UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structure/Function in the CD site of parvalbumin : understanding calcium affinity using synthetic single… Franchini, Patrick Lorenzo Angelo 1999

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1999-463443.pdf [ 16.44MB ]
Metadata
JSON: 831-1.0089275.json
JSON-LD: 831-1.0089275-ld.json
RDF/XML (Pretty): 831-1.0089275-rdf.xml
RDF/JSON: 831-1.0089275-rdf.json
Turtle: 831-1.0089275-turtle.txt
N-Triples: 831-1.0089275-rdf-ntriples.txt
Original Record: 831-1.0089275-source.json
Full Text
831-1.0089275-fulltext.txt
Citation
831-1.0089275.ris

Full Text

STRUCTURE / FUNCTION IN THE CD SITE OF PARVALBUMIN: UNDERSTANDING CALCIUM AFFINITY USING SYNTHETIC SINGLE SITE EF-HAND PEPTIDES. by PATRICK LORENZO ANGELO FRANCHINI B.Sc. (Pharm.), The University of British Columbia, 1993 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES Division of Pharmaceutical Chemistry Faculty of Pharmaceutical Sciences We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A July 1999 © Patrick Lorenzo Angelo Franchini 1999 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y a v a i l a b l e for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It i s understood that copying or p u b l i c a t i o n of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed^without my written permission. Department o The University of B r i t i s h Columbia Vancouver, Canada ABSTRACT Abstract This study examined the influence of the flanking helices and non-chelating loop residues in the CD site of carp parvalbumin 4.25 (PCD) on calcium ion (Ca2+) affinity using synthetic single site EF-hand peptide chimeras based on the sequences of PCD and bovine brain calmodulin site III (Cam3). The peptides are observed to dimerize [Shaw, G.S., Hodges, R.S., and Sykes, B.D. (1990) Science 249, 280-283.] and a mathematical model was developed that described the Ca binding process taking into account dimerization. The model PCD site had 105-fold lower Ca 2 + affinity than the native site and did not bind magnesium ions. A Glu to Asp replacement in the -X position of the Ca 2 + binding loop in the PCD model site increased Ca 2 + affinity suggesting repulsion is a factor in the low Ca 2 + affinity of the PCD model site. The N-terminal PCD helix gave higher Ca 2 + affinity than the helix from Cam3 in both the monomer and dimer forms possibly through alterations in the N-terminal helix dipole. The C-terminal PCD helix gave higher Ca 2 + affinity in the monomer forms and increased monomer negative mean residue ellipticity. This higher Ca 2 + affinity may be due to enhanced PCD C-terminal helix structure. Both the Cam3 N-terminal helix and the PCD C-terminal helix promoted dimerization possibly through electrostatic interactions. The Gin in non-chelating loop position 2 from PCD does not alter dimerization or Ca 2 + binding to the monomer or dimer forms compared to Lys. The Lys found in PCD loop position 4 appears to have negative effects on Ca 2 + affinity in the monomer and dimer forms compared to Gly possibly through restrictive phi/psi angles or a decrease in the negative charge density in the loop. The Glu and Asp residues found in position 10 and 11 of the PCD loop ii Abstract promote C a 2 + affinity over Ala. This increase may be the result of increased negative charge density in the loop or increased C-terminal helix stability. Preliminary N M R studies support the contention that the PCD model site peptide is a dimer in solution. X-ray diffraction data from a PCD model site peptide crystal has been collected to 1.7 A resolution. i i i Table of Contents T A B L E OF C O N T E N T S A B S T R A C T i i T A B L E OF CONTENTS iv LIST OF T A B L E S xi LIST OF FIGURES .. xi i i LIST OF ABBREVIATIONS . xvi LIST OF A M I N O ACID CODES xvii PEPTIDE N O M E N C L A T U R E xviii A C K N O W L E D G E M E N T xix C H A P T E R 1 I N T R O D U C T I O N 1.1. BIOLOGICAL FUNCTION OF C A L C I U M 1 1.2. C a 2 + BINDING PROTEINS 2 1.3. E F - H A N D C a 2 + BINDING PROTEINS 5 1.3.1. Description of the EF-Hand Motif 5 1.3.2. Parvalbumin 11 1.3.3. Troponin C 15 1.3.4. Calmodulin 18 1.3.5. CalbindinD 9 K 21 1.3.6. Other EF-Hand Proteins 24 1.4. C a 2 + AFFINITY IN THE EF-HAND MOTIF , 25 iv Table of Contents 1.4.1. C a 2 + Binding Theory 25 1.4.2. Factors Influencing Ca Affinity 27 1.5. SINGLE SITE M O D E L EF-HAND MOTIF 33 1.5.1. Model Sites 33 1.5.2. Dimerization 34 1.6. C A R P P A R V A L B U M I N 4.25 CD SITE 36 1.6.1. S equence and Structure 36 1.6.2. Factors Influencing C a 2 + Affinity in the Paralbumin CD site 41 1.7. PROJECT B A C K G R O U N D 42 1.8. HYPOTHESIS 43 1.9. PROJECT OBJECTIVES 44 CHAPTER 2 USING CIRCULAR DICHROISM TO MEASURE DIMERIZATION AND Ca 2 + BINDING IN SYNTHETIC EF-HAND PEPTIDES 2.1. Overview 45 2.2. Mathematical Models 48 2.2.1. Circular Dichroism Theory 48 2.2.2. Monomer Model 51 2.2.3. Dimer Model - Dimerization 53 2.2.4. Dimer Model - C a 2 + Binding 54 2.3. Materials 58 2.4. Methods ;...60 v Table of Contents 2.4.1. tBOC Solid Phase Peptide Synthesis 60 2.4.2. Purification of PPPQKED 64 2.4.3. Dimerization Dissociation Constant of PPPQKED 67 2.4.4. C a 2 + Binding Measurement 68 2.4.5. Amino Acid Analysis 71 2.5. Results 71 2.5.1. Peptide Purification 71 2.5.2. Dimerization Study 72 2.5.3. Comparison of Monomer and Dimer C a 2 + Binding Models 73 2.6. Discussion 78 2.6.1. Comparison of Monomer and Dimer Models 78 2.6.2. Model Validity 81 2.6.3. Comparison of Different Dimer Models 83 CHAPTER 3 A -X GLUTAMATE TO ASPARTATE SUBSTITUTION IN THE CD SITE OF PARVALBUMIN: EFFECTS ON Ca2+ AFFINITY AND DIMERIZATION 3.1. Overview 87 3.2. Materials 87 3.3. Methods 88 3.3.1. tBOC Solid Phase Peptide Synthesis 88 3.3.2. Purification of PPPQKED-XD 88 3.3.3. Dimerization of PPPQKED-XD 88 vi Table of Contents 3.3.4. C a z + Binding Measurement 89 3.3.5. M g 2 + Sensitivity Studies 89 3.3.6. Tests of Significance 90 3.3.7. Molecular Models 91 3.4. Results 91 3.4.1. Peptide Purification 91 3.4.2. Dimerization and C a 2 + binding to PPPQKED-XD 91 3.4.3. Comparison of PPPQKED and PPPQKED-XD 94 3.4.4. M g 2 + study of PPPQKED and PPPQKED-XD 96 3.5. Discussion 97 3.5.1. The CD Site Model Peptide - PPPQKED 97 3.5.2. Effect of-X Glu to Asp Replacement 100 CHAPTER 4 N AND C TERMINAL a-HELIX REPLACEMENT IN THE PARVALBUMIN CD SITE MODEL PEPTIDE 4.1. Overview 108 4.2. Materials 108 4.3. Methods 109 4.3.1. Fmoc Peptide Synthesis 109 4.3.2. tBOC Peptide Synthesis 113 4.3.3. Peptide Purification 114 4.3.4. Dimerization 114 vii Table of Contents 4.3.5. C a 2 + Binding Measurement 115 4.3.6. Tests of Significance 115 4.3.7. Molecular Models 115 4.4. Results 116 4.4.1. Peptide Purification 116 4.4.2. Dimerization and C a 2 + Binding 116 4.4.3. N-terminal Helix Replacement 120 4.4.4. C-terminal Helix Replacement 122 4.5. Discussion 124 4.5.1. N-terminal PCD Helix Replacement - Dimerization 124 4.5.2. N-terminal Helix Replacement - C a 2 + Affinity 127 4.5.3. C-terminal PCD Helix Replacement - Dimerization 132 4.5.4. C-terminal PCD Helix Replacement - C a 2 + Affinity 134 CHAPTER 5 NON-CHELATING LOOP RESIDUE SUBSTITUTION IN THE CD SITE OF PARVALBUMIN 5.1. Overview 137 5.2. Materials 137 5.3. Methods 138 5.3.1. Synthesis 138 5.3.2. Purification 139 5.3.3. Dimerization ...139 viii Table of Contents 5.3.4. Ca 2 +Binding 140 5.3.5. Significance 140 5.3.6. Molecular Models 140 5.4. Results 141 5.4.1. Peptide Purification 141 5.4.2. C a 2 + Binding and Dimerization 141 5.4.3. Non-chelating Loop position 2 - Gin to Lys substitution 148 5.4.4. Non-chelating Loop Position 4 - Lys to Gly Substitution 150 5.4.5. Non-chelating Loop Position 10 - Glu to Ala Substitution 152 5.4.6. Non-chelating Loop position 11 - Asp to Ala substitution 154 5.5. Discussion 156 5.5.1. The Peptide PPPQKAA 156 5.5.2. Q2K Loop Substitution 156 5.5.3. K4G Loop Substitution 159 5.5.4. E10A Loop Substitution 161 5.5.5. D11A Loop Substitution 166 CHAPTER 6 NMR AND CRYSTALLIZATION STUDIES ON THE CARP PARVALBUMIN 4.25 CD SITE MODEL PEPTIDE 6.1. Overview 168 6.2. NMR Theory 168 6.3. X-ray Crystallography Theory 170 ix Table of Contents 6.4. Materials 172 6.5. Methods 173 6.5.1. N M R 173 6.5.2. X-ray Crystallography 174 6.6. Results 177 6.6.1. N M R 177 6.6.2. X-ray Crystallography 178 6.7. Discussion 181 6.7.1. N M R 181 6.7.2. X-ray Crystallography 182 CONCLUSIONS 184 FUTURE STUDIES 190 REFERENCES 192 BIBLIOGRAPHY 225 APPENDIX A 226 APPENDIX B 238 APPENDIX C 239 APPENDIX D 241 APPENDIX E 245 x List of Tables LIST OF TABLES Table I. Summary of 567 EF-Hand Loop Sequences 7 Table II. C a 2 + Dissociation Constants for Selected EF-Hand Proteins 28 Table III Hydrogen Bonds to Side-Chain Atoms in the CD site of Carp Parvalbumin 4.25 38 Table IV. Summary of the Sequences of 55 Identified Parvalbumin CD sites or CD Site Fragments 40 Table V HPLC Elution Gradient for Peptide Using Buffer System A / B 65 Table VI HPLC Elution Gradient for Peptide Using Buffer System C/D 66 Table VII. Fit Parameters for PPPQKED Fit to the Monomer Model 74 Table VIII. Fit Parameters for PPPQKED Fit to the Dimer Model 77 Table IX. Dimerization and C a 2 + Dissociation Parameters for PPPQKED and PPPQKED-XD 92 Table X . Comparison of C a 2 + Dissociation Constants for EF-hand Peptides and Proteins in which a - X Glu is Substituted with Asp 103 Table XI. Change on Free Energy (AAG) of C a 2 + Chelation on Glu to Asp Substitution in the - X Position in Different EF-hand Sites 104 Table XII. Dimerization and C a 2 + Dissociation Parameters for Peptides PPPQKED, PPCQKED, CPPQKED, and CPCQKED 119 Table XIII. Change in Free Energy on Dimerization and C a 2 + Binding in the Peptides PPPQKED, PPCQKED, CPPQKED and CPCQKED 120 xi List of Tables Table XIV. Dimerization and Ca Dissociation Parameters for Peptides PPPQKED, PPPQKAD, P P P K K A A , and PPPKGAD, PPPKGAA, PPPKGED and P P P K K E D 146 Table X V . Change in Free Energy on Dimerization and C a 2 + Binding in Peptides PPPQKED, PPPQKAD, P P P K K A A , PPPKGAA, PPPKGAD, PPPKGED, and P P P K K E D 148 Table X V I . Data collection parameters for PPPQKED crystal 179 Table XVII . Peptide Molecular Weights Determined by Mass Spectrometry 238 Table XVIII .Ca 2 + Titration Data Fit to the Single Site Model 239 Table X I X . C a 2 + Titration Data Sets Fit Individually to the Dimer Model 241 xii List of Figures LIST OF FIGURES Figure 1. The EF-hand 5 Figure 2. Schematic of EF-hand Loop 6 Figure 3. Primary sequences of EF-hand proteins 12 Figure 4. Structural model of carp parvalbumin 4.25 14 Figure 5. Structural model of chicken skeletal muscle troponin C 17 Figure 6. Structural model of bovine brain calmodulin 20 Figure 7. Structural model of bovine calbindin D9k 23 Figure 8. Structural model of troponin C site III dimer 35 Figure 9. CD Site of Parvalbumin 36 Figure 10. Proposed mechanism for C a 2 + binding and dimerization in an EF-hand peptide 46 Figure 11. Generation of circularly polarized light 48 Figure 12. Secondary structure CD spectra 50 Figure 13. Plot of molar ellipticity versus total peptide concentration of PPPQKED 72 Figure 14. C a 2 + titration plots of the synthetic EF-hand peptide PPPQKED using the monomer model 73 Figure 15. C a 2 + titration plots of the synthetic EF-hand peptide PPPQKED using the dimer model 75 Figure 16. Plot of total ellipticity BTOT versus free C a 2 + generated using the simultaneous fit dimer model 76 Figure 17. The effect of changing K i and K2 on shape of Ca titration plot 82 xii i List of Figures Figure 18. The effect of changing K i and K2 on shape of C a 2 + titration plot when B=C/2=D/2 84 Figure 19. Plot of molar ellipticity versus total peptide concentration and total titration 94-point ellipticity versus free Ca concentrations of PPPQKED-XD 93 Figure 20. - X Glu to Asp substitution in PPPQKED 95 Figure 21. Addition of M g 2 + to PPPQKED and PPPQKED-XD 96 Figure 22. Dimer interaction in PPPQKED and PPPQKED-XD 101 Figure 23. 12 residue loop sequences of the site pairs PPPQKED / PPPQKED-XD, Cam3:(DSE)/Cam3:(DSD), rOM D59E / rOM and GBP Q142E / GBP Q142D... 105 Figure 24. Stick models of the C a 2 + chelating loop of PCD and TnC3 107 Figure 25. Plots of molar ellipticity versus total peptide concentration for PPCQKED, CPPQKED and CPCQKED 117 Figure 26. Plot of total titration point ellipticity versus free C a 2 + concentrations for PPCQKED, CPPQKED and CPCQKED 118 Figure 27. Bar charts illustrating N-terminal PCD to Cam3 helix substitution 121 Figure 28. Bar charts illustrating C-terminal PCD to Cam3 helix substitution 123 Figure 29. Possible electrostatic interaction between Arg in position -3 and Glu in loop position 10 126 Figure 30. Dimer interaction change on N-terminal PCD to Cam3 helix replacment... 127 Figure 31. Ribbon structures highlighting potential interaction between residues in positions -1 and 16 in the peptides PPPQKED, CPPQKED, PPCQKED and C P C Q K E D 129 Figure 32. Dipole moment of N-terminal PCD and Cam3 Helices 131 xiv List of Figures Figure 33. Proposed electrostatic interaction between residues in positions -9 and 21.133 Figure 34. Average predicted helical propensity of the PCD and Cam3 C-terminal helices 135 Figure 35. Plots of molar ellipticity versus total peptide concentration for PPPQKAD, P P P K K A A , PPPKGAA, PPPKGAD, PPPKGED and PPPKKED 142 Figure 36. Plot of total titration point ellipticity versus free C a 2 + concentrations for PPPQKAD, P P P K K A A , PPPKGAA, PPPKGAD, PPPKGED, and P P P K K E D . . . 144 Figure 37. Q2K non-chelating loop substitution 149 Figure 38. K 4 G non-chelating loop substitution 151 Figure 39. E l OA non-chelating loop substitution 153 Figure 40. DI 1A non-chelating loop substitution 155 Figure 41. Effect of Q2K substitution on N-terminal helix dipole 158 Figure 42. Effect of E l OA substitution on C-terminal helix dipole 164 Figure 43. Relative position of non-chelating loop residues 2 and 4 and the N-terminus of the C-terminal helix 165 Figure 44. Proposed electrostatic interaction in PPPKGAD dimer 166 Figure 45. 500 M H z ' H N M R spectra of PPPQKED 177 Figure 46. Crystals of PPPQKED 178 Figure 47. Sample diffraction pattern from PPPQKED crystal 180 xv List of Abbreviations LIST OF A B B R E V I A T I O N S [0] Molar ellipticity (8) Mean residue ellipticity Cam3 Bovine brain calmodulin site 3 C D Circular dichroism C E Capillary electrophoresis Fmoc 9-fluorenylmethyloxycarbonyl H P L C High performance liquid chromatography KC1 Potassium chloride K D Dissociation constant K i Dissociation constant for monomer K2 Dissociation constant for dimer K D I M Calcium dependent dimerization dissociation constant M r Relative molecular weight M W Molecular weight N M R Nuclear magnetic resonance P C D Carp parvalbumin 4.25 CD site p i Isoelectric point tBOC / Boc tert-butyloxycarbonyl T F A Trifluoroacetic acid TnC3 Troponin C site 3 U V Ultraviolet xvi List of Amino Acid Codes LIST OF AMINO ACID CODES Amino Acid 3 Letter Code 1 Letter Code Alanine Ala A Arginine Arg R Asparagine Asn N Aspartic Acid Asp D Cysteine Cys C Glutamic Acid Glu E Glutamine Gin Q Glycine G l y G Histidine His H Isoleucine He I Leucine Leu L Lysine Lys K Methionine Met M Phenylalanine Phe F Proline Pro P Serine Ser S Threonine Thr T Tryptophan Trp W Tyrosine Tyr Y Valine Val V xvii Peptide Nomenclature PEPTID E N O M E N C L A T U R E Peptide Sequence N-Terminal Helix C-Terminal Helix Ca Binding Loop -12 -1 1 12 13 21 Native PCD Site S A D D V K K A F A I I D Q D K S G F I E E D E L K L F L Q N F K Native Cam3 Site SEEEIREAFRVF DKDGNGYISAAE LRHVMTNL6 PPPQKED PPPQKED-XD CPPQKED CPCQKED PPCQKED PPPKKED PPPKGED PPPKGAD PPPKGAA PPPKKAA PPPQKAA PPPQKAD S A D D V K K A F A I I S A D D V K K A F A I I SEEEIREAFRVF SEEEIREAFRVF S A D D V K K A F A I I S A D D V K K A F A I I S A D D V K K A F A I I S A D D V K K A F A I I S A D D V K K A F A I I S A D D V K K A F A I I S A D D V K K A F A I I S A D D V K K A F A I I D Q D K S G F I E E D E D Q D K S G F I E E D E D Q D K S G F I E E D E D Q D K S G F I E E D E D Q D K S G F I E E D E D Q D K S G F I E E D E G I S G F I E E D E G S G F I E H D E GJSGFIEAATE EjKjDKSGFIE A A E D Q D K S G F I E S A J E D Q D K S G F I E E D E L K L F L Q N F K L K L F L Q N F K L K L F L Q N F K LRHVLTNLG LRHVLTNLG KE KE KE L K L F L Q N F K L K L F L Q N F K L K L F L Q N F K L K L F L Q N F K L K L F L Q N F K L K L F L Q N F K L K L F L Q N F K Nomenclature Explained: I I C P P Q K E D Residues in Non-chelating loop positions 2, 4, 10 and 11 N-terminal Helix Loop C-terminal Helix C - Bovine Brain Calmodulin site III P - Carp Parvalbumin 4.25 CD site x v i n Acknowledgement ACKNOWLEDGEMENT I would like to thank my research supervisor Dr. Ron Reid for his guidance and support throughout my research. I would also like to thank my supervisory committee members, Dr. Frank Abbott, Dr. Gary Brayer, Dr. Keith McErlane, Dr. Wayne Riggs and Dr. John Sinclair for their constructive suggestions. I would also like to thank the colleagues I have worked with, Dr.Xiaochun Wu, Dr. Ric Procyshyn, Dr. Spyros Potamianos and Mr. Scott Loucks for their assistance and stimulating discussion. Final thanks go to the most important people in my life, my family. I would like to thank my wife Roberta for the patience and support she has shown over the last 6 years, my daughter Julianna for all the hugs and kisses as I wrote up my thesis, baby number 2 for giving me the incentive to finish and my parents for their support, encouragement and understanding throughout the years. Financial assistance was provided by a University Graduate Fellowship and a Medical Research Council of Canada Studentship. This research was funded by the Medical Research Council of Canada. E X LIBRIS P.L.A. FRANCHINI P.S. To any descendant that should chance upon this document I say "Hello". (August 13, 1999) xix CHAPTER 1 Chapter 1 - Introduction INTRODUCTION 1.1. BIOLOGICAL FUNCTION OF CALCIUM The role most commonly associated with calcium in biological systems is that involving biomineralization in teeth, bones and shells. In these structures calcium exists in the form of crystals of hydroxyapatite, calcite or aragonite (reviewed in Mann, 1988). In addition to this skeletal role, calcium ions (Ca ) play an important role in intracellular 94-regulation. Some regulatory roles of Ca include eukaryotic cell division (Lu and Means, 1993), contraction of cardiac and skeletal muscle (Ashley et al., 1991) and regulation of visual transduction in rod cells (Hurley, 1993). C a 2 + can act as an • 94-intracellular second messenger because in the resting cell the concentration of Ca is 104-fold lower than outside the cell (0.2 p M versus 2 mM) (McPhalen et al., 1991). C a 2 + regulation of intracellular processes begins with an extracellular stimulus that activates a surface receptor or channel protein. Some receptors involved in a Ca response include tyrosine-kinase-linked receptors, G-protein-linked seven-helix receptors and Ca channels (Falke et al., 1994). Activation of the surface receptor results in a release of Ca from intracellular stores via a second messenger or direct electrical contact. The intracellular stores are located in mitochondria, the endoplasmic reticulum or in muscle cells, the sarcoplasmic reticulm. The C a 2 + flux that results regulates C a 2 + binding proteins that control key points in physiological processes (reviewed in Berridge, 1990). 1 Chapter 1 - Introduction 1.2. Ca 2 + BINDING PROTEINS Proteins bind C a 2 + to stabilize secondary structure or regulate protein action (reviewed in McPhalen et al., 1991). Protein stabilization by C a 2 + occurs for the most part in extracellular proteins due to the relatively constant extracellular C a 2 + concentration. Proteins that use Ca to stabilize structure are largely enzymatic in nature. This includes the proteins P-trypsin (Bode and Schwager, 1975), elastase (Meyer et al., 1988), proteinase K (Betzel et al., 1988), thermitase (Gros et al., 1989), and D -galactose-binding protein (Vyas et al., 1987), among others. The highest C a 2 + affinity sites have been observed in these extracellular proteins (e.g. K D = 10"1 0 M for thermitase (Briedigkeit and Frommel, 1989)). At extracellular concentrations of C a 2 + (2 mM) proteins with such high C a 2 + affinity would be permanently in the C a 2 + bound state. In these proteins Ca appears to protect the enzymes against thermal, chaotropic or proteolytic degradation (Olafson and Smillie, 1975). Ca regulation of protein function occurs for the most part in intracellular C a 2 + binding proteins that share a C a 2 + binding motif termed the EF-hand or helix-loop-helix C a 2 + binding motif. This class of proteins includes troponin C, calmodulin, parvalbumin and calbindin Dok. In all Ca binding proteins the coordination of Ca is by oxygen atoms. This preference arises from the low polarizability of Ca resulting in ligand interactions dominated by ionic forces with little covalency. In such cases, ligands with low polarizability such as oxygen are preferred over polarizable atoms such as phosphorus 2+ and sulfur. In a survey of 182 ligands in 27 protein Ca binding sites it was found that 53 (29 %) are Asp, 11 (6 %) are Asn, 33 (18 %) are Glu, three (2 %) are Gin, 43 (23 %) are peptide carbonyl oxygens, 37 (20 %) are water molecules, one is a Ser and one is an 2 Chapter I - Introduction oxygen from a phosphate (McPhalen et al., 1991). The residue Thr has also been observed in protein Ca binding sites. The large number of Asp and Glu residues can be explained by the fact that in addition to Ca chelation, these residues provide charge balance and are able to participate in hydrogen bonding networks that can stabilize the sites. The Asn and Gin residues can also participate in hydrogen bonding networks via their side-chain amide nitrogen but do not provide charge balance to the Ca 2 + . The low number of Ser and Thr residues may be due to the lack of side-chain functional groups that can form hydrogen bonding networks. The high number of carbonyl ligands could be due to the permanent dipole of the carbonyl which could result in stronger C a 2 + interaction (McPhalen et al., 1991). C a 2 + coordination by oxygen containing ligands has no preferred coordination geometry. Instead, coordination is optimized by maximizing the number of ligand atoms interacting with the Ca . Studies on small-molecule Ca complexes reveal that the 9+ coordination number for Ca can vary from 4-12 (Brown, 1988). In proteins, the observed range of coordinating ligands is from five to eight with an average of seven (McPhalen et al., 1991). The average Ca 2 +-ligand distance is around 2.4 A (Einspahr and Bugg, 1984). The restricted number of ligands in protein C a 2 + binding sites results in a chelating geometry that can be best described as octahedral or pentagonal bipyramidal. In proteins that display pentagonal bipyramidal geometry but have only 5 or 6 ligands it is assumed that the missing ligands are disordered water molecules. In C a 2 + binding sites many of the chelating ligands come from the same part of the polypeptide chain. In the EF-hand sites all the chelating residues come from a continuous 12 residue stretch of polypeptide chain. In pseudo-EF-hand sites, such as that 3 Chapter 1 - Introduction found in calbindin DQK site I, the chelating residues come from a 14 residue polypeptide segment. These stretches of polypeptide chain that chelate the C a 2 + are perfect examples of structures termed Q loops (Leszczynski and Rose, 1986). A Q. loop is a segment of polypeptide chain between six and 16 residues in length where the termini of the loop lie close together and there is no regular secondary structure (oc-helices or P-sheets). Some extracellular C a 2 + binding sites have continuous chelating loops such as P-trypsin, elastase, thermitase site 2, thermolysin sites 2 and 3, oc-lactalbumin and concanavalin A . However, these sites do not share the same loop structure as found in the EF-hand sites. Other extracellular C a 2 + binding sites can be described as semi-continuous including subtilisin carlsberg site 1, thermitase site 1, thermolysin site 1, and the C a 2 + binding site in D-galactose binding protein. The C a 2 + binding sites found in extracellular proteins generally have irregular secondary structure with the exception of the D-galactose binding protein site. This site contributes five of six ligands from the same part of the polypeptide chain (Vyas et al., 1987) and the first 9 residues of the loop region have a structure that is highly homologous to that observed in EF-hand sites. 4 Chapter 1 - Introduction 1.3. EF-HAND Ca 2 + BINDING PROTEINS 1.3.1. Description of the EF-Hand Motif The EF-hand C a 2 + binding motif was first identified in carp parvalbumin 4.25 by Kretsinger and Knockolds (1973). The motif consists of an 8-10 residue a-helix, a 12 residue loop region that binds the metal ion, followed by another 8-10 residue a-helix. The term EF-hand is derived from the fact that one of the C a 2 + binding sites in parvalbumin is flanked by the E and F helices. This site resembles a hand with the index finger and thumb representing the E and F helix and the curled fingers the C a 2 + binding loop (Figure 1). The sites are found in pairs with a 2-fold rotation axis about the loops. Figure 1. The EF-hand. Schematic showing the resemblance of the backbone of the EF Ca 2 + binding site of carp parvalbumin 4.25 (right) to a folded hand (left). The index finger and thumb represent the E and F helices respectively, and the loop is represented by the folded fingers. Adapted from (Kretsinger, 1980). 5 Chapter 1 - Introduction The two sites form a globular domain with a small three residue anti-parallel (3-sheet between the loops of the paired sites. The loop region of an EF-hand binds the C a 2 + via the side chains of residues in positions 1, 3, 5, 9 and 12 of the loop and via the backbone carbonyl oxygen of the residue in position 7 (Figure 2). The residue in position 12 is predominantly Glu and chelates via both carboxylate oxygens. The seven coordinating atoms form a pentagonal bipyramidal arrangement around the Ca ion with three axes of chelation designated X , Y and Z. The positions of the chelating residues are thus identified by their axis of +z Figure 2. Schematic of EF-hand Loop. Schematic representation of EF-hand Ca 2 + binding loop showing the three axes of chelation and the pentagonal bipyramidal coordination geometry. The chelating residues are in loop positions 1, 3, 5, 7, 9 and 12 and are designated +X, +Y, +Z, -Y, -X and - Z respectively. 6 Chapter 1 - Introduction coordination and are given the designations +X, - X , +Y, - Y , +Z and -Z. The positional frequency of the different amino acids in the 12 residue loop region of EF-hand sites has been examined by Marsden et al. (1990) and Falke et al. (1994). The results from Falke et al. (1994) are summarized in Table I. Table I. Summary of 567 EF-Hand Loop Sequences . Coordinate: +X +Y +Z -Y -X -Z Loop Position: 1 2 3 4 5 6 7 8 9 10 11 12 Preferred: D K D G D G T I D F E E Preferred % : 100 29 76 56 52 96 23 68 32 23 29 92 Observed: D567 K163 D432 G319 D295 G541 T130 1384 D181 F131 E164 E523 A 6 7 N130 K 6 9 S131 D9 F90 V 9 4 SI 16 Y 6 4 D108 Q54 S5 R48 N123 N 8 K 7 0 L 7 4 T79 A 5 9 K 7 0 T54 N47 T9 K 4 Q54 M i l E65 T53 A 5 4 V 4 6 Q22 G8 R2 Y51 C 4 N 5 7 L 4 4 P47 139 A 1 5 E l H2 E36 G56 V 4 3 N 3 5 S36 H13 R27 Q9 E37 Q24 E32 S l l S26 C4 K 3 5 S20 R30 D7 115 S19 R7 L16 E7 C13 P17 G 7 F9 T5 D l l 114 T7 M 8 M 3 L l l R13 Y 7 Y 5 C l V l l G10 L 5 N 3 A 8 W 7 V 5 C3 H6 N 5 H4 D I M 5 Q5 M 3 G l N 3 M 5 D3 C 2 H I *The amino acid distribution o f the loop residues are summarized with the consensus sequence highlighted in bold. The table is taken from (Falke et al., 1994). 7 Chapter I - Introduction A description of the characteristics of each position in the EF-hand loop is summarized as follows: Position 1: The residue in loop position 1 is an invariant Asp and directly chelates the Ca 2 + with one of its carboxylate oxygens and with the other forms a strong hydrogen bond with the main-chain NH of the Gly in position 6. The side-chain oxygen that chelates the Ca 2 + is also involved in a hydrogen bond with the main-chain NH of position 4. This residue has (b,\f/ angles characteristic of a y-turn terminating the incoming helix. The extensive hydrogen bonding of this residue explains the high conservation. Position 2: This residue is not highly conserved but is a basic residue (Lys or Arg) approximately 50% of the time. Position 3: This residue directly chelates the Ca 2 + ion and is most commonly an Asp though Asn and Ser are also observed in this position. Position 4: This position is most commonly occupied by a Gly though 13 different amino acids are observed in this position. The main chain NH forms a bifurcated hydrogen bond with the carbonyl oxygen of the Asp in position 1 (a Type I (3-turn) and the side chain carboxylate oxygen that chelates the Ca ion. Position 5: This residue directly chelates the Ca ion and is most commonly an Asp, Asn or Ser. If this residue is Ser and if the residue in position 9 is Glu a strong hydrogen bond is observed between the side chains. Position 6: The Gly in position 6 is highly conserved and allows the loop to make a 90° turn orienting the other ligands into chelating position. The main chain NH 8 Chapter 1 - Introduction forms a strong hydrogen bond with the non-coordinating carboxylate oxygen of the Asp in position 1. P o s i t i o n 7: This residue chelates the Ca via its backbone carbonyl oxygen, and is most commonly Thr, although 16 different amino acids have been observed. This residue is also involved in the initiation of a small P-strand that extends for three residues between the paired sites. P o s i t i o n 8: The residue in position 8 is the central residue in the small P-sheet and is almost always hydrophobic with He, Leu and Val the most commonly observed. This residue is buried in the paired EF-hand domain for loop stabilization and shows anti-parallel backbone hydrogen bonding with the same residue in the other EF-hand in the paired domain. P o s i t i o n 9: The residue in position 9 appears to have several roles. When it is a Glu it directly chelates the C a 2 + (Kretsinger and Nockolds, 1973). However, when 2+ it is a Ser, Thr, Asp or Asn it cannot chelate the Ca and the ligand in this position is a water molecule. This residue initiates the exiting helix and can form two stabilizing hydrogen bonds; from the residue side chain to the main chain N H of the invariant Glu in position 12, and from the residue main chain N H to the side chain of the Glu in position 12. This interaction between the residues in loop positions 9 and 12 is termed an N-capping box and can stabilize the N-terminus of a helix (Serrano et al., 1992). P o s i t i o n 10: The residue in position 10 is the first of the exiting helix and is commonly aromatic though almost all residues have been observed in this position. P o s i t i o n 11 : This residue commonly has a negative charge and may electrostatically 9 Chapter 1 - Introduction stabilize the N-terminus of the exiting helix. Position 12: The residue in position 12 is almost always Glu and binds the C a 2 + via both side chain oxygens. The side chain carboxylate of this residue is also involved in hydrogen bonds with main-chain N H groups of residues in position 2, 3 and 9. The backbone of EF-hand site loop regions have high structural homology with deviations ranging from 0.28 A to 0.68 A rms (Strynadka and James, 1989). Within each EF-hand loop there are a conserved Type I P-turn and Asx turns. The Type I p-turn involves loop positions 1 through 4 with a hydrogen bond between the main-chain carbonyl of the residue in position 1 and the main-chain N H of the residue in position 4. The importance of the Asx turns in the C a 2 + binding site was first described by Herzberg and James (1985) on crystallographic analysis of turkey troponin C. An Asx turn involves the side chain of an Asp, Asn, Ser or Thr residue in the n position making a hydrogen bond with the main-chain N H of a residue in the n+2 position (Baker and Hubbard, 1984; Rees et a l , 1983; Richardson, 1981). The Asx turns in the EF-hand occur between side-chains of residues in loop positions 3 and 5 and the main-chain N H of the residues n+2 along the chain. These turns appear to position the side-chain oxygens of the residues in positions 3 and 5 for chelation of the Ca 2 + , stabilizing the negatively charged ligand oxygen atoms (McPhalen et al., 1991). Another important structural feature of the loop is the highly conserved Gly in position 6 which allows chain reversal. The residues in position 9 through 12 are part of the exiting helix and have a-helical hydrogen bonds between their main-chain carbonyls and the main-chain amide hydrogens n+4 residues along the chain. One observed exception is in the CD site of parvalbumin 10 Chapter I - Introduction which demonstrates a hydrogen bond between the main-chain carbonyl of the residue in position 9 and the main-chain amide hydrogen of the n+3 residue. This is a Type III p-turn characteristic of a 310 helix. The helices that flank the chelating loop are generally amphipathic in nature and display regular n to n+4 hydrogen bonding characteristic of a-helices. Both helices contain four conserved hydrophobic residues in positions -1, -4, -5 -8 of the N-terminal helix and positions 13, 16, 17, and 20 of the C-terminal helix (numbering relative to the 12 residue loop) (Falke et al., 1994). These residues are on the same side of the helices and are buried in the hydrophobic core of the 2 site domains or, in the case of the central helix of troponin C or calmodulin are involved in target recognition. The most extensively studied members of the EF-hand family of proteins include parvalbumin, troponin C, calmodulin and calbindin (representative sequences are summarized in Figure 3). Multiple three-dimensional structures have been determined for all four proteins using both crystallographic and N M R techniques. In addition, the 2_j_ Ca binding characteristics of many species variants of these proteins have been determined. Other proteins such as calcineurin (Kissinger et a l , 1995) and calpain (Lin et al., 1997), are observed to have EF-hand sites. However, the study of these sites is limited and they will be discussed only briefly. 1.3.2. Parvalbumin Parvalbumins are small acidic proteins ( M r 10000-12000 pi 3.9-5.5) that are found mainly in white and red skeletal muscle tissue though appreciable amounts have also been detected in brain tissue (Baron et al., 1975) (See Figure 3 for representative 11 Chapter I - Introduction 1 5 10 TnC Ac A S M T D Q Q A E A R A F L CaM A D Q L 1 15 20 25 30 35 40 45 50 TnC S E E M I A E F K A A F D M F D A D G G G D I S T K E L G T V M R M L G Q N P T CaM T E E Q I A E F K E A F S L F D K D G D G T I T T K E L G T V M R S L G Q N P T 5 10 15 20 25 30 35 40 55 60 65 70 75 80 85 90 TnC K E E L D A I I E E V D E D G S G T I D F E E F L V M M V R Q M K E D A CaM E A E L Q D M I N E V D A D G N G T I D F P E F L T M M A R K M K D T D 45 50 55 60 65 70 75 80 95 100 105 110 115 120 125 130 TnC K G K S E E E L A N C F R I F D K N A D G F I D I E E L G E I L R A T G E H V T CaM S E E E I R E A F R V F D K D G N G Y I S A A E L R H V M T N L G E K L T 85 90 95 100 105 110 M»3 117 135 140 145 150 155 160 TnC E E D I E D L M K D S D K N N D G R I D F D E F L K M M E G V Q CaM D E E V D E M I R E A D I D G D G Q V N Y E E F V Q M M T A K -120 125 130 135 140 145 1 5 10 15 20 25 30 35 Parv Ac A F A G V L N D A D I A A A L E A C K A A D S F D H K A F F A K V G L T S K S 40 45 50 55 60 65 70 Parv A D D V K K A F A I I D Q D K S G F I E E D E L K L F L Q N F K A D A 75 80 85 90 95 100 105 Parv R A L T D G E T K T F L K A G D S D G D G K I G V D E F T A L V K A 1 5 10 15 20 25 30 35 CalD K S P E E L K G I F E K Y A A K E G D P N Q L S K E E L K L L L Q T E F P S L 40 45 50 55 60 65 70 CalD L K G P S T L D E L F E E L D K N G D G E V S F E E F Q V L V K K I S Q Figure 3. Primary sequences of EF-hand proteins. The one letter amino acid sequences of chicken skeletal troponin C (Wilkinson, 1976) (TnC), bovine brain calmodulin (Watterson et al., 1980) (CaM), carp parvalbumin 4.25 (Coffee and Bradshaw, 1973) (Parv) and bovine calbindin D 9 k (Fullmer and Wasserman, 1977) (CalD). The sequences of CaM and TnC are aligned. Helical sequences are colored blue, Ca 2 + binding loops red and the chelating residues bold. 12 Chapter 1 - Introduction sequence). The name parvalbumin originates from the fact these proteins have similar solubility to albumins. There are two evolutionary distinct lineages of parvalbumins designated a and p. The a-lineage includes parvalbumins from rat and shark whereas the P-lineage includes carp and pike parvalbumin and the protein oncomodulin. The three dimensional structures of a number of species variants of parvalbumin have been determined by crystallographic techniques (Declercq et al., 1991; Kretsinger and Nockolds, 1973; Kumar et al., 1990; McPhalen et al., 1994; Roquet et al., 1992; 94-Swain et al., 1989). Parvalbumins contain three EF-hand Ca binding motifs (Kretsinger and Nockolds, 1973) designated the A B , CD and EF sites (Figure 4). The A B site is inactive with a truncated 10-residue loop whereas the CD and EF sites bind both C a 2 + and magnesium ions (Mg 2 +). The CD and EF sites are paired forming a hydrophobic patch into which a hydrophobic patch on the A B site is folded. 94- * • • * * * The parvalbumins are high Ca affinity sites with dissociation constants in the range 10"8-10"9 M (reviewed in Kretsinger, 1980; Pauls et al., 1996a). The highest C a 2 + affinity is observed in carp parvalbumin 4.25 which demonstrates a K D for C a 2 + of 3.7 x 10"10 M and a K D for of M g 2 + of 1.1 x 10"5 M for both the CD and EF sites (Moeschler et 9 + 94-al., 1980). In addition to the two EF-hand Ca / Mg binding sites, there is evidence for a third low affinity non-EF-hand C a 2 + / M g 2 + binding site adjacent to the CD site in P lineage parvalbumins (Ahmed et al., 1990; Declercq et a l , 1988; McPhalen et a l , 1994). This third site is thought to stabilize the crystal forms of these proteins. 94-Parvalbumins are thought to act as Ca buffers, increasing the speed of relaxation in muscle fibers (Pechere et al., 1977) or preventing C a 2 + overload in neurons (Nitsch et al., 1989). The proposed mechanism for parvalbumin increasing the speed of relaxation 13 Chapter I - Introduction Figure 4. Structural model of carp parvalbumin 4.25. Backbone ribbon model of parvalbumin showing inactive AB site (grey), CD site (blue) and EF site (magenta). Model based on the Ca 2 + bound crystal structure at 1.5 A resolution (Kumar et al., 1990). 14 Chapter I - Introduction in muscle fibers has the protein in the Mg bound state in resting muscle. On the release of the C a 2 + to initiate muscle contraction, the C a 2 + binds to troponin C or calmodulin first because of slow M g 2 + off rates from parvalbumin. As the M g 2 + is released from 94-parvalbumin it takes up the Ca released from troponin C or calmodulin preventing re-initiation of contraction. The protein oncomodulin is classified by tertiary structure as a member of the (3-lineage of parvalbumins. It was first extracted from rat hepatoma (MacManus, 1979) and demonstrates great similarity in sequence (50%)(Berchtold, 1989) and three-dimensional structure (Ahmed et al., 1990; McPhalen et al., 1994) to rat parvalbumin. What is unique about oncomodulin is that it is the only member of the parvalbumin family with a C a 2 + specific (the CD site) and C a 2 + / M g 2 + mixed site (the EF site) (Cox et a l , 1990; Hapak et al., 1989; MacManus et al., 1984). The C a 2 + affinity of the CD site is 10 to 15 fold lower than the EF site (Cox et al., 1990; Hapak et al., 1989). 1.3.3. Troponin C Troponin C is the C a 2 + binding component of the tripartite muscle contraction regulating complex troponin found in both skeletal and cardiac muscle (Greaser and Gergely, 1973; Potter and Gergely, 1975).s The troponin complex also includes troponin I and troponin T. In muscle contraction, myosin in the thick filament binds to an actin containing thin filament forming a M g 2 + activated ATPase which hydrolyses ATP. This causes the filaments to slide past one another generating contractile force. The regulation of interaction between the thick and thin filament is regulated by the troponin / tropomyosin complex (Kay et al., 1987; Leavis and Gergely, 1984; Zot and Potter, 1987). 15 Chapter 1 - Introduction Troponin binds at regular intervals along the actin containing thin filament and prevents formation of the actomyosin complex in the absence of Ca 2 + . The C a 2 + released to initiate muscle contraction is bound by troponin C exposing a hydrophobic patch on the protein that alters the interaction with troponin I (Ingraham and Swenson, 1984; Wang and Cheung, 1985) resulting in a change of the disposition of the troponin / tropomyosin complex relative to actin. This results in either the removal of the steric hinderance to the approach of myosin heads (Wakabayashi et a l , 1975), or the release of the phosphate (Pi) from the actin / myosin/ADP:Pi complex (Chalovich et al., 1981) and initiates contraction. Troponin C is approximately 18,000 daltons in size, is acidic (pi 4.25) and is composed of a single polypeptide chain 159-162 residues in length (See Figure 3 for representative sequence). The structure of troponin C has been determined using crystallographic (Herzberg and James, 1988; Houdusse et al., 1997; Satyshur et al., 1988; Strynadka et al., 1997; Vassylyev et al., 1998) and N M R (Findlay et al., 1994; Gagne et al., 1997; Gagne et al., 1995; Sia et al., 1997; Slupsky and Sykes, 1995; Spyracopoulos et al., 1997) techniques and reveals a dumbbell shaped protein that contains four EF-hand sites arranged in pairs connected by a nine-turn central helix (Figure 5). In addition there is an N-terminal helix unique to troponin C that is not part of an EF-hand. In skeletal troponin C, the two EF-hand sites in the N-terminal domain are low C a 2 + binding affinity (KD = 3 X 10"6 M) (Potter and Johnson, 1982) and the C-terminal domain EF-hand sites demonstrate high C a 2 + affinity (KD = 5 X 10"8 M) and also bind M g 2 + (KD = 2 x 10"4 M) (Potter and Johnson, 1982). In cardiac troponin C there are only three active EF-hand sites with site I inactive due to multiple mutations including the elimination of 16 Chapter 1 - Introduction Figure 5. Structural model of chicken skeletal muscle troponin C. Backbone ribbon model of troponin C showing N-terminal helix (grey), site l(blue) site 2 (magenta), site 3 (light blue) and site 4 (yellow). Model based on the C a 2 + bound crystal structure at 1.78 A resolution (PDB 1TOP). 17 Chapter 1 - Introduction the conserved Asp in loop position 1 (Holroyde et al., 1980; Johnson et al., 1980; Leavis and Kraft, 1978; Potter et al., 1977). The C-terminal sites in cardiac troponin C are high affinity Ca 2 + / M g 2 + binding sites and the lone N-terminal site is a low affinity C a 2 + specific site. Troponin C exhibits two hydrophobic patches, one in each domain. The hydrophobic patch in the C-terminal domain is exposed whether or not C a 2 + or M g 2 + is bound to sites III and IV. In contrast, the hydrophobic patch in the N-terminal domain is only exposed on binding of Ca 2 + . The exposure of the N-terminal patch is due to a large change in interhelical angles in the sites which alters the tertiary structure of the protein (Gagne et al., 1995). A study of mutants of skeletal troponin C reveals that this structural transition is dependent on the ability of site I to bind C a 2 + (Gagne et al., 1997). 1.3.4. Calmodulin 9-4-Calmodulin is a small, acidic (MW 16.7 kDa) Ca binding protein found in all 9+ eukaryotic cells (Means et al., 1982). Calmodulin is involved in the Ca mediated activation of intracellular enzymes (reviewed in Klee et al., 1980) such as phosphodiesterase (Lin et al., 1974; Teo and Wang, 1973), myosin light chain kinase (Dabrowska et al., 1978; Hathaway and Adelstein, 1979), calcineurin (Stewart et al., 1982), erythrocyte Ca 2 +-ATPase (Hanahan et a l , 1978; Lynch and Cheung, 1979), brain adenylate cyclase (Brostrom et al., 1978; Westcott et al., 1979), phosphorylase kinase (Cohen et al., 1978) and nicotinamide dinucleotide kinase (Anderson and Cormier, 1978). When C a 2 + binds to calmodulin the protein undergoes a conformational change that 18 Chapter 1 - Introduction results in the exposure of hydrophobic surfaces which in turn interact with the target enzymes (Ikura et al., 1992; Meador et al., 1992). Calmodulin is structurally very similar to troponin C with a 51% identical amino acid sequence (McPhalen et al., 1991) (See Figure 3). The structure of calmodulin has been extensively studied by both crystallographic (Babu et al., 1988; Chattopadhyaya et al., 1992; Rao et al., 1993; Tabernero et al., 1997; Taylor et al., 1991; Wall et al., 1997) and N M R techniques (Finn et al., 1995; Kuboniwa et al., 1995; Zhang et al., 1995) and reveals, like troponin C, four EF-hand Ca binding motifs arranged in pairs connected by a long central helix giving the protein a "dumbbell" shape (Figure 6). The sites are designated I through IV with sites I and II located in the N-terminal domain and sites III and IV in the C-terminal domain. The central helix that connects the two globular domains is three residues shorter than that of troponin-C resulting in a different orientation of the domains (reviewed in Strynadka and James, 1989). The central helix of calmodulin has been shown to be flexible in the solution structure (Barbato et al., 1992; Ikura et al., 1991). This flexible helix folds on binding to target proteins resulting in interaction of hydrophobic surfaces on calmodulin with hydrophobic surfaces on the target proteins (Ikura et al., 1992; Meador et al., 1992; Meador et al., 1993). 9 + In the crystal structure all four sites are occupied by Ca (Babu et al., 1985). The N-terminal sites are low Ca affinity ( K D =10" M) and the C-terminal sites have ~10-fold higher C a 2 + affinity (K D = 10"6 M) (Linse et al., 1991a). Work on the whole protein (Kuboniwa et al., 1995; Zhang et al., 1995) and the carboxyl terminal domain (Finn et al., 1995) has shown that in the absence of C a 2 + the calmodulin structure is similar to that observed in the C a 2 + bound state. However, addition of C a 2 + results in the 19 Chapter I - Introduction Figure 6. Structural model of bovine brain calmodulin. Backbone ribbon model of calmodulin showing site l(blue) site 2 (magenta), site 3 (light blue) and site 4 (yellow). Model based on the Ca 2 + bound crystal structure at 2.2 A resolution (Babu et al., 1988). 20 Chapter I - Introduction re-orientation of the helices in each of the EF-hands and exposes a hydrophobic patch essential for protein target interaction (Finn et al., 1995; Zhang et al., 1995). The rearrangement of helices is thought responsible for the cooperative effects observed between paired sites (Zhang et al., 1995). C a 2 + binding studies on intact calmodulin and tryptic fragments comprising each domain reveal that isolated N and C terminal domains retain their C a 2 + binding characteristics (Linse et al., 1991a; Minowa and Yagi, 1984). However, there is some evidence of interdomain interactions (Kilhoffer et al., 1992; Pedigo and Shea, 1995; Shea et al., 1996). 1.3.5. Calbindin D<>K Calbindin DQK, also known as vitamin D dependent intestinal C a 2 + binding protein, is a member of the SI00 family of C a 2 + binding proteins that also includes the proteins calcyclin, S100B and SI00a (reviewed in Schafer and Heizmann, 1996). Calbindin DQK is thought to act as an intracellular C a 2 + buffer (Christakos et al., 1989) and aid in the transport of C a 2 + through the brush border membrane of the small intestine (Staun, 1991), through the placental trophoblast epithelium (Nikitenko et al., 1998), and through renal basolateral membranes (Bindels et a l , 1991). It is a small globular protein (9 kD) 75 residues in length with two Ca binding sites (For sequence see Figure 3). The structure of calbindin DOK has been determined using x-ray crystallography (Szebenyi and Moffat, 1986; Szebenyi et al., 1981), and N M R (Akke et al., 1992; Akke et al., 1995; Drakenberg et al., 1989; Johansson et al., 1993; Kordel et al., 1989; Kordel et al., 1993; Skelton et al., 1994; Skelton et a l , 1995; Skelton et a l , 1990b) and reveals four helices numbered I through IV connected by 3 loops, two of which bind Ca 2 + . Bovine calbindin DQK. displays 21 Chapter I - Introduction cis-trans isomerism due to the peptide bond between Gly-42 and Pro-43 resulting in multiple conformations for the folded protein in solution (trans (75%) and cis (25%)) (Chazin et al., 1989). The trans is in the majority and is shown in Figure 7. Calbindin Dak, like all SI00 proteins, has a C-terminal C a 2 + binding site that is a true EF-hand site and an N-terminal site that is a pseudo EF with 14 residues in the loop region compared to the usual 12. A pseudo EF-hand site chelates C a 2 + with four peptide carbonyls (residues 1, 4, 6 and 9) and one side-chain carboxyl oxygen (residue 14) (numbering relative to loop region). The - X position is occupied by a water molecule. The linker between the two sites is 10 residues in length, in contrast to paired sites in calmodulin, troponin-C and parvalbumin which have a 4-5 residue linker. This linker has hydrophobic residues that interact with a hydrophobic patch formed by helices I and IV. The surface of the protein has no hydrophobic patch that would allow interaction with - other proteins. The C a 2 + dissociation constants for calbindin DOK have been measured at 0.31 and 0.16 u M at physiological salt concentrations (Linse et al., 1991b). Binding of Ca results in little change in the overall structure of calbindin DOK (Skelton et al., 1990b) but there are subtle structural changes involving helices III and IV (Skelton et al., 1995) and there is a decrease in the dynamics of the protein (Skelton et al., 1990a). The binding of C a 2 + to calbindin DOK is cooperative with the first C a 2 + binding to the N-terminal site 94-(Wimberly et al., 1995). The structural changes on binding the first Ca are more pronounced then on binding the second C a 2 + (Akke et al., 1995). The cooperativity of Ca binding to calbindin Dak is reduced by high salt concentrations suggesting that cooperativity is dependent on electrostatic interactions (Kesvatera et al., 1994). 22 Chapter 1 - Introduction Figure 7. Structural model of bovine calbindin D9k. Backbone ribbon model of calbindin D 9 k showing pseudo-EF-hand site l(blue), 10 residue linker (grey) and EF-hand site 2 (magenta). Arrow indicates location of Gly42-Pro43 cis-trans isomerism. Model based on the NMR structure (Kordel et al., 1997). 23 Chapter 1 - Introduction Other SI00 proteins have been examined for their structural and C a 2 + binding properties including calcyclin, psoriasin (Brodersen et al., 1998), and SlOOB(pP) (Drohat et al., 1996). The protein calcyclin is 90 residues in length and the solution structure reveals that it exists as a homodimer (Potts et a l , 1995). The distribution of secondary structural elements is very similar to that seen in calbindin Dgk though the tertiary structure differs in the interhelical angles (Potts et al., 1995; Sastry et al., 1998). The two sites in calcyclin bind Ca with a K D of 0.32 m M (Pedrocchi et al., 1994). Psoriasin is a 11.4 kDa protein that was initially extracted from psoriatic keratinocytes (Madsen et al., 1991). Psoriasin lacks a critical chelating residue in the N-terminal pseudo EF-hand (Brodersen et al., 1998; Brodersen et al., 1999) and is thought to bind only one C a 2 + binding per monomer. However, equilibrium dialysis experiments suggests it may bind up to seven C a 2 + ions, with the first C a 2 + ion binding with a K D = 0.063-0.077 m M (Vorum et al., 1996). The structure of the protein S100B(pp) in the absence of C a 2 + reveals interhelical angles for helix III in C a 2 + binding site 2 that differ significantly from other EF-hand sites (Drohat et a l , 1996). Therefore, in order for the protein to adopt a Ca bound structure typical of SI00 proteins it has been proposed that a large conformational change would have to occur (Drohat et al., 1996). C a 2 + binding studies on S100B(pP) reveal a K D for C a 2 + of 0.02 - 0.5 m M (Baudier et al., 1986). 1.3.6. Other EF-Hand Proteins Other sites that have EF-hand hand motifs include the myosin light chains, calcineurin B, calbindin D28k, calpain, calretinin, a-actinin, recoverin, sorcin and reticulocalbin, among others (Ikura, 1996). Of these sites, only the structures of 24 > Chapter 1 - Introduction calcineurin B and calpain have been determined. Examination of the crystal structure of the regulatory subunit of human calcineurin (Calcineurin B) reveals four EF-hand C a 2 + binding sites and a structure that resembles calmodulin folded about the central helix (Balendiran et al., 1995; Kissinger et a l , 1995). A study of the C a 2 + binding characteristics of calcineurin B reveals one high affinity site ( K D = 0.01 uM) and three lower affinity sites ( KD = 15 uM) (Kakalis et al., 1995). The crystal structures of porcine and rat calpain have been determined and reveal five EF-hand motifs (Blanchard et al., 1997; Lin et al., 1997). The apparent dissociation constant of human erythrocyte calpain has been measured at 60 u M (Vilei et al., 1997) 1.4. Ca 2 + AFFINITY IN THE EF-HAND MOTIF 1.4.1. Ca 2 + Binding Theory 9+ Ca binding to a single EF-hand site is an equilibrium process that can be described by the equation: P + C a 2 + ± ; P C a (1) 2+ . . . . . 24" The affinity of a Ca binding site is commonly described using a Ca dissociation constant K D or C a 2 + association constant K a . These constants are defined by the following equation: K D = 1/Ka = [P][Ca 2 +] / [PCa] (2) Underlying the process of C a 2 + chelation by a protein is a change in the free energy of the system, with the unbound Ca and protein in a higher energy state than the Ca2 +-protein complex. The change in Gibbs free energy (AG 0) on C a 2 + chelation is related to the C a 2 + dissociation and association constants by the following equation: 25 Chapter 1 - Introduction A G 0 = RT l n K D = -RT lnK a (3) where R is the gas constant and T is the temperature in Kelvins. The change in free 94-energy on Ca chelation is related to changes in enthalpy (AH°), temperature (T) and entropy change (AS0) by the equation: AG° = AH°-TAS° • (4) 2_|_ In order to understand factors influencing Ca binding to proteins one must understand the factors that influence changes in the enthalpy and entropy of chelation. The greatest contribution to Ca affinity in the EF-hand arises from charge neutralization on chelation. Charge neutralization would result in a large increase in entropy (Linse and Forsen, 1995). This increase in entropy is thought to be due to the fact that two separated charges will restrict the motion of the surrounding water to a greater degree than complexed charges (Linse and Forsen, 1995).. In addition to electrostatics, the C a 2 + affinity of an EF-hand site is influenced by the chelate and macrocycle effects (described in Falke et al., 1994). The chelate effect is the increase in complex stability observed when ligands are coupled in a multidentate chelator. This effect is primarily entropic due to the fact that multidentate chelators lose less translational entropy on complex formation then free ligands (Schwarzenbach, 1952). The macrocycle effect describes the increase in complex stability observed on cyclization of a linear chelator (Cox and Schneider, 1992). Cyclization would sacrifice some free energy to organize the chelating ligands into a preformed site. In EF-hand proteins, stabilization by the whole protein structure would provide the macrocycle effect. The increase in stabilization from the macrocycle effect is due to enthalpic rather than entropic contributions as demonstrated for organic cryptand chelators (Buschmann, 1986; 26 Chapter 1 - Introduction Buschrnann, 1986a). Cyclization is thought to decrease the enthalpic cost of chelation by partially desolvating the chelating ligands and pre-straining the chelator (Buschrnann, 1986; Buschrnann, 1986a; Cox and Schneider, 1992). 1.4.2. Factors Influencing C a 2 + Affinity EF-hand Ca binding sites demonstrate a high degree of structural conservation in different proteins but display a wide range of C a 2 + affinities (Table II), (reviewed in Falke et al., 1994; Linse and Forsen, 1995). Both factors within the EF-hand (i.e. the chelating residues, non-chelating loop residues and residues in the flanking helices) and outside of it have been shown to influence C a 2 + affinity. As the chelating residues make direct contact with the C a 2 + ion a number of studies have examined the effect of these residues on EF-hand C a 2 + affinity. It has been demonstrated that the Asp in position 1 (+X) and the Glu at position 12 (-Z) of the C a 2 + binding loop are critical for C a 2 + binding (Babu et al., 1992; Beckingham, 1991; Drake et a l , 1997; Haiech et al., 1991; Maune et al., 1992; Negele et al., 1992; Putkey et a l , 1989). In contrast, studies on the chelating residues in the +Y, +Z, - Y , and - X positions demonstrate variable effects. For example, an Asp to Asn mutation in the +Y position in different sites of bovine calmodulin resulted 2_j_ in opposite effects on Ca affinity (Waltersson et al., 1993). In troponin C site II, this substitution had no effect on C a 2 + affinity (Babu et a l , 1992) however replacement with a Glu greatly reduced Ca affinity (Babu et al., 1992; Wu and Reid, 1997a). Replacement of the +Y residue with Ala also resulted in a significant loss of Ca affinity (Dotson and Putkey, 1993; Linse and Chazin, 1995). In the +Z position, Asp or Asn give higher C a 2 + affinity than Ser in single site calmodulin site 3 peptides (Procyshyn and Reid, 1994b). 27 Chapter I - Introduction Table II. C a 2 + Dissociation Constants for Selected EF-Hand Proteins. Protein Solvent Salt" / Buffer (mM) (mM pH) K D (U M) (sites) Reference Carp PV 4.25 Frog Parv 4.50 Pike Parv 4.2 Rabbit Parv Rat Parv Rat Onco Bovine Brain CaM Scallop testis CaM Wheat germ CaM Rabbit skeletal TnC Bovine cardiac TnC Pig Ca lD 9 k Bovine CalDpk 80/25 7.4 1 m M EGTA 150/50 7.55 -/50 8.1 150/25 7.55 150/10 7.5 150/25 7.4 100/10 7.55 100/20 7.0 100/20 7.0 300/25 1 m M EGTA 100/20 7.0 0.1 m M EGTA -12 7.5 100/2 7.5 0.00037 (2 sites) (Moeschler et al., 1980) 0.002 (2 sites) 1.45,0.0038 0.0067 (2 sites) 0.006 (CD site) 0.011 (EF site) 0.8 (CD) 0.045 (EF) 61, 7.3, 2.0, 1.9 16.7, 7.7, 2.8,2.9 7.7,6.1, 3.9, 4.0 10, 2.5 (Sites I, II) 0.1,0.025 (Sites III, V) 50 (Site II) 0.08 (Sites III, IV) 0.0125, 0.0056 0.31,0.16 (Haiech et al., 1979) (Permyakov et al., 1983) (Haiech etal., 1979) (Williams etal., 1986) (Henzl e ta l , 1998) (Haiech et al., 1981) (Minowa and Yagi, 1984) (Minowa and Yagi, 1984) (Wang and Cheung, 1985) (Potter et al., 1977) (Linse etal., 1987) (Linse etal., 1991b) Abbreviations for proteins are: TnC - troponin C, CaM - calmodulin, Parv - Parvalbumin, CalD 9 k calbindin D 9 k . Salt is either NaCl or KC1. 28 Chapter 1 - Introduction The - Y residue, as it chelates via the backbone carbonyl, demonstrates a high degree of variability (Falke et al., 1994; Marsden et al., 1990). Substitution of a Phe for Tyr in the - Y position in the CD site of rat oncomodulin had no effect on C a 2 + affinity (Palmisano et al., 1990). However a Glu to Asp mutation in this position in calbindin Dok site 2 led to a significant decrease in C a 2 + affinity (Linse et a l , 1994). The decrease in C a 2 + is thought to be due to the fact the side chain of the residue in the - Y position is involved in chelation in site 1. The - X residue has a dual role chelating the C a 2 + ion (directly or indirectly) and initiating the C-terminal helix. Because of this dual role, it has been observed that residues that can form an N-cap structure (See page 9 position 9 for description of N-cap structure) at the N-terminus of a helix (e.g. Asp, Asn and Ser) can 94- V increase Ca affinity up to three-fold (Trigo-Gonzalez et al., 1993). However, a Glu in the - X chelating position has been observed to increase Ca affinity modestly (Hapak et al., 1989) or decrease Ca affinity significantly (Drake et al., 1996; Procyshyn and Reid, 1994a) compared tp Asp. The variable effect of different residues in the +Y, +Z, - Y and - X chelating positions indicate that factors other than the chelating residues influence C a 2 + affinity in the EF-hand. Examination of the effect of the non-chelating loop residues (loop positions 2, 4, 6, 8, 10 and 11) indicate variable influence on C a 2 + affinity, much like that observed for the chelating residues. A study by Drake et al. on non-chelating residues in the C a 2 + binding site of E.Coli D-galactose binding protein suggests that, with the exception of the residue • 9+ in loop position 8, the non-chelating residues have little effect on Ca affinity (Drake et al., 1997). The importance of a hydrophobic residue in position 8 is supported by a study in which substitution of the Val in loop position 8 of Paramecium calmodulin site IV with 29 Chapter 1 - Introduction a Thr resulted in a 3-fold decrease in C a 2 + affinity (Han and Roberts, 1997). In contrast to the work by Drake et al., other studies have shown that non-chelating loop residue substitution can influence C a 2 + affinity. For example, replacement of the Gly in position 6 with an Ala in a 25 residue peptide corresponding to rabbitt troponin C site II resulted in a four-fold decrease in Ca affinity (Malik et al., 1987). Work on the CD site of rat oncomodulin demonstrated that a Gin to Lys substitution in position 2 decreased C a 2 + affinity, a Lys to Gin substitution in position 4 had no effect on C a 2 + affinity and a Gly to Glu substitution in position 10 increased C a 2 + affinity (Palmisano et al., 1990). In loop position 2 of troponin C site II, substitution of a Glu for an Ala had no effect on C a 2 + affinity (Babu et al., 1992). However, the non-chelating loop residues of troponin C site II give 2.5-fold higher Ca affinity that those from site III in a synthetic 34 residue peptide corresponding to troponin C site III (Shaw et al., 1991b). The influence of the flanking helices on C a 2 + affinity was first investigated with the synthesis and study of a single site EF-hand peptide corresponding to rabbitt troponin C site III (Reid et al., 1981). This study demonstrated that the loop region of an EF-hand has little intrinsic Ca affinity and addition of the C terminal helix does little to enhance affinity. It is only on addition of the N-terminal helix that there is a 1000-fold increase in 2_j_ Ca affinity. Another study on synthetic EF-hand peptides in which the loop regions of calmodulin sites I through IV were sandwiched between the helices of sites I or IV it revealed that site IV helices give higher Ca affinity then site I helices (Sharma et al., 1997). Work on troponin-C site III synthetic analogs has shown that removal of hydrophobic residues within the helices can alter the C a 2 + affinity of the site (Monera et al., 1992). In mutants of whole calmodulin replacement of the helices in the fourth site 30 Chapter I - Introduction with those from the fourth site in troponin C modestly decreased C a 2 + affinity (George et al., 1996). The Ca affinity of an EF-hand site is also influenced by factors outside of the site. When an EF-hand is extracted from its native environment the C a 2 + affinities of these isolated sites are 18 to 700 fold lower than the sites in the native proteins (Procyshyn and Reid, 1994b; Reid, 1987b). In addition, studies on chimeric proteins reveal that EF-hand sites engineered into other EF-hand proteins retain the ability to bind C a 2 + but have different C a 2 + affinities (Durussel et al., 1996; George et al., 1993; George et al., 1996; Matsuura et al., 1991; Persechini et al., 1996). Cooperativity between EF-hand sites in the paired domains has also been shown to influence C a 2 + affinity (Carlstrom and Chazin, 1993; Forsen et al., 1991; George et al., 1996; Golosinska et al., 1991; Linse et a l , 1994; Linse and Chazin, 1995; Linse et al., 1991a; Trigo-Gonzalez et al., 1993; Waltersson et al., 1993). Factors distant from the EF-hand such as surface charges (Linse et al., 1988) and hydrophobic residues (da Silva et al., 1993; Pearlstone et al., 1992) can also affect C a 2 + affinity. In calbindin D9K alteration of the hydrophobic core residues can alter C a 2 + binding by enhancing pre-formation of the C a 2 + binding site and by stabilizing the C a 2 + -bound state (Kragelund et al., 1998). In addition to factors within the EF-hand proteins, it has been demonstrated with calbindin that increasing the KC1 and protein concentration can decrease the C a 2 + affinity (Linse et al., 1991a; Linse et al., 1995). The loss of C a 2 + affinity is attributed to 94- 94-electrostatic screening. The presence of Mg can also alter Ca affinity through competition resulting in lower C a 2 + affinity (Drabikowski et al., 1982; Haiech et al., 1979; Moeschler et al., 1980; Ogawa, 1985). 31 Chapter 1 - Introduction Attempts have been made to develop predictive techniques for C a 2 + affinity in the EF-hand sites. Sekharadu and Sundaralingam developed a quantitative structure-activity relationship (QSAR) which predicts that the C a 2 + affinity of the EF-hand is influenced by the hydrophobicity of the P-sheet and the four helices of the paired domain and will increase with the number of acidic chelating residues (Sekharudu and Sundaralingam, 9 + 1988). The role of acidic chelating residues in predicting Ca affinity was further expanded in the "acid-pair hypothesis" (Reid and Hodges, 1980) which proposed that the Ca affinity of an EF-hand site is dictated not only by the number of acidic residues but their position on the axes of chelation. This was demonstrated in work on bovine brain calmodulin site III as a synthetic 33 residue EF-hand peptide (Procyshyn and Reid, 1994b) and in the whole calmodulin protein (Wu and Reid, 1997b) which showed that C a 2 + affinity is greater with four acidic chelating residues than three. In sites with four 9 + chelating residues, Ca affinity was maximal when they were paired on the X and Z axes of chelation. Subsequently it has been shown that increasing the number of acidic 94-chelating residues to five further increases Ca Z T affinity (Henzl et al., 1996). 32 Chapter I - Introduction 1.5. SINGLE SITE MODEL EF-HAND MOTIF 1.5.1. Model Sites The use of single site model peptides to study C a 2 + binding to the EF-hand motif began with work on analogs of rabbit skeletal troponin C site III by Reid et al. (1980). It was found that C a 2 + binding was not detectable in a peptide corresponding to the 12 residue loop. However, addition of the 9 residues corresponding to the C-terminal helix resulted in a C a 2 + induced change in circular dichroism spectra which allowed the C a 2 + dissociation constant for the peptide to be measured ( K D = 4348 pM) (Reid et al., 1980). A subsequent study by Reid and co-workers demonstrated that 13 additional residues at the N-terminus increased C a 2 + affinity 1000-fold ( K D - 4 uM) (Reid et al., 1981). Since the initial work on troponin C site III, other studies have used model peptides to examine C a 2 + binding to EF-hand motifs in troponin C (Gariepy et al., 1982; Malik et al., 1987; Monera et al., 1992; Shaw et al., 1991b), calmodulin (Procyshyn and Reid, 1994a; Procyshyn and Reid, 1994b; Reid, 1990; Reid and Procyshyn, 1995; Sharma et al., 1997), parvalbumin (Revett et a l , 1997) and calbindin9 k (Finn et a l , 1992; Tsuji and Kaiser, 1991). Subsequent to the initial work on troponin C site III by Reid et al., it has been observed that isolated 12 residue loops can bind trivalent lanthanum ions (Marsden et al., 1988) or trivalent terbium ions (Clark et al., 1993; Kanellis et al., 1983) and can undergo C a 2 + induced circular dichroism changes in a 98% trifluoroethanol solution (Borin et al., 1985; Marchiori et al., 1983). 33 Chapter 1 - Introduction 1.5.2. Dimerization The use of single site EF-hand peptides to study factors influencing C a 2 + affinity in the EF-hand in theory simplifies analysis of C a 2 + binding data by eliminating cooperative effects in the whole protein. However, synthetic EF-hand peptides have been shown to dimerize (Shaw et al., 1990) complicating interpretation and analysis of C a 2 + binding data. Dimerization of single EF-hand site peptides was first proposed on examination of the x-ray diffraction pattern of crystals of a 34 residue peptide corresponding to rabbit skeletal troponin C site III (TnC3) (Delbaere et al., 1989). N M R studies on the same EF-hand site demonstrated evidence of a P-sheet and interactions between hydrophobic residues consistent with dimerization (Shaw et a l , 1990). The solution structure of a 39 residue proteolytic fragment of rabbit skeletal troponin C site IV was determined by Kay et al. and revealed a dimer with a two-fold axis of symmetry similar to the C-terminal domain in the crystal structure of chicken troponin C (Kay et al., 1991). Subsequently the solution structures of TnC3 (Shaw et al., 1992) (Figure 8) and a heterodimer comprising troponin C sites III and IV (Shaw and Sykes, 1996) were determined. Recently it has been observed that the 12 residue loop region of calmodulin site III can dimerize in the presence of the trivalent lanthanum ion (Wojcik et al., 1997). In addition to EF-hand peptides from troponin C, EF-hands from hake parvalbumin (Revett et a l , 1997), calmodulin (Sharma et al., 1997) and calbindin Dak (Finn et al., 1992) appear to dimerize. Attempts to quantitate the dimerization of the EF-hand peptides have included techniques such as N M R (Shaw et al., 1991a), circular dichroism (Monera et al., 1992), sedimentation equilibrium (Revett et al., 1997), the van't Hoff plot (Wojcik et al., 1997), gel electrophoresis (Finn et al., 1992) and HPLC gel filtration (Sharma et al., 1997). 34 Chapter I - Introduction Figure 8. Structural model of troponin C site III dimer. a-carbon backbone model of a homodimer of a 34 residue synthetic peptide analog of chicken skeletal muscle troponin C site I I I . Helical residues are colored red, Ca 2 + ions green. (3-sheet H-bonds are indicated by dashed green lines. Model based on the NMR structure (Shaw et al., 1992). 35 Chapter 1 - Introduction 1.6. CARP PARVALBUMIN 4.25 CD SITE 1.6.1. Sequence and Structure In carp parvalbumin 4.25, the C D site stretches from position 39 to 70, has a 13 residue N-terminal helix, a 12 residue loop and a 12 residue C-terminal helix (Figure 9) (Kumar et al., 1990). The loop shares its first and last four residues with the N and C terminal helices, respectively. Ca 2 + Binding Loop 39 • • • • • • • • • • • • • • • • 7 0 Parvalbumin CD Site S A D D V K K A F A I I D Q D K S G F I E E D E L K L F L Q N F N-Terminal Helix C-Terminal Helix Figure 9. CD Site of Parvalbumin. Stereo stick model and sequence of the CD site of carp parvalbumin 4.25 residues 39 to 71 (cross-eyed). The backbone is highlighted with a blue ribbon. The model is based on the crystal structure at 1.5 A resolution (Kumar et al., 1990). 36 Chapter 1 - Introduction The hydrogen bonding observed within the CD site is summarized in Table III. Hydrogen bonds of note in the N-terminal helix include an N-capping hydrogen bond between the Ser in position 39 (n) and main chain nitrogen of Asp42 (n+3). This capping interaction would act to terminate the N-terminus of the helix. In addition, a salt bridge between the side chains of Asp42 and Lys45 is observed. Within the loop an extensive network of stabilizing hydrogen bonds are observed characteristic of an EF-hand type site. One hydrogen bond specific to the CD site of parvalbumin (PCD) is observed between the side chains of Glu59 and Ser55 (Kretsinger and Nockolds, 1973; Kumar et al., 1990). This hydrogen bond only occurs with a Glu in the - X chelating position and is thought to stabilize the C a 2 + bound form. Examination of the C-terminal helix of the CD site reveals a 124° bend at Leu65 (Strynadka and James, 1989). This bend allows enhanced hydrophobic interaction with the inactive A B site of parvalbumin. Because of the bend there is a disruption of hydrogen bonding between the carbonyl oxygen of Lys64 and the main chain N H of Leu67. A water molecule bridges this disruption in hydrogen bonding. What makes the PCD site unique among EF-hand sites is the presence of the highly conserved Glu in the - X chelating position that binds the C a 2 + ion directly (Declercq et al., 1991; Kretsinger and Nockolds, 1973; Kumar et al., 1990). Other sites with residues such as Asp or Ser in the - X chelating position bind C a 2 + via an intermediate water molecule. There is evidence that this - X Glu is responsible for the ability of this site to bind M g 2 + (MacManus et al., 1989). Examination of 55 identified PCD sites (Table IV) reveals areas of sequence conservation. In the N-terminal helix the residue that "caps" the N-terminus is Ser 96%. 37 Chapter 1 - Introduction Table III Hydrogen Bonds to Side-Chain Atoms in the CD site of Carp Parvalbumin 4.25*. Amino Acid Atom 1 / Donor Atom 2 / Acceptor1 Ser-39 42-OD2 42-N Asp-42 45-NZ 39-OG Lys-45 42-OD1 Asp-51 54-N, 55-N 56-N Asp-53 55-N Lys-54 48-0,51-0 1 Ser-55 57-0, 59-OE2 57-N Glu-59 55-OG Glu-62 53-N 59-N Hydrogen bonding data table taken from 1.5-A model (Kumar et al., 1990). The atoms involved in hydrogen bonds are identified by the code used in the Brookhaven Data Bank. For the «-th residue. n-O and H -N are the main chain carbonyl oxygen and amide nitrogen, respectively. The electronegative side-chain atoms are H - O D I and «-0D2 for aspartate, «-OEl and «-0E2 for glutamate, «-NZ for lysine and n-OG for serine. The atoms listed in this column are hydrogen bonded to the first electronegative atom (Asp, 0 D 1 ; Glu, OE1) or the only electronegative atom functioning as a hydrogen bond donor (Lys, NZ; Ser, OG) of the side chain of the residue in the left-hand column. The atoms listed in this column are hydrogen bonded to the second electronegative atom (Asp, OD2; Glu, 0E2) or the only electronegative atom functioning as a hydrogen bond acceptor (Ser, OG) in the side-chain of the residue in the left-hand column. 38 Chapter 1 - Introduction of the time which can participate in stabilizing hydrogen bonds, as observed in the crystal structure of carp parvalbumin 4.25 (see above, (Kumar et al., 1990)). In fact, this position is occupied 100% of the time by residues that can form N-cap side-chain hydrogen bonds. In N-terminal helix positions -10 and -11 (relative to loop) the residues are acidic (59% and 69%, respectively) and could participate in stabilizing salt bridges with residues -7 and -6 which are Lys 89% and 78% of the time, respectively. The residues in positions -1, -4, -5 and -8 are generally hydrophobic or non-polar in nature as observed in most EF-hand sites (Falke et al., 1994). In addition, the residue in position -2 is observed to be hydrophobic 78%) of the time. Within the loop there are the conserved Asp, Gly and Glu residues in positions 1, 6 and 12 common to all EF-hands. In addition, PCD has the following conserved residues; Asp in position 3 (100%), Ser in position 5 (98%), and Glu in positions 9 (92%) and 10 (88%). The residue in position 7, which coordinates via the backbone carbonyl, is aromatic (Phe or Tyr) 100% of the time. In positions 8 the residue is 100% hydrophobic and in position 11 the residue is 100% acidic. The residue in loop positions 2 is not highly conserved though position 4 is occupied by a basic residue 78% of the time. This is in contrast to most EF-hand sites which have a Gly in position 4. Examination of the C-terminal helix of the PCD site reveals a conservation of hydrophobic residues in positions 13, 16, 17 and 20 as observed in most EF-hand sites (Falke et al., 1994). In addition, PCD has an aromatic or hydrophobic residue (Phe or Leu) in position 15, 85% of the time. This residue is observed to be highly variable in other EF-hand sites (Falke et al., 1994). The reason for conservation of a hydrophobic residue in this position may be linked to the bend in the C-terminal helix. 39 Chapter I - Introduction Table IV. Summary of the Sequences of 5 5 Identified Parvalbumin CD sites or CD Site Fragments. 1 s t Helix -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 EF-Loop Ser Ala Asp Asp Val Lys Lys Val Phe His lie l ie/ 96% 67% 59% 69% 51% 89% 78% 44% 98% 24% 55% Leu 49% N 1 A 32 A 3 D 38 A 1 A 1 A 1 A 23 F 49 A 8 A 2 A 1 S 45 K 2 D 32 E 3 I 16 E 1 D 4 I 7 I 1 D 1 F 7 I 24 T 1 L 2 E 10 I 1 L 8 K 49 E 4 L 1 E 2 I 27 L 24 N 1 G 1 Q 10 M 2 N 1 K 43 V 24 F 1 M 2 P 5 N 3 T 1 V 28 Q 1 N 2 G 6 V 11 S 4 S 5 V 2 T 2 Q 1 H 12 T 2 K 5 N 1 Q 1 R 5 S 1 V 3 Y 4 1 2 3 4 5 6 7 8 9 10 11 12 +x +Y +Z -Y -X -Z Asp Gin Asp Lys Ser Gly Phe lie Glu Glu Asp Glu 98% 49% 100% 65% 98% 98% 81% 90% 92% 88% 65% 100% A 1 A 2 D 49 A 3 E 1 D 1 F 39 I 43 D 4 D 1 D 31 E 48 D 48 E 3 K 32 S 47 G 47 Y 9 L 4 E 44 E 42 E 17 G 1 Q 8 V 1 G 1 K 11 R 6 K 1 N 6 Q 1 Q 24 V 2 R 2 2 n a Helix 13 14 15 16 17 18 19 20 21 Preferred: Observed: Leu Lys Leu Phe Leu Gin Asn Phe Lys 100% 69% 58% 69% 100% 60% 45% 100% 56% L 4 8 C 2 F 13 F 33 L 48 I l A 1 F 47 E 4 E 1 K 1 I 9 K 18 C 1 K 9 G 8 L 28 M 1 Q 29 G 12 Q 2 K 33 N 1 V 5 K 3 R 1 Q 4 S 1 N 21 Y 4 R 3 S 2 V 4 "Positional numbering is relative to loop. The consensus sequence is given in bold with the amino acid distribution for each residue underneath. Sequences obtained using a non-redundant NCBI BLASTP search (Altschul et al., 1997) using the sequence of the CD site of carp parvalbumin 4.25 residues 39 to 71 as a template. 40 Chapter I - Introduction 1.6.2. Factors Influencing C a 2 + Affinity in the Parvalbumin C D site The Ca affinity of the CD and EF sites of parvalbumin range from a K D of 3.7xl0" 1 0 M in carp parvalbumin 4.25 (Moeschler et a l , 1980) to a K D of 4.2 xlO" 8 M for rat a-parvalbumin (Pauls et al., 1993). It has been demonstrated that both the EF site and the inactive A B site contribute positively to the C a 2 + affinity of the native parvalbumin CD site (Pauls et al., 1994; Permyakov et al., 1991). The contribution is through cooperativity between the inactive and two active sites in parvalbumin, the cooperative effect being more pronounced for the CD site than the EF site (Pauls et al., 1994). Factors influencing ion specificity and affinity within the CD site of rat parvalbumin have been examined using mutants of rat oncomodulin (Durussel et al., 1996; Hapak et al., 1989; MacManus et a l , 1989; Palmisano et a l , 1990; Pauls et al., 1996b; Trevino et al., 1991). Rat oncomodulin is used because it is part of the (3-parvalbumin family of proteins and has 50% identical sequence to rat parvalbumin (Pauls et al., 1996a). Despite the similarities, the CD sites have been observed to have up to a 70 fold difference in 9+ 9+ Ca affinity and have different Mg binding properties (Eberhard and Erne, 1994; Hapak et al., 1989). One residue of particular interest in the CD site of oncomodulin is the - X Asp. Replacement of the Asp in the - X chelating position of the CD site of oncomodulin with a Glu found in the CD site of parvalbumin results in a modest increase 9-1- 9-1- 9-t- 9-+-in Ca affinity and turns a Ca specific site in to a Ca / Mg site (Hapak et al., 1989). A similar effect is observed when the Gly residue C-terminal to the - X residue in oncomodulin is replaced with a Glu found in PCD (Palmisano et al., 1990). Substitution of Asn and Gin in non-chelating loop positions 2 and 4 of the rat oncomodulin CD site with Lys and Gly found in the CD site of rat parvalbumin resulted in a decrease in 41 Chapter 1 - Introduction affinity and no effect, respectively (Palmisano et al., 1990). Modifications to the hydrophobic core of parvalbumin have also been shown to affect the M g 2 + antagonism and possibly the selectivity of the CD site (Pauls et al., 1996b). A recent study on oncomodulin mutants in which the chelating residues of the CD-site were switched with those of the EF site resulted in a pseudo EF-site with lower C a 2 + affinity than the native EF-site (Henzl et al., 1998). This indicates that factors outside the chelating residues in the CD site can influence Ca affinity and could explain why the CD site of oncomodulin, with the chelating residues of the CD site of rat parvalbumin, has C a 2 + affinity two orders of magnitude lower than the native rat parvalbumin (Palmisano et al., 1990). 1.7. PROJECT BACKGROUND The CD site of carp parvalbumin 4.25 (PCD) shares the highest C a 2 + affinity observed in EF-hand proteins (Moeschler et al., 1980). Work on pike parvalbumin fragments (Permyakov et al., 1991) and mutants of rat oncomodulin have demonstrated that factors in the whole protein and within the site can influence Ca affinity. However, the nature of the factors influencing C a 2 + affinity in the site have yet to be fully characterized. In addition, it has been demonstrated that the chelating residues of the PCD site do not confer high affinity when placed in a 33 residue bovine brain calmodulin site 3 (Cam3) model peptide (Procyshyn and Reid, 1994a). This fact presented a model based on synthetic peptides that could be used to study factors within the parvalbumin CD site that contribute to C a 2 + affinity. It was hypothesized that peptide chimeras of PCD and Cam3 with identical chelating residues could be used to examine the effect on 42 Chapter 1 - Introduction 2+ Ca affinity of the helices and the non-chelating loop residues that differ between the sites. 2"b In order to examine Ca binding to these sites a suitable mathematical model is needed to describe the C a 2 + binding process. In the past a monomer model has been used to describe Ca binding to synthetic EF-hand sites (Monera et al., 1992; Procyshyn and Reid, 1994a; Procyshyn and Reid, 1994b; Prod'hom and Karplus, 1993; Reid, 1990; Reid et al., 1980; Reid et al., 1981; Reid and Procyshyn, 1995; Shaw et al., 1990; Shaw et al., 1991b; Tsuji and Kaiser, 1991). It was decided that this model was insufficient to describe the binding process in light of evidence of dimerization of the single site EF-hand peptides (section 1.5.2). A model had been developed for N M R monitored C a 2 + titration of the 34 residue peptide corresponding to TnC3 (Shaw et al., 1991a) and there was evidence that circular dichroism could be used to assess EF-hand dimerization (Monera et al., 1992). Ca dependent dimerization dissociation constants for proteins had previously been determined using circular dichroism (Azuma and Hamaguchi, 1976) but the technique had not been applied to EF-hand peptides. Therefore it was hypothesized that a model based on circular dichroism could be developed that would 2"i" 2+ allow determination of the Ca dependent dimerization dissociation constant and Ca dissociation constants for the monomer and dimer forms of the peptide, respectively. 1.8. HYPOTHESIS 2*F The high Ca affinity of the native parvalbumin CD site is in part due to the non-chelating loop residues or flanking a-helices. The role of the non-chelating loop residues or flanking a-helices can be examined using synthetic single site peptide chimeras of 43 Chapter 1 - Introduction PCD and Cam3. C a 2 + dependent dimerization and C a 2 + binding to the monomer and dimer forms of the peptides studied can be determined using circular dichroism based techniques. 1.9. PROJECT OBJECTIVES The first objective of this project was to develop a mathematical model that described C a 2 + binding to synthetic EF-hand peptides that took into account C a 2 + dependent dimerization using circular dichroism techniques. The second objective was to successfully synthesize and study the C a 2 + binding and C a 2 + dependent dimerization characteristics of an EF-hand peptide corresponding to the CD site of carp parvalbumin 4.25 residues 39-71. Synthetic chimeras of PCD and Cam3 were then to be constructed to examine the effect of the flanking helices and non-chelating loop residues in positions 2, 4, 10 and 11 of PCD on Ca affinity and Ca dependent dimerization. The third objective of this project was to determine the solution or crystal structure of the synthetic peptides in order to aid in the interpretation of the C a 2 + binding and dimerization results. The mathematical model developed to describe C a 2 + binding and C a 2 + dependent dimerization to EF-hand peptides is presented in chapter 2. The peptide used to test the model is the synthetic PCD model site. Chapter 3 examines the PCD model site and the role of the - X Glu in the PCD model site C a 2 + affinity. Chapters 4 and 5 examine the effect of the PCD flanking helices and non-chelating loop residues, respectively on C a 2 + affinity and dimerization. Finally, chapter 6 presents ongoing structural determination studies on the synthetic CD site of parvalbumin using N M R and x-ray crystallographic techniques. 44 Chapter 2 - Model CHAPTER 2 USING CIRCULAR DICHROISM TO MEASURE DIMERIZATION AND Ca 2 + BINDING IN SYNTHETIC EF-HAND PEPTIDES 2.1. Overview Assessing the Ca affinity of an EF-hand site involves the addition of a known amount of C a 2 + to a known quantity of protein and measurement of the amount of C a 2 + that binds. Measurement of bound Ca can be done directly, as in the case of equilibrium dialysis, or indirectly by measuring proton shifts in N M R spectra or changes in fluorescence or circular dichroism spectra. Once the amount of bound C a 2 + is known at various Ca concentrations, the data is fit to a mathematical model that describes the Ca binding process and generates a binding constant that is independent of protein concentration. The constants generally reported are association or dissociation constants. For synthetic EF-hands peptides the simplest model that would describe the process of C a 2 + binding would be the "monomer model". This model assumes one C a 2 + ion binds to 94-one peptide. The dissociation constant for this model would be the free Ca concentration at which half the peptides are in the C a 2 + bound state. This model would be valid except for the fact that EF-hand peptides dimerize (Shaw et al., 1990). If dimerization alters the binding or spectroscopic properties of the peptide then the single site model will not be a valid descriptor of Ca binding to the model peptide. The monomer model will not "fit" and one will observe a concentration dependence of the association or dissociation constant derived. This chapter presents a novel mathematical model that attempts to address the 2_|_ phenomenon of Ca dependent dimerization in synthetic EF-hand peptides and generate 4 5 Chapter 2 - Model Ca dissociation constants for the monomer and dimer forms of the peptides. The dimer model was developed because it has been shown that EF-hand peptides similar to the ones under study dimerize (Shaw et al., 1992). In addition, the peptides under study demonstrated an increase in negative molar ellipticity with increasing peptide concentration consistent with dimerization (results sections 2.5.2, 3.4.2, 4.4.2, and 5.4.2). The dimer model used to describe the C a 2 + binding process assumes that the monomer peptide binds C a 2 + and then associates with either another C a 2 + bound monomer or an unbound monomer. If the dimer is composed of a C a 2 + bound and unbound monomer, the unbound monomer can then bind another C a 2 + ion (Figure 10). The C a 2 + dissociation constants for the dimer model are independent of peptide concentration as in the monomer model. However, the extent of dimerization is dependent on peptide concentration. Figure 10. Proposed mechanism for Ca binding and dimerization in an EF-hand peptide. The peptide (P) binds one Ca 2 + to form PCa with a dissociation constant K|. The PCa can then bind an unbound monomer to form P 2Ca with a dimerization dissociation constant K D I M ' or can bind another PCa to form P 2 Ca 2 with a dimerization dissociation constant K D I M . P 2 Ca 2 can also form if P 2Ca binds another Ca 2 + with a dissociation constant K2. 46 Chapter 2 - Model The dimer model presented in this chapter is applied to circular dichroism monitored Ca titrations but could be applied to any spectroscopic technique. Circular dichroism can be used to monitor C a 2 + titrations because when C a 2 + is added to an EF-hand peptide solution an increase in negative ellipticity at 222 nm has been observed (Procyshyn and Reid, 1994a; Procyshyn and Reid, 1994b; Reid, 1987a; Reid, 1987b; Reid, 1990; Reid and Procyshyn, 1995). The increase in negative ellipticity at 222 nm is assumed to be due to a Ca induced change in the structure of the EF-hand. The structural change is most likely a folding of the helical residues in the EF-hand. The dimer model presented in this chapter has two experimental components. The first involves measurement of the change in molar ellipticity with increasing peptide concentration under C a 2 + saturating conditions and allows estimation of the C a 2 + dependent dimerization dissociation constant. This technique is similar to that used to derive a dimerization constant for immunoglobulin G described by Azuma and Hamaguchi (1976). The second experimental component involves measurement of the change in ellipticity of an EF-hand peptide solution with increasing amounts of Ca 2 + . The mathematical model used to analyze the EF-hand C a 2 + titration data uses the dimerization constant to derive Ca binding constants for the monomer and dimer form. The EF-hand peptide PPPQKED which comprises the CD site of carp parvalbumin 4.25 (residues 39-71) (Coffee and Bradshaw, 1973) is used to demonstrate the dimer model. The C a 2 + binding parameters determined over an 8-fold range in peptide concentration using the dimer model are compared to those determined with the monomer model. 47 Chapter 2 - Model 2.2. Mathematical Models 2.2.1. Circular Dichroism Theory Circular dichroism (CD) spectroscopy measures the difference in the absorption of right and left circularly polarized light by a substance. This differs from commonly used spectroscopic techniques such as U V spectroscopy which measure the absorption of isotropic light. Circularly polarized light is generated from two electromagnetic waves that are perpendicular and one quarter out of phase (Figure 11) . The electromagnetic radiation in circularly polarized light, like all forms of electromagnetic radiation, can be thought of as a combination of two waves perpendicular to each other, one electric (E) and the other magnetic (M). In circularly polarized light the vector sum of the E-components (and M-components) rotates in a helical path. Therefore the magnitude of the electric component remains constant but the direction oscillates. This differs from linearly polarized light in which the magnitude oscillates and the direction remains constant. Figure 11. Generation of circularly polarized light. The E-components of two electromagnetic waves are one-quarter wavelength out of phase and are perpendicular. The vector sum of the E-components rotates in a helical path (dotted line). Taken from (Freifelder, 1976). 48 Chapter 2 - Model CD spectroscopy is used to analyze substances with asymmetric chromophores (uncommon) or symmetric chromophores in asymmetric environments that interact differently with right and left circularly polarized light. In such substances the extinction coefficients for right and left circularly polarized light will be different at certain wavelengths resulting in a measurable difference in the absorption of right and left circularly polarized light. This frequency dependent differential absorption of light polarized in two directions is termed dichroism. This measured difference in absorption (AA) is measured by CD spectrometers and converted to ellipticity 9 by the equation, 6 = 32.98-(AA). (5) Ellipticity is the unit of measurement and is reported in degrees. The measured difference in absorption (AA) obeys the Beer-Lambert law and is equal to the difference in extinction coefficients ( S L - £ R , A S ) at a given wavelength multiplied by the molar concentration (c) and pathlength (1). As is related to molar ellipticity [0] by the equation [6] =32.98 ( As). The units of As are liter-moL'-cm"1 and the units of [9] are 2 1 deg-cm -dmol" . The measured ellipticity at a specific wavelength X can be expressed by the following equation. 9 , = [ 9 ] x - c - l (6) Proteins and peptides are suited to analysis using circular dichroism because each amino acid in the backbone (except Gly) has an asymmetric alpha-carbon. This creates an asymmetric environment in which chromophores such as the amide bond absorb circularly polarized light. The CD spectra in the near U V region (180-260 nm) is very sensitive to the secondary structure of proteins and can be used to assess the different 49 Chapter 2 - Model structural types present, i.e. a-helix, parallel and anti-parallel P-sheet, turn, p-turn and random coil (Figure 12). Figure 12. Secondary structure CD spectra. CD spectra for the a-helix (solid line), P-sheet (dots and dashes), P-turn (dotted line) and random coil (dashed line). Redrawn from (Brahms and Brahms, 1980). 50 Chapter 2 - Model A particularly powerful application of circular dichroism is in the monitoring of changes in secondary structure in proteins. Changes in the backbone structure can be observed between 178 and 230 nm. Above 230 nm one can observe changes in the CD due to environment changes of the aromatic side-chains. For proteins the CD measurements are commonly reported as mean residue ellipticity (Q)\. This is calculated from the ellipticity measurement Ox. or molar ellipticity [0]^ using equation 7. ©» T0L c - l -n r n r where X is the wavelength, c is the molar concentration, 1 is the pathlength in m and n r is the number of residues in the polypeptide chain. The units of mean residue ellipticity are deg-cm2-dmor' -residue"1. 2.2.2. Monomer Model In the monomer model the peptide exists in two forms, the free peptide [P] and the « 2+ . . . . . . . 2+ • Ca bound peptide [PCa] which exist in equilibrium with free Ca [Ca] and is described by the equation: K D = ™ ^ 1 (8) D [PCa] where K D is the C a 2 + dissociation constant. Since [PTOT] = [P] + [PCa], equation 8 can be rewritten as: K _([P T O T ]-[PCa])[Ca] D [PCa] Equation 9 can be rearranged to give equation 10. 51 Chapter 2 - Model gga= ™ -f oo) [PTOTI K D +[Ca] where / is the molar fraction of peptide bound to Ca 2 + . On titration with C a 2 + the measured ellipticity at a given wavelength X at any point i can be described by: e ^ V t P W + V f P C a ] , - / (11) where A and B are the molar ellipticity constants for P and PCa and / is the pathlength. On titration the total amount of peptide is constant i.e. PJOT = P + PCa. However, the volume (vol) changes. Therefore equation 11 can be rewritten as: 9i • vol, = A, • P T 0 T .l + (Bx-Axy PCa ( • / (12) where voli is the volume at point i and P TOT and PCai are the total moles of peptide and moles of C a 2 + bound peptide at point i , respectively. In the absence of C a 2 + the measured ellipticity Bo is expressed as; e0 = v ^ r - z (13) where volo is the initial volume. Likewise when saturated with C a 2 + the measured ellipticity 9F at the final volume ( V O I F ) is e F = ^ . Z i 2 L . / ( 1 4 ) volIT Using equations 12, 13 and 14, the following relationship between the ellipticity at point i and the ellipticities in the absence and saturated with C a 2 + can be derived. 0i -vol, -9 0 -vol 0 Ax • P T 0 T - l + (Bx - ^ ) - P C a f -l-Ax - P T 0 T •/ 9 F • vo/ F - 9 0 • vol0 (Bx -Ax)-PT0T • / Equation 15 can be simplified to: (15) 52 Chapter 2 - Model e.-vo / j-eo-vo/o pc a i 9 F • vo/F - 9 0 • vol0 P T 0 T which is equal to the fraction of peptide bound to C a 2 + if). Combining equations 16 and 10 results in the following equation. e,-vO/,-e 0-vO/Q [Ca] 0 F • vo/F - 0 0 • vol0 K D + [Ca] Equation 17 can be fit to a plot of the fractional change in volume corrected ellipticity versus free C a 2 + concentration to determine the C a 2 + dissociation constant (Kn) of the peptide under study. 2.2.3. Dimer Model - Dimerization Under C a 2 + saturating conditions the EF-hand peptide can exist as a C a 2 + bound monomer, PCa, or Ca bound dimer, P2Ca2, with the concentrations of P and P2Ca minimized. Thus the equilibrium can be expressed as follows: K-DIM 2 P C a ^ p 2 C a 2 (18) The dimerization dissociation constant (KDIM) is described by equation 19. D I M ~ [ P 2 C a 2 ] ( 1 9 ) The measured ellipticity 6, for a particular peptide concentration can be expressed as: 9, =5 l - [ P C a ] - / + D x - [ P 2 C a 2 ] - / (20) where B and D are the molar ellipticity constants for each species. The total peptide concentration under Ca saturating conditions can be expressed by the equation: [PTOT] = [PCa] + 2-[P2Ca2] (21) 53 Chapter 2 - Model Substituting equation 21 into equation 20 results in equation 22. 0, =BX- [PCa] • / +1 Dx • ( [ P T 0 T ] - [PCa]) • / (22) Equation 22 can be rewritten in terms of total measured molar ellipticity by dividing both sides by [PTOTH resulting in equation 23. ^ • [ P C a ] + ^ X - ( [ P T 0 T ] - [ P C a ] ) [eix = V l ( 2 3 ) L R TOT J Where [PTOT] is the total peptide concentration. The term [PCa] can be expressed in terms of [PJOT] as follows. Substituting equation 21 into equation 19 results in equation 24. _ 2 • [PCa] 2 D I M " [ P T 0 T ] - [ P C a ] ( 2 4 ) Solving for [PCa], equation 24 becomes; [PCa] = ^Knu^^l K D I M - [ P T O T ] K _|_ " D I M L^TOTJ _ " U I M (25) V t J which can be substituted into equation 23. Fitting equation 23 to a plot of [PTOT] VS molar ellipticity, will result in the determination of KDIM, B and D. 2.2.4. Dimer Model - C a 2 + Binding The technique to determine the Ca binding constants for both the monomer (Ki) and dimer (K2) forms of an EF-hand peptide is based on principles similar to those used to determine KDIM- The mathematical model assumes that four different forms are present during the Ca titration, P, PCa, P2Ca and P2Ca2, that have molar ellipticity constants termed A, B, C and D, respectively. Each form should additively contribute to the 54 Chapter 2 - Model measured ellipticity value at a given wavelength. The peptide species in the presence of 2_f_ Ca can be described by the following equilibrium; K i KDIM ' K 2 P + Ca ti PCa + P zz P 2 Ca + Ca tz P 2 Ca 2 (26) The equilibrium constants for the different forms, PCa, P 2 Ca and P 2 Ca 2 can be expressed by the following equations; 1 [PCa] K D I „ . = E £ i f f l ( 2 8 ) D I M [P2Ca] K [ P ^ a H C a ] [P 2Ca,] where K i and K 2 are the Ca dissociation constants for the monomer and dimer forms, respectively and KDIM ' is the dimerization constant for a Ca bound peptide and an unbound monomer. The total peptide concentration in the presence of Ca can be expressed as; [PTOT] = [P] + [PCa] + 2-[P2Ca] + 2-[P2Ca2] (30) The ellipticity measurement at wavelength X elicited by Ca for each data point i can be described in terms of the peptide species present and the molar ellipticity constants for each by the following equation: 6, = A •[?],-l + Bx -[PCa], -l + Cx -[P.Ca],-l + Dx - [P .CaJ , •/ (31) Substituting equations 27, 28 and 29 and the fact that Ax = (6o-volo) / (PTOT-0 into equation 31 results in equation 32. 55 e, = Chapter 2 - Model e, vol. [PL + t k m, [Ca], •/ + & [P],' [Ca], + _ [P],' .[Ca],' •/ 0 2 ) TOT 1^ K , - K 2 - K D I M where Go is the initial measured ellipticity, volo is the initial volume, PTOT is the total peptide in moles, [P]j is the free peptide concentration at a specific point i and [Ca]j is the free Ca concentration. [P]i can be expressed in terms of PTOT, vol] (volume of solution at point i) and [Ca]j through the following series of algebraic manipulations. Substitution of equation 30 into equation 29 results in equation 33. K ' = B 2[P 2 Ca][Ca] TOT vol - [P]-[PCa]-2-[P 2 Ca] Solving for [PiCa] results in equation 34. (33) 1 TOT [P2Ca] = ~ K 2 vol [P]-[PCa] K 2 +[Ca] Equation 34 can be substituted into equation 28 resulting in equation 35. D I M ~ [PCa].[P] TOT 1 vol - [P]-[PCa] 2 K 2 +[Ca] Solving for [PCa] results in equation 36. [PCa] = K D I M ' - K 2 ( ^ - [ P ] ) vol ( K D I M ' - K 2 + 2 - K 2 - [ P ] + 2.[P].[Ca] Substituting equation 36 into equation 27 results in equation 37. (34) (35) (36) 56 [P][Ca] i<r ' K DIM - J V 2 ( % - [ P ] ) vol Chapter 2 - Model (37) ( K D I M ' - K 2 + 2 - K 2 - [ P ] + 2-[P]-[Ca] Equation 37 can be solved for [P] resulting in equations 38 and 39. [P] = K DIM K 2 4 - [ C a ] - ( K 2 + [ C a ] ) ( K , + [ C a ] ) 2 + -8 - K , - [ C a ] - ^ L - ( K 2 + [ C a ] ) vol IN" 2 ' DIM • K , - [ C a ] (38) [P] = DIM I V 2 4 - [ C a ] - ( K 2 + [ C a ] ) (K, + [ C a ] ) 2 +• 8 . K l . [ C a ] . - ^ - ( K 2 + [ C a ] ) vol 1 V 2 I V DIM K , - [ C a ] (39) Of the two equations only equation 38 can have positive values. Since KDIM' is related to KDIM by equation 40 KDIM'= KDIM • K1/K2 (40) equation 39 can be rewritten as equation 41. [P] = KDIM ' K-i 4-[Ca]-(K 2 +[Ca]) (K,+[Ca]) 2+-8 - [ C a ] - ^ . ( K 2 + [ C a ] ) vol K K , - [ C a ] DIM (41) Equations 40 and 41 can be substituted into equation 32 to create a equation which describes the measured ellipticity of a EF-hand peptide on C a 2 + titration in terms of the variables vol and [Ca], and in terms of the constants B, C, D, K\, K 2 , KDIM and PTOT-57 Chapter 2 - Model 2.3. Materials tBOC Solid Phase Peptide Synthesis • Benzhydrylamine resin (BHA) ( 1 % crosslink), N-Boc-L-alanine, N-Boc-Ny-p-tosyl-L-arginine, N-Boc-L-aspartic acid-P-cyclohexyl ester, N-Boc-L-glutamic acid-y-benzyl ester, N-a-Boc-N-Tt-benzyloxymethyl-L-histidine, N -Boc-L-isoleucine hemihydrate, N-Boc-L-leucine, N-Boc-N-e-(2-chloro-CBZ)-L-lysine, N-Boc-L-phenylalanine, N-Boc-O-benzyl-L-threonine, and N-Boc-L-valine were obtained from Bachem, Torrance, CA, U.S.A. • Tetraethylammonium chloride hydrate was obtained from Aldrich Chemical Company, St. Louis, M O , USA. • Acetic anhydride, N,N-dimethylformamide (DMF) HPLC grade, ninhydrin, pyridine (distilled over ninhydrin prior to use), toluene, and phenol were obtained from B D H Chemicals Canada Limited, Toronto, ON, Canada. • Acetic acid (glacial), dichloromethane (DCM) (distilled over CaCOs), 2 -propanol (IPA) and ethyl ether were from Fisher Scientific Co., Ottawa, Canada. • Trifluoroacetic acid (TFA) was from Halocarbon Product Corporation, River Edge, NJ, USA. • N-Boc-glycine was from Peninsula Laboratories, San Carlos, CA, USA. • Benzotriazol-l-yl-oxy-tris-(dimethylamino) phosphonium hexafluorophos-phate (BOP) was from Richelieu Biotechnologies, Montreal, PQ, Canada. • Anisole, N,N-diiso-propylethylamine (DIEA) (distilled over NaH and then over ninhydrin prior to use), Na-t-Boc-N-a-xanthyl-L-glutamine, 1-58 Chapter 2 - Model hydroxybenzotriazole (HOBt), indole, piperidine, N-t-Boc-O-benzyl-L-serine were from the Sigma Chemical Co., St. Louis, MO, USA. Peptide Purification • Acetonitrile HPLC grade, formic acid, o-phosphoric acid and triethylamine H P L C grade were from Fisher Scientific Co., Ottawa, Canada. • Trifluoroacetic acid (TFA) was from Halocarbon Product Corporation, River Edge, NJ, USA. • Potassium phosphate dibasic was obtained from J.T Baker Chemicals, Phillipsburg, NJ, USA . Dimerization and Ca 2 + Titration Studies • d-Camphorsulfonic acid and murexide were obtained from Aldrich Chemical Company, St. Louis, MO, USA. • Calcium chloride dihydrate, potassium chloride, potassium hydroxide, were obtained from B D H Chemicals Canada Limited, Toronto, ON, Canada. • Ethylene glycol-bis-(p-Aminoethyl ether) N , N , N ' , N'-tetra-acetic acid (EGTA), and 3-[N-morpholino) propanesulfonic acid (MOPS) were from the Sigma Chemical Co., St. Louis, M O , USA. • Chelex (sodium form, 200-400 Mesh) was from Bio-Rad Laboratories, Hercules, CA, USA. 59 Chapter 2 - Model 2.4. Methods The mathematical models for dimerization and C a 2 + binding to an EF-hand peptide were tested on the peptide corresponding to the carp parvalbumin 4.25 CD site, PPPQKED. The synthesis and purification of this peptide is described below. 2.4.1. tBOC Solid Phase Peptide Synthesis The peptide PPPQKED which corresponds to carp parvalbumin 4.25 residues 39-71 (Coffee and Bradshaw, 1973) was synthesized using solid phase methodology and tBOC chemistry on a Vega 1000 peptide synthesizer (Vega Biotechnologies, Tucson, A Z , USA).The peptide is N-acetylated and C-amidated. The sequence from the N-terminus is: 1 33 PPPQKED S A D D V K K A F A I I D Q D K S G F I E E D E L K L F L Q N F K Solid phase peptide synthesis involves the sequential addition of a-amino protected amino acids to a growing peptide chain anchored via its carboxyl terminus to a solid resin support (reviewed in Barany and Merrifield, 1980; Bodanszky, 1984; Stewart and Young, 1984). Peptides are synthesized from the C to N terminus and the addition of an amino acid involves 3 basic steps; coupling of an a-amino / side-chain protected amino acid, washing away of any unreacted amino acid, and removal of the a-amino protecting group. This is repeated for each amino acid. Once the peptide has reached the desired length, side-chain protecting groups are removed and the peptide is cleaved from the resin support. A description of the individual steps involved in the synthesis of PPPQKED on the Vega 1000 peptide synthesizer are outlined below. 60 Chapter 2 - Model a) Selective Coupling of the first amino acid Approximately 3.0 g of B H A resin (1% crosslinked) was placed in a 50 mL glass reaction vessel and allowed to swell overnight in D C M . B H A resin produces a C-terminal amide group on cleavage. After swelling, the D C M was drained and replaced with 5% DIEA / D C M (v/v) to activate the resin. Following base treatment the resin was washed and selectively coupled with 0.3 mEq N-Boc-N-s-(2-chloro-CBZ)-L-lysine per gram of resin (373.5 g). Selective coupling was accomplished using the BOP/HOBt method (Hudson, 1988). The lysine and 0.3 mEq HOBt (137.7 mg) were placed in one 25 ml beaker and 0.3 mEq BOP (398.1 mg) was placed in another. Both were dissolved in 15 mL D M F . To the BOP solution, 0.54 mEq of DIEA per gram resin were added (282 uL). The contents of both beakers were added to the drained resin in the reaction vessel. The coupling reaction proceeded for 30 minutes with constant shaking. After coupling the reaction vessel was drained, washed three times with 80% IPA / D C M , followed by six washes with D C M . b) Acetylation of the resin. After the first amino acid was coupled, the remaining free amino groups on the B H A resin were acetylated to prevent coupling of subsequent amino acids to the B H A resin. To acetylate the resin, a mixture of sodium dried toluene, ninhydrin distilled pyridine and acetic anhydride in a 3:3:1 ratio was prepared (70 mL). To the resin, 35 mL of this solution was added and the mixture shaken for 5 minutes. The reaction vessel was drained and another 35 mL was added. Acetylation with constant shaking occurred for 60 minutes. After acetylation, the Lys-resin was washed six times with D C M , three 61 Chapter 2 - Model times with 80% IPA / D C M , once with 5% DIEA / D C M followed by a final six washes with D C M . c) Deprotection After coupling of the amino acid, the amino terminal protecting group tBOC was removed to allow coupling of the next amino acid. Deprotection consisted of washing the resin with a 50% TFA in D C M (v/v) solution for 1, 5 and 20 minutes with shaking. After deprotection the peptide-resin was neutralized by washing with 5% DIEA / D C M twice. Finally, the peptide was washed six times with D C M . d) Active ester coupling Following deprotection the growing peptide chain is ready for coupling of the next amino acid. Coupling of the amino acid is much like selective coupling of the first amino acid with the exception that the amount of amino acid and reagents used are 3 fold higher. Therefore 0.9 mEq per gram resin of the appropriate amino, HOBt (413.1 mg) and BOP (1.194 g) are weighed out. The amino acid and HOBt are placed in one beaker and the BOP in another. Both are dissolved in 15 mL DMF and to the BOP reagent 1.59 mEq per gram resin DIEA (830 uL) is added. Beyond this point the coupling procedure for this amino acid is identical to that for the first amino acid. The deprotection-coupling cycle was repeated until the peptide of desired length was created. The final step in synthesis was the acetylation of the peptide after final deprotection. Acetylation was carried out as described as for the B H A resin. 62 Chapter 2 - Model e) Monitoring of Solid Phase Peptide Synthesis Coupling and deprotection were monitored during peptide synthesis using a modified quantitative ninhydrin reaction which measures free amino groups (Sarin et al., 1981). The following solutions were prepared for the ninhydrin reaction: Solution 1: 20 g Phenol 5mLEthanol(100%) 1 mL l O m M K C N 50 mL Pyridine Solution 2: 5% Ninhydrin in Ethanol (w/v) Solution 3: 0.5 M tetraethyl ammonium Chloride in D C M (w/v) Approximately 10 mg of dried peptide-resin was collected after each coupling and deprotection. The resin was weighed and placed in a glass test tube and to it 200 uL of solution 1 and 50 uL of solution 2 were added. A blank with no resin was also prepared. The tubes were placed in an oven at 100 °C for 10 minutes. The reaction mixture was then filtered through a pasteur pipette containing glass wool to remove the resin and collected in a 2 mL volumetric flask. The resin was rinsed with 60% ethanol and i f the color was concentrated a small amount of solution 3 was added. The eluent was added to the volumetric flask and made up to 2 ml with 60% ethanol. The absorbance was measured at 570 nm against the blank to detect the presence of Ruhemann's purple which is produced in the reaction of ninhydrin with amino groups. The amount of free amino groups (mole/g) was calculated using the equation; Free N H 2 = (Abs57o)-(Volume)/ (Weight Resin)-(1.5 x 104) This technique allowed qualitative assessment of the coupling reaction and also the amount of growing peptide after deprotection. When used to estimate the amount of 63 Chapter 2 - Model growing peptide, a 25 mL volumetric flask was used instead of a 2 mL because of the high chromophore concentration. f) HF Cleavage After synthesis the peptide-resin was washed with D C M and dried under vacuum. Approximately one gram of peptide resin was placed in a teflon reaction vessel of a type 1 HF apparatus along with 1 mL anisole and a stir bar. The reaction vessel was attached to the type 1 HF apparatus and was cooled in a liquid nitrogen bath. HF gas was condensed in the reaction vessel to a volume of approximately 10 mL. The contents of the reaction vessel were then stirred for 45 minutes in an ice-water bath. The HF was then removed by vacuum through a calcium oxide trap. g) Peptide Extraction The cleaved peptide mixture was removed from the teflon reaction vessel and washed with ethyl ether (2x10 mL). The cleaved peptide was then extracted with 50% (v/v) acetic acid (5 x 10 mL). The resulting solution was reduced to dryness using a Savant AS290 Speedvac concentrator. 2.4.2. Purification of PPPQKED Purification of PPPQKED was done using reverse phase HPLC on two different Waters H P L C systems, (Milford, M A , USA). System 1 - Model 510 Dual pump - Model 481 lambda Max U V detector - Model 721 Programmable system controller - Model 730 Data module 64 Chapter 2 - Model System 2 - Model 626 Pump - Model 600s Controller - Model 486 Tunable absorbance detector Two buffer systems were used to purify the peptides. The first buffer system (TEAP) utilized buffers A and B as outlined below. A: 0.2% triethylamine phosphate, pH 6.5 B: 0.2% triethylamine phosphate, pH 6.5 : Acetonitrile (60:40) The column used for this buffer system was a Zorbax C8 or C18 column (9.4 x 250 mm) (Hewlett-Packard, Palo Alto, CA, USA). The peptide was dissolved in buffer A , filtered with a 0.45 u M disposable filter, and aliquots were injected onto the HPLC and run according to the linear gradient as outlined in Table V . Table V HPLC Elution Gradient for Peptide Using Buffer System A/B. Time (minutes) % Buffer A % Buffer B 0 70 30 30 20 80 35 20 80 36 70 30 The initial aliquot was 50 to 100 uL and the run was monitored at 210 nm to identify the main peptide peak. Subsequent injections were 0.5 mL to 3 mL and were monitored at 254 nm to decrease sensitivity. The fractions corresponding to the main peptide peak were collected and freeze dried using a modified Savant model AS290 concentrator or a Labconco model 77510 lyophilizer (Kansas City, M O , USA). The purification procedure was repeated until a single peak was observed. 65 Chapter 2 - Model The final purification was done using TFA based buffers on a Zorbax C8 Stablebond column (9.4 x 250 mm) (Hewlett-Packard, Palo Alto, CA, USA). The buffers C and D are described below. C: 0.1 % TFA in Nanopure Water D: 0.1% TFA in Acetonitrile The peptide was dissolved in a 90:10 mixture of buffers C and D, filtered with a 0.45 u M disposable filter, and a 50 to 100 uL aliquot injected into the H P L C and monitored at 210 nm to determine the retention time of the main peak. The linear gradient of buffers C and D is described in Table VI. Subsequent injections were 0.5 to 3.0 mL and were monitored at 254 nm. The fractions were collected and freeze dried using a Labconco lyophilizer. Table VI HPLC Elution Gradient for Peptide Using Buffer System C/D. Time (minutes) % Buffer C % Buffer I) 0 90 10 35 40 60 40 40 60 41 90 10 Peptide molecular weights were confirmed using a V G Quattro Quadrupole electrospray mass spectrometer, (Fisons, Altrincham, England). The peptides were dissolved in a 0.1% formic acid in 50% mefhanol/50% nanopure water and 5-50 pL were injected into the mass spectrometer using the autoinjector on a HP model 1090 liquid chromatograph, (Hewlett-Packard, Palo Alto, CA, USA) without a column. The flow 66 Chapter 2 - Model was 50 )j.L/min. Positive ions were monitored with a scan speed of 100 mass units/s over a mass range of 600 to 1600. Myoglobin (Sigma ml882) was used for mass calibration of the spectrometer. The purity of the peptide was assessed using a Beckman P A C E System 5000 capillary electrophoresis apparatus, (Fullerton, CA, USA). The running buffer for the electrophoresis was 25 m M K2HPO4, pH 2.0. A peptide sample was dissolved in the running buffer to make a peptide solution of approximately 1.0 mg/mL. The capillary used in the electrophoresis was 50 cm with a 40 cm distance to the detection window and an internal diameter of 75 um. The pressure injection of the peptide sample was for 4 seconds. Electrophoresis occurred for 40 minutes at 12 kV and was monitored at 200 nm. 2.4.3. Dimerization Dissociation Constant of PPPQKED Dimerization was assessed by dissolving 7.91 mg of lyophilized PPPQKED peptide in 500 uL of a buffer containing 200 m M CaCl 2 , 100 m M KC1 and 50 m M MOPS, pH 7.5. The peptide solution was prepared and used immediately. The peptide concentration was determined using amino acid analysis (See section 2.4.5). A serial dilution of this peptide solution was made using the same buffer. The 6222 of each peptide solution was measured using a quartz cell of path-length 1, or 0.1 mm with a Jasco J720 spectropolarimeter in the laboratory of Dr. Grant Mauk (Department of Biochemistry, UBC) and the [9]222 was calculated. The study was done at room temperature. The dimerization constant K D I M and the B and the D values were determined by fitting this data to equation 23 (page 54) using Sigmaplot® software from SPSS Inc, (Chigago, IL, 67 Chapter 2 - Model USA). To obtain a satisfactory curve fit the data was weighted to l/([6]222)2- The parameters were obtained by adjusting the value of B manually and allowing the curve fitting software to generate KDIM and D values. This process was repeated until the total of the standard weighted residuals for the fit were equal to zero. The errors reported for the parameters are the asymptotic standard errors generated from the fit +5% for peptide concentration error (or just ±5% for B). 2.4.4. Ca 2 + Binding Measurement Titrating a series of peptide solutions with C a 2 + and monitoring the change in ellipticity at 222 nm with a Jasco J720 spectropolarimeter, (Easton, M D , USA) assessed C a 2 + binding. Circular dichroism measurements were done in the laboratory of Dr. Grant Mauk (Department of Biochemistry, UBC). A quartz cell with a pathlength of 1.0 mm was used for the titration. Measurements were done at room temperature. A stock peptide solution was prepared by dissolving 4.14 mg of lyophilized PPPQKED peptide in 2 mL of a buffer (50 m M MOPS and 100 mM KC1, pH 7.5). The peptide solution was prepared and used immediately. The peptide concentration was determined using amino acid analysis. A l l solutions used were prepared with Chelex 100 treated NANOpure® water using Nalgene® laboratory equipment. The stock peptide solution was serially diluted with the buffer to make solutions with relative peptide concentrations of 2, 1, 0.5 and 0.25. Stock C a 2 + solutions used in the C a 2 + titration experiments were prepared using CaCl 2 -2H 2 0 and contained 50 mM MOPS, 100 m M KC1, pH 7.5. Calcium concentrations of the stock solutions used were determined by titration with 0.01 M E G T A against a standard 0.1 M Ca solution (Orion Research Incorporated, Boston, M A ) using murexide 68 Chapter 2 - Model as the primary indicator. The data fitting described below was done using Sigmaplot from SPSS Inc. The data obtained from the C a 2 + titration was fit to both the monomer and dimer models. For the monomer model, the fraction bound /was plotted against total C a 2 + and fit to equation 17 (page 53). This generated an initial value for KD . The K D generated was then used to calculate a set of free C a 2 + concentrations from the total C a 2 + concentrations and total moles of peptide using the Comics 2.0 computer program written by Robert Boyko (http://diadem.biochem.ualberta.ca/pence.html). The process was repeated until there was convergence or the r 2 of the fit was maximal. For the dimer model two methods of fitting the data were used. In the first method, the ellipticity data for each peptide concentration was analyzed separately. Each set of ellipticity data was plotted against the volume and total C a 2 + concentration and this was fit to equation 32 (page 56) with the KDIM, B and D values taken from the dimerization study. The PTOT constants for the four different solutions were replaced with peptide concentrations based on the concentration of the stock peptide solution (PSTOCK) i.e. PSTOCK, PSTOCK/2, PSTOCK/4 and PSTOCK/8. The fit of the data generated initial values for the constants PSTOCK, KI, K 2, and C. The C value was constrained to be less than D. The Ki, K 2 and PSTOCK constants generated were then used to generate free C a 2 + concentrations using the Comics computer program. The free Ca concentrations then 2_|_ replaced the total Ca concentration. The fitting process was repeated until there was no longer any change in the Ki and K 2 constants. If there was failure to converge, the C value was set to equal D which led to convergence. The values for A were calculated from the Go, volo and the peptide concentration derived from the PSTOCK value. The 69 Chapter 2 - Model standard errors reported for K i , K 2 and A were generated from the four data sets +5% for peptide concentration error. The second method fit all four data sets simultaneously to generate the K i , K 2 and PSTOCK constants. This method is based on the following facts: C a 2 + is added to each peptide solution in equal molar aliquots; the same volume changes occur for each peptide solution on titration; ellipticity values at each titration point are additive; and the PTOT for the peptide solutions are related by serial dilution. Therefore, the measured ellipticities at each titration point for the different peptide concentrations can be added together as expressed by the equation: QTOT = ©2.0 + 01.0 + 6o.5 + 00.25 (42) where each of the four ellipticities 02.o, 0i.o, 0o.5, ©0.25 are described by equation 32 (page 56). The expanded equation would describe 0TOT. at each point in terms of the constants K i , K 2 , KDIM, B, C, D, volo, the 0o for each data set, PSTOCK (the four peptide solutions are based on a serial dilution of PSTOCK and have concentrations PSTOCK, PSTOCK/2, PSTOCK/4 and PSTOCK/8) and the variables vol, [Ca]2.o, [Ca]i.o, [Ca]o.s, [Ca]0.25 (the free C a 2 + concentrations that would exist for a given total amount of C a 2 + for each peptide concentration). To fit the data, the KDIM, B, and D came from the dimerization study. In this equation, C was limited to being less than D. If convergence did not occur, C was made to equal D. The 0TOT was plotted against vol, [Ca]2.o, [Ca]i.o, [Ca]o.5, and [Ca]o.25 (six dimensional plot) using equation 42 to generate the K i , K 2 and PSTOCK constants. These values were used to generate new free C a 2 + concentrations for each point in the titration of the four different peptide concentrations and used to calculate new K i , K 2 and PSTOCK values. This iteration process continued until convergence occurred. The errors 70 Chapter 2 - Model reported for K i , K 2 and PSTOCK were the asymptotic errors generated from the curve fit +5% for peptide concentration error. 2.4.5. Amino Acid Analysis Amino acid analysis was performed by taking a 50 uL aliquot of a peptide solution, placing it in a glass tube (Corning, Cat. No. 9820) and reducing it to dryness under vacuum. To the dried sample 100 uL of 6 N constant boiling HC1 was added. The tube was placed in a glass vial with a teflon valve and repeatedly evacuated and flushed with pre-purified nitrogen gas. The vial was left in an evacuated state and hydrolyzed at 110 °C for 24 hours. After hydrolysis the sample was reduced to dryness. To the dried sample 200 uL of NANOpure water was added to dissolve the contents. The sample was analyzed by Dr. Krystyna Piotrowska at the Protein Service Lab at the University of British Columbia in triplicate on an Applied Biosystems Model 420A/H amino acid analyzer, (Foster City, CA, USA) using 1000 pmol Norleucine as the internal standard. The peptide concentration was calculated from the pmol amount of Ala recovered, the dilution factor and the volumes used in the amino acid analysis. 2.5. Results 2.5.1. Peptide Purification An H P L C chromatogram of unpurified PPPQKED and a capillary electrophoresis chromatogram of purified PPPQKED are given in Appendix A . The molecular weight determined by electrospray mass spectrometry is given in Appendix B. 71 Chapter 2 - Model 2.5.2. Dimerization Study The peptide PPPQKED demonstrated increasing negative molar ellipticity with increasing peptide concentration consistent with dimerization (Figure 13). CD spectra over the wavelength range 200-250 nm for each peptide concentration studied are shown in Appendix E. The peptide concentration for the stock solution was determined to be 14.12 ± 0.25 mg/mL from amino acid analysis. The fitting of the molar ellipticity and peptide concentration data to equation 23 (page 54) resulted in a K D i M = 56 ± 14 uM and molar ellipticity constants B and D of -1.88xl05 ± 0.09 xlO5 and -8.30xl05 ± 0.61xl05 9 1 9 deg-cm -dmol" , respectively. The fit had a r = 0.9987. o E xp CM* E o 6) <D b CD 1e-5 1e-4 1e-3 Total Peptide (M) 1e-2 Figure 13. Plot of molar ellipticity versus total peptide concentration of PPPQKED. Ellipticity measurements were made at 222 nm using a 1 or 0.1 mm quartz cell. Solid line indicates the best fit curve. 72 Chapter 2 - Model 2.5.3. Comparison of Monomer and Dimer Ca 2 + Binding Models The peptide concentration for the stock solution was determined to be 1.80 ± 0.05 mg/mL from amino acid analysis. The fit of the C a 2 + titration data to the monomer model using peptide concentrations based on amino acid analysis resulted in the apparent 2+ Ca dissociation constant KQ increasing with decreasing peptide concentration (Figure 14 and Table VII). The increase in the KD value is approximately 4 fold over the course of an 8-fold dilution, going from 72 uM to 264 pM. This increase in K D is also observed when the data is fit using the monomer model and peptide concentrations derived from the dimer model fit (Table VII). Examination of the r 2 values and coefficients of variation reveals that as the peptide concentration decreases there is a decrease in the quality of the fit and an increase in the percentage error of KD. 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 Free C a 2 + (M) Figure 14. Ca 2 + titration plots of the synthetic EF-hand peptide PPPQKED using the monomer model. Peptide concentrations are 1.80 mg/mL(»), 0.90 mg/mL ( • ) , 0.45 mg/mL (A) and 0.23 mg/mL (T) from amino acid analysis. 73 Chapter 2 - Model Table VII. Fit Parameters for PPPQKED Fit to the Monomer Model. Cone. (mg/mL)* l.SO 0.90 0.45 0.23 K D (uM) (Ca 2 +) 72 98 172 264 r 2 0.9993 0.9999 0.9989 0.9975 3.65 1.49 4.32 6.29 Cone. (mg/mL)** 2.01 1.06 0.56 0.30 K D ( u M ) ( C a 2 + ) 62 85 159 255 r 2 0.9991 0.9998 0.9992 0.9978 C V % f 4.27 2.02 3.76 5.86 * Concentrations based on amino acid analysis. ** Concentrations from dimer study in which data sets were fit individually. 1 Coefficients of variation in the K D value expressed as a percentage. The dimer model in which each of the four data sets were fit individually resulted in an average K , = 421 + 43 uM and K 2 = 47 ± 10 uM (Figure 15 and Table VIII). A l l fitted plots demonstrated r >0.9995. The C value generated for each data set converged with D due to the limitation in the curve fitting that C < D. The K i values display no concentration dependence but the K 2 values increase ~2 fold over the 8 fold dilution. The peptide concentration of the stock peptide solution (PSTOCK) derived from the fit of each data set increases approximately 5% after each dilution. The A values decrease approximately 5% after each dilution due to the fact they are calculated from the PSTOCK values. 74 Chapter 2 - Model 200 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 Free Calcium (M) Figure 15. C a 2 + titration plots of the synthetic EF-hand peptide PPPQKED using the dimer model. The titrations were carried out for four different peptide concentrations with relative peptide concentrations of 2 ( • ) , 1 ( • ) , 0.5 ( • ) and 0.25 ( T ) . The dimer model in which all the data sets were fit simultaneously resulted in a K i = 431 ± 33 u M and K 2 = 37 ± 6 u M (Figure 16 and Table VIII). The PSTOCK that best described all four data sets was 0.489 ± 0.025 umoles. The fit demonstrated a r of 75 Chapter 2 - Model 0.9999. The calculated A values demonstrate a 5% decrease in negative molar ellipticity over the course of the 8-fold dilution. 400 100 J-. , , , • 1 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 Free C a 2 + (M) Figure 16. Plot of total ellipticity 0 T O T versus free Ca + generated using the simultaneous fit dimer model for PPPQKED. Representative plot showing the 9 T O T fit (solid line) through each of the four data sets. Relative peptide concentrations are 2 ( • ) , 1 ( • ) , 0.5 ( A ) and 0.25 ( T ) . 76 Chapter 2 - Model Table VIII. Fit Parameters for PPPQKED Fit to the Dimer Model Parameter Relative Amount of Peptide Mean ± SE 2 1.0 0.5 0.25 Fit of C a 2 + Titration Data Sets** K , (uM) (Ca 2 +) 476 386 388 435 421 ± 43 K 2 (uM) (Ca 2 +) 33 40 52 63 47 ± 10 PSTOCK* Qunole) 0.470 0.496 0.522 0.557 0.511 ±0.045 A -1.51xl0 5 -1.40xl0 5 -1.33xl0 5 -1.21xl0 5 -1 .36x l0 5 ±0 .13x l0 5 e -8.301xl0 5 -8.301xl05 -8.301xl05 -8.301xl0 5 -8.30xl0 5 r 2 0.9998 0.9999 0.9996 0.9995 ~ Simultaneous Fit of Ca 2 + Titration Data Sets K i (uM) § (Ca 2 +) 431 431 ± 3 3 K 2 ( u M ) § (Ca 2 +) 37 37 ± 6 PSTOCK* (urnole)8 0.489 0.489 ± 0.025 A -1.45xl0 5 -1.42xl0 5 -1.42xl0 5 -1.38xl0 5 -1 .42x l0 5 ±0 .08x l0 5 r 2 0.9999 ~ Dimerization Parameters* B -1.88xl0 5 -1 .88x l0 5 ±0.09x l0 5 D -8.30xl0 5 -8 .30x l0 5 ±0.61x l0 5 KDIM (UM) 56 56 ± 14 * From dimerization study. Error is the asymptotic standard error from the curve fitting +5%. Error is the standard error from the average of the four values +5%. f In the curve fitting C was limited to being less than D. I PSTOCK is the total peptide in umoles that would be in 900 uL of undiluted stock solution. § Error is the asymptotic standard error from the curve fit +5%. 77 Chapter 2 - Model 2.6. Discussion 2.6.1. Comparison of Monomer and Dimer Models Of the two models used to examine C a 2 + binding to the peptide PPPQKED, the dimer model is the most consistent with the measured data. Fitting the C a 2 + titration data with the monomer model results in a 4-fold increase in the C a 2 + dissociation constants over the 8-fold dilution (Table VII). This increase in the KD is consistent with the concentration dependent peptide dimerization suggested by the molar ellipticity versus peptide concentration plot (Figure 13). When the C a 2 + binding data is fit to the dimer model using the KDIM, B and D constants generated from the molar ellipticity versus peptide concentration plot, two different binding constants are revealed, K i = 421 ± 43 u M and K 2 = 47 + 10 uM. Thus dimerization appears to increase the binding affinity of PPPQKED 10-fold. Analysis of the set of four K i and K 2 values reveals a constant K i value. However, the K 2 values increase two-fold over the 8-fold dilution (Table VIII). In addition, the PSTOCK values increase approximately 5% per dilution when they should be constant. The A values, that are calculated from the PSTOCK values, also decrease 5% per dilution. The deviations in the K 2 , PSTOCK and A values indicate that the dimer model does not perfectly describe the data. The root of the deviations observed in the K 2 , PSTOCK and A values could lie in a peptide concentration dependent non-linearity in the measured ellipticities for each data set. Increasing the peptide concentration would change the refractive index of the solution. If the refractive index (n) increases then the absorptivity (E) of the right and left circularly polarized light will decrease according to the equation: & = SmuE^2f (43) 78 Chapter 2 - Model This would result in a decrease in the ellipticity measurements. Decreased ellipticity measurements will have the effect of making the peptide solution appear less concentrated. This would result in a PS T O C K value generated from the fit that is lower than expected as the peptide concentration increases. This would explain the increase in the calculated PS T O C K values over the 8-fold peptide solution dilution. The question arises as to whether this concentration dependent non-linearity could be the cause of the concentration dependent nature of the C a 2 + dissociation constants calculated with the single site model. This was tested by fitting the C a 2 + titration data to the monomer model using the peptide concentrations from the dimer model. The calculated values are summarized in Table VII and demonstrate little change from the values calculated using the peptide concentrations based on amino acid analysis. This indicates that the deviations in the PS T O C K parameters are not responsible for the 4-fold increase in the single site KD values over the 8-fold dilution. In an attempt to minimize the concentration dependent deviations between the 94-individual Ca titration data sets, a mathematical model based on the dimer model was developed that describes additive changes in ellipticity at each titration point in all four data sets simultaneously. The curve fit resulted in K i , K 2 and PS T O C K values very similar to the average values generated by fitting each curve separately (Table VIII). The advantage of using this model is that it leverages the fact that the peptide concentrations are related by serial dilution. This results in smaller errors in the generated values due in part to the fact that deviations from the ideal in each data set will be averaged by the single curve fit. The single PS T O C K value generated represents all four peptide concentrations studied related by serial dilution. The generation of one PS T O C K that 79 Chapter 2 - Model describes all four data sets is reflected in the calculated A values which show a 5% drop over 8 fold dilution (compared to a -20% drop in A observed for the individually analyzed data sets). However, this PS T O C K value is equivalent to a peptide concentration of 2.09 mg/ ml which is -10% higher than that determined by amino acid analysis. This discrepancy in the calculated and measured stock peptide concentration for the Ca titration studies can be explained by examining the origin of the PS T O C K parameter. The PS T O C K parameter is the theoretical peptide stock concentration that best describes the Ca titration data sets. The fit value of the PS T O C K parameter is determined by the B and D molar ellipticity parameters from the dimerization study. It is this dependence that is the possible cause of the discrepancy. It is assumed that the B and D parameters in both the dimerization equation and the calcium titration equation are the same. However, i f the B and D parameters are not exactly the same for both data sets then deviations in the parameter that is dependent on B and D in the C a 2 + titration, namely PS T O C K , will occur. What could cause this difference? One difference is that the dimerization study is conducted at a CaCb concentration of 200 m M for all measurements whereas the C a 2 + titration study goes from 0 to 20 m M CaCl2. The higher CaCl2 concentration in the dimer study could increase the refractive index resulting in a decrease in the absorptivities (s) of the right and left circularly polarized light and a decrease in the difference between the two, As (EL - £R)- A S the As and molar ellipticity [9] are related by the equation [6] = 32.98- (As), the B and D parameters would decrease proportionally. If these B and D parameters from the dimer study are then used in the C a 2 + titration equation (as is the case in this method) it would appear as though the peptide concentrations for the C a 2 + titration studies have increased. In effect, by allowing the PS T O C K value to be fit in the 80 Chapter 2 - Model C a 2 + titration equation we compensate for lower B and D parameters from the dimerization study. This discrepancy in the measured and fit peptide concentrations for the Ca titration data underscores the difficulty in combining parameters from two data sets that have different experimental conditions. Even with this discrepancy the simultaneous fit dimer model best describes the data and will be the basis of all C a 2 + binding studies presented. 2.6.2. Model Validity Examination of the fitted molar ellipticity constants reveals little difference between the values of the constants A and B as well as C and D. If this is the case, how can C a 2 + binding to the monomer or dimer be detected when little or no change in the molar ellipticity constants occurs on Ca binding? It turns out that the shape of the titration curve allows determination of the Ki and K 2 parameters even if A=B and C=D as long as there is a measurable difference between the two sets of constants. This is illustrated in Figure 17 which shows how the shape of the curves change with changing Ki and K 2 values. Figure 17A demonstrates how with a constant K 2 and KDIM, a change in the Ki value will act to shift the plot along the x-axis. A similar change in K 2 (with Ki and KDIM constant ) acts to shift the curve until K 2 < Ki at which point the curve starts to converge (Figure 17B). When K 2 is 1000-fold smaller than Kj , convergence occurs. At this point the observed change in ellipticity is dependent on Ki and KDIM only. This limits the dimer model to estimating K 2 only if K] and K 2 are less than 1000 fold apart. 81 Chapter 2 - Model 260 -r 240 -1e-12 1e-11 1e-10 1e-9 1e-8 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 Free Ca2 + (M) 260 -r 240 -1e-10 1e-9 1e-8 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 Free Ca 2 + (M) Figure 17. The effect of changing K] (A) and K2 (B) on shape of Ca 2 + titration plot. Parameters used for modeling were P T OT = 0.5, A=B=-1.5xl0 6 and C=D=-8.5xl0 6 deg-cm^dmol"1, K D I M = 5xl0" 5 M, K 2 = lxlO" 6 M (A) and K, = lxlO" 5 M (B). K, and K 2 were modeled over the free Ca2 + range 1 to lxlO" 8 M. Solid line indicates Ki=K 2 . Modeling was done with Mathcad by Mathsoft Inc. 82 Chapter 2 - Model 2.6.3. Comparison of Different Dimer Models Dimerization in synthetic single site EF-hand peptides complicates their use as models to study Ca binding to the native sites. This study uses circular dichroism to estimate the extent of dimerization which in turn allows estimation of C a 2 + dissociation constants for the monomer and dimer forms. Other groups have attempted to measure these equilibrium constants (Monera et al., 1992; Shaw et al., 1991a). Shaw et al. examined a 34 residue peptide based on the sequence of TnC3. Their technique had two experimental components as in this study. The first was a N M R monitored C a 2 + titration. This was combined with an experiment in which N M R was used to measure the amount of unbound and dimerized peptide where the ratio of peptide to C a 2 + was kept constant but the amount of peptide was decreased. Using both sets of data an estimate of KDIM' (the dimerization of the Ca bound monomer to an unbound monomer) as well as the Ca dissociation constants for the monomer and dimer were reported. The mathematical models used by Shaw et al. assume that the N M R resonance that corresponds to C a 2 + bound peptide is a combination of the individual resonances of the three peptide forms PCa, P 2 Ca and P 2 Ca 2 . In the mathematical models used to describe the circular dichroism monitored titrations this would be equivalent to the molar ellipticities of the peptides in each of the three Ca bound forms being equal (i.e. B = CI2 = DI2). (Note: The molar ellipticity of each peptide in the dimer form is half that of the molar ellipticity of the dimer as a whole). Figure 18 demonstrates how the shape of the C a 2 + binding curve monitored using circular dichroism will still allow determination of the constants K i and K 2 . As in Figure 17A, the K] constant shifts the curve along the x axis (Figure 18A). A decreasing K 2 constant also shifts the curve until K 2 is 1000 fold smaller than K i at which 83 Chapter 2 - Model 260 -r 240 -60 -j 1 1 1 1 1 1 1 1 1 1 1 1 1e-12 1e-11 1e-10 1e-9 1e-8 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 Free C a ^ (M) 260 i 240 -60 H 1 1 1 1 1 1 1 1 1 1 1e-10 1e-9 1e-8 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 1e+0 Free C a 2 + (M) Figure 18. The effect of changing Kj (A) and K2 (B) on shape of Ca 2 + titration plot when B=C/2=D/2. Parameters used for modeling were P T 0 T = 0.5, A=-1.5xl06, B=-4.25xl0 6 andC=D=-8.5xl0 6 degcm2-dmol" K D 1 M = 5xl0"5 M, K 2 = lxlO"6 M (A) and K, = lxlO" 5 M (B). K, and K 2 were modeled over the free Ca2 + range 1 to lx l0" 8 M. Solid line indicates K]=K 2 . Modeling was done with Mathcad by Mathsoft Inc. 8 4 Chapter 2 - Model point convergence occurs (Figure 18B). Thus, the technique described by Shaw et al., though it used a different mathematical model and a different experimental approach, appears to be valid and represents an alternative to the method described here. The study by Monera et al., like this study, used circular dichroism to estimate dimerization and C a 2 + binding constants (Monera et al., 1992). The peptides examined were 34 residue peptides based on the sequence of TnC3. The observation that the negative molar ellipticity of EF-hand peptides can increase with increasing peptide concentration was first reported by Monera et al. However, the ellipticity data was not used to estimate dimerization constants. Instead, they fit a CD monitored C a 2 + titration using model curve fitting techniques developed for N M R data (Shaw et al., 1991a) to generate KDIM ' and K i and K 2 . This would have been a valid technique i f the molar ellipticity of the peptide in all three C a 2 + bound forms was equal. The study by Monera et al. clearly demonstrates, in a concentration versus molar ellipticity plot, that this is not the case with the peptide in the C a 2 + bound dimer contributing a greater amount to the measured molar ellipticity then the C a 2 + bound monomer (the [9] of the dimer is > 2 times the [9] of the monomer). This fact is not represented in the equations used to fit the data. For the CD data reported by Monera et al. the relationship between the fractional intensity /and the molar amounts of the different peptide forms is not as follows; Instead, because of the difference in molar ellipticity for the monomer and dimer Ca bound forms, the equation should be : QjVol, -9 0 vo/ 0 PCa + 2-P 2 Ca + 2 P 2 C a 2 Qjvolj-Q0vol0 P T O T (44) / = 9,-vo/,. -9 0 vo/ 0 QyVolf -Q0volQ (B-A)-PCa, + (C-2-A)-V2Ca{ + ( D - 2 - ^ ) - P 2 C a (D-2-B)-?2Ca2F + (B-A)-PT0T 2i (45) 85 Chapter 2 - Model Even i f C = D, as is observed for PPPQKED, this equation would only be simplified to: e,.vo/,.-fy^o _ ( ^ ) - P C a i + ( D - 2 ^ ) - ( P 2 C a i + P 2 C a 2 i ) Qfvolf-Q0vol0 (D-2-B)-F2Ca2F+(B-A)-FT0T ^ ) It is only i f B = C/2 = D/2 that equation 45 would be reduced to equation 44. Another problem with the Monera et al. study is that the C a 2 + titration data alone was used to calculate the C a 2 + binding and dimerization constants. Using only the C a 2 + titration data was found to be inadequate to derive the constants by the original authors of the model and required an additional experiment as described above (Shaw et al., 1991a). It can thus be concluded that the interpretation of the C a 2 + binding data by Monera et al. may not be valid. Another study in which dimerization and C a 2 + binding was examined in an EF-hand motif was with a synthetic 26 residue fragment of the extracellular glycoprotein BM-40 94-(Maurer et al., 1995). In the Ca binding experiment the change in fluorescence was monitored (in contrast to CD used in this study). The mathematical model used to analyze the data assumed no C a 2 + dissociation from the dimerized form and thus the constants calculated described C a 2 + binding to the monomer and the dimerization dissociation 24" constant for the Ca bound monomer. Though a simplified model, the technique is mathematically sound. The dimer model presented here represents an alternative method to N M R for determining dimerization and C a 2 + dissociation constants for EF-hand peptides. The principles outlined in this study would also be applicable to any other spectroscopic techniques used to analyze C a 2 + binding EF-hand peptides e.g. fluorescence. 86 Chapter 3 -X Substitution CHAPTER 3 A -X GLUTAMATE TO ASPARTATE SUBSTITUTION IN THE CD SITE OF PARVALBUMIN: EFFECTS ON Ca2+ AFFINITY AND DIMERIZATION 3.1. Overview This chapter examines the C a 2 + and M g 2 + binding characteristics of a synthetic peptide based on the CD site of parvalbumin 4.25 resiudes 39-71, (PPPQKED) and compares it to a peptide with the same sequence in which the - X Glu was replaced with Asp (PPPQKED-XD). The CD site of parvalbumin has not previously been examined in peptide form although the EF site of parvalbumin has (Revett et al., 1997). Replacement of a - X Glu with Asp has been examined in synthetic peptides based on bovine brain clmodulin site III (Procyshyn and Reid, 1994a), in the EF-hand like site in E.Coli D-galactose binding protein (Drake et al., 1996), and in rat oncomodulin (Hapak et al., 1989). Only in oncomodulin does a - X Glu give higher C a 2 + affinity than Asp. The 94-effects on dimerization and Ca binding of the - X replacement in PPPQKED are assessed and compared to the effects observed in the other EF-hand sites in which this substitution has been examined. 3.2. Materials t-BOC Solid Phase Peptide Synthesis, Peptide Purification, Dimerization and Ca 2 + Titration Studies • See chapter 2 section 2.3 87 Chapter 3 -X Substitution M g 2 + Studies • Magnesium chloride hexahydrate, potassium chloride, potassium hydroxide, were obtained from B D H Chemicals Canada Limited, Toronto, ON, Canada. • 3-(N-Morpholino) propanesulfonic acid (MOPS) was from the Sigma Chemical Co., St. Louis, MO, USA. • Chelex (sodium form, 200-400 Mesh) was from Bio-Rad Laboratories, Hercules, CA, USA. 3.3. Methods 3.3.1. tBOC Solid Phase Peptide Synthesis The peptide PPPQKED-XD was synthesized using solid phase methodology and tBOC chemistry on a Vega 1000 peptide synthesizer as described in section 2.4.1. The sequence was based on the CD site of parvalbumin 4.25 resiudes 39-71 (Coffee and Bradshaw, 1973) with the Glu in position 59 replaced with Asp. The peptide is N-acetylated and C-amidated. The sequence from the N-terminus is: 1 33 PPPQKED-XD S A D D V K K A F A I I D Q D K S G F I D E D E L K L F L Q N F K 3.3.2. Purification of PPPQKED-XD The purification, confirmation of molecular weight and assessment of purity of PPPQKED-XD was accomplished as described in section 2.4.2. 3.3.3. Dimerization of PPPQKED-XD Dimerization of PPPQKED-XD was assessed by dissolving 7.61 mg of lyophilized 88 Chapter 3 -X Substitution peptide in 500 uL of a buffer containing 200 m M CaCl 2 , 100 m M KC1 and 50 m M MOPS, pH 7.5. The peptide concentration was determined using amino acid analysis (section 2.4.5). A serial dilution of this peptide solution was made using this same buffer and examined using a Jasco J720 spectropolarimeter to generate the dimerization constant KDIM and the B and the D values as described in section 2.4.3. The mean residue ellipticity values for the C a 2 + bound monomer PCa and C a 2 + bound dimer P2Ca2 were calculated from the B and D values using equation 7 (page 51). 2+ 3.3.4. Ca Binding Measurement A stock peptide solution of PPPQKED-XD was prepared by dissolving 4.14 mg of lyophilized peptide in 2 mL of a buffer containing 50 mM MOPS and 100 m M KC1, pH 7.5. The stock solution was serially diluted and each solution was titrated with C a 2 + while monitoring at 222 nm with a Jasco J720 spectropolarimeter. A quartz cell with a pathlength of 1.0 mm was used. The titration data was fit to the simultaneous fit dimer model as described in section 2.4.4 to generate the parameters K i , K 2 , PSTOCK- The A parameter was generated from the fit parameter PSTOCK , the serial dilution factors, the volume of the solution (900 uL) and the ellipticity measurement of each peptide solution 24" in the absence of Ca . The mean residue ellipticity value for the unbound monomer P was calculated from the A value using equation 7 (page 51). 3.3.5. Mg2+ Sensitivity Studies 2+ Mg sensitivity studies were carried out by dissolving the peptides in a buffer containing 100 m M MOPS, 150 m M KC1, 1 m M nitrilotriacetic acid, pH 7.2. The peptide 89 Chapter 3 -X Substitution solutions were prepared and used immediately. The concentrations determined by amino acid analysis are; PPPQKED - 0.407 mg/mL, PPPQKED-XD - 0.355 mg/mL. To 900 uL of the peptide solutions, 200 pL of a 1.18 M M g C l 2 solution was added resulting in a final M g C l 2 concentration of 214.5 mM. The ellipticity over the wavelength range 210-250 nm was measured in the absence and presence of M g 2 + . The pathlength of the quartz cell was 1 mm. The spectra were converted to mean residue ellipticity (0)^ at each wavelength X to correct for dilution effects using equation 7 (page 51). 3.3.6. Tests of Significance Differences in the parameters generated for the peptides PPPQKED and PPPQKED-XD were assessed using the one tailed t-test. The difference in the dimerization dissociation constants (KDIM) and the mean residue ellipticity of the dimer forms P2Ca2 were assessed using asymptotic standard errors calculated from the curve fit +5% for peptide concentration error and residual degrees of freedom of 7 (taken from curve fit results by Sigmaplot®). In the case of the C a 2 + bound monomer (PCa) the error was taken to be ±5% (peptide concentration error). Differences in the C a 2 + dissociation constants of the monomer (Ki) and dimer (K 2 ) peptide forms were assessed using asymptotic standard errors calculated from the curve fit +5% (peptide concentration error). However, the residual degrees of freedom used was 6. The difference in the average mean residue ellipticity of the unbound monomer forms P was assessed using the standard errors calculated from the SD + values +5% (peptide concentration error) and degrees of freedom of 3 (a total of four measurements were used). 90 Chapter 3 -X Substitution 3.3.7. Molecular Models Molecular models of the peptide dimers were constructed from the N M R structure of TnC3 (Shaw et al., 1992) using the Swiss-PDB Viewer by Nicolas Guex (http ://www.expasy.ch/spdbv/mainpage.htm 1) and rendered using Weblab Viewer Pro from MSI Inc, (San Diego, CA, USA). 3.4. Results 3.4.1. Peptide Purification The HPLC elution profile of the unpurified PPPQKED-XD and the CE elution profile of the purified peptide are given in Appendix A. See Appendix B for measured molecular weight. 3.4.2. Dimerization and Ca 2 + binding to PPPQKED-XD The stock peptide concentration was determined to be 14.48 ± 0 . 1 8 mg/mL from amino acid analysis. The plot of peptide concentration versus molar ellipticity and the C a 2 + titration plots for PPPQKED-XD are given in Figure 19 with the calculated C a 2 + dissociation and dimerization dissociation constants summarized in Table IX along with those of PPPQKED. The mean residue ellipticity values at 222 nm for the four forms, unbound monomer (P), C a 2 + bound monomer (PCa), dimer bound to single C a 2 + (P2Ca) and C a 2 + bound dimer (P2Ca2) of both PPPQKED and PPPQKED-XD were calculated with equation 7 (page 51) from the parameters A, B, and C / D parameters and are included in Table IX. The C a 2 + titration data sets fit individually to the single site and dimer models are given in Appendices C and D, respectively. 91 Chapter 3 -X Substitution Table IX. Dimerization and Ca Dissociation Parameters for PPPQKED and PPPQKED-XD. Parameters* PPPQKED P P P Q K E D - X D K D I M ± SE (uM) 56 ± 14 33 ± 9 Ki ± SE (uM) (Ca2+) 431+33 228 ± 35 K 2 ± SE (uM) (Ca2+) 37 + 6 24 ± 11 P S T O C K (pinole)* 0.489 + 0.025 0.447 + 0.024 t ! V O CD B * — Q w tio A B C/D 1.42 ±0.08 1.88 ±0.09 8.30 ±0.61 1.75 ±0.12 2.38 ±0.12 8.66 ± 0.59 <^  s g s 5 1 O In bO tD T 3 P PCa P 2Ca / P 2 Ca 2 4300 ±300 5700 ±300 12600 ±900 5300 ±400 7200 ±400 13100±1000 * The errors reported are the asymptotic SE generated from the curve fit + 5% except for the A which has its SE generated from the four ellipticity values in the absence of Ca 2 + + 5%. "Molar ellipticity for the peptide forms P - A, PCa - B and P 2Ca / P 2 Ca 2 -C/D. * Mean residue ellipticity for the peptide forms P, PCa, and P 2Ca / P 2 Ca 2 calculated from the A, B, and Cl D constants using equation 7 (page 51). * Total amount of peptide (umoles) in 900 uL of peptide stock solution. 92 Chapter 3 -X Substitution 1e-5 1e-4 1e-3 1e-2 T o t a l P e p t i d e ( M ) 400 T 1 B. 150 J . , . , . 1 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 Free C a 2 + (M) Figure 19. Plot of molar ellipticity versus total peptide concentration (A) and total titration point ellipticity versus free C a 2 + concentrations (B) of PPPQKED-XD. The Ca2 + titrations were carried out on four different peptide concentrations with relative peptide concentrations of 2 (•), 1 (•), 0.5 ( A ) and 0.25 ( T ) . 93 Chapter 3 -X Substitution ^ 3.4.3. Comparison of PPPQKED and PPPQKED-XD The dimerization dissociation constant for -PPPQKED is 56 ± 14 u M and for PPPQKED-XD is 33 ± 9 uM. Replacement of the - X Glu with Asp resulted in a 41% decrease in the KDIM value (p = 0.10). The C a 2 + dissociation constants for the monomer and dimer forms of PPPQKED are 431 ± 3 3 uM and 37 ± 6 uM, respectively. The C a 2 + dissociation constants for the monomer and dimer forms of PPPQKED-XD are 228 ± 3 5 u M and 24 ± 1 1 uM, respectively. Thus replacement of the - X Glu with Asp results in a 47%) decrease in K i (p <0.01) and a 35% decrease in K 2 (p = 0.17). The decrease in K i is significant but the decrease in K 2 is not. See Figure 20A for a bar chart illustrating the changes in KDIM, K J and K 2 . Replacement of the - X Glu with an Asp resulted in a 26% increase in the negative mean residue ellipticity of P, increasing from -4300 ± 300 to -5300 ± 400 deg-cm2-dmol"1-residue"1 (p = 0.07). The negative mean residue ellipticity of PCa increased 27% from -5700 ± 300 to -7200 ± 400 deg-cm2-dmol"1-residue"' (p <0.01). The mean residue ellipticity values of P2Ca and P2Ca2 are equal because of constraints placed on the curve fitting process. The negative mean residue ellipticity of P2Ca2 increased 4% from -12600 ± 900 to -13100 ± 1000 deg-cm2-dmol"1-residue"1 (p = 0.36). Figure 20B illustrates the mean residue ellipticities for the peptide forms PPPQKED and PPPQKED-XD. 94 Chapter 3 -X Substitution A. P P P Q K E D P P P Q K E D - X D 500 400 300 200 100 DIM B. 16000 v 14000 •g <g 12000 ^ 10000 E "O £ o cn -a 8000 6000 j 4000 2000 Figure 20. -X Glu to Asp substitution in PPPQKED. A) The effect on K D I M , K, and K 2. B) The effect on mean residue ellipticity of the peptide forms P (monomer), PCa (Ca 2 + bound monomer) and P 2Ca / P 2 Ca 2 (dimer forms). The peptide pair examined was PPPQKED/PPPQKED-XD. 95 Chapter 3 -X Substitution 3.4.4. M g 2 + study of PPPQKED and PPPQKED-XD The peptides PPPQKED and PPPQKED-XD demonstrate little change in the mean residue ellipticity over the wavelength range 210 to 250 nm on addition of > 200 m M of M g 2 + (Figure 21). This suggests that neither peptide exhibits M g 2 + induced conformational change. PPPQKED 210 220 230 240 250 Wavelength (nm) PPPQKED-XD 210 220 230 240 250 Wavelength (nm) Figure 21. Addition of M g 2 to PPPQKED and PPPQKED-XD. Change in mean residue ellipticity measured over the wavelength range 210-250 nm in the absence of M g 2 + ( ) and in the presence of > 200 mM M g 2 + ( ). 96 Chapter 3 -X Substitution 3.5. Discussion 3.5.1. The CD Site Model Peptide - PPPQKED The peptide PPPQKED, which corresponds to the CD site of Parvalbumin, demonstrates a K | of 431 ± 33 u M which is > 105 fold lower than the measured C a 2 + affinity of the site in the native protein ( K D = 3.7 x 10"10 M) (Moeschler et al., 1980). The lower Ca affinity of the model site is not unexpected, as both the EF site and the inactive A B site have been shown to positively contribute to the C a 2 + affinity of the native parvalbumin CD site (Pauls et al., 1994; Permyakov et al., 1991). The lower affinity of the PPPQKED model site may be partially explained by a lower conformational stability. This is reflected by the 10 fold higher C a 2 + affinity of the dimer form of PPPQKED (K 2= 37 ± 6 uM), which would be expected to have increased stability. However, the C a 2 + affinity of the dimer form remains >104 fold lower than the site in the native protein. This difference in Ca affinity is greater than that seen in other model sites studied. Work on bovine brain calmodulin site III reveals a model site with 9+ 50-700 fold lower Ca affinity than the native site (compare CaM:3(DNS) from (Procyshyn and Reid, 1994b) and F92W from (Wu and Reid, 1997a)). A 34 residue model site based on rabbit skeletal troponin C site III demonstrates 18 fold lower C a 2 + affinity than the native site (compare STnC (Reid, 1987b) and troponin C high affinity domain (Ogawa, 1985)). Thus it appears that the high C a 2 + affinity of the native parvalbumin CD site is more dependent on its protein environment than other EF-hand sites. The basic assumption in studying model peptides is that there is a correlation between the C a 2 + affinity of a model site and the same site in the native protein. 97 Chapter 3 -X Substitution Mutational studies involving the chelating residues of bovine brain calmodulin site III demonstrate just such a correlation (Wu and Reid, 1997a; Wu and Reid, 1997b). If the parvalbumin CD site is like calmodulin site III then one would expect the peptide PPPQKED to have the highest C a 2 + affinity observed in a model site owing to the high Ca affinity of the native site. However, this is not the case, with other EF-hand model peptides exhibiting higher C a 2 + affinity than PPPQKED (Procyshyn and Reid, 1994b; Reid, 1987b) (Note: these earlier studies do not take into account the effects of dimerization). The cause of this discrepancy may lie in the thermodynamics of chelation by the - X Glu residue. A Glu residue is rarely found in the - X position in EF-hand C a 2 + binding sites but is invariant in identified parvalbumin CD sites (Table IV, page 40) (excluding oncomodulin CD sites which have an Asp in the - X position). In the native parvalbumin CD site the - X Glu directly chelates the Ca ion and displaces a water molecule (Kretsinger and Nockolds, 1973). This is in contrast to residues such as Asp or Ser in the - X position which chelate via an intermediate water molecule (Strynadka and James, 1989). In order to understand the effect of direct versus indirect chelation by a - X Glu the 9+ thermodynamic contribution to Ca chelation by this residue must be discussed. 9-4-The free energy change of Ca binding to EF-hand peptides (AG°) is governed by both enthalpy changes (AH 0), temperature (T), and entropy changes (AS°) as described in equation 4, page 26. In order to understand the effect on C a 2 + affinity of chelation by a - X Glu in an EF-hand site, one must examine the potential thermodynamic costs and gains associated with it. On the cost side there is the energy required to bring the - X Glu into chelating position and to remove the water from the C a 2 + ion and side chain. 98 Chapter 3 -X Substitution Additional cost arises from the loss of the Glu side chain entropy. The thermodynamic gain is due to the freeing of water molecules from C a 2 + and the Glu side chain, increasing the entropy of the system. In both the native parvalbumin CD site and the PPPQKED model site, one would expect the thermodynamic gains and losses for chelation by the - X Glu to be roughly equal with the exception of the thermodynamic cost associated with bringing the residue into chelating position. The difference in the thermodynamic cost of bringing the - X residue into chelating position would arise from repulsive forces within the chelating loop. On direct chelation of the C a 2 + ion by the - X Glu, the side chain would be expected to experience repulsive forces from the negative charges of the other chelating ligands. Work on D-galactose binding protein demonstrates that negatively charged Glu and Asp residues in the - X position confer lower C a 2 + affinity than their neutral counterparts Gin and Asn (Drake et al., 1996). This repulsion by the - X Glu would induce strain in the loop region of the EF-hand and necessitate an enthalpic cost to maintain direct chelation. In the native parvalbumin CD site one would expect greater stability and thus the enthalpic cost to maintain direct chelation would be less then in the PPPQKED model site. The high C a 2 + affinity of the native parvalbumin CD site suggests that the direct chelation of the C a 2 + and displacement of water by the - X Glu is thermodynamically favorable, i.e. the thermodynamic benefit of direct chelation is greater than the enthalpic costs. The model site would have lower stability than the native site and would require a higher enthalpic cost to overcome the repulsive forces and maintain direct chelation. This enthalpic cost could negate partially or fully the thermodynamic benefit of direct chelation resulting in the lower than expected C a 2 + affinity of PPPQKED. A high 99 Chapter 3 -X Substitution enthalpic cost due to - X Glu repulsion could also explain the lack of M g 2 + binding in the PPPQKED peptide. The M g 2 + has a higher charge density than C a 2 + resulting in higher enthalpic costs of water removal from the ion. Thus the net thermodynamic benefit of direct chelation would be smaller for M g 2 + than for Ca 2 + . In addition, the smaller radius of the M g 2 + ion would mean closer ligands and increased repulsion. 3.5.2. Effect o f - X Glu to Asp Replacement. To assess the extent of negation of the thermodynamic benefit of direct chelation of C a 2 + in the PPPQKED model site, the - X Glu was replaced with Asp resulting in the peptide PPPQKED-XD. This peptide should experience decreased repulsion by the - X residue and should lose the thermodynamic benefit of direct chelation. Replacement of the PPPQKED - X Glu with an Asp resulted in a 2-fold increase in dimerization and a 2-fold increase in C a 2 + affinity of the monomer form. In the dimer form this replacement also increases calcium affinity though this increase is not significant (p = 0.17). The peptide PPPQKED-XD remained unable to bind M g 2 + . An increase in dimerization is not unexpected as the N M R structure of a 34 residue TnC3 dimer (Shaw et al., 1992) shows the - X position at the dimer interface (Figure 22). Thus a change in the - X residue could alter the interface and result in the change in dimerization. The higher C a 2 + affinity for the monomer forms of PPPQKED-XD compared to PPPQKED indicate that a - X Asp is thermodynamically more favorable than Glu in the model site. Two different processes could be responsible for the - X Asp being more thermodynamically favorable. First, the enthalpic costs associated with repulsion by the - X Glu in PPPQKED may be greater than the thermodynamic benefit of direct chelation. 100 Chapter 3 -X Substitution PPPQKED PPPQKED-XD Figure 22. Dimer interaction in PPPQKED and PPPQKED-XD. Dimer model of PPPQKED (top) and PPPQKED-XD (bottom) with one peptide in each dimer displayed as the a-carbon backbone and the other as a surface model. The loop regions are highlighted in red with the side chain of the residue in the - X position shown. Areas highlighted in green are within 3 A of atoms found with the loop region of the complementary peptide. Models based on the N M R structure of the TnC3 dimer (Shaw et al., 1992). 101 Chapter 3 -X Substitution Examination of the effect of the - X Glu to Asp substitution in other peptide / protein systems demonstrates how site stability appears to decrease the enthalpic cost of direct chelation. As stated in the introduction (section 1.4.2), a model site corresponding to Calmodulin site III and the EF-hand like C a 2 + binding site in the D-galactose binding protein demonstrate an increase in C a 2 + affinity on replacement of a - X Glu with an Asp (Drake et al., 1996; Procyshyn and Reid, 1994a). The same substitution in the parvalbumin-like protein oncomodulin decreases C a 2 + affinity (Hapak et al., 1989) (See Table X for dissociation constants of different sites). Thus it appears that Glu confers lower C a 2 + affinity than Asp in synthetic peptides and the D-galactose binding protein, which has a non-contiguous loop region (Vyas et al., 1987), but confers higher C a 2 + affinity in oncomodulin. In the synthetic sites and D-galactose binding protein one would expect lower site stability and thus higher enthalpic costs. The difference in the free energy change on a - X Glu to Asp replacement (AAG°) in different sites is a measure of the thermodynamic benefit of the - X Glu to Asp substitution. Using the binding constants, the AAG° values for the different peptides were calculated using equation 3 (page 26) and are summarized in Table XI . Of the different sites, D-galactose binding protein has the largest AAG° at -7.63 KJ/mol. In D-galactose binding protein, the -Z residue is not part of a contiguous chelating loop. In addition, D-galactose binding protein does not have a Ser in the +Z position which is shown to hydrogen bond with a - X Glu in parvalbumin and would potentially stabilize the site and reduce the effects of repulsion (Kumar et a l , 1990) (Figure 24B). Both the lack of a continuous loop and the lack of the stabilizing interaction between the +Z Ser and - X Glu could be the cause of the large AAG° in the D-galactose binding protein. In 102 Chapter 3 -X Substitution Table X. Comparison of Ca 2 + Dissociation Constants for EF-hand Peptides and Proteins in which a -X Glu is Substituted with Asp. Peptide K i SK ( u M ) ( C a 2 Y K 2 ± SE (pM) (Ca2+)* PPPQKED 431 ± 3 3 37 ± 6 PPPQKED-XD 228 ± 35 24 ± 11 K D (pM) (Ca2+) Cam3:(DSE)** 2806±185 ~ Cam3:(DSD)** 407 ± 27 — GBP Q142ET >500 ~ GBP Q142D 1 23 ± 2 — r O M D59E* 0.55 ±0.08 ~ rOM* 0.78 ± 0.07 ~ The SE reported is the Asymptotic SE generated from the curve fit +5%. ** Cam3:(DSE) is from (Procyshyn and Reid, 1994a) and Cam3:(DSD) is from (Procyshyn and Reid, 1994b). The K D values were calculated using the single site model and are ± SD. T Mutants of E.Coli D-galactose binding protein from (Drake et al., 1996). * Rat oncomodulin mutants from (Hapak et al., 1989). the three contiguous EF-hand sites, the Cam3 peptides demonstrate the greatest negative AAG° at -4.78 KJ/mol. Next is the monomer forms of the PCD peptides at -1.58 KJ/mol and then the dimer forms at -1.08 KJ/mol. Finally there is the CD site of oncomodulin in which the AAG° is +0.87 KJ/mol. In this site the - X Glu gives higher C a 2 + affinity than Asp. Since all these sites have identical chelating residues (save the - Y residue that chelates via the backbone) the differences in AAG° must be due the non-chelating loop 103 Chapter 3 -X Substitution Table XI. Change on Free Energy (AAG°) of Ca 2 + Chelation on Glu to Asp Substitution in the -X Position in Different EF-hand Sites. Peptide AG° (Ki) * AAG°(K,) KJ/mol KJ/mol AG 0 (K2) * AAG° (K2) KJ/mol KJ/mol PPPQKED PPPQKED-XD -19.20 -1.58 -20.78 -25.28 -1.08 -26.36 AG" (KD)* A AG" (K|>) K.l/mol KJ/mol l I l A l I l l ^ ^ GBP Q142E1" Q142D r -18.83 -7.63 -26.46 Cam3:(DSE)** Cam3:(DSD)** -14.56 -4.78 -19.34 r O M D59E* rOM* -35.71 +0.87 -34.85 All AG 0 values calculated using equation 3 (page 26) with a T of 298 °K. Cam3:(DSE) is from (Procyshyn and Reid, 1994a) and Cam3:(DSD) is from (Procyshyn and Reid, 1994b). The AAG value was calculated from K D values calculated with a single site model. f Calculated from K D values of E.Coli D-galactose binding protein mutants from (Drake et al., 1996). * Calculated from K D values of rat oncomodulin mutants (Hapak et al., 1989). residues, flanking helices or whole protein effects (See Figure 23 for loop sequences). The different A A G 0 values indicate different thermodynamic costs of - X Glu chelation. One way in which factors outside of the chelating residues can alter the thermodynamic cost of - X Glu chelation is through site stabilization. Site stabilization would effect the whole site and one should see an increase in the relative C a 2 + affinity as the A A G 0 decreases. This is exactly what is observed, with the Cam3 peptides having the lowest 104 Chapter 3 -X Substitution EF-Loop Position +x +Y +z - Y - X -z 1 12 PPPQKED D Q D K S G F I E E D E PPPQKED-XD D Q D K S G F I D E D E Cam3:(DSE) D K D G S G Y I E A A E Cam3:(DSD) D K D G S G Y I D A A E rOM D59E D N D Q S G Y L E G D E rOM D N D Q S G Y L D G D E GBP Q142E* D L N K D G Q I E E GBP Q142D* D L N K D G Q I D E Figure 23. 12 residue loop sequences of the site pairs PPPQKED / PPPQKED-XD, Cam3:(DSE) / Cam3:(DSD) (Procyshyn and Reid, 1994a; Procyshyn and Reid, 1994b), rOM D59E / rOM (Hapak et al., 1989), GBP Q142E / Q142D (Drake et al., 1996). *The Glu in the -Z position in GBP is not part of the contiguous loop. Identical chelating residues are boxed. C a 2 + affinities and the largest AAG° and the oncomodulin CD site having the highest C a 2 + affinities and a positive AAG°. Therefore, increasing site stabilization may be increasing the C a 2 + affinities and decreasing the thermodynamic cost of - X chelation until the thermodynamic benefit of direct chelation is observed. Evidence of a second factor contributing to the higher calcium affinity of PPPQKED-XD comes from examination of the mean residue ellipticity data. The peptide PPPQKED-XD demonstrates -26% greater negative mean residue ellipticity in the unbound and calcium bound monomer forms than PPPQKED. This suggests that in the monomer forms the higher calcium affinity of PPPQKED-XD may be due in part to enhanced structural stability. Increased structural stability in PPPQKED-XD could arise 105 Chapter 3 -X Substitution from a hydrogen bond between the side chain of the - X Asp and the hydrogen of the backbone nitrogen of the -Z Glu. This is observed in site III in the crystal structure of troponin C which naturally has a - X Asp (Figure 24A). This interaction could stabilize the N-terminus (N-Cap) of the C-terminal helix and increase the negative mean residue ellipticity of the monomer forms. Examination of the crystal structure of the parvalbumin CD site (Kumar et al., 1990) reveals no capping interaction though a hydrogen bond between the - X Glu and the +Z Ser is observed (Figure 24B). This hydrogen bond could potentially stabilize the site as well but not be evident as increased negative mean residue ellipticity. Though both sites potentially have stabilizing hydrogen bonds, the capping interaction in PPPQKED-XD may provide greater stabilization and be the cause of the higher C a 2 + affinity of this site. A n indication as to the maximum contribution of the - X Asp capping interaction to the C a 2 + affinity of the peptide PPPQKED-XD comes from comparison of the AAG° values of the monomer and dimer forms. In the more stable dimer form, the negative mean residue ellipticity difference between PPPQKED and PPPQKED-XD is only 4% suggesting the higher calcium affinity for the dimer form of PPPQKED-XD is for the most part the result of decreased repulsion. The AAG° for the dimer forms is 68% of what occurs in the monomer. Therefore a maximum of 32% (0.50 KJ/mol) of the difference in calcium binding free energy between PPPQKED and PPPQKED-XD in the monomer forms can be attributed to enhanced stability from an N-cap hydrogen bond. The actual contribution is most likely less than 31% because of the potential for a stabilizing hydrogen bond between the - X Glu and the +Z Ser in PPPQKED and/or thermodynamic stabilization provided by the dimer environment. 106 Chapter 3 -X Substitution Figure 24. Stick models of the C a 2 + chelating loop of PCD and TnC3. A) Model showing the hydrogen bond between the -X Asp and the backbone amide hydrogen of the -Z Glu. Coordinates taken from the crystal structure of chicken skeletal muscle troponin C (Satyshur et al., 1988). B) Model showing the hydrogen bond between the -X Glu and the sidechain of the +Z Ser. Coordinates taken from the crystal structure of carp parvalbumin 4.25 (Kumar et al., 1990). 107 Chapter 4 - Helix Substitution CHAPTER 4 N AND C TERMINAL a-HELIX REPLACEMENT IN THE PARVALBUMIN CD SITE MODEL PEPTIDE 4.1. Overview This chapter examines the influence of the N and C-terminal PCD helices on the dimerization and Ca affinity of the PCD model site. The series of peptides studied are based on the sequences of PCD and Cam3. The peptides have identical loop regions from PCD flanked by sequences corresponding to the N and C terminal oc-helices of PCD or Cam3. The peptides are termed PPPQKED, CPPQKED, PPCQKED and C P C Q K E D (See Peptide Nomenclature page xvii). Structural models based on the N M R structure of TnC3 (Shaw et al., 1992) are used to interpret changes in dimerization and C a 2 + binding. 4.2. Materials Fmoc Solid Phase Peptide Synthesis • 4-(2',4'-Dimethoxyphenyl-Fmoc-aminomethyl)-phenoxy resin (Rink resin), Fmoc-L-alanine monohydrate, Na-Fmoc-NY-2,2,5,7,8-pentamethylchroman-6-sulfonyl-L-arginine, Fmoc-N-P-trityl-L-asparagine, Fmoc-L-aspartic acid p-t-butyl ester, Fmoc-L-glutamic acid y-t-butyl ester, Fmoc-N-y-trityl-L-glutamine, Fmoc-glycine, N a-Fmoc-N'm-trityl-L-histidine, Fmoc-L-isoleucine, Fmoc-L-leucine, Na-Fmoc-NE-t-butyloxycarbonyl-L-lysine, Fmoc-L-phenylalanine, Fmoc-O-t-butyl-L-serine, Fmoc-O-t-butyl-L-threonine, Fmoc-L-valine and 0-benzotriazole-N,N,N',N'-tetramethyl-uronium-hexafluoro-108 Chapter 4 - Helix Substitution phosphate (HBTU) were from Advanced Chemtech, Inc. (Louisville, K Y , USA). • l-Methyl-2-pyrrolidinone (NMP) was obtained from E M Science, Cincinnati, OH, USA. • Ethyl ether, trifluoroactetic acid (TFA) and dichloromethane (DCM) were from Fisher Scientific Co., Ottawa, Canada. • Acetic anhydride, m-Cresol, N,N-diisopropylethylamine (DIEA), 1,2 ethanedithiol (EDT), 1-hydroxybenzotriazole (HOBt) and piperidine were from Sigma Chemical Co., St. Louis, MO, USA. • Dimethylformamide (DMF) and thioanisole were obtained from the Aldrich Chemical Co., St. Louis, M O , USA. tBOC Solid Phase Peptide Synthesis, Peptide Purification, Peptide Ca 2 + and Dimerization Studies • See materials section 2.3 4.3. Methods 4.3.1. Fmoc Peptide Synthesis Synthesis of the peptide PPCQKED was carried out on an Applied Biosystems 43 3 A Peptide synthesizer (Foster City, CA, USA) using Fmoc chemistry. Fmoc chemistry employs amino acids protected on the a-amino group by the Fmoc (9-fluorenylmethyloxycarbonyl) moiety. These protecting groups are base labile in contrast 109 Chapter 4 - Helix Substitution to tBOC synthesis in which the tBOC group is acid labile. The sequence of the synthesized peptide PPCQKED is given below. 1 33 PPCQKED S A D D V K K A F A I I D Q D K S G F I E E D E L R H V L T N L G A Met naturally found in position 109 of bovine brain calmodulin (peptide position 29) is replaced with a Leu in PPCQKED to avoid oxidation during HF cleavage. The peptide is acetylated at the N-terminus and amidated at the C-terminus. A description of the steps involved in peptide synthesis on a 433A peptide synthesizer is outlined below. a) Peptide Synthesis Modules Fmoc solid phase peptide synthesis of PPCQKED has the same core steps as peptide synthesis described using tBOC synthesis, namely amino acid coupling, washing away of any residual amino acid and removal of the a-amino protecting group. Once the peptide of a desired length is created the N-terminus is acetylated and the peptide is cleaved from the resin support. Fmoc synthesis of PPCQKED required the use of six solvents or solutions which are listed below with their functions. N M P - Washing, dissolving amino acid D C M - Washing H B T U - Activation of the a-carboxyl group on the Fmoc amino acid DIEA - Catalysis Piperidine - Removal of a-amino Fmoc group Acetic anhydride - Acetylation of amino groups. The use of the above solutions or solvents on an Applied Biosystems 433A peptide synthesizer are organized into chemistry modules, each of which perform a step in the 110 Chapter 4 - Helix Substitution synthesis. The chemistry modules are a series of steps performed by the peptide synthesizer. The modules used in the synthesis of PPCQKED have letter codes and are designated as follows: A - Dissolving amino acid in cartridge B - Piperidine deprotection of amino groups C - Acetylation of amino groups with acetic anhydride c - Resin D C M washes D - Resin N M P washes E - Addition of DIEA to dissolved amino acid and transfer to reaction vessel. F -Clean cartridge, couple amino acid, drain reaction vessel, and wash with N M P I - Wait module Peptide synthesis of PPCQKED was carried out at 0.25 mmol scale using 0.7 mmol /g substituted Rink resin. Rink resin results in a C-terminal amide group when cleaved. The resin (357 mg) was placed into the teflon reaction vessel and allowed to soak overnight in D C M . The method modules used were based on FastMoc chemistry. Four-fold excess amino acid (1 mmol) packed into sealed cartridges was used for all cycles. These cartridges were loaded into the synthesizer in the reverse sequence of the peptide PPCQKED. The addition of each amino acid to the growing peptide chain involves a cycle of modules. The cycle for attaching the first amino acid to the resin is as follows: Initial Amino Acid Cycle: c D B A D E F This cycle begins with a D C M wash of the resin (c) followed by a wash with N M P (D). The resin is then treated with piperidine to neutralize the amino groups on the resin (B). The amino acid in the cartridge is dissolved in a mixture of N M P and H B T U 111 Chapter 4 - Helix Substitution solution (A). The resin is again washed with N M P (D). To the cartridge 0.5 ml of DIEA solution is added, mixed and the contents of the cartridge transferred to the reaction vessel (E). The cartridge is then rinsed with NMP, the rinsings added to the reaction vessel and the amino acid allowed to couple for 13 minutes. The amino acid solution is then drained from the reaction vessel. For the next 32 cycles this process is repeated for each amino acid without the initial washing steps. Module B removes the a-amino Fmoc protecting group from the previously added amino acid with piperidine. Thus the amino acid cycle is: Subsequent Amino Acid Cycles: BADEF The conductance of the piperdine eluent was monitored to ensure complete deprotection in each cycle. After synthesis is complete, the terminal Fmoc group is removed and the N-terminus acetylated using the following cycle. Terminal Acetylation Cycle: BIDCDc This cycle involves piperidine deprotection (B), followed by a ten minute wait, followed by N M P washes (D), addition of acetylation solution to the resin (C), more N M P washes (D), and a final D C M wash (c). The peptide-resin is then removed from the reaction vessel and dried in preparation for cleavage. b) Peptide Cleavage From the Resin The dried peptide-resin is placed in a 50 mL round-bottom flask with a stirbar and cooled in an ice bath. The cleavage mixtures that can be used are: 112 Chapter 4 - Helix Substitution A: 0.75 g crystalline phenol B: 0.25 mL EDT 0.25 mL Distilled H 2 0 9.5 mL TFA 0.25 mL EDT 0.5 mL thioanisole 0.5 mL Distilled H 2 0 10 mL TFA Cleavage mixture A is used i f the peptide contains the Arg with a 2,2,5,7,8-pentamethylchroman-6-sulfonyl (PMC) side chain protecting group. Cleavage mixture B is used i f the peptide has residues with a trityl side chain protecting group (e.g. Gin or Asn). Cleavage mixture A was used for the peptide PPCQKED. This mixture was cooled in an ice bath and then added to the peptide-resin. The reaction mixture was stoppered, allowed to warm to room temperature, and stirred for 3.0 hours. After cleavage, the reaction mixture was filtered through a sintered glass funnel into a second round-bottomed flask to remove the resin. The reaction flask was rinsed with ~1 mL of TFA and with ~10 mL D C M which was added to the second round bottomed flask. This mixture was reduced to -1-2 mL with a rotary evaporator with a water bath of less than 40° C. To this concentrated reaction mixture 50 mL of cold ethyl ether was added to precipitate the peptide. The peptide was filtered from the reaction mixture and dried under vacuum. 4.3.2. tBOC Peptide Synthesis The peptides CPPQKED and CPCQKED were synthesized using solid phase methodology and tBOC chemistry on a Vega 1000 peptide synthesizer as described in section 2.4.1. 1 33 CPPQKED S E E E I R E A F R V F D Q D K S G F I E E D E L K L F L Q N F K CPCQKED S E E E I R E A F R V F D Q D K S G F I E E D E L R H V L T N L G 113 Chapter 4 - Helix Substitution A Met naturally found in position 109 of bovine brain calmodulin (peptide position 29) is replaced with Leu in CPCQKED. The peptides are N-acetylated and C-amidated. 4.3.3. Peptide Purification The peptides CPPQKED, PPCQKED and CPCQKED were purified using H P L C and the molecular weight and purity were confirmed as described in section 2.4.2. 4.3.4. Dimerization Dimerization of the peptides CPPQKED, PPCQKED and CPCQKED was assessed by dissolving lyophilized peptide in a buffer containing 200 m M CaCL., 100 m M KC1 and 50 m M MOPS, pH 7.5 to make stock solutions with concentrations based on peptide mass of; CPPQKED - 15.42 mg/mL, PPCQKED - 14.36 mg/mL, CPCQKED - 15.90 mg/mL. The peptide solutions were prepared and used immediately. Serial dilutions of these stock peptide solutions were made using this same buffer. The peptide concentrations used in the calculations were determined using amino acid analysis as described in section 2.4.5. Each set of solutions were examined using a Jasco J720 spectropolarimeter to generate the dimerization constants KDIM and B and D values as described in section 2.4.3. The mean residue ellipticity values for the C a 2 + bound 94-monomer PCa and Ca bound dimer P2Ca2 for each of the peptides were calculated from the B and D values using equation 7 (page 51). 114 Chapter 4 - Helix Substitution 4.3.5. Ca 2 + Binding Measurement Stock peptide solutions of CPPQKED, PPCQKED, and CPCQKED were prepared by dissolving lyophilized peptide in a buffer containing 50 m M MOPS and 100 m M KC1, pH 7.5. The peptide solutions were prepared and used immediately. The peptide concentrations based on mass were as follows: CPPQKED - 2.00 mg/mL, PPCQKED -2.03 mg/mL, CPCQKED - 2.03 mg/mL. The stock solutions were serially diluted using the buffer resulting in four peptide concentrations each for study. Each set of solutions • 2"t* was titrated with Ca while monitoring at 222 nm with a Jasco J720 spectropolarimeter. The quartz cell used had a pathlength of 1.0 mm. From the titration data the C a 2 + dissociation constants (Ki and K 2 ) and the A constant were determined as described in section 2.4.4 using the simultaneous fit dimer model. The C a 2 + titration data sets are also individually fit to the single site and dimerization model. 4.3.6. Tests of Significance The tests of significance for changes in the constants KDIM, K I , K 2 j and the mean residue ellipticities of the unbound monomer (P), C a 2 + bound monomer (PCa) and dimer forms (P2Ca / P2Ca2) were calculated as described in section 3.3.6. 4.3.7. Molecular Models Molecular models of the peptide dimers were constructed as described in section 3.3.7. Dipole moments were determined using GRASP (Graphical Representation and Analysis of Structural Properties) written by Anthony Nicholls (http ://trantor. bioc. columbi a. edu/grasp/). 115 Chapter 4 - Helix Substitution 4.4. Results 4.4.1. Peptide Purification The HPLC elution profiles of the unpurified peptides PPCQKED, CPPQKED and C P C Q K E D and the CE elution profiles of the purified peptides are given in Appendix A . The molecular weights for the peptides PPCQKED, CPPQKED and C P C Q K E D confirmed using mass spectrometry are in Appendix B. 4.4.2. Dimerization and Ca 2 + Binding The stock peptide concentrations for the dimerization studies determined by amino acid analysis are as follows; PPCQKED - 12.82 ± 0.54 mg/mL, CPPQKED - 14.56 ± 0.27 mg/mL and CPCQKED - 20.34 + 0.15 mg/mL. The plots of peptide concentration versus molar ellipticity and the plots of 6 T O T versus free C a 2 + for PPCQKED, CPPQKED and C P C Q K E D are given in Figure 25 and Figure 26, respectively. CD spectra over the wavelength range 200-250 nm from the dimerization and C a 2 + titration studies are given m Appendix E. The calculated Ca dissociation and dimerization dissociation constants are summarized in Table XII. These dimerization and dissociation constants are expressed in terms of free energy change (AG°) in Table XIII. The parameters A, B, and C / D as well as the mean residue ellipticity values at 222 nm for the four forms, P, PCa, and P 2Ca / P 2 Ca 2 are included in Table XII. The parameters from PPPQKED are also included for comparison purposes. The C a 2 + titration data sets fit individually to the monomer and dimer models are given in Appendices C and D, respectively. 116 Chapter 4 - Helix Substitution CPPQKED CPCQKED 1e-5 1e-4 1e-3 1e-2 1e-5 1e-4 1e-3 1e-2 Total Peptide (M) Total Peptide (M) Figure 25. Plots of molar ellipticity versus total peptide concentration for PPCQKED, CPPQKED and CPCQKED. 117 Chapter 4 - Helix Substitution ro CD T3 PPCQKED 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 Free Ca 2 + (M) CPPQKED CPCQKED 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 Free Ca 2 + (M) ro (U •o E Free Ca 2 + (M) Figure 26. Plot of total titration point ellipticity versus free Ca + concentrations for PPCQKED, CPPQKED and CPCQKED. The Ca 2 + titrations were carried out on four different peptide concentrations with relative peptide concentrations of 2 ( • ) , 1 ( • ) , 0.5 ( A ) and 0.25 ( T ) . 118 Chapter 4 - Helix Substitution Q c OH C H OH •4H s 93 L. 93 OH 00 +1 r-oo +1 o co vo +1 +1 VO U +1 00 T t +1 O r-ON T t co CN +1 ro T t T t VO ro +1 oo 00 ro C O ro +1 ro in CN OO O o o o in ro o © o © +1 +1 +1 +1 +1 VO 00 00 VO oo VO 00 ON ro p T t © 00 © ro CN r—1 T—, T t oo o in © © © © +1 +1 +1 +1 +1 I—1 in VO ro VO 00 ro in T—, p T t CN od O T t V VO +1 ro CN o © +1 00 CN T t © in CN o © +1 ON oo T t © es U u S i w W +i +1 J2 <J O H OH r-~ vo o © © VO © © © +1 +1 +1 © ON ro "I °1 oo ON T—1 © O ' VO © © © +1 +1 +1 CN 00 © T t O O ro 1—1 1—1 od «5 * -.2 0 0 © © ro +1 © © © ro © © ro +1 © © T t © © ro +1 O o ro ro © © ro +1 © © ro T t © © CN +1 © © ^ I-H © © C O +1 o O in vo © © CN +1 © © ON ro © O ro +1 © © in 03 u OH « « a * a © ' S a V o 93 = a 0 0 T 3 © © 00 +1 © © CN CN © © O O +1 © © CN CN © © ON +1 © o r-© © ON +1 © © VO CN 93 U 93 U GO T3 w P. ^ fi 6 0 02 .fi .fi o i "T^ +-> ( H PL, (U o X o ! a . • in TJ + i <D , * ^ f s o oo ° (D w g c o 1 3 a ~ > T3 O 1) • • - I P< ^ 1-, o ™ ^ H p 2 -53 3 o119 Chapter 4 - Helix Substitution Table XIII. Change in Free Energy on Dimerization and Ca 2 + Binding in the Peptides PPPQKED, PPCQKED, CPPQKED and CPCQKED. A ( ; ° ( K M M ) KJ/mol* AG°(K,) KJ/mol A G ° (K2) KJ/mol* PPPQKED -24.27 -19.20 -25.25 PPCQKED -23.65 -13.76 -30.87 CPPQKED -25.80 -16.15 -23.36 C P C Q K E D -24.22 -13.14 -23.84 Values calculated from dissociation constant using equation 3 (page 26). Temperature used in the calculation was 298 °K. 4.4.3. N-terminal Helix Replacement The changes in K D 1 M , K i and K 2 on substitution of the N-terminal PCD helix with that from Cam3 for the peptide pairs PPPQKED/CPPQKED and PPCQKED/CPCQKED are illustrated in Figure 27A. Examination of the peptide pair PPPQKED/CPPQKED reveals that replacement of the PCD N-terminal helix with that from Cam3 decreases the KDIM 46% (p=0.08). This substitution in this peptide pair also increases both K i 242%) (pO.Ol) and K 2 115% (p=0.03). In the peptide pair PPCQKED/CPCQKED, replacement of the PCD N-terminal helix with that from Cam3 decreases the KDIM 21% (p=0.23). In this peptide pair the substitution also increases K i 28% (p=0.04) and increases K 2 at least 1606%). The uncertainty in the increase in the K 2 value arises from the fact that the K i and K 2 values in PPCQKED are >1000 fold apart and the mathematical model used cannot solve such a difference. The difference in A G ° ( A A G ° ) for dimerization is -1.51 KJ/mol in the pair PPPQKED/CPPQKED and -0.57 KJ/mol in the pair 120 Chapter 4 - Helix Substitution A . 100 PPPQKED CPPQKED I PPCQKED ] CPCQKED CD ZS TJ '</> <D O E E u b) 0 TO 14000 12000 10000 8000 A 6000 A 4000 2000 6000 5000 4000 3000 2000 1000 100 B. PCa P2Ca / P 2Ca 2 Figure 27. Bar charts illustrating N-terminal PCD to Cam3 helix substitution. A) The effect on KDIM , K, and K 2. B) The change in the mean residue ellipticity (8) at 222 nm in the peptide forms P -unbound monomer, PCa - Ca 2 + bound monomer, P2Ca / P 2Ca 2 - dimer. The peptide pairs in which this substitution was examined are PPPQKED/CPPQKED and PPCQKED/CPCQKED. 121 Chapter 4 - Helix Substitution PPCQKED/CPCQKED. The AAG° for C a 2 + binding to the monomer is +3.02 in the pair PPPQKED/CPPQKED and +0.61 for the pair KJ/mol PPCQKED/ CPCQKED. In the dimer forms the AAG° for Ca binding to the dimer is +1.88 KJ/mol for the pair PPPQKED/CPPQKED and +6.96 KJ/mol for the pair PPCQKED/CPCQKED. The changes in the mean residue ellipticity of the peptide forms P, PCa and P2Ca/P2Ca2 on substitution of the N-terminal PCD helix with that from Cam3 in the peptide pairs PPPQKED/CPPQKED and PPCQKED/CPCQKED are illustrated in Figure 27B. In the peptide pair PPPQKED/CPPQKED substitution of the N-terminal PCD helix with that from Cam3 results in a 10% increase in the (6)222 of P (p=0.21), a 14%) increase in the (6)222 of PCa (p=0.05) and a 3% decrease in the (6)222 of P2Ca/P2Ca2 (p=0.37). In the peptide pair PPCQKED/CPCQKED substitution of the N-terminal PCD helix with that from Cam3 results in a 11% decrease in the (6)222 of P (p=0.27), a 5% increase in the (6)222 of PCa (p=0.25) and a 5% increase in the (6)222 of P2Ca/P2Ca2 (p=0. 35). 4.4.4. C-terminal Helix Replacement The changes in KDIM, K I and K2 on substitution of the C-terminal PCD helix with that from Cam3 in the peptide pairs PPPQKED/PPCQKED and CPPQKED/CPCQKED are illustrated in Figure 28A. In the peptide pair PPPQKED/PPCQKED, replacing the C-terminal PCD helix with that from Cam3 increases the KDIM 28% (p=0.25). In this same peptide pair, replacing the PCD C-terminal helix with the helix from Cam3 increases K i 800% (p<0.01) and decreases K2 at least 90%. The uncertainty of the decrease in the K2 value arises from the fact that the K i and K2 values in PPCQKED are >1000 fold apart and the mathematical model used cannot solve such a difference. In the 122 Chapter 4 - Helix Substitution A . 100 PPPQKED I I PPCQKED I—I CPPQKED I i CPCQKED o E "9 E o cr> cu 14000 V 12000 $ 10000 8000 6000 4000 <3L- 2000 6000 5000 4000 3000 2000 1000 100 80 60 40 A 20 Figure 28. Bar charts illustrating C-terminal PCD to Cam3 helix substitution. A) The effect on K D i M , K! and K 2. B) The change in the mean residue ellipticity (8) at 222 nm in the peptide forms P - unbound monomer, PCa - Ca 2 + bound monomer, P2Ca / P 2Ca 2 - dimer. The peptide pairs in which this substitution was examined are PPPQKED/PPCQKED and CPPQKED/CPCQKED. 123 Chapter 4 - Helix Substitution peptide pair CPPQKED/CPCQKED, the substitution of the PCD C-terminal helix with the helix from Cam3 increases the K D I M 89% (p=0.02). In this peptide pair the replacement also increases K\ 237% (p<0.01) and decreases K 2 18% (p=0.35). The AAG° for dimerization is +0.62 KJ/mol in the pair PPPQKED/PPCQKED and +1.56 KJ/mol in the pair CPPQKED/CPCQKED. The AAG for C a 2 + binding to the monomer is +5.39 in the pair PPPQKED / PPCQKED and +2.98 for the pair KJ/mol CPPQKED/CPCQKED. In the dimer forms the AAG° for C a 2 + binding to the dimer is -5.56 KJ/mol for the pair PPPQKED/PPCQKED and -0.48 KJ/mol for the pair CPPQKED/CPCQKED. The changes in mean residue ellipticity of the peptide forms P, PCa and P2Ca/P2Ca2 on substitution of the C-terminal PCD helix with that from Cam3 in the peptide pairs PPPQKED/PPCQKED and CPPQKED/CPCQKED are illustrated in Figure 28B. In the peptide pair PPPQKED/PPCQKED substitution of the C-terminal PCD helix with that from Cam3 results in a 22% decrease in the (9) 2 2 2 of P (p=0.05), a 31% decrease in the (9)222 of PCa (pO.Ol) and a 7% decrease in the (9) 2 2 2 of P2Ca/P2Ca2 (p=0.25). In the peptide pair CPPQKED/CPCQKED substitution of the C-terminal PCD helix with that from Cam3 results in a 37% decrease in the (9) 2 2 2 of P (p=0.01), a 36% decrease in the (9) 2 2 2 of PCa (pO.Ol) and no change in the (9) 2 2 2 of P2Ca/P2Ca2. 4.5. Discussion 4.5.1. N-terminal PCD Helix Replacement - Dimerization The replacement of the PCD N-terminal helix with that from Cam3 increases dimerization in both the pairs PPPQKED/CPPQKED and PPCQKED/CPCQKED. 124 Chapter 4 - Helix Substitution However, the negative A A G 0 is three-fold greater for the pair PPPQKED/CPPQKED than for the pair PPCQKED/CPCQKED. Insight into a possible cause for the increase in dimerization comes from examination of structural models based on the N M R structure of the TnC3 dimer (Shaw et al., 1992). Analysis of the sequences reveals that substitution of the N-terminal PCD helix with that of Cam3 introduces an Arg in the -3 position (relative to loop position 1) where previously there was an Ala (See Figure 30 for residue positions). This positively charged Arg could form a salt bridge with a negatively charged Glu in loop position 10 of the other EF-hand (Figure 29). This interaction could act to stabilize the dimer. As this interaction is between the Cam3 N-terminal helix and the invariant loop this increase in dimerization should be observed for both peptides with C-terminal PCD helices (PPPQKED/CPPQKED) and for peptides with C-terminal Cam3 helices (PPCQKED/CPCQKED). This does appear to be the case. The larger increase in dimerization (as reflected in the larger negative AAG 0 ) observed in the peptide pair PPPQKED/CPPQKED could indicate more favorable interactions between the Cam3 N-terminal and PCD C-terminal helices in the dimer. Such an interaction is described in section 4.5.3 between the positively charged Lys in position 21 of the PCD C-terminal helix and a Glu in position -9 of the Cam3 N-terminal helix. Therefore substitution of the PCD N-terminal helix with that from Cam3 introduces two potentially stabilizing interactions when the C-terminal helix is from PCD but only one stabilizing interaction when the C-terminal helix is from Cam3 (Figure 30). This would explain the difference in the magnitude of the AAG° of dimerization in the two peptide pairs. 125 Chapter 4 - Helix Substitution Figure 29. Possible electrostatic interaction between Arg in position -3 and Glu in loop position 10. Backbone model based on the TnC3 dimer (Shaw et al., 1992) showing possible electrostatic interaction between an Arg in position -3 of one monomer and a Glu in position 10 of the other. This interaction would occur in peptides with an N-terminal Cam3 helix, i.e. CPPQKED and CPCQKED. 126 Chapter 4 - Helix Substitution PPPQKED CPPQKED -9 -3 10 21 S A D D V K K A F A I I D Q D K S G F I E E D E L K L F L Q N F K SEEEIREAFRVF D Q D K S G F I E E D E L K L F L Q N F K 4 I PPCQKED CPCQKED S A D D V K K A F A I I D Q D K S G F I E E D E LRHVLTNLG SEEEIREAFRVF D Q D K S G F I E E D E LRHVLTNLG Figure 30. Dimer interaction change on N-terminal PCD to Cam3 helix replacment. Proposed change in electrostatic dimer interactions in the pairs PPPQKED/CPPQKED and PPCQKED/CPCQKED. Arrow indicates interaction between monomers in the dimer. 4.5.2. N-terminal Helix Replacement - Ca 2 + Affinity The N-terminal PCD helix gives higher C a 2 + affinity than the Cam3 N-terminal helix in both the monomer and dimer forms of the peptide pairs PPPQKED/CPPQKED and PPCQKED/CPCQKED. However, the AAG° for the change in C a 2 + affinity in the monomer forms on the PCD to Cam3 N-terminal helix substitution is ~5-fold greater when the C-terminal helix is from PCD then when the C-terminal helix is from Cam3. A n explanation for this discrepancy in the A A G 0 of the PCD to Cam3 helix substitution is illustrated in Figure 31 which shows four models based on one of the EF-hand peptides from the TnC3 dimer. The residue sidechains shown in the ribbon models are in peptide positions -1 and 16 which are located in the N and C terminal helices, respectively. The residue in position -1 of the PCD N-terminal helix is He whereas in the Cam3 N-terminal helix it is 127 Chapter 4 - Helix Substitution Phe. Position 16 is a Phe in the C-terminal helix of PCD and is a Val in the C-terminal helix of Cam3. What is important about these two positions is that the residues can sterically conflict. It should be noted that all the residues listed above can be accommodated. However, it would be expected that the larger Phe side chains would have the greatest frequency of steric conflict. Therefore, when the residues in both positions -1 and 16 are Phe, which only occurs with CPPQKED, one would expect the greatest degree of conflict. This increased steric conflict would result in a greater decrease in C a 2 + affinity in the peptide pair PPPQKED/CPPQKED, than in the pair PPCQKED/CPCQKED, explaining the difference in the AAG° observed in the peptide pairs. Analysis of the K 2 values reveals that the PCD N-terminal helix gives higher C a 2 + affinity than the Cam3 N-terminal helix independent of the nature of the C-terminal helix. However, in the dimer forms the AAG° for the change in C a 2 + affinity on the PCD to Cam3 N-terminal helix substitution is ~4-fold greater when the C-terminal helix is from Cam3. The root of the discrepancy in effect lies in the high dimer C a 2 + affinity of the site PPCQKED. The dimer C a 2 + dissociation constant is >=10-fold lower than any of the other peptides. However, in the monomer form this PPCQKED has the lowest C a 2 + affinity observed in the peptide series. Therefore, the high C a 2 + affinity must be due to the dimerization process. One possibility is conformational changes specific to the dimer of PPCQKED that prevent access of the C a 2 + to the solvent. Whatever the cause, the net result is that the AAG° for the pair PPCQKED/CPCQKED (which have a C-terminal Cam3 helix) is greater than the AAG° for the pair PPPQKED/CPPQKED (which have a PCD C-terminal helix). 128 Chapter 4 - Helix Substitution rj ra a. a. at • 3 (1 •S O c. u B. u •9 jg tu e « .S T H ra _ JS a D 03 3 « .9 d D . O o c o (U o tu 5 , "S S S3 o CL, 6Jj tu -CJ 3 -B O > .9 jo £ 3 o u a u • a 83 a u o u c a. d © 2-Cu U as ON Xi B u m U cu J3 o 4= 0 d o £ o c O 129 Chapter 4 - Helix Substitution Overall the N-terminal PCD helix increases C a 2 + affinity relative to an N-terminal Cam3 helix. Thus some characteristic of the PCD N-terminal helix must be responsible 9+ for enhancing Ca affinity. The mean residue ellipticities at 222 nm change in an inconsistent manner on substitution of the N-terminal PCD helix in the two peptide pairs suggesting greater site stability is not the cause of the increased C a 2 + affinity. Another possibility is alteration of the negative charge density of the Ca binding loop through changes in the helical dipole of the N-terminal helix. The helix dipole arises from the directional alignment of the carbonyl groups and amide hydrogens in the helix backbone resulting in a net positive charge at the N-terminus and a net negative charge at the C-terminus. Differences in the distribution of charged residues along a helix can alter the direction and magnitude of the dipole. Calculation pf the dipole moments of the PCD and Cam3 N-terminal helices reveals differences in direction and magnitude (Figure 32). The PCD N-terminal helix has a small dipole pointing away from the loop. This would introduce a small negative charge into the loop. The Cam3 helix in comparison has a large dipole perpendicular to the helix. The extra negative charge from the PCD N -terminal helix could in theory increase the negative charge density of the loop and increase the electrostatic attraction of the loop for Ca . The perpendicular dipole of the Cam3 N-terminal helix would not have this effect. Therefore, the higher C a 2 + affinity observed for peptides with PCD N-terminal helices could be the result of a dipole induced increase in the negative charge density in the binding loop. 130 Chapter 4 - Helix Substitution PPPQKED Figure 32. Dipole moment of N-terminal PCD and Cam3 Helices. Dipole moment calculated for residues in peptide positions -1 through -13 in the peptides PPPQKED and CPPQKED. Models based on the cc-carbon backbone of TnC site 3 dimer (Shaw et al., 1992). Residues corresponding to PCD are colored blue, Cam3 colored pink and Ca 2 + colored green. Red dipole moment arrow is proportional to magnitude and points in the direction of positive charge. Calculations and modeling were done with GRASP, (http://trantor.bioc.columbia.edu/grasp/). 131 Chapter 4 - Helix Substitution 4.5.3. C-terminal PCD Helix Replacement - Dimerization The PCD C-terminal helix gives significantly higher dimerization than the Cam3 C-terminal helix only when the N-terminal helix is from Cam3 (CPPQKED/CPCQKED). Again, the use of structural models based on the TnC3 dimer can help explain the higher dimerization. The N-terminal Cam3 helix has an acidic Glu residue in N-terminal helix position -9. The PCD C-terminal helix has a positively charged Lys at the C-terminus (position 21) whereas this residue is a Gly for the Cam3 C-terminal helix. Examination of structural models reveals that the Lys in position 21 of the PCD C-terminal helix in one monomer can interact electrostatically with a negatively charged Glu in position -9 of the other monomer (Figure 33). Such an interaction should stabilize the dimer. When the C-terminal PCD helix is replaced with that from Cam3 the Lys is replaced with Gly eliminating the electrostatic interaction and decreasing dimerization. However, the PCD N-terminal helix has a negatively charged Asp in position -9 that could participate in an electrostatic interaction with the Lys in position 21. So why does the PCD C-terminal helix not stabilize the dimer when the N-terminal helix is from PCD? The lack of stabilization when the N-terminal helix is from PCD could be due to the shorter side chain of the Asp in position -9. The distance between the charged groups of an Asp in position -9 and a Lys in position 21 would be greater, resulting in a weaker electrostatic interaction with a Lys in position 21. The loss of a weaker interaction would have less impact on dimerization. 132 CPPQKED Chapter 4 - Helix Substitution Figure 33. Proposed electrostatic interaction between residues in positions -9 and 21. Structural backbone models of CPPQKED and CPCQKED based on the TnC3 dimer (Shaw et al., 1992). The sidechain of the residue in position -9 highlighted in Red (Glu) and in position 21 highlighted in blue (Lys). 133 Chapter 4 - Helix Substitution 4.5.4. C-terminal PCD Helix Replacement - Ca 2 + Affinity Examination of the C a 2 + binding constants for the monomer forms of the peptides reveals that the C-terminal PCD helix confers greater C a 2 + affinity than the Cam3 C-terminal helix independent of the N-terminal helix. However, the AAG° for the change in Ca affinity on C-terminal PCD to Cam3 replacement is ~2-fold greater when the N -terminal helix is from PCD. The cause of this difference could be due to the steric conflict between the N and C-terminal helices proposed for CPPQKED (section 4.5.2 and Figure 31). This steric conflict would act to decrease the C a 2 + affinity of CPPQKED, reducing the impact of C-terminal PCD to Cam3 substitution. The peptides with PCD C-terminal helices have consistently greater negative mean • . . . 94-residue ellipticity in the unbound and Ca bound monomer forms than peptides with Cam3 C-terminal helices. This suggests that in the monomer forms the PCD C-terminal helix is more stable than the helix from Cam3. Increased stability in the PCD C-terminal helix may be responsible for the higher C a 2 + affinity observed in the monomer forms of peptides with PCD C-terminal helices. Increased helix stability can increase Ca z _ r affinity by stabilizing the C a 2 + bound form, enhancing the macrocycle effect (see section 1.4.1). The higher stability of the PCD helix could be related to the positively charged Lys at the C-terminus (position 21). A Lys in this position could stabilize the helix through neutralization of the negative charge at the C-terminus that arises from the helix dipole. Both negatively charged residues at the N-terminus (Huyghues-Despointes et al., 1993) and positively charged residues at the C-terminus of a helix (Sali et al., 1988) have been shown to stabilize helices by neutralizing the helix dipole. 134 Chapter 4 - Helix Substitution Another factor that could influence stability in the PCD C-terminal helix is helix propensity. Helix propensity is a measure of the tendency of a residue to fold into helical structure. Numerous helix propensity scales have been developed in an attempt to quantitate these helix forming tendencies (Blaber et al., 1994; Chakrabartty et al., 1994; Horovitz et al., 1992; O'Neil and DeGrado, 1990). Using four normalized propensity scales, the average helical propensity for the residues in the PCD and Cam3 C-terminal helices are illustrated in Figure 34. From this plot the PCD C-terminal helix appears to have greater overall helix forming tendency. This increased tendency to form helical structure suggests the PCD C-terminal helix would be more stable than the Cam3 C-terminal helix explaining the higher negative mean residue ellipticity observed. 13 14 15 16 17 18 19 20 21 Residue Position Figure 34. Average predicted helical propensity of the PCD and Cam3 C-terminal helices. PCD helix is indicated by solid line and Cam3 helix by dashed line. Propensity scales used are from (Blaber et al., 1994; Chakrabartty et al., 1994; Horovitz et al., 1992; O'Neil and DeGrado, 1990) and are normalized to between 0 and 1. 135 Chapter 4 - Helix Substitution Examination of dimer forms reveals that the Cam3 C-terminal helix gives higher C a 2 + affinity than the PCD C-terminal helix in the peptide pair PPPQKED/PPCQKED. This is opposite to that observed in the monomer forms. In the other peptide pair CPPQKED/CPCQKED there is no change in C a 2 + affinity or negative mean residue ellipticity on PCD C-terminal helix substitution. The lack of a change in the peptide pair CPPQKED/CPCQKED supports the notion that the increase in C a 2 + affinity in the monomer forms is due to site stabilization by the PCD C-terminal helix. The dimer form would be more stable decreasing the effect of C-terminal helix stabilization. The increase in dimer C a 2 + affinity due to the C-terminal Cam3 helix in the pair PPPQKED/PPCQKED can be traced to the high dimer C a 2 + affinity of PPCQKED. As discussed in section 4.5.2 the high C a 2 + affinity of this site is due to dimerization and is only observed when the N-terminal helix is from PCD and the C-terminal helix is from Cam3. Some factor in the dimer, possibly a conformational change that prevents C a 2 + access to the solvent, is the cause of this high C a 2 + affinity. However, such a conformational change is not site stabilizing as the negative mean residue ellipticity of the dimer forms decreases in the pair PPPQKED/PPCQKED. Also, it does not involve interactions that stabilize the dimer as no change in dimerization is observed. 136 Chapter 5 - Non-chelating Loop Residue Substitution CHAPTER 5 NON-CHELATING LOOP RESIDUE SUBSTITUTION IN THE CD SITE OF PARVALBUMIN 5.1. Overview In this chapter, the PCD non-chelating loop residues in peptide positions 2, 4, 10 and 11 were sequentially replaced with those from Cam3. This involved synthesis and study of a series of seven peptides termed PPPQKAD, PPPQKAA, P P P K K A A , P P P K G A A , PPPKGAD, PPPKGED and PPPKKED (see peptide nomenclature page xvn). The dimerization and Ca binding characteristics for each peptide were examined using circular dichroism based techniques. The design of the peptide series allows examination of the non-chelating loop residue substitutions in two different loop environments. Comparison of the dimerization and Ca binding characteristics of the different peptides allows assessment of the effect of each of the non-chelating loop residues in the PCD site model peptide. 5.2. Materials Fmoc Solid Phase Peptide Synthesis • See materials section 4.2. tBoc Solid Phase Peptide Synthesis, Peptide Purification, Peptide Ca 2 + and Dimerization Studies • See materials section 2.3 137 Chapter 5 - Non-chelating Loop Residue Substitution 5.3. Methods 5.3.1. Synthesis a) tBOC Chemistry The peptides P P P K G A A and PPPKGED were synthesized using solid phase methodology and tBOC chemistry on a Vega 1000 peptide synthesizer as described in section 2.4.1. The peptides are N-acetylated and C-amidated. The sequences from the N -terminus are: 1 33 PPPKGAA S A D D V K K A F A I I D K D G S G F I E A A E L K L F L Q N F K PPPKGED S A D D V K K A F A I I D K D G S G F I E E D E L K L F L Q N F K b) Fmoc Chemistry Synthesis of the peptides PPPQKAD, P P P K K A A , PPPKKED, PPPKGAD and P P P Q K A A were carried out on an Applied Biosystems 431 or 433A Peptide synthesizer using Fmoc chemistry as described in section 4.3.1. Synthesis on the Applied Biosystems 431 synthesizer was carried out in the laboratory of Dr. John Sherman (Department of Chemistry, University of British Columbia). The sequences of the synthesized peptides are given below. 1 33 PPPQKAD • S A D D V K K A F A I I D Q D K S G F I E A D E L K L F L Q N F K PPPKKAA S A D D V K K A F A I I D Q D K S K F I E A A E L K L F L Q N F K PPPKKED S A D D V K K A F A I I D K D K S G F I E E D E L K L F L Q N F K PPPKGAD S A D D V K K A F A I I D K D G S G F I E A D E L K L F L Q N F K PPPQKAA S A D D V K K A F A I I D Q D K S G F I E A A E L K L F L Q N F K The peptides are acetylated at the N-terminus and amidated at the C-terminus. 138 Chapter 5 - Non-chelating Loop Residue Substitution 5.3.2. Purification The peptides PPPKGAD, PPKGED, PPPKKED, PPPQKAD, PPPQKAA, P P P K K A A and P P P K G A A were purified using HPLC and had their molecular weight determined and purity confirmed as described in section 2.4.2. 5.3.3. Dimerization Dimerization of the peptides PPPKGAD, PPPKGED, PPPKKED, PPPQKAD, P P P K K A A and P P P K G A A were assessed by dissolving lyophilized peptide in a buffer containing 200 m M CaCl 2 , 100 m M KC1 and 50 m M MOPS, pH 7.5 to make solutions with the following approximate concentrations; PPPKGAD - 16.32 mg/mL, PPPKGED -16.84 mg/mL, PPPKKED - 16.16 mg/mL, PPPQKAD - 17.60 mg/mL, P P P K K A A -17.20 mg/mL, P P P K G A A - 15.60 mg/mL (based on mass). The peptide solutions were prepared and used immediately. The peptide concentrations used in the calculations were determined using amino acid analysis (section 2.4.5). These stock solutions were serial diluted and each set of solutions were used to generate dimerization constants (KDIM) and B and D values as described in section 2.4.3. The mean residue ellipticity values for the C a 2 + bound monomer PCa and C a 2 + bound dimer P 2 C a 2 for each of the peptides were calculated from the B and D values using equation 7 (page 51). A solution of P P P Q K A A was prepared with a concentration of 17.72 mg/mL but no dimerization study was done due to the fact it came out of solution on titration with Ca 2 + . 139 Chapter 5 - Non-chelating Loop Residue Substitution 5.3.4. Ca 2 + Binding Stock peptide solutions of PPPKGAD, PPPKGED, PPPKKED, PPPQKAD, P P P K K A A and P P P K G A A were prepared by dissolving lyophilized peptide in a buffer containing 50 m M MOPS and 100 m M KC1, pH 7.5. The peptide solutions were prepared and used immediately. The peptide stock solution concentrations based on peptide mass were as follows: PPPKGAD - 2.00 mg/mL, PPPKGED - 2.06 mg/mL, PPPKKED - 2.06 mg/mL, PPPQKAD - 2.12 mg/mL, P P P K K A A - 2.09 mg/mL, P P P K G A A - 2.05 mg/mL. The stock solutions were serially diluted and each set of solutions titrated with C a 2 + to generate the C a 2 + dissociation constants (Ki and K 2 ) and PSTOCK values as described in section 2.4.4. The A values were determined as described in section 2.4.4. The mean residue ellipticity values for the unbound monomer P were calculated from the A value using equation 7 (page 51). A stock solution of PPPQKAA was prepared with a concentration of 2.01 mg/mL but came out of solution on C a 2 + titration and therefore the C a 2 + binding was not assessed. The C a 2 + titration data sets were also fit to the single site model and dimer model as described in sections 2.2.2 and 2.2.4. 5.3.5. Significance The tests of significance for changes in the constants KDIM? KI , K 2 , and the mean residue ellipticities of the unbound monomer (P), C a 2 + bound monomer (PCa) and unbound / C a 2 + bound dimer (P2Ca/P2Ca2) were calculated as described in section 3.3.6. 5.3.6. Molecular Models Molecular models of the peptide dimers were constructed as in section 4.3.7. 140 Chapter 5 - Non-chelating Loop Residue Substitution 5.4. Results 5.4.1. Peptide Purification The HPLC elution profiles of the crude peptides and CE elution profiles of the purified peptides for PPPQKAD, PPPQKAA, P P P K K A A , PPPKGAA, PPPKGAD, PPPKGED and PPPKKED are given in Appendix A and the molecular weights confirmed using mass spectrometry are given in Appendix B. 5.4.2. Ca 2 + Binding and Dimerization The stock peptide concentrations for the dimer peptide solutions determined using amino acid analysis are; PPPQKAD - 13.34 ± 0.19 mg/mL, P P P K K A A - 14.27 + 0.39 mg/mL, P P P K G A A 15.38 ± 0.29 mg/mL, PPPKGAD - 13.60 + 0.42 mg/mL, PPPKGED -15.17 ± 0.09 mg/mL and PPPKKED - 12.26 ± 0.01 mg/mL. The plots of peptide concentration vs molar ellipticity for PPPQKAD, P P P K K A A , PPPKGAA, PPPKGAD, PPPKGED and PPPKKED are given in Figure 35. Plots of total titration point ellipticity vs the four free C a 2 + concentration data sets for the peptides studied are given in Figure 36. CD spectra over the wavelength range 200-250 nm from the dimerization and C a 2 + titration studies are given in Appendix E. The peptide PPPQKAA could not be studied 94- 94-because it came out of solution on addition of Ca . The calculated Ca dissociation and dimerization dissociation constants, the parameters A, B, and C/D and the mean residue ellipticity values at 222 nm for the peptide forms P, PCa, and P2Ca / P2Ca2 are summarized in Table XIV. The dimerization and dissociation constants expressed in terms of free energy change (AG°) are given in Table X V . The Ca titration data sets fit individually to the monomer and dimer models are in Appendices C and D, respectively. 141 Chapter 5 - Non-chelating Loop Residue Substitution PPPQKAD -20 , 1 -22 -24 -26 -28 -j | -30 "48 J , , , 1 1e-5 1e-4 1e-3 1e-2 Total Peptide (M) Figure 35. Plots of molar ellipticity versus total peptide concentration for PPPQKAD, PPPKKAA, PPPKGAA, PPPKGAD, PPPKGED and PPPKKED. 142 Chapter 5 - Non-chelating Loop Residue Substitution P P P K G E D -20 , -22 -24 -26 -28 -46 -48 J 1 1 1 1e-5 1e-4 1e-3 1e-2 Total Peptide (M) ure 3 5 . (Cont) P P P K K E D -20 -, -22 -24 --26 --28 • 1e-5 1e-4 1e-3 1e-2 Total Peptide (M) 143 Chapter 5 - Non-chelating Loop Residue Substitution PPPQKAD 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 Free Ca 2 + (M) PPPKKAA PPPKGAA CD (D X> o H 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 Free C a 2 + (M) 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1 Free Ca 2 + (M) Figure 36. Plot of total titration point ellipticity versus free Ca concentrations for PPPQKAD, PPPKKAA, PPPKGAA, PPPKGAD, PPPKGED, and PPPKKED. The Ca 2 + titrations for each peptide were carried out on four different peptide concentrations with relative peptide concentrations of 2 ( • ) , 1 ( • ) , 0.5 ( A ) and 0 .25 ( T ) . 144 Chapter 5 - Non-chelating Loop Residue Substitution Figure 36. (cont.) 145 Chapter 5 - Non-chelating Loop Residue Substitution 3 w P H P M < < P H P H C H Q c P M P H . P H y u o s ca u 93 C-O N +1 C O V O +1 V O W 05 +1 5 T t m +1 O N o C O C O +1 C O T t r-CN V O CN in CN o r—1 in o" © © © +1 +1 +1 +1 +1 C O •—1 V O , — 1 O N V O CN in T t CN © C O oo O N 1—1 C O +1 +1 +1 C O CN C O T—1 oo *—H CN V O T T r—1 +1 +1 +1 wo C O CN 1/0 00 CN O © +1 t -» o T r © © C N © © +1 T t OO C O V O +1 + + 93 u 93 U s w W 05 +1 +1 o a H O H P H as © +1 V O OO C O +1 >/0 C N VO in © +1 O N O N od i in in l—1 l—< V O © © © +1 +1 +1 r~- co © r-^  © r-i C O O N CN 00 O N © © © V O © © © © +1 +1 +1 +1 0^ CN 00 © 00 T t oo C O T t , 4 , 4 od © 0 I o J3 D O O CD © © T t +1 © © T t © © T t +1 © © V O in © © T t +1 © © T t I/O © © C O +1 © © C O T t © © T t +1 © © C O r-© © T t +1 © © V O r-© © >/o +1 © © C N O N © © co +1 © © l /O © © 00 +1 © © oo © © O N +1 © © V O C O © © Os +1 © © r--T t © © O N . +1 © © vo C N 93 u P H 93 U 93 u <U « £j T3 v t d &o 2 -13 CD fi 'oO CD T3 ID •s CD fi CD OO w G O oo fi O o j3 U 00 .fi o I CD Hfi -«-» PH (L> O X CD in + (D I O (D . G <D •B B o ~o CD 3 o "rt o rs 93 U f N P H 93 U f N P H -a 93 U P H on fi ^ <D a) O (D + & O P H fi ^ <D O O O CD W g rt CD > in CD "JU ti — CD l-i a CD ^- bO P , rt -=2 3< CD , 3 •-1 IP O ^ 00 _ CD fcl <p ^ fi O fi 5 S | fi° * <H H - fi 146 Chapter 5 - Non-chelating Loop Residue Substitution Q w o P H P H P H Q < O a ' P H P H OH S-CU •«H = P H O O +1 cn +1 CN 00 i—i o o o ^ H © o +1 +1 +1 o <N VO 00 cn * © +1 ON VO cn +1 VO 00 CN VO m o »—1 VO © o © © +1 +1 +1 +1 +1 VO oo CN o cn cn o © ON CN od o 00 m ON +1 +1 +1 VO vo , — i CN in vo CN © © +1 cn o o +1 5 + + (Ca: (Ca H H tZ5 +1 +1 a H O H P H oo +1 CN VO o vo © +1 00 o +1 vo © +1 VO CN o r -o +1 o in ON 0 O , <S £3 /—^  .5 6o O CD ^ T3 O O +1 O o VO m o o VO +1 o o o o o oo +1 o o cn o o o o o o o in in +1 +1 +1 o o o o o o 1—1 1 in VO ON cn T—1 o o +1 o o cn in o © +1 © © ON © © © +1 © © u P H at U CU § a ' 9 O U OA " CD S T3 o ^ S 'cu I C/3 CD to a o i CD <g -*-> CD o X CD in + -t-» «4H T 3 CD a o t + H 1 3 CD o "c3 o n U CN P i fa °^ * § j ? W 9 + P H + 3 cd - f l a cj_, U S P H -+-> CD cd CO _ " C «^  52 CD CD 0 0 ^ g W C eg CD cj U X ) g ^ & • H (D ^ tD 3 O OH s <g .a .2 l a 1 3 8* VH C CD CD -a ra CD CD H CD ^ s s 6 0 ^ -iJ to O 147 Chapter 5 - Non-chelating Loop Residue Substitution Table XV. Change in Free Energy on Dimerization and Ca Binding in Peptides PPPQKED, PPPQKAD, PPPKKAA, PPPKGAA, PPPKGAD, PPPKGED, and PPPKKED. AG° (K,)m) KJ/mof AG°(K,) KJ/mol* AG° (K2) KJ/mol* PPPQKED -24.27 -19.20 -25.25 PPPQKAD -24.30 -17.63 -23.65 PPPKKAA -25.54 -15.26 -21.27 PPPKGAA -23.98 -17.97 -23.98 PPPKGAD -25.13 -18.16 -25.00 PPPKGED -25.13 -20.21 -25.34 PPPKKED -24.83 -19.24 -25.11 Values calculated from dissociation constant using equation 3. Temperature used in the calculation was 298 °K. 5.4.3. Non-chelating Loop position 2 - Gin to Lys substitution The changes in K D I M , K I and K 2 on a Gin to Lys substitution in non-chelating loop position 2 for the peptide pair PPPQKED/PPPKKED are illustrated in Figure 3 7A. The Q2K substitution results in a 20% decrease in K D I M (p=0.24), a 2% decrease in K i (p=0.45) and a 6% increase in K 2 (p=0.40). The difference in the changes in free energy (AAG°) for dimerization is -0.55 KJ/mol. For the C a 2 + binding to the monomer and dimer forms the AAG° is -0.04 and +0.14 KJ/mol, respectively. The changes in negative mean residue ellipticity for the peptide forms P, PCa and P2Ca/P2Ca2 on the Q2K substitution in the peptide pair PPPQKED / PPPKKED is illustrated in Figure 37B. The Q2K substitution results in a 31% increase in the (6) 2 2 2 of P (p=0.04), a 93% increase in the (9) 2 2 2 of PCa (p<0.01) and a 7% increase in the (9) 2 2 2 of P2Ca/P2Ca2 (p=0.26). 148 Chapter 5 - Non-chelating Loop Residue Substitution Figure 37. Q2K non-chelating loop substitution. A) The effect on K D I M , K< and K 2. B) The effect on mean residue ellipticity of the peptide forms P (monomer), PCa (Ca 2 + bound monomer) and P 2Ca / P 2Ca 2 (dimer forms). The peptide pair examined was PPPQKED / PPPKKED. 149 Chapter 5 - Non-chelating Loop Residue Substitution 5.4.4. Non-chelating Loop Position 4 - Lys to Gly Substitution Substitution of a Lys in non-chelating loop position 4 with a Gly is examined in the peptide pairs P P P K K A A / P P P K G A A and PPPKKED/PPPKGED. The changes in KDIM, KI and K 2 are illustrated in Figure 3 8A. In the peptide pair P P P K K A A / P P P K G A A , the K 4 G substitution increases the KDIM 88% (p=0.09), decreases Ki 66% (p<0.01) and decreases K 2 66% (p=0.01). In the peptide pair P P P K K A A / P P P K G A A the AAG°'s are +1.44 KJ/mol, -2.68 KJ/mol and -2.68 KJ/mol for dimerization, the C a 2 + binding to the monomer and the Ca binding to the dimer, respectively. In the other peptide pair, PPPKKED/PPPKGED, the substitution decreases the KDIM 11% (p=0.38), decreases Ki 33% (p=0.02) and decreases K 2 9% (p=0.40). This corresponds to a AAG° for dimerization of -0.30 KJ/mol and AAG values for C a 2 + binding to the monomer and dimer forms of -0.96 and -0.22 KJ/mol, respectively. The changes in the negative mean residue ellipticity of the peptide forms P, PCa and P2Ca/P2Ca2 on the K 4 G substitution in the peptide pairs P P P K K A A / P P P K G A A and PPPKKED/PPPKGED are illustrated in Figure 38B. In the peptide pair P P P K K A A / P P P K G A A this substitution results in a 16% decrease in the (0) 2 2 2 of P (p=0.10), a 4% decrease in the (0) 2 2 2 of PCa (p=0.31) and a 13% decrease in the (6) 2 2 2 of P2Ca/P2Ca2 (p=0.09). The peptide pair PPPKKED/PPPKGED displays a 9% increase in the (6) 2 2 2 of P (p=0.25), a 17% decrease in the (G) 2 2 2 of PCa (p=0.02) and a 1% increase in the (9) 2 2 2 of P2Ca/P2Ca2 (p=0. 47) on the K 4 G substitution. 150 Chapter 5 - Non-chelating Loop Residue Substitution A. 100 PPPKKAA ! I PPPKGAA PPPKKED PPPKGED o E E o D) <D T3 16000 14000 = 12000 10000 2500 2000 1500 1000 500 250 K 2 B. Figure 38. K4G non-chelating loop substitution. A) The effect on K D I M ) K! and K 2. B) The effect on mean residue ellipticity of the peptide forms P (monomer), PCa (Ca 2 + bound monomer) and P2Ca / P 2Ca 2 (dimer forms). The peptide pairs examined were PPPKKAA / PPPKGAA and PPPKKED / PPPKGED. 151 Chapter 5 - Non-chelating Loop Residue Substitution 5.4.5. Non-chelating Loop Position 10 - Glu to Ala Substitution The Glu to Ala substitution in non-chelating loop position 10 is assessed in the peptide pairs PPPQKED/PPPQKAD and PPPKGED/PPPKGAD. The changes in K D I M , K i and K 2 are illustrated in Figure 39A. In the peptide pair PPPQKED / PPPQKAD, the E10A substitution decreases the K D I M 1% (p=0.48), increases K i 89% (pO.Ol) and increases K 2 91% (p=0.03). This corresponds to a AAG for dimerization of -0.03 KJ/mol and a A A G 0 for Ca binding to the monomer and dimer forms of +1.56 and +1.59 KJ/mol, respectively. The peptide pair PPPKGED/PPPKGAD displays a 33% decrease in K D I M (p=0.22), a 129% increase in K i (pO.Ol) and a 14% increase in K 2 (p=0.38). The corresponding AAG°'s are -0.98 KJ/mol, +2.04 KJ/mol and +0.33 KJ/mol for dimerization, the C a 2 + binding to the monomer and C a 2 + binding to the dimer forms, respectively. The changes in negative mean residue ellipticity of the peptide forms P, PCa and P2Ca/P2Ca2 on the E l OA substitution in the peptide pairs PPPQKED/PPPQKAD and PPPKGAD/PPPKGED are illustrated in Figure 39B. In the peptide pair PPPQKED/PPPQKAD the E l OA substitution results in a 26% increase in the (9)222 of P (p=0.06), a 62% increase in the (9)222 of PCa (pO.Ol) and a 17% increase in the (9)222 of P2Ca / P2Ca2 (p=0.07). The peptide pair PPPKGED/PPPKGAD displays a 13% decrease in the (9)222 of P (p=0.15), a 13% decrease in the (9)222 of PCa (p=0.05) and a 6% increase in the (9)222 of P2Ca/P2Ca2 (p=0. 27) on the E l OA substitution. 152 Chapter 5 - Non-chelating Loop Residue Substitution A. I PPPQKED 1 PPPQKAD I PPPKGED ] PPPKGAD 16000 Figure 39. E10A non-chelating loop substitution. A) The effect on K D I M , K, and K 2. B) The effect on mean residue ellipticity of the peptide forms P (monomer), PCa (Ca 2 + bound monomer) and P2Ca / P2Ca2 (dimer forms). The peptide pairs examined were PPPQKED / PPPQKAD and PPPKGED / PPPKGAD. 153 Chapter 5 - Non-chelating Loop Residue Substitution 5.4.6. Non-chelating Loop position 11 - Asp to Ala substitution Figure 40A illustrates the changes in K D IM, K I and K 2 on an Asp to Ala substitution in non-chelating loop position 11 for the peptide pair PPPKGAD/PPPKGAA. The DI 1A substitution results in a 138% increase in KDIM (p=0.06), an 8% increase in K i (p=0.26) and a 51% increase in K 2 (p=0.10). The AAG° for dimerization is +2.13 KJ/mol and for 94-Ca binding to the monomer and dimer forms the AAG°'s are +0.19 and +1.01 KJ/mol, respectively. The changes in negative mean residue ellipticity for the peptide forms P, PCa and P2Ca/P2Ca2 on the D11A substitution in. the peptide pair P P P K G A D / P P P K G A A are illustrated in Figure 40B. The DI 1A substitution results in a 11% decrease in the (9) 2 2 2 of P (p=0.18), a 7% decrease in the (6) 2 2 2 of PCa (p=0.16) and a 18% decrease in the (9) 2 2 2 of P2Ca/P2Ca2 (p=0.04). 154 Chapter 5 - Non-chelating Loop Residue Substitution Figure 40. D11A non-chelating loop substitution. A ) The effect on K D I M . K I and K 2 . B ) The effect on mean residue ellipticity of the peptide forms P (monomer), PCa (Ca 2 + bound monomer) and P 2Ca / P 2Ca 2 (dimer forms). The peptide pair examined was P P P K G A D / P P P K G A A . 155 Chapter 5 - Non-chelating Loop Residue Substitution 5.5. Discussion 5.5.1. The Peptide P P P Q K A A The peptide PPPQKAA was observed to come out of solution on C a 2 + titration and therefore could not be studied. This peptide did dissolve in the dimerization buffer which has a CaCl2 concentration of 200 mM. The peptide precipitation during C a 2 + titration may be due to the fact that the net charge of the peptide is -1. C a 2 + chelation by the monomer would result in a Ca2+-peptide complex with a net charge of +1. This C a 2 + bound peptide could then bind to an unbound monomer resulting in a complex with a net charge of 0. Aggregation of this complex may follow. Such precipitation following charge neutralization has been observed in soya protein (Virkar et al., 1982). In the dimerization buffer all the peptide would be forced into the C a 2 + bound state and the neutral complex would not form. 5.5.2. Q 2 K Loop Substitution The first non-chelating loop residue examined was in peptide position 2. This residue is a Gin in PCD and a Lys in Cam3. The Q2K substitution changes the side chain volume and alters the charge from neutral to positive. In identified PCD sites Gin is found more than twice as often as Lys (section 1.6.1, Table IV). This is in contrast to EF-hand sites in general in which Lys is 3 times more likely to be found in this position than Gin (Falke et al., 1994) (section 1.3.1, Table I). This preference for Gin in this position in PCD suggests it may be involved in the high C a 2 + affinity observed in the native site. This contention is supported by the observation that an Asn to Lys substitution in this position in the CD site in rat oncomodulin decreased C a 2 + affinity (Palmisano et al., 156 Chapter 5 - Non-chelating Loop Residue Substitution 1990). However, work on synthetic EF-hand sites from Troponin C suggests that a Lys in this position could in fact increase C a 2 + affinity by interacting with the dipole of the N-terminal helix to stabilize it (Shaw et al., 1991b). In contrast to the work on the CD site in oncomodulin and the synthetic EF-hand sites from Troponin C, the Q2K substitution in the peptide pair PPPQKED/PPPKKED appears to have no significant effect on C a 2 + affinity of either the monomer or dimer forms. In addition, this substitution does not appear to significantly alter dimerization. However, the Q2K substitution does appear to increase negative mean residue ellipticity suggesting the Lys may in fact be stabilizing the N-terminal helix through interaction with the dipole. However, i f the Lys increases helix stabilization compared to Gin why does this stabilization not increase C a 2 + affinity? A possible explanation comes from examination of changes in the electrostatics on the Q2K substitution. As stated above, a positively charged Lys in position 2 could potentially negate the partial negative charge from the dipole moment at the C-terminus of the N-terminal helix, stabilizing it. This is evident in Figure 41 which shows dipole neutralization on the Q2K substitution. However, this dipole neutralization also decreases the negative charge density in the loop which would be expected to have a negative impact on Ca affinity. Therefore, any increase in Ca affinity from N-terminal helix stabilization by a Lys in loop position 2 may be negated by the decrease in negative charge density in the loop. If this negation of effect is occurring in the PCD model sites then why was a decrease in Ca affinity observed on a N2K substitution in loop position 2 in the CD site of oncomodulin? This discrepancy in effect can be explained by taking into account the fact that the CD site in the intact oncomodulin would be expected to have greater intrinsic 157 Chapter 5 - Non-chelating Loop Residue Substitution PPPQKED Figure 41. Effect of Q2K substitution on N-terminal helix dipole. Dipole moment calculated for residues in peptide positions 2 through -13 in the peptides PPPQKED and PPPKKED. Models based on the a-carbon backbone of TnC3 dimer (Shaw et al., 1992). Residues corresponding to PCD are colored blue, Cam3 colored pink and Ca 2 + colored green. Red dipole moment arrow is proportional to magnitude and points in the direction of positive charge. Calculations and modeling was done with GRASP. 158 Chapter 5 - Non-chelating Loop Residue Substitution stability. Therefore, any helix stabilization through interaction of a positively charged Lys with the N-terminal helix dipole would be negligible. Only the negative effects of introducing this residue into the loop, namely decreasing negative charge density, would be observed. It is possible that in the CD site in the whole parvalbumin protein, a Lys in 94-non-chelating loop position 2 would have similar detrimental effects on Ca affinity. 5.5.3. K4G Loop Substitution 94-The K 4 G substitution increases the Ca z affinity of the monomer forms in both peptide pairs P P P K K A A / P P P K G A A and PPPKKED/PPPKGED. Examination of the negative mean residue ellipticity changes in the peptide forms P and PCa on the K 4 G replacement reveals that for the most part this substitution decreases the negative mean residue ellipticity or has no effect on it. This indicates that increasing structural stability is most likely not involved in the observed increase in monomer C a 2 + affinity. A n alternative explanation for this increase in monomer C a 2 + affinity involves the fact that the K 4 G substitution eliminates a sidechain and a positive charge from the loop. Elimination of the Lys sidechain on K4G substitution would decrease steric interference 94-as the loop folds around the Ca ion. In addition, a Lys residue in this position would have more restrictive phi/psi angles than a Gly residue. The positive charge of a Lys 2"T" could also decrease Ca affinity by decreasing the negative charge density in the loop. 9.4" • In the dimer forms the K 4 G substitution increases the Ca affinity only in the peptide pair P P P K K A A / P P P K G A A . There is no change in the dimer C a 2 + affinity in the pair PPPKKED/PPPKGED. This discrepancy in effect can be explained by examining the factors influencing C a 2 + affinity in the dimer forms. First, the dimer forms would be 159 Chapter 5 - Non-chelating Loop Residue Substitution stabilized and would not have to "fold" around the Ca 2 + . Therefore the only thing that can alter C a 2 + affinity would be a change in the loop negative charge density. The negative charge density of the loop could potentially be altered by other residues in the loop. The only difference between the pairs P P P K K A A / P P P K G A A and PPPKKED/PPPKGED is the nature of the residues in loop positions 10 and 11. In the pair PPPKKED/PPPKGED these residues are an acidic Glu and Asp, respectively whereas in the pair P P P K K A A / P P P K G A A these residues are Ala. Therefore, the loop regions in the peptide pair PPPKKED/PPPKGED would have a much higher negative charge density and may be less prone to changes in loop charge. In contrast the peptide pair P P P K K A A / P P P K G A A may be more sensitive to changes in the loop negative charge density. This would explain the increase in dimer C a 2 + affinity in the pair P P P K K A A / P P P K G A A and the lack of change in the pair PPPKKED/PPPKGED. The fact that Gly in loop position 4 increases C a 2 + affinity begs the question as to why in the native PCD site, a site with the highest C a 2 + affinity observed in an EF-hand site, this residue is a Lys? Examination of the amino acid frequency in this position in different PCD sites reveals a strong preference for positively charged residues (Lys or Arg) 78% of the time with no instance of Gly (section 1.6.1, Table IV). This is in contrast to EF-hands in general in which Gly is the preferred residue in loop position 4 (Falke et al., 1994) (section 1.3.1, Table I). The conservation of a positively charged residue in this position in PCD suggests that this type of residue must be performing a conserved role in the site. It is observed in the crystal structure of parvalbumin that the side chain of the Lys in position 4 hydrogen bonds to the backbone carbonyls of the Asp in loop position 1 and the Ala of the -3 position of the PCD N-terminal helix. Thus it 160 Chapter 5 - Non-chelating Loop Residue Substitution appears that in the native protein this residue is involved in site stabilization. How can this high degree of conservation and apparent stabilizing role of a Lys in non-chelating loop position 4 be reconciled with fact that it has detrimental effects on C a 2 + affinity in the synthetic peptides? One possible explanation is that the observed increase in C a 2 + affinity on K 4 G substitution may be model site specific. It would be expected that the model site and native site differ in stability. To bind C a 2 + the model site would have to go from a random-coil state and "fold" around the C a 2 + ion incurring a thermodynamic cost. In the parvalbumin protein, the CD site is structured (Williams et al., 1986) and one • • 94-would expect minimal cost of "folding" about the Ca ion. Therefore the thermodynamic cost of folding around the Ca ion in the model site may be larger then any benefit of hydrogen bonding by the Lys in position 4. This would result in the Lys in loop position 4 having detrimental effects on C a 2 + affinity in the model PCD sites. In the 9 + native PCD a Lys in loop position 4 may very well increase Ca affinity. 5.5.4. E10A Loop Substitution The peptide pairs PPPQKED/PPPQKAD and PPPKGED/PPPKGAD examined the effect of a Glu to Ala substitution in non-chelating position 10. The Glu found in PCD 94-loop position 10 does not affect dimerization but does give higher monomer C a Z T affinity 94-than Ala. The higher Ca affinity for Glu in position 10 explains the high degree of conservation of this residue in this position in PCD sites (section 1.6.1, Table IV). In the native site this residue may be partially responsible for the high C a 2 + affinity of the site. In addition, the higher Ca affinity for the Glu in position 10 correlates with the 161 Chapter 5 - Non-chelating Loop Residue Substitution observation that a Glu in loop position 10 in oncomodulin gives higher C a 2 + affinity than a Gly (Palmisano et al., 1990). In the dimer forms of the peptide pairs, the Glu in position 10 gives higher C a 2 + affinity in the pair PPPQKED / PPPQKAD but has little effect on the pair PPPKGED / PPPKGAD. Therefore, it appears that in the dimer form the effect of a Glu in position 10 can be moderated by other factors in the site. The following discussion will examine 94-possible causes for the changes in Ca affinity on the E l OA substitution in the two different peptide forms of the different peptide pairs. The residue in position 10 is both a loop residue and part of the C-terminal helix. Therefore, the higher Ca affinity observed for a Glu in position 10 in the monomer forms could be the result of two different effects. The first possible cause of the higher Ca affinity is that the negatively charged Glu residue contributes a higher negative charge density to the loop than Ala. If increased negative charge in the loop is responsible 94-for the higher monomer Ca affinity then the monomer mean residue ellipticity should not change and the dimer should experience the same changes in Ca affinity. This appears to be what is happening with the peptide pair PPPQKED/PPPQKAD. Not only does the monomer negative mean residue ellipticity not change but it actually increases on the E l OA substitution. In addition, the A A G 0 values for C a 2 + binding are nearly identical for the monomer and dimer forms indicating the E l OA substitution is having the same thermodynamic effect in both forms. The second possible cause of the higher C a 2 + affinity is that the negatively charged Glu is interacting with the dipole of the C-terminal helix to stabilize the N-terminus and 9+ increase Ca affinity through an increased macrocycle effect (see section 1.4.1 for 162 Chapter 5 - Non-chelating Loop Residue Substitution description of macrocycle effect). An uncharged Ala would not interact with the dipole and would provide no stabilization. Figure 42 illustrates the change in the helix dipole on the E l OA substitution. In the peptide with a Glu in loop position 10 the dipole is greater in the direction associated with helix stabilization. If an increase in stabilization was occurring one should see higher monomer negative mean residue ellipticity for a Glu in position 10. This occurs for the peptide pair PPPKGED / PPPKGAD. The dimer form should also have a much smaller change in Ca affinity because it is intrinsically more stable. This is evident in the 6 fold higher AAG° in the monomer forms then in the dimer 94-forms. Thus, it appears that the same E l OA substitution is decreasing the monomer Ca affinity by different mechanisms in the different peptide pairs. The answer as to how this could occur may lie in the residues that differ between the pairs of sites, namely the residues in positions 2 and 4. Examination of the sequences of the different peptide pairs reveals that they differ in the position of a positively charged Lys. In the pair PPPQKED/PPPQKAD it is in loop position 4 and in PPPKGED/PPPKGAD it is in loop position 2. In a model based on the TnC3 dimer the residue in position 2 is in close proximity to the N-terminus of the C-terminal helix (Figure 43). Positive charges at the N-terminus of a helix are in theory destabilizing. In the monomer forms of the peptide pair PPPKGED/PPPKGAD the Lys in position 2 could potentially destabilize the N-terminus of the C-terminal helix by neutralizing the negative charges at the N-terminus of the C-terminal helix. 163 Chapter 5 - Non-chelating Loop Residue Substitution PPPQKED Figure 42. Effect of E10A substitution on C-terminal helix dipole. Dipole moment calculated for residues in peptide positions 2 through -13 in the peptides PPPQKED and PPPQKAD. Models based on the cc-carbon backbone of TnC3 dimer (Shaw et al., 1992). Residues corresponding to PCD are colored blue, Cam3 colored pink and Ca 2 + colored green. Red dipole moment arrow is proportional to magnitude and points in the direction of positive charge. Calculations and modeling were done with GRASP. 164 Chapter 5 - Non-chelating Loop Residue Substitution Figure 43. Relative position of non-chelating loop residues 2 and 4 and the N-terminus of C-terminal Helix. Structural model showing the position of a positively charged Lys in position 2 and 4 (blue) in relation to the concentration of negatively charged residues at the N-terminus of the C-terminal helix (red). Models based on the a-carbon backbone of TnC3 dimer (Shaw et al., 1992). In the peptide pair PPPQKED/PPPQKAD the residue in position 2 is a neutral Gin. This residue would have no effect on the charge at the N-terminus of the C-terminal helix. The net effect would be less negative charge at the N-terminus of the C-terminal helix for the pair PPPKGED/PPPKGAD compared to PPPQKED/PPPQKAD. Thus, the C-terminal helix in the peptide P P P K G A D may not be fully stabilized and thus an additional negative charge in loop position 10 (as in PPPKGED) would increase stability. In contrast, the peptide PPPQKAD may be fully stabilized (there is no Lys in position 2) 165 Chapter 5 - Non-chelating Loop Residue Substitution and thus, the extra negative charge from a Glu in loop position 10 (PPPQKED) would enhance Ca affinity by only increasing the loop negative charge density. 5.5.5. DI 1A Loop Substitution Examination of the peptide pair PPPKGAD/PPPKGAA reveals that an Asp in position 11 in the PCD site promotes dimerization compared to Ala. This could be due to the presence of an electrostatic interaction between the Asp in position 11 and a Lys in position -7 of the N-terminal helix (Figure 44). Figure 44. Proposed electrostatic interaction in PPPKGAD dimer. Backbone model of dimer showing the position of the negatively charged Asp in peptide position 11 (red) and positively charged Lys in position -7 (blue). Model based on the a-carbon backbone of TnC3 dimer (Shaw et al., 1992). 166 Chapter 5 - Non-chelating Loop Residue Substitution The D11A substitution would break this electrostatic interaction decreasing dimerization. However, in the previous section an E l OA substitution had no significant effect on dimerization yet a Glu in position 10 could also potentially interact with the Lys in position -7. This can be explained i f only one electrostatic interaction is occurring though both are possible. Thus removal of one (e.g. in the peptides PPPQKED/PPPQKAD) would still allow the other to occur and no change in dimerization would be observed. However, removal of the final interaction in P P P K G A D / P P P K G A A would result in an observable decrease in dimerization. A n Asp in position 11 gives modestly higher C a 2 + affinity than Ala in the peptide pair P P P K G A D / P P P K G A A in both the monomer and dimer forms. As described for a Glu in position 10, an Asp in this position can potentially increase C a 2 + affinity through increased negative charge density in the loop and/or stabilization of the C-terminal helix. Stabilization by an Asp in position 11 appears to be minimal as reflected in the mean residue ellipticities of the monomer forms. This leaves increased negative charge density by the Asp in position 11 as responsible for the higher Ca affinity. If increased negative charge density in the loop was responsible for the increase in calcium affinity then one should see an higher dimer Ca affinity for the site with an Asp in position 11 9+ with no change in dimer mean residue ellipticity. In fact, the Ca affinity of the dimer form is observed to be greater with an Asp in position 11. However, the negative mean residue ellipticity is also greater. Perhaps this increase in dimer negative mean residue ellipticity is due to dimer stabilization by the electrostatic interaction between the Asp in position 11 and the Lys in position -7. 167 Chapter 6 - Structural Studies CHAPTER 6 NMR AND CRYSTALLIZATION STUDIES ON THE CARP PARVALBUMIN 4.25 CD SITE MODEL PEPTIDE 6.1. Overview The two methods used to obtain detailed structural information on peptides and proteins are nuclear magnetic resonance (NMR) spectroscopy and x-ray crystallography. This chapter presents preliminary structural analysis of the peptide PPPQKED using N M R and x-ray crystallographic techniques. The N M R study involves examination of the change in a ID H-NMR spectra on addition of Ca to a peptide solution. The x-ray crystallographic studies present the crystallization and initial diffraction results for PPPQKED. As of the writing of this chapter the structure of the peptide has not been determined by either method. 6.2. NMR Theory A detailed description of protein N M R techniques can be found in Cavanagh et al., (1995) and Markley and Opella, (1997). N M R spectroscopy is based on the fact that certain atomic nuclei such as ! H , 1 3 C , 1 5 N and 3 I P have a magnetic moment or spin. When molecules containing these atoms are placed in a strong magnetic field the spins of the atoms align with the magnetic field. Application of a radio frequency (RF) pulse to a sample in such a magnetic field will alter the spin equilibrium raising the nuclei to an excited state. As the nuclei return to the equilibrium state they emit RF radiation that can be measured. The frequency of the emitted radiation is dependent on the molecular 168 Chapter 6 - Structural Studies environment of the atoms in the molecule. Fourier transform techniques are applied to the measured RF radiation to generate a plot of the component frequencies that make up the RF radiation. The frequencies of the different peaks relative to a reference signal are termed chemical shifts. The magnitude of a peak with a particular chemical shift is dependent on the number of atoms that share that chemical environment. For example, the protons of the CH3 side chain in Ala will have the same environment and thus the peak will have three times the intensity of the hydrogen on the a-carbon on the same amino acid. In proteins the most commonly examined nuclei are ' H . In theory it is possible to obtain a unique chemical shift for every hydrogen atom in a protein (except for those that are chemically equivalent) though in practice the differences between the chemical shifts can be very small. This results in spectra with multiple overlapping peaks. To complicate spectra, one peak can be split into multiple peaks by neighboring nuclei. Two dimensional (2D) N M R techniques have been developed to resolve these complicated spectra. One such 2D technique is termed COSY (Correlation Spectroscopy) which gives information about which nuclei are covalently connected through one or two atoms (Aue et al., 1976; Bax and Freeman, 1981; Nagayama et al., 1980). Another 2D technique is termed N O E S Y (Nuclear Overhauser Effect Spectroscopy) and gives information about hydrogen atoms that are coupled through space (Kumar et al., 1980). Using such 2D techniques (and techniques not discussed here) it is possible to get detailed structural information on proteins such as hydrogen bonding and secondary structural characteristics. The information gained can be also used to generate distance 169 Chapter 6 - Structural Studies constraints for hydrogen atoms in the protein. Using such distance constraints three-dimensional structures can be constructed. 6.3. X-ray Crystallography Theory X-ray crystallography is used to determine the three-dimensional structure of proteins in the crystal state. A detailed description of x-ray crystallographic techniques can be found in McRee, (1993). The first step in the technique involves growing a protein crystal usually by bringing a solution of protein to a state of super-saturation. Once a sufficiently large crystal has been obtained, x-ray analysis can begin. When a crystal is subjected to a narrow x-ray beam, diffraction of the beam occurs. Diffraction occurs because the protein molecules in the crystal are arranged in a regular three dimensional lattice. The x-rays interact with the electrons in each atom of each molecule in the lattice. This interaction causes the atoms to oscillate which in turn results in the emission of new x-rays. These x-rays are emitted in all directions and most will cancel. However, the regular arrangement of the molecules in the crystal results in some of the emitted x-rays adding together in certain directions resulting in diffracted beams that can be recorded as a pattern of spots of varying intensity. The diffraction pattern can be recorded on film or an electronic area detector. The spacing of the spots in the diffraction pattern can be related to the crystal lattice using Braggs Law (Bragg, 1931a; Bragg, 1931b). This law states that diffraction by a crystal can be regarded as reflection of the primary beam by a set of parallel planes that pass through the unit cell. The unit cell is the basic building block of the crystal and can consist of one or more molecules. X-rays diffracted from different planes will travel different distances and when the difference is equal to a multiple of the wavelength of the 170 Chapter 6 - Structural Studies x-ray beam, constructive interference will occur. The relationship between the distance between planes d, the reflection angle 0 and the wavelength X is given by the following equation. 2dsin6 = X (44) Using this equation and the positions of the diffraction spots, the unit cell dimensions can be determined. Information on the three dimensional structure of the protein that makes up the crystal is contained in the intensity pattern of the diffraction spots. Each intensity spot is the result of interference from the x-rays emerging at the same angle from all atoms in the molecule. Solving the structure from the diffraction pattern requires knowledge of the amplitude, wavelength and phase of each diffraction spot. The amplitude can be measured from the intensity of the spot and the wavelength is set by the x-ray source. However, the phase of each spot is lost in the x-ray experiment. Two techniques are commonly used to get around the "phase problem". One is termed multiple isomorphous replacement and involves introduction of a new x-ray scatterer into the unit cell of a crystal. This is done by diffusing a heavy metal into the crystal. Measurement of the diffraction pattern of this crystal and comparison with the original diffraction pattern allows estimation of the phase of each intensity from the original protein crystal. Using these phases and the diffraction data collected, an electron density map of the unit cell can be constructed. The protein can then be fit to this electron density map to produce a three dimensional structure. The accuracy of this fit can be estimated by a R-factor which for most structures varies between 0.15 and 0.20, with a smaller value indicating a more optimal fit to the diffraction data. 171 Chapter 6 - Structural Studies Another technique used to overcome the phase problem is termed molecular replacement. Molecular replacement requires a previously solved protein structure similar to the one under study. The difficulty in this technique is placing the phasing structure in the unit cell in the proper orientation and position to generate the proper phases. In theory a search of every possible position and orientation can be done but in practice this is impractical. To simplify the search, determination of the best position and best orientation are done separately. The search for orientation is done using Patterson maps (Patterson, 1935) of the phasing model. These maps are matched to the Patterson map of the protein under study. Once the rotational orientation has been determined, the search for the translational position can be initiated. The initial orientation of the model determined by this search is then refined until a minimal R-factor is obtained. 6.4. Materials PPPQKED Crystallization • Magnesium chloride hexahydrate, polyethylene glycol 400-4000, calcium chloride dihydrate and 4-(2-hydroxyethyl)piperazine-l-ethanesulfonic acid sodium salt (HEPES-Na) were obtained from Fluka Chemie A G , Buchs, Switzerland. • The peptide PPPQKED was prepared and purified as described in sections 2.4.1 and 2.4.2. 172 Chapter 6 - Structural Studies PPPQKED NMR • D 4 imidazole was obtained from Aldrich Chemical Company, St. Louis, M O , USA. • D 2 0 was from MSD Isotopes, Rahway, NJ, USA. • The calcium chloride dihydrate and potassium chloride were obtained from B D H Chemicals Canada Limited, Toronto, ON, Canada. • 3-(Trimethylsilyl)-l-propanesulfonic acid (DSS) was from Sigma Chemical Co., St. Louis, M O , USA. • Chelex (sodium form, 200-400 Mesh) was from Bio-Rad Laboratories, Hercules, C A , USA. 6.5. Methods 6.5.1. NMR Nuclear magnetic resonance (NMR) analysis of the peptide PPPQKED was carried out using a Varian Unity 500 MHz instrument in the laboratory of Dr. Lawrence P. Mcintosh (Departments of Biochemistry and Molecular Biology and Department of Chemistry, University of British Columbia) with the assistance of Dr. Carolyn M . Slupsky. A 1 m M solution of the peptide PPPQKED was prepared in a buffer with the following composition: 50 m M KC1, 30 m M D 4 Imidazole, pH 7.2 in D 2 0 . The peptide solution was treated with Chelex as described by Shaw et al. (1991b). 1.0 g of Chelex® was added to 4 mL of H 2 0 and 14.92 mg of KC1 and allowed to equilibrate for 10 minutes. The Chelex® was filtered through a sintered glass funnel and washed three times with ~0.5 mL of D 2 0 . The Chelex® was suspended in 3 mL of D 2 0 and 300 uL of 173 Chapter 6 - Structural Studies this suspension was added to the peptide solution. This solution was heated at 63° for 45 minutes. The suspension was centrifuged and the peptide solution transferred to a clean 94-microcentrifuge tube. A Ca solution was prepared with the following composition: 0.05 M CaCl 2 , 50 mM KC1 and 30 m M D 4 Imidazole, pH 7.2 in D 2 0 . The ID ' H N M R spectra was recorded for 500 uL of the peptide solution with 10 uL of 0.1 M DSS added as a standard. To the peptide solution, 52 uL of the Ca solution was added and the ID ' H N M R spectra recorded. The final concentration of CaCl 2 was 4.71 mM. The spectra were obtained at a temperature of 30 °C with a spectral width of 6000 Hz and a 1 sec acquisition time. For each spectra 12K data points were collected and plotted with the same vertical scale. 6.5.2. X-ray Crystallography The initial screen for crystallization of PPPQKED was carried out using selected conditions outlined in the Crystal Screen™ kit from Hampton Research, (Laguna Hills, CA, USA). To prepare the crystallization buffers a stock 2X buffer of 0.2 M NaHEPES, 0.2 M CaCl 2 , pH 7.5 was prepared. This 2X buffer was used to prepare 5 mL of each of the following crystallization buffers with compositions as follows. A 0.1 M NaHEPES, 0.1 M CaCl 2 0.2 M M g C l 2 , 30% v/v 2-Propanol B 0.1 M NaHEPES, 0.1 M CaCl 2 28% v/v PEG 400 C 0.1 M NaHEPES, 0.1 M CaCl 2 0.2 M M g C l 2 , 30% v/v PEG 400 D 0.1 M NaHEPES, 0.1 M CaCl 2 10% v/v 2-Propanol, 20% w/v PEG 4000 174 Chapter 6 - Structural Studies Crystallization was done in the laboratory of Dr. Gary Brayer (Department of Biochemistry and Molecular Biology, University of British Columbia). The peptide solution was prepared by dissolving 1 mg of PPPQKED in 50 uL of a 50% solution of the 2X buffer for a final concentration of ~20 mg/mL. Crystallization was attempted by the hanging drop method as described below. Samples of peptide solution (2 to 3 uL) were placed on siliconized (Sigmacote, Sigma Chemical Co., St. Louis, M O , USA.) microscope cover slips. To the peptide droplets, 2 uL of one of the available crystallization buffers was added and mixed. Each cover slip was then placed inverted over a well in a Linbro plate (Flow Laboratories, McLean, V A , USA.) in which lmL of crystallization buffer had been added. The well was sealed by placing high vacuum silicon grease (Dow Corning Co., Midland, MI, USA) on the lip of the well prior to cover slip placement. Irregular star shaped crystals appeared in wells with crystallization buffer C after 5 weeks. Using buffer C conditions, the dependence of crystallization on PEG 400 was examined. A series of six crystallization buffers were prepared with the composition 0.1 M NaHEPES, 0.1 M CaCl 2 , 0.2 M M g C l 2 pH 7.5 and a PEG 400 concentration that varied from 5% to 30% in 5% increments. The peptide solutions were plated as described above and crystals grew in the PEG range 25%-30% after 1 month. Some crystals appeared to be well formed but subsequent x-ray diffraction analysis revealed they were twinned. At this point crystallization was taken over by Nham Thi Nguyen in the laboratory of Dr. Gary Brayer (Department of Biochemistry and Molecular Biology, University of British Columbia). Crystallization of the peptide was successfully repeated using the techniques outlined above and a crystallization buffer with a composition of 0.1 175 Chapter 6 - Structural Studies M NaHEPES, 0.1 M CaCl 2 , 0.2 M M g C l 2 , pH 5.5 with PEG 4000 18-20% as the precipitant. The crystals used in the x-ray diffraction analysis were grown from a crystal that had been broken and hair seeded. The crystals were allowed to grow for six months. X-ray diffraction was carried out in the laboratory of Dr. Gary Brayer (Department of Biochemistry and Molecular Biology, University of British Columbia) by Nham Thi Nguyen and Dr. Gunnar Olovsson from which the following information was obtained. A crystal was mounted in a thin walled glass capillary (diameter - 0.7 mm) with mother liquor placed above and below the crystal and sealed with dental wax. The capillary was placed on a goniometer head and diffraction data was collected on a Rigaku R-AXIS IIC imaging plate area detector system using CuKa radiation supplied by a Rigaku RU300 rotating anode generator operating at 50 kV and 100 mA. Exposure time for each frame was 45 minutes and 200 frames were collected. Data frames were collected using an oscillation angle of 1.0°. The crystal to detector distance was set to 70 mm. Data was collected over 6.25 days. The diffraction data intensities were integrated via a profile fitting algorithm using the software H K L (Otwinowski and Minor, 1997). The amount of background radiation was determined from the intensity directly surrounding each measured and integrated peak. The individual diffraction intensities on each frame were corrected for background, Lorentz, and polarization effects, then scaled to compensate for crystal decay and absorption effects. A n estimate of the absolute scale for this diffraction data set was obtained using the CCP4 suite of software (Collaborative Computational Project Number 4, 1994) and the method of Wilson (1942). 176 Chapter 6 - Structural Studies 6.6. Results 6.6.1. NMR The N M R spectra of PPPQKED in the absence and presence of C a 2 + are shown in Figure 4 5 . A) B) Figure 45. 500 MHz *H NMR spectra of PPPQKED. A) In the absence of Ca 2 +. B) In the presence of 4.71 mM CaCl2. Resonance that may be due to (3-sheet formation in the dimer is indicated by ( • ) . This peak is of a similar magnitude and chemical shift as observed for the cc-carbons of the residues involved in the p-sheet in the TnC3 dimer (Shaw et al., 1991a; Shaw et al., 1990). 177 Chapter 6 - Structural Studies 6.6.2. X-ray Crystallography Crystallization trials for the peptide PPPQKED resulted in the growth of well formed crystals, of which two examples are illustrated in Figure 46. Figure 46. Crystals of PPPQKED. Crystal forms observed for the synthetic peptide PPPQKED grown using the hanging drop method and crystallization buffer with a composition of 0.1 M CaCl2, 0.1 M NaHEPES, 0.2 M MgCl 2 , pH 7.5 and PEG 400 in the range of 25-30%. 178 Chapter 6 - Structural Studies The x-ray diffraction data collection parameters for a PPPQKED crystal are summarized irt Table X V I . A sample diffraction pattern from this crystal is given in Figure 47. Table XVI. Data collection parameters for PPPQKED crystal. Parameter Space group C2 Crystal system monoclinic Cell dimensions (A) a, b, c 46.3,28.1,49.0 Cell angles (deg) - a, (3, y 90, 96.7, 90 Number of measurements 79318 Number of unique reflections 7107 Mean IIol * 20.6 Completeness (%) 96.3 Merging R-factor (%) ** 8.0 Resolution (A) 1.7 The average reflection intensity divided by the average standard deviation of the intensity. A measure of the level of agreement among the different data frames after scaling. 179 Chapter 6 - Structural Studies Figure 47. Sample diffraction pattern from a PPPQKED crystal. Space group for this crystal is C2 with the following unit cell parameters: a=46.3 A, b=28.1 A, c=49.0 A, oc=90°, (3=96.7°, y=90°. Resolution is 1.7 A with a merging R-factor of 8.0%. 180 Chapter 6 - Structural Studies 6.7. Discussion 6.7.1. NMR The N M R spectra of PPPQKED in the absence of C a 2 + (Figure 45 A) has narrow spectral lines and displays groups of resonances that fall into regions of chemical shift characteristic of "random coil" peptides (Bundi and Wuthrich, 1979). However, as indicated in the previous chapters the peptides studied have a measured mean residue ellipticity in the absence of C a 2 + in the range -3000 to -6000 deg-cm2-dmor]-residue"1. It is possible that the structure observed in these peptides is due to factors such as residual Ca which is estimated from 0.1 to 1 u M in buffers (Linse and Forsen, 1995). However, if this was the case then one would expect peptides with higher C a 2 + affinity to have higher apo (absence of Ca 2 + ) negative mean residue ellipticities. Examination of the apo mean residue ellipticities and K\ binding constants for the peptides in the non-chelating loop residue study (Table XIV) reveals that with the exception of PPPKGED all the peptides have higher apo negative mean residue ellipticities and lower C a 2 + affinities than PPPQKED. Thus it appears that the mean residue ellipticities of the apo forms of the peptides must be influenced by other factors. It is quite possible that in the absence of C a 2 + the peptides adopt transitory structures detectable with circular dichroism but buried in the complex N M R spectra. Addition of Ca to the PPPQKED results in a dramatic change in the N M R spectra (Figure 45B) with a broadening of the spectra lines. This loss of resolution indicates that 2D N M R techniques would be difficult for this peptide under these experimental conditions. In spite of this, the spectra does provide some indication of the folding process of PPPQKED on the addition of C a 2 + . , Note the appearance of a small peak 181 Chapter 6 - Structural Studies between 5.0 and 5.2 ppm indicated by a • in Figure 45B. In studies by Shaw et al. on the TnC3 dimer, peaks of similar magnitude in this range have been assigned to oc-carbon protons in residues that form the small p-sheet in the dimer (Shaw et al., 1991a; Shaw et al., 1990). It is quite possible that the appearance of this peak, though small, could also be due to P-sheet formation. The small size of the peak can be explained by the fact it represents 1-3 protons out of 269 total protons in each monomer. Other peaks of note that appear on addition of Ca include the two between 0.0 and 0.4 ppm. In the N M R spectra of the TnC3 dimer, resonances in this area were assigned to CH3 groups from He residues in loop position 8 and C-terminal helix position 16 (Shaw et al., 1991a). The peptide PPPQKED does possess an He in loop position 8 so it would be safe to assume that one of these resonances arises from this residue. The identity of the other resonance remains unknown until resonance assignments can be made. 6.7.2. X-ray Crystallography Successful crystallization and x-ray diffraction analysis of PPPQKED is the first step in determining the structure of this peptide. Only one other EF-hand peptide, a synthetic 34 residue TnC3 peptide, has been successfully crystallized and analyzed using x-ray crystallographic techniques (Delbaere et al., 1989). The TnC3 crystal, though an EF-hand like PPPQKED, crystallizes in a different space group -141. This contrasts with the C2 space group of PPPQKED. The crystallographic 2-fold axis of the I4i space group led Delbaere et al. (1989) to conclude that the peptide might be present as a dimer. Unfortunately, the structure was not solved by x-ray diffraction methods. This structure was subsequently determined using N M R techniques and was indeed shown to be a 182 Chapter 6 - Structural Studies dimer (Shaw et al., 1992). Currently, work is proceeding in the laboratory of Dr. Gary Brayer to determine the structure of PPPQKED from the collected diffraction data using molecular replacement techniques. The structural models being used in the molecular replacement include the TnC3 dimer determined using N M R (Shaw et al., 1992) as well as single site and two site domain fragments from the crystal structures of carp parvalbumin 4.25, troponin C and calmodulin. 183 Conclusions CONCLUSIONS The goal of this study was to identify factors responsible for the high C a 2 + affinity of the native parvalbumin CD site using model single site EF-hand peptides. Specifically the role of the flanking helices and non-chelating loop residues in the PCD site were examined. The use of single site peptides to examine C a 2 + binding to an EF-hand site eliminates factors associated with the whole protein that can complicate interpretation of the results. This includes cooperative interactions between paired EF-hand sites as well as distal conformational, hydrogen bonding and electrostatic interactions. However, 94-analysis of Ca binding to EF-hand peptides is complicated by the fact they dimerize. To address this phenomenon, a mathematical model was developed for this study that 2_i_ described the Ca binding process taking into account dimerization. This "dimer model" was found to more completely describe the C a 2 + titration data from the synthetic EF-hand peptides studied than a model that does not take dimerization into account. Once a model was developed to describe the observed C a 2 + titration data, factors that influence Ca affinity in the synthetic PCD model site were investigated. The first observation was that the model PCD site had 105-fold lower C a 2 + affinity than the native 94-site. This is lower then expected. In addition, it did not bind Mg . This stresses the • 9 + 94-importance of the whole protein in the Ca affinity and Mg binding ability of the site. 94-Also it indicates that the Ca affinity of synthetic EF-hands does not necessarily reflect C a 2 + affinity in the native site. The lower than expected C a 2 + affinity and loss of M g 2 + binding in the model PCD site could be due to electrostatic repulsion by the - X Glu in the loop region. In theory a - X Glu should increase C a 2 + affinity through displacement of a 184 Conclusions water molecule. However, electrostatic repulsion could negate this effect. Repulsion by the - X Glu in the PCD model site as the cause of the lower than expected C a 2 + affinity is supported by the observation that replacement of the - X Glu with an Asp in the model PCD site increases C a 2 + affinity. Increasing site stability would decrease the negative 94-effects of repulsion on Ca affinity. This explains why, in the CD site of the protein oncomodulin which would be expected to have greater site stability than the model PCD site, a - X Glu gives higher Ca affinity than Asp. Increased site stability would decrease the effect of repulsion until the benefits of direct chelation by the - X Glu, namely displacement of a water molecule, would be observed as increased C a 2 + affinity. Though the model PCD site has lower C a 2 + affinity than expected, it did demonstrate Ca affinity at least -10 fold higher than a peptide based on Cam3 with the chelating residues of PCD (Procyshyn and Reid, 1994a). This indicated that the flanking helices and/or the non-chelating loop residues were enhancing Ca affinity in the PCD model site. To investigate the role on the PCD helices and non-chelating loop residues, synthetic chimeras of PCD and Cam3 were synthesized and studied for their dimerization and C a 2 + binding characteristics. Within the PCD model site it was observed that the N-terminal PCD helix gives higher C a 2 + affinity than the N-terminal helix from Cam3 in both the monomer and dimer 94-forms. The higher Ca affinity attributed to the PCD N-terminal helix may be due to a 94-helix dipole induced increase in the negative charge density in the Ca binding loop. In the monomer forms the magnitude of the increase in Ca affinity from the N-terminal PCD helix appears to be dependent upon the identity of the C-terminal helix. Analysis of structural models based on the TnC3 dimer structure suggest that this difference in 185 Conclusions magnitude could be the result of a steric conflict between the N-terminal Cam3 helix and PCD C-terminal helix. • 2"T" * • The C-terminal PCD helix gave higher Ca affinity in the monomer forms compared to Cam3. The higher C a 2 + affinity attributed to the PCD C-terminal helix in the monomer forms is accompanied by a higher negative mean residue ellipticity, suggesting increased structural stability. This increased structural stability could be the result of a stabilizing Lys at the C-terminus of the PCD C-terminal helix and/or a greater helical propensity of the residues that make up the C-terminal PCD helix. In the dimer forms, which would be expected to be more stable, substituting the PCD C-terminal helix for Cam3 results in no change in Ca affinity when the N-terminal helix is from Cam3. In addition, no change in mean residue ellipticity is observed. This supports the contention that the higher Ca affinity attributed to the PCD C-terminal helix in the monomer forms is the result of increased helix stability. However, when the N-terminal helix is from PCD, a Cam3 C-terminal helix gives higher C a 2 + affinity than a PCD C-terminal helix. This reversal of effect can be traced to the very high dimer C a 2 + affinity 9+ of the peptide PPCQKED. The monomer of this peptide has low Ca affinity therefore this increase in C a 2 + affinity is the result of dimerization. The high C a 2 + affinity of the 9+ dimer form could be due to conformational changes that prevent access of the Ca to the solvent. In addition to the effects on Ca affinity, the helices also altered dimerization. Both the Cam3 N-terminal helix and the PCD C-terminal helix promote dimerization. This increase in dimerization can be explained using dimer models based on TnC3 which demonstrate electrostatic interactions present only when these helices are present. 1 8 6 Conclusions The four non-chelating loop residues influence C a 2 + affinity to varying degrees. The Gin in loop position 2 from PCD does not alter dimerization or C a 2 + binding to the monomer or dimer forms compared to a Lys from Cam3. The Lys found in PCD loop position 4 appears to have negative effects on C a 2 + affinity in the monomer forms compared to a Gly from Cam3. However, in the dimer forms the Lys in position 4 has 2+ negative effects on Ca affinity only when the residues in loop positions 10 and 11 are Ala. The negative effects of Lys in position 4 in the monomer forms could be the result of restrictive phi/psi angles for this residue on folding around the C a 2 + ion and/or a decrease in negative charge density in the loop. In the dimer forms it may be only the decrease in negative charge density involved which would explain why no change in C a 2 + affinity is observed when the residues in positions 10 and 11 are the negatively charged Glu and Asp found in PCD. Negatively charged residues in loop positions 10 and 11 would increase the net charge of the loop potentially making it less sensitive to the introduction of a positively charged Lys in loop position 4. This Lys in position 4 appears to have no significant effect on dimerization compared to Gly. The Glu found in loop position 10 of PCD promotes C a 2 + affinity in the monomer forms but promotes C a 2 + affinity in the dimer forms only when the residues in loop positions 2 and 4 are from PCD. In the monomer forms the Glu in loop position 10 could promote Ca affinity by increasing C-terminal helix stability and/or increasing negative charge density in the loop. In the more stable dimer forms the increase in C a 2 + affinity may be due only to increasing negative charge density. Analysis of the dimer models reveals that the negative charge from a Glu in loop position 10 could potentially be neutralized by a positively charged residue in loop position 2. This would explain why a 187 Conclusions 9+ Glu in position 10 increases the Ca affinity of the dimer forms only when the residue in position 2 is from PCD (i.e. a neutral Gin). When the residue in position 2 is from Cam3 (i.e. a positively charged Lys) the increasing negative charge density could be negated. The Asp in loop position 11 of PCD appears to increase C a 2 + affinity a'small amount in both the monomer and dimer forms. Dimerization also increases with an Asp in position 11. This increase in dimerization could be due to an electrostatic interaction between the Asp in position 11 and a Lys in the N-terminal helix. No change in the monomer mean residue ellipticity is observed on Asp substitution with Ala in position 11. Therefore, the higher C a 2 + affinity observed for the monomer form of the peptide with an Asp in position 11 could be due to an increase in the negative charge density in the loop. The higher Ca affinity observed for an Asp in position 11 in the dimer form is accompanied by higher dimer negative mean residue ellipticity. The higher negative mean residue ellipticity could be due to the electrostatic stabilization proposed to increase dimerization. From work on the model PCD site it appears that both the flanking helices and the non-chelating loop residues in positions 10 and 11 could be partially responsible for the high Ca affinity of the native parvalbumin CD site. However, knowing which factors increasing Ca affinity in the model PCD site are applicable to the whole protein 9+ requires examination of the effect on Ca affinity of the helical and non-chelating loop residue substitutions outlined in this study in the native parvalbumin CD site. In this study structural models based on the TnC3 dimer were used to interperet changes in dimerization and C a 2 + affinity. Though the models were useful, the actual structures may differ from those used to explain the observed changes in dimerization 188 Conclusions and C a 2 + binding. In an attempt to get a more accurate model of the sites used in this study, structural analysis of the model PCD site using the techniques of N M R and x-ray crystallography has begun. So far preliminary work using N M R on the synthetic PCD site indicate that these peptides may be dimerizing on addition of Ca 2 + . Unfortunately the resolution N M R spectra of the Ca bound form of the peptide is not good enough to permit more detailed analysis. Crystals of the model PCD site have been successfully grown and diffraction data has been collected for them. Determination of the peptide structure from this data is, at the time of writing, proceeding using molecular replacement techniques. 189 Future Studies FUTURE STUDIES In this study, synthetic EF-hand peptides were used to identify factors responsible 94-for the high Ca affinity of the native parvalbumin CD site using model single site EF-hand peptides. To do this the flanking helix and non-chelating loop residues of PCD were replaced with those of Cam3. However, the work was limited to examining intact flanking helices and only selected combinations of non-chelating loop residues. In addition, the structural model used to interpret the changes, though useful, were based on a different EF-hand site, TnC3. Finally, the relevance of the work in the synthetic peptides to the intact parvalbumin protein is not known. Some specific projects that would address the limitations of the present work are outlined below. 1. Helical Residues - The effect on C a 2 + affinity of individual residues in both the N and C terminal helices of PCD and Cam3 could be studied. In the Cam3 N-terminal helix the Arg in peptide position -3 and the Phe in position -1 could be replaced with the residues found in these positions in the PCD N-terminal helix and their effect on dimerization and C a 2 + affinity assessed. In the PCD C-terminal helix, the Lys found at the C-terminus could be replaced with the Gly found in the Cam3 C-terminal helix and the effect on Ca affinity and helix stability assessed. 2. Loop Residues - In the loop region of the PCD model peptide different combinations of non-chelating loop residues from PCD and Cam3 could be examined to further explore interactions that could be occurring between these residues. 190 Future Studies 3. Structural Models - A structural model of PCD should be determined using NMR or x-ray crystallography techniques. This would improve the interpretations presented in this and any future studies. Another specific peptide to determine the structure of would be PPCQKED. Determination of the structure of the dimer form of this 9-4-peptide could answer why Ca does not dissociate from it. 4. Whole Protein Model - The changes made to the model sites in this study could be examined in a whole protein model. The whole parvalbumin protein would probably not be a good candidate due to its high Ca 2 + affinity. A more suitable model would involve insertion of the PCD site into a protein with overall lower Ca 2 + affinity such as calmodulin. This would allow the effects to be more easily monitored. Specific things to examine include any substitution that affected model site stability. 5. Kinetic Studies - The peptides examined in this study could be examined for Ca 2 + binding kinetics, specifically the on and off rates. This could be done using a stopped-flow apparatus. The information could enhance the understanding of the 94-mechanism behind the changes in Ca affinity observed. 6 . Peptide Stability - The stability of the peptides in this study could be examined using thermal or guanidine H C 1 denaturation to confirm that site stability is involved in some of the Ca 2 + affinity changes observed. 191 References REFERENCES Ahmed, F. R., Przybylska, M . , Rose, D. R., Bimbaum, G. I., Pippy, M . E. and MacManus, J. P. (1990). Structure of oncomodulin refined at 1.85 A resolution. A n example of extensive molecular aggregation via Ca2+. J Mol Biol 216, 127-40. Akke, M . , Drakenberg, T. and Chazin, W. J. (1992). Three-dimensional solution structure of Ca(2+)-loaded porcine calbindin D9k determined by nuclear magnetic resonance spectroscopy. Biochemistry 31, 1011-20. Akke, M . , Forsen, S. and Chazin, W. J. (1995). Solution structure of (Cd2+)1-calbindin D9k reveals details of the stepwise structural changes along the Apo~>(Ca2+)IIl-->(Ca2+)I,II2 binding pathway. J Mol Biol 252, 102-21. Altschul, S. F., Madden, T. L. , Schaffer, A . A . , Zhang, J., Zhang, Z., Miller, W. and Lipman, D. J. (1997). Gapped B L A S T and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 3389-402. Anderson, J. M . and Cormier, M . J. (1978). Calcium-dependent regulation of N A D kinase. Biochem Biophys Res Commun 84, 595-602. Ashley, C. C , Mulligan, I. P. and Lea, T. J. (1991). Ca2+ and activation mechanisms in skeletal muscle. Q Rev Biophys 24, 1-73. Aue, W. P., Bartholdi, E. and Ernst, R. R. (1976). Two-dimensional spectroscopy. Application to nuclear magnetic resonance. J. Chem. Phys. 64, 2229-46. Azuma, T. and Hamaguchi, K. (1976). The mechanism of reassembly of immunoglobulin G. J Biochem (Tokyo) 80, 1023-38. 192 Babu, A. , Su, H. , Ryu, Y . and Gulati, J. (1992). Determination of residue specificity in the EF-hand of troponin C for Ca2+ coordination, by genetic engineering [published erratum appears in J Biol Chem 1992 Nov 15;287(32):23434]. J Biol Chem 267, 15469-74. Babu, Y . S., Bugg, C. E. and Cook, W. J. (1988). Structure of calmodulin refined at 2.2 A resolution. J Mol Biol 204, 191-204. Babu, Y . S., Sack, J. S., Greenhough, T. J., Bugg, C. E., Means, A . R. and Cook, W. J. (1985). Three-dimensional structure of calmodulin. Nature 315, 37-40. Baker, E. N . and Hubbard, R. E. (1984). Hydrogen bonding in globular proteins. Prog Biophys Mol Biol 44, 97-179. Balendiran, K. , Tan, Y . , Sharma, R. K. and Murthy, K. H. (1995). Preliminary crystallization studies of calmodulin-dependent protein phosphatase (calcineurin) from bovine brain. Mol Cell Biochem 149-150, 127-30. Barany, G. and Merrifield, R. B. (1980). "The Peptides: Analysis, Synthesis, Biology.". Academic Press, New York, New York. Barbato, G., Ikura, M . , Kay, L . E., Pastor, R. W. and Bax, A . (1992). Backbone dynamics of calmodulin studied by 15N relaxation using inverse detected two-dimensional N M R spectroscopy: the central helix is flexible. Biochemistry 31, 5269-78. Baron, G., Demaille, J. and Dutruge, E. (1975). The distribution of parvalbumins in muscle and in other tissues. FEBS Lett 56, 156-60. Baudier, J., Glasser, N . and Gerard, D. (1986). Ions binding to S100 proteins. I. Calcium-and zinc-binding properties of bovine brain SI 00 alpha alpha, SI 00a (alpha beta), 193 and SI00b (beta beta) protein: Zn2+ regulates Ca2+ binding on SI00b protein. J Biol Chem 261,8192-203. Bax, A . and Freeman, R. (1981). Investigation of complex networks of spin-spin coupling by two-dimensional N M R . J. Magn. Reson. 44, 542-61. Beckingham, K. (1991). Use of site-directed mutations in the individual Ca2(+)-binding sites of calmodulin to examine Ca2(+)-induced conformational changes. J Biol Chem 266, 6027-30. Berchtold, M . W. (1989). Structure and expression of genes encoding the three-domain Ca2+- binding proteins parvalbumin and oncomodulin. Biochim Biophys Acta 1009, 201-15. Berridge, M . J. (1990). Calcium oscillations. J Biol Chem 265, 9583-6. Betzel, C , Pal, G. P. and Saenger, W. (1988). Synchrotron X-ray data collection and restrained least-squares refinement of the crystal structure of proteinase K at 1.5 A resolution. Acta Crystallogr B 44, 163-72. Bindels, R. J., Timmermans, J. A. , Hartog, A . , Coers, W. and van Os, C. H . (1991). Calbindin-D9k and parvalbumin are exclusively located along basolateral membranes in rat distal nephron [published erratum appears in J A m Soc Nephrol 1992 Mar;2(9):1461]. J Am Soc Nephrol 2, 1122-9. Blaber, M . , Zhang, X . J., Lindstrom, J. D., Pepiot, S. D., Baase, W. A. and Matthews, B. W. (1994). Determination of alpha-helix propensity within the context of a folded protein. Sites 44 and 131 in bacteriophage T4 lysozyme. J Mol Biol 235, 600-24. Blanchard, H. , Grochulski, P., L i , Y . , Arthur, J. S., Davies, P. L. , Elce, J. S. and Cygler, M . (1997). Structure of a calpain Ca(2+)-binding domain reveals a novel EF-hand 194 References and Ca(2+)-induced conformational changes [see comments]. Nat Struct Biol 4, 532-8. Bodanszky, M . (1984). "Principles of Peptide Synthesis". Springer-Verlag, Berlin. Bode, W. and Schwager, P. (1975). The refined crystal structure of bovine beta-trypsin at 1.8 A resolution. II. Crystallographic refinement, calcium binding site, benzamidine binding site and active site at pH 7.0. J Mol Biol 98, 693-717. Borin, G., Pezzoli, A . , Marchiori, F. and Peggion, E. (1985). Synthetic and binding studies on the calcium binding site I of bovine brain calmodulin. II. Synthesis and CD studies on the cyclic 20-31 sequence. Int JPept Protein Res 26, 528-38. Bragg, W. L . (1931a). The diffraction of short electromagnetic waves by a crystal. Proc. Cambridge Phil. Soc. 17, 43. Bragg, W. L . (1931b). The reflection of X-rays by crystals. Proc. Roy. Soc. Lond. A38, 428. Brahms, S. and Brahms, J. (1980). Determination of protein secondary structure in solution by vacuum ultraviolet circular dichroism. J Mol Biol 138, 149-78. Briedigkeit, L . and Frommel, C. (1989). Calcium-ion binding by thermitase. FEBS Lett. 253,83-7. Brodersen, D. E., Etzerodt, M . , Madsen, P., Celis, J. E., Thogersen, H . C , Nyborg, J. and Kjeldgaard, M . (1998). EF-hands at atomic resolution: the structure of human psoriasin (S100A7) solved by M A D phasing. Structure 6, 477-89. Brodersen, D. E., Nyborg, J. and Kjeldgaard, M . (1999). Zinc-binding site of an S100 protein revealed. Two crystal structures of Ca2+-bound human psoriasin (S100A7) in the Zn2+-loaded and Zn2+-free states. Biochemistry 38, 1695-704. 195 References Brostrom, M. A., Brostrom, C. O. and Wolff, D. J. (1978). Calcium-dependent adenylate cyclase from rat cerebral cortex: activation by guanine nucleotides. Arch Biochem Biophys 191, 341-50). Brown, I. D. (1988). What factors determine cation coordination numbers? Acta Crystallogr., Sect. B 44, 545-53. Bundi, A. and Wuthrich, K. (1979). Proton NMR parameters of the common amino acid residues measured in aqueous solutions of the linear tetrapeptides H-Gly-Gly-X-L-Ala-OH. Biopolymers 18, 285. Buschrnann, H.-J. (1986). The macrocyclic and cryptate effects. Complexation of alkali ions by monocyclic and bicyclic ligands in methanol. Inorg. Chim. Acta 125, 31-5. Buschrnann, H.-J. (1986a). The macrocyclic and cryptate effects. Influence of structural changes on the complexation behavior of aza crown ethers and cryptands in different solvents. Inorg. Chim. Acta 120, 125-9. Carlstrom, G. and Chazin, W. J. (1993). Two-dimensional 1H nuclear magnetic resonance studies of the half-saturated (Ca2+)1 state of calbindin D9k. Further implications for the molecular basis of cooperative Ca2+ binding. J Mol Biol 231, 415-30. Cavanagh, J., Fairbrother, W. J., Palmer, A. G. and Skelton, N. J. (1995). "Protein NMR Spectroscopy: Principles and Practice". Academic Press, San Diego, U.S.A. Chakrabartty, A., Kortemme, T. and Baldwin, R. L. (1994). Helix propensities of the amino acids measured in alanine-based peptides without helix-stabilizing side-chain interactions. Protein Sci 3, 843-52. 196 References Chalovich, J. M . , Chock, P. B. and Eisenberg, E. (1981). Mechanism of action of troponin . tropomyosin. Inhibition of actomyosin ATPase activity without inhibition of myosin binding to actin. J Biol Chem 256, 575-8. Chattopadhyaya, R., Meador, W. E., Means, A . R. and Quiocho, F. A . (1992). Calmodulin structure refined at 1.7 A resolution. J Mol Biol 228, 1177-92. Chazin, W. J., Kordel, J., Drakenberg, T., Thulin, E., Brodin, P., Grundstrom, T. and Forsen, S. (1989). Proline isomerism leads to multiple folded conformations of calbindin D9k: direct evidence from two-dimensional 1H N M R spectroscopy. Proc Natl Acad Sci U S A 86, 2195-8. Christakos, S., Gabrielides, C. and Rhoten, W. B. (1989). Vitamin D-dependent calcium binding proteins: chemistry, distribution, functional considerations, and molecular biology. Endocr Rev 10, 3-26. Clark, I. D., Hi l l , I., Sikorska-Walker, M . , MacManus, J. P. and Szabo, A . G. (1993). A novel peptide designed for sensitization of terbium (III) luminescence. FEBS Lett 333, 96-8. Coffee, C. J. and Bradshaw, R. A . (1973). Carp muscle calcium-binding protein. I. Characterization of the tryptic peptides and the complete amino acid sequence of component B. J Biol Chem 248, 3302-12. Cohen, P., Burchell, A. , Foulkes, J. G. and Cohen, P. T. (1978). Identification of the Ca2+-dependent modulator protein as the fourth subunit of rabbit skeletal muscle phosphorylase kinase. FEBS Lett 92, 287-93. Cox, B. G. and Schneider, H. (1992). "Coordination and transport properties of macrocyclic compounds in solution.". Elsevier, Amsterdam. 197 References Cox, J. A . , Milos, M . and MacManus, J. P. (1990). Calcium- and magnesium-binding properties of oncomodulin. Direct binding studies and microcalorimetry. J Biol Chem 265, 6633-7. da Silva, A . C , de Araujo, A . H. , Herzberg, O., Moult, J., Sorenson, M . and Reinach, F. C. (1993). Troponin-C mutants with increased calcium affinity. Eur J Biochem 213, 599-604. Dabrowska, R., Sherry, J. M . , Aromatorio, D. K. and Hartshorne, D. J. (1978). Modulator protein as a component of the myosin light chain kinase from chicken gizzard. Biochemistry 17, 253-8. Declercq, J. P., Tinant, B., Parello, J., Etienne, G. and Huber, R. (1988). Crystal structure determination and refinement of pike 4.10 parvalbumin (minor component from Esox lucius). JMol Biol 202, 349-53. Declercq, J. P., Tinant, B., Parello, J. and Rambaud, J. (1991). Ionic interactions with parvalbumins. Crystal structure determination of pike 4.10 parvalbumin in four different ionic environments. J Mol Biol 220, 1017-39. Delbaere, L . T., Vandonselaar, M . and Reid, R. E. (1989). Crystallization of a synthetic analog of calcium-binding site III of rabbit skeletal troponin C. J Biol Chem 264, 10261-3. Dotson, D. G. and Putkey, J. A . (1993). Differential recovery of Ca2+ binding activity in mutated EF-hands of cardiac troponin C. JBiol Chem 268, 24067-73. Drabikowski, W., Brzeska, H. and Venyaminov, S. (1982). Tryptic fragments of calmodulin. Ca2+- and Mg2+-induced conformational changes. J Biol Chem 257, 11584-90. 198 References Drake, S. K. , Lee, K. L . and Falke, J. J. (1996). Tuning the equilibrium ion affinity and selectivity of the EF-hand calcium binding motif: substitutions at the gateway position. Biochemistry 35, 6697-705. Drake, S. K. , Zimmer, M . A. , Miller, C. L. and Falke, J. J. (1997). Optimizing the metal binding parameters of an EF-hand-like calcium chelation loop: coordinating side chains play a more important tuning role than chelation loop flexibility. Biochemistry 36, 9917-26. Drakenberg, T., Hofmann, T. and Chazin, W. J. (1989). 1H N M R studies of porcine calbindin D9k in solution: sequential resonance assignment, secondary structure, and global fold. Biochemistry 28, 5946-54. Drohat, A . C , Amburgey, J. C , Abildgaard, F., Starich, M . R., Baldisseri, D. and Weber, D. J. (1996). Solution structure of rat apo-S100B(beta beta) as determined by N M R spectroscopy. Biochemistry 35, 11577-88. Durussel, I., Pauls, T. L. , Cox, J. A . and Berchtold, M . W. (1996). Chimeras of parvalbumin and oncomodulin involving exchange of the complete CD site show that the Ca2+/Mg2+ specificity is an intrinsic property of the site. Eur J Biochem 242,256-63. Eberhard, M . and Erne, P. (1994). Calcium and magnesium binding to rat parvalbumin. Eur J Biochem 222, 21-6. Einspahr, H . and Bugg, C. E. (1984). "Metal Ions in Biological Systems". Dekker, Basel, Switzerland. Falke, J. J., Drake, S. K. , Hazard, A . L . and Peersen, O. B. (1994). Molecular tuning of ion binding to calcium signaling proteins. Q Rev Biophys 27, 219-90. 199 References Falke, J. J., Snyder, E. E., Thatcher, K. C. and Voertler, C. S. (1991). Quantitating and engineering the ion specificity of an EF-hand-like Ca2+ binding. Biochemistry 30, 8690-7. Findlay, W. A. , Sonnichsen, F. D. and Sykes, B. D. (1994). Solution structure of the TR1C fragment of skeletal muscle troponin-C. J Biol Chem 269, 6773-8. Finn, B. E., Evenas, J., Drakenberg, T., Waltho, J. P., Thulin, E. and Forsen, S. (1995). Calcium-induced structural changes and domain autonomy in calmodulin. Nat Struct Biol 2, 777-83. Finn, B . E., Kordel, J., Thulin, E., Sellers, P. and Forsen, S. (1992). Dissection of calbindin D9k into two Ca(2+)-binding subdomains by a combination of mutagenesis and chemical cleavage. FEBS Lett 298, 211-4. Forsen, S., Linse, S., Drakenberg, T., Kordel, J., Akke, M . , Sellers, P., Johansson, C , Thulin, E., Andersson, I., Brodin, P. and et al. (1991). Ca2+ binding in proteins of the calmodulin superfamily: cooperativity, electrostatic contributions and molecular mechanisms. Ciba Found Symp 161, 222-36. Freifelder, D. (1976). "Physical Biochemistry". W.H. Freeman and Co., San Francisco, C A . Fullmer, C. S. and Wasserman, R. H . (1977). In "Calcium Binding Proteins and Calcium Function" (R. H . Wasserman, R. A . Corradino, E. Carafoli, R. H . Kretsinger, D. H . MacLennan and S. F X . , eds), pp. 303-12. Elsevier North Holland, Inc., New York, New York. 200 References Gagne, S. M . , L i , M . X . and Sykes, B. D. (1997). Mechanism of direct coupling between binding and induced structural change in regulatory calcium binding proteins. Biochemistry 36, 4386-92. Gagne, S. M . , Tsuda, S., L i , M . X . , Smillie, L . B. and Sykes, B. D. (1995). Structures of the troponin C regulatory domains in the apo and calcium- saturated states. Nat Struct Biol 2, 784-9. Gariepy, J., Sykes, B. D., Reid, R. E. and Hodges, R. S. (1982). Proton nuclear magnetic resonance investigation of synthetic calcium- binding peptides. Biochemistry 21, 1506-12. George, S. E., Su, Z., Fan, D. and Means, A . R. (1993). Calmodulin-cardiac troponin C chimeras. Effects of domain exchange on calcium binding and enzyme activation. JBiol Chem 268, 25213-20. George, S. E., Su, Z., Fan, D., Wang, S. and Johnson, J. D. (1996). The fourth EF-hand of calmodulin and its helix-loop-helix components: impact on calcium binding and enzyme activation. Biochemistry 35, 8307-13. Golosinska, K. , Pearlstone, J. R., Borgford, T., Oikawa, K., Kay, C. M . , Carpenter, M . R. and Smillie, L . B. (1991). Determination of and corrections to sequences of turkey and chicken troponins-C. Effects of Thr-130 to He mutation on Ca2+ affinity. J Biol Chem 266, 15797-809. Greaser, M . L . and Gergely, J. (1973). Purification and properties of the components from tropinin. J Biol Chem 248, 2125-33. Gros, P., Betzel, C , Dauter, Z., Wilson, K. S. and Hoi, W. G. (1989). Molecular dynamics refinement of a thermitase-eglin-c complex at 1.98 A resolution and 201 References comparison of two crystal forms that differ in calcium content. J Mol Biol 210, 347-67. Haiech, J., Derancourt, J., Pechere, J. F. and Demaille, J. G. (1979). Magnesium and calcium binding to parvalbumins: evidence for differences between parvalbumins and an explanation of their relaxing function. Biochemistry 18, 2752-8. Haiech, J., Kilhoffer, M . C , Lukas, T. J., Craig, T. A. , Roberts, D. M . and Watterson, D. M . (1991). Restoration of the calcium binding activity of mutant calmodulins toward normal by the presence of a calmodulin binding structure. J Biol Chem 266,3427-31. Haiech, J., Klee, C. B. and Demaille, J. G. (1981). Effects of cations on affinity of calmodulin for calcium: ordered binding of calcium ions allows the specific activation of calmodulin- stimulated enzymes. Biochemistry 20, 3890-7. Han, C. H . and Roberts, D. M . (1997). Altered methylation substrate kinetics and calcium binding of a calmodulin with a Vall36~>Thr substitution. Eur J Biochem 244, 904-12. Hanahan, D. J., Taverna, R. D., Flynn, D. D. and Ekholm, J. E. (1978). The interaction of Ca2+/Mg2+ ATPase activator protein and Ca2+ with human erythrocyte membranes. Biochem Biophys Res Commun 84, 1009-15. Hapak, R. C , Lammers, P. J., Palmisano, W. A. , Birnbaum, E. R. and Henzl, M . T. (1989). Site-specific substitution of glutamate for aspartate at position 59 of rat oncomodulin. J Biol Chem 264, 18751-60. 202 Hathaway, D. R. and Adelstein, R. S. (1979). Human platelet myosin light chain kinase requires the calcium-binding protein calmodulin for activity. Proc Natl Acad Sci USA 76, 1653-7. Henzl, M . T., Hapak, R. C. and Goodpasture, E. A . (1996). Introduction of a fifth carboxylate ligand heightens the affinity of the oncomodulin CD and EF sites for Ca2+. Biochemistry 35, 5856-69. Henzl, M . T., Hapak, R. C. and Likos, J. J. (1998). Interconversion of the ligand arrays in the CD and EF sites of oncomodulin. Influence on Ca2+-binding affinity. Biochemistry 37,9101-11. Herzberg, O. and James, M . N . (1985). Common structural framework of the two Ca2+/Mg2+ binding loops of troponin C and other Ca2+ binding proteins. Biochemistry 24, 5298-302. Herzberg, O. and James, M . N . (1988). Refined crystal structure of troponin C from turkey skeletal muscle at 2.0 A resolution. J Mol Biol 203, 761-79. Holroyde, M . J., Robertson, S. P., Johnson, J. D., Solaro, R. J. and Potter, J. D. (1980). The calcium and magnesium binding sites on cardiac troponin and their role in the regulation of myofibrillar adenosine triphosphatase. J Biol Chem 255, 11688-93. Horovitz, A. , Matthews, J. M . and Fersht, A . R. (1992). Alpha-helix stability in proteins. II. Factors that influence stability at an internal position. J Mol Biol 227, 560-8. Houdusse, A . , Love, M . L. , Dominguez, R., Grabarek, Z. and Cohen, C. (1997). Structures of four Ca2+-bound troponin C at 2.0 A resolution: further insights into the Ca2+-switch in the calmodulin superfamily. Structure 5, 1695-711. 203 References Hudson, D. (1988). Methodological implications of simultaneous solid-phase peptide synthesis. 1. Comparison of different coupling procedures. J Org Chem 53, 617-24. Hurley, J. B. (1993). Visual transduction. Sensing calcium in rod cells [news; comment]. Nature 361,20-1. Huyghues-Despointes, B. M . , Scholtz, J. M . and Baldwin, R. L. (1993). Effect of a single aspartate on helix stability at different positions in a neutral alanine-based peptide. Protein Sci 2, 1604-11. Ikura, M . (1996). Calcium binding and conformational response in EF-hand proteins. Trends Biochem Sci 21, 14-7. Ikura, M . , Clore, G. M . , Gronenborn, A . M . , Zhu, G., Klee, C. B. and Bax, A . (1992). Solution structure of a calmodulin-target peptide complex by multidimensional N M R . Science 256, 632-8. Ikura, M . , Spera, S., Barbato, G., Kay, L. E., Krinks, M . and Bax, A . (1991). Secondary structure and side-chain 1H and 13C resonance assignments of calmodulin in solution by heteronuclear multidimensional N M R spectroscopy. Biochemistry 30, 9216-28. Ingraham, R. H . and Swenson, C. A . (1984). Binary interactions of troponin subunits. J Biol Chem 259, 9544-8. Johansson, C , Ullner, M . and Drakenberg, T. (1993). The solution structures of mutant calbindin D9k's, as determined by N M R , show that the calcium-binding site can adopt different folds. Biochemistry 32, 8429-38. 204 References Johnson, J. D., Collins, J. H. , Robertson, S. P. and Potter, J. D. (1980). A fluorescent probe study of Ca2+ binding to the Ca2+-specific sites of cardiac troponin and troponin C. J Biol Chem 255, 9635-40. Kakalis, L . T., Kennedy, M . , Sikkink, R., Rusnak, F. and Armitage, I. M . (1995). Characterization of the calcium-binding sites of calcineurin B. FEBS Lett 362, 55-8. Kanellis, P., Yang, J., Cheung, H. C. and Lenkinski, R. E. (1983). Synthetic peptide analogs of skeletal troponin C: fluorescence studies of analogs of the low-affinity calcium-binding site II. Arch Biochem Biophys 220, 530-40. Kay, C. M . , McCubbin, W. D. and Sykes, B. D. (1987). Biophysical studies on the calcium trigger of muscle contraction. Biopolymers 26, SI 23-44. Kay, L . E., Forman-Kay, J. D., McCubbin, W. D. and Kay, C. M . (1991). Solution structure of a polypeptide dimer comprising the fourth Ca(2+)- binding site of troponin C by nuclear magnetic resonance spectroscopy. Biochemistry 30, 4323-33. Kesvatera, T., Jonsson, B., Thulin, E. and Linse, S. (1994). Binding of Ca2+ to calbindin D9k: structural stability and function at high salt concentration. Biochemistry 33, 14170-6. Kilhoffer, M . C , Kubina, M . , Travers, F. and Haiech, J. (1992). Use of engineered proteins with internal tryptophan reporter groups and pertubation techniques to probe the mechanism of ligand-protein interactions: investigation of the mechanism of calcium binding to calmodulin. Biochemistry 31, 8098-106. 205 Kissinger, C. R., Parge, H. E., Knighton, D. R., Lewis, C. T., Pelletier, L . A. , Tempczyk, A. , Kalish, V . J., Tucker, K. D., Showalter, R. E., Moomaw, E. W. and et al. (1995). Crystal structures of human calcineurin and the human FKBP12-FK506-calcineurin complex. Nature 378, 641-4. Klee, C. B. , Crouch, T. H . and Richman, P. G. (1980). Calmodulin. Annu Rev Biochem 49, 489-515. Kordel, J., Forsen, S. and Chazin, W. J. (1989). 1H N M R sequential resonance assignments, secondary structure, and global fold in solution of the major (trans-Pro43) form of bovine calbindin D9k. Biochemistry 28, 7065-74. Kordel, J., Pearlman, D. A . and Chazin, W. J. (1997). Protein solution structure calculations in solution: solvated molecular dynamics refinement of calbindin D9k. JBiomol NMR 10, 231-43. Kordel, J., Skelton, N . J., Akke, M . and Chazin, W. J. (1993). High-resolution structure of calcium-loaded calbindin D9k. J Mol Biol 231, 711-34. Kragelund, B. B. , Jonsson, M . , Bifulco, G., Chazin, W. J., Nilsson, H. , Finn, B. E. and Linse, S. (1998). Hydrophobic core substitutions in calbindin D9k: effects on Ca2+ binding and dissociation. Biochemistry 37, 8926-37. Kretsinger, R. H . (1980). Structure and evolution of calcium-modulated proteins. CRC Crit Rev Biochem 8, 119-74. Kretsinger, R. H. and Barry, C. D. (1975). The predicted structure of the calcium-binding component of troponin. Biochim Biophys Acta 405, 40-52. Kretsinger, R. H. and Nockolds, C. E. (1973). Carp muscle calcium-binding protein. II. Structure determination and general description. J Biol Chem 248, 3313-26. 206 References Kuboniwa, H. , Tjandra, N . , Grzesiek, S., Ren, H. , Klee, C. B. and Bax, A . (1995). Solution structure of calcium-free calmodulin. Nat Struct Biol 2, 768-76. Kumar, A. , Ernst, R. R. and Wuthrich, K. (1980). A two-dimensional nuclear Overhauser enhancement (2D NOE) experiment for the elucidation of complete proton-proton cross-relaxation networks in biological macromolecules. Biochem. Biophys. Res. Comm. 95, 1-6. Kumar, V . D., Lee, L . and Edwards, B. F. (1990). Refined crystal structure of calcium-liganded carp parvalbumin 4.25 at 1.5-A resolution. Biochemistry 29, 1404-12. Leavis, P. C. and Gergely, J. (1984). Thin filament proteins and thin filament-linked regulation of vertebrate muscle contraction. CRC Crit Rev Biochem 16, 235-305. Leavis, P. C. and Kraft, E. L. (1978). Calcium binding to cardiac troponin C. Arch Biochem Biophys 186, 411-5. Leszczynski, J. F. and Rose, G. D. (1986). Loops in globular proteins: a novel category of secondary structure. Science 234, 849-55. Lin, G. D., Chattopadhyay, D., Maki, M . , Wang, K. K., Carson, M . , Jin, L. , Yuen, P. W., Takano, E., Hatanaka, M . , DeLucas, L. J. and Narayana, S. V . (1997). Crystal structure of calcium bound domain VI of calpain at 1.9 A resolution and its role in enzyme assembly, regulation, and inhibitor binding [see comments]. Nat Struct Biol 4, 539-47. Lin, Y . M . , Liu, Y . P. and Cheung, W. Y . (1974). Cyclic 3':5'-nucleotide phosphodiesterase. Purification, characterization, and active form of the protein activator from bovine brain. J Biol Chem 249, 4943-54. 207 Linse, S., Brodin, P., Drakenberg, T., Thulin, E., Sellers, P., Elmden, K. , Grundstrom, T. and Forsen, S. (1987). Structure-function relationships in EF-hand Ca2+-binding proteins. Protein engineering and biophysical studies of calbindin D9k. Biochemistry 26, 6723-35. Linse, S., Brodin, P., Johansson, C , Thulin, E., Grundstrom, T. and Forsen, S. (1988). The role of protein surface charges in ion binding. Nature 335, 651-2. Linse, S., Bylsma, N . R., Drakenberg, T., Sellers, P., Forsen, S., Thulin, E., Svensson, L . A. , Zajtzeva, I., Zajtsev, V . and Marek, J. (1994). A calbindin D9k mutant with reduced calcium affinity and enhanced cooperativity. Metal ion binding, stability, and structural studies. Biochemistry 33, 12478-86. Linse, S. and Chazin, W. J. (1995). Quantitative measurements of the cooperativity in an EF-hand protein with sequential calcium binding. Protein Sci 4, 1038-44. Linse, S. and Forsen, S. (1995). Determinants that govern high-affinity calcium binding. Adv Second Messenger Phosphoprotein Res 30, 89-151. Linse, S., Helmersson, A . and Forsen, S. (1991a). Calcium binding to calmodulin and its globular domains. J Biol Chem 266, 8050-4. Linse, S., Johansson, C , Brodin, P., Grundstrom, T., Drakenberg, T. and Forsen, S. (1991b). Electrostatic contributions to the binding of Ca2+ in calbindin D9k. Biochemistry 30, 154-62. Linse, S., Jonsson, B. and Chazin, W. J. (1995). The effect of protein concentration on ion binding. Proc Natl Acad Sci USA 92, 4748-52. Lu, K . P. and Means, A . R. (1993). Regulation of the cell cycle by calcium and calmodulin. Endocr Rev 14, 40-58. 208 Lynch, T. J. and Cheung, W. Y . (1979). Human erythrocyte Ca2+-Mg2+-ATPase: mechanism of stimulation by Ca2+. Arch Biochem Biophys 194, 165-70. MacManus, J. P. (1979). Occurrence of a low-molecular-weight calcium-binding protein in neoplastic liver. Cancer Res 39, 3000-5. MacManus, J. P., Hutnik, C. M . , Sykes, B. D., Szabo, A . G., Williams, T. C. and Banville, D. (1989). Characterization and site-specific mutagenesis of the calcium-binding protein oncomodulin produced by recombinant bacteria. J Biol Chem 264, 3470-7. MacManus, J. P., Szabo, A . G. and Williams, R. E. (1984). Conformational changes induced by binding of bivalent cations to oncomodulin, a paravalbumin-like tumour protein. Biochem J220, 261-8. Madsen, P., Rasmussen, H . H. , Leffers, H. , Honore, B. , Dejgaard, K. , Olsen, E., K i i l , J., Walbum, E., Andersen, A . H. , Basse, B. and et al. (1991). Molecular cloning, occurrence, and expression of a novel partially secreted protein "psoriasin" that is highly up-regulated in psoriatic skin. J Invest Dermatol 97, 701-12. Malik, N . A . , Anantharamaiah, G. M . , Gawish, A . and Cheung, H. C. (1987). Structural and biological studies on synthetic peptide analogues of a low-affinity calcium-binding site of skeletal troponin C. Biochim Biophys Acta 911, 221-30. Mann, S. (1988). Molecular recognition of in biomineralization. Nature 332, 119-24. Marchiori, F., Borin, G., Chessa, G., Cavaggion, G., Michelin, L . and Peggion, E. (1983). Synthetic and binding studies on the postulated calcium binding site I of calmodulin. Hoppe Seylers Z Physiol Chem 364, 1019-28. Markley, J. L . and Opella, S. J. (1997). . Oxford University Press. 209 References Marsden, B. J., Hodges, R. S. and Sykes, B. D. (1988). 1H-NMR studies of synthetic peptide analogues of calcium-binding site III of rabbit skeletal troponin C: effect on the lanthanum affinity of the interchange of aspartic acid and asparagine residues at the metal ion coordinating positions. Biochemistry 27, 4198-206. Marsden, B. J., Shaw, G. S. and Sykes, B. D. (1990). Calcium binding proteins. Elucidating the contributions to calcium affinity from an analysis of species variants and peptide fragments. Biochem Cell Biol 68, 587-601. Matsuura, I., Ishihara, K. , Nakai, Y . , Yazawa, M . , Toda, H . and Yagi, K . (1991). A site-directed mutagenesis study of yeast calmodulin. J Biochem (Tokyo) 109, 190-7. Maune, J. F., Klee, C. B. and Beckingham, K. (1992). Ca2+ binding and conformational change in two series of point mutations to the individual Ca(2+)-binding sites of calmodulin. J Biol Chem 267, 5286-95. Maurer, P., Hohenadl, C , Hohenester, E., Gohring, W., Timpl, R. and Engel, J. (1995). The C-terminal portion of BM-40 (SPARC/osteonectin) is an autonomously folding and crystallisable domain that binds calcium and collagen IV. J Mol Biol 253, 347-57. McPhalen, C. A. , Sielecki, A . R., Santarsiero, B. D. and James, M . N . (1994). Refined crystal structure of rat parvalbumin, a mammalian alpha-lineage parvalbumin, at 2.0 A resolution. J Mol Biol 235, 718-32. McPhalen, C. A . , Strynadka, N . C. and James, M . N . (1991). Calcium-binding sites in proteins: a structural perspective. Adv Protein Chem 42, 77-144. McRee, D. E. (1993). "Practical Protein Crystallography". Academic Press, Inc, San Diego, CA. 210 References Meador, W. E., Means, A . R. and Quiocho, F. A . (1992). Target enzyme recognition by calmodulin: 2.4 A structure of a calmodulin-peptide complex. Science 257, 1251-5. Meador, W. E., Means, A . R. and Quiocho, F. A . (1993). Modulation of calmodulin plasticity in molecular recognition on the basis of x-ray structures. Science 262, 1718-21. Means, A . R., Tash, J. S. and Chafouleas, J. G. (1982). Physiological implications of the presence, distribution, and regulation of calmodulin in eukaryotic cells. Physiol Rev 62, 1-39. Meyer, E., Cole, G., Radhakrishnan, R. and Epp, O. (1988). Structure of native porcine pancreatic elastase at 1.65 A resolutions. Acta Crystallogr B 44, 26-38. Minowa, O. and Yagi, K. (1984). Calcium binding to tryptic fragments of calmodulin. J Biochem (Tokyo) 96, 1175-82. Moeschler, H . J., Schaer, J. J. and Cox, J. A . (1980). A thermodynamic analysis of the binding of calcium and magnesium ions to parvalbumin. Eur J Biochem 111, 73-8. Monera, O. D., Shaw, G. S., Zhu, B. Y . , Sykes, B. D., Kay, C. M . and Hodges, R. S. (1992). Role of interchain alpha-helical hydrophobic interactions in Ca2+ affinity, formation, and stability of a two-site domain in troponin C. Protein Sci 1, 945-55. Nagayama, K. , Kumar, A. , Wuthrich, K . and Ernst, R. R. (1980). Experimental techniques of two-dimensional correlated spectroscopy. J. Magn. Reson. 40, 321-34. 211 References Negele, J. C , Dotson, D. G., Liu, W., Sweeney, H. L . and Putkey, J. A . (1992). Mutation of the high affinity calcium binding sites in cardiac troponin C. J Biol Chem 267, 825-31. Nikitenko, L. , Morgan, G., Kolesnikov, S. I. and Wooding, F. B. (1998). Immunocytochemical and In situ hybridization studies of the distribution of calbindin D9k in the bovine placenta throughout pregnancy. J Histochem Cytochem 46, 679-88. Nitsch, C., Scotti, A. , Sommacal, A . and Kalt, G. (1989). GABAergic hippocampal neurons resistant to ischemia-induced neuronal death contain the Ca2(+)-binding protein parvalbumin. Neurosci Lett 105, 263-8. Ogawa, Y . (1985). Calcium binding to troponin C and troponin: effects of Mg2+, ionic strength and pH. J Biochem (Tokyo) 97, 1011-23. Olafson, R. W. and Smillie, L . B. (1975). Enzymic and physicochemical properties of Streptomyces griseus trypsin. Biochemistry 14, 1161-7. O'Neil, K . T. and DeGrado, W. F. (1990). A thermodynamic scale for the helix-forming tendencies of the commonly occurring amino acids [published erratum appears in Science 1991 Aug 30;253(5023):952]. Science 250, 646-51. Otwinowski, Z. and Minor, W. (1997). Processing of X-ray diffraction data collected in oscillation mode. Methods in Enzymology 276, 307-26. Palmisano, W. A. , Trevino, C. L . and Henzl, M . T. (1990). Site-specific replacement of amino acid residues within the CD binding loop of rat oncomodulin. J Biol Chem 265, 14450-6. 212 Patterson, A . L . (1935). A direct method for the determination of components of interatomic distances in crystal. Z Krist. 90, 517. Pauls, T. L. , Cox, J. A . and Berchtold, M . W. (1996a). The Ca2+(-)binding proteins parvalbumin and oncomodulin and their genes: new structural and functional findings. Biochim Biophys Acta 1306, 39-54. Pauls, T. L. , Durussel, I., Berchtold, M . W. and Cox, J. A . (1994). Inactivation of individual Ca(2+)-binding sites in the paired EF-hand sites of parvalbumin reveals asymmetrical metal-binding properties. Biochemistry 33, 10393-400. Pauls, T. L. , Durussel, I., Clark, I. D., Szabo, A . G., Berchtold, M . W. and Cox, J. A . (1996b). Site-specific replacement of amino acid residues in the CD site of rat parvalbumin changes the metal specificity of this Ca2+/Mg(2+)-mixed site toward a Ca(2+)-specific site. Eur J Biochem 242, 249-55. Pauls, T. L. , Durussel, I., Cox, J. A. , Clark, I. D., Szabo, A . G., Gagne, S. M . , Sykes, B . D. and Berchtold, M . W. (1993). Metal binding properties of recombinant rat parvalbumin wild-type and F102W mutant. J Biol Chem 268, 20897-903. Pearlstone, J. R., Borgford, T., Chandra, M . , Oikawa, K. , Kay, C. M . , Herzberg, O., Moult, J., Herklotz, A. , Reinach, F. C. and Smillie, L. B. (1992). Construction and characterization of a spectral probe mutant of troponin C: application to analyses of mutants with increased Ca2+ affinity. Biochemistry 31, 6545-53. Pechere, J. F., Derancourt, J. and Haiech, J. (1977). The participation of parvalbumins in the activation-relaxation cycle of vertebrate fast skeletal-muscle. FEBS Lett 75, 111-4. 213 Pedigo, S. and Shea, M . A . (1995). Quantitative endoproteinase GluC footprinting of cooperative Ca2+ binding to calmodulin: proteolytic susceptibility of E31 and E87 indicates interdomain interactions. Biochemistry 34, 1179-96. Pedrocchi, M . , Schafer, B. W., Durussel, I., Cox, J. A . and Heizmann, C. W. (1994). Purification and characterization of the recombinant human calcium- binding S100 proteins C A P L and C A C Y . Biochemistry 33, 6732-8. Permyakov, E. A . , Medvedkin, V . N . , Kalinichenko, L . P. and Burstein, E. A . (1983). Comparative study of physiochemical properties of two pike parvalbumins by means of their intrinsic tyrosyl and phenylalanyl fluorescence. Arch Biochem Biophys 227, 9-20. Permyakov, E. A . , Medvedkin, V . N . , Mitin, Y . V . and Kretsinger, R. H . (1991). Noncovalent complex between domain A B and domains CD*EF of parvalbumin. Biochim Biophys Acta 1076, 67-70. Persechini, A . , Stemmer, P. M . and Ohashi, I. (1996). Localization of unique functional determinants in the calmodulin lobes to individual EF hands. J Biol Chem 271, 32217-25. Potter, J. D. and Gergely, J. (1975). The calcium and magnesium binding sites on troponin and their role in the regulation of myofibrillar adenosine triphosphatase. J Biol Chem 250, 4628-33. Potter, J. D. and Johnson, J. D. (1982). In "Calcium and Cell Function" (W. Y . Cheung, ed.), vol. 2, pp. 145-73. Academic Press, New York. Potter, J. D., Johnson, J. D., Dedman, J. R., Schreiber, W. E., Mandel, F., Jackson, R. L . and Means, A . R. (1977). In "Calcium Binding Proteins and Calcium Function" 214 (R. H . Wasserman, R. A . Corradino, E. Carafoli, R. H. Kretsinger, D. H. MacLennan and S. F.L., eds), pp. 239-250. Elsevier North Holland, Inc., New York, New York. Potts, B. C , Smith, J., Akke, M . , Macke, T. J., Okazaki, K., Hidaka, H. , Case, D. A . and Chazin, W. J. (1995). The structure of calcyclin reveals a novel homodimeric fold for SI00 Ca(2+)-binding proteins [published erratum appears in Nat Struct Biol 1995 Oct;2(10):912]. Nat Struct Biol 2, 790-6. Procyshyn, R. M . and Reid, R. E. (1994a). A n examination of glutamic acid in the - X chelating position of the helix-loop-helix calcium binding motif. Arch Biochem Biophys 311, 425-9. Procyshyn, R. M . and Reid, R. E. (1994b). A structure/activity study of calcium affinity and selectivity using a synthetic peptide model of the helix-loop-helix calcium-binding motif. JBiol Chem 269, 1641-7. Prod'hom, B. and Karplus, M . (1993). The nature of the ion binding interactions in EF-hand peptide analogs: free energy simulation of Asp to Asn mutations. Protein Eng 6, 585-92. Putkey, J. A. , Sweeney, H. L. and Campbell, S. T. (1989). Site-directed mutation of the trigger calcium-binding sites in cardiac troponin C. JBiol Chem 264, 12370-8. Rao, S. T., Wu, S., Satyshur, K. A. , Ling, K. Y. , Kung, C. and Sundaralingam, M . (1993). Structure of Paramecium tetraurelia calmodulin at 1.8 A resolution. Protein Sci 2, 436-47. Rees, D. C , Lewis, M . and Lipscomb, W. N . (1983). Refined crystal structure of carboxypeptidase A at 1.54 A resolution. J Mol Biol 168, 367-87. 215 References Reid, R. E. (1987a). A synthetic 33-residue analogue of bovine brain calmodulin calcium binding site III: synthesis, purification, and calcium binding. Biochemistry 26, 6070-3. Reid, R. E. (1987b). Total sequential solid phase synthesis of rabbit skeletal troponin C calcium binding site III. Int JPept Protein Res 30, 613-21. Reid, R. E. (1990). Synthetic fragments of calmodulin calcium-binding site III. A test of the acid pair hypothesis. J Biol Chem 265, 5971-6. Reid, R. E., Clare, D. M . and Hodges, R. S. (1980). Synthetic analog of a high affinity calcium binding site in rabbit skeletal troponin C. J Biol Chem 255, 3642-6. Reid, R. E., Gariepy, J., Saund, A. K . and Hodges, R. S. (1981). Calcium-induced protein folding. Structure-affinity relationships in synthetic analogs of the helix-loop-helix calcium binding unit. J Biol Chem 256, 2742-51. Reid, R. E. and Hodges, R. S. (1980). Co-operativity and calcium/magnesium binding to troponin C and muscle calcium binding parvalbumin: an hypothesis. J Theor Biol 84, 401-44. Reid, R. E. and Procyshyn, R. M . (1995). Engineering magnesium selectivity in the helix-loop-helix calcium- binding motif. Arch Biochem Biophys 323, 115-9. Revett, S. P., King, G., Shabanowitz, J., Hunt, D. F., Hartman, K. L. , Laue, T. M . and Nelson, D. J. (1997). Characterization of a helix-loop-helix (EF hand) motif of silver hake parvalbumin isoform B. Protein Sci 6, 2397-408. Richardson, J. S. (1981). The anatomy and taxonomy of protein structure. Adv Protein Chem 34, 167-339. 216 References Roquet, F., Declercq, J. P., Tinant, B., Rambaud, J. and Parello, J. (1992). Crystal structure of the unique parvalbumin component from muscle of the leopard shark (Triakis semifasciata). The first X-ray study of an alpha-parvalbumin. J Mol Biol 223, 705-20. Sali, D., Bycroft, M . and Fersht, A . R. (1988). Stabilization of protein structure by interaction of alpha-helix dipole with a charged side chain. Nature 335, 740-3. Sarin, V . K. , Kent, S. B., Tarn, J. P. and Merrifield, R. B. (1981). Quantitative monitoring of solid-phase peptide synthesis by the ninhydrin reaction. Anal Biochem 117, 147-57. Sastry, M . , Ketchem, R. R., Crescenzi, O., Weber, C , Lubienski, M . J., Hidaka, H . and Chazin, W. J. (1998). The three-dimensional structure of Ca(2+)-bound calcyclin: implications for Ca(2+)-signal transduction by SI00 proteins. Structure 6, 223-31. Satyshur, K . A. , Rao, S. T., Pyzalska, D., Drendel, W., Greaser, M . and Sundaralingam, M . (1988). Refined structure of chicken skeletal muscle troponin C in the two-calcium state at 2-A resolution. J Biol Chem 263, 1628-47. Schafer, B. W. and Heizmann, C. W. (1996). The SI00 family of EF-hand calcium-binding proteins: functions and pathology. Trends Biochem Sci 21, 134-40. Schwarzenbach, G. (1952). Der chelateffekt. Helv Chim Acta 35, 2344-59. Sekharudu, Y . C. and Sundaralingam, M . (1988). A structure-function relationship for the calcium affinities of regulatory proteins containing 'EF-hand' pairs. Protein Eng 2, 139-46. Serrano, L. , Sancho, J., Hirshberg, M . and Fersht, A . R. (1992). Alpha-helix stability in proteins. I. Empirical correlations concerning substitution of side-chains at the N 217 and C-caps and the replacement of alanine by glycine or serine at solvent-exposed surfaces. J Mol Biol 227, 544-59. Sharma, Y . , Chandani, S., Sukhaswami, M . B., Uma, L. , Balasubramanian, D. and Fairwell, T. (1997). Modified helix-loop-helix motifs of calmodulin—The influence of the exchange of helical regions on calcium-binding affinity. Eur J • Biochem 243, 42-8. Shaw, G. S., Golden, L . F., Hodges, R. S. and Sykes, B. D. (1991a). Interaction between paired calcium-binding sites in proteins: N M R determination of the stoichiometry of calcium binding to a synthetic troponin-C peptide. J Am Chem Soc 113, 5557-63. Shaw, G. S., Hodges, R. S. and Sykes, B. D. (1990). Calcium-induced peptide association to form an intact protein domain: 1H N M R structural evidence. Science 249, 280-3. Shaw, G. S., Hodges, R. S. and Sykes, B. D. (1991b). Probing the relationship between alpha-helix formation and calcium affinity in troponin C: 1H N M R studies of calcium binding to synthetic and variant site III helix-loop-helix peptides. Biochemistry 30, 8339-47. Shaw, G. S., Hodges, R. S. and Sykes, B. D. (1992). Determination of the solution structure of a synthetic two-site calcium- binding homodimeric protein domain by N M R spectroscopy. Biochemistry 31, 9572-80. Shaw, G. S. and Sykes, B. D. (1996). N M R solution structure of a synthetic troponin C heterodimeric domain [published erratum appears in Biochemistry 1996 Sep 17;35(37):12220]. Biochemistry 35, 7429-38. 218 References Shea, M . A. , Verhoeven, A . S. and Pedigo, S. (1996). Calcium-induced interactions of calmodulin domains revealed by quantitative thrombin footprinting of Arg37 and Argl06. Biochemistry 35, 2943-57. Sia, S. K. , L i , M . X . , Spyracopoulos, L. , Gagne, S. M . , Liu, W., Putkey, J. A . and Sykes, B. D. (1997). Structure of cardiac muscle troponin C unexpectedly reveals a closed regulatory domain. JBiol Chem 272, 18216-21. Skelton, N . J., Forsen, S. and Chazin, W. J. (1990a). 1H N M R resonance assignments, secondary structure, and global fold of Apo bovine calbindin D9k. Biochemistry 29,5752-61. Skelton, N . J., Kordel, J., Akke, M . , Forsen, S. and Chazin, W. J. (1994). Signal transduction versus buffering activity in Ca(2+)-binding proteins. Nat Struct Biol 1, 239-45. Skelton, N . J., Kordel, J. and Chazin, W. J. (1995). Determination of the solution structure of Apo calbindin D9k by N M R spectroscopy. J Mol Biol 249, 441-62. Skelton, N . J., Kordel, J., Forsen, S. and Chazin, W. J. (1990b). Comparative structural analysis of the calcium free and bound states of the calcium regulatory protein calbindin D9K. J Mol Biol 213, 593-8. Slupsky, C. M . and Sykes, B. D. (1995). N M R solution structure of calcium-saturated skeletal muscle troponin C. Biochemistry 34, 15953-64. Spyracopoulos, L . , L i , M . X . , Sia, S. K. , Gagne, S. M . , Chandra, M . , Solaro, R. J. and Sykes, B. D. (1997). Calcium-induced structural transition in the regulatory domain of human cardiac troponin C. Biochemistry 36, 12138-46. 219 References Staun, M . (1991). Calbindin-D of human small intestine and kidney. Purification, molecular properties and clinical significance. Dan Med Bull 38, 271-82. Stewart, A . A . , Ingebritsen, T. S., Manalan, A. , Klee, C. B. and Cohen, P. (1982). Discovery of a Ca2+- and calmodulin-dependent protein phosphatase: probable identity with calcineurin (CaM-BP80). FEBS Lett 137, 80-4. Stewart, J. M . and Young, J. D. (1984). "Solid Phase Peptide Synthesis". Pierce Chemical Company, Rockford, Illinois. Strynadka, N . C , Cherney, M . , Sielecki, A . R., L i , M . X . , Smillie, L . B. and James, M . N . (1997). Structural details of a calcium-induced molecular switch: X-ray crystallographic analysis of the calcium-saturated N-terminal domain of troponin C at 1.75 A resolution. J Mol Biol 273, 238-55. Strynadka, N . C. and James, M . N . (1989). Crystal structures of the helix-loop-helix calcium-binding proteins. Annu Rev Biochem 58, 951-98. Swain, A . L. , Kretsinger, R. H. and Amma, E. L . (1989). Restrained least squares refinement of native (calcium) and cadmium-substituted carp parvalbumin using X-ray crystallographic data at 1.6-A resolution. J Biol Chem 264, 16620-8. Szebenyi, D. M . and Moffat, K. (1986). The refined structure of vitamin D-dependent calcium-binding protein from bovine intestine. Molecular details, ion binding, and implications for the structure of other calcium-binding proteins. J Biol Chem 261, 8761-77. Szebenyi, D. M . , Obendorf, S. K. and Moffat, K. (1981). Structure of vitamin D-dependent calcium-binding protein from bovine intestine. Nature 294, 327-32. 220 References Tabernero, L. , Taylor, D. A. , Chandross, R. J., VanBerkum, M . F., Means, A . R., Quiocho, F. A . and Sack, J. S. (1997). The structure of a calmodulin mutant with a deletion in the central helix: implications for molecular recognition and protein binding. Structure 5, 613-22. Taylor, D. A . , Sack, J. S., Maune, J. F., Beckingham, K. and Quiocho, F. A . (1991). Structure of a recombinant calmodulin from Drosophila melanogaster refined at 2.2-A resolution. JBiol Chem 266, 21375-80. Teo, T. S. and Wang, J. H . (1973). Mechanism of activation of a cyclic adenosine 3':5'-monophosphate phosphodiesterase from bovine heart by calcium ions. Identification of the protein activator as a Ca2+ binding protein. J Biol Chem 248, 5950-5. Trevino, C. L. , Boschi, J. M . and Henzl, M . T. (1991). Interactions between residues in the oncomodulin CD domain influence Ca2+ ion-binding affinity. J Biol Chem 266, 11301-8. Trigo-Gonzalez, G., Awang, G., Racher, K., Neden, K. and Borgford, T. (1993). Helix variants of troponin C with tailored calcium affinities. Biochemistry 32, 9826-31. Tsuji, T. and Kaiser, E. T. (1991). Design and synthesis of the pseudo-EF hand in calbindin D9K: effect of amino acid substitutions in the alpha-helical regions. Proteins 9, 12-22. Vassylyev, D. G., Takeda, S., Wakatsuki, S., Maeda, K. and Maeda, Y . (1998). Crystal structure of troponin C in complex with troponin I fragment at 2.3-A resolution. Proc Natl Acad Sci USA95, 4847-52. 221 References Vilei , E. M . , Calderara, S., Anagli, J., Berardi, S., Hitomi, K. , Maki, M . and Carafoli, E. (1997). Functional properties of recombinant calpain I and of mutants lacking domains III and IV of the catalytic subunit. J Biol Chem 272 , 25802-8. Virkar, P.D., Hoare, M . , Chan, M . Y . Y . , and Dunnill, P. (1982). Kinetics of the acid precipitation of soya protein in a continuous-flow tubular reactor. Biotech Bioeng 24,871-87. Vorum, H. , Madsen, P., Rasmussen, H. H. , Etzerodt, M . , Svendsen, I., Celis, J. E. and Honore, B. (1996). Expression and divalent cation binding properties of the novel chemotactic inflammatory protein psoriasin. Electrophoresis 17, 1787-96. Vyas, N . K. , Vyas, M . N . and Quiocho, F. A . (1987). A novel calcium binding site in the galactose-binding protein of bacterial transport and chemotaxis. Nature 327 , 635-8. Wakabayashi, T., Huxley, H. E., Amos, L . A . and Klug, A . (1975). Three-dimensional image reconstruction of actin-tropomyosin complex and actin-tropomyosin-troponin T-troponin I complex. J Mol Biol 93, All-91. Wall, M . E., Clarage, J. B. and Phillips, G. N . (1997). Motions of calmodulin characterized using both Bragg and diffuse X-ray scattering. Structure 5, 1599-612. Waltersson, Y . , Linse, S., Brodin, P. and Grundstrom, T. (1993). Mutational effects on the cooperativity of Ca2+ binding in calmodulin. Biochemistry 32 , 7866-71. Wang, C. K . and Cheung, H. C. (1985). Energetics of the binding of calcium and troponin I to troponin C from rabbit skeletal muscle. Biophys J AS, 727-39. 222 Watterson, D. M . , Sharief, F. and Vanaman, T. C. (1980). The complete amino acid sequence of the Ca2+-dependent modulator protein (calmodulin) of bovine brain. JBiol Chem 255, 962-75. Westcott, K. R., La Porte, D. C. and Storm, D. R. (1979). Resolution of adenylate cyclase sensitive and insensitive to Ca2+ and calcium-dependent regulatory protein (CDR) by CDR-sepharose affinity chromatography. Proc Natl Acad Sci U S A 76, 204-8. Wilkinson, J. M . (1976). The amino acid sequence of troponin C from chicken skeletal muscle. FEBSLett 70, 254-6. Williams, T. C , Corson, D. C , Oikawa, K. , McCubbin, W. D., Kay, C. M . and Sykes, B. D. (1986). 1H N M R spectroscopic studies of calcium-binding proteins. 3. Solution conformations of rat apo-alpha-parvalbumin and metal-bound rat alpha-parvalbumin. Biochemistry 25, 1835-46. Wilson, A . J. C. (1942). Determination of absolute scale from relative x-ray intensity data. Nature 150, 151-2. Wimberly, B. , Thulin, E. and Chazin, W. J. (1995). Characterization of the N-terminal half-saturated state of calbindin D9k: N M R studies of the N56A mutant. Protein Sci 4, 1045-55. Wojcik, J., Goral, J., Pawlowski, K. and Bierzynski, A . (1997). Isolated calcium-binding loops of EF-hand proteins can dimerize to form a native-like structure. Biochemistry 36, 680-7. 223 References Wu, X . and Reid, R. E. (1997a). Conservative D133E mutation of calmodulin site IV drastically alters calcium binding and phosphodiesterase regulation. Biochemistry 36, 3608-16. Wu, X . and Reid, R. E. (1997b). Structure/calcium affinity relationships of site III of calmodulin: testing the acid pair hypothesis using calmodulin mutants. Biochemistry 36, 8649-56. Zhang, M . , Tanaka, T. and Ikura, M . (1995). Calcium-induced conformational transition revealed by the solution structure of apo calmodulin. Nat Struct Biol 2, 758-67. Zot, A . S. and Potter, J. D. (1987). Structural aspects of troponin-tropomyosin regulation of skeletal muscle contraction. Annu Rev Biophys Biophys Chem 16, 535-59. 224 B I B L I O G R A P H Y Bibliography The following is a list of publications in press or submitted that are based on the work presented in this study. Franchini, P .L.A. and Reid, R.E. (1999) A Model for Circular Dichroism Monitored Dimerization and Calcium Binding in an EF-Hand Synthetic Peptide. In Press. J Theor Biol. Franchini, P .L.A. and Reid, R.E. (1999) Investigating Site Specific Effects of the - X Glutamate in a Parvalbumin CD Site Model Peptide. Submitted. Arch Biochem Biophy. Other publications include: Reid, R.E. and Franchini, P.L.A. (1999) Chapter 1: Practical Principles of Protein Design. In "Protein and Peptide Drug Analysis", Reid, R.E., Editor, In Press, Marcel Dekker Inc. 225 APPENDIX A Appendix A PEPTIDE PURIFICATION (A) HPLC chromatograms of crude peptides using a TEAP buffer system on a C18 column (9.4x250 mm) monitored at 254 nm. Arrow indicates collected peak. (B) Capillary electrophoresis chromatograms of purified peptides in 25 m M K2HPO4 pH 2.0 buffer monitored at 200 nm. Purified peptides appeared as a single peak on HPLC. PPPQKED B . 1.20E-01 1.00E-01 8.00E-02 g 6.00E-02 \ crs X I o Jg 4.00E-02 < 2.00E-02 0.00E+00 -2.00E-02 10 J 20 Time (min) 30 40 15 20 25 Time (min) 226 PPPQKED-XD Appendix A — i — 10 1 20 Time (min) 30 40 B . £ 4.00E-02 in 3.00E-02 227 Appendix A P P C Q K E D J ^ 15" 10 20 Time (min) 40 B. 4.50E-02 4.00E-02 3.50E-02 3.00E-02 S 2.50E-02 2.00E-02 1.50E-02 1.00E-02 5.00E-03 O.OOE+OO -5.00E-03 10 15 20 25 Time (min) 30 35 40 228 CPPQKED Appendix A 10 20 Time (min) 3 0 4 0 B . C 4.00E-02 CO ° 3.00E-02 < 20 25 Time (min) 30 35 40 229 Appendix A 10 15 20 25 Time (min) 30 35 40 230 Appendix A P P P Q K A D 10 20 Time (min) 30 40 g 5.00E-02 15 20 25 Time (min) 35 40 231 Appendix A PPPQKAA A. —i— 10 20 Time (min) —i— 30 40 B. 15 20 25 Time (min) 232 PPPKKAA Appendix A — i — 10 1 20 Time (min) 30 40 6.00E-02 5.00E-02 -4.00E-02 3.00E-02 2.00E-02 1.00E-02 0.00E+00 -1.00E-02 10 15 20 25 Time (min) 30 35 233 P P P K G A A Appendix A A . 10 20 Time (min) — i — 30 l b (5 2.50E-02 15 20 25 Time (min) 30 35 40 234 Appendix A PPPKGAD A. 0 10 20 30 40 Time (min) B. 1.00E-01 9.00E-02 • 8.00E-02 • 7.00E-02 6.00E-02 -I a 5.00E-02 o in Si < 4.00E-02 -| 3.00E-02 2.00E-02 1.00E-02 0.00E+00 -1.00E-02 -I 10 15 20 25 Time (min) 30 35 40 235 PPPKGED Appendix A 10 20 Time (min) 30 40 o £ 8.00E-02 re X) v> 6.00E-02 < 15 20 25 Time (min) 30 35 40 236 PPPKKED Appendix A L — I — 10 1 20 Time (min) —i— 30 40 B. g 5.00E-02 c CO • Q k_ o in Si < 10 15 20 25 Time (min) 30 35 40 237 Appendix B APPENDIX B PEPTIDE MOLECULAR WEIGHTS Table XVII. Peptide Molecular Weights Determined by Mass Spectrometry. Peptide Calculated MAY'. Measured M.W.* PPPQKED 3844.1 3843.2 PPPQKED-XD 3830.0 3830.1 CPPQKED 4078.3 4078.4 PPCQKED 3715.9 3716.3 C P C Q K E D 3950.0 2949.9 PPPQKAD 3786.2 3786.5 P P P Q K A A 3742.2 3742.0 P P P K K E D 3844.3 3844.1 P P P K G A A 3670.9 3671.0 PPPKGED 3773.0 3772.5 PPPKGAD 3715.1 3714.6 P P P K K A A 3742.3 3742.0 *Peptide molecular weights were confirmed using a Fisons V G Quattro Quadrupole electrospray mass spectrometer. The peptides were dissolved in 0.1% formic acid in 50%) methanol / 50% nanopure water. Positive ions were monitored with a scan speed of 100 mass units/s over a mass range of 600 to 1600. 238 Appendix C APPENDIX C SINGLE SITE MODEL Table XVIII. C a z + Titration Data Fit to the Single Site Model. Relative Peptide Concentration Peptide Parameter* 2.0 1.0 0.5 0.25 PPPQKED K D ( u M ) 72 98 172 264 K D C V (%) 3.44 1.40 4.07 5.93 r 2 0.9994 0.9999 0.9990 0.9978 PPPQKED-XD KD(UM) 18 20 31 72 K D C V (%) 5.28 2.37 2.66 4.19 r 2 0.9985 0.9997 0.9995 0.9984 CPPQKED K D (UM) 189 313 534 786 K D C V (%) 2.91 2.29 8.55 9.40 r 2 0.9994 0.9995 0.9937 0.9924 PPCQKED K D (UM) 1123 1942 3147 4256 K D C V (%) 9.38 11.7 12.8 12.0 r 2 0.9913 0.9956 0.9816 0.9819 CPCQKED K D (UM) 1242 1896 2590 3646 K D C V (%) 8.38 11.0 12.8 14.0 r 2 0.9933 0.9878 0.9828 0.9786 PPPQKAD K D (UM) 167 228 388 666 K D C V (%) 2.61 1.91 6.44 12.9 r 2 0.9995 0.9997 0.9968 0.9856 239 Appendix C Table XVIII. Ca 2 + Titration Data Fit to the Single Site Model (cont.) Relative Peptide Concentration Peptide Parameter 2.0 1.0 0.5 0.25 PPPKKAA K D ( u M ) 408 430 615 1523 K D C V (%) 2.81 4.74 6.54 5.94 r 2 0.9994 0.9983 0.9963 0.9958 PPPKGAA K D (UM) 168 194 222 395 K D C V (%) 4.498 3.467 1.837 4.497 r 2 0.9987 0.9991 0.9998 0.9984 PPPKGAD K D (uM) 101 80 159 228 K D C V ( % ) 2.65 2.26 3.07 4.25 r 2 0.9995 0.9997 0.9992 0.9984 PPPKGED K D (UM) 25 43 57 80 K D C V (%) 4.38 3.12 3.11 1.90 r 2 0.9991 0.9994 0.9994 0.9998 PPPKKED K D ( u M ) 73 94 173 257 K D C V (%) 3.15 2.82 4.39 3.78 r 2 0.9992 0.9993 0.9981 0.9985 KD is the calcium dissociation constant for the peptide. The KD values are generated from a plot of/versus free C a 2 + as described in section 2.2.4. KD C V % is the percentage convergence error in the KD value derived for the curve fit. Peptide concentrations used for the calculation were based on peptide mass with the exception of PPPQKED which was based on amino acid analysis. 240 APPENDIX D Appendix D DIMER MODEL Table XIX. Ca^ Titration Data Sets Fit Individually to the Dimer Model. Parameter Relative Amount of Peptide Mean ± SE 2 1.0 0.5 0.25 PPPQKED K , (uM) 476 386 388 435 421 ± 2 2 K 2 (uM) 33 40 52 63 47 ± 7 P S T O C K 1 (umole) 0.470 0.496 0.522 0.557 0.511 ±0.019 A (xlO"5) -1.51 -1.40 -1.33 -1.21 -1.36 ±0.06 C1" (xlO"5) -8.30 -8.30 -8.30 -8.30 -8.30 r 2 0.9998 0.9999 0.9996 0.9995 ~ PPPQKED-XD K , (uM) 2866 171 114 125 174 ± 3 7 K 2 (uM) 11 15 26 47 25 ± 8 P S T O C K 1 (pmole) 0.438 0.449 0.461 0.479 0.457 ± 0.009 A (xlO"5) -1.83 -1.77 -1.70 -1.58 -1.72 ±0.06 C T (xlO"5) -8.66 -8.66 -8.66* -8.66* -8.66 r 2 0.9978 0.9991 0.9997 0.9985 ~ * Indicates C was made equal to D to allow convergence. t In the curve fitting C was limited to being less than D. I PSTOCK is the total peptide in umoles that would be in 900 uL of undiluted stock solution. 241 Appendix D Table XIX. Dimer Model Fit Parameters for Synthetic EF-Hand Peptides (cont). Parameter Relative Amount of Peptide Mean ± SE 2 1.0 0.5 0.25 PPCQKED K , (pM) 3240 4048 5325 6364 4744 ± 689 K 2 (pM) 3 4 5 6 5 ± 1 P S T O C K 1 (umole) 0.414 0.433 0.459 0.485 0.448 ±0.016 A (xlO"5) -1.11 -1.08 -1.04 -0.99 -1.05 + 0.03 C T (xlO"5) -7.73* -7.73* -7.73* -7.73* -7.73 r 2 0.9996 0.9992 0.9968 0.9947 ~ CPPQKED K i (uM) 1542 1423 1246 1390 1400 ± 6 1 K 2 (uM) 68 148 33 50 75 ± 2 5 P S T O C K 1 (umole) 0.424 0.435 0.453 0.479 0.448 ±0.012 A (xlO"5) -1.62 -1.47 -1.52 -1.47 -1.52 ±0.03 d fx lO" 5 ) -8.06 -8.06 -8.06 -8.06* -8.06 r 2 0.9999 0.9997 0.9998 0.9997 — CPCQKED K i (uM) 4655 4807 5066 5806 5083 ±255 K 2 ( p M ) 80 5 5 6 24 ± 1 9 PSTOCK* (umole) 0.465 0.498 0.529 0.564 0.514 ±0.021 A (xlO"5) -0.948 -0.982 -0.928 -0.880 -0.935 ± 0.021 C f (xlO" 5 ) -8.08 -8.08* -8.08* -8.08* -8.08 r 2 0.9998 0.9999 0.9997 0.9995 — 242 Appendix D Table XIX. Dimer Model Fit Parameters for Synthetic EF-Hand Peptides (cont). Parameter Relative Amount of Peptide Mean ± SE 2 1.0 0.5 0.25 PPPQKAD K i (uM) 865 771 745 982 841 ± 54 K 2 (uM) 102 96 1 1 50 ± 2 8 PSTOCK* (pmole) 0.378 0.388 0.395 0.410 0.393 ± 0.007 A (xlO"5) -1.86 -1.73 -1.72 -1.64 -1.74 ±0.06 C f (xlO"5) -9.70 -9.70* -9.70* -9.70* -9.70 r 2 0.9998 0.9998 0.9992 0.9900 — PPPKKAA K , (pM) 2835 1688 1377 2276 2044 ± 323 K 2 (uM) 339 115 42 513 252 ±108 P S T O C K 1 (pmole) 0.405 0.418 0.405 0.388 0.404 ± 0.006 A (xlO"5) -1.99 -1.82 -1.84 -1.85 -1.88 ±0.04 C f (xlO"5) -8.99* -8.99* -8.99* -8.99* -8.99 r 2 0.9998 0.9998 0.9998 0.9978 ~ PPPKGAA K i (pM) 713 821 631 666 708 ± 4 1 K 2 (pM) 15 203 171 77 117 + 43 PSTOCK* (pmole) 0.504 0.532 0.547 0.568 0.538 ±0.013 A (xlf/ 5 ) -1.83 -1.56 -1.39 -1.32 -1.52 ± 0.11 C1 (xlO - 5) -7.79* -7.79* -7.79 -7.79 -7.79 r 2 0.9989 0.9990 0.9997 0.9997 — 243 Appendix D Table XIX. Dimer Model Fit Parameters for Synthetic EF-Hand Peptides (cont). Parameter Relative Amount of Peptide Mean ± SE 2 1.0 0.5 0.25 PPPKGAD K i (pM) 850 481 492 590 603 ± 86 K 2 (pM) 45 25 35 142 62 + 27 PSTOCK* (pmole) 0.396 0.403 0.418 0.430 0.412 ±0.008 A (xlO"5) -1.90 -1.78 -1.65 -1.56 -1.72 ±0.07 d fx lO" 5 ) -9.50 -9.50 -9.50* -9.50* -9.50 r 2 0.9993 0.9998 0.9997 0.9981 ~ PPPKGED K i (pM) 381 213 176 189 240 ± 48 K 2 (pM) 25 32 34 48 35 ± 5 PSTOCK* (pmole) 0.437 0.456 0.463 0.482 0.4594 ± 0.009 A (xlO' 5) -2.17 -2.07 -1.93 -1.74 -1.98 ±0.09 C f (xlO"5) -8.93 -8.93 -8.93 -8.93* -8.93 r 2 0.9989 0.9991 0.9995 0.9998 ~ PPPKKED K i (pM) 469 373 402 422 416 ± 2 0 K 2 (pM) 38 47 63 0.422 37 ± 13 PSTOCK* (pmole) 0.340 0.342 0.347 0.348 0.344 ± 0.002 A (xlO"5) -1.90 -1.84 -1.80 -1.83 -1.84 ±0.02 C* (xlO"5) -8.87 -8.87* -8.87* -8.87* -8.87 r 2 0.9996 0.9997 0.9988 0.9975 ~ 244 Appendix E APPENDIX E ELLIPTICITY SPECTRA FOR DIMERIZATION AND CALCIUM BINDING STUDIES 94-A) Ellipticity spectra for serial diluted peptide solutions under Ca saturating conditions. B) Peptide solution of -1.0 rng/mL (except PPPKKED ~0.5 mg/mL) in the absence of Ca + (Apo) and saturated with C a 2 + . * Spectra are pathlength and volume corrected. PPPQKED o -500 A. 1 -iooo <X> -1500 -2000 -1 1 1 1 200 210 220 230 240 250 Wavelength (nm) o -80 A -100 -C a 2 + -120 - o a -140 --160 -I 1 1 1 1 200 210 220 230 240 250 Wavelength (nm) 245 PPPQKED-XD Appendix E 2 4 6 Appendix E 247 Appendix E CPPQKED A. 200 210 220 230 Wavelength (nm) 240 250 B . cn T3 E, CD 200 210 220 230 Wavelength (nm) 240 250 248 Appendix E 2 4 9 Appendix E PPPQKAD Appendix E 251 Appendix E 252 Appendix E 253 Appendix E PPPKGED o -500 CD ] | -1000 <x> -1500 -2000 -200 210 220 230 240 250 Wavelength (nm) -140 --160 -I 1 1 1 1 200 210 220 230 240 250 Wavelength (nm) 254 Appendix E 255 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0089275/manifest

Comment

Related Items