Preparation and Characterization of and Intramolecular Electron Transfer in a Pentaammineruthenium Derivative of Candida krisei Cytochrome c Citation Selman, Mary (1989) Preparation and Characterization of and Intramolecular Electron Transfer in a Pentaammineruthenium Derivative of Candida krisei Cytochrome c. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/hpxj-n706. https://resolver.caltech.edu/CaltechTHESIS:08262013-151906759 Abstract A semisynthetic binuclear metalloprotein has been prepared by appending the pentaammineruthenium moiety to histidine 39 of the cytochrome c from the yeast Candida krusei . The site of ruthenium binding was identified by peptide mapping. Spectroscopic and electrochemical properties of the derivative indicate the protein conformation is unperturbed by the modification. A preliminary (minimum) rate constant of 170s -1 has been determined for the intramolecular electron transfer from ruthenium(II) to iron(III), which occurs over a distance of at least 13Å (barring major conformational changes). Electrochemical studies indicate that this reaction should proceed with a driving force of ~170mV. The rate constant is an order of magnitude faster than that observed in horse heart cytochrome c for intramolecular electron transfer from pentaammineruthenium(II)(histidine 33) to iron(III) (over a similar distance, and with a similar driving force), suggesting a medium or orientation effect makes the Candida intramolecular electron transfer more favorable. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: Chemistry Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Gray, Harry B. Thesis Committee: Campbell, Judith L. (chair) Baldeschwieler, John D. Gray, Harry B. Hopfield, John J. Defense Date: 1 March 1989 Funders: Funding Agency Grant Number NSF UNSPECIFIED Earle C. Anthony Foundation UNSPECIFIED Record Number: CaltechTHESIS:08262013-151906759 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:08262013-151906759 DOI: 10.7907/hpxj-n706 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 7946 Collection: CaltechTHESIS Deposited By: INVALID USER Deposited On: 27 Aug 2013 15:21 Last Modified: 21 Oct 2021 22:36 Thesis Files Preview PDF - Final Version See Usage Policy. 29MB Repository Staff Only: item control page

PREPARATION AND CHARACTERIZATION OF AND INTRAMOLECULAR ELECTRON TRANSFER IN A PENT AAMMINER UTHENIUM DERIVATIVE OF CANDIDA KRUSE/ CYTOCHROME c Thesis by Mary Selman In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1989 (Submitted March 1, 1989) -11- Acknowledgments First and foremost, I Harry B. Gray, for support years. I am especially grateful his would like and to to thank my encouragement him for research over allowing me advisor, these several to set my own pace and standards for the work presented here. The collaboration development of acknowledged. at Caltech, their the of Cindy St. Clair FMOC amino acid in the analysis initial method stages is of gratefully I would also like to thank Cindy, and many other women especially friendship Emily and Carter, Adel encouragement. Naylor, Adel and also Kathy deserves Kanes, for mention for helping me understand the ins and outs of Biograf, and for proofreading much of this thesis. I am grateful to Walther Ellis for introducing me to the tech- niques of electrochemistry. Certain members of the support staff deserve special mention: I am indebted to Tom Dunn and Dave Malerba for their patience and expertise in answering numerous questions like thank related to electronics and com- puters, respectively. I would also to my parents and friends, Bernhard, Monica, Susan, Keith, and Sunny and Jim for ment over the years. I generous hospitality and thank Tio Antonio love, and for and providing Tia especially their encourageRebeca for their us with some semblance of family in Southern California. And has finally, generously mathematics, I shared and thank his my husband, knowledge electronics. As a of As Fernando. computers, friend, he a scientist, programming, has given me he physics, suppo rt, -111- encouragement, and love. Thanks, chum! This work was supported by grants from the National Science Foundation. An Earle C. Anthony Graduate academic year is gratefully acknowledged. Fellowship for the 1982-1983 -ivAbstract A semisynthetic appending the cytochrome c binding trochemical tion is metalloprotein pentaammineruthenium from was binuclear the identified properties by of peptide the The krusei. site indicate the modification. A preliminary 170s- 1 has been determined for the from ruthenium(II) of least 13A at chemical studies to (barring iron(III), major this reaction -170mV. The rate constant horse heart force of faster than that observed in electron transfer from iron(III) (over suggesting a a similar medium of ruthenium and occurs conforma- (minimum) is an over a cytochrome electron Electro- proceed order and orientation intramolecular electron transfer more favorable. with effect a with a of magnitude c for similar makes rate distance changes). should elec- intra- pen taa mmi neru then i um(II)(his tidi ne distance, or the of intramolecular conformational that driving molecular which indicate 39 protein the transfer by Spectroscopic mapping. derivative prepared histidine by of force), to been unperturbed constant to moiety Candida yeast has the 33) driving Candida -v- Table of Contents Acknowledgments 11 Abstract iv List of Figures Vll List of Tables Xll List of Abbreviations XIV Chapter 1. Introduction 1. Introduction 2 1.1 Theoretical Background 4 1.2 Experimental Approaches to the Study of Elec- tron Transfer 6 2. Covalent Labeling 2.1. The Experimental System 2.2. Results to Date from 2.3. 10 10 Covalent Labeling Exper- iments 16 Aim of this Project 17 3. Previous Investigations of Medium and Orientation Effects Chapter 2. Chapter 3. Preparation and Isolation of the Cytochrome Derivative 26 29 Materials and Instrumentation 30 Methods 32 Results 37 Characterization of the Derivative 47 Materials and Instrumentation 48 Methods 53 Results 66 -viEPR Spectroscopy 66 NMR Spectroscopy 68 UV -visible Difference Spectroscopy 77 Peptide Mapping 80 Spectroelectrochemistry 90 Differential Pulse Polarography 98 104 Discussion Chapter 4. Chapter 5. Electron Transfer Studies 116 Materials and Instrumentation 117 Methods 120 Results and Discussion 124 Discussion and Conclusion 140 Appendix 1. Amino Acid Analysis Method Development Appendix 2. Tryptic Digestion of the Modified Derivative 156 and Related Studies 184 Appendix 3. Spectroelectrochemical Data 195 Appendix 4. Problems with the Flash Photolysis Instrumentation 200 Appendix 5. Programs for Fitting Kinetic Data 209 Appendix 6. Characterization of the Second Modified Product 225 -viiList of Figures ChaQter 1 Figure 1.1. Energy diagram for photosynthesis Figure 1.2. Potential reactions of the [A 5 Ru(OH 2 )] 2 + with common amino acid side chains 3 complex II Figure 1.3. Cytochrome prosthetic group, protoporphyrin IX 19 Figure 1.4. Cytochrome c prosthetic group (Biograf photo) 20 Figure 1.5. Sequences of horse heart Figure 1.6. Figure 1.7. the cytochromes c Histidine residues of C. krusei cytochrome c from C. krusei and 22 interest for modification in 24 Cytochrome c fold 24 ChaQter 2 Figure 2.1. Figure 2.2. Figure 2.3. Figure 2.4. Elution profiles of commercial C. krusei c from (a) Amberlite CG-50, and (b) CM-52 Time analyses of the modification reaction by trophoresis cytochrome 39 C. krusei cytochrome c polyacrylamide gel elec42 Chromatogram for the elution of the products of the reaction of pentaammineruthenium(II)aquo and C. krusei cytochrome c from CM-52 44 Chromatogram for the analytical FPLC separation of the oxidized modification reaction mixture 46 Front and side view of an optically thin layer spectroelectrochemistry cell transparent 51 Dual-jacketed electrochemical pulse polarography differential ChaQter 3 Figure 3.1. Figure 3.2. Figure 3.3. Figure 3.4. cell Schema tic representation of [A 5 Rupy] 2+/ 3+ as a redox mediator EPR spectra of native cytochrome and modified cytochrome c for 51 the role of 59 c, [A 5 RuHis 3 +], 67 -viii- Figure 3.5. Figure 3.6. Figure 3.7. Figure 3.8. Figure 3.9. Figure 3.10. Figure 3.11. Figure 3.12. Figure 3.13. Figure 3.14. Figure 3.15. Figure 3.16. Figure 3.17. Figure 3.18. Figure 3.19. Figure 3.20. Figure 3.21. NMR spectra of native and A 5 Ru-modified cytochrome c, 66.5-9.5ppm C. krusei krusei ferricyto- Titrating chrome c resonances of native C. 70 73 NMR spectra of native and cytochrome c, 6-38 to -22ppm modified krusei C. 76 Clarification of the discrepancy proton resonance chemical shifts in histidine C-2 78 UV -visible difference spectrum of A 5 Ru-modified krusei cytochrome c minus native cytochrome c Peptides expected upon digestion cytochrome c by submaxillaris protease HPLC chromatograms detection of HPLC chromatograms detection of SMP c. of C. 79 krusei 81 peptides with 220nm 82 SMP peptides with 300nm 85 HPLC chromatogram of SMP peptides modified derivative with 400nm detection of ruthenium86 UV -visible spectrum of ruthenium-containing recovered from the SMP digest peptide 87 Family of visible spectra recorded spectroelectrochemistry experiment on krusei cytochrome c, 24.4 °C. during native N ernst plot for spectroelectrochemistry native C . krusei cytochrome c, 24.4 °C data a C. 91 on 94 Temperature dependence of the reduction potential of the iron site in native and A 5 Ru-modified C. krusei cytochrome c DPP traces cytochrome for native and A 5 Ru-modified C. krusei 100 Temperature dependence of the ruthenium potential in A 5 Ru-C. krusei cytochrome c reduction Residues which are potential ligands mineru theni um in C. krusei cytochrome c pentaam- His26 hydrogen-bonding network to 96 102 106 109 -ixFigure 3.22. Local environments of histidine residues est in C. krusei and horse cytochromes c of inter11 2 Chapter 4 Figure 4.1. Schema tic ment diagram of the flash photolysis instru119 125 Figure 4.2. Flow chart for kinetic data collection and fitting Figure 4.3. Sequence of reactions initiated by a microsecond pulse of 450nm light in the electron transfer 127 experiment Figure 4.4. Stern-Vol~er plot for the quenching [Ru(bpy) 3 ] +• by C. krusei cytochrome c of Postulated mechanism of action of flash photolysis experiment Na 2 EDT A in Flash photolysis chrome c data C. Flash photolysis (112s) of Figure 4.5. Figure 4.6. Figure 4.7. Figure 4.8. Figure 4.9. for native 128 krusei the 130 cyto133 C. krusei modified cytochrome c 135 Flash photolysis of modified (12ms; fit to all data) C. krusei Flash photolysis of modified (12ms; fit to 7-12ms range) C. krusei cytochrome c 136 cytochrome c 138 Chapter 5 Figure 5.1. Possible pathways of electron (His39)-C. krusei cytochrome c transfer in A 5 Ru- 150 Figure 5.2. Reaction of imidazole with diethylpyrocarbonate Figure 5.3. Sequences of cytochromes krusei and S. cerevisiae c Elution profile standard for derivatized Peak area as a za tion reaction function of the pH of the deri vati- from the yeasts 144 C. 154 Appendix 1 Figure A 1.1. Figure A 1.2. the amino acid 168 173 -xFigure A1.3. Figure A1.4. Area counts reaction time as a function of derivatization 175 178 Calibration data for FMOC-amino acid derivatives Appendix 2 Figure A2.1. Figure A2.2. Figure A2.3. Peptides expected upon chrome c with trypsin digestion of C. krusei cyto186 UV -visible spectrum of a from a tryptic digest cytochrome c modified peptide recovered of A 5 Ru(His39)-C. krusei for odd peptide DPP traces A 5R u(His39)-C. krusei cytochrome and tyrosine c recovered tryptic 189 from digest 192 Appendix 4 Figure A4.1. Signal resulting probe lamp from cessation of air flow cooling 203 Figure A4.2. Relationship of output-to-bleeder chain current ratio and light intensity for a typical Hamamatsu photomultiplier tube 203 Figure A4.3. Signal resulting from detection at 549nm flashing a blank sample, with 205 Figure A4.4. Signal resulting from flashing a blank sample after the installation of capacitors in the PMT housing to maintain the voltage difference between the final dynodes 205 Figure A4.5. Signal resulting from 450nm bandpass filters flashing a blank sample using 206 Appendix 6 Figure A6. 1. Chromatogram for oxidized product 2 the analytical FPLC elution of 228 Figure A6.2. EPR spectra of [A 5 RuHis]Cl 3 , native C. krusei cytochrome c, A 5 Ru(His39)-C. krusei cytochrome c, and the second A 5 Ru-cytochrome modification product 229 Figure A6.3. NMR 1H spectra for native, A 5 Ru(His39)-modified, and A 5 Ru-modified (product 2) C. krusei ferricytochromes c, 66-IOppm 232 -xiFigure A6.4. NMR 1H spectra for native, A 5 Ru(His39)-modified, and A 5 Ru-modified (product 2) C. krusei ferricytochromes c, 6-22 to -38 ppm 233 Figure A6.5. UV-visible difference spectrum: A 5 Ru-modified krusei cytochrome c (product 2) minus native krusei cytochrome c c. c. 234 -Xll- List of Tables Chapter I Table 1.1 Results of covalent prosthetic groups) labeling experiments to date (native 13 Chapter 2 Table 2.1. FPLC programs for analytical of the oxidized reaction mixture and preparative separation 38 Chapter 3 Table 3.1. Chemical shifts of titrating histidine C-2 protons 72 Table 3.2. Parameters for least-squares fit of NMR titration data 72 Table 3.3. Total amino acid analyses of c SMP peptides Table 3.4. Table 3.5. Table 3.6. native krusei cytochrome C. 83 Amino acid analysis of a ruthenium-containing from the SMP digest of A 5 Ru-C. krusei cytochrome c Reduction potentials of the iron A 5 Ru-modified C. krusei cytochromes c site in peptide 89 native and 95 Thermodynamic parameters for the reduction of the site i n native and A 5 Ru-modified C. krusei cytochrome c iron 97 Table 3.7. Differential pulse polarography data Table 3.8. Thermodynamic parameters for the reduction of the ruthenium site in A 5 Ru-C. krusei cytochrome c and other environments 103 Table 3.9. Surface areas of tuna cytochrome residue side which are potential pentaammineruthenium ligands krusei cytochrome c 101 chains in C. 108 Chapter 4 Table 4.1. Table 4.2. [Ru(bpy) 3 ]2 +* tryptophan lifetimes Monoexponential cytochrome c in the presence and absence of 131 fits to the kinetic data of modified 139 -xiiiAppendix 1 Table Al.l. Table A1.2. Gradient program reequilibration for FMOC-AA elution and 162 Retention times and amino acid derivatives standard deviations Area counts as a function of derivatization pH Table Al.4. Area counts tion time Table A1.5. a for FMOC 170 Table Al.3. as column function of 172 derivatization reac174 Calibration data centration peak area as a function of con179 Table Al.6. Coefficients for calibration data analysis 180 Table Al.7. Analyses of known peptides 182 Appendix 2 Table A2.1. Gradient peptides program Native C. stry data krusei for the separation of tryptic 188 Appendix 3 Table A3.1. Table A3.2. Table A3.3. Table A3.4. cytochrome c spectroelectrochemi196 Fitted parameters to spectroelectrochemical plots for native C. krusei cytochrome c Modified A 5 Ru(His39)-C. troelectrochemistry data Fitted parameters plots for modified c krusei cytochrome Nernst 197 c spec- to spectroelectrochemical N ernst A 5 Ru(His39)-C. krusei cytochrome 198 199 -xivList of Abbreviations A 5Ru pen taammineru theni urn ASCII American standard code for information interchange bpy bipyridine CM carboxymethylcell ulose DEPC dieth y lpyrocar bonate DPP differential pulse polarography DSS 2,2'-dimethyl-2-silapen tane-5-sulf on ate EDTA ethylenediaminetetraacetic acid FID free induction decay FMOC-Cl 9-fluorenylmethoxycarbonyl chloride FPLC fast protein liquid chromatography FWHM full width half maximum HEPES N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid HPLC high performance liquid chromatography ID inner diameter NaAc sodium acetate buffer NaPi sodium phosphate buffer NHE normal hydrogen electrode OD outer diameter OTTLE optically transparent thin layer electrochemistry PAGE polyacrylamide gel electrophoresis PMT photomultiplier tube py pyridine SCE saturated calomel electrode SMP submaxillaris protease -xvTEA triethylamine TIMED N ,N ,N' ,N' -tetrameth y leth ylenediamine TFA trifluoroacetic acid TMK trimethyllysine TMS tetramethylsilane TPCK L-1-tosylamido-2-phenylethyl chloromethyl ketone Tris tris(hydroxymethyl)aminomethane Standard three letter abbreviations Appl. Chern. 56, 595 (1984)). are used for the amino acids (Pure -1- CHAPTER 1 INTRODUCTION -2- 1. Introduction Electron transfer damental processes fixation. 1 Energy to is the elementary released another in transfer higher reduction protein the production for cellular processes of interest with regard and nucleic acid 3 degradation of of the dox adenosine triphosphate (figure to underlying of (ATP), 1.1). Protein from one re- is harnessed in is then other biological caused by which electron fun- nitrogen electrons potential such and photosynthesis, respiration, as reaction utilized transfer is also processes, e.g., protein 2 radicals and ionizing free radiation. Photosynthetic contain one or ions proteins), more copper (e.g., and respiratory redox-active in azurins metalloporphyrins chain prosthetic and ganic moieties (e.g., flavins). 4 proteins have revealed that which may plastocyanins, iron in posed at the protein surface. This across large distances (e.g., lOA), Structural these generally groups, studies prosthetic be of metal iron-sulfur chlorophyll), cytochromes, (e .g ., proteins redox or or- numerous redox often ex- groups are suggests that electrons barring large not must protein travel con forma- tiona! changes, in order to migrate between redox centers. It electron is of transfer interest to rates, controlled as understand (a) Hatefi, Y. Ann. Rev. Biochem. Photosynthesis, J. Wiley (New York), 1984. those charge separation 1015 which is the govern key to ( 1985); (b) Foyer, C. 2 Henriksen, T., Melo, T., Saxebol, G. in Free Radicals Pryor, ed., vol. 2, Academic Press (New York), 1976, p. 213. in Biology, w. 3 Myers, L. in Free Radicals in Academic Press (New York), 1980, p. 95. 4 Dreyer, J. Experentia 40, 653 (1984). 54, factors Biology, w. Pryor, ed., vol. 4, -3- Ferredoxin P430 ........... -0.6 , ',, Shunt pathway '...., _ of cyclic electron "'\.flow -0.4 \ NADPH' ~ c., g_ ., "E + 0 .2 1l a; j e- Plastoquinone \ P , +\ ADP Z 1: + 0.4 a: -,;, c "' + 0 .6 Vi Figure 1.1. Energy diagram for photosynthesis. Absorption of light results in charge separation which leads to redox protein reduction. As electrons are transferred to proteins of higher reduction potential, the released energy is coupled to ATP production (figure 23-12 from A. Lehninger, Principles of Biochemistry, Worth Publishing Co. (New York), 1982, p. 657). -4- photochemical energy understanding of oxygen reduction devices to storage and utilization. Nature's exquisite control reaction may efficiently derive provide energy Ultimately, of a thorough the highly exoergic for the guidelines from exceedingly design exothermic of reac- tions. 1.1. Theoretical background theory 5 Marcus rates of organic This has theory reasonably outer-sphere 6 intermolecular coordination been complexes predicts that rates electron in transfer self -exchange from of successful electron transfer predicting the for in- many constants. 7 rate will be dependent on the following factors: -nuclear frequency, v -free energy of activation, llG* -donor and acceptor electronic coupling, /3 -distance of donor-acceptor separation, d . The relevant equation is ket = v exp(-/3d) exp(-llG*/ RT) where ket classified is into the electron contributions transfer of rate. either an These influences electronic can be (exp(-/3d)) or 5 (a) Marcus, R. J. Chern. Phy s. 43, 679 (1965); (b) Marcus, R . J. Chern. Ph ys . 24, 966 (I 956). 6 Outer sphere electron transfer is defined as electron transfer be tween complexes with no change in ligation, as opposed to inner sphere electron transfer, in which donor and acceptor are linked by a bridging ligand prior to electron transfer. 7 (a) Siders, P ., Marcus, R. J . Amer. Chern. Soc. 103, 741 Chou, M, Creutz, C., Sutin, N. J. Amer. Chern. Soc. 99, 5615 (1977). (1981); (b) -5- nuclear exp(-AG*/ RT)) (11 The nature. free energy of activation, AG*, can be expressed in terms of the free energy of the reaction, AG 0 , and the reorganization energy i\ : The reorganization energy is geometry accompanying the that redox (inner-sphere reorganization) (outer-sphere reorganization). a semi-classical or process at and in the Several oped which differ from Marcus in energy required for changes in the donor and acceptor sites surrounding devel- theory in that they treat nuclear motion mechanical theories have molecules been quantum other solvent manner, rather than in a experimental results for biolog- classical sense. 8 Predictions of Marcus theory and ical systems have been compared in a review by Marcus and Sutin. 9 present, to is there applicable parameters 11, are insufficient to biological ~. and data transfer, 10 electron will i\ determine vary from whether and protein if to this At theory so, how the protein (or even electron trans- travels. The within a single protein). One outstanding i.e., the electron may fer, pected acceptor 8 to path question through is the mechanism of the which the electron (or hole) migrate through "space", in which case be related orbitals, to which the overlap of only is, of course, the critically the coupling is ex- participating donor and dependent on DeVault, D. Quantum-Mechanical Tunneling in Biological Cambridge University Press (New York), 1984, and references therein. donor- Systems, 9 Marcus, R ., Sutin, N. Biochem. Biophys. Acta 811, 265 (1985). 10 in terms of absolute rate seem to apply (vide infra). predictions; some aspects of Marcus theory -6acc ep tor orientation. While rates on orientation is this variable through which making the the distance these two likely to mechanisms, be required recognized, of orbitals transfer closest of distance is donor-acceptor which in dependence Alternatively, use electron of likely generally explicitly. 11 bonds, case the are proteins. of electron transfer few theories have treated the electron may travel the intervening medium, in likely to be greater than approach. A combination of different efficiency, are than single of probably of Hence, more a value ~ would be necessary to describe the rate theoretically. 1.2 Experimental approaches to the study of electron transfer Until recently, predominantly proteins or protein, 12 investigations of studies of bimolecular between a single and investigations metal of this protein electron transfer were reactions, either between redox complex and single redox sort continue. 13 a One difficulty 11 (a) Cave, Robert J., "A Theoretical Model for Orientation Effects in Electron Transfer Reactions.", Ph.D. Dissertation, California Institute of Technology, 1986; (b) Cave, R., Siders, P., Marcus, R., J . Phys. Chern. 90, 1436 (1986); (c) Cave, R., Klippenstein, S., Marcus, R. J . Chern. Phys. 84, 3089 (1986); (d) Siders, P., Cave, R ., Marcus, R. J . Chern. Phys., 81, 5613 (1984); (e) Siders, Paul, "Theory of Outer Sphere Electron Transfer Reactions.", Ph.D. Dissertation, California Institute of Technology, 1983, ch. 5. 12 (a) Brunschwig, B., DeLaive, P., English, A ., Goldberg, M., Gray, H., Mayo, S., Sutin, N . lnorg. Chern. 24, 3743 (1985); (b) Mauk, A., Bordignon, E ., Gray, H . J. A mer. Chern. Soc. 104, 7654 ( 1982); (c) English, A., Lum, V ., DeLaive, P., Gray, H. J. Amer. Chern. Soc. 104, 870 (1982); (d) Schichman, S., Gray, H . J. Amer. Chern. Soc. 103, 7794 (1981); (e) McArdle, J., Yocom, K ., Gray, H . J. Amer. Chern. Soc. 99, 4141 ( 1977); (f) Wherland, S., Gray, H . Proc. Nat/. Acad. Sci. 73, 2950 (1976); (g) Coyle, C., Gray, H. Biochem. Biophys. Res. Comm., 73, 1122 ( 1976); (h) McArdle, J., Gray, H., Creutz, C., Sutin, N . J . A mer. Chern. Soc. 96, 5737 ( 1974). 13 (a) 7102 Reid, L., Dalvit, G ., ( 1987); (b) McGinnis, Wright, C., Saltman, J., Sinclair-Day, J., P. Biochemistry 26, Sykes, A., Powls, -7- encountered in complex(es) in has recently tations in of tion). 14 there the lies electron in defining that single protein pair, differing in transfer (as to being be of several sites there small of place. Evidence several suggests case of takes which the nature event obtained opposed the transfer been In may studies which a electron these subtle ways, one molecule-protein interaction, suggested that the ligands of metal protein structure, making definition and may of electron orien- may result unique orienta- electron transfer, some complexes the possible the experiments intercalate have into transfer the distance difficult. 15 To have a void resorted ticularly those Numerous donors, to the difficulties using more involving organic acceptors, and linkage of the systems have been donor various amino acids as mediators peptides bridged by but containing two more synthesized, sufficiently acceptor, and those flexible direct to acceptor. 16 with complexes evaluate par- synthesized inorganic been and complexes, To also researchers donor-acceptor bridges. 17 transport, donor well-defined reactions, and electron are bimolecular covalent model of residues of or allow close outer-sphere electron have amino approach acid of the transfer. 18 Poly- R ., Moore, J ., Wright, P. Inorg. Chern. 27, 2306 (1988). 14 15 Poulos, (1987). Mauk, T., Sheriff, A., Scott, R., S., Howard, Gray, H. J. A. Bioi. J. Chern. Arner. Chern. Soc. 16 (a) Closs, G., Miller, J. Science 240, Mauk, A., Gray, H . J. Chern. Ed. 62, 932 (1985). 440 17 (a) Miller, J ., Calceterra, L., Closs, G. 3047 (1984); (b) Joran, A., Leland, B., Dervan, P. J . Arner. Chern. Soc. 106, 6090 (1984). J. 262, 13881 102, 4360 (1980). (1988); (b) Scott, R ., Arner. Chern. Soc. 106, Geller, G., Hopfield, J., -8- proline isomerization is intramolecular electron conformation) under iments has do also sufficiently slow transfer these in conditions of study high not permit the of been studied in bimetallic approach has been to allow measurement complexes driving medium (in an fast extended force, 19 but effects. Electron complexes of such expertransfer with other types of redox protein partners, bridges. 20 Another i.e., proteins pairs of between redox which proteins to electron work with transfer (nonphysiological) in occurs whose vivo, interactions or other have been modeled and / or defined experimentally. 21 Yet another tack has been taken by Hoffman et al. in the preparation of "hybrid for iron in with the electron 18 hemoglobins", either other hemoglobin native transfer. 2 2 where subunit Because the zinc subunit, to or and study donor magnesium the product photoinduced and is acceptor substituted reconstituted intramolecular sites (a) Isied, S., Vassilian, A., Magnuson, R., Schwarz, H. Chern. Soc. 107, 7432 (1985); (l)) Isied, s., Vassilian, A. Chern. Soc. 106, 1732 (1984); (c) lsied, S., Vassilian, A. Chern. Soc. 106, 1726 ( 1984). 19 By high driving force thermodynamically. is meant a reaction are fixed, J. J. J. Arner. Arner. Arner. which is strongly favored 20 (a) Kendrick Geno, M, Dawson, J. Inorg. Chern. Anderes, B., Lavallee, D. Inorg. Chern. 22, 2665 (1983). 21 (a) Liang, N ., Kang, C., Ho, P., Margoliash, E., Hoffman, B. J . A mer. Chern. Soc. Taylor, K., 108, 4665 ( 1986); (b) Cheung, E ., Kornblatt, J ., English, A., McLendon, G., Miller, J. Proc. Nat/. A cad. Sci. 83, 1330 ( 1986); (c) Simolo, K., McLendon, G ., Mauk, M., Mauk, A. J. Arner. Chern. Soc. 106, 5012 (1984). 24, 1731 (1985); (b) 22 (a) Natan, M., Hoffman, B. Book of Abstracts, 196th American Chemical Society National Meeting, Inorganic Abstract #357, Los Angeles, CA, September 25-30, 1988; (b) Peterson-Kennedy, S., McGourty, J., Kalweit, J., Hoffman, B. J. Arner. Chern. Soc. 108, 1739 (1986); (c) Peterson-Kennedy, S., McGourty, J., Hoffman, B. J. Arner. Chern. Soc. -9- this native hemoglobin orientation introduce tution and electron involve transfer perturbations in with regard changes distance the does not effects. 23 medium uncertainty may system a to protein ligation and is uncertain, chain, not but only force electron involve different mechanisms accessible in reactions of driving force, that absolute great lar tures, as driving of electron affect the of these systems is transfer in advantage electron high may initiation of the the solid reaction transfer redox of not moiety level of the coupling. 25 One ability to study i.e., at subzero tempera- involve any bimolecular have been covalently moiety, which then serves forming a state, does extent aromatic e.g., donor-acceptor the possible porphy- state the of the cation excited believed Also, radical for low substi- the may electron effects. unknown or the because transfer transferring as also Systems is systems often is it porphyrin because rins. 24 and studying secondary delocalization participation, of structure, electron involving means Metal-substituted in polypeptide provide not intramolecu- process. Finally, modified as the with a single-site metalloproteins single organic 26 or inorganic counterpart of the prosthetic 106, 5010 (1984); (d) McGourty, Chern. Soc. 105, 4470 (1983). group J., in Blough, N., donor-acceptor Hoffman, B. J. Arner. 23 Site-directed mutagenesis would be one way to overcome this problem, and has been demonstrated in the /3 chain of human hemoglobin. Nagai, K., Perutz, M., Poyart, C. Philos. Trans. R. Soc. London A, 317, 443 ( 1986). 24 Atamian, M, 1510 (1988). Wagner, R., Lindsey, J., Bocian, D. Inorg. 25 Beratan, D. J . Arner. Chern. Soc. 108, 4321 (1986). 26 Faraggi, M., Klapper, M. J. Arner. Chern. Soc. 110, 5753 (1988). Chern. 27, -10- is It pair. a modification of the latter type which is the basis of this work. 2. Covalent Labeling 2.1 The experimental system The approach taken in this work involves the covalent binding of an inert ruthenium well-characterized complex to single-site Thermodynamic similar those to of defined amino meta lloprotein, acceptor complex bridged by a orientation. a acid to side create chain of a a donor- known medium at a defined distance and parameters, biological which electron expected to be reactions, can be are transfer determined from electrochemical studies of the metal sites. Aquopentaammineruthenium(II), [A 5 Ru(OH 2 )] 2 +, is known from model studies the to react with imidazole, 2 7 thiols, 28 and thioethers, 29 functional cysteine, moieties the side chains of the amino figure 1.2). polypeptide chains, while methionine is more frequently encountered, but it is often found protein of the (see in proteins single interior. globular Modification pen taammineru theni urn histidine both respectively histidine, is residue methionine, acids Cysteine rare and of the complex +2 of and 27 Sundberg, R., of these complexes amino are pentaammineruthenium +3 Bryan, oxidation R., 1., is rather is states. 27 Taylor, acids Taube, H. 28 Kuehn, C., Taube, H. J. Amer. Chern. Soc. 98, 689 (1976). 2 9 Stein, C., Taube, H . lnorg. Chern. 18, 1168 (1979). the undesirable, as 28 29 • The to inert in complexes of Amer. Chem. Soc. Carboxylato- 96, 381 (I 974). in labile. expected J. a be -11- Cys > Figure 1.2. Potential reactions of the aquopentaammineruthenium(ll) complex with common amino acid side chains. -12pentaammineruthenium(III) are also known, nor pH has been stability at neutral but neither their demonstrated. 30 synthesis Reaction with lysine to form hexammine complexes can be discouraged by the appropriate choice of reaction pH. first Modification of proteins demonstrated with by the pentaammineruthenium moiety was the utilized to ribonuclease; 31 chromophore was monitor conformational changes, 32 to aid in the assignment of NMR onances,33 and to perform distance measurements by fluorescence Proteins in which electron transfer been studied res- energy transfer. 34 (table 1.1) using this procedure include myoglobin, horse heart cytochrome c, 37 has sperm whale 35 Pseudomonas to date and human 36 aeruginosa azurin ' 38 30 Stritar, J., Taube, H. lnorg. Chem. 8, 2281 (1969). 31 Matthews, C., Erickson, Chem. Soc. 100, 2260 (1978). P., Van 32 Matthews, 499 (1980). C., Erickson, P., Froebe, 33 Matthews, (1981). C., Recchia, J ., Froebe, 34 Recchia, J., Matthews, Acta 702, 105 (1982). C., Rhee, Vliet, M., C. C. D ., Petersheim, M. J. Am. Biochim. Biophys. Acta 624, Anal. Biochem. 112, 329 Horrocks, W. Biochim. Biophys. 35 (a) Cowan, J., Gray, H. Chem. Scripta 28A, 21 (1988); (b) Karas, J., Lieber, C., Gray, H . J. Amer. Chem. Soc. 110, 599 (1988); (c) Crutchley, R., Ellis, W., Gray, H. in Frontiers in Bioinorganic Chemistry, A. Xavier, ed., VCH Verlagsgesellschaft (Weinheim, Fed. Rep. Germ.), 1986, p. 679; (d) Crutchley, R., Ellis, W., Gray, H . J. Am. Chem. Soc. 107, 5002 (1985). 36 Zewert, T., unpublished 37 (a) Elias, H., Chou, M., Winkler, J. J. Amer. Chem. Soc. 110, 429 (1988); (b) Bechtold, R., Kuehn, C., Lepre, C., lsied, S. Nature 322, 286 (1986); (c) Bechtold, R ., Gardineer, M, Kazmi, A ., van Hemelryck, B., lsied, S. J . Phys. Chem. 90, 3800 (1986); (d) Nocera, D ., Winkler, J., Yocom, K., Bordignon, E., Gray, H. J. Amer. Chem. II II 39 47 83 II II II C. krusei P. stutzeri P. aeruginosa II cyt. C551 azurin II II 33 II 48 sperm whalec II His Species horse heart cytochrome c myoglobin Protein II II II (D) A5Ru2+ (A) A4isnRu3+ (A) A4pyRu3+ (D) A5Ru2+ (A) A4pyRu3+ (A) A5Ru3+ (D) A5Ru2+ Label b II Cu II II II II II II II Fe(heme) Prosthetic Group 0 II 1.9(0.4) 2.5(0.8) II 13(2) >170e lle 2e 30(3) 2.5(0.5) 0.041(0.003) 0.019 (0.002) k(s- 1 ) 11.8 7.9 -200[ 280 -13.0 II II 11.8 II II 12 d(A) 170 -160 -90-100 170 275 20 -20 0 ~E (mV) Table 1.1. Results of covalent labeling experiments to date (native prosthetic groups) a n m k J J 1 h g g ref. I .... l.f 20 42 II C. vinosum II II Label II II II II 0 6.9 7.9 270f 9.2 10-12 -11.9 d(A) 289(7) 258(7) -3oor -26fl 0 ~E (mV) k(s- 1) His48 derivative. d. HiPIP = high potential iron-sulfur protein e. tentative; measured under conditions described in appendix 4 f. assuming a reduction potential of 80mV for the pentaammineruthenium histidine g. Crutchley, R., Ellis, W., Gray, H. 1. Amer. Chem. Soc. 110, 599 (1988). h. Karas, J., "Long Range Electron Transfer in Ruthenium-Labelled Myoglobin.", Ph.D. Dissertation, California Institute of Technology (1989). i. Nocera, D., Winkler, J., Yocom, K., Bordignon, E., Gray H. 1. Amer. Chern. Soc. 106, 5145 (1984). temperature for these systems. Distances are in the 19-22A range. Driving force is expected to be similar to that of the c. His18-, His112-, and His116-A5Ru myoglobin derivatives have been prepared. No electron transfer is observed at room p 0 14(2)e 18 0 I, n I, n ref. 14(2)e 0.04 (0.22) 0.024(0.058) b. D or A preceding label indicates whether electron donor or acceptor, py=pyridyl, isn=isonicotinamide; 4Fe-4S II Cu Group Prosthetic II II (D) A5Ru2+ a. only 1:1 (Ru:M) derivatives are included; HiPIPd II 59 A. variabilis plastocyanin S. obliquus His Species Protein Table 1.1 continued f" I - j. Meade, T., unpublished results k. this work l. Osvath, P., Salmon, G., Sykes, A. J. Arner. Chern. Soc. 110, 7114 (1988). m. Kostic, N., Margalit, R., Che, C., Gray, H. J. Arner. Chern. Soc. 105, 7765 (1983). n. Jackman, M., McGinnis, J., Powls, R., Salmon, G., Sykes, A. J. Arner. Chern. Soc. 110, 5880 (1988). o. Cowan, J., unpublished results p. Jackman, M., Lim, M.-C., Salmon, G., Sykes, A. J.C.S. Chern. Cornrn. 179 (1988). I 1-' Ul I -16plastocyanins, 39 two algal high paten tial In addition been iron-sulfur to done the with Rhus protein results the (HiPIP) presented same stellacyanin,40 vernicifera in table modifications vinosurn. 41 Chrornatiurn from 1.1, using many the and studies nonnative, have photoactive prosthetic groups. 2.2 Results to date from covalent labeling experiments Studies proteins of have intramolecular yielded some electron transfer in important results. Results from c the ammineruthenium-modified horse demonstrate that electron transfer measurable rate and biologically transfer in zinc-su bsti tu ted ruthenium-modified strongly support the view "through space" rather than a electron myoglobin derivatives occurring by a heart can cytochrome ruthenium-modified occur relevant over were large driving of first distances force. at Rates sperm electron "through penta- a of whale transfer bond" to as pathway, 42 Soc. 106, 5145 (1984); (e) Isied, S., Kuehn, C., Worosila, G. J. Arner. Chern. Soc. 106, 1722 (1984); (f) Yocom, K., Winkler, J., Nocera, D., Bordignon, E., Gray, H. Chern. Scripta 21, 29 (I 983); (g) Isied, S., Worosila, G., Atherton, S. J. Arner. Chern. Soc. 104, 7659 (1982); (h) Yocom, K., Shelton, J. B., Shelton, J. R., Schroeder, W., Worosila, G., Isied, S., Bordignon, E., Gray, H. Proc. Nat/. Acad. Sci. 79, 7052 ( 1982). 38 (a) Pecht, Margalit, R., Kostic, N., Che, C.-M, Blair, D., Chiang, I., Shelton, J. B., Shelton, J. R., Schroeder, W., Gray, Proc. Nat/. Acad. Sci. 81, 6554 (I 984); (b) Kostic, N., Marga lit, Che, C.-M, Gray, H. J . Arner. Chern. Soc. 105, 7765 (1983). 39 Jackman, M, McGinnis, J., Chern. Soc. 110, 5880 (1988). Powls, R., Salmon, G ., Sykes, A . J. H., H. R., Arner. 40 Farver, 0., Pecht, 1., submitted for publication 41 Jackman, M., Lim, M., Salmon, G., Sykes, A. J .C.S. Chern. Cornrn. 179 (I 988). 42 By "through space" it is not meant without orbital participation, -17and have provided an estimate of 0.9A -I for f!>, assuming an exponential distance dependence of the rates. 43 Using two derivatives of sperm whale myoglobin modified at His48, it was transfer demonstrated rates predicted More extensive driving data on the by rates A 5 RuHis48 His33 the exhibit force theory dependence is obeyed dependence curves have ruthenium-modified The driving Marcus force metal-substituted substituted atives. that of proteins. 44 in been electron obtained with sperm whale myoglobin 45 and zinc- horse heart cytochrome c46 deriv- driving force, un- predicted dependence on but the "inverted region" has yet to be found in these systems. 47 2.3 Aim of this project In dertaken this project, the to investigate modification medium and yeast cytochrome c was orientation effects on electron of but rather a relatively direct (i.e., shortest distance) pathway. "Through bond" is used here to denote participation of all residues as they occur in the peptide chain between the modified histidine and and the point of ligation or covalent attachment of the prosthetic group. 43 (a) Axup, A., Albin, M., Mayo, S., Crutchley, R., Gray, H. J. Arner. 110, 435 (1988); (b) Axup, Andrew W. "Laser Flash Spectroscopy of Zinc/Ruthenium Myoglobins: An Investigation of Distance and Medium Effects of Photoinduced Long-Range Intraprotein Electron Transfers.", Ph.D. Dissertation, California Institute of Technology, 1987. Chern. Soc. 44 Karas, J., Lieber, C., Gray, H. J. Amer. Chern. Soc. 110, 599 (I 988). 45 Cowan, J., Gray, H., Chern. Scripta 28A, 21 (1988). 46 Meade, T., Gray, H., Winkler, J., unpublished. 47 The "inverted region" is that regime rates are predicted by Marcus theory to othermicity, i.e., -h.G 0 >A.. in which electron transfer decrease with increasing ex- -18transfer rates. c Cytochrome protein responsible terminal electron duction of is a member for the transfer oxygen to of the reduction protein and of the Studies water. respiratory chain, and is the c oxidase, the cytochrome enzyme which effects long-range electron of the re- transfer involving cytochrome c have recently been reviewed by Pielak et a!. 48 c Cytochromes containing a prosthetic group thioether linkages and are monomeric proteins protoporphyrin single is covalently formed by 1.4). His 18 and Met80 oxidation states in the electronic configuration. IX bound at residues, 103-111 of heme per the 1-vinyl The molecule. positions via cysteine residues 14 and 17 (figures 1.3 serve ligands in both the 2+ and 3+ range, resulting in a low-spin pH The as 4-8 structural aspects of cytochromes have been reviewed. 49 krusei Cytochromes from the yeasts both histidine residues have Saccharomyces at positions cerevisiae and Candida 33 39. 50 Horse and heart cytochrome c has been modified with the pentaammineruthenium(III) moiety and lacks for this derivative histidine at position 39). 37 d A comparison of 48 Pielak, (1987). G ., 49 (a) Mathews, F., Prog. Biophys. Molec. Bioi. 45, (1985); (b) Timkovich, R . in The Porphyrins, D. Dolphin, ed., vol. 7, Biochemistry, part B, Academic Press (San Francisco), 1979, p. 241; (c) Dickerson, R., Timkovich, R . in The Enzymes, P. Boyer, ed., vol. XI, part A, 3rd ed., Academic Press (New York), 1975. D., Moore, G., transfer protein observed Concar, electron 30s- 1 intramolecular been His33, of an has at (the horse this Williams, R. rate rate with Protein that of Eng. 50 numbering used is that based on the horse heart cytochrome sequence 1, a 83 -19- COOH COOH Figure 1.3. Cytochrome prosthetic group, protoporphyrin IX. Vinyl side chains are actually saturated, and form thioether linkages with Cys14 and Cys17 side chains (at the 1 positions, indicated by the arrows). The iron is ligated by the protein residues His18 and Met80 (not shown). Figure 1.4. The cytochrome c prosthetic group is a protoporphyrin IX (red) with a central iron ion. The polypeptide anchors the prosthetic group by both ligation to the iron via His18 and Met80 (yellow), and by bonds to the 1-vinyl side chains via the Cys14 and Cys17 (green) side chains. -21yeast c, cytochrome singly-modified at are medium dependence, as sequence (see 1.5). figure the proteins second A this same -50% residue, identical derivative of a might in reveal amino yeast acid cytochrome, singly-modified at His39, might allow for comment on the role of donoracceptor orientation histidines are the His33 heme. heme, 51 governing dramatically lies -11.7A in electron transfer rates, different orientations front the of ring (see figure similar, as the 1.6). 52 intervening Medium protein Hisl8-ligated effects (in as with a might through the two respect to of the plane of the be expected to while His39 is -13A below the heme edge, in the porphyrin be in in face space pathway) contains no aromatic residues in both cases. Initial work was begun on its gene delete been cloned 53 insert histidine has and the cytochrome c from S. cerevisiae, as and mutagenesis residues) were presence of a cysteine residue introduced ization, and commercial supplies of the experiments (to selectively planned. However, complications, e.g., protein were the dimer- unreliable. The work reported here was done exclusively on the Candida cytochrome. Although been 51 determined, a crystal that of structure the iso-1 of the Candida cytochrome from cytochrome has not S. cerevisiae 54 con- The triad of the Ccx of residue 33, the heme iron atom, and the pyrrole ring IV nitrogen (the nearest porphyrin atom) form an angle of 51.7° in the reduced tuna cytochrome crystal structure. 52 distances are measured from the C'Y of the residue to the heme or ligand atom of closest approach; there is some uncertainty as to what distance to consider, but the heme is generally thought of as an extended donor /acceptor due to mixing of the porphyrin molecular orbitals with the iron d orbitals. 53 Montgomery, D., Hall, B., Gillam, S., Smith, M. Cell 14, 673 (1978). 54 Louie, G., Hutcheon, W., Brayer, G . J. Mol. Bioi. 199, 295 (1988). -22- 10 1 Pro-Ala-Pro-Phe-Glu-Gln-Gly-Ser-Ala-Lys-Lys-Gly-Ala-Thr-Leu-Pheacetyl-Asp-Val-Glu-Lys -Lys- Ile- 20 Lys-Thr-A rg-Cys-Ala-Gln-Cys-His- Thr-Ile-G lu-Ala-G ly-Gly-Pro-HisVal- Gln - Lys-Val-Lys-Lys40 30 Lys- Val-G ly-Pro-Asn-Leu-His-G ly-Ile-Phe-Ser-Arg-His-Ser-G ly-G In-Thr- -Leu- -Gly- -Lys-Thr- 50 Ala-G I u-Gly-Tyr-Ser-Tyr-Thr-Asp-Ala-Asn-Lys-Arg-Ala-Gly-Val-Giu-Pro- -Phe -Thr- -Asn- Lys- -Ile - Thr- 60 70 Trp-Ala-Glu-Pro-Thr-Met-Ser-Asp-Tyr-Leu-Glu-Asn-Pro-TMK-Lys-Lys-Glu-Leu- Met-Glu-Lys- 80 Tyr-Ile-Pro- G ly-T h r-Lys- Met-Ala -Phe-G ly-G ly-Leu-Lys-Lys-Ala-Lys-lle-Ala-Ile-Lys-Thr- 90 100 Asp-A rg-Asn-Asp-Leu-Val-Thr-Tyr-Met-Leu-Glu-Ala-Ser-Lys Glu- -Glu- -Ile- Ala- -Leu -Lys- Lys- -Thr-Asn-Glu Figure 1.5. Sequences of the cytochromes c from Candida krusei8 and horse hearth (the horse heart residues are identical to those of yeast unless specified). The residue numbers are assigned according to the horse sequence. The two proteins have 59 identical residues; approximately half of the 44 substitutions are conservative. a. Lederer, F. Eur. J. Biochem. 31, 144 (1972); Narita, K., Titani, K. J. Biochem. 63, 226 (1968). b. Margoliash, E., Smith, E., Kreil, G., Tuppy, H. Nature 192, 1125 (1962). -23- Figure 1.6. Histidine residues (yellow) of interest for modification in C. krusei cytochrome c. His33 is located at 11.7A from the heme (red), as measured from the Cy atom to the N8 of His18. His 39 is located at 13.0A from the heme, as measured from the Cy atom to the C atom of pyrrole ring IV to which the propionic acid is bound. Figure 1.7. The cytochrome c fold. The polypeptide backbone (-N-C-C-)n is shown here in blue. Cytochrome c is a globular protein with no J3-sheet, and several stretches of a-helix. -24- Figure 1.6 Figure 1.7 -25firms the general trend found and fold are similar for species of cytochrome [reduced 5 5 tuna Saccharomyces utili zing and cerevisiae, nuclear for species all have cytochromes thus far Furthermore, NMR see 1.7). have (five crystallographically: bonito, 57 a horse ,57 enhancement conformation rice ,58 oxidized 56 ], Overhauser the investigated characterized been figure i.e., c, demonstrated and studies the sim- between the cytochromes c are the of ilarity of the crystal and solution cytochrome structures. 59 Studies native not of the iron site and necessarily electron the expected transfer intramolecular label in to new insight into is known from provide this protein. It with physiological c reacts and cytochrome c oxidase, 55 modified in cytochrome the protein transfer ruthenium that bypassing electron medium its through its altogether. 60 exposed numerous partners, heme mechanism studies cytochrome c1 edge, seemingly These experiments have some (a) Takano, T ., Dickerson, R . J. Mol. Bioi. 153, Takano, T ., Kallai, 0 ., Swanson, R ., Dickerson, R . 248, 5234 (1973). 79 J. (1981); (b) Chern. Bioi. 56 Takano, T., Dickerson, R. J . Mol. Bioi. 153, 95 (1981). 57 (a) Dickerson, R., Takano, T ., Eisenberg, D., Kallai, 0., Samson, L., Cooper, A ., Margoliash, E . J . Bioi. Chern. 246, 1511 (1971); (b) Dickerson, R., Kopka, M, Borders, C., Varnum, J., Weinzierl, J . J . Mol. Bioi. 29, 77 (1967). 58 Ochi, H ., Hata, Y., Tanaka, N ., Kakudo, M., Morita, Y. J. Mol. Bioi. 166, 407 (1983). 59 (a) Feng, Y., Roder, H ., Englander, S., Wand, A., DiStefano, D . Biochemistry 28, 195 (1989); (b) Pielak, G., Boyd, J ., Moore, G., Williams, R . Eur. J. Biochem. 177, 167 (1988). 60 Rieder, R ., Bosshard, H. J. Bioi. Chern. 255, 4732 (1980). However, evidence has recently been obtained to suggest that electrostatically-stabilized protein complexes are not necessarily optimum for electron transfer efficiency: Hazzard, J., McLendon, G ., Cusanovich, M., Das, G ., Sherman, F ., Tollin, G . Biochemistry 27, 4445 (1988). Furthermore, electron transfer through the heme edge is Sakurai, T ., Aihara, S., -26bearing, in however, other on proteins the mechanism whose redox of intramolecular centers are electron separated by transfer tens of angstroms (e.g., cytochrome c oxidase). 61 3. Previous investigations of medium and orientation effects A number of organic model systems have been developed intramolecular electron transfer, and rate depends on electronic structure donor and fer acceptor. For from dimethylaniline whose A the bridge similar includes example, suggested the medium et al. found pyrene is -25% 1,4-naphthyl 1,5-naphthyl-bridged have of Heitele to excited a several as opposed compound to study a undergoes the bridging the electron faster to that trans- in compounds moiety. 62 phenyl electron transfer at a rate -60% of that of the phenyl-bridged compound. The cytochrome idase electron transfer rates of c mutants with zinc-substituted (cyt) (Zn-CCP) appear to display photoactivated electron transfer similar (-200s- 1 ) for rate containing Tyr or Ser at value when Gly is series significant from the a to protein not necessarily expected to (see Marcus and Sutin, p. 265). Intermediate be adiabatic, Fe 3 +cyt (Phe) values as c iso-1 perox- dependence. 63 residue position 82, but falls substituted. cerevisiae S. cytochrome medium 3 znCCP native of the proceeds and the The at a mutants to -10% of this are observed for heme is recessed 61 Brudvig, G ., Blair, D ., Chan, S. J. Bioi. Chern. 259, 11001 (1984). 62 Heitele, H ., 273 (1987). 63 (a) Liang, N., Mauk, A., Pielak, G., Johnson, J ., Smith, M., Hoffman, B. Science 240, 311 (1988); (b) Liang, N., Pielak, G., Mauk, A ., Smith, M, Hoffman, B. Proc. Nat/. Acad. Sci. 84, 1249 (1987). Michel-Beyerle, M., Finckh, P. Chern. Phys. Lett. 134, -27Leu and Ile cy tochrome mutants. and Tyr residues (Ser, Leu, v ariance with what In contrast, mutant than the in thermal those back mutants electron transfer containing aliphatic These one predict on the basis of results reduction at are potentials: the Ser and Gly mutants have a reduction potential -50mV less than that of the results WT is and that the Tyr mutant. the aromatic One residues possible mediate interpretation electron of these transfer by a the reaction superexchange mechanism. 64 While centers the of the recent x-ray photosynthetic structural bacteria R. sphaeroides R-26 66 have revealed the acceptor not evident at pairs, geometric it is arrangement serves to anal y ses of Rhodopseudomonas viridis 65 and exact of donor and what extent this this facilitate orientation point to "forward" electron transfer vis-a-vis preventing charge recombination. Relatively few bound donor the redox of those and of acceptor participants that do the in have numerous moieties a rigid, demonstrated model systems featuring that have defined how been synthesized contain However, studies geometry. critical covalently orientation may be in determining electron transfer rates. The rate of between stacked porphyrin dimer and a cofacial pyromellitimide times faster than that from a acceptor has been found to be one thousand transfer electron 6 4 Miller, J ., Beitz, J. J. Che rn. Phys. 74, 6746 (1981). 65 (a) Deisenhofer, J ., Epp, 0., Miki, K ., Huber, 318, 618 (1985); (b) idem, J. Mol. Bioi. 180, 385 (1984). 66 Chang, C., Tiede, D ., F EBS L ett. 205, 82 ( 1986). Tang, J., Smith, U ., R., Michel, Norris, J., H. Nature Schiffer, M -28- a side-by-side isomeric dimer to the same acceptor in a perpendicular is attempt to investigate and medium orientation. 67 The two project influences, making use of reported on donor-acceptor what is expected here an orientation to be a these effects, by minor perturbation to a native J.C.S. Chern. Comm. biological system. 67 Cowan, J., 55 (1987). Sanders, J., Beddard, G., Harrison, R. -29- CHAPTER 2 PREPARATION AND ISOLATION OF THE CYTOCHROME DERIVATIVE -30MA TERIALS and INSTRUMENTATION All was buffers further (model and solutions purified 02794). by The were passage specific prepared through resistance a of with distilled Barnstead the water water Nanopure was that system greater than 18MQ-cm. Candida krusei cytochrome c was purchased from Sigma (Type VII). Chloropen taammineru theni u m(III) was chloride used as supplied by Strem Chemicals, Inc. Mossy zinc, mercuric ether were of reagent grade. oxide, sulfuric acid, ethanol, and diethyl Ammonium hexafluorophosphate (NH 4 PF 6 ) was obtained from Pennwalt. All buffer salts (Na 2 HP0 4 , NaH 2 Po 4 ·H 2 0, Tris, Tris·HCl, glycine, and HEPES acid and sodium salt) were reagent grade. CM-52 carboxymethylcellulose from Whatman. Amberlite CG-50, cation-exchange 100-200 resin was obtained mesh, was purchased from Alfa Products. Biorad Econocolumns (2.5cm ID) were purifications involving less than lOOmg made Caltech glass shop, were in the utilized protein. glass tubes for separations All other columns, with medium fitted and sintered glass frits and Teflon stopcocks. Flow rates were usually maintained 2 model Rl or a Pharmacia model P-1 with either a Gilson Minipuls peristaltic pump fitted with poly- vinylchloride manifold tubing. Fractions tionator in were 13mm x collected with a Gilson model lOOmm Pyrex glass test tubes. FC-80 Micro-Frac- Elution profiles were recorded using a Gilson Model HM Holochrome and Gilson model N2 DC recorder, or constructed by plotting the absorbance reading (A-=41 Onm) of -31(at least) every fraction third during apparent band elution and (at least) every fifth fraction otherwise. pH measurements were made with a Brinkmann Instruments model pH101 or Beckman model +32 pH meter, using a combination microelectrode. Beckman buffer standards were used for calibration. The pH values quoted for buffer solutions are those measured at room temperature. Conductivity measurements were made with a Markson model 10 porta ble conductivity meter. Beckman solution calibrating KCl O.OlN (1410J,JMH0/cm) was used. Ultrafiltration cutoff), and cells, type Centricon-10 YM-5 membranes microconcentrators (5000 (10000 molecular weight molecular weight cutoff) were purchased from the Amicon Corporation. UV-visible spectra were recorded on either a Cary 219, Cary 17, or Shimadzu UV -260 spectrophotometer. Solutions were degassed scrubbed oxygen a by vacuum-argon Argon was tower. Solutions were transferred anaerobically fitted at end with a stainless or LKB acrylamide each of on 6" passage through by steel dual-manifold line. a oxide manganese means of 20-gauge a cannula deflected-tip needle. BioRad methylenediacrylamide Acrylamide up to were solutions two were weeks. used to filtered and tion. (EM Reagents) were Trichloroacetic sulfosalicylic acid prepare stored in from prepared acid Kodak. was N,N'- Reagents EM electrophoresis amber N,N,N',N'-tetramethylethylenediamine obtained in 1 ml ampules from LKB. fate and bottles at (TEMED) was Solutions of ammonium peroxydisul- immediately prior to obtained from Sigma, and 5- was supplied by Coomassie Blue R250 gel polymeriza- -32All Scientific Chemical Company. Analytical horizontal electrophoresis Multiphor and DC 2197 were obtained circulated a chemicals was carried out were reagent using an 2117 paper wicks Glass supports and Water cooled by a Brinkmann RM3 support plate to maintain the gel grade. LKB power supply. from LKB. within other bath was temperature below 10°C. The fast protein liquid chromatography (FPLC) instrumentation, manufactured by Pharmacia, included two P-500 pumps, a UV -M UV -visible detector and an recorded simultaneously on (Pharmacia REC-482) carried on 5/ 5 (Hg out or lamp), 10/10, LCC-500 a Elution controller. strip two-channel and the LCC-500 integrator. Pharmacia cation exchange Mono for analytical and preparative S were recorder chart Separations columns, runs, tions were collected during preparative runs using a profiles were either size Frac- respectively. Pharmacia FRAC-100 instrument in a time-based peak-only mode. METHODS Na[Co(EDT A)] was prepared by the method of Kirschner, 1 substituting sodium carbonate for barium carbonate. The prepared hexafluorophosphate by the method of salt of Callahan aquopentaammineruthenium(II) et a/. 2 The compound was was stored under vacuum and used within one week of preparation. CM-52 resin was removed from each column after a single use and suspended in 1M NaCl for a minimum of 24 hours. If the supernatant was Kirschner, S. !nor g. Syntheses 5, 186 (1957). 2 Callahan, R., Brown, G., Meyer, T. Inorg. Chern. 14, 1443 (1975). -33especially dirty, 1M NaCI. The resin was rinsed by repeated decanting of the supernatant and it resuspension in was decanted water. off Once and the the ionic resin was strength resuspended was reduced resin while in suffi- ciently, the resin was suspended in the buffer to be used. Chromatography All All columns were chromatography Columns were packed with a slurry of 4°C, unless was performed at considered equilibrated when the pH stated and flowing. otherwise. conductivity of the reservoir and effluent buffers were identical. Purification of commercial Candida krusei cytochrome c The protein was dissolved in a minimum of pH 7 sodium phosphate buffer (<lOOmM}, and an aliquot of a concentrated Na[Co(EDTA}] solution was added. to insure cations dant The cytochrome solution was stored at 4°C for several hours that from all the redox protein isomers was sometimes removed sufficiently column in was oxidized, subsequent ionic strength, the directly (the negatively charged thus avoid chromatography. by ultrafiltration, low and or cytochrome compli- Excess oxi- if the sample was of was onto the bind to CoEDT A- loaded does not cation-exchange resins, but elutes immediately in the void volume). The cytochrome was routinely 85mM, purified by elution from CM-52 pH 7 sodium phosphate (NaPi) buffer. 3 with However, before deciding upon this method, a comparison was made with a second method, which had 3 Brautigan, 128 (1978). D ., Ferguson-Miller, S., Margoliash, E. Met h. Enzym . 53, -34been reported specifically from Candida krusei. 4 for This the purification of the cytochrome method involved the elution of c 100mg cytochrome c from a 4cm x 23cm column of 100-200 mesh Amberlite CG-50, equilibrated in pH 7 100mM NaPi, with a linear gradient from 50mM to 250mM NaCl in the same buffer {1280mL each), at a flow rate of 52mL/hr. Concentration of protein solutions Concentration of protein solutions was achieved tion in an Amicon apparatus with YM-5 membranes. were required, were were soaked Amicon several YM-10 hours Centricons prior to use to by ultrafiltra- When small volumes utilized. remove New membranes preservatives. Mem- branes were stored in 10% ethanol at 4°C to prevent bacterial growth. Estimates of protein concentration An aliquot of cytochrome solution was diluted with a known volume of buffer, and £=106100M- 1 cm- 1 its absorbance for horse was heart read at 410nm cytochrome c) (cytochrome against the isosbestic; appropriate blank. 5 Optimization of the reaction conditions to achieve a 1:1 derivative mine Polyacrylamide gel the time reaction nium:cytochrome derivative. 250mm; T=3.6%, C=2.6%) 6 electrophoresis (PAGE) for optimum 3.5% polyacrylamide with was production utilized to deter- of 1:1 ruthe- a gels (1.5mm 51-11 sample wells were 4 Lederer, F . Eur. J. Biochem. 31, 144 (1972). 5 Margoliash, E., Frohwirt, N . Biochem. J. 71, 570 (I 959). x 125mm x prepared according -35to LKB Application Note 306 using pH 7 phosphoric acid) or pH S.9 Tris-Gly at least 12 trophoresed hours at least lOOmM imidazole (titrated with buffer and were stored at 4°C for prior to use. Imidazole-buffered gels one half hour at were SOmA; samples were preelec- concentrated for 10 minutes at 20mA, and run for 2 hours, 50 minutes at SOmA. Tris- Gly con- gels were preelectrophoresed 1.5 hours at 3SmA; samples were centrated for 5 minutes at 20mA, and run for I hour 45 minutes at 3SmA. A solution of Candida krusei cytochrome c (>0.2mM) 7.5 HEPES buffer was degassed Care was taken to result in protein denaturation. 7 avoid cient to result in a the the cycles serum saturation complex of bottle with was a also. argon and evacuation, the solution anaerobically to the l.2ml aliquots and diluted (NaAc) buffer. to approximately Ultrafiltration of the removal, was carried dilution, and and of the bottle were pH mixture An aliquot Henson, A., 32, 162 (1970). Mussell white, C=crosslinker P. J. Colloid of ruthenium the removed was least by equiv- cycles sodium 5 reconcentration. (w j w), with containing at 7 J., more to suffi- degassed thirty for T=total acrylamide concentration as a percentage of the total acrylamide known buffer argon, several with reaction of out 6 Mitchell, with After lOml is thoroughly evacuated. proceeded, temperature. when combined again the reaction room volume Once saturation IOOmM pH which was syringe concentration, solution, and As of the Meanwhile, was degassed transferred upon of bottle at evacuation cytochrome. mediately foaming serum protein concentration of 0.2mM cytochrome solution several in a in acetate begun three of via im- cycles CoEDTA- concentration Inter fac . S ci. -36solution was added to the concentrated protein to oxidize the ruthenium Removal of the oxidant and exchange of the buffer were accom- label. plished by repeated ultrafiltration and dilution with 10mM points: 17 elec- trophoresis buffer in Centricons. Aliquots were removed at the following min., 50 min., 1 hr., 1.5 2 hr., 3 hr., and hr. A second tion run was also monitored by electrophore- sis, with hr., 18 under time hr. similar hr., conditions points of and 24 min., hr., Proteins were 15 hr. 2 time 4 hr., fixed 4 hr., and 6 min., hr., stained 8 and 34 reac- hr., 12 the gels preserved according to the instructions of LKB Application Note 306. Modification of Candida krusei cytochrome c on a preparative scale The reaction thirty tion minutes, of vessel was Several to a the was carried out as described usually on ruthenium filled milliliters 50-100mg complex was often argon and the with of the scale of the supernatant spectrophotometric cell which had An extinction coefficient of 44M- 1 cm- 1 concentration, 23:1 [Ru]:[cyt], after the supernatant ruthenium protein solution, which the generally dissolving turbid, in was were transferred allowed was used The solu- case the to settle. anaerobically the buffer blank). to calculate the ruthe- that the molar ratio was to -30 equivalents. 8 As mass equal solution was transferred changed period of with argon , and measured (versus was such a which solid a cytochrome for cytochrome. been flushed absorbance of the solution at 414nm was nium above, from anaerobically red 8 Ford, P., Kuempel, J., Taube, H . lnorg. Chem. 1, 1976 (1968). to to the orange, an -37indication of its reduction. Separation of the pentaammineruthenium modification reaction products The oxidized reaction mixture was applied to a column of CM-52 equilibrated in 100mM pH 7 sodium phosphate, and eluted with same. The gradient 2.1. programs for for analytical control and of FPLC preparative functions runs are and the presented solvent in table The binary solvent system included 50mM pH 7 NaPi ("A") and 50mM pH 7 NaPi, 1M in NaCl (''B"). 9 Product elution was monitored by absorp- tion at either 280nm or 405nm. RESULTS Purification Purification remove of deamidated asparagine and the commercial protein, glutamine) cytochrome i.e., protein in have hydrolyzed been which product is required to amide side chains (of to the corresponding acid (Asp or Glu). 10 that The purification achi eved on CM-52 was judged to obtained the numerous and with better Ambe rlite CG-50, as former resolved bands for cytochrome bands and the be superior to resulted of in more identical lot (f i gure 2. 1). Although the product on CM-52 9 number of elution pattern of the varied from lot to lot, a minimum of four bands was the method of preparation took into account the dependence of the sodium phosphate ionization constant; Enzym es: Physical Principles, Wiley (New York), 1972, p . 34. 10 Harding, J. Adv. Prot. Chern. 37, 247 (1985). ionic strength Gutfreund, H . -38- Table 2.1. FPLC programs for analytical (a) and preparative (b) separation of the oxidized reaction mixture. a. volume (mL) 0 2 32 34 36 38 42 b. volume (mL) 0 40 296 312 328 344 360 instruction %B=O 1rnUmin 0.2cm/mL valve position = 1.1 2rnl)mark monitor clear data level%= 3 integration on valve position = 1.2 %B=50 %B=100 integration off %B = 100 %B=O %B=O valve position = 1.1 instruction %B=O 4rnl)min 0.05cm/mL valve position = 1.1 8rnUmark monitor clear data level%= 4 integration on valve position = 1.2 %B=O port open %B=50 %B=100 integration off port closed %B = 100 %B=O %B=O valve position = 1.1 0 so SO 100 100 - ---:::;.-------------- ------ . fr1ct ton ISO fraaioo ISO 200 200 lSO I ~ o.os 0.10 2SO 441 ·/ /- t O I.IOmM 300 J 9 i o.u ~ - - - - - --l IOOmM ~----------- - - - - - - - -- - - -- - - (b) (a) Figure 2.1. Elution profiles of commercial C. krusei cytochrome c from (a) Amberlite CG-50, and (b) CM-52. See text for conditions. ~ -f 0 ~ ~ ] < .<> ll J ~:!. O.l.l I I <:0 w -40- always seen: the major two bands, one brown, likely to the top great be of reduced the resin affinity of component was usually and the other protein). A fourth bed, this but could material immediately pinkish-orange for the with anionic (the by latter is band remained at 150mM NaPi. The suggests this reddish-pink be eluted preceded support band may be polymeric cytochrome. 11 In more impure lots, and in cases where aged repurified, an cytochrome band was observed. samples were additional unretained The purified cytochrome was found to be homogeneous by PAGE and FPLC. Time analyses In the electrophoresis product of mobility After one cytochrome (figure half hour than taken immediately after mixing, a that of native cytochrome c is evident. (figure 2.2a, lane reacted to form this product. products of even greater has 2.2b), higher sample 8), most At mobility of longer are the native reaction times observed, and a concommitant decrease of the initial product is observed. Interpretation ambiguity. species. ular What appears Electrophoretic dipole absolute of moment charge, as pentaammineruthenium His116-Mb) under electrophoretic results to band be a single mobility may or perhaps demonstrated myoglobin conditions be more a sample-gel by the to may not be function interactions, resolution derivatives identical is of one of described (data not shown). 1 1 Margoliash, E ., Lustgarten, J . J. Bioi. Chem. 237, 3397 (1962). or the rather more molec- than the singly-modified (A RuHis12-Mb 5 those without in and A Ru5 figure 2.2b -41Figure 2.2. reaction Time by krusei analyses of the C. cytochrome polyacrylamide gel electrophoresis. c Migration modification is from top to bottom, towards the (negative) cathode. (a) Upper: pH 8.9 Tris-glycine buffer, 5 hr. reaction Lanes 20 1-4: mg/ ml, C. native krusei 12mg/ ml, cytochrome c, 3mg/ ml, lOmg/ ml, Lane 5: modified product respectively; from a previous experiment; Reaction time points: Lanes 6, 7, 8, 9: 0, 17, 34, and 50 min., respectively; Lanes 10, 11, 12, 14, 15, 16: I, 1.5, 2, 3, 4, and 5 hours, respectively; Lane 17: as for lane 4 No sample was applied to the defective slot of lane 13. Splotch- iness of the lane 9 sample is due to overfilling of the slot. (b) Lower: pH 7 imidazole-phosphate buffer, 24 hr. reaction Lane 1: native C. krusei cytochrome c, reduced; Lanes 2, 8, 15: native C. krusei cytochrome c, oxidized; Reaction time points: Lanes 3, 4, 5, 6, 7: 15 min., hr., 2 hr., 4 hr., and 24 hours, 6 hr., respectively; Lanes 10, 11, 12, 13: 16: produc t 8, 12, 18, respectively; Lanes 9, from previous experiments (thought to be singly modified) Lane 14: mixture, native and modified cytochromes -42- 1 1 2 3 2 4 3 5 4 6 5 7 6 8 7 9 8 10 9 11 10 12 11 13 14 15 16 17 12 13 14 15 16 -43Modification fact that the might be of protein expected the krusei C. molecule to repel has cytochrome c high positive a the positively While protein charge has been suggested cation of horse heart cytochrome c, tion of two of local factors, prominently species including in net rapid, charged and for the despite charge, 12 ruthenium to account for plastocyanin, 13 charge is which complex. the slow modifi- relative rates of reacthat this result demonstrates steric constraints, as well as rate with which determining the three were evident 2.3). Additional the modification figure progresses (vid e infra). Chromatography Typically, bands reaction mixture observed, particularly when presumed to be (figure redox column volumes for to elute the elution of bands period of oxidation isomers. the first upon Elution with lOOmM buffer, and buffer solution once elution of Typical yields of the three recovery was as high as occasionally were were -3.7 and -7.8 The third band was slow often removed the second component components were 12%, 50%, and were 95%, was were but half -hour was short, and volumes two components. the tails of bands separation of the with a was stronger complete. 12% (total excluded from determinations of yield). An 12 elution profile for products by FPLC is estimated to be +6 at pH 7 for the reduced form on the basis of the amino acid sequence, assuming all Lys and Arg are protonated, all Glu and Asp are deprotonated, and a pKa of 7 for all His (i.e., half are protonated at pH 7) 13 Jackman, M., McGinnis, J ., Chem. Soc. 110, 5880 (1988). Powls, R., Salmon, G., Sykes, A. J. Amer. -44- r--- ---- 'E c 250 200 0 v ......... . 3 0 ~ c z 0 .D .... 0 0 "' I .D 0 :P. I 0.5 I ,.-------' 150 0 0 100 - -- -------------------- -- -- ------ -- --------- ~ 0 100 200 300 400 500 600 fraction Figure 2.3. Chromatogram for the elution of the products of the reaction of pentaamrnineruthenium(ll)aquo and C. krusei cytochrome c from CM-52. -45shown in figure from CM-52 tional and Individual 2.4. FPLC runs of each product recovered confirmed that the order of elution is identical for convenfast protein chromatography. Reduced native protein is also resolved by FPLC, and elutes prior to the oxidized form (as expected on the basis of total products are resolved Retention times minutes. Baseline ative could tion volumes broad and on a be protein for charge). also, the but four resolution of Redox coelute isomers of the native with modified cytochrome. species are 12.8, 14.6, 15.9, the native and first modified deriv- at 18%B, reten- However, bands were achieved under other conditions 9.5mL and 15mL, respectively). of the (elution and 19.3 resolution was compromised when the procedure was attempted preparative basis. No evidence was obtained from FPLC for the first two significant amounts of additional products. Efforts bands eluted NMR, and toward characterization from CM-52. EPR to be The first were focused on band was shown by electrophoresis, native cytochrome c. Characterization experiments for the second band are described in chapter 3, and this product may be referred to as the modified cytochrome. The third band tially characterized, and this work is described in appendix 6. has been par- -46- i .... 100 0 .04- 1- 0.03- 80 ~ -E c ~ 0 co 60 C\J Q) <..> I 0 .02- ) c I Cl3 ..0 .... ~ 0 ~ CD " I I 0 / (/) ~ " ..0 Cl3 I 40 " " 0.01 I I " / - 20 I / I \)\. -\. ~ 0 - - / I / I 0 ll I I I I I 0 10 20 30 time Figure 2.4. 0 (minutes) Chromatogram for the analytical FPLC separation of the oxidized modification reaction mixture(--- gradient, 0-50%B in 32 minutes at lmL/rnin; injection at 2 minutes; unretained product is CoEDTA -; other products in order of elution are reduced native cytochrome c, oxidized native cytochrome c, and the two modified products). -47- CHAPTER 3 CHARACTERIZATION OF THE DERIVATIVE -48MATERIALS and INSTRUMENTATION Protein All of the in chapter 2. CM-52 native protein samples used The modified cytochrome was were purified as described used as recovered column, except for peptide mapping, in which from the case the derivative was further purified by FPLC (see chapter 2 for details). Electron Paramagnetic Resonance Spectroscopy EPR spectra Band were recorded spectrometer operating in on the a Varian E-line absorption mode. Century Sample series Xtemperature was maintained with an Air Products Heli-Tran cryostat. The chloride salt of plex, [A 5 Ru(His)]CI 3 ·H 2 0, Sundberg and Gupta. 1 the pentaammineruthenium(III)histidine com- was prepared by Kathy Yocom by the method of Nuclear Magnetic Resonance Spectroscopy 1H NMR spectra instrument operating acquired stored and were at as recorded at on a Bruker AM-500 Typically, 128 transients were samples) 32000 sampled points 500.13MHz. 16000 (pH 5.2 25°C or of the FID. UV-visible Difference Spectroscopy Spectra Quartz were cuvettes recorded (1 em on pathlength) a Shimadzu were UV -260 obtained Cells. 1 Sundberg, R ., Gupta, G . Bioinorg. Chem. 3, 39 (1973). from spectrophotometer. NSG Precision -49Peptide Mapping Submaxillaris protease (SMP) was obtained from the Pierce Chemi- cal Company. The ReactiTherm heating module was purchased from the Pierce Chemical Company. The pumps an chromatograph controlled Altex mixer, consists by a Model and several of two 201JL Model standard Scientific flow maintain the 110-10 analytical restrictor, effluent Microprocessor columns connected variable flow part under II OA 420 by two Rheodyne 7060 switching valves. Altex/ Hitachi Altex pressure metering controller/programmer, in parallel and selected Effluent was monitored with an wavelength detector fitted with (model 155-00). An Upchurch cell U-444, solvent is (40 placed psi), after and a the detector to prevent degassing in the detector cell. All tubing diameter 0.02" is prior type to 316, the 1/16" injector system is also equipped with a valve, an Alltech #9299 in-line O.D. and stainless 0.01" Rheodyne of the injector. The 7037 pressure relief after model pressure gauge, steel, and an SSI internal Prime/Purge Valve (#02-0292). Column temperature temperature was maintained control system. Samples constant (:!: 0.5°C) were injected by a manually Waters via a Rheodyne model 7125 syringe loading sample injector, which accommodated loops up to 2001JL Shimadzu C-R5A Sample injection in volume. Chroma topac and was A Hitachi / EM used to Science record and integrator integrate or peaks. the recorder were synchronized triethylamine (TEA), and trifl uoroacetic activation of manually. Acetonitrile, methanol, -50acid (TFA) viously. type were HPLC 4mm Millipore HA filters held grade. Water was purified type HV 0.451-1 filter units, and 45mm Millipore filtration apparatus, were in a glass as described pre- used for samples and aqueous solvent mixtures, respectively. The HPLC column utilized was a 4.5mm x 15cm silica-based reverse phase C4 type, manufactured by Vydac (catalog no. 214TP5415, no. 270028, #21). A Savant SpeedVac, model SVCIOOH, containing a RH40-ll rotor, was attached to a vacuum pump via a liquid nitrogen trap, and was always operated with the sample compartment maintained at 40°C. Spectroelectrochemistry Optically-transparent thin-layer electrochemistry (OTTLE) cells were constructed by Walther R . Ellis using two 17mm x 35mm rectangles of fine Paul, gold MN) structed as with mesh (500 lines/inch; the working electrode a copper-constantan 60% transmittance; (figure 3.1 ). 2 Interconics, The microthermocouple cell (Omega was St. con- Engineering) which protruded into the sample compartment and, in conjunction with a Fluke 2175A digital thermometer (:!: 0.2°C), allowed for the sample Platinum the counter electrode, miniature saturated calomel electrode cell held the compart- and temperature. the reference electrode wire was a was used as monitoring of (SCE) obtained from Sargent-Welch. ment 2 The assembled of a Cary electrochemical was 219 spectrophotometer by a Ellis, Walther Robert and Copper Proteins.", Technology, 1986. special in mount cell built in the Jr., "Spectroelectrochemical Studies of Heme Ph.D. Dissertation, California Institute of -51- electrode \ ports I\ I I \ I \ -~ LJ~ 1- "" "" 1- ""' ,.. ,. r ""' ,., :"' > optical window (gold minigrid) Figure 3.1. Front (left) and side (right) view of an optically transparent thin layer spectroelectrochernistry cell. reference compartment and jacket--~ fine f r it sample compartment and r---.....;..;:::,..._jacket ports for circulating water Figure 3.2. Dual-jacketed electrochemical cell for differential pulse polarography. -52Caltech machine shop. Sample temperature was maintained by water from a bath constant-temperature The reference electrode which (SCE) circulated through maintained at was this cell ambient holder. temperature, which was measured with an ordinary mercury thermometer (graduated in 0.2°C). The search cell Model potential 174A was controlled polarographic with a analyzer, and IJ=0.1M, pH Princeton was Applied measured to Re- O.lmV with a Keithley 177 digital voltmeter. The cytochrome, 0.4mM been purified twice by in cation exchange pen taammineru thenium(III)pyridine 7.0 sodium phosphate, chromatography. A had "pinch" of [R u(NH 3 ) 5 py](Cl0 4 ) 3, perchlorate, prepared by Walther R. Ellis, 3 was added as a mediator (E 0 = 298 mV). 4 Differential Pulse Polarography shop, The electrochemical cell and consists compartments of two 3.2). sintered frit (figure jackets. Each compartment of a thermocouple, and Teflon stopcock to a was The manufactured which are compartments contains a port in to of nitrogen or Caltech connected by independent allow for argon the Glass a have the sample compartment can line the fine water insertion be connected via a degassing. Incuba- for tion of the two compartments was achieved with Forma Scientific models 2095 and 2132 circulating baths. A Systems. 4mm gold Miniature button electrode saturated calomel was purchased electrodes were 3 Cummins, D., Gray, H. J. Arner. Chern. Soc. 99, 5158 (1977). 4 Matsubara, T., Ford, P. lnorg. Chern. 15, 1107 (1976). from Bioanalytical obtained from -53Sargent Welch. Platinum wire was used as the counter electrode. Equipment to scan the potential and record the current included a Princeton Applied Research Houston Instrument 2000 Model XY 174A recorder. Polarographic Temperature Analyzer and and potential a were monitored by a Fluke 2175A digital thermometer and 8010A digital multimeter, respectively. Thermocouples, catalog number STC-TT-T-36-36-STD (type T copper-constantan) were obtained from Omega Engineering. Aldrithiol-4 (4,4'-dipyridyl disulfide) was used as supplied by surface areas were Aldrich. Modeling Studies All conducted modeling using and the calculations program of Biograf distances and (Biodesign, Pasadena, CA), version EPR spectra were recorded with a modulation amplitude of 16G, a 1.5. METHODS Electron Paramagnetic Resonance Spectroscopy modulation frequency of 100KHz, and a microwave power of 0.2mW. after The temperature the spectra at were the sample acquired position using a was measured gold-chromel before and thermocouple. Spectra were recorded at -12K. NMR Spectroscopy pH 5.2 (lyophilized lyophilization. NMR from samples prepared to two cycles samples were exchanged water) Other were of by subjecting dissolution in n 2o by into cytochrome D 20 and diafiltration -54in Centricon-10 units. o 2 o. NaOD and DCI in measured with a Solution pH was adjusted using solutions of pH values quoted are direct meter readings, as micro-combination electrode within one hour after the spectra were recorded. was The sodium present in values are salt of samples at quoted (DSS) 2,2'-dimethyl-2-silapentane-5-sulfonate concentrations in parts per million data were fit by of (ppm) -I-2mM. Chemical downfield of shift the methyl resonance of DSS. Titration a least-squares analysis to the of interest for and 6 the expression pH = pKa + log [(6 +-6 )/(6-6 °)] where the 6° and 5+ are the chemical shifts of the deprotona ted and protona ted species, proton respectively, is observed chemical shift for the proton at a pH where both forms are in equilibrium. The derivation assumption is made of the above expression is as that the chemical shift a linear is follows. The function of concentration, i.e., where and [His+) [His] represent the is is the the concentration total fractions of histidine of histidine [His 0 ] ([His 0 )/[His]) (where deprotonated form histidine the protonated concentration. deprotona ted of the residue), chemical shift can be expressed as follows: 6 = 6+(x) + 6°(y) Since x + y = I, Letting protona ted is the histidine x residue, and ([His +)/ [His]) concentration respectively, any of y and the observed -55o+(I-y) + o 0 (y) 0 = o+(x) + 6°(1-x) Solving for each variable, and Therefore, using the Henderson-Hassel balch equation, 5 pH = pKa + log [His 0 ]/ [His+] = pKa + log (y j x) pH = pKa + log [(o + -o )/ (o-o 0 )] UV-visible Difference Spectroscopy Water or matically buffer subtracted was used from the to record spectra a baseline, recorded which was thereafter. auto- Solutions of -1.7mM of the native cytochrome and modified product (to correspond to a maximum spectra were memory. absorbance recorded Digital subtracted corded. The from Any different the 2) were separate prepared, and their of the spectrophotometer 410nm (an isosbestic channels readings at 450-250nm point) The spectrum of the native protein was multiplied A 410 (modified)/ A 410 (native) concentration. of into absorbance were noted for each. by of corrected proportions of correct spectrum spectrum absorbance at to of of the for slight differences in the native cytochrome was the re- derivative, and result 416nm in the difference spectrum, indicative oxidized and reduced cytochrome in each Row (New sample, was used to correct the 300nm reading, 6 as follows: 7 5 Boikess, R., Edelson, York), 1978, p. 481. E. Chemical Principles, Harper and 6 this neglects any contribution from the ruthenium absorbance at 416nm 7 Margoliash, E., Frohwirt, N. Biochemical J. 71, 570 (1959). -56c = llA 416 + (129100 - 88800) M- 1 cm- 1 llA 300(corrected) = llA 300 (observed) - c(21000 - 13000)M- 1 cm- 1 where c is the cytochrome concentration of the reduced species that native protein. An minus in the in the extinction modified coefficient of 2000M- 1 cm- 1 was u sed to estimate the ruthenium concentration. 8 Peptide Mapping Both The strategy the native known and of the peptide mapping and modified proteins are highly selective specificity, experiments digested a protease of pep tides, each of which can be separated by reverse phase HPLC. 9 The enzymatic the carried out under very mild two follows. of is produce as sets reaction to with is conditions, and A 5 Ru moiety remains bound. A comparison of the peptide chromatograms should reveal peptides which are shifted in retention (in either tion, depending on net charge of peptide) due to the the the direc- presence of trivalent ruthenium. Confirmation of the presence of ruthenium on a peptide UV -visible tide(s) the upon can then be identified protein) by total amino the course of stroyed 8 depends in (i.e., acid acid spectral related analysis The pep- sequence of characterization. to (the hydrolysis). 10 a specific A 5RuHis complex It is important is de- that no An extinction coefficient of 1880M- 1 cm- 1 has been reported for the pentaammineruthenium imidazole complex: Sundberg, R ., Bryan, R., Taylor, 1., Taube, H . J. ~mer Chem. Soc. 96, 381 (1974). An extinction coefficient of 2100M- em- 1 has been reported for the pentaammineruthenium histidine complex: Sundberg & Gupta, p. 39. 9 Hermodson, M., Mahoney, W. Meth. En zym. 91, 352 (1983). 10 Matthews, (1981}. C., Recchia, J., Froebe, J. Anal. Biochem . 112, 329 -57more than present one on complete, potential a single and the site of peptide. peptides modification In are principle, (of if well-resolved, each the ligand protein only the type) be cleavage is shifted peptides need be analyzed. 250 units ( -1 mg) SMP were dissolved in 2.5mL 1 mM HCl, and stored in 5001-lL aliquots in liquid nitrogen until needed. A 251-lL aliquot of SMP solution was added to each of eight microcentrifuge tubes containing 5001-lL of I %(w/ v) NaHC0 3 (-pH 8.3). 6 hours. a -2mg/ mL cytochrome solution in A second 251-lL aliquot of SMP was added after The reaction mixtures were incubated at -37°C for a period of 32 hours; 331-lL 2M HCI were added to halt the digestion. Samples were then frozen until analyzed. The reaction mixtures were analyzed by The HPLC. peptide products were separated on a C4 column, maintained at 27°C, at a rate of 1 mL/ min., O%B, 0-1 O%B flow using the following elution method: 5 minute hold at in 12 min., 10-15%B in 18 min., 5 min. hold at 15%B, 15- 40%B in 64 min., 40-100%B in 10 min., where A = O.l%TFA, O.l%TEA in H 2 0 , and B prepared = fresh daily). aliquot of Peptides were detected ing at 0.1%TFA, 0.1%TEA, ruthenium least Samples 1001-lL were 10%H2 0 by was were used to in CH 3CN (mobile phase thawed flush their absorption at identified as such by and and filtered, fill the 220nm, and their and 201-lL was an loop. those contain- absorption at 300nm an milligram (i\max) and lack of absorption at 400nm. For preparative runs, the 32-hour digest of entire of cytochrome was taken to dryness in the SpeedVac, and then dissolved in centrifugation 1251-lL of solvent "A". through a 4mm type HV 0.451-l Following filtration (by Millipore filter in a microcentrifuge tube) -58the entire sample was loaded into a 2001-lL loop, and eluted as described The above. effluent was typically delay of 5001-lL to allow collected for the dead peptides in lmL fractions, with a volume between the detector cell and tubing outlet. UV -visible spectra of recorded. Spectra were were also suspected of recorded of containing the ruthenium peptide solutions made basic by the addition of 0.1% (v jv) triethylamine. Fractions appropriate further peptides including and concentrated purified by interest in the SpeedVac. elution from the same C4 Pooling of conditions similar to the original. tion, purification were repeated and of until a were pooled as The peptides were column, using elution fractions, single, concentra- symmetric peak was evident in the chromatogram. The peptide solution, as eluted from a hydrolysis tube and lyophilized. The residue (Pierce}, and milliliter of constant-boiling jected at least three freeze-pump-thaw vacuum, and placed off to under ll0°C. HCl the HPLC, was transferred to in a the cycles. dissolved solution The Reacti-Therm tube block in one was sub- was closed maintained at After 24 hours, the sample was removed from the block and taken to dryness on a vacline or in the SpeedVac. residue, were was and this repeated. was again The residue were derivatized in the taken to 1mL H 2 0 dryness. was dissolved in presence of at least was added to the Dissolution and drying borate/CH 3 CN and aliquots five equivalents of FMOC- Cl. Amino acid analyses were carried out as described in appendix I. A native of the peptide mixture resulting from a 32 hour utilized to optimize krusei cytochrome c was digestion products. An initial C. five minute SMP hold digestion of the separation at O%B was -59chosen to might otherwise tion, preparative permit however, interfere can be samples with to the obtained be washed peptide without this free of salt resolution (better hold the in which resolu- absence of requires the large quantities of salt). Spectroelectrochemistry Spectroelectrochemistry presence of a mediator, similar to that this to case the of serves protein, the to of a krusei C. redox active cytochrome species which with the electrode, is slow to itself reduction potential The [A 5 Rupy] 2 +/ 3+ used in protein under study. interact of c and transfer shuttles electrons electrons with the using a non- and the ref- This is slow to potential of electrode. 2 Ru + ) ( Fe3+-cytochrome Ru 3+ Fe2+-cytochrome electrode Figure 3.3. Schematic representation of the role of [A 5 Rupy] 2 +/ 3+ (="Ru 2 +/ 3 +") as a redox mediator. The isothermal erence done configuration, electrode for equilibrate the spectroelectrochemical are reference which maintained convenience, following in as the changes electrode studies was were the working electrode at different temperatures. reference in performed electrodes temperature. corrected for The slight are formal very fluctuations in the -60room temperature using the following equation: 11 E 0 (T) = 0.2444 - 0.00066 (T - 25°C) description A follows. the direction) through a extremes, i.e., oxidation eight to ten tion of time, either commencing ative as of total increments. with the potential is where step. UV-visible the the for presence each covers optically stepped the protein, as a absorbance is a absorbance constant spectrum the as is points. relative in func- redox-sensitive, equilibration) isosbestic potential, is electrochemical monitored electrode of both of Once of (Eapplied) reduction its and for reading total mediator, portion electrode which protein wavelength check absorbance The a a to range and experiment spectroelectrochemistry working at protein, large corded potential typical a of (in The of is (indic- obtained, feasible Using amounts is the of refinal reduced and oxidized cytochrome can be calculated. The relevant expression is: where Ar and A0 totally oxidized dividual potential tion, a plot the are the absorbances of the totally (ox) protein, and A·1 of interest. For systems obeying versus the of the applied potential molar ratio of reduced to 2.303RT /n1 and an intercept equal is oxidized to the species the formal reduced absorbance (red) and the in- at the Nernst logarithm (base yields line a reduction equaI 0) of of slope potential of the redox couple of interest: 12 II Ives, D., Janz, G . in Reference York), 1961, p. 161, cited by Lin, 1766 (1965). 12 Bard, A., 1980, p. 51. Faulkner, L. Electrodes, J., Breck, Electrochemical Methods, Academic Press (New W. Can. J. Chern. 43, J. Wiley (New York ), -61E = E 0 - (RT /n1) In ([red]/[ ox]) where R, number T, of n, 1 and equivalents are of the universal electrons gas constant, involved in the of formal temperature, redox the process, and the Faraday constant, respectively. The yields temperature dependence information regarding nonisothermal cell the the thermodynamics reduction potential of reduction the process: For ture a gradients can be arrangement neglected, the in which value the of effects the of entropy temperadifference derived from a temperature dependence study is that of the redox couple only, Using the value 15.6eu of for the hydrogen half reac- tion, 0 6S cell = 6Src - 15.6eu a value for the complete reaction (vs. NHE) can be obtained. 14 The visible lamp of the spectrophotometer was one hour prior to commencement of the experiment. corded (i .e., automatically OTTLE cell filled with subtracted water. from A subsequent lit for at least baseline was re- scans) the with C. krusei cytochrome c was monitored at 548nm; the analogous band in horse heart cytochrome c (i\max=550nm) has 13 Yee, E., Cave, R., Soc. 101, 1131 (1979). 14 Guyer, K., Tyma, P., Weaver, M. J . Amer. Chem. T aniguc . h'I, V . , Ell'IS, W., C ammarata, V ., We bb , J ., A nson, F ., G ray, H . Adv. Chem. Ser. 201, 51 (1982), K. Kadish, ed., Electrochemical and Spectrochemical Studies of Biological Redox Components, American Chemical Society, Washington D.C., and references therein. -62a 18500M- 1 cm· 1 in extinction for difference of this wavelength. 15 The mediator wavelength. 16 The 600-500nm (0.5nm/ second) after has higher equilibration nated The energies. with each negligible absorption at direction temperature. The each potential step; strong of meaningful spectra the recording of the potential data for the native cytochrome lowest temperature and steps with a single sample, starting from the working upward to protein. Two obtain all same prolong preparative this the recorded the at range of the visible spectrum was scanned Soret absorption at 409nm prevented at the two redox states at the "life" of the batch) were required to is to denote reduction cannot be to site the was alterwas samples data (of for the modified derivative. The E0 ' symbol used potentials determined at pH 7, as per convention. Differential Pulse Polarography Spectroelectrochemistry potential of the ruthenium in used determine modified protein, the reduction as absorption by the heme dominates at all wavelengths of absorbance of the ruthenium chromophore. Differential pulse polarography (DPP) has been used while the instead. In current would 15 DPP, is the potential monitored. produce signals Hence, which £(oxidized)=9100M-l cm· 1 ; Frohwirt, p. 570. 16 Ford, P ., Rudd 1187 (1968). D ., range a would of interest mediator is cannot obscure the scanned be data. £(reduced)=27600M- 1 cm·l; Gaunder, R ., Taube, H . J. employed, as it Some success has Margoliash & Amer. Che rn. Soc. 90, -63been achieved facilitate heart the electron been between redox protein inactive and molecules electrode, which and horse reported on extensively in this regard. 17 postulated that 4,4'-bipyridine enhances the rate of the while also residues. 18 A similar required in the case of C. to denature on the the cytochrome by adsorbing via its surface to surface disulfide electrode lysine (Aldrithiol-4), c, cytochrome electrode promoters, been transfer krusei of has tron 4,4'-dipyridyl use transfer c cytochrome has It in as more is the yeast protein readily than the tends horse elec- hydrogen-bonding cytochrome. molecule, This process is delayed somewhat by Aldrithiol, which seems to have a greater affinity for the electrode by virtue of its disulfide moiety. 19 Like the spectroelectrochemical a nonisothermal by polishing arrangement. the gold surface on fine alumina, and then electrode was through the -1.0 a +1.3V (vs. piece thoroughly electrochemically to DPP The button electrode slurry of then experiments, cleaned SCE) of was performed was cleaned felt by potential manually moistened rinsing it with in with a water. The sweeping (20mV /s) twice range IJ=0.1M pH 7 0.3mM in in NaPi. This process was repeated for each complete scan. Samples Aldrithiol-4 with a were 0.25mM in cytochrome NaPi. in 1J=0.1M pH 7 gentle stream of argon, The potential range of -0.4V and Samples except to 0.35V approximately were while continuously scanning was scanned at a the purged potential. rate of 2mV j s 17 Eddowes, M., Hill, H. J. Amer. Chern. Soc. 101, 4461 (1979). 18 Eddowes, (1979). 19 Niwa, (1988). K., M., Hill, H., Furukawa, Uosaki, M, Niki, K. J. Arner. Chern. Soc. 101, 7113 K. J. Electroanal. Chern. 245, 275 -64drop time of 0.5s and a using a pulse amplitude (fl.E) of 2SmV. For a reversible system, Parry and Osteryoung have shown that the peak potential, is EP' related to the half-wave potential, by the following correction, the reduction E 112 , expression: 20 for a cathodic wave. potentials were maximum current The diffusion assumed to Rather determined from to apply this be the average forward coefficients be than equal, of (anodic) the and of the backward potentials (cathodic) of scans. reduced and oxidized species are i.e., the half-wave potential is potential were made at the beginning hence equal to the formal reduction potential. 21 Three determinations of temperature and Temperature and of each scan, thermocouple was never rise to end presence gave averaged. and reduction upon present reversing in disturbances was in measured at the scan direction, but the while scanning, as its solution the each signal. Data was collected approximately every five degrees, from 5 to 40°C. Modeling Studies The starting point for structures was the resolution) coordinates reduced data C. krusei model cytochrome both horse and tuna cytochrome set, as obtained crystal from the structure 22 (1.5A Brookhaven Data 20 Parry, E., Osteryoung, R . Anal. Chern. 37, 1634 (1965). 21 Bard and Faulkner, p. 160. 22 Takano, T., ences therein. Dickerson, R. J. Mol. Bioi. 153, 79 (1981), and refer- -65Bank. 23 The sibility, as occurs. The reduced the six-residue N-terminal C. of the C. krusei is not expected Residue 103 of 104 further extension N-terminal to alter the the closest of c but extend krusei extension in result: both N-terminal cytochrome removed, which the histidine acces- 104 of horse cytochrome, and polypeptide which of sidues of tuna were excluded (residue from in study re- of state the 103 portions redox for beyond is the used reaction 17A is was modification least this state the resides calculations and 39 are at -6 through -1). of residue 33; result in an overestima- cytochrome, 33 (residues 5A may Inclusion area residues within its exclusion cytochrome). surface residue the residue tion of the His33 accessibility in horse heart cytochrome. To construct cytochrome the appropriate chain was placed. 24 then No of another species from tuna made, and each residue substitutions were aligned as possible with that which energy minimization of the resultant particular residues solvation or well as new side it re- struc- tures was performed. Surface areas of were determined by the method of Connolly 25 using all atoms (except water molecules) within 7A to define to define the local the topology. Probe solvent-accessible radii and of 1.4A and reagent-accessible 3.5A were used surfaces, re- specti vel y. 26 23 Bernstein, F., Koetzle, Rodgers, J., Kennard, 0., 112, 535 (1977). Brice, M., T., Williams, G., Meyer, E., Shimanouchi, T ., Tasumi, M J. Mol. Bioi. 24 A recent survey of homologous proteins has found this to be a valid assumption: Summers, N., Carlson, W., Karplus, M J . Mol. Bioi. 196, 175 (1987). 2 5 Connolly, M. Science 221, 709 (1983). -66RESULTS Electron Paramagnetic Resonance Spectroscopy EPR spectra C. modified signals are of 3.12 and for krusei the model complex, c cytochromes evident in 2.2. The the are [A 5 RuHis) 3+, presented native spectrum, third, high field and signal in These Brautigan et a/., 28 distinct 3.2, classes 2.05, visiae & iso-1 results are was 1.39. fair by g values not observed, and latter cytochrome c and cytochromes c 2 and c SSO· the His18 ligand. The author hydrogen could find with those of into two 1.25, and S. cere- cerevisiae iso-2 be grouped and the includes is highly concentrated tuna while the values horse, cytochromes, in g to & includes differences Two 2.26, class from 3.4. 3.06, former result figure of The to and agreement cytochromes c could who found characterized in native correspond reported to be broad and evident only in spectra of solutions. 27 and S. The difference was suggested bonding no or protonation of reported EPR previously data for the Candida cytochrome. The low field signal at a g value of 2.89 for the ruthenium model compound, [A 5 RuHis] 3 + is similar to that observed by Margalit et a/. (g=2.95). 2 9 These workers did not report any higher field signals. 26 The Ru-N bond length in hexaammineruthenium(II), [Ru(NH 3) 6 ) 3 +, is 2.1A: Stynes, H ., Ibers, J. lnorg. Chern. 10, 2304 (1971). Ru-N bond lengths of 2.07-2.12A have been reported for (Lhistidinato)pentaammineruthenium(III) chloride: Krogh-Jesperson, K., Westbrook, J ., Potenza, J ., Schugar, H. J. Amer. Chern. Soc. 109, 7025 (1987). 27 Palmer, G. in The Porphyrins, D . Dolphin, ed., vol. Chemistry, Part B, Academic Press (San Francisco), 1979, p. 313. IV, Physical 28 Brautigan, D ., Feinberg, B., Hoffman, J., Blumberg, W. J . Bioi. Chern. 252, 574 (1977). E., Peisach, Chiang, H.-J., 29 Margalit, R ., Kostic, N., Che, C.-M., B., Margoliash, Blair, D., --67- (a) (c) 2000 3000 4000 5000 Gauss Figure 3.4. EPR spectra of (a) 2mM native C. krusei cytochrome c, (b) 7.73mM [AsRuHis]Cl3, (c) 1.85 mM AsRu-modified C. krusei cytochrome c. Protein samples in j..l=O. l MpH 7 NaPi, complex in O.lM NaAc, pH 5. Initial sample temperatures (rose as much as 0.5K during a run) : (a) 12.6K, (b) 12.8K, (c) 13.3K. Other conditions: modulation amplitude, 16G; power, 0.2mW; gain 1.6 x 104; frequency: (a) 9.245GHz, (b) 9.25GHz, (c) 9.247GHz. -68The spectrum of the modified cytochrome a superposition suggesting more, of the modification it confirms native cytochrome c at histidine residue that a the integrity of derivative appears to and the model has complex spectra, Further- occurred. cytochrome be fold and iron coordination sphere are maintained after modification. This EPR spectrum does not rule out methionine modification, as a pentaammineruthenium complex would be expected to have a reduction potential of -500mV, and therefore would not be oxidized by the CoEDTAThe low spin d 6 ruthenium(II) configuration would be complex employed. EPR silent. NMR Spectroscopy Two sharp resonances are evident in the NMR spectrum of native C. krusei c, ferricytochrome presumed to (figure 3.5a). assignment be the The to at 58.62 signals pH of and the 58.76 C-2 dependence of residues: both histidine (ppm protons these with in resonances are not found spectrum, they contact-shifted to protons to iron ion. 30 ferricytochromes, the resonances expected found to pseudocon tact the be shifts 31 ' and By upfield are Senn, (1983). H., Eugster, A., analogy the to this support their of -6.4 the these horse and tuna conserved His26 are current or pH 5-8 the to ring to titrate in K. His39 of due WUthrich, and proximity (56.8-7.8) R., are region of highly not and pK a 's the expected Pecht, 1., Shelton, J., Shelton, Nat!. Acad. Sci. 81, 6554 (1984). 30 due of His33 titrate The Hisl8 are of DSS), resonances (vide infra). as vs. Schroeder, W., Biochem. Biophys. the Gray, Acta H. Proc. 743, 58 -69- Figure 3.5. NMR spectra of (a) native and (b) AsRu-modified C. krusei ferricytochrome c, o6.5-9.5ppm (vs. DSS). Conditions: pH 5.2, 25oc. Note that the sharp native resonance at 88.62 is absent in the spectrum of the modified protein. -70- (a) native J .--. - , - . - - T 'J ~ ~~~ 'J ------r·-,...._.., - II '} IH!I (b) modified 8 tt lll ~ ~~- 8 b il T • b a;.t - ....... 1 - -r----r•·--.- · T - . - 1 _,.... --r _.,...._. r8 / Ill 1:1 r'it 1 8 1! J I ! r till I II 1 1!9 PO . • 1 r-- I - .. • I 6 !J ill I• t -Il -71range. NMR spectra were recorded for at six pH values. This pH data (table 3.1), and possible cytochrome c relate high and low estimation of pKa values. Data plotted the two histidine resonances the for the native number is insufficient to cor- but does for the titration curves are 3.6 each in figure C. krusei allow (for a second can be drawn through the remaining data points). 9.1 region, only these two resonances were seen to shift ferri- curve, In the 66.7and could be cleanly followed throughout the pH 4.2-. 8 range. Data analysis does low pH data. terms of fit to lines of one. The not allow While both curves correlation with results in more any the correlation of the high and and 2 give good fits to the data in coefficients slopes for (table closely case are 3.2), curves approaching similar: 3 the pKa and 4 are theoretical values for best value the two histidines are approximately 6.55 and 6.33. The values observed reported by histidine resonances protein under upfield of possibility of inappropriate chemical Cohen at identical 68.48 shifts, and and 68.62 conditions,33 chemical shift a systematic error, as the with those workers reported C-2 to TMS for same which are both observed. This suggests Hayes. 32 the peak however, standard. which in These relative values values are might There is, conflict arise from however, no the 0.14ppm using the an resonance evident at 60.14ppm vs. DSS which might be mistaken for the standard in 31 Dobson, C., Moore, G ., Williams, R. FEBS Lett. 51, 60 (1975). 32 Cohen, J., Hayes, M. J. Bioi. Chern. 249, 5472 (1974). 33 Chemical shifts are as read from figure 6 of reference 32. -72- Table 3.1. Chemical shifts 8 of titrating histidine C-2 protons 12H upfield downfield 4.2 8.6430 8.75680 4.9 8.6110 8.74850 6.1 8.4043 8.47490 6.7 8.1340 8.33216 7. 1 7.8977 8.28370 7.9 7.6570 8.19930 a. parts per million downfield of the methyl resonance of 2,2'dimethyl-2-silapentane-5-sulfonate Table 3.2. Parameters for least-squares fit of NMR titration data8 Upfieldb Down field m pKa fl m pKa r2 1.212 6.337 0.995 0.8844 6.542 0.993 2 1.236 6.266 0.986 0.8859 6.562 0.996 3 0.9600 6.574 0.969 0.9425 6.335 0.973 4 0.9858 6.537 0.979 0 .9555 6.371 0.987 Curve a For curves 1 and 2, 8+=8.643 and 0°=8.18 for the upfield resonance, and o+=8.76 and 0°=7.63 for the downfield resonance. The fits to curves 3 and 4 were evaluated using 8+=8.643 and 0°=7.630 for the upfield resonance, and 8+=8.76 and 0°=8.18 for the downfield resonance. b Labels describe resonances according to their relative positions at pH 5.2. - 73- 8.8,------------------------------------------------------. • 8.6 • ...-... (f) (f) 0 vi :::.. 8.4 E a. a. .__. 8.2 ....... £ en ~ u E 8.0 (l) £ u 7.8 7 . 64-------~-----r------~----~------~-----,,-----~----~ 4 5 6 7 8 pH Figure 3.6. Titrating resonances of native C. krusei ferricytochrome c. Four possible titration curves are shown (to delineate what points were included in each fit); complementary curves can be drawn through the remaining points. -74the cytochrome spectrum (TMS has a chemical shift identical to DSS; the former is aqueous solution). The workers internal standard, and incorrectly to refer them methylene in soluble the in organic to DSS. protons samples while may this to adjusted the the actually Resonances of (adjacent m solvent, have the residual sulfonate work served as might consider latter is soluble used a chemical water (54.8) moiety) of in different shift values and of the DSS (54.67) 34 in identifying reference points that chemical the DSS methyl resonance. Alternatively, is identical is in to one that disagreement, reported but in it is the one literature difficult to (58.62), imagine how shift value the other and such a situation would result. The spectrum sharp is indicating signal abolished that the at in 1=5/2) 35 is adjacent histidine that of the in the native cytochrome NMR modified cytochrome (figure for resonance is responsible that a 3.5b), site of The paramagnetic and quadrupolar nature of the ( 99 Ru: ruthenium found residue ruthenium modification. bound 58.62 expected 12.7% abundant, to broaden and residues. The second with 10 1Ru: 1=3/2; split the proton abundant, resonances of proton resonance appears the ruthenium reagent. A in the spectrum of the modified likely to be a contact-shifted histidine et al. for pentaammineruthenium(III) 34 Pouchert, C., ed. Aldrich Library of Chemical Co. (Milwaukee), 1983, p. 1003. NMR Spectra, 35 Becker, E. High Resolution Press (New York), 1980, p. 285. Magnetic to be unaffected by the reaction evident at -5-32.5ppm resonance is protein (figure resonance, as 3.7). reported It is by Toi Nuclear C-2 17.1% vol. Resonance, 2, new Aldrich Academic -7 5- Figure 3.7. NMR spectra of (a) native and (b) AsRu-modified C. krusei ferricytochrome c, 8-22 to -38 ppm (vs. DSS). Conditions: pH 5.0, 25oc. A new resonance is evident at 8--32.5 in the spectrum of the modified protein. -76- (a) native -22 -23 -24 -25 -26 -27 -28 -29 -30 PPM -31 -32 -33 -34 -35 -36 -37 -38 (b) modified - 22 -23 -24 -25 -26 -27 -28 -29 -30 PPM -31 - 32 -33 -34 -35 -36 -37 -38 -77- derivatives of sperm whale myoglobin. 3 6 Robinson et al. the other to His39 reduced state. 37 onances were positions have in No the very chemical reported. between assigned the similar shift Assuming oxidation upfield the His33 and cytochrome in the for ferricytochrome res- the resonances and states, to cerevisiae S. values resonance maintain assuming their relative disagreement the between chemical shift values from the work of Cohen and Hayes and this work arises would from interpret indication of taken be one of histidine systematic error in the spectrum of the the (vide former modified at His33. Alternatively, the evidence for His39 modification the is assignment, extensively it is clear modified by supra), cytochrome modification as Regardless a from the the A 5 Ru one as an spectrum could (see figure 3.8). NMR spectrum that moiety, while the other is not. UV-visible Difference Spectroscopy The UV -visible krusei ferricytochromes 528.5, 409, 360, absorption spectra of native and c and are exceedingly 280 om. similar, UV-visible A 5 Ru-modified with maxima difference at C. 696, spectroscopy reveals a chromophore with a wavelength of maximal absorbance at 300nm (figure 3.9), consistent with [A 5 RuHis] 3 + model that [ruthenium]:[cytochrome] a 1:1 complex. The the intensity ratio 36 Toi, H ., LaMar, G ., Margalit, Chern. Soc. 106, 6213 (1984). R., 37 Robinson, M, Boswell, Biochemical J . 213, 687 ( 1983). Huang, A., properties spectral of is Che, this absorption indicated. C.-M., Z ., the is such Differences Gray, Eley, of c ., H. J. Moore, in Amer. G. -78- I His39 His33 I I I (b) I 8.76 (a) (c) 8.62 8.48 chemical shift (ppm) Figure 3.8. Clarification of discrepancy in histidine C-2 proton resonance chemical shifts: (a) as read from figure 6 of reference 32, with assignments given by Robinson eta/. for S . cerevisiae ferrocytochrome above; (b) this work, as observed for native C. krusei ferricytochrome c at pH 5.2; (c) this work, as observed for A 5Ru-C. krusei ferricytochrome at pH 5.2. An assumption could be made that a systematic error is the cause of the discrepancy between (a) and (b), in which case one would assign the modification site to the residue responsible for the upfield resonance, His33. Alternatively, one could assume 88.62 corresponds to the His39 proton, with no explanation for the discrepancy of the second resonances. One would then conclude His39 had been modified. -79- (a) (b) 0 .04- 0.03Q) (,) c C\1 ..0 ..... 0 (/) ..0 C\1 0.02- ~ 0 .01- 250 350 450 wavelength (nm) Figure 3.9. (a) UV-visible difference spectrum of AsRu-modified C. krusei cytochrome c minus native cytochrome c (-l8J.!M); (b) UV-visible spectrum of -l8J.!M [AsRuHis]Cl3. -80absorption between the difference spectrum and a spectrum of the model complex states can in be attributed native the spectrum in more the of figure to and 3.9(a) reduced differences form the than the the proportions of redox samples, e.g., the contained 3.5% hence higher cytochrome modified indicates in modified native absorption is observed below 280nm and sample sample, and in the -410-440nm region (horse cytochrome isosbestic points occur at 339, 410, and 434nm). 38 Peptide Mapping Because of peptide mapping ments, to which The the uncertainty was in the undertaken histidine residue is su bmaxillaris protease (SMP) NMR to modified histidine C2- 1H assign- obtain definitive evidence the derivative of interest. very specific, in is reported to be as cleaving peptide bonds involving the carboxyl group of arginine res- idues. 39 Five peptide products are expected (see figure 3.10), and His33 and His39 should reside on different peptides due to the of a cleavage numerous 3.11) site peptides suggested in at residue the native incomplete Cleavage at sites other native peptides. correspond ing to unexpected (Arg) 38. cytochrome excessive or than arginine Tyr46-Arg54 cleavage after 32-hour and Gly45 the SMP cleavage was confirmed Peptides of -4 and -7 residues However, presence presence digest had of (figure occurred. by analysis of two minutes retention were found to His39-Tyr48 (table 3.3), reveal- and respectively. Such Tyr48, 38 Margoliash & Frohwirt, p. 570. 39 Schenkein, 1., Levy, Biophys. 182, 64 (1977). M., Franklin, E., Frangione, B. Arch. Biochem. -81- 1 10 Pro-Ala-Pro-Phe-Giu-Gin-Gly-Ser-Ala-Lys-Lys-Giy-Ala-Thr-Leu-PhePl ------------------------- 20 Lys-T hr-A rg-C ys-Ala-G In-C ys- His-Th r-Ile-G Iu- Ala-G ly-G ly-Pro-His--------~~~--------------------------------------P2~ 30 40 Lys-Val-Gly-Pro-Asn-Leu-His-Gly-Ile-Phe-Ser-Arg-His-Ser-Gly-Gln- 50 Ala-G lu-G ly-Tyr-Ser-Tyr-Thr-Asp-Aia-Asn-Lys-Arg-Ala-G ly-Val-G lu------------P3--------------------------~~------------- 60 70 Trp-Ala-G lu-Pro-Thr- Met-Ser-Asp-Tyr-Leu-G lu-Asn-Pro-TMK-Lys- 80 Tyr-Ile-Pro- G ly-Th r-Lys-Met-Ala -Phe-G ly-G ly- Leu-Lys-Lys-Ala- LysP4 ----------------------------------------------------.- 90 100 Asp-Arg-Asn-Asp-Leu-Val-Thr-Tyr-Met-Leu-Glu-Ala-Ser-Lys -------+-----------------------P5------------------~ Figure 3.10. Peptides expected upon digestion of C. krusei cytochrome c by submaxillaris protease. The points of cleavage (after Arg residues) are indicated by vertical lines. -82- 0 '" "' nJ .,,.., "' "· "' 0) 0 0 •::> Figure 3.11. HPLC chromatograms of SMP peptides with 220nm detection. Elution profiles of -40~g SMP digest of native (upper) and ruthenium-modified (lower) C . krusei cytochrome c, as monitored at 220nm (x axis is time, in minutes). -83- Table 3.3. Total amino acid analyses of native C. krusei cytochrome c SMP peptides peptide la peptide 2b Amino acid Found Exwctedc Arg 1.07 1 Ser 1.03 1 Asp 1.89 2 Glu Thr 1.14 His/Lys Tyr e e Expectedd 1.40 2 1.97 2 2.18 2 1 Gly Ala Found 1.07 1 0.84 1 + -/1 + 1/- + 2 + 2 a. retention time of -4 minutes b. retention time of -7 minutes c. on the basis of the peptide being Tyr46-Arg54 (fyr-Ser-Tyr-Thr-Asp-Ala-AsnLys-Arg) d. on the basis of the peptide being His39-Tyr48 (His-Ser-Gly-Gln-Ala-Glu-GlyTyr-Ser-Tyr) e. + indicates that the presence of these amino acids was confirmed by HPLC resolution of their FMOC derivatives, but their amounts could not be quantified (see appendix 1) -84adventitious cleavage may have resulted from contamination of the SMP with another protease. The most dramatic difference between the chromatograms of the SMP digests of the the native significant and reduction ruthenium-modified in a tion, and the appearance of a native krusei C. peptide with new peak of 3-4 cytochromes c is -7 minutes reten- minutes retention in the digest of the modified protein (figure 3.11). Chromatograms (figure fied 3.12) of indicate cytochrome absent in the (figure 3.13) the that peptide there that native cytochrome the with several components are digest of mixtures absorb at this digest. ruthenium-modified detection in at 300nm the modi- wavelength and The elution 400nm cytochrome which digest are profile indicates that the 300nm-absorbing components common to both digests are probably heme fragments (native cytochrome c is retained at least 90 minutes). The peptide of -3-4 minutes retention was observed to have a 300nm absorption, which shifted to -365nm upon the addition of base to the sample (figure 3.14). These spectral features are identical to for the model [A 5 RuHis] 3 + complex; upon deprotona tion of coordinated imidazole moiety the ligand-to-metal (pKa=8.9), 40 those reported the charge transfer tion, upon due to acid shifts 40 41 the alkalinization presence to -295nm Sundberg, R., 96, 381 (1974). Johnson, (I 984). bands41 (LMCT) C., of the Bryan, R., Henderson, shifted absorbance tyrosine; upon are the at W., I., of energy. increased, absorption the Taube, Shepherd, lower 300nm -275nm deprotonation Taylor, to R. phenolic H. J. lnorg. of In addi- presumably this side amino chain. 42 Amer. Chern. Soc. Chern. 23, 2754 -85- cr> ('U "'..,. 0 "' ., 0 -"' 0 Figure 3. 12. HPLC chromatograms of SMP peptides with 300nm detection. Elution profiles of -401-!g SMP digest of native (upper) and ruthenium-modified (lower) C. krusei cytochrome c, as monitored at 300nm (x axis is time, in minutes). -86- -' .. 0 - ' .,; "' -' .. ... " ' .,; "" 0 '" .. ' ' ,.; '-" "' ",;, 0 0 - Figure 3.13. HPLC chromatogram of SMP peptides with 400nm detection. Elution profile of -40j.lg SMP digest of ruthenium-modified C. krusei cytochrome c , as monitored at 400nm (x axis is time, in minutes). These fragments also absorb at 300nm (figure 3.12) due to their heme content. -87- 0.035 ,,,··\ t'l' t I . I r,, f \ ll I ,~ I t '.!'' ""\' 1 I \ \ \ \ ' \ \ \ \ \ \ \ ' \ \ \ \ \ \ ', 0 200 400 300 wavelength (nm) Figure 3.14. UV-visible spectrum of the ruthenium-containing peptide recovered from the SMP digest of the modified protein: being made basic by the addition of TEA. as eluted from the HPLC column; --- after -88The composition of analysis 3.4), uniquely defines only one portion of (table corresponds to this new peptide, as determined the site by total amino acid of the C. krusei modification, as it cytochrome c sequence, i.e., His39 through Tyr46. The additional peaks eluting near the void volume had intense UV absorption below 250nm, but lacked the pentaammineruthenium histidine complex. was observed upon The spectrum (in alkalinization, but 0.1% is TFA) pentaammineruthenium(III) difficult to understand is not a nium(II) cal in the strong (E0 =380 methionine A general red no maximum bands shift of emerged at similar to that reported complex (in 4N HC1), 43 of the 10-25nm -365nm. for the but it is how this species might have formed, as CoEDT A- sufficiently studies provide residue methionine characteristic vs. no evidence intact oxidant oxidize pentaammineruthe- 500mV, respectively). 44 Electrochemi- support modification a to (vide protein to infra). The at methionine early-eluting com- ponents may have been decomposition products of some sort (see appendix 2). Upon attempting analysis, retention ible, suggesting to repurify volumes and the peaks were the components for were found to be irreproduc- extent artifacts related to retention remaining in the modi- (resulting from peak areas to some amino acid the high salt concentration of the samples. The fied components cytochrome decomposition 42 of of digest the seven-minute may be unmodified ruthenium complex) or Cantor, C., Schimmel, P. Biophysical Freeman (San Francisco), 1980, p. 377. His39-Tyr46 another peptide Chemistry, 43 Stein, C., Taube, H . lnorg. Chem. 18, 1168 (1979). 44 Hin-Fat, L., Higginson, W. J. Chem. Soc. A 298 (1967). Part (analysis of 2, w. H. -89- Table 3.4. Amino acid analysis of the ruthenium-containing peptide recovered from the SMP digest of A5 Ru-modified C. krusei cytochrome c (3.9 minutes retention) Found Expecteda Tyrb + 1 Gly 1.88 2 Glu 1.49 2 Ala 1.14 1 Ser 0.91 1 Hisb + 1 a. on the basis of the sequence of the peptide being that of residues 39 through 46: His-Ser-G ly-Gln-Ala-Glu -Gly-Tyr b. + indicates that the presence of these amino acids was confirmed by UV spectroscopy of the peptide and by HPLC resolution of their FMOC derivatives, but their amounts could not be quantified (see appendix 1). -90the seven-minute two peptides, note that Tyr48) of the only His39 and fraction one was of the native which found modified ruthenium of on alters suggested was identified). different peptides (His39-Tyr46) label digest digests, the site it is It in contained interesting to the native (His39- that the presence suggesting of cleavage, though reduction potential serves not the two pur- protease specificity. Spectroelectrochemistry Determination poses: it not dynamics, only but in the the allows also environment. 45 fold of is for the characterization of the a sensitive indicator of changes any major disruption of the protein is Hence, modified heme expected to result redox in native thermo- the heme polypeptide in an alteration the reduction of the cytochrome reduction potential. Knowledge tials of the of the temperature donor and acceptor dependence are of critical in poten- understanding the temperature dependence of electron transfer rates. Equilibration of the cytochrome with the electrode was usually Four isosbestic 600-500nm 525, attained points region; 540-541, and within were their five observed positions 554-556 nm fell minutes of for the in the (figure 3.15). setting was rapid, and the potential. native cytochrome in ranges of 523.5- lsosbestic 503-505, points of the the modified protein were observed in the same ranges as three of the four 45 (a) Varadarajan, R., Zewert, T., Boxer, S., Gray, H. Science 243, 69 (1989); (b) Santucci, R., Reinhard, H ., Brunori, M. J. Amer. Chem. Soc. 110, 8536 (1988); (c) Walker, F., Emrick, D., Rivera, J., Hanquet, B., Buttlaire, D. J. Amer. Chem. Soc., 110, 6234 (1988); (d) Pielak, G., Mauk, A., Smith, M. Nature 313, 152 (1985). -91- 0.04 0.03 G) (.) c::: tU -e 0.02 0 en .c tU 0.01 0 500 520 540 560 580 600 wavelength (nm) Figure 3.15. Family of visible spectra recorded during a spectroelectrochemistry experiment on native C. krusei cytochrome c, 24.40C. Spectra were recorded at applied potentials of -230.9, -25.5, -6. 1, 14.0, 34.8, 55.0, 79.9, 109.8, and 230.7 mV (vs. SCE), which correspond to the curves of the most to the least absorbance at -550nm. -92native points; the fourth, 503nm. 46 Isosbestic points steps for native gross alteration presence of of the of the at points, results perature the was at was or had 54.8 °C occurring. observed throughout 55°C conformation vs. pendix 3). maintained maintained protein and energy, protein cipitously (226.6mV data), highest were samples isosbestic denaturation of The at the 502- potential less, indicating no occurred. Despite the indicate some potential dropped pre- 247.3mV (vs. NHE) predicted by fit the Nernst relation was poor (see has been reported to This temperature is similar to that which the thermal lower ternap- to result in the partial denaturation of horse ferricytochrome c. 47 At chrome elevated were temperatures, lost, suggesting isosbestic an points unfolding of of the modified the protein. The cytoexact temperature varied from 36.6°C upward, depending on the history of the sample. The temperature recovery suggests the isosbestic points upon reducing the denaturation process is reversible one. that the obtain data for the modified protein utilized for the native cytochrome The inability to as high those as of a at temperatures suggests the cytochrome is somewhat destabilized by the modification. in The data for the appendix 3. The total fell to 80% of The minimum attributed protein 46 to spectroelectrochemical experiments change in absorbance its original value by the absorbance denaturation precipitate onto observed, and the ta hula ted typically for a sample completion of the however, precipitation gold are of the the Both protein: may be should the transmittance of the The significance of this difference is questionable, spectrophotometer chart drive was not working properly at times. as the 47 Myer, Y. Biochemistry 1, 765 (1968). minigrid, increased. experiment. -93electrochemical compartment might be reduced, thus giving rise to an apparent increase in absorption for the fully oxidized protein. A is typical excellent, Nernst r2 with cytochrome. plot is values shown of in figure 0.9994 or 3.16. Data correlation greater for the native The slopes of most of the Nernst plots agree well with the theoretical values ( <4% error in most cases). The tabulated reduction potentials determined in 3.5 and plotted parameters are presented table thermodynamic the native (E 0 '=276mV protein ( vs. kcal/mol, 6S 0 '=-37.8eu) are Margalit and (E 0 '=264mV, Schejter in in the range in figure 3.17. in table 3.6. 5-60°C The The agreement with those derived for 6H 0 '=-17 .6 reported 6H 0 '=-17.8 6G 0 '=..,6.0kcal/mol, are values 6G0 '=-6.36kcal/mol, NHE), good of by kcal/ mol, 6S 0 '=-39eu, IJ=0.01M). 4 8 The parallel nature native and modified 6S 0 ' rc for tion processes changes. is each. The indicative of has to been of by the have trivalent been a ti ves. 49 reduction This increase of 13mV can reduction is more favorable presence of It ruthenium. observed can be redox-linked in the for increase additional Increases other concluded for the many protein reduc- temperature modest of for the the to attributed displacement plots is an indication of the similarity of found slight due temperature-dependence The large negative 6S0 electrostatics: cytochrome the cytochromes c modified cytochrome. basis of of no potential in for the on the modified charge provided reduction potential A 5 Ru-protein deriv- in alteration 48 Margalit, R ., Schejter, A. Eur. J . Biochem. 32, 492 (1973). plots the positive 6-12m V major dependence be rationalized singly-modified that conformational of the heme -94- -w 0 (f) CJ) > 100 > E w ~ -~ c Q.) ...... 0 a.. -o Q.) 0 a.. a.. ~ - 1 2 0 log [0]/[R] Figure 3.16. Nernst plot for spectroelectrochernical data on native C. krusei cytochrome c, 24.40C. The data is best fit to the equation y = 59.3x + 32.15, r2 = 0.9999. -95- Table 3.5. Reduction potentials of the iron site in native and A5Ru-modified C. krusei cytochromes c as determined by variable temperature spectroelectrochemical studies (a) native temperature ('C) E 0 ' (mV vs. NHE) 9.1 15.2 19.8 24.4 29.6 34.2 39.8 44.8 50.2 54.8 290.1 285.9 281.8 275.8 271.4 269. 1 262.3 256.7 250.4 226.5 temperature C'C) E 0 ' (mV vs. NHE) 6.8 a 7.6 12.6 19.6 26.4 31.0 36.6 36.6 a 40.6 a 47.0 a 304.7 307.2 300.5 294.2 287.5 282.6 277.2 273.7 271.0 258.6 (b) modified a. sample 2 of same preparative batch -96- 0 .31 ,-----~--------------------------------------------------, 0.30 1:1 native • modified .--.. w I z (/) 0.29 > > ~ ·.;:: 0.28 c (]) 0 a.. c - .Q 0.27 ( .) ::I "0 (]) '- 0.26 0.25;---~----~---r----r---~---,----~--~----~---.--~~--~ 270 280 290 300 310 320 330 temperature (K) Figure 3.17. Temperature dependence of the reduction potentials of the iron site in native and AsRu-modified C. krusei cytochromes c, as determined by spectroelectrochemistry. The data are best fit by the equations y = -0.964 x 10-4x + 0.5634 for the native cytochrome, andy= -1.04 x 10-3x + 0.5967 for the modified. -97- Table 3.6. Thermodynamic parameters for the reduction of the iron site in native and A5Ru-modified C. krusei cytochromes c, as determined from variable temperature spectroelectrochemical studies native A5Ru-modified 276.0 ± 0.4 287.9 ± 0.4 ~G ' (kcal/mol, 25°C) -6.36 ± 0.01 -6.60 ± 0.01 ~Srco, (eu) -22.2 ± 0.2 -23.9 ±0.2 ~S ' (eu) -37.8 ± 0.2 -39.5 ±0.2 Ml0 ' (kcal/mol, 25°C) -17.6 ±0.1 -18.4±0.1 E 0 ' (25°C, mV vs. NHE) 0 0 -98environment occurs in this derivative as a result of appending the pentaammineruthenium moiety to the C. krusei cytochrome molecule. Differential Pulse Polarography Typical traces recorded iments on the native and figure 3.18. potential A of single -290mV -120mV dine (vs. residue, is NHE), (vs. A pulse suggests the reduction exper- C. krusei cytochrome c are shown observed and for the corresponds that the site potential is of native to the similar NHE) region demonstrates protein at in a oxidation / reduc- modified derivative modification ported for the model compound [A 5 RuHis] 2 +/ 3 +. 50 in the 500mV (vs. polarography new signal evident in the NHE) as differential modified peak tion of the heme iron. at in to is the a histi- 106mV re- The absence of signals the lack of modification at methionine. Reduction and modified 3.19. Using 119mV (vs. potentials protein the are best-fit NHE) is determined reduction 3.8), is not necessarily studies in table 3.7 parameters for the ruthenium obtained potentials DPP presented Deviation of this value from from from that found an for the reduction reported for for other indication and of the native plotted in figure site, value potential at of 25°C. the model compound, and A 5 Ru-modified of a error, but proteins may (table reflect the 49 (a) Crutchley, R., Ellis, W., Gray, H. J. Amer. Chern. Soc. 107, 5002 (1985); (b) Margalit, R ., Kostic, N., Che, C.-M., Blair, D., Chiang, H.-J., Pecht, I., Shelton, J ., Shelton, R., Schroeder, W., Gray, H. Proc. Nat/. Acad. Sci. 81, 6554 (I 984); (c) Nocera, D ., Winkler, J ., Yocom, K., Bordignon, E., Gray, H. J. Arner. Chern Soc. 106, 5145 (1984). 50 Yocom, K ., Winkler, Scripta 21, 29 ( 1983). J ., Nocera, D., Bordignon, E., Gray, H. Chern. -99- Figure 3.18. Differential pulse polarography traces (current vs. potential) for anodic scans of (a) native and (b) AsRu-modified C. krusei cytochromes c. Conditions: 0.25mM cytochrome in pH 7 j..t=O.lM NaPi, -0.3mM in Aldrithiol-4; (a) 26.3°C, (b) 2o.ooc. - 100- (a) native (b) modified -0.2 -0.1 0 0.1 potential (V vs. SCE) 0.2 -101- Table 3.7. Differential pulse polarography data Native C. krusei cytochrome c 8 T {°C} Eo' {mV vs. NHE} 6.6 10. 1 15.1 21.2 26.2 31.0 302.1 ± 1.4 297.1 ± 1.4 295.0 ± 1.4 290.4 ± 1.4 283.5 ±2.2 282.9 ±2.0 Modified (A5RuHis39) C. krusei cytochrome cb T {°C} Eo' {Ru 2 mV vs. NHE} Eo' {Fe2 mV vs. NHE} 6.4 10.2c 14.2 20.4c 24.0 35.6 40.4 96.5 ± 1.9 112.3 ± 1.5 103.9 ± 1.4 117.7 ±1.4 119.0 ± 1.4 129.4 ±2.1 132.8 ± 3.6 300.5 ± 1.9 310.6 ± 1.5 308.6 ± 1.4 303.7 ± 1.4 299.3 ± 1.4 291.7 ± 2.1 289.4 ±3.6 a. Linear regression of the data results in a fit to the equation y = -0.799x + 306.5 with correlation coefficient fl-=0.97. b. Linear regression of the data for the ruthenium site results in a fit to the equation y = 0.963x + 95.1 with correlation coefficient fl-=0.89. The data for the iron site is best fit to the line y=0.528x + 312.0 with correlation coefficient r 2=0.71. c. sample 2 of same preparative batch -102- u:J0.13 I z (f) > ~ 0.12 ....... c (1) ....... g_ 0.11 c 0 ....... (.) :J "0 ~ 0.10 0.09~----~----~----r----,r----,----~----~-----r----~--~ 270 280 290 300 310 320 temperature (K) Figure 3.19. Temperature dependence of the ruthenium reduction potential in AsRumodified C. krusei cytochrome c, as determined by differential pulse polarography. The data are best fit by the equation y=9.63e-4x- 0.168. -0.78 ±0.95 -0.7 ± 0.4 4.2 ± 1.2 -1.98 ±0.05 85.8 ±2.0 A5Ru-myoglobin -3.0 ±0.8 -3.4 ±2.0 -1.96 ± 0.12 80.0 ±5 [A5RuHis] 2+!3+ a. Data for myoglobin from Crutchley, R., Ellis, W., Gray, H. J. Amer. Chern. Soc. 107, 5002 (1985). Data for the model compound were recorded in 0.1M NaC104 using cyclic voltammetry(CV); Nocera, D., Winkler, J., Yocom, K., Bordignon, E., Gray, H. J. Arner. Chern. Soc. 106, 5145 (1984). Ml0 ' (kcal/mol, 25°C) !:lS ' (eu) 6.6 ± 1.6 -2.75 ±0.48 !:lG0 ' (kcal/mol, 25°C) 0 119.2 ±21 E0 ' (mV vs. NHE) A5Ru-C.k. cytochrome Table 3.8. Thennodynamic parameters for the reduction of the pentaammineruthenium histidine site (as determined by DPP or CV) in C. krusei cytochrome c and other environments8 I w 0 I ....., -104influence of the local protein environment in stabilizing one oxidation state over the other. The DPP data is used only to derive the thermodynamic parameters for the ruthenium trochemistry. electrode tron in The as whereas between difficulties cident junctions, ments is displaced meet (see figure the reference in the technique is inferior results promoters mediators the allow to for fully reversible elec- electrode. Additional error and the the design of the electrochemical constructing a glass the sample interface of from the point where 3.2). The slight overlap compartment gives rise to spectroelec- quasi-reversible in protein DPP results from technical the of use kinetics, exchange the site, cell the with and cell: multiple reference independent of the fluctuations due to coin- compart- water jackets sample jacket onto in the temperature can potentially of the latter, to which the reference electrode is slow to respond. DISCUSSION krusei C. react with cytochrome the c has numerous sites aquopentaammineruthenium(II) which reagent. Attempts to modi- fy tuna cytochrome c were unsuccessful (IOOmM NaAc, pH5, 0.2mM cytochrome, 30:1 [ruthenium]:[cytochrome], hours) as determined routine modification by of gel for reaction electrophoresis carboxylate times (data not moieties up to 10 shown). Thus, (of Glu, and amine Asp, present in C. kruse i cytochrome the covalent binding of therefore contain thioether which are and Lys side chains) can be ruled out. The only cysteine residues are those involved Cysl7, and these (figure 3.20a) Cysl4 and sulfhydryl moieties. Other amino in acids c the heme, rather than potential sites of -105- Figure 3.20. Residues which are potential ligands to pentaammineruthenium in C. krusei cytochrome c (in addition to the His33 and His39 targets). Dotted regions are matched in color to the line drawing of the residues whose side chain surface they represent (3.5A probe; 8 dots/A2). (a) Cys14 and Cys17 (green) are the sites of covalent linkage of the porphyrin to the polypeptide. Only the latter (forward) is exposed. Histidine residues (pink) are present at positions 18 and 26. The former serves as a ligand to the iron, and is buried within the protein. (b) Methionine residues in C. krusei cytochrome c. Three methionine (yellow) residues are present, Met64, Met80 (the sixth ligand), and Met98, none of which are expected to be solvent-accessible. -106- Figure 3.20 (a) Figure 3 .20 (b) -107modification b, are the histidine and methionine respectively). c. krusei has three and four histidine Met98, residues (figure methionine residues, residues, Hisl8, 3.20a & Met64, Met80, and His26, His33, and His39. His 18 and Met80 are ligands to the iron and are not expected to react as long as the protein conformation is maintained. The was accessibility evaluated using of these Biograf. residues The in reduced results are tuna cytochrome presented in table c 3.9. Aside from His33 and His39, none of the residues except Cysl7 and His26 are of concern, as they are internal and are inaccessible to solvent. Cysl7 modification can be ruled out, as it was not observed to occur m tuna c. cytochrome extensive residue His26, hydrogen-bonding be modified, a though solvent-accessible, network substantial (figure change is 3.21), in the involved and in an should this polypeptide conforma- would the tion is likely to result. Thus, potential one might sites studies of histidine that idues of the predict histidine mod if ica tion. derivative complex has been formed. one of two normally-titrating is modified. The by UV -visible difference of 1:1 modified in the all Indeed, modified the residues number suggest derivative. pentaammineruthenium support the histidine (His33 or sites spectroscopy, a of which only spectroscopic the data NMR of of be modification reveals Peptide mapping His33 is a conclusion His39) is narrowed [Ru]:[Fe] assigns res- the ratio site of modification definitively to His39. The lack heart cytochrome react with the of modification reaction at kinetics: after consistent allowing aquopentaammineruthenium(II) the reagent with the horse protein (0.2mM horse to cytochrome, lOmM [A Ru(OH 2 )] 2 +, IOOmM HEPES buffer, pH 7.0) for twenty four hours, 5 -108- Table 3.9. Surface areas of tuna cytochrome residues which are potential pentaamrnineruthenium ligands in C. krusei cytochrome c. a PROBE RADIUS 0 residue t.4A. 3.5A Cys14 Cys17 Leu64b Met80 Leu98b His18 His26 0 10.72 0 0 0 0 37.28 0 5.82 0 0 0 0 28.36 a. as defined by Connolly, M. Science 221,709 (1983); given in A 2; based on the 1.5A crystal structure of reduced tuna cytochrome c: Takano, T., Dickerson, R. J . Mol. Bioi. 153, 79 (1981) and Takano, T., Kallai, 0 ., Swanson, R., Dickerson, R. J. Bioi. Chern. 248, 5234 (1973). Calculations with the 1.4A probe radius define the solvent(water)-accessible surface whereas those of radius 3.5A reveal area accessible to the pentaammineruthenium reagent. Residues 33 and 39 are omitted, as these are known to be at the protein surface, and are not histidine in the tuna cytochrome. See text for further details. b. Met in C. krusei cytochrome c -109- Figure 3.21. His26 hydrogen-bonding network. His26 is hydrogen-bonded through both imidazole nitrogen atoms, and modification of this residue is expected to disrupt the polypeptide conformation. The residue 44 (purple) carbonyl serves as a hydrogen bond acceptor to the Ne of His26 (yellow). The NB of His26 is hydrogen-bonded to the amine of residue 31 (green). The adjacent carbonyl (of the residues 30-31 peptide bond) is hydrogen-bonded to the NB of the iron ligand, His18 (red). -110a 30% of yield this of His33-modified reaction mobility higher by than product electrophoresis that of the obtained. 51 is indica ted native that protein) was A time analysis no product observed (of until the reaction had progressed for four hours {data not shown). orate Very recent the results studies and cerevisiae S. with interpretation cytochrome presented mutants Attempts here. corrob- to modi- fy His58 in a mutant cytochrome containing His33 but not His39 at pH 7 (30:1 [Ru]:[cytochrome], pH 7, lOOmM HEPES) result formation after 45 minutes, as judged by FPLC. 52 seem to lead fore, the residue to the formation which of in product Longer reaction multiply-modified in little modifies rapidly the the lack of modification reaction of His33 Modeling of products. times There- Candida cytochrome at or is likely to be His39. The reason for appropriately, the to hindrance. be steric slower ducted to investigate the environment 3.22a). No evidence was found pulsion might be the cause histidine is surrounded each positive vealed and negative, more constrained imidazole ring of protein, side 51 and chain is His33 accessible extends freely surrounding suggest for the discrepancy by two or a 7A radius. environment is only via the into solution, con- electrostatic re- the reaction rates: charged residues, both to between (figure of His33 re- that of His39: the two portions In contrast, approach by Nocera, D., Winkler, J ., Yocom, K., Bordignon, E., Chern. Soc. 106, 5145 (I 984). 5 2 Chang, J., personal communication was local Inspection edge. appears residue in three and His39, more each that relative "sandwiched" to krusei cytochrome to within a C. the relative His33, the of the the His39 ruthenium Gray, H . J. Amer. -111- Figure 3.22. Local environment of histidines of interest in (a) C. krusei and (b) horse heart cytochrome c. Negatively (Asp, Glu) and positively (Lys, Arg) charged side chains are shown in red and blue, respectively. Histidine side chains are shown in yellow. Dotted regions indicate the surface accessible to a 3.5A probe for each side chain (8 dots/A2), and are colored appropriately. (a) C. krusei cytochrome c. Note that there is no preponderance of one type of charge surrounding either histidine. Lys103 provides additional negative charge, as it is the Cterminal residue. Note the greater surface area of the His39 residue (71.7 A2) as opposed to His33 (43.4A2). (b) Horse heart cytochrome c. There are no indications of excessive positive charge surrounding His33 which would slow the modification reaction. Note that the imidazole side chain is accessible only at the edge (33.7 A2 total surface area). -112- Figure 3.22 (a) Figure 3.22 (b) -113- is complex ring. possible Calculated from the surface edge areas and for either of the imidazole and His39 are 43.4 and horse and yeast His33 may reaction of His33 54 is His33 face 71.7A 2, respectively (3.5A probe). 53 The also be relative rates of explained observed C. in c cytochrome by modification steric of hindrance. krusei cytochrome shows no evidence of (under similar conditions). solvent in horse before four hours is less exposed to protein (surface areas of c Some within the 33.7 one modification hour, (by while horse electrophoresis) Modeling cytochrome 43.4A 2 , and half suggests than in respectively, His33 the yeast were calcu- contrasted with lated (see figure 3.22b)). The previous c. results and at of histidine Kostic both His33 isolated by inorganic and two found and modification horse can modification C. His39 singly-modified Horinishi et al. found fied this investigations Brothers readily of be in yeast cytochrome c krusei cytochrome be labeled by terpyridylplatinum(II) reagent, equal yield. 55 a derivatives in to and yeast (Bakers') cytochromes were diazonium-1-H-tetrazole reacting most readily, followed in three steps, with by His26, and then His33 Hisl8. 56 and modiHis39 The second 53 For comparison: an unencumbered histidine side chain has an area of 82.9A 2 accessible to a 3.5A probe. 54 Reaction at His33 is inferred from the further reaction of the major C. krusei cytochrome c A 5 Ru-derivatization product with time (as observed by electrophoresis), and the recovery of a minor product after one half hour which seems to be modified at two histidine residues (see appendix 6). 55 56 , Brothers, H., Kostic, N. lnorg. Chern. 27, 1761 (1988). H OrtnlS . . h"1, H ., 111, 520 (1965). Kurihara, K., Shibata, K. Arch. Biochem. Biophy s. -114and third steps supporting the were accompanied assumption that by the changes His26 by these reagents than the visible hydrogen-bonding integral to maintaining the cytochrome fold. modified in spectrum, network is His33 may be more readily by aquopentaammineruthenium(II) because of reduced steric hindrance. Preliminary characterization thenium-modified product of suggests the it second is a pentaammineru- Candida doubly-modified derivative, and is discussed in appendix 6. The results reported C-2 1H NMR chrome c. the upfield of the residue demonstrated cation by resonance modified to be (1>8.62) product, the site at pH 5, can be assigned of which mapping. There is evidence to of yeast cytochrome reduction potential the the ionization of His39. 5 7 is to absent in His39, the pentaammineruthenium peptide dependence the unambiguous assignment of the resonances for histidines 33 and 39 in C. krusei ferricyto- The spectrum here allow for suggest that is modifithe pH related to This may be readily confirmed by the use of A 5 Ru(His39)-cytochrome derivative: the heme reduction potential is expected to be pH-independent in the pH 5 to 8 range, as the modified histidine would not undergo any ionization. tion Taken together, the electrochemical of iron site is favored preparation of a Ru 2+-Fe 3 + intermediate iron i.e., Hence, tron the transfer to the site, over studies that of indicate the that reduc- ruthenium center. should Ru3+_Fe 2 + result in production, should proceed with a driving force of -170mV at room temperature. thermodynamic parameters 5 7 Robinson et a/., p. 687. for the overall reaction (at 25°C) are elecwhich The 0 f1G '= -115- -3.85(± 0.49)kcalj mol, the temperature become narrows. less is exothermic l1H0'=-17 .6(± 1.1 )kcaljmol, l1S 0 '=-46.1 (± 1.8)eu. As increased, transfer will as the the gap electron between the two reaction reduction potentials -116- CHAPTER 4 ELECTRON TRANSFER STUDIES -117MA TERIALS and INSTRUMENTATION [R u(bpy) 3 ]Cl 2 ·6H 2 0 , hydrate, was used as supplied by Strem Chemicals Inc. reagent grade. in an inert Degassing All manipulations atmosphere was chloride hexa- Na 2 EDT A·2H 2 0 was (2,2'-bipyrid yl)ru thenium(II) of degassed solutions box (Vacuum/ Atmospheres Co.) accomplished on the dual-manifold were filled conducted with argon. vacuum-argon line in chapter The described in chapter 2. Prote in Native cytochrome c modified cytochrome studies, and was was further was purified as described used as from purified eluted CM-52 by FPLC prior to the 2. for quenching flash photolysi s studies. [ Ru( bpy) 3 ] 2 +'* lifetime studies Cells needle for valves to quenching experiments allow evacuation, for were and test with tubes side fitted arms with terminating in 1 mm x 1Omm x 30mm quartz cells. The narrow side-arm cell was used for permitted lifetime measurements, to be made. and also absorbance Lifetimes were measured at ambient measurements temperature; no special provisions were made to maintain the sample temperature constant. The excitation source was a Quanta Ray DCR-1 quency-doubled with a Quanta Ray HG-1 Quanta Ray PHS-1 8os (fwhm) collected at ( f / 1.3), then Nd:Y AG laser fre- harmonic generator, followed by a prism harmonic separator to produce a 532nm pulse of duration at a angle to by second right focused 10Hz. a Emitted the light, monitored laser beam with a lens (f / 6.5) through at 650nm, was collimating lens a 3-67 Corning -118filter (transmittance<0.5% at i\<524nm) onto entrance the slits of a MacPherson monochromator. The light was detected by a Hamamatsu R928 photomultiplier the tube, and resultant signal was passed through LeCroy VVlOlATM amplifier to the 50fi impedance input of a 6500 waveform firing of the a Biomation recorder. A Digital PDPll/03-L computer controlled the laser the data acquisition. The acquired for and data each sample were averages of at least 200 laser pulses. Analysis of the data was performed TOOBIGl, written by Michael Albin. using the BASIC program This program allows one to fit data to a mono- or bi-phasic exponential decay. Flash photolysis The flash photolysis apparatus (figure 4.1) consisted of a Xenon Corporation N851D flash lamp fired by a Xenon Corporation model 457A micropulser through (36-lOOJ/ flash; Corning transmittance 3-73 at T 112 -51JS). filters i\>511 (37% nm) to The flash transmittance prevent at electronic light was filtered 416<i\<436nm; >80% excitation the of protein by ultraviolet light. The sample in length, lamp. A which was contained in an fit aluminum within constant sample an evacuated temperature could cylindrical cell, -15cm containing flash cavity be maintained the by circulating thermosta tted water through the cell jacket. The tungsten probe Oriel 6324 housing, powered by a PowerMate and cooled with a gentle flow of filtered the data throughout filtered through a lamp (Osram HLX64 collection Corning 15 process. filter (zero 625) was contained in a power supply, Corporation air The which was monitoring transmittance at maintained beam was i\<523nm to -119- .------· • power• :supply: •------· ,- • - - - • - I : power: • supply ,...,- - - - - - . , ·------FLASH LAMP ftlters _n MONOCHROMATOR er! :I ~ PMT fl A L shutter ftlt D ~ ,{). hou se air u SAMPLE CELL i" ---- -, , power. 1 , supp1Y•I optical fibers L----- -' I TRIGGER BOX J : VOLTME1ER 1 Voltage AMPLIFIER &OFFSET J -, i----, power, , supp1Y•I 1 L----- -' I ·J COMPUTER I OFFSET VOLTMETER Figure 4 .1. Schematic diagram of the flash photolysis instrument. I PROBE LAMP -120minimize guided continuous to the cell excitation of by fibers optical [Ru(bpy) 3 ]2 + the (Oriel). The the far end of the cell was carried by optical grating monochromator (i\.=549 nm). The light was detected (PMT), contained in a The slit complex), light fibers to width was by a Hamamatsu R928 Pacific Photometric and then emerging from the Oriel 7240 typically 2.3mm. photomultiplier tube Instruments model SOB housing, and powered by a Pacific Precision Instruments model 204 power supply. The output current of the PMT was fed monitored voltage 2865 by a offset digital Keithley circuit. panel. 177 microvoltmeter, through a 5KQ resistor, and The offset voltage was Release 2.25, the Kbyte 256 then fed displayed version into a DC on a Simpson of UnkelScope (Unkel Software, Lexington, MA), running on a ATronics ANI 8T computer with a Princeton lamp and data Graphic System collection. For screen, details was used to trigger about the shortcomings the flash of this H 2 0) by arrangement, see appendix 4. METHODS [Ru(bpyhJ 2 +* lifetime measurements Absorption charge tion transfer are of visible light (400-580nm; i\max=452nm results in an electronically excited metal-to-ligand pathways of (MLCT) available: the ground state with a its ground state by state. 1 Two compound may competitive emit light as it in returns to the return to constant of rate constant of k .2.• or the complex may nonradiative relaxation with a de-excita- rate The photophysical properties of ruthenium(II) polypyridyl complexes have recently been reviewed: Juris, A., Balzani, V ., Barigelletti, F., Campagna, S., Belser, P ., von Zelewsky, A. Coord. Chem. Rev. 84, 85 (1988). -121knr· The rate expression for the de-excitation process is given by -d[Ru*]/dt = k .Q.[Ru*] + knr[Ru*] where Ru* represents concentration of the excited the excited the initial excited state, state state of the will ruthenium complex. decrease according concentration of the is by the The to the excited state. The inverse of sum following expression: where [Ru*] 0 lifetime of of de-excitation the cules in react such molecular is the with cases the the process T 0 , given the Certain excited complex and excited state proceeds with is a said rate return it to be constant to the mole- ground state; "quenched". If this bi- of rate of de- kq, the excitation is given by -d[Ru*]/dt = k .Q.[Ru*] + knr[Ru*] + kq[Q][Ru*] and [Ru*] = [Ru*] 0 {exp -(k .2.+knr+kq[Q])t} if [Q]>>[Ru*] 0 , lifetime in where the [Q] is the presence of a quencher of concentration [Q], A of the ratio of the lifetimes of the of quencher, /T, versus (k.2.+knr+kq[Q]f 1 . state in quencher the plot absence and concentration presence concentration should give a of the T quencher. 0 line with an intercept of The T, is excited the 1 and a slope of kq T 0 : 'to 't = =1+ This is known as the Stern-Volmer relation. 2 2 Barltrop, J., Coyle, York), 1978, p. 110. J. Principles of Photochemistry, J. Wiley (New -122All quenching experiments were done with -6S1-1M [Ru(bpy) 3)C\ 2 in IJ=0.1M pH 7 sodium phosphate buffer. termined from ence successively of of the a highest series Quenching rate constants were de- of emission lifetime lower quencher concentrations. (practical) concentrations measurements in the Quencher attainable were pres- solutions maintained in the test tube side of the cell and were degassed and purged with argon for at least three cycles, then sealed. After the lifetime measurement was made, the cell was transferred to an inert atmosphere box, where it was diluted solution. with This a known procedure volume allowed of deoxygenated [Ru(bpy) 3 ] 2 +-buffer a more estimation for accurate of the quencher concentration by avoiding the reduction in sample volume which would have accompanied evacuation with each dilution. For tryptophan quenching studies, 51.4mg acid standard kit 22) was added to ionic sodium pH 7 to allow strength phosphate from (Pierce lOmL 63.21-lM [Ru(bpy) 3 ]Cl 2 overnight, protected tryptophan concentration was estimated absorbance at using an 280nm, light, tryptophan buffer. to the The mixture solid to be -33.5mM extinction amino in O. lM was left dissolve. The from the solution of 5600M-l cm- 1 coefficient (recorded after the lifetime measurement). 3 Flash photolysis Flash vated 3 photolysis react.ions. 4 The is a sample technique is for exposed the study of fast a short (llS or ns), intense Part 2, w. to Cantor, C., Schimmel, P. in Biophysical Freeman (San Francisco), 1980, p. 443. Chemistry , 4 Norr.ish, R ., Thrush, B. Q. Re v. Chern. Soc. 10, 149 (1956). photoacti- H. -123of light which librium situation. A of flash, flash the initiates probe allows for the reaction beam, set at a monitoring of by right the creating a non-equi- angle the direction progress to of the reaction by changes in absorbance. ~J=O.lM A -98mM cytochrome solution in degassed and saturated with argon on a pH 7 NaPi was thoroughly dual manifold Ar-vacline. second solution of 6S1-1M [Ru(bpy) 3 ]Cl 2 and 5mM Na 2 EDTA buffer was similarly inert atmosphere pared by pipetting cally lOmL (2 x cytochrome prepared. Both along with a aliquots of box 5mL) of solution were were transferred dry cell. Samples flash solution into the the [Ru(bpy) 3 ]2 + solution to the solutions each used, in yield a final flash and A same into an were pre- cell; typi- <2001-!L of the cytochrome concentra- concentration dependence tion of -1.05~JM. This of the a variation procedure electron degassing studies have mentation tions above transfer in from volume each not (see was adopted rate so could that be the investigated without introducing which would (and therefore concentration) sample. However, rigorous been conducted appendix 4). due Those to concentration-dependence problems with reported arise at the flash cytochrome instru- concentra- l1-1M were conducted on individually degassed, complete reac- tion mixtures. It is preferable to a void reduction of the cytochrome degas by any all light the reactants ambient light separately present to during the sample preparation. foil) Samples were prior placing to protected them from in the probe (by wrapping beam. A delay the cells of in several minutes was required to allow the PMT to stabilize prior to flashing. Data reported here for native cytochrome were obtained at room -124temperature; no special temperature constant. provisions Modified were made to maintain cytochrome samples were 1024 64 of the sample maintained at which defined the by Unkel- 22°C by a circulating water bath. Data were collected as baseline voltage prior to scope binary form, was in BINLOT, also values of values using the flash) flash. The data file, recorded converted to ASCII form using the program product of Unkel Software. The intensity (voltage) resulting file were converted into with the baseline (i.e., that to zero. Data were plotted and x=l,2,3... A a the the the points, program absorbance BLT2DAT, being set equal the parameters estimates were made of parameter file was created, MEXPFIT DP. The fit was made to an equation of the form and used A(x), as absorbance relative where input to before the program absorbance values y = A(l) + A(2)[1 - exp(-A(3)t)] Parameters were for output the to best fit separate found files. and the Programs predicted which Unkel Software can be found in appendix 5. are not products of A flow chart for the data analysis procedure can be found in figure 4.2. Data for the kinetics of the once using all data above 2ms, and 7ms. The latter that which region is modified again cytochrome were fit twice, using collected beyond should data be relatively unaffected by the problem of flash breakthrough (see appendix 4). RESULTS and DISCUSSION The transfer cytochrome result rate c) presented here for constant of the derivative which the subject is the of this intramolecular electron (A 5 R u(His39)-C. krusei thesis is preliminary. -125- . Unkelscope .. fl as h expenment OOl.t.nt ! BINLOT (binary to ASCII conversion) OOl.blt BLT2DAT (intensity to relative absorbance) OOl.par + (estimated parameters) 1 OOl.dat MEXPFIT DP (fitting routine) OOl.fit (fitted parameters) OOl.the (theoretical absorbance values using best fit) Figure 4.2. Flow chart for kinetic data collection and fitting. -126Several problems covered by the author, and remedy the necessary with the linear (although would be flash photolysis instrument time to be made. caveat: if one and politics Therefore, dis- permit assume the PMT response such that the definitely nonlinear for large changes to be a lower in all presented can was likely not is operation is did data instrument presented were the the rate design several changes were made in an attempt to However, changes following the the situation. the then with light was response intensity), limit. A discussion of sequence of reactions ex- the instrumental problems is presented in appendix 4. Scheme (figure 4.3) to occur following flashing of the sample at duration, serves to excite pected microseconds in inducing formation state. the The longest tion is 685ns, 5 able to participate in absorbance. lengthened in of after directly Implicit the a reported hence rate (five any presence of cytochrome A -2.4~-ts the of infra). of is trisbipyridylruthenium(II) as the constant the 4.4 -450nm. The flash, [Ru(bpy) 3 ] 2 + complex, charge transfer (MLCT) [Ru(bpy) ] 2 +* 3 in aqueous half it should not lives), solube reaction causing an observed change assumption that the lifetime is not protein. In fact, the lifetime of state is reduced an oxidative quencher (vide has determined for the excited latter of the metal-to-ligand lifetime in the c, delineates serves x as 10 8M-I s- 1 been in the presence this quenching reaction (figure 4.4). 6 5 Harriman, A. J.C.S. Chern. Comm. 777 (1977). 6 Ihis ii somewhat faster than the quenching rate constant of 2.5 x 10 M-l sreported for horse heart cytochrome c; Nocera D., Winkler, J ., Yocom, K., Bordignon, E., Gray, H. J. Am. Chern. Soc. 106, 5145 (1984). -127- hv Scheme 1 Figure 4.3. Sequence of reactions initiated by a microsecond pulse of visible (-450nm) light in the electron transfer experiment (the stoichiometry is not necessarily such that the ruthenium and iron sites of the modified protein are reduced in equal proportions). -128 - 4 .e0 3 ~ 2 0 15 Figure 4.4. Stern-Volmer plot for the quenching of the [Ru(bpy)3]2+* by C. krusei cytochrome c. The solid line is the least squares fit to the data: y = 0.263x + 1.09 (r2 = 0.996). -129- heme When ruthenium is present on the cytochrome, it, as well as the iron, may as serve an molecules reduced at molecular electron transfer, the site form iron tetraacetic to oxidizing the ruthenium as (EDT A) present are expected electron electron donor, reducing the electron transfer with reduced oxidation by in is [Ru(bpy) 3 ] 3+ has been one like to monitor verify that migrate to as a can sacrificial undergo back mechanism for the EDT A and in scheme A proposed, to intra- Ethylenediamine- serves it Protein undergo expected before protein. to product. solution [Ru(bpy) 3] 3 + [Ru(bpy) 3 ] 2 +*. to thermodynamic acid the label the the center is outlined 2 (figure 4.5). Ideally, would the acceptor reduction, the observed electron activity of the to transfer ruthenium both no is donor oxidation and other groups contribute to Monitoring process. center the precluded in required wavelength of the probe beam (280-450nm) of the redox this system, as the would result in con- tinuous excitation of the photosensitive [Ru(bpy) 3 ] 2+ complex. The quenching investigate the the [Ru(bpy) 3 ]2 +* of possibility tryptophan was studied to of acid radicals in of participation Lifetime measurements in presence electron transfer process. trisbipyridyl complex of Ru(II) are state observed, suggesting that the reduction was low allow for tryptophan excited state result is with reports 8 7 Butler, (1986). consistent J., too in table the to 4.1. No amino tryptophan was presented by of made and quenching 0.87(± O.l)V, 7 with the value of 0.64V for the tryptophan radical Land, E., Prtitz, W., Swallow, J .C.S. A. of for the absence of the potential excited of oxidation. the This and inconsistent (Trp·) reduction Chern. Comm. 348 -130- RR'N-CH 2 -COOH "+ RR'NH + HCOH Scheme 2 Figure 4.5. Postulated mechanism of action of Na2EDTA in the flash photolysis experiment (D. Miller, G . McLendon, Inorg. Chern. 20,950 (1981)). EDTA2(R=CH2COOH, R'=CH 2CH2N(CH2COO-h) acts to reduce [Ru(bpy))] 3+ (A=acceptor) and prevent its reduction by the protein under study. The radical intermediate is believed to be responsible for the slow reduction of the native cytochrome, i.e., the second equivalent of acceptor may be ferricytochrome rather than [Ru(bpy))] 3+. -131Table 4.1. [Ru(bpy)3J2+* lifetimes in the presence and absence of tryptophan sample [Ru(bpy) 3 ]2 + [Ru(bpy) ]2 + 3 [Ru(bpy) 3]2 + + Trp a. average of two values, each 591.5 573.7 581.5 determined by fits to averages of at might be least 200 pulses. potential,9 assuming a value of 0.84V 10 for [Ru(bpy) 3]2 +* reduction. 11 Since the reduction expected to vary studies should be Trp· radical reduction. residue potential depending conducted upon of a on its particular test for the to [Ru(bpy) 3 ]2 + tryptophan excitation C. krusei cytochrome contains (number 59). 12 The absorbance c can protein transient and residue environment, formation of the subsequent cytochrome a tryptophan single be monitored at 526.5nm, Chern. 90, a 8 Jovanovic, (1986). 9 Once oxidized, tryptophan may lose a proton, depending on the solution pH. The pKa has been reported to be 4.3 by Posener, M., Adams, G ., Ward man, P., Cundall, R. J.C.S. Faraday Soc. I 72, 2231 (1976). S., Harriman, A., Simic, M. J. Phys. 1935 10 Creutz, C., Sutin, N . Inorg. Chem. 15, 496 (1976). 11 Butler, (1987). J., Land, E., Swallow, A., Prlitz, 12 Narita, K., Titani, K. J. Biochem. 63, 226 (1968). W. J. Phys. Chem. 91, 3114 -132krusei C. cytochrome c isosbestic point, near the 485-510nm absorbance maximum of Trp·. 13 In an described intramolecular above, where the E 0 (Fe)>E0 (Ru)), formation of transfer rate constant molecular redox for systems, other formation and electron transfer experiment such as that fully oxidized protein is utilized (i.e., not interfere with the provided no Trp· would determination reaction occurred where the per fully Fe 2 + -+Trp· subsequent more cytochrome reduced electron than molecule. protein transfer electron one bi- However, is utilized, might be Trp· observed in addition to Fe 2 +-+ Ru 3 + electron transfer. Reduction possibility, alanine as of the amino amino potentials of acid/radical preliminary experiments -0.84V been has acid residues the anion tryptophan, couples to be more reported for the was not tyrosine, have negative been and phenyl- estimated -1 V. 14 than a considered A [Ru(bpy) ]3+ /[Ru(bpy) 3 ]2 +* 3 from value of reduction potential. 15 Upon flashing [Ru(bpy) 3 ]2 +, This a confirms chrome is the rapid native rise C. krusei in cytochrome in absorbance that the mechanism indeed an electron of is observed [Ru(bpy) ] 2 +* 3 transfer the process. presence (figure quenching Ideally, the of 4.6). by cytoexcited ruthenium complex would transfer an electron to the heme of native cytochrome c, and no further change in absorbance would be observed. 13 (a) Redpath, J., Santus, R., Ovadia, J ., Grossweiner, L. Int. Radial. Bioi. 27, 201 (1975); (b) Santus, R., Grossweiner, Photochem. Photobiol. 15, 101 ( 1972). 14 Butler, J ., Land, 705, 150 ( 1982). E., Pri.itz, W., Swallow, 15 Lin, C., Sutin, N. J. Phys. Chern. 80, 97 (1976). A. Biochim. Biophys. J. L. Acta -133- Sxl0- 3 .---.---~----r---~----.---~----.---~~---r----~--~----~ 4x10- 3 Q) 0 c 0 .D \.... 0 (I) .D 0 Q) > ·;, 0 Q) \.... 2x 10- 3 0~--~--~----~--~----J---~----~----~--~----~--~----~ 0 20 40 60 80 100 time (seconds) Figure 4.6. Flash photolysis data for native C. krusei cytochrome c (A.=549nm). The initial jump in absorbance at the time of the flash (Os) is attributed to reduction of the heme iron by the excited ruthenium(II) tris(bipyridyl) complex. The slower reduction is attributed to EDTA radicals which are formed by reaction with the oxidized ruthenium(III) tris(bipyridyl) complex. The data can be approximated by a monoexponential function (k=0.05/s, an average of three experiments). -134However, the data reveal a slow process which can be fitted to a monoexponential tions function identical to studied. The by EDT A the with those process a at constant which the modified be the is believed radicals -O.OSs- 1 rate to which result of protein reduction from for of the concentra- kinetics were the heme iron of the change is after the time reduction [Ru(bpy) ]3+. 3 On observed a similar upon initial increase scale reveal time scale flash photolysis (figure 4.7). an absorbance the Data collected increase and much shorter leveling after approximately direct reduction iron site tion of electrons by oxidation of [Ru(bpy) 3 ]2 +*) to rate constant of -170s- 1 . concentration from This the ruthenium(II) The l-6~M the an intramolecular that this is present, further jump expected from 4.8). over absorbance on range process the is attributed centers changes are to (created iron(III), and supports of the migra- upon flashing proceeds independence ([cytochrome]) process. protein a the (figure initial absorbance modified addition rate the of little 16ms in served to (IOOs), of with the a ob- the assumption Although EDT A radicals may be minimal because the protein is almost completely reduced. l6 If the intramolecular electron transfer data is fit beyond 7ms, so as to eliminate that which might be affected by the problem of flash 16 The similar values of the absorbance change for the native and modified samples suggests that a similar number of bimolecular electron transfer events take place. Hence, the EDT A radical concentration is expected to be similar in each case. Direct EDT A radical reduction of the iron site, or rate-determining radical reduction of the ruthenium site, followed by fast intramolecular electron transfer, may be responsible for the slow rise observed in the 0-Ss range following flash photolysis of the modified derivative. -13 5- Cl) 0 6x 1o- 3 c 0 ..0 .... 0 Ul ..0 0 Cl) > ~ 0 Q) 4x 1o- 3 .... 0~--~--~-----L----~--~----~----~--~~---L----~--~~--~ 0 20 40 60 80 100 time (seconds) Figure 4 .7. Flash photolysis of modified C . krusei cytochrome c (A..=549nm). Reduction is rapid, and little of the observed signal is due to EDTA radical reduction. -136- 6x10- 3 r-,---~----~----~---r----~--~----~---,----~----r----r----r-• 4x1o- 3 Q) 0 c: 0 .D .... 0 1/) .D 0 Q) :g> Q) .... 2x1 o-3 o~~--~---L~~----L-~~--~~~--~~~~--~--~--~~ 0 2x1o- 3 4x1o- 3 6x1o- 3 8x1o- 3 0.01 0 .012 time (seconds) Figure 4.8. Data for flash photolysis of the modified C. krusei cytochrome c. The average of monoexponential fits to two runs for the 2-12ms range is 170/s. -137breakthrough, value ably is a constants an average rate of is obtained much more uncertain, lower limit to actual electron transfer for the intramolecular electron transfer the but 295s- 1 as stated modified protein are tabulated in table 4.2. (figure previously, 4.9). 170s- 1 This is prob- rate constant. Rate in the ruthenium- -138- C) 0 c: 0 .D 0 II) .D .... 0 Q) > :;; 0 ~ 2x10- 3 o~~--~----~~-L----~~~----~~~----~~~----~--~--~~ 0 2x10- 3 4x10- 3 6x10-J 8x10- 3 0 .01 0 .012 time (seconds) Figure 4.9. Data for flash photolysis of the modified C. krusei cytochrome c. The average of rnonoexponential fits to two runs for the 7-12rns range is 295/s. -139- Table 4.2. Monoexponential fits to the kinetic data of modified cytochrome c a A. t>2ms parameter A(1) 0.0096 ± 0.0002 0.0013 ± 0.0001 A(2) 0.0041 ± 0.0001 0.0049 ± 0.0001 194 ± 20 145 ± 16 parameter Run 1 Rlin..2 A(1) -0.0027 ± 0.009 -0.0022 ± 0.006 A(2) 0.0075 ± 0.009 0.0078 ± 0.006 k (s-1) 320±210 270 ± 139 B. t>7ms a. Data were fit to an equation of the form y = A(1) + A(2)[1-exp(-A(3)t)], where A(3) is the rate constant k. -140- CHAPTER 5 DISCUSSION AND CONCLUSION -141The rate constant (His39)-C. krusei cytochrome This the is transfer the fastest in a those of In c has been measured this determined result, for to it is electron found date protein horse Pseudomonas and cally rate intramolecular ruthenium-modified significance with for to for (see table A 5 Ru- least 170s- 1 . state electron To evaluate at 1.1). to compare this rate c (ket= A 5 Ru(His47)cytochrome stutzeri three cases, the equivalent, and so are expected to have The driving force is energies. in A 5 Ru(His33)cytochrome all reorganization be ground appropriate heart transfer donor and acceptor sites are chemi- identical inner sphere equivalent (within ex- perimental error) in each case (-170mV). These tance, derivatives medium, and farthest removed removed 11.7 the greater significantly in three orientation. His39 krusei cytochrome 13A the at and 7.9 distance result in a tronic coupling differ from A, in of the and f1G* are identical using a value of donor for 0.9A -l 3 Osvath, (1988). P., Salmon, and G., while His33 Contrary to derivative transfer rate acceptor. and what the His47 3 are is observed, would be due decreased to expected Assuming the values and cytochromes, krusei {3, an electron Yocom, K ., Bordignon, E ., Gray, H. Sykes, A. J. Amer. Soc. horse transfer rate Chern. dis- c is C. both for Nocera, D., Winkler, J ., Chern. Soc. 106, 5145 (1984). 2 Candida electron C. heme, respectively. the diminished of respects: to elecof v and constant of J. Amer. 110, 7114 His47 of cytochrome c 551 is homologous to residue 67 in horse and yeast cytochrome c· (a) de Silva, D., Sykes, A. Biochirn. Biophys. Acta 952, 334 (1988); (b) Almassy, R., Dickerson, R. Proc. Nat!. A cad. Sci. 75, 2674 (1978). . -142-lOs- 1 is horse cytochrome Candida derivative). Using for ket of 170s- 1 estimated for this observed is the predicted significantly stra tes horse heart observed if for i.e., Distance one a distance of krusei derivative electron (as the opposed to a modified C. the is mechanism however, ket value coupling explain through-space the assumptions, derivative, cannot transfer, (from same donor-acceptor greater derivative. actual the Candida the derivative, assumes derivative demonthe than the rate difference operative. if electron medium in a the the krusei two) of and 0.64A-l be other the of may the of shorter in The the C. transfer occurs along a through-bond pathway. When pathway, the considering none the mechanism, nature three However, others. bond" of the derivatives is significantly different one considers the possibility of Candida derivative must if the presence of an aromatic moiety, Trp59, necting His39 and "through-space" the heme. The be singled a "through- out for in the vicinity of the edge of the from tryptophan the path conindole ring lies <2A off the direct line connecting the C.y of His39 and the closest porphyrin atom, and (relative the plane) toward heme cation of enhanced myoglobin derivative, lar electron other present ring the inclined heme (figure a -22A distance shorter the direct 4 Axup, A., Albin, M, Soc. 110, 435 (1988). path at a 70° angle There 5.1a). coupling [A 5 Ru(His12)-ZnMb], involving in is donor-acceptor transfer over derivatives residue is the in a is some to indi- zinc-substituted which undergoes intramolecu- at a rate similar to that of distances (l9-20A ). 4 A tryptophan between His12 and the heme. Gray, H. J. Mayo, S., Crutchley, R., To Amer. Chern. -143- Figure 5.1. Possible pathways of electron transfer in AsRu(His39)-C. krusei cytochrome c. (a) Trp59 (green) is located above the plane containing the porphyrin (red) and His39 (yellow), and <2A from the direct pathway between the Cy of His39 and the carbon atom of pyrrole ring IV to which the propionate is bound. (b) A potential through-bond pathway: the amine nitrogen (yellow) of residue 41 forms a hydrogen bond with the propionic side chain of pyrrole ring IV of the heme (red), a possible 13-bond route of electron transfer. -144- Figure 5.1 (a) Figure 5.1 (b) -145investigate the role tryptophan coupling via another residue substituted However, Trp59 is of it acid, the of in S . cerevisiae phenylalanine, position at evolutionarily iso-1 has possibly mechanism, 5 superexchange an in been found mutant cytochrome 59 would be residue, and by to donor-acceptor a invariant cytochrome enhanced even lead the to interest. of most instability with replacement similar amino and reduced transfer biological activity.6 If for one each attempts derivative, to delineate specific routes of electron the results do point to any Candida enhanced electron transfer in electron transfer pathway may bone the to propionic nitrogen acid of (figure 5.1 b). 7 A horse cytochrome along the the exist not cytochrome. Candida, in of residue 41, which pyrrole ring IV in thirteen bond electron polypeptide is tuna via backbone from for A thirteen bond the peptide back- hydrogen-bonded and pathway has also transfer: 8 reason rice been cytochromes proposed electron transfer the to His33 to for could residue 30, the c the occur whose carbonyl is hydrogen-bonded to the His 18 ligand (see figure 3.21 ). A third donor-acceptor factor to orientation. consider His33 in and comparing His47 the heme, the former at an angle of -52° are these on derivatives opposite faces is of with respect to the porphyrin 5 Miller, J., Beitz, J . J. Chern. Phys. 74, 6746 (1981). 6 Schweingruber, (1978). M, Stewart, J ., Sherman, F. J. Mol. Bioi. 118, 481 7 (a) Ochi, H., Hata, Y., Tanaka, N., Kakudo, M, Sakurai, T ., Aihara, S., Morita, Y. J. Mol. Bioi. 166, 407 (1983); (b) Takano, T., Dickerson, R. J. Mol. Bioi. 153, 79 (1981). 8 Meade, T ., Gray, H., Winkler, J ., in press. -146plane. 9 His39 is situated exactly in the plane of the heme. With the programs have protein pairs. come of and computer several Models c cytochrome advent graphics proposals for been proposed have molecular interactions for peroxidase, 10 c cytochrome and of the modeling physiological complexes cytochrome of and b5 met- hemoglobin, 11 and cytochrome c and cytochrome b 5 , 12 and are based to a large extent on optimized electrostatic mentary charged surface residues. near coplanarity tive from the of the hemes, the a other, pair) and being this in to pairs of a protein feature common suggesting heme edge as opposed cytochrome cd 1 contains two (within A interactions that the perpendicular has been to coupling heme hemes c between face. and suggested to them all is the is effec- more Conversely, d 1 , with orientation compte- each with function the heme respect to store re- to ducing equivalents. 13 The discussion thus far has been based on the assumption that the free energy cytochromes. three of 9 of While derivatives, activation activation may ~G 0 the not is is similar equivalent reorganization be. As for electron (within energy, mentioned the and transfer uncertainty) both for hence the free previously, the similarity measured as the angle formed by the pyrrole IV nitrogen, atom, and the c'Y of residue 33 in reduced tuna cytochrome c 10 in (a) Poulos, T., Finzel, B. Pept. Prot. Rev. Poulos, T ., Kraut, J. J. Bioi. Chern. 255, 10322 (1980). all energy of the iron 4, 115 ( 1984); (b) H. Science 222, 929 11 Poulos, T., Mauk, A. J . Bioi. Chern. 258, 7369 (1983). 12 Salemme, F. J. Mol. Bioi. 102, 563 (1976). 13 Makinen, (1983). M, Schichman, S., Hill, S., Gray, -147the coordination spheres sphere reorganization outer sphere components: of the and acceptor indicate the inner to be similar in case. The can be considered reorganization of energy is reorganization there is the donor likely energy solvent each as two molecules separate around ruthenium donor and the reorganization of the medium around acceptor. et Pielak al. have the suggested that the reorganization not be heme of the for each solvent surrounding the ruthenium center may protein, and depending the local environment. 14 Little accessible, and the is con- solvent re- the heme stitutes may is about vary actually solvent 3% the total to be of on surface, In to expected ative of flavinmononucleotide semiquinone Meyer et a/. the (although not necessarily cytochrome c on proposed the small. order is c, that more basis of edge cytochrome organization rates identical the rationalize some reducing relaxation Evidence which outer in sphere response suggests to the a that in horse heart While the peptide reorganization, it change the energy reorganization cytochrome less models. 15 in the itself Even reactants is at 1M expected salt, to in be 14 Pielak, ( 1987). 15 Meyer, T ., Watkins, J., Biochemistry 23, 4761 (1984). 16 Churg, A ., 1683 (1983). G., Concar, Weiss, D ., R., which the may undergo heme oxidation state. 16 may be different electrostatic Candida cytochrome Moore, Williams, R. Przysiecki, Warshel, A., c., Tollin, Takano, T. Prot. for rate con- repulsion suppressed, G., recessed polar solvent, the two cytochromes comes from measurements of self-exchange stants. rei- is medium surrounding the heme serves to shield it from thereby of the heme than space-filling so reduction Candida exposed) that of of the exhibits a Eng. 1, 83 G., Cusanovich, M. J. Phys. 87, Chern. -148self -exchange rate heart cytochrome energy is plaining an c. 17 higher the order for of magnitude This would yeast cytochrome, fast rate the anomalously smaller suggest of than that and that the would horse reorganization not intramolecular of aid in electron ex- transfer observed. Studies of cytochrome c are not exist reduction cytochromes peroxidase have demonstrated apparent between reduced of from current c cytochrome at a rate cytochrome is reduced 12)s-1 ). 18 This electron transfer how subtle different horse species: within the protein rate one order of trend parallels the one observed between the at a differences, structural of heme site has been and yeast zinc-substituted thermodynamic and 17('! 3)s- 1 of by information, c cytochrome complex, while magnitude an which faster for is yeast (138(! intramolecular appended ruthenium in horse and Candida cytochromes c. The that foregoing electron discussion transfer action. It is tions exist, cies. In tional change, possible, each such is a acceptor, which crystallizes in the case, the necessary precedes slightly rate-determining however, having that different observed to a predicated two step (or in rate might the fast electron the the more) donor-acceptor enhance different on observed conforma- coupling efficien- that of coupling of the transfer. 19 in re- protein be conformations assumption a conformadonor an d Cytochrome the reduced c and 17 Gupta, R. Biochirn. Biophys. Acta 292, 291 (1973). 18 Ho, P ., Sutoris, C., Liang, N., Chern. Soc. 107, 1070 (1985). Margoliash, M., Hoffman, 19 Hoffman, B., Ratner, M J. Arner. Chern. Soc. 109, 6237 (1987). B. J. Arner. -149oxidized state, and course is therefore reduction. 20 of The assumed rate of to undergo some electron transfer motion in in A 5 Ru(His33)- horse heart cytochrome c has been demonstrated by Bechtold et al. independent of occur similar on a the methionine ligand scale. 21 time dissociation, However, which other the is to be known rate-determining to con- formational changes cannot be ruled out at present. Directions for future research An krusei intramolecular derivative would discussed above. (His39)-C. krusei electron help to Specifically, is due A 5 Ru(His33)cytochrome to a may transfer sort if out the the fast favorable undergo rate for effects rate pathway electron a A 5 Ru(His33)-C. of the observed or transfer variables in A 5 Ru- medium effect, at rate a the similar to that of the horse heart derivative. To date, only one pair of homo- logous at proteins have variabilis and S. fied His59 with at intramolecular been obliquus a a driving atives, force was identical plastocyanin. 22 Each pentaammineruthenium electron -0.03s- 1 (6E 0 =300mV). modified transfer (copper interpreted to as that in evidence protein moiety, reduction) The slow rate, at a comparable residues, that of been modi- has and each at a A. undergoes rate of moderate distance (-l2A) and azurin and supporting cytochrome the c deriv- existence of specific pathways which are more conducive to electron transfer. 20 Taka no, T., Dickerson, R . J . Mol. Bioi. 153, 95 (1981 ). 21 Bechtold, R., Gardineer, M, Kazmi, A., van Hemelryck, B., Isied, S. J . Phys. Chern. 90, 3800 (1986). 22 Jackman, M, McGinnis, J., Chern. Soc. 110, 5880 (1988). Powls, R ., Salmon, G., Sykes, A. J. Amer. -150- Several options are available to obtain a rate derivative might be for the His33 derivative: a (1) His33-modified taking advantage of the greater His39 with a redox-inactive synthesized reactivity of His39 moiety which would to by modify prevent reac- tion with [A 5Ru(OH 2 )] 2 +; (2) differences the two to control in residues, deprotonated the if ionization significant constants enough, the relative proportions forms of two the (pKa could of residues, values) be exploited protonated and of and therefore the relative concentrations of the reactive species; (3) a rate could be obtained for the doubly-modified could be employed deriv- ative; (4) site-directed protein lacking mutagenesis His39, making modification of to the create His33 a site straightforward. Each of these options will be discussed in turn. Option lacking. is (1) Currently, viable, the most but the specific ideal and modification widely used modification of histidine residues is diethylpyrocarbonate, DEPC. 23 Figure 5.2. Reaction of imidazole with diethylpyrocarbonate. 23 Miles, E., Meth. Enzym. 47, 431 (1977). reagent is reagent for -151- The suitability may react equally readily size (see discussion of it is of reversible this for this with both His33 and chapter 3). The modification doubt, as His39 due to small is attractive, react with tyrosine, the desired cysteine, specificity, and lysine drawback is that DEPC can react with produce product (i.e., reagent to tempts a stable to use tion of The usual this krusei C. result histidine content reagent. This each may and been the be or a verified that has been residues as well. as histidine before c [A 5 Ru(OH 2 )] 2 + at content despite to the reaction other changes at sites, in both of the Another atoms reac- unsuccessful. than the small excesses to to At- after nitrogen leading unreactive product of absorption and were greater use to irreversible). 24 quantify with found nitrogen is attributed the and reaction [A 5 RuHis] 3+ is that period both imidazole histidine reaction a the protein, concommitant studies suggest not indicated of imidazole, turation cytochrome over its it How- lacks decays in day s at pH 7, or the reversal can be induced by hydroxylamine). reagent compound is several the model application of ever, (the reagent known atoms protein properties. of of dena- UV-visible towards DEPC, but reaction of it has histidine with determinations of DEPC is unreactive toward [A 5 Ru(OH 2 )] 2 +. NMR studies prior constants and histidine residues, indicated that of respectively). 25 24 His39 in Hence, this C2- 1H ionization than to at assignments that the pH work, the reduced 5.3, His33 state two full including for the pKa of was the pH yeast significantly protein units cytochrome (6.7 below vs. the c less 7.3, His39 A doubly-modified hybrid might be used, i.e., His39 modified with DEPC and His33 with pentaammineruthenium, but it would be preferable to use a protein resembling the native cytochrome as closely as possible. -152pKa, 1% only of this Approximately 4% of form at this pH. residue would be expected His33, in contrast, would Therefore, the effective to exist be unionized. in the deprotonated concentration of His33 could be raised relative to that of His39 by decreasing the pH of the modi fication reaction mixture. The third option evidence indicates that cation is reaction a is the the second derivative appendi x 6). The observed expected to be stants most rate biexponential, corresponding and Preliminary feasible. the [A 5 Ru(OH 2 )] 2 + modifi- product of modified at of iron reduction in such a should be by and the to desirable two both His33 fitted and His39 (see protein the rate is con- intramolecular electron transfer available will produce processes. Finally, the exact whether a fourth alternative result desired, but all Richards Gray. electron mutants at which Caltech, transfer of S. and which allow for the investigation c derivatives will display kinetics. Work cerevisiae cytochrome is done being in c in progress in the lab collaboration histidines genotype could than 25 of with on Jack Harry A gene in which His39 has been replaced with a glutamine residue introduce ative. of similar is has been prepared, and has been used as the basis for to not will A 5 Ru(His33)-cytochrome intramolecular site-specific is The the locations of interest. 26 be prepared and modified to obtain a result horse at would be cytochrome Robinson, M, Boswell, J. 213, 687 ( 1983). a better basis for the as the two yeast Huang, Z ., Eley, C., rate, A., 26 Casimiro, D., Richards, J ., Gray, H., unpublished results further mutations Protein of A 5 Ru-His33 Candida deriv- comparison proteins Moore, this G. are 77% Biochem . -153identical in amino cytochrome and in common. acid krusei cytochrome c C. The author knows of 5.3),27 (figure composition have only no report of whereas horse their residues 50% of the isolation of the C. krusei cytochrome c gene to date. The dependence of the electron transfer rate on driving force been studied in atives. C. krusei reacts readily studies a is to should thermodynamic series of a superior give be A 4 RuL(His33)-horse a candidate heart cytochrome deriv- for such studies, as His39 modified product. Such spectroscopic and sizeable yield of accompanied by detailed characterization of the has protein, as the majority of the voluminous amount of work on cytochrome c to date has been done with that of between horse heart. the yeast Such and a study horse may indicate cytochrome whether are justified, the growing comparisons i.e., whether their reorganization energies are similar. Concluding Remarks Another number molecular electron unusually fast acceptor coupling nature of the rate has been added transfer rates in may very resulting intervening well from ruthenium-modified be their medium, to but attributed relative a firm to list of intra- proteins. The enhanced orientation donor- and / or interpretation of the this number must await further experimentation. One cannot draw any conclusions about the mechanism of electron transfer from a single intraprotein parisons with others suffer from a 27 on a residue-by-residue basis electron transfer rate, and com- problem which plagues man y electron -15410 1 Pro-Ala-Pro-Phe-Giu-Gln-Gly-Ser-Ala-Lys-Lys-Gly-Ala-Thr-Leu-PheThr-Glu-Lys -Ala- 20 Lys-Th r-Arg-Cys-Ala-G ln-Cys-His-Thr-lle-G lu-Ala-G ly-Giy-Pro-His-Leu-Val-Lys- 40 30 Lys- Val-Gly-Pro-Asn-Leu-His-Giy-Ile-Phe-Ser-Arg-His-Ser-Gly-Gln-Gly50 Ala-Glu-Giy-Tyr-Ser-Tyr-Thr-Asp-Ala-Asn-Lys-Arg-Ala-Gly-Val-Glu-lle -Lys- Lys- Asn-Leu- 60 70 Trp-Ala-G lu-Pro-Thr-Met-Ser-Asp-Tyr-Leu-Glu-Asn-Pro-TMK-Lys-Asp-Asn-Asn-Glu-Thr- 80 Ty r-Ile- Pro- G ly- Th r- Lys- Met- Ala- Phe- G ly- G ly- Leu-Lys- Lys-Ala-Lys-Glu- 90 100 Asp-Arg-Asn-Asp-Leu- Val-Thr-Tyr-Met-Leu-Glu-Ala-Ser-Lys -lie-Leu -Lys- Lys-Cys-Glu 8 Figure 5.3. Sequences of the cytochromes c from the yeasts Candida krusei and Saccharomyces cerevisiaeb (the residues of the latter are identical to those of the former unless specified). The residue numbers are assigned according to the horse sequence. Of the 108 residues the two proteins have in common, 83 are identical. a. Lederer, F. Eur. J . Biochem. 31, 144 (1972); Narita, K., Titani, K. J . Biochem. 63 , 226 (1968). b. Lederer, F. , Simon, A., Verdiere, J. Biochem. Biophys. Res. Comm., 47, 55 (1972); Narita, K., Titani, K. J . Biochem. 65, 259 (1969); Yaoi, Y. J . Biochem. 61 , 54 (1967). -155transfer studies: the inability to investigate one variable at a time. In comparing this number to that of horse cytochrome and c 551 , the nonequivalence of the donor-acceptor coupling /3), as well information is available, electron transfer (reflected in as of 'A must be borne in mind. The horse cytochrome cytochrome studies. Mutant versatility effects, in and s. terms about is of which limited cerevisiae the use iso-1 most for in-depth cytochromes offer of probing distance, medium, comparisons between rates electron transfer appropriate than ruthenium-labeled comparisons. c, c derivatives may be of more and the most orientation in their interspecies -156- Appendix 1 Amino Acid Analysis Method Development -157INTRODUCTION In for order to amino identify proteolytic analysis acid was fragments. quantitative developed. The reacts rapidly (FMOC-Cl) fl uoren ylmeth yl)-chlorof orma te a method reagent with (9- primary and secondary amines at alkaline pH to form stable derivatives: 1 + CH 2 -0-C-NH-CHR-COO . II 0 Such derivatization provides a facile means of detection: amino acids derivatized with FMOC-Cl (FMOC-AA's) absorb UV light of 200-300nm (i\max = 300. 289. 265, 227 and 218 nm, with the highest extinction of -20000- 30000 M- 1 cm- 1 at 265nm). 2 and have an emission maxima in the region of 310nm. 1a The derivatives are separated on reverse phase resins. de- tected by UV absorbance. and quantified. Precolumn derivatization with post-column derivatization. with available instrumentation. with FMOC-Cl was fact that the reagent form stable (a) Einarsson. 282, 609 ( 1983 ); (1979). do chosen as the The because of ion-exchange separation former method compatible method involving its over potential reacts with both primary and monosubstituted phthaldialdehyde. 2 chosen was not J .• (b) derivatives. react with was derivatization sensitivity and the secondary amines to Other reagents. such secondary amines, i.e.. as o- quantita- Josefsson, B.. Lagerkvist, s. J. Chromatography Moye. H.. Boning. A. Anal. Lett. 12(BI). 25 An extinction coefficient of 34000 M- 1 cm- 1 is reported in reference 3 for the FMOC derivatives in pentane; values obtained for solutions of both reagent and derivatives in -40% CH 3CN in this lab were -18500 M -1 em-1 . -158tion of dabsyl proline is chloride or Other methods, e.g., reaction phenyl isothiocyana te, result in nonfluorescent not possible. with derivatives, and so are less sensitive. 3 The method has been developed to the but not absolute, amounts tinguished firmed from of most amino phenylalanine. but not quantified. confirmed, but the two The The are acids. presence presence not point of providing relative, of Tryptophan is not tyrosine can be lysine can of histidine distinguished or from one discon- another, be nor quantified. MATERIALS and INSTRUMENTATION The chromatographic instrumentation is described in chapter 3. Derivatives were detected by their absorption at 265nm. n-Pentane was of HPLC grade, and was obtained from Fisher Scientific. Other solvents used are described in chapter 3. The amino acid derivatives were separated on Vydac 218TPS (4.6mm x 250mm) reverse phase, silica-based CIS columns. (9-fl uoren ylmethyl)-chlorof orma te and obtained from Fluka; the former Solutions of FMOC-CI in acetonitrile used several acids for glycine, methionine, Dynamics were The valine, alanine, leucine, tryptophan were and and FMOC stored weeks. proline, FMOC-L-lysine was Individual, at of argon. periodically and L-isomer amino the isoleucine, obtained were phenylalanine, from purchased underivatized were under 0°C prepared derivatives N-£ -FMOC-L-Iysine Corporation. 9-fluorenylmethanol Fluka. from amino 3 Ogden, G., Foldi, P. LCGC 5, 28 (1987), and references therein. N-cx- Chemical acids were -!59supplied in the Pierce Amino Acid 22. 1-adamantanamine were reagent grade. Borate buffer was Standard Kit was purchased from Aldrich. NaHC0 3 and Na 2 B4 0i IOH 2 0 prepared by titrating a solution of -50mM Na 2 B4 0i IOH 2 o to the desired pH with 2N HCI. Amino acid AA-S-18 (two valine, ,S-alanine, II, oxidized standards different lots) and norleucine, insulin chain from Sigma. employed in the calibration included Pierce Amino Acid cysteic acid, bradykinin, B and hen egg white Standard lysozyme Sigma H. Nor- angiotensin were obtained Sperm whale myoglobin, obtained from Sigma and purified by elution from CM-52 with 50mM pH 7.2 Tris buffer, was kindly provided by Dr. James Cowan. Materials for hydrolyses of test peptides are de- scribed in chapter 3. Filtration chapter 3. for solvents amino acid derivatization The were carried into new dispensed materials out lOmm using and P-1000) been recently a glass, x 75mm Gilson P-200, and in calibrated. 22G Rheodyne screw-capped borosilicate Pipetman or an and samples and described subsequent vials. Samples extractions were filtered All reagents were microliter pipets (models P- 20, Eppendorf model 4710 pipet, all of which New Becton blunt-tipped in tubes. test digital are Dickinson needle lee disposable were used to column had syringes inject calibration samples. METHODS Method Calibration development and actual #870205-10 at 35°C. was carried out on sample analysis was done on #500354 column at #60 25°C. of lot -160When quantitation was 15% (v / v) CH 3CN in H 2 0 important, the needle port was rinsed with to prevent sample carryover from one injection to the next. No data was utilized from runs which were the first of the day. Derivatization The following procedure was used initially during development of the method to provide samples containing most amino acid derivatives of interest, any agent. plus An possible aliquot (-2.51-lmol / mL of of side known products, volume each amino acid impurities, (-101JL) dissolved in and of amino O.lN HCl) excess acid re- standard was added to -2301JL 150mM NaHC0 3 (pH>7). An aliquot of -2501JL FMOC-Cl, -IOmM in CH 3CN, was added to give a reagent-to-amino ten minutes, the reaction volumes of n-pen rane. diluted with to concentration was was ratio of -5: l. extracted twice After with equal An aliquot of the aqueous phase was removed and sufficient concentration mixture acid 15% mobile (the phase "A" to assumption was made unaffected by the reduce extraction the acetonitrile the acetonitrile procedure). Derivatiza- that tions were performed immediately prior to injection. Separation The mobile phase derivatives was containing equimolar phan, to developed asparagine, be C ysteine separation of the amino basis, using a standard (except trypto- a trial-and-error amounts of all twenty amino acids glutamine, the latter two which forms during to present for on and hydrolyzed was gradient their as the acid disulfide, cystine. of peptide are acid expected h yd rol ys is). Assignments of th e -161peaks were made by spiking the derivatization reaction mixture with a single amino acid, or with the preformed FMOC-AA derivative. Minor variations in conditions were necessary with new columns. The mobile phase used is as follows: "A" = 2.55mL acetic acid was made up to 925mL with water, 1mL triethylamine was added, and the solution was made up to liter with methanol; "B" "A", using acetonitrile in place of water (prepared in 500mL batches). Fresh solutions were prepared daily. The gradient program used to separate the amino acid derivatives and reequilibrate the column is described in table ALI. Standardization of reaction conditions Several runs 8.0, 8.4, and 9.0 were conducted under identical conditions borate buffer: 2101JL 11.7mM FMOC-Cl with pH was added to a solution of 9.61J L standard in 2301JL -50mM borate buffer. The reaction was allowed to proceed for 20 minutes, at which time l251JL of the reaction mixture CH 3 CN. was removed to a The entire volume, after vial containing filtration, H 20 9381JL was used to and flush 2751JL and fill a 2001-1 L loop. using A time pH 9 analysis -50mM of borate. the derivatization Reaction conditions reaction was were follows: amino acid standard was expelled into 2301JL pH 9 2101JL aliquot of -llmM FMOC-Cl was added. proceed for a specified time (5, 10, 20, as performed 9.61J L 50mM borate, and a The reaction was allowed to 30, or 40 minutes). l251JL of the reaction mixture was diluted with 2751JL CH 3CN and 9381JL H 2 0 . After -162Table ALI. Gradient program for FMOC-AA elution and column re-equili- bration time(minutes) event flow rate climbs to 1.3 mL/ min. in 0.5 min. 0 %B = 15 filtration, 0.9 alarm sounds signalling beginning of gradient 1.0 %B climbs to 37 in 42.3 min. 43.3 %B climbs to 60 in 17.7 min. 61 o/oB climbs to 100 in 15.4 min. 76.4 flow rate climbs to 2mL/ min. in 81.4 flow rate is reduced to 1.3mL/ min. in 2 min. 118 flow rate is reduced to 0.3mL/ min. in 2 min. the entire volume was used to rinse min. and load a used for all 2001JL loop. Four runs were conducted for each reaction time. Calibration The run s: following reaction conditions calibration 2251JL 11.2mM FMOC-Cl was added to a solution of l01-1L amino acid standard in 2251JL pH 9 -52mM borate. ceed were for 45 minutes before aliquots of n-pentane. The reaction was allowed to pro- the mixture was extracted twice with 4601JL An aliquot of the aqueous layer was removed and diluted with the appropriate volume of CH 3 CN and "A" to achieve a final CH 3 CN concentration of 15%, and a final volume of -12501JL. One milli- li ter of the filtered solution was used to flush and load a 2001JL loop. Three individual runs of derivatized cysteic acid were conducted -163to obtain a rough dissolved in 50mL H 2 0 . 2mL to obtain were diluted 2.25mM with 225~-tL runs volumes of 1250~-tL, tra tion, was up to IOmL. by IO~o~L acid monohydrate Aliquots of IO~o~L of borate derivatized with 225~-tL and 60~-tL and of the aqueous this solution FMOC-Cl. phase, in total A second solution, 2.5mM in concen- 980~-t L ma k 1a g was aliquot of this solution was made up to were conducted. prepared L-cysteic solution. 200~-tL using A 460~-tL A a Calibration 91.5mg calibration. of a solution, O. ll92g in 25mL, aliquot of this solution was derivatized and run to give a third calibration point. Since the extent of extraction of each of the FMOC-AA is unknown, the concentration i.e., 200~-tL aqueous concentration, aliquot the (lx, 0.75x, aqueous 0.5x, expressed layer derivative of was after extraction layer. 0.25x, At 0.125x, only in 0.25x, etc., where x is dilution, for a 400~-tL and least two runs 0.0625x, and 0.03125x) and average peak areas were used to construct a derivative, which were fitted a by derivatives at relative each fashion, the concentration were performed response curve for each a least squares for actual sample of deterioration. The new with methanol linear regression method. Adjustment of conditions for new column A new first was showing five hours and then at column was signs lmL/min washed eleven obtained (to react hours at 0.4mL/ min analysis, column with was as the washed any chlorosilane), with i-propanol (to dissolve of Biological Compounds , any polymerized silanizing reagent). 4 4 Hancock, W., Sparrow, Dekker (New York), 1984. J. HPLC Analysis -164- Evaluation of possible internal standards An as aliquot of -15mM usual. amino acid standard solutions (in O.lN were prepared, potential internal standards used separate derivatization in was HCl) reactions. derivatized of and extracted compounds which were and aliquots of these were Equal volumes of the aque- ous phases of both the standard and the test compound reaction mixtures were combined, injected, and eluted as usual. Evaluation of the method The accuracy of the method was evaluated using sperm whale myoglobin, lysozyme, bradykinin, angiotensin II, and oxidized insulin chain B. Myoglobin, several water an cycles to provided in pH 7.2 50mM Tris buffer, was subjected of remove extinction concentration buffer salts. coefficient of in The and Centricons-1 0 concentration 157mM- 1 cm- 1 was 409.5nm. 5 at to dilution with determined using An aliquot con- taining lmg of protein was placed in a clean hydrolysis tube and lyophilized. One milliliter frozen in then thawed. liquid HCI nitrogen, The was the sample added tube was to protein, the solution was was evacuated, and the solution was subjected the to pump-thaw cycles before the at for 20 minutes. A second sample manner, but was several before being 155°C stored subjected to at tube was sealed. days hydrolysis at least two more freeze- The sample was incubated was treated in below 0°C after 110°C for twenty an identical lyophilization, four After digestion, the acid was removed by evaporation under vacuum. 5 Rothgeb, T., Gurd, F . Meth. Enzym. 50, 473 (1978). hours. The -165resid ue of each 9~-tL filtered. hydrolysates, sample was 10~-tL and aliquots, respectively, in -200~-t L pH 9 the 155°C and l I0°C dissolved were of diluted 226~-tL with respectively, and derivatized with 225~-tL FMOC-Cl. and (using 200~-tL of three filled-loop runs the borate, digestion 225~-tL and Area counts aqueous then borate, for four for the layer), l55°C and l I0°C hydrolysates, respectively, were averaged. One milligram transferred to prepared lmL hours. in a of hydrolysis HCl as tube, for was and dissolved in The lyophilized. myoglobin, and was 0.5mL water, sample digested at was li0°C 24 The acid was removed on an evacuated line, while the sample was warmed to 45°C with a times are production by serine). 6 For the to be from side The first 45 was 20~-tL 113~-tL "A". result as in significant esterification of artifact glutamic acid in 200~-tL borate and filtered . was diluted with 97~-tL borate, and extracted twice with 230~-tL pentane aqueous was dissolved hydrolysate with 140~-tL The second sample was prepared as follows: 20~-tL of the 215~-tL hydrolysate was diluted with Cl, and extracted they such FMOC-Cl, 100~-tL 1010~-tL as reactions, residue minutes. CH 3 CN and water bath to speed the evaporation (long drying a voided, run, derivatized with after bradykinin twice with layer diluted borate, derivatized with 460~-tL pentane after 45 225~-tL minutes. FMOC- 200~-tL of the aqueous layer was diluted with 90~-tL CH 3 CN and 960~-tL "A". One B, and milligram lysozyme samples were of prepared angiotensin II, oxidized and as for digested aliquots of the filtered angiotensin diluted with 209~-tL borate, and II derivatized insulin chain bradykinin. 26~-tL hydrolysate solution 225~-tL FMOC-Cl. with 6 Ika wa, M., Snell, E. J . Bioi. Chern. 236, 1955 ( 1961 ). were -166aliquots of the diluted with filtered insulin borate, 2201-!L and chain hydrolysate B derivatized with 2251-!L solution were FMOC-Cl. Two runs of each hydrolysate were conducted, and peak areas for each amino acid derivative were averaged. The divisor (to arrive at integer determined as follows: concentrations glutamic acid, glycine, alanine, ratios of amino arginine, for proline, acids) was aspartic acid, phenylalanine, methionine, and leucine were summed, and divided by the total number of equivalents of these amino acids expected. undergo Serine threonine were omitted, decomposition during acid hydrolysis. 7 these residues Valine and isoleucine were omitted, residues are known be particularly resistant residues are often underestimated tryptophan was known therefore these analyses. 8 If excluded from found to For samples visual the coelute of inspection some and to calculation, as after of the peptide to be derivatives completion unknown as of the composition, analyses of to bonds involving these acid hydrolysis, and amino acid in present, phenylalanine was of amino were these development integer the as ratios of were above-mentioned acids the method. derived by amino acids, composition and for which results were the most reliable. RESULTS Resolution Despite gradient, no numerous variations conditions were in found the mobile with this column 7 Osterman, L. Meth. Protein Nucl. Acid Res. 3, 464 (1984). 8 Hare, P . Meth. Enzym. 47, 3 (1977). phase type that would -167resolve all amino acid derivatives. No other columns were evaluated. FMOC-Ala and FMOC-OH could not be resolved on the original column on which the separation was developed of other derivatives, but the was achieved with the without partial resolution column. 9 second compromising of the these temperature A resolution two of species 35°C was required to resolve (FMOC) 2 -His and (FMOC) 2 -Lys with the second column (variations and in the improved leucine gradient the alone resolution derivatives. Higher of and glycine derivatives, of lysine and histidine program). At 35°C, and resulted lysine height. was In rarely the and peaks actual practice, achieved with resulted no of which in (for fused isoleucine, in to and of the the resolution original gradient amounts resolution containing derivatives), coelution the equimolar however, these improvement were samples resolve phenylalanine, derivatives derivatization in not temperatures threonine the could one of of histidine third these histidine the peak derivatives and lysine tn unequal amounts. Figure Al.l i.e., that which shows the had been resolution achieved used derivatives eluted consistently serine aspartic acid, and assignment of and for calibration pairs as threonine of and amino acids for times were highly variable. the mobile phase real with the second column, and analyses. Several fused peaks, including glycine. pep tides, This aided retention as in times the were unreliable. Retention the 9 volatile nature of This is most likely components (and therefore due to con- The lack of reproducibility from one column to the next, of the same t ype and manufacturer, is not unusual; Dolan, J . LC Magazine 3, 316 ( 1985); Ibid., p. 422. ca .0 (/) .0 '-0 ca c (.) <D lO <D C\J c E - 25 "'~ Q. 30 "'<5 35 t-C!J '->- .!:. - (.) "'>- (.) "' >- 45 0 .... ll. ::2 50 nl ...Q) > time (minutes) 40 <( -nl 0.. Q) 55 ~ ~=1.3 (.) "0 60 "0 "' =i 65 "0 "'>- _J "0 >f- .... 70 75 Figure A 1.1. Elution profile for the derivatized amino acid standard (column is that used for actual analyses). The prefix di- refers to doubly-derivatized amino acids, i.e., (FMOChAA's. 20 .1 .... <( C) (/) Q) .... u. ~ "'>"'>- (.) CP I en ~ -169tinually changing a result of mobile phase composition), and variations general, the was the range each derivative the period of the retention time of a derivative, elution volumes observed for it. greater of listed in to some extent in table Al.2 are column equilibration. the of at be In narrower Retention averages may times for 17 runs least (data from calibration runs). FMOC-Trp was found to coelute with FMOC-Phe. fere minimally with peptide analysis, (C. krusei cytochrome c and when present, expected is contains to as This should inter- tryptophan occurs residue infrequently one tryptophan be destroyed in after derivatization the per molecule), course of acid hydrolysis. 10 Determination of reaction conditions Injection tial, as of histidine the samples forms a soon disubstituted the derivative which is essen- decays to the monosubstituted derivative with a half -life of -5 days at pH 7.7. 11 Initial work indicated that also reacted with a equivalent Both the reaction singly was Conditions and carried were second doubly out sought in the phenolic oxygen atom of FMOC-Cl, to derivatized tyrosine bicarbonate buffer in which would result in bicarbonate buffer was abandoned pH and ionic strength. of form tyrosine an ester. formed when the the 7-8 pH reproducible range. analyses and complete deri va tiza tion. The in use of maintaining it at constant 10 Osterman, p. 465. 11 Einarsson, Josefsson, and Lagerkvist, p. 609. due to difficulties Borate buffer was -170- Table Al.2. Retention times and standard deviations for FMOC-AA derivatives FMOC-AA Retention time (minutes) cr Arg 28.76 0.92 Ser 30.19 0.83 Asp 30.59 0.80 Glu 32.68 1.06 Thr 34.87 0.96 Gly 35.33 0.95 Ala 39.89 0.82 CysCys 42.67 0.72 Pro 45.28 0.66 Met 48.18 0.71 Val 50.33 0.63 Phe 53.80 0.41 lle 54.54 0.40 Leu 54.95 0.37 diCysCys 57.98 0.32 diHis 63.60 0.32 diLys 63.86 0.35 diTyr 67.05 0.30 -171- selected as it is one of the few alkaline buffering agents lacking an amine functionality (which would react with the reagent). The extraction was omitted in the study of pH and time dependence of the reaction to avoid complications arising from the lack of reproducibility this procedure introduces. The indicate data that derivatives aspartic presented modest in table Al.3 and increases in peak area as the reaction pH is acid, and glutamic acid showed attributed to column derivatives was observed). derivative concommitant decrease in the A 1.2 FMOC-AA from to 9. 8 anomalous behavior, of in seen with monotyrosine figure for increases are in observed (doubling Dramatic tyrosine are increased deterioration disu bsti tu ted illustrated peaks the Serine, which for these formation increasing derivative is is of a pH. A observed, but does not appear to be as dramatic. An evaluation of the pH dependence of the rates of histidine and lysine derivative formation could be performed if a control study was done for the coeluting FMOC byproduct, but this was not pursued, as the results for tyrosine derivative formation demonstrated the advantage of performing the reaction at pH 9. Data are presented showing the analytical results for reaction times spanning time der iva tiza tion table Al.4 and of the derivatization dependence many reaction are stable. derivative increases over for reason 45 derivatives 5 is derivatives this in to 40 are plotted constant minutes, in minutes, Formation of the range of reaction minutes was chosen as figure Al.3 reaction. The a range of that the that the for indicating complete the in and disubstituted time the tyrosine investigated, reaction time and for -172- Table A1.3. Area counts as a function of derivatization pHa Amino Acid pH 8 pH 8.4 pH 9 Arg 90558 90856 94881 Ser 90674 141154 153809 Asp 84318 41168 27703 Glu 82824 24051 44130 Thr 91198 90062 93400 Gly 95068 97720 104515 Tyr 50231 15272 21314 CysCys 111672 107875 112234 Pro 79643 84666 92130 Met 82621 95187 90950 Val 82647 86360 93266 Phe 70949 80443 91981 Ile 67937 77945 89049 Leu 75114 83106 91439 diTyr 11794 87808 180262 a. A quantitative evaluation of the histidine and lysine derivatives was not possible, as they coelute with an FMOC byproduct (which is completely removed in routine analyses by pentane extraction). -173- 110000,----------------------------------------------------, 100000 ~ 90000 ::l 0 a • • • • u "'~ 80000 70000 Arg Thr D Gly Pro Met Val .. • • Phe lie Leu 60000~-r~--.-~~-,r-~~-.--r-~-r~--~.-~~~.-~~~ 7.8 8.0 8 .2 8 .4 8 .6 8 .8 9 .0 9 .2 pH 200000~r================,-------------------------------~ --t:r- FMOC-Tyr di(FMOC)Tyr "' E ::l 8 100000 "'~ "' 0+-~~-,.-~~-.~--r-.-~~-,.-~~-.-,--r-,-~~~ 7 .8 8 .0 8.2 8 .4 8 .6 8.8 9.0 9 .2 pH Figure A1.2. Peak area as a function of the pH of the derivatization reaction (20 minutes duration). 129878 183459 197625 85080 208917 127906 95812 107872 102315 99468 100695 105770 115137 141728 180284d 128464 Ser/Aspb Glue Thr/Giyb CysCys Pro Met Val Phe lie Leu diCysCys diHis diLys diTyr 179705 183208e 134642 146952 92717 89866 92389 93923 92103 94035 152789 199855 85587 186823 93442 20 min. 179193 185924 131703 105852 93707 90378 92596 93646 96070 94362 128055 200303 85560 173414 92782 30 min. 190157 180985 134939 90677 93353 89259 93247 93885 92307 %838 122541 203166 84884 183925 91697 40 min. a. averages of four runs, unless specified otherwise b. Due to column deterioration, resolution of the components of these pairs of fused peaks was poor. Hence, the sum of the areas for each pair only were considered. c. Due to column deterioration, FMOC-Giu eluted as a doubled peak; the sum is reported here. d. single run (unresolved from FMOC byproduct in other runs) e. average of three runs 152135 105322 97164 89877 93067 93159 93380 93945 107949 199101 84033 171985 92819 101126 Arg 10 min. 5 min. Amino acid Table A1.4. Area counts as a function of derivatization reaction time8 I I .t:> -..J ..... C'O ~ Q) C'O u 0 :::J c (/) - a. ~ 0 0 v Pro CysCys Glu Thr/Giy - Phe -+--lie -Leu diCysCys diHis dilys diTyr 1 - Met -Val -+--o-- · - Arg - + - Ser/Asp 10 30 reaction time (minutes) 20 Figure A1.3. Area counts as a function of derivatization reaction time. 0 40 50 50000+-------------~--------~--~~--~--~--------~ 100000 150000 200000 ~ 250000 I ~ I CJl ....:J -176derivatization of real peptide hydrolysates. Several amino acids were derivatized to a suggesting the derivatives minutes than after 10 minutes, However, the peak areas were consistent time and pH for larger reaction extent after are times 5 unstable. equal to or variability is exceeding l 0 minutes. For both the expected for runs tion was not stopped but rather was slowed injection. Also, upon derivative stability may of carried under abruptly by the from baseline studies, some identical reagent considerably vary the out dependence by conditions, removal dilution as and as the with extraction, exposure to method of integration for one to the next, depending extent of coelution and run the a of reac- -pH 4 particular on the other com- pounds. Alternatives to extraction is readily hydrolyzed, and the hydrolysis product must be FMOC-Cl extracted from the chromatography. product the the solution required be 12 Pentane effectively derivatives. 12 cantly reaction found in The peak has with pentane area mixture counts been to interference the remove extraction the of amino found to (a direct comparison was not the extraction of CH 3 CN, by solution to solubilize the fluorenyl coefficient Cunico, R ., Mayer, A., Instruments at Work Publication Creek, CA), 1986. of Dimson, P., Varian #163, in the hydrolysis was changed partition its extraction is the minimize reported minimal volume on to reduce acid signifi- possible, which as is moiety; no data acetonitrile in n-pentane). Zimmerman, C. Instrument Group could Varian (Walnut -177- The extraction introduces error. difficult to pentane, it sure the small volumes and is complete removal of extraction the rep rod uci bili ty. of Because pipet the high vapor pressure quantitatively; the high vapor pres- reaction make involved of in the derivatization the first pentane step ought to lead to thought that reduction Also, it was layer Hence, difficult. of omission sensitivity greater of the and FMOC-OH concentration might permit resolution of FMOC-Ala. The possible reagent 13 extreme was use of an investigated. hydrophobicity, so additional amine to 1-amino-adamantane that elution of its scavenge was the selected FMOC excess for derivative its would not interfere with that of the FMOC-AA's. Several problems were encountered with this approach, among the solubility problems with FMOC-aminoadamantane the scavenging (once 45 reagent reaction minutes is beneficial, left as to have to it compete elapsed sea venge). removed with for the Pentane impurities the and the inability hydrolysis of derivatization extraction was well excess as as them of the reagent reaction, little found to reagent be and hydrolysis product. Calibration Peak deri va ti ves, areas were except found diHis, to increase diLys, and linearly for diTyr, all with FMOC-AA correlation coefficients of 0.9973 or greater (figure Al.4, tables Al.5 and A1.6). A -250pmol, conservative considering estimate the for the sensitivity concentrations used 13 Betner, I., Foldi, P. Chromatographia 22, 381 (I 986). of the with method the is lowest -178- 500000 Ill •a • 400000 (/) ...... c 300000 ::J •c • 0 (.) ~ Q) ..... 200000 Arg Ser Asp Gly Ala lie Val ~ 100000 0 0.0 (/) ...... 0.2 1.0 0.8 0.4 0.6 relative concentration 1.2 300000 c ::J 0 (.) Ill 200000 • ~ a Q) ..... ~ • • c • 100000 0.0 0.2 0.4 0 .6 0.8 Glu Thr Pro Met Phe Leu cysteic acid 1.0 1.2 relative concentration Figure A1.4. Calibration data for FMOC-amino acid derivatives. Solid lines indicate results of least squares fitting. 340894 418708 356943 349891 393015 269276 399302 350651 350453 349247 331005 366826 Ser Asp Glu Thr Gly Ala Pro Met Val Phe lie Leu 265530 256216 266724 254842 254482 283050 202622 284026 259188 260554 281939 260433 252230 0.75x 200384 174357 182624 179526 179092 208176 121036 198914 174867 180428 202017 179443 180217 O.Sx 102627 88229 97920 92127 92834 99423 64407 100391 90044 88552 99891 90759 90243 0.25x 48880 43706 47858 44311 44122 51211 23738 49616 42199 43188 49663 45221 42558 0.125x Data for three single runs of cysteic acid: 0.45x (144929), 0.25x (87244), 0.135x (43733). 355470 l.Ox Arg Amino Acid Relative Concentration 25060 22330 27196 22572 22619 24106 18623 24617 22222 21223 23142 22087 22327 0.0625x Table A1.5. Calibration data - peak area (average of at least 4 runs) as a function of concentration 15789 13223 15383 13662 13413 14564 7290 14371 13263 12247 13246 12650 12711 0.0375x I f--0 I co -..] -180- Table A 1.6. Coefficients for calibration data analysis a Amino acid m (x10 6 ) b r2 Arg 2.85 2.53 e-5 0.9983 Ser 2.92 -7.72e-3 0.9993 Asp 2.44 9.62e-3 0.9954 Glu 2.81 2.51 e-3 0.9996 Thr 2.87 -9.33 e-4 0.9998 Gly 2.57 -2.18e-3 0.9992 Ala 3.67 1.90 e-2 0.9968 Pro 2.54 -1.81 e-3 0.9976 Met 2.89 -5.39 e-3 0.9990 Val 2.89 -5.50e-3 0.9991 Phe 2.89 -1.77e-2 0.9993 Ile 3.00 -8.99 e-3 0.9992 Leu 2.76 -1.55 e-2 0.9964 Cysteic acid 3.13 -9.67 e-3 0.9946 a. concentration = m(area counts) + b; r = correlation coefficient -181attenuation, and assuming no extraction ,S-alanine, and of acetonitrile or the amino acid deri va ti ves. Internal standards Norvaline, norleucine were evaluated as potential internal standards. FMOC-,S -alanine incompletely resolved chromatogram is is (the crowded retention elutes after from, FMOC-cystine. undesirable for FMOC-alanine, elution monosubstituted characteristics) of but This an tyrosine and dominated by elutes immediately after before, region and of the internal standard, as derivative has similar the tailing it peak of the FMOC-leucine, and the reagent hydrolysis product. FMOC-norleucine peaks of of these two FMOC-leucine (for attempt was made gradient due to components are a -4:1 to improve the lack of fused ratio to of of one-third norleucine the resolution success about by this to variations method in the height leucine). in No the elution improving the resolution of the histidine and lysine derivatives. FMOC-norvaline was found to elute immediately after FMOC-valine, with minimal fusion of the peaks (-5% of the FMOC-valine Provided it is stable under the conditions of acid peak height). hydrolysis, norvaline would be a good choice for an internal standard, as its FMOC derivative is well-resolved from the compounds of interest. Evaluation Data for total amino acid analyses of known peptides are compiled m table Al.7. The determinations agree quite well with the theore- 1.65 (2) 4.05 (5) 11.6(11) 19.8 (17) 3.85 (4) 1.99 (2) 0.90 (1) 2.91 (3) 2.37 (2) 0.94 (1) Thr 2.03 (2) 3.02 (3) 0.97 {1) + + (1) (1) + (3) + (2) (2) + (19) + (!) 2.21 (2) + (12) + (2) (2) + (3) + (19) + (12) 17.4 (18) 8.15 (9) 5.42 (9) 15.1 (18) 6.55 (6) 8.07 (8) 6.38 (6) 3.65 (4) 2.74 (3) 1.01 (1) 0.93 (I) 2.90 (3) 1.00 (1) 1.02 (1) 18.6 (17) 10.8 {11) 4.70 (5) a. Found (theoretical); moles per mole of peptide b. Theoretical values from the catalog of the Sigma Chemical Company, 1987. c. Theoretical values: Edmundson, A. Nature 205, 883 (1965). d. Theoretical values: CRC Handbook of Biochemistry, 2nd edition, H. Sobcn, cd., Chemical Rubber Co. (Cleveland), 1970, p. C-237. Trp Cys~H Tyr Lys His Leu lie Phe Val Met Pro Ala Gly 5.08 (8) 3.82 (4) 19.1 (19) 3.34 (3) Glu 19.0 (19) 7.67 (8) 8.11 (8) 1.00 (1) Asp 0.94 (1) 5.10 (6) 5.91 (6) 1.06 (1) Ser 3.94 (4) Sperm Whale Myoglobin c 155°C l10°C 0.95 (1) Insulin, chain B b (oxidized) 3.36 (4) 1.98 (2) Arg 1.02 (1) Angiotensin II b 0.98 {I) Bradykininb Table A 1.7. Analyses of known peptides • + (6) + (3) + (6) + (!) 6.79 (8) 3.69 (6) 5.23 (3) 4.09 (6) 1.44 (2) 1.89 (2) 11.8 (12) 12.2 {12) 7.10 (7) 5.53 (5) 22.5 (21) 11.0 (10) 10.8(11) Hen Egg d Lysozyme I I N CXl ...... -183tical values for small the analysis with peptides. increasing The protein deterioration of size puzzling is a the quality of trend also found for several other precolumn derivatization methods. 14 The FMOC-amino i.e., purpose, sequence. the Its identification shortcomings method and the system might better ity would chemistry resolve be greatly However, the problems His) derivative and acid analysis method is adequate for its intended are a peptides from a protein result of both the of the the FMOC-AA enhanced of of by multiple instability system. Another column derivatives, and the use sites of (especially of of experimental or solvent the sensitiv- fluorescence derivatization FMOC 2 His) cannot known detection. (Tyr, be Lys, readily eliminated. 14 (a) Bergman, T., Carlquist, M., Jornval, H . Advances in Methods in eds., a/., Wittman-Lie bold, et B. Analysis, Protein Microsequence Springer-Verlag (New York), 1986; (b) Oray, B., Lu, H., Gracy, R. J. Chrom. 270, 253 (1983). -184- Appendix 2 Tryptic Digestion of the Modified Derhathe and Related Studies -185- INTRODUCTION SMP If arginine, the krusei to cleave C. at peptide bonds whose five peptide products specifically were fragment P2, peptide chain linkage to Cysl4-Arg38 includes the Cysl4 heme cytochrome carboxy would c completely and is contributed by moiety result. His33 would occur (see figure 3.10). This and Cys17, which provide the group. Hence, should these prosthetic stretch on of the covalent bonds remain intact (as they are expected to do under the mild conditions of enzymatic digestion), the UV -visible spectrum of the His33-containing peptide would be dominated by the porphyrin bands, and the presence of [A 5 RuHis] 3 + would be difficult to confirm by this technique. In an effort to eliminate conducted using the protease whose carboxy moiety is specific digestion expected to in reside fragment on 16 problem, trypsin, contributed Complete, result this C. of peptides which by krusei (see T6, Val28-Arg38, peptide mapping cleaves lysine or c A2.1), independent of peptide arginine cytochrome figure was by and the also bonds residues. trypsin His33 heme is would fragment T5, Cys14-Lys27. The results results, but action, which dent studies are rather reported to may of all relate have here the not as a occurrence important confirmation of implications an of the SMP side re- unexpected for pentaammineruthenium-modified temperature-depenproteins in mildly basic solution. MATERIALS AND INSTRUMENTATION TPCK-treated Corporation and trypsin was obtained used as supplied. from United Digestion of native States Biochemical and modified C. -186- 1 10 Pro-Aia-Pro-Phe-Giu-Gin-Giy-Ser-Aia-Lys-Lys-Gly-Aia-Thr-Leu-Phe- - - - - - - - T1 T21· T3--- ·I 20 Lys-Thr-Arg-Cys-Ala-G ln-Cys-His- Thr-lle-Glu-Ala-G ly-G ly-Pro-His•1• T4--=t T5------30 40 Lys- Val-Gly-Pro-Asn-Leu-His-Gly-lle-Phe-Ser-Arg-His-Ser-Gly-Gln--~~--------------- T6----------------~~------------- 50 Vai-G lu-G ly-Tyr-Ser-Tyr-Th r-Asp-Aia-Asn-Lys-Arg-Ala-Gly-Vai-G luT81....,•------- ------17 60 70 Trp-Aia-Giu-Pro-Thr-Met-Ser-Asp-Tyr-Leu-Glu-Asn-Pro-TMK-Lys- - - - - - - - T9 T101 80 Tyr-Ile-Pro- G ly-Th r-Lys- Met- Ala -Phe-G ly-G ly-Leu-Lys-Lys-Ala-Lys1• Tll • T12 ~13rT14--J 1 90 100 Asp-Arg-Asn-Asp-Leu-Val-Thr-Tyr-Met-Leu-Glu-Ala-Ser-Lys J--T15 T16--------t ·I· Figure A2.1. Peptides expected upon digestion of C. krusei cytochrome c with trypsin. The points of cleavage (after Lys and Arg residues) are indicated by vertical lines. -187- krusei cytochrome c was carried out as described by Mauk and Mauk. 1 Sources of HPLC solvent components and a description of the HPLC instrumentation can be found in Chapter 3. column (model 218TP54, #755577) Na[CoEDTA] was synthesized A Vydac 4.6mm x 250mm Cl8 was utilized to resolve the peptide mixtures. [Ru(NH 3 ) 5 Im]Cl 3·2H2 0 as described was synthesized by K. chapter in 2. Yocom. 2 The synthesis of a identical [Ru(NH 3 ) 5 His]Cl 3 is described in chapter 3. Control digestion, experiments without were analyzed the were use carried out in trypsin. The of manner CoEDT A- reaction to the products by UV -visible spectroscopy and HPLC after separation on Biogel P-2 (BioRad, 200-400 mesh). DPP studies were conducted as described in chapter 3. METHODS Peptides were A 0.1%TFA tion: = gradient program presented in resolved in water; B for table using peptide A2.1. The the = following O.l%TFA, mobile phase composi- 10%H 2 0, in CH 3CN. separation and column reequilibration column was maintained at 27°C, The is and peptides were detected by their absorption of 220nm light. RESULTS In the course of concentrating HPLC fractions to recover peptides 1 Mauk, M, Mauk, A. J. Chrom. 439, 408 (1988). 2 Sundberg, R ., 96, 381 (1974). Bryan, R ., Taylor, I., Taube, H. J. Amer. Chern. Soc. -188Table A2.1. Gradient program for the separation of tryptic peptides time (minutes) event 0 for possible 12.5%B; flow rate climbs to l.OmL/min. 0.5 min. in 0.9 alarm 1.0 injection 6.0 %B climbs to 25 in 50 minutes 56 %B climbs to 50 in 50 minutes 106 %B climbs to 100 in 5 minutes 111 flow rate climbs to 2mL/min. in minute 116 flow rate decreases to 1mL/min. in 1 minute; %B returns to 12.5 in 15 minutes 156 flow rate decreases in 1 minute analysis, it was noted that a to peptide 0.3mL/min. of -30 minutes elution time from the digest of the modified cytochrome was purple in color. No eluted in such a was found in the native peak(s) in the modified digest but due lack region where a one of the native retention, it was relative to ibility in in two the peptide digests. No digest, not other certain which differences were digest. to peaks The peptide was shifted of reproduc- were correlated observed between the revealed two elution profiles of the native and modified digests. A UV -visible spectrum of the peptide (figure A2.2) visible bands with maxima at -450nm and -550nm, as well as an absorption band at 280nm. While the [A 5 RuHis] 3+ complex has an absorbance maximum at -450nm, it also has a more intense absorption centered at -189- ,, --, I I 1 I 0.16 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I \ I 'J I I I I I I I I I I I I I I I I I ~ <.I c co ... &I Q II) &I 0.08 I co I I I I I I I I I I I \ \ ',tl 0 200 300 400 soo 600 700 wavelength (nrn) Figure A2.2. UV-visible spectrum of a modified peptide recovered from a tryptic digest of the AsRu(His39)-C. krusei cytochrome c : - in O.lM sodium phosphate;--- after being made basic by the addition of sodium hydroxide. -190which 300nm, observation parative is lacking was made batch of in with the a spectrum second of tryptic digest of a krusei A 5 Ru(His39)-C. peptide. 1 the cytochrome spectrum had maxima at 275.6, 321, 450, and 550 nm. peak was split, indicating that probably two A similar different pre- c: the peptide In that case, the peptides were present. No further purification of either peptide was attempted. The 550nm (J\max=384nm, absorption 538nm), suggested might be that involved oxidant, the in a CoEDT A- reaction. 2 side The presence of oxidant alone, or the presence of any product which might result from such small, hydrophobic CoEDT A- its reaction with the ruthenium charged molecules are not column. Control complex might studies undergo reagent expected were to done substitution be to are unlikely, as retained by the determine if the analogous to reactions those of the aquopentaammine complex:3 [A 5 Co(OH 2 )] 3 + + RCOOH -+ [A 5 Co-OOCR] 2 + + H 3 o+ Very small yields of products were obtained upon incubation of CoEDT Ain the presence of lysine, but none had spectral properties similar to ruthenium model complexes to those of the peptide. Control confirm studies the inert were performed nature of with the [A 5 Rulm] 3+ (lm=imidazole) was found absorption but spectroscopy, noticeable changes. while the These ruthenium to be quite robust [A 5 RuHis] 3+ changes coordination spectrum included 1 Sundberg, R., Gupta, G. Bioinorg. Chern. 3, 39 (1973). 2 Seaman, G., Haim, A. J. Am. Chern. Soc. 106, 1319 (1984). 3 Hawkins, C., Lawson, P. Inorg. Chern. 9, 6 (1970). by sphere. UV-visible showed broadening minor of the -191300nm band, but no features resembling those of the peptide were observed. Differential pulse ducted to to associated be determine figure A2.3. to be due polarography the reduction with the studies of the peptide potential of the metal results are peptide. The were site con- presumed presented in The intense anodic peak near 0.85V (vs . SCE) was thought to redox of the DPP studies processes free of amino tyrosine, acid. and Further this was evidence confirmed by indicate the to presence of tyrosine is the pH dependence of the UV absorption (figure A2.2): the and occurs as red only features shift a result found not attributable (vs. SCE). The -20nm the to with its presence on DPP behavior as that the side reaction ruthenium was present base DPP trace Tyr) -150m V couple, in of ionization of in (and [A 5RuHis] 3 +/ 2+ of are peak but the described is characteristic of the phenolic side chain. 4 which anodic is in are UV-visible peptide. The in at region the undigested 3 (figure resulted from the digestion on protein as -0.15V and 0.4V expected for the spectra chapter the a The unique to the peptide maxima the tyrosine, are inconsistent protein exhibited 3.18), demonstrating process, and pentaammine that imidazole complex. DISCUSSION A literature survey of ruthenium possibilities 4 as to the identity of ammine compounds suggested the complex Cantor, C., Schimmel, P. Biophysical Freeman (San Francisco), 1980, p. 377. on Chemistry, the peptide. Part 2, some Rudd w. H. -192- ....~\ .\ . ; \ '' \ f ' / ' :• '' ··- ~~ \. I ~ . I .......... ---- .. ------ ... ' '\··· ... •. ·, ........_-......... .. ' ' ' I //··-...... .... -- .. ··. '• ' \ ·-. ... __ ...... ..~'' ' ' ........... _____________ , .... ,,"'' .. .......... ---- -----............... - -0 .5 0 0 .5 1 .0 potential (V vs. SCE) Figure A2.3. Differential pulse polarography traces for (upper) odd peptide recovered from AsRu(His39)-C. krusei cytochrome c tryptic digest and (lower) tyrosine; legend: anodic and -- cathodic wave. -193and Taube have proportionates i.e., isomer, reported in basic and a that pentaammineruthenium(III) solution to give the ruthenium(II) compound. 5 ruthenium(IV) pyridine The latter dis- compound, exhibited an absorption maximum at 558nm in carbonate buffer. It was proposed that the deprotonation of an can be reported to disproportionation results from an initial ammine ligand. Aquation ruled out, of as an the ammine spectral ligand to yield [A 4 Ru(Im)(H 2 0)]n+ features for the 3+ species are be Amax=297 and 385 nm,6 and the 2+ complex has no visible absorption as well. 7 an amino acid Aquation of an carbonyl ammine ligand, nitrogen to form or followed a by coordination peptide chelate is of pos- sible, and absorption bands in the region of 295nm and 555nm have been reported for complexes. 8 such is It difficult to rationalize, how- ever, why these reactions might occur on the peptide but not with the histidine model complex. of The observed properties of the product recovered the modified cytochrome indicate it probably is from a the digest ruthenium-con- taining peptide which includes His39, and which has undergone some reaction to alter the coordination sphere of the ruthenium complex. this is observed illaris protease in the tryptic is puzzling, as digestion the and sole not that difference of the Why submax- (aside from the Polyhedron 4, 1059 5 Rudd, D., Taube, H. lnorg. Chern. 10, 1543 (1971). 6 Isied, S., Kuehn, C. J. Arner. Chern. Soc. 100, 6754 (1978). 7 lsied, S., Taube, H. lnorg. Chem. 15, 3070 (1976). 8 Kohata, (1985). S., Sagara, K., Takada, H ., Ohyoshi, A. -194- identity of the protease) used (NH 4 co 3 vs. can be drawn between NaHC0 3 , from these the should be borne in procedures respectively). An results side is that ruthenium complex do occur at elevated this two mind when the buffer important conclusion reactions temperatures "recycling" after variable-temperature experiments is practiced. is of involving in mild a that the base, and modified protein -195- Appendix 3 Spectroelectrochemical Data -196- Table A3.1. Native Candida krusei cytochrome c spectroelectrochemistry dataa T=9.1 (TscE=24.5) T=15.2 (TscE=23.9) T=19.8 (TscE=24.75) E A54s E A54s E A548 -231.0 -25.2 -4.4 15.2 35.4 56.2 86.1 115.0 230.1 0.815 0.785 0.755 0.699 0.608 0.498 0.372 0.318 0.289 232.3 109.6 80.4 55.2 34.8 15.1 -5.1 -25.3 -230.6 0.287 0.314 0.371 0.472 0.573 0.663 0.724 0.758 0.791 -230.4 -24.1 -4.4 15.4 35.6 56.1 81.1 110.7 230.3 0.782 0.749 0.700 0.633 0.542 0.446 0.359 0.311 0.288 T=24.4 (TscE=26.25) T=29.6 (TscE=25.9) T=34.2 (TscE=25.1) E A54s E A54s E As4s 230.7 109.8 79.9 55.0 34.8 14.0 -6. 1 -25.5 -230.9 0.291 0.314 0.355 0.431 0.519 0.618 0.685 0.727 0.774 -230.9 -24.4 -4.3 15.2 30.5 51.9 85.9 115.2 230.0 0.768 0.712 0.660 0.589 0.523 0.432 0.344 0.314 0.299 229.9 109.3 79.8 54.6 34.2 14.8 -5.9 -25.6 -229.6 0.309 0.323 0.353 0.418 0.498 0.587 0.665 0.715 0.769 T=39.8 (TscE=24.25) T=44.8 (TscE=24.25) T=50.2 (TscE=24.0) E A548 E A548 E A548 -229.7 -23.8 -4.5 15.7 36.2 57.3 83.8 113.2 242.6 0.769 0.685 0.627 0.548 0.466 0.399 0.351 0.328 0.313 242.6 109.8 79.3 54.4 35.0 12.8 -6.9 -26.3 -234.9 0.320 0.331 0.353 0.398 0.453 0.534 0.608 0.666 0.754 -234.9 -23.9 -4.6 15.6 36.3 56.0 81.7 110.3 239.9 0.748 0.637 0.576 0.500 0.432 0.387 0.352 0.336 0.328 a. Data is expressed as follows: potentials are given in mV vs. SCE, [0] and [R] are the concentrations of oxidized and reduced forms of the protein, respectively, and T and T seE are the temperatures of the working and reference electrodes, respectively, in degrees Celsius. -197- Table A3.1. Native Candida krusei cytochrome c spectroelectrochemistry data, continued T=54.8 (TscE=24.0) Table A3.2. E A548 239.8 75.4 54.9 34.3 14.8 -5.2 -25.5 -45.9 -85.0 - 150.7 -231.1 -351.1 0.343 0.361 0.379 0.409 0.452 0.504 0.552 0.590 0.629 0.649 0.656 0.662 Fitted parameters to spectroelectrochemical Nernst plots for native Candida krusei cytochrome c 8 Temperature _!L Il'lexptl mtheor %error in m ..c. 9.1 15.2 19.8 24.4 29.6 34.2 39.8 44.8 50.2 54.8 45.44 40.80 37.21 32.15 27.55 24.76 17.41 11.84 5.34 -18.48 56.7 56.2 55.9 59.3 59.6 56.7 64.4 62.5 62.3 96.2 56.0 57.2 58.1 59.0 60.1 61.0 62.1 63.1 64.1 65.1 1.25 1.75 3.79 0.51 0.75 7.05 3.74 1.01 2.80 48.0 0.9998 0.9998 0.9994 0.9999 0.9999 0.9999 0.9999 0.9998 0.9997 0.9710 a. b, ffiexptl• and flare the intercept, slope, and correlation coefficient, respectively, of a plot of the potential (vs.SCE) vs. the log of the ratio of concentrations of oxidized to reduced forms of cytochrome c; mtheor is the theoretical value of the slope, and is equal to RT/;;c, assuming a one electron reduction. The error in miscalculated as X 100 -198- Table A3.3. Modified A 5 Ru(His39)-Candida krusei cytochrome c spectroelectrochemistry data (units and abbreviations as in Table A3.1) Sample 1 T=7.6 (TscE=19.4) T=12.6 (TscE=21.0) T=19.6 (TscE=22.35) E A54s E A548 E A548 299.8 129.4 97.4 71.8 46.9 29.4 12.1 -11.5 -40.8 -202.0 0.201 0.223 0.278 0.369 0.484 0.552 0.598 0.629 0.643 0.651 -202.0 -34.9 -6.4 19.6 43.8 64.5 87.1 110.8 299.4 0.643 0.631 0.608 0.558 0.468 0.375 0.292 0.241 0.202 299.6 112.9 87.0 62.5 41.4 20.0 -8.3 -35.4 -205.3 0.207 0.236 0.280 0.356 0.443 0.521 0.586 0.611 0.625 T=26.4 (TscE=22.5) T=31.0 (TscE=22.0) T=36.6 (TscE=22.0) E A54s E A54g E A548 -205.4 -61.2 -34.9 -7.1 15.7 38.0 54.9 75.3 101.7 300.5 0.611 0.602 0.591 0.560 0.507 0.426 0.362 0.296 0.246 0.211 300.4 115.6 89.4 64.0 38.3 7.6 -16.1 -45.8 -246.1 0 .213 0 .230 0.258 0.313 0.402 0.504 0.552 0.581 0.598 -246.2 -78.8 -38.8 -13.3 8.9 34.1 50.6 97.8 299.7 0.589 0.579 0.560 0.528 0.477 0.394 0.338 0.242 0.212 Sample 2 T=6.8 (TscE=24.6) T=36.6 (TscE=27.5) T=40.6 (TscE=27.75) E A548 E A54s E A54g 295.8 111.8 88.9 70.8 49.0 29.0 6.1 -20.4 -241.4 0.284 0.333 0.392 0.468 0.574 0.659 0.716 0.747 0.761 -240.7 -70.6 -50.1 -18.2 7.0 31.2 54.9 73.0 145.5 298.6 0.697 0.688 0.679 0.646 0.587 0.5 11 0.416 0.367 0.300 0.296 298.7 122.4 90.9 70.5 47.4 23.9 -0.7 -30.8 -233.9 0.303 0.314 0.337 0.368 0.428 0.509 0.587 0.644 0.682 -199- Table A3.3. Modified A5 Ru(His39)-Candida krusei cytochrome c spectroelectrochemistry data, continued T=47 .0 (TscE=26.5) E A548 -331.6 -78.5 -49.4 -19.5 7.1 31.9 56.2 84.8 298.4 0.670 0.658 0.641 0.601 0.536 0.455 0.387 0.338 0.320 Table A3.4. Fitted parameters to spectroelectrochemical Nernst plots for modified A 5Ru(His39)-Candida krusei cytochrome c 8 Temperature ___Q_ Illexptl mtheor %error in m £ 7.6 12.6 19.6 26.4 31.0 36.6 59.13 53.80 48.06 41.48 36.22 30.81 55.6 56.6 57.3 60.6 60.4 66.3 55.7 56.7 58.1 59.4 60.3 61.4 0.07 0.10 1.30 1.95 0.09 7.91 0.9993 0.9999 1.0000 0.9989 0.9997 0.9966 6.8 36.6 40.6 60.07 30.95 28.42 54.0 59.7 61.8 55.5 61.4 62.2 2.81 2.86 0.69 0.9996 0.9981 1.0000 a. b, mexptl• and r 2 are the intercept, slope, and correlation coefficient, respectively, of a plot of the potential (vs. SCE) vs. the log of the ratio of concentrations of oxidized to reduced forms of cytochrome c; mtheor is the theoretical value of the slope, and is equal to RT/7, assuming a one electron reduction. The error in miscalculated as Imtheor - ffiexptll mtheor The blank line separates the results of two samples. X 100 -200- Appendix 4 Problems with the Flash Photolysis Instrumentation "A wise sage has commented more serious disadvantages of always gives an answer.'." J. Gunn, (1981 ). J. Westphal, Soc. (in another context) 'One of the this photometer is that it nearly Photoopt. lnstr. Eng. Proc. 290, 16 -201Flash photolysis data were collected with two of PMT in design, (1) nonlinear response, and, existed for the labeling other tion these problems, it was at one time common practice to the air data cooling The resulting drift, which as as presented the may probe of have figure the flash eighteen warming well (see results, past (2) cooling air a breakthrough. region m of the possible problems These have months, and hence are the reason for in 1.1 preliminary also. In addi- table lamp immediately prior to flashing the lamp gives rise to some of those contributed A4. 1). the flaws instrument to operation major alone error to This problem does was left on not apply throughout a a sample. measurable measurements to each turn off the Candida experiment. A better solution to this problem is discussed in a review by Porter. 1 The tubes is state that maintained current. 3 The operated with sion specif ica tions 2 manufacturer provided linear at or R928 a by operation can below level PMT PMT for the in housing 200KQ the be assured of flash containing resistors. photomultiplier Hamamatsu l / 20th photolysis ten the output current the "bleeder chain" if apparatus stages, with an applied Typically, has been voltage divi- voltage of 800V is used, hence the bleeder chain current would be i = V / R = 800V / (2x 10 6 Q) = 0.4mA The output current is passed through a 5KQ resistor, and monitored by a voltmeter. Typically, a 3V signal is observed, with a 50mV change re- 1 Porter, G. Techniques of Organic Chemistry 8, 1055 (1963). 2 Hamamatsu catalog no. SC-1-4-1, Hamamatsu TV Co., to voltage Ltd., Hamamatsu, Japan. 3 "Bleeder chain" refers the divider circuit in housing which maintains the voltage difference between each dynode. the PMT -202suiting after flashing the modified The output chain current derivative. current is is -1.5 approximately v / R = 3V; 5ooon = 0.6mA ; = Therefore, the ratio (0.6mA / 0.4mA), 0.05 for output to significantly higher than the operation. Hence, the possibility linear of bleeder maximum recommended of exists that the PMT current ratio to light output signal is not directly related to light intensity. Figure A4.2 depicts the relationship intensity. Although a value of predicts nonlinearity of the PMT small changes in light much, but all, of not 1.5 intensity the of this for the current ratio definitely response for large intensity changes, may behavior result for in current a linear ratios response, exceeding as 0.05 is linear. The second event itself. light of Because from major With almost of the the problem relates the the 3-73 entire is readily IO~s broader signals seven to The baseline. typical of 50mV which possible, as is the (-426nm cutoff) the flash lamp and conducted down the optical of to m the flash use, flash on the cell holder, fiber sample. light leading to A blank sample (water only) gives rise not be related milliseconds are signal observed to required cannot in PMT an be direct reduction Subtraction of signal intensity this was found overload before ignored electron immediate, signal response impinges to 1Om V + signal PMT spike expected from the flash itself, but also to presumed Approximately the spectrum the monochromator and PMT. only to the FWHM filters visible configuration flash to the in transfer of from to A4.3 ). signal returns rei a tion to the experiment, half site by data is not dependent on the the actual be (figure iron -203- S'-0.05 .......... ~ "iii c: cu ~ .s -0.1 ~ cu (I) 0 I -0. 15 0 40 20 time (seconds) Figure A4.1. Signal resulting from cessation of air flow cooling probe lamp (no flash, water in cell, A.=549nm). -e c ._ , 10r-------~----~------~------~ ,, ,,' :I u._ cu ~ ,,, .......~=~ 1.0 ~---L----+-_,_ .~ 0 -0 ~ ._ ._ 0.1 r-------~----~~------~------~ - 0. 01 !'------::~~----+------;----------1 :I u :I 0.. -; 0 0 0 - ~ 0.001 ______...__ _ _......._____~------___. 0 .001 0.01 0.1 1.0 10 Light Intensity (arb. unit) Figure A4.2. Relationship of output-to-bleeder chain current ratio and light intensity for a typical Harnamatsu photomultiplier tube (page 6 of reference 2). -204sample absorbance, a property which is expected electron transfer life data of Assuming experiment. to this i.e., problem, one to change throughout an could "sacrifice" t 1; 2 =-7ms, a one maximum half- rate of approximately IOOs- 1 could be measured with this instrument. As recommended by the Hamamatsu Company, 2 capacitors were installed in the last four PMT among the final dynodes, incident upon but reduction no the housing stages to maintain tends to be lost rise to a which This PMT. gave in the time necessary to eliminate the voltage division when intense light "ringing" of for the return problem of flash is the signal, to baseline (figure A4.4). In another effort 450nm (:!: 40nm) bandpass boro, VT) to transmission at reduce filters the 452nm; the were -549nm 0.01% at obtained (Omega component of 550nm). definite A the noted, in that a lapse of -Sms was required for baseline on problem has strong is blank been whatever sufficient to (figure the A4.5). identified, -550nm overload overloaded omitting flash properly that already and a light the PMT by the first half-life the manages to (which, Optical, flash Brattle- light (-72% improvement the signal to Therefore, but breakthrough, it flash was return to that the seems intensity is pass through the as discussed above, so filters is beam). Using these i.e., t 112 =Sms, a is not an population of [Ru(bpy) 3]2 + molecules molecules. Increasing decreasing the intense probe of data, filters, maximum rate of 140s- 1 could be measured. A reduction in necessary to excite to reduce a sizeable the concentration intensity) would the a of result flash intensity sufficient portion the in of the cytochrome ruthenium complex a background larger (and option, absorbance as it is flash (background -205- ........... > .......... ~ ·u; .,c 0.02 +J .s ., +J C/1 0 I 0 0 sx1o-3 0.01 time (seconds) Figure A4.3. Signal resulting from flashing a blank (water) sample, with detection at 549nm. ........... > .......... ~ ·u; .,c 0.02 +J .s ., +J C/1 0 I 0 0 5x10-3 0.01 time (seconds) Figure A4.4. Signal resulting from flashing a blank sample after the installation of capacitors in the PMT housing to maintain the voltage difference between the final dynodes (1..=549nm). -206- >. ~ c 0.02 r- ..... Q) .s ..... ..... ..... 0 Q) (/) I a~ 0 5x10-3 0.01 time (seconds) Figure A4.5. Signal resulting from flashing a blank sample using 450nm bandpass filters (A.=549nm). -207absorbance is -0.4 at 549nm for the 65JJM (Ru(bpy) 3 ]2 + routinely used). The problems associated with the flash photolysis instrumentation may be rectified by the changes suggested below. 1. The probe neutral and beam density light intensity filter or output-to-bleeder iris should be reduced by that the output current ratio is reduced. so chain current a Simply operating the lamp at a reduced voltage was found to increase noise. below 0 .05 response Operation may lead curve is of the PMT at a current ratio as the slope of the portion to larger signals, steeper here than on any linear fiber and impinging above y=0.5. 2. The probe beam the face of the to a diameter would exiting flash less the optical cell should than that seem that insufficient that losses could be the light impinging be collimated of on the cell width. is not a is important light tolerated, the and it on focused While problem, to it and reduce sample. citation band centered at 452nm extends as far as 580nm, so that the probe beam of 'A> 530nm does itself excite the complex somewhat, and will result in some loss of signal. 4 4 For a typical sample of 1.3JJ M cytochrome, the absorbance change occurring at 550nm upon reduction (through a pathlength of l5cm) is ( 1.3x 10-6 M)(I 5cm)((29000 - 7 800)M-l em -l) = 0.41 Absorbance tensities any is defined incident absorbance to as log(I 0 / I), and exiting change log(I 0 / Ii), where from accompanying where If exiting light intensities, respectively: and a 10 a and sample, reaction Ii I are are the light respectively. can the be final Hence, expressed and in- as initial -2083. Fittings which of connecting hold the cell the optical in place and provide a means fibers should be made of a black material to reduce light scatter. of cell where the probe to distance the optical the longer, light rays originating this connector would from The beam exits fiber the further fitting on the side should be made from flash the Baffles lamp. discourage the path of within propagation of flash light to the PMT. These suggestions are but a few of the many possible changes which would improve the operation of the possibilities, which would require include the use a with output current the PMT of PMT without with a changing photodiode, a flash photolysis apparatus. in basic changes fewer the stages light the (dynodes) intensity, solid state device or Other components, to reduce the replacement of whose response is less likely to be perturbed by intense incident light. llA = log(I 0 / If)- log(I 0 / Ii) = log[(I 0 / Ir)/(1 0 / Ii)] = log (1/If) For an absorbance intensities initial is change expected intensity to reading of 0.4, the be 2.51 (=10°.4). of 3.0V, a ratio final of initial and final light Therefore, for a typical 1.20 ( =3.0 / 2.51) voltage of is expected. However, a change of only -50mV is observed. dicative operation of c urve (as that probe behavior. predicted light in the nonlinear by the current which bypasses the This is in- portion of the PMT response ratio). It is important to note sample may also contribute to this -209- Appendix 5 Programs for Data Fitting -210- All programs included in this appendix were written by Fernando J. Selman. Subroutines Recipes, W. not Press, explicitly B. written Flannery, University Press (New York), 1986. S. here can Teukolsky, be found in Numerical W. Vettering, Cambridge -211 - PROGRAll BLT2DAT C••••****************************************************** c I c I C Used to reformat data from flash experiment {voltage I C readings) into a format understandable by the WEXPFIT_DPI C program {relative absorbance values). I c•••••••••••••••••••••••••••••••••••••••••••••••••••••••••• c IMPLICIT NONE REAL TIME{2096), ABSORBA{2096), SIGYAO, VOLTAGE{2096), RVOLT{2096), SIGYAV, VI, VHAT, V3I, GAIN, TSPIKE, T, Dmffil, AVER, SUY, SIGYA, NXWIND, LOGlO, XFACTOR c INTEGER LEN, !SPIKE, !START, NDATA, NVALUES, LENAYE, I,J !Time data vector (seconds) !Absorbance vector !Estimate of error in absorbance !Offset amplified voltage data !vector !Raw voltage data vector !Estimate of error in voltage !Baseline raw voltage !Voltage offset constant !Baseline offset amplified voltage !Gain (will divide offset voltage) !Time of flash !Time (after flash) !Dummy variable !Moving average !Sum of baseline voltages !Moving standard deviation !Length of window for AVER !ln{lO) !Number of standard deviations to define flash spike. !Fortran integer library function !Position of spike in data vector !Length of window for moving avg . !Number of data points !Number of points after spike !Length of filename !Integer loop variables !Filename of input (BINLOT output) !Filename of output !Dummy character !Prefix of input filename !End of file CHARACTER•16 FILEl , FILE2, DUYCH•l, PREFIX LOGICAL EOFIL, FOUND PARAMETER (LOG10=2 . 302585093) c C Ini tial values c c c EOFIL=.FALSE. FOUND=. FALSE . XFACTOR=5 . C -------------------- ---------- INPUT ----------------------------1 - 212- 100 c c WRITE (6, 100) FORMAT(' Enter name of input file READ(5, '(A) ')FILE1 --> ',1) C If filename has an extension , use as is . If not, append .BLT. c 110 c LENAYE=LEN(FILE1) I=O DO WHILE ( ( .NOT.FOUND) .AND. (I.LT .LENAME) ) I=I+1 DUYCH=FILEl(I:I) IF (DUMCH.EQ.' . ') THEN FOUND= . TRUE. PREFIX=FILE1 (: I-1) END IF END DO IF (.NOT .FOUND) THEN PREFIX=FILE1 (: LENAME) FILE1=FILE1 ( :LENAliE) I I, .BLT' ENDIF WRITE(6,110)FILE1 FORMAT(' Opening file ',A16) OPEN (UNIT=11 , FILE=FILE1 , TYPE=' OLD' , FORll= 'FORMATTED') c C Get gain c 120 WRITE(6,120) FORllAT(' Reading GAIN from first line and skipping 2 lines .. ') c 125 READ(11,125) DUVCH,GAIN FORYAT(A1,F3 .0) WRITE(6,•)' GAIN= ',GAIN READ(11,'(A)') DUYCH READ(11,'(A)') DUMCH c C Reading rest of file c 130 150 WRITE(6,130) FORMAT(' Reading rest of file . . ') I=O DO WHILE (.NOT. EOFIL) 1=!+1 READ(11,•,END=150) DUYR1 , TIME(!) , VOLTAGE(!) , DUYR1 DUYR1 , DUMR1 , RVOLT(I) END DO CONTINUE NDATA=I- 1 CLOSE(ll) C-------------------------------------------------------------------1 0-------- - ---- ------------ - PREPARING DATA -------------------------1 c -2 13- C Finding spike c 160 c c c c c 200 205 206 WRITE(6,160) FORMAT(' Finding position of spike .. ') !START= 20 NXWIND = FLOAT(ISTART) SUY=O. DO J=1,ISTART SUY=SUY+VOLTAGE(J) END DO AVER=SUY I NXWIND SIGYA=O . DO J=1,ISTART SIGWA=SIGWA+(VOLTAGE(J)-AVER)••2 ENDDO SIGYA=SQRT(SIGWA)ISQRT( NXWIND -1. ) !=!START DO WHILE ( .NOT.FOUND .AND. I.LT.NDATA) !=!+1 SIGMA = SIGMA - (VOLTAGE(I-ISTART)-AVER)••2 I (NXWIND- 1.) AVER =AVER + (VOLTAGE(I)-VOLTAGE(I-ISTART)) I NXWIND SIGMA= SIGMA + (VOLTAGE(I)-AVER)••2 I ( NXWIND - 1. ) IF ( ABS(VOLTAGE(I)-AVER) .GT.XFACTOR•SIGMA) FOUND=.TRUE. END DO IF ( FOUND ) THEN !SPIKE=I WRITE(6,200) !SPIKE FORMAT(' . . at data pair ',!4) ELSE WRITE (6 , 205) !FIX (XF ACTOR) FORMAT(' .. no ',!4,' sigma spike in data') END IF WRITE(6,206) FORMAT(' Enter spike position--> ',$) READ (5,•) !SPIKE c C Calculating average and dispersion of baseline values. c C AVERAGE . . . WRITE(6,210) 210 FORMAT('Calculating average and dispersion of initial values .' ) V3I=O. VI=O . NXWIND=FLOAT(ISPIKE-2) DO I=1,ISPIKE-2 V3I=V3I+VOLTAGE(I) VI=VI+RVOLT(I) END DO -214- V3I= V3IINXWIND VI = VIINXWIND C DISPERSION ... SIGYAV=O . DO I=l,ISPIKE-2 SIGYAV=SIGYAV+(VOLTAGE(I)-V3I)••2 END DO SIGYAV=( SQRT(SIGYAVIGAIN ) I SQRT( NXWIND ) ) VRAT=(V3IIGAIN-VI) SIGYAO=ABS( SIGYAV I ( VI • LOGlO ) ) c C Calculating absorbances c WRITE(6,220) FORMAT(' Calculating absorbances .. ') DO I=l,NDATA VOLTAGE(I)=VOLTAGE(I)IGAIN-VHAT C Offset amplified voltage variable redefined as a C raw voltage (dumb!). ENDDO 220 c DO I=l,ISPIKE ABSORBA(I)=O. END DO DO I=ISPIKE+l,NDATA IF ( VOLTAGE(I).LT. 0. ) THEN ABSORBA(I) = ALOGlO( ( VIIVOLTAGE(I) ) ELSE ABSORBA(I) = 10.0 END IF END DO ) c C------------------------------------------------------------------1 C-------------------------- OUTPUT --------------------------------1 C Storing results c c FILE2=PREFIX(:LENAYE)II'.DAT' OPEN(UNIT=12,STATUS='NEW',FILE=FILE2,FORW='FORYATTED') WRITE (6 300) FILE2 FORMAT(' Storing reformatted data in ',A16) I 300 c c NVALUES=NDATA-ISPIKE TSPIKE=TIYE(ISPIKE) DO I=l, NVALUES T=TIME(I+ISPIKE)-TSPIKE WRITE(12,•) T,ABSORBA(I+ISPIKE),SIGWAO,VOLTAGE(I+ISPIKE) ENDDO CLOSE(12) C Storing some useful information in file FILEl.HDR c WRITE(6,350) PREFIX(:LENAYE)II'.HDR' -2 15- 350 400 FORMAT(' Storing useful information in ',A16) OPEN (UNIT=13,STATUS='NEW',FILE=PREFIX(:LENAME)//' .HDR') WRITE(13,400)GAIN,VI,VRAT,ISPIKE,TSPIKE,NVALUES,SIGYAO,SIGMAV, TIME(ISPIKE+l)-TSPIKE,ABSORBA(ISPIKE+l),VOLTAGE(ISPIKE+l), TIME(NDATA)-TSPIKE,ABSORBA(NDATA),VOLTAGE(NDATA) FORYAT('l',Tll,' GAIN= ',FQ.2/ Tll,' VI= ',F9.3/ Tll,'OFFSET = ',F9.3/ Tll,'ISPIKE = ',I9/ Tll,'TSPIKE = ',F9.3/ Tll,' NDATA = ',I9/ Tll,'SIGYAO = ',E9.3/ Tll , 'SIGYAV = ',E9.3/ Tll,' TO= ',FQ .3,' AO = ',F11.4,' YO= ',F9.3/ Tll,' TF = ',F9 .3,' AF = ',F11.4,' VF = ',F9 . 3) STOP END -216- c PROGRAM WEXPFIT DP C------------------------------------------------------------\ C DOUBLE PRECISION VERSION. I C This program is used to calculate the fitting parameters I C of a nonlinear model; written after subroutine func_fit tol C specify physical sizes of arrays, as required by FORTRAN. I C------------------------------------------------------------C c IMPLICIT INTEGER PARAMETER c NONE NYAX, NPMAX ( NYAX=2096, NPMAX=20 ) !max. no. of data points !max. no. of parameters X(NYAX), Y(NYAX), SIG(NYAX), xo, XE, !initial x value for fit !final x value for fit XDUll, yy c A(NPYAX), !initial values for parameters DYDA (NPMAX) , ALPHA (NPYAX, NPllAX) , CHISQ, COVAR(NPYAX,NPYAX), EPS, Q, SIGMA_A(NPYAX) c INTEGER INTEGER c YA, liFIT, NDATA, NVALUES, NlUN COUNTS, NYAX-STEPS c c c c LISTA(NPYAX), INTEGER LENAYE, ICH LEN IO, IE, I, J CHARACTER•64 FILEl, FILE2 !parameters to be varied !no. of parameters in model !no. of variable parameters !Integer function. !index for data pair at start of fitting window !index for data pair at end of fitting window !data filename !parameter filename -2 17- c CHARACTER•64 FILE3, FILE4, PREFIX !filename for fitted parameters !filename for theor. y values LOGICAL !not end of file NOTEOF, FOUND c C--------------------------- INPUT ---------------------------------1 c 100 WRITE (6 ,100) FORMAT(' Enter file with (x,y,sig) data [.DAT] --> ',$) READ(5,'(A)') FILEl c C Determining if file has extension; if not, .dat appended. c C C c 200 LENAYE=ICH LEN(FILEl) Ich len is a subroutine which determines the length of a filename. I=O DO WHILE ( (.NOT.FOUND) .AND. (I.LT.LENAWE) ) I=I+1 IF (FILEl(I:I).EQ.'.') THEN FOUND=. TRUE. PREFIX=FILEl (: I-1) END IF END DO IF (.NOT. FOUND) THEN PREFIX = FILEl (: LENAYE) FILE1 = FILEl(:LENAME)//'.DAT' END IF OPEN (UNIT=10,FILE=FILE1,STATUS='OLD') NOTEOF=.TRUE. I=O DO WHILE (NOTEOF) I=I+1 READ(10,•,END=200) X(I),Y(I),SIG(I) ENDDO CONTINUE NDATA=I-1 CLOSE(UNIT=lO) c C Reading parameters for fit c 210 c c FILE2=PREFIX(:LENAYE)//'.PAR' WRITE(6,210) FILE2 FORMAT(' Reading parameters from ',A16) OPEN(UNIT=11,FILE=FILE2,STATUS='OLD') READ (11, •) MA READ (11, •) llFIT DO I=1,MA !Number of parameters in model !Number of free parameters - 218- c READ (11, •) A(I) END DO DO I=1,WFIT READ(11,•) LISTA(I) END DO READ(11,•) XO READ (11 , •) XE CLOSE(UNIT=11) !Initial guesses for parameters !List of parameters to be varied !Starting x-value for fitting window . !Ending x-value for fitting window. 0-----------------------PERFORMING FIT -----------------------------0 WRITE(6,•)' Performing fit .. ' CALL FUNCT FIT(X,Y,SIG,NDATA,XO,XE,YA,WFIT,A,LISTA, -COVAR,ALPHA,NPYAX,Q,CHISQ,SIGYA_A) WRITE(6,•)' OK. ' c OUTPUT 0-----------------------c C Storing fitting parameters c 300 c c FILE3=PREFIX(:LENAYE)I/'.FIT' WRITE(6,300)FILE3 FORMAT(' Storing fit results in ',A16) OPEN(UNIT=12,STATUS='NEW',FILE=FILE3,FORM='FORMATTED') DO I=1,YA DO J=1,YA WRITE(12,400)I,J,COVAR(I,J),I,J,ALPHA(I,J) ENDDO ENDDO 400 FORMAT(' COVAR(',I2,',',I2,') = ',D12.6, . , ALPHA(' ,I2,, I, ,I2, ') = , ,D12.6) WRITE(12,500) Q,CHISQ 500 FORMAT(' Goodness of fit= ',D12.6,' chi-square= ',D12.6) DO I=1,YA WRITE(12,600) I,A(I),SIGWA_A(I) END DO 600 FORMAT(' A(',I2,') = ',D12.5,' SIGllA A= ',D12.6) CLOSE(12) - c C Storing a tabulation of the fitting function c 650 FILE4=PREFIX (: LENAYE) I I ' . THE' WRITE (6 , 650) FILE4 FORMAT(' Storing fitting curve (x,y pairs) in ',A16) c OPEN(UNIT=12,STATUS='NEW',FILE=FILE4,FORM='FORMATTED') c 710 WRITE(12,•)' OK. ' WRITE (12, 710) FORMAT(5X,' X ',5X,' CALL LOCATE(X,NDATA,XO,IO) y ') -219- c CALL LOCATE(X,NDATA,XE,IE) C LOCATE is a subroutine (Numerical Recipes) which finds the indices C of the data pairs most closely matching the desired fitting window . c c IF (IO.LE.1) I0=1 IF (IE . GE . NDATA) IE=NDATA NVALUES=IE-I0+1 DO I=1, NVALUES XDIDl=X(IO+I-1) CALL FUNCT (XDUY, A, YY, DYDA, llA) C FUNCT subroutine calculates theoretical y values for each x C value using specified parameters. c WRITE(12,•) XDUK,YY END DO CLOSE(12) STOP END c SUBROUTINE FUNCT FIT(X,Y,SIG,NDATA,XO,XE,YA,MFIT,A,LISTA, - COVAR,ALPHA,NP,Q,CHISQ,SIGMA_A) C+ C----------------------------------------------------------------1 C DOUBLE PRECISION VERSION . I C Determines indices of data pairs most closely matching desired! C fitting window. Calls subroutines FUNCT, LOCATE, and I C NONLINFIT (the last itself calls many subroutines from I C Numerical Recipes). I c C INPUT: c c c c c c c c c c c c c c c c c c c c c >X >y > SIC > NDATA > xo >XE >llA I Vector of dimensions NDATA containing the independent, error free abscissas. Vector of dimensions NDATA containing the dependent variable. Vector of dimensions NDATA containing the standard deviations of the Y's. INTEGER•4 number of data points REAL•4 The starting value of the X coordinate: to be used for the fitting. REAL•4 The last value of the X coordinate to be used for the fit . INTEGER•4 The number of parameters for the fit > MFIT > A(llA) > LISTA(llA) INTEGER•4 The number of parameters to be varied during the fit. REAL•4 Vector of dimension WA, containing the initial guesses for the parameters. It will contain the final values INTEGER•4 Vector containing the list of parameters to be varied in its first MFIT -220- c slots. INTEGER•4 Physical dimensions of arrays c >NP c COYAR and ALPHA. c C OUTPUT: c c REAL•4 Vector containing the desired < A(lLA) best fit parameters in its first lLA c c slots . REAL•4 Vector containing the standard c < SIGlLA_A(lLA) deviations of the calculated parameters c REAL•4 Covariance Matrix c < COYAR (NP, NP) c REAL•4 Curvature Matrix < ALPHA(NP ,NP) REAL•4 Goodness of fit c <Q c < CHISQ REAL•4 Chi-square c C ----------------------------------------------------------------' C+ c IMPLICIT NONE INTEGER NllAX STEPS, NYIN COUNTS, NDATA,YA,YFIT,NP, I,J,IO,IE,NVALUES c EPS, X(NDATA),Y(NDATA),SIG(NDATA), A(WA),SIGYA A(YA),LISTA(lLA), XO,XE, ALPHA(NP,NP),COVAR(NP,NP), Q,CHISQ, X1(2096),Y1(2096),S1(2096) c C c c PARAllETER (NllAX STEPS=lOOO,NlliN COUNTS=4,EPS=l.OOE-04) (Default values for max~ and min . no. oi iterations, epsilon . ) EXTERNAL FUNCT C Determining indices of array marking position of window to C be used for fitting. c IF ( (XO.GT.X(l)) .AND . (XE .LT .X(NDATA)) ) THEN CALL LOCATE(X,NDATA,XO,IO) CALL LOCATE(X,NDATA,XE,IE) ELSE IF ( (XO.GT .X(l)) .AND. (XE . GT .X(NDATA)) ) THEN CALL LOCATE(X,NDATA,XO,IO) XE=X(NDATA) IE=NDATA ELSE IF ( (XO.LE.X(l)) .AND. (XE.LT.X(NDATA)) ) THEN CALL LOCATE(X,NDATA,XE,IE) XO=X(l) IO=l ELSE IO=l -221- IE=NDATA XO=X(1) XE=X (NDATA) ENDIF NVALUES=IE-I0+1 c 100 WRITE(5,100) NVALUES,X(IO),X(IE),IO,IE FORMAT(' Using ',I3,' values for fit.' I I • 'From X= ',D10.4,' to X= ',D10.4,'.'11 • ' From IO = ' , 13, ' to IE= ' , 14) c DO I=1 ,NVALUES X1(I)=X(IO+I-1) Yl(I)=Y(IO+I-1) S1(I)=SIG(IO+I-1) ENDDO c CALL NONLINFIT(X1,Yl,Sl,NVALUES,A,YA,LISTA,WFIT, COVAR,ALPHA,NP,CHISQ,FUNCT,Q,NYAX STEPS,NYIN COUNTS,EPS) (Summons actual fitting routine) - C c DO I=l,YA SIGMA A(I)=DSQRT(ABS(COVAR(I,I))) END DORETURN END c : SUBROUTINE NONLINFIT ( X,Y,SIG,NDATA,A,WA,LISTA,WFIT, COVAR,ALPHA,NCA,CHISQ,FUNCS, Q,NYAX STEPS,NWIN COUNT,EPS) 0-----------------------------------~----------~----------------------- C DOUBLE PRECISION VERSION. C This subroutine does a nonlinear fitting of the function FUNCS to C the data X,Y,SIG. The parameters of the fit are in the array A(WA). C The list of WFIT parameters to be varied are in the first WFIT slots C of LISTA(MA). The remaining WA-WFIT parameters are held fixed at C their input values. The method used is the Levenberg-Marquardt C method as implemented in Numerical Recipes. The user also supplies C the maximum number of iterations allowed in the variable NYAX STEPS, C the minimum number of iterations with a negligible change in CHISQ C required to stop the iteration in the variable NWIN COUNT, and the C meaning of 'negligible change in CHISQ' in the variable EPS, it C being the maximum allowable percentual change in CHISQ(e.g., 0.001). C The user must supply the first guess at A(WA), the parameters of C the fit. The output is the optimal set of parameters A(WA), C the covariance matrix COVAR(NCA,NCA), the curvature matrix C ALPHA(NCA,NCA), the goodness of the fit Q, and chi-square CHISQ. C--------------------------------------------------------------------- c FJS I I CIT 1988 C==================================================== c IliPLICIT NONE c -2 22- c SUBROUTINE VARIABLES DECLARATIONS c INTEGER c NDATA 1 WA 1 NCA 1 LISTA(YA) 1 YF'IT 1 NYAX_STEPS, NYIN_COUNT X(NDATA),Y(NDATA),SIG(NDATA),A(WA), COVAR(NCA,NCA),ALPHA(NCA,NCA), Q,EPS,FUNCS,CHISQ c C GENERAL VARIABLES DECLARATIONS c DCHISQ,PREV_CHISQ,ALAYBDA,GAMYQ c INTEGER NSTEPS,ICOUNT c C Initialization call to MRQYIN c ALAYBDA=-1.0 CALL MRQMIN(X,Y,SIG 1 NDATA 1 A,WA,LISTA,YFIT,COVAR,ALPHA, NCA,CHISQ,FUNCS 1 ALAYBDA) c C Iterative procedure for optimization c NSTEPS=O ICOUNT=O DO WHILE (ICOUNT .LT .NlliN COUNT) PREY CHISQ=CHISQCALL-MRQMIN ( X, y, SIG, NDATA A, WA LISTA, llFIT, COVAR,ALPHA,NCA,CHISQ,FUNCS,ALAWBDA) DCHISQ=DABS(CHISQ-PREV CHISQ)/CHISQ IF (DCHISQ.LT.EPS) THEN ICOUNT=ICOUNT+l WRITE(6,•)' Iteration #',!COUNT,' going for ',NlliN_COUNT ELSE ICOUNT=O END IF NSTEPS=NSTEPS+l IF (NSTEPS .GT.NYAX_STEPS) PAUSE ' NO CONVERGENCE ' END DO I I c C FINAL CALL TO MRQMIN WITH ALAYBDA=O. c ALAYBDA=O .O CALL WRQMIN(X,Y,SIG,NDATA 1 A1 WA,LISTA,llFIT,COVAR,ALPHA, NCA,CHISQ,FUNCS,ALAYBDA) c C OBTAINING GOODNESS OF FIT c Q=GAMYQ( (FLOAT(NDATA) - 2.)/2. RETURN END c I CHISQ/2 . ) -223- SUBROUTINE FUNCT(X,A,Y,DYDA,NA) C-------------------------------------------------------------------1 C DOUBLE PRECISION VERSION. I C Given an independent variable x and parameters for a multiexponen-1 C tial fit, calculates theoretical values for the dependent variable! C y and dy/d[A(NA)]. Function is of the form I C y = A(1) + A(2){1-exp[-A(3)t]} + A(4){1-exp[-A(5)t]} etc. I C To be called from subroutine FUNC FIT or NONLINFIT. I C----------------------------------=--------------------------------' c c IYPLICIT NONE INTEGER NA, !Number of parameters • NEXP !Number of exp. components INTEGER I DIMENSION A(NA),DYDA(NA) REAL•8 X,A, Y,DYDA,SUY,EXPARG LOGICAL LARGE LARGE=. FALSE. NEXP = ( NA-1 ) / 2 DO I=1,NEXP EXPARG = A(2•I+1)•X IF ( EXPARG . GT. 20.) THEN LARGE=. TRUE. END IF END DO c C FOR LARGE ARGUMENT (to prevent underflow) c IF (LARGE) THEN SUY=O. DO I=1,NEXP SUY=SUY+A(2•I) DYDA(2•I)=1. DYDA(2•I+1)=0. END DO DYDA(1)=1. Y=A(1)+SUY RETURN END IF c C FOR REGULAR SIZE ARGUMENT c SUY=O. DO I=1,NEXP SUW=Silll+A(2•I)•( 1. - DEXP( -A(2•I+1)•X ) ) ENDDO Y = A(1)+SUll DYDA(1)=1. DO I=1,NEXP DYDA(2•I) = 1. - DEXP( -A(2•!+1)•X ) DYDA(2•I+l) = A(2•I)•X•DEXP( -A(2+!+1)•X) ENDDO -224RETURN END -225- Appendix 6 Characterization of the Second Derivatization Product -226In traduction Described tion of here the are second data relating to the product resolved on CM-52 aquopentaammineruthenium(II) and C. krusei preliminary characteriza- of the reaction cytochrome c. This are identical of species will be referred to as "product 2". All those instrumentation. described in materials. chapters 2 and and 3 methods used for preparation the and to char- acterization of A 5 Ru(His39)-C. krusei cytochrome c. Purification Because of the slow mobility of modification reaction mixture on high strength buffer (made ionic the third CM-52. so band resolved it was often by the addition from eluted the with of a NaCl). Although there was no indication of inhomogeneity of this product from its fied shape. 1 it band products attempts were should of was formed at known from electrophoresis long reaction times that (figure several modi- Therefore, 2.2b). made to resolve any constituent components by FPLC. be noted that the second product native seems to formed the be at known to be A 5 Ru(His39)-cytochrome of mobility expense of higher product I, than It that which is c, indicating there is at least one site modified in addition to His39 in product 2. Product 2 was eluted from an analytical MonoS cation-exchange FPLC column with a variety of gradients. In each case. a single product was observed. which eluted on top of a broad band. major Using the The shoulder on the tail of the third band seen in figure 2.3 is likely to be a result of the abrupt switch to a higher ionic strength elution buffer. -227method described from to 12.8 mL, broad band The include native those fully of chapter 26.4 A6.1). as in and the oxidized with 2, retention the peak singly-modified major (reduced major which native is broad at band ranged 19.2-l9.3mL (figure component cytochrome component, species this occurring the preceding of likely to isomers, as well presumed to elutes 12.8mL). redox itself cytochrome is at be a Such products may have formed from degradation of the major component, as evidence from the FPLC studies of the ruthenium label is much cates that state. Components modification His26, or at higher of higher sites other retention than modifications, more ~3 labile may be and His39 His33 i.e., c A 5 Ru(His39)-cytochrome in the reduced of double products ruthenium infra), (vide ions indi- per e.g., cytochrome molecule. Samples utilized for NMR were purified All other experiments were conducted on the using this FPLC method. product as it eluted from the CM-52 column, after desalting and the appropriate buffer change. Results EPR spectra for the native and two modified products are presented in figure signals of indicative product recorded the A6.2. at compared of a Since in the spectrum of A 5 Ru(His39)-cytochrome histidine the complex spectra of evident both modified temperature, from the iron and ruthenium apparent ratio of the intensity in product to the iron signal a are identical signals greater 2. As seen comparison centers of 2 is the the spectrum products were intensities appropriate. ruthenium of The signal as vs. A 5 Ru(His39)-cytochrome c suggests more than one histidine has been modified. the of in c, -228- 1.0 0.04 0 .8 0.03 0 .6 'E c: z Ill 0 () co ~ 0 .02 a> (.) c: «< ~ , -e0 , VJ .0 , , , I 3 0 iii' ~. I 0 .4 :g I «< , , I I 0 .01 , I , I I I I 0.2 0 , I ~r------r------r------r----~0 0 10 20 30 time (minutes) Figure A6.1. Chromatogram for the analytical FPLC elution of the oxidized product 2, the third band eluting from a CM-52 separation of the modification reaction mixture (--gradient, 0-50%B in 32 minutes at lmL/min; injection at 2 minutes; unretained product is CoEDTA-). -229- (a) (b) (c) (d) 2000 4000 3000 5000 Gauss Figure A6.2. EPR spectra of (a) 7.73mM [AsRuHis]Cl3, (b) 2mM native C. krusei cytochrome c, (c) 1.85mM AsRu(His39)-C. krusei cytochrome c, (d) 2.1mM AsRucytochrome modification product 2. Protein samples in J.l=0. 1M pH 7 NaPi, complex in 0.1M NaAc, pH 5. Initial sample temperatures (rose as much as 0.5K during a run) : (a) 12.8K, (b) 12.6K, (c) & (d) 13.3K. Other conditions: modulation amplitude, 160; power, 0.2mW; gain 1.6 x 104; frequency: (a) 9.25GHz, (b) 9.245GHz, (c) 9.247GHz, (d) 9.2485GHz. -230NMR spectra for are presented Both sharp absent in His39 is in figure histidine the the native cytochrome and A6.3 C-2 spectrum of two modified (o6-10ppm) and A6.4 (o-22 resonances in the o8.7ppm product 2, confirming modified, and indicating His33 resonance is evident at -o-3lppm (in observed for is modified addition A 5 Ru(His39)-cytochrome), the which to as the may be a to products -38 ppm). region are suspicion that well. new A o-32.5ppm signal histidine proton resonance contact-shifted by a proximal ruthenium complex. A UV -visible difference spectrum krusei cytochrome is shown in figure expected for a spectrum are those complex, and ferences in for product 2 The A6.5. minus native C. features of the pentaammineruthenium(III) histidine the intensity of the 300nm absorbance (corrected redox state concentrations) is such that a 2:1 for dif- ruthenium- to-cytochrome ratio is indicated. Discussion and Conclusion The UV -visible spectrum and NMR spectrum of product 2 together indicate that experiments both His33 (DPP) and should His39 be have been modified. performed on this Electrochemical derivative to verify with fact that reaction mix- that no other (non-histidine) sites have been modified. The native ture, production protein suggest is of this recovered that some derivative, from product, the together A 5 Ru-modification singly-modified be present. No evidence was found for such a sible it that cytochrome exists product. as a minor contaminant NMR, 2 EPR, and at the His33, should product, but it to UV -visible the also is pos- A 5 Ru(His39)- difference spectra -231- Figure A6.3. NMR lH spectra for (a) native, (b) AsRu(His39)-modified, and (c) AsRumodified (product 2) C. krusei ferricytochromes c, 86-lOppm vs. DSS. -232- (a) 9 . 00 8 . 50 8 . 00 7 . 50 7 . 00 6 . 50 7 50 7 . 00 6.50 7 . 50 7.00 6 50 PPM (b) 9 . 50 9 . 00 8 . 50 8 . 00 9 . 50 9 . 00 8 5c 8 . 00 PPM (c) PPM -233- (a) -23 . 0 - 24 .0 -25 . 0 -26 . 0 -27 . 0 -28 . 0 -29 . 0 - 30 . 0 - 31.0 - 32 . 0 - 33 . 0 -34 . 0 - 35 . 0 -36 . 0 -37 . 0 PPM (b) - 23 . 0 -24 . 0 -25 . 0 - 26 . 0 - 27 . 0 - 28 . 0 - 29 . 0 - 30 . 0 - 31 . 0 - 32 . 0 -33 . 0 - 34 . 0 - 35 . 0 -36 . 0 -37 . 0 PPM (c) - 23 0 - 24 . 0 -25 . 0 - 26 .0 27 . 0 - 28 . 0 - 29 . 0 - 30 0 -31.0 -32 . 0 -33 . 0 -34 . 0 -35 . 0 - 36 . 0 - 37 0 PPM Figure A6.4. NMR lH spectra for (a) native, (b) AsRu(His39)-modified, and (c) AsRumodified (product 2) C. krusei ferricytochromes c, 8-22 to -38 ppm vs. DSS. -234- 0 .04 0.03 0.02 Q) u c ctS .0 '- 0 (/) .0 ctS 0.01 300 400 500 600 700 wavelength (nm) Figure A6.5. UV -visible difference spectrum: 8.74J..l.M AsRu-modified C. krusei cytochrome c (product 2) minus native C. krusei cytochrome c; absorption of -0.038 at 300nm indicates two histidine side chains have been modified. -235would fied not necessarily derivatives, but reveal such the a presence situation of a should mixture of singly-modibe evident from peptide mapping. No such evidence was found. The reaction of the C. krusei cytochrome ~ 12% hour (pH 7.5, 30 equivalents of ruthenium, pared (pH with 7, discussed that 50 the equivalents in His33 reactivity chapter is less of 3, of this same ruthenium, residue 24 hours, computer-modeling accessible cytochromes its reaction with a in the horse of His33 within one half modified) can be comin horse 10-30% these cytochrome c modified). As proteins cytochrome, and suggests in both bulky reagent such as aquopentaammine- ruthenium(II) is expected to be severely hindered. 2 provided the concentration of one component is minor; an equimolar mixture of singly-modified products might not appear to be modified at all unless quantitative comparisons with native spectra were made.