7777 77777 77777 7777 7777777 77777777777 "'7 L ., ‘ 3 29 0069 47554 This is to certify that the dissertation entitled STRUCTURE FUNCTION RELATIONSHIPS OF METALLOPORPHYRINS presented by Brian Ward has been accepted towards fulfillment of the requirements for Ph. D . degree in Chemisczy (‘44 /< aw Major professor d Date November 7, 1983 MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771 MSU LlBRARlES n RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. JUL 2 7193-2 STRUCTURE FUNCTION RELATIONSHIPS 0F METALLOPORPHYRINS BY Brian Ward A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Chemistry 1983 ABSTRACT STRUCTURE FUNCTION RELATIONSHIPS OF METALLOPORPHYRINS BY Brian Ward The research presented here endeavors to shine light on specific questions in functional porphyrinoid biochemistry. There are three main topics of investigation: Chromophore Selection: Comparative studies on C0 kinetics and equilibrium constants have been carried out for iron porphyrins, chlorins and isobacteriochlorins. These studies in conjunction with 13 C NMR, N-methyl imidazole binding, autoxidation of Fe-O2 and nitrite binding have lead to the conclusion that the iron hydroporphyrins are electron rich, and thus should be superior to iron porphyrin for the catalysis of substrate reductions. Ligand Specificity; CO and O2 binding to hemes with a steric encumbrance was studied (Chapter 2). The results indicate that if a steric effect can differentiate CO and 02 it does so by association rate constant modulation. Studies of CO and O2 bindings to Mb reconstituted with hemes lacking peripheral side groups (Chapter 3) suggest that peripheral methyl and vinyl groups play a cooperative role in orienting the heme in the protein and maintaining protein integrity. Synthetic hemes equipped with substituents of varying polarity on the ligand binding side were shown to have 02 Ward, Brian association and dissociation rates which correlated with a quadratic equation in the dipole moment of the local group (Chapter 4). Hydrogen bonding to oxy-heme was shown to affect only 02 dissociation. CO kinetics did not correlate with distal polarity, but rather with the size of the distal substituent. Chlorophyll optical shifts: The Optical spectra of protonated Schiff's base porphyrin, chlorin and bacterio- chlorin were characterized (Chapter 5). Related derivatives, solvent and counterion effects in conjunction with NMR and resonance Raman spectroscopies indicate that the spectral shifts associated with Schiff's base protonation are due to a combination of molecular symmetry and peripheral electron withdrawing effects. To My Family and Friends -11.. ACKNOWLEDGEMENTS I would like to thank Professor C.K. Chang for his encouraging attitude which allowed me to Pursue the work presented here. I am also very grateful to R. Young, C.B. Wang, S. Ebina and C.K. Chang for supplying synthetic materials and P.M. Callahan and G.T. Babcock for resonance Raman data and interpretation. I would also like to thank all of the above and M.A. Meador and P.J. Wagner for helpful discussions and friendship. Appreciation is extended to the National Science Foundation, United States Department of Agriculture, Amway Corporation and Dow Chemical Company for financial support in the form of research assistantships. -iii- TABLE OF CONTENTS PAGE LIST OF TABLES. . . . . . . . . . . . . . . . . . . . . vii LIST OF FIGURES. . . . . . . . . . . . . . . . . . . . . ix PART A CHAPTER 1 - COORDINATION REACTIONS OF IRON HYDROPORPHYRINS. . . . . . . . . . . . . . . 1 Introduction. . . . . . . . . . . . . . . . . . . . . 1 Results and Discussion. . . . . . . . . . . . . . . . 3 Four Coordinate Hemes. . . . . . . . . . . . . . . . . 3 Five Coordinate Chelated Hemes. . . . . . . . . . . . 12 Imidazole Binding. . . . . . . . . . . . . . . . . . .20 Oxygen Binding. . . . . . . . . . . . . . . . . . . . 24 Nitrite Binding. . . . . . . . . . . . . . . . . . . .26 Conclusion. . . . . . . . . . . . . . . . . . . . . . 35 Materials and Methods. . . . . . . . . . . . . . . . .35 Materials. . . . . . . . . . . . . . . . . . . . . .35 Reduction of Fe(III) hemes. . . . . . . . . . . . . 36 Kinetic and Equilibrium Measurements. . . . . . . . 36 IR. . . . . . . . . . . . . . . . . . . . . . . . . 38 Carbon-13 NMR. . . . . . . . . . . . . . . . . . . .40 Cyclic Voltammetry. . . . . . . . . . . . . . . . . 40 CHAPTER 1: SUPPLEMENT - A CONVENIENT PHOTOCHEMICAL METHOD FOR REDUCTION OF FERRIC HEMES. . . . . 42 Introduction. . . . . . . . . . . . . . . . . . . . . 42 Results and Discussion. . . . . . . . . . . . . . . . 42 Materials and Methods. . . . . . . . . . . . . . . . .50 Solvents. . . . . . . . . . . . . . . . . . . . . . 50 Hemes. . . . . . . . . . . . . . . . . . . . . . . .50 Photoreduction in Non-Aqueous Systems. . . . . . . .50 Photoreduction in Aqueous Systems. . . . . . . . . .51 PART B -— ENVIRONMENTAL INFLUENCES ON co AND 02 BINDING TO HEME, CHAPTER 2 - KINETICS OF CO AND 02 BINDING TO IRON- COPPER COFACIAL DIPORPHYRINS AND STRAPPED HEMES. . . . . . . . . . . . . . . 52 -iv- Introduction. . . . . . . . . . . . . . . . . . . smary O O I C O O O O O O O 0 O O O O O O O O 0 Materials and Methods. . . . . . . . . . . . . . . CHAPTER 3 - KINETIC STUDY OF CO AND 02 BINDING TO HORSE HEART MYOGLOBIN RECONSTITUTED WITH SYNTHETIC HEMES LACKING METHYL AND VINYL SIDE CHAINS. . . . . . . . . . . Introduction. . . . . . . . . . . . . . . . . . . . Results and Discussion. . . . . . . . . . . . . . . Conclusion. . . . . . . . . . . . . . . . . . . . . Materials and Methods. . . . . . . . . . . . . . . Myoglobins. . . . . . . . . . . . . . . . . . . . Kinetic Measurements. . . . . . . . . . . . . . . pK3 Titrations. . . . . . . . . . . . . . . . . . CHAPTER 4 - POLARITY CONTROL OVER LIGAND BINDING TO HEMOPROTEINS. KINETICS OF OXYGEN AND CARBON MONOXIDE BINDING TO HEME MODELS EQUIPPED WITH POLAR GROUPS NEAR THE COORDINATE SITE. . . . . . . . . . . . . . Introduction. .,. . . . . . . . . . . . . . . . . . Results and Discussion. . . . . . . . . . . . . . . Oxygenation Kinetics. . . . . . . . . . . . . . . Carbonylation Kinetics. . . . . . . . . . . . . . Distal Steric Effect. . . . . . . . . . . . . . . Summary. . . . . . . . . . . . . . . . . . . . . . Theoretical Section. . . . . . . . . . . . . . . . Materials and Methods. . . . . . . . . . . . . . . PART C - SPECTRAL SHIFTS UPON REVERSIBLE MODIFICATIONS OF CHO PERIPHERAL SUBSTITUENTS IN PORPHYRIN, CHLORIN AND BACTERIOCHLORIN CHAPTER 5 - A PHENOMENOLOGICAL EXPLANATION FOR THE RED SHIFT OF PROTONATED SCHIFF'S BASE. . Introduction. . . . . . . . . . . . . . . . . . . Results and Discussion. . . . . . . . . . . . . . . Environmental Effects. . . . . . . . . . . . . . Redox Potentials. . . . . . . . . . . . . . . . Summary. . . . . . . . . . . . . . . . . . . . . . Materials and Methods. . . . . . . . . . . . . . . Materials. . . . . Nickel 2, 6- di-n- penty1-4- -viny1- 8- -formyl- 1,3, 5,7——tetramethylporphine (1b).. . . . Nickel 6, 7- di- -n- pentyl- -1, 4- -diformyl- -2, 3, 5, 8-. tetramethylporphine (2b) and Nickel 2, 6- di- n-pentyl- 4, 8- -diformy1— l, 3, 5, 7- -tetramethy1- porphine (3b). . . . . . . . . . . . . . —V- PAGE .62 .63 . 68 . 69 . 8O .81 . 81 . 81 .100 .103 106 107 111 .114 .114 .116 140 .157 161 164 164 .165 166 PAGE 2,6-di-n-penty1-4-vinyl-7-hydroxyl-8- acroleinyl-l,3,5,7-tetramethylchlorin (4a) and 2,6-di-n-penty1-3,7-dihydroxy-4,8- diacroleiny1-1,3,5,7—tetramethylbacterio- chlorin (5a). . . . . . . . . . . . . . . . . . . .167 Schiff's Base Formation. . . . . . . . . . . . . . 168 Pyrrolidinium Salt (4d). . . . . . . . . . . . . . 169 Malononitrile adduct (4f). . . . . . . . . . . . . 170 Ethyl Cyanoacetate Adduct (4e). . . . . . . . . . .171 Pyrrolidine Hemiaminal. . . . . . . . . . . . . . .171 Schiff's Base Protonation/Deprotonation. . . . . . 171 Borohydride Reduction. . . . . . . . . . . . . . . 172 REFERENCES AND NOTES. . . . . . . . . . . . . . . . . . .173 -vi- LIST OF TABLES PAGE CO Binding Constants to 4-Coordinate Hemes (ZO-ZZOC) o o o o o o a o a o o o o o o c o o 0 Absorption Spectra of Hemes. . . . . . . . . . . 13C Chemical Shifts of CO and VCO of CO-Hemes. . 13 CO Binding Constants to Chelated Hemes (20-22°C). . . . . . . . . . . . . . . . . . . . 14 N-Methyl Imidazole Binding Constants to Ferrous Hemes (20-22°C). . . . . . . . . . . . . 21 Formation Constants of Hemin Nitrites. . . . . . 26 Isosbestic Points for Photochemical Heme Reduction in Toluene. . . . . . . . . . . . . . . 37 Oxidation Potentials vs. SCE of Organic Free Radicals in H20. . . . . . . . . . . . . . . 47 Kinetic and.Equilibrium Constants for Binding of CO and 02 to Sterically Hindered Hemes (ZO-ZZOC) o o o o o o o o o o o o o o o a o o o o 57 VCO of Sterically Hindered Hemes. . . . . . . . . 61 Kinetic and Equilibrium Constants for CO and 02 Binding 0 o o o o o o o o o o o o o o o o o 71 Absorption Spectral Maxima of Synthetic Hemes and Myoglobins. . . . . . . . . . . . . . . . . . 75 CO and 02 Binding Constants of Diphenyl Hemes with Groups of Varying Polarity Situated Near the Ligand Binding Site (20-22°C). . . . . . 88 CO and 02 Binding Constants of Diphenyl Hemes with Remote Polar Groups (20-22°C). . . . . . . . 89 Vibrations Observed and Normal Coordinate Assignments NL(II) Porphyrin Schiff's Base Species. . . . . . . . . . . . . . . . . . . . . 133 UV-Visible Spectral Data for Protonated Schiff's Base 1c as a Function of Counterion and Solvent. . . . . . . . . . . . . . . . . . . 145 -vii- TABLE 5-3 PAGE UV-Visible Spectral Data for Protonated Schiff's Base 4c as a Function of Counterion and Solvent. . . . . . . . . . . . . . . . . . . 146 UV-Visible Spectral Data of 4d as a Function of Counterion and Solvent Composition. . . . . . 147 -Viii- LIST OF FIGURES FIGURE PAGE 1-1 Absorption spectra of FeII etioporphyrin I. Toluene -—-; THF ----. . . . . . . . . . . . . . 4 1—2 Absorption spectra of FeMeOEC (left) and FeDMeOEiBC (right). FeII (top), Fe-CO (-—-) and OC-Fe-CO (---). . . . . . . . . . . . . 8 1-3 Reaction coordinates for reaction of CO with four coordinate hemes. . . . . . . . . . . . 11 1-4 Structures of chelated chlorin and isobacteriochlorin. . . . . . . . . . . . . . .. 15 1-5 0 (upper) and n (lower) contributions to M-CO bonding. . . . . . . . . . . . . . . . . . . 18 1-6 13C NMR resonant frequency vs. “CC for N-Methyl Imidazole-CO hemes; TPP's ; B-substituted porphyrines ; Hydrohemes . Solid line traces Fe-CO in order of decreasing electron density (left to right) from 3C NMR data. Dashed curve represents a leveling of n-backbonding. (Possible) linear correlations dependent on macrocycle: TPP's -°-; OEP's ----; and hydrohemes -----. . . . . . . . . . . . . . . . . 19 1-7 0 (upper)and n (lower) donation to Fe from coordinated imidazole. . . . . . . . . . . . . . 23 1-8 Absorption spectra of FeIIMeOEC in 5% HZO/DMF containing excess N-Methyl Imidazole (-45°C); Im-Fe-Im -—-, Im-Fe-Oz -—-. The oxyheme has a half life of approximately 10 min. . . . . 25 1—9 Absorption spectra monitoring the titration of Fe IIMeOEc-Cl with tetrabutylammonium nitrite in THF. Arrows indicate spectral shifts with increased nitrite concentration. . . 27 1-10 Absorption spectra monitoring the titration of Fe II2,3-DMeOEch-Cl in THF. The arrows indicate the spectral shifts associated with FIGURE PAGE increased tetrabutylammonium nitrite concen- trations O O O O O O O O O O I O O O O O O I O O 0 28 1-11 Difference infrared spectrum (FeNOz/FeCl) of Etioheme nitrite in AgCl pellet. . . . . . . . 31 1-12 Cyclic voltammograms of Etioheme species under argon (-—-); CO (----) in THF containing 0.1 M TBAP at a scan rate of 100 mV/sec. . . . . 32 1-13 Experimental technique for generation of ferrous hemes. . . . . . . . . . . . . . . . . . 39 l-l4 13C NMR spectrum of FeIIT(F5)PP in 0.1 M N-methyl imidazole/CDC13 and excess 13CO (90%). o o o o o o o o o o o o o o o o o o o o o 41 18-1 Photoreduction of Etioheme chloride. A: 0.1 mM benzophenone/toluene; irradiation time==0, 3, 10 sec. B: 0.1 mM benzophenone/THF; irradiation time==0, 5 sec. Arrows indicate progress of reduction. . . . . . . . . . . . . . 43 IS-2 Photoreduction of FeIIIetiochlorinzO in 0.2 M pyridine/toluene containing 0.2 mM benzophenone. Irradiation time==0, 5, 10, 20 sec. . . . . . . . . . . . . . . . . . . . . . 45 13—3 Photoreduction of metmyoglobin in 0.1 M potassium phosphate buffer (pH 7.0) containing 0.008% acetophenone and 2% isopropyl alcohol. Irradiation time =0, 5, 10, 25 sec. . . . . . . . . . . . . . . . . . 46 2-1 Structural formula of sterically hindered hemes. . . . . . . . . . . . . . . . . . . . . . 54 2-2 Absorption spectra of the various forms of FeCu-du and the corresponding oscilloscope traces for regeneration of Fe-Oz (A. 0.2 ms/div.; B. 2 ms/div.) and Fe-CO (C. 23 /division). . . . . . . . . . . . . . . . . . . . 56 2-3 I.R. Cell for anaerobic generation of FeII-CO porphyrins. . . . . . . . . . . . . . . . 65 2-4 Infrared spectrum of FeIISP-l3(CO) in 0.1 M N-methyl imidazole/CHZClz. . . . . . . . . . . . 67 3-1 Structural formula of synthetic hemes used for myoglobin reconstitutions. . . . . . . . . . 70 3-2 Optical spectra of "stripped" heme myoglobin; Oxy (—- -—), Deoxy (---), CO(--), in 0.1 M (pH 7.0) potassium phosphate buffer. . . . 76 -X- FIGURE 5-1 Autoxidation of reconstituted myoglobins. . . . . pH titration of "bald" porphyrin in 2.5% sodium dodecyl sulphate. . . . . . . . . . . . . Structural formula of diphenyl hemes. . . . . . . Correlation between2 1nk' calc and lnk'obs. 1nk' 0. 0925112 -0. 745ug+18. 0. CorreIation coeffigient: 0.994. Slope:l. 06. . . Correlation between lnkc a and ln Robs: lnkcalc =-0. 01481192 -0. 261115" "".+9 73 Correlation coefficient: 0.994. Slope: 0.995 (calculated without 1c,1e). . . . . . . . . . . . Relative orientation of FeOz dipole and dipole of a 3,5 disubstituted benzamide (A). Scale drawing of the relative orienta— tions and distances of an Fe02 dipole and an unconstrained o-phenyl amide dipole (B). . . . Schematic representation of a proposed simplified reaction coordinate of heme oxygenation. Hypothetical unperturbed coordinate (---), plus an interacting dipole (a...) and hydrogen bonded oxyheme complex (b -—-). . . . . . . . . . . . . . . . . . The change in dipole orientation upon introduction of a tight strap across the heme face. (-——) unconstrained o-phenyl amide. (---) sterically encumbered model. . . Structures and substitution patterns of formyl porphyrins (1-3), acroleinyl chlorin (4) and bacteriochlorin (5). . . . . . . . . . . Spectral shifts associated with protonation and deprotonation of Schiff's Base 1c in CH2C12. The arrows indicate the direction of change of the absorption spectrum upon dropwise addition of 70% HClOE-saturated CH2C12. The dashed spectra are those obtained upon addition of Et3N. . . . . . . . . . Absorption spectra of Schiff's base 4c protonated with HF vapor in CH2C12, and conversion of 4c-HF to 4c~HBFu (inset). . . . . . Absorption spectra of bacteriochlorin Sb (inset), 5c (-——) and 5c (CF3C02H)2 (---) in THF. . . . . . . . . . . . . . . . . . -xi- PAGE 77 78 87 91 92 95 101 104 118 121 122 124 FIGURE 5-6 5-8 5-9 5-10 5-11 5-12 5-13 A): Absorption spectra monitoring the reaction of 4b with pyrrolidine°HCqu in THF (total elasped time7~l hr). B): Absorption spectra of ethyl cyanoacetate adduct 4e (---) and malononitrile adduct 4f (-——) in THF. . . . . . . . . . . . . . . . . . Absorption spectra monitoring the reaction of 4b with excess pyrrolidine in CH2C12 (R2==C4H3). Reduction of 4b with tetrabutyl- ammonium borohydride in CH2C12 (inset). . 250 MHz NMR spectra of 1c plus hydrogen chloride in CDClg. Free base, intermediate (0.5 eq. HC1) and complete protonation (1.2 eq. HC1) from bottom to top respectively. Resonance Raman spectra in CH2C12 with 406.7 nm laser excitation of 1c (top). lc-HCl (bottom) and lc-DCl (middle 1550- 1700 cm' ). . . . . . . . . . . . . . . . Resonance Raman spectra of 4c-HC1 (CHZClz and THF) and pyrrolidinium perchlorate adduct 4d (THF) with 406.7 and 488.0 nm laser excitation. . . . . . . . . . . . Perchloric acid titrations (70% in CHZClz) of di-Schiff's bases 2c and 3c. Upper left and right are the first protonation step of 2c and 3c respectively. Lower spectra are the second protonation step. . . . Counter anion and solvent dependence of the absorption spectrum of SBH+lc. A): lc-HCL (-——) and lc-HClOu (---); B): lc-HC1OE in THF (——-) and CH2C12 (---). . . . . . . . Solvent and counterion dependence of the spectrum of SBH+4c. A): 4c-HC1 (--r) and 4C'HClOu (-—-) in CH2C12. B): 4C‘HClOu in THF (---) and CH2C12 (-——). . . . . . Solvent and counterion dependence of the spectrum of 4d-x' (RQi=CuH8). A): 4d°Cl’ (---) and 4d-c101.‘ (—) in CHZClz. B): 4d-Clou' in THF (---) and CH2C12 (-—-). . Absorption spectra of 4b in THF (-—-) and CH2C12 (---). Inset shows the spectral shifts observed upon dropwise addition of pyridine to 4b in CH2C12. . . . . . . . . -xii- PAGE 126 127 130 131 136 139 142 143 144 151 FIGURE PAGE 5-15 Soret region resonance Raman spectra of aldehyde 4b and Schiff's base 4c in CHZClz and THF with 406.7 nm laser excitation. . . . . . 154 5-16 Absorption spectra of n-butyl Schiff's base 4c in THF (-——) and CH2C12 (---). Conversion of 4b to 4c in THF with excess n-butylamine and catalytic amounts of HCl (inset). . . . . . . 155 5-17 Cyclic voltammogram of 4b in THF containing 0.1 M tetrabutylammonium perchlorate (TBAP). Scan rate was 100 mV/sec. . . . . . . . . . . . . 158 5-18 Cyclic voltammograms of ethyl cyanoacetate adduct 4e in THF containing 0.1 M TBAP at a scan rate of 100 mV/sec. Scan direction was reversed after first, second and third reduction wave (top to bottom respectively). . . 159 5-19 Cyclic voltammograms of malononitrile adduct 4f in THF containing 0.1 M TBAP (100 mV/sec). Scan direction was reversed after first (top). second (middle) and third (bottom) reduction wave. . . . . . . . . . 160 5-20 Half-wave redox potentials of aldehyde 4b, ethyl cyanoacetate adduct 4e and malononitrile adduct 4f (measured in THF vs. SCE). . . . . . . . . . . . . . . . . . . . . 162 R-l Infrared spectra monitoring Schiff's base formation between cinnamaldehyde and n- butylamine in CHZClz. Arrows indicate spectral changes with time. . . . . . . . . . . . 185 PART A CHAPTER 1 COORDINATION REACTIONS OF IRON HYDROPORPHYRINS Introduction There are a number of heme proteins which do not utilize the ubiquitous protoheme as prosthetic group, and as such, the supposition is that nature chooses different hemes which are better than protoheme at providing a specific function. Among these are: cytochrome oxidase which contains two molecules of heme a per functional protein,1 sulfite and assimilatory nitrite reductases which use siroheme as the prosthetic group2 and a dissimilatory nitrite reductase which contains two heme g and two heme d; moieties. Cytochrome oxidase functions to reduce dioxygen to water,1 sulfite and assimilatory nitrite reductases reduce sulfite to H284 and nitrite to ammonia,2 while in the absence of O2 dissimilatory nitrite reductase reduces N02- to nitrous oxide or dinitrogen.3b-d In the presence of molecular oxygen dissimilitory nitrite reductase functions as an oxidase, reducing O2 to H20. In all three systems there is substantial evidence that ligand binding and activation la / C02" COZH . COZH 002" Protohemo Home 9 HO \\ COZH COZH Home 9 Sirohomo (TENTATIVE STRUCTURE) occur at the unique heme site. Since cytochrome oxidase is a much studied system further discussion of it are not presented. An obvious method of investigation to probe what seems to be an obligatory role of these unique hemes is to synthesize model macrocycles (presented elsewhere5,6 ) and study the physicochemical properties of these in comparison with porphyrin model compounds.7 Since iron chlorins (heme d analogs) and iron isobacteriochlorins (siroheme analogs) are 8 position reduced porphyrins, a better understanding of their properties should result from a comparison with porphyrins on ligand binding with various peripheral substituents. Therefore, this work centers on gig electronic effects of ligand binding to iron porphyrins and hydroporphyrins. We and others8 have encountered difficulty in reducing FeIII chlorins to the ferrous state, as a result, a novel reduction method was developed.9 This method and its implications to previous reports of hemeprotein photoreduc- tion are given as a supplement to this chapter. In view of the multielectron catalytic processes for which hydrohemes are employed in nature, investigations into the redox chemistry of these hemes is necessary and can be found elsewhere.7’10 Results and Discussion A property of hydro-hemes which possibly makes them superior to iron porphyrins for the roles they play in nature is their coordination chemistry. To investigate this, the binding parameters of CO to four coordinate and five coordinate hydro-hemes have been investigated and compared with porphyrin models containing different peripheral substituents. The merits of this approach are twofold. It provides data for these hemes which might reveal any peculiarities in their binding behavior of CO (which may infer effects on substrate binding). Also, any variations in CO binding found for these hemes when compared to porphyrins with various peripheral substituents puts a perspective on the variation. Four Coordinate Hemes: Carbon monoxide binding to four coordinate ferrous heme was studied in toluene. Evidence for a four- coordinate heme under these conditions is provided II by the optical spectrum of Fe Etioporphyrin I. As shown in Figure 1-1, the Soret region displays the splitting features typical of four-coordinate ferrous heme,ll which are not present in coordinating solvents (i.e., THF Figure l-l). This insures that CO association and dissociation occur by a direct pathway. This is of importance since in coordinating solvents (coordinating FeuEtio I Toluene 1 1 I L 1 "1"- ----— -300 400 500 600 Anm Figure 1—1. Absorption spectra of FeIIetioporphyrin I. Toluene o I THF ligands) CO binding may be controlled by competing equilibria, which complicates binding data: | hV L I L-Fe-CO -——‘ L-Fe ==== L-Fe-L 7 co ' L L L = solvent I lie-CO = Fe I go I Four coordinate ferrous heme reacts with CO in two successive steps.12 At moderate CO pressures (generally less than 200 torr) the five coordinate mono CO adduct predominates and at high pressures the his adduct, according to: CO CO Fe = Fe-CO 1-——- OC-Fe-CO The kinetic and affinity constants for the first step and the affinity constant for the second step are contained in Table 1-1. To insure the observed CO recombination rates for all compounds were indeed due to mono carbonyl heme formation, CO pressures were used which should lead to little or no bis adduct. The reaction was monitored in the Soret region of the mono adduct (where mono and his carbonyl heme are spectrally distinct, Table 1-2 and Figure 1-2). Formation of Fe(CO)2 would have resulted in biphasic kinetic behavior. A slow phase was observed for Fe 2,3-DMeOEiBC which was more pronounced at higher CO .vH mucouomwmm “muHEHH :ofluoouop mowsomoummm mucmumcoo oumu Hocno umufimuocsmnm cw mcauasmou mummmoooc muoz mcoflumuucoocoo OD :mfln .auflcfimwm 30H mnu on waaw uwcmucmmu “DNH mocmnmmmmu “Ema mocmumwmmU «Amoaw Hound pmumfifiummv mmfifiu N no EszcfiE m pwusmmolm, umcosaoes omm Hp oax o.m on n.o mmxmmva i. I. mlomavomoa mim.acm.s v3x >.~ moax H.e Ems cams m.m v3x H.m moax m.m mm18207dve omm m.H v3x ~.H moax m.h onmusma~o<7v.m omm n.8m m 43.. v H moax v m amzo mem ovm ~.~ oax m.H oax A.m mzo oumusma u as am umoax o m omzo 0mm: omm m.~ v3x H.H moax m.v H onum ems m.H v3x ~.H moax m.e omomz om m.o v3x H.H moaxno.ea omflmomzoum.~ Indore Annouc 1H7m. Aaum H72. x m . Nxooc m co m a .H mama a.s.AUoNNIomV mmEmm mumawpnooolv on mucmumcou UQACCwm OU .HIH OHDMB Ahvwvm.ammavhav Amvomm Aoavamm.AN©vamv 1624mm Amcqmm.ronavmmv Amavmmm.AHmvav Aaavhhm Awavmvm.A0mavav Am.mvmom Ao.mvhmm.Ammvaov Amvvhao.amhvadq Aomvhmm.Amvvmmm OUOmUO Amvmwm.Amvavmov Amavham.amomvmov Avalanm.20mmcmav Aoacaom.lmocmnv Ammvmmm.fimmavoov Acacovm..ommvamm Ammcmno.rehcmmm .mmvamm.ravcnhm 000m .mcoom AOHVOmm.AomVNHv .oacmmm Aqmcqu.rooacmnq ANHVOmm Aoncmvv.xomavamv AHHVOBm Aoavovm.ammvhav Amachom Amacomm.xmnveoq Amacmmm.xmavhmm Amncoov.rom.omm Amavomo.rho.mmm romvmao.lo¢chmm 0 HH an .mcooaoec Anacoom.xooacoav Amachom.xmoacoav Amacoam.rmmacmmv Ahvmmm.aomvmav Amvovm Amvmom.Avmvomm Amavvmm AHvaom.Amoav©hm Ahacmoo.xmmvmnm Aomcaom.xmmcomm Iii-I‘ll Ii mmxmmve Ede mmA6207dve N ouousoc omlv.m WEQmem H chum omomz omhmomzoum.~ o HUHHH m KM:— 81722 s K oEmm 'Ililli . d it .moEom mo muuoomm coflumuomnfi .iiii Ill i .NIH manna I l l T T l l l T E . . w MeOEC DIMeOEIBC .40 «.40 ”7 Fen N0 LIGAND .1 ms —60 d-30 1 m9 0 620 ..20 F4 x2 T .20 1.10 —7 l I L l l_ l U T I T I T I U M? 'l I l q l l I - l ‘ \ L \ l L l ‘ l 1 I ‘ 400 500 600 700 400 500 600 700 nm Figure 1-2. Absorption spectra of FeMeOEC (left) and FeDMeOEiBC (right). ( ) and OC-Fe-CO (top), Fe-CO (----). concentrations. The observed rates of CO recombination were linear with CO concentration. As shown in Table 1-1 variations in P$iCO are primarily the result of association rate constant modulation. In coordinating solvents Smith13a and Sono et al.13b found that 2' decreased with increasing electron withdrawing capability of the peripheral substituent (2,4- diformal7<2(4)-formal-4(2)-vinyl7 deutero > proto > etio =meso) . Since the rate constants measured by the above authors are approximately an order of magnitude less than those in toluene indicates that CO binding in coordinating solvents is controlled by solvent coordination processes. It is not known which of the equilibria is of major importance. The trend that CO association rates increase with increased electron withdrawing capability of porphyrin 8 substituents is reversed for the porphyrin-chlorin- isobacteriochlorin series. The essential difference between these two series of compounds is that the substituted porphyrin's electronic effect is attributable to resonance while that of pyrrole saturation is inductive. Since the phenyl ring of tetraphenyl porphyrins is approximately perpendicular to the porphyrin plane, substituents on the phenyl ring should primarily exert an inductive electronic effect. Indeed, CO association rates increase with increasing electron donating ability of phenyl 10 substitution (T(p-OMe)PP>’TPP>'T(F5)PP). These effects are summarized in Figure l-3. The trends observed for first CO binding do not apply to Fe(CO)2 formation. For the porphyrin-chlorin- isobacteriochlorin series the second CO binds more strongly (affinity constant decreases) with pyrrole saturation. The TPP series shows the opposite behavior (T(F5)PP>'TPP>’T(p-OMe)PP) and the B-substituted hemes show little or no effect. These results indicate either the trends observed for Fe-CO are coincidental or Fe(CO)2 formation is dependent on a trans effect which is peculiar to the type of macrocycle. Further discussion of this is deferred to the section on imidazole binding. Table 1-2 contains the infrared stretching frequencies of mono and his adducts of the above hemes in CHZClZ. The frequencies assigned to mono and his CO complexes were shown by the dependence of the relative intensity of their absorption bands as a function of CO pressure. At low CO pressures a peak assigned to Fe-CO appeared at ca. 1950-1970 cm-l. At higher CO pressures the monocarbonyl heme peak was replaced by an Fe(CO)2 vibration at 2010- 2040 cm-1, in agreement with previously reported frequencies.14 The stretching frequency observed for CO bound heme results from populating COn* orbitals with metal dn 1 electrons reducing v from the unbound value of 2143 cm- . CO vCO of the his adduct is higher than FeCO due to 11 FoCO* Foco* sumnmu sumflmu comma ooumuc Fe+-CO FoCO FeCO Inducuvo Resonance Figure 1-3. Reaction coordinates for reaction of CO with four coordinate hemes. 12 competition for metal dfl electrons. If competition for metal dn electrons were independent of the ring system, then AvCO should be constant throughout the 3 series. There is little variation (:3 cm-l) in Av except for CO the TPP's. Previous work has revealed little correlation between VCO and CO binding data.15 However, AVCO for the TPP's stands alone, suggesting that this system may be subject to an electronic or trans effect not present in the others. Five Coordinate Chelated Hemes CO binding was studied in 2.5% aqueous myristyltri- methylammonium bromide (MTAB) and benzene at ambient temper- ature. Table 1-4 contains the binding parameters for CO binding to N-alkylimidazole chelated isobacteriochlorin (iBC), chlorin (Chl) and 8-substituted hemes. (SeeFflgu 1-4 for structures of Chl and iBC.) Variations in the kinetic and affinity constants are relatively small for the chelated relative to the four coordinate hemes. In fact, in benzene 2' for B-substituted hemes is within experimental 7 M-1 5-1.150 error of 1.1.X10 The association rate under similar conditions for the chelated chlorin is 7 M"1 s-l. These results indicate that occupation 2.0 Xl0 of the fifth coordination site by imidazole.far outweigh effects caused by substituents or ring saturation. This is evidenced by comparing the data of four-coordinate 13 noo.ooc «can can 106.771 mean 00 > HHDE nonsz :Hw “canonomcn .va mononmwomm “mmn ouconowom u nanommoa “mno mo\mno~mcn2n nanumzuz z n.oa uanoooEnonensn nerumz7z z n.oe me anon mean aaan m.enaa aanmmve a.wao ~77 a77 mean a.eaov may as omom oaan moan m.oaoa aa18207ave mo anom mman aman ua.naaa onmuzmomonua.a so «now aaan «man ncmaann mzo xnaa no anon aaan mman a.mnnv n ennm nm anom naan oman e.aanv omomz nm mnoa nman eman mama omnmowzaum.~ .mmqmm. oo.77 .mmqam. oo.sn :Onbmann \ use: nlummw nmnmmnI Alummn nmze.nmn are noovumnoovpn n oo> oouznpn .noo> oo o o a 8 mn 77 ,7 oo 77. .mmsmmnoo no > can 00 mo muwnnm HmOHEwnu U . l m M ma m H HQ 9 14 Table 1-4. CO Binding Constants to Chelated Hemes (20-2200).“ Chelated 1' 1 0 C0 -1 -1 -1 3 Heme (M s ) (s ) (torr) chb 3.0><107 0.03d 0.00065 Chlb 2.0><107 0.025d 0.00093 Mesoc 8.2 x106 0.0142 0.0013 Protoc 3.6x106 0.0058 0.0010 2,4-Diacety1c 5.6 x106 0.00859 0.0010 “2% aqueous MTAB. bThis work, pH 7.0 potaSsium phosphate (0.1 M). cReference 15c, pH 7.3 potassium phosphate (0.05—0.l M). dCalculated from L==z'/2. 2. Measured by stopped flow. 15 Figure 1-4. Structures of chelated chlorin and isobacteriochlorin. 16 (Table 1-1) to that in Table 1—4. Five coordinate imidazole hemes bind CO lO-lOO times slower than four coordinate, while in general the gig effect allows for less than an order of magnitude variation. Similarly, CO dissociation decreases ca. 106 for the imidazole chelates vs. the vacant Egags ligand CO complexes; which displayed little variation. Therefore, the presence of imidazole in the fifth coordination site has the effect of dampening out any electronic effect on C0 binding. The CO stretching frequencies of N-methyl imidazole-CO hemes are given in Table 1-3. As with CO binding to four- coordinate heme, there appears little correlation between 0 and CO binding behavior.15 According to the synergistic CO interpretation of v on the binding nature of M-CO CO complexes the extent of n backbonding parallels the Fe-C bond strength predicting the order: Etio=zprotoz; MeOEC ~2 ,3-DMeOEiBC3 2 , 4-Diacetyldeutero > T (p-OMe) PP > TPP>»T(F5)PP. Unfortunately, it is not possible to study CO binding to Im-FeTPP's, since covalently attached imidazoles are not available with these hemes and addition of external base results in competion between CO and base. 13 The chemical shift of a C nucleus is approximated by the sum of a diamagnetic and paramagnetic screening 16 . . . . . tensor. For d1amagnet1c screening an increase in electron density causes an upfield shift, while for paramagnetic screening increased electron density causes a downfield shift. In general, for transition metal 17 carbonyl complexes it has been found that electron donation to the metal results in a downfield shift of the carbonyl 17 carbon resonance. Thus, from theoretical and 16 experimental considerations paramagnetic screening dominates. Table 1-3 contains the 13C NMR resonant frequencies relative to TMS for N-methyl imidazole l3CO hemes. If as with other metal carbonyl complexes the paramagnetic screening tensor dominates, then the electron richness of the carbonyl carbon increases in the order: T(F5)PP <2,4-diacetyldeutero ’TPP>’T(F5)PP). Sigma donation from the first CO then increases the electron donating ability of the iron and allows the second CO to bind more easily (T(F5)PP>'TPP>’T(p-0Me)PP). The inherent electron 23 Figure 1—7. 0 (upper) and n (lower) donation to Fe from coordinated imidazole. 24 richness of hydrohemes allow the second CO to bind more easily. It would seem that octa-alkyl type hemes are electronically neutral. That is, Fe(CO)2 formation is not altered by either gig electronic effects or Egagg CO effects. Oxygen Binding Oxygen binding to unprotected hemes can be conveniently studied by kinetic CO replacement21 (below) or at low temperatures.22 hv 2 B-F II-CO n——— B-FeII ==== B-FeII-O2 CO -02 Reliable O2 binding constants have not been obtained for either the chelated chlorin or iBC. Flashing off CO in the presence of O2 resulted in facile oxidation of the heme. The hemes were found to be completely oxidized within a relatively few flashes. Thus, it was not possible to determine whether the observed kinetic traces were due to O2 binding or oxidation. At -45°C in 5% HZO/DMF, N-methyl imidazole FeII-O2 porphyrins are almost indefinitely stable.22 Under the same conditions oxy MeOEC had a half life of approximately 10 minutes (spectrum of MeIm-Fe-OZMeOEC, Figure 1-8). Autoxidation of oxy-iBC was so rapid that attempts at measuring the optical spectrum were unsuccessful. These 25 l T l— V T l T T T 2 E_ n m MeOEC BOI -45°DMF ‘ 60~ « 40- - "’ '7 20-- ~ Mnm) 400 500 800 Figure 1-8. Absorption spectra of FeIIMeOEC in 5% HZO/DMF containing excess N-Methyl Imidazole (-45°C); Im-Fe-Im ----, Im-Fe-02 The oxyheme has a half life of approximately 10 min. 26 results are not totally surprising since stability of oxy-heme decreases as the macrocycle becomes electron rich.23 Nitrite Binding Of principal importance in assessing the obligatory role of hydrohemes in denitrification reactions is their comparative nitrite binding. Table 1-6 contains the formation constants of nitrite ferric hemes in THF. Table 1-6. Formation Constants of Hemin Nitrites.a'b Ferric Heme ‘gf (m'l) Etio 770 MeOEC 33o 2,3-DMeOEiBC 1600 “Reaction of hemin chlorideanuitetrabutylammonium nitrite. bIn THF. Kf FeIIIC1+nBu4NNO2 '——“- FeIII(N02) Figures 1-9 and 1-10 show the optical spectral changes III I upon titration of Fe MeOEc-Cl and Fe II-2,3-DMeOEch-Cl with tetrabutylammonium nitrite. Hill plot slopes for FeIIIEtio I, MeOEC and 2,3-DMeOEiBC were 0.95, 0.95 and 0.93, respectively. Etioheme nitrite was also prepared by refluxing the hemin chloride with one equivalent of 27 Fe MeOEC Cl -—-—* Fe MeOEC NO2 Figure 1—9. Absor tion spectra monitoring the titration of Fe IIMeOEC'Cl with tetrabutylammonium nitrite in THF. Arrows indicate spectral shifts with increased nitrite concentration. 28 Fe DMOOEIBC Cl ——-—> Fe DMeOEIBC N02 600 X2 588 1 1 1 300 400 600 700 Anm Figure 1-10. Absorption spectra monitoring the titration of Fe II2,3-bMe0Ech-c1 in THF. The arrows indicate the spectral shifts associated with increased tetrabutylammonium nitrite concentrations. 29 ‘ AgNO2 in butyronitrile. The products from both reactions were spectrally identical. As evident from Table 1-6 displacement of chloride by nitrite increases in the order porphyrin ¢-C-R'-+R. Fe(III) Fe(II) R'==¢,CH3 RH and R as shown in Table lS-l. Heme photoreduction has previously been observed for a number of proteins. Among these are: cytochrome oxidase,34 horse heart cytochrome 3,35 cytochrome 9,36 T-state hemoglobin37 and cytochrome 3552.38 Furthermore, heme photoreduction has purposefully been demonstrated with cyt P450 using a proflavin/EDTA system39a and cyt 3 using flavin/EDTA.39b The reduction presumably involved the prior photoreduction of a chromophore (e.g. flavin) 34b,38,40 followed by heme reduction. In fact cyt b and cyt 9552 do not photoreduce in the absence of flavin.38'40 Therefore, the mechanism of hemoprotein photoreduction apparently is akin to scheme 1. One common feature to all of the protein photoreductions, as well as to our system, is a wavelength dependence. For instance, cytochrome g oxidase is not photoreduced with visible 34 excitation41 but is reduced in the UV region, cytochrome g photoreduces 8 times faster at 410 nm than at 535 nm,35 49 as well T-state Hb displays a photoreductive wavelength dependence.37 In our system, irradiation through pyrex vs. quartz results in a very sluggish reduction, attributable to the low absorptivity above 310 nm of benzophenone ‘ or acetophenone. It is also the experience of many workers in the field that ferric hemes in pyridine under CO atmosphere can be photoreduced to the carbon monoxide complex. We have observed, however, that the rate of this reaction is' related to pyridine purity. Using highly purified pyridine this photoreduction, in fact, did not proceed to completion; but addition of trace amounts of 02CO/i-PrOH resulted in completed reduction. Therefore, it seems there are impurity chromophores present in commercial grade pyridine which are responsible for initiating the photoreduction. In conclusion, we have develOped a simple and reliable technique for reducing ferric heme under a variety of conditions. This method is especially suitable for studies carried out in organic solvents as well as aqueous protein work. One advantage of this method is that tedius reductant titration may be avoided when excess reducing agents may be deleterious (i.e. 0 binding). The fact that 2 the ketyl radicals are more powerful reducing agents than dithionite also suggests that some hemoproteins, e.g. cytochrome 93, which cannot be completely reduced by dithionite may do so with this technique. 50 Materials and Methods solvents: Toluene was purified by stirring with several changes of concentrated sulfuric acid followed by distillation from lithium aluminum hydride. Tetrahydrofuran (THF) was distilled from lithium aluminum hydride. Pyridine was distilled from calcium hydride. Ewmes: Etioporphyrin and etiochlorin were prepared by standard procedures.27b Iron insertion was accomplished by the FeSO4 method. Horse heart myoglobin (Sigma, type III) was used as received. Photoreduction in Hen-Aqueous Systems: Ferric hemes either in the chloride or u-oxo dimer form (~10‘5 M) and benzophenone (~10.4 M) were dissolved in dry toluene, THF, or 0.2 M pyridine/toluene mixture. The solution was rendered oxygen-free by either flushing with argon gas for 30 minutes using a syringe needle or by freeze-thaw cycles under vacuum. The deoxygenated solution was then irradiated with a Hg lamp (Hanovia "Utility" UV lamp, 140 W). In general, reduction was complete within 20 5. Spectral changes were monitored using a Cary 219 spectrophotometer. In all experiments 51 further irradiation after the heme reduction was complete resulted in 29 spectral change. Photoreduction in Aqueous system: Myoglobin was dissolved in 0.1 M potassium phosphate buffer (pH 7.0) containing 2% isopropanol and 0.008% acetophenone. The solution was degassed and irradiated as above. PART B ENVIRONMENTAL INFLUENCES ON CO AND 02 BINDING TO HEME CHAPTER 2 KINETICS OF co AND 02 BINDING TO IRON-COPPER COFACIAL DIPORPHYRINS AND STRAPPED HEMES Introduction X-ray crystallography showed that the structures of carbon monoxide liganded hemoglobins (Hb) and myoglobins (Mb) exhibit a bent or tilted FeCO linkage with respect 42-46 to the porphyrin ring, whereas in heme model compounds the FeCO bond is linear and perpendicular to the heme plane.47’48 The origin of the distorted configuration in the proteins is attributed primarily to nonbonding steric interactions of the axial ligand with residues at the distal side. An assumption is made that ligands such as 02 and NO, which preferentially form bent complexes, should encounter less steric hindrance when bound in the heme pocket.49'50 It has been proposed that in Hb and Mb, the distal steric effect would decrease the affinity ratio of CO vs. 02, and is responsible for the detoxification of 51-54 . A compar1son CO poisoning in respiratory systems. of ligand binding constants of proteins and model compounds often shows that many heme models have a larger CO vs. 02 52 53 affinity ratio (M value) than the proteins. However, such a comparison does not necessarily constitute a correlation between the distal steric effect and affinity as the ligand binding constants of heme models can be drastically altered by medium effects.28b’55 Indeed, Traylor and coworkers have shown that a five-coordinate protoheme-imidazole model binds both 02 and CO in aqueous suspensions with equilibrium and kinetic parameters almost identical to 15c,28,55,56 In other R—state isolated hemoglobin chains. cases, for example, T-state hemoglobin and notably myoglobins have very small M values which cannot be duplicated with simple heme compounds. It is therefore of importance to examine the steric effects on ligand affinity using synthetic models equipped with varying degrees of steric hindrance at the distal side. Several porphyrin models of this kind have been prepared57-60 and recently an iron complex with a bent CO has been shown.61 Here is presented the equilibria and kinetic rates of CO and O2 binding to two hindered heme systems. One is mixed metal cofacial diporphyrins in which an inert copper porphyrin62 is tightly linked to the heme thereby providing a compression from above to the coordinating ligand. The second system is iron cyclophane porphyrins where a hydrocarbon chain is strapped across one face of the heme. Depending on the chain length, the strap would mostly exert a side-way shearing strain to the gaseous ligand. 54 Fe-Cu-4 X=(CH2) FeSP-l3 n=5 Fe-Cu-S X=(CH2)2 . FeSP-l4 n=6 FeSP-lS n=7 Figure 2-1. Structural formula of sterically hindered luames. 55 CO and O2 binding to the ferrous hemes were studied in benzene solution containing excess N-alkyl imidazoles. The nitrogen base was chosen such that it can only form five-coordinate heme. N-Methylimidazole was bulky enough to meet this criterium only with the very tightly gapped Cu-Fe-4 and FeSP-13 but was not satisfactory for Cu-Fe-S nor FeSP-lS as considerable competition of CO binding at the hindered site by a second imidazole can be observed. We therefore used a "tall" bicyclic 63a imidazole for the dimers and a "fat" dicyclohexyl- imidazole63b for the strapped hemes; using these bases no competition was observed.64 Typical relaxation curves along with the corresponding absorption spectra of different complexes are shown in Figure 2-2. The CO and O2 association rates were obtained under pseudo-first order conditions. The oxyheme complex formed in the Cu-Fe dimers was so stable (no oxidation detectable even after 12 hrs at room temperature) that the Pa02 values can be measured directly by gas titration and as such, they provided an independent check on the 02 off rates 28’65 All rates derived from the kinetic equations. and equilibrium constants for models and relevant heme proteins are tablulated in Table 2—1. The most striking result shown by Table 2-1 is that indeed, distal steric hindrance can affect ligand binding but this effect is manifested only in the ligand association rate constants and has almost no effect on 56 in i l . I f ”— 'nlCO I ' , , a 7 i :lfifi . 1, ; 7 I 7 f 1 l“ i .i . i ...; I :7 1.; i l l l C l i l I i 7 3 I i , . 7 7.4 : g 7 7 3 <7 ; 5 i ; I, l I ’ i L __ -i ' o i : '3’. :.L‘+~A:‘: A‘: we 1 5 . J A ALLA hf VTV e - ABSORBANCE C) a) 02 Figure 2-2. Absorption spectra of the various forms of FeCu-d4 and the corresponding oscilloscope traces for regeneration of Fe-Oz (A. 0.2 ms/div.; B. 2 ms/div.) and Fe-CO (C.23/division). 57 .mmmm mcfionHOM :0 Hum manna HON mmuoz E E HEUD €03 35 oom 3 omm o3 x a; moo woo «3 x do 33:3 z m .o + maummmo oEmoo x83 3.: m3 x m m .o woo mod x o 83:3 z o4 + «Tampa EHmzna x33 35 NH no .o mod x o 33:3 z m .o + Summon oeHmme x33 mot omm on 3 mg x o4 moo moo «S x o .o 33:8 2 mo + musoumm 022:9 fioz 35 mod x o.m 2o 86 «S x «d 33:3 z o4 + Tsonmm EHmSIE x33 m2» o: o8 o3 mod x m .m 1o moo «3 x o.m 83:8 z m.o + Tsonmm EHmzla mooo.o omoo o3 x S mcmucmn + mswsoxmnomo 3 6m: S x o Na 85 nod x m6. ooooo moo o3 x o 2833 22035 @3220 3 .9: omv mm .o mm hS x ~.m flood goo o3 x 3 oom: $283 833:0 3 oom o A S boa x o. N o8o.o moo.o o3 x 9m oo~m megououo 8336 o o2 mo om o3 x o.m «mood Zoo o3 x o of mod omuflomfinm mmooz m R ovzom Hzmo omon honmza Zoo ooonmoo mo: mum o m n: .mmm sz AHIommv Aaum HIZV Annouv Adlommv Aalm alzv ucm>Hom pcsomaoo x x No a x .x oo o a .H 3.AUoNNIomv mmEmm thmficflm >HHm0flumum on NO can 00 mo mcwccflm Ho“ mucmumcou Edwunflawsvm can Dflumzflm .HIN mHQMB 58 .wom cflnufl3 owmumm mucmEmusmmmE Uflumcwx Eoum paw mcoflumuUHu uomuwo Scum owcwmuno Noxm .maoumoflsfiaxxmooaosoao-m.H "eHmoo .mcooflnxo-muum.HHonoAEnouoxsmuumuuo.a.o.m "eHmma .mdaz Ho m¢BU wm CH cmccmmmsm .Aommo 2 wuoax o.H .Amcmsaou .wcmucmnv z mloax numo mo anon H “AONmV z loavxmm.a .chmsaou mcmucwnv z mloavxanuoo mo when H "mmfluwawnsaom mewsoHHOM own means pmumasoamo mumz mwumms QUfiN .HIN manna How mmuoz 59 the off rates, in agreement with the predictions made earlier by Moffat et a1.49 and consistent with the isocyanide binding results of Traylor.66 Since it has previously been shown that the CO and O2 association rates are nearly independent of medium and heme electronic effects and that the 02 off rates are very much affected by the local polarity of the ligand binding site,67’68 it is futile to directly compare the Oz/CO affinity ratio, M, of different model compounds. However, when we compare only the association rate data we find, relative to chelated mesoheme, FeCu-S or FeSP-lS a CO reduction of 90—fold while an 02 of 30 (a reduction ratio of 3) and for Fe-Cu-4 a CO reduction of 400-fold with 02 being reduced by 100 (a reduction ratio of 4). This unequal reduction of CO and O2 association rates may be considered as an evidence for the steric differentiation of O2 and CO. This steric selectivity nonetheless does not explain why we cannot obtain the degree of differentiation observed for Mb, i.e., chelated protoheme or R-Hb vs. Mb has a reduction ratio of at least 5, even though our model compounds have more steric hindrance built into them than does Mb, as reflected by the CO on rates. Neither can we reconcile the fact that there is essentially no change in the on rate reduction ratio nor the M value going from Fe-CuS to FeCu-4 while the structural data as well as the CO on rates indicate clearly that the FeCu-4 has a tighter gap than FeCu-S. If the bending of CO is responsible for 60 the differentiation, it would have to show in the 4 to 5 comparison. One possibility is that the differentiation is not proportional to the steric hindrance; it reaches a maximum then decreases as the steric effect becomes too great. Unfortunately, in the present study we found it is difficult to have a system whose CO on rate is in the 5 M.1 3.1, to compare with Mb. neighborhood of 5 XIO Cofacial diporphyrins with longer linkages, e.g., FeCu-6 and FeCu-7, exhibit kinetic rates similar to FeCu-S since the two porphyrin rings have a tendency to assume a slipped conformation and maintain a tight gap, as shown by X-ray studies,69 thus these compounds offered no insight. On the other hand, hemes equipped with longer straps tend to form six-coordinate hemochromes with the excess base. Although the present study does not provide a definitive answer as to whether or not steric bulk at the ligand binding site can selectively reduce the affinity of CO vs. 02, surely the kinetic results imply that models which bind CO 2 to 3 orders of magnitude slower than Mb should decisively indicate whether there is any relation between ligand affinity and v Table 2-2 summarizes the CO' VCO of some of the synthetic compounds measured in different solvents. It is evident that the influence of medium is far greater than the steric effect. That is, the variation in vCO observed for the strapped hemes in 0.1 M N-Methyl imidazole/CH2C12 can be completely removed in neat N-Methyl imidazole. Thus, it would be dangerous to 61 Table 2-2. vCO of Sterically Hindered Hemes. Compound Medium v v 12CO l3CO (cm‘l) (cm‘l) FeCu d4 N-MeIma 1960 1915 0.1 M TImb/ 1967 1924 CHzBr2 FeSP-13 N-Melma 1962 0.1 M TImb/ 1967 CHzBr2 0.1 M N-MeIm/ 1932 1888 Ch2C12 FeSP-14 0.1 M N-MeIm/ 1939 1894 CH2C12 FeSP-lS 0.1 M N-MeIm/ 1945 1901 CH2C12 c a Heme 5 N-MeIm 1955 1910 0.1 M N-MeIm 1954 1910 CH2C12 “Neat N-Methyl imidazole. bN-(triphenylmethyl)imidazole. cIron 2,6-di-n-penty1-l,3,5,7-tetramethy1-4,8- d dimethacetamido porphine. 62 use v as an indication of CO affinity for comparison CO between different systems, i.e. comparing proteins and model systems. The unequal reduction of the CO and O2 association rates by the steric bulk implies that such differentiations must be related to the bond forming processes. Szabo70 has suggested that CO-heme transition state resembles product while 02-heme has a more reactant- like transition state. That is to say since the Fe-CO bond formation requires shorter contact, the CO molecules must be in closer proximity than 02 to attain transition state. Any steric barricade at the heme binding site therefore would hinder C0 coordination more than 02 coordination. The present study also indicates that it would be a unique synthetic challenge to prepare heme models71 that match Mb's kinetic behavior. So long as we showed that bending of CO cannot be solely responsible for the large differentiation observed in Mb, other factors such as the basicity of the proximal base, pre-equilibrium of the heme conformation inside the protein pocket, etc. have to be taken into consideration.72 The synthesis of other sterically hindered, five-coordinate hemes is underway. Summary The work presented does not provide definitive evidence either for or against a steric effect 63 differentiating CO vs. 02. It does suggest that if a steric effect can differentiate small ligands (CO, 02) it does so primarily by ligand association rate modulation. This work also shows that v is not a reliable indicator of CO ligand affinities. Materials and Methods Strapped and cofacial diporphyrins were synthesized 73 by previously described methods and characterized by UV-visible, NMR and mass spectroscopies. Iron insertions were accomplished by the ferrous sulfate method.27b Benzene was purified by stirring with several changes of conc. H2804 followed by washing with water, drying over anhydrous Na2c03 and distillation from lithium aluminum hydride. 5,6,7,8-tetrahydroimidazo[1,5-a]pyridine was prepared by hydrogenation 0f 2.33 diazaindene,63a and purified by vacuum distillation from KOH. N-methyl imidazole was vacuum distilled from CaH. 1,2 dimethyl imidazole was vacuum distilled from Na. All other reagents were of purest commercial grade and used as received. Kinetic rates were measured in benzene containing 0.1 M base by flash photolysis28 according to: hv k'(02) B—Fe 13(CO) k B-Fe-CO B-Fe-O 2 64 Under pseudo-first order conditions l'CO was monitored spectrally at 398 nm. A plot of l'CO vs. CO yielded a straight line with slope 1'. CO dissociation rates were calculated from L==l'/l, which was determined by direct titration of the ferrous heme with CO. Oxygen dissociation rates were measured directly as shown in trace A of Figure 2-2. Oxygen dissociation was calculated from K =k'/k which was determined from the Gibson equation:65 l/R = 1/k-+K(Oz)/1'(CO) where R is the observed rate of heme-CO formation in the presence of 02; l'(CO) was the pseudo-first order rate constant before introduction of 02. When possible (see results and discussion) K was measured by direct titration of the ferrous heme with 02. The results of each method were in excellent agreement. FeIIICl porphyrins were reduced to the ferroheme by addition of a slight excess of Vitride (sodium bis-(2-methoxyethoxy)aluminum hydride). Use of this reagent for heme reduction requires that solutions be absolutely dry since reaction with water can produce hydroxide, which can compete with ligand binding to ferroheme.15b Infra-red spectra were recorded on a Perkin-Elmer 283B spectrometer in a NaCl cavity cell in solvents as described II_12 in Table 2-2. Samples of Fe CO porphyrins were prepared anaerobically as shown in Figure 2-3. 13CO bound samples were prepared by exchange with 12co. 0.3-0.5 m1 92.1 atom 65 / deuclng Agent 1 b —(.o.°.°( ;\ Sample .1 ”'1— GC Septum E '1— Cavlty Cell '. ------ i Figure 2-3. I.R. cell for anaerobic generation of FeII-CO porphyrins. 66 percent 13CO was bubbled through the solution using a gas tight syringe and an outlet needle to allow excess gases to escape. As shown in Figure 2-4 this method allows 13CO to completely replace 12CO. Iron reduction was accomplished by addition of a couple of crystals of tetrabutyl ammonium borohydride to the sample before allowing it to flow into the cell. Optical spectra were recorded on a Cary 219 spectro- meter. 67 1950 1900 cm" ‘T I I l I I”‘T I I I I I FeSP-IS *- 32 I2 00 IS CO |932 I888 Figure 2-4. Infrared spectrum of FeIISP-13(CO) in 0.1 M N-methyl imidazole/CHZClZ. CHAPTER 3 KINETIC STUDY OF CO AND 02 BINDING TO HORSE HEART MYOGLOBIN RECONSTITUTED WITH SYNTHETIC HEMES LACKING METHYL AND VINYL SIDE CHAINS Introduction Variations in heme function among hemoproteins often arise from specific interactions between the heme prosthetic group and the apoprotein. An obvious method of probing the heme-protein relationship is to investigate structural modifications in the heme moiety and relate these to alterations of the functional properties of reconstituted hemoproteins. Indeed, reconstitutions with protohemes 67’73'74 and iron chloro- modified at the 2,3 positions phyllides75 have revealed significant changes in the ligand binding or oxidative stabilities of hemeglobins and myoglobins. While the variation in properties can often 67'74’75 effects, be attributed to electronic73 and/or steric precise interpretation of these results is often difficult for there is a very limited number of modified hemes capable of distinguishing between a steric and electronic 67,74 effect. As part of an effort to probe structure function relationships of oxygen binding hemoproteins, we 68 69 report here the oxygen and carbon monoxide binding behavior of horse heart myoglobin reconstituted with hemes deprived of peripheral side chains. As shown in Figure 3-1, the hemes employed were 2,4-dimethyl deuteroheme; deuteroheme; 1,3-didemethyl deuteroheme ("bald" heme); and iron-6,7- dipropionic acid porphine ("stripped" heme). These synthetic hemes all possess the two propionic acid groups necessary for specific binding to the protein heme pocket but they differ from each other in the number of ring methyl groups. In the absence of large variations in their electronic properties, this series of hemes should allow a more accurate assessment on the effect brought about by structural and shape perturbations of the heme group. Results and Discussion The basic supposition of hemoprotein reconstitution is that after reconstitution the protein tertiary structure regains its native form. Indeed, the ligand binding properties of protoheme reconstituted Mb's and Hb's are essentially identical to the native proteins.67’73-75 Table 3-1 summarizes the kinetic and equilibrium constants for CO and O2 binding to the reconstituted Mb's of this study. Comparison of our results with those of Sono et al.739 for deuteroheme-reconstituted horse heart Mb are in good agreement. As well, 2,4-dimethyl deuteroheme Mb has similar CO and O2 binding properties to mesoheme 7O HOZC 2 ,4 Dimethyl deutero Deutero 1,3 Didemethyl deutero 6,7Dipropionic acid porphine Figure 3-1. Structural formula of synthetic hemes used for myoglobin reconstitutions. 71 .mom mm.N CH o .oomm-om .o.e mo .Hmmmsn mumramora cHe mew: MH Hm 53 x o.H omo Zoo moH x mo omoo o.m eomaaHfime mo Hm eoH x 5m ov.o mHo.o ooH x m .m Hvooo m .m 9:91 eonm: Hm HH eoH x Him mmo oHo.o ooH x o.H 386 m .m mamroxmusmo mamnonmusmo o Tm eonoa oe.o «moo moneo omoo o6 erumsHo-e.m A.umcoomu USN m>Humcv 3 3 a3 x To 2.6 mmoo moH x To mmoo o6 mamsouoxa z 1 me x A m 5 .x 3.83 re A 3 H A m 5 .H 383 mm nod 233%»: H- H- H- H- H- H- o . NO nuw3 coHuomwm oo saw: coHuommm c.msH©ch No can 00 MOM mucmumcoo Eswunflawsvm can oaumcflx .H-m mHnoB 72 Mb, as obtained by the same authors.73g The results of the present study clearly indicate that the removal of peripheral methyl and vinyl groups of heme can change the ligand binding constants, particularly the carbon monoxide association of reconstituted myoglobins. As shown in Table 3-1, carbon monoxide association rates underwent a five-fold increase upon removal of peripheral methyl groups on going from 2,4—dimethyl deuteroheme Mb to the "bald" heme myoglobin, while 02 association rates, for the same MB's increased only 20%. The behavior of "stripped" heme Mb, however, does not fit any trend. That the ligand binding behavior of dealkylated heme reconstituted Mb's is not primarily caused by an electronic effect is evidenced by a comparison with the work of Sono et a1.739 These workers have shown that ligand dissociation rates (CO and 02) correlate with pk3 of the porphyrin for horse heart Mb reconstituted with protoheme derivatives modified at positions 2 and 4. Also, with the exception of deuteroheme Mb, a similar correlation exists for ligand association (k' and 1') rates-and equilibrium constants (K and L). Their results predict that k, k', l, l' and L decrease and K increases with increasing pK3 of the free base porphyrin. Since porphyrin dealkylation resulted in a slight decrease in pK3 (see Table 3-l), from an electronic perspective, changing from "stripped" heme Mb to 2,4-dimethyl deuteroheme Mb, nearly all parameters 0 (k, k', P;5 2, 1 and 1') would decrease. From Table 3-1 73 only 1 seems to follow this prediction, even so, the very small decrease in 1 does not support such a correlation without question. The inability of electronic effects to account for the anomalous behavior of deuteroheme previously nor the results in Table 3-1 argues strongly that the modified ligand binding prOperties of our reconstituted Mb's are due to effects other than electronic. Previously we15d and others60 have shown that steric crowdedness near the ligand binding site in heme models greatly modifies the ligand association rate. For myoglobin, the steric hindrance from residues at the distal side has been postulated to cause a geometric distortion of the bound CO and therefore is responsible for differentiating CO and O2 binding.49 Mb reconstitution with dealkylated hemes which have a smaller bulk than protoheme would result in a looser fit of the heme into the heme pocket. This could result in a relaxing of the steric constraints surrounding the heme group. As noted earlier, any steric effect affecting CO and O2 binding should principally be reflected in ligand association rate data. With the exception of "stripped" heme Mb, Table 3—1 reveals an apparent correlation between heme steric bulk and 1' and to a much lesser degree, k'. Thus, if one assumes that the native tertiary structure remains intact for Mb's reconstituted with 2,4-dimethyldeutero, deutero, and "bald" hemes, then these results allow for the possibility of a steric effect differentiating CO and 0 binding. 2 74 With the exception of the CO dissociation rate, the ligand binding properties of "stripped" heme Mb are anomalous when compared with the other dealkylated heme- containing myoglobins. It seems certain that the heme in the protein is still bound to the proximal imidazole since its CO dissociation is consistent with the other hemes (1 is known to be dependent on the trans ligand effectslza’15b ) and the spectral shifts upon ligation and oxidation are similar to the other myoglobins (Table 3—2 and Figure 3-2). However, the absorption spectrum of this Mb is remarkable in that all absorption maxima shift bathochromically relative to the "bald" heme Mb. In contrast, visible absorption peaks of pyridine hemochrome of both hemes are essentially identical. This can only be attributed to modulations by local environment. A 73a noticed that similar case reported by Sono et a1. Mb's reconstituted with spirographis and isospirographis hemes exhibit different absorption maxima even though the two hemes have the same spectrum outside the protein. In view of the fact that the association rates are slow in comparison with the other demethylated systems, it would appear that the lack of position 5 and 8 methyl groups would allow a protein conformational change producing a more crowded ligand binding site. To probe this heme substituent effect further, the stability of the various oxymyoglobins was examined and their autoxidation kinetics are presented in Figure 3-3. 75 .NHommo - meHonsa eH No.5 ma .wmmwsn mumramoro :He MP mmm mom Hoe mEonnOOEmn osflowuhm mom amm aoe e as oo mom mmm new a: No mvm mmv n2 axomp mmo mow mam a: H85 mam: roommfluuma mmm mom mow meonnooamn OCHpfluhm omm mmm ooe e as oo omm own How a: mo mmm oav n: axomp omm «me man as Hoe mam: =onm= Hem MHm mos «sowroosmr mcHonxa mmm mam doe e as oo mom omm Noe as No mvm qu n: mxoop mmw mmv mom a: umE mfimnoumusma mom mHm wow mEoucooEm: mchHuxm omm omm doe e as oo pom «mm mos a: No eem «me n: mxomo mmo mae mam a: pas mEosoumusmp HHerEmoHo-e.~ c5: modem mHnHmH> lees umwom msmoxnz 1| I'Ilul' '1 5 I0 I... s.m:HQono>z can mmEmm oHumnucwm mo oEmez Hmuuommm :oHumHOmQ4 .msm manna 76 Figure 3-2. Optical spectra of "stripped" heme myoglobin; Oxy (--—-), Deoxy (—-—--), CO ( )- in 0.1 M (pH 7.0) potassium phosphate buffer. 77 A L2__ “- < I S .‘H a 'f OJ!— .£ A s ‘ I A A n 0.4 __ A I I A I _ A -“II’ A ‘/-’. . . . . C /' . O . . H‘s-””“ I . I I I I I J I 2 3 ‘9 5 6 '7 8 Thus (hr) Figure 3-3. Autoxidation of reconstituted myoglobins. Protoheme slope: 0.39 ><10-3 s.1 corr. coef.: 0.996 (native) Proto _3 -l (reconst.) 0.56 XlO 5 0.995 2,4 Me2 1.25 ><10‘3 5'1 0.981 Deutero 1.04 x 10'3 s“1 0.997 "Bald" 1.56 x 10’3 5‘1 0.997 "Stripped" 2.79 x10 3 0.994 78 5 r 1 l r I r 1 s I 1 60 -< I' K3 + _ 2HBH———"' 3HB H 120 - d _ x4 - -I I I ‘ I 300 400 500 600 A nm Figure 3-4. pH titration of "bald" porphyrin in 2.5% sodium dodecyl sulphate. 79 As is apparent, the "stripped" heme Mb stands alone in that the other myoglobins (2,4-dimethy1 deutero, deutero and "bald") are approximately within experimental error of one another. The rapid oxidation of the "stripped" Mb cannot be attributed to electronic effects; on the contrary, autoxidation rate is usually slowest in acidic (electron deficient) hemes.23 In a recent communication LaMar76 has shown by 1H NMR that myoglobin reconstituted with protohemin or carbonylheme results immediately in a statistical (1:1) heterogeneous Mb mixture, which slowly equilibrates to a ratio of 12:1. The heterogeneity is believed to be from a rotational disorder of the holo- protein. The equilibration (heme reorientation) probably occurs via a heme-protein dissociation-association pathway, during which the proximal imidazole-heme iron bond must rupture. This finding provides for the possibility that one pathway of myoglobin autoxidation77 may involve heme- protein dissociation. Thus, a heme pocket conformational change could result in a faster heme-protein dissociation and hence accelerated autoxidation. In any event that the "stripped" Mb stands alone with regard to autoxidation lends support to our speculation that Mb reconstitution with this particular heme produced a holoprotein comprising a modified heme environment which is at variance with the native myoglobin. 80 Conclusion If one assumes that the protein tertiary structure of various reconstituted Mb's retains their native conforma- tion, our results suggest that there is an apparent ligand specificity associated with heme size. To a first approximation, this may be interpreted as evidence for the operation of a steric effect regulating ligand specificity since if the above assumption is true then the dealkylated hemes should fit the heme crevice in a looser manner, allowing residues near the oxygen binding site to become less constrained and thereby decreasing distal steric controls on ligand binding. However, this cannot be considered as a general rule in Mb reconstitution experiments since it is not known to what extent structural modifications of the prosthetic groups would cause functionally significant alterations on the protein. It is indeed unexpected that the removal of the 5 and 8 methyl groups should bring about a sudden turn in the trend. If the anomalous behavior of the "stripped" heme Mb is due to a protein conformational effect, it would suggest that the 5 and/or 8 methyl groups are critical for maintaining the nativeness of the protein. This result opens the question as to what effect side chain positions have on protein function. Previous reconstitution studies on b-type hemoproteins invariably centered on variations in the 2,4 positions of protoheme. The present 81 work reveals a seemingly obligatory role of the other methyl groups in determining the properties of the protein. Materials and Methods Myoglobins: Horse heart myoglobin (Sigma, type III) was used as received. Heme extraction and myoglobin reconstitution 78 2’4_ dimethyl deuteroporphyrin Ix,67 deuteroporphyrin Ix,79a 79b and porphine 6,7- were carried out using standard procedures. 1,3-didemethyl deuteroporphyrin IX, dipropionic acid79b were synthesized according to previously described methods. Iron was inserted by the ferrous sulfate method.80 For kinetic and equilibrium measurements, meth's in pH 7.0, 0.1 M potassium phosphate buffer were reduced with a minimum amount of aqueous sodium dithionite in an argon atmosphere. The previously described photochemical reduction method was employed for spectral characterization.9 Kinetic Measurements: The flash photolysis technique was used to measure CO and O2 association rates. The laser/xenon flash photolysis apparatus, the tonometer, and procedures for sample CO preparations have been previously described.28 P3 was determined by titrating deoxyMb's with a 0.82% CO in 82 argon gas mixture. The M values were measured by titrating oxyMb's with CO which were previously equilibrated with CO 8 I values 0 O 500 torr 02' P;5 2 was calculated by PS 2==M>

.H mhoo.o omo.o moHvx~.H ma oom.m ooavxm.m o.H memmoum.m 0H ooo.mH mooo.o mmo.o moax_H.H mm ooa.m hoacxw.m m.H mSONmUIm.m QH oom.m smoo.o ea.o ooHvxm.m mm oom.ma N.0.ch h.v mm.o kusmlunc mH HHO m m HHO A mt HH- V AH- 75 A t AH-mo AH-m TE 6 x o: co m H .H .I- No a III-Ni, .6. Fa pcsoasoo c.AUoN~|oNV muwm mCHUCHm pcmqu map Hmmz omumsufim wuHHmHom mcH>Hm> mo museum £uw3 mmEmm chman9 m0 mucmumcou mcwpsHm No can 00 .Hlv mHnt 89 loosmmm ou GCHUMOH . x x 00 m\NO m w .HN wocmnwmmmn .mchfluhm pmsomuum >Hucmam>oo m mMB mmmn Hoaxmo .cofluomump mo muHEHH may mcflzomonmmo mucmumcoo mums Hmpno umHHm >Hmmmmoms mum3 mcowumuucmocoo cmmhxo 50H: huwcflmmm «0 BOA map OH 050 .ln-l-‘I o .mcmsaou cHs N.oH x m.m m ooo.m MWoH x o.m Hcamwnm 63%. >3 x no o.oH ooo.oe ooH x o.m Hemmnum Scum omm.m meooo ~85 ooH x o.H oH ooHJ 53 x o.~ 6633666 28.8 no ooo.m~ emooo «Ho oon m.~ omH ooo.om eeon m 635386 36 6N 4mg 172 17... 75 3.83 17...; AT... H-E oz ouxm H .H Norm x .x ocsomeov d.AUo-Io~V masonu Hmaom mHOEmm baa: mmEmm H>smanQ mo musmumcou mafiosfim NO can 00 .mlv anme 90 lJik'c 2 alc = 0.092511g ‘-0.748u -+18.0 g 2 1nk =-O.014811g ~0.263ug+9.73 calc Figures 4-2 and 4-3 show plots of the calculated rate constants vs. the observed rate constants. The dipole moments used were those of unsubstituted phenyl models or reasonably close approximations.90 The striking linearity of Figure 4-2 (correlation coefficient 0.994, lepe 1.06) shows that oxygenation behavior can be affected in a predictable manner by the magnitude of a nearby dipole. The linearity of Figure 4-3 (correlation coefficient 0.967, slope 0.995) is not quite as ideal as Figure 4-2, however, the most striking feature is the very large deviations from the calculated line of the compounds capable of hydrogen bonding, namely the 3,5-CONHnBu and 3,5-CH OH. 2 Prior to 1940 quadratic equations correlating reaction rates and acidity constants with dipole moment data were emperically developed.91 These equations have the same form as Equations 7 and 9 (see theoretical section). Concurrent with the earlier work, the Hammett free energy relationships were being developed. Since little work has appeared after about 1940, it seems as though the success of the Hammett correlations left such dipole moment correlations by the wayside. Thus, there appears to be no theoretical explanation for the observed quadratic dependence of reaction rates on substituent dipole moments. 91 l7.5 U! .D . O x 3.5-CHzOMe E: I? - 3.5-CH20H is-coun,// I6.5 // 3,5-c0NHc.H, / I L I I6.5 I7 I7.5 In I"calc Figure 4-2. Correlation between Ihik' and 1nk' . . 2 calc obs 1nkzcalc-0.0925u -O.748ug-+18.0. Correlation coefficient: 0.994. SlOpe: 1.06. 92 4-t-Bu 3,5-cozC‘ I 3.s--cov~I(-..I=-)z f 3-5'CHzOMe 3,5 - con-(rat)2 8 _ I 3.5-CH,0H I 3,5-CONHC,H, Ink calc Figure 4-3. Correlation between 1n kcalc .. _ 2 _ lnkcalc— 0.014811g 0.263ug+9.73. Correlation coefficient: 0.994. Slope: 0.995 (calculated without 1c,1e). and ln kobs . 93 The quadratic functional form for the dependence of ln]<(k') on a local dipole moment can be obtained without including dipole-dipole polarization terms. In effect, replacing the polarization scalers by unity and the polarization vectors by zero in Equations 6 and 8 (theoretical section) yields such equations. However, the "unpolarized" equations are not intuitively satisfying. The signs of the coefficients on ugz obtained from fitting Equations 7 and 9 to the observed 02 rate data dictate that the radius of the deoxyheme (rD) is greater than the transition state (r*) which is greater than the oxyheme complex (rB). This does not constitute a proof that the model presented in the theoretical section is necessarily complete, it does suggest that the interaction between FeO2 and a local dipole is more than the vector sum of their respective dipoles. Unfortunately, the large number of parameters in Equations 6 and 8 makes physical interpretation of them difficult. The dipole- dipole interaction thought to occur in the oxyheme complexes is shown pictorially in Figure 4-4. The unmistakable linearity in Figure 4—1 indicates that dipole-dipole interactions, interpreted here as polarization, in the transition state plays a dominant role in 02 binding to the models studied. Assuming an 70 the increased scatter in early transition state, Figure 4-3 is not surprising since oxygen dissociation should be more dependent on orientation parameters than Figure 4-4. 94 Relative orientation of FeOZ dipole and dipole of a 3,5 disubstituted benzamide (A). Scale drawing“ of the relative orientations and distances of an FeOz dipole and ag unconstrained o—phenyl amide dipole (B). “From ref. 94b; bFor acetanilide ref. 105. .96 association. Comparison of the 02 rate data for diethyl amide (1e) with that of diisopropyl amide (1f) indicates that the difference in 02 binding is primarily in 02 dissociation. It is rationalized that the bulkier diisopropyl amide may not be able to assume an orientation in which the amide dipole and FeO2 dipole achieve maximal head-to-tail alignment. As well, it may also restrict the distance at which the two dipoles approach one another. Indeed CPK models predict that such a situation might exist. The electrostatic potential relationshipg2 (Equation i) predicts that both processes should result in a lower 02 affinity for 3,5-diisopropyl benzamide substituted heme relative to the diethyl amide. 2 . I) = 11 cos e/r , (1) where w is the electrical potential; u, the dipole moment; r, the distance between dipoles; and 6, the angle between dipoles. In addition, the 4-t-butyl benzamide heme deviates from a simple rule of thumb relationship consisting of "the more polar, the higher the affinity". This could be rationalized as a dipole-dipole orientation and/or distance effect as above. However the unmistakable linearity of Figure 4-2 as well as the congruence to the trend in Figure 4-3 argue against this rationalization and suggest that oxygen affinity need not adhere to such a simple relationship. This is understandable since 02 association 97 is dependent on the dipolar forces in the transition state while dissociation is dependent in both the transition state and oxyheme complex. Thus the O2 affinities in Table 4-1 reveal that Pk reaches a maximum for the 3,5- dibutyl ester (neglecting the di—isopropyl amide). In fact a TPP based capped porphyrin in which the cap was linked to the porphyrin via ether and ester linkages actually 60b'93 From this then the discriminates against oxygen. polarization of the transition state and the oxyheme complex seem to be unequal, i.e. aaéa# (see theoretical section). Somewhat disconcerting is that upon increasing the polarity of the distal side results in slower O2 association. This too is consistent with an early oxyheme transition state. Increased polarization of O2 in the transition state results in a more product-like activated complex, hence decreasing O2 dissociation rates. If this is indeed the case then by microscopic reversibility O2 association would also be reduced. The results obtained here indicate that placing a dipole at close proximity to the heme center produces kinetic and thermodynamic control of the reaction between 02 and ferroheme. That this control is due to polarization of FeO2 is further evidenced by the data in Table 4—2 which shows the binding parameters of O2 and CO to hemes equipped with groups of differing dipole moments at greater than 4 A94 from the heme center on one face and covalently linked 98 bases on the other. In order to assure no electronic" effects, cis-trans rotomers were chosen, in which the polar acetamido group is on the same (Eggs) and opposite (gig) of the OZ/CO binding site. The other two entries in Table 4-2 are of Momenteau and Lavalette95 in which amide and ether linkages were used to supply an alkyl strap across the distal heme face. The results are essentially identical. These data suggest that the enhanced O2 affinity of the Erangfamides relative to the gisfacetamide or ether wouldlxadue to a head-to-tail alignment of the Fe-OO and amide dipoles. That this may be the case is evidenced by the strong dependence of k which increases by 8-900% on going from the same side amide system to the cis— acetamide or ether strap. Furthermore, we believe that this enhanced affinity is derived from a constructive dipole-dipole interaction and not a direct interaction such as hydrogen bonding. In the crystal structure of FeO2 (TpivPP) (l MeIm)94 there exists a four way statistical disorder of the terminal oxygen atom which bisects the NPYREFeENPYR angle and thus points toward the amide "picket" moieties. The crystallographic structure shows the terminal oxygen to N distance to be approximately AMIDE 4 A, which is too far for effective H—bonding. Assuming in solution the structure remains for the oxyheme complex as in the crystal than an Fe-OO dipole would align in a constructive mode with the amide dipole (see Figure 4-4B) 99 resulting in a net increase of the FeO2 dipole moment. Thus, it seems reasonable to conclude that the stabilization is the result of increased polarization of FeO2 producing a more ferric superoxide-like compound rather than an Fe(II)02. In this regard, deviations between experimen- 96,97 tal and theoretical98 considerations may become unified since calculations do not consider the effects of the local environment resulting in polarization of the FeO2 moiety. In the absence of other model heme-O2 complexes of known structure, a dipole-dipole alignment mechanism provides an alternative rationale for O2 orientation relative to the heme axis. We notice that in oxy 49,87,99 88,100 hemoglobin and oxy myoglobin the 0-0 axis eclipses an NPYROLE 94 heme it bisects the NPYR-Fe-NPYR angle. Inspection 88 89 of stereoviews of the active sites for oxy Hb and oxy Mb -Fe bond whereas in picket fence reveals that an FeO2 dipole may be somewhat aligned with the distal histidine dipole in a head-to-tail fashion besides being hydrogen bonded. The most striking feature in Figure 4-3 is the large deviations from the calculated line of the 3,5-di-n—butyl amide and 3,5-dibenzylic alcohol substituted hemes. These deviations are accounted for in terms of hydrogen bonding to iron bound 02' Comparison of Figures 4-2 and 4-3 reveals that the primary amide and alcohol lie on the calculated line for O2 association whereas significant deviations occur in oxygen dissociation. This is the 100 result of hydrogen bond formation occurring subsequent to a rate limiting process. Estimated hydrogen bond strengths for the n-butyl amide (1d) and benzyl alcohol (1c) are 1.6 and 0.6- 0.8 Kcal/mole, respectively. These are the AAG values for the n-butyl vs. diethyl amide and benzyl alcohol vs. benzyl methyl ether or butyl ester where AG==-RTIh1K. Such a calculation should be regarded as qualitative since these calculations assume that the magnitude of the local dipole moment and that the FeO2 dipole and local dipole orientation are independent of hydrogen bonding. Figure 4-5 shows the effect of dipole-dipole interactions and hydrogen bonding on a simple reaction coordinate for heme oxygenation. Carbonylation Kinetics: Since free carbon monoxide has a dipole moment (“CO 101) and bond formation between Fe and CO 85 increases the dipole moment of CO, it would be expected =-0.112 D that dipole-dipole interactions should play some role in carbonylation kinetics. Consistent with previous solvent 28’55 we find little correlation between CO kinetics studies and the magnitude of a local dipole moment. The lack of correlation is explainable from Equation 1 in the theoreti- cal section. It would appear that electrostatic perturba- tions on carbonylation are outweighed by non-electrostatic Figure 4-5. 101 Foo; // I I / I / I ‘ I / I / I / / Fe +02 Fe 02 ’ \ / \\’/ o 4——-—>- +> FeO2 + R b 4—I- F902. ' H B Schematic representation of a proposed simplified reaction coordinate of heme oxygenation. Hypothetical unperturbed coordinate (---), plus an interacting dipole (a———) and ydrogen bonded oxyheme complex (b-—-). 102 factors. This dependence could be explained in terms of the size and position of the phenyl substituent. Previously we15d and others60 have shown that a distal steric effect differentiating CO and O2 binding will be most pronounced in the association rate constants. Indeed, a comparison of the 3 amides in Table 4-1 shows that increasing the size of N substitution regularly decreases l' with almost no effect on 1. Oxygenation however remains almost constant. Furthermore, compounds lb-d have very similar CO association rates which differ significantly from 1a. This may be attributed to the position of substitution on the benzamide distal substituent. CPK models reveal that at the heme center compounds lb-d have similar steric bulk which is not present in la. Although this does not constitute a proof for the existence of a steric effect differentiating CO and 02, it is noteworthy that with the exception of the O2 dissociation rate, the 3,5-diisopropyl amide (1g) has CO and O2 kinetic 5 l -1 rate constants very similar to Mb (1' =3-5 XlO M- s , 1 =0.001s-0.04 s’l, '=1-2 x107 M'l s’l, k' =10-3o {1.55 The possibility of hydrogen bonding to carbonylated 102 It is hemoproteins has previously been discussed. evident that compounds 1c and 1e can provide a hydrogen bond to FeO2 yet the possibility of hydrogen bonding has no effect on carbon monoxide kinetics. That is to say, if hydrogen bonding is occurring in our models is has no effect on carbonylation kinetics. 103 Distal Steric Effect: Recently much effort has been put forth in the synthe- sis and characterizationlSd’6o’93 of models aimed to test the distal steric effect hypothesis.49'50’95’103 It is well-established that a distal steric effect does 15d,60 modify oxygen and carbon monoxide association rates but the effect on dissociation rates, primarily oxygen, is in question. Conflicting results have been obtained for oxygen dissociation rates as a function of steric encumbrance.15d'60 From the work presented here it becomes evident that the lack of a clear cut answer to steric differentiation could mainly be due to dipole- dipole and/or hydrogen bonding variability within a seemingly congruent series of compounds. Collman's picket fence-based compounds showed marked decreases in 02 dissociation rates upon shrinking the cavity of the ligand binding site. The reported oxygen dissociation rates are, respectively, 2900, 71, and 9 s.1 for FePiv35CIm, FeMedPoc(l MeIm) and FePocPiv (l MeIm).60b “From CPK models it appears that the introduction of the phenyl cap on top of the heme face (pocket hemes) results in a pulling of the amide moieties toward the heme center (Figure 4-6). This would bring about a closer head-to-tail dipole—dipole interaction between the amides and FeOZ, relative to Fe picket fence. In fact for FePocPivP it appears that the amide proton may come close enough for hydrogen bonding. Figure 4-6. 104 The change in dipole orientation upon introduction of a tight strap across the heme face. (--) unconstrained o-phenyl amide. (---) sterically encumbered model. 105 The alkyl amide linkages of Traylor's cyclophane hemes602 should be flexible enough to rotate and it is not difficult to find conformations in which the amide dipole may align itself in a head-to-tail manner with FeOZ. On expanding the length of the anthracenyl strap, the amide dipole would become more distant from the heme center. Oxygen dissociation rates for Fe 6,6-cyclophane heme (k=800 s‘l) and Fe 7,7-cyclophane heme (k=1000 s‘l) are consistent with this interpretation. Our linear chain strapped hemes also showed this correlation;15d on increasing the strap length we obtained an increase in oxygen dissociation rate: 250, 175, 130 s.1 for FeSP-15, -l4, and -13 respectively. I Among all sterically hindered models which contain amide linkages, the only inconsistency is the O2 dissocia- tion rate data for our Fe-Cu cofacial diporphyrins where k for Fe-Cu-4 and Fe-Cu-S is 160 and 91 5.1, respectively.15d We believe that this is due to the relatively large distance between the amide group and the heme center, which renders the dipole interaction relatively insignificant. Expanding the diporphyrin gap may also result in an increased accessibility to FeO2 by polar molecules present in the system, i.e. imidazoles. This is not the case with the strapped hemes since the linear hydrocarbon chain provides much less shielding than does the Cu porphyrin cap. Our contention is that with models so far available, it is still not possible to judge the steric differentiation 106 between CO and 02 based upon the affinity ratio M since one cannot ascertain how much change in 02 dissociation rates comes from pure ligand distortion and how much is due to dipolar effects, let alone the effects of bending and ruffling of the porphyrin plane introduced by the encumbrance. Summary Our work presented definitive evidence showing that both short and long range dipolar forces as well as hydrogen bonding can play a significant role in regulating oxygen affinities to heme proteins. While this general conclusion is consistent with previous solvent studies,28’55 the use of covalently attached polar groups offered much greater advantages in probing the micro-environment of the heme coordination site. We have demonstrated that while dipolar forces can produce kinetic and thermodynamic control, hydrogen bonding provides an additional path for thermodynamic control of heme oxygenation. In contrast, CO binding to hemes is little affected by distal polarity effects. 107 Theoretical Section In general92 the reaction constant between two dipolar molecules in a medium of dielectric constant e is related to an unperturbed rate constant (k0) by I; .3 u: 1nk '-'- lnkO " (E-l)/kT(2€-1) 3+3-3 __a B t + £0/kT (1) where and u* are dipole moments of the reacting “A! ”B species and transition state respectively, rA, r and r* B are molecular radii, 20 is the perturbation due to non— electrostatic forces (i.e. H-bonding, steric effects, etc.) and kT has its usual meaning. Equation 1 is usually used to investigate solvent effects of a particular reaction (a is a variable and ui's are constants). If Equation 1 is taken to be valid, then it seems reasonable that 6 could be held constant and ui's varied. For this approach it is necessary that a series of compounds be used in which the varying dipoles location, relative to the point of reaction, is essentially constant. It is felt that most of the compounds in Table 4-1 meet this requirement. Thus, setting (€-1)/kT(2€+1) equal to a constant (c) for reaction rates measured under conditions of constant temperature and solvent composition, the electrostatic perturbation (E_ from Equation 1) may be L written for unimolecular ligand dissociation as: 108 E U2 u 2 .29.: _§H._§:_ (2) c r3 r 43 B B where the subscript B refers to the ligand bound state. It is assumed that the net dipole interacting with solvent is a dipole aggregate of the liganded heme and transition state and is represented as the vector sum of the individual dipoles of the complex, that is, the Fe-L dipole (uc) and the interacting dipole (ul). Equation 2 is rewritten as: 2 2 E-L = [uc'+ul] ‘_[uc'+ul*) (3) c r3 r 3 ' B 3* Since the ligand does not chemically react with the local dipole it seems reasonable that the dipole moment of the local group in the activated complex may become polarized, changing its magnitude and direction in the extreme. The local groups intrinsic dipole moment in the activated complex is related to the unpolarized dipole moment (mg) as: = +16 4 “9* aug ( ) where a is a magnitude sealer and b is an orientation vector. Similarly, it is expected that the interaction between the 109 Fe-L and local dipoles results in polarization of the respective intrinsic dipoles: uc = 0011 +8 (5a) 11* = equis +84. (5b) “1 = Yug +6 (5c) ult = Yiugt'+5# (5d) where “i is the unpolarized intrinsic dipole moment of Fe-L, d's and y's are polarization scalers due to changes in dipole magnitude and 8's, Y's are polarization vectors due to changes in dipole direction. Substitution of Equations 4 and 5 into Equation 3 yields the electrostatic perturbation term for ligand dissociation (Equation 6). 2 2 2 . E-L Y Y*a 2 2Y(5+aui+8) 2aY*(Y*b+5*+a*ui*+B*) c r3 r' 3 g r3 r 3 g B 3* B 3* (5+8)2 +a(6+B)u. +0211? 1 l + (6) r3 B 2 2 Ytb -+2Y*b(6*+a*ui¢+8¢)-t(0*+B*)2-+20*(6*+B*)ui¢-+G§ui¢2 r 3 3* For a series of closely related compounds (as in Table 4-1) it is conceivable that the polarization sum terms 110 (coefficients of “g in Equation 6) approximate constants. The kinetic expression for ligand dissociation as a function of a local dipole moment is then simply a quadratic in the local groups dipole moment. 2 1nk = A +B +C 7 -L “9 Mg ( ) Equation 7 could be made more general by including dipole-dipole interaction energy terms (Equation i in results and discussion), however for the models studied it is assumed that the distance and angles between Fe-L and the local dipole are essentially constant. A more complete theoretical treatment will be presented elsewhere.104 Identical treatment of ligand association results in the following expression for the dipole-dipole perturbation term. E 2 2 '2 Za'Y (Y b'+6 +a u +8 ) _L_ _ 111+ _1_ Hat 2_ 1|: t 4: t 1* I: c - r3 r3-. 3 ’ug 3 pg L D r13“ rB* (8) 2 2 Y2b2-+2Y*b'(6*+o*ui¢+8#)-+(6*+B*)2-+2a*(6*+8*)ui¢-ta*ui¢ r 3 5* where “L’ rL and rD are the free ligands dipole moment, radius and deoxy heme radius respectively. As above the coefficients of “g are taken to be constants for the 111 compounds in Table 4-1. For ligand association the kinetic expression is: 2 JJlk = A' -+B' '+C' 9 L “9 ug ( ) The coefficients in Equations 7 and 9 can be evaluated by fitting observed k_ and k to a quadratic in L L pg. USing the coeffiCients so obtainedlnkL calc and Ih1k_L calc can be determined for particular values of pg. A plot of calculated vs. observed rate constants, in the absence of other dominating perturbations, should be linear with slope equal to unity. Materials and Methods Compounds la-g and 2a,b were prepared according to reference 89. Toluene was purified by stirring at 0°C with several changes of conc. H280 followed by drying over 4 anhydrous sodium carbonate and distillation from Lithium Aluminum Hydride just prior to solution preparation. Sample solutions for kinetics and CO titrations were prepared by dissolving the ferric compounds in approximately 4 4-ml of toluene (~lo‘5 M) containing 10' M of benzo- phenone. The solutions were degassed in an 80 mL tonometer by freeze-pump-thaw cycles at 10-5 torr. The hemin chlorides were reduced by photolysis according to the previously described method.9 Flash photolysis was carried out with either a xenon photographic flash gun 112 (Braun 2000) or a flash lamp pumped dye laser (Phase-R DL2100) with rhodamine 6G dye. Decay constants were calculated from transmittence vs. time measurements at 432 nm (five-coordinate heme disappearance) or 410 nm (oxyheme appearance). CO association was monitored at 418 nm and the output of the photomultiplier was recorded on a Bascom-Turner recorder through a log amplifier in absorbance units then directly computed as pseudo-first order rate constants. CO and 02 concentrations ranged 5 3 4 to 6 >coo can .NHUNmU Homm> mm :uH3 omumcououm 0v anon m.mmw£om mo mnuommm coHumHOmnm -n_+.._mm III mm N n no. to +-.._+zmm _ muc 00m CON 000 com CO? :34 a _ I _ q _ _ _ _ u - on fl ( - #00 mm? 9. nzw may .m-m opson 123 can HI-Iv 0m .sms cH 1--o mimmoommoc.om .Aummcflv am GHHOHQOOHHOHUMD mo mnuommm GOHHQHOmn¢ .vnm mHsmHm 124 00m , CON 000 00m 00? E: < .. _ _ _ _ L _ L H _ \\ III /// 7/ \\\. I- x /,\:mm ,, , moo m+ . one 1, H z. \ ooo ooe ooo ooh ooe 1:. I/ \ — fl _ _ _ _Omr\ / .s L «o... oo» n.x mm 98 nos 81. SOUBCIJOSQV 125 coulombically neutral ethyl cyanoacetate and malononitrile adducts, e.g. 4e and 4f, and other carbonyl derivatives were studied. As shown in Figure 5-5, the fact that the spectral characteristics of 4e and 4f are very similar to those for 4c~H+ or 4d almost immediately rules out the charge delocalization model. Substitution of an ester for nitrile moiety, as in 4e and 4f, results in a less electron withdrawing substituent, and consequently a smaller degree of red shift (4e: 1520 cm"1 vs. 4f: 2056 cm.1 shifted from 4b in THF) as well as less split- ting in the Soret region (Av4 : 3630 cm-1 vs. Av : e 4f 4280 cm-l). Likewise, the substituent effect can also be evidenced by "saturating" the carbonyl group. Figure 5-6 shows the spectral changes associated with addition of pyrrolidine to 4b in CH2C12. The resultant spectrum is essentially identical to that observed after addition of tetrabutylammonium borohydride to 4b in CH2C12 (Fig. 5—6, inset). Acidification of the borohydride-treated compound resulted in quantitative 113,114 conversion to a copper porphyrin. Acidification of the pyrrolidine adduct converted it to the pyrrolidinium salt. Based on these observations a hemiaminal115 structure is proposed as the pyrrolidine adduct 4g. Thus, irreversible reduction or reversible hemiaminal formation produces a blue shift of the visible band by approximately 700 cm".1 along with the Soret region coalescing to a single sharp peak. Similarly, as observed for all our 50 25 Figure 5-5. B 403 403 \ 47 - l/ \b/ \ 437 737 .. I’ \\ 709’- CHC(CN)2_ / l \ v, \‘ cuc(c~)c0251/ \ I ,’ \\ r- ‘X x /’ \\ 7 \ ’ \ I I Al I I I I .J I\‘t:- Anm 400 500 600 700 800 A): Absorption spectra monitoring the reaction of 4b with pyrrolidine-HC104 in THF (total elapsed time-l hr). B): Absorption spectra of ethyl cyanoacetate adduct 4e (---) and malononitrile adduct 4f (-—) in THF. 127 .Huonch NHommo cH.oonoxconon EsficosfimH>usnmuuou cuw3 he no sowuosomm .Amvanummv «HUNmU cH wchHHouuhm mmmoxm suwz be NO sofluomwu on» mcwnouflcoE mnuommm cofiumuomnd com com com com 00v 8: x l . . _ mm -0m .6. as p .1239on 28 - oomI ooc I 1 ////v , . #mWN mx Imm z. A - a man Ion Imh Ev IOOF ozo- I-Ilc 98- To -_oo. 1 I W L .mlm wusmwm 128 Schiff's bases, replacing O with a less electronegative N during Schiff's base formation leads to a blue shift of the visible band relative to their parent formyl or acroleinyl compounds. From a phenomenological perspective, the above data combined with proton NMR and resonance Raman studies below leave little doubt that the unusual spectral properties of protonated Schiff's base porphyrinoid derivatives is brought about by the presence of a conjugating electron deficient substituent on the ring. The proton NMR spectra during addition of anhydrous HCl to 1c in CDCl3 are reproduced in Figure 5-7. As shown, progressive acidifi- cation caused all protons near the C=N group to shift and broaden due to exchange. While the deshielding effect of protonation of the C=N double bond undoubtedly could produce the upfield shift for the -CH=N- proton, the adjacent meso proton, and the adjacent pyrrolic methyl group, the protonation caused the a-imino methylene protons to shift downfield owing to increased electro- negativity. That the other peripheral substituents on the porphyrin ring were little perturbed indicates that the positive Charge hsnot substantially delocalized throughout the porphyrin n-system. The resonance Raman spectra of lc(SB) and lc-HCl(SBH+) in CHZCl2 using 406.7 nm excitation are shown in Figure 5-8. The Raman spectrum of 1c shows strong enhancement of totally symmetric modes at 1598 cm-1 129 .»H0>Huommmmu mou ou Eouuon Eonm Ado: .vm N.Hv soflumcououm mumHmEoo can Haum .vm m.ov OHMHcmfiumusH .mmon mmum .mHooU :H mpHHoHno cmoonchs msam OH HO muuommm mzz um: omm .oIm musoflm 130 Eon O _ N n c n m - m m 0. .. F*_LLDPFPrp-r-FLIIrPFI—rrlr-PIPI—FPFPb>+>PPHFPP_FF+PFLP>+FNFFPb‘NFFF J as. as: 131 .HH-so ooeH-ommH mHooHso Hoo.oH one Asouuono Ho:.oH .Haouo 0H mo :oflumufloxm HmmmH E: h.mov nua3 NHUNmU :H muuommm :mEmm mosmsommm .mum musmflm race 08. 00.. 8m. 8m. 00¢. com. 00m. 00.... _ _ a _ 4 . ._ _ _ ._. 0 I. C. m. H by _- _ m u _ _ m an m- l l 6 6 _ 8m. U a x m - m m _ ._._ u 2mm I 9| W + . +68 m on m 2 w u .A _ . _ .. _ _ w. _ _ _ n _. mm _ . I I __ I H 2 II L I-I- % mm W I m m «tows. Ra m ._- 62 C. _ - _ mm m a m. a % 132 1 (v 1518 cm.1 (v3) and 1383 cm- (v4), whose assignments 116 2)- are in analogy with those of Abe et al. The line observed at 1652 cm-1 corresponds to a Blg mode (v10) which is most likely enhanced through a Jahn-Teller or intramanifold coupling mechanism;117 it is commonly observed when studying hemes and heme proteins by Soret excitation Raman.118 The Schiff's base -C=N-stretching vibration is responsible for the line observed at 1639 cm-1. For the protonated Schiff's base, we note little change in the frequencies of the observed ring vibrations,119 indicating that protonation effects are localized at the Schiff's base and do not strongly perturb the basic porphyrin bonding pattern. However, the decrease in symmetry, which is apparent in the optical spectrum (Figure 5-2), is reflected in the Raman spectrum of lc-HCl in that the scattered intensity from non-totally symmetric modes (Blg’ A2 32g) is much stronger, while 9’ that from the totally symmetric modes (A19) is decreased, relative to the free Schiff's base. Protonation of the Schiff's base shifts the -S=CH- stretching frequency into the 1650 cm"1 region where it overlaps strongly with le‘ In order to determine its frequency more precisely, we carried out analogous IR experiments which showed 3 (-3H=CH-)==1650 cm-l; these data form the basis for the assignments of the two vibrational frequencies in Table 5-1. If DCl is used to deuterate the porphyrin Schiff's base, the stretching frequency decreases to 133 .m¢ mEoowm_HWNm .mmd .mam .mHmw .« use m How msoum 5N0 uomuuoo OHOE may CH NH.HM um and Eonm mumnenc woos 0cm mmfluumfiewmo o 6 II Ammao ms NNHH mmHH AHHao Hm memH I- o o Homav mm emHH I- Anal Ha NooH oomH o m AH~>V ma ooMH moMH H-monmz-v a osoH II a m m Heal He HomH oan H-moumz-o > omoH I- a m m Hmao Hm onH sHmH H-moIz-V > -- omoH s 0 Homes m - somH AoHao Hm emoH mmoH Lev ucmscmHnmm +mmm mm Hes ucmscoHnmm +mmm mm .mmHommm mmmm m.mmw£om cwuhsmnom AHHVHZ muswficmflmmfi mumcflouoou HMEHOZ can pm>HmeO mcowumnnfi> .Hlm manna d 134 1640 cm.1 (Figure 5-8, inset). A similar pattern of -C=X- stretching vibtation frequency shifts is observed in retinal Schiff's bases upon protonation and deuteration and the physical mechanism underlying these shifts has 120 From this been discussed in detail by Aton et al. analysis it appears as if the interaction between the C=N stretching vibration and the C=N-H bending mode is somewhat less in the aromatic porphyrin case than in the linear polyene retinal case. The high frequency (1000-1800 cm-l) resonance Raman Cl spectra of 4d-HC1'in CH and THF as well as lc-ClO4 in 2 2 THF with 406.7 and 488.0 nm excitation are reproduced in Figure 5-9. The low frequency (0-1000 cm-l) region spectra were masked by solvent vibrations and are not shown here. Comparison of the SB and SBH+ resonance Raman spectra with 406.7 nm laser excitation reveals similar vibrational frequencies for the two except that the peripheral stretching frequency (C=N) for SBH+ is absent. With 488.0 nm laser excitation of SB 4c-HCl results in a resonance Raman spectrum displaying the typical -C=NHR stretching vibration at 1646 cm-1, in analogy with lc-HCl. Further comparison of the resonance Raman spectra of 4c-HC1 with 406.7 and 488.0 nm excitation reveals large intensity differences between the two spectra. With 488.0 nm excitation, the totally symmetric A modes at 1592, 1502, and 1375 cm_1 are decreased 19 in intensity relative to the nontotally symmetric 135 Figure 5-9. Resonance Raman spectra of 4c‘HCl (CH2C12 and THF) and pyrrolidinium perchlorate adduct 4d (THF) with 406.7 and 488.0 nm laser excitation. 136 H H m + as? + $20. + FM. 4 a. a .3 D H 26 W 2 mHm i. 1.: «sow % S w ST 4 a. I W =8- ET .I-.. =qu .5? I .3... 5.7 m :unI _ I sou- . UN .60 I- Wmm m ST 1 D... a .2 t 8. a .muOl Baal I .3 Baal m 2.:- .u.o I .ufi - a:- .u.oI .gu- B .uuoI Go? 5.6- .8.- .uqu .wax- .uqm - .uquI ad. 3qu .927 ouI m 38- a8- .3... I 6qu W..- I 4 . x a. x m. l. .m. 33-. m. 3.3- ahwu- 3.3-: .m. 3.3m m. n .u. x .m. I W _uo~I 8.8 I 68- - .b - I a - .uao - .uuu - aoo - 83 .86 .3» I «.0 GNUI gOI .uflul .uflb I .ugl m coo - .8» I .So- .3. I 3.3- I 6 .3- .80- 5: I II .ouo I 6.9- .96 - 60.7 3.8- .2» I . O a.- m t.wzm._.z_ Z<2H0>Huommmmu On one Um no noun coHumsouonm umHHm may one Haven 0cm uuoH momma .om can on momma m.mmanomIHp mo .mHommU cw mobv mcowumnuwu neon OHHOHnonmm .OHIm mHDmHm 000 00¢ 00m A a I. H J ON 0? 00 N + row-c. . TEE-Tomi oo WW3 < uco>aom Ix .usm>aom can sceumucsoo mo coeuocsm II o mo 0H mmmm m.mme:om pmumsouonm Hem mama Hmuuommm maneme>l>b .mlm manna 146 .mmmmsocmm Ammo dmem Ammo mmm Ammo emm Ammo mme mmmm zommo mnemo Ammo dmem Ammo mmm “mm. emm Ammo mme mmmm zommo -mo Aemv mom Ammo omm Ammo mme mmom mam memo Ammo mom Ammo mem Ammo moe mmmm mme Imo Ammo eem Ammo mom Ammo mmm Ammo mme memm mmommo mmmi.um Ammo eem Ammv mom Ammo mmm Ammo mme oemm mmommo «IOHU Ammo eem Ammo mom Ammo mmm Ammo mme oemm mmommo IH Ammo mmem Ammo mom Ammo mmm Aemc mme mmmm mmommo -um Ammo deem Ammo mom Ammo emm Ammo mme mmmm mmommo -mo Ammo meem Ammo mom Ammo mmm Ammo mme mmom mmommo um EU AT V AHIEU HIEE my E: K Omo>< ucm>aom x .uco>Hom 6cm COmHmucoou mo GOmuocom m mm om mmmm m.mmmsom omumcouomm mom mama mmuuommm mmnmmm>I>a .mum mmnme 147 .mmmmsonme Ammo emem Ammo mmm Ammo mmm Ammo mme mmmm zommo «IOHQ Ammo dmem Ammo mmm Ammo mmm Ammo mme mmom zommo -um Ammo mmem Ammo mmm Amm. mmm Ammo mme mmmm zommo Imo Ammo mom Ammo «mm Ammo eve mmmm may mIOHU Ammo mom Ammo mmm Ammo Hue ommm ems Ium Ame. mom meme mmm Ammo mme mmom mme Imo Ammo mem Aemv mom Ammo emm Ammo mmm mmmm mmommo mnemo Ammo mem Ammo mom Ammo mom Ammo mee meem mmommo -um Ammo mmm Amm. mom Ammo mom Ammo mee mmmm mmommo Imo EU AT c AHIEU HIZE wv me K omo>< ucm>aom x .soHuwmomEou ucm>aom can coflumucsou mo cowuocsm 0 mm we no muoo Hmuuommm manmmw>I>D .va manna 148 observed essentially no counter anion effect for either 4c-HX or 4d-X-. The anion dependence in the Soret region of chlorins 4c-HX and 4d:x- also paralleled that of porphy- rin lc-HX. There is however, a difference in the protona- tion of Schiff‘s base 1c vs. 4c with HF. The spectrum of 4c-HF in CHZCl2 is essentially identical to 4c-HC1 (Table 5-2) and 4c-HBF4 (obtained from addition of BF3-0Et2 vapor to 4c-HF123b ) is identical to 4c-HClO4 (see Figure 5-3 inset and Table 5-2). This could be the result of two effects: the pKa of 4c is higher than the HFZ- equilibrium constant or more likely, the presence of the 7-hydroxyl moiety aids in fluoride association with 4c HF, thereby lowering the "free" fluoride concentra- tion.123C The anion dependence on the spectral properties of pyrrolidinium salt 4d is clearly the result of ion pairing, however, the anion dependence of protonated Schiff's bases (lC'HX and 4c:HX) cannot be regarded so simply. Compound 4c-HX and 4d-X- have very similar anion dependencies in CH2C12 yet different in THF. Since both solvents have relatively low dielectric constants we expect that the smaller anions are closely associated with the positive charge, producing the observed relative‘ blue shifts with decreasing anion size. For 4d-x- this accounts for only 90 cm-1 in THF while for 4c~HX, there is l 600 cm- difference. This phenomenon may be the result 149 of Cl- H-bonded to SBH+. The diminished anion effect of 4d-X- in THF relative to CH2C12 is rationalized as arising from positive charge solvation. Since using perchlorate as the counter ion for pro- tonated 1c and 4c as well as 4d produced the largest red shifts, we assume that in the 3 solvents used, ion pairing with perchlorate is minimal. In THF and acetonitrile, we expect that a positive charge will be much more solvated than in CHZClz. For pyrrolidinium perchlorate 4d this 1 1 produces a Av of 1130 cm- vs. THF) and 1270 cm- CHO 2 (CH2C12 vs. CH3CN); for 4c-HC104, the solvent dependence 1 l (CHZCI vs. THF) and 590 cm- 1 is 550 cm (CHZCl (CH2C12 vs. 2 CH3CN); and for lc-HCIO 1 433 cm- (CHZCl2 vs. THF) and 4: 459 cm“ (CHZCl2 vs. CH3CN). However again, we cannot ascribe a single dominating effect to the shifts observed for SBH+. Intuitively,we would expect that hydroqen bonding from a protonated Schiff's base would result in a blue shift relative to a non-H-bonded SBH+. Since the pyrrolidinium perchlorate spectral data suggests large contributions from solvation on the extent of visible band 4 to 4d-C104- spectral data in H—bond accepting solvents (i.e. THF) vs. non- red shifting, comparison of 4c:HClO accepting solvents (i.e. CHZClZ) would lead to no useful conclusions. The presence of a hydroxyl group in chlorin 4 offers a unique opportunity to study the effect of an 150 intramolecular H-bonding to CHO or CH=NR. In CHZCl2 we observed that the visible absorption maximum of 4b is red shifted by 169 cm‘1 with further splitting of the Soret region relative to THF (Figure 5-14). This effect can be titrated away with a variety of hydrogen bond acceptors, i.e. amines and ethers. The inset in Figure 5-14 shows such a titration with pyridine. We observed no solvent dependence on the spectrum of lb. From the previous discussion the visible absorption band position and Soret splitting are a function of the electron withdrawing strength of the peripheral substituent. Intuitively it is expected that a H-bonded -CHO is more electron withdrawing than a "free" CHO. We believe the spectral difference of 4b in CH2C12 vs. THF are principally due to intramolecular hydrogen bonding between -CH0 and the 7-OH group, where in THF the OH primarily interacts with solvent. Similar red shifting has been observed for copper porphyrin a on addition of H-bond donors in CH2C12.2b Soret region resonance Raman spectroscopy also offered a deeper insight on the H-bonding effects on 4. The highest frequency vibrations of the aldehyde 4b in 1 CHZCl2 and THF occur at 1656 and 1663 cm- , respectively. These are assigned to the v stretching frequencies by CO analogy to formyl substituted metalloporphyrins.124 The 7 cm-1 variation in VCO between CHZCl2 and THF represents a hydrogen bond strength of approximately 1 kcal/mole, 151 .mmommo cw av ou mcmpwuhm mo :oHuwopm mmezmoup com: pm>uomno mumenm muuommm GS“ m30r~m “meH .AIIIV mmommo can AIIIV mee am am no muuommm :omumu0mnm coo com 00¢ mm on awe Imus; .mmum musmmm 152 which predicts an Optical red shift of 93 cm-l.124a This is in qualitative agreement with out observations. The resonance Raman spectra of aldehyde 4b in CHZCl2 and THF (Figure 5-15) show that the n electron 1 density sensitivity marker band at 1373 cm- ) 1 (V4, A19 and core-size marker bands at 1641 cm- 1582 cm"1 (v2, A (v10. Blg)’ -1 3, AZg) and 1507 cm (v3, A ) lg; “1 are not affected by hydrogen bonding at the ring 19 periphery. Vibrations that are affected are: i) 1552 cm-1 1 (THF), ii) 1451 cm'1 1 (CH2C12) to 1543 cm- 1 (CHZClZ) to (THF), iii) 1352 cm“ 1 l 1469 cm- (CH2c12) to 1342 cm- 1 (THF), and iv) 1150 cm' (CH2C12) to 1141 cm‘ (THF). These vibrations may involve C8-C8, CB-Cs stretching character or CHO bending,125 however, further analysis is not possible at the present time. The mode observed at 1622 cm_1 is assigned to the C=C stretch of the acroleinyl substituent by analogy to similar group frequencies.126 As with aldehyde 4b Schiff's base 4c also forms an intramoleclar hydrogen bond. In the absorption spectrum 1 red shift (Figure 5-16) this is reflected as a 225 cm- of the visible band and an overlapped Splitting of the Soret. Like the aldehyde, the Schiff's base ring vibrations of the Soret region resonance Raman spectrum are not significantly perturbed (Figure 5-15). The major changes upon Schiff's base formation is the disappearance of the C=O frequency at approximately 1660 cm-1 and the C=C 153 Figure 5-15. Soret region resonance Raman spectra of aldehyde 4b and Schiff's base 4c in CHZClZ and THF with 406.7 nm laser excitation. 154 2 0 ll mm : fr HIC H mm _ rt MW =~mI 4. =uOI a \A fiamI .2uI .uunI. .uflm- .umal * .5u7. .uoq.. .uun- .mmm- .amnI . .oav. .mumn .ONN I NM 5mm: . Hulk.» m 53 I .50: :00 I .85 I final .23 I Go» I 5qu 59.7 t .2 3.3. I 38 I Gnu .I I 00 I .muoI an» I 58! on t 3.... . mm m m le.» 4 a... HIC T _ 2!: .=:. =qu .mnou a... .mmmI .maou .mam- aGI afin .uamI 500.. .umuu .37 an? t .9 1.3 3mm- .AmoI * w. .OOOI 50ml .uaaI no .uoau .3»- .aa». I .aauI .moo ._um I m. * .3 0‘! ”()C) I000 I200 I300 |500 I400 Aficm" I I600 I700 >._..mzm._.z. Zcoo can AIIIV has cw 0v mmmn m.mww£om H>usnlc mo muuoomm GOquHOmnd n:Umw _ .AIIIV mmommo mmw nxw “HUS .mmxm vmsmmm 156 vibration at 1622 cm.1 with the appearance of a band at approximately 1630 cm-1.128 This new vibration is assigned to the C=NR vibration in analogy to previously reported assignments.110a Due to overlap with a band at approximately 1640 cm”1 the exact position of the C=NR vibration was not possible to determine. Comparison of the 1640-1630 cm.1 region of SB-4c in CH2C12 and THF reveals that in THF the overlapping bands are more intense and narrower- than in CH2C12. We interpret this as an increase of VC=N due to an intramolecular hydrogen bond to the 7-OH group in CHZClz. Other evidence is obtained in the 1150-1130 cm-1 region. For CH0 and CHNR there are two 1 bands in CH2C12 (C30: 1150, 1128 cm- , CHNR+: 1155, 1138 cm-l), while in THF these bands apparently collapse (CHO: 1141, 1131 cm‘1 1 , CHNR: 1138 cm- ). The other solvent sensitive bands of CHO are not very much perturbed fOr CHNR . Besides solvent and anion effects demonstrating the variability of the spectral shifts associated with Schiff's base protonation, further evidence for separation of the in-plane polarized transition dipoles is also provided. The effect of solvent and anion on the Soret region in general is that the position of the higher energy Soret component remains essentially invariant while the lower energy component parallels the shifts observed by the visible region absorbance maximum (see Figures 5-11, 5-12, 5-13 and Tables 5-2, 5-3, 5-4). Since the lower 157 energy Soret component is sensitive to ion pairing, solvation and hydrogen bonding, it seems logical to conclude that the lower energy Soret band is predominantly the in- plane polarized transition dipole along the protonated Schiff's base axis, while the invariability of the higher energy Soret band suggests it is the other polarization transition dipole. These assignments were confirmed by resonance Raman spectroscopy (above) for 4C‘HC1 and 4d°C104 . Redbx Potentials: Electrochemical redox potentials of porphyrinoid compounds often give useful information concerning the orbital energies of the system since, to a first approximation, the redox span of the monocation to mono- anion radical formation corresponds to the HOMO-LUMO energy gap.129 Unfortunately, cyclic voltammograms for 4c, 4c-HX or 4d gave only ill-defined redox waves. Therefore, the aldehyde 4b, ethyl cyanoacetate 4e, and malononitrile 4d adducts were investigated as models. The CV of these compounds in THF are reproduced in Figures 5-17 to 5-19. The reversible formation of cation and anion radical species can be readily observed. The apparently coupled irreversible waves at -ll60 and 40 mV for 4e and at -1150 and 130 mV for 4f, which were absent in 4b, are probably due to redox reactions occurring at the ring periphery. 158 -I3OO -l020 I 760 < 1 2 I CHO F I I I I I I000 500 o -500 -I000 mV vs.SCE Figure 5-17. Cyclic voltammogram of 4b in THF containing 0.1 M tetrabutylammonium perchlorate (TBAP). Scan rate was 100 mV/sec. 159 —850 790 T -I460 .3 I .. CHCICNICOZEt I I I F I I I000 500 o -500 -I000 vas.SCE Figure 5-18. Cyclic voltammograms of ethyl cyanoacetate adduct 4e in THF containing 0.1 M TBAP at a scan rate of 100 mV/sec. Scan direction was reversed after first, second and third reduction wave (top to bottom respectively). Figure 5-19. -7IO -II5O 1 In -I430 I30 820 CHCICNI2 I I I I I I I000 500 O -500 -IOOO mV vs. SCE Cyclic voltammograms of malononitrile adduct 4f in THF containing 0.1 M TBAP (100 mV/sec). Scan direction was reversed after first (top), second (middle) and third (bottom) reduction wave. 161 While it may be difficult to ascertain to what extent coulombic forces will perturb the redox potentials of SBH+, the model studies indicate it seems certain that SBH+ will be easier to reduce but somewhat harder to oxidize than the parent aldehyde (Fig. 5-20). As is the case with mono-formyl vs. di-formyl porphyrins,114a the effect of increasing the electron withdrawing power of the substituents serves to decrease the HOMO-LUMO gap, indicating the presence of a strong resonance effect. Also similar to formyl porphyrins, the electron withdrawing substituents seem to mostly affect the energy of the LUMO. This has been interpreted as an evidence that the substituent lowest vacant n* orbital lies very close to the macrocycle lowest n* orbital to allow mixing.114a Therefore, protonation of Schiff's bases further splits the excited states degeneracy and lowers the fi* orbital energies, but leaving the n orbitals essentially unaltered. Indeed, resonance Raman and NMR spectra of lc-HC1 (above and reference 110a) showed that the ground state is not very much perturbed. Detailed theoretical discussions of orbital structures are given in reference 112. Summary By measuring the optical spectra of derivatives produced by reversible modifications at the -CH0 group on porphyrinoid macrocycles in conjunction with proton NMR and resonance Raman spectroscopies, we have shown that -1.5 -1.0 X\ 1.0 Figure 5-20. CHO 162 -~- ~ ~— CHCICNICOZEt CHCICNI —-- ~~u — -——_ __—- 2 I Half-wave redox potentials of aldehyde 4b, ethyl cyanoacetate adduct 4e and malononitrile adduct 4f (measured in THF VS. SCE). 163 the unusual red shift of absorption maxima in the visible region and the Soret band splitting observed for Schiff's base porphyrins and chlorins upon protonation is due to the resonance effect of a strong electron withdrawing group on the ring periphery, not because of a delocaliza- tion of the positive charge onto the ring. We have further demonstrated that the conversion of a carbonyl to a Schiff's base peripheral group would subject the spectral properties of chlorin and porphyrin to a greater degree of environmental control than otherwise possible. In view of the large variations in the visible absorption maxima of photosynthetic chlorophylls, our result would certainly make a Schiff's base chlorophyll an interesting model for reaction centers. Even more intriguing is the fact that there are large differences in oxidation potentials of P700 and P680. Titrations of P700 yield a midpoint potential ranging between +0.4 and +0.5 V versus the normal hydrogen electrode (NHE) whereas the minimum potential needed to oxidize water to oxygen at physiological pH sets a lower limit of +0.8 V for P680, which makes an electron withdrawing system such as 4c very attractive. Further investigations, particularly EPR and electrochemical studies of Schiff's base chlorophylls are needed to verify the validity of these proposals. 164 Materials and Methods Visible spectra were recorded on a Cary 219 spectro- photometer interfaced to a Bascom—Turner recorder. Spectra shown here were recalled directly from flOppy diskettes. NMR spectra were obtained using a Bruker WM-250 instrument. Elemental analyses were performed by Spang; C, H, N analyses were within 0.5%. Cyclic voltammetry was performed using a Bioanalytical Systems CV-lA unit or a Pine Instrument RDE-3 potentiostat in a specially constructed glass cell which contains two platinum spherical electrodes sealed through the cell wall. All measurements were carried out in THF containing 0.1 M tetrabutylammonium perchlorate at a scan rate of 100 mV/sec. Resonance Raman spectra were recorded using a Spex 1401 double monochromator and the associated Ramalog electronics. Laser excitations at 406.7 and 488.0 nm were obtained with a Spectra Physics 164-ll krypton ion laser equipped with a high-field magnet. Incident powers were 20-40 mW. All spectra were collected at 90° scattering geometry at room temperature. To ensure no sample decomposition, optical spectra were recorded before and after each RR experiment. Materials: CHZClZ' from CaH, tetrahydrofuran from lithium aluminum hydride CH3CN and triethylamine were freshly distilled before use. Pyrrolidine hydroperchlorate and hydrobromide were prepared by addition of the concentrated acid to 165 pyrrolidine in THF until the solution was just acidic enough to wet pH paper. Water was azeotroped out with benzene on a rotary evaporator followed by three crystallizations from THF/ethyl acetate. Pyrrolidine hydrochloride was prepared by bubbling an etheral solution of pyrrolidine with anhydrous HCl followed by crystallizations (3X) from THF/ethyl acetate. All other commercially obtained chemicals were used without further purification. Nickel 2,6-di-n:pentyZ-4-v§gyZ-8-fbrmyZ-1L5;QLZ-tetramethyl- gogghine (1b): Chlorin 4a (vide infra, 100 mg) in CH2C12 was reduced by addition of sodium borohydride (50 mg) in (100 ml) methanol (2 mL) followed by quenching with dilute acetic acid after 5 min. The resultant diol porphyrin was dissolved in pyridine (100 mL) followed by addition of aqueous sodium periodate (5%, 30 mL), heated on a steam bath for 30 min, cooled, diluted with CH C12 and extracted 2 with 15% aqueous HC1. The crude product was purified by column chromatography using silica gel, crystallized from CHZClz-MeOH and characterized by NMR and UV—vis spectro- scopies. Yield of la: 60 mg. Nickel insertion was accomplished by refluxing 1a (60 mg) in CH ClZ/MeOH with 2 excess Ni(OAC)2 for approximately 2 hours followed by crystallization from CH2C12/MeOH. Yield 1b: 55 mg. NMR 6(CDC13) pentyl: 0.89 (6 H, m), 1.95 (4 H, m), 3.53 (4 H, m); ring Me: 3.16 (3 H, s), 3.20 (3 H, s), 166 3.37 (3 H, s), 3.42 (3 H, 3); vinyl: 6.09 (2 H, m), 7.94 (l H, m); meso: 8.97 (l H, s), 9.13 (l H, s), 9.29 (l H, s), 9.96 (1 H, s); CH0: 11.06 (1 H, s). Amax (emM) 1n CH2C12 589(21). : 409 nm (114), 516(5.9), 542(6.9), Nickel 6,7-di-nepentyl-llfl-difbrmyl-§,§,5,8-tetrametflyl- porphine (2b) and Nickel 2,6-di-nepentyl-4L8-difbrmyl-JJ3,5,7- tetramethylpgrphine (3b): The appropriate divinyl porphyrin114 (100 mg) dissolved in pyridine (100 mL) was added to a solution of osmium tetroxide (100 mg) in pyridine (10 mL) and stirred at room temperature for 1 hr. To this was added an aqueous sodium sulfite solution (15%, 30 mL) and heated on a steam bath for 30 min. followed by partition between CHZClz/HZO‘ The porphyrin glycols were then oxidized with sodium periodate and purified as with la. The yield for either porphyrin: ~80%. Nickel insertion was accomplished as with la except that overnight reflux was necessary for completion. NMR 2b 6(CDC13)pentyl: 1.01 (6 H, t), 1.58 (8 H, m), 2.00 (4 H, m), 3.52 (4 H,t); ring Me: 2.52 (6 H, s) 3.52 (6 H, s); meso: 8.82 (l H, s), 9.39 (2 H, 3), 10.54 (1 H, s); CH0: 10.53 (2 H, s). Amax (e ): 423 nm (125), 535(ll.5), 578(21). NMR 3b mM 5(CDC13)pentyl: 0.92 (6 H, t), 1.55 (8 H, m), 1.90 (4 H, m), 3.5 (4 H, m); ring Me: 3.10 (6 H, s), 3.42 167 (6 H, s); meso: 8.83 (2 H, s), 9.76 (2 H, s); CH0: 11.1 (2 Ii, s). l (a max mM>= 412 nm (119), 509(5.5, 607(35). 2,6-di-Q:pentyl-4-vinyl-7-hydrogyl-8—acroleinyl-1L3,5L7- tetramethylchlorin (4a) and 2,6-di-nepentyl-3l7-dihydrcmy: 4‘1 8-diacro leigy l- 1, g, 5, 7— te frame thglbczcteriochlcrin (5a) : 2,6-Di-n-pentyl-4,8-diviny1-l,3,5,7-tetramethyl- porphine9 (200 mg) in CH2C12 was photolyzed with aeration for 30 min in a water cooled photolysis apparatus with a 250 W tunsten halogen lamp. The reaction mixture was then concentrated and chromatographed on silica gel. CH2C12 elution afforded unreacted divinyl porphyrin, followed by the monooxygen adduct then the trans-di- adduct. The gis-di-adduct was obtained by elution with 5% MeOH-CH2C12. Pure epimeric mono-adduct and gis- di-adduct were crystallized from MeOH-CH2C1 NMR of the 2. chlorin is obtained (90 mg) in CDC13: 6 pentyl: 0.89 (3 H, t), 9.06 (3 H, t), 1.5 (4 H, m), 1.98 (2 H, q), 2.12 (2 H, q), 3.7 (4 H, m); ring Me: 1.60 (3 H, s), 3.22 (3 H, s), 3.45 (3 H, s), 3.51 (3 H, s); OH: 2.9 (l H, b); vinyl: 6.18 (2 H, m), 8.1 (1 H, m); =CHCHO: 6.8 (1 H, d), 10.2 (1 H, d); meso-H: 8.17 (1 H, s), 8.56 (l H, s), 9.60 (1 H, s), 9.72 (1 H, 5); NH: -3.43 (1 H, s), -3.61 (l H, s), Amax (EmM (24), 391(80), 411(91), 423(89), 504(6.l) 568(18), 601(7.9), in THF): 336 nm 660(46). The cis-bacteriochlorin 2a 5(CDC13)penty1: 0.88 (6 H, t), 1.5 (16 H, m), 1.93 (4 H, q), 3.76 (4 H, m); 168 ring Me: 1.63 (6 H, s), 3.14 (6 H, s); 0H: 5.84 (2 H,s); =CHCHO: 7.14 (2 H, d), 10.58 (2 H, d); meso-H: 8.30 (2 H, s), 8.55 (2 H, 5); NH: -4.41 (2 H, s); xmax (e in THF): 350(23), 419(55), 443(75), 581(12), 659(6.1), mM 692(5.6), 729(70). The transisomer had identical spectral Properties. Copper was inserted by standard procedures.27b Schiff's Base Formation: i) Schiff's bases ic, 2c, and 3c: Nickel formyl porphyrins lb, 2b and 3b were refluxed in benzene containing excess n-butylamine for 3 hr. Water produced was removed by allowing the condensate to filter through a silica gel pad prior to returning to the flask. Lypholyzation afforded pure 1c, 2c and 3c. The Schiff's bases were each characterized by NMR and visible spectroscopies. NMR 1c 5(CDC13)pentyl: 0.94 (6 H, t); 1.50 (8 H, m); 2.09 (4 H, m), 3.71 (4 H, m); butyl: 1.18 (3 H, t), 1.63 (2 H, m), 1.82 (2 H, m), 4.07 (2 H, t); ring Me: 3.29 (3 H, s), 3.34 (3 H, s), 3.47 (3 H, s), 3.54 (3 H, 3); vinyl: 6.08 (2 H, m), 8.06 (l H, m); Meso: 9.50 (3 H, m), 9.65 (1 H, s); CHN: 10.64 (1 H, s), A (6 max mM (149), 515(4.6), 538(5.5), 577(16). NMR 2c 6(CDC13) ): 404 nm pentyl: 0.96 (6 H, t), 1.52 (8 H, m), 2.09 (4 H, m), 3.72 (4 H, t); butyl: 1.20 (6 H, t), 1.65 (4 H, m), 1.83 (4 H, m), 4.07 (4 H, t); ring Me: 3.34 (6 H, s), 3.42 (6 H, s); meso and CHN: 9.38 (1 H, s), 9.42 (3 H,s) 10.58 (2 H, 3). NMR 3c 6(CDC13) pentyl: 0.96 (6 H, t), 169 1.50 (8 H, m), 2.08 (4 H, m), 3.73 (4 H, t); butyl: 1.18 (3 H, t), 1.64 (2 H, m), 1.83 (2 H, m), 4.06 (2 H, t); ring Me: 3.25 (6 H, s), 3.56 (6 H, s); meso: 9.51 (2 H,s), 9.54 (2 H, s); CHN: 10.68 (2 H, s), Xmax (EmM): 407 nm (162), 515(7.2), 541(8.7), 584(29). ii) Schiff's base 4c: To 4b (2 mg) in CHZCl2 (3 ml) was added 5 drOps n-butylamine and allowed to stand 15 minutes followed by evaporation under a stream of dry argon. The absorption spectrum was identical to that from (iii) in CH C12 or THF. 2 iii) Schiff's base by spectrophotometric method: To an ~1o'5 M solution of 1b, 2b, 3b (CH2C12) 4b or 5b (THF) was added 1 drop of n-butylamine and then 1 mL of air equilibrated over conc. HCl was bubbled through the solution. Reactions were complete within 10 min. Isosbestic points: 1b to lc (nm); 375, 409, 500, 582; 2b to 2c: none; 3b to 3c: none; 4b to 4c (nm): 395, 436, 5 500, 591, 612, 635; 5b to Sc: none. To 4b (~1o' M) in CH2C12 of n-butylamine in 5 mL CH2C12 and a catalytic amount of was added 2 drops of a solution containing 3 drops HC1. Reaction required about 2 hr for completion. Isosbestic points: 396, 445, 517, 647 nm. Pyrrolidinium Salt (4d): To 5 mg 4b in CH2C12 (5 mL) was added 1 equivalent of pyrrolidine hydroperchlorate and 1 drop of trimethyl orthoformate, allowed to stand 48 hours at room 170 temperature, diluted with benzene and lypholyzed. The UV-vis spectrum was identical to that obtained below. To ~10-5 M 4b in CH2C12 or THF was added a couple of crystals of pyrrolidine HX and the spectrum monitored. Isosbestic points (THF): 4b to 4d-C104: 383, 459, 538, 662 nm. Malononitrile adduct (4f): i) To ~10"5 M 4b in THF was added 1 drop of malo— nonitrile and 1 drop of triethylamine. Reaction was complete within 10 min. Isosbestic points: 385, 457, 564, 659 nm. ii) 7 mg 4a was dissolved in 30 mL THF. To this solution 4 drops of malononitrile and 3 drops of triethyl- amine were added. This was refluxed 2 hr followed by dilution with ether. The ether phase was extracted with 20% acetic acid (4X), washed with H 0 (2X), brine (2X), 2 dried over anhydrous NaZSo4, and evaporated in vacuo; yield: quantitative. NMR 6(CDC13)penty1: 0.89 (3 H, t), 1.07 (3 H, t), 1.3 (4 H, m), 1.6 (6 H, m), 2.3 (2 H, m), 3.7 (2 H, m), 4.0 (2 H, m); ring Me: 1.27 (3 H, s), 3.44 (3 H, s), 3.58 (3 H, s), 3.68 (3 H, s); OH: 6.7 (l H, 5); vinyl: 6.26 (2 H, m), 6.18 (l H, m); =CHCH=C(CN)2: 2.84 (1 H, d), 7.67 (1 H, s); meso H: 8.26 (l H, s), 9.86 (2 H, s), 9.89 (1 H, s); NH: -3.6 (1 H, s), -4.1 (1 H, s). 171 Ethyl cyanoacetate Adduct (4e): This was prepared as in (i) for malononitrile. Reaction required about 10 hours for completion. Isosbestic points: 380, 455, 535, 655 nm. Eyrrclidine Hemiaminal: To ~10.5 M lb in THF or CH2C12 was added 1 drop of pyrrolidine. Reaction was completed within 30 minutes. Isosbestic points (THF): 321, ~370, 423, 496, 578, 602, 628 nm. Schifffs Base Protonation/Deprotonaticn: 2C12: To ~10.5 M 1c or 4c in CH2C12 was bubbled air which had been equilibrated over 1) HF, HCl, HBr in CH the respective concentrated acid. (This was easily accomplished by withdrawing the air inside a bottle of acid with a small syringe and then passing the air into the cuvette.) The resultant SB-HCl spectra were identical as in (ii). BFBOEt2 was introduced to SB-HF by bubbling BF3OEt2-saturated air through the solution. 5 M Schiff's base in CH2C12, CH3CN was added dropwise an anhydrous HC1-saturated CH2C12 ii) To ~1o' THF or solution. iii) HI: To ~1o'5 M 1c or 40 in CH c1 2 2 a small amount of HI vapor prepared by adding conc. was injected sulfuric acid to KI. 172 iv) HC1O4: To ~10“5 M Schiff's base in CH2C12, THF, or CH CN was added dropwise a 70% HClO4-saturated 3 methylene chloride solution. v) SBH+ were returned to the original SB by bubbling triethylamine-saturated air through the acidified solution. Borohydride Reduction: To ~10-5 M 4b in CH2C12 was added a couple of crystals of tetrabutylammonium borohydride and the UV-vis spectrum monitored. Isosbestic points: 321, 370, 423, 496, 578, 602, 628 nm. Addition of 2 drops of 1:1:1 CH3OH:TFA:HZO solution yielded a typical copper porphyrin spectrum. Isosbestic points: 313, 345, 369, 409, 544, 557, 528 nm. Amax (Cu porphyrin): 400, 528, 570 nm. REFERENCES AND NOTES 10. REFERENCES AND NOTES Wilkstrom, M; Krab, K.; Saraste, M. "Cytochrome Oxidase", Academic Press:New York, 1981. a. Murphy, M.J.; Siegel, L.M.; Tove, S.R.; Kamin, H. (1974) Proc. Nat'l. Acad. Sci USA 21, 612; b. Vega, J.M.; Garrett, R.H.; Siegle, L.M. (1975) J. Biol. Chem. 250, 7980; c. Vega, J.M.; Kamin, H. (1977) J. Biol.‘EHem. 252, 896. a. Barrett, J. (1956) Biochem. J. 64, 626; b. Yamanaka, T.; Kihimoto, S.; Okunuki, K. (1963) J. Biochem. 53, 416; c. Yamanaka, T.; Ota, A.; Okunuki, K. (1960) Biochem. Biophys. Acta 44, 397; d. Kuronen, T.; Ellfolk, N. (1972) Biochem. Biophys. Acta 215, 308. a. Siegle, L.M.; Murphy, M.J.; Kamin, H. (1973) J. Biol. Chem. 248, 251; b. Murphy, M.J.; Siegle, L.M.; Kamin, H.'TI973) J. Biol. Chem. 248, 2801;” c. Murphy, M.J.; Siegle, L.M. (1973) JT_§iol. Chem. 248, 6911; d. Siegle, L.M.; Davis, P.S.; Kamin, H. (1974) J. Biol. Chem. 249, 1572; e. Murphy, M.J.; Siegle, L.M.; Kamin, H. (1974) J. Biol. Chem. 249, 1610. Chang, C.K. (1980) Biochem. 1971. Young, R.; Chang, C.K.; Ward, 3., manuscript in preparation. See also: Stolzenberg, A.M.; Strass, S.H.; Holm, R.H. (1981) J. Am. Chem. Soc. 103, 4763. Jacob, G.S.; Orme-Johnson, W.H. (1979) Biochem. 2967. Ward, B.; Chang, C.K. (1982).Photochem. Photobiol. 35, 757. Chang, C.K. in "The Biological Chemistry of Iron" Dunford, H.B. et al. (eds.) D. Reide1:Boston 1982, pp. 313-334 and references therein. 173 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 174 Brault, D.; Rougee, M. (1973) Nature (New Biol.) 241, 19. a. White, D.K.; Cannon, J.B.; Traylor, T.G. (1979) J. Am. Chem. Soc. 101, 2443; b. Rougee, M.; Brault, D. (1975) Biochem. 14, 4100. a. Smith, M.H. (1959) Biochem. J. 73, 90; b. Sono, M.; McCray, J.A.; Asakura, T. (1977T J. Biol. Chem. 252, 7475. Wayland, H.B.; Mehne, L.F.; Swartz, J. (1978) J. Am. Chem. Soc. 100, 2379. a. Caughey, W.S.; Alben, J.O.; McCoy, S.; Boyer, S.H.; Charache, 8.; Hathaway, P. (1969) Biochem. 8, 59; b. Chang, C.K.; Dolphin, D. (1976) Proc. Nat'l. Acad. Sci. USA 13, 3338; c. Traylor, T.G.; Campbell, D.; Sharma, V.; Geibel, J. (1979) J. Am. Chem. Soc. 191, 5376; d. Ward, 3.: Wang, C.B.; Chang, C.K. (1981) J. Am. Chem. Soc. 191, 5236. Gansow, O.A.; Vernon, W.D. in "Topics in Carbon-13 NMR Spectroscopy", Levy, G.C. (ed.) Vol. 2, John Wiley:New York, 1976, pp. 270-325 and references therein. Karplus, M.; Pople, J.A. (1963) J. Chem. Phys. 38, 2803. Phillips, J.N. (1960) Rev. Pure & Appl. Chem. 19, 35. Parmely, R.C.; Goff, H.M. (1980) J. Inorg. Chem. 102, 841. Buchler, J.W.; Kokisch, W.; Smith, P.D. (19 ) Structure and Bonding 34, 79._ Geible, J.; Chang, C.K.; Traylor, T.G. (1975) J. Am. Chem. Soc. 91, 5924. Brinigar, W.S.; Chang, C.K. (1974) J. Am. Chem. Soc. 26, 5595. Chang, C.K. (1979) in "Biochemical and Clinical Aspects of Oxygen", Academic Press:New York, pp. 449-451. Nakamoto, K. (1970) "Infrared Spectra of Inorganic and Coordination Complexes", Wiley-Interscience: New York. - 25. 26. 27. 28. 29. 30. 31. 32. 33. 175 Olson, L.W.; Schaeper, D.; Lancon, D.; Kadish, K.M. (1982) J. Am. Chem. Soc. 104, 2042. a. Lancaster, J.R.; Vega, J.M.; Kamin, H.; Orme—Johnson, N.R R.J.; Siegel, L.M .; Orme-Johnson, W.H.; . (1979) J. Biol. Chem. 254, 1268; b. Johnson, M.K.; Thompson, A.J.; Walsh, T.A.; Barber, D.; Greenwood, C. (1980) Biochem. J. 189, 1393. a. Adler, A.D.; Longo, F.R.; Finarelli, J.D.; Goldmacher, J.; Assour, J.; Korsakoff, L. Kreuger, (1967) J. Org. Chem. 32, 476; b. Fuhrrop, J.H.; Smith, K.M. (1975) in_"Porphyrins and Metalloporphyrins" Smith, K.M. (ed.) Elsevier:New York, p. 803. a. Chang, C.K.; Traylor, T.G. Acad. Sci. USA 72 a. Brault, D.; Morliere, P.; Santus, R.; Swallow, A.J. (1980) J. Amer. (1975) Proc. Nat'l. __, 1166; b. Traylor, T.G.; Chang, C.K.; Geibel, J.: Berzinis, A.; Mincey, T.; Cannon, J. (1979) J. Am. Chem. Soc. 181, 6716. Rougee, M.; Land, Chem. Soc. 102, 1015; b. Shafferman, A.S.; Stein, G. (1974T—ECience 188, 428; c. van Leeuwen, J.W.; H. (1979) Biochem. Biophys. .___ Goff, H.; Simic, M.G. (1975) Biochem. Biophys. Acta 888, 201; e. Simic, M.G.; Taub, I.A.; Tocci, J.; Hurwitz, P.A. (1975) Biochem. Biophys. Res. Comm. 62, 161; f. Simic, M.G.; Taub, I.A. (1978) Bioph§§. J. 31, 285. Tromp, J.; Nauta, Acta 577, 394; d. Neckers, D.C. (l9 Corp.:New York. E.J.; 67) "Mechanistic Organic Photo- chemistry", Chapter 7, Reinhold Publishing a. Dolphin, D.; Sams, J.R.; Tsin, T.B.; Wong, K.L. (1976) J. Amer. Chem. Soc. 98, 6970; b. Reed, C.A.; Mashiko, T.; Schel'dt, W.R.; Spartalian, K.; Lang, G. Soc. 102, 2302. a. Castro, C.B.; Robertson, Bioorg. Chem. 3, (1980) J. Amer. Chem. (1977) Faraday—Discuss. Chem. Soc. 88, 270. a. Rao, P.S.; Hayon, E. (1974) J. Amer. Chem. 88, 1287; b. Bansal, K.M.; Henglein, A. (1974) J. Phys. Chem. 78, 160; c. Bansal, K.M.; Henglein, A.; Sellers, R.HT Chem. 18, 569. (1974) Ber. Bunsenges Phys. C.; Davis, H. (1974) 343; b. Simic, M.G.; Taub, I.A. Soc. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 176 a. Kitagawa, T.; Orii, Y. (1978) J. Biochem. 88, 1245; b. Adar, F.; Yonetani, T. (1978) Biochem. Biophys. Acta 502, 80; c. Salmeen, I.; Rimai, L.; Babcock, G.T. (1978) Biochem. 11, 800. VOrkink, W.P.; Cusanovich, M.A. (1974) Photochem. Photobiol. 18, 205. Poff, K.L.; Butler, W.L. (1975) Plant Physiol. 88, 427. Kitagawa, T.; Nagai, K. (1979) Nature 281, 503. Kitagawa, T.; Fukumori, Y.; Yamanaka, T. (1980) Biochem. 18, 5721. a. Lipscomb, J.D.; Sligar, S.G.; Namtvedt, M.J.; Gunsalas, I.C. (1976) J. Biol. Chem. 881, 1116; b. Schmidt, W.; Butler, W.L. (1976) Photochem. Photobiol. 88, 71. Manabe, K.; Poff, K.L. (1978) Plant Physiol. 81, 961. Bocian, D.F.; Lemeley, A.T.; Peterson, N.O.; Brudvig, G.W.; Chan, S.I. (1979) Biochem. 18, 4396. Huber, R.; Epp, 0.; Formanek, H. (1970) J. Mol. Biol. 23: 349. Heidner, E.J.; Ladner, R.C.; Perutz, M.F. (1976) J. Mol. Biol. 104, 707. Norvell, J.C.; Nunes, A.C.; Schoenborn, B.P. (1975) Science 190, 568. Steigemann, W.; Weber, E.; (1979) J. Mol. Biol. 127, 309. Padlan, E.A.; Love, W.E. (1975) J. Biol. Chem. 249, 4067. Peng, S.M.; Ibers, J.A. (1976) J. Am. Chem. Soc. 88, 8032. Hoard, J.L. (1975) in "Porphyrins and Metallo- porphyrins", Smith, K.M. (ed.) Elsevier:New York, p. 351. Moffatt, K.; Deatherage, J.F.; Seybert, D.W. (1979) Science 206, 1035. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 177 Perutz, M.F. (1979) Ann. Rev. Biochem. 88, 327. Caughey, W.S. (1970) Ann. N.Y. Acad. Sci. 174, 148 O Perutz, M.F. (1976) Br. Med. Bull. 88, 195. Collman, J.P.; Brauman, J.I.; Halbert, T.R.; Suslick, K.S. (1976) Proc. Nat'l. Acad. Sci. USA 18, 333. Collman, J.P.; Brauman, J.I.; Doxsee, K.M. (1979) Proc. Nat'l. Acad. Sci. USA 18, 6035. Traylor, T.G.; Berzinis, A.P. (1980) Proc. Nat'l. Acad. Sci. USA 11, 3171. Geibel, J.; Cannon, J.; Campbell, D.; Traylor, T.G. (1978) J. Am. Chem. Soc. 100, 3575. Traylor, T.G.; Campbell, D.; Tsuchiya, S. (1979) J. Am. Chem. Soc. 101, 4748. Battersby, A.R.; Hamilton, A.D. (1980) J. Chem. Soc., Chem. Commun., 117. Collman, J.P.; Brauman, J.I.; Collins, T.J.; Iverson, B.; Sessler, J.L. (1981) J. Am. Chem. Soc. 102, 2450. a. Traylor, T.G.; Michell, M.J.; Tsuchiya, S.; Campbell, D.H.; Stynes, D.V.; Koga, N. (1981) J. Am. Chem. Soc. 188, 5234; b. Collman, J.P.; Brauman, J.I.; Iverson, B.L.; Sessler, J.L.; Morris, R.M.; Gibson, Q.H. (1983) J. Am. Chem. Soc. 188, 3052. Busch, D.H.; Zimmer, L.L.; Gryzbowski, J.J.; Olszanski, D.; Jackels, S.; Callahan, R.; Christoph, G.G. (1981) Proc. Nat'l. Acad. Sci. USA 18, Copper(II) porphyrins do not coordinate any ligand and cannot be reduced to Cu(I) state; instead they reduce to Cu(II) porphyrin anion radicals below -1.4 V (SCE). a. 5,6,7,8-Tetrahydroimidazo[1,5-a]pyridine was synthesized by hydrogenation of 2,3a-diazaindene (Bower, J.D.; Ramage, G.R. (1955) J. Chem. Soc., 2384);b. 1,5-Dicyclohexylimidazole was a generous gift from Professor T.G. Traylor. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 178 However, we did observe a positive shift in l' as a function of N-methylimidazole concentration for FeSP-13 and FeSP-l4. At N-MeIm concentration ranging from 0.1-3.0 M, FeSP-13 had an 1' increase of about 200%. On the other hand, CuFe-4 had a corresponding shift of less than 20% (0.1-1.0 M). This indicates that with 1' less than about 10“, equilibria on the proximal side (base-off37) present a somewhat serious problem. For this reason, any 02 vs. CO comparisons for FeSP-13 and FeSP-14 are probably not reliable and therefore, are not discussed. With 1' greater than 2><10”, proximal base equilibria should present little difficulty. Indeed, 1' of FeCu-5 or FeSP—lS is unaffected by base concentration. Antonini, B.; Brunori, M. (1971) in "Hemoglobin and Myoglobin in Their Reactions with Ligands", North-Holland:Amsterdam, p. 219. Traylor, T.G.; Stynes, D.V. (1980) J. Am. Chem. Soc. 102, 5938. Seybert, D.W.: Moffat, K.: Gibson, Q.H.; Chang, C.K. (1977) J. Biol. Chem. 252 4225. Traylor, T.G.; White, D.K.; Campbell, D.H.; Berzinis, A.P., J. Am. Chem. Soc., in press. Hatada, M.H.; Tulinsky, A.; Chang, C.K. (1980) J. Am. Chem. Soc. 102, 7115. Szabo, A., (1978) Proc. Nat'l. Acad. Sci. 88, 2108. Ortho-substituted TPP's cannot be compared directly since kinetically they behave differently from nomal hemes or even unsubstituted TPP (Traylor, T.G. and Hambright, P.; private communication). Traylor, T.G. (1981) Acc. Chem. Res. 18, 102. a. Sono, M; Asakura, T. (1975) J. Biol. Chem. 250, 5227; b. Asakura, T.; Sono, M. (1974) J. Biol. Cfien. 249, 7087; c. Mamoru,fr,;Woodrow III, G.V.; YonetaHIT T. (1973) Biochem. BiOphys. Acta 317, 34; d. Antonini, T.; Brunori, M.; Caputo, AT?— Chiancone, B.; Fanelli, A.R.; Wyman, J. (1964) Biochem. Biophys. Acta 88, 284; e. Seybert, D.W.; Moffat, K.; Gibson, Q.H. (1975) Biochem. Biophys. Res. Comm. 88, 43; f. Antonini, E; Gibson, Q.H. (1960) Biochem. J. 76, 534; g. Sono, M.; Smith, P.D.; McCray, J.A.;—Asakura, T. (1976) J. Biol. Chem. 881, 1418. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 179 Seybert, D.W.; Moffat, K.; Gibson, Q.H. (1976) J. Biol. Chem. 251, 45. Iizuka, T.; Watanabe, K.; Makino, R.; Sakaguchi, K.; Mitani, P.; Isimura, Y.; Kawabe, K.; Yoshida, Z.; Ogoshi, H. (1982) "Oxygenases and Oxygen Metabolism", Academic Press:New York, pp. 445-450. Jue, T.; Krishnamoorthi, R.; Lamar, G.N. (1983) J. Am. Chem. Soc. 105, 5701. Wallace, W.J.; Houtchens, R.A.; Maxwell, J.C.; Caughey, W.S. (1982) J. Biol. Chem. 257, 4966 and references therein. a. Yonetani, T.; Asakura, T. (1969) J. Biol. Chem. 244, 4580; b. Teale, F.W.J. (1959) Biochem. Biophys. Acta 88, 543. a. Dinello, R.K.; Chang, C.K. (1978) "The Porphyrins" Dolphin, D. (ed.) Vol. 1, Academic Press:New York, p. 305; b. Chang, C.K. (1981) J. Org. Chem. 88, 4610. Falk, J.E. (1964) "Porphyrins and Metalloporphyrins", Elsevier:New York, p. 28. Perutz, M.F. (1980) Proc. R. Soc. London, Ser. B 208, 135. Autsin, R.H.; Beeso, K.W.; Eisenstein, L.; Fraunfelder, H.; Gunsalus, I.E. (1975) Biochem. 18 5355. Alberding, N.; Chan, S.S.; Eisenstein, L.; Fraunfelder, H.; Good, D.; Gunsalis, I.C.; Nordlund, T.M.; Perutz, M.F.; Reynolds, A.H.; Sorenson, L.R. (1978) Biochem. 18, 43. Baldwin, J.M.; Chothia, C.J. (1979) J. Mol. Biol. 129, 175. Doster, W.; Beece, D.; Bowne, S.F.; Dilorio, E.E.; Eisenstein, L.; Fraunfelder, H.; Reinisch, L.; Shyamsunder, E.; Winterhalter, K.H.; Yue, K.T. (1982) Biochem. 88, 4831. Traylor, T.G.; Campbell, D.; Sharma, V.; Geibel, J. (1979) J. Am. Chem. Soc. 101, 5376. Shaanan, B. (1982) Nature 296, 683. Phillips, S.E.V.; Schoenborn, B.P. (1981) Nature 292, 81. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 180 Young, R.; Chang, C.K., manuscript in preparation. McClellan, A.L. "Tables of Experimental Dipole Moments", Rahara Enterprises:El Cerito, 1974. t-Butyl t-Butyl Benzene 0.51 3,5-CH20M3 . MeOMe 1.35 EtOEt 1.27 PhCHZOCHZPh 3,5-CH20H PhCHZOH 1.67-1.73 3,5-C02Bu PhCOzEt 1.85-1.93 3,5-CONHBu PhCONHEt 3.6 3,5-CONEt2 PhCONMe2 3.8-3.9 3,5-CONiPr2 PhCONMe2 Dippy, J.F.J. (1939) Chem. Rev. 88, 151. Glasstone, S.; Laidler, K.J.; Eyring, H. (1941) McGraw Hill:New York. Hushimoto, T.; Basolo, F. (1981) Comments Inorg. Chem. 1, 199 and references therein. a. Collman, J.P.; Gagne, R.R.; Reed, C.A.; Robinson, W.T.; Rooley, G.A. (1974) Proc. Nat'l. Acad. Sci. USA 81, 1326; b. Jamesson, G.B.: Rodley, G.A.; Robinson, W.T.; Gagne, R.R.; Reed, C.A.; Collman, J.P. (1978) Inorg. Chem. 18, 850. Momenteau, M.; Lavalette, D.J. (1982) J. Chem. Soc., Chem. Commun., 341. a. Collman, J.R.; Brauman, J.I.; Halbert, T.R.; Suslick, K.S. (1976) Proc. Nat'l. Acad. Sci. USA 88, 3333; b. Barlow, C.H.; Maxwell, J.C.; Wallace, W.J.; Caughey, W.S. (1973) Biochem. Biophys. Res. Commun. 88, 91; c. Maxwell, J.C.; Volpe, J.A.; Barlow, C.H.; Caughey, W.S. (1974) Biochem. Biophys. Res. Commun. (1974) 88, 166; d. Maxwell, J.C.; Caughey, W.S. (1974) Biochem.Bi0phys. Res. Commun. 88, 1309. Pin, S.; Alpert, B.; Michalowicz, A. (1982) FEBS Lett. 147, 106. Case, D.A.; Huynh, B.H.; Karplus, M.J. (1979) J. Am. Chem. Soc. 101, 4433. ‘ 181 99. Perutz, M.F. (1965) J. M01. Biol. 18, 646. 100. Phillips, S.E.V. (1978) Nature 273, 247. 101. Moody, G.J.; Thomas, J.D.R. (1971) "Dipole Moments in Inorganic Chemistry", Edward Arnold:London. 102. Hayashi, Y.; Yamada, H.; Yamazaki, I. (1976) Biochem. Biophys. Acta 427, 608. 103. a. Collman, J.P.; Brauman, J.I.; Doxee, K.M. (1979) Proc. Nat'l. Acad. Sci. USA 88, 6035; b. Tucker, P.W.; Phillips, S.E.V.; Perutz, M.F.; Houtchens, R.; Caughey, W.S. (1978) Proc. Nat'l. Acad. Sci. USA 75, 1076; c. Wallace, W.J.; Volpe, J.A.; Maxwell, 37C.; Caughey, W.S.; Charache, S. (1976) Bidchem. Biophys. Res. Commun. 88, 1739; d. Caughey, W.S. (1970) Ann. N.Y. Acad. Sci. 188, 148. 104. Ward, B.; Chang, C.K.; Hunt, K., manuscript in preparation. 105. Minkin, V.I.; Osipov, O.A.; Zhdanov, Y.A. (1970) "Dipole Moments in Organic Chemistry", translated by Hazzard, B.J., Plenum Press:New York, p. 119. 105. a. Katz, J.J.; Norris, J.R.; Shipman, L.L.; Thurnauer, M.C.; Wasielewski, M.R. (1978) Ann. Rev. Biophys. Bioengineer. 8, 393; b. Maggiora, G.M. (1979) Int. J. Quantum Chem. 18, 331; c. Sauer, K. (1979) Ann Rev. Phys. Chem. 88, 155. 107. a. Callahan, P.M.; Babcock, G.T. (1983) Biochem. 22, 452; b. Babcock, G.T.; Callahan, P.M. (1983) Hiochem. 88, 2314. 108. a. Fajer, P.J.; Chang, C.K.; Wang, C.B.; Bergkamp, M.A.; Netzel, T.L. (1982) J. Phys. Chem. 88, 3754; b. Netzel, T.L.; Bergkamp, M.A.; Chang, C.K. (1982) J. Am. Chem. Soc. 188, 1952. 109. a. Wasielewski, M.R.; Norris, J.R.; Shipman, L.; Lin, C.-P.; Svec, W.A. (1981) Proc. Nat'l. Acad. Sci. USA 78, 2957; b. Davis, R.C.; Ditson, S.L.; Fentiman,—A'.F.; Pearlstein, R.M. (1981) J. Am. Chem. Soc. 188, 6823. 110. a. Ward, B.; Callahan, P.M.; Young, R.; Babcock, G.T.; Chang, C.K. (1983) J. Am. Chem. Soc. 105, 634; b. Pearlstein, R.M.; Ditson, S.L.; Davis, R.C.; 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 182 Tentiman, A.F. (1982) Biophys. J. 88, 112a c. Maggiora, L.L.; Maggiora, G.M., J. Am. Chem. Soc., manuscript submitted. a. Lewis, A.; Marcus, M.A.; Ehenberg, E.; Crespi, H. (1978) Proc. Nat'l. Acad. Sci. 88, 4642; b. Motto, M.G.; Sheves, M.; Tsujimoto, K.; Balogh-Nair, V.; Nakanishi, K. (1980) J. Am. Chem. Soc. 188, 7947. Hanson, L.K.; Chang, C.K.; Ward, B.; Callahan, P.M.; Babcock, G.T.; Head, J.D., J. Am. Chem. Soc., manuscript submitted. DiNello, R.K.; Chang, C.K. (1978) in "The Porphyrins" Dolphin, D. (ed.) Academic Press:New York, p. 305. a. Chang, C.K.; Hatada, M.H.; Tulinsky, A. (1983) J. Chem. Soc. Perkin. Trans. II, 371; b. Chang, C.K., unpublished work. March, J. (1968) in "Advanced Organic Chemistry: Reactions, Mechanisms and Structure", McGraw- Hill:New York, p. 667. Abe, M.; Kitagawa, T.; Kyogoku, Y. (1978) J. Chem. Phys. 88, 4526. Shelnutt, J.A. (1981) J. Chem. Phys. 88, 6644. Callahan, P.M.; Babcock, G.T. (1981) Biochem. ‘88, 452. The 3-4 cm.1 increases which are apparent in v2 and in v10 upon formation of the protonated Schiff's base most likely arise from a slight porphyrin core contraction accompanying the protonation reaction.1“ Aton, B.; Doukas, A.G.; Narva, D.; Callender, R.H.; Dinur, U.; Honig, B. (1980) Biophys. J. 88, 79. Leonard, N.J.; Paukstelis, J.V. (1963) J. Chem. Soc. 88, 3021. Felton, R.H.; Yu, N.-T. (1978) in "The Porphyrins", Dolphin, D. (ed.) Vol. 3, Part A, Academic Press:New York, p. 347. 123. 124. 125. 126. 127. 128. 183 a. O'Donnell, T.A. (1973) in "Comprehensive Inorganic Chemistry", Vbl. 2, Trotman-Dickenson, A.P. (ed.), Pergamon Press:New York, p. 1042; b. Emeleus, H.F. (1969) in "The Chemistry of Fluorine and It's Compounds" Academic Press: New York, p. 33; c. The formation of SBH+HF2- is described by the two equilibria: K1 + _ SB-tHF ==== SBH -+F - K2 - HF +F =3 HF2 In order that protonation yields SBH+HF K2>'K1, resulting in an upper limit for SBH+ pKa of ca. 15, (assuming aqueous solution values). For SB4c a third equilibrium is introduced due to the presence of the 77-OH moiety, namely: K 7-OH+F —3- 7-OHF Due to coloumbic attraction between SBH+ and F- it does not seem unreasonable to expect that K3 > K2. a. Babcock, G.T.; Callahan, P.M. (1983) Biochem. 88, 2314; b. Van Steelandt-Frentrup, J.; Salmeen, I.; Babcock, G.T. (1981) J. Am. Chem. Soc. 103, 5981; c. VC=O for as methyl cinnamaldehyde occurs at 1658 cm'1 (CHClz) (Ref. a. Choi, S.; Lee, J.J.; Wei, J.H.; Spiro, T.G. (1983) J. Am. Chem. Soc. 105, 3692; b. Green, J.H.S.; Harrison, D.J. (1976) Spectro. Chemica Acta 888, 1265. VC=C of 0,8 unsaturated aldehydes ranges from from 1618-1695 cm'1 (ref. 127a). a. Dolphin, D.; Wick, A. (1977) "Tabulation of Infrared Spectral Data", Wiley Interscience: New York, p. 37; b. ibid. p. 170. By analogy with cinnamaldehyde where VC=O and “C=C occur at 1674 and 1622 cm"1 respectively. Monitoring n-butyl Schiff's base formation of cinnamaldehyde by IR absorption in CH2C12 (Fig. R-l) resulted in disappearance of VCHO 129. 130. 184 and vC=C with the concomitant appearance of a 1 band at 1632 cm- 1614 cm"1 (C=C). (C=N) and a very weak band at Felton, R.H. (1978) in "The Porphyrins", Dolphin, D. (ed.) Academic Press:New York, Vol. 1, p. 53. a. Govindjee (ed.) (1978) "Bioenergetics of Photosynthesis", Academic Press:New York; b. Bearden, A.J.; Malkin, R. (1975) Quart. Rev. Biophysics 8, 131; c. Evans, M.C.; Shira, C.H.; Slibus, A.R. (1977) Biochem. J. 188, 75. 185 cm”1700 1650 1600 IIIIIIIlI—IIIII ‘%T Figure R-l. Infrared spectra monitoring Schiff's base formation between cinnamaldehyde and n-butylamine in CH2C12. Arrows indicate spectral changes with time.