show a full PSB formation. Dr.Wang noticed the presence of a lysine resid ue Lys40 in proximity to the introduced lysine residue Q108K. In order to remove the possible positive charge of Lys40 adjacent to the formed PSB, which might be the cause of the low pKa of Q108K mutant, this residue was mutated to a hydrophobic Leucine. Q108K:K40L (KL) showed significant PSB formation in 10 minutes in the physiological pH with a !max = 508 nm. The pKa titration of this protein gives a pKa of 8.3 for the PSB (5). Figure 2 shows the crystal structure of this protein with retinal covale ntly bound to the Q108K residue (6) (Figure IV-2). By extensive mutagenesis, spectroscopy and crystallization studies we were able to regulate the wavelength of this protein from 425 to 646 nm, more than Q38 Q108K T53 T51 Q4 Q128 W W W W Figure IV -2) Retinal bound covalently inside the binding pocket of the hCRBPII Q108K:K40L mutant (PDB ID: 4RUU) . !"#$!50 nm red shifted compared to the known rhodopsin s, and unveil the mechanism of wavelength regulation of this protein (6-8). Overlay of more than twenty crystal structures of hCRBPII shows conserved water molecules inside the binding pocket of this protein. These water molecules are the determinant factors that regulate the wavelength of the protein (9). These include 4 water molecules in the vicinity of the retinal chromophore that play a critical role in the wavelength regulation of this protein (Figure IV-3). The mutation of the interacting residues with these waters changes the wavelength of the protein at least by 20 nm. These water molecules are : (1) Gln4 water molecule W1 (2) Thr51 and Thr53 water W2 (3) Gln38 and Gln128 water molecules W3 and W4. Figure IV -3) Crystal structure of the hCRBPII Q108K:K40L mutant (PDB ID: 4RUU) . The important water molecules and the interacting residues with these waters are highlighted. 3.0 † 2.9 † 3.1 † 3.3 † 2.8 † 3.0 † 2.9 † 2.9 † Q128 Q38 Q108K T53 T51 Q4 2.8 † 2.7 † W1 W3 W2 W4 !"#$!IV.2 Gln4 water molecule Gln4 water molecule is in close proximity to the Schiff Base and is hydrogen bonded to the iminium . The Gln4 orientation is fixed in place by the main chain carbonyl of the neighbor residues to hold it in one fixe d conformation to apply a certain dipole moment to the water molecule , which will interact with the chromophore (6) (Figure IV-4). As expected the mutation of the Gln4 residue, especially to positively charged residues like Arginine or Lysine red shif ts the protein by 25 nm . This is expected as the removal of the localized charge on the iminium and distributing it on the poly ene will red shift the protein (6) (Table IV-1). Protein !max (nm) pKa KLVCWLW: Q4 591 8.2 KLVCWLW: Q4A 612 7.0 KLVCWLW: Q4L 614 7.9 KLVCWLW: Q4F 613 7.5 KLVCWLW: Q4W 613 7.7 KLVCWLW: Q4T 608 7.8 KLVCWLW: Q4K 618 7.2 KLVCWLW: Q4R 622 6.5 KLVCWLW: Q4H 585 7.9 KLVCWLW: Q4E 590 n.d. T1 I110 D91 Q108K Q4 3.1 † 2.9 † 2.9 † 2.7 † 3.1 † 2.8 † Figure IV -4) Gln4 is fixed in its conformation through a water molecule that is in interaction with the main chain carbonyl of Asp91 and Thr1. (PDB ID: 4RUU) Table IV -1. Gl n4 mutation summary . ¥KLVCWLW is the abbreviation for: Q108K:K40L:T51V:T53C:R58W:T29L:Y19W !"#$!The crystal structure of the KLVCWLW:Q4R mutant shows that the introduced Arginine residue faces away from the positive charge o f the iminium and the Gln4 water gets eliminated by this mutation (6) (Figure IV-5). IV.3 Thr51 and Thr53 water molecule The water molecule in between the residues Thr51 and Thr53 is close to the beginning of the polyene tail of the retinal. One of the interesting findings during the mutagenesis of these residues is that a T51V mutation causes 25nm red shift while a T53C mutation counts for just 5nm red shift. Dr.NossniÕs crystal structures of the mutants that have T51V mutation show that the water molecule is removed by introduction of this m utation, and Val 51 takes a new conformation to face the hydrophobic polyene. However, the crystal structure of the mutant KL:T53C sh ows that the water molecule is still present. Therefore, the removal of the polarized water is important in changin g the wavelength of the protein (9) (Figure IV-6). R58W Q108K T1 Y19 W K40L Q4 Q4R Figure IV -5) The overlaid structure o f KL:T51V:T53C:R58W:Y19W:T29L (C yan, PDB ID: 4EFG, 591 nm) and of KL :T51V:T53C:R58W:Y19W:T29L:Q4R (G reen, PDB ID: 4EEJ, 622 nm) . !!""#! Mutagenesis efforts prove s the importance of this water molecule in red shifting the protein by 25 nm (Table 2 ) (5). IV.4 Gln 38 and Gln 128 water molecules There are two water molecules in interaction with Gln 38 and Gln 128, which are located close to the retinal , toward the end of the polyene. These two glutamine residues interact with the retinal chromophore through the two water molecules ( Figure IV -2). These two waters are highly polarized as close nearby hydrogen bonding interactions force Gln38 and Gl n128 to face the waters through their amide carbonyl groups ( Figure IV-7). Protein !max (nm) pKa Q108K:K40L 508 8.3 Q108K:K40L: T51V 533 8.3 Q108K:K40L:R58Y 535 9.5 Q108K:K40L:R58Y: T51V 563 10.0 Q108K:K40L:T53C 513 7.5 Q108K:K40L:T53C: T51V 539 8.4 T53 T51 V T51 Q108K T53C 3.2 † 2.6 † Figure IV -6) a) Crystal structure of the mutant Q108K:K40L:T51V shows the loss of W2 which counts for the 2 5 nm red shift of the protein. b ) Crystal struct ure of the mutant Q108K:K40L:T53 C shows the presence of the W2, which counts for the wavelength of the protein not to change. Table IV-2. T51V mutation causes 25 -30 nm red shift. W3 W4 !"""! The main chain carbonyl groups of residues Ala 33 and Leu 36 act as hydrogen bonding acceptor for a neighbor water molecule that directly interacts with Gln 38 and indirectly through Asn 13 to Gln 128. This will force the two glutamine residues to face the wa ter molecule toward the polyene through amide carbonyl group and eventually the water (W3) will face the retinal through lone pairs to imply a negative dipole moment toward the retinal polyene. The orientation of the W4 can not be determined exactly, because the amidic ÐNH2 group can act as a hydrogen bonding donor or acceptor. Gln 38 Gln 128 Asn 13 Leu 36 Val 34 R111K 2.7 † 3.2 † 2.8 † 3.0 † 2.9 † 3.1 † 2.9 † 3.3 † 2.5 † 3.1 † 2.7 † 3.0 † Ala 33 NHOOOHHNOHONHHNOHHHOHHQ108KGln38Gln128Asn13Leu36Ala33a) b) Figure IV-7) a) The Orientation of the Gln38 and Gln128 is defined by adjacent residues that fix the amide side chain of these residues in place. Crystal structure of Q108K:K40L. (PDB ID: 4RUU) b) Chemic al representation of the orientation of these two rsidues. !""#!Mutation of these residues has proved to wipe out the whole wavelength regulation of the protein by blue shifting it to the start point mutant of Q108K:K40L, which absorbs at 508 nm. A single Q38L mutation blue shifts the wavelength of the protein by up to 80 nm. The mutation of both of these residues to more hydrophobic residues produces proteins that counteracts the effects o f all of the other mutations that red shifts the protein to above 600 nm back to 500 nm (Table IV-3 and IV-4). Gln 38 muta tion is more effective than Gln 128 mutation in blue shifting the mutants , but it does not drop the pKa of the pr otein more than 1.5 units . On the other hand, Gln128 mutation can drop the pKa of the protein by as much as 4 units , but is not as effective in blue shi fting the protein as Gln 38 is (Table IV-4). Protein !!max (nm) !pKa!KL:R58Y 533 9.0 KL:R58Y :Q38A!513 8.2 KL:R58Y :T51V!556 10.0 KL:R58Y :T51V: Q38A:Q128A!510 8.2 KL:R58Y :T51V: Q38L:Q128L!503 7.9 KL:R58Y :T51V:Y19W !565 10.2 KL:R58Y :T51V:Y19W: Q38L 522 8.7 KL:R58Y :T51V:Y19W: Q38A:Q128A!510 8.2 Table IV-3. Gln38 and Gln128 mutation on the shorter hCRBPII mutatnts. ¥KL is the abbreviation for Q108K:K40L. !""#! Unfortunately, the drop in the pKa that accompanies the Gln38 and Gln128 mutations makes it hard to get good quality diffracting crystals from these mutants. The highest pKa mutant of hCRBPII series Q108K:K40L:T51V:R58Y:Y19W with a pKa of 10.2, af ter mutating both Gln38 and Gln 128 produces blue shifted p roteins that absorb 500 -510 nm (Table IV-3). Although the pKa of these mutants is as high as 8.3, but still the crystallization trials provide crystals that do not diffract anywhere better than 3.5 †. However, the crystallization of the Q108K:K40L:T51V:R58Y:Y19W:Q38L mutant provided good quality diffracting crystals in the same crystal packing form of the parent mutant to enable a comparison between the two crystal structures. Interestingly, the comparis on of the two crystal structure s shows that in the crystal structure of Q108K:K40L:T51V:R58Y:Y19W in the two chains of the protein in the asymmetric unit the water mole cules are present both in one and the other chain shows just one water, while in the cry stal structure of Protein !!max (nm) !pKa!KLW:T51V:T53C:T29L:Y19W !591 8.2 KLW:T51V:T53C:T29L:Y19W :Q128L!531 4.6 KLW:T51V:T53S :T29L:Y19W !600 9.0 KLW:T51V:T53S :T29L:Y19W :Q38L!520 8.1 KLW:T51V:T53S :T29L:Y19W :Q38N!571 8.3 KL:T51V:T53S :T29L:Y19W :Q38L:Q128L!502 - KLW:T51V:T53C:T29L:Y19W :Q4R:A33W 644 7.2 KLW:T51V:T53C:T29L:Y19W :Q4R:A33W: Q38L 510 - KLW:T51V:T53C:T29L:Y19W :Q4F:A33W 646 8.7 KLW:T51V:T53C:T29L:Y19W :Q4F:A33W: Q128L 608 KLW:T51V:T53C:T29L:Y19W :Q4F:A33W: Q38N 610 4.7 KLW:T51V:T53C:T29L:Y19W :Q4F:A33W: Q128N 565 - KLW:T51V:T53C:T29L:Y19W :Q4F:A33W: Q38L: Q128L!538 - Table IV -4. The effect of the Gln38 and Gln 128 mutations on hCRBPII longer series proteins. ¥KLW is the abbreviation for Q108K:K40L:R58W. !""#!Q108K:K40L:T51V:R58Y:Y19W:Q38L mutant one of the chains shows the density for one water molecule and the other chain does not show any wate r molecules between the two glutamine residues proving the partial occupancy of water after the Q38L mutation. Additionally, the introduced hydrophobic leucine residue faces away from the water molecule in the chai n that shows the density for it , proving the decreased dipole moment of the water compared to when that bo th glutamines are present . More importantly the water molecule in the Q38L structure is able to rotate in its place and is not fixed. Therefore , there is no specific dipole moment direction from the water to the retinal polyene (Figure IV-8). 3.0 † 3.4 † Gln 128 Q38L Gln 128 Q38L 2.7 † 3.4 † 2.6 † Gln 128 Gln38 2.9 † 2.9 † Gln 128 Gln38 3.6 † 2.8 † 3.3 † 3.6 † a) b) Figure IV -8) a) Gln38 -Gln128 Water molecules in the two chains of crystal structure of KL:T51V:R58Y:Y19W. b) Gln38 -Gln128 Water molecules in the two chains of crystal structure of KL:T51V:R58Y:Y19W:Q38L. !""#!IV.5 Arg 58 and Ala33 mutations In hCRBPII b esides the residues that directly interact with waters and their mutations affect the wavelength of the protein; two residues on the entrance of the binding pocket play an important ro le in the wavelength regulation of this protein as well . These include Arg5 8 and Ala33 . M utation of each one of these residues to aromatic hydrophobic residues can increase the absorbe d wavelength of the protein by 25 nm (5, 6) (Table IV-5). Interestingly , the addition of R58W mutation can increase the effect of the internal mutations on the wavelength of the protein by a factor of 2 or 1.5 respectively (6) (Table IV-6). Protein !max (nm) pKa Q108K:K40L #$%!%&'!Q108K:K40L: R58W #"(!%&)!Q108K:K40L :R58F #*'!%&)!Q108K:K40L :R58 Y #'#!(&#!Q108K:K40L:T51V #''!%&'!Q108K:K40L:T51V: R58W #+#!%&,!Q108K:K40L:T51V: R58F #+"!%&)!Q108K:K40L:T51V: R58 Y #+'!"$!Q108K:K40L :T51V:T53C #'(!%&,!Q108K:K40L :T51V:T53C: R58 W #%#!)&,!Q108K:K40L :T51V:T53C:T29L:Y19W #''!)&(!Q108K:K40L :T51V:T53C:T29L:Y19W: R58W #("!%&*!Q108K:K40L :T51V:T53C: R58W :T29L:Y19W:Q4R +**!+&)!Q108K:K40L :T51V:T53C: R58W :T29L:Y19W:Q4R: A33W +,,!+&#!Table IV -5. The effect of R58W and A33W on red shifting the hCRBPII mutants. protein. !""#! The mutation of R58F also has similar effects, and increases the effect of the intern al residues by a factor of 1.5 (5) (Table IV -7). A surface representation of the crystal structures of the mutants which have the R58W mutation with the ones that do not, show s that there is not that much difference in how much the binding pocket is covered or in case of the Arg58 mutants the binding pocket is more covered compared to when this residue is mutated to a Tryptophan . Besides the fact that Tryptophan is more hydrophobic than an Arginine residue , Phe57 also flips out of the binding Entry hCRBPII mutant !max (nm) R58 Protein Shift (nm) Protein Shift (CM-1) !max (nm) R58 W Protein Shift (nm) Protein Shift (CM-1) Enhancement (nm) 1 KL 508 0 0 527 0 0 - 2 KL:T51V 533 25 923 570 43 1431 18 (1.7 !) 3 KL:T53C 513 5 192 540 13 457 8 (2.6 !) 4 KL:T51V:T53C 539 31 1132 585 58 1881 27 (1.9 !) 5 KL:Y19W 513 5 192 538 11 388 6 (2.2 !) 6 KL:T51V:Y19W 537 29 1063 577 50 1644 21 (1.7 !) 7 KL:T51V:T53C: Y19W 538 30 1098 590 63 2026 33 (2.1 !) Entry hCRBPII mutant !max (nm) R58 Protein Shift (nm) Protein Shift ( CM-1) !max (nm) R58F Protein Shift (nm) Protein Shift (CM-1) Enhancement (nm) 1 KL 508 0 0 523 0 0 - 2 KL:T51V 533 25 923 561 38 1296 13 (1.5 !) 3 KL:T53C 513 5 192 537 14 499 9 (2.8 !) 4 KL:Y19W 513 5 192 537 14 499 9 (2.8 !) 5 KL:T51V:T53C 539 31 1132 571 48 1608 17 (1.5 !) Table IV -7. R58F mutation enhances the effect of the internal mutations by a factor of 1.5 -2. Table IV -6. Protein shift with reference to Q108K :K40L, R58W enhancement effect. ¥wavenumbers provide a direct correlation to the change in energy. €Protein shift with reference to Q108K:K40L:R58W. Enhancement is calculated as the difference in protein shift between KL -R58W mutants and the KL mutants, and reflects the overall increased red shift in excess of that anticipated from a purely additive effect of R58W. For example, the T51V mutation leads to a 25 -nm bathochromic shift (KL versus KL:T51V). A 25 -nm red shift would be expected for KL:T51V:R58W versus KL:R58; however, a 43 -nm shif t is observed. The 18 -nm difference in the level of enhancement (factor of 1.7 increase) is a result of the R58W mutation. Numbers in parentheses are relative increases of the protein shift of the KL -R58W mutant series with respect to the KL mutant series. !""#!pocket by the introduction of R58W mutation or R58F or R58Y. This causes the formation of a hydrophobic barrier that is made by R58 WFY and Phe57. This hydrophobic barrier does not allow the water molecules to path e asily into the binding pocket; T herefor e, the effects of the high dielectric constant of the solvent , which wipes out the effect of the mutations , are furthe rmore shiel ded by this mutation (9) (Figure IV-9). The addition of A33W mutation to the mutants that have t he R58W mutation also adds 15 -20 nm red shift t o the absorption of the protein . The overlay structure of the two proteins without and with the A33W mutation shows the effect of this residue in effec tively covering the entrance of the binding cavity (6) (Figure IV-10). Figure IV -9) a) Surface representation of the Q108K:K40L:T51V:T53C mutant, !max = 539 nm (left) b) Surface representation of Q108K:K40L:T51V:T53C:R58W:T29L:Y19W (Right) mutant, !max = 591 nm. c) R58W mutation forces the Phe57 to flip out of the binding pocket. a) b) c) !""#! IV.6 Counter anion in the hCRBPII system One of the unique characteristics of hCRBPII retinal pigments is the pKa stabilization without a counteranion. For all of the mutants of this protein that have an absorption above 508 nm the retinal PSB has been achieved via the introduction of no counter anion. As mentioned in previous chapters i n rhodopsin , residue E113 acts as a direct counter anion for the PSB rendering it with a pKa of 16 and in bacteriorhodopsin multipl e Aspartic acid residues, which include the D85, 96 and 212 are involved in indirect hydrogen bonding to the PSB that increases the pKa of the PSB to 13 (10). Figure IV -10) a) Surface representation of the Q108K:K40L:T51V:T53C :R58W :T29L:Y19W mutant, !max = 591 nm (PDB ID: 4EFG). b) Q108K:K40L:T51V:T53C:R58W:T29L:Y19W :A33W mutant, !max = 606 nm (PDB ID: 4EDE) . c) A33W mutation helps to cover the entrance of the binding pocket more effectively. a) b) c) !""#!The water networks in hCRBPII not only govern its wavelength regulation but also the pKa of this system as well. Specially , the Q38 and Q128 residues play the most sign ificant effect in both the pKa and wavelength absorbed by the protein. As mentioned above Q 4 residue contributes to 1 -1.5 pKa unit, Gln38 1 -1.5 unit, Gln128 up to 4 units and R58Y mutation can increase the pKa by 1 unit. All of these mutations have a direct effect on the water binding pattern of the binding pocket (Tables IV-1 to IV-5) (5, 6, 9). In hCRBPII system pKas of up to 10.2 have been achieved by introducing no counter anion for the Schiff base , which is pretty novel (5) (Table s IV-3 and IV-5). However, achieving retinal pigments with absorption below 508 nm was not possible without the introduction of counter anion . The most blue shifted proteins in these series are the Q108K:T51D mutant absorbing at 474 nm and Q108K:T51D:L117E mutant absorbing at 425 nm. A crystal structure of the Q108K:K40D protein showed the introduced Aspartic acid residue in hydrogen bonding distance to the PSB. This residue can act as a counter anion or it probably accumulates waters n ear the PSB that causes the blue shift of the protein. Residue K40 stays close to the C12 of the retinal in some chains . In other chains it points right at the C14 carbon of retinal. It can be concluded that there might be some of the K40 retinal PSB in the structure as well with partial occupancy. If proven to be true with more mutagenesis and structural studies, study ing the K40 PSBs can unveil novel findings on wavelength regulation and retinal binding in this system (Figure IV-11). !"#$! KD structure has a unique fold different than all of the other retinal bound structures of hCRBPII. Unlike , the other retin al bound structures that show the hCRBPII fold as a monomer, KD folds as a dimer. Two protein molecules each make an open monomer tha t fold together. An open monomer is the result of the severe torsion of the Phi and Psi angles of residues 50 -60 which confers a different tertiary structure to the protein (9). IV.7 Crystal growth for Neutron Diffraction In order to understand the complete orientation of the water molecules in the binding pocket of the hCRBPII protein and their effect in the wavelength regulation of this protein, we aimed for neutron diffraction studies of this protein. Since the contribution of the hydrogens to the overall X-ray diffraction pattern is small, it is not possible to find the absolute orientation of the hydrogens from an X -ray electron diffraction pattern. However, unlike X -ray 2.6 † 2.4 † 3.7 † 3.5 † Thr53 T51D Q108K Q108K 3.5 † 3.0 † 2.8 † 2.3 † Thr53 T51D a) b) Figure IV -11) a) The retinal density in chain G of the Q108K:T51D crystal structure, Lys 40 points at C12 of retinal. b) The retinal density in chain L of the Q108K:T51D crystal structure, Lys 40 points at C14 of retinal. Map is contoured at ! = 1 .0. !"#"!crystallography, hydrogen Õs contribution to the overall neutron diffraction map is co nsiderable and the position of hydrogens can be determined by this method. The bottleneck of the Neutron diffraction is the requirement for the growth of big crystals (11, 12). hCRBPII grows as 0.2*0.1*0.2 mm crystals or smaller in hanging drops , which are quite small for Neutron Diffraction. For Neutron diffraction the volume of the crystals should be at least 0.15 mm 3. Besides this hCRBPII has a tendency of growing multiple lattices into each other during crystallization. Multiple lattices growing into each other become more problematic , when the formation of bigger crystals is required. Three mutants were chosen for Neutron diffraction crystal growth studies based on their regular crystallization in H 2O hanging drops and X-ray diffraction quality. These include Q108K:K40L, Q108K:K40L:T51V:R58F and Q108K:K40L:T51V:R58Y:Y19W . Mutants with Q4X mutation were not chosen. In spite of better diffraction and better retinal density, they tend to produce less crystals that are clusters. Different crystallization tech niques including hanging drop, s itting dro p, crystallization under oil, micro seeding and macro seeding into hanging drop and sitting drop were tried. IV.7 -1 Q108K:K40L mutant of hCRBPII In order to make sure that the binding of retinal to hCRBPII proteins in D 2O is not significantly different than H 2O, retinal binding assay in D 2O was done for this mutant, and as it is shown the kinetics of the binding is considerably slower, but there is no change in the protein absor ption. In D 2O retinal binds slower and the PSB loss is also slower ( Figure IV-12). !"##!The first crystallization trials were done with 35 microliter polystyrene sitting drop micro -Bridges. 24 well VDX Hampton boxes were used and all of the solutions and buffers were prepared in D 2O. The protein solution was buffer exchanged with 10mM Tris.HCl D 2O pH=8.0 and 150 mM NaCl prio r to retinal incubation and then concentrated to 6 -8 mg/ml. The crystals were grown in 1 -2 weeks . Unfortunately , the crystals grown in these drops were even smaller than the regular crystals in small hanging drops (Figure IV-13a). So, I decided to go back to the hanging drops and repeat the same crystallization strategies but this time with 2 -3 microliter 00.050.10.150.20.250.30.350.4300400500600700800KL - PSB loss in D2O 10min 20min 1h2h4h8h15hAbsorption Wavelength (nm) Figure IV -12) PSB loss of the mutant Q108K:K40L of hCRBPII in a) H 2O b) D2O c) The kinetic of the PSB loss comparison in H 2O and D 2O. 0.050.10.150.20.250100200300400500KL - PSB loss in H2O vs D2O H2OD2OAbsorption Wavelength (nm) 00.050.10.150.20.250.3300400500600700800KL - PSB loss in H2O Retinal-10min 20min 1h2h4h8h12hAbsorption Wavelength (nm) !"#$!hanging drop s. The crystals grown this time were bigger than the sitting drop ones, but still not big enough for the studies (Figure IV-13b to 13 d). The next plan was to increase the protein concentration (12 mg/ml) , but screening int o lower precipitation reagent ( PEG 4000). This strategy proved to be effective in increasing the size of the crystals two -three times bigger than the ones with lower prote in concentration and higher PEG 4000 (Figure IV-13e). Incre asing the protein concentration to 20 -25 mg/ml and screening into lower PEG with bigger drops provided a l ot of crystals with the same size in 24 hours, but we were doubtful, if they are big enough for the Neutron diffraction studies (Figure IV-13f). Since the original sitting drop trials were not successful , micro seeding and macro seeding experiments were done to see if we can get eve n bigger crystals with these methods. Unfortunately, this mutant provided just small size crystals with the micro seeding and a lot of nucleation around the original crystal in the macro seeding (Figure IV-13g). Crys tallization of Q108K:K40L mutan t under AlÕs oil (Micro batch ) in D 2O does not provide any better crystals compared to hanging drop or sitting drop. At this point the only way to improve the size of the crystals seemed to be the bigger drops, which can provide more protein for a larger amount of protein deposition onto the crystals. Since the original sitting drop trials were just at 6 -8 mg/ml, this time 20 -25 mg/ml protein in 10,20 and 40 microliter sitting drops with polystyrene micr o bridges were tried and it proved to be successful in making big size crystals. Also, it should be mentioned that a lot of these crystals still had multiple lattices grown into each other and not bigger than the hanging drop crystals . The next try was the crystallization of the protein !"#$!with larger sitting drops. For this purpose crystallization of the protein at 25 -30 mg/ml with 80 microliter sitting drops on glass rods were tried. The crystals grown this way were the biggest crystals grown. They we re as big as 1*1*0.5 mm in the size (Figure IV-13h and IV-13i). Some of these crystals were shot at Oak Ridge National Laboratory (ORNL). Although, the diffraction pattern of these crystals shows more than one lattice , because of the time limitations smaller crystals , which tend to be more single, were not tried (Figu re IV-13j and IV-13k). Further data collection on smalle r but single crystals can provide better diffraction data. Another experiment that was not tried for this mutant was the sitting drops on glass rods in H 2O rather than D 2O. !"#$! a) c) g) j) b) h) e) f) d) i) k) Figure IV -13) Crystallization trials of Q108K:K40L in D 2O. a) crystals from sitting drop at 6 mg/ml and 35 µl size. b,c,d) Crystals from hanging drop at 6 -8 mg/ml. e) Crystals from hanging drop at 12 mg/ml by screening into lower PEG4000. f) Crystals from hanging drop at 20 -25 mg/ml by screening into lower PEG4000. g ) Macro seeding of the crystal in panel e. h and i) crystals from sitting drop at 25 -30 mg/ml and 80 µl size. j and k) Neutron diffraction pattern of the crystal in panel h. !"#$!IV.7 -2 Q108K:K40L:T51V:R58F mutant of hCRBPII All t he experimental procedures tried with Q108K:K40L mutant were tried with Q108K:K40L:T51V:R58F mutant as well. This mutant produces huge clusters with 40 and 80 microliter sitting drops in D 2O. Crystallization of this mutant under AlÕs oil also does not seem to produce any better crystals compared to the hanging drop crystallization. 10-15 microliter hanging drops wit h 20-25 mg/ml protein concentration with suitable PEG4000 concentration can provide single crystals in D 2O (Figure IV-14a). Micro seeding experiments in hanging drops also provided very small crystals (Figure IV-14b). Also, the crystallization of this mutant with sitting drops in H2O might improve the quality of these crystals , but it was not tried . IV.7 -3 Q108K:K40L:T51V:R58Y:Y19W mutant of hCRBPII Crystallization of this protein in D 2O with 40 and 80 microliter sitting drops at 20-25 mg/ml in polystyrene micro bridges and glass rods mainly is accompanied with a lot of protein precipitation . However, in H 2O with glass a) b) Figure IV -14) Crystallization of Q108K:K40L:T51V:R58F mutant of hCRBPII in D 2O: a) with 20 -25 mg/ml protein in 10 µl hanging drops. b) microseeded drops provides smaller crystals. !"#$!rods and 80 microliter drops suitable size single crystals are formed. The Neutron diffraction of these crystals was not tried (Figure IV-15). IV.7 -4 hCRBPII Expression in Minimal M edia In order to deuterate the back bone of the protein for improving the signal to noise in neutron diffraction , the expression of the protein in a deut erated carbon source is required. The first step is to check the expression level in the same carbon source with hydrogen and in H 2O rather than D 2O. The expression protocol described in (11) was used. Colonies were grown in regular plates and transferred to minimal media plates for 36 -48 h ours for cell adaptation before transferring to liquid minimal media. Adaptation of the cells to minimal media on plates is important for protein expression. Cells were grown in 50 ml minimal media before 1 liter expression, and the OD 600 was 2 -3 before induction with 1mM IPTG for 24 hour s at room temperature. The protein was then purified and analyzed by SDS -PAGE gel ( Figure IV-16). Figure IV-15) Crystal of Q108K:K40L:T51V:R58Y:Y19W mutant of hCRBPII in H 2O. !"#$! IV. 8 Wavelength Regulation in second g eneration hCRABPII mutants IV.8 -1 R59W and A32W mutations Most of the wavelength regulation in hCRABPII comes fro m mutating residues toward the mouth of the binding pocket, which as described in hCRBPII makes a hydrophobic barrier for the passage of the waters into the binding pocket. Position s Arg 59 and Ala 32 are the ones that contribute significantly to red shift ing the wavelength of this protein . An R59W mutation to enclose the binding cavity i s accompanied by 22 nm red shift. The crystal structure of the R111K:R132L:Y134F:T54V:R59W mutant of hCRABPII showed that the entrance of the binding pocket of the protein i s still wide open. To further enclose the binding cavity of the protein A32W mutation was added. Interestingly, this mutation red shifts the protein by 54 nm and the protein absorbs at 610 nm (Table IV-8). 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 9 16a) 16b) Figure IV -16) a) Expression of hCRBPII in regular and H 2O minimal media for mutant: Q108K:K40L:T51V:T53C:R58W:T29L:Y19W 1) MW marker 2,3,4) regular media, elutions from FastQ each 30ml 5,6,7) H 2O minimal media, e lutions from FastQ, each 30ml 15 b) Expression of Q108K:K40L in H 2O minimal media without cell adaptation 1) MW marker 2) Fast Q elution 3 -9) Source Q elutions. Fraction 4 provides crystals. !"#$! The crystal structure of the R111K:R132L:Y134F:T54V:R59W:A32W mutant shows that the entrance of the binding pocket is still open and not efficiently closed (Figure 17 ). Therefore, the importance of these mutations is not just in closing the entrance of the binding pocket. We hypothesized that the introduction of these mutat ions makes a hydrophobic barrier that inhibits the movement of water molecules into and out of the binding pocket of the protein, which is essential for wavelength tuning. In fact, these findings in hCRABPII has been in parallel helping us to mutate hCRBPI I rationally as well as described. These residues do not seem to be directly interacting with water molecules, but they decrease the global access of the waters fro m the solution into the protein (Figure IV-17). Protein !max (nm) R111K:R132L:Y134F:T54V %&'!R111K:R132L:Y134F:T54V: R59W %%(!R111K:R132L:Y134F:T54V: R59W :A32W (")!Table IV -8) R59W and A32W mutations contribute significantly to the wavelength regulation of hCRABPII . !"#$! IV.8 -2 Y134F and T54 V mutations Originally the formation of PSB in second generation hCRABPII was achieved by four mutations R111K:R132L:Y134F:T54V. I tried to find the role of these mutations in lieu of P39Q and R132Q mutation s described in chapter2 , which improves the pKa of the protein, more stable PSBs and better diffracting crystals. The tetra mutant of R111K:R132 Q:P39Q:R59Y provides complete PSB formation at pH=5 .0 after 23 hours of retinal incubation with !max = 516 nm. The addition of Y134F mutation to this mutant does not change the absorption of the protein, but accelerates the PSB formation and eventually R59W A32 W Figure IV -17) a) Crystal structure of R111K:R132L:Y134F:T54V: R59W at 2.6 † (Green, PDB ID: 4I9S ) !max = 556 nm and crystal structure of R111K:R132L:Y134F:T54V: R59W :A32W at 2.6 † (Cyan , PDB ID: 4I9R ) !max = 610 nm of hCRABPII overlaid. b) Surface representation of these mutants: Left: R111K:R132L:Y134F:T54V: R59W (magenta) and Right: R111K:R132L:Y134F:T54V: R59W :A32W (cyan) . a) b) !"#"!the mutant R111K:Y134F:T54V:R132Q:P39Q:R59Y forms the PSB in 5 hours, and interestingly the addition of T54V mutation, which is the equal to Thr 51 in hCRBPII, red shifts the protein by 25 nm to absorb at 546 nm. The addition of A32W mutation also red shifts the protein by 34 nm ( Figure IV-18). 00.10.20.30.40.50.60.70.8300400500600700800PSB formation at pH 5 - Absorption maxima at 516 nm 1h2h4h8h12h16h20h23hAbsorption Wavelength (nm) 00.050.10.150.20.250.30.350.4300400500600700800PSB formation at pH 5 - Absorption maxima at 546 nm 0h1h2h3h4hAbsorption Wavelength (nm) 00.050.10.150.20.250.30.350.4300400500600700800PSB formation at pH 5 - Absorption maxima at 580 nm 20min 40min 1hAbsorption Wavelength (nm) Figure IV -18) Retinal incubation of hCRABPII mutants in citr ate buffer at pH5 for mutants: a) R111K:R132Q:P39Q:R59Y b) R111K:Y134F:R132Q:P39Q:R59Y c) R111K:Y134F:T54V: R132Q:P39Q:R59Y d) R111K:Y134F:T54V:R132Q:P39Q:R59Y:A32W . 00.10.20.30.40.50.60.70.8300400500600700800PSB formation at pH 5 - Absorbtion maxima at 519 nm 20min 1h2h3h4h5h6h7h8h9hAbsorption Wavelength (nm) c) b) a) d) !"#$!It should be mentioned that the Y134F and T54V mutations favor the formation of the PSB at the physiological pH ( Figure IV-19). This might be because of the overall decrease in the polarity of the binding pocket. Y134F mutation in the context of R132Q:P39Y mutation has an important effect in red shifting the protein by 55 nm , but no effect in the context of R132Q:P39Q (Figure IV-18 and IV-19). 00.10.20.30.40.50.60.70.8300400500600700800Retinal Incubation at pH 7 -Absorption maxima at 571 nm 20min 1h2h3h4h5h6h7h8h9h10hAbsorption Wavelength (nm) 00.10.20.30.40.50.60.7300400500600700800Retinal Incubation at pH 7 - Absorption maxima at 500 nm 2h8hAbsorption Wavelength (nm) 00.20.40.60.811.2300400500600700800PSB formation at pH 7 - Absorbtion maxima at 555 nm 1h2h3h4h5h6h8hAbsorption Wavelength (nm) Figure IV -19) The retinal incubation of hCRABPII mutants at physiological pH: a) R111K:R132Q:P39Y:R59Y b) R111K:Y134F: R132Q:P39Y:R59Y c) R111K:Y134F: T54V:R132Q:P39Y:R59Y . !"##!Also the acid titration of this protein shows different PSB absorbing in a range from 510 nm to 555 nm. A satisfactory answer to the effects of the Y134F cannot be achieved without further experiments . Comparing Figure 18 and 19 , it is apparent that a P39Y mutation can red shift the protein by 25 -30 nm compared to the mutants that have the P39 Q mutation in the same position . IV.8 -3 P39Q, R132Q and F3Q mutations The mutation of Gln 38 and Gln 128 in hCRBPII causes big blue shifts as the negative dipole moment toward the end of the polyene is removed. Since the equivalent positions in hCRABPII are Pro39 and Arg132, the mutations of these residues to Gln supposedly should increase the negative dipole moment toward the end of the polyene by the introduction of these polar residues and red shift the absorption , but counter intuitively these mutations do not cause any wavelength shift or usually blue shift the protein by 10 -15 nm (Table IV-9). We have learned in our hCRBPII system that equal distribution of the charge is the key to get super red shifted for the mutants of this protein, the blue shift by the introduction of the P39Q and R132Q mutations in hCRABPII might be Protein !max (nm) R111K:R132L:Y134F:T54V: R59W $$%!R111K:Y134F:T54V: R132Q:P39Q :R59W $$$!R111K:R132L:Y134F:T54V:R59Y $%"!R111K:Y134F:T54V: R132Q:P39Q :R59Y $$%!R111K:R132L:Y134F:T54V:R59Y: A32W $&"!R111K:Y134F:T54V: R132Q:P39Q :R59Y: A32W $'(!R111K:R132L:Y134F:T54V:R59W: A32W %")!R111K:R132L:Y134F:T54V: R132Q:P39Q :R59W: A32W $&'!Table IV -9. The effect of the R132Q:P39Q mutation on the wavelength absorbed by the hCRABPII mutants. !"#$!because of the fact that compared to the parent mutant the P39Q and R132Q mutations localize the charge and therefore get slightly blue shifted. Additionally , the crystal structure s of the mutants with the P39Q and R132Q muta tions shows that unlike hCRBPII, the two glutamines are separated via the water molecules in between them (Figure IV-20d), here the glutamine residues are close to each other for hydrogen bonding to one another. The electrostatic effects of the two glutamines on retinal will more likely cancel out as they face each other from the opposite ends of their amidic groups (Figure IV-20). The addition of the F3Q mutation to these mutants blue shifts the protein by 25 nm, which is expected as the polar glutamine localizes the charge around the retinal iminium PSB (Table IV-10). R111K P39Q R132Q 3.0 † 3.5 † R111K P39Q R132Q 3.3 † 3.2 † 2.6 † 3.0 † R111K Gln 128 Gln 38 4.0 † 4.4 † 2.7 † 2.8 † 3.3 † 3.0 † 2.8 † 2.7 † 3.3 † 3.0 † 3.2 † Leu121 P39Q R111K R132Q Cys 130 Ala 36 3.1 † d) c) b) a) Figure IV -20) a) The orientati on of the P39Q and R132Q in the R111K:Y134F:T54V:R132Q: P39Q:R59Y mutant of hCRABPII (PDB ID: 4YBP) b,c) The orientation of the P39Q and R132Q in the R111K:Y134F: T54V:R132Q:P39Q:R59Y:A32W:F3Q mutant of hCRABPII in th e two chains of the protein (PDB ID:4YKM). d) The orientation of the Gln 38 and Gln 128 in the Q108K:K40L mutant of hCRBPII (PDB ID: 4RUU) . !"#$! IV.9 Conclusion Water molecules play a crucial role in the wavelength regulation of reengineered hCRBPII. Mutating the interacting residues with these waters around the Schiff base or at the beginning of the polyene combined with mutations at the entrance of the binding pocket to decrease the global access of the waters to the binding pocket can red shift the protein from 508 to 644 nm. Mutating the interacting residu es with the water molecules toward the end of the polyene can wipe out the whole red shift of the protein from 644 to 510 nm. Neutron diffraction studies are in progress to find the absolute orientation of these waters inside the binding pocket of the prot ein. We have shown that the wavelength regulation in hCRABPII follows the same principles . Most of the wavelength regulation of this protein comes from the mutation of the residues at the entrance of the binding pocket of the protein. While mutating the po lar residues or introducing them toward the Schiff base or at the beginning of the polyene like Y134F and T54V and F3Q changes the wavelength absorbed by the protein, introducing glutamine mutations toward Protein !max (nm) R111K:Y134F:T54V: R132Q:P39Q :R59Y: A32 Y $%$!R111K:Y134F:T54V: R132Q:P39Q :R59Y: A32 Y:F3Q $&$!R111K:Y134F:T54V: R132Q:P39Q :R59Y: A32W $'&!R111K:Y134F:T54V: R132Q:P39Q :R59Y: A32W :F3Q $$%!R111K:R132L:Y134F:T54V: R132Q:P39Q :R59W: A32W $('!R111K:R132L:Y134F:T54V: R132Q:P39Q :R59W: A32W :F3Q $%"!Table IV -10. The effect of the F3Q mutation on the wavelength absorbed by the hCRABPII mutants. !"#$!the end of the retinal polyene in hCRABPII does no t change the wavelength of the protein by more than 10 nm. The crystal structures of hCRABPII shows that unlike hCRBPII, which glutamines are separated by waters and do not interact with each other, these residues directly hydrogen bond to one an other and probably cancel each otherÕs dipole moment to have a slight effect on the wavelength of the protein. IV.10 Other Structural Studies of hCRABPII and hCRBPII Reengineering During my PhD, I also studied the structural studies of hCRABPII and hCRBPII in collaboration with Dr.Yapici and Dr.Berbasova. These studies included : (1) The crystallization of hCRABPII proteins w ith Merocyanine in the development of an in -vivo fluorescent prote in. The crystal structures provided invaluable understanding of the Quantum Yield of these pigments that correlates to the mode of the binding of the fluo rop hore with the protein (Figure IV-21) (13). !R132K Merocyanine Leu121 Ser12 2.7 † 3.3 † a) b) Figure IV -21) a) Crystal structure of the R132K:K111L mutant of hCRABPII with Merocyanine (PDB ID: 4QGV) . b) Different modes of the binding of Merocyanine inside the binding pocket of hCRABPII (PDB IDs: Merocyanine: 4QGV, 4QGX and 3FEP, Retinal: 2G7B). !"#$! (2) The structural studies of the wavelength regulation of retinal analogs with hCRBPII. Th e retinal analogs were synthesized and assayed with proteins by Dr.Yapici and the structural data confirmed that these analogs bind in the predicted conformation . The focus was on the rotation around the ionone ring of the retinal and its effect on the wav elength regulation of hCRBPII protein. The 6s -trans retinal and 6s -cis retinal conformations were confirmed by high resolution crystal structures (Figure IV-22) (14). (3) The structural s tudies of the hCRBPII protein with Fluorophore ligand Julolidine for the development of an in -vivo pH sensor. The introduction of a glutamic acid residue in the vicinity of the Julolid ine PSB proved to make a 1,1-di demethyl r etinal K40L Q108K 5-demethyl r etinal K40L Q108K a) b) Figure IV -22) The crystal structure of the Q108K:K40L:T51V:R58Y:Y19W mutant of hCRBPII with: a) 1,1 -di demethyl retinal in 6s -trans retinal conformation and b) 5 -de methyl retinal in 6s -cis conformation. !"#$!titrat able residue that in the anionic and protonated form can have different absorptions. Different emission ratios based on the excitation of the protein at two wavelengths that it absorbs can be used to address pH. This can be the development of a single protein ratiometric fluor escent pH sensor (Figure IV-23) (15). Q4H Q108K Leu117 2.8 † 3.0 † a) b) c) Figure IV -23) a) Crystal structure of the Q108K:K40L:T51V:T53C:R58W:T29L: Y19W:Q4H mutant of hCRBPII with Julolidene b) pH dependent change of the absorption of the protein with the introduction of L117E mutation c) pH dependent fluorescence ratio from excitation at the two absorption peaks of the protein. !"#$! REFERENCES !"#$!REFERENCES "%!&'()*+),-!. !"#$"%&' !/0$$12!34,5+)6!7+()869!:++68)6++4)68!.+**-*'4!:+5)6,);! <;)7!=)67)68!34,5+)6!>>!)65,!'!4?,7,@()6!@4,5+)6!A)A);%! ()*+,%&")-"$.# "/0#+12%,"3.#012%&"4)21#$5 !"0B/"B29C"#$DC"#E%!!0%!.4,F!G>!/?.:=3>>2!O,-67!5,!4+5)6,*!'67!4+5)6'*%! /2$%" 2+59$%&&)8+%>.12%'"4#2$1),"?!"@1)&)812%&"2+59$%&&)8+%>.5 !1$/35!"029Q00C DQ0Q0%!!L%!U'68!U!/0$"02!34,5+)6!7+()869!:++68)6++4)68!,M!;+**-*'4!4 +5)6,*!O)67)68! @4,5+)6!>>!/.:=3>>2!)65,!'!4?,7,@()6!A)A);V!M-6;5),6'*)N'5),6!,M!.:=3>>! )65,!'!M*-,4+(;+65!@4,5+)6!5'8!'67!7+()86!,M!'!@?,5,(F)5;?'O*+!@4,5+)6!5'8%! 3?J!/W);?)8'6!R5'5+!X6)Y+4()5P2% !!C%!U'68!UG!"#$"%&' !/0$"02!Z-6)68!5?+!K*+;54,6);!0#,$%&"=1)&)85 !QC/"29CLD1#%!!B%!S,((,6)!T!/0$"#2!RZ:X.ZX:<_!>SR>\]Z!>S!ZI!Z]K!WK.]RW!I`! U<&K_KS\Z]!ZXS>S\!>S!S!W>W>.!S\_K! WXZIS!:KRX _ZKJ!&K!QJ!JIW<>S!RU<33KJ! J>WK:>TIS!>S!].:=3>>%!3?J!/W);?)8'6!R5'5+!X6)Y+4()5P2% !!"$%!K46(5!I3 !"#$"%&' !/0$"#2!W);4,O)'*!'67!<6)A'*!:?,7,@()6(9!R54-;5-4+(V! `-6;5),6(V!'67!W,*+;-*'4!W+;?'6)(A(%! 3.#012%&"+#D1#C9 !""#/"29"0CD"CQ%!!""%!W+)**+-4!`V!U +)((!b_V!H!WP*+(!J<8!?+>:'(8! 9:+*@:*+'!A8B!8'C!@A?AD()(:('9!=>+!8'*:+>8!;A@+>;>)' @*)A+! @+E9:A))>F+A?2E%! ()$%"*+,-$%&&./+"0 !GH<$"IJK$"G.%!!"/%!LA?(@(!3 !"#$"%&' !-$."I0!M1*+8K>8M!?+>:'(8!=)*>+'9@'8@'+;A:(>8!>=! @EA8(8'!BE'9%!1.2+3%&".4"$5#"(6#+7)%3"*5#67)%&"8.)7#$, !"/J-/0<".J/K".N.%!!"#%!LA?(@(!3!-$."I0!O'9(F8!>=!8>P')! ?+>:'(8Q@2+>;>?2>+'!@>;?)'R'98!>=!@2+>;>?2>+'9T!=)*>+>?2>+'9T!A8B!'8F(8''+(8F! >=!?+>:'(8!2>9:!9E9:';9%!U2O!-&(@2(FA8!S:A:'!V8(P'+9(:E0% !!"I%!W'+DA9>PA!1!-$."#0!78F(8''+(8F!>=!:2'!9;A))!@E:>9>)(@!+':(8>(B!D(8B(8F! ?+>:'(89!(8:>!?X K9'89(8F!?+>D'9!A8B!8>P')!=)*>+'9@'8:!?+>:'(8!:AF9%!U2%!O%! -&(@2(FA8!S:A:'!V8(P'+9(:ET!U+>Y*'9:T!V&3!O(99'+:A:(>89!U*D)(92(8F0% !! !"#$!Chapter V: Stabilizing the interaction between Cellular Retinoic Acid Binding Protein II and Retinoic Acid Receptor V.1 Introduction Retinoic Acid Receptors (RARs) belong to the superfamily of Nuclear Receptors. They are transcription factors that play a significant role in cell growth, differentiation , homeostasis and apoptosis (1-3). They are one of the targets for the suppression of different types of cancer (3-9). Structurally these proteins have multiple subunits in their sequence. The protein binds to the DNA through the DNA Binding Domain (DBD) of the protein and the Ligand Binding Domain (LBD) of the protein mediates the transcriptional activity (4, 10) (Figure V-1). The Ligand Binding d omain of this protein includes eleven !-helices, which forms an antiparallel sandwich. RARs as their name indicates accept all -trans -Retinoic Acid as the ir ligand (10) (Figure V-2). Figure V-1) Structural organization of Retinoic Acid Receptors. H2NCOOHAF1AF2LBDDBD!"#$! RARs supply of Retinoic Acid comes from Cellular Retinoic Acid Binding Proteins (CRABP). Noa Noy studied the transfer of the Retinoic acid from CRABPs to RARs and multiple publications elucidated how this process happens. The main focus of these studies were: (1) Elucidating the mechanism of the transfer (11-13). (2) Identifying the important residues for protein -protein interactions (12). (3) Stabilizing the protein -protein interaction (13). (4) Identifying the nuclear localization signal (14) and the mechanism of the transfer from cytosol to the cell nucleolus (15). They indicated that from the two isoforms of the CRABP, CRABPI transfers the Retinoic acid to RAR via a dissociation mechanism, independent from the RAR concentration. However, CRABPII goes through direct protein -protein interaction to deliver the ligand to RA R. Different biological assays were Figure V -2) a) Crystal structure of the Ligand Binding Domain of holo -human RAR -! shows the Retinoic Acid inside the binding pocket of the protein (PDB ID: 2LBD). OOHAll-trans-Retinoic Acid!"##!performed to characterize CRABPII and RAR interaction for Retinoic acid transfer which includ ed the chemical cross linking, e lectrophoresis under non-denaturing condition , fluorescence anisotropy titrations and electroph oretic mobility shift assays . The complex formation was not detected under any of these conditions (11). Electrostatic calculation s highlighted three residues that differ significantly in CRABPI and CRABPII surface potential from one CRABP isoform to the other. These include the Q75, P81 and K102 in CRABPII sequence, which the equivalent residues in CRABPI are E75, K81 and E102. Mutating these residues of CRABPI to equivalent residues in CRABPII provided a protein that was able to directly interact with RAR. On the other hand, mutating these residues of hCRABPII to equivalent residues in hCRABPI provided a p rotein that was not able to directly interact with RAR (12). The translocalization of CRABPII to the cell nucleus for retinoic acid delivery to RAR was further proved by in -vivo fluorescence imaging, while the result of this experiment f or CRABPI was negative. Eventually, a novel idea came in handy to be the only known experiment so far to visualize the two proteins together on an SDS -PAGE gel . The rationalization was that the CRABPII -RAR interaction is so transient that is hard to trap the complex of the two proteins , and it comes from the higher affinity of the RAR for Retinoic acid compared to CRABPII. The two proteins interact transiently that CRABPII delivers the Retinoic acid to RAR immediately and leaves. The idea was to design a ligand with high affinity for CRABPII and lower affinity for RAR . That way the two proteins will interact, but since the ligand will not partition to RAR easily, the two proteins will stay together longer. The ligand of choice was a !"#$!molecule named CD -270, which has 100 times more affinity for CRABPII compared to RAR, while Retinoic acid has nanomolar or sub nanomolar affinity for the both proteins. It should be mentioned that the in -vivo fluorescence imaging experiment of CRABPII with CD -270 shows the trans localiztion of the CRABPII to the nucleus as well. A pull down assay in the presence of the Retinoic Acid or CD -270 was performed. The two proteins were pulled in the same fraction while CD -270 was used and different fractions when the Retinoic Acid was used (Figure V-3) (13). SOOHCD-270Figure V-3) a) The result of the pull down assay for the Retinoic acid and CD 270 M) MW marker 1) CRABPII at 50% percent of total input . 2) RAR !-LBD on the Ni -NTA resin. 3) RAR !-LBD incubated with CRABPII in the absence of ligand 4,5) RAR !-LBD incubated with CRABPII in the presence o f 7.5 and 18.75 µM Retinoic acid, respectively. 6,7) RAR !-LBD incubated with CRABPII in the presence of 7.5 and 18.75 µM of CD -270, respectively. b) C hemical structure of CD -270. a) b) !"#$!V.2 Synthesis of CD -270 We were interested to visualize the interaction of the CRABPII and RAR with an atomic resolution description. To this end the synthesis of the CD -270 was targeted. The synthesis of the ligand from original patent pathway tried by Dr.Berbasova was not successful (Figure V-4) (16). A closer search in the literature showed the synthesis of benzothiophene molecules similar to CD -270 via the progress in the Suzuki cross -coupling reactions (17). A retro synthesis strategy was designed and the CD -270 was bro ken to two halves (Figure V-5). COOHCH3CH3ClCl/ AlCl3KMnO4HSCOOHH2NCOOHi) HNO2ii) NaHSO3, NaOHOSCOOHBr2OSCOOHBri) PPh3ii) DBUSCOOHHOOHHCl11Inconclusive mess65 %99 %99 %5 %Figure V -4) CD -270 synthesis trial with t he original proposed synthesis was not successful (With Dr.Tetyana BerbasovaÕs permission) . !"#$! The synthesis of the first half was performed successfully via an iridium C-H activation reaction developed in Prof.Smit h lab over the last two decades in the chemistry Department at Michigan State University (Figure V-6) (18-22). OHOHClCl4N HClAlCl3[Ir(Cod)OMe]2, 2 mol%dtbpy, 4 mol%B2Pin2 (1.5 equiv)Cyclohexane, 60 ¡C,48hBOO178 %85 %dtbpy =NN70 %Figure V-6) The synthesis of the first fragment of the CD -270. BrBrSHRB(OH)2PdCl2 (3 mol%)SPhos (3 mol%)K3PO4/Et3N (3 equiv)dioxane, 110 ¡CSRR1R1SPhos =PCy2OMeMeOSOOHCD-270SOOHSuzuki cross couplingBOOSHBrBrEtOO12a) b) Figure V -5) a) Reported modified Suzuki cross coupling reaction b) proposed retrosynthesis for CD -270. !"#$!The synthesis of the second half proved to be more challenging. The first two steps of the synthesis of this fragment worked. Also, the protection of the aldehyde worked the best with the reagent shown (23) (Figure V-7). The rest of the synthetic plan is as shown in the following scheme , which was not tried as the first step did not work . The Grignard formation followed by carboxylation or organo lithiation followed by ethyl chloroformate addition did not work to carboxylate the aromatic ring . At this point the synthesis of the sec ond fragment was not troubleshot extensively to find the way for the synthesis of this fragment ( Figure V-8). HSBrOMeMeO1) nBu-Li, -78 ¡C3) H3O+2)OOClHSOMeMeOOOPPh3, CBr4CH2Cl20 ¡Cr.t.TSOH, H2O/Toluene110 ¡C, OvernightSHHOOBrBrSOOBrBrHClO4-SiO2( 5 mol % )MeOH/H2O 1:1r.t., 1hHSOOO2Figure V-8) The rest of the synthetic plan for the second fragment. Figure V-7) The first two steps in the synthesis of the second fragment. FBrHO tBuSH (1 equiv)NaH (1 equiv)/ DMFStBuBrHO0 ¡Cr.t.1hOver nightHSBrOMeMeOHClO4-SiO2( 5 mol % )75 %TMOF (1.2 equiv)MeOH61 %!"#$!It should be mentioned that CRABP -RAR interaction was studied via a transporter to trap conversion as well. It was suggested that the second !-helix and "C-D sheet of CRABP protein s goes through conformational changes as the Retinoic acid binds to this pro tein (Figure V-9). Two residues, which go through the most conformational changes Ala35 and Thr57, were highlighted and mutated to Cysteine residues to convert the CRABP I protein from a transporter of Retinoic acid to a trap for it via the formation of a disulfide bond . The f ormation of the disulfide bond was confirmed by prot eolysis fragmentation of CRABPI . Further studies showed that CRABPI penta mutant E75Q:K81P:E102K:A35C: T57C can stabilize the interaction of the CRABP -RAR complex (24, 25). !"#$! Figure V-9) a) Overlay of th e structure of the Apo -CRABPII -Chain A (PDB ID: 2FS7) with the Holo structure (PDB ID: 2FR3) does not show the difference in position of Ala 35 and Thr 57. b) Overlay of the structure of the Apo -CRABPII -Chain B and Holo -CRABPII shows the two residues of Ala35 and Thr57 get close to each other upon Retinoic acid binding. c) Overlay structure of the Apo ( PDB ID: 1CBI) and Holo CRABPI (PDB ID: 1CBR) shows that the A35 and Thr57 get close to each other upon Retinoic acid binding. 3.7 † T57 A35 A35 T57 3.7 † 3.1 † 4.2 † T57 A35 7 a) c) b) T57 !"#"!Dr.Xia tried a pull down assay in Prof.GeigerÕs lab as well. The idea was to use the R111K:R132L:L121E mutant of hCRABPII which effectively binds to retinal and take the protein -Retinal complex through a reductive amination reaction to make the retinal non -hydrolysable and non -transfer able to RAR. The pull down assay was performed and the result did not show the two proteins together (26). Also the synthesis of the CD -270 or its close analogs, and trying to crystallize the complex of C RABP and RAR is one way of approaching this question, using the library of molecules that now we have available f or hCRABPII, trying them with first and second generation mutants of hCRABPII , measuring their affinity for RAR and CRABPII and trying a pull d own assay with these ligands is another alternative approach. V.3 Conclusion The interaction between the CRABPII and RAR proteins is essential in the delivery of Retinoic acid to RARs and regulation of their transcriptional activity. We aimed at synthesizing molecule CD-270 that is the only molecule already shown to stabilize the interaction between the two proteins. We aimed for the advantage of using C-H activation and cross -coupling reaction s in our synthesis . The molecule was broken into two p recursors. The first precursor was synthesized via an iridium C-H activation reaction. The synthesis of the second fragment was hampered by difficulties in carboxylation of the aromatic ring. !"#$! REFERENCES !"#$!REFERENCES "%!&'()!*+!,-./0'1!2+!3!45-6780!2!9"::;/(?>(@(/A!B'0'1-/'.!)(C'1.(/A!(0! /5'!1'/(08(@!-@()!.(B0->>(0B!?-/5D-A.%! !"#$%&'($')(*+,#-(+./'&+(#$+#& !"E9"F'-1! 1'@'?/81.%! 0"*1#($'2"*3(/# !;9""'-1!1'@'?/81.!(0!5'-)! -0)!0'@Z!@-0@'1G!@=11'0/!Z08D>')B'!-0)!?'1.?'@/(C'.%! :$1#"$.1(*$./' ;*6"$./'*3'+.$+#"<'=*6"$./'($1#"$.1(*$./'%6'+.$+#" !";V9H-1!@5-0B'.!(0! B'0'!']?1'..(80!-0)!(0)=@'!.'0'.@'0@' I>(Z'!B18D/5!-11'./!(0!*4^ IE!71'-./!@-1@(086-!@'>>.%! >.$+#"'"#&#."+, !VV9"E>!B 18D/5!7A!1'/(08(@!-@()G!/5'!@'>>!@A@>'!@80/18>!B'0'!P/B;!(.!-!)(1'@/! /-1B'/!N81!1'/(08(@!-@()!1'@'?/81!.(B0->(0B%! >.$+#"'"#&#."+, !VE9;>=>-1!1'/(08(@!-@() I7(0)(0B!?18/'(0!\\!.(B0->(0B%! >.$+#"'"#&#."+, !V#9"O.$+#"'"#&#." +,!V#9"O-(0!W ?'#1'./< !9;FF:>(0B!(0!/5'!5=6-0!>=0B%! @6$7'+.$+#" !VV9"!./1=@/=1'!8N!/5'!UYU IB-66-!>(B-0) I7(0)(0B!)86-(0!78=0)!/8!->> I/1-0.!1'/(08(@!-@()%! A.16"#!$EO9V##O![T+!3!Q8A!Q!9":::'.!N81!@'>>=>-1! 1'/(08(@!-@() I7(0)(0B!?18/'(0.!\!-0)!\\!(0!1'B=>-/(0B!.(B0->(0B!7A!1'/(08(@! -@()%! !,#'=*6"$./'*3')(*/*7(+./'+,#-(&1"8 !;EH9$H>(>-0 !U+!3!Q8A!Q!9;FF"(X-/(80!8N!/5'!UYU!(0/'1-@/(80! )86-(0!8N!@'>>=>-1!1'/(08(@!-@()!7(0)(0B!?18/'(0 I\\%!=*6"$./'*3'-*/#+6/."' )(*/*78 !$F#9H!/?!78:6!C7/:867/A:*! ;778E:&! !"#$%&#'()'*+)%$##&#'(),-"#"./ !221H4I2J%2B2J$"&!!"$&!,8EE=87!KL!-!./0!.!1233#4!+!=6>;<) B;9:6M;:8)!<(9=8;7!=/9;=6F;:6/<;=!6!C7/:86< BDD&!!"#$%&#'()%$## !"H1%4I%$%B%#%&!!"#&!N;O(G);7!+P!Q8:78E9(!+5P!R6/<>!SP!-!./0!.!123""4!.(9=8;7!:7; !C7/:86(=;:8)!@0!78:669;=!;9:6M6:0!/?!78:6!C7/:86<&! 9:-*) ;1'(4'%"#"./)<)61$)"33-%-'#)="&(*'#)"3)61$)9:-*)>1'(4'%"#"./)9"%-$6/ !"1%4I"J"B"W#!1;<)!78?878<98E!:*878!687!L+P!-!Y;(:8&! ?*.$@'*+6$)A1$4-$ !$H1%H4IW3J$BW3JH&!!"H&!N;=89F[;!KZP!L7&P!,*6!\P!]/=G8E!5P!-!,G6:*!NKP!%7)!1233%4!X B]!;9:6M ;:6/<^@/70=;:6/<^/_6);:6/!/7:*/ B^C;7; B)6789:6<>!>7/(CE&! 2"&(*'#)"3)61$) ?4$(-%'*)A1$4-%'#)9"%-$6/ !"2#12J4IWWV2BWWV%&!!"V&!Q;(=!, 7)$6)'#8 !1233J4!D7 B9;:;=0F8)!?(<9:6/<;=6F;:6/8< B)6789:8)!;7/G;:69!@/70=;:6/<&! 2"&(*'#)"3)61$) ?4$(-%'*)A1$4-%'#)9"%-$6/ !"2H1$V4I"###2B"###%&!!23&!`;<9*(7;!'+P!2<) 7)$6)'#8 !123"34!Z=89:7/<69!8??89:E!6*:E!?7/G!(<8<9(G@878)!E(@E:7;:8E!;<)!M;76; :6/;<)!E(@E:6:(8<:E&! A1$4-%'#)%"44&*-%'6-"*5 !$J1$"4IWW2$BWW2J&!!2"&!a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hapter VI: Materials and Methods UV-Vis spectra were recorded using a Cary 300 Bio WinUV, Varian spectrophotometer. All -trans -retinal was purchased from TRC and was used as received. All the other chemicals were purchased from Sigma -Aldrich unless otherwise specified. Fast Q anion exchange resin was purchased from Agilent Technologies. BioLogic DuoFlow (BioRad) was used for protein purifications. Source 15Q resin was purchased from Am ersham Bioscience. Sonication of bacterial cells was performed using Ultrasonic Homogenizer from Biologics, Inc. VI.1 Site -directed mutagenesis The CRABPII -pET17b plasmid described in (1) was used for mutagenesis following the QuickChange Site -directed Mutagenesis Kit protocol (Agilent Technologies). PCR conditions for amplification of mutants are specified below: Total Reaction Volume 50 L Template ( DNA plasmid ) 70ng ( x L ) Primer Forward 20 pmol ( y L ) Primer Reverse 20 pmol ( z L ) dNTP 1 L 10x pfu Buffer 5 L Pfu Turbo (DNA Polymerase) 1 L DI water 50-x-y-z-7 L Table VI-1. PCR protocol for hCRABPII and hCRBPII mutagenesis. !!"#$!Table VI -1 (con tÕd) The PCR product (5 L) was transformed into DH5 cells (50 L) and grown on Luria -Bertani (LB) -agar plates supplemented with Ampicillin (75 g/mL) for 15h. A single colony was inoculated in 15 mL LB medium containing 100 g/mL ampicillin and grown at 37 ¼C while shaking, for 12 hours. DNA purification was performed using a QIAGEN Miniprep DNA purification kit. The average isolated plasmid DNA concentration was 120 ng/ l in 60 L solution. The DNA sequence was verified by the MSU gene sequencing facility using T7 primer. Primers F3Q Forward: 5Õ- CCAAAC CAA TCTGGCAACTG GAAA -3Õ Reverse: 5Õ -TTTCCAGTTGCCAGA TTGGTTTGG -3Õ !A32Y Forward: 5Õ-GAGGAAGATT TAT GTGGCTGC -3Õ Reverse: 5Õ -GCAGCCAC ATA AATCTTCCTC -3Õ!!A32W Forward: 5Õ -GTGATGCTGAGGAAGATT TGG GTGGCTGC -3Õ Reverse: 5Õ -GCAGCCAC CCA AATCTTCCTCAGCATCAC -3Õ PCR Program 1x 95 C 30 min 95 C 30 sec 20x Temprature 3 -5 C lower than primer melting temprature 1 min 72 C 4min 30 sec 1x 72 C 10 min 1x 25 C 10 min !"#$!P39Y For: 5' GCAGCGTCCAAG TAT GCAGTGG 3' Rev: 5' CCACTGC ATA CTTGGACGCTGC 3' ! P39Q For: 5' GCAGCGTCCAAG CAA GCAGTGG 3' Rev: 5' CCACTGC TTGCTTGGACGCTGC 3' ! T54V Forward: 5Õ -CTACATCAAA GTC TCCACCACCGTGCG -3Õ Reverse: 5Õ - CGCACGGTGGTGGA GAC TTTGATGTAG -3Õ! R59Y Forward: 5Õ - CCTCCACCACCGTG TAC ACCAC AGAG -3Õ Reverse: 5Õ - CTCTGTGGT GTA CACGGTGGTGGAGG -3Õ! R59W Forward: 5Õ - CCTCCACCACCGTG TGG ACCACAGAG -3Õ Reverse: 5Õ - CTCTGTGGT CCA CACGGTGGTGGAGG -3Õ! R111K Forward: 5Õ -CCCAAGACCTCGTGGACC AAA GAACTGACCAACGATGGG -3Õ Reverse: 5Õ -CCCATCGTTGGTCAGTTC TTTGGTCCACGAGGTCTTGGG -3 R132Q: Y134F Forward: 5Õ - GTTGTGTGCACC CAG GTC TTC GTCCG -3Õ Reverse: 5Õ -CGGAC GAA GAC CTG GGTGCACACAAC -3Õ R132Y:Y134F Forward: 5' GTTGTGTGCACC TAT GTC TTCGTCCGAGAG 3' Reverse: 5' CTCTCGGAC GAA GAC ATA GGTGCACACAAC 3' VI. 2 Protein Expression and Purific ation of hCRABPII and hCRBPII/pET -17b Mutants The target gene was transformed into competent E. coli BL21(DE3)pLysS compet ent cells (100 ng of DNA, for 10 0 L of cell solution) following standard protocols and the cells were grown on Luria -Bertani ( LB) -agar plates supplemented with antibiotics (Chloram phenicol: 34 g/mL; Ampicillin: 100 g/mL) at 37 ¼C for 12 hours. A single colony was used to inoculate 50 mL of !"#$!LB medium containing 34 g/mL chloramphenicol and 100 g/mL ampicillin and was grown at 37 ¼C, while shaking overnight. The resulting culture was used to in oculated 1 L of LB containing 27 g/mL chloramphenicol and 100 g/mL ampicillin and was grown at 37 ¼C while shaking till OD 600 reached 0.7 -0.9. The expression was induced with addition of isopropyl --D-thiogalactopyranoside (IPTG, Gold Biotechnology, 1 mM). The culture was shaken at 19 ¼C for 36h for hCRABPII and at room temperature for overnight for hCRBPII. The cells were harvested by centrifugation (5000 rpm, 20 min, 4 ¼C) and resuspend ed in Tris -binding buffer (10 mM Tris, pH=8.0, 50 mL). The cells were lysed by sonication (VWR Scientific , power 70%, 3 x 1 min). The solution was spun down (12000 rpm, 30 min, 4 ¼C) and the supernatant was loaded on a FastQ anion exchange column pre equil ibrated with Tris -binding buffer. The column was washed twice with Tris -binding buffer (2 x 100 mL) and the protein was eluted with Tris -elution buffer (10 mM Tris, 150 mM sodium chloride, pH = 8.0, 70 mL). The eluent underwent buffer exchange with Tris -binding buffer using an unltrafiltration cell under nitrogen pressure (~20 psi) (Millipore TM, Regenerated Cellulose membrane YM10, NMWL: 10,000). The desired protein was further purified using a BioLogic DuoFlow system (BioRad) equipped with a 15Q anion exc hange column, according to the following protocol: !"#$! The buffer used for above protocol is 50 mM Tris, pH is adjusted automatically. The proteins elute with 4% 2M NaCl. VI. 3 Extinction Coefficient Determination The absorption extinction coefficients ( ) for the various CRABPII mutants were determined according to the method described by Gill and von Hippel (2). VI. 4 UV-vis Measurements The CRABPII -PSB formation ( max > 450 nm) was followed by UV -vis in phosphate buffer (4 mM NaH 2PO4, 16 mM Na 2HPO 4, 150 mM NaCl, pH=7.3) or 50 mM citrate buffer at pH=5.0. The experiment was performed with a final protein concentration of 50 M, and 0.5 equivalent of retinal was added (from a stock solution of 1 mM in ethanol). Peaks with max > 450 nm are considered PSB peaks, while deprotonated imine peaks (SB) appear at ~ 360 nm. Non -covalently bound retinal absorbs at ~380 nm. VI. 5 pKa determination s: hCRABPII mutants : The pKa value s were determined based on the plot of the absorbance change vs pH by using the curve fit as described previously for bacteriorhodopsin pKa determination. Description Parameters Isocratic flow pH=8.1, 0% 2M NaCl 10.00 ml, 3.00 ml/min Linear Gradient pH=8.1, 0 -4% 2M NaCl 20.00 ml, 3.00 ml/min Isocratic flow pH=8.1, 4% 2M NaCl 20.00 ml, 3.00 ml/min Linear Gradient pH=8.1, 4 -8% 2M NaCl 10.00 ml, 3.00 ml/min Isocratic flow pH=8.1, 8% 2M NaCl 20.00 ml, 3.00 ml/min Table VI -2. Anion Exchange purification protocol for hCRABPII and hCRBPII. !"#"!The total absorbance change of PSB during each acidification or basification in comparison to the point with minimum PSB absorption () is plotted against the pH (3, 4). 0 [pH Ð pKa] ) hCRBPII mutants : max versus pH were plotted in KaleidaGraph. The curve was fit to a third degree polynomial and the pKa was determined from the second derivative of the polynomial expression set to 0 (point of inflection): y = ax 3 + bx 2 + cx + d yÕ = 3ax 2 + 2bx + c yÕÕ = 6ax + 2b Base titrations in citrate buffer : A 50 M solution of hCRABPII proteins was incubated with 0.5 equivalent of retinal in 50 mM citrate buffer at pH=5.0. The UV-Visible spectrum was followed till maximum PSB formation. The leftover SB was converted to PSB by acidifying the solution with 2M H Cl before titration. Then a 1M NaOH solution was used to basify the protein solution till complete PSB to SB conversion. !Acid titrations : A 50 M solution of hCRABPII proteins was incubated with 0.5 equivalent of retinal in PBS buffer. The UV -Visible spectrum was followed till maximum PSB loss (12 -24h). Then a saturated citric acid buffer solution was used to acidify the protein for the first few points by further acidification with a 6M HCl solution till complete SB to PSB conversion. For F3Q mutants of hCRABPII the acid titration was performed at PSB maximum formation. !"#$!Base titrations in PBS buffer : A 50 M solution of hCRABPII proteins was incubated with 0.5 equivalent of retinal in PBS buffer. The UV -Visible spectrum was followed till maximum PSB fo rmation. Then a 1M NaOH solution was used to basify the protein solution till complete PSB to SB conversion. UV irradiation - titrations in PBS buffer : A 50 M solution of hCRABPII proteins was incubated with 0.5 equivalent of retinal in PBS buffer. The UV -Visible spectrum was followed till maximum PSB formation. Then a UV machine with a 340 -380 light filter pass was used to convert the leftover SB to PSB (1 -2 minute irradiation). A 1M NaOH solution was used to basify the protein solution till complete PSB to SB conversion. In case of M2 the UV irradiation was after complete PSB loss. hCRBPII titrations in PBS buffer : A 25 M solution of hCRBPII proteins was inc ubated with 0.5 equivalent of retinal in PBS buffer. The UV -Visible spectrum was followed to track the PSB formation and loss. For base titration a 1M NaOH solution was used to basify the protein solution till complete PSB to SB conversion. For acid titra tion a saturated citric acid buffer solution was used to acidify the protein for the first few points by further acidification with a 6M HCl solution till complete SB to PSB conversion. All of the graphs were recorded with a UV -Vis spectrophotometer. Graphs were recorded as Excel and Kaleidagraph files, and absorption at max vs different pHs were plotted. pKa values were det erm ine as described in section VI-5. !"#$! Table VI-3. hCRABPII mutants. a: Measurements were done in 50 mM in 50 mM citrate buffer, pH = 5.0. !"#$!Table VI -3 (contÕ d) a: Measurements were done in 50 mM in 50 mM citrate buffer, pH = 5.0. !"#$!Table VI -3 (con tÕd) a: Measurements were done in 50 mM in 50 mM citrate buffer, pH = 5.0. !"##!Table VI -3 (con tÕd) a: Measurements were done in 50 mM in 50 mM citrate buffer, pH = 5.0. !"#$!Table VI -3 (con tÕd) a: Measurements were done in 50 mM in 50 mM citrate buffer, pH = 5.0. !"#$! Protein 280nm (M-1.cm -1) KL:R58Y :Q38A 31431 KL:R58Y:T51V: Q38A 28945 KL:R58Y:T51V: Q38A :Q128A 28680 KL:R58Y:T51V: Q38L :Q128L 28431 KL:R58Y:T51V:Y19W: Q38A 33765 KL:R58Y:T51V:Y19W: Q38 L 32920 KL:R58Y:T51V:Y19W: Q38 M 34002 KL:R58Y:T51V:Y19W: Q38 F 33982 KL:R58Y:T51V:Y19W: Q38A :Q128A 34556 KL:R58Y:T51V:Y19W: Q38C :Q128L 33909 KL:R58Y:T51V:Y19W: Q38F :Q128L 34006 KL:R58W:T51V:T53C:T29L: Q38 N 33718 KL:R58W:T51V:T53C:T29L:Y19W: Q128L 35604 KL:T51V:T53C:T29L:Y19W: Q128L 31262 KL:R58W:T51V:T53S:T29L:Y19W: Q38L 37670 KL:R58W:T51V:T53S:T29L:Y19W: Q38N 37326 KL:R58W:T51V:T53S:T29L:Y19W: Q38L :Q128L 38579 KL:R58W:T51V:T53C:T29L:Y19W:Q4R: Q128L 35936 KL:R58W:T51V:T53C:T29L:Y19W:Q4R:A33W: Q38L 41977 KL:R58W:T51V:T53C:T29L:Y19W:Q4R:A33W: Q128L 45062 KL:R58W:T51V:T53S:T29L:Y19W:Q4F:A33W: Q128L 42875 KL:R58W:T51V:T53S:T29L:Y19W:Q4F:A33W: Q38N - KL:R58W:T51V:T53S:T29L:Y19W:Q4F:A33W: Q128N - KL:R58W:T51V:T53S:T29L:Y19W:Q4F:A33W: Q38L :Q128L - Table VI -4. hCRBPII Proteins Extinction Coefficient. KL is the abbreviation for Q108K:K40L !"#$! Q38 and Q128 mutations on the parent mutants absorbing below 550 nm. Table VI-5. hCRBPII mutants. Q38 and Q128 mutations on Q108K:K40L:R58Y:T51V:Y19W mutant ( highest pKa mutant of hCRBPII ). !"#$!Table VI -5 (con tÕd) Q38 and Q128 mutations on the parent mutants absorbing between 550 -600 nm. Q38 and Q128 mutations on the parent mutants absorbing above 600 nm. !"#"!VI.8 Protein Crystallization of hCRABPII VI.8 -1 R!111K:P39Y:R132Q:Y134F:T54V:R59Y For kinetic PSB product : To the concentrated protein solution at 20 mg/ml was added four equivalents of retinal and incubated for 20 minutes. A 24 well crystallization plate was used for vapor diffusion crystallization with the reservoir solution of 1 ml, containing 12 % PEG3350 and 0.1 M malonate at Crystals appeared and grew after 24 hours at 4 ¡C. Crystals we re flash frozen in a solution containing the mother liquor and 20 % glycerol. For Thermodynamic SB product : To the protein solution at 1.1 mg/ml was added four equivalents of retinal. The spectrum of the protein -ligand mixture was tracked until PSB loss completion. Then the mixture was concentrated to 20 mg/ml using concentrator cells. A twenty -four well crystallization plate was used for vapor diffusion crystallization with the reservoir solution containing 12 % PEG 3350 and 0.1 M malonate at pH = 6.0. 1 These crystals had a light blue color indicating the presence of PSB in the crystals. In order to deprotonate the PSB in the crystals, they were soak ed in 12% PEG 3350 and 0.1 M malonate at pH = 7.5 for 30 minutes until crystals became colorless. The soaked crystals were flash frozen in the same solution containing 20 % glycerol. For UV irradiated PSB product : Soaked crystals described above were UV irradiated with a hand held TLC UV lamp for 30 minutes in dark. The UV irradiated crystals showed a dark blue color at the end of the irradiation !"#$!period. These crystals were transferred to a solution of 12% PEG 3350, 0.1 M malonate at pH = 7.5 and 20 % gly cerol and flash frozen immediately under dim microscope light, while still being irradiated with the UV light. VI.8 -2 R!111K:P39Q:R132Q:Y134F:T54V:R59Y For Thermodynamic SB product : To the protein solution at 1 mg/ml was added four equivalents of retinal . The spectrum of the protein -ligand mixture was tracked until PSB loss completion. Then the mixture was concentrated to 20 mg/ml using concentrator cells. A twenty -four well crystallization plate was used for vapor diffusion crystallization with the reser voir solution containing 12 These crystals had a light red color indicating the presence of PSB in the cry stals. In order to deprotonate the PSB in the crystals, they were soaked in 12% PEG 3350 and 8 % Tacsimate at pH = 7.5 for 45 minutes untill crystals became colorless. The soaked crystals were flash frozen in the same solution containing 20 % glycerol For UV irradiated PSB product : Soaked crystals described above were UV irradiated with a hand held TLC UV lamp for 1 hour in dark. The UV irradiated crystals showed a red color at the end of the irradiation period, and were flash frozen immediately under dim m icroscope light in 12% PEG 3350, 8 % Tacsimate at pH = 7.5 and 20 % glycerol. VI.8 -3 R111K:P39Q:R132Q:Y134F:T54V:R59Y:A32Y:F3Q To the concentrated protein solution at 20 mg/ml was added four equivalents of retinal and incubated for three hours. A 24 wel l crystallization plate was !"#$!used for vapor diffusion crystallization with the reservoir solution of 1 ml, hours and grew in 5 days at 4 ¡C. Crystals were flash frozen in a solution containing the mother liquor and 20 % glycerol. The same method was used for crystallization and cryo protection of kinetic PSB of mutant M7 in the tables after 20 minutes of retinal incubation . VI.8 -4 R!111K:P39Q:R132Q:Y134F:T54V:R59Y:A32W:F3Q To the concentrated protein solution at 20 mg/ml was added four equivalents of retinal and incubated for three hours. A 24 well crystallization plate was used for vapor diffusion crystallization with the reservoir solution of 1 ml, after 24 hours and grew in 5 days at 4 ¡C. Crystals were flash frozen in a solution containing the mother liquor and 20 % glycerol. The same method was used for the crystallization and cryo protection of thermodynamic -SB of mutant M5 in the tables after 24 hours of retinal incubation. VI.9 Pr otein crystallization of hCRBPII To the concentrated protein solution at 8 -12 mg/ml is added four equivalents of retinal in dark. The solution is allowed to sit at room temperature until maximum PSB formation. A 24 well crystallization plate was used for vapor diffusion crystallization with the reservoir solution of 1 ml, containing 25 % PEG 4000 and 0.1 M Ammonium Acetate and 0.1 M Sodium Acetate at pH !"#$!range 4 - Crystals appeared after 24 hours and grew in 5 days at 4 ¡C. Crystals were flash frozen in a solution containing the mother liquor and 20 % glycerol %!Most of hCRBPII Proteins also show PSB loss overtime. In some cases for crystallizing more of a homogenous solution that can improve diffraction and provide better electron density maps a solution of hCRBPII protein at 1 mg/ml protein was incubated with 4 equivalent retinal and the UV spectrum was followed until maximum PSB loss. Then this solution was concentrated to 8 -12 mg/ml for cr ystallization trials. Waiting method was tried for all of the hCRBPII proteins reported here except Q108K:K40L:T51V:R58Y:Y19W, Q108K:K40L:T51V:R58Y:Y19W:Q38L and structures crystallized with Julolidene. ! VI. 10 Data Collection and Refinement Diffraction da ta were collected at the Advanced Photon Source (APS) (Argonne IL) at the LSCAT (sector 21) at 1.00 † wavelength radiation at 100 K. Data reduction and scaling were performed usin g the HKL2000 program package (5). The structures were solved using Molecular Replacement in CCP4 progra m suite (6), and refined in PHENIX program packages (7). The search model was R132K:R111L:L121E mutant of Cellular Retinoic Acid Binding P roteinII (protein data bank code 2G7B). Three cycles of refinement was considered for each run and placement of ordered water molec ules was done in COOT (0.6.1) (8). The chromophore was manually fitted in the electron density at the end of the refinement. For hC RBPII protein the 2RCT was used as the model for refinement. !"#$! a) c) d) b) Figure VI -1) Retinal incubation for crystallization of: a) R111K:P39Q:R132Q:Y134F:T54V:R59Y mutant of hCRABPII b) R111K:P39Y:R132Q:Y134F:T54V:R59Y mutant of hCRABPII . c) Q108K:K40L mutant of hCRBPII d) Q108K:T51D mutant of hCRBPII . e) A photo of R111K:P39Y:R132Q:Y134F:T54V:R59Y crystal. f) A photo of R111K:P39Q :R132Q:Y134F:T54V:R59Y :A32W:F3Q. e) f) !"#$! Substrate M1b-SB-cycle1 M1-UV-cycle1 M1-SB-cycle2 M1-UV-cycle2 M1-SB-cycle3 M1-UV-cycle3 Space group P3121 P3121 P3121 P3121 P3121 P3121 a (†) 58.552 58.326 58.599 58.488 58.509 58.655 b (†) 58.552 58.326 58.599 58.488 58.509 58.655 c (†) 99.392 99.936 99.854 100.402 99.688 100.352 (¡) 90 90 90 90 90 90 (¡) 90 90 90 90 90 90 120 120 120 120 120 120 Molecules per Asymmetric Unit 1 1 1 1 1 1 Total reflection 189007 163428 161415 161005 155486 145510 Unique Reflection 17990 15647 14985 14868 14966 13508 Completeness (%) 99.6 (99.6) a 99.4 (85.3) 99.5 (99.3) 100 (100) 99.4 (99) 100 (99.8) Average I/ 36.1 (3.78) 34.86 (3.65) 33.94 (4.69) 39.06 (4.38) 26.69 (2.00) 38.81 (4.38) Rmerge (%) 7.1 (52.3) 8.3 (40.5) 8.9 (44.6) 8 (54.4) 9.8 (57.3) 6.7 (55.7) Resolution (†) (Last Shell) 50-1.83 (1.86 -1.83) 50-1.92 (1.95 -1.92) 50-1.95 (1.98 -1.95) 50-1.96 (1.99 -1.96) 50-1.95 (1.98 -1.95) 50-2.03 (2.07 -2.03) Rcryst/Rfree (%) 18.73/24.20 !19.74/24.45 !18.99 /23.80 20.52/24.93 19.16/24.05 21.02/25.73 RMSD From Ideal Values Bond Length ( †) 0.007 0.008 0.007 0.008 0.008 0.008 Bond Angle (¡) 1.192 1.228 1.106 1.104 1.201 1.152 Average B factor 26.90 29.8 29.76 34.71 29.99 36.24 Number of water molecules 199 141 161 123 147 106 PDB IDs 4YBP 4YBU 4YCE 4YCH 4YDA 4YDB Table VI -6. The hCRABPII crystal table. a Values in the parenthesis refer to the last resolution shell. b M1: R111K:P39Q:R132Q:Y134F:T54V:R59Y !"##!Table VI -6 (con tÕd) Substrate M2b-Kinetic PSB M2-SB- cycle1 M2-UV- cycle1 M2-SB- cycle2 M2-UV- cycle2 M2-SB- cycle3 M2-UV- cycle3 Space group P3121 P3121 P3121 P3121 P3121 P3121 P3121 a (†) 58.935 58.869 58.865 59.199 58.779 58.975 59.031 b (†) 58.935 58.869 58.865 59.199 58.779 58.975 59.031 c (†) 99.954 99.71 100.441 100.038 101.228 100.443 101.123 (¡) 90 90 90 90 90 90 90 (¡) 90 90 90 90 90 90 90 120 120 120 120 120 120 120 Molecules per Asymmetric Unit 1 1 1 1 1 1 1 Total reflection 164293 281303 163969 179720 132920 140437 126296 Unique Reflection 15102 25940 15237 16619 12329 13048 11730 Completeness (%) 99.5 (99.4) a 99.4 (98.8) 99.6 (99.6) 99.8 (99.2) 99.8 (99.3) 99.9 (99.8) 100 (99.9) Average I/ 34.6 (3.0) 62 (4.44) 49.5 (4.41) 48.04 (4.53) 41.94 (4.62) 40.72 (5.12) 38.92 (4.1) Rmerge (%) 9.4 (63.9) 4.7 (47.9) 6 (52.6) 5.9 (53) 6.5 (51.6) 6.8 (53.3) 7.5 (54.2) Resolution (†) (Last Shell) 50-1.95 (1.98 -1.95) 50-1.62 (1.65 -1.62) 50-1.95 (1.98 -1.95) 50-1.90 (1.93 -1.90) 50-2.10 (2.14 -2.10) 50-2.06 (2.10 -2.06) 50-2.14 (2.18 -2.14) Rcryst/Rfree (%) 19.97/24.88 !19.47/22.74 !20.01 /26.53 20.43/23.81 20.38/23.65 20.16/25.45 19.87/23.46 RMSD From Ideal Values Bond Length (†) 0.008 0.006 0.007 0.007 0.007 0.009 0.009 Bond Angle (¡) 1.301 1.158 1.144 1.123 1.097 1.228 1.167 Average B factor 35.2 29.74 35.97 32.94 39.96 36.02 40.63 Number of water molecules 108 215 116 135 90 123 92 PDB IDs 4YFP 4YFQ 4YFR 4YGG 4YGH 4YGZ 4YH0 a Values in the parenthesis refer to the last resolution shell. b M2: R111K:P39Y:R132Q:Y134F:T54V:R59Y !"#$!Table VI -6 (con tÕd) Substrate M6b-Thermodynamic SB M8c-Kinetic PSB M11d M3e Space group P1 P21 P1 P1 a (†) 39.716 82.839 36.737 36.574 b (†) 65.13 45.455 39.21 39.54 c (†) 103.743 93.963 58.402 58.183 (¡) 80.446 90 86.775 87.078 (¡) 88.391 116.004 79.881 80.016 (¡) 71.303 90 70.831 70.511 Molecules per Asymmetric Unit 6 4 2 2 Total reflection 200218 206425 135505 169219 Unique Reflection 73056 45694 42116 44746 Completeness (%) 97.5 (97.2) a 95 (65.2) 93.3 (70.2) 96.0 (93.4) Average I/ 15.69 (4.46) 22.66 (1.92) 31 (2.55) 36.41 (2.97) Rmerge (%) 9.1 (25.9) 7 (45.7) 6.9 (28.8) 6.1 (47.0) Resolution (†) (Last Shell) 50-2.10 (2.17 -2.10) 50-1.96 (1.99 -1.96) 50-1.57 (1.60 -1.57) 50-1.54 (1.57 -1.54) Rcryst/Rfree (%) 24.19/28.66 !20.77/26.23 !19.63 /22.53 18.20/21.52 RMSD From Ideal Values Bond Length ( †) 0.008 0.009 0.006 0.006 Bond Angle (¡) 1.161 1.188 1.118 1.149 Average B factor 29.83 40.51 25.56 26.85 Number of water molecules 702 355 297 340 PDB IDs - - 4YKO 4YKM a Values in the parenthesis refer to the last resolution shell. b M6: R111K:P39Q:R132Q:Y134F:T54V:R59Y:A32W c M8: R111K:P39Y:R132Q:Y134F:T54V:R59Y:A32Y d M11 : R111K:P39Q:R132Q:Y134F:T54V:R59Y:A32Y:F3Q e M3: R111K:P39Q:R132Q:Y134F:T54V:R59Y:A32W:F3Q !"#$! Substrate Q108K:K40L -Retinal Q108K:K40L:T51V:R58F -Retinal Space group P1 P1 a (†) 29.479 30.234 b (†) 36.57 35.704 c (†) 64.093 64.104 (¡) 86.54 91.284 (¡) 86.601 90.912 (¡) 64.827 114.19 Molecules per Asymmetric Unit 2 2 Total reflection 171006 136653 Unique Reflection 47535 36188 Completeness (%) 95 (92.9) a 96.1 (94) Average I/ 43.69 (3.72) 31.67 (2.17) Rmerge (%) 4.5 (29.2) 5 (46.8) Resolution (†) (Last Shell) 50-1.40 (1.42 -1.40) 50-1.54 (1.57 -1.54) Rcryst/Rfree (%) 20.74/23.78 !19.34/22.88 !RMSD from ideal values Bond Angle 1.096 1.081 Bond Length ( †) 0.007 0.007 Average B factor 27.09 23.76 Number of water molecules 308 237 PDB IDs 4RUU - Table VI -7. The hCRBPII crystal table. a Values in the parenthesis refer to the last resolution shell. !"#$!Table VI -7 (con tÕd) Substrate Q108K:K40L:T51V: R58Y:Y19W -Retinal Q108K:K40L:T51V: R58Y:Y19W:Q38L -Retinal Space group P1 P1 a (†) 29.953 29.97 b (†) 35.721 35.722 c (†) 64.937 64.455 (¡) 90.846 90.866 (¡) 88.451 91.328 (¡) 65.68 114.36 Molecules per Asymmetric Unit 2 2 Total reflection 184981 160420 Unique Reflection 59171 43022 Completeness (%) 94.6 (89.2) a 95.5 (93) Average I/ 38.98 (1.9) 37.26 (3.14) Rmerge (%) 3.6 (41.4) 5.3 (46.5) Resolution (†) (Last Shell) 50-1.31 (1.33 -1.31) 50-1.5 (1.53 -1.50) Rcryst/Rfree (%) 20.17/23.58 !20.33/25.32 !RMSD from ideal values Bond Angle 1.138 1.158 Bond Length ( †) 0.007 0.007 Average B factor 25.30 26.22 Number of water molecules 347 219 PDB IDs - - a Values in the parenthesis refer to the last resolution shell. !"#"!Table VI -7 (con tÕd) Substrate Q108K:T51D Q108K:K40L:T51V Space group P212121 P1 a (†) 66.199 29.898 b (†) 73.792 36.323 c (†) 353.4 63.987 (¡) 90 87.17 (¡) 90 92.480 (¡) 90 115.28 Molecules per Asymmetric Unit 12 2 Total reflection 527001 180765 Unique Reflection 108095 48161 Completeness (%) 97.2 (98.8) a 95.5 (92.7) Average I/ 22 (1.92) 37.85 (1.84) Rmerge (%) 7.5 (72.3) 4.5 (58.1) Resolution (†) (Last Shell) 50-2.06 (2.10 -2.06) 50-1.40 (1.42 -1.40) Rcryst/Rfree (%) 22.23/27.35 !21.70/25.54 !RMSD from ideal values Bond Angle 1.139 1.157 Bond Length ( †) 0.009 0.007 Average B factor 40.00 23.76 Number of water molecules 652 319 PDB IDs - - a Values in the parenthesis refer to the last resolution shell. !"#$! Table VI -7 (con tÕd) Substrate Q108K:K40L:T51V:T53C:R58W: T29L:Y19W:Q4H -Julolidene Q108K:K40L:T51V: R58F -Julolidene Space group P1 P1 a (†) 29.941 29.787 b (†) 36.14 35.947 c (†) 64.502 64.089 (¡) 91.78 90.506 (¡) 88.209 92.446 (¡) 115.116 113.305 Molecules per Asymmetric Unit 2 2 Total reflection 59395 150975 Unique Reflection 21865 45939 Completeness (%) 92.2 (79) a 94.9 (79.4) Average I/ 38.65 (8.38) 42.2 (4.11) Rmerge (%) 4.5 (15) 4.5 (19.4) Resolution (†) (Last Shell) 50-1.82 (1.85 -1.82) 50-1.42 (1.44 -1.42) Rcryst/Rfree (%) 21.83/28.54 !22.19/25.63 !RMSD from ideal values Bond Angle 1.203 1.144 Bond Length ( †) 0.007 0.007 Average B factor 25.30 23.76 Number of water molecules 137 194 PDB IDs - - a Values in the parenthesis refer to the last resolution shell. !"#$!VI. 11 CD-270 Synthesis !!VI.11 -1 C-H activation reaction An oven dried flask was purged with nitrogen gas. B 2Pin 2 1.9 g (7.5 mmol) and then Catalyst 66 mg (0.1 mmol, Yellow solid) were added and dissolved in 5 ml dry cyclohexane. Small amount of HBPin was added to accelerate the borylation of the catalyst. The dtbpy ligand 0.2 mmol (54 mg) was added and the change of the color of the solution to bloody red was tracked. The mixture was stirred for 5 minutes and then the solid substrate 0.941 g (5 mmol) was directly added to the flask and stirred for 48h in reflux at 60 ¡C. The product was purified from the starting mat erial left over by column chromatography. Spectral data: 1HNMR (300MHz, CDCl 3): [PPM] = 1.26 (s, 6H), 1.29 (s, 6H), 1.31 (s, 12H), 1.66 (s, 4H), 7.30 (dd, J = 7.8 Hz), 7.56 (dd, J = 7.8 Hz, J = 1.2 Hz, 1H), 7.75 (d, J = 1.2 Hz, 1H). 13CNMR (125MHz, CDCl 3): [PPM] = 24.86, 31.76, 31.85, 34.2, 34.5, 35.09, 35.32, 83.5, 125.92, 131.93, 133.23, 144.08. Figure VI -2) Borylation reaction . !"#$! VI.11 -2 SNAr reaction To an oven dried Flask was added 1.03 g NaH (43 mmol) and 30 ml distilled DMF. The Flask was purged with nitrogen and cooled to 0 ¡C; 4.37 ml tBu -SH (43 m mol) was added slowly until all of the solid was dissolved (1h). In another flask the substrate was added 8.73 g in 30 ml DMF and cooled down to 0 ¡C. The c ontent of the first flask was added to the second flask slowly and the reaction mixture was allowed to warm up to room temperature and stir overnight. The product was purified by extraction in 10% HCl and diethyl ether. The organic phase was dried with MgS O4. ( With distilled DMF more than one equivalent reagent will give the disubstituted product. ) Spectral data: 1HNMR (300MHz, CDCl 3): [PPM] = 1.29 (s, 9H), 7.4 (dd, J = 5.1 Hz, J = 0.9 Hz, 1H), 7.78 (d, J = 0.9 Hz, 1H), 7.83 (d, J = 5.1 Hz, 1H). 13CNMR (125MHz, CDCl 3): [PPM] = 31.21, 48.49, 128.58, 129.75, 133.15, 138.5, 18.72, 142.32, 192.74. Figure VI -3) Nucleophilic aromatic substitution reaction. !"#$!VI.11 -3 Protection Reaction To 0.45 g (1.65 mmol) of the aldehyde from the aromatic substitution reaction was added 1.2 equivalent trimethyl orthoformate (TMOF), one equivalent MeOH and 5 mol% dried HClO 4-SiO 2 reagent. The reaction was stirred for 1h and then purified. Spectral data: 1HNMR (300MHz, CDCl 3): [PPM] = 1.3 (s, 9H), 3.33 (s, 6H), 5.86 (s, 1H), 7.5 (d, J = 0.6 Hz, 2H), 7.69 1H (d, J = 0.6 Hz). 13CNMR (125MHz, CDCl 3): [PPM] = 31.13, 47.68, 54.07, 101.93, 121.75, 128.71, 131.9, 133.99, 140.26, 141.79 Figure VI -4) CD-270 synthesis: protection reaction of the aldehyde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`,?['A@!=R!G;,55!]FR!H!.,Z5'6!V!/01"12!a+'5-4+(!'67! 7+X+*,@A+65!,J!.,,5%! /2$%"3+57$%&&) :+"> !CC9D#C EY1"% !