Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The study of homo- and hetero-substituted de novo four-helix bundle proteins Huttunen, Heidi Esther Katrina 2006

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2007-267295.pdf [ 34.98MB ]
Metadata
JSON: 831-1.0059769.json
JSON-LD: 831-1.0059769-ld.json
RDF/XML (Pretty): 831-1.0059769-rdf.xml
RDF/JSON: 831-1.0059769-rdf.json
Turtle: 831-1.0059769-turtle.txt
N-Triples: 831-1.0059769-rdf-ntriples.txt
Original Record: 831-1.0059769-source.json
Full Text
831-1.0059769-fulltext.txt
Citation
831-1.0059769.ris

Full Text

T H E STUDY OF H O M O - AND HETERO-SUBSTITUTED DE NOVO F O U R - H E L I X B U N D L E PROTEINS by HEIDI ESTHER KATRINA HUTTUNEN B.Sc, University of British Columbia, 2000 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Chemistry) THE UNIVERSITY OF BRITISH COLUMBIA December 2006 © Heidi Esther Katrina Huttunen, 2006 Abstract The ability to design, synthesize and characterize de novo proteins can help facilitate the understanding of how individual amino acids contribute to the stability and structure of a protein. The de novo approach can be extended to include the use of templates, which assist in the organization of the peptides to form predetermined three-dimensional structures. These template assembled de novo proteins have been named "caviteins" (cov/tand + protein). One of the challenges in this area of research is the ability to design and synthesize native-like de novo proteins. Previously two caviteins, LG2 and LG3, had shown some native-like characteristics, although the evidence for a completely native-like structure remained debatable. The approach to identify a native-like structure was to design a corresponding protein that would exhibit non native-like properties. The leucine residues of LG2 and LG3 were replaced with norleucine residues in NG2 and NG3, respectively. The norleucine-based caviteins were less native-like in structure, as speculated, than their leucine-based counterparts. 11 In the past, the designed caviteins were limited to having only one type of peptide sequence attached within one bundle. Here, the design of a hetero-TASP, i.e. two different sequences within one bundle, was explored, and provided a means to create various de novo proteins, including an anti-parallel four-helix bundle. The hetero-TASPs were characterized and found to exhibit different native-like properties depending on the attached peptide sequences, and helix orientations. Lastly, the N - and C-capping efficiency of glycine was examined. Caviteins having, peptides linked to the cavitand template via their N - and C-terrnini, and with and without glycine caps were synthesized and characterized. It was found that the caviteins lacking glycine caps at their respective helix termini were comparable in stability with their capped-counterparts, which was contrary to what was hypothesized. It has been shown that subtle changes in the peptide sequence, linker and helix orientation have dramatic effects on the overall cavitein structure and stability. Since many of the factors underlying the stability and structure of four-helix bundles are now well understood, it would be exciting to undertake the challenges of designing caviteins with specific applications. i i i Table of Contents Abstract i i Table of Contents iv List of Tables ix List of Figures x i i List of Schemes xx List of Abbreviations xxi Acknowledgements xxiii CHAPTER 1: Introduction 1 1.0 The Significance of Proteins 1 1.1 The Protein-Folding Problem 1 1.1.1 Attempts Towards Solving the Protein-Folding Problem 2 1.1.1 Thesis Goals :..2 1.1.2 Thesis Overview 4 1.2 Protein Folding 5 1.2.1 Protein Folding in V ivo 5 1.2.2 The Thermodynamic Hypothesis and Levinthal's Paradox 6 1.2.3 Protein Folding Models and the "New V i e w " 7 1.2.4 Molten Globules: What are They? 10 1.2.5 Protein Folding using Molecular Chaperones 12 1.3 Protein Structure • 13 1.3.1 Primary Structure ....15 1.3.2 Secondary Structure ' 17 1.3.2.1 Thep-Strand 18 1.3.2.2 Thea -He l ix 19 1.3.3 Tertiary Structure ' 21 1.3.4 Quaternary Structure 21 1.3.5 a-Helical Motifs 22 1.3.5.1 Coiled Coils 22 1.3.5.1.1 The Leucine Zipper 25 1.3.5.2 The Square Bundle 25 iv 1.4 Factors Contributing to Protein Folding and Stability .2 1.4.1 Electrostatic Interactions ....2 1.4.2 Hydrogen Bonds..... .....2 1.4.3 Hydrophobic Interactions 2 1.5 Factors Contributing to the Stability of a-Helical Structures.. .2 1.5.1 Hel ix Capping.. .-. -3 1.5.2 Amino A c i d Propensities ...................3 1.5.3 Hel ix Chain Length..... ..3 1.5.4 The Hel ix Macrodipole... ...3 1.6 Factors Contributing to the Stability of Four-Helix Bundles 3 1.6.1 Side Chain Packing.. ... ....,-.3 1.6.2 Interhelical Side Chain Interactions. 3 1.6.3 Hel ix Orientation •• 3 1.7 De Novo Protein Design 3 1.7.1 DeGrado...:.. r...4 1.7.2 Hodges ................4 1.7.3 Hecht ......4 1.8 Template Use .............................4 1.8.1 Mutter ...A 1.8.2 Jensen .....5 1.8.3 Additional Templates ,'.5 1.9 Conclusion —~ 1.10 References... * CHAPTER 2: Design, Synthesis and Characterization of Caviteins LG3, LG2 NG3, and NG2: An Investigation of Native-like Structure.. t 2.0 Introduction. .............t 2.1 Rationale for the Investigation of Native-like Structure........... t 2.2 Results and Discussion t 2.2.1 Rationale for Using Cavitands as Templates for De Novo Design........... t 2.2.1.1 The Synthesis of Aryl thiolCavi tand 5. .t 2.2.2 Peptide Design ,....................'/ 2.2.2.1 Peptide Synthesis 2.2.3 Template Assembled Synthetic Protein (TASP) or Cavitein Synthesis 2.2.4 Characterization of the Cavitieins ."i 2.2.4.1 Fa r -UV Circular Dichroism (CD) Spectroscopy 76 2.2.4.2 Near-UV C D Spectroscopy .". 80 2.2.4.3 Oligomeric States Evaluated by G u H C l Denaturation Experiments.... 82 2.2.4.4 Analytical Ultracentrifugation ( A U C ) 86 2.2.4.4.1 Theory Behind Sedimentation Equilibrium .87 2.2.4.4.2 Determination of the Partial Specific Volume....... 89 2.2.4.4.3 Oligomeric States Evaluated by Sedimentation Equilibrium A U C .92 2.2.4.4.4 Sedimentation Velocity A U C ....97 2.2.4.5 ' H Nuclear Magnetic Resonance ( N M R ) Spectroscopy... ...101 2.2.4.6 One-Dimensional ( ID) *H N M R Spectroscopy 101 2.2.4.7 Hydrogen/Deuterium Amide Exchange ....108 2.2.4.8 Variable Temperature ' H N M R Spectroscopy 115 2.2.4.9 Two-Dimensional (2D) Homonuclear J H N M R Spectroscopy. 121 2.2.4.10 A N S Binding.... . . . . . . . . . . . . 135 2.3 Summary and Conclusions '. 137 2.4 Experimental .141 2.4.1 Aryl thiol Cavitand Synthesis 141 2.4.1.1 General........ .... 141 2.4.1.2 Synthesis of Methyl-Footed Aryl thiol (or Cavitand) 5 ...142 2.4.2 Peptide Synthesis..... 143 2.4.2.1 General..... 143 2.4.2.2 Synthesis of Peptides 6 (lg2), 7 (lg3), 8 (ng2), and 9 (ng3) 144 2.4.3 Caviteins Synthesis. 147 2.4.3.1 General 147 2.4.3.2 Synthesis of Caviteins 10 (LG2), 11 (LG3), 12 (NG2), and 13 (NG3). . . 147 2.4.4 C D Studies........ 148 2.4.5 Analytical Ultracentrifigation ( A U C ) Experiments.... .....151 2.4.6 Nuclear Magnetic Resonance ( N M R ) Experiments 173 2.4.6.1 I D ' H N M R Spectroscopy ...174 2.4.6.2 2D ! H N M R Spectroscopy.......... « ....176 2.4.7 A N S B ind ing . .....182 2.5 References • 184 CHAPTER 3: Design, Synthesis and Characterization of Hetero-TASPs .... 188 3.0 Introduction 188 3.1 Rationale for Synthesizing Hetero-TASPs 189 v i 3.2 Results and Discussion 191 3.2.1 Peptide Synthesis ................. 191 3.2.2 General Hetero-TASP Synthesis via Approach One ., ........192 3.2.3 General Hetero-TASP Synthesis via Approach Two 194 3.2.4 Characterization of the LG3-Substituted Cavitein Variants........... ...198 3.2.4.1 Fa r -UV C D Spectroscopy .............198 3.2.4.2 Near -UV C D Spectroscopy :.. 200 3.2.4.3 Oligomeric States...... ....202 3.2.4.3.1 G u H C l Denaturation Experiments. 202 3.2.4.3.2 A U C Sedimentation Equilibrium Experiments ..................205 3.2.4.4 One-Dimensional ( ID) *H N M R Spectroscopy Cavitein Variants .207 3.2.5 Characterization of the Hetero-TASPs .......210 3.2.5.1 F a r - U V C D Spectroscopy... ..: .212 3.2.5.2 Near -UV C D Spectroscopy .......217 3.2.5.3 Oligomeric States 221 3.2.5.3.1 G u H C l Denaturation Experiments......................... ..........221 3.2.5.3.2 A U C Sedimentation Equil ibrium Experiments ..........229 3.2.5.4 One-Dimensional ( I D ) ' H N M R Spectroscopy .....234 3.2.5.5 Hydrogen/Deuterium Amide Exchange 238 3.2.5.6 A N S Binding ...............243 3.3 Summary and Conclusions 246 3.4 Experimental.. — ...253 3.4.1 General..... 253 3.4.2 Peptide Synthesis....... 253 3.4.3 Hetero-TASP Synthesis,... .255 3.4.4 C D Studies... 263 3.4.5 Analytical Ultracentrifigation ( A U C ) Experiments................ 263 3.4.6 Nuclear Magnetic Resonance ( N M R ) Experiments 273 3.4.7 A N S Binding 273 3.5 References........: : 274 CHAPTER 4: Evaluating the C- and N-Capping Efficiency of Glycine.......276 4.0 Introduction......... 276 4.1 Background and Rationale for Studying N - and C-Capping of T A S P s 276 4.2 Results and Discussion ....277 4.2.1 Peptide Synthesis.... .278 4.2.2 Cavitein Synthesis .................. .........279 4.2.3 Characterization of the Capping Caviteins....... ....280 v i i 4.2.3.1 Fa r -UV C D Spectroscopy ;........i..280' 4.2.3.2 Near-UV C D Spectroscopy 283 4.2.3.3 Oligomeric States. ...................285 4.2.3.3.1 G u H C l Denaturation Experiments..... ...285 4.2.3.3.2 A U C Sedimentation Equil ibrium Experiments 288 4.2.3.4 One-Dimensional ( ID) l H N M R Spectroscopy .....291 4.2.3.5 Hydrogen/Deuterium Amide Exchange 292 4.2.3.6 ANS Binding-.....;.;.........; .....v.. ...................295 4.3 Summary and Conclusions .........297 4.4 Experimental — ••• .....300 3.4.1 General...... ..300 3.4.2 Peptide Synthesis............. ........300 3.4.3 Hetero-TASP Synthesis... 301 3.4.4 C D Studies... '.. 302 3.4.5 Analytical Ultracentrifigation ( A U C ) Experiments ......302 3.4.6 Nuclear Magnetic Resonance ( N M R ) Experiments .....308 3.4.7 A N S Binding ........308 4.5 References ......309 CHAPTER 5: Summary, Conclusions, and Future Work 310 4.1 Thesis Summary .....310 5.1 Experimental Conclusions........ ......311 5.2 Future Work............ .....319 5.3 References ...323 Appendix A. The One- and Three-Letter Codes for the Amino Acids Mentioned in this Thesis ; ..... .... 325 v i i i List of Tables Table 1.1. DeGrado's Peptide Series for the Synthesis of Four-Helix Bundles .43 Table 2.1. Complete Sequences Including Modified Termini for Peptides 6,1,8,9 73 Table 2.2. Names and Sequences for Caviteins 10,11,12 and 13...... .........75 Table 2.3. Molar Ellipticity at 222 nm ([^ 222) and Percent Helicity for Caviteins L G 2 , L G 3 , N G 2 , and N G 3 . . 379 Table 2.4. Guanidine Hydrochloride-Induced Denaturation Data Calculated for Caviteins L G 2 , L G 3 , N G 2 , and N G ; .........385 Table 2.5. Experimentally Determined Partial Specific Volumes by Sedimentation Equilibrium for A l l Cavitein Variants at 20 °C at Concentrations of 10 | i M and a Rotor Speed of 40000 rpm......... 92 Table 2.6. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For A l l Cavitein Variants at 20 °C in 50 m M p H 7.0 Sodium Phosphate Buffer at Concentrations of 10, 50 and 80 | i M with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section 2.4.5) ........97 Table 2.7. Tabulated Data from the Amide H / D Exchange Experiments on L G 2 , L G 3 , N G 2 and N G 3 in a 50 m M pD 5.02 CD3COOD/CD 3 COO"Na + Buf fe r at 20 °C... . . . . . 114 Table 2.8. Tabulated Data from the C O S Y and N O E S Y Experiments of L G 2 in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (note that sequence numbering began at the first glycine residue of the linker) 132 Table 2.9. Tabulated Data from the C O S Y and N O E S Y Experiments of L G 3 in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (note that sequence numbering began at the first glycine residue of the linker) .....133 Table 2.10. Tabulated Data from the C O S Y and N O E S Y Experiments of N G 3 in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (note that sequence numbering began at the first glycine residue of the linker).................. .........134 Table 2.11. Summary of Results for L G 2 , N G 2 , L G 3 and N G 3 138 Table 2.12. % Yields and M A L D I - M S Characterization of the "Activated" Peptides ....146 Table 2.13. % Yields and M A L D I - M S Characterization of the Caviteins made from "Activated" Peptides. .... ......148 Table 2.14. Experimental Parameters for Sedimentation Equil ibrium Studies on Caviteins L G 2 , L G 3 , N G 2 , and N G 3 Run at 20 °C ........152 Table 3.1. Complete Sequences from N - to C-termini Using One Letter Abbreviated Amino Acids Including Modified Termini for Peptides 6,7,14, and 15 192 Table 3.2. Names and Sequences for Caviteins 10,11,16-25. ,...197 Table 3.3. Molar Ellipticity at 222 nm ([^ 222) for the L G 3 / L G 2 Substituted Caviteins......213 Table 3.4. Molar Ellipticity at 222 nm ([6\22i) for the L G 3 / L G 2 C Substituted Caviteins...215 Table 3.5. Molar Ellipticity at 222 nm ([65222) for the L G 3 / A G 3 Substituted Caviteins .....217 Table 3.6. Guanidine Hydrochloride-Induced Denaturation Data Calculated for the L G 3 / L G 2 Substituted Caviteins ....223 Table 3.7. Guanidine Hydrochloride-Induced Denaturation Data Calculated for the L G 3 / L G 2 C Substituted Caviteins. 226 ix Table 3.8. Guanidine Hydrochloride-Jjiduced Denaturation Data Calculated for the L G 3 / A G 3 Substituted Caviteins... :.. 227 Table 3.9. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For L G 3 / L G 2 Substituted Caviteins at 20 °C in 50 m M p H 7.0 Sodium Phosphate Buffer at Concentrations of 10, 50, and 80 | i M with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section 3.4.5)'..'.:.'.'.....::....:„....,, .......:..........;232 Table 3.10. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For L G 3 / L G 2 C Substituted Caviteins at 20 °C in 50 m M p H 7.0 Sodium Phosphate Buffer at Concentrations of 10, 50, and 80 p:M with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section 3.4.5)............... „......233 Table 3.11. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For L G 3 / A G 3 Substituted Caviteins at 20 °C in 50 m M p H 7.0 Sodium Phosphate Buffer at Concentrations of 10, 50, and 80 [iM with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section 3.4.5)........... ................................ 233 Table 3.12. Tabulated Data from the Amide H / D Exchange Experiments of the L G 3 / L G 2 Substituted Caviteins in a 50 m M pD 5.02 C D 3 C O O D / C D 3 C O O " N a + Buffer at 20 °C ...240 Table 3.13. Tabulated Data from the Amide H / D Exchange Experiments of the L G 3 / L G 2 C Substituted Caviteins in a 50 m M pD 5.02 CD3COOD/ C D 3 C O O N a + Buffer at 20 ° C . : ................241 Table 3.14. Tabulated Data from the Amide H / D Exchange Experiments of the L G 3 / A G 3 Substituted Caviteins in a 50 m M pD 5.02 CD3COOD/ C D 3 C O O " N a + Buffer at 20 °c...... 242 Table 3.15. % Yields and M A L D I - M S Characterization of the "Activated" Peptides ..........254 Table 3.16. M A L D I - M S Characterization of the lg3 Substituted Hetero-TASP Intermediates ...............256 Table 3.17. % Yields and M A L D I - M S Characterization of the Hetero-TASPs ...262 Table 3.18. Experimental Parameters for Sedimentation Equil ibrium Experiments for the L G 3 / L G 2 Hetero-TASP Family ......264 Table 3.19. Experimental Parameters for Sedimentation Equilibrium Experiments for the L G 3 / L G 2 C Hetero-TASP Family............... ..„................265 Table 3.20. Experimental Parameters for Sedimentation Equil ibrium Experiments for the L G 3 / A G 3 Hetero-TASP Family266 Table 4.1. Complete Sequences Using One Letter Abbreviated Amino Acids Including Modified Termini for Peptides 6, 7,14, and 26-28 278 Table 4.2. Names and Sequences for Caviteins 10,11,16 and 29-31 , ...279 Table 4.3. Molar Ellipticity at 222 nm ([65222) for the Capping Caviteins............ 283 Table 4.4. Guanidine Hydrochloride-Indueed Denaturation Data Calculated for the Capping Caviteins .......287 Table 4.5. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For the Capping Caviteins at 20 °C in 50 m M p H 7.0 Sodium Phosphate Buffer at Concentrations of 10 pJVI, 50 pJvI and 80 ( i M with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section 4.4.5).... ..............290 Table 4.6. Tabulated Data from the Amide H / D Exchange Experiments of the Capping Caviteins in a 50 m M pD 5.02 C D 3 C O O D / C D 3 C O O N a + Buffer at 20 °C .......294 Table 4.7. % Yields and M A L D I - M S Characterization of the "Activated" Peptides 301 Table 4.8. % Yields and M A L D I - M S Characterization of the Caviteins.... 302 Table 4.9. Experimental Parameters for Sedimentation Equil ibrium Experiments for the Caviteins 304 x i List of Figures Figure 1.1. The Energy Landscape Representation of Protein Folding . . . . . . . . . . . . L .8 Figure 1.2. Schematic Representation of the Protein Folding Funnel Mode l ...9 Figure 1.3. A n Illustration of Side Chain Packing in Native Versus Molten Globule States ..11 Figure 1.4. Schematic Representation of Protein Organization ..14 Figure 1.5. Illustration Showing the Resonance of Peptide Bonds...... 15 Figure 1.6. A Ramachandran Plot Outlining the Al lowed Regions of Protein Secondary Structural Elements...... 17 Figure 1.7. The Hydgrogen Bonding Pattern of a fi-Sheet (hydrogen bonds are shown with pink dotted lines) ....18 Figure 1:8. Three-Dimensional Representations of a Right-Handed a-Hel ix (hydrogen bonds are shown with pink dotted lines) 20 Figure 1.9. Schematic Representations of "Knobs-Into-Holes" Packing in a (a) Parallel Dimeric Coiled C o i l (view into the dimeric interface), and (b) Trimeric Coiled C o i l (view of one packing layer) .........23 Figure 1.10. Helical Wheel Diagram Outlining Heptad Repeat of a Dimeric Coiled C o i l . (note that residues a, d, a ' , and d ' form the hydrophobic core) .24 Figure 1.11. Illustration of the Non-Hydrogen Bonded Atoms in the a-Hel ix Amide Backbone 31 Figure 1.12. Ranking of the Helical Propensities of the 20 Naturally Occurring Amino Acids.. . . . . . . . . 33 Figure 1.13. Simplified Two-Dimensional Illustration of the Packing Between Two a-Helices. Side Chains of Each Hel ix are Shown by Open and Closed Circles. (only one side chain of each helix is shown for clarity) (a) "Knobs-Into-Holes" Packing (b) "Ridges-Into-Grooves" 37 Figure 1.14. Schematic Representation of a (a) Parallel and an.(b) Anti-Parallel Four-Helix Bundle Protein •...,.....39 Figure 1.15. Diagram Displaying Template Assisted Folding...... .'. 48 Figure 1.16. A n Example of Mutter's Peptide Template .......50 Figure 1.17. Jensen's Carboproteiri Using a Methyl 2,3,4,6-tetra-O-Aoa-a-D-Galp Template.............. * .53 Figure 2.1. Primary Sequence for Peptide 6 Used in the Synthesis of Cavitein 10 70 Figure 2.2. Helical Wheel Diagram of Four Strands of Peptide 6 with the Helices Oriented in Parallel, (reader is looking down the helical axes from C - to N-termini) ........71 Figure 2.3. The Reference C D Spectra for Poly-L-Lysine in an a-Helix, (3-Sheet, and Random C o i l Conformation 76 Figure 2.4. F a r - U V C D Spectra for Caviteins L G 2 , L G 3 , N G 2 , and N G 3 at -40 [iM in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C .....78 Figure 2.5. Near -UV C D Spectra for Caviteins L G 2 , L G 3 , N G 2 , and N G 3 at -40 ( i M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C ..........81 Figure 2.6. Effect of G u H C l on the Helicity ([G[2 2) of Caviteins L G 2 , L G 3 , N G 2 , and N G 3 at -40 pJVI in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C .83 x i i Figure 2.7. Differential Sedimentation Plot of L G 3 in H2O and D2O, respectively, at a Concentration of 10 uJvl at 20 °C, and Rotor Speed of 40000 rpm................ 91 Figure 2.8. Sedimentation Equilibrium Concentration Distributions of L G 3 at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 \iM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit.. . . . 94 Figure 2.9. Sedimentation Equilibrium Concentration Distributions of L G 3 at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 50 pM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer-Trimer Equilibrium. The Upper Panel Represents the Residuals for the Fit........96 Figure 2.10. Sedimentation Velocity Raw Data from SEDFLT for L G 3 at a Concentration of 1.0 mg/mL, and a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C. The Upper Panel Represents the Residuals for the Fit. The Lower Picture Shows the Fitting Obtained from the S E D F I T Program ........99 Figure 2.11. Sedimentation Velocity Concentration Distributions of L G 3 at a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C 100 Figure 2.12. Fu l l 500 M H z lH N M R Spectrum of L G 2 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (* = cavitand signals) .....102 Figure 2.13. Fu l l 500 M H z / H N M R Spectrum of L G 3 at-1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (* = cavitand signals) 103 Figure 2.14. Fu l l 500 M H z ' H N M R Spectrum of N G 2 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (* = cavitand signals) 104 Figure 2.15. Fu l l 500 M H z ! H N M R Spectrum of N G 3 at -1.5 m M in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (* = cavitand signals) 104 Figure 2.16. Expanded Amide Regions of the 500 M H z ' H N M R Spectra of (a) L G 2 (b) L G 3 (c) N G 2 , and (d) N G 3 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (* = cavitand signals) ........105 Figure 2.17. Stack Plot of the 500 M H z ' H N M R Spectra Illustrating the Time Dependent Amide H / D Exchange of L G 2 at -1.5 m M in 50 m M pD 5 . 0 2 C D 3 C O O D / C D 3 C O O ~ N a + Buffer at 20 °C. (a) 5 min (b) 20 min (c) 1 h 4 min (d) 6 h 10 min (e) 25 h 17 min (f) 74 h 36 min .(* = cavitand signals)...... 110 Figure 2.18. Stack Plot of the 500 M H z *H N M R Spectra Illustrating the Time Dependent Amide H / D Exchange of L G 3 at-1.5 m M in 50 m M p D 5.02 C D 3 C O O D / C D 3 C O O N a + Buffer at 20 °C. (a) 6 min (b) 20 min (c) 1 h 5 min (d) 6 h 10min (e) 23 h 19 min (f) 72 h 53 min (g) 9 d 1 h 21 min (* = cavitand signals) I l l Figure 2.19. Stack Plot of the 500 M H z ! H N M R Spectra illustrating the Time Dependent Amide H / D Exchange of N G 2 at -1.5 m M in 50 m M pD 5.02 C D 3 C O O D / C D 3 C O O N a + Buffer at 20 °C. (a) 4 min (b) 18 min (c) 1 h 15 min (d) 6 h 20 min (e) 22 h 26 min (* = cavitand signals) ...........112 Figure 2.20. Stack Plot of the 500 M H z *H N M R Spectra Illustrating the Time Dependent Amide H / D Exchange of N G 3 at-1.5 m M in 50 m M pD 5.02 C D 3 C O O D / C D 3 C O O " N a + B u f f e r at 20 °C. (a) 5 nnn (b) 28 rmn (c) 1 h 3 min (d) 6 h 8 min (e) 22 h 15 min (f) 4 d 21 h 10 min (* = cavitand signals) ...113 x i i i Figure 2.21. Stack Plot of the 500 M H z XK N M R Spectra Illustrating the Temperature Dependent Amide Regions of L G 2 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (a) 20 °C (b) 30 °C .(c) 40 °C (d) 50 °C (e) 60 °C (* = cavitand signals).... 116 Figure 2.22. Stack Plot of the 500 M H z ' H N M R Spectra Illustrating the Temperature Dependent Amide Regions of L G 3 at -1.5 m M in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (a) 20 °C (b) 30 °C (c) 40 °C (d) 50 °C (e) 60 °C (* = cavitand signals).. ....117 Figure 2.23. Stack Plot of the 500 M H z lH N M R Spectra Illustrating the Temperature Dependent Amide Regions of N G 2 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (a) 20 °C (b) 30 °C (c) 40 °C (d) 50 °C (e) 60 °C ( * = cavitand signals) 118 Figure 2.24. Stack Plot of the 500 M H z ! H N M R Spectra Illustrating the Temperature Dependence of the Upfield Regions of N G 2 at -1.5 m M in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (a) 20 °C (b) 30 °C (c) 40 °C (d) 50 °C (e) 60 °C......... 119 Figure 2.25. Stack Plot of 500 M H z ! H N M R Spectra Illustrating the Temperature Dependent Amide Region of N G 3 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (a) 20 °C (b) 30 °C (c) 40 °C (d) 50 °C (e) 60 °C (* = cavitand signals) ........120 Figure 2.26. Diagram Showing a Polypeptide Segment with C O S Y Connectivities Shown in Dotted Lines, and N O E Connectivities Shown with Arrows 123 Figure 2.27. 2D 500 M H z ! H N M R C O S Y Spectrum of L G 2 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (expanded in the aliphatic/ amide region) 124 Figure 2.28. 2D 500 M H z ' H N M R N O E S Y Spectrum of L G 2 at -1.5 m M in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (expanded in the amide region). 125 Figure 2.29. Fu l l 2D 500 M H z ] H N M R C O S Y Spectrum of L G 3 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C 126 Figure 2.30. 2D 500 M H z ' H N M R C O S Y Spectrum of L G 3 at -1.5 m M in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (expanded in the aliphatic/amide region) 127 Figure 2.31. 2D 500 M H z ] H N M R N O E S Y Spectrum of L G 3 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (expanded in the amide region) 128 Figure 2.32. 2D 500 M H z ] H N M R C O S Y Spectrum of N G 3 at -1.5 m M in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (expanded in the aliphatic/ amide region) 130 Figure 2.33. 2D 500 M H z } H N M R N O E S Y Spectrum of N G 3 at -1.5 m M in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (expanded in the amide region) ..................131 Figure 2.34. Fluorescence Emission Spectra of 2 p:M A N S in the Presence of 95 % Ethanol, 100 % Methanol, 50 | i M L G 2 , L G 3 , N G 2 and N G 3 at 20 °C in p H 7.0 50 m M Sodium Phosphate Buffer .136 Figure 2.35. Schematic Representation of the FastMoc™ Protocol on the A B 1 431 A Peptide Synthesizer ....145 xiv Figure 2.36. Sedimentation Equilibrium Concentration Distributions of L G 3 at a Rotor Speed of 35000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 1 0 | i M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit .158 Figure 2.37. Sedimentation Equilibrium Concentration Distributions of L G 3 at a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °.C at 80 [iM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit......... 159 Figure 2.38. Sedimentation Equilibrium Concentration Distributions of L G 2 at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 p:M. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit... . . . . . 160 Figure 2.39. Sedimentation Equilibrium Concentration Distributions of L G 2 at a Rotor Speed of 35000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 [iM.In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fi t . . 161 Figure 2.40. Sedimentation Equilibrium Concentration Distributions of L G 2 at a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 [iM. In the Lower Panels The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit 162 Figure 2.41. Sedimentation Equilibrium Concentration Distributions of N G 2 at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 50 [iM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. . . . . 163 Figure 2.42. Sedimentation Equilibrium Concentration Distributions of N G 2 at a Rotor Speed of 35000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 50 ' u M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit... . . . . . . ....164 Figure 2.43. Sedimentation Equilibrium Concentration Distributions of N G 2 at a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0,20 °C at 50 | i M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit .............165 Figure 2.44. Sedimentation Equilibrium Concentration Distributions of N G 3 at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 | i M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit 166 Figure 2.45. Sedimentation Equilibrium Concentration Distributions of N G 3 at a Rotor Speed of 35000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0,20 °C at 50 | i M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit 167 Figure 2.46. Sedimentation Equilibrium Concentration Distributions of N G 3 at a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 [iM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit .....168 Figure 2.47. Sedimentation Velocity Concentration Distributions of L G 2 at a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C 171 xv Figure 2.48. Sedimentation Velocity Concentration Distributions of N G 2 at a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C 172 Figure 2.49. Sedimentation Velocity Concentration Distributions of N G 3 at a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C 173 Figure 2.50. Fu l l 2D 500 M H z *H N M R N O E S Y Spectrum of N G 2 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C „.......177 Figure 2.51. Fu l l 2D 500 M H z ^H N M R T O C S Y Spectrum of N G 2 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (mixing time - 10 ms)...... 178 Figure 2.52. Fu l l 2D 500 M H z *H N M R T O C S Y Spectrum of N G 2 at -1.5 m M in 10 % D 2 0 , : 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (mixing time =15 ms)...... 179 Figure 2.53. Fu l l 2D 500 M H z ! H N M R T O C S Y Spectrum of N G 2 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (mixing time = 25 ms) 180 Figure 2.54. Fu l l 2D 500 M H z ! H N M R T O C S Y Spectrum of N G 2 at -1.5 m M in 10 % D 2 0 , 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (mixing time = 35 ms) ......181 Figure 3.1. Schematic Representation Highlighting the a,c and a,b Disubstituted Hetero-T A S P Intermediates, (note only two peptides are shown for clarity)........ 190 Figure 3.2. Fa r -UV C D Spectra for Ig3-Substituted Cavitein Variants at -40 | i M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C ...........................199 Figure 3.3. Near -UV C D Spectra for Ig3-Substituted Cavitein Variants at -40 | i M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C. . . ..........201 Figure 3.4. Effect of G u H C l on the Helicity ([^ 222) of the Ig3-Substituted Cavitein Variants at -40 | j M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C .......203 Figure 3.5. Sedimentation Equilibrium Concentration Distributions of LG3/2pep_ab at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C, and at a Concentration of 10 | i M . In the Lower Panel The Solid Lines Represents Theoretical Fits to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit .206 Figure 3.6. Fu l l 500 M H z *H N M R Spectra Overlay of (a) 2 m M LG3/2pep_ab (b) 1 m M LG3/2pep_ab in 200 m M KC1 salt (c) 0.5 m M LG3/2pep_ab (d) 2 m M LG3/2pep_ac, in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 6 C 208 Figure 3.7. Fa r -UV C D Spectra for the L G 3 / L G 2 Substituted Caviteins at -40 |±M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C ......212 Figure 3.8. Fa r -UV C D Spectra for the L G 3 / L G 2 C Substituted Caviteins at -40 ( i M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C. . . . . . . 214 Figure 3.9. Fa r -UV C D Spectra for the L G 3 / A G 3 Substituted Caviteins at - 40 \xM in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C 216 Figure 3.10. Near -UV C D Spectra for the L G 3 / L G 2 Substituted Caviteins at -40 u M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C 218 Figure 3.11. Near -UV C D Spectra for the L G 3 / L G 2 C Substituted Caviteins at -40 |xM in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C 219 Figure 3.12. Near-UV C D Spectra for the L G 3 / A G 3 Substituted Caviteins at - 40 ujvl in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C ..220 Figure 3.13. Effect of G u H C l on the Helicity ([0\222) of the L G 3 / L G 2 Substituted Caviteins at -40 ,uM in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C 222 xv i Figure 3.14. Effect of G u H C l on the Helicity ( [ f e ) of the L G 3 / L G 2 C Substituted Caviteins at -40 pjvl in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C 224 Figure 3.15. Effect of G u H C l on the Helicity ([^222) of the L G 3 / A G 3 Substituted Caviteins at -40 | J M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C ...227 Figure 3.16; Sedimentation Equilibrium Concentration Distributions of 2LG3»2LG2_ab at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C, and at a Concentration of 10.joM. In the Lower Panel The Solid Lines Represents Theoretical Fits to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit.......... ...................230 Figure 3.17. Sedimentation Equilibrium Concentration Distributions of 2LG3»2LG2_ac at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C, and at a Concentration of 10 , | i M . In the Lower Panel The Solid Lines Represents Theoretical Fits to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit.............. .". ........231 Figure 3.18. Expansions of the Amide Regions of 500 M H z ! H N M R Spectra of the L G 3 / L G 2 Substituted Caviteins at-1.5 m M in 10 % D2O, 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (a) L G 3 (b) L G 2 (c) 2LG3«2LG2_ab (d) 2LG3«2LG2_ac (•* = cavitand signals).....; ....234 Figure 3.19. Expansions of the Amide Regions of 500 M H z ! H N M R Spectra of the L G 3 / L G 2 C Substituted Caviteins at 1.5 m M in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (a) L G 3 (b) L G 2 C (c) 2LG3«2LG2C_ab (d) 2LG3»2LG2C_ac (* = cavitand signals) ...........236 Figure 3.20. Expansions of the Amide Regions of 500 M H z lU N M R Spectra of the L G 3 / A G 3 Substituted Caviteins at 1.5 m M in 10 % D2O, 45 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (a) L G 3 (b) A G 3 (c) 2LG3«2AG3_ab (d) 2LG3»2AG3_ac (* = cavitand signals).............. ...237 Figure 3.21. Stack Plot of 500 M H z ! H N M R Spectra Illustrating the Time Dependent Amide H / D Exchange of 2LG3»2LG2_ab in 50 m M pD 5.02 C D 3 C O O D / C D 3 C O O N a + Buffer at 20 °C. (a) 4 min (b) 18 min (c) 1 h 3 min (d) 6 h 9 min (* = cavitand signals) ......239 Figure 3.22. Fluorescence Emission Spectra of 2 | i M A N S in the Presence of 95 % Ethanol, 100 % Methanol, 50 | i M L G 3 / L G 2 Substituted Caviteins at 20 °C in p H 7.0 50 m M Sodium Phosphate Buffer .......243 Figure 3.23. Fluorescence Emission Spectra of 2 uJVI A N S in the Presence of 95 % Ethanol, 100 % Methanol, 50 |xM L G 3 / L G 2 C Substituted Caviteins at 20 °C in p H 7.0 50 m M Sodium Phosphate Buffer 244 Figure 3.24. Fluorescence Emission Spectra of 2 | i M A N S in the Presence of 95 % Ethanol, 100 % Methanol, 50 p:M L G 3 / A G 3 Substituted Caviteins at 20 °C in p H 7.0 50 m M Sodium Phosphate Buffer 245 Figure 3.25. The Purification of the Crude Reaction Mixture of the lg3 Substituted Hetero-T A S P s by Preparatory Reversed-Phase H P L C Using a Gradient of 20-70 % Acetonitrile (with 0.05 % T E A ) in Water (with 0.1 % T E A ) Over 60 Minutes ..257 Figure 3.26. M a l d i - M S Spectrum of Peak (1) = LG3/4pep.... . . . .........258 Figure 3.27. M a l d i - M S Spectrum of Peak 2 = LG3/3pep ........259 Figure 3.28. M a l d i - M S Spectrum of Peak 3 = LG3/2pep_ac .259 Figure 3.29. M a l d i - M S Spectrum of Peak 4 = LG3/2pep_ab....... .....................................260 xv i i Figure 3.30. M a l d i - M S Spectrum of Peak 5 = LG3/ lpep . . ...260 Figure 3.31. Sedimentation Equilibrium Concentration Distributions of L G 2 C at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 | j M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the F i t . . . 267 Figure 3.32. Sedimentation Equilibrium Concentration Distributions of 2LG3»2LG2C_ab at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 | i M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. . . . . .......268 Figure 3.33. Sedimentation Equilibrium Concentration Distributions of 2LG3»2LG2C_ac at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 p:M. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit . . . 269 Figure 3.34. Sedimentation Equilibrium Concentration Distributions of A G 3 at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 \iM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit 270 Figure 3.35. Sedimentation Equilibrium Concentration Distributions of 2LG3»2AG3_ab at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 | i M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit... . . . . . 271 Figure 3.36. Sedimentation Equilibrium Concentration Distributions of 2LG3«2AG3_ac at a Rotor Speed of 27 000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0,20 °C at 10 pJVl. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit . . . .......272 Figure 4.1. F a r - U V C D Spectra for the Capping Caviteins at -40 | i M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C..... . . . ...281 Figure 4.2. Near -UV C D Spectra for the Capping Caviteins at - 40 | j M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C ...284 Figure 4.3. Effect of G u H C l on the Helicity ([6^ 222) of the Capping Caviteins at -40 u.M in 50 m M p H 7.0 Sodium Phosphate Buffer at 20 °C 286 Figure 4.4. Sedimentation Equilibrium Concentration Distributions of LG2_nocap at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 jaM. In the Lower Panel The Solid Lines Represents Theoretical Fits to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. . . . . ...289 Figure 4.5. Expansions of the Amide Regions of 500 M H z ! H N M R Spectra of the Capping Caviteins at -1.5 m M in 10 % D 2 0 , 4 5 m M Sodium Phosphate Buffer, p H 7.0 at 20 °C. (a) L G 2 (b) LG2_nocap (c) L G 3 (d) LG3_nocap (e) L G 2 C (* = cavitand signals) 291 xv i i i Figure 4.6. Stack Plot of 500 M H z ! H N M R Spectra Illustrating the Time Dependent Amide H / D Exchange of LG2_nocap in 50 m M pD 5.02 CD3COOD/CD3COO N a + Buffer at 20 °C. (a) 4 min (b) 18 min (c) 1 h 3 min (d) 6 h 9 min (* = cavitand signals).... ..293 Figure 4.7. Fluorescence Emission Spectra of 2 |0M A N S in the Presence of 95 % Ethanol, 100 % Methanol, 50 | l M Capping Caviteins at 20 °C in p H 7.0 50 m M Sodium Phosphate Buffer 296 Figure 4.8. Sedimentation Equilibrium Concentration Distributions of LG3_n0cap at a Rotor Speed of 27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 | j M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit ...............305 Figure 4.9. Sedimentation Equilibrium Concentration Distributions of LG2C_nocap at a Rotor Speed of27000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C at 10 |JtM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit.. . . . . ......306 xix List of Schemes Scheme .1.1. Synthesis of a Template Assembled De Novo Four-Helix Bundle ..3 Scheme 2.1. The Synthesis of Methyl-Footed Aryl thiol Cavitand 5 69 Scheme 2.2. Diagram Illustrating the Synthesis of Cavitein 10. The Linkage Between the Cavitand and the Peptides is "cavitand-(S-CH 2CO-NH-peptide 6)4" ......74 Scheme 3.1. Schematic Representation of Approach One Outlining the General Synthesis for the Desired a,c Disubstituted Cavitein Intermediate, and a,c Hetero-TASP 193 Scheme 3.2. Schematic Representation of Approach Two Outlining the General Synthesis for an a,c Hetero-TASP where P G is the 2-(4-nitrophenyl)ethyl Group 195 xx List of Abbreviations A c acetyl A angstrom(s) A N S l-anilinonaphthalene-8-sulfonate A U C analytical ultracentrifugation C D circular dichroism C O S Y correlation ( N M R ) spectroscopy C T B cyclotribenzylene (template) 8 chemical shift d day(s) da dalton(s) D C M dichloromethane D H B 2,5-dihydroxybenzoic acid D I P E A diisopropylethylamine D M A N-N-dimethylacetamide D M F N-dimethylformamide D P D S 2,2'-dipyridyl disulfide D Q F double quantum filter Et ethyl ( - C H 2 C H 3 ) E t O A c ethyl acetate E t O H ethanol equiv. equivalents G u H C l guanidine hydrochloride h hour(s) H B T U 2-(lH-benzotriazol-l-yl)-l,l,3,3-tetramethyluroniumhexafluorophosphate H O B t 1 -hydroxybenzotriazole H P L C high performance (pressure) liquid chromatography IR infrared J coupling constant L litre m/z mass-to-charge ratio M parent mass (mass spectra) or moles per litre M A L D I matrix assisted laser desorption ionization M e methyl ( -CH 3 ) M e C N acetonitrile M e O H methanol M e O N a sodium methoxide M G molten globule min minute(s) m L millilitre m M millimolar ms millisecond(s) mt mixing time M S mass spectrometry or spectrum M W molecular weight x x i N a O M e sodium methoxide n - B u L i n-butyllithium N native state N B S Af-bromosuccinimide nm nanometer(s) N M P Af-methylpyrrolidone N M R nuclear magnetic resonance N O E S Y nuclear Overhauser enhancement spectroscopy npe 2-(4-nitrophenyl)ethyl group I D one-dimensional P G protecting group P D B protein data bank Ph phenyl group ppm parts per million R P - H P L C reverse-phase high performance (pressure) liquid chromatography rpm revolutions per minute it room temperature s seconds Spy S-pyridyl group U unfolded state [6\ molar ellipticity per residue tj/2 half-life T A S P template assembled synthetic protein T F A 2,2,2-trifluoroacetic acid T F E 2,2,2-trifluoroethanol T H F tetrahydrofuran T L C thin layer chromatography T O C S Y total correlation spectroscopy 2D two-dimensional U V ultraviolet vs. versus A l l of the one- and three-letter codes for the amino acids mentioned i n this thesis, including the twenty commonly occurring amino acids can be found in Appendix A . xx i i Acknowledgements I would first like to thank my supervisor, John Sherman, for his continual support and guidance throughout my graduate degree. He has heightened my interest for chemistry and my desire to pursue a job in academia. I am also grateful for all of the help from current and past members of the Sherman group. We have shared many unforgettable experiences both in and outside of the lab. A special thank you goes to Diana Wallhorn, who shared her extensive knowledge on de novo proteins, Jon Freeman, for proofreading sections of this thesis, and Emi ly Seo, whom I met in the lab and became a great friend with. I would also like to thank Mark Okon for acquiring the N M R spectra and sharing his expertise in protein N M R , Elena Polishchuk for the use of biological services, and all of the other service departments at U B C who have helped along the way. Finally, I would like to thank my family and friends, especially my parents for their unconditional love, support and dedication to their children, and my husband Patrick Hennelly for his love and encouragement. xx i i i CHAPTER 1: Introduction 1.0 The Significance of Proteins Proteins are biopolymers that participate in virtually every biological process. Numerous proteins operate as enzymes, as hormones, as carriers and as ion channels and others are found in structural elements of cells and organisms. The function of a protein is ultimately determined by its three-dimensional, or tertiary, structure. Insulin, for instance, is a small hormone that fits into complementary receptors to start a series of protein activation cascades, hemoglobin's globular shape allows it to bind and carry oxygen, actin's and myosin's long filaments enable Our muscles to contract, and collagen's structure makes it a useful connective tissue. It was Anfinsen's early studies on ribonuclease which concluded that a protein's tertiary structure is encoded in its linear sequence of amino acids. 1 From this transpired one of the greatest unanswered questions of the life sciences to which much effort has been expended over the past 25 years. How do linear polypeptides that carry all the necessary information for folding take on native three-dimensional shapes? 1.1 The Protein-Folding Problem Proteins put themselves together in a process known as "folding". The details of this process remain unclear, and its study is now referred to as the "protein-folding problem". The essential fact of folding entails that the amino acid sequence alone contains the necessary information that specifies both the native structure and the pathway to attain that state. A long-term goal in this area of research has been to deduce the parameters controlling the relationship between sequence, tertiary structure, and function. Solving the protein-folding problem w i l l have practical consequences in biology, drug development, and medicine and w i l l provide new insights into life's basic units. 1.1.1 Attempts Towards Solving the Protein-Folding Problem Pauling posed the protein-folding problem over 60 years ago, but considerable advances have been made with respect to solving it. Empirical secondary structure prediction schemes and conformational analysis algorithms 4 have been developed for structure prediction in proteins. In addition, systematic synthetic approaches (i.e. de novo design and template assembly) have been devised to elucidate the mechanisms by which proteins fold into their three-dimensional structures. Furthermore, Rose and Srinivasan at Johns Hopkins School of Medicine have created a computer program, L I N U S , with notable preliminary achievement that predicts how proteins w i l l fold. 5 1.1.2 Thesis Goals The design of new synthetic proteins in order to help elucidate the protein^folding problem, the prediction of a protein's tertiary structure from its primary amino acid sequence, has received great attention over the past few decades. One approach used to illustrate the 2 interactions that govern protein structure is to design and characterize simple de novo proteins. De novo proteins are small proteins designed from first principles, retaining the crucial interactions of natural proteins, but otherwise lacking their complexity. The use of templates to assist in the organization of the peptides to form pre-determined three-dimensional structures has emerged as a useful tool in this area of research. In our group, cavitands, that are rigid organic macrocycles with an enforced cavity, are used as such a template in the formation of four-helix bundle proteins (Scheme 1.1). This new family of template assembled de novo proteins has been named "caviteins" (cav/tand + protein). Scheme 1 .1 . Synthesis of a Template Assembled De Novo Four-Helix Bundle. This thesis w i l l explore protein folding and stability through the design, synthesis, and characterization of Template Assembled Synthetic Proteins, T A S P s . 6 A challenge in this area of research is the ability to create native-like structures. This is further hampered by the lack of information available to diagnose such structures. The design and characterization of a native-like protein w i l l be approached by first attempting to create and diagnose what constitutes a noh native-like structure. Furthermore, the synthesis of a four-helix bundle containing two different 3 peptide sequences attached within one bundle w i l l be undertaken. Our caviteins have previously been limited to having only one type of peptide sequence attached, but the development of hetero-TASPs opens an opportunity to create a variety of potentially native-like de novo proteins, including an anti-parallel four-helix bundle. 1.1.3 Thesis Overview Chapter 1 w i l l introduce the process of protein folding, explain the levels of protein organization, and expand on the factors contributing to the stability of a-helical structures and four-helix bundles. Furthermore, it w i l l present and expand on the concept of using de novo proteins and template assembly in the study of protein folding. Chapter 2 w i l l focus on the study of native-like structure through the design and synthesis of leucine and norleucine-based T A S P s , respectively. It w i l l discuss the choice of cavitand as template for the synthesis of the T A S P s , and explain the rationale for the peptide and four-helix bundle design. Chapter 3 w i l l present the efforts towards synthesizing hetero-TASPs with two different peptide sequences attached within one bundle, including the first anti-parallel T A S P . Chapter 4 w i l l focus on evaluating the effectiveness of glycine as an N - and C-capping residue for our caviteins, respectively. Lastly, Chapter 5 w i l l provide a thesis summary, followed by a review of the experimental data and overall experimental conclusions, and finally outline areas for future research. The process of protein folding w i l l be presented below, followed by a discussion on protein structure. 4 1.2 Protein Folding Protein folding is the process by which an unfolded polypeptide chain folds into a specific native and functional structure. Studying protein folding is of utmost biological relevance since misfolded protein aggregates have been shown to lead to Familial Amyloidotic Polyneuropathy (FAP) , Alzheimer's disease, M a d Cow disease, an inherited form of emphysema, and even many cancers.7 Furthermore, understanding the mechanisms of protein folding is essential in solving the protein-folding problem. Some of the key questions that arise in studying protein folding is (1) do proteins fold co- or post-translationally in vivo, (2) what is the correct mechanism and folding model, (3) and what is the explanation for the fast folding of proteins (Levinthal's Paradox)? This Section w i l l begin with a short discussion on the co- and post-translational folding of proteins (not a focus of this thesis), and w i l l then expand on the thermodynamic hypothesis of protein folding accompanied by Levinthal's paradox. A n explanation of additional protein folding models w i l l follow with an elucidation of the molten globule, and lastly illustrate the role of molecular chaperones in protein folding. 1.2.1 Protein Folding in Vivo Two opposing post-translational and co-translational approaches have extensively been debated as the means for the folding of proteins in vivo. The post-translational approach proposes that a polypeptide chain starts to fold immediately after it has been synthesized and left 5 the ribosome, whereas, the co-translational approach suggests that the native conformation of a protein is formed during the biosynthesis on the ribosome. 8 The post-translational approach was initially supported by two main kinds of in vitro experiments in the early 1970's: the familiar denaturation and renaturatiOn studies by Anfinsen 1 ' 9 , and then more recently by the synthesis of protein analogs using Merrifield's solid phase synthetic procedures. 1 0 The ability to make chemically synthesized (i.e. de novo) active proteins has provided support that proteins can fold independent of a ribosome. However, sufficient experimental data exists to support the co-translational folding of proteins. 1 1 In a recent review, Basharov argues that the post-translational theory of protein folding is neither precise nor suitable for solving the protein-folding problem, and he proposes a mechanism for co-translational protein folding. Polypeptides enter the folding compartments on the ribosome leading with the N-terminus, and gradually fold there. 1 2 He concludes that after each folding stage, a definite conformation is produced. Since studying protein folding in vivo is limited, it is still not definite whether proteins fold co- or post-translationally. Experimental data is available to support both theories, and therefore protein folding remains an area of extensive research. 1.2.2 The Thermodynamic Hypothesis and Levinthal's Paradox It was Anfinsen's early work on the reversible folding and unfolding of bovine pancreatic ribonuclease (RNase) that led to the thermodynamic hypothesis of protein folding. He suggested that since protein folding in vitro is evident, the final conformation is predetermined by the amino acid sequence, with a single energetically favored conformation that w i l l always be attained at equilibrium. 1 Anfinsen theorized that the natural or native conformation occurs because this unique three-dimensional structure is thermodynamically the most stable alternative (global Gibbs free energy minimum). Anfinsen's experimental studies demonstrated that the unfolded RNase enzyme refolded spontaneously back into a native conformation after returning the denaturing chemical environment back to natural conditions. 1 ' 9 Overall, the thermodynamic hypothesis supports the notion that proteins reach their global free energy minimum (native structure) in a pathway-independent manner under thermodynamic control. In 1969, Levinthal pointed out that it is impossible for an unfolded protein to fold into its native state by randomly sampling all possible conformations. 1 3 He estimated that the number of different conformations available to a 150-residue protein to be on the order of 10 3 0 0 , whereas the number of conformations sampled by a natural protein before reaching its final configuration was 10 8. This led h im to postulate that a protein must fold via a specific path to its final configuration guided by the rapid formation of local interactions, and he concluded that folding must be under kinetic control. There has been constant debate over these two proposed viewpoints with experimental data supporting one theory over the other. Some small monodomain proteins 1 4 obey the thermodynamic hypothesis whereas some native proteins are not in the most thermodynamically stable conformation. 1 5 1.2.3 Protein Folding Models and the "New View" Additional folding models, which follow the "classical" folding view that a specific path exists between the unfolded and folded state with a predominance of local interactions and stable 7 intermediates, include: the framework, diffusion-collision, hierarchic, nucleated collapse, 1 9 and the hydrophobic collapse models. 2 0 21 A "new view" of protein folding, the energy landscape theory, emerged a decade ago. Onuchic states "that protein folding is a progressive self-organization process whereby multiple pathways can lead to the native structure". 2 2 In this way, the protein is funneled toward fewer low energy states from higher energy states until it reaches the favored energy state. The energy landscape representation of protein folding can be seen in Figure l . l , 2 3 Figure 1 .1 . The Energy Landscape Representation of Protein Folding. (Figure Adapted with Permission from Reference 22.) Protein Folding Native Structure 8 The bumps in the funnel reflect folding traps of local free energy minima, possible misfolded protein intermediates, i.e. molten globules (expanded in Section 1.2.4), which are kinetically trapped. 2 1 This suggests that the molten globule state may not be a productive state at all . However, recent studies by Ara i and Kuwajima et al. provide strong evidence that the molten globule is a productive on-pathway intermediate in protein folding, and they further explain that it can be effectively represented using the folding funnel model (Figure 1.2).24 Figure 1.2. Schematic Representation of the Protein Folding Funnel Model . (Figure Adapted with Permission from Reference 23.) C onformational Entropy Funnel II Native Structure The first stage (Funnel I) involves the formation of a molten globule state, a flexible intermediate with secondary structural elements but lacking specific side chain interactions, from the unfolded state.2 4 The formation of the native state, characterized by specific hydrophobic and side chain packing interactions, from the molten globule state, is the second stage in the 9 folding pathway (Funnel II). This folding often requires the help of specific molecules called molecular chaperones (Section 1.2.5) to help in the accurate formation of the native conformation. In this new view, the molten globule is viewed as a productive structurally diverse intermediate that is formed in the folding funnel, and furthermore, the molten globule has been shown to lead to the native state via multiple pathways. 1.2.4 Molten Globules: What are They? There has been much debate over the correct usage of the term molten globule and whether it is a productive intermediate in the folding of proteins to their active native state. The molten globule structure possesses all of the elements of secondary structure distinctive of the native state but lacks specific tertiary interactions, such as explicit side chain packing. 2 5 B y contrast, the native state is well defined and does not interconvert between accessible low energy conformations. It wasn't until the late 1980's, when Ptitsyn's first review 2 5 on the molten globule and Kunihiro and Kuwajima's review 2 6 explaining an important prerequisite for the discovery of the molten globule state was published, that the scientific community recognized the validity of this intermediary state. In the 1970's, Ptitsyn suggested the concept of an intermediate along the protein-folding pathway possessing properties in between those of the native and unfolded states. He postulated that a protein first folds into a flexible state that has the native-like positions of secondary structural units of a-helices and (3-strands, but lacks the tight hydrophobic packing of the side chains characteristic of the native state.2 7 Ptitsyn's predictions were finally supported, through 10 experimental evidence, by Wong and Tanford's studies on the unfolding of carbonic anhydrase, • 28 which proceeded through an intermediate with looser packing than the native state. The term "molten globule" emerged much later in 1983 2 9 after Ptitsyn's extensive studies on alpha-lactalbumins showed that the intermediate state in the acid-induced unfolding was compact with native-like secondary structure, but was molten (i.e. no cooperative melting temperature) at room temperature.3 0 Ptitsyn's model of the molten globule state is shown in Figure 1.3. Figure 1.3. A n Illustration of Side Chain Packing in Native Versus Molten Globule States. Native State Molten Globule State Ptitsyn suggested that molten globules are universal productive folding intermediates formed on the kinetic folding pathways of all proteins, and the molten globule was assigned the kinetic role 3 1 : U ^ ^ M G — - > N . 11 This kinetic equation implies that the molten globule, M G , is required for the formation of the native state, N , from the unfolded state, U . In addition, Wright et al. have used hydrogen exchange pulse labeling N M R and stopped-flow circular dichroism (CD) to establish that the earliest detectable kinetic intermediate in the re-folding pathway of apomyoglobih has many of the characteristics of a molten globule. 3 2 The C D data indicates the helix content to be around 35 % with sufficient secondary structural stability to allow complete protection of the amide protons from exchange. 3 2 Furthermore, numerous studies on the structure and nature of the molten globule have shown that it can be a general intermediate in protein folding. 3 3 However, research on the relevance and role of the molten globule continues since it has been argued that the molten globule is a productive stable intermediate separated from the native and unfolded states by significant potential energy barriers. 3 4 A s was mentioned in Section 1.2.3, other experimental data is consistent with the molten globule being a non-productive species that is simply the energetically preferred form of the unfolded protein under refolding conditions (i.e. M G ^ ^ U > N , where M ^ ~" U is fast). 1.2.5 Protein Folding using Molecular Chaperones Molecular chaperones were first described by Laskey et al. in 1978, and are defined as a group of proteins that recognize and bind non-native proteins. Their primary function is to facilitate folding by both preventing the aggregation of unfolded proteins, and by interacting with unfolded proteins to give them an opportunity to fold correctly. There is uncertainty about how many proteins use chaperones during the folding process, although there is evidence that 12 numerous proteins use chaperones even for a short period of time to assist in their self-organization to their final native state. Ruddon and Bedows explain that "chaperones function by binding to particular hydrophobic regions on the polypeptide chain, which are more likely to misfold from the molten globule state, by massaging them until the polypeptide is correctly folded." During this "massage" period the conformation of the polypeptide changes followed by the protein being released from the chaperone. 3 9 In some cases the protein may bind to another chaperone to further assist the protein in attaining its native state. In conclusion, there still is debate over which folding model provides the most suitable explanation for protein folding. However, the achievements in the field of protein structural research give hope that the protein-folding problem wi l l be solved in the near future. In many cases, the amino acid sequence alone is enough to predict the tertiary structure of a protein, and the actual mechanisms involved in the folding of the de novo proteins presented in this thesis w i l l not be explored. However, factors contributing to the folding and stability of proteins, and more specifically to a-helices are fundamental to the research presented here, and these factors w i l l be described in the following Sections. 1.3 Protein Structure Proteins are amino acid chains (polypeptides) that fold into unique three-dimensional structures (see Figure 1.4).40 Protein sizes can vary from -50 to several hundred residues in length, and proteins smaller than 50 residues are commonly referred to as peptides. The information necessary for a protein's unique three-dimensional shape is encoded in the linear •13. ; • sequence of amino acids. Proteins may adopt more than one stable folded conformation, but only one conformation is considered to be the native state.41 Protein structural organization can be broken down into four components: primary (1°), secondary (2°), tertiary (3°) and quaternary structures (4°), and each level w i l l be discussed separately. Figure 1.4. Schematic Representation of Protein Organization. Pleated Sheet Amino A c i d s A l p h a Helix Pleated Sheet A l p h a Helix Primary Protein Structure is a sequence of a chain of amino acids Secondary Protein Structure occurs when the sequence of amino acids are linked by hydrogen bonds Tertiary Protein Structure occurs when certain attractions are present between alpha helices and pleated sheets Quaternary Protein Structure is a protein consisting of more than one amino acid chain 14 1.3.1 Primary Structure The primary structure of a protein is the linear sequence of amino acids of the polypeptide chain, without regard to spatial arrangement.4 1 Polypeptides are made up from the twenty commonly occurring natural amino acids, designated in short-form by either a three- or one-letter code (see Appendix A for a complete list of amino acids and their short-form Codes). Generally, polypeptide chains are written from left to right with the amino or N-terminus, dictated by a free amino group, on the left-hand side, and a carboxyl or C-terminus characterized by a free carboxyl group on the right. The amino acid backbone residues are covalently linked by peptide bonds formed by a condensation reaction between an amino group of one amino acid and the free carboxyl group of an adjacent residue. The peptide bond has partial double bond character due to resonance (see Figure 1.5),42 where the N H - C O dihedral angle, GO, is typically planar and 180° 4 3 Figure 1.5. Illustration Showing the Resonance of Peptide Bonds. 15 This partial double bond character shortens the average C-N bond distance from 1.45 A in a single bond to 1.33 A, and more importantly reduces the rotational freedom of the polypeptide during folding.43 Most amino acids in the polypeptide chain adopt a trans configuration since it is more stable than the cis conformation due to fewer steric repulsions and dipole-dipole interactions. Like any double bond, rotation about the peptide bond angle, co, is restricted with an energy barrier of -55-100 kJ/mol between the cis and the trans configurations.44 The backbone dihedral angles between HN-C a and C a -CO are denoted by § and \|/, respectively. The allowed regions of (j) and Xf space for each amino acid are displayed on Ramachandran plots,45 Al l residues have access to the two most favorable $ and Wf dihedral angles (located on the upper and lower left of the Ramachandran plot), but the allowed regions differ for each amino acid because of restriction due to C a and its substituents. The substituent or side chain dihedral angles are referred to as %i, %2, %3 starting from the atom closest to the amide backbone and the atoms of the side chains bound to the backbone C a are termed Cp, C Y, and C5 etc. Furthermore, it is the two favorable regions of the Ramachandran plot (Figure 1.6)46 which correspond to combinations of the ()) and \|/ angles that characterize secondary structural elements that can be adopted by the polypeptide backbone, namely the P-strand and the a-helix, discussed in Sections 1.3.2.1 and 1.3.2.2, respectively. 16 Figure 1.6. A Ramachandran Plot Outlining the Al lowed Regions of Protein Secondary Structural Elements. 180 +psi -psi •180 Right handed alpha-helix -180 -phi A Left handed alpha-helix + phl 180 1.3.2 Secondary Structure Most proteins fold into two regular, highly organized configurations designated either a a-helix or a (3-strand, although some small protein segments can remain less ordered in a random coil conformation. A sequence of residues that all have similar allowed (J) and \|/ space can influence the polypeptide chain to develop a- or P-structures. In general terms, the secondary structure of a protein is considered to be the spatial arrangement of the backbone atoms without regard to the conformation of its side chains. 4 1 Secondary structures are locally defined, indicating that there can be many different secondary motifs within one protein. Secondary conformations are mobile and can become distorted from the ideal structure depending on solution conditions in which the protein is dissolved. A n introduction to the 2 ° structural 17 features of P-strands and oc-helices will follow, but the focus of the remainder of this thesis will be on a-helical structures. 1.3.2.1 The P-Strand The P-strand structure (Figure 1.7) is essentially flat with the side chains sticking out on alternate sides.47 Successive dipole moments alternate along the chain and therefore the P-strand has no permanent dipole. Unlike a-helical segments, p-strands form intermolecular hydrogen bonds between two polypeptide chains; thus individual P-strands do not exist independently.41 It is P-sheets that result from two hydrogen-bonded P-strands, and are present in either a "parallel" or "anti-parallel" arrangement. In the parallel configuration the two interacting sheets run in the same direction and in the anti-parallel arrangement the sheets run in opposite directions (i.e. the N-terminal to C-terminal direction of one chain being the reverse of the other). Figure 1.7. The Hydrogen Bonding Pattern of a P-Sheet. (hydrogen bonds are shown with dotted lines) 18 In many cases, a single amino acid chain looping back on itself can form an anti-parallel P-sheet. Proline, because of its inherent cyclic side-chain which blocks the backbone N H group, is disruptive to secondary structure and in result produces a kink or turn in the polypeptide backbone, and is therefore commonly found in loop segments. 4 8 Loop segments or turn structures are also classified as 2° structural elements, but unlike helices and strands, they do not have regular, repeating geometries. However, turns are essential to allow the polypeptide backbone to fold back on itself in the formation of tertiary structures (Section 1.3.3). 1.3.2.2 The a-Helix The a-helix is a right-hand (clockwise) spiral that was suggested by Pauling and coworkers in the early 1950's. 4 9 Alpha-helices are classified as repetitive secondary structures, where the helix makes a turn every 3.6 amino acid residues (Figure 1.8),50 and has backbone dihedral angles of - 6 0 ° and - 4 0 ° for <J) and \|/, respectively. 5 1 Helix formation is driven by favorable enthalpy, resulting from both the formation of backbone intramolecular hydrogen bonds and van der Waals interactions between the atoms in the helix. On the other hand, fixing the backbone into a helical conformation has an associated entropic cost, but is still overall thermodynamically favorable. The intramolecular hydrogen bonds form between the backbone carbonyl (/) and the amide hydrogen atom of the fifth amino acid in the chain (i+4).49 The hydrogen bonds run parallel to the helical axis and have approximate N - 0 and H - 0 distances of 2.86 A and 1.93 A, respectively. 4 1 From this directional bonding pattern develops a helix macrodipole with a charge of +0.5 and -0.5 on the N - and G - termini, respectively. 5 3 Although the intramolecular '19 hydrogen bonds do provide substantial stabilization for the a-helical structure, there are many other factors which contribute to the stability of the helical arrangement that w i l l be discussed in Section 1.5. Alpha helices have the ability to be amphiphilic, 5 4 possessing one polar and one nonpolar face. A n example of a classical amphiphilic oc-helix is found in the four-helix bundle protein cytochrome Z?562-55 It is common for the nonpolar faces of independent amphiphilic a-helices to aggregate through hydrophobic interactions to exclude water in aqueous solution and arrange themselves in bundles. Thus, amphiphilic a-helices have been a natural starting point for the study of protein structural motifs in de novo design projects (Section 1.7). Figure 1.8. Three-Dimensional Representations of a Right-Handed a-Helix. (hydrogen bonds are shown with dotted lines) The side chains or R-groups extend to the outside of the a-helix backbone like rungs on a ladder. When multiple helices interact it is the noncovalent interactions (Section 1.4) between 20 the side chains such as hydrophobic packing, electrostatic, and hydrogen bonding that are vital for stabilizing a protein's final conformation, in addition to being responsible for binding ligands, and catalyzing biochemical reactions. 1.3.3 Tertiary Structure The tertiary structure of a protein is the overall three-dimensional structure of the polypeptide chain(s) held together primarily by hydrophobic interactions. 4 1 Hydrogen bonds, disulfide bonds, and ionic interactions also contribute. The tertiary structure of a protein is of utmost importance since the function of a protein depends on its overall shape. Tertiary and secondary structures are interconnected since proteins exhibiting well-defined topologies generally have a well-developed core, consisting of hydrophobic residues with restricted conformational freedom, arising from the secondary structural restraints. In this way, it is believed that both secondary and tertiary structures, because of their interdependence, are lost simultaneously upon changes to the environment that disrupts the folded state. When the tertiary structure is disrupted, the protein is denatured and loses its activity. 1.3.4 Quaternary Structure Quaternary structures result from the association of independent tertiary structural units through surface interactions, which function together as part of a larger protein complex. A common example is the formation of hemoglobin from myoglobin-like monomers. 21 1.3.5 a-Helical Motifs tt-Helices are a common structural element found in native proteins and are known to exist in a variety of topologies. 5 6 A structural motif is a three-dimensional structural element or "fold" within the chain. In the context of proteins it is sometimes used interchangeably with "structural domain." Motifs exhibit both secondary and tertiary structure and may be regarded as an arrangement of secondary structures. Coiled coi l s , 5 7 four-helix bundles, 5 8 D N A binding motifs including: zinc fingers, 5 9 helix-turn-helix ( H T H ) , 6 0 leucine zippers, 6 1 and basic-helix-loop-helix ( b H L H ) 5 6 are examples of a-helical motifs. Protein motifs containing two or three secondary structural elements have been classed as a family of super-secondary structures initially introduced by Weber and Salemme. 5 8 b Helical bundles encompass only a fraction of known protein helical motifs. Four-helix bundles, more specifically, can be categorized into two main classes. The first class contains near-parallel helix-helix interactions and is made up of two subclasses: coiled coils and square bundles, the latter being the focal point of this thesis. The second class contains both parallel and perpendicular helix interactions, the latter is not entirely clear and w i l l not be discussed. 1.3.5.1 Coiled Coils Crick was the first person to propose the idea of multi-stranded coiled coils back in the early 1950's. 6 3 Coiled coils consist of two to five amphipathic, right-handed a-helices that wrap around each other to adopt a left-handed supercoil where the nonpolar face of each a-helix is continually adjacent to that of the other helix, like strands on a rope. 6 4 This supercoiling 22 (twisting) changes the effective number of residues per turn (along the coiled coil axis) from 3.6 to 3.5 residues per turn and allows the helices to stay in continuous contact over long distances. 6 5 Crick proposed that the dimeric structure was stabilized by the packing of "knobs" formed by the hydrophobic side chains of one helix into "holes" formed by the spaces between the side chains of the residues in the adjacent helix (Figure 1.9).66 Figure 1.9. Schematic Representations of "Knobs-Into-Holes" Packing in a (a) Parallel Dimeric Coiled C o i l (view into the dimeric interface), and (b) Trimeric Coiled C o i l (view of one packing layer). (Figure Adapted with Permission from Reference 64.) (a) (b) In nature, two-stranded coiled coils are most common including muscle proteins, oc-keratin, bacterial surface proteins, tumor suppressors, and intermediate filaments, 6 7 although the protein R O P is an example of a naturally occurring four-helix bundle with a coiled coil topology. 6 8 From the crystal structure of G C N 4 , a two-stranded coiled coi l , it was deduced that two right-handed helices wrap around each other to form approximately a quarter turn of a left-23 handed supercoil, with a crossing angle of 20° between the helices. The X-ray structure of G C N 4 also showed that the super helical axis of the helices is nearly straight, whereas the individual a-helices are smoothly bent because the main chain hydrogen bonds at the interface are shorter than those on the outside of the helices. In 1972 Hodges was able to obtain the first amino acid sequence of a coiled coi l , tropomyosin, 6 9 the smallest and simplest of these proteins postulated to contain this structure at that time. From this sequence, Hodges and coworkers identified that the "hydrophobic repeat" was responsible for the formation and stabilization of the coiled coil structure. The hydrophobic repeat consists of a repeating heptad of seven amino acid residues [abcdefg] n, where positions a and d are occupied by hydrophobic residues and oppositely charged residues at e and g of adjacent heptads to form interhelical ion pairs (Figure 1.10). 6 7 b ' 6 9 From this sequence repetition it follows that every seventh amino acid is in the same position with regards to the helical axis. 7 1 Figure 1.10. Helical Wheel Diagram Outlining Heptad Repeat of a Dimeric Coiled C o i l , (note that residues a, d, a ' , and d ' form the hydrophobic core). 24 1.3.5.1.1 The Leucine Zipper These repetitive patterns of hydrophobic and charged residues are also evident in "leucine zipper" sequences.7 2 The term "leucine zipper" was suggested in 1988 as a result of the leucine repeat at seven residue intervals observed in D N A binding proteins. This model incorrectly proposed that the leucine side chains in adjacent helices were interdigitated like teeth in a zipper. It was soon realized that these "leucine zipper" sequences contained the same 3-4 hydrophobic repeat characteristic of coiled coils and folded as a parallel two-, three-, or four-stranded coiled c o i l . 5 8 b > 6 4 ' 7 4 The X-ray structure of G G N 4 further weakened the leucine zipper model in terms of side chain packing and demonstrated that the conserved leucine residues at position d were not interdigitated but pack like "knobs-into-holes", in a fashion similar to the hydrophobes at position a. Furthermore, the X-ray structure of the G C N 4 leucine zipper clearly identified two main differences with classical coiled coils (1) leucine zippers are much shorter than most coiled coils and (2) that leucine occurs almost invariably at position d, whereas in traditional coiled coils only a quarter to a half of the residues at position d are leucine. 1.3.5.2 The Square Bundle The square four-helix bundle, also abbreviated four-helix bundle, is a common a-helical motif found in native proteins. Some examples of currently known proteins which incorporate 7S the four-helix bundle as their dominant structural motif include: haemerythrin subunit, 1f\ 77 appoferritin monomer, tobacco mosaic virus ( T M V ) coat protein, monomeric haem protein 25 cytochrome £ 5 6 2 , and the dimeric cytochrome c. The helices of a square bundle pack at 20° with respect to each other and are most stable in the anti-parallel conformation, but diverge at their termini owing to the absence of significant supercoiling. 6 5 ' 7 9 These open regions can serve as binding pockets typically occupied by a prosthetic group, as is seen in cytochrome £ 5 6 2 with the binding pocket occupying a heme group. 5 5 Square bundles, in comparison to coiled coils, tend to contain a wider hydrophobic face, owing to a greater incidence of hydrophobes at the e and g positions, whereas coiled coils contain a narrow hydrophobic face (hydrophobic residues primarily at the a and d positions in the heptad repeat). A s mentioned in Section 1.3.5.1 on Coiled coils, they are characterized by left-handed supercoiling of the helices, allowing the helices to stay in continuous contact over long distances. In contrast to the coiled coi l structure favoring a two-helix bundle, the wider hydrophobic face of helical square bundles naturally promotes higher oligomerization states, in which more surface area can be buried, and so the four-helix bundle is the most common. 6 4 The square four-helix bundle topology represents a natural structural motif, which has extensively been a target of de novo protein design (Section 1.7).8 0 A n interesting variation of this de novo design concept is offered by the template assembled synthetic protein (TASP) approach introduced by Mutter's group (Section 1.8).6 1.4 Factors Contributing to Protein Folding and Stability A protein must adopt a specific folded three-dimensional structure in order to be active. Different protein conformations arise from rotation about the bonds in the polypeptide backbone and amino acid side chains. Protein folding involves sampling different conformations before 26 arriving at one of the thermodynamically most stable, i f not the most stable, native conformations (global energy minimum). 1 It is noncovalent forces that act on the primary structure to cause a protein to adopt secondary, tertiary, and quaternary structures. Individually, these noncovalent interactions are weak (~5 kcal/mol) as compared with a covalent bond (-100 kcal/mol), and the relative importance of each of these noncovalent interactions is not yet clear. 4 1 However, many of these noncovalent interactions have a substantial effect on the overall 81 topology and stability of a protein making it 5-15 kcal/mol more stable than the unfolded state. There are three main kinds of noncovalent forces: ionic or electrostatic interactions, hydrogen bonding, and hydrophobic interactions, which w i l l be described in the following Sections. In addition, the factors contributing to the stability of a-helices and to four-helix bundles w i l l be presented in Sections 1.5 and 1.6, respectively. 1.4.1 Electrostatic Interactions Proteins have many charged groups that can participate in binding them to each other or stabilizing their own secondary structures, through ionic interactions. For example, glutamic acid has a carboxyl side chain group, which is negatively charged and may be attracted to the positively charged free ammonium group on a lysine side chain. The strength of this interaction has been estimated at -0.5 kcal/mol in model G i n - A l a based peptides. 8 3 Electrostatic interactions and salt-bridges can exist between separate helices or on a single a-helix between residues three to four amino acids apart. However, these ionic interactions are sensitive to changes in p H and salt concentration. 27 The p H of the environment solvating the protein alters the charge of the side chain groups and of the whole protein, and in result the p H greatly affects the availability of ionic interactions. For example, as the p H drops, negatively charged carboxyl groups and neutral amino groups become protonated, and end up neutral and positively charged, respectively. A s the p H rises, protonated carboxyl groups lose their proton and develop a negative charge, and positively charged amino groups end up neutral. Increasing the salt concentration results in a competition for the ionic interactions between the solute and the charged residues of the protein itself, leaving charged residues in the protein without electrostatic partners. This results in an overall reduction of the favorable ionic interactions within the protein. 8 2 1.4.2 Hydrogen Bonds Hydrogen bonds are critical in stabilizing a-helices. 8 4 Hydrogen bonds were first proposed in the early 1920's and form between a highly electronegative atom (N, O, F) and the DC hydrogen covalently linked to another highly electronegative atom. The strength of an average hydrogen bond lies between 2-10 kca l /mol . 1 4 In proteins, hydrogen bonds mainly develop between the N - H and the C = 0 groups on separated amino acids in the peptide backbone, giving rise to a-helices and between adjacent peptide strands in (3-sheets.86 Hydrogen bonds are not limited to backbone interactions and can exist between secondary structural motifs to help stabilize a protein in its active three-dimensional conformation. 28 1.4.3 Hydrophobic Interactions It is generally accepted that hydrophobic interactions are the major forces involved in 87 initializing protein folding and stabilizing the tertiary structures of proteins, although some experimental data supports that hydrogen bonds may be as important for stabilizing protein structures.8 6 Amino acids such as leucine, isoleucine, and phenylalanine have nonpolar side 88 chains, which interact through hydrophobic interactions with other nonpolar groups. Hydrophobic residues interact poorly with water and thus usually point toward the interior of the protein. It is energetically more favorable for nonpolar surfaces to approach each other and cluster together, in result, excluding water molecules from between them. This phenomenon is known as the hydrophobic effect and is a major contributor to the tertiary structure and stability of proteins. 1 4 It is clear that when two hydrophobic residues or regions approach each other and cluster that there is a decrease i n the disorder of the system and hence a loss of entropy, the main energetic factor opposing protein folding. 8 9 However, the solvating water molecules from between the hydrophobic residues get excluded as the cluster forms and result in an increase in the entropy of the surroundings. Overall this process is energetically favorable. 1.5 Factors Contributing to the Stability of a-Helkal Structures a-Helices are dominant structural elements found in proteins. 5 6 Numerous experimental studies have focused on a-helix folding and its sequence-structure relationship. Several groups have worked on deciphering factors that stabilize short, synthetic peptides as models for 29 monomeric helix formation in aqueous solution. ' Electrostatic interactions, hydrogen bonding (discussed in Sections 1.4.1 and 1.4.2, respectively) and intrinsic conformational preferences are important determinants of helix stability. One complication with the study of isolated monomeric helices is that they may not be representative of helices in folded proteins. A n a-helix within a protein may have non-uniform solvent exposure, and varying dielectric constants. In this Section additional factors responsible for a-helix stability including helix capping effects, amino acid propensities, helix chain length, and helix macrodipole effects w i l l be discussed. 1.5.1 Helix Capping A s was mentioned in Section 1.3.2.2 the a-helix contains internal hydrogen bonds between the carbonyl of residue / and the amide proton of residue i+4.49 This network, however, leaves four carbonyl groups at the C-terminus and four amide protons at the N-terminus without hydrogen bonding partners (Figure 1.11).91 Statistical analyses 8 4 and physical-chemical arguments9 1 propose that there is a preference for certain residues to reside at the N - and C -termini that can satisfy the hydrogen bonding requirements. Richardson and Richardson predicted that polar side chains on amino acids near the ends of helices hydrogen bonded to the exposed amides at the ends of the helix. These interactions, known as "capping" interactions, provide thermodynamic stability to alpha helices. N-caps have been shown to stabilize monomeric helices by up to 2 kcal /mol . 9 2 Furthermore, about 70 % of the N-termini of alpha helices take part in capping interactions. Amino acids that serve as useful N-caps include Ser, Asn , Asp and Thr, whereas, A l a , Leu, V a l , l ie, Trp, Arg , G in and Glu are rarely found. 8 4 ' 9 3 30 Figure 1 . 1 1 . Illustration of the Non-Hydrogen Bonded Atoms in the a-Hel ix Amide Backbone. A special form of N-cap utilizing a reciprocal hydrogen-bonding motif is called the "capping box" . 9 4 In this arrangement the hydroxyl oxygen of the N-capping residue / forms a standard N-cap hydrogen bond with the unsatisfied amide nitrogen atom of residue i+3, and the side chain carbonyl atom of the i+3 residue forms a hydrogen bond with the amide of the N-cap. In this assembly two of the four non-hydrogen bonded helix N-terminal amides are satisfied. 9 4 B y comparison, C-capping interactions at the carboxy terminus occur in only 30 % of the helices in proteins of known structure. 8 4 Furthermore, the stabilizing effects of C-capping residues such as His , Lys and A r g were found to be much less than N-caps with stabilization 31 energies of only 0.1-0.4 kca l /mol . 9 2 a Glycine is most frequently found at the C-cap position. Helices terminating in G l y can be classified into two primary motifs based on their hydrogen bonding patterns, the Schellman motif and the alpha-L motif. The Schellman motif involves two "capping" hydrogen bonds in contrast to the one hydrogen bond formed in the alpha-L motif. However, because of glycine's inherent flexibility it is believed to be a better helix-terminating amino acid, than helix stabilizer. 8 4 ' 9 5 1.5.2 Amino Acid Helix Propensities Each amino acid has distinct conformational preferences that lead to stabilization or destabilization of protein secondary and tertiary structures.3 Fasman and Chou established from their studies on proteins of known structure that certain amino acids were commonly found in a -helices while others were more often found in P-sheets.96 Wi th the development of site-directed mutagenesis, it became possible to measure the contributions of individual amino acids to the conformational stability of secondary structural elements and entire proteins themselves. 9 7 Others have developed model systems to measure amino acid propensities in oc-hehces. DeGrado and coworkers adopted a minimalistic approach to design a de novo model for helical proteins that contained a single site into which several amino acids were substituted. They were able to obtain a complete thermodynamic scale for the helix forming propensities of the 20 naturally occurring amino acids. 9 8 Kallenbach and coworkers have used monomeric salt-bridge stabilized peptides as model systems to evaluate amino acid propensities. 9 9 The peptides contained G l u and Lys residues spaced three to four amino acids apart to form ion pairs in the helical conformation. Baldwin's group designed short alanine based peptides as models to - ' ' • -32, . calculate intrinsic helical propensities.100 The diverse experimental methods were in agreement with the qualitative order of helical propensities (outlined below in Figure 1.12), but had poor quantitative agreement between them, in addition to disagreeing with the propensities measured in native proteins.101 Figure 1.12. Ranking of the Helical Propensities of the 20 Naturally Occurring Amino Acids. (strongest helix conformer) Ala > Arg > Leu >Lys > Met > Gin > Glu > De > Phe > Trp > Ser > Tyr > His > Asp > Cys > Asn > Val > Thr > Gly > Pro (weakest helix conformer) Alanine was found to have the largest stabilizing effect and is considered to be the only amino acid to contribute positively to helix stability.1003 It was proposed that the methyl side chain of Ala was sufficient to reduce the conformational freedom of the main chain in the unfolded state. Residues with larger side chains including Leu, Met, and Phe have a similar effect* but their own conformational freedom is restricted in the a-helical state. However, it is the burial of their nonpolar surfaces (hydrophobic effect) that contributes to their helical propensities.102 (3-branched amino acid such as Val, Thr, and lie further destabilize helix formation because of potential clash between a substituent at the y-position in the side chain and the carbonyl oxygen in the polypeptide backbone. Glycine possesses considerable conformational flexibility and is considered to be energetically unfavorable due to the entropic cost of fixing the residue in ahelical conformation. Proline has the lowest helical propensity and is known to induce a kink in helices.104 However, the propensity differences between Ala and Gly fall within a narrow range of about 1 kcal/mol. 33 In order to understand the disagreement between the previously measured helical propensities in peptides and proteins, Scholtz and Pace measured helical propensities in the small protein Rnase-Ti and a monomeric peptide segment (residues 13-29) containing the same helical sequence. 1 0 5 Site-directed mutagenesis and mutant peptides both with residue 21 being substituted individually with the twenty commonly occurring amino acids afforded very similar propensities in each system. The nonpolar amino acids showed excellent agreement, whereas the polar residues, namely His and Asp , were to some extent conflicting. Penel et al. have recently shown that some amino acid residues favored hydrogen bonding with neighboring residues via either main chain-to-side chain or side chain-to-side chain interactions suggesting a potential neighbor-dependent preference for residues in the a -he l ix . 1 0 6 Feng and coworkers, based on the neighbor-dependent propensity values obtained by Penel et al, developed a series of amino acid dyads (a-b) that had a predominant preference for a-helical conformation, whereas their dyad pairs (b-a) had little or no preference. 1 0 7 They also expanded their study to determine propensity variations among amino acids in short, medium, and long a-helices. A l l analyses were performed using data from the Brookhaven Protein Data Bank ( P D B ) . 1 0 8 They chose a subset of helices with chain lengths between 4 and 22 residues. The most significant finding was that the helical composition of the short helices (4-7 residues) was quite different from that of the medium (8-13 residues) and of the long helices (14-22 residues), with Pro being the strongest helix conformer in the short helices. Furthermore, strong helix conformers including A l a , G lu , Leu, and G i n were found to have a strong influence on the preference of neighboring residues adopting helical conformations. Recently Engel and Degrado concluded that amino acids also have very strong position-dependent propensities throughout the length of the helix, propagating at least 15 residues from 34 the N-terminus. 1 0 9 It was previously believed that beyond the first few residues from the N -terminus the average environment would become uniform, leading to isotropic distributions. 1 1 0 1.5.3 Helix Chain Length Helix chain length has a dramatic effect on helix stability. Generally, isolated a-helices contain between 8 and 12 residues, and usually many more are required for stable secondary structures. It was found that increasing the length of the helices from 21 to 35 residues increased the stability of the four-helix coiled coil by 9 kcal/mol per he l ix , 1 1 1 Proteins are stabilized by many other noncovalent interactions and therefore shorter helices have been observed. A recent study by Feng et al. found that the length of a-helices in known proteins from the Brookhaven Protein Data Bank ( P D B ) 1 0 8 varied between three and 77 residues with a total of 10643 helices in the P D B . 1 0 7 The average length of the a-helices was found to be 12.1 residues. Barlow and Thornton have found a similar average of 12.0 residues. 1 1 2 The number of helices containing more than 13 residues in the P D B gradually decreased with only a handful of helices containing more than 47 residues. A number of proteins were found to contain a-helices three residues in length, but were l ikely half-turn helices or irregular helices. 1.5.4 The Helix Macrodipole The hydrogen bonding pattern of a-helices is directional and results in a cumulative helix macrodipole with a charge of +0.5 and -0.5 on the N - and C - termini, respectively. 4 1 This • -35. ' explains the observed preferences of negatively charged amino acids like G l u near the N -terminus and positively charged residues like Lys at the C-terminus of helices to enhance alpha-helix stability. 1 1 3 The presence of a negatively charged amino acid at the N-cap can add some 1.6 kcal/mol of stabilization energy by alleviating the unfavorable effects of the helix macrodipole. 1 1 4 In addition, phosphate moieties have been found to bind frequently at the N -termini of helices to help reduce the destabilizing consequences of the helix macrodipole. 1.6 Factors Contributing to the Stability of Four-Helix Bundles In Sections 1.4 and 1.5 the elements influencing the folding and stability of proteins and a-helices was explained. The stability of four-helix bundles is interdependent on the stability of the constituent helices, however other important contributors include side chain packing, side chain interactions, and helix orientation. 1.6.1 Side Chain Packing When a-helices pack together, the side chains at their interface become buried. For optimum hydrophobic interactions the two interfaces should have complementary surfaces. The "knobs-into-holes" packing (see Figure 1.14) proposed by Crick is one of the universal models used to depict the packing arrangement of protein hydrophobic cores. 6 4 A s was mentioned in Section 1.3.5.1, the side chains of one helix generate "knobs" that can fit into the "holes" formed between side chain residues at positions /, i+3, i+4, and i+7 in a neighboring helix. However, 36. the "knobs-into-holes" model could not explain all of the helical packing interactions observed in some proteins. A n additional "ridges-into-groves" model (see Figure 1.13) was proposed in 1981 to help explain the observed variety of packing arrangements in proteins. 1 1 5 This representation suggests that the surface of an a-helix consists of grooves and ridges, like a screw thread. Wi th 3.6 residues per turn, the side chains of every fourth residue form a ridge with an angle of 26° from the direction of the helical axis. When two helices pack together such that the ridge of one helix fits into the other's groove, the predicted angle is 52° between the two. There are supplementary ridges-into-grooves stacking patterns involving every third residue or even every residue. However, the " i+4" ridge is the most common because the side chains of every fourth residue are more closely aligned than the " i+3" or " i + i " ridges. 1 1 5 Figure 1.13. Simplified Two-Dimensional Illustration of the Packing Between Two a-Helices. Side Chains of Each Hel ix are Shown by Open and Closed Circles, (a) "Knobs-Into-Holes" Packing (b) "Ridges-Into-Grooves" (a) (b) 37 Although both models do explain the packing geometries observed in many proteins, there is still much debate over which model is more accurate, and no model has been entirely accepted. 1.6.2 Interhelical Side Chain Interactions As was mentioned in Section 1.4.1, electrostatic interactions between side chains can exist within helices to help stabilize secondary structures or between side chains of adjacent helices to aid in stabilizing the tertiary conformation. Fairman and coworkers report a 0.6 kcal/mol stabilization energy in heterotetrameric coiled coils by Glu"-Lys+ ion-pair interactions.116 Hodges et al. state a similar stability of 0.5 kcal/mol from a Glu"-Lys+ interhelical ion-pair.117 Several X-ray crystal structures have confirmed ion-pair electrostatic interactions and have shed light on their relevance to coiled coil structure and stability.57'64 1.6.3 Helix Orientation Anti-parallel four-helix bundles represent the most stable tertiary conformation, but only with an additional stabilization energy of about 1 kcal/mol over the parallel assembly.118 The anti-parallel configuration minimizes the unfavorable effects of the helix macrodipole, and is the more common arrangement of helices in helical bundles (see Figure 1.14). 38 Figure 1.14. Schematic Representation of a (a) Parallel and an (b) Anti-Parallel Four-Helix Bundle Protein. (a) C (b) C N N N N C N N N Parallel = Less Favorable Anti-Parallel = More Favorable In summary, noncovalent interactions help in stabilizing protein secondary and tertiary Structures. Many of the above factors need to be considered when designing and synthesizing de novo proteins. Section 1.7, w i l l explain some of the efforts towards de novo protein design, and highlight some of the aspects necessary to create native-like structures. 1.7 De Novo Protein Design Synthesizing and characterizing de novo proteins is another approach used to study the process of protein folding and structure. One of the objectives Of de novo protein design is to design artificial proteins in an attempt to unravel the protein-folding problem, which is the uncertain relationship between the primary and the tertiary structure, by predicting a protein 39 sequence that w i l l adopt a particular fold, and then experimentally testing the prediction. Another area of protein design focuses on obtaining not only models for understanding the challenges of the folding process but to create novel molecules with useful functional activities for medical or industrial applications. 1 1 9 A quantitative understanding of the relative contributions that all of the noncovalent interactions have in controlling protein folding, conformation and stability can be achieved by using this approach. M u c h of the effort has been devoted to measuring the effects of various noncovalent forces, the interactions that are largely responsible for protein folding and stability, in simple model protein systems. 1 2 0 One of the advantages of using de novo proteins is that they can be designed to be much smaller than naturally occurring proteins, yet still maintain many of the elements necessary for folding. De novo design affords the potential to use an incremental approach and to make individual amino acid modifications in the design to evaluate contributions of specific residues to the three-dimensional structure of a protein. This is a significant and essential task since there is a general absence of residue-specific probes that are sensitive to analyzing protein conformations. The daunting circumstance of de novo protein design is choosing "correct" amino acid sequences, since a sequence of 100 residues composed of the 20 commonly occurring amino acids, yields 2 0 1 0 0 possible sequence combinations. Furthermore, the folding of such a protein containing 100 residues has 1 0 1 0 0 possible backbone conformations, assuming 10 possible conformations per residue. 1 2 1 If the polypeptide converts between conformations in only 10"1 3 seconds, a typical time for bond rotation, then it would take 10 7 7 years to test all possible conformations. Therefore, the principles underlying the processes by which proteins adopt a relatively well-defined set of conformations, out of such a large number of possibilities, are only beginning to be understood. 40 This Section w i l l outline some of the main research groups using de novo protein design to study protein folding and stability, and w i l l briefly mention a few potential applications of their designs. 1.7.1 DeGrado Degrado and coworkers began studying de novo design almost two decades ago, and are one of the leaders in this area of research. One of his early goals was to design a native, stable anti-parallel four-helix bundle using an incremental approach. 1 2 2 In the mid 1980's Degrado's group, in collaboration with Eisenberg and his coworkers, designed a four-helix bundle protein that is an idealized version of a naturally occurring class of proteins, which includes myohemerythrin, apoferritin, tobacco mosaic virus coat protein and cytochrome c".5 4'5 8 b'1 2 3 These proteins are composed of four helices connected by three loops with very similar folding patterns even though the protein sequences are quite varied. This suggests that hidden beneath the complexities of different amino acid sequences of these proteins is a common and highly degenerate code that allows them to adopt similar topologies. The first step included designing and synthesizing a helix keeping in mind the characteristics of protein folding and stability. For a four-helix bundle these aspects include helix formation, helix termination, helix-helix packing and loop formation. DeGrado's approach included a minimalistic design in which a minimum number of different residues were used, which strongly favor helix formation. 1 2 2 The helices were designed to be amphiphilic (i.e. when folded they would possess a nonpolar face to interact with neighboring helices and a polar face to maintain water solubility). Leucine was the sole hydrophobic residue and Glu and Lys were incorporated as the hydrophilic residues. In addition, an acetyl group was added at the N -41 terminus and an amide to the C-terminus in order to stabilize the unfavorable helix macrodipole and a glycine residue was included as the first and last residue to break the helix and aid with helix capping. Glycine is regularly found at the ends of helices,84 and also provided an attachment point for the future incorporation of the loop segments. They initially designed and Optimized a peptide, afB (see Table 1.1 for peptide sequences),124 aggregated into homotetramers forming a four-helix bundle protein With the hydrophobic faces packing towards the interior of the bundle. Hydrophobic interactions were determined to be the dominant force in stabilizing the tertiary structure. The second stage of the design process involved the design of a loop to connect two helices of the optimized sequence, otiB. A single proline residue was included into the loop region because of its high tendency to break helices. 1 0 4 This subsequent protein, 0C2B, self-associated into homodimers to form another four-helix bundle. Finally, the last stage included ' ' 1 2 5 connecting two o^B peptides by loop segments to form a single stranded polypeptide, 0C4. 0 4 folded into a stable four-helix bundle and provided a basis for the future of de novo protein design and synthesis. 42 Table 1 . 1 . DeGrado's Peptide Series for the Synthesis of Four-Helix Bundles. Sequence Amino Acid Sequence8 Name Helix: [ G E L E E L L K K L K E L L K G ] Loop: [PRR] a i B : A c - [ H e l i x ] - C O N H 2 a 2 B : Ac-[Hel ix] - [Loop]- [Hel ix] -CONH 2 0C4: M-[Helix]-[Loop]-[Helix]-[Loop]-[Helix]-[Loop]-[Helix]-COOH a 2 G : A c - [ G E V E E L L K K F K E L W K G ] - [ L o o p ] - [ G E T E E L F K K F K E L I K G ] - C O N H 2 a 2 D : A c - [ G E V E E L E K K F H E L W K G ] - [Loop] - [ G E I E E L H K K F K E L I K G ] - C O N H 2 underlined letters represent change from previous sequence. The previously designed proteins appeared to adopt folded states with loosely packed hydrophobic cores such as those found in molten globules. They were found to be stable to chemical denaturants and did fold with the desired topology. However a compact, globular protein that shows all the features of a native-like protein had not yet been synthesized. In order to design a more native-like interior, they replaced a number of the leucine residues in oc2 with more conformationally constrained, (^-branched (Val and He) and aromatic amino acids (Phe and Trp) and synthesized a 2 C . 1 2 4 a oc2C did possess many of the characteristics of native-like proteins, including a cooperative thermal transition between a native-like state at low temperatures and a molten globule-like state at higher temperatures. However, oc2C retained a few properties that were not entirely consistent with the native state, and modifications were made in an attempt to alleviate those characteristics. 43 The incorporation of metal-binding sites and the introduction of polar residues in the original hydrophobic positions to promote specific interactions within the core was the next stage of the design generating oc^D.126 Briefly, the successful incorporation of a Z n + 2 ion complexed by the histidine residues decreased the conformational mobility within the core of (X2D. Titration with l-anilinonaphthalene-8-sulfonate (ANS) provided evidence for a well-packed hydrophobic core, as no binding was detected using fluorescence spectroscopy as was seen with previous models. (X2D also exhibited a cooperative thermal-denaturation cuve, which is coincident with the loss of both secondary and tertiary structure. Although this redesigned variant was more native-like in nature, it did not exist as a true monomer, and thus a truly native designed monomeric protein still remained elusive. Degrado's group has not only focused on de novo designed square bundles, but has also designed several four-stranded coiled coils. Only small differences existed in the peptide sequences of their helical bundles compared to those of their coiled coils, which were categorized into two groups, " c o i l - X L " 1 2 7 and " R L P - X " 1 2 8 resulting from differences in strand numbers. The former of these designs aggregated into tetrameric four-stranded coiled coils, while the latter formed a series of dimeric four-stranded coiled coils. Their designs demonstrated the importance of incorporating Vat , 'Ala and Leu residues into the peptide sequence in order to achieve native-like structures. In addition, the positions of the hydrophobic residues had a significant effect on coiled coil stability. Coiled coil designs will be expanded on in the following Section. DeGrado's early work formed a foundation for the future of de novo design. DeGrado and coworkers have designed some of the most native-like proteins to date,1 2 4 1 5'1 2 9 and have stated the necessity of destabilizing possible alternate folding products for the successful design of a native-like protein. Furthermore, they have focused on introducing functionality into their • . 44 protein models with the incorporation of redox-active cofactors, 1 3 0 D N A binding sites, and transition metals. 1 3 1 Additionally, they have been interested in the function and structure of 132 membrane-active peptides and have synthesized artificial ion channels. 1.7.2 Hodges Hodges and coworkers implemented a minimalistic approach to their de novo design of two-stranded a-helical coiled coils. Because of the simplicity of the coiled coil structure synthesizing and characterizing de novo proteins can be used to study it. The issues that must be addressed in the de novo design of coiled coils are (1) controlling parallel versus anti-parallel chain orientation, (2) controlling the number of helical strands in the assembly (3) maximizing the stability in the shortest possible chain length, and (4) the ability to have selective heterodimerization without homodimenzation. Hodges suggests that the main advantages to studying coiled coils is that they are stabilized by both intrachain and interchain interactions, and only two interacting a-helices are required to introduce the tertiary and quaternary structure. Furthermore, it is these interchain side chain interactions between the a-helices that are the main feature responsible for the folding and stabilization of the three-dimensional structure. A s was explained in Section 1.3.5.1, coiled coils can be considered as a repeating heptad of seven amino acid residues [abcdefg]n, where positions a and d are occupied by hydrophobic residues. Hodges and coworkers discovered that this 3-4 (or 4-3) hydrophobic repeat was continuous throughout the entire 284-residue polypeptide chain of tropomyosin. 6 9 45 From the amino acid sequence of tropomyosin, it was clear that hydrophobicity was a critical issue in the formation and stability of coiled coils. Hodges and coworkers designed and synthesized the first model coiled coil protein and demonstrated that the minimum chain length required for the formation and stabilization of the tertiary structure was 28 residues, based on the heptad repeat Lys-Leu-Glu-Ala-Leu-Glu-Gly. 1 3 4 (3-branched amino acids are generally conserved at the a positions and Leu residues are conserved at the d positions. They went on to study the effects on coiled coil formation and stability by placing hydrophobic amino acids such as Leu, lie, and Val at either the a or d position, in an attempt to address whether (3-branched amino acids at position a make similar 133 contributions to the formation and stability of coiled coil compared to those at position d? From the early studies, it could be concluded that at position a, P-branched amino acids actually stabilize the coiled coil structure relative to Leu residues in terms of the packing effect. In contrast, when mutations were carried out at position d, both lie and Val destabilized the 136 coiled coil structure. This was mainly the result of the hydrophobic effect. Other research included studying the effects of Ala substitutions in the hydrophobic positions in an attempt to alter the aggregation state of a two-stranded coiled coil to a four-stranded one. They established that a very subtle positional effect of a single Ala residue could have a very dramatic effect on molecular aggregation. Furthermore, the effects of additional amino acid substitutions on protein stability and oligomerization state were more recently evaluated.137 46 1.7.3 Hecht Hecht and coworkers have focused on synthesizing de novo proteins from designed combinatorial libraries.138 Success in de novo design of proteins ultimately lies oh the ability to choose amino acid sequences that will generate proteins possessing native-like properties. With numerous sequences available for even a moderately sized polypeptide, composed of the 20 naturally occurring amino acids, Hecht states "that it would not be practical to design and characterize proteins one by one", and that combinatorial methods are to a large extent more 139 attractive. They focused on combinatorial libraries with specific binary patterning of polar and nonpolar residues within the polypeptide in order to favor the formation of amphiphilic folded structures. The locations for the polar and nonpolar residues were defined, however, their identities were varied expansively. Some restrictions were required in the combinatorial library since random sequences will rarely yield proteins with desired native-like characteristics.139'140 The designed a-helical structures followed a sequence periodicity of polar and nonpolar residues that matched the a-helical structural repeat of 3.6 residues per turn. Most of the proteins in the initial collection formed tertiary structures characteristic of molten globules with fluctuating internal cores.141 The second generation library contained proteins of 102 residues in length in comparison to the 74 amino acid chain length of the first generation. This new family contained an abundance of proteins possessing native-like properties including well-resolved signals with good dispersion in the ! H NMR spectrum, indicative of specific tertiary interactions.142 In addition to creating well-ordered a-helical de novo proteins using a binary patterned combinatorial library, Hecht et al. extended this approach to generate cofactor binding 47 proteins, 1 4 3 catalytically active enzymes, self-assembled monolayers, amyloid-like nanofibrils, 1 4 6 and template-directed assemblies of two-dimensional biomaterials. 1 4 7 1.8 Template Use Mutter and Vulleumies have introduced the approach known as template assembly in which peptides are covalently linked to a rigid scaffold. 6 ' 1 4 8 In this method, potential units of secondary structure are attached to a rigid template to minimize the unfavorable loss of entropy upon folding. The template serves to preorganize the peptides and promote folding (see Figure 1.15). The resulting proteins are termed Template Assembled Synthetic Proteins (TASPs) . Figure 1.15. Diagram Displaying Template Assisted Folding. 48 The covalent attachment of the peptides to the template can be achieved by a number of different chemoselective reactions of two mutually reactive functionalities. Some methods rely on the formation of a non-natural (other than amide) bonds at the site of ligation, such as the oxime and hydrazone ligations, and even thioether and thioester ligations have been used. 1 4 9 The most common approach is the use of native chemical ligation. Another advantage of attaching peptides to a template is that it eliminates the need for designing loop regions between the peptide strands. This Section w i l l introduce the work of Mutter and Jensen in detail, and w i l l highlight some additional templates used by other research groups in the synthesis of T A S P s . 1.8.1 Mutter Mutter and coworkers introduced the idea of template assembly to overcome the notorious folding problem in de novo protein design. Their original template (Figure 1.16) was an acyclic peptide template, 1 5 0 which included a Pro-Gly peptide fragment to induce a type JJ j i -turn into the polypeptide chain (template sequence = A c - [ K K K P G K E K ] - C O O H ) . ' 5 0 This turn was essential in preorganizing the peptides for four-helix bundle formation. Peptide sequences (i.e. E A L E K A L K E A L E K L G 1 5 1 ) were designed using the guidelines 1 5 2 established by DeGrado and coworkers by incorporating amino acids with high helical propensity, appropriately spacing hydrophilic and hydrophobic residues to promote amphiphilicity, and positioning Glu and Lys residues four residues apart to encourage intrahelical salt-bridges. The corresponding T A S P s displayed molten globule-like characteristics that were attributed to the intrinsic flexibility of the acyclic peptide template. 49 Figure 1.16. A n Example of Mutter's Peptide Template. K = Lysine E = Glutamic A c i d G = Glycine P = Proline A second generation template was designed to overcome the inherent flexibility of the original acyclic template. After assembling the linear template it was cyclized via the formation of a disulfide bond between the terminal cysteine residues to generate a cyclic peptide (template sequence = A c - [ C K A K P G K A K C ] - C O N H 2 ) . 1 5 1 Anti-parallel four-helix bundles were synthesized in addition to the all-parallel T A S P s , by first modifying the linker moieities on the template and then employing protection and subsequent deprotection methodologies. These new cyclic scaffolds yielded four-helix bundles with more native-like characteristic and since then, additional cyclic peptide templates have been created using conformationally restricted non-natural amino acids. 1 5 4 More recently, new generations of regioselectively addressable functionalized templates (RAFTs) for use in protein design have been synthesized and crystal l ized. 1 5 4 b These novel templates also incorporate non-natural amino acids to reduce side chain conformational 50 flexibility, which aid in creating native-like tertiary structures and represent a powerful tool in protein design. 1.8.2 Jensen Since his group began in 1997, their research interests have been in synthetic bioorganic chemistry, including the de novo design and total chemical synthesis of model proteins, the development of new reagents and reactions, especially for solid phase synthesis, solid-phase oligosaccharide synthesis, solid-phase assisted analysis of oligosaccharides, and combinatorial chemistry. Jensen's group has developed the use of carbohydrates as templates for the de novo design of proteins, and has suggested the name "carboproteins" for the members of this novel class of model proteins. 1 5 5 They suggested that carbohydrates would be promising candidates for templates, as monosaccharides are polyfunctional molecules, pyranose ring forms are relatively rigid, and epimers of sugars are often accessible. They have approached the synthesis of carboproteins in two different ways. Their first generation carboproteins employed a methyl oc-D-galactopyranoside (Galp) derivative as a template. The synthesis of the four-helix bundle was accomplished by first modifying the Galp template such that it could be attached to a resin. Then the peptides were individually built up amino acid by amino acid at the four reactive moieties in a solid phase synthesis. 1 5 5 3 51 There was no characterization of their initial products and they put their efforts toward alternate routes to synthesize larger carboproteins by surpassing the peptide size limit encountered with solid-phase synthesis. The carbopeptide concept was extended in a second generation approach with the goal to prepare protein-sized structures. 1 5 6 A s compared to their first generation design a new strategy was adopted to increase the freedom of the design. The peptides and the templates were prepared and purified prior to the final coupling step. This facilitated not only the preparation of larger structures in good purity, but also allowed the attachment of non-identical peptides and an anti-parallel arrangement of helices to the template. The peptide aldehydes were linked to an aminooxy-functionalized carbohydrate template. The first template they synthesized belonging to this family was methyl 2,3,4,6-tetra-O-Aoa-a-D-Galp (Figure 1.17). 1 5 7 This tetra-functionalized template gave access to four-stranded carbopeptides and ultimately four-helix carboproteins. They synthesized a number of carbopeptides and proteins, and the longer peptide sequences were based on a peptide sequence designed by Mutter . 1 5 1 Characterization of one of Jensen's larger 64 amino acid carboproteins was found to be 67 % a-helical, stable towards chemical denaturants, and possessed some native-like characteristics. 52 Figure 1.17. Jensen's Carboprotein Using a Methyl 2,3,4,6-tetra-O-Aoa-a-D-Galp Template. More recently they have studied the effects of carbohydrate templates on the stability and structure of the resulting carboproteins, and have found that the templates do have a significant effect on tertiary structure. 1 5 8 Furthermore, they have synthesized a carboprotein with one thiol moiety in the aglycon. They have shown by electrochemical studies and in situ Scanning: Tunneling Microscopy (STM) that they form a self-assembled monolayer on gold surfaces. 1 5 9 They report that this is fundamental research to establish electron transport processes of immobilized functional redox proteins and enzymes, mechanisms of coupled chemical and biochemical catalytic processes, and two-dimensional molecular interactions including phase transitions. Future biotechnological perspectives include protein-membrane interactions and drug delivery, biological corrosion, biocompatibility of metallic implants, and biofilms. 53, 1.8.3 Additional Templates The diversity of templates used in the synthesis of T A S P s goes beyond those of Mutter and Jensen, and additional examples of molecular scaffolds include porphyrins, 1 6 0 steroids, 1 6 1 calixarenes, 1 6 2 glycosides, 1 6 3 transition metals, 1 6 4 and cavitands. 1 6 5 1.9 Conclusion This Chapter introduced the basics of protein folding and structure with an emphasis on a-helices and a-helical motifs. The relationship between the primary amino acid sequence and the folded native state has received significant attention but remains unclear, and is known as the protein-folding problem. The ability to design, synthesize and characterize de novo proteins can help us understand how individual amino acids contribute to the stability and structure of a protein, and has emerged as a useful tool in the design of proteins with specific functions. Furthermore, the de novo concept can be expanded by using the T A S P approach, which helps overcome the entropic costs of peptide association. This thesis w i l l present the efforts towards the design and synthesis of native-like de novo proteins using a cavitand macrocycle as template. It w i l l also expand on the synthesis of hetero-T A S P s including an anti-parallel four-helix bundle, and investigate the efficiency of glycine as an N - a n d C-capping residue. 54 1.10 References 1. Anfinsen, C . B . Science 1973,181, 223-230. 2. Mirsky , A . E . ; Pauling, L . Proc. Natl. Acad. Sci. USA 1936, 22,439-447. 3. Chou, P . Y . ; Fasman, G . D . Adv. Enzymoly. 1978,47, 145-148. 4. (a) Karplus, M . Adv. Biophys. 1984, 18, 165-190. (b) Levitt, M . Annu. Rev. Biophys. Bioeng. 1982, 77,251-271. 5. (a) Srinivasan, R.; Fleming, P.J.; Rose, G . D . Meth. Enzym. 2004,383, 48-66. (b) Srinivasan, R.; Rose, G . D . Proteins 2002,47, 489-495. 6. Mutter, M . ; Vuilleumier, S. Angew. Chem. Int. Ed. Engl. 1989, 28,535-554. 7. (a) Taubes, G . Science 1996, 277, 1493-1495. (b) Thomas, P.J.; Bao-He, Q.; Pedersen, P .L . TIBS 1993, 20,456-459. 8. Federov, A . N . ; Baldwin, T .O. J. Biol Chem. 1991, 272, 32715-32718. 9. Anfinsen, C . B . J. Biol. Chem. 1961, 236, 1361-1363. 10. Merrifield, R . B . Science 1986, 232, 341-347. 11. (a) Jaenick, R. Prog. Biophys. Mol. Biol. 1987, 49, 117-237. (b) Komar, A . A . ; Kommer, A . ; Krashenininnikov, I .A.; Spirin, A S . / . Biol. Chem. 1997, 272, 10646-10651. (c) Kramer, G . ; Ramachandrian, V . ; Hardesty, B . Int. J. Biochem. & Cell Biol. 2001, 33, 541-553. 12. Basharov, M . A . J. Cell. Mol. Med. 2003, 7, 223-237. 13. Levinthal, C . J . J . Chem. Phys. 1968, 65, 44-45. 14. D i l l , K . A . Biochem. 1990, 29, 7133-7155. 15. Berkenpas, M . B . ; Lawrence, D . A . ; Ginsburg, D . EMBO J. 1995,14, 2969-2977. 16. Ptitsyn, O . B . FEBS Lett. 1991,285, 176-181. 17. Karplus, M . ; Weaver, D . L . Protein Sci. 1994,3, 650-668 18. Baldwin, R . B . ; Rose, G . D . Trends Biochem. Sci. 1999, 24, 77-83. 19. G\io,Z.;Thimmdlm,D. Folding and Design 1997,2, 377-391. 55 20. D i l l , K . A . Biochemistry 1985, 24, 1501-1509. 21. D i l l , K . A . , Chan, H.S. Nature Struct. Biol. 1997,4, 10-19. 22. Bryngelson, J .D.; Onuchic, J .N . ; Socci, N . D . ; Wolynes, P . G . Proteins 1995,21, 167-195. 23. Onuchic, J .N. ; Nymeyer, H . ; Garcia, A . E . ; Chahine, J . ; Socci, N . D . In Advances In Protein Chemistry - Protein Folding Mechanisms. Matthews, C R ; Academic Press: San Diego, C A , 2000; V o l . 53, pp 88-89. 24. Ara i , M . ; Kuwajima, K . In Advances In Protein Chemistry — Protein Folding Mechanisms. Matthews, C R ; Academic Press: San DiegO, C A , 2000; V o l . 53, pp 209-210. 25. Ptitsyn, O B . J. Protein Chem. 1987, 6, 273-293. 26. Kuwaj ima ,K. Proteins Struct. Fund. Genet. 1989, 6, 87-103. 27. Ptitsyn, O .B . Curr. Opinion Struct. Biol. 1995, 5, 74-78. 28. Wong, K . - P . ; Tanford, C.J. Biol. Chem. 1973,248, 8518-8523. 29. Ohgushi, M . ; Wada, A . FEBS Lett. 1983,164,21-24. 30. Dolgikh, D . A . Gilmanshin, R. I.; Brazhnikov, E . V . ; Bychkova, V . E . ; Semisotnov, G . V . ; Venyaminov, S . Y . ; Ptitsyn, O. B . FEBS Lett. 1981,136, 311-315. 31. (a) Ptitsyn, O .B : Adv. Protein Chem. 1995, 47, 83-229. '(b) Ptitsyn, O.B Trends in Bioch. Sciences 1995, 20, 376-379. 32. Jennings, P . A . ; Wright, P .E. Science 1993, 262, 892-896. 33. Peng, Z . - Y . ; and mm, P.S. Biochemistry 1994,33, 2136-2141. 34. (a) Privalov, P .L . ; G i l l , S.J. Adv. Protein Chem. 1988, 39, 191-234. (b) Ewbank, J.J.; Creighton, T . E . Biochemistry 1993, 32, 3677-3707. 35. Laskey, R . A . ; Honda, B . M . ; M i l l s , A . D . ; Finch, J T . Nature'1978,275,416-420. 36. Jaenicke, R.; L i l i e , H . Arfv. Praf. Cnem. 2000,53,329-401. 37. Feldman, D . E . ; Frydman, J. Curr. Opin. Struct. Biol. 2000,10, 26-33. 38. Ruddon, R . W . ; Bedows, E . J. Biol. Chem. 1997,272, 3125-3128. 39. Kuznetsov, G . ; Nigam, S .K. N. Engl. J. Med. 1998,339, 1688-1695. 56 40. Taken from http://www.genome.g0v/Pages/Hvperior1/D protein.cfm?key=protein (accessed 01/14/06). 41. Creighton, T. Proteins: Structure and Molecular Properties, 2 n d ed.; W . H . Freeman: New York , 1993. 42. Taken from http://www.arikah.com/encvclopedia/iniages/thumb/6/6c/400px-RotationAroundPeptideBond.ipg (accessed 01/14/06). 43. MacArthur, M . W . ; Thorton, J . M . J. Mol. Biol. 1996, 264, 1180-1195. 44. (a) Fischer, G . Chem. Soc. Rev. 2000, 29, 119-127, (b) Reimer, U . ; Fischer, G . Biophys. Chem. 2002, 96,203-212. 45. Ramachandran, G . N . Adv. Protein Chem. 1968, 23, 283-438. 46. Taken from http://employees.csbsju.edu/hjakubowski/classes/ch331/prOtstructure/ rama.gif (accessed 04/21/06). 47. Taken from http://www.sciencecollege.co.uk/SC/biochemicals/bsheet.gif (accessed 01/14/06). 48. (a) Aurora, R.; Srinivisan, R.; Rose, G . D . Science 1994,264, 1126 - 1130. (b) Ishikawa, K . ; Nakamura, H . ; Morikawa, K . ; Kimura, S.; Kanaya, S. Biochemistry 1993, 32, 7136-7142. (c) Ueda, T.; Tamura, T.; Maeda, Y . ; Hashimoyo, Y . ; M i k i , T.; Yamada, H . ; Imoto, I. Protein Eng., 1993, 6, 183-187. 49. Pauling, L . ; Corey, R . B . ; Branson, H .R . Proc. Natl. Acad. Sci. U.S.A. 1951, 37, 205 -210. 50. Taken from http://www.sciencecollege.co.uk/SC/biochemicals/ahelix.gif (accessed 05/02/06). 51. Richarson, S.S.Adv. Protein Chem. 1981,34,167-339. 52. Scholtz, J . M . ; Marqusee, S.; Baldwin, R . L . ; York , E . J . ; Stewart, J . M . , Santoro, M . ; Bolen, D . W . Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 2854-2858. 53. (a) Wada, A . Adv. Biophys. 1976, 9, 1-63. (b) H o i , W . G . J . ; van Duijnen, P.T.; Berendsen, H.J .C . Nature 1978,273,443-446. 54. (a) DeGrado, W.F . ; Kezdy F.J . ; Kaiser, E .T . J.Am. Chem. Soc. 1981,103, 679-681. (b) Eisenberg, D . ; Weiss, R . M . ; Terwilliger, T . C . Nature 1982, 299, 371-374. 55. Mathews, F.S. ; Bethge, P . H . ; Czerwinski, E.W.J. Biol. Chem. 1979,254,1699-1706. 56. Kohn, W . D . ; Mant, C.T. ; Hodges, R S . J. Biol. Chem. 1997,272, 2583-2586. 57 57. (a) Glover, J.N.M.; Harrison, S.C. Nature 1995, 373, 257-261. (b) G'Shea, E.K.; Klemm, J.D ; Kim, P.S.; Alber, T.A. Science 1991,254, 539-544. 58. (a) Presnell, S.R.; Cohen, F.E. Proc. Natl. Acad. Sci. USA 1989, 86, 6592-6596. (b) Weber, P. C ; Salemme F.R. Nature 1980, 287, 82-84. 59. (a) Berg, J.M. Ann. Rev. Biophys. Chem 1990, 19, 405-420. (b) Pavletich, N.P.; Pabo, C O . Science 1991, 252, 809-817. 60. (a) Brennan, R.G.; Matthews, B.W. J. Biol: Chem. 1989, 264, 1903-1906. (b) Matthews, B.J.; Ohlendorf, D.H.; Anderson, W.F.; Takeda, Y. Proc. Natl. Acad. Sci. USA 1982, 5, 1428-1432. 61. Ellenberger, T.E. , Brandl, C.J.; Struhl, K.; Harrison, S.C. Cell 1992, 71, 1223-1237. 62. Harris, N.L.; Presnell, S.R.; Cohen, F.E. J. Mol. Biol. 1994,236, 1356-1368. 63. Crick, F.H.C. Nature 1952,170, 882-883. J 64. Crick, F.H.C. Acta Crystallogr. 1953, 6, 689-697. 65. Harbury, P.B.; Zhang, T.; Kim, P.S.; Alber, T. Science 1993, 262, 1401-1407. 66. Walshaw, J.; Woolfson, D.N. Prot. Sci. 2001,10, 668-673. 67. (a) Cohen, C ; Parry, D.A.D. Trends Biochem. Sci. 1986, 11, 245-248. (b) Cohen, C ; Parry, D.A.D. Proteins: Struct. Func. & Genet. 1990, 7, 1-15. (c) Martin-Zanca, D.; Hughes, S.H., Barbacid,M. Nature 1986,319, 743-748. 68. Banner, D.W.; Kokkinidis, M . ; Tsernoglou, D. J. Mol. Biol. 1987,196, 657-675. 69. Hodges, R.S.; Sodek, J.; Smillie, L.B.; Jurasek, L. Cold Spring Harbor Symp. Quant. Biol. 1972,37,299-310. 70. Taken from http://en.wikipedia.Org/wiki/Image:Coiledcoil-wheelcartoon.gif (accessed 06/09/06). 71. McLachlan, A.D.; Stewart, M . J. Mol. Biol. 1975, 98, 293-304. 72. 0'Shea,E.K.; Rutkowski, R.; Kim, P.S. Science 1989,243, 538-542. 73. Landschulz, W.H.; Johnson, P.F.; McKnight, S.L. Science 1988,240, 1759-1764: 74. Harbury, P.B.; Kim, P.S.; Alber, T. Nature 1994, 37i, 80-83. 75. (a) Ward, K.B.; Hendrickson, W.A.; Kippenstein, G.L. A/a^re 1975, 257, 818-821. (b) Stenkamp, R.E.; Sieker, L .C. ; Jensen, L .H. ; McQueen, J.E. Jr Biochemistry 1978,17, 2499-2504. 58 76. Banyard, S.H.; Stammers, D.K.; Harrison, P.M. Nature 1978, 271, 282-284. 77. (a) Bloomer, A .C. ; dampness, J.N.; Bricogne, G.; Staden, R.; Klug, A. Nature 1978, 276, 362-368. (b) Stubbs, G.; Warren, S.; Holmes, K. Nature 1977,267, 216-221. 78. Weber, P.C.; Bartsch, R.G.; Cusanovich, M.A.; Hamlin, R.C.; Howard, A.; Jordan, S.R.; Kamen, M.D.; Meyer, T.E.; Weatherford, D.W.; Xuong, N.-H.; Salemme, F.R. Nature 1980,256,302-304. 79. Harrison, C.J.; Hayer-Hartl, M . ; Di Liberto, M . ; Haiti, F.U.; Kuriyan, J. Science 1997, 276,431-435. 80. (a) DeGrado, W.F.; Wasserman, Z.R.; Lear, J.D. Science 1989, 243, 622-628. (b) Heeht, M.H.; Richardson, J.S.; Richardson, D . C , Ogden, R.C. Science 1990, 249, 884-891. 81. Pace.C.N. Trends Biochem. Sci. 1990,15, 14-17. 82. Marqusee S.; Baldwin R.L. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 8898-8902. 83. (a) Scholtz, J.M.; Qian, H.; Robbins, V .H. ; Baldwin, R.L. Biochemistry 1993, 32, 9608-9678. (b) Smith, J.S.; Scholtz, J.M. Biochemistry 1998, 37, 33-40. 84. Richardson, J.S.; Richardson, D.C. Science 1988, 240, 1648-1652. 85. Huggins, M.L. Angew. Chem. Int. Ed. Engl. 1971,10, 147-208. 86. Pace, C.N.; Shirley, B.A.; McNutt, M . ; Gajiwala, K. FASEB J. 1996,10, 75-83. 87. DeGrado, W.F.; Lear, J.D. J. Am. Chem. Soc. 1985,107, 7684-7689. 88. Privalov, P.L.; Gill, S.J. Adv. Prot. Chem. 1988,39, 191-234. 89. Doig, A.J.; Sternberg,M.J.E. Prot. Sci. 1995,4, 2247-2251. 90. (a)Padmanabhan S. et al. Nature 1990, 344, 268-270. (b) Merutka G.; Stellwagen E. ; Biochemistry, 1990, 29, 894-898. 91. Presta, L .G. ; Rose, G.D. Science 1988, 240, 1632-1641. 92. (a) Doig, A.J.; Baldwin, R. L . Protein Sci. 1995, 4, 1325-1335. (b) Lyu, P . C ; Zhou, H.X.; Jelveh, N.; Wemmer, D.E.; Kallenbach, N.R. J. Am. Chem. Soc. 1992,114, 6560-6562. (c) Padmanabhan, D.; Baldwin, R.L. Protein Sci. 1994, 3, 1992-1997. 93. Parker, M.H.; Hefford, M.A. Protein Eng. 1997, 70,487-496. 94. Harper, E.T.; Rose, G.D. Biochemistry 1993,32, 7605-7609. 95. Dasgupta, S.; Bell, J.A. Int J. Pept. Protein Res. 1993, 41,499-511. 59 96. Chou, P.Y.; Fasman, G.D. Annu. Rev. Biochem. 1978, 47, 251-276. 97. (a) Fersht, A.R.; Serrano, L. Curr. Opin. Struct. Biol. 1993, 3, 75-83. (b) Matthews, B. W.Anm*. Rev. Biochem. 1993,62, 139-160. 98. O'Neil, K.T.; DeGrado, W.F. Science 1990,250, 646-651. 99. Lyu, P.C.; Liff, M.I.; Marky, L.A.; Kallenbach, N.R. Science 1990,250,669-673. 100. (a) Chakrabartty, A.; Kortemme, T ; Baldwin, R.L. Protein Sci. 1994,3, 843-852. (b) Rohl, C.A.; Chakrabarty, A.; Baldwin, R.L. Protein Sci. 1996, 5, 2623-2637. 101. (a) Chakrabartty, A.; Baldwin, R.L. Adv. Protein Chem.1995, 46, 141-176. (b) Qian, H.; Chan, S.I. J. Mol. Biol. 1996, 261, 279-288. 102. Blaber, M . ; Zhang, X.J.; Matthews, B.W. Science 1993, 260, 1637-1640. 103. Cornish, V.W.; Kaplan, M.I.; Veenstra, D.L.; Kollman, P.A.; Schultz, P.G. Biochemistry 1994,33, 12022-12031. 104. Piela, L . ; Nemethy, G.N.; Scheraga, H.A., Biopolymers 1987,26, 1587-1600. 105. Myers, J.K.; Pace, C.N.; Scholtz, J.M. Biochemistry 1997, 36, 10923-10929. 106. PeneLS.; Hughes, E. ; Doig, A.J. J. Mol Biol. 1999, 287, 127-143. 107. Wang, J.; Feng, J.-A. Protein Eng. 2003,16, 799-807. 108. Bernstein, F.C.; Koetzle,T.F.; Williams, G.J.B.; Meyer, E.F.; Brice, M.D.; Rodgers, J.R.; Kennard, O.; Shimanouchi,T.; Tasumi, M . J. Mol. Biol. 1977,11, 535-542. 109. Engel, D.E.; DeGrado, W.F.; J. Mol. Biol. 2004, 337,1195-1205. 110. Kumar, S.; Bansal, M . Proteins: Struct. Fund. Genet. 1998, 37,460-476; 111. Fairman, R.; Chao, H.-G.; Mueller, L . ; Lavoie, T.B.; Shen, L ; Novotny, J.; Matsueda, G.R. Protein Sci. 1995,4,1457-1469. 112. Barlow, D.J.; Thornton, 7.M. J. Mol Biol. 1988, 207,601-619. 113. Armstrong, K . M . ; Baldwin, R.L. Proc. Natl. Acad. Sci. USA 1993, 90, 11337-11340. 114. Serrano, L.; Fersht, A.R. Science 1989,342,296-299. 115. (a) Chothia, C ; Levitt, M . ; Richardson, D. J. Mol Biol. 1981,145, 215-250. (b) Chothia, C. Ann. Rev. Biochem. 1994, 53, 537-572. 60 116. Fairman, R.; Chao, H.-G.; Lavoie, T.B.; Villafranca, J.J.; Matsueda, G.R.; Novotny, J. Biochemistry 1996,35, 2824-2829. 117. Zhou, N.E ; Kay, C M . ; Hodges, R.S. Protein Eng. 1994, 7, 1365-1372. 118. Robinson, C.R.; Sligar, S.G. Protein Sci. 1993, 2, 826-837. 119. (a) Kaplan, J.; DeGrado, W.F. Proc. Natl. Acad. Sci. USA 2004,101,11566-11570. (b) Mason, J.M.; Arndt, K . M . Chem. Biochem. 2004, 5, 170-176. (c) Pinto, Dieckmann, G.R.; Ghandi, C.S.; Papworth, C.G.; Braman, J.; Shaughnessy, M.A.; Lear, J.D.; Lamb, R.A.; DeGrado, W.F. Proc. Natl. Acad. Sci. USA 1997, 94, 11301-11306. (d) Lear, J. D.; Gratkowski, H.; DeGrado, W.F. Biochem. Soc. Trans. 2001, 29, 559-564. (e) Rossi, P.; Tecilla, L. ; Baltzer, P. Chem. Eur. J. 2004,10,4163-4170. 120. (a) Lai, J.R.; Fisk, J.D.; Weisblum, B.; Gellman, S.H. J. Am. Chem. Soc. 2004, 126, 10514-10515. (b) Chen, Y.; Mant, C.T.; Hodges, R.S.J. Peptide Res. 2002, 59, 18-33. (c) Hill, R.B.; Hong, J . -K; DeGrado, W.F. J. Am. Chem. Soc. 2000, 722, 746-747. (d) Lu, S.M.; Hodges, R.S. Protein Sci. 2004,13, 714-726. 121. Creighton, T.E. Trends in Biochem. Sci. 1997, 22, 6-10. 122. Ho, S.P.; DeGrado, W.F. J. Am. Chem. Soc. 1987,109, 6751-6758. 123. Argus, P.; Rosmann, M.G. , Johnson, J.E; Biochem. Biophys. Res. Commun. 1977, 75, 83-86. 124. (a) Raleigh, D.P.; DeGrado, W.F. J.Am. Chem. Soc. 1992, 114, 10079-10081. (b) Hill, R.B.; DeGrado, W.F. J. Am. Chem. Soc. 1998,720, 1138-1145. 125. (a) DeGrado, W.F.; Raleigh, D.P.; Handel, T .M. Curr. Opin. Struct. Biol. 1991, 1, 984-993. (b) DeGrado, W.F.; Wasserman, Z.R.; Lear, J.D. Science 1989,243, 622-628. 126. (a) Raleigh, D.P.; Betz, S.F.; DeGrado, W.F. J. Am. Chem. Soc. 1995, 117, 7558-7559. (b) Handel, T .M. ; DeGrado, W.F. J. Am. Chem. Soc. 1990,112, 6710-6711. 127. Betz, S.F.; DeGrado, W.F. Biochemistry 1996, 35, 6955-6962. 128. Betz, S.F.; Liebman, P.A.; DeGrado, W.F. Biochemistry 1997,36, 2450-2458. 129. Betz, S.F.; Raleigh, D.P.; DeGrado, W.F. Curr. Opin. Struct. Biol. 1993, 3, 601-610. 130. DeGrado, W.F. Proc. Natl. Acad. Sci. USA 2000, 97, 6298-6305. 131. DeGrado, W.F. J. Am. Chem. Soc. 2001,123, 12749-12757. 132. (a) Pinto, L .H. ; Dieckmann, G.R.; Ghandi, C.S.; Papworth, C.G.; Braman, J.; Shaughnessy, M.A.; Lear, J.D. Proc. Natl. Acad. Sci. USA 1997, 94, 11301-11306. (b) DeGrado, W.F. J. Am. Chem. Soc. 1997, 779,3212-3217. 61 133. Kohn, W.D.; Hodges, R.S. Trends in Biotech. 1998, io~, 379-389, 134. (a) Hodges, R.S.; Saund, A.K.; Chong, P.C.S.; St.-Pierre, S.A.; Reid, R.E. J. Biol. Chem. 1981, 256, 1214-1224. .(b) Talbot, J.A.; Hodges, R.S. Acc. Chem. Res. 1982,15, 224-230. 135. Hodges, R.S.; Zhou, N.E.; Kay, C M . ; Semchuk, P D . Peptide Res. 1990, 3, 123-137. 136. Hodges, R.S. Biochem. Cell. Biol. 1996,14, 133-154. 137. (a) Tripet, B.; Wagschal, K.; Lavigne, P.; Mant, C.T.; Hodges, R.S. J. Mol. Biol. 2000, 300, 377-402. (b) Wagschal, K.; Tripet, B.; Hodges, R.S. / . Mol. Biol. 1999, 285, 785-803. (c) Kwok, S .C; Hodges, R.S. J. Biol. Chem. 2003, 278, 35248-35254. 138. Beasley, J.R.; Hecht, M.H. J. Biol. Chem. 1997, 272, 2031 -2034. 139. West, M.W.; Hecht, M.H. Prot. Sci. 1995, 4, 2032-2039. 140. Moffet, D.A.; Hecht, M . H. Chem. Rev. 2001,101, 3191-3204. 141. (a) Roy, S.; Helmer, K.J.; Hecht, M.H. Fold. Des. 1997, 2, 89-92. (b) ROsenbaum, D.M.; Roy, S.; Hecht, M.H. J.Am Chem. Soc. 1999, 121, 9509-9513: (c) Roy, S.; Hecht, M.H. Biochemistry 2000, 39, 4603- 4607. 142. (a) Wei, Y.; Kim, S.; Fela, D ; Baum, J.; Hecht, M.H. Proc. Natl. Acad. Sci. USA 2003, 100, 13270-13273. (b) Wei, Y.; Lui, T.; Sazinsky, S.L.; Moffet, D.A., Pelczer, I.; Hecht, M.H. Prot. Sci. 2003,12, 92-102. 143. (a) Rojas, N.R.; Kamtekar, S.; Simons, C.T.; McLean, J.E., Vogel, K . M . ; Sprio, T.G.; Farid, R.S., Hecht, M.H. Prot. Sci. 1997, 6, 2512-2524. (b) Moffet, D.A.; Certain, L.K.; Smith, A.J.; Kessel, A.J.; Beckwith, K.A.; Hecht, M.H. J. Am. Chem: Soc. 2000, 122, 7612-7613. (c) Moffet, D.A.; Case, M.A.; House, J.C.; Vogel, K.; Williams, R., Spiro, T.G.; McLendon, G.L.; Hecht, M.H. J. Am. Chem. Soc. 2001,123, 2109- 2115. (d) Moffet, D.A., Foley, J.; Hecht, M.H. Biophys. Chem.2003,105, 231-239. 144. Wei,Y:,llecht,M.H.. Protein Eng. Des,Sel. 2003, 77,67-75. 145. Xu, G.; Wang, W.; Groves, J.T.; Hecht, M.H. Proc. Natl. Acad. Sci. USA 2001, 98, 3652-3657. 146. (a) West, M.W.; Wang, W.; Patterson, J.; Mancias, J.D.; Beasley, J.R.; Hecht, M.H. Proc. Natl. Acad. Sci. USA 1999, 96, 11211-11216. (b) Wang, W.; Hecht, M.H. Proc. Natl. Acad. Sci. USA 2002, 99, 2760-2765. 147. Brown, C.L.; Aksay, LA.; Saville, D.A.; Hecht, M.H. / . Am. Chem. Soc. 2002, 124, 6846-6848. 148. Mutter, M . Trends Biochem. Sci. 1988,13, 260-265. 62 Rose, K. J. Am. Chem. Soc. 1994,116, 30-33. (a) Mutter, M ; Tuchscherer, G. Macromol. Chem. Rapid. Commun. 1988, 9, 437-443. (b) Mutter, M . ; Altmann, E.; Altmann, K.; Hersperger, R.; Koziej, P.; Nebel, K.; Tuchscherer, G.; Vuilleumier, S. Helv. Chim. Acta 1988, 71, 835- 847. Mutter, M ; Tuchscherer, G.; Miller, C ; Altmann, K.; Carey, R.L; Wyss, D.F.; Labhardt, A . M . ; Rivier, J.E. J. Am. Chem. Soc. 1992,114, 1463-1470. DeGrado, W.F. Adv. Prot. Chem. 1998,39, 51-124. Nyanguile, O; Mutter, M . ; Tuchscherer, G. Peptide Sci. 1994,7,9-16. (a) Tuchscherer, G.; Domer, B,; Sila, U.; Kamber, B.; Mutter, M . Tetrahedron 1993,49, 3559-3575. (b) Peluso, S.; Ruckle, T.; Lehmann, C ; Mutter, M . ; Peggion, C ; Crisma, M . Chem. Biochem. 2001, 2, 432-437. (a) Jensen, K.J.; Barany, G. J. Pept. Res. 2000, 56, 3.-11. (b) Jensen, K.J.; Brask, J. Cell. Mol. Life. Sci. 2002, 59, 859-869. Brask, J.; Jensen, K.J. J Pept. Sci. 2000 6, 290-299. Brask, J.; Jensen, K.J. Bioorg. & Med. Chem. Lett. 2001,77,697-700. Brask, J.; Dideriksen, J.M.; Nielsen, J.; Jensen, K.J. Org. & Biomol. Chem. 2003, 1, 2247. 159. (a) Brask, J.; Wackerbarth, H.; Jensen, K.J. Bioelectrochemistry 2002, 56", 27-32. (b) Brask, J.; Wackerbarth, H.; Jensen, K.J. J. Am. Chem. Soc. 2003, 725, 94. 160. (a) Sasaki, T.; Kaiser, E.T.; J. Am, Chem. Soc. 1989, 111, 380-381. (b) Sasaki, T.; Kaiser, E.T. Biopolymers 1990, 29, 79-88. (c) Akerfeldt, K.S.; Kim, R.S.; Camac, D.; Groves, J.T.; Lear, J.D., DeGrado, W.F. J.Am. Chem. Soc. 1992,114, 9656-9657. (d) Mihara, H., Nishino, N.; Hasegawa, R.; Fujimoto, T. Chem. Lett. 1992, 9, 1805-1808. 161. (a) Hirschmann, R.; Sprengeler, P.A.; Kawasaki, T.; Leahy, J.W.; Shakespeare, W.C.-Smith III, A.B. J. Am. Chem. Soc. 1992, 114, 9699-9701. (b) L i , H.; Wang, L.X. Org. Biomol. Chem. 2003, / , 3507-3513. 162. (a) Hamuro, Y.; Calama, M.C.; Park, H.S.; Hamilton, A.D. Angew. Chem. 1997, 109, 2797-2800. (b) Wong, K.A.; Jacobsen, M.P.; Winzor, D.J.; Fairlie, D.P. J. Am. Chem. Soc. 1998,120, 3836-3841. 163. Hirschmann, R.; Nicolaou, K.C.; Pietranico, S.; Leahy, E . M . ; Salvino, J.; Arison, B.; Cichy, M.A.; Spoors, P.G.; Shakespeare, W.C.; Sprengeler, P.A.; Hamley, P.; Smith III, A.B.; Reisine, T.; Raynor, K.; Maechler, L. ; Donaldson, C ; Vale, W.; Freidinger, R.M.; Cascieri, M.R., Strader, C.D.R. J. Am. Chem. Soc. 1993,115, 12550-12568. 63 164. (a) Ghadiri, M.R.; Soares, C ; Choi,C. J. Am. Chem. Soc. 1992,114,4000-4002. (b) Mutz, M.W.; Case, M.A.; Wishart, J.F.; Ghadiri, M.R.; McLendon, G.L. J. Am. Chem. Soc. 1999,121, 858-859. 165. (a) Gibb, B. C ; Mezo, A.R.; Causton, A.S.; Fraser, J.R.; Tsai, F.C.S.; Sherman, J.C. Tetrahedron 1995, 51, 8719-8732. (b) Mezo, A.R.; Sherman, J.C. J. Am. Chem. Soc. 1999,121, 8983-8994. (c) Causton, A.S.; Sherman, J.C. J. Pep. Sci. 2002,6, 275-282. 64 CHAPTER 2: Design, Synthesis and Characterization of Caviteins LG3, LG2, NG3, and NG2: An Investigation of Native-like Structure* 2.0 Introduction Chapter 1 introduced the basics of protein folding and structure with an emphasis on oc-helices and a-helical motifs. It also illustrated how designing, synthesizing and characterizing de novo proteins can be used to study the protein-folding problem. De novo protein design provides a model system in which the contributions of individual amino acids to the stability of the tertiary structure can be studied. This Chapter will describe the design, synthesis and characterization of several caviteins, which were designed in order to deduce and analyze native-like structures. Section 2.1 of this Chapter will outline the background and rationale for the study of native-like structure. Section 2.2 will be focus on the synthesis of the caviteins and a discussion of the experimental results. Section 2.3 will summarize the results from this Chapter and provide some conclusions from the analysis of the experimental data. Lastly, Section 2.4 is the experimental, and Section 2.5 contains a list of references for this Chapter. * "A version of this Chapter will be submitted for publication. Huttunen, H. and Sherman, J.C. An Investigation into the Native-Like Properties of De Novo Designed Four-Helix Bundle Proteins." 65 2.1 Rationale for the Investigation of Native-like Structure As was explained in Chapter 1, a native-like protein is defined by a unique tertiary structure with specific side chain packing in the hydrophobic core. A molten globule-like protein, on the other hand, is defined by a more mobile hydrophobic core which results in several protein structures sampling different conformations of similar energy. One main obstacle encountered in the design and synthesis of native-like de novo proteins, which is a challenging task in itself, is the lack of information available to diagnose the respective native-like properties. There are few experiments which can provide nearly precise evidence to distinguish between a native-like and a molten globule-like protein. For example, the primary data used to support a native-like tertiary structure is by the examination of sharp, dispersed signals in the amide region of the 'H NMR spectrum. Therefore, the development of decisive characterization techniques for distinguishing between native-like and molten globule-like proteins is much needed. Previously in our lab numerous de novo four-helix bundles have been synthesized with a variety of peptide sequences attached via their N- or C-termini resulting in tertiary structures with varying degrees of native-like properties. A former member of the Sherman lab, Mezo, had previously synthesized and partially characterized two caviteins, LG2 and LG3, by CD, GuHCl denaturation and *H NMR spectroscopy.18 Mezo was interested in investigating the effects of the linker, between peptide helices and the cavitand template, on the tertiary structures of the resulting caviteins. Although, LG2 and LG3 exhibited many native-like characteristics, the conclusive evidence for a native-like structure remained nebulous! The goals of this work were to determine whether the caviteins, LG2 and LG3, were indeed native-like in structure, and to investigate methods to characterize a native-like tertiary structure. More specifically, we were 66 interested in probing the subtle motion of the side chains within the hydrophobic cores of LG2 and LG3, which was not a simple task. Our approach, that could be used to substantiate the native-like structures of the LG2 and LG3, was to design corresponding caviteins that would exhibit non native-like properties, without significantly altering the original peptide sequences (lg2 and lg3). In this way, we sought to perturb the putative native-like structures of LG2 and LG3 to see if we could diagnose this loss in conformational specificity in the synthesis and characterization of NG2 and NG3. By replacing the leucine residues of LG2 and LG3 with norleucine residues in NG2 and NG3 respectively, one should be able to detect a reduction in native-like character. Norleucine units have unbranched side chains and should result in a more poorly packed protein, and hence a molten globule-like structure. The potentially reduced native-like structures of NG2 and NG3, could then by characterized by the observation of broad signals in the amide regions of their ] H NMR spectra. 2.2 Results and Discussion This Section will detail the design and synthesis of cavitand 5, peptides 6, 7, 8, and 9, and of caviteins 10, 11, 12, and 13. The caviteins will be characterized by physical techniques followed by a discussion of the characterization data. 67 2.2.1 Rationale for Using Cavitands as Templates for De Novo Protein Design As was mentioned in Chapter 1, template assembly can be applied to de novo protein design in order to help overcome the entropic costs of peptide association. An efficient template should preorganize the peptides into a desired folding pattern and be easily integrated into the design. Cavitands were first introduced by Cram and coworkers in 1982.2 They are rigid organic macrocycles with an enforced cavity that have been shown to bind guests. Cavitands serve as useful templates for de novo protein synthesis for a number of reasons. First, the extreme rigidity of the cavitand macrocycle would be expected to limit the degrees of freedom of the constituent peptides, and, in result, stabilize the overall tertiary structure. Some of the problems associated with other templates in the synthesis of native-like four-helix bundles has been their inherent flexibility.4 Secondly, the thiol functional groups located on the upper rim of the cavitand are approximately 7A apart,5 which are fitting for four-helix bundle proteins where the interhelical distances extend between 7-14 A. 6 Furthermore, the thiol functionalities on cavitand 5 are nucleophilic, and thus serve as viable moieties for peptide attachment to efficiently afford a template assembled synthetic protein (TASP). As was mentioned in the introduction, the use of a template also overcomes the large entropic costs of bringing peptide strands close together in space. Furthermore, the cavitand template promotes the peptides strands into a desired parallel bundle and the size of a cavitand can be modified to accommodate different sized bundle structures. The focus of this thesis is on the four-helix bundle motif, and thus only cavitands to which four peptides can be attached will be described. 68 2.2.1.1 The Synthesis of Methyl-Footed Arylthiol Cavitand 5 The synthesis of methyl-footed arylthiol 5 was accomplished through four steps following literature procedures (see Scheme 2.1).2a'7 Briefly, the first reaction included stirring a mixture of resorcinol and acetaldehyde in the presence of methanol and hydrochloric acid. The reaction mixture was stirred for a week to precipitate resorcinarene 2 (77 %). The bromination of macrocycle 2 with Af-bromosuccinimide (NBS) in 2-butanone gave tetrabromoresorcinarene 3 (61 %). This macrocycle was then treated with bromochloromethane in the presence of potassium carbonate to yield bromocavitand 4 (48 %). Bromocavitand 4 was then treated with n-butyllithium and quenched with Sg to give methyl-footed tetrathiol 5 (60 %) and approximately 15 % of the corresponding tris-thiol (not shown in Scheme 2.1). Methyl-footed arylthiol 5 was used in impure form for subsequent cavitein reactions, and the resulting caviteins were purified by reversed phase HPLC. Scheme 2.1. The Synthesis of Methyl-Footed Arylthiol Cavitand 5. 69 2.2.2 Peptide Design The peptide sequences were designed by following many of the design features outlined in Chapter 1. DeGrado's research in the area of de novo protein design, particularly in the synthesis and study of four-helix bundles, provided a foundation for the design of the peptides sequences. DeGrado addressed the complications of protein folding with a minimalistic approach.9 In a few words, this included designing peptides that had a minimum number of different amino acid residues with high helical propensities/The peptides were also designed to be amphiphilic (i.e. the helix would contain a hydrophobic core and hydrophilic surfaces when folded). The primary sequence for peptide 6 to be covalently linked to cavitand 5 is shown in Figure 2.1, and the design features will be addressed below it-Figure 2.1. Primary Sequence for Peptide 6 Used in the Synthesis of Cavitein 10. N — ; • — > C [ G G] - E E L L K K L E E L L K K G 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Apart from glycine, the design included only amino acids with high helical propensities to promote a stable a-helix when folded. To maintain simplicity, no more than four different types of amino acids were used, and the peptide sequence was only fourteen residues in length. The minimum chain length required for stable a-helix formation is r8 -12 residues.10 In accord with DeGrado's design, the helices were designed to be amphiphilic, and this would result in a . -70 ' ' • • hydrophobic core and a hydrophilic exterior in the subsequent four-hehx bundle (see Figure 2.2 for the helical wheel diagram showing the formation of a hydrophobic core in the four-helix bundle). As was explained in Chapter 1, a-helices develop a helix macrodipole in the direction of the C-terminus and thus placing positively charged lysine residues near the C-terminus would help reduce the effect of the helix macrodipole. Two or three glycine residues were added to the N-terminus to serve as a linker between the template and the a-helical segment of the peptide (the effects of the linker on cavitein stability and structure were evaluated by Mezo and Seo, previous members of the Sherman group),18'11 and the glycine at the C-terminus was incorporated as a C-capping residue. Furthermore, a-helices are composed of 3.6 residues per turn and thus oppositely charged residues (E and K) were placed three and four amino acids apart to encourage intrahelical salt-bridges. The main driving force for the formation of the four-helix bundle (four peptide 6 molecules linked to cavitand 5) was expected to be hydrophobic bundling, since the resulting cavitein was studied in an aqueous environment. Figure 2.2. Helical Wheel Diagram of Four Strands of Peptide 6 with the Helices Oriented in Parallel, (reader is looking down the helical axes from C- to N-termini) H 9 0 The helical wheel diagram clearly illustrates the formation of a hydrophobic interior with the Leu residues lying on top of each other in individual helices and the hydrophilic Glu and Lys residues exposed to the aqueous environment aiding in water-solubility. The diagram also suggests the formation of possible interhelical salt-bridges between Glu and Lys residues of adjacent helices. Peptide 6 was designed to promote the formation of a stable a-helix and was modified slightly for synthetic purposes. 2.2.2.1 Peptide Synthesis Peptides 6, 7, 8, and 9 (see Table 2.1 for peptide sequences) were synthesized using standard Fmoc techniques on an automated Applied Biosystems peptide synthesizer following literature procedures.12 For a schematic representation of the peptide synthesis see Section 2.4.2.2. Side chain protected amino acids were used for chemoselective synthesis of the peptide, which was in turn bound by its C-terminus to a resin developed by Rink to afford a C-terminal amide upon cleavage.13 After the peptides were removed from the automated peptide synthesizer, manual chloroacetylation of the free N-termini of the peptides was achieved by treatment with chloroacetyl chloride, resulting in an activated form of the peptide. Then the individual peptides were manually treated with 2,2,2-trifluoroacetic acid (TFA), which cleaved the peptide from the resin in addition to removing the side chain protecting groups simultaneously. This reaction afforded the desired peptides 6, 7, 8, and 9 ready for reaction with cavitand 5 to give caviteins 10,11,12, and 13, respectively. 72, Table 2.1. Complete Sequences Including Modified Termini for Peptides 6,7,8,9. Peptide Peptide Peptide Sequence Number Name 6 ig2 C1CH 2 C0-NH-[GG-EELLKKLEELLKKG]-C0-NH 2 7 lg3 ClCH 2 CO-NH- [GGG-EELLK K L E E L L K K G ] -CO-N H 2 8a ng2 ClCH 2 CO-NH-[GG-EENNKKNEENNKKG]-CO-NH 2 9a ng3 C1CH 2 C0-NH-[GGG-EENNKKNEENNKKG]-C0-NH 2 a note "N" in ng2 and ng3 refers to norleucine and not asparagine. The nomenclature used in this thesis includes using small letters for naming peptide sequences and the same name is then capitalized to refer to the corresponding four-helix bundle caviteins. For this thesis the use of the letters "n" and "N" refer to norleucine and not asparagine. In this way, the first letter in either the peptide or cavitein name refers to the hydrophobic residue in the helical sequence, following by the number of glycine linkers between the peptide helices and cavitand template. Consider LG3, the capitals signify that it is a cavitein, L = leucine (the hydrophobic residue in the peptide sequence), and G3 = three glycine linkers between the peptide helices and the cavitand. It should be noted that in Chapter 3 and Chapter 4 some peptide sequences were attached to the cavitand template via their C-termini indicated by the letter "c" or "C" at the end of the peptide or cavitein names, respectively. For example cavitein LG2C, L = leucine, G2 = two glycine linkers, and C = peptides linked to the cavitand via their C-termini. Furthermore, in Chapter 4 some peptides and caviteins were designed without helix capping residues and those names end with "nocap". For example cavitein LG2nocap, L = leucine, G2 = two glycine linkers, and nocap = no C-terminal cap. 73, 2.2.3 Template Assembled Synthetic Protein (TASP) or Cavitein Synthesis The synthesis of cavitein 10 entails the coupling of activated peptide 6 and cavitand template 5 (see Scheme 2.2 for the cavitein reaction). Similar couplings were undertaken to produce caviteins 11,12,13 with peptides 7,8,9, respectively (see Table 2.2 for a list of cavitein products and their corresponding names). Scheme 2.2. Diagram Illustrating the Synthesis of Cavitein 10. The Linkage Between the Cavitand and the Peptides is "cavitand-(S-CH2CO-NH-peptide 6)4". Cavitein 10 was synthesized by following literature procedures a' by reacting 4.4 equivalents of peptide 6 with one equivalent of arylthiol cavitand 5 in the presence of diisopropylethylamine (DIPEA) in dimethylformamide (DMF) solvent at 25 °C. The reaction was monitored by analytical reversed-phase HPLC and was complete after 4 hours with a yield of 68 % with the tris-cavitein being the major byproduct. Caviteins 10, 11, 12, and 13 were purified by preparatory reversed phase HPLC and characterized by matrix assisted laser Cavitand 5 Peptide 6 Cavitein 10 74 desorption ionization (MALDI) mass spectrometry, circular dichroism (CD) and nuclear magnetic resonance (NMR) spectroscopy, analytical ultracentrifugation (AUC), and the binding of a hydrophobic dye monitored by fluorescence spectroscopy (see Section 2.2.4 for the characterization of LG2, LG3, NG2, and NG3). Table 2.2. Names and Sequences for Caviteins 10,11,12 and 13. Cavitein Cavitein Sequence Number Name 10 LG2 Cavitand 5-(S-CH 2 CO-NH-[GG-EELLKKLEELLKKG]-CO-NH2)4 11 LG3 Cavitand 5 - (S-CH 2 CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 ) 4 12a NG2 Cavitand 5-(S-CH 2 CO-NH-[GG-EENNKKNEENNKKG]-CO-NH 2 )4 13a NG3 Cavitand 5-(S-CH 2 CO-NH-[GGG-EENNKKNEENNKKG]-CO-NH 2 )4 a note "N" in NG2 and NG3 refers to norleucine and not asparagine. 2.2.4 Characterization of the Caviteins The caviteins were synthesized in good yield and with high levels of purity. Physical studies, which will be discussed below, were conducted to assess and compare the relative stabilities and tertiary structures of LG2, LG3, NG2, and NG3. 75 2.2.4.1 Far-UV Circular Dichroism (CD) Spectroscopy Far-UV (190-250 nm) CD spectroscopy is regularly used to quantify the extent of secondary structure present in peptides and proteins.14 Different types of protein secondary structures (helices, sheets, turns, and coils) give rise to characteristic CD spectra. Therefore, the approximate fraction of the kind of secondary structure that is present in any protein can be determined by analyzing its far-UV CD spectrum as a sum of fractional multiples of the reference spectra depicted in Figure 2.3.1 5 Figure 2.3. The Reference CD Spectra for Poly-L-Lysine in an a-Helix, P-Sheet, and Random Coil Conformation. 180 190 200 210 220 230 240 250 wavelength (nm) 76 As can be seen in the CD spectrum in Figure 2.3 an a-helix displays two distinctive negative bands near 222 nm and 208 nm and one positive band near 195 nm. P-sheets, on the other hand, have a single minimum at 218 nm and a maximum at 200 nm, and random coil conformations exhibit a solitary minimum at 200 nm. The CD spectra for LG2, LG3, NG2 and NG3 at concentrations of -40 |J,M are shown in Figure 2.4. CD spectra for each of the caviteins were also obtained at -4 J J M to evaluate whether concentration has an effect on the a-helicity of the caviteins. However, the effects of T F E on the a-helicity of the caviteins was not investigated in this thesis, since it has previously been shown that T F E results only in a marginal increase in the a-helicity of our caviteins.1 For all of the caviteins, the CD spectra at concentrations of -4 and -40 )0,M agree within experimental error, respectively, and therefore the -4 | i M curves for each of the caviteins are not shown in Figure 2.4. Since there was no observed increase in helicity with an increase in concentration this supports that the caviteins were monomeric. However, if the caviteins were already highly helical at low concentrations, an increase in helicity would not necessarily be observed at higher concentrations. Thus, concentration experiments by monitoring the molar ellipticity at 222 nm in the presence of a denaturant were performed to more thoroughly study the oligomeric states of the caviteins (Section 2.2.4.4.3). Analytical ultracentrifugation (AUC) experiments were also performed. A U C is one of the most common methods used to assess the oligomeric states of proteins in solution (Section 2.2.4.4). 77 Figure 2.4. Far-UV CD Spectra for Caviteins LG2, LG3, NG2, and NG3 at -40 \iM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. -30000 J Wavelength (nm) The molar ellipticity at 222 nm ([0\222), arising from the n—>n amide transition, is used as a gauge of a-helical structure. The [c7J2 2 values for the caviteins and an estimate of the maximum value of [t7J2 2 using the chain-length-dependent equation developed by Chen et al. are shown in Table 2.3 below.16 However, for the case of our caviteins [t7j2 2 should not be over interpreted since the aromatic contributions by the cavitand template near 220-260 nm may influence the absorbance at 222 nm. It should be mentioned that Mezo has previously studied the effect of the template on the a-helicity of the resultant cavitein, and it was deduced that the cavitand template does increase the a-helicity of the corresponding cavitein as compared to a single-stranded peptide.12 78 The ratio of [O^n'Whos has often been used to estimate the extent of protein tertiary structure. Interacting coiled coil-like structures are characterized by [0]222/[#]2O8 values greater 17 than 1, whereas ratios less than Tare indicative of non-interacting helices. LG2 and NG2 display [0]222/[#]2O8 > 1, while LG3 and NG3 display [f?]222/[#]208 < 1- Although it is known that the cavitand template does influence the absorbance of the caviteins at 222 nm, do the values of [0]222/[#]2O8 < 1 for LG3 and NG3 advocate less tertiary structure compared to LG2 and NG2? Baldwin and coworkers have shown that aromatic residues can contribute both positively and negatively to the [f?]222 values of single-stranded a-helices, and they found that inserting 18 from one to three glycine residues diminished the effect of the aromatic residues on [dhu-Therefore, one should not conclude that LG3 and NG3 exhibit less tertiary structure than LG2 or NG2 from the far-UV GD spectra alone, because it is likely that changing the number of glycine residues in the linker between the cavitand template and the peptides would result in a non-uniform effect on [0]222. Table 2.3. Molar Ellipticity at 222 nm ([^222) and Percent Helicity for Caviteins LG2, LG3, NG2, and NG3. Cavitein Concentration (HM) Experimental [^ 222 (deg cm 2 dmol"1) Calculated Maximum [0)222 (deg cm 2 dmol"1) Percent Helicity (%) LG2 39 -20000 -33200 -60 LG3 39 -18000 -33500 -54 NG2 40 -19000 -33200 ~57 NG3 38 -16000 -33500 -48 79 Although the experimental molar ellipticities at 222 nm should riot be quantitatively analyzed given that the exact effect of the template on the individual cavitein oc-helicities is unknown, the far-UV CD data does suggest that a-helicity does decrease when the leucine residues in LG2 and LG3 were replaced with norleucine units in NG2 and NG3, respectively. This is expected since the norleucine amino acid has an unbranched side chain and should have an intrinsically lower helical propensity value when compared to leucine. 2.2.4.2 Near-UV C D Spectroscopy In addition to determining the extent of a-helical secondary structure by interpreting the far-UV region of a CD spectrum, information about protein tertiary structure can be gathered by examining the near-UV (250-350 nm) spectral region.14b In this range the aromatic amino acids and disulfide bonds are the chromophores, and the CD signals they produce are sensitive to the overall tertiary structure of the protein. LG2, LG3, NG2, and NG3 do not have aromatic residues or disulfide bonds but the arenes of the cavitand template do absorb in this region. Furthermore, proteins lacking well-defined three-dimensional structures (e.g. a molten globule or misfolded protein) produce nearly no signal in the near-UV spectral region due to the time-averaged fluctuating structures.19 On the other hand, enhanced near-UV signals are indicative of a well-defined protein structure due to the asymmetric environments of their aromatic chromophores.20 The near-UV CD spectra for LG2, LG3, NG2 and NG3 are graphed in Figure 2.5. 80 Figure 2.5. Near-UV CD Spectra for Caviteins LG2, LG3, NG2, and NG3 at -40 (iM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. -2100 -J Wavelength (nm) It is observed that the signal strength in the near-UV CD region is much weaker than that in the far-UV CD region. Furthermore, the replacement of the leucine residues with norleucine units had little effect on the near-UV CD signal. This was to some extent surprising since a hydrophobic core consisting of norleucine amino acids should give rise to a molten globule-like tertiary structure and hence a reduced near-UV signal due to time-averaged fluctuating structures. Therefore, it appears as though the length of the linker (i.e. 2 Gly versus 3 Gly) in the caviteins has a stronger effect on the tertiary structures of the proteins than does the substitution of leucine residues with norleucine residues, respectively. 8 1 The signs of the signals in the near-UV region can be examined to gain some information about the supercoiling of the helices near the cavitand chromophore. The 2 Gly-linked caviteins both show negative absorptions around 250 nm and positive absorptions around 270 nm suggesting that they are supercoiling in the same direction. The 3 Gly-linked caviteins exhibit a reversal in the signs of the absorptions at -250 and -270 nm, respectively. The reversal of the positive and negative absorptions would support that the 3 Gly-linked caviteins are supercoiling in the opposite direction compared to the 2 Gly-linked caviteins. 2.3.3 Oligomeric States Evaluated by GuHCl Denaturation Experiments The oligomeric states of the caviteins were studied by monitoring the concentration dependence of their unfolding in the presence of guanidine hydrochloride (GuHCl), and by analytical ultracentrifugation (AUC) experiments explained in Section 2.2.4.4.3. As was mentioned in Section 2.2.4.1 LG2, LG3, NG2, and NG3 produced concentration independent CD spectra, however this is not sufficient enough data to conclude the existence of monomers in solution. Therefore, the stabilities of the caviteins in the presence of the denaturing salt, GuHCl, were determined at different concentrations. GuHCl is a salt and is thought to preferentially bind to protein surfaces and disrupt hydrogen bonding, hydrophobic packing, and electrostatic interactions, although the exact mechanism is still not known. Other chemical denaturants such as urea could also be used, however, urea only affects hydrogen bonding and hydrophobic interactions, and therefore GuHCl was selected as the chemical denaturant. The stabilities of oligomeric proteins are expected to vary with concentration, whereas monomeric proteins should give rise to identical denaturation curves. The assumption is that self-association would be accompanied with an increase in structural stability. Figure 2.6 '. 82 displays the unfolding curves of caviteins LG2, LG3, NG2, and NG3 monitored at 222 nm in the presence of 0-8.0 M GuHCl at concentrations of -40 (iM. An additional denaturation experiment was also carried out for each of the caviteins at a concentration of -4 p:M. The data for each of the caviteins at both concentrations overlapped within experimental error, respectively, as can be seen with their similar A G ° H 2 O values listed in Table 2.4. In this way, the unfolding curves obtained for the -4 J I M concentrations are not included on Figure 2.6 for clarity's sake. Furthermore, the stability of a single-stranded peptide has previously been measured and was found to be completely unstable toward GuHCl and unfolded non-cooperatively.1 Figure 2.6. Effect of GuHCl on the Helicity {[0\222) of Caviteins LG2, LG3, NG2, and NG3 at -40 uM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. [GuHCl] (M) 83 The cooperativity of the unfolding reaction can be measured qualitatively by the width and shape of the unfolding transition. A highly cooperative unfolding curve indicates that the protein existed originally as a compact, well-folded structure. On the other hand, a very gradual, non-cooperative unfolding transition indicates that the protein existed initially as a flexible, only partially folded (e.g. molten globule) protein. Examining the unfolding curves for LG2, LG3, NG2 and NG3 one can notice that the caviteins were completely unfolded by 8.0 M GuHCl, the unfolding transitions were cooperative, and that the norleucine based caviteins were less stable toward the chemical denaturant. Furthermore, the unfolding curves for NG2 and NG3 seem to be more gradual, starting to denature at -3.0 M GuHCl, as compared to the curves for their leucine based counterparts which begin to unfold at-4.5 M GuHCl. When comparing the stabilities of our caviteins towards GuHCl one can take the concentration of GuHCl required to unfold the cavitein halfway ([GuHCl] m) as a rough estimate of their stabilities. A more accurate method for determining the stabilities of proteins is via a "linear extrapolation method", in which the free energy of unfolding is assumed to vary linearly with GuHCl concentration.22 To determine the free energy of unfolding of the caviteins, we used the expanded method described by Santoro and Bolen which assumes the unfolding transition to be a reversible, two-state process.23 This method uses a non-linear least squares fitting of data to derive at a A G ° H 2 O value. The extrapolated values for A G ° H 2 O , m , and the denaturation midpoint for each of the caviteins are listed in Table 2.4. 84 Table 2.4. Guanidine Hydrochloride-Induced Denaturation Data Calculated for Caviteins LG2, LG3, NG2, and NG3. Cavitein Cavitein Concentration (|J.M) [GuHCl] 1 / 2 a (M) m (kcal/molM) A G ° H 2 O (kcal/mol) LG2 39 5.7 ± 0 . 1 -1.8 ± 0 . 1 -10.4 ± 0 . 3 LG2 4 5.8 ± 0 . 1 -1.8 ± 0 . 1 -10.2 ± 0 . 3 LG3 39 5.6 ± 0 . 1 -1.9 ± 0 . 1 -10.8 ± 0 . 4 LG3 4 5.7 + 0.1 -1.8 ± 0 . 1 -10.7 ± 0 . 4 NG2 40 4.7 ± 0 . 1 -1.4 ± 0 . 1 -7.0 ± 0.3 NG2 4 4.6 ± 0 . 1 -1.5 ± 0 . 1 -6.8 ± 0 . 3 NG3 39 4.6 ± 0 . 1 -1.6 ± 0 . 1 -7.5 ± 0 . 3 NG3 4 4.6 ± 0 . 1 -1.5 ± 0 . 1 -7.4 ± 0 . 3 [GuHCl] 1/2 is the concentration of GuHCl required to unfold half of the helical structure. The A G ° H 2 O values of unfolding for caviteins L G 2 , L G 3 , N G 2 , and N G 3 at concentrations of ~4 JJLM and -40 uM fall within experimental error of each other, respectively. L G 3 and L G 2 have comparable A G ° H 2 O values, and are the most stable of the four caviteins. It was hypothesized that L G 2 would be more stable than L G 3 since L G 2 has one less glycine linker between the template and peptide helices than does L G 3 , and therefore the stability of the bundle in L G 2 should be more influenced by the stabilizing effects of the cavitand template. In addition, the norleucine-based caviteins were less stable than their corresponding leucine-based caviteins, possibly due to the entropic cost of restricting the number of conformations of the norleucine side chains within the hydrophobic core. Furthermore, the stabilities of the caviteins were found to be more stable when compared to other template assembled synthetic proteins.24 85 Larger m values correspond to a more cooperative unfolding transition, and can therefore be used as an estimate of native-like structure.25 From Table 2.4, it is clear.that the m values of LG2 and LG3 are similar and slightly higher than those observed for NG2 and NG3. This is consistent with the above-mentioned data suggesting that LG2 and LG3 have a similar tertiary structure to each other as do NG2 and NG3, respectively. Furthermore, since the m values for LG2 and LG3 are slightly higher than for NG2 and NG3 this would support that the leucine based caviteins are slightly more native-like. Furthermore, the m values for the four caviteins were comparable to the m values found in native proteins.26 In addition to assessing the stability of the caviteins towards the chemical denaturant GuHCl, the stability of the caviteins was monitored as a function of temperature at 222 nm. All of the caviteins were highly stable towards temperature, with a linear decrease in a-helicity of approximately 10 % from 5 to 90 °C. The experiments were also repeated in 3.0 M GuHCl and the curves were superimposible in all cases. Since cooperative unfolding transitions were not observed the data is not shown, and further analyses were not performed. As was observed with the concentration independent CD spectra, the concentration independent unfolding curves of the caviteins support a monomeric species in solution. The oligomeric states of the caviteins were more accurately assessed using A U C . 2.2.4.4 Analytical Ultracentrifugation (AUC) Analytical ultracentrifugation is a powerful tool for the study of proteins. It allows for the determination of both hydrodynamic and thermodynamic behavior of macromolecules in solution, without the interaction with any matrix or surface.27 A U C relies on the principles of 86 ' ' 98 mass and the fundamental laws of gravitation. Sedimentation is a nondestructive technique that can be used to analyze the solution behavior of many molecules at varying solute concentrations and in a variety of buffers. The combination of new instrumentation and powerful software for data analysis has led to major advances in the characterization of proteins and protein complexes by analytical ultracentrifugation (see Section 2.4.5 for complete details on sample preparation, instrumentation, equations, and data analysis). Analytical ultracentrifugation had previously been narrowly studied in our lab (only sedimentation equilibrium), and thus an introduction to the theory of A U C will follow. Two parallel Views, sedimentation equilibrium and sedimentation velocity, provide information on molecule behavior, both of which were explored in detail. The next Section will focus on sedimentation equilibrium, describe its use for the determination of protein molecular weight in solution and present the experimentally determined molecular weights of LG2, LG3, NG2 and NG3 in solution. 2.2.4.4.1 Theory Behind Sedimentation Equilibrium Sedimentation equilibrium can provide thermodynamic information of molecules including their molar mass, association constants, and stoichiometrics. The basis behind this technique is that at equilibrium the flux of the sedimenting molecules is exactly balanced by the flux of the diffusing molecules at each position within the cell. Therefore, at equilibrium a concentration gradient is established, and no concentration boundaries can be observed within the cell. 87 The equations used to describe sedimentation equilibrium have been derived from a number of approaches, and the most common approach is based on the balance of fluxes, which will briefly be described.2915 In the context of sedimentation equilibrium the term flux refers to the rate of flow of a material through a given surface. As was mentioned previously at equilibrium a concentration gradient is established where the flux of sedimentation (concentration of material x velocity) is equal to the flux due to diffusion (derived by applying Tick's First Law and is equal to the diffusion coefficient x the rate of change of concentration with respect to the radius).29b From the balance of these two fluxes a reduced molecular weight, a, can be derived and is outlined in equation l ; 3 0 o-=Mhar/RT (1) Where Mb refers to the buoyant mass, (O is the angular velocity determined from the rotor speed in radians per second, R is the gas constant, and T is the temperature in Kelvin. Integration of this function leads to the Lamm equation31 describing the movement of molecules in a centrifugal field, from which equation 2 3 2 can be derived for a single, thermodynamically ideal solute: Ln(cr)/r2=M( 1 - v p)(£>2/2RT (2) Where cr is the concentration of the solute at radial distance r, r is the radial distance from the center of rotation, M is the monomer molecular weight, v is the partial specific volume of the solute, and p is the solvent density. 88 Using a data analysis program such as NONLXN, equation 2 is solved by a nonlinear least-squares fitting of the raw data, and the molecular weight of the macromolecule in solution can be determined. However, interpretation of this data requires information about the rotor speed, the temperature, the partial specific volume and the density of the solvent, which need to be determined experimentally or calculated, and inputted into the NONLJN program prior to analysis. 2.2.4.4.2 Determination of the Partial Specific Volume The partial specific volume, v, is defined as the change in volume (in mL) of the solution per gram of added solute, and is temperature dependent. It is essentially the partial volume of the protein corrected for hydration, solute binding, and electrostriction effects.34 It can be measured from the density increment,35 or from density perturbation sedimentation equilibrium,36 or can be calculated from the composition.37 Most commonly, the partial specific volume is estimated from the amino acid composition using equation 3 derived by Cohn and Edsall. 3 7 v c = IWi V j / ZWi = ZNiMt v i/NtMi (3) Where v c is the calculated partial specific volume, Wi is the weight percent of the ith component, TV, is the number of residues, Mi is the molecular weight of the corresponding component, and v, is the partial specific volume of the component. One observation is that the specific volumes of all proteins must fall within the upper and lower limits for the amino acid 89 residues. However, significant errors in this method may come about from proteins that exhibit preferential hydration, proteins that have non-globular structure, and proteins with substantial non-amino acid components in their structure. It is essential to have an accurate value for v since small errors in v manifest themselves into large errors in the molecular weight.278 With our caviteins, the cavitand template cannot be accounted for in the simple calculation method for the determination of v, and thus v was determined experimentally using density perturbation sedimentation equilibrium studies developed by Edelstein and Schachman.36 The method of density perturbation equilibrium permits the determination of v with microgram quantities, and involves parallel sedimentation experiments in solutions of H2O and D 2 0 . Although it requires very little sample, it does require very accurate sedimentation equilibrium data, and multiple determinations are necessary in order to reduce the error associated with v . Sedimentation experiments in the two solvents can be related by equation 4, 3 6 and a plot of ln c vs r 2 should result in a straight line where the slope is proportional to v . V = k-[(d \nc/dr2)D20/(d \nc/dr2)H20] : ' (4) pD20 - pH20[d ln cldr2)D20 l{d ln c I dr2)H20] Where k is the ratio of the molecular weight of the protein in the deuterated to that in the non-deuterated solvent, and p is the solvent density. Figure 2.7 below shows the plots for of ln c vs. r for LG3. 90 Figure 2.7. Differential Sedimentation Plot of LG3 in H 2 0 and D 2 0 , respectively, at a Concentration of 10 [iM at 20 °C, and Rotor Speed of 40000 rpm. An analysis of the partial specific volume data resulted in straight-line plots for all of the caviteins where v was calculated from the slopes. Table 2.5 outlines the partial specific volume data for the caviteins. 91 Table 2.5. Experimentally Determined Partial Specific Volumes by Sedimentation Equilibrium for All Cavitein Variants at 20 °C at Concentrations of 10 (iM and a Rotor Speed of 40000 rpm. Cavitein Experimental Partial Specific Volume (mL/g)a Calculated Partial Specific Volume (mL/g) LG2 0.78 ± 0 . 0 1 0.7814 LG3 0.75 ± 0 . 0 1 0.7770 ; / NG2 0.78 ± 0 . 0 1 0.7814 NG3 . 0.79 ± 0.01 0.7770 A G 3 b 0.75 ± 0 . 0 1 0.7310 a average value from three separate trials. b AG3 cavitein introduced in Chapter 3. It is clear from Table 2.5 that the experimental and calculated partial specific volumes are very close to each other for a variety of caviteins, respectively. Since an extreme amount of work is required to experimentally determine the partial specific volumes of our caviteins, using the partial specific volumes calculated from the amino acid composition alone would suffice for the calculation of the molecular weight of our caviteins using A U C . 2.2.4.4.3 Oligomeric States Evaluated by Sedimentation Equilibrium AUC The sedimentation equilibrium data for LG2, LG3, NG2 and NG3 were analyzed by NONLIN and the exponential plot of absorbance versus radius for LG3 at 10 p M and 27000 rpm 92 is shown in Figure 2.8. The sedimentation experiments for each of the caviteins included three different cavitein concentrations (10, 50, and 80 p:M) and three different rotor speeds (27000, 35000, 40000 rpm), for a total of nine exponential plots of absorbance versus radius per cavitien, and thus only three representative plots for each of the caviteins are included in this thesis (the remaining representative plots for LG3 and the other caviteins are included in Section 2.4.5). 93 Figure 2.8. Sedimentation Equilibrium Concentration Distributions of LG3 at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 uM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. 5.000 o i < o o I D p 2.500 L -2.500 L -5.000 17.500 10.000 10.500 Radius Squared/2 (cmA2) 22.500 o I o CU O a o O 12.500 L 7.500 17.500 10.000 10.500 Radius Squared/2 (cmA2) 94 The data for each of the caviteins were also fit to monomer-dimer, monomer-trimer, and monomer-tetramer equilibria in order to check for the best theoretical fit to the experimental data. In all cases the monomer fits were most accurate, assessed by the even distribution of the residuals about zero as was seen, for example, in the upper panel of Figure 2.8. For comparison, a fit to a monomer-trimer equilibrium for LG3 is included in Figure 2.9 to show an uneven distribution of the residuals, and hence poor fit. 95 Figure 2.9. Sedimentation Equilibrium Concentration Distributions of LG3 at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 50 | iM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer-Trimer Equilibrium. The Upper Panel Represents the Residuals for the Fit. CM o • < I o fi 1.250 L 0.000 L -1.250 L 17.250 17.750 10.250 Radius Squared/2 (cmA2) 10.750 0.500 o o O o 0.375 L 0.250 L 0.125 L 0.000 17.250 17.750 18.250 Radius Squared/2 (cmA2). 96 18.750 The results from the sedimentation equilibrium data for all of the caviteins are consistent with the CD and GuHCl denaturation data for the presence of a monomeric species. Table 2.6 summarizes the sedimentation data and for caviteins LG2, LG3, NG2 and NG3. Table 2.6. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For All Cavitein Variants at 20 °C in 50 mM pH 7.0 Sodium Phosphate Buffer at Concentrations of 10, 50 and 80 uM with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section 2.4.5) Cavitein Experimentally Calculated M W Predominant Species Estimated M W (Da) (Da) _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ 8 Q 1 6 ~ Monomer LG3 80001300 8240 Monomer NG2 8600 ± 5 0 0 8016 Monomer NG3 85001 400 8240 Monomer 2.2.4.4.4 Sedimentation Velocity AUC Sedimentation velocity provides hydrodynamic information about the size and shape of a molecule.27 It differs from sedimentation equilibrium in that a high rotor speed and a long solution column are recommended to maximize the resolution of the species. Furthermore, sedimentation velocity is regularly used to determine how many different species are present in solution, but if an accurate estimate of the molecular weight or association constant is desired 97 sedimentation equilibrium experiments should be performed. For the purpose of our cavitiens the sedimentation data was only used to determine whether a single species was present in solution, and thus only a brief introduction to the theory of sedimentation velocity is provided below. In velocity experiments the sedimentation coefficient, s, is determined from the rate of boundary movement, where as the rate of boundary spreading is used to determine the diffusion coefficient, D . 2 8 ' 3 8 These two quantities describe the hydrodynamic behavior and their ratio, s/D, can be used to determine its buoyant mass, Mb. As was explained previously in Section 2.2.4.4.1 the molecular weight can be calculated from Mb knowing the density of the solvent and the partial specific volume of the solute. Detailed analyses of s and D can be acquired by curve-fitting methods based on solutions to the Lamm equation:39 The velocity data are commonly analyzed using a software program called SEDFIT. 4 0 This program determines the sedimentation coefficient, and clearly illustrates whether one or more species is present in solution. Figure 2.10 shows the raw data obtained from SEDFIT for a single concentration and rotor speed for LG3. The plot of the residuals at the top of the Figure 2.10 represent a good fit of the data represented by a consistent gray display. Figure 2.11 contains the plots of the C(s) distribution versus the sedimentation coefficient, s, for LG3 at varying concentrations. 98 Figure 2.10. Sedimentation Velocity Raw Data from SEDFTT for LG3 at a Concentration of 1.0 mg/mL, and a Rotor Speed of 40000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C. The Upper Panel Represents the Residuals for the Fit. The Lower Picture Shows the Fitting Obtained from the SEDFLT Program. 99 Figure 2.11. Sedimentation Velocity Concentration Distributions of L G 3 at a Rotor Speed of 40000 rpm in 50 m M Sodium Phosphate Buffer, p H 7.0, 20 °C. 12 Sedimentation Coefficient (s) The velocity data for all of the caviteins supports that only one species is present in solution through the inspection of only a single peak at each concentration in Figure 2.11. The studies were done at different concentrations to monitor the concentration dependence of the caviteins in solution and only one predominant species was detected. The velocity data is again consistent with the previous findings obtained from the C D , G u H C l denaturation experiments and sedimentation equilibrium data that all of the caviteins exist as monomers in solution. 100 2.2.4.5 ^  Nuclear Magnetic Resonance (NMR) Spectroscopy Although, CD spectroscopy is commonly used to gather information on the secondary structure and aggregation state of proteins, NMR spectroscopy is a useful tool for studying the global fold (i.e. tertiary structure) of proteins.41 One-dimensional ! H NMR spectroscopy is a simple diagnostic method used to differentiate between native-like and molten globule-like structures. However, one of the limitations of standard ! H NMR spectroscopy is the inability to assign every peak to nuclei of specific amino acid residues, and therefore, ID ! H NMR spectroscopy is restrictive in the area of structure determination. Therefore, two-dimensional NMR experiments were performed in an attempt to assign the amide proton signals of the *H NMR spectra to specific amino acid residues (Section 2.2.4.9). 2.2.4.6 One-Dimensional (ID) 'H NMR Spectroscopy The 'H NMR spectra of proteins can be very complex and the spectra substantially increase in complexity as the size of the macromolecule increases. It has been demonstrated by McDonald and Phillips that the amino acid side chain XH NMR signals in a denatured or random coil polypeptide chain correspond closely to the sum of the resonances of the constituent amino acid residues.42 This observation can be rationalized by the assumption that all amino acid side-chains in a random coil polypeptide chain are exposed to the same solvent environment, so that equivalent amino acids have identical chemical shifts. 101 Chemical shift dispersion arises because the interior peptide regions in native proteins are shielded from the solvent, and so are neighbors of other peptide segments, so that different residues experience different microenvironments. In this way, natural proteins exhibit considerable chemical shift dispersion, which is indicative of native-like structure.43 In contrast, molten globule-like structures portray more broad, less disperse signals due to the looser packing arrangement of the hydrophobic core.44 Usually, the spectral region between 7 and 11 ppm is examined because distinct sharp and dispersed signals are observed for native-like proteins corresponding to the slowly exchanging amide protons. The full lH NMR spectra for caviteins LG2, LG3, NG2 and NG3 are shown in Figures 2.12, 2.13, 2.14 and 2.15, respectively. Figure 2.16 has an overlay of the amide regions for the four caviteins to more clearly illustrate the spectral region of interest, and in following Chapters only the amide region will be displayed. Figure 2.12. Full 500 MHz *H NMR Spectrum of LG2 at -1.5 mM in 10 % P 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (* = cavitand signals) 9 8 7 6 5 4 3 2 1 0 ppm 102 Figure 2.13. Full 500 MHz ! H NMR Spectrum of LG3 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (* = cavitand signals) Figure 2.14. Full 500 MHz 'H NMR Spectrum of NG2 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (* = cavitand signals) I 1 1 1 1 I 1 1 1 1 I . 1 1 1 1 I 1 1 1 1 I ! 1 1 1 I ' 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 '• 1 ! I 1 1 l ; l | i i i i i i | i i i i | 1 0 9 8 7 6 5 4 3 2 1 0 -1 - 2 - 3 - 4 ppm Figure 2.15. Full 500 MHz l H N M R Spectrum of NG3 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (* = cavitand signals) T — i — i — i — | — i — i — i — i — | ' i i—i—i—j. i • i — i — i — | — i — i — i — i — | — | — i — i — i — | — i — r ~ i — I — | — I — i — i — i | r i i i | i I i r — p 9 8 7 6 5 4 3 2 1 0 ppm 104 Figure 2.16. Expanded Amide Regions of the 500 MHz *H NMR Spectra of (a) LG2, (b) LG3 (c) NG2, and (d) NG3 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (* = cavitand signals) * I I i I | I ! I I | I I I ! j I I I I | I I I I j I I I I | I I ! I | I ! I I | I I I I | I I I I j I I I I | ! I I ! j I I I I j I I I I j I ! I I | I I I I | I I I I | I I I I j I I I ! j I I I I | I I I I | I I I ! | I I 9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 ppm The spectrum of LG2 shows -12 distinguishable dispersed amide signals indicative of a well-defined amide backbone with a high content of secondary structure. The presence of only 12 amide signals suggests that many of the amino acid residues are in a degenerate environment and therefore indistinguishable from each other, likely due to the four-fold symmetry of the cavitein. Although, the primary sequence is degenerate with only four different types of amino acid residues 12 signals can still be resolved. Upon examination of the amide region of the ! H NMR spectrum of NG2 it is as dispersed as LG2 but more broad. Many of the amide signals of NG2 are less defined, but the difference compared with LG2 is not compelling, except for the H o u t cavitand peak of NG2 at -6.1 ppm 105 which is significantly broader. A possible explanation for the broad signal is that the bound norleucine side chain (discussed in the following paragraph) is breaking the symmetry of the cavitein, and thus results in a few overlapping non-equivalent H o u t signals. Besides the effect on the cavitand signal, it appears that the substitution of leucine residues in LG2 with norleucine units did not greatly affect the tertiary structure of the cavitein, suggesting that perhaps LG2 was not as native-like as previously thought, or that perhaps NG2 is as native-like as LG2. A distinctive observation with the ! H NMR spectrum of NG2 is the presence of signals upfield at -0.5 and -3.0 ppm. Signals in this region were not observed for any of the other three caviteins, and such signals had previously never been detected for other caviteins in our group. Signals in this region generally correspond to a guest bound within the confines of a cavitand template, and the more the signals are upfield the more deeply the guest is resides inside the template. The shielding effects of the arenes, creating the framework of the cavitand, are the cause for the bound guest signals to shift upfield (i.e. lower ppm values). The assignments of these upfield signals in NG2 were determined from 2D total correlation spectroscopy (TOCSY) and nuclear Overhauser enhancement spectroscopy (NOESY) experiments, and are attributed to the side chains of the first norleucine residue in the peptide sequence. The signal at -0.5 ppm was determined to be from the hydrogens of CYH2 and the signal at -3.0 ppm to be from the hydrogens of C e H 3 . The TOCSY and NOESY spectra of NG2 are shown in the experimental, in addition to a brief explanation of how they were analyzed. Section 2.3.5.4 includes an introduction to 2D NMR spectroscopy focusing only on double-quantum-filtered (DQF)-correlation spectroscopy (COSY) and NOESY experiments. The amide region for LG3 shows -14 relatively sharp and disperse amide signals, although not as dispersed as LG2. Coiled coil structures generally exhibit less dispersion in the amide region than do square bundles, and thus the lower dispersion does not preclude native-like 106 structure.45 Comparing LG3 to NG3, the amide region of LG3 is sharper than NG3 which only has -10 distinguishable amide signals, in addition to a slightly more broad cavitand H o u t signal at -6.15 ppm. An interpretation is that the side chains in LG3 are packed well, characteristic of native-like structure, while the hydrophobic core of NG3 may have slight more motion leading to average signals. Thus, the norleucine residues, with their unbranched side chains, did affect the tertiary structure of NG3 with a slight loss of native-like structure. However, since the differences in the 'H NMR spectra of LG3 and NG3 are not significant, another interpretation is that NG3 is packing almost as well as LG3, and the entropic cOst of restricting the norleucine side chain to a similar number of conformations as the leucine side chain, is reflected in the lower A G ° H 2 O value of NG3. To further probe the presence of a well-defined tertiary structure, the aliphatic region of a J H NMR spectrum can be examined. Specific side chain packing interactions generally give rise to sharp and disperse signals. Comparing the aliphatic regions of Figures 2.12, 2.13, 2.14 and 2.15, it is evident that LG2 exhibits slightly more dispersed signals in this region than do the other three caviteins, although the difference is not compelling. Furthermore, the signals at -0.8 ppm in the *H NMR spectra of all four caviteins arising from the side chain packing leucine/norleucine core residues show relatively poor dispersion. It can be concluded that the overall dispersion of the signals in the ''H NMR spectra LG2 and LG3 suggest that both caviteins exhibit some native-like tertiary interactions, with the characteristics Of LG3 being even more so definitive. The substitutions of leucine residues with norleucine units in LG3 caused a slight change in the resultant *H NMR spectrum of NG3, supporting that LG3 has a well-defined tertiary structure. However, changing the number of glycine linkers between the peptide sequence and the cavitand template (2 Gly vs. 3 Gly) results in the most dramatic change in the *H NMR spectra of the caviteins, even more so than the 107 substitution of leucine residues with norleucine residues in the peptide sequence. The ! H NMR data further support previous experimental findings that the linker between the cavitand template and the peptide sequences plays a fundamental role in influencing the overall tertiary structure of the caviteins. Furthermore, the binding of the norleucine side chain within the confines of the cavitand template in NG2 proposes a possibility of binding other substrates in the future. The binding of substrates within the template of our caviteins, had never previously been observed. 2.2.4.7 Hydrogen/Deuter ium A m i d e Exchange Before we consider protons, which can undergo exchange with the solvent, we must understand that all proteins have some conformational flexibility or dynamics. Fully folded proteins or native-like proteins are dynamic and exhibit a wide variety of motions, for example, bond vibrations, side chain rotations, hinge bending and even local unfolding.41 The flexibility of a protein is important for many reasons, including folding, enzymatic activity or ligand binding. One common form of protein dynamics is seen in the motions of the side chains, such as rapid interconversion between different rotamers. Having stated that proteins are flexible, it is now interesting to consider the labile protons that can be studied by NMR, and are defined as the protons that can be readily exchanged with the solvent protons in aqueous media. The rate of exchange of the amide protons in native-like proteins has been found to be much slower when compared to such exchange in molten globule-like structures, and therefore, H/D exchange is a valuable diagnostic technique in the study of • 46 protein tertiary structures. 108 The exchange of labile protons is always acid or base catalyzed. In aqueous solution OH and H 2 0 are potential base catalysts, and H 3 0 + and H 2 0 are potential acid catalysts. The primary steps involve the formation of a hydrogen-bonded complex with the catalyst, transfer of the labile proton to the catalyst in this H-bonded complex, followed by dissociation of the complex. However, this process does not always result in a net transfer. The observed rate of exchange of an amide proton in a structured environment ( k o ^ ) can be divided by a calculated intrinsic rate of exchange for an amide in an unstructured environment at a given temperature and pH ( k i n ) to give a number called a "protection factor".466 The larger the protection factor, the slower the rate of exchange of the amide proton with the solvent and hence the proton is said to be "protected" from exchange. Typical protection factors for native proteins is in the range of 105—108, where as molten globule-like structures exhibit protection factors in the range of 101-JQ3 46d,e It is necessary to maintain constant pH and temperature during amide exchange experiments, and it is also known that the rate of amide exchange is also dependent on the peptide sequence.47 Some of the amide signals were assigned (see Section 2.2.4.9) to specific amino acids in the peptide sequence, but because many of the signals in the amide region were overlapping, not all assignments were possible. In this way, the effect of different side chains and neighboring groups were not taken into account. A context-independent equation was used to calculate the intrinsic rate of exchange for an "unprotected" amide proton.463 As a result, the calculated protection factors for LG3, LG2, NG3 and NG2 cannot be directly compared to literature values. However, Hecht and coworkers have characterized a de novo four-helix bundle using this same method for calculating protection factors, and thus can still be used as a method to study dynamics of de novo proteins 4 8 The 109 stack plots of ! H NMR amide regions for LG2, LG3, NG2 and NG3 are shown in Figures 2.17, 2.18, 2.19 and 2.20, respectively. Figure 2.17. Stack Plot of the 500 MHz *H NMR Spectra Illustrating the Time Dependent Amide H/D Exchange of LG2 at -1.5 mM in 50 mM pD 5.02 CD 3 COOD/CD 3 COO"Na + Buffer at 20 °C. (a) 5 min .(b)' 20 min (c) 1 h 4 min (d) 6 h 10 min (e) 25 h 17 min (f) 74 h 36 min (* = cavitand signals) * 9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 ppm 110 Figure 2.18. Stack Plot of the 500 MHz ! H NMR Spectra Illustrating the Time Dependent Amide H/D Exchange of LG3 at -1.5 mM in 50 mM pD 5.02 CD 3 COOD/CD 3 COO"Na + Buffer at 20 °C. (a) 6 min (b) 20 min (c) 1 h 5 min (d) 6 h 10 min (e) 23 h 19 min (f) 72 h 53 min (g) 9 d 1 h 21 min (* = cavitand signals) (a) (b) (c) (d) (e) A , - A . .(f) (g) • " T T T T j' 'I I ' l l | M ' H j I'l ''1 . p . . r r r j . . . r T T . r j"! 1 i | | | | | | | | 1 1 1 1 | | | 1 f I I' I' |' I I II j 1 1 I [ | I I I'l j' 1 ' I "I I "j " i - r T T p "TT T J I ' T I "I "j".! I 'I I | "i" 1 ' I" | " | - [ i - r T T * y T T T T ' y T T T T " j ' T T 9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 ppm 111 Figure 2.19. Stack Plot of the 500 MHz ' H NMR Spectra Illustrating the Time Dependent Amide H/D Exchange of NG2 at -1.5 mM in 50 mM pD 5.02 CD 3 COOD/CD 3 COO"Na + Buffer at 20 °C. (a) 4 min (b) 18 min (c) 1 h 15 min (d) 6 h 20 min (e) 22 h 26 min (* = cavitand signals) 112 Figure 2.20. Stack Plot of the 500 MHz ' H NMR Spectra Illustrating the Time Dependent Amide H/D Exchange of NG3 at -1.5 mM in 50 mM pD 5.02 CD 3 COOD/CD 3 COO~Na + Buffer at 20 °C. (a) 5 min (b) 28 min (c) 1 h 3 min (d) 6 h 8 min (e) 22 h 15 min (f) 4 d 21 h 10 min (* = cavitand signals) 9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 ppm The amide H/D exchanges of all of the caviteins were studied over the course of 1. to 10 days depending on the cavitein. Many of the amide protons exchanged before the first scan could be acquired (-5 min) and may be characteristic of molten globule-like protons. However, several amides were visible for much longer, and their calculated protection factors are outlined in Table 2.7 below. The remaining signals at -6.1 and -7.1 ppm correspond to H o u t and H p a r a of 113 the cavitand and do not change over time, and were used to normalize the peak heights of the amide protons. Table 2.7. Tabulated Data from the Amide H/D Exchange Experiments on LG2, LG3, NG2 and NG3 in a 50 mM pD 5.02 CD3COOD/CD 3COO"Na +Buffer at 20 °C. Cavitein Name Amide Proton Chemical Shift (ppm) Resultant Residue3 First-Order Rate Constant (V) Half-Life (h) Protection Factor6 LG2 7.98 LI 3 2.31 x.10"1. 3.0 (1.0 ± 0.3) x 103 LG2 7.83 E l l 1.54 xlO" 1 4.5 (1.4 ± 0.4) x 103 LG2 8.39 L9 3.45 x 10"2 20 (6.3 ± 0.4) x 103 LG3 8.10 L14 1.67 xlO" 1 4.2 (1.3 +0.4) x 103 LG3 8.58 E l l 1.07 xlO" 1 6.5 (2.0 ± 0.4) x 103 LG3 8.55 L10 2.97 x 10"2 23 (7.3 ± 0.5) x 103 NG2 8.4 - 2.31 xlO" 1 3.0 (1.0,± 0.3) x 103 NG2 8.0 - 1.73 x 10"1 4.0 (1.3 ± 0.4) x 103 - NG2 7.8 1.16 x 10"1 6.0 (1.9 ± 0 . 4 ) x 103 NG3 8.42 E l l 1.61 x 10"1 4.3 (1.4 ± 0.4) x 103 NG3 8.10 Nle7 1.54 xlO" 1 4.5 (1.4 ± 0.4) x 103 NG3 8.50 NlelO 3.85 xlO" 2 18 (5.7 ± 0.5) x 103 a the proton assignments were made using 2D NOESY and COSY experiments in the C H 3 C O O H / CH 3 COO"Na + Buffer. The assignments for NG2 were not possible. b these values are based on the half-life of an unprotected proton at 20 °C at pD 5.02 to be 3.18 x 10"3h. 114 Several of the amide protons exhibit substantial protection from exchange, although all of the protection factors tabulated in Table 2.7 are characteristic of molten globule-like structures. For all of the caviteins, however, it appears as though the most protected proton was a middle hydrophobic residue. It is interesting that for all of the caviteins the middle residue was the most protected from exchange given that there is experimental evidence from another project in our research group which has shown that the middle Leu residue in the lg2 peptide sequence acts as a linchpin, holding both the secondary and tertiary structure of the LG2 cavitein intact.49 2D COSY and NOESY experiments were performed in C H 3 C O O H / CH 3 COO"Na + buffer for each of the caviteins to get accurate assignments for the N-H/D exchange experiments (data not shown). Note that the assignments for the amino acid residues listed in Table 2.7 may differ slightly in the ppm values for similar residues outlined in Tables 2.8, 2,9, and 2.10 for the respective caviteins because different buffers were used. For NG2 the individual proton assignments were not possible since the 2D COSY and NOESY spectra were too broad, and therefore, not resolvable. 2.2.4.8 Variable Temperature *H NMR Spectroscopy The stabilities of the caviteins towards temperature, with respect to secondary structure, were probed by monitoring [$222 as a function of temperature by CD spectroscopy, and a minimal decrease in helicity was observed in all cases. To determine whether temperature induced a change in the tertiary structures of the caviteins, variable temperature ! H NMR studies were completed. It was hypothesized that an increase in temperature would cause an increase in the motion of the side chains within the hydrophobic core especially, and hence sharpen the 115 signals in the ! H NMR spectrum due to averaged signals. Furthermore, in the case of NG2 the peaks at -0.5 and -3.0 ppm were also examined to see if an increase in temperature would affect these signals. Figures 2.21, 2.22, 2.23, 2.24, and 2.25 show the stack plot ! H NMR spectra at various temperatures for LG2, LG3, NG2 (Figure 2.23 and 2.24) and NG3, respectively. Figure 2.21. Stack Plot of the 500 MHz *H NMR Spectra Illustrating the Temperature Dependent Amide Regions of LG2 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (a) 20 °C (b) 30 °C (c) 40 °C (d) 50 °C (e) 60 °C (* = cavitand signals) I I [ I I I I [ II I I | I I M j 1 . 1 I I | I I I T p T ' T ' T | T I I I [ M I I J T T T T j T T l"l | I I' I 'i j ! ! r ' " | ' ! T~T T—1 T I T I 1.1 1 1 I I | I' M I | I I M | I 1 1 I j I 1 1 1 | I T T T | T T 9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 116 ure 2.22. Stack Plot of the 500 MHz *H NMR Spectra Illustrating the Temperature Dependent Amide Regions of LG3 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (a) 20 °C (b) 30 °G (c) 40 °C (d) 50 °C (e) 60 °C (* = cavitand signals) (a) (b) (c) (d) (e) r p -m - i - r r r i p , 11 p i i i p m p . n i p n i p i i i - r , n 1.i i i i | M . M | n M | , M i p X M p I M p I M p I M | . M 6.2 5.8 9.8 9.4 9.0 8.6 8.2 7.8 ppm 7.4 7.0 6.6 117 Figure 2.23. Stack Plot of the 500 MHz J H NMR Spectra Illustrating the Temperature Dependent Amide Regions of NG2 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (a) 20 °G (b) 30 °C (c) 40 °C .(d) 50 °C (e) 60 °C (* = cavitand signals) T - r r r yrrr "['I'T I i i ' j r r n [ i ]• T " i " | " T ' I ••|iT"|"n"T'.T|'*T"T"r*r"pr"T"T"T"pT"T"T'T | i \ i r | rr I T"|" i-iVryrri-ry'r'T'TT'i r i ' l i [ i_ i i i j i i i i j )T-T^jT"rrry"r:rT"T-|--T"T"rr-|-"1--|-9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 ppm 118 Figure 2.24. Stack Plot of the 500 MHz 'H NMR Spectra Illustrating the Temperature Dependence of the Upfield Regions of NG2 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (a) 20 °C (b) 30 °C (c) 40 °C (d) 50 °C (e) 60 °C (a) (b) (c) (d) (e) | M I I | M I I | I M I | I I I I | I I I I | I M I | I .M l . | I I I I | I I I I | l-l I I | . M I I | I M l . | I I I I | I II l . | I I M | I I I I | I I I I | I M I | M I I | I I I | | 0.0 -0.4 -0.8 -1.2 -1.6 -2.0 -2.4 -2.8 -3.2 -3.6 -4.0 ppm 119 Figure 2.25. Stack Plot of 500 MHz J H NMR Spectra Illustrating the Temperature Dependent Amide Region of NG3 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (a) 20 °C (b) 30 °C (c) 40 °C .(d) 50 °C (e) 60 °C (* = cavitand signals) (a) (b) (c). (d)" '(e) M | I I I I | M I I | M I I | M M | I I I I | I I I I | I I I I | I M I | I I I I | I I I I | I I I I | I I M | I I I I | I I I I | M I I | I I I I | I I I I | I I I I | M I I | I I I I I.I I I I | M 9.8 9.4 9.0 8.6 8,2 7.8 7.4 7.0 6.6 6.2 5.8 ppm For all the caviteins an increase in temperature did result in fewer and less disperse signals in the amide regions of their *H NMR spectra. It can be concluded that as the temperature was increased, the motion of the side chains within the hydrophobic core also increased, leading to overlapping signals. This data supports that the thermodynamics of the caviteins are changing. An increase in temperature results in the caviteins populating more conformations, which in result causes the amide protons to experience similar environments and hence overlapping signals. 120 Furthermore, in the case of NG2 the signals at -0.5 and -3.0 ppm disappeared into the baseline and were no longer detectable at a temperature of 60 ° C . After the variable temperature experiments were completed, an additional ! H NMR spectrum for each of the caviteins at 20 °C was acquired, and this spectrum was indistinguishable with the original spectrum acquired at 20 ° C . ' . ' 2.2.4.9 Two-Dimensional (2D) Homonuclear *H N M R Spectroscopy In ID NMR spectroscopy, qualitative information can be acquired about the.tertiary structures of proteins by inspecting the amide and aliphatic regions. However, to obtain data on amino acid connectivities or through-space residue relationships, double or multiple irridation techniques must be used. Using 2D NMR spectroscopy, information about the protein sequence can be attained and used to label NMR signals to specific amino acid residues within a protein.41 The two most common techniques are NOESY, nuclear Overhauser enhancement spectroscopy, which gives information about spatial coupling and hence about the structure of a protein, and PQF-COSY, double-quantum-filtered-correlation spectroscopy which provides information about J-coupling or through bond coupling.4 1'5 0 It is important to realize, however, that no coupling is observed through a peptide bond. A short introduction to COSY and NOESY techniques is provided in the following paragraphs. Cross peaks in a 2D protein COSY spectrum can occur between the following hydrogens N H / C a H , CaH/CpH, and C p H / C 7 H 4 1 The J-coupling between the above protons depends on the dihedral angle formed by these bonds, and this dependence is usually described by an empirical expression, known as the Karplus equation. Evaluating the coupling between NH/CaH is - : • i2i dependent on the dihedral angle 0, and thus provides information about the backbone conformation of a protein. When looking at a COSY spectrum, chemical shift information is contained in the diagonal starting from the upper right hand corner to the lower left and corresponds to the ID ! H NMR spectrum. Scalar coupling connectivities between individual protons are manifested by cross peaks located at the intersections of straight lines parallel to the axes.503 A main advantage of a single COSY spectrum is that it can in principle outline all scalar spin-spin coupling connectivities between protons in a protein, and is therefore much more efficient than ID spin-decoupling experiments. Furthermore, since the resonance signals are spread out in two dimensions, signal separation is largely improved. NOESY techniques are similar to those for COSY, and numerous COSY signals show up in a NOESY spectrum since hydrogen atoms connected by scalar couplings are also usually located at a short distance through space. One of the advantages of a 2D nuclear Overhauser enhancement (NOE) experiment is that the irridation at one position and the intensity in the response at another position is symmetrical (i.e. the irridation in the reverse order is equivalent), which may not always hold for ID NOE experiments. Furthermore, the NOE depends On 1/r6 where r is the distance between the protons.41 Generally, the stronger the NOESY cross peak, the closer the two interacting protons. However, due to this strong distance dependence, we are limited to studying protons separated by 1.8-5 A. Based upon the stereochemistry of. a dipeptide, there are three nearest-neighbor distances to consider, denoted dNN, d a N, dpN (Figure 2.26).41 For an oc-helix, strong NOEs are observed between NH,/NH,+y, Cp;/NH,+/, and sometimes between NH,/NH,+.< and N H , / N H , + / ' 122 Figure 2.26. Diagram Showing a Polypeptide Segment with COSY Connectivities Shown in Dotted Lines, and NOE Connectivities Shown with Arrows. In a NOESY spectrum the area below the diagonal can be divided into six areas, each containing specific types of signals. For example, one area contains "NH aromatic-NH aromatic signals", whereas another area contains "NH aromatics - aliphatic side chains" signals.41 Using the information from the location of the peaks within a NOESY spectrum aids in making individual assignments to specific amino acid residues. COSY and NOESY spectra were collected for all of the caviteins by Okon in the laboratory of Mcintosh, and both spectra for each of the caviteins were used to label the observed proton signals to specific amino acid residues. Figures 2.27, 2.28 display the COSY and NOESY spectra for LG2, respectively. Figures 2.29, 2.30, and 2.31 display the COSY (full), COSY (expanded) and NOESY spectra for LG3, respectively. Lastly, Figures 2.32, and 2.33 display the COSY and NOESY spectra for NG3, respectively. The first set of assignments for LG3 (explained after Figure 5.31) was done with the help of Okon, and the assignments for caviteins LG2 and NG3 were completed in a similar manner single-handedly. 123 Figure 2.27. 2D 500 MHz *H NMR COSY Spectrum of LG2 at ~1.5 mM in 10 % D 2 0,45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (expanded in the aliphatic/amide region) OO O N Q OO S O : ; |oo g 1^ -oo oo in oo oo 00 I N oo OS o o cn r-—i 1H ppm 124 Figure 2.28. 2D 500 MHz 'H NMR NOESY Spectrum of LG2 at -1.5 mM in 10 % D 2 0,45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (expanded in the amide region) 1H ppm 125 Figure 2.29. Full 2D 500 MHz ' H NMR COSY Spectrum of LG3 at -1.5 mM in 10 % D 2 0,45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. Figure 2.30. 2D 500 MHz ! H NMR COSY Spectrum of LG3 at -1.5 mM in 10 % D 2 0,45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (expanded in the aliphatic/amide region) :^  ^  ^ : U^F ^  ^  M ^ ! H M M M M Ui U ^  ^  :=^! ^  ^  H=M rT^r^i^^^^^^^^^' ^  a. :>y^^^ ::^^{^: :: :: :: :: :: :: O O OS p od os od co od od CL. o . m od od oo co co 'oo, oo oo o\ oo ON O oo O CO oo CO O O C S co CO 1H ppm 127 Figure 2.31. 2D 500 MHz ! H NMR NOESY Spectrum of LG3 at -1.5 mM in 10 % D 2 0,45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (expanded in the amide region) 1H ppm 128 Looking at the amide region on the COSY spectrum in Figure 2.29 there is a set of cross peaks at -7.2 ppm which are attributed to the NH/NH through bond interaction of the amide cap protons on the C-terminal glycine (G17) residue. The subsequent assignments were achieved by examining the amide region on the expanded NOESY spectrum in Figure 2.31. The first step involved identifying the amide cap protons on the NOESY spectrum again at -7.2 ppm, which have a NOE interaction with the C-terminal Gl7 amide proton. The other cross peak for G17 must correspond to an adjacent amino acid which leaves K16. The assignment for these two residues can be confirmed by examining the COSY spectrum to determine the specific locations of the C K H of G17 and C a H of K16 signals. It is clear from the COSY that the signal at 3.92 ppm is from Gl7 and the other at 4.25 ppm is from K16. Then referring again back to the NOESY the cross peak at 7.85 ppm must be the NH of K16. This signal at 7.85 ppm is connected to another signal at 7,8 ppm. The signal at 7.8 ppm is determined to be an NH from K15 from the COSY spectrum. Following these steps, it is possible to continue moving down the peptide chain from C- to N-terminus and assign all of the amide protons to specific residues. This process was carried out for all of the caviteins in order to make the proton assignments. 129 Figure 2.32. 2D 500 MHz 'H NMR COSY Spectrum of NG3 at ~1.5 mM in 10 % D 2 0,45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (expanded in the aliphatic/amide region) Figure 2.33. 2D 500 MHz 'H NMR NOESY Spectrum of NG3 at ~1.5 mM in 10 % D 2 0,45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (expanded in the amide region) A tabulation of the proton assignments for LG2, LG3 and NG3 are shown in Tables 2.8, 2.9 and 2.10, respectively. The assignments for NG2 were not possible due to poor resolution in the 2D COSY and NOESY 'H NMR spectra. Table 2.8. Tabulated Data from the COSY and NOESY Experiments of LG2 in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (note that sequence numbering began at the first glycine residue of the linker) Residue Label N H Chemical Shift (ppm) C«H Chemical Shift (ppm) NH_-cap 7.20 • -NH_-cap . 7.43 - • K8 7.67 4.16 L5 7.70 4.07 K14 7.72 4.14 K7 7.81 4.11 E l l 7.84 4.00 LI 2 7.93 4.10 L6 7.94 3.90 G16 8.15 3.91 L9 8.40 4.05 E10 8.42 3.85 E4 8.90 4.08 132 Table 2.9. Tabulated Data from the COSY and NOESY Experiments of LG3 in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (note that sequence numbering began at the first glycine residue of the linker) Residue Label N H Chemical C«H Chemical Shift (ppm) Shift (ppm) NH2-cap 7.18 NH2-cap 7.42 E12 7.71 4.00 L6 7.75 4.09 K15 7.81 4.16 K8 7.84 4.10 K16 7.85 4.26 E4 7.91 3.96 L13 7.91 4.10 L7 8.00 3.90 G17 8.12 3.93 E5 .. 8.15 4.04 L14 8.22 4.08 E l l 8.48 3.82 L10 8.56 4.03 133 Table 2.10. Tabulated Data from the COSY and NOESY Experiments of NG3 in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (note that sequence numbering began at the first glycine residue of the linker) Residue Label NH Chemical Shift (ppm) C«H Chemical Shift (ppm) NH2-cap 7.18 -. NH2-cap 7.42 -K15 7.77 4.16 E12 7.78 4.02 Nle6 7.79 3.96 E4 7.82 3.95 K16 7.86 4.25 K8/K9 7.87 4.13 Nlel3 7.99 4.02 Nle7 8.10 3.88 G17 8.10 3.93 Nlel4 8.17 4.01 E5 8.23 4.03 E l l 8.37 3.84 NlelO 8.50 3.90 Having the ability to label the signals in the NMR spectra of our caviteins to specific residues in the peptide sequence is especially helpful in assigning protection factors to specific protons in the N-H/D exchange experiments. As was mentioned previously, for LG2, LG3, and 134 NG3 the most protected protons belonged to the middle leucine or norleucine residues in the peptide sequence, as was hypothesized. Comparing the resonance amide assignments for the middle residue of the leucine-based caviteins (i.e. L6 for LG2 and L7 for LG3) one can notice that L6 of LG2 is located more upfield at 7.92 ppm. L7 for LG3 is located at 8.00 ppm, respectively. These findings again support that the linker between the cavitand template and the peptide sequence has a strong influence on the overall folded structure of the caviteins. 2.2.4.10 ANS Binding l-anilinonapthalene-8-sulfonate (ANS) is a hydrophobic, nonpolar dye, which binds preferentially to exposed hydrophobic surfaces of a protein, and can be effectively detected using fluorescence spectroscopy.51 Molten globule-like structures typically expose more hydrophobic surfaces through their mobility, and therefore typically bind nonpolar molecules more strongly than native-like states.52 Furthermore, ANS seldom binds to unfolded proteins. When ANS binds to a nonpolar surface the ANS fluorescence intensity increases greatly and shifts to lower wavelengths. ANS binding was studied by fluorescence spectroscopy for LG2, LG3, NG2 and NG3 and the emission spectra are shown in Figure 2.34. Under standard experimental conditions53 negligible binding was observed for any of the caviteins. The studies were completed using three different concentrations per cavitein of 50, 100, and 150 uM, and the curves were indistinguishable. Figure 2.34 only shows the fluorescence emission spectra for the caviteins at a concentration of 50 uM for the sake of clarity. 135 Figure 2.34. Fluorescence Emission Spectra of 2 u\M ANS in the Presence of 95 % Ethanol, 100 % Methanol, 50 uM LG2, LG3, NG2 and NG3 at 20 °C in pH 7.0 50 mM Sodium Phosphate Buffer. No appreciable binding of ANS was observed for any of the caviteins. It was hypothesized that some binding of ANS would be observed for the norleucine-based caviteins due to their molten globule-like behavior reflected by the poor dispersion and broad signals in their 'H NMR spectra. Out of the four caviteins NG3 was found to bind the most, which is still considered negligible compared to the binding of ANS in MeOH or EtOH. Comparable ANS binding experiments were done for the lg2 peptide alone to determine whether the cavitand template was in some way inhibiting the binding of ANS, although no binding was observed (data not shown). 136 The presence of a denaturing agent, like GuHCl, has been shown to induce ANS binding in some native-like proteins. Furthermore, studying ANS binding in the presence of a denaturant can yield information about the transition from a native-like state to the unfolded state.54 Therefore, comparable ANS binding experiments were conducted in the presence of 0-6.0 M GuHCl to detect if binding would be observed under unfolding conditions, and similarly no binding was observed suggesting that there were no observable unfolding intermediates near the unfolding transition. It has also been suggested that ANS binds preferentially to hydrophobic pockets and not exclusively to hydrophobic surfaces in molten globule-like structures, and if our caviteins lack a complementary binding pocket for ANS perhaps this is could be a reason for no observable binding. 2.3 Summary and Conclusions This Chapter presented the design, synthesis and characterization of four caviteins, LG2, LG3, NG2 and NG3, in the hopes of designing native-like proteins, and also diagnosing native-like structure. The peptide sequences were designed and synthesized using a minimalistic approach, and were then covalently linked on to a template to overcome the entropic costs of peptide association. All four caviteins had similar amounts of secondary structure with the a-helices being rigid and well-structured. The oligomeric states of the caviteins were studied by CD spectroscopy and A U C and were found to exist as monomers in solution. Experimental determination of the partial specific volumes of the caviteins, by density perturbation A U C 137 experiments, was completed, and it appears as though the cavitand template does not have much of an influence on the partial specific volumes of the caviteins (experimental and calculated partial specific volumes were similar for all caviteins). The main focus of this project was to further investigate the apparently native-like characteristics of LG2 and LG3 by creating NG2 and NG3, which should manifest greater side chain mobility, and hence poorer packing. In other words, we were attempting to deduce whether we could induce a molten globule-like structure into seemingly native-like proteins, and to study the dynamics of both systems. Norleucine amino acids were substituted into the all leucine-based cavitiens, LG2 and LG3, to produce NG2 and NG3, respectively. Upon examination of all of the experimental data (Table 2.11), it can be concluded that LG3 exhibited the most definitive native-like properties, sharp dispersed amide signals, greatest stability and largest protection factor. Table 2.11. Summary of Results for LG2, NG2, LG3 and NG3. Cavitein Far-UV Near-UV 'H NMR Protection Factor .ANS binding GuHCl ( A G ° H 2 O ) kcal/mol LG2 helical min/max broad, dispersed (6.3 ± 0.4) x 103 no binding -10.4 ±0 .3 NG2 helical min/max broad, dispersed (1.9 +0.4) x 103 no binding -7.0 ± 0.3 LG3 helical max/mi n sharp, less dispersed (7.3 ± 0.5) x 103 no binding -10.8 ±0 .4 NG3 helical max/min sharp, less dispersed (5.7+ 0.5) x l O 3 no binding -7.5 ± 0.3 138 The ! H NMR amide region of the LG2 spectrum was the most dispersed of all caviteins but the signals were relatively broad, where as the signals in LG3 were sharpest although slightly less dispersed. Normally, the amide regions of native-like proteins are sharp and disperse, except for coiled coil structures, which have a lower dispersion in the amide region.45 It has been discovered by molecular dynamics simulations,55 performed by Scott and Straus, that LG3 does show evidence of supercoiling, and thus the lower dispersion in the amide region supports a native-like coiled coil structure. Two-dimensional NMR spectroscopy was used to assign the protons signals in the amide region of the lH NMR spectra to specific amino acid residues for each of the caviteins (because four identical peptide sequences were attached to a single cavitand template, the identities of the residues could not be assigned to specific peptide strands). Only minor differences in the labeling patterns were observed for the four caviteins, which was expected since the four peptides sequences were similar. Although there were differences in the structural characteristics of the four caviteins, the largest differences still appear to be a result of the linker between the cavitand template and the peptide sequence. Overall, it appeared as though the 2 Gly-linked caviteins behave similar to each other, as did the 3 Gly-linked caviteins, regardless of the hydrophobic residue in the peptide sequence. Overall, the substitution of the leucine residues to norleucine units did slightly reduce the native-like character of their corresponding caviteins, for example, lower protection factors and reduced stability for the norleucine-based caviteins. However, since the ! H NMR spectra for LG3 and NG3, and LG2 and NG2 were similar, respectively, it can be concluded that the tertiary structures of LG2 and LG3 were marginally affected by the substitution with norleucine residues. A possible explanation is that the norleucine residues pack as well as the leucine 139 residues within the hydrophobic core, although there is an entropic penalty associated with restricting the total number of conformations of the norleucine side chains. The entropic cost of burying the norleucine side chains was evident in the less negative A G ° H 2 O values of unfolding. It would be reasonable to conclude that the leucine-based caviteins contain hydrophobic cores that are well-packed, supported by their highly negative A G ° H 2 O values of unfolding. The packing in the hydrophobic cores of the norleucine-based caviteins was slightly less favorable, and hence resulted in the slightly more molten globule-like characteristics. Overall, the CD, NMR, and fluorescence data in this Chapter was consistent with the leucine-based caviteins exhibiting an appreciable extent of native-like structure. Furthermore, it was evident that the number of glycine linkers between the cavitand template and the peptide sequence has a huge effect on the overall tertiary structure of the caviteins. It should be noted that L G 2 and L G 3 have been typified as some of the most native-like de novo proteins synthesized to date.la Our caviteins have been limited to having only one type of peptide sequence attached within one bundle. The following Chapter will present methodologies towards creating TASPs with two different peptides sequences attached within one bundle, including an anti-parallel derivative. 140 2.4 Experimental 2.4.1 Arylthiol Cavitand Synthesis 2.4.1.1 General All chemicals used for the synthesis tetrathiol cavitand 5 were reagent grade and purchased from Aldrich. HPLC grade A/,N-dimethylacetamide (DMA) was dried under 4 A molecular sieves. Reagent grade THF was distilled under nitrogen from sodium benzophenone ketyl. Matrix Assisted Laser Desorption Ionization (MALDI) mass spectra were recorded on a Bruker Biflex IV in reflectron mode using 2,5-dihydroxybenzoic acid (DHB) as the matrix. The 'H NMR spectrum of cavitand 5 was recorded on a Bruker A V A N C E 400 MHz spectrometer at ambient temperature using the residual ' H NMR signal (7.24) from deuterated CDCI3 as a reference. Column chromatography was used for purification using silica gel (230-400 mesh, BDH) and silical gel glass backed analytical plates (0.2 mm) were used for thin layer chromatography (TLC) with U V detection. All products were dried at least overnight at room temperature and 0.1 torr. The synthesis of methyl-footed arylthiol 5 was accomplished through four steps following literature procedures,7 and a slight modification to the final purification step is highlighted in the next Section. 141 2.4.1.2 Synthesis of Methyl-Footed Arylthiol (or Cavitand) 57 A modification to the literature procedure was done in the final purification step, which involves recrystallization and separation steps in order to purify the arylthiol (tetrathiol) cavitand from a small amount of the tris byproduct. This purification was not carried out, as it was easier to use the arylthiol cavitand 5 in impure form in subsequent cavitein reactions, and then purify the caviteins by reversed-phase HPLC after the incorporation of the peptide strands. *H NMR (400 MHz, CDC13) 5 6.96 (s, 4H, H p a r a ) , 5.95 (d, J = 7.0 Hz, 4H, H o u t ) , 4.94 (q, J = 7.4 Hz, 4H, H m e t h i n e ) , 4.36 (d, J = 7.0 Hz, 4H, H i n), 3.76 (s, 4H, SH), 1.71 (d, J = 7.4 Hz, 12H, CH 3 ) PPm-MS (MALDI, DHB) m/z: 721 (M + H) + 142 2.4.2 Peptide Synthesis 2.4.2.1 General All reagents for the peptide syntheses were reagent grade (except where noted otherwise), and purchased from Advanced Chem Tech. The peptides were purified by preparative reversed-phase HPLC using a Perkin-Elmer Biocompatible Pump 250 with a PE LC90 BIO Spectrophotometric U V detector and a KJPP and ZONEN chart recorder. A Phenomenex Selectosil Ci8 reversed-phase HPLC column (preparative: 250 mm x 10 mm, 10 | i M particle size, 300 A pore size) was used. The wavelength for the U V detection was set at 229 nm for recognition of the amide chromophore. The peptide samples were filtered through a 0.45 | i M Nylon™ syringe filter (Phenomenex) prior to injection and run at a flow rate of 10 mL/min using helium sparged filtered water (0.1 % TFA)/HPLC-grade acetohitrile (0.05 % TFA) gradient. Peptides samples were analyzed by analytical reversed-phase HPLC and were filtered prior to injection onto a Varian 9010 Pump with a Varian 9050 U V detector and a Varian 4290 Integrator. A Phenomenex Selectosil Cig reversed-phase HPLC column (analytical: 250 mm x 4.5 mm, 5 pJVI particle size, 100 A pore size) was used. Analytical samples were run at a flow rate of 1 mL/min using the same solvents as for the preparative purification. The purified samples were evaporated in vacuo and lyophilized. The mass spectra were run on a MALDI-MS Bruker Biflex fV in reflectron mode using 50 |J,M cinnamic acid in 1:1, H2O: MeCN, as a matrix. 143 2.4.2.2 Synthesis of Peptides 6 (lg2), 7 (lg3), 8 (ng2), and 9 (ng3) The peptides were synthesized following to a large extent literature procedures, with peptides 6 and 7 having previously been synthesized.13 The synthesis of peptide 6 is described below, and similar procedures were followed for the synthesis of peptides 7 (lg3), 8 (ng2), and 9 (nleg3). The peptide synthesis involved using standard Fmoc techniques on an automated Applied Biosystems peptide synthesizer attached to an Apple Usi Macintosh computer. All Fmoc protected amino acids, solvents and coupling reagents were purchased from Advanced Chemtech (Louisville, K Y , USA). The peptides were synthesized on a 0.25 mmol scale using the FastMoc™ protocols. Side chain protected amino acids were used for chemoselective synthesis of the peptide^ which was in turn bound by its C-terminus to a resin developed by Rink to afford a C-terminal amide upon cleavage. A single amino acid coupling cycle included a: (1) 13 min Fmoc deprotection using piperidine, (2) 6 min wash step using N-methylpyrrolidone (NMP), (3) 30 min coupling step to 1.0 mmol of the next Fmoc amino acid using 2-(lH-benzotriazole)l,l,3,3-tetramethyluronium hexafluorophosphate (HBTU) and 1-hydroxybenzotriazole (HOBt) as coupling reagent (note activation of the amino acid started with DEPEA), and lastly (4) 6 min wash with NMP. NMP was the solvent throughout the synthesis with each cycle having an approximate time of 55 minutes. For a schematic representation of a peptide synthesis see Figure 2.35, and for more details for peptide synthesis see the ABI 431A manual. 144 Figure 2.35. Schematic Representation of the FastMoc™ Protocol on the AB1 431 A Peptide Synthesizer. R, \ X ^ C-terminus (t-Butyl) ot peptide 1. Deprotection with Piperidine 2. Wash Fmoc Protecting Group (t-Butyl) Hi O O Ri \ (t-Butyl) / / / (t-Butyl) O R, 1 (t-Butyl) Deprotection with Piperidine (t-Butyl) I. Coupling ot next amino acid with HOBt/HBTU and DIEA 2. Wash HBTU = f| P F B ft , i HOBt = 0 > Fmoc-NH-Amino Acid = O O—I— NH-Amino Acid Peptide on Resin with Fmoc Protecting Group Removed 1. Treatment with Chloroacetylchloride ; Cleaved Peptide with 2. 95% TFA Protecting Groups Removed Thereafter, chloroacetylation of the free N-terminus was achieved through manual treatment of the resin (600 mg peptide resin, -300 mg peptide 6, -0.160 mmol) with 145 chloroacetyl chloride (75 uL, 0.96 mmol, 6 equiv.) and DIPEA (165 uL, 0.95 mmol, 6 equiv.) in D M F for 1 hour at room temperature under nitrogen. The last step included cleavage of the peptides from the resin in addition to removing the side chain protecting groups simultaneously using a 2 hour treatment with 95 % T F A / H 2 0 . An ice bath was used for the first 10 minutes of the reaction. After completion, the resin was removed by suction filtration through a medium frit filter with a CH2CI2 wash. The TFA/CH 2 C1 2 filtrate was evaporated to a few mLs in vacuo and the crude peptide was precipitated using ice-cold diethyl ether. The peptide was recovered by suction filtration using a fine frit filter. The peptide was then dissolved in distilled water, filtered, and purified by re versed-phase HPLC. The peptide was lyophilized until a fluffy white solid peptide was obtained (105 mg, 23 %). The peptides were characterized for purity by the inspection of a single peak by analytical reversed-phase HPLC (>95 % pure), and using MALDI-MS. Table 2.12. % Yields and MALDI-MS Characterization of the "Activated" Peptides. Peptide % Yield Mass (Da.) lg2 • 23 1860 lg3 26 1916 ng2 28 1860 ng3 25 1916 146 2.4.3 Cavitein Synthesis 2.4.3.1 General All reagents for the cavitein syntheses were reagent grade (except where noted otherwise), and purchased from Aldrich. HPLC-grade Af,N-dimethylformamide (DMF) was stored under 4 A molecular sieves. GuHCl was electrophoresis grade. The caviteins were purified in a similar manner to the peptides. Purified cavitein samples were evaporated in vacuo then lyophilized and their masses confirmed by MALDI-MS. M A L D I spectra were acquired as described for the peptides above. Cavitein concentrations were determined using a Bradford Assay 5 6 measured on a C A R Y UV-visible spectrophotometer. The pH's of the buffers were determined using a Fisher Scientific Accumet pH meter 915 calibrated with two purchased buffer standards (pH = 4.0 and 10.0). 2.4.3.2 Synthesis of Caviteins 10 (LG2), 11 (LG3), 12 (NG2), and 13 (NG3) Caviteins 10,11,12, and 13 were synthesized by following literature procedures. • The synthesis of cavitein 10 is described below, and similar procedures were followed for the synthesis of caviteins 11, 12 and 13. A solution of cavitand 5 (1.1 mg, 1.4 (imol, 1 equiv.) and peptide 6 (21 mg, 11.3 (imol, 8 equiv.) were stirring in degassed D M F under N 2 . DIPEA (2.5 pL, 15 p:mol, TO equiv.) was added in excess until the solution turned cloudy. The reaction was monitored (appearance of the cavitein peak) by analytical reversed-phase HPLC and was complete after 4 hours. The crude reaction mixture was evaporated in vacuo, dissolved in water, 147 filtered, and purified by reversed-phase HPLC to yield cavitein 10 as a fluffy white solid (5.5 mg, 49 %) after lyophilization. The additional unwanted tris-cavitein byproduct was separated and removed during purification. The caviteins were characterized using MALDI-MS and the masses of the caviteins are outlined in Table 2.12. Table 2.13. % Yields and MALDI-MS Characterization of the Caviteins made from "Activated" Peptides. Cavitein % Yield Mass (Da.) LG2 49 8016 LG3 54 8240 :.NG2 51 8016 NG3 46 8240 2.4.4 C D Studies All CD spectra were recorded on a JASCO J-710 spectropolarimeter except for the temperature-dependence experiments which were run on a JASCO J-810 equipped with a computer-directed water-bath at the U B C Laboratory of Molecular Biophysics (UBC-LMB) supervised by Rosell located in the Life Sciences Center. The J-710 had a circulating water bath set to 25 °C, a 400W xenon lamp, and an IBM-compatible PC computer for data acquisition. Some of the parameter settings include: 0.1 nm step resolution, 2 nm bandwidth, and 50 nm/min scanning speed. The J-710 spectropolarimeter was calibrated routinely using d-l0-(+)-148 camphorsulfonic acid. 5 7 Each spectrum was an average of three scans subtracted from a reference background scan. Individual samples were run three different times to ensure reproducibility. The caviteins were monitored at 4 jiM and 40 uM in 50 mM sodium phosphate buffer (pH = 7.02) to check for concentration effects in a 1 cm and a 1 mm quartz cuvette, respectively. The raw spectra were normalized to a mean residue ellipticity [9] using the following equation: [0\ = dobsl 10/cn where 60bs is the observed ellipticity measured in millidegrees, / is the path length in cm, c is the cavitein concentration in mol/L, and n is the number of residues in the cavitein. Errors were on average ± 5 %. For comparison, maximum value molar ellipticities at 222 nm were calculated using the following equation developed by Chen et al.:16 [ f e - M A x = 39500 [ 1 - ( 2 . 5 7 / « ) ] where n is the number of residues per helix. Note that 39500 and 2.57 are wavelengths constants at 222 nm only. Guanidine Hydrochloride (GuHCl) denaturation experiments were performed between 0 and 8.0 M GuHCl in a 50 mM sodium phosphate buffered (pH = 7.0) cavitein solution. The exact concentration of the 8.0 M GuHCl stock solution was determined by the refractory method 149 outlined by Pace22 using an Abbe refractometer (Physical Chemistry Lab of Ben Clifford at UBC), and found to be to within 0.8 % of 8.0 M GuHCl. Data points were collected at 1 molar units between 0 and 8.0 M to generate a rough unfolding curve. 0.25 interval denaturation studies were then completed to achieve accuracy in the unfolding region, and repeated three times to ensure reproducibility. Likewise, cavitein samples were monitored for unfolding at 4 pM and 40 p M to study concentration effects in a 1 mm and a 1 cm quartz cuvette, respectively. Samples were prepared immediately before data acquisition and equilibrated for 10 min (previously determined that any effect Of GuHCl is immediate). The mean residue ellipticity was again monitored at [6\ = 222 nm. Protein unfolding was analyzed using the linear extrapolation method of Santoro and Bolen. 2 3 According to this method, unfolding is a reversible, two-state process and that the free energy of folding is a linear function of the GuHCl concentration. The GuHCl denaturation data were fit using a nonlinear least-squares analysis to fit the pre-transitional baseline using the following equation: 0obs = ^ (/*)(! - afGuHCl]) + So (1 -fN) where 60bs is the mean residue ellipticity at 222 nm at a certain concentration of GuHCl, is the mean residue ellipticity of the folded state in the absence of GuHCl, &u is the mean residue ellipticity of the unfolded state, a is a constant andfy is the fraction of the protein in the folded state, / J V is related to the free energy of unfolding, A G ° H 2 O , by the following equation: r ((AG0H')O-m [GuHCl])/RT) / r i . f ( A G ° H , 0 - m [GuHCt\)IRT) n IN= & / I + ? z I 150 where A G ° H 2 O is the free energy of unfolding in the absence of GuHCl, m is the free energy change with respect to the concentration of GuHCl, R is the universal gas constant, and T is the temperature. A Macintosh compatible computer program, KaleidaGraph V. 3.G8d was used to calculate the values for A G ° H 2 O by a nonlinear least-squares regression analysis. The value of fa was normalized to one. The software analysis program calculated the reported errors in this thesis. 2.4.5 Analytical Ultracentrifugation (AUC) Experiments Sedimentation equilibrium and velocity studies were carried out on a temperature-controlled Beckman Coulter Optima™ XL-I analytical ultracentrifuge (located in the laboratory of Creagh in the N C E building on the UBC campus). Sedimentation equilibrium experiments were run using either an An60 Ti rotor, or an An50 Ti rotor (4 sample holders and 8 sample holders, respectively) and a U V photoelectric scanner. A six sector cell, equipped with a 12 mm Epon centerpiece and quartz windows, was loaded with 3 x 120 fiL of sample at 3 different concentrations made up in 50 mM sodium phosphate buffer at pH = 7.0, and 3 x 130 U.L of reference solvent (see Table 2.13 below for exact experimental details). Data were collected at 20 °C and at rotor speeds of 27000, 35000, and 40000 rpm until equilibrium was established. Samples were equilibrated for 40 hours and single scans 3 hours apart were overlaid to determine that equilibrium had been reached. Scanning parameters included: radial step size of 0.001 cm, step mode, 10 replicate scans, radial scan range between 5.8 cm and 7.3 cm, and U V detection at 270 nm. The solution density of the samples in sodium-phosphate buffer was taken to be 1.000 g/mL. Partial specific volumes, v , of the caviteins were experimentally determined 151 using parallel sedimentation experiments in solutions of H2O and D2O, and calculated based on amino acid composition.37 Table 2.14. Experimental Parameters for Sedimentation Equilibrium Studies on Caviteins LG2, LG3, NG2, and NG3 Run at 20 °C. Cavitein Cavitein Concentration (|JM) Partial Specific Volume (mL/g) Predominant Species LG2 10 0.78 ± 0 . 0 1 Monomer LG2 50 0.78 ± 0 . 0 1 Monomer LG2 80 0.78 ± 0 . 0 1 Monomer LG3 10 0.75 ± 0 . 0 1 Monomer LG3 50 0.75 ± 0 . 0 1 Monomer LG3 80 0.75 ± 0.01 Monomer . NG2 10 0.78 ± 0 . 0 1 Monomer NG2 50 0.78 ± 0.01 Monomer NG2 80 0.78 ± 0 . 0 1 Monomer NG3 - 10 0.7.9 ± 0 . 0 1 Monomer NG3 50 0.79 ± 0 . 0 1 Monomer NG3 80 0.79 ± 0.01 Monomer In the next few pages a detailed description of A U C sedimentation is provided with a focus on solution properties, method of analyses, and solution behavior. 152 The rotor speed is expressed as the angular velocity, co, in rad/s. The rotor speed is easily determined from the period of rotation (rpm), and is carefully maintained throughout the experiment. Rotor speeds are chosen to straddle the optimum rotor speed for the sedimentation equilibrium run. A set of at least three speeds is chosen to yield varied data for the computer analyses.28 For the three speeds, the ratio of the squares of the two slower speeds should be 1.4 or greater, and the ratio of the squares of the fastest and slowest speeds should be three or greater.32 As for the temperature, it must be known and stable and commonly centrifugation studies are performed at room temperature. With respect to time in the case of sedimentation equilibrium, it is the time necessary to reach an equilibrium concentration gradient that must.be achieved before scans to be analyzed are acquired. Buoyancy is also relevant in analytical ultracentrifugation since the net or buoyant mass, Mb, always appears in sedimentation equations where Mb = Af(l-vp). M is the molecular mass of the solute, and the displaced solvent mass is M v p , where v is the solute's partial specific volume (in mL/g), and p is the solvent density (in g/mL). Therefore by definition Mb is the apparent net mass of a particle in solution, which is equal to the anhydrous particle mass less the mass of the solvent it displaces.28 It is of utmost importance to have great accuracy in the buoyancy term since any error in v or p propagates a threefold error in M. The density of the solvent is simply the mass (in g) of one mL of solvent and is dependent on temperature and composition. Solvent densities can be measured by pycnometery, but published data permits the accurate calculation of p that account for both temperature and buffer composition.278 As for analysis of the sedimentation equilibrium data, it can be done so by methods that fit into one of two groups: graphical analysis58 or by nonlinear least-squares fitting.33 Nonlinear 153 least-squares analysis is the most common method for analyzing sedimentation equilibrium experiments. NONLIN 3 3 is the most widely used program for this purpose. In the NONLIN analysis, a series of curves is calculated to locate a "best" fitting model of the data. Each successive iteration leads to a better approximation of the curve parameters until the approximations lead to a stable value for the parameters being varied. Programs such as NONLIN permit the analysis of several data sets simultaneously, and with enough data acquired over a wide range of concentrations and rotor speeds, it is possible to obtain very accurate estimates of molecular weight, association constants, and stoichiometries. Four common situations can arise from the behavior of the protein in solution: 1) homogeneity (ideal), 2) non-ideality, 3) self-association, 4) polydispersity. When NONLIN analyzes the sedimentation data a plot of absorbance versus molecular weight can provide insight into how the substance of interest is behaving in solution. If the molecular weight decreases with increasing absorbance, this indicates thermodynamic non-ideality, whereas, if the molecular weight increases with increasing concentration, self-association could be the result. The difference between self-association and polydispersity is that self-association is specific. It can be a reversible equilibrium between monomer-dimer, monomer-trimer, monomer-tetramer etc. On the other hand, polydispersity results from an irreversible equilibrium or contamination by non-specific aggregation, which yields a heterogeneous system. Another distinction between ideal and even self-associating samples with polydispersed samples is that the latter will show a systematic decrease in molecular weight with increasing rotor speed. At higher rotor speeds, the larger aggregates in a polydispersed sample may precipitate from solution and result in an apparent molecular weight lower than expected. Non-ideality can be explained in terms of excluded volume or charge-charge interactions, 9R and these interactions tend to increase the apparent concentration. Non-ideality can be 154 corrected for in the analysis of the data by including the second virial coefficient, P, in the exponential Lamm equation. Setting P equal to zero removes all non-ideality, and the model is the same as that described for an ideal self-associating system. If non-ideality is observed in the sedimentation data, one method used to correct this is to run the samples in high salt concentrations.59 In the studies by Dekker, he explained that the salt helped eliminate artificial aggregation, and provided a fit to an ideal species model. The sedimentation equilibrium data was analyzed on a PC compatible software program called NONLIN explained above. This program uses a nonlinear least-squares analysis in order to generate a reduced molecular weight, <r, from which the actual experimental molecular weight, M, can be calculated. Nine sets of data (3 different concentrations at the 3 different rotor speeds) per cavitein were analyzed at a time. The data were initially fit to a single non-associating ideal species model using the Lamm 3 1 equation below: A r = Exp [In (A0) + Mo}( 1 - v p I RT) (r2 - r02)] + E where At is the absorbance at radius r, A0 is the absorbance at a reference radius r0 (the meniscus), M is the molecular weight in g/mol, a) is the angular velocity of the rotor in rad/sec, v is the partial specific volume of the peptide, p is the density of the solvent in g/mL, R is the universal gas constant, T is the temperature in K, and E is the baseline correction factor or baseline offset. To test whether a single species is present in solution a plot of In A vs. r should yield a straight line with a slope proportional to the molecular weight. For non-ideal or associating systems, or when multiple species are present, a straight line will not be obtained. The initial fit of the data using NONLIN was completed as follows: (1) let delta y, baseline offset and reduced molecular weight vary, (2) fix delta y, let baseline offset and reduced 155 molecular weight vary (note: a series of iterations followed until a reduced molecular weight value was converged upon) (3) evaluate error limits. The experimental molecular weight can then be calculated from: <r=(M(l - vp)or)/RT where cris the reduced molecular weight. When the data did not fit well with the equation for a single non-associating ideal species, fits were made to the equation below describing possible oligomeric species: A r •= A ( m o n o m e r i r o ) exp [ ((1 - v p)G? I 2RT) (M(r2 - r02))] .+ A(monomer> r o ) n l Ka>2 exp [-((1 - vp)o} 12RT) n2(M(r2 - r02))] + Aononomer, ro"4 KaA exp [ ((1 - vp)al 12RT) n4(M(r2- r02))] + E where A(monomer< r0) is the absorbance of the monomer at the reference radius r 0, n2 is the stoichiometry for species 2, Ka,2 is the association constant for the monomer -n- mer equilibrium of species 2, n3 is the stoichiometry for species 3, ^ 3 is the association constant for the monomer -n- mer equilibrium of species 3, is the stoichiometry for species 4, is the association constant for the monomer -n- mer equilibrium of species 4. The second fit included a monomer-dimer equilibrium and was completed as follows: (1) let delta y, baseline offset and Ka<2 vary, and fix the reduced molecular weight for a monomer (if unknown let vary), (2) fix delta y, let baseline offset and K&<2 vary; and fix the reduced molecular weight for a monomer (if unknown let vary) (note: a series of iterations followed until 156 a value for was converged upon) (3) evaluate error limits. For thoroughness the data can be fit to the equations describing a monomer-trimer and a monomer-tetramer equilibrium respectively, to eliminate other possible association states. Figures 2.36, 2.37 show the remaining representative fits for LG3. Figures 2.38, 2.39 and 2.40 show the representative fits for LG2. Figures 2.41, 2.42 and 2.43 show the representative fits for NG2. Lastly, Figures 2.44, 2.45 and 2.46 show the representative fits for NG3. 157 Figure 2.36. Sedimentation Equilibrium Concentration Distributions of LG3 at a Rotor Speed of 35000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 | iM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. 5.000 17.500 18.000 18.500 Radius Squared/2 (cmA2) 158 Figure 2.37. Sedimentation Equilibrium Concentration Distributions of LG3 at a Rotor Speed of 40000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 80 j i M . In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. O 1 < O o ft -1.000 23.750 24.250 24750 Radius Squared/2 (cmA2) 25.250 I e o O a o U 0.875 L 0.625 L 0.375 L 0.125 23.750 24.250 24.750 Radius Squared/2 (cmA2) 25,250 159 Figure 2.38. Sedimentation Equilibrium Concentration Distributions of LG2 at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 uM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. CM O i < O c o ft 5.000 2.500 S « J 0.000 -2.500 -5.0001 17.250 o o o ° o © §> o o ° •Sb o o o o $o o ° O o o o ° © ° o o ° o O © 0 _ o ° o 0 O o O 0 <S>~ 0 o . o ©o 17.750 18.250 Radius Squared/2 (cmA2) 18.750 1.250 17.250 17.750 18.250 Radius Squared/2 (cmA2) 18.750 160 Figure 2.39. Sedimentation Equilibrium Concentration Distributions of LG2 at a Rotor Speed of 35000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 pJVI. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. C O o I o I O ID ft 5.000 2.500 L 0.000 L -2.500 L -5.0001 17.250 o o o © O o o o v o e o * ° O o 0 O _ 0 a O O o o ° o O o o O J L . 17.750 18.250 Radius Squared/2 (cmA2) OO 0 o 18.750 I O o o 0.100 L 0.050 L 0.000 17.250 17.750 18.250 Radius Squared/2 (cmA2) 10.750 161 Figure 2.40. Sedimentation Equilibrium Concentration Distributions of LG2 at a Rotor Speed of 40000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 uM. In the Lower Panels The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. O <' O a o <L> Q 0.000 -5.000 o 9> 17.500 . o <*© o o 0 S 5" o ° o o o O 0 0 o © O ° 0 O o o ° o ---0 u © o o o © 0 o © 18.000 18.500 Radius Squared/2 (cmA2) o o e o <u o CI O o 0.100 L 0.050 U 0.000 17.500 10.000 10.500 Radius Squared/2 (cmA2) 162 Figure 2.41. Sedimentation Equilibrium Concentration Distributions of NG2 at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 50 uM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. C M O < o a o ID Q 2.500 0.000 -2.500 17.250 o o o ° ^ o 0 o ° o o o o 17.750 18.250 Radius Squared/2 (cmA2) 10.750 Figure 2.42. Sedimentation Equilibrium Concentration Distributions of NG2 at a Rotor Speed of 35000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 50 uM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. o < o o p 2.500 L 5; 0.000 -2.500 17 250 o o o o o * o o <fe o 0 o 0 O » <j o o o •30 o •© 17.750 10.250 Radius Squared/2 (cmA2) 18.750 0.500 L a o £ 0.250 o O 0.000 © O / -ofo -© 0 ^ _ 0 I i i i • 17.250 17.750 18.250 Radius Squared/2 (cmA2) 164 18.750 Figure 2.43. Sedimentation Equilibrium Concentration Distributions of NG2 at a Rotor Speed of 40000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 50 | iM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. 2.500 L o < o <^  0.000 a o •« ID Q -2.500 17.250 o o ° p o * 0 0 ° o ° ° • o o o o © o • o ° 17.750 18.250 Radius Squared/2 (cmA2) 18.750 o fi <D O C O 0.625 0.500 L 0.375 L 0.250 L 0.125L 0.000 17.250 17.750 18.250 Radius Squared/2 (cmA2) 165 18.750 Figure 2.44. Sedimentation Equilibrium Concentration Distributions of NG3 at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 (xM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. o i < o o •s P (D Q 5.000 L 0.000 •5.000 17.250 o o • 0 O O ©<S> ° o o ° o O o o o o ° o <> o rt P 6 ' " 0 ' o o o 17.750 10.250 Radius Squared/2 (cmA2) 18.750 Figure 2.45. Sedimentation Equilibrium Concentration Distributions of NG3 at a Rotor Speed of 35000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 50 u.M. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. 1.250 L CM O i < o <j| 0.000 0 o ° o l_ _£ o_ coo ti o Q J o o o o o o -1.250 17.500 18.000 18.500 Radius Squared/2 (cmA2) 1.250 I ti o fi ID O ti O o 0.750 0.250 17.500 18.000 18.500 Radius Squared/2 (cmA2) 167 Figure 2.46. Sedimentation Equilibrium Concentration Distributions of NG3 at a Rotor Speed of 40000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 uJVL In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. 20.250 20.750 21.250 21.750 Radius Squared/2 (cmA2) 168 The partial specific volume sedimentation equilibrium experiments used an An50 Ti rotor loaded with 7 samples and the counterbalance (note the 7 t h sample was made up of water). A double sector cell, equipped with a 12 mm Epon centerpiece and quartz windows, was loaded with 120 \ih of sample and 130 | iL of reference solvent. Each cavitein was lyophilized and then dissolved in either distilled H 2 0 or D2O and loaded into separate sample cells. Data were collected at 20 °C and at a rotor speed of 40000 rpm until equilibrium was established. Samples were equilibrated for 60 hours and single scans 3 hours apart were overlaid to determine that equilibrium had been reached. Scanning parameters included: radial step size of 0.001 cm, step mode, 10 replicate scans, radial scan range between 5.8 cm and 7.3 cm, and U V detection. Duplicate scans were then collected between 80 and 120 hours with 5 hour intervals. The plot of InA versus r2 leads to a straight line from where the partial specific volumes were then calculated using the following equation: V =k-[(d In cldr2)D20 /(Jin cldr2 )H2Q] pD20 - pH20[d ln c I dr2)D20 l(d ln c I dr2)H20] Where k is the ratio of the molecular weight of the protein in the deuterated to that in the non-deuterated solvent, and /? is the solvent density. Since errors associated with v. lead to three times the error in the determination of molecular weight, the partial specific volume sedimentation equilibrium experiments were repeated 3 times (on different days) per sample for reproducibility and error determination (95 % confidence interval). Sedimentation velocity experiments were run using an An50 Ti rotor, and an interference scanner. A double sector cell, equipped with a 12 mm Epon centerpiece and sapphire windows, was loaded with 120 uJ_ of sample and 130 U.L of reference buffer. Data were collected at 20 °C 169 and at rotor speeds of 40000 rpm. Samples were equilibrated for 5 hours at 20 °C to allow for an accurate temperature equilibration. Scanning parameters included: radial step size of 0.001 cm, continuous mode, 10 replicate scans, radial scan range between 5.8 cm and 7.3 cm, and interference detection. Scans were then taken from the time the rotor started to spin until 24 hours later. Data was processed using a PC compatible processing program called SEDFIT, 4 0 and was mainly analyzed in order to determine how many species were present in solution. For all of the caviteins it was clear that only a single species existed. Figures 2.47, 2.48 and 2.49 show the plots of the C(s) distribution versus the sedimentation coefficient, s, for LG2, NG2, and NG3, respectively. 170 Figure 2.47. Sedimentation Velocity Concentration Distributions of LG2 at a Rotor Speed of 40000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C. 0.5 1 1.5 2 2.5 Sedimentation Coefficient (s) 171 Figure 2.48. Sedimentation Velocity Concentration Distributions of NG2 at a Rotor Speed of 40000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C. 10 -2 J Sedimentation Coefficient (s) 172 Figure 2.49. Sedimentation Velocity Concentration Distributions of NG3 at a Rotor Speed of 40000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C. Sedimentation Coefficient (s) 2.4.6 Nuclear Magnetic Resonance (NMR) Experiments The ] H NMR spectra of the caviteins were recorded on a 500 MHz Varian Unity NMR Spectrometer by Okon in the laboratory of Mcintosh at UBC. 173 2.5.6.1 ID J H N M R Spectroscopy The ID ! H NMR spectra were run at 20 °C, and the samples were dissolved in 45 mM sodium phosphate buffer at pH = 7.0 (90 : 10, H2O : D2O) to a final concentration of approximately 1.5 mM. Spectra were processed using a PC "Windows XP" compatible NMR processing program, MestRe-C 2.3. The ID 'H NMR N-H/D exchange spectra were run at 20 °C, and the samples were prepared as follows: -1.5 mM cavitein solutions in a 50 mM acetic acid/acetate buffer at pH = 4.62 were lyophilized to a white solid. D2O was then added to the lyophilized samples in the NMR room to the previous volume before lyophilization of 0.5 mL. The resulting sample in a deuterated acetic acid/acetate buffer at pD = 5.02 was transferred quickly to an NMR tube. The pH was re-checked after the exchange experiments were completed to ensure a correct reading of pD = 5.02, since pH has a dramatic effect on exchange rates. The pD was corrected for isotope effects using the equation:60 pD = pH r e ad + 0.4 where p H r e a d is the reading of the pH electrode. The first scan was acquired 5 minutes after the addition of D2Q and subsequent scans were collected at various time intervals until all of the amide protons had completed exchanged with deuterium. The spectra were analyzed using the same processing program mentioned above. The peak heights were integrated and normalized with the non-exchangeable cavitand proton (H o u t) at 6.1 ppm. (Note: proton assignments were made for three residues per cavitein except for NG2). The first-order rate constants were calculated using the first-order rate equation: 174 In ([H0]7[Ht]) = kobs t where k0bs is the first-order rate constant, t is the time at which the scan was taken, [H0] is the integration of the proton at time zero, and [Ht] is the integration of the same proton at time t. The half-lives, ha, of the amide protons were then calculated using the equation: tin = In 2 / k0bs Protection factors were then calculated using the equation: P = knt / k0bs Where P is the protection factor, k0bS is the experimental first-order rate constant, and km is the first-order rate constant for an "unprotected" amide proton at pH = 4.62 at 20 °G. k[nt can be calculated from the intrinsic half-life, t\a - mX, which is determined using the following equation;463 / , / 2 - i n , = 200 / [10 ( / , / /" 3 ) + 10^-^'j [100057"] Where t\a - intrinsic is the intrinsic half-life for an unprotected proton, and T is the temperature in °C. Errors represent one standard deviation from three rate constant estimates. The ID *H NMR variable temperature spectra were run from 20. to 60 °C, at 10 °C intervals. The caviteins were dissolved in a 45 mM sodium phosphate buffer at a pH = 7.0 to a 175 final concentration of -1.5 mM. Spectra were processed using a PC "Windows XP" compatible NMR processing program, MestRe-C 2.3. 2.4.6.2 2D lH N M R Spectroscopy The 2D 'H NMR COSY, NOESY, and TOCSY spectra were also collected on the 500 MHz Varian Unity NMR Spectrometer. The caviteins were dissolved in a 45mM sodium phosphate buffer at a pH = 7.0 to a final concentration of -1.5 mM. An additional set of 2D ! H NMR COSY, and NOESY spectra were run in 45 mM acetic acid/acetate buffer at pH = 4.62 in order to determine chemical shift assignments of the amino acid residues in the N-H/D exchange experiments (spectra not shown but a similar procedure for signal assignment was undertaken for the spectra acquired in 45 mM acetic acid/acetate buffer as was for LG3 in sodium phosphate buffer describe below). The 2D spectra were processed by Okon using NMRpipe, and a polynomial baseline correction was carried out. No line-broadening functions were used, and the spectra were calibrated using the water signal at 4.78 ppm. Furthermore, Okon helped with one set of amide proton assignments and from there the remaining assignments were made. The 2D NOESY spectrum for NG2 is shown in Figure 2.50 followed by the 2D TOCSY spectra for NG2 in Figures 2.51, 2.52 and 2.53. An explanation of how the assignments were made to the signals at-0.5 ppm and-3.0 ppm is included after Figure 2.53. 176 Figure 2.50. Full 2D 500 MHz 'H NMR NOESY Spectrum of NG2 at -1.5 mM in 10 % D 2 0 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. Figure 2.51. Full 2D 500 MHz 'H NMR TOCSY Spectrum of NG2 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (mixing time = 10 ms) Figure 2.52. Full 2D 500 MHz ] H NMR TOCSY Spectrum of NG2 at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (mixing time = 15 ms) Figure 2.53. Full 2D 500 MHz r H NMR TOCSY Spectrum of NG2 at ~1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (mixing time = 25 ms) - L c o o o - ^ j o c n j ^ c o r o — ^ o • 1H ppm 180 181 A TOCSY spectrum shows the spin-spin propagation via scalar coupling. The advantage of using TOCSY is that information about all of the protons related to a specific proton in the same spin system can be obtained. More specifically, a 2D TOCSY NMR experiment provides information on through bond connectivites moving from C a H to CpH, C Y H, C5H, and even C e H as the mixing time (mt) is increased. Therefore, it was possible to identify the peaks at -0.5 ppm and -3.0 ppm in the ID '.H NMR spectrum of NG2 using TOCSY. The analyses of the NOESY spectra in Figure 2.50 shows C«H of Nle5 at 4.0 ppm with respect to the y-axis followed by CpH, C Y H, and CsH from the norleucine side chain of Nle5 at 1.5, 1.25, 0.75, respectively. Using the first TOCSY spectrum at a mt of 10 ms, one can see the connections between C a H and CpH of Nle5 with the signal at 0.5 ppm with respect to the y-axis being indicative of CpH. Looking at the TOCSY spectrum with a mt of 25 ms, one can see the connections between C«H and C Y H of Nle5 with the signal at -0.5 ppm with respect to the y-axis being indicative of C y H . The final TOCSY spectrum with a mt of 35 ms, one can see the connections between C a H and CsH of Nle5 with the signal at 0.75 ppm with respect to the y-axis being indicative of C5H. The signals of C e H could not be seen (folded over at 10 ppm) leaving it to be the only proton that could produce at signal at-3.0 ppm. 2.4.7 ANS Binding ANS fluorescence measurements were made on a Varian GARY Eclipse Fluorescence Spectrophotometer at the UBC Laboratory of Molecular Biophysics (UBC-LMB) supervised by Rosell located in the Life Sciences Center. The spectrophotometer is equipped with a Xenon 182 Arc lamp. Samples were run at 25 °C using a 1 cm path length with concentrations of 50 p M and 100 pM, and contained 2 p M ANS respectively, in 50 mM sodium phosphate buffer at pH = 7.02. An additional set of data was collected on using 50 mM cavitein solutions with the addition of 0-6.0 M GuHCl in 1.0 M increments, respectively, and no binding of ANS was observed. Reference emission spectra were collected for 95 % ethanol and 100 % HPLC-grade methanol with 2 p,M ANS. Excitation was at 370 nm and emission was recorded between 385 and 600 nm. 183 2.5 References 1. (a) Mezo, A. R.; Sherman, J.C. J. Am. Chem. Soc, 1999, 121, 8983-8994. (b) Causton, A.S.; Sherman, J.C. Bioorg. Med. Chem. 1999, 7, 23-27. 2. (a) Moran, J.R.; Karbach, S.; Cram, D.J. / . Am. Chem. Soc. 1982,104, 5826-5828, (b) Cram, D.J.; Cram, J.M. Container Molecules and Their Guests; Royal Society of Chemistry: Cambridge, 1994. 3. Tucker,'JA.; Knobler, C.B.; Trueblood, K.N. , Cram, D.J. J. Am. Chem. Soc. 1989, 111, 3688-3699. 4. Tuchscherer, G.; Mutter, M . J. Biotechnol. 1995,41, 197-210. 5. According to a crystal structure, the distance between adjacent bromines when Br's replace SH's of 5 are 6.8 A and the distance between opposing bromines is 9.6 A. See: Cram, D.J.; Karbach, S.; Kim, H.-E.; Knobler, C.B.; Maverick, E.F.; Ericson, J.L.; Helgeson, R.C. J. Am. Chem. Soc. 1988,110, 2229-2237. 6. Reddy, B.V.; Blundell, T.L. J. Mol. Biol. 1993, 233, 464-479. 7. Gibb, B.C.; Mezo, A.R.; Causton, A.S.; Fraser, J.R.; Tsai, F.C.S.; Sherman, J.C. Tetrahedron 1995, 51, 8719-8732. 8. A synthesis of tetrathiol 5 with pentyl feet was also reported in 1995: Helgeson, R . C ; Knobler, C.B.; Cram, D.J. / . Chem. Soc. Chem. Commun. 1995, 307-308. 9. DeGrado, W.F.; Wasserman, Z.R.; Lear J.D. Science 1989,243, 622-628. 10. Chakrabartty, A.; Baldwin, R.L. Adv. Prot. Chem. 1995,46,141-176. 11. Seo, E.S. Ph.D. Thesis, University of British Columbia, Vancouver, Canada, 2006. 12. Gibb, B.C., Mezo, A.R.; Sherman, J.C. Tetrahedron Lett. 1995, 36, 7587-7590. 13. Rink, H. Tetrahedron Lett. 1987, 28, 3787-3790. 14. (a)Johnson Jr., W.C. Proteins Struc. Func. Genet. 1990, 7, 205-214. (b) Woody, R.W. Methods Enzymol. 1995, 246", 34-70. (c) Fasman, G.D. Circular Dichroism and the Conformational Analysis of Biomolecules; Plenum Press: New^ork, 1996. 15. CD spectra taken from http://www.proteinchemist.com/cd/cdspec.html (accessed 05/19/06). 16. Chen, Y. -H. ; Yang, J.T.; Chau, K.H. Biochemistry 1974,13, 3350-3359. 184 17. (a) Kohn, W. D.; Kay, C. M . ; Hodges, R. S. Protein Sci. 1995,4, 237-250. (b) Gibney, B. R.; Rabanal, F.; Skalicky, J. J.; Wand, A. J,; Dutton, P L . J.Am. Chem. Soc. 1997,119, 2323-2324. (c) Woody, R. W. Biopolymers 1978,17, 1451-1467. 18. Chakrabartty, A.; Kortemme, T.; Padmanabhan, S.; Baldwin, R.L. Biochemistry 1993, 32, 5560-5565. 19. Kuwijima, K. Proteins Struc. Fund. Genet. 1989, 6, 87-103. 20. Kahn, P.C. Methods Enzymol. 1979, 61, 339-377. 21. Creighton, T.E. Proteins: Structure and Molecular Properties, 2 n d ed.; W.H. Freeman: NewYork, 1993. 22. Pace, C. N. Methods Enzymol. 1986,131, 266-280. 23. (a) Santoro, M . M . ; Bolen, D. W. Biochemistry 1988,27, 8063 - 8068. (b) Santoro M . M . ; Bolen D.W. Biochemistry 1992,31,4901-4907. 24. Mutter, M . ; Tuchscherer, G.G.; Miller, C , Altmann, K.H. ; Carey, R.L; Wyss, D.F.; Labhart, A . M . ; Rivier, J.E. 7. Am. Chem. Soc. 1992,114, 1463-1470. 25. Myers, J.K.; Pace, C.N.; Sholtz, J.M. Protein Sci. 1995,4, 2138-2148. 26. Ahmad, F.; Bigelow, C.C. Biopolymers 1996, 25, 1623-1633. 27. (a) Harding, S.E.; Rowe, A.J.; Horton, J.C. Analytical Ultracentrifugation in Biochemistry arid Polymer Science; Royal Society of Chemistry: Oxford, 1992. (b) Hansen, J . C ; Lebowitz, J.; Demeler, B. Biochemistry 1994,33, 13155-13163. 28. Laue, T .M.; Stafford IH, W.F. Annu. Rev. Biophys. Biomol. Struct. 1999, 28, 75-100. 29. (a) Casassa, E.F.; Eisenberg, H. Adv. Protein Chem. 1964, 19, 287-395. (b) Fujita H. Foundations of Ultracentrifugal Analysis; Wiley: New York, 1993. (c) Goldberg, R.J. J. Phys. Chem. 1953, 57, 194-202. 30. (a) Yphantis D A . J. Phys. Chem. 1956, 60, 623-635. (b) Yphantis D.A. Ann. NY Acad. Sci. 1960, 88, 586-601. (c) Yphantis D.A. Biochemistry 1964, 3, 297-317. 31. Lamm, O. Die Differentialgleichung der Ultrazentrifugierung. Ark. Mat. Astr. Fys. 1929, 2IB, 1-4. [in germanj 32. McRorie, D.K.; Voelkeri P.J. Self-Associating Systems in the Analytical Ultracentrifuge, Beckman Instruments: Fullerton CA, 1993. 33. Johnson, M.L.; Correia, J.J.; Yphantis, D.A.; Halvorson, H.R. Biophys. J. 1981, 36, 575-588. 185 34. Perkins, S J . Eur. J. Biochem. 1986,157, 169-180. 35. Durchschlag, H. Specific volumes of biological molecules and some other molecules of biological interest. Thermodynamic Data for Biochemistry and Biotechnology; Springer-Verlag: Berlin-Heidelberg, 1986. 36. Edelstein, S.J.; Schachman, H.K. J. Biol. Chem. 1967,242, 306-311. 37. Cohn, E.J.; Edsall, J.T. Amino Acids, Peptides and Proteins as Ions and Dipolar Ions; Reinhold Publishing Corporation: New York, 1943. 38 Lebowitz, J.; Lewis, M.S.; Schuck P. Prot. Sci. 2002,11, 2067-2069. 39. (a) Behlke, J.; Ristau O. Biohys. J. 1997, 72, 428-434. (b) Holloday, L .A. Biophys. Chem. 1979,10, 187-190. (c) Schuck, P. Prog. Colloid Polym. Sci. 1994, 94, 1-13. 40. (a) Schuck P. Biophys. J. 2000, 78, 1606-1619. .(b) Schuck P. Aria. Biochemistry 2003, 320,104-124. 41. Wuthrich, K. NMR of Proteins and Nucleic Acids; Wiley: New York, 1986. 42. MacDonald, C.C.; Phillips, W.D. J. Am. Chem. Soc. 1967, 89, 6332-6341. 43. (a) Cavanagh, J.; Fairbrother, W.J.; Palmer III, A .G. ; Skelton, N.J. Protein NMR Spectroscopy: Principles and Practice; Academic Press: San Diego, 1996. (b) Roy, S.; Helmer, K.J.; Hecht, M.H. Folding & Design 1997, 2, 89-92. 44. Ptitsyn, O.B. Adv. Prot. Chem. 1995,47, 83-229. 45. (a) Wiltscheck, R.; Kammerer, R.A.; Dames, S.A.; Schulthess, T.; Blommers, M.J.; Engel, J.; Alexandrescu, A.T. Prot. Sci. 1997, 6, 1734-1745. (b) Greenfield, N.J.; Montelione, G.T.; Farid, R.S.; Hitchcock-DeGregori, SB. Biochemistry 1998, 37, 7834-7843. 46. (a) Englander, S.W.; Downer, N.W.; Teitelbaum, H. Annu. Rev. Biochem. 1972, 41, 903-905. (b) Raschke, T.M.; Marqusee, S. Curr. Opin. Biotech. 1998, 9, 80-86. (c) Englander, S.W.; Mayne, L.; Bai, Y.; Sosnick, T.R. Protein Sci. 1997, 6, 1101-1109. (d) Jeng, M.-F.; Englander, S.W.; Elove, G.A.; Wand, A.J.; Roder, H. Biochemistry'1990, 29, 10433-10437. (e) Hughson, F.M.; Wright, P.E.; Baldwin, R.L. Science 1990,249, 1544-1548. 47. Bai, Y.; Milne, J.S.; Mayne, L.; Englander, S.W- Proteins Struc. Funct. Genet. 1993,17, 75-86. 48. Roy, S.; Ratnaswamy, G.; Boice, J.A.; Fairmah, R.; McLendon, G.; Hecht, M.H. J. Am. Chem. Soc. 1997, 779,5302-5306. 49. Freeman, J.; Wallhorn, D.; Sherman, J.C. "Four-Helix Bundle Cavitein Reveals Middle Leucine as Linchpin" (in preparation) 186 50. (a) Schwalbe, H. Chem. Bio. Chem. 2003, 4, 135-142. (b) Grzesiek, B.A. Acc. Chem. Res. 1993,26, 131-137. 51. Kuwajima,K. Proteins Struc.Funct. Genet. 1989, 6~, 87-103. 52. Semisotnov, G.V.; Rodionova, N.A.; Razgulyaev, O.I.;. Uversky, V.N. ; Gripas, A.F.; Gilmanshin, R.I. Biopolymers 1991,31, 119-128. 53. (a) Roy, S.; Ratnaswamy, G.; Boice, J.A.; Fairman, R.; McLendon, G.; Hecht, M.H. J. Am. Chem. Soc. 1997, 119, 5302-5306. (b) Raleigh, D.P.; Betz, S.F.; DeGrado, W.F. J. Am. Chem. Soc. 1995,117, 7558-7559. 54. Sherman, M.A.; Beechem, J.M.; Mas, M.T. Biochemistry 1995,34, 13934-13942. 55. Scott, W.R.P.; Seo, E.S.; Huttunen, H.; Wallhorn, D.; Sherman, J . C ; Straus, S.K. Proteins: Struc. Func. Bioinfor. 2006 (in press). 56. Bradford, M . M . Anal. Biochem. 1976, 72, 248-254. 57. Chen, G . C ; Yang, J.T. Ana. Lett. 1977, 10, 1195-1207. 58. Haschemeyer, R.H.; Bowers, W.F. Biochemistry 1970, 9,435-445. 59. Dekker, C ; Agianian, B.; Weik, M . ; Zaccai, G.; Kroon, J.; Gros, P.; de Kruijff, B. Biophys. J. 2001, 81, 455-462. 60. Glasoe, P.K.; Long, F.A. / . Phys. Chem. 1960, 64, 188-190. 187 CHAPTER 3: Design, Synthesis and Characterization of Hetero-TASPs 3.0 Introduction Chapter 2 focused on designing native-like TASPs, and on investigating native-like structure. The TASPs synthesized in our lab to date, including those described in Chapter 2, have resulted from the simultaneous ligation of four peptides strands to the four thiol moieties on the cavitand template, and have thus been limited to having only one type of peptide sequence attached within one bundle (i.e., four identical helices).1 Having the ability to synthesize hetero-TASPs, TASPs having different peptide sequences within one bundle, would open an opportunity to create various de novo proteins, including an anti-parallel four-helix bundle, and perhaps even more native-like synthetic proteins, one of the goals of this project. This Chapter will describe the design, synthesis and characterization of several hetero-TASPs. Section 3.1 of this Chapter will outline the rationale for synthesizing hetero-TASPs. Section 3.2 will be broken down into sub-sections describing the peptide synthesis (Section 3.2.1) and the two different synthetic approaches used to synthesize a variety of hetero-TASPs (Sections 3^ 2.2 and 3.2.3, respectively). Furthermore, a discussion of the characterization of the lg3-subsituted cavitein intermediates (Section 3.2.4) and hetero-TASPs (Section 3.2.5) will follow. Section 3.3 will summarize the experimental findings for this Chapter, in addition to highlighting some conclusions that may be drawn from the experimental results. * " A version of this Chapter will be submitted for publication. Huttunen, H . and Sherman, J.C.The Design, Synthesis and Characterization of De Novo Hetero-TASPs." 188 3.1 Rationale for Synthesizing Hetero-TASPs As was explained in Chapter 1, the native-like structure of a protein ultimately depends on the packing of the hydrophobic core. The hydrophobic cores of native-like proteins commonly have their side chains packing like "knobs-into-holes".2 Briefly, this hydrophobic packing arrangement is composed of the side chain (i.e. knob) of one amino acid residue filling a hole formed by the adjacent amino acid residues. Although our previous caviteins have possessed a high degree of native-like character, it is uncertain whether the side chains of the hydrophobic core were packing in a specifically defined manner, and even more so whether they were able to pack like "knobs-into-holes". Synthesizing a hetero-TASP having peptides with linkers of differing lengths (i.e. between the cavitand template and the peptide sequences), could provide a means to having the side chains of adjacent helices "out of register". If our previously designed caviteins were not completely native-like due to an overpacked hydrophobic core, an "out of register" arrangement of the helices could enable the side chains to pack in the preferred "knobs-into-holes" conformation. An initial goal for this project was simply the synthetic challenge of creating the first hetero-TASP. The most interesting hetero-TASPs would be those having different peptide sequences on opposing sites (i.e. a,c substitution) on the cavitand template. For comparison's sake, the a,b disubstituted caviteins would also be synthesized. It was hypothesized that the packing of the side chains within the hydrophobic core would be more efficient in the a,c derivatives, and hence would display more native-like characteristics than the a,b disubstituted hetero-TASPs. The general design for all of the hetero-TASPs synthesized was to have one type of peptide sequence at the a,c positions on the cavitand template, and a different peptide sequence at the remaining two positions to afford the four-helix bundle. The first family of hetero-TASPs 189 was designed with peptides having different linker lengths, the second family including anti-parallel four-helix bundles, and the third family included two peptides with the same linker lengths but differing hydrophobic residues in the peptide sequences. For example, consider a hetero-TASP made up of two different peptides sequences (e.g. D and E), where peptides D had shorter linkers than peptides E (see Figure 3.1). This could result in the nonpolar residues of peptides D and peptides E lying in different planes (i.e. "out of register") within the hydrophobic core by design, and perhaps interacting like "knobs-into-holes". Figure 3.1. Schematic Representation Highlighting the a,c and a,b Disubstituted Hetero-TASP Intermediates, (note only two peptides are shown for clarity) Peptide E a,b a,c The anti-parallel hetero-TASPs which will be synthesized for the first time in our laboratory also presented the possibility of having the side chains of adjacent helices "out of register" as depicted in Figure 3.1. The synthesis of an anti-parallel hetero-TASP would also 190 provide an opportunity to study the stability of an anti-parallel structure, which was predicted to be highly stable due to the complementary helix macrodipoles. Other examples of anti-parallel de novo proteins exist in the literature, and it also is a frequent topology found in natural proteins.4 3.2 Results and Discussion This Section will start by describing the peptide designs and syntheses followed by the two different synthetic approaches which were used to create the hetero-TASPs 16-25. 3.2.1 Peptide Synthesis Peptides 14 and 15 (see Table 3.1 for peptide sequences) were designed following the principles outlined in Chapter 2 (Section 2.2.2), arid were synthesized using standard Fmoc techniques on an automated Applied Biosystems peptide synthesizer following literature procedures.5 Slight modifications to the synthesis of peptide 14 which differs from the peptides synthesized in Chapter 2 includes the incorporation of a C-terminal cysteine residue, and the N-terminus of peptide 14 was acetylated since future attachment of the peptide to the cavitand template would take place from the C-terminal cysteine residue via a disulfide bond to cavitand template. Peptide 14 was then cleaved from the Rink resin and of protecting groups in one step including the S-trityl side chain protecting group of the cysteine residue with TFA (95 %), H2O (2.5 %) and 1,2 ethanedithiol (2.5 %). The C-terminal cysteine was then activated using 2,2'-191 dipyridyl disulfide (DPDS) to afford the activated form of the peptide ready for incorporation into the caviteins via its C-terminus. Table 3.1. Complete Sequences from N- to C-termini Using One Letter Abbreviated Amino Acids Including Modified Termini for Peptides 6,7,14, and 15. Peptide Peptide Peptide Sequence Number Name 6a lg2 C I C H 2 C O - N H - [ G G - E E L L K K L E E L L K K G ] - C O - N H 2 T lg3 C I C H 2 C O - N H - [ G G G - E E L L K K L E E L L K K G ] - C O - N H 2 14b lg2c C H 3 C O - N H - [ G E E L L K K L E E L L K K G G C ] -Spy 15 ag3 C I C H 2 C O - N H - [ G G G - E E A A K K A E E A A K K G ] C O - N H 2 a peptide sequences repeated from Chapter 2. b lg2c will be attached to the cavitand via a C-terminal cysteine residue; Spy is the S-pyridyl group. 3.2.2 General Hetero-TASP Synthesis via Approach One The first approach developed to synthesize a hetero-TASP involved using suitably experimentally determined peptide equivalents via Scheme 3.1. 192 Scheme 3.1. Schematic Representation of Approach One Outlining the General Synthesis for the Desired a,c Disubstituted Cavitein Intermediate, and a,c Hetero-TASP. 1. 2.5 eq. Peptide D DMF/DIPEA 1 2. RP-HPLC Purification Note: LG3/LG2 hetero-TASPs = cavitand-[(S-CH2CO-NH-[Ig3peptide]-CO-NH2)2 (S- CH 2CO-NH-[lg2 peptide]-CO-NH2)2] LG3/LG2C hetero-TASPs = cavitand-[(S-CH2CO-NH-[lg3peptide]-CO-NH2)2 (S-[lg2c peptide]-NH-COCH3)2] LG3/AG3 hetero-TASPs = cavitand-[(S-CH2CO-NH-[lg3peptide]-CO-NH2)2 (S- CH 2CO-NH-[ag3 peptide]-CO-NH2)2] 1. xs. Peptide E DMF/DIPEA 2. RP-HPLC Purification The first step entailed reacting 2.5 equivalents of peptide D with cavitand template 5 in the presence of excess DIPEA base, and D M F solvent. The subsequent peptide mixture was purified using RP-HPLC. The isolated products included two different disubstituted 193 intermediates where the two peptides attached were either at the a,b (not shown in Scheme 3.1) or a,c positions on the template, respectively, in addition to mono, tri, and tetrasubstituted products (not shown in Scheme 3.1). The masses of the products were confirmed by MALDI-mass spectrometry.6 The product distribution contained mainly the a,b disubstituted product (57 %) followed by the a,c disubstituted product (24 %) and minimal amounts of the Other three possible products. The purified and separated mono, di, trisubstituted products were then individually subjected to peptide E in the presence of DJPEA baseband D M F solvent, to yield the corresponding four-helix hetero-TASPs. The overall yield of the desired disubstituted intermediates was low, and therefore an alternate synthetic route was explored (approach two) in an attempt to increase the yields of the precursors (disubstituted intermediates) for the formation of the hetero-TASPs. 3.2.3 General Hetero-TASP Synthesis via Approach Two The yields of the disubstituted intermediates obtained via the first approach were poor, and thus another route was explored using protecting group methodology in an attempt to improve the product yield. A number of different potential thiol protecting groups were considered, but because of limited literature procedures the only protecting group used to protect the cavitand thiol moieties was 2-(4-nitrophenyl)ethyl (npe) group.7 Scheme 3.2 provides a general synthesis of an a,c hetero-TASP. 194 Scheme 3.2. Schematic Representation of Approach Two Outlining the General Synthesis for an a,c Hetero-TASP where PG is the 2-(4-nitrophenyl)ethyl Group. 1. xs. Peptide D DMF/DfPEA 2. RP-HPLC Purification • The first step entailed reacting 2.5 equivalents of 2-(4-nitrophenyl)ethyl bromide with cavitand 5 in the presence of excess DIPEA base, and D M F solvent. The subsequent protecting 195 group-template mixture was not purified but rather directly reacted with excess DIPEA base and excess peptide D. The water-soluble product mixture was then filtered, and purified by RP-HPLC. Four different products were isolated. Two disubstituted products, where the two protecting groups attached were either at the a,b (not shown in Scheme 3.2) or at the a,c positions on the template, respectively, and peptide D's in the remaining two positions. In addition a monoprotected product with three peptide D's, and a triprotected product with one peptide D were also isolated (not shown in Scheme 3.2). The identities of the products were confirmed by MALDI-MS. 6 The product distribution contained mainly the a,b disubstituted product followed by the a,c disubstituted product, and minimal amounts of the other two possible products. The purified and separated mono, di, triprotected products were then individually subjected to NaOMe in MeOH to cleave the protecting groups, which is thought to follow an E 2 mechanism.8 The cleaved products were then reacted with peptide E in the presence of DIPEA base, and D M F solvent, to yield the corresponding four-helix hetero-TASPs. Approach one and two were both investigated for determining the most convenient method for the synthesis of a hetero-TASP. Unfortunately, both approaches resulted in rather low yields, but because of synthetic simplicity approach one was used to synthesize the hetero-TASPs (caviteins 16-25) outlined in Table 3.2. 196 Table 3.2. Names and Sequences for Caviteins 10,11,16-25. Cavitein Cavitein Name Sequence Number 10a LG2 5-(S-CH 2 CO-NH-[GG-EELLKKLEELLKKG]-CO-NH2)4 l l a LG3 5-(S-CH2CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 )4 16 2LG3«2LG2_ ab 5[-(S-CH2CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 )2 - (S-CH 2 CO-NH-[GG-EELLKKLEELLKKG]-CO-NH 2 ) 2 ] 17 2LG3*2LG2_ ac 5[ - (S-CH 2 CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 ) 2 -(S-CH2CO-NH-[GG-EELLKKLEELLKKG]-CO-NH 2 )2] 18b LG2C 5-(S-[CGG-KKLLEELKKLLEEG]-NH-COCH3) 4 19 2LG3-2LG2C _ab 5[ - (S-CH 2 CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 ) 2 - (S- [CGG-KKLLEELKKLLEEG]-NH-COCH 3 ) 2 ] 20 2LG3-2LG2C _ac 5[- (S-CH 2 CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 ) 2 - (S-LCGG-KKLLEELKKLLEEG]-NH-COCH 3 ) 2 ] 21c AG3 5-(S-CH 2CO-NH [GGG-EEAAKKAEEAAKKG]-CO-NH 2 )4 22 3LG3-1AG3 5[- (S-CH 2 CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 ) 3 -(S CH 2 CO-NH-[GGG-EEAAKKAEEAAKKG]-CO-NH 2 ) i ] 23 2LG3»2AG3_ ab 5[ - (S-CH 2 CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 ) 2 - (S-CH 2 CO-NH-[GGG-EEAAKKAEEAAKKG]-CO-NH 2 ) 2 ] 24 2LG3»2AG3_ ac 5[- (S-CH 2 CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 ) 2 - (S-CH 2 CO-NH-[GGG-EEAAKKAEEAAKKG]-CO-NH 2 ) 2 ] 25 1LG3»3AG3 5[-(S-CH 2 CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH 2 ) i - (S-CH 2 CO-NH-[GGG-EEAAKKAEEAAKKG]-CO-NH 2 ) 3 ] a cavitein sequences from Chapter 2. b reference cavitein for anti-parallel caviteins 19, and 20 with the lg2c peptides are attached via their C-termini. 0 reference cavitein for caviteins 22-25 having alanine residues in the peptide sequences. 197 Caviteins 16-25 were synthesized with high levels of purity, and their corresponding masses were confirmed by MALDI-MS. 6 The caviteins were further characterized by CD and NMR spectroscopy, A U C , and the binding of a hydrophobic dye monitored by fluorescence spectroscopy. 3.2.4 Characterization of the Ig3-Substituted Cavitein Variants In the syntheses of the disubstituted hetero-TASPs outlined in Table 3.2 two interesting disubstituted intermediates were observed along the way. The disubstituted caviteins having the lg3 peptide sequence at the a,b and a,c positions were isolated and characterized using MALDI-MS, CD and NMR spectroscopy, and A U C . It was believed that the lg3 disubstituted intermediates would possibly aggregate into anti-parallel dimers, which could be diagnosed by A U C . In some cases further comparisons were made to the mono and trisubstituted lg3 variants. 3.2.4.1 Far-UV CD Spectroscopy As was explained in Chapter 2, far-UV (190-250 nm) CD spectroscopy is regularly used to quantify the extent of secondary structure present in peptides and proteins.9 For all of the caviteins depicted in this thesis it was expected that CD curves characteristic of a-helical structure would be observed. Figure 3.2 compares the CD curves observed for the lg3-subsituted cavitein variants at concentrations of-40 p:M. 198 Figure 3.2. Far-UV CD Spectra for Ig3-Substituted Cavitein Variants at -40 p:M in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. .25000 J Wavelength (nm) It is clear from the CD curves for the lg3-subsituted cavitein variants that they were all a-helical by the observation of the characteristic minima at 222 and 208 nm, and the maximum at 195 nm. Since the cavitand template affects the absorbance of our caviteins around 222 nm, the experimental molar ellipticities at 222 nm should not be quantitatively analyzed, however, the CD data does suggest that a-helicity does not change (per mole) when more lg3 peptide helices are added to the cavitand template. The mono and disubsituted variants are slightly more helical than the tri and tetrasubstituted caviteins. This small increase in helicity for the mono and disubstituted variants could be the result of slight aggregation, which would be consistent with the idea that the cavitein variants should be susceptible to aggregation. 199 CD spectra for each of the lg3 cavitein variants were also obtained at ~4 jiM to evaluate whether concentration has an effect on the a-helicity of the protein. For all of the caviteins, the CD spectra at concentrations of ~4 and ~40 | i M were within experimental error, respectively, and therefore the ~4 | i M curves for each of the caviteins are not shown in Figure 3.2. It was expected that the lg3 caviteins variants (excluding the 41g3 regular cavitein) would show concentration dependent CD spectra since it was expected that these caviteins would aggregate. Since there was no observed increase in helicity with an increase in concentration this supports that the proteins were either monomeric or a higher order mixture that does not change with an increase in concentration. Furthermore, if the proteins were already highly helical at low concentrations, an increase in helicity would not necessarily be observed at higher concentrations. For these reasons, concentration experiments by monitoring the molar ellipticity at 222 nm in the presence of a denaturant were performed to more thoroughly analyze the oligomeric states of the caviteins (Section 3.2.4.3.1). Analytical ultracentrifugation studies were also performed to assess the oligomeric states of proteins in solution (Section 3.2.4.3.2). 3.2.4.2 Near-UV C D Spectroscopy In addition to determining the extent of a-helical secondary structure by interpreting the far-UV region, information about protein tertiary structure can be gathered by examining the near-UV (250-350 nm) spectral region.9"5 The presence of aromatic residues or disulfide bonds are necessary in order to detect an absorption in the near-UV region. The lg3 cavitein variants do not have aromatic residues or disulfide bonds, but the arenes of the cavitand template provide 200 a suitable chromophore for detection of a near-UV signal. The near-UV CD spectra for the lg3 cavitein variants are graphed in Figure 3.3. Figure 3.3. Near-UV CD Spectra for Ig3-Substituted Cavitein Variants at -40 pM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. 1000 -1500 -* Wavelength (nm) It is observed that increasing the number of lg3 peptides from one to four on the cavitand template resulted in only a slight increase in the near-UV CD signal. This was to some extent surprising since it would be assumed that a four-helix bundle would have a largely more defined tertiary structure and hence a much more pronounced near-UV CD signal. 201 For comparison, the signals observed in Figure 3.3 are weaker than that observed for LG2, which shows a relatively strong near-UV signal. Furthermore, the near-UV CD curves of all the lg3 cavitein variants display a positive absorption at ~250 nm, and a negative absorption at -280 nm, which is consistent with the LG3 four-helix bundle. The signs for the absorptions for the cavitein variants are opposite to those for LG2, which supports the idea that the supercoiling of the cavitein variants is in reverse direction to that of LG2. These observations are consistent that it is the linker between the template and the peptide helices which influences the tertiary structure of the caviteins, even more so that the residues within the peptide sequence or how many helices are attached to the cavitand template. 3.2.4.3 Oligomeric States The oligomeric states of the lg3 cavitein variants were studied by monitoring the concentration dependence of their unfolding in the presence of guanidine hydrochloride (GuHCl) and by A U C . 3.2.4.3.1 GuHCl Denaturation Experiments As was mentioned in Section 3.2.4. T the lg3 cavitein variants produced concentration independent CD spectra. However, this is not sufficient enough data to conclude the existence of monomers in solution. Therefore, the stabilities of the caviteins in the presence of the denaturing salt, GuHCl, were determined at different concentrations. 202 The stabilities of oligomeric proteins are expected to vary with concentration, whereas monomeric proteins should give rise to identical denaturation curves. Figure 3.4 displays the unfolding curves of the lg3 cavitein variants at concentrations of -40 | iM, monitored at 222 nm in the presence of 0-8.0 M GuHCl. The denaturation experiments were also performed at concentrations of -4 (iM for each of the cavitein variants, and surprisingly the unfolding curves were identical at both concentrations. (The unfolding curves obtained for the -4 uM concentrations are not included on Figure 3.4 for clarity's sake.) Figure 3.4. Effect of GuHCl on the Helicity ([(^ 222) of the Ig3-Substituted Cavitein Variants at -40 | i M in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. •—LG3/4pep » -LG3/3pep _—LG3/2pep_ab LG3/2pep_ac LG3/1pep GuHCl (M) 203 Upon examination of the unfolding curves in Figure 3.4, it is apparent that LG3/4pep displays a highly cooperative unfolding curve. On the other hand, non-cooperative unfolding transitions are observed for the other lg3 cavitein variants, suggesting that they existed only as molten globule-like proteins before any denaturant was added. It is also apparent upon inspection of the unfolding curves that there is a systematic decrease in stability as peptide helices are removed from the cavitand template. Comparing the LG3/4pep cavitein to the LG3/3pep cavitein, a significant loss in structural stability is encountered with the removal of the first peptide helix, even after accounting for the expected -25 % loss in helicity. This was not necessarily surprising since the cavitand template and the lg3 peptide sequence were designed to accommodate a four-helix bundle structure. Subsequent removal of the lg3 peptide sequences even further destabilized the overall protein structure. Due to the lack of pre- and post-transitional baselines on the GuHCl denaturation curves for lg3 cavitein variants no A G ° H 2 O values could be evaluated, except for the four-helix bundle (LG3) which was evaluated in Chapter 2. As was observed with the concentration independent CD spectra, the concentration independent unfolding curves (data not shown) for the lg3 cavitein variants support monomeric species in solution. However, it is common that protein aggregation would be broken up in the presence of guanidine hydrochloride, and therefore, concentration independent denaturation curves alone cannot be used to confirm the presence of a monomeric protein. 204 3.2.4.3.2 AUC Sedimentation Equilibrium Experiments A U C was used to assess the oligomeric states of the lg3 cavitein variants in solution. The lg3 cavitein Variants were prepared similar to caviteins in Chapter 2, and were analyzed at concentrations of 10, 50, and 80 jiM and at rotor speeds of 27000, 35000, 40000 rpm. The sedimentation equilibrium data for the Tg3 cavitein variants were analyzed by NONLIN, 1 0 and the initial fits for each of the caviteins were to that of a monomeric species in solution. The monomer fits were very poor for all of the cavitein variants excluding LG3/4pep. The sedimentation data for each of the caviteins was subsequently fit to a monomer-dimer, dimer, monomer-trimer, trimer, monomer-tetramer, tetramer etc., and none of the fits were acceptable. Figure 3.5 shows an example of the best fit of the sedimentation equilibrium data for the LG3/2pep_ab cavitein variant. The remaining plots for LG3/2pep_ab and those for LG3/2pep_ac are not shown since they look very similar to the plot displayed in Figure 3.5. 205 Figure 3.5. Sedimentation Equilibrium Concentration Distribution of LG3/2pep_ab at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C, and at a Concentration of 10 (iM. In the Lower Panel The Solid Lines Represents Theoretical Fits to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. C M o < o ti o Pi 2.500 U 0.000 -2.500 17.250 <*>° o 8 o ° l & o 0 o p o 0 o O ° o ° o . . O O o <> o 0 o 0 % o © o s o o v 0 °o a 5 ° o O oO 17.750 18.250 Radius Squared/2 (cmA2) 18.750 8.500 I ti o •s fi <D O ti o U 0.000 17.250 17.750 18.250 Radius Squared/2 (cmA2) 18.750 206 It was concluded that all of the cavitein variants excluding that of the four-helix bundle were aggregating in a non-specific manner; This data is consistent with the ! H NMR spectra of the lg3 cavitein variants to follow in Section 3.2.4.4, but is inconsistent with the CD and GuHCl experiments previously mentioned, which suggested monomers in solution for all of the cavitein variants. Taking into account all of the experimental data, a possible explanation for the behavior of the cavitein variants is that they are only slightly aggregating in solution, or that the aggregation is not accompanied by an increase in a-helicity. As was mentioned previously, GuHCl could break up an aggregate, and thus the denaturation data should not be taken as conclusive on its own. 3.2.4.4 One-Dimensional (ID) X H N M R Spectroscopy One-dimensional ' H N M R spectroscopy is routinely used to differentiate between protein native-like structures, molten globule-like structures, and also protein aggregation. Native-like proteins display a high degree of chemical shift dispersion especially in the amide region.11 On the other hand, molten globule-like structures or aggregating proteins portray broad and less disperse signals, due to the looser packing arrangement of the hydrophobic core.12 The ! H NMR spectra for the disubstituted (LG3/2pep_ab, LG3/2pep_ac) cavitein variants are shown in Figure 3.6. 207 Figure 3.6. Full 500 MHz 'H NMR Spectra Overlay of (a) 2 mM LG3/2pep_ab (b) 1 mM LG3/2pep_ab in 200 mM KG1 salt (c) 0.5 mM LG3/2pep_ab (d) 2 mM LG3/2pep_ac, in 10 % D 2 0,45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. 14 12 ppm From the ! H NMR spectra of the disubstituted caviteins it is apparent that they are exhibiting either molten-globule type properties, or non-specific aggregation demonstrated by the broad and poorly defined signals in the amide spectral region. The possibility of the disubstituted caviteins existing as some type of higher-order aggregate would be consistent with 208 the irresolvable A U C sedimentation data. An additional *H NMR spectrum for the LG3/2pep_ab cavitein was acquired in the presence of 200 mM KC1 salt, in order to determine whether the aggregation was resulting from electrostatic interactions. No significant change in the NMR spectrum for this a,b disubstituted cavitein was observed in the presence of the KC1 salt, or even at a lower concentration, suggesting that an undefined aggregate was probable. The LG3/2pep_ac cavitein also exhibited poorly defined and broad signals in the amide spectral region, also suggesting that it was associating in an unspecified manner. The studies on the disubstituted cavitein variants were indicative toward the formation of non-specifically aggregating species in solution. The formation of specific anti-parallel dimers was not observed, however, some information on the stability of a designed four-helix bundle as compared to a three- or two-helix bundle with equivalent peptide sequences was determined. It was observed that a large decrease in thermodynamic stability results from the removal of one or more peptide helices from the designed four-helix bundle structure. The lg3 peptide sequence was designed to accommodate a four-helix bundle structure, and therefore, it was not surprising that the mono, di or trisubstituted caviteins resulted in molten globule-like tertiary structures compared to LG3/4pep. In this way, the suitability of the designed lg3 peptide sequence for a four-helix bundle was confirmed by forcing the lg3 peptide into smaller bundle structures. Furthermore, the lg3 mono, di, and trisubstituted intermediates provided the opportunity to synthesize hetero-TASPs. The lg3 cavitein intermediates were further reacted with different peptide sequences to produce numerous hetero-TASPs with varying goals in mind. 209 3.2.5 Characterization of the Hetero-TASPs Approach one was used to synthesize the hetero-TASPs listed in Table 3.2. A complete table of the caviteins with their corresponding masses confirmed by MALDI-mass spectrometry, along with the percent yields is given in the Section 3.4.3. Three hetero-TASP families were designed and named accordingly: LG3/LG2, LG3/LG2C, and LG3/AG3. The LG3/LG2 family was designed in order to model a potential "knobs-into-holes" packing within the hydrophobic core by using two peptides sequences with differing linker lengths. As was mentioned in the Section 3.1, such a design would potentially enable the side chains of adjacent peptide helices to be "out of register''^  which may in fact increase the efficacy of the hydrophobic packing within the four-helix bundle core. If our previously designed caviteins were unable to pack efficiently due to an Overpacked hydrophobic core, for example by side chains clashing in the same plane, then creating a 2LG3»2LG2_ac hetero-TASP may overcome the packing limitations of the previous design. The LG3/LG2C hetero-TASPs represent the first anti-parallel caviteins to be synthesized, which was one main goal in itself. Furthermore, perhaps the LG3/LG2C caviteins could accommodate a stable "knobs-into-holes" packing arrangement, and result in highly native-like caviteins. The design of native-like caviteins is a major challenge, and only few examples of native-like de novo proteins exist. An anti-parallel design further permitted the evaluation of whether an increase in structural stability could be diagnosed when the peptide helices are oriented with opposing helix macrodipoles with respect to one another. Lastly, the LG3/AG3 family was created to examine the effects of decreasing the hydrophobicity of an entire peptide helix, as opposed to simply altering the hydrophobicity of individual residues within a peptide helix. Additional research focusing on the effects of the 210 alanine residue on the stabilities and structures of our caviteins has been performed in our group. Causton synthesized an all alanine peptide sequence linked via disulfide bonds to a benzylthiol cavitand template.13 He found that his all alanine-based cavitein exhibited highly molten globule-like characteristics. Wallhorn and Freeman have investigated the effects of single leucine for alanine substitutions on the overall cavitein structure and stability.14 They concluded that it was the middle leucine residue of the all-leucine based peptide sequence (lg2) which acted as a linchpin, crucial for holding the secondary and tertiary structure of the cavitein intact. If our previously designed leucine-based caviteins were not entirely native-like due to an overpacked hydrophobic core, then designing a hetero-TASP with peptide helices having residues with much smaller side chains (e.g. alanine) may allow the four-helix bundle structure to pack more efficiently. It was believed that a four-helix bundle cavitein composed of two ag3 peptides and two lg3 peptides could result in a native-like cavitein. The creation of native-like de novo proteins is one of the major goals in the area of de novo protein design, and the LG3/AG3 family was hoped to be provide a promising candidate. Furthermore, without a crystal structure of our caviteins, it is unclear as to the relative size of the helices compared to the cavitand template. If the interhelical distances of our helices are much larger than the diameter of the template, synthesizing an all alanine-based cavitein or an Ala/Leu derivative again may be more suitable for obtaining a native-like structure. 211 3.2.5.1 Far-UV C D Spectroscopy Far-UV CD spectroscopy was used to evaluate the secondary structures of the hetero-TASPs. The far-UV CD spectra for the LG3/LG2 variants, LG3/LG2C (anti-parallel) variants, and the LG3/AG3 variants at concentrations of -40 | i M are shown in Figures 3.7, 3.8 and 3.9, respectively. CD spectra for each of the caviteins were also obtained at -4 jiiM to evaluate whether concentration has an effect on the a-helicity of the protein. For all of the caviteins, the CD spectra at concentrations of -4 and -40 \iM were within experimental error, respectively, and therefore the -4 p M curves for each of the caviteins are not shown. Figure 3.7. Far-UV CD Spectra for the LG3/LG2 Substituted Caviteins at -40 uJVI in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. 50000 -_—(4)LG3 —»-(4)LG2 * - 2LG3_2LG2_ab O 30000 A *-2LG3_2LG2_ac CM cn .g 10000 -re 5 JZ -10000 A -30000 Wavelength (nm) 212 From examining the CD curves it is clear that all of the caviteins are a-helical in structure. Furthermore, the a-helicities of the 2LG3»2LG2_ab and the 2LG3»2LG2_ac caviteins seem to be an average of the helicities observed for their parent LG3 and LG2 derivatives. Furthermore, there is little difference in the helicities between the a,b and a,c derivatives with their CD spectra being almost identical. In this case, designing a hetero-TASP, which incorporated both the lg3 and lg2 peptide sequences, did not result in an increase in the a-helicity of the four-helix bundle protein. Table 3.3. Molar Ellipticity at 222 nm ([0\2 2) for the LG3/LG2 Substituted Caviteins. Cavitein Concentration (HM) Experimental \_0\222 (deg cm 2 dmol"1) Calculated Maximum [0\222 (deg cm 2 dmol"1) Percent Helicity (%) LG2 39 -20000 -33200 -60 LG3 39 -18000 -33500 -54 2LG3»2LG2_ab 40 -19000 -33400 : -57 2LG3«2LG2_ac 39 -19000 -33400 -57 213 Figure 3 . 8 . Far-UV CD Spectra for the LG3/LG2C Substituted Caviteins at -40 uM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. -30000 J Wavelength (nm) All of the caviteins in the LG3/LG2C family are a-helical in structure, and the a-helicities of the LG3»LG2C_ab and LG3«LG2C_ac anti-parallel caviteins again seem to be an average of the helicities observed for their parent LG3 and LG2C derivatives. Similarly, there is little difference in the helicities between the a,b and a,c derivatives with their CD spectra being almost identical. Unfortunately, the design and synthesis of an anti-parallel hetero-TASP did not result in an observed increase in a-helicity. 214 Table 3.4. Molar Ellipticity at 222 nm ([^222) for the LG3/LG2C Substituted Caviteins. Cavitein Concentration (M-M) Experimental [^ 222 (deg cm 2 dmol"1) Calculated Maximum [^ 222 (deg cm 2 dmol"1) Percent Helicity (%) LG3 39 -18000 -33500 ; -54 LG2C 39 -21000 -33500 -63 2LG3«2LG2C_ab 40 -19500 -33500 -58 2LG3«2LG2C_ac 38 —20000 -33500 -60 The LG2C cavitein shows the highest percent helicity for this family of caviteins. The main structural difference between LG2C and LG3 is that LG2C has its peptide helices covalently linked to the cavitand template via C-terminal disulfide bonds. As has been explained previously, the linker has a dramatic effect on the structure of the caviteins, and thus it may be the disulfide bonds which are responsible for the slightly greater helicity observed for LG2C. 215 Figure 3.9. Far-UV CD Spectra for the LG3/AG3 Substituted Caviteins at -40 pJVl in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. 50000 -10000 -30000 - * - (4 )LG3 _ » - ( 4 ) A G 3 - * -3LG3_1AG3 - X - 2LG3_2AG3_ab —2LG3_2AG3_ac —•—1LG3 3AG3 250 Wavelength (nm) From examining the CD curves in Figure 3.8 it is apparent that all of the caviteins are a-helical, with a-helicity decreasing as the lg3 peptide helices are replaced with ag3 peptide helices. It is apparent that the entire CD curves seem to shift to lower wavelengths as the number of ag3 helices increases (notice the minimum at -203 nm for AG3, compared to -208 nm for LG3). The shift in the CD curve toward lower wavelengths is indicative of random coil contributions, and has been observed by Kwok and Hodges in studying the effects of changing the hydrophobicity in two-stranded coiled coils.15 It is interesting that the caviteins containing any number of ag3 peptides were not more a-helical than the leucine-based caviteins, since the alanine residue has the highest helical propensity.16 Furthermore, the a,c cavitein derivative 216 appears more helical than does it's a,b counterpart, and this difference between the a,b and a,c derivatives was not observed for the other hetero-TASP families. Table 3.5. Molar Ellipticity at 222 nm {\0\222) for the LG3/AG3 Substituted Caviteins. Cavitein Concentration (uM) Experimental \_6\222 (deg cm 2 dmol"1) Calculated Maximum[^222 (deg cm 2 dmol"1) Percent Helicity ••(%). LG3 39 -18000 -33500 -54 AG3 39 -9000 -33500 -27 3LG3»1AG3 39 -15000 -33500 -45 2LG3«2AG3_ab 40 -12000 —33500 -36 2LG3«2AG3_ac 39 -15000 -33500 -45 1LG3-3AG3 38 -10000 -33500 -30 Overall, the AG3 cavitein is the least helical of the caviteins, and this suggests that decreasing the hydrophobicity of the peptide helices results in a decrease in the a-helicity of the caviteins. It appears as though a decrease in a-helicity will correspond to a decrease in structural stability, as will be demonstrated in Section 3.2.5.3.1 3.2.5.2 Near-UV C D Spectroscopy The near-UV CD spectra for the LG3/LG2 variants, LG3/LG2C (anti-parallel) variants, and the LG3/AG3 variants at concentrations of -40 | i M are shown in Figures 3.10, 3.11 and 217 3.12, respectively. The near-UV region was used to analyze the tertiary structures of the hetero-TASPs. Figure 3.10. Near-UV CD Spectra for the LG3/LG2 Substituted Caviteins at -40 pM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. 1500 -2000 Wavelength (nm) The replacement of lg3 peptide helices with lg2 peptide helices had little effect on the near-UV CD signal when compared to the LG2 cavitein. The near-UV signals for 2LG3»2LG2_ab and 2LG3«2LG2_ac lie in between the curves observed for LG3 and LG2. However, the signs for the absorptions of 2LG3»2LG2_ab and 2LG3»2LG2_ac correspond to the signs of LG2, and are opposite to LG3. This supports the idea that the LG2 and 2LG3»2LG2_ab and 2LG3»2LG2_ac caviteins are all supercoiling in the same direction, while LG3 is supercoiling in the reverse direction. However, it could be that the 2 Gly-linked peptide 218 helices which are held more tightly to the bowl result in a stronger near-UV signal. The stronger signals for the 2 Gly-linked peptides could overshadow the signals from the 3 Gly-linked peptides, which are further removed from the cavitand chromophore. Figure 3.11. Near-UV CD Spectra for the LG3/LG2C Substituted Caviteins at -40 p:M in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. 1000 .2000 ~ : Wavelength (nm) Examining the near-UV region for the anti-parallel caviteins one can notice an entire loss in the near-UV CD signal. Even the LG2C reference cavitein has very little signal in this spectral region. In this case, the lack of a near-UV signal for the anti-parallel caviteins provides possible evidence for reduced tertiary structural specificity. However, another possible explanation for the absence of a near-UV signal is that perhaps the lg2c peptides are too far 219 removed from the disulfide or cavitand chromophore. It is not possible to make any definite conclusions from the near-UV data, although it could be that these disulfide linked caviteins exhibit more molten globule-like characteristics resulting in no near-UV signal. Figure 3.12. Near-UV CD Spectra for the LG3/AG3 Substituted Caviteins at -40 p M in 50 rhM pH 7.0 Sodium Phosphate Buffer at 20 °C. Wavelength (nm) The near-UV spectral region displays a relatively strong signal for both the LG3 and the 2LG3»2AG3_ac caviteins. The near-UV signal for 2LG3«2AG3_ac is clearly differentiable from the other variants of the LG3/AG3 family. This is the first time that a distinction has been observed between the a,b or a,c derivatives, and this further supports the fact that the a,c derivative was hypothesized to provide more efficient packing within the hydrophobic core. • ' 220 The near-UV spectral regions for any of the hetero-TASP provided no conclusive information on the tertiary structural properties, although some differences were noteworthy. 3.2.5.3 Oligomeric States The oligomeric states of the each of the hetero-TASPs were studied by monitoring the concentration dependence of their unfolding in the presence of guanidine hydrochloride (GuHCl) and by A U C sedimentation experiments. 3.2.5.3.1 GuHCl Denaturation Experiments As was previously mentioned all of the hetero-TASPs produced concentration independent CD spectra, however this is not sufficient enough data to conclude the existence of monomers in solution. Therefore, the stabilities of the hetero-TASPs in the presence of the denaturing salt, GuHCl, were determined at different concentrations. Figures 3.13, 3,14 and 3.15 display the unfolding curves of the LG3/LG2, LG3/LG2C and LG3/AG3 families, respectively, monitored at 222 nm in the presence of 0-8.0 M GuHCl at -^ -40 u M concentrations. The GuHCl experiments were also run at 4 u M concentrations (data not shown) for each of the caviteins, but the unfolding curves at both concentrations overlapped within experimental error. The A G ° H 2 O values of unfolding for the hetero-TASPs can be found in Tables 3.6, 3.7 and 3.8, and were calculated using the method described by SantorO and Bolen.'7 .221- ' Figure 3.13. Effect of GuHCl on the Helicity ( [ f e O of the LG3/LG2 Substituted Caviteins at -40 uM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. 1.1 [GuHCl] (M) Examining the unfolding curves for the LG3/LG2_ab and LG3/LG2_ac caviteins one can notice that the caviteins were completely unfolded by 8.0 M GuHCl, and that the unfolding transitions were cooperative. Furthermore, LG3/LG2_ab and LG3/LG2_ac were less stable towards the chemical denaturant than were LG3 and LG2. The a,b and a,c substituted caviteins began to denature at -3.0 M GuHCl, as compared to the curves for their reference caviteins which begin to unfold at -4.5 M GuHCl. 222 Table 3.6. Guanidine Hydrochloride-Induced Denaturation Data Calculated for the LG3/LG2 Substituted Caviteins. Cavitein Concentration (m [GuHCl] 1 / 2 (M) m (kcal/molM) A G ° H 2 O (kcal/mol) LG2 39 5.7 ± 0 . 1 -1.8 ± 0 . 1 -10.4 ± 0 . 3 LG2 4 5.8 ± 0 . 1 -1.8 ± 0 . 1 -10.2 ± 0 . 3 LG3 39 5.6 ± 0 . 1 -1.9 + 0.1 -10.8 ± 0 . 4 LG3 4 5.7 ± 0 . 1 -1.8 ± 0 . 1 -10.7 ± 0 . 4 2LG3»2LG2. .ab 40 5.4 ± 0 . 1 -1.2 ± 0 . 1 -6.6 ± 0 . 3 2LG3«2LG2. .ab 4 5.4 ± 0 . 1 -1.2 ± 0 . 1 -6.6 + 0.3 2LG3«2LG2. _ac 39 5.2 ± 0 . 1 -1.4 ± 0 . 1 -7.4 ± 0 . 5 2LG3«2LG2. _ac 3 5.1 ± 0 . 1 -1.4 ± 0 . 1 -7.1 ± 0 . 4 The A G ° H 2 O values of unfolding for the LG3/LG2 caviteins at concentrations of 4 | i M and 40 JJ.M fall within experimental error of each other, respectively. Both the LG2 and LG3 caviteins are the most stable, and within experimental error of each other. By altering the four-helix bundles of LG3 or LG2 to the hetero-TASP counterparts, a loss in structural stability of 3-4 kcal/mol is encountered. This provides clear evidence that they thermodynamic stability of the four-helix bundle is reduced as the helices within the bundle are different. This further suggests that changing the linker lengths of the constituent helices of the bundle did not accommodate a more efficient packing arrangement. In this way, it appears that a cavitein where the peptide sequences have equivalent linkers is a more efficient design, or then a suitable combination of peptide sequences to be incorporated into a hetero-TASP has yet to be designed. 223 By comparing the m values for the caviteins listed in Table 3.6 it is clear that LG3 and LG2 have higher m values, suggesting a more cooperative unfolding transition, and hence more native-like character. A simple conclusion would be that 2LG3»2LG2_ab and 2LG3»2LG2_ac are more molten globule-like in structure, than are LG3 or LG2. Figure 3.14. Effect of GuHCl on the Helicity ([^222) of the LG3/LG2C Substituted Caviteins at -40 uM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. TJ Oi TJ 1.1 0.9 0.7 c 0.5 O o TO LL 0.3 0.1 -0.1 - » - ( 4 ) L G 3 - • - ( 4 ) L G 2 C - * - 2 L G 3 _ 2 L G 2 C _ a b - * - 2 L G 3 2LG2C ac [GuHCl] (M) Examining the unfolding curves for the LG3/LG2C substituted caviteins one can notice that all of the caviteins were completely unfolded by 8.0 M GuHCl, and the unfolding transitions were very similar and cooperative. The anti-parallel caviteins (LG3»LG2C_ab and LG3»LG2C_ac) were less stable toward the chemical denaturant as compared to their respective 224 reference caviteins (LG3 and LG2C). One would believe that the anti-parallel caviteins with a reduced helix macrodipole should be more stable than their reference LG3 and LG2C caviteins. Perhaps the hydrophobic packing in the anti-parallel caviteins was not as optimal as their reference caviteins rendering them less stable. Furthermore, the unfolding curve for the LG2C reference cavitein was slightly steeper than the other denaturation curves, and resulted in the greatest value for the A G ° H 2 O values of unfolding. Two possible reasons for this increase in stability observed for LG2C could be due to the peptides being linked via disulfide bonds to the cavitand template, or because they are linked via their C-termini,13 and the bowl serves as a better C-cap than N-cap. The latter explanation would be a more likely since other disulfide linked caviteins synthesized in our lab l b have exhibited much lower stabilities towards the same chemical denaturant, than that observed for LG2C. 225 Table 3.7. Guanidine Hydrochloride-Induced Denaturation Data Calculated for the LG3/LG2C Substituted Caviteins. Cavitein Concentration (UM) [GuHCl] 1/2 (M) m (kcal/molM) A G ° H 2 O (kcal/mol) LG3 39 5.6 ± 0 . 1 -1.9 ± 0 . 1 -10.8 ± 0 . 4 LG3 • 4 ."• 5.7 ± 6 . 1 -1.8 + 0.1 -10.7 ± 0 . 4 LG2C 39 5.4 ± 0 . 1 -2.3 ± 0 . 1 -11.8 ± 0 . 4 LG2C 3 5.4 ± 0 . 1 -2.2 ± 0 . 1 -11.6 ± 0 . 3 2LG3-2LG2C. .ab 40 5 . 5 ± 0 . 1 -1.6 ± 0 . 1 -8.8 ± 0 . 4 2LG3»2LG2C. .ab 4 5.6 ± 0 . 1 -1.7 ± 0 . 1 -8.6 ± 0 . 3 2LG3«2LG2C_ _ac 38 5.4 ± 0 . 1 -1.5 ± 0 . 1 -8.4 ± 0 . 4 2LG3»2LG2C _ac 3 5.3 ± 0 . 1 -1.5 + 0.1 -8.4 ± 0.3 The A G ° H 2 O values of unfolding for the LG3/LG2C caviteins at concentrations of 4 .uM and 40 uM fall within experimental error of each other, respectively. LG2C is the most stable cavitein, most likely resulting from the favorable C-capping effects of the cavitand template. The m value is the largest for LG2C, followed by LG3 and LG2, and then by the substituted caviteins. This data would suggest that LG2C has the most native-like character of these caviteins, which complements the high stability of LG2C. It appears that the synthesis of the anti-parallel cavitiens did not increase the native-like properties or stability of the caviteins. 226 Figure 3.15. Effect of GuHCl on the Helicity ([0)222) of the LG3/AG3 Substituted Caviteins at -40 uM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. GuHCl (M) Examining the unfolding curves for the LG3/AG3 substituted caviteins one can notice that all of the caviteins were completely unfolded by 8.0 M GuHCl, and the unfolding transitions were non-cooperative except for the reference cavitein LG3. It is clear from the denaturation curves that the replacement of a single lg3 helix with an ag3 helix results in a large loss in: structural stability. This could be explained by the fact that the side chains of the alanine residues of a single helix within the four-helix bundle (3LG3»1AG3) cannot fill the hydrophobic core as efficiently as can the side chains of the leucine residues. The increased loss in hydrophobicity as the lg3 helices get replaced with ag3 helices has a dramatic effect on structural stability rendering the AG3 cavitein the least stable. 227 Table 3.8. Guanidine Hydrochloride-Induced Denaturation Data Calculated for the LG3/AG3 Substituted Caviteins. Cavitein Concentration OiM) [GuHCl] 1 / 2 (M) m (kcal/molM) A G ° H 2 O (kcal/mol) LG3 39 5.6 ± 0 . 1 -1.9 ± 0 . 1 -10.8 ± 0 . 4 LG3 4 5.7 ± 0 . 1 -1.8 + 0.1 -10.7 ± 0 . 4 AG3 39 1.1 ± 0 . 1 - • ••' AG3 ' ' 4 •1.0 ± 0 . 1 -3LG3«1AG3 39 3.3 ± 0 . 1 -0.8 ± 0 . 1 -3.7 ± 0 . 4 3LG3»1AG3 3 3.4 ± 0 . 1 -0.8 ± 0 . 1 -3.7 ± 0 . 3 2LG3»2AG3_ab 40 2.1 ± 0 . 1 -2LG3«2AG3_ab 4 2.1 ± 0 . 1 -2LG3»2AG3_ac 39 2.2 ± 0 . 1 - -2LG3«2AG3_ac 3. 2.1 ± 0 . 1 - • -1LG3-3AG3 38 1.6 ± 0 . 1 - '-1LG3»3AG3 . 3 . ' 1.6 ± 0 . 1 - - • The A G ° H 2 O values of unfolding for the LG3/AG3 caviteins at concentrations of 4 u M and 40 u M fall within experimental error of each other, respectively. LG3 is the most stable of the six caviteins. Replacing any of the lg3 helices with ag3 helices has an unfavorable effect on the stability of the four-helix bundle. No A G ° H 2 O values could be evaluated for 2LG3»2AG3_ab, 2LG3«2AG3_ac, 1LG3»3AG3, and AG3 due to the lack of pre- and post-transitional baselines in their GuHCl denaturation curves. 228 From inspection of Table 3.8 it is obvious that even the replacement of a single lg3 helix with an ag3 helix reduces the m value substantially. This reduction in the m value further supports that the packing within the hydrophobic core is sacrificed when an all-alanine helix is incorporated into the four-helix bundle, reflecting a reduction in the native-like character. 3.2.5.3.2 AUC Sedimentation Equilibrium Experiments The hetero-TASP samples were prepared similarly to caviteins outlined in Chapter 2, and were analyzed at concentrations of 10, 50, and 80 pM and at rotor speeds of 27000, 35000, and 40000 rpm. The sedimentation equilibrium data for the hetero-TASPs were analyzed by NONLIN, 1 0 and the exponential plots of absorbance versus radius for 2LG3»2LG2_ab and 2LG3«2LG2_ac are shown in Figures 3.16 and 3.17, respectively. The other hetero-TASP plots were very similar and can be found in Section 3.4.5. 229 Figure 3.16. Sedimentation Equilibrium Concentration Distributions of 2LG3«2LG2_ab at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C, and at a Concentration of 10 uM. In the Lower Panel The Solid Lines Represents Theoretical Fits to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. CO o I < o c o •£ Q 5.000 L 0.000 L -5.000 U 17.625 17.875 18.125 18.375 Radius Squared/2 (cmA2) o i < o c o ••s u ti o U 11.250 U 8.750 U 6.250 17.625 17.875 18.125 Radius Squared/2 (cmA2) 18.375 230 Figure 3.17. Sedimentation Equilibrium Concentration Distributions of 2LG3«2LG2_ac at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C, and at a Concentration of 10 pM. In the Lower Panel The Solid Lines Represents Theoretical Fits to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. CO o • < o I o ••s <D fi 5.000 L 0.000 L -5.000 17.500 C O o % °o * o o ° Q O O O ° •=> o t> Q -° ° o o * o ° ° o o o °0 O 0 © ° o © o o © o 18.000 18.500 Radius Squared/2 (cmA2) o i < o I a o fi (U o a o O 11.250 L 8.750 L 6.250 17.500 18.000 18.500 Radius Squared/2 (cmA2) 231 The data for each of the hetero-TASPs were also fit'to monomer-dimer, dimer, monomer-trimer, trimer, monomer-tetramer, and tetramer in order to check for the best theoretical fit to the experimental data. In all cases the monomer fits were most accurate, assessed by the even distribution of the residuals about zero as can be seen in the upper panels of Figures 3.15 and 3.16. The results from the sedimentation equilibrium data for all of the hetero-TASPs are consistent with the CD and GuHCl denaturation data for the presence of a monomeric species. Tables 3.9, 3.10 and 3.11 summarize the sedimentation data for all of the hetero-TASP families, respectively. Table 3.9. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For LG3/LG2 Substituted Caviteins at 20 °C in 50 mM pH 7.0 Sodium Phosphate Buffer at Concentrations of 10, 50, and 80 \iM with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section 3.4.5) Cavitein Experimentally Estimated M W (Da) Calculated M W (Da) Predominant Species LG2 8500 ± 6 0 0 8016 Monomer LG3 8000 ± 3 0 0 8240 Monomer 2LG3»2LG2_ab 7800 ± 5 0 0 8128 Monomer 2LG3»2LG2_ac 7900 ± 400 8128 Monomer 232 Table 3.10. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For LG3/LG2C Substituted Caviteins at 20 °C in 50 mM pH 7.0 Sodium Phosphate Buffer at Concentrations of 10, 50, and 80 p M with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section . 3.4.5) Cavitein Experimentally Estimated M W (Da) Calculated M W (Da) Predominant Species LG3 LG2C 2LG3»2LG2C_ab 2LG3»2LG2C_ac 8000 ± 3 0 0 8700 ± 6 0 0 8800 + 500 8700 + 400 8240 8424 8332 8332 Monomer Monomer Monomer Monomer Table 3.11. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For LG3/AG3 Substituted Caviteins at 20 °C in 50 mM pH 7.0 Sodium Phosphate Buffer at Concentrations of 10, 50, and 80 pJVI with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section 3.4.5) : Cavitein Experimentally Calculated M W Predominant Estimated M W (Da) Species (Da) LG3 AG3 2LG3-2AG3 ab 2LG3»2AG3 ac 8000 ± 300 7500 ± 2 0 0 8000 ± 400 7900 ± 3 0 0 8240 7400 7820 7820 Monomer Monomer Monomer Monomer 233 3.2.5.4 One-Dimensional (ID) *H N M R Spectroscopy In 'Hi NMR spectroscopy usually the spectral region between 7 and 11 ppm is examined because distinct sharp and dispersed signals are observed for native-like proteins corresponding to the slowly exchanging amide protons.11 The *H NMR spectra for hetero-TASP families are shown in Figures 3.18, 3.19 and 3.20, respectively. The figures display an overlay of the amide regions for the hetero-TASPs belonging to each family to more clearly illustrate the spectral region of interest. Figure 3.18. Expansions of the Amide Regions of 500 MHz *H NMR Spectra of the LG3/LG2 Substituted Caviteins at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (a) LG3 (b) LG2 (c) 2LG3«2LG2_ab (d) 2LG3»2LG2_ac (* = cavitand signals) * M | I I. U | M I I | I I M | I I I I | I M . I | I I I I | M II | I I I I | M M | I M I | M I I | I I M | I I 1 I | I I I I | I I I I | I I I I | r i 1 I | r I I I | I I - M | II I I | I I 9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 ppm 234 The ' H NMR spectra of LG2 and LG3 each show -13 distinguishable dispersed amide signals indicative of a well-defined amide backbone with a high content of tertiary structure. The presence of only 13 amide signals for the LG2 and LG3 caviteins suggest that many of the amino acid residues are in a degenerate environment and therefore indistinguishable from each other, likely due to the four-fold symmetry of the cavitein. Looking at the spectra of 2LG3»2LG2_ab and 2LG3«2LG2_ac the amide signals are more broad and less resolved than the amide signals of the parent reference caviteins, LG3 and LG2. More specifically, the dispersion and sharpness for the amide signals of the LG3/LG3 substituted caviteins are comparable to the amide dispersion and sharpness exhibited by LG2. It is perhaps the influence of the lg2 peptide helices, which are more closely linked to the cavitand template as compared to lg3 peptide helices, which largely dictate the overall tertiary structure. Lastly, the amide signals in the ! H NMR spectrum for the 2LG3«2LG2_ac cavitein are only slightly sharper than the amide signals in the spectrum of the 2LG3«2LG2_ac cavitein. 235 Figure 3.19. Expansions of the Amide Regions of 500 MHz ! H NMR Spectra of the LG3/LG2C Substituted Caviteins at 1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (a) LG3 (b) LG2C (c) 2LG3»2LG2C_ab (d) 2LG3«2LG2C_ac (* = cavitand signals) (b) I I 11111; 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I I 111 r 1111 I I 11111 i 111 M 1111 M 11111111 M 11 I I 111 : 11111 I I 111111111111 .i i I I 111111' I I 11' 111111 9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 . P P m The *H NMR spectrum of LG3 is considerably sharper than any of the other caviteins in this family. Unfortunately, the spectrum of the LG2C cavitein is of poor quality due to a lack of sufficient protein sample (synthesis hampered by an S-pyridyl impurity making purification difficult). Looking at the spectrum of 2LG3«2LG2C_ac the amide signals are slightly sharper than the amide signals in the 2LG3»2LG2C_ab cavitein. It is difficult to make any strong conclusions about the tertiary structures of the anti-parallel caviteins, although they do exhibit less native-like character than LG3. One interesting observation in the ' H N M R spectrum for the 236 2LG3»2LG2C_ab cavitein is the splitting of the cavitand signal (H o ut) at -6.1 ppm into three signals in a 1:2:1 ratio. A 1:2:1 splitting of the H o u t cavitand signal supports the a,b symmetry. Figure 3.20. Expansions of the Amide Regions of 500 MHz *H NMR Spectra of the LG3/AG3 Substituted Caviteins at 1.5 mM in 10 % D2O, 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (a) LG3 (b) AG3 (c) 2LG3»2AG3_ab (d) 2LG3«2AG3_ac (* = cavitand signals) * •(c) (d) • I I M j I M I I M I 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 I I I I I I I M I I 1 1 1 I I I I I I I I I I I I I I I I I I I I I I I I M I I I I I I I I I' I I I I I I I I I I I I I I . I I I I I 9.8 9.4 9.0 8.6 8.2 7.8 .7.4 7.0 6.6 6.2 5.8 ppm Analyzing the *H NMR spectra in Figure 3.20 shows that the amide region of LG3 and AG3 are considerably sharper than those of 2LG3«2AG3_ab or 2LG3«2AG3_ac. The amide region for AG3 however, is less dispersed than the amide region of LG3. The broad spectra for 2LG3«2AG3_ab and 2LG3»2AG3_ac suggest that these caviteins are less native-like than their parent reference caviteins, and exhibit some molten globule-like characteristics. The broad 237 signals in the amide region of the J H NMR spectra of 2LG3»2AG3_ab and 2LG3»2AG3_ac complement the non-cooperative unfolding curves observed in the GuHCl denaturation experiments, again supporting that the tertiary structures of 2LG3»2AG3_ab and 2LG3»2AG3_ac were possibly molten globule-like. A similar argument could not be made for AG3, which has very sharp, although poorly dispersed, amide signals in its *H NMR spectrum, characteristic of native-like structure, whereas a non-cooperative unfolding curve was observed in the GuHCl denaturation experiment. 3.2.5.5 Hydrogen/Deuterium Amide Exchange A representative stack plot for the *H NMR spectra amide region for the 2LG3«2LG2_ab hetero-TASP is shown in Figure 3.21. All of N-H/D exchange data were analyzed for the other hetero-TASPs, and all of the protection factors are tabulated Tables 3.12, 3.13 and 3.14. All of the amide N-H/D exchanges for the AG3, 2LG3«2AG3_ab, and 2LG3»2AG3_ac caviteins were complete before the first spectrum could be recorded, and therefore no protection factors could be calculated. 238 Figure 3.21. Stack Plot of 500 MHz *H NMR Spectra Illustrating the Time Dependent Amide H/D Exchange of 2LG3»2LG2_ab in 50 mM pD 5.02 CD 3 COOD/CD 3 COO"Na + Buffer at 20 °C. (a) 4 min (b) 18 min (c) 1 h 3 min (d) 6 h 9 min (* = cavitand signals) * 9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 ppm The amide H/D exchanges of all of the hetero-TASPs were studied over the course of about four hours depending on the hetero-TASP. Many of the amide protons exchanged before the first scan could be acquired. 239 Table 3.12. Tabulated Data from the Amide H/D Exchange Experiments of the LG3/LG2 Substituted Caviteins in a 50 mM pD 5.02 C D 3 C O O D / CD 3 C00"Na + Buffer at 20 °C. Cavitein Name Amide Proton Chemical Shift" (ppm) First-Order Rate Constant Half-Life (h) Protection Factorb LG3 8.5 2.97 x 10"2 23 (7.3 ± 0.5) x 103 LG2 8.5 3.37 x 10"2 20 (6.3 ± 0.5) x 103 2LG3»2LG2_ab 8.5 4.62 x 10'1' 1.0 (3.1 ± 0 . 3 ) x 102 2LG3«2LG2_ac 8.5 6.93 x l O 1 1.5 (4.7 ± 0.3) x 102 a only the data on the most protected proton is included. b these values are based on the half-life of an unprotected proton at 20 °C at pD 5.02 to be 3.18 x 10"3 h. Upon examination of Table 3.12 it is clear that the proton at 8.5 ppm for 2LG3«2LG2_ab and 2LG3«2LG2_ac exhibits less protection from exchange, than the proton at 8.5 ppm for LG3 or LG2. In this way, the design of the 2LG3«2LG2_ab and 2LG3»2LG2_ac hetero-TASPs did not appear to increase the specificity of the hydrophobic core. However, all of the protection factors outlined in Table 3.12 fall in the range of molten globule-like structures, and therefore perhaps an entirely native-like de novo protein has yet to be designed. 240 Table 3.13. Tabulated Data from the Amide H/D Exchange Experiments of the LG3/LG2C Substituted Caviteins in a 50 mM pD 5.02 C D 3 C O O D / CD 3 COO"Na + Buffer at 20 °C. Cavitein Name Amide Proton Chemical Shift3 First-Order Rate Constant (h-1) Half-Life (h) Protection Factor15 LG3 8.5 2.97 xlO" 2 23 (7.3 ± 0.5) x 103 LG2C 8.5 1.39 0.50 (1.6 ± 0.2) x 102 2LG3»2LG2C_ab 8.5 2.77 0.25 (7.9 ± 0.2) x 101 2LG3«2LG2C_ac 8.5 2.10 0.33 (1.0 ± 0.3) x 102 a only the data on the most protected proton is included. b these values are based on the half-life of an unprotected proton at 20 °C at pD 5.02 to be 3.18 x 10"3h. The proton at 8.5 ppm for the 2LG3«2LG2C_ab hetero-TASP is the least protected from exchange with the solvent according to the protection factors tabulated in Table 3.13, although it does lie within experimental error of the protection factor calculated for 2LG3«2LG2C_ac. The N-H/D exchange data should not be over interpreted, however, it does seem reasonable to conclude that the proton at 8.5 ppm in LG3 does exhibit the most protection from exchange compared to similar protons in LG2C, 2LG3»2LG2C_ab, and 2LG3«2LG2C_ac. The anti-parallel hetero-TASPs namely 2LG3»2LG2C_ab, and 2LG3»2LG2C_ac have hydrophobic cores that are clearly more mobile than that of LG3. 241 Table 3.14. Tabulated Data from the Amide H/D Exchange Experiments of the LG3/AG3 Substituted Caviteins in a 50 mM pD 5.02 C D 3 C O O D / CD 3 COO"Na + Buffer at 20 °C. Cavitein Name Amide Proton Chemical First-Order Rate Constant Half-Life (h) Protection Factor6 Shift1 .a (h") LG3 8.5 2.97 x 10 ,-2 23 (7.3 ± 0.5) x 103 AG3 C - - - -2LG3»2AG3_ab c - - - -2LG3»2AG3_ac c - - - -a only the data on the most protected proton is included. b these values are based on the half-life of an unprotected proton at 20 °C at pD 5.02 to be 3.18 x 10"3h. c the exchanges of the amide protons were too fast in order to calculate any protection factors. All of the alanine-substituted caviteins outlined in Table 3.14 had amide protons, which exchanged with the solvent before the first NMR spectrum could even be acquired. All of the experimental data acquired so far (CD and GuHCl denaturation experiments, ! H NMR spectra and N-H/D exchange data) support that the replacement of leucine-based peptide sequences with those containing alanine residues reduces the stability of the four-helix bundle, resulting in molten globule-like properties. Overall LG3 has the largest protection factor determined from the N-H/D exchange experiments compared to all of the caviteins. However, the protection factor for LG3 still falls in the range of a molten globule-like protein. The protection factor is only one piece of experimental evidence, and from the other physical studies LG3 seems to be a highly native-like protein. 242 3.2.5.6 ANS Binding ANS binding was studied by fluorescence spectroscopy for LG3/LG2, LG3/LG2C, and LG3/AG3 cavitein families, and their emission spectra are shown in Figures 3.22, 3.23 and 3.24, respectively. Under standard experimental conditions negligible binding was observed for any of the caviteins. The studies were completed using three different concentrations per cavitein of 50, 100 and 150 pJvf and the curves were indistinguishable (note Figures 3.22, 3.23 and 3.24 for clarity's sake only show data for the 50 u M cavitein concentration). Figure 3.22. Fluorescence Emission Spectra of 2 pM ANS in the Presence of 95 % Ethanol, 100 % Methanol, 50 u M LG3/LG2 Substituted Caviteins at 20 °C in pH 7.0 50 mM Sodium Phosphate Buffer. (A c CD O c o (A <D w o 3 385 (4)LG3 •(4)LG2 •2LG3_2LG2_ab •2LG3_2LG2_ac •EtOH MeOH 435 485 535 Wavelength (nm) 585 243 No binding of ANS was observed for the LG3/LG2 substituted caviteins. The amide regions of the ! H NMR spectra of LG3, LG2, 2LG3»2LG2_ab, and 2LG3»2LG2_ac support native-like tertiary structures. This is in agreement with no appreciable binding of ANS. Figure 3.23. Fluorescence Emission Spectra of 2 uM ANS in the Presence of 95 % Ethanol, 100 % Methanol, 50 uM LG3/LG2C Substituted Caviteins at 20 °C in pH 7.0 50 mM Sodium Phosphate Buffer. ( 4 ) L G 3 ( 4 ) L G 2 C 385 435 485 535 585 Wavelength (nm) A very slight binding of ANS is observed in the fluorescence emission spectra for 2LG3«2LG2C_ab, and 2LG3»2LG2C_ac. However, the binding of ANS observed for 2LG3»2LG2C_ab, and 2LG3»2LG2C_ac is neglible compared to the binding of ANS to the reference compounds, EtOH and MeOH. However, if ANS binds preferentially to hydrophobic 244 pockets and not exclusively to hydrophobic surfaces in molten globule-like structures, then the lack of observed binding of ANS for our caviteins would not preclude the possibility of a molten globule-like structure. Figure 3.24. Fluorescence Emission Spectra of 2 pM ANS in the Presence of 95 % Ethanol, 100 % Methanol, 50 uM LG3/AG3 Substituted Caviteins at 20 °C in pH 7.0 50 mM Sodium Phosphate Buffer. (4)LG3 (4)AG3 385 435 485 535 585 Wavelength (nm) For the 2LG3«2AG3_ab hetero-TASP a small amount of ANS binding is detected. However, no appreciable ANS binding is observed for the 2LG3»2AG3_ac hetero-TASP suggesting that the two LG3/AG3 disubstituted caviteins differ at least slightly in their tertiary structures. 245 3.3 Summary and Conclusions This Chapter presented the design, synthesis and characterization of different hetero-TASP families in the hopes of creating native-like proteins, and more generally to develop a methodology to synthesize TASPs made up of different peptide sequences. Two different approaches were developed and tested. The first one used experimentally determined ideal peptide equivalents, and the second one employing protecting groups. Both approaches were successful in the formation of the four-helix bundle hetero-TASPs, although approach one was used due to synthetic simplicity. In the synthesis of the hetero-TASPs, two lg3 a,b and a,c disubstituted intermediates were isolated and studied along the way. It was hypothesized that the disubstituted intermediates would potentially aggregate into anti-parallel dimers, which would have opened the door to studying peptide-peptide interactions. Unfortunately, the a,b and a,c disubstituted intermediates exhibited molten globule-like properties, and were found to exist as non-specific aggregates by A U C and'H NMR. Three different hetero-TASP families which were then synthesized from the lg3 a,b and a,c disubstituted intermediates were designed with specific reasons and goals in mind. The conclusions for each family will be discussed separately. The first family included the LG3/LG2 substituted caviteins in addition to the two reference caviteins, LG3 and LG2. These hetero-TASPs were designed in an attempt to model "knobs-into-holes" packing. Incorporating two different peptide sequences (lg3 and lg2) onto the cavitand template with linkers differing in length by one glycine residue would hopefully have placed the side chains of the lg3 and lg2 peptide helices "out of register". Physical studies were then performed to assess oligomeric state, protein stability, and secondary and tertiary structures. 246 The oligomeric states of the LG3/LG2 substituted caviteins were studied by CD spectroscopy and A U C , and were found to exist as monomers in solution. Furthermore, 2LG3»2LG2_ab, and 2LG3«2LG2_ac were found to be highly helical from their the CD spectra, and had very similar \6\222 values. The shapes of the CD curves of 2LG3»2LG2_ab, and 2LG3»2LG2_ac appeared to be intermediate between the CD curves of the parent reference compounds, LG3 and LG2. The tertiary structures of 2LG3»2LG2_ab, and 2LG3«2LG2_ac were assessed by their stability toward GuHCl, amide dispersion by *H NMR spectroscopy, protection factors determined by N-H/D exchange and by the binding of ANS. 2LG3«2LG2_ab, and 2LG3»2LG2_ac were found to be -3 kcal/mol less stable than LG3 and LG2, and 2LG3«2LG2_ac was only slightly more stable than 2LG3»2LG2_ab. It was predicted that 2LG3«2LG2_ac would be the more stable of the two hetero-TASPs because the lg2 peptides helices and lg3 peptides helices would be opposite from one another on the template, respectively, and this was predicted to be a more favorable orientation for the formation of ideal hydrophobic packing. The amide regions of the ' H NMR spectra for 2LG3»2LG2_ab, and 2LG3»2LG2_ac are only slightly broader than the amide region of the ! H NMR spectrum of LG2. The amide region of the 5 H NMR spectrum of LG3 is the sharpest, although slightly less dispersed than that of LG2. It is interesting to note that although 2LG3«2LG2_ab, and 2LG3«2LG2_ac are -3 kcal/mol less stable than LG2, their ' H NMR spectra are very similar. The experimentally determined protection factors for the amide protons at -8.5 ppm for 2LG3»2LG2_ab, and 2LG3»2LG2_ac were lower by a factor of ten, than LG3 or LG2. Lastly, neither 2LG3»2LG2_ab or 2LG3»2LG2_ac were found to bind ANS in any considerable 247 amount. Taking all of the experimental data into account the anti-parallel caviteins did not result in a more stable or native-like tertiary structure than LG3. A possible explanation for the reduced native-like structures of the LG3/LG2 substituted caviteins is that the hydrophobic cores were not able to efficiently pack with specific tertiary interactions. The formation of the hydrophobic core is main driving force for protein folding, and it is the specific side chain contacts within the hydrophobic core that provide stability and a high degree of native-like structure. The LG3/LG2 substituted caviteins were not likely hampered by any steric restrictions since the reference caviteins, LG3 and LG2, show a high content of native-like structure. Since it is very difficult to conclude why the LG3/LG2 hetero-TASPs did not exhibit native-like properties, it would be useful to use molecular dynamic simulations in order to study these hetero-TASPs further. Furthermore, perhaps a more successful design could include designing a LG3/LG1 hetero-TASP, where there is a larger difference in the linker length than what was used for the LG3/LG2 hetero-TASPs. Chapter 5 will expand on the suggestions for future work of this project. The second hetero-TASP family, which was designed and studied, consisted of the LG3/LG2C substituted caviteins. The idea behind synthesizing an anti-parallel four-helix bundle was to make a cavitein where the helix macrodipoles of the attached helices would complement each other, and potentially create a more stable cavitein. Furthermore, an anti-parallel bundle was also thought to be able to accommodate more efficient hydrophobic packing, since the peptide helices within the bundle likely had different linker lengths. The synthesis of 2LG3»2LG2C_ab, and 2LG3»2LG2C_ac involved the incorporation of two peptide helices linked via their C-termini to the cavitand template by means of a disulfide bond. Disulfide linked Caviteins have previously been designed in our group, and many of those caviteins have exhibited molten globule-like characteristics.15 248 The oligomeric states of the LG3/LG2C substituted caviteins were studied by CD spectroscopy and A U C , and were found to exist as monomers in solution. Furthermore,' 2LG3«2LG2C_ab, and 2LG3«2LG2C_ac were found to be highly helical from their the CD spectra, and had very similar {6^222 values. The tertiary structures of 2LG3»2LG2C_ab, and 2LG3«2LG2C_ac were assessed by their stability toward GuHCl, amide dispersion by ' H NMR spectroscopy, protection factors determined by N-H/D exchange and by the binding of ANS. 2LG3*2LG2C_ab, and 2LG3»2LG2C_ac were found to be ~2 kcal/mol less stable than LG3 and LG2C, and the A G ° H 2 O values for 2LG3»2LG2C_ab and 2LG3«2LG2C_ac were within experimental error of each other. It was hypothesized, however, that 2LG3»2LG2C_ac would be more stable since the anti-parallel peptides would be opposite one another due to the a,c positions on the cavitand template. Lastly, the LG2C reference cavitein was found to be the most stable of the caviteins of this family according to the GuHCl denaturation data. A possible explanation for the higher stability observed for LG2C could be that the cavitand template serves as a better C-cap than N-cap. Chapter 4 focuses on further exploring the N- and C-capping effects of the cavitand template, and a commonly used capping residue, glycine. The amide regions of the J H NMR spectra for 2LG3«2LG2C_ab, and 2LG3«2LG2C_ac are broader than the amide region of the J H NMR spectrum of LG3. Unfortunately, the spectrum of the LG2C cavitein is of poor quality due to a lack of sufficient protein sample. Furthermore, the amide signals in the spectrum of 2LG3»2LG2C_ac are more resolved than the amide signals in the 2LG3»2LG2C_ab cavitein, suggesting that it could be slightly more native-like in character. It is difficult to make any definite conclusions about the tertiary structures of the anti-parallel caviteins, although they do exhibit less native-like character than LG3. It is interesting 249 to note that although 2LG3«2LG2C_ab, and 2LG3»2LG2C_ac are comparable in stability, their spectra do indicate that their tertiary structures are different. The mobility of the hydrophobic cores of 2LG3«2LG2C_ab, and 2LG3«2LG2C_ac were studied by monitoring the rate of exchange of their amide protons with the D2O. The experimentally determined protection factors for the amide protons at -8.5 ppm in 2LG3»2LG2C_ab, and 2LG3«2LG2C_ac were similar to each other and much lower than observed for LG3. Furthermore, neither 2LG3»2LG2C_ab or 2LG3«2LG2C_ac were found to bind any ANS. It can be concluded that 2LG3«2LG2C_ab, and 2LG3«2LG2C_ac were less native-like in character than the reference caviteins LG3 and LG2C. It appears as though the introduction of peptide helices with different helix orientation on the cavitand template did not result in the formation of a more stable or more native-like protein. Possible reasons could be the disulfide or the C-terminal linkage of the lg2c peptides to the cavitand template that were inhibiting the formation of a well-pack hydrophobic core, and therefore, other linkages should be explored. No definite conclusions can be made as to the exact reason for the more molten globule-like properties of the anti-parallel TASPs, although more sequences should be tested to gather more experimental data. The final hetero-TASP family, which was designed and studied, consisted of the LG3/AG3 substituted caviteins. The rationale behind designing these alanine-substituted caviteins was to determine whether the hydrophobic core of LG3 was overpacked, and whether a less hydrophobic sequence could enhance the native-like properties. Alanine residues, with smaller hydrophobic side chains than leucine residues, were introduced as the sole hydrophobic residues into the peptide sequence. Once the hetero-TASPs were synthesized, physical studies were performed to evaluate the properties of these caviteins. 250 It was evident from all of the experimental data acquired that the replacement of even a single lg3 peptide helix with that of an ag3 helix greatly reduced the oc-helicity, stability, and native-like structure of the cavitein. The CD spectra for the alanine-substituted caviteins show a consistent decrease in rx-helicity as more lg3 helices are replaced with ag3 helices. The CD curve for the AG3 cavitein is entirely shifted toward a lower wavelength indicative of more random coil contributions. The LG3/AG3 substituted caviteins provided the first example of a non-cooperative unfolding of a four-helix bundle, when subjected to GuHCl, and therefore the A G ° H 2 O values for AG3, 2LG3«2AG3_ab, 2LG3»2AG3_ac, and 1LG3-3AG3 could not be determined. Again, a decrease in stability toward the chemical denaturant was observed as more lg3 peptides helices were replaced with ag3 helices. Furthermore, the A G ° H 2 O value for 3LG3»1AG3 was ~7 kcal/mol less than that of LG3, resulting from the replacement of a single lg3 peptide helix. The amide regions in the ! H NMR spectra for 2LG3»2AG3_ab, and 2LG3»2AG3_ac were very broad indicative of molten globule-like structure. On the other hand, the amide region in the ! H NMR spectrum of AG3 was very sharp, although not very dispersed. A possible reason for the lower dispersion yet distinct sharpness of the amide signals for AG3 could be that the peptide helices in AG3 were able to supercoil. As was mentioned in Chapter 2, native-like coiled coil proteins generally exhibit sharp, less dispersed signals in their amide regions, with the lower dispersion being attributed to the supercoiling of the helices.19 Even more conclusive evidence that the alanine-substituted caviteins exhibit molten globule-like properties is the instantaneous amide proton exchange with the solvent. The slight binding of ANS to 2LG3»2AG3_ab is supportive of a molten globule-like structure as well. It seems as though four-helix bundles with any number of alanine-based peptide helices results in a poorly packed hydrophobic core. It is likely that the methyl side chain of an alanine residue is 251 much too small to fill the hydrophobic pocket created when peptide helices are covalently linked onto the arylthiol cavitand template. It is known that hydrophobicity is a driving force for folding and for creating a stable hydrophobic core. Perhaps the ag3 peptide sequence should be tested on a cyclOtribenzylene (CTB) template to determine whether an all alanine peptide sequence would be hydrophobic enough to accommodate the formation a stable three-helix bundle. Overall this Chapter presented the first attempts at designing and synthesizing four-helix bundles containing different peptide sequences within one bundle. Although, the first generations of hetero-TASPs did not result in an increased native-like structure when compared to LG3 or LG2, it did result in the development of a new methodology for synthesizing novel caviteins. By simply synthesizing a hetero-TASP, one of the main goals of this project was realized. Furthermore, a greater understanding of the factors influencing the tertiary structures and stabilities of our caviteins was achieved. Perhaps, there are more suitable sequences for the formation of a native-like hetero-TASP, and additional peptide sequences should be designed, and the subsequent hetero-TASPs studied, before this methodology is dismissed. In the future, other hetero-TASPs could be designed and synthesized using the new methodology, especially if one was interested in modeling a specific native four-helix bundle, undoubtedly made up of non-Identical helices. The final studies, which were undertaken for the completion of this thesis investigated the C- and N-capping abilities of the bowl and of the glycine residue, and will be explained in Chapter 4. 252 3.4 Experimental 3.4.1 General All reagents used for the synthesis of the hetero-TASPs were purchased from Aldrich. Short descriptions of the general experimental details and modifications to the studies performed in Chapter 2 will be noted, but for a thorough explanation of reversed-phase HPLC purification, MALDI-MS procedures, CD and ANS binding studies, GuHCl denaturation and NMR experiments, and sedimentation equilibrium analyses see the experimental of Chapter 2. 3.4.2 Peptide Synthesis All of the peptides were synthesized by the solid-phase synthesis using the automated peptide synthesizer. Peptides lg3 and lg2 had been synthesized previously in Chapter 2, and the synthesis of ag3 followed similar experimental procedures as those described for lg2 and lg3. In the case of lg2c, the C-terminal cysteine residue was added on the automated peptide synthesizer containing an S-trityl side chain protecting group. Each peptide was cleaved from the resin with a 2 h treatment with 95 % T F A / H 2 0 except for lg2c for which a mixture of TFA (95 %), H 2 0 (2.5 %) and 1,2 ethanedithiol (2.5 %) was used. Each of the peptides was purified by reversed-phase HPLC as described in Chapter 2. The peptides were characterized for purity by the inspection of a single peak by analytical HPLC (>95 % pure), and using MALDI-MS. 253 Table 3.15. % Yields and MALDI-MS Characterization of the "Activated" Peptides. Peptide % Yield Mass (Da.) Tg2 a 85 1860 lg3a 82 1916 lg2c 68 2037 ag3 87 1706 a peptide sequences from Chapter 2. Peptide 14 (lg2c) After peptide lg2c was removed from the automated peptide synthesizer the free N-terminus was acetylated through a manual treatment of lg2c (-300 mg resin, -100 mg peptide 14,-0.6 mmol) with excess acetic anhydride (3.5 mL) in 1 m L NMP for 1 hour at room temperature. The peptide resin was then filtered through a medium frit filter with D C M . The lg2c peptide was then cleaved off the resin and of protecting groups using a mixture of T F A (95 •%), H 2 0 (2.5 %) and 1,2 ethanedithiol (2.5 %). The last step involved activating the free C-terminus of lg2c by adding lg2c (20 mg, 10 |imol) in 3 mL ethanol to a rapidly stirring solution of 2,2'-Dipyridyl disulfide (12 mg, 55 umol) in 2 mL ethanol. The reaction was stirred at room temperature for 1 hour. The ethanol was reduced to 1 mL in vacuo, and the solution was pipetted onto ice-cold diethyl ether. The resulting solid was re-dissolved in water (1.5 mL) containing 0.1 % TFA, and filtered using a 0.45 |im nylon filter. Subsequent purification of lg2c by RP-HPLC and lyophilization afforded peptide 14 as a white solid (50 mg, 19 %). 254 MS (MALDI, Cinnamic Acid) m/z: 2037 (M + H) + Peptide 15 (ag3) After ag3 was synthesized on the automated peptide synthesizer, chloroacetylation of the free N-terminus was achieved through manual treatment of the resin (350 mg resin, -140 mg peptide 15, -0.6 mmol) with chloroacetylchloride (25 mL, 0.32 mmol, 5 equiv.) and DIPEA (55 pL, 0.32 mmol, 5 equiv.) in D M F for 1 h at room temperature. Subsequent cleavage from the resin and purification by reversed-phase HPLC afforded peptide 15 as a white solid (-100 mg). MS (MALDI, Cinnamic Acid) m/z: 1706 (M + H) + 3.4.3 Hetero-TASP Synthesis Two different experimental approaches were developed for the synthesis of the hetero-TASPs. Approach One described below was used more frequently due to synthetic simplicity. Approach one entailed mixing a solution of cavitand 5 (5 mg, 6.9 jimol, 1 equiv.) with lg3 peptide (33.3 mg, 17.4 jxmol, 2.5 equiv.) in degassed D M F (5 mL) under N 2 . DIPEA (100 pL) was then added in excess and the reaction was left to stir for 4 hours. The crude reaction mixture was then evaporated in vacuo and purified by reversed-phase HPLC to isolate the mono, di (a,b and a,c), tri, and tetrasubstituted caviteins. 255 The masses of the isolated products were then confirmed by MALDI-MS (see Table 3.16), and were characterized by physical studies explained in the following pages. Table 3.16. MALDI-MS Characterization of the lg3 Substituted Hetero-TASP Intermediates. Substituted Hetero- Mass (Da.) TASP Intermediates LG3/3pep 6360 LG3/2pep_ab 4480 LG3/2pep_ac 4480 LG3/lpep 2600 256 Figure 3.25. The Purification of the Crude Reaction Mixture of the lg3 Substituted Hetero-TASPs by Preparatory Reversed-Phase HPLC Using a Gradient of 20-70 % Acetonitrile (with 0.05 % TFA) in Water (with 0.1 % TFA) Over 60 Minutes. Peak(1) = LG3/4pep Peak (2) = LG3/3pep Peak(3) = LG3/2pep_ac Peak (4) = LG3/2pep_ab Peak (5) = LG3/2pep_ac From the HPLC spectrum it was deduced that peaks 3 and 4 correspond to the a,c and a,b disubstituted cavitein variants, respectively. Peak 4 is about two times as large as peak 3 which supports the expected 2:1 product ratio of the a,b:a,c disubstituted cavitein variants. Furthermore, in the ! H NMR spectrum for the 2LG3«2LG2C_ab cavitein in Figure 3.18, the 257 cavitand signal (Ho ut) at -6.1 ppm is split into three signals in a 1:2:1 ratio. A .1:2:1 splitting of the Hout cavitand signal supports the a,b symmetry. In this way, the HPLC and ! H NMR spectra provide evidence for the supposed identities of the a,c and a,b disubstituted products. Figure 3.26. Maldi-MS Spectrum of Peak (1) = LG3/4pep. v - i l B i » : : ma 8241.14 258 Figure 3.27. Maldi-MS Spectrum of Peak 2 = LG3/3pep. 6361.01 •304 4* Figure 3.28. Maldi-MS Spectrum of Peak 3 = LG3/2pep_ac. 4481.84 ::;.iMi.oa::X;:;:;:;«0i.8i:::: 44tSM' :: :: :: 259 Figure 3.29. Maldi-MS Spectrum of Peak 4 = LG3/2pep_ab 4481.11 Iweaie Figure 3.30. Maldi-MS Spectrum of Peak 5 = LG3/lpep. Xstoaa : 1.000 : 2601.96 1U2M -x :x ::iJoi.M:> :: ::i :: :: :: :: :: :»7i.ai :' 1000 ' ' ' 2000 5000 8000 7000 MOO 10000 260 Briefly, approach two involved the protection of two thiol moieties on cavitand 5, followed by peptide attachment at the remaining two thiols, selective deprotection of the protecting groups, and subsequent different sequence peptide attachment. A solution of cavitand 5 (5 mg, 6.9 umol, 1 equiv.) with 2-(4-nitrophenyl)ethyl bromide (4 mg, 17.4 umol, 2.5 equiv.) in degassed D M F (5 mL) was stirring under N2 to Which DJPEA (100 uL) was then added in excess. The reaction was left to stir for 4 hours. The crude reaction mixture was then evaporated in vacuo and not purified further. A crude reaction mixture of protected cavitand 5 (5 mg) and excess lg3 peptide (30 mg) was stirring in degassed D M F under N2. DIPEA (100 uL) was then added in excess and the reaction was left to stir for 4 hours. The reaction mixture was then evaporated in vacuo, dissolved in H 2 O , filtered, and purified by reversed-phase HPLC to isolate the mono, di (a,b and a,c), tri, and tetrasubstituted caviteins with npe protecting groups at the remaining thiol moieties. The isolated a,b and a,c disubstituted caviteins (2 mg, 0.4 umol, 1 equiv.) were then individually subjected to a solution of excess MeONa/MeOH (0.03 mL) in degassed D M F under N2 to cleave the npe protecting groups. After 2 hours the reaction mixture was acidified with 0.1 M HC1 until acidic (acidity tested with litmus paper). The mixture was then evaporated in vacuo, dissolved in H 2 O , filtered and purified by RP-HPLC to afford pure a,b and a,c lg3 disubstituted caviteins, respectively. The isolated disubstituted a,b and a,c hetero-TASP intermediates were then individually subjected to an excess of appropriate peptide to yield the a,b and a,c disubstituted hetero-TASPs, Note the LG3/AG3 family included the synthesis of 3LG3»1AG3 and 1LG3»3AG3 hetero-TASPs in addition to the disubstituted counterparts. The crude products were then purified by RP-HPLC and lyophilized to afford the desired hetero-TASPs and their masses confirmed by 261 MALDI-MS (see Table 3.17). The synthesis of hetero-TASP 16 (2LG3»2LG2_ab) is outlined below and similar procedures were followed for the other hetero-TASPs. Table 3.17. % Yields and MALDI-MS Characterization of the Hetero-TASPs. Cavitein % Yield Mass (Da.) LG2 a 29 8016 LG3 a 34 8240 2LG3»2LG2_ab b 83 8128 2LG3»2LG2_ac b 81 8128 LG2C 32 8424 2LG3»2LG2C_ab b 65 8332 2LG3«2LG2C_ac b 57 8332. AG3 45 7400 3 L G 3 » l A G 3 b 78 8030 2LG3«2AG3_ab b 75 7820 2LG3«2AG3_ac b 77 7820 lLG3»3AG3 b 83 7610 a cavitein sequences from Chapter 2. b hetero-TASPs made from their corresponding lg3 substituted caviteins. 262 Hetero-TASP 16 (2LG3«2LG2_ab) The lg3 a,b disubstituted cavitein intermediate (2 mg, 0.4 uinol, 1 equiv.) and excess peptide lg2 (15 mg, 8.0 umol, 20 equiv.) were dissolved in degassed D M F and stirring under N 2 . DIPEA (50 (imol) was added in excess. The reaction was left to stir for 4 h. The crude reaction mixture was evaporated in vacuo and dissolved into distilled H 2 0 . It was filtered using a 45 |im nylon filter, and purified by reversed-phase HPLC to yield hetero-TASP 16 (3 mg, 83 %). The mass was confirmed by MALDI-MS and determined to be >95 % pure by analytical reversed-phase HPLC. MS (MALDI, Cinnamic Acid) m/z: 8128 (M + H) + 3.4.4 CD Studies All CD spectra were acquired as described in Chapter 2. The GuHCl denaturation experiments were also performed as described in Chapter2. 3.4.5 Analytical Ultracentrifugation (AUC) Experiments Sedimentation equilibrium experiments were performed as described in Chapter 2. No partial specific volume or sedimentation velocity studies were carried out for the cavitiens in this Chapter. Each sample was run at 20 °C at concentrations of 10, 50, and 80 p M and at rotor 263 speeds of 27000, 35000, and 40000 rpm. Tables 3.18, 3.19 and 3.20 list specific experimental details for each experiment. Table 3.18. Experimental Parameters for Sedimentation Equilibrium Experiments for the LG3/LG2 Hetero-TASP Family. Cavitein Cavitein Concentration (MM) Calculated Partial Specific Volume (mL/g) Predominant Species LG2 a 10 0.7814 Monomer LG2 a 50 0.7814 Monomer LG2 a 80 0.7814 Monomer LG3 a 10 0.7770 Monomer L G 3 a 50 0.7770 Monomer L G 3 a 80 0.7770 Monomer 2LG3»2LG2_ .ab 10 0.7802 Monomer 2LG3-2LG2. .ab 50 0.7802 Monomer 2LG3-2LG2. .ab 80 0.7802 Monomer 2LG3»2LG2. _ac 10 0.7802 Monomer 2LG3-2LG2. _ac 50 .0.7802 Monomer 2LG3-2LG2. _ac 80 0.7802 Monomer data from Chapter 2. 264 Table 3.19. Experimental Parameters for Sedimentation Equilibrium Experiments for the LG3/LG2C Hetero-TASP Family. Cavitein Cavitein Concentration (UM) Calculated Partial Specific Volume (mL/g) Predominant Species LG3 a 10 0.7770 Monomer LG3 a 50 0.7770 Monomer LG3 a 80 0.7770 Monomer LG2C 10 0.7740 Monomer LG2C 50 0.7740 Monomer LG2C 80 0.7740 Monomer 2LG3-2LG2C. .ab 10 0.7763 Monomer 2LG3-2LG2C. .ab 50 0.7763 Monomer 2LG3-2LG2C. .ab 80 0.7763 Monomer 2LG3»2LG2C _ac 10 0.7763 Monomer 2LG3»2LG2C _ac 50 0.7763 Monomer 2LG3»2LG2C _ac 80 0.7763 Monomer a data from Chapter 2. 265 Table 3.20. Experimental Parameters for Sedimentation Equilibrium Experiments for the LG3/AG3 Hetero-TASP Family. Cavitein Cavitein Concentration (UM) Calculated Partial Specific Volume (mL/g) Predominant Species L G 3 a 10 0.7770 Monomer LG3 a 50 0.7770 Monomer LG3 a 80 0.7770 Monomer AG3 10 0.7310 Monomer AG3 50 0.7310 Monomer AG3 80 0.7310 Monomer 2LG3»2AG3. .ab 10 0.7563 Monomer 2LG3-2AG3. .ab 50 0.7563 Monomer 2LG3-2AG3. .ab 80 0.7563 Monomer 2LG3»2AG3. _ac 10 0.7563 Monomer 2LG3-2AG3. _ac 50 0.7563 Monomer 2LG3-2AG3. ac 80 0.7563 Monomer data from Chapter 2. Figures 3.31, 3.32, 3.33, 3.34, 3.35 and 3.36 show the remaining plots of the sedimentation equilibrium data for LG2C, 2LG3«2LG2C_ab, 2LG3«2LG2C_ac, AG3, 2LG3«2AG3_ab, and 2LG3»2AG3_ac, respectively. Note that only the data for one concentration at one rotor speed is shown for each of the hetero-TASP as a representative plot since a total of 90 plots were acquired for the caviteins in this Chapter. 266 Figure 3.31. Sedimentation Equilibrium Concentration Distributions of LG2C at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 pM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. 5.000 o • < o ^ 0.0001 » ti o ••a fi C M O l < o I ti o •£ fi (D O ti o U -5.000 ft o o "o o o °, o °0 O O O *> ° " < o o o 0 o ° o° o o o o o 17.500 17.500 12.500 U 7.500 2.500 17.500 ~© v o © O e> O 18.000 18.580 Radius Squared/2 (cmA2) -—s© -«foO $&4> © 18.000 18.500 Radius Squared/2 (cmA2) 2 6 7 Figure 3.32. Sedimentation Equilibrium Concentration Distributions of 2LG3«2LG2C_ab at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 uM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. 17.500 17.750 18.000 18.250 Radius Squared/2 (cmA2) 10.500 CM O i < o I c o QJ u o O 1.000 17.500 17.750 18.000 18.250 Radius Squared/2 (cmA2) 18.500 268 Figure 3.33. Sedimentation Equilibrium Concentration Distributions of 2LG3»2LG2C_ac at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 uM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. ro o i < o o Q 5.000 U 0.000 L -5.000 L 17.500 18.000 18.500 Radius Squared/2 (cmA2) 1.250 17.500 18.000 18.500 Radius Squared/2 (cmA2) 269 Figure 3.34. Sedimentation Equilibrium Concentration Distributions of AG3 at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 uM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. CO o I < o ti o •« ID Q 5.000 L U © 0.000 -5.000 o o o o o ° 0 ° o 0 O o O O o o o ° 0 Q O o o _Q c g 2 2. oo o o 0 O m o o CO O 0 o O ° CP 00 17.250 ° „ O „ O O O 17.750 18.250 Radius Squared/2 (cmA2) 10.750 0.263 I ti o <D o ti o O 0.238 U 0.213 0.188 0 y© "— O * o S * " ^ %*y * - <*> o o c^e{ o o o o 1 1 1 17.250 17.750 10.250 Radius Squared/2 (cmA2) 270 18.750 Figure 3.35. Sedimentation Equilibrium Concentration Distributions of 2LG3»2AG3_ab at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 pM. In the Lower Panei The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. o I < o o P; Q 17.750 18.250 18.750 Radius Squared/2 (cmA2) Figure 3.36. Sedimentation Equilibrium Concentration Distributions of 2LG3»2AG3_ac at a Rotor Speed of 27 000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 pM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. o • < o ti o ••s •p Q 5.000 L 0.000 L -5.000 17.750 o o o 0° o o o 8° o o o o o O O ° o o O O 0 o o 0 o o O O OOO o o© o ° o , o o 0 o © o 18.250 18.750 Radius Squared/2 (cmA2) oo o i < o I c o •£ fi o ti o o 22.500 L 17.500 L 12.500 L 7.500 17.750 18.250 18.750 Radius Squared/2 (cmA2) 272 3.4.6 Nuclear Magnetic Resonance (NMR) Experiments The ID *H NMR (including N-H/D exchange) spectra of the caviteins were recorded as described in Chapter 2. No variable temperature or 2D NMR experiments were performed on the caviteins in this Chapter. 3.4.7 ANS Binding ANS fluorescence experiments were performed as described in Chapter 2, although no fluorescence measurements were made in the presence of GuHCl. 273 3.5 References 1. (a) Mezo, A. R.; Sherman, J.C. J. Am. Chem. Soc, 1999, 121, 8983-8994. (b) Causton, A.S.; Sherman, J.C. Bioorg. Med. Chem. 1999, 7, 23-27. 2. Crick, F.H.C. Acta Crystallogr. 1953, 6, 689-697. 3. (a) Betz, S.F.; DeGrado, W.F. Biochemistry 1996, 35, 6955-6962. (b) Betz, S.F.; Liebman, P. A.; DeGrado, W.F. Biochemistry 1997, 36, 2450-2458. (c) Hill, R.B.; DeGrado, W.F. J. Am. Chem. Soc. 1998,120, 1138-1145. 4. (a) Cusack, S.; Berthet-Colominas, C ; Hartlein, M . ; Nassar, N.; Leberman, R. Nature 1990, 347, 249-255. .'(b) Soisson, S.M.; MacDougall-Shackleton, B.; Schleif, R.; Wolberger, C Science 1997, 276~, 421-425. (c) Biou, V.; Yaremchuk, A.; Tukalo, M . ; Cusack, S. Science 1994, 263, 1404-1410. 5. Gibb, B.C., Mezo, A.R.; Sherman, J.C. Tetrahedron Lett. 1995, 56,7587-7590. 6. See Section 3.4.3 for an explanation of how the a,b and a,c disubstituted cavitein intermediates were identified and distinguished from each other. 7. Royo, M . ; Garcfa-Echeverria, C ; Giralt, E . ; Eritja, R.; Albericio, F. Tetraheron Lett. 1992,33,2391-2394. 8. Gandler, J.R.; Yokoyama, T. J. Am. Chem. Soc. 1984,106, 130-135. 9. (a)Johnson Jr., W.C. Proteins Struc. Func. Genet. 1990, 7, 205-214. (b) Woody, R W . Methods Enzymol. 1995, 246, 34-70. (c) Fasman, G.D. Circular Dichroism and the Conformational Analysis of Biomolecules; Plenum Press: New York, 1996. 10. Johnson, M.L.; Correia, J.J.; Yphantis, D.A.; Halvorson, H.R. Biophys. J. 1981,36, 575-588. 11. Wiithrich, K. NMR of Proteins and Nucleic Acids; Wiley: New York, 1986. 12. (a) Cavanagh, J.; Fairbrother, W.J.; Palmer IJJ, A .G. ; Skelton, N.J. Protein NMR Spectroscopy: Principles and Practice; Academic Press: San Diego, 1996. (b) Roy, S.; Helmer, K.J.; Hecht, M.H. Folding & Design 1997,2, 89-92. 13. Causton, A.S. Ph.D. Thesis, University of British Columbia, Vancouver, Canada, 2001. 14. Freeman, J.; Wallhorn, D.; Sherman, J.C. "Four-Helix Bundle Cavitein Reveals Middle Leucine as Linchpin" (in preparation) 15. Kwok, S .C; Hodges, R.S. J. Biol. Chem., 2003,278, 35248-35254. 274 16. Chakrabartty, A.; Kortemme, T.; Baldwin, R.L. Protein Sci.'1994,3, 843-852. 17; (a) Santoro, M . M . ; Bolen, D. W. Biochemistry 1988,27, 8063 - 8068. (b) Santoro M . M . ; Bolen D.W. Biochemistry 1992,31,4901-4907. 18. (a) Roy, S.; Ratnaswamy, G.; Boice, J.A.; Fairman, R.; McLendon, G.; Hecht, M.H. J. Am. Chem. Soc. 1997, 119, 5302-5306. (b) Raleigh, D P . ; Betz, S.F.; DeGrado, W.F. J. Am. Chem. Soc. 1995,117, 7558-7559. 19. (a) Wiltscheck, R.; Kammerer, R.A.; Dames, S.A.; Schulthess, T.; Blommers, M.J.; Engel, J.; Alexandrescu, A.T. Prot. Sci. 1997, 6, 1734-1745. (b) Greenfield, N.J.; Montelione, G.T.; Farid, R.S.; Hitchcock-DeGregori, S.E. Biochemistry 1998, 37, 7834-7843. 275 CHAPTER 4: Evaluating the C- and N-Capping Efficiency of Glycine* 4.0 Introduction Chapter 2 focused on designing native-like TASPs, and on in investigating native-like structure. Chapter 3 introduced the design and synthesis of the first anti-parallel TASP, as well as, additional hetero-TASPs. The hetero-TASPs incorporated two different kinds of peptides within one bundle that were linked on to the cavitand template separately. Chapter 4 will explore the C- arid N-capping efficiencies of the arylthiol cavitand template, and of glycine. Section 4.1 of this Chapter will outline the background literature and rationale for studying N- and C-capping. Section 4.2 will describe the peptide and cavitein syntheses, and provide a structural analysis of the caviteins. Finally, Section 4.3 will summarize the experimental results, and provide some conclusions for this Chapter. 4.1 Background and Rationale for Studying N- and C-Capping of TASPs As was explained in Chapter 1, the a-helix contains internal hydrogen bonds between the carbonyl of residue i and the amide proton of residue i+4,1 however, this network leaves four carbonyl groups at the C-terminus and four amide protons at the N-terminus without hydrogen bonding partners.2 It has been proposed that there is a preference for certain residues to reside at * " A version of this Chapter will be submitted for publication. Huttunen, H . and Sherman, J.C. An Evaluation into the Efficiency of Glycine as a C- and N-Capping Residue for De Novo Four-Helix Bundle Proteins." 276 the N- and C- termini that can satisfy the hydrogen bonding requirements.2,3 The hydrogen bonding interactions, known as "capping" interactions, provide thermodynamic stability to alpha helices. N-caps have been shown to stabilize monomeric helices by up to 2 kcal/mol.4 Amino acids that serve as useful N-caps include Ser, Asn, Asp and Thr, whereas, Ala, Leu, Val, lie, Tip, Arg, Gin and Glu are rarely found.3'5 By comparison, C-capping interactions at the carboxy terminus occur in less frequently than N-capping interactions of the helices in proteins of known structure. Furthermore, the stabilizing effects of C-capping residues such as His, Lys and Arg were found to be much less than N-caps with stabilization energies of only 0.1-0.4 kcal/mol.4a The most common residue found at the C-cap position is glycine. This project aimed at determining whether the cavitand template or the glycine residue would serve as more efficient N- and/or C-caps. The approach involved synthesizing caviteins having peptides linked to the cavitand template via their N- and C-termini. Furthermore, these caviteins were synthesized with and without capping residues at their respective helix termini. It was hypothesized that the presence of a capping residue would increase the stability of the four-helix bundle. It was further presumed that glycine would serve as the best C-cap, and the cavitand template as the best N-cap. 4.2 Results and Discussion This Section will start by describing the peptide designs and syntheses, which were used . to create the caviteins with and without capping residues. The synthesis of caviteins 29-31 will then be described, followed by a discussion of the characterization data. 277 4.2.1 Peptide Synthesis Peptides 26, 27, and 28 (see Table 4.1 which contains all relevant peptide sequences for this Chapter) were designed following the principles outlined in Chapter 2 (Section 2.2.2). Furthermore, peptides 26, 27, and 28 were synthesized using standard Fmoc techniques on an automated Applied Biosystems peptide synthesizer following literature procedures.6 Peptide 28, which contains a C-terminal cysteine residue for subsequent linkage to the cavitand template via a disulfide bond was synthesized according to the same procedure outlined for peptide 14 in Chapter 3. Table 4.1. Complete Sequences Using One Letter Abbreviated Amino Acids Including Modified Termini for Peptides 6, 7,14, and 26-28. Peptide Peptide Peptide Sequence Number Name 6a lg2 C I C H 2 C O - N H - [ G G - E E L L K K L E E L L K K G ] - C O - N H 2 7a lg3 C 1 C H 2 C 0 - N H - [ G G G - E E L L K K L E E L L K K G ] - C 0 - N H 2 1 4 b , c lg2c C H 3 C O NH-[GEELLKKLEELLKKGGC]-Spy 26 lg2_nocap C1CH 2 C0-NH-[GG-EELLKKLEELLKK]-C0-NH 2 27 lg3_nocap C I C H 2 C O - N H - [ G G G - E E L L K K L E E L L K K ] C O - N H 2 28c lg2c_nocap C H 3 C O - N H [EELLKKLEELLKKGGC]-Spy a peptide sequences repeated from Chapter 2. b peptide sequences repeated from Chapter 3. c lg2c will be attached to the cavitand via a C-terminal cysteine residue; Spy is the S-pyridyl group. 278 4.2.2 Cavitein Synthesis The synthesis of caviteins 29, 30, and 31 entailed the coupling of activated peptides 26, 27, and 28 to cavitand template 5, respectively. Peptides 26 and 27 were linked to cavitand 5 via their N-termini by following literature procedures.7 Peptide 28 was linked to cavitand 5 via its C-terminus by also following literature procedures.713 A complete list of the caviteins and their corresponding names are outlined in Table 4.2. Table 4.2. Names and Sequences for Caviteins 10,11,16 and 29-31. Cavitein Number Cavitein Name Sequence 10a LG2 Cavitand 5-(S-CH 2 CO-NH-[GG-EELLKKLEELLKKG]-CO-NH2)4 l l a LG3 Cavitand 5-(S-CH 2 CO-NH-[GGG-EELLKKLEELLKKG]-CO-NH2)4 16b LG2C Cavitand 5 - (S - [CGG-KKLLEELKKLLEEG]-NH-COCH 3 ) 4 29 LG2_nocap Cavitand 5 - (S-CH 2 CO-NH-[GG-EELLKKLEELLKK]-CO-NH 2 )4 30 LG3_nocap Cavitand 5-(S-CH2CO-NH-[GGG-EELLKKLEELLKK]-CO-NH2)4 31 LG2C_nocap Cavitand 5 - (S- [CGG-KKLLEELKKLLEE]-NH-COCH 3 ) 4 a cavitein sequences from Chapter 2. b cavitein sequences from Chapter 3. Caviteins 29-31 were synthesized with high levels of purity. The caviteins were further characterized by CD and NMR spectroscopy, A U C experiments, and by the binding of a hydrophobic dye monitored by fluorescence spectroscopy. 279 4.2.3 Characterization of the Capping Caviteins Physical experiments were conducted to assess and compare the relative stabilities and tertiary structures for each of the caviteins involved in the study of helix capping. The caviteins of interest in this Chapter were divided into two families those having peptides linked from the N-termini to the cavitand template and C-termini, respectively. 4.2.3.1 Far-UV CD Spectroscopy The far-UV CD spectra for the caviteins at concentrations of -40 p M are shown in Figure 4.1, and were performed in order to assess the extent of a-helicity of our caviteins. CD spectra for each of the caviteins were also obtained at -4 uM to evaluate whether concentration has an effect on the a-helicity of the protein. For all of the caviteins, the CD spectra at concentrations of -4 and -40 'uM overlapped, respectively, and therefore only the high concentration data is shown. Since there was no observed increase in helicity with an increase in concentration this supports that the proteins were monomeric. 280 Figure 4.1. Far-UV CD Spectra for the Capping Caviteins at -40 uM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. -30000 J Wavelength (nm) The characteristic shape of the CD curves for all of the caviteins are indicative of oc-helical structure, however, the oc-helicities of the uncapped caviteins in all cases are lower than their respective capped counterparts. The LG2C cavitein, which has its peptides linked to the template via their C-termini, seems to be the most helical of the six caviteins, while LG2C_nocap is the least helical. For LG3 and LG2 an equivalent loss in helicity was observed when the glycine caps at C-termini of these caviteins were removed, respectively. The last observation, which is distinct when comparing the CD curves, is that the CD curves of LG2C and LG2C_nocap (which have peptides linked to the cavitand via their C-termini) have minima at 208 and 222 nm of similar 281 magnitude, respectively. The arenes of the cavitand template have a positive absorption at -240 nm, which in many cases artificially reduces the minimum at 222 nm for our caviteins. Since, LG2C and LG2C_nocap have peptides linked to the cavitand template via disulfide bonds, perhaps they are further removed from the cavitand template rendering the far-UV signals at 222 nm of LG2C and LG2C_nocap undistorted. Whereas for caviteins LG2, LG3, LG2_nocap, and LG3_nocap it is likely that the cavitand template influences the CD signal at 222 nm, and thus the signal seems to be smaller than expected. The reduction in the oc-helicities of LG2_nocap, LG3_nocap, and LG2C_nocap by 5-10 % compared to their reference caviteins, could be a result of removing their capping residues from their helix termini which aid in holding the ends of the helices intact. Therefore, with regards to the a-helical designs it is fundamental to incorporate a capping residue at the helix termini. 282 Table 4.3. Molar Ellipticity at 222 nm ([$222) for the Gapping Caviteins. Cavitein Concentration (|IM) Experimental \6\222 (deg cm 2 dmol"1) Calculated Maximum [^ 222 (deg cm 2 dmol"1) Percent Helicity (%) LG2 a 39 -20000 -33200 -60 L G 3 a 39 -18000 -33500 -54 L G 2 C b 39 —21000 -33500 -63 LG2_nocap 39 -18000 —32700 -55 LG3_nocap 40 -15000 -33200 -45 LG2C_nocap 39 —15000 -33200 -45 a data taken from Chapter 2. b data taken from Chapter 3. In summary, removing the glycine C-caps from LG2 and LG3, or the glycine N-cap from LG2C does result in a loss in the oc-helicities of the caviteins. Furthermore, the observed loss in helicity when removing the glycine N-cap from LG2C is the most significant compared to the other caviteins suggesting that glycine can indeed serve as an important N-cap. 4.2.3.2 Near-UV C D Spectroscopy The near-UV CD region provides information on the tertiary structures of proteins, and it is the disulfide or aromatic chromophores which absorb in this region. LG2C and LG2C_nocap have disulfide bonds which absorb in the near-UV region, in addition to all of the caviteins 283 having the arenes in the cavitand template, which also absorb in this region. The near-UV CD spectra for the capping caviteins at concentrations of -40 uM are shown in Figure 4.2. Figure 4.2. Near-UV CD Spectra for the Capping Caviteins at -40 uM in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. ••—LG2 .2000 1 Wavelength (nm) The removal of the capping residues in LG2, LG3, and LG2C had little effect on the near-UV CD signal, respectively. The enhanced near-UV signal of LG2 and LG2_nocap could be a result of the peptides being located in closer proximity to the cavitand template, which is the chromophore in the near-UV region. From the signs of the near-UV signals it can be concluded that LG2, LG2_nocap are supercoiling in the same direction, and opposite in direction to that of LG3 and LG3_nocap. Furthermore, one could conclude that LG2C and LG2C_nocap appear to exhibit molten globule-like properties from the absence of a CD signal in the near-UV region. 284 This would be in accordance with other disulfide linked caviteins previously studied within our group, which also exhibited molten globule-like characteristics. The near-UV spectral regions for any of the caviteins lacking capping residues provided no convincing information on a change in the tertiary structural properties of the caviteins. In all cases the capped and uncapped caviteins produced nearly identical near-UV CD curves. 4.2.3.3 Oligomeric States The oligomeric states of the capping caviteins were studied by monitoring the concentration dependence of their unfolding in the presence of GuHCl and by A U C sedimentation experiments. 4.2.3.3.1 GuHCl Denaturation Experiments As was previously mentioned all of the capping caviteins produced concentration independent CD spectra, however this is not sufficient enough data to conclude the existence of monomers in solution; Therefore, the stabilities of the caviteins in the presence of the denaturing salt, GuHCl, were determined at different concentrations. Figure 4.3 displays the unfolding curves of the capping caviteins monitored at 222 nm in the presence of 0-8.0 M GuHCl at -40 uM concentrations. The GuHCl experiments were also run at 4 uM concentrations for each of the caviteins, but the unfolding curves at both concentrations overlapped within experimental error as can be seen with their similar A G ° H 2 O 285 values, listed in Table 4.4. (The unfolding curves obtained for the ~4 \iM concentrations are not included on Figure 4.3 for clarity's sake.) Figure 43. Effect of GuHCl on the Helicity ([$222) of the Capping Caviteins at -40 | i M in 50 mM pH 7.0 Sodium Phosphate Buffer at 20 °C. 1.1 i [GuHCl] (M) As was explained in Chapter 2, the cooperativity of the unfolding reaction can be measured qualitatively by the width and shape of the unfolding transition. All of the unfolding curves for the caviteins outlined in Figure 4.3 support that the caviteins existed as a well-folded structures before the GuHCl salt was added. Furthermore, all the caviteins were completely unfolded by 8.0 M GuHCl, and the unfolding transitions were cooperative. 286 Table 4.4. Guanidine Hydrochloride-Induced Denaturation Data Calculated for the Capping Caviteins. Cavitein Concentration (uM) [GuHCl] 1 / 2 (M) m (kcal/molM) AG°„ 2 o (kcal/mol) LG2 a 39. 5.7 + 0.1 -1.8 ± 0 . 1 -10.4 ± 0 . 3 LG2 a 4 ''• 5.8 ± 6 . 1 -1.8 ± 0 . 1 -1.0.2 ± 0 . 3 L G 3 a 39 5.6 ± 0 . 1 -1.9 ± 0 . 1 -10.8 + 0.4 LG3 a 4 .'- 5.7 + 0.1 -1.8 ± 0 . 1 -10.7 ± 0 . 4 L G 2 C b 39 5.4 ± 0 . 1 -2.3 + 0.1 -11.8 ± 0 . 4 L G 2 C b 3 5.4 ± 0 . 1 -2.2 ± 0 . 1 -11.6 + 0.3 LG2_nocap 39 5.4 + 0.1 -1.9 ± 0 . 1 -10.5 ± 0 . 4 LG2_nocap 3 5.4 ± 0 . 1 -1.8 + 0.1 -10.2 ± 0 . 4 LG3_nocap 40 5.5 ± 0 . 1 -2.2 ± 0 . 1 -10.9 ± 0 . 4 LG3_nocap 4 5.4 ± 0 . 1 -2.2 ± 0 . 1 -10.7 ± 0 . 3 LG2C_nocap 39 5.4 ± 0 . 1 -2.3 + 0.1 -11.2 ± 0 . 4 LG2C_nocap 4 5.3 ± 0 . 1 ., - 2 . 2 ± 0 . 1 -11.1 +0.4 a data taken from Chapter 2. b data taken from Chapter 3. The AG°H 2O values of unfolding for the caviteins at concentrations of 4 uM and 40 uM fall within experimental error of each other, respectively. Higher m values are indicative of a highly cooperative unfolding curve, and hence more native-like structure. Not very much information can be deduced from the m values in Table 4.4, other than that LG2C_nocap and LG3_nocap have the highest m values. It would be difficult to 287 infer that LG2C_nocap and LG3_nocap are more native-like in structure, because they exhibited less enhanced near-UV signals than did LG2, for example. 4.2.3.3.2 AUC Sedimentation Equilibrium Experiments A U C sedimentation equilibrium experiments has been used to analyze the solution behavior of various biological molecules.9 The capping cavitein samples were prepared similar to caviteins outlined in Chapter 2, and were analyzed at concentrations of 10, 50 and 80 p M and at rotor speeds of 27000, 35000 and 40000 rpm. The sedimentation equilibrium data for the capping caviteins were analyzed by NONLIN, 1 0 and the exponential plot of absorbance versus radius for LG2_nocap is shown in Figure 4.4. The exponential plots of absorbance for LG3_nocap, and LG2C_nocap were very similar to that of LG2_nocap, and can be found Section 4.4.5. 288 Figure 4.4. Sedimentation Equilibrium Goncentration Distributions of LG2_nocap at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 uM. In the Lower Panel The Solid Lines Represents Theoretical Fits to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. o i < o a o P fi 2.500 L <^  0.000 -2.500 17.500 © O © 0 O <b 0 o °o o V ° „ o © ° © © ° © O o o © % © O © © © O © 10.000 10.500 Radius Squared/2 (cmA2) 0.250 17.500 18.500 Radius Squared/2 (cmA2) 289 The data for each of the caviteins were also fit to a monomer-dimer, dimer, monomer-trimer, trimer, monomer-tetramer, and tetramer in order to check for the best theoretical fit to the experimental data. In all cases the monomer fits were most accurate, assessed by the even distribution of the residuals about zero as can be seen in the upper panel of Figure 4.4. The results from the sedimentation equilibrium data for all of the capping caviteins are consistent with the CD and GuHCl denaturation data for the presence of a monomeric species in solution. Table 4.5 summarizes the sedimentation data, and the solution conditions for the capping caviteins. Table 4.5. Experimentally Estimated Molecular Weights (MWs) Determined by Sedimentation Equilibrium For the Capping Caviteins at 20 °C in 50 mM pH 7.0 Sodium Phosphate Buffer at Concentrations of 10 pM, 50 pM and 80 p M with Rotor Speeds of 27000, 35000 and 40000 rpm. (note the partial specific volumes used for the determination of the molecular weights are shown in Section 4.4.5) Cavitein Experimentally Calculated M W Predominant Species Estimated M W (Da) (Da) LG2 a 8500 + 600 8016 Monomer L G 3 a 8000 ± 3 0 0 8240 Monomer L G 2 C b 8700 ± 6 0 0 8424 Monomer LG2_nocap 8100 ± 4 0 0 7784 Monomer LG3_nocap 8500 ± 5 0 0 8016 Monomer LG2C_nocap 8300 ± 4 0 0 8200 Monomer a data taken from Chapter 2. b data taken from Chapter 3. 290 4.2.3.4 One-Dimensional (ID) *H N M R Spectroscopy The *H NMR spectra for the capping caviteins are shown in Figure 4.5. Figure 4.5 represents an overlay of the amide regions for the caviteins to more clearly illustrate the spectral region of interest. Sharp and disperse signals in the amide region are indicative of native-like structure.11 Figure 4.5. Expansions of the Amide Regions of 500 MHz ] H NMR Spectra of the Capping Caviteins at -1.5 mM in 10 % D 2 0 , 45 mM Sodium Phosphate Buffer, pH 7.0 at 20 °C. (a) LG2 (b) LG2_nocap (c) LG3 (d) LG3_nocap (e) LG2C (* = cavitand signals) * I I | il M | I I.I I | II M | I I I I | I I I I | I. I I I | I I I I | I n-T | , , I . |.l M . | II I I | I I II | I u I I I I II I I I I I I I I II I M M | I I I I | I I , I | II II | M 9.8 9.4 9.0 8.6 8.2 7.8 7.4 7.0 6.6 6.2 5.8 ppm 291 The ! H NMR spectra of LG2, LG2_nocap, LG3, and LG3_nocap each show -13 distinguishable dispersed amide signals indicative of a well-defined amide backbone with a high content of tertiary structure. The presence 6f only 13 amide signals for these caviteins suggest that many of the amino acid residues are in a degenerate environment and therefore indistinguishable from each other, likely due to the four-fold symmetry of the cavitein. The ! H NMR spectrum of LG2C_nocap could not be acquired due to loss of the sample. It could be expected, however, that the spectrum of LG2C_nocap would resemble the spectrum of LG2G with very broad amide signals indicative of molten globule-like structures. The lH NMR spectra of LG2 and LG2_nocap and of LG3 and LG3_nocap are similar, respectively. It is surprising that the signals in the ! H NMR spectrum of LG2_nocap are sharper than comparable signals in LG2. The ] H NMR spectrum of LG3 remains to be the sharpest of the caviteins, although still slightly less dispersed than LG2 and LG2_nocap. However, as was mentioned in Chapter 2, having a lower dispersion does not preclude a native-like structure, since it has been shown that coiled coil proteins generally exhibit less dispersion in the amide region than do square bundles.12 It can be concluded that the tertiary structures of the caviteins without caps are most likely similar to the tertiary structures of their reference capped counterparts since their *H NMR spectra are virtually indistinguishable, respectively. 4.2.3.5 Hydrogen/Deuterium Amide Exchange As was explained in Chapter 2 proteins have some conformational flexibility or dynamics. It is common to use NMR to study protein amide protons that can be readily exchanged with solvent protons in aqueous media.13 292 A representative stack plot of the 'H NMR amide region for LG2_nocap is shown in Figure 4.6. All of N-H/D exchange data were analyzed for the other caviteins including LG3_nocap and LG2C_nocap, and all of the protection factors are tabulated Table 4.6. Figure 4.6. Stack Plot of 500 MHz r H NMR Spectra Illustrating the Time Dependent Amide H/D Exchange of LG2_nocap in 50 mM pD 5.02 CD 3 COOD/CD 3 COO"Na + Buffer at 20 °C. (a) 4 min (b) 18 min (c) 1 h 3 min (d) 6 h 9 min (* = cavitand signals) * 9.8 9.4 9.0 8.6 8.2 7.8 7,4 7.0 6.6 6.2 5.8 ppm The amide H/D exchanges of all of the capping caviteins were studied over the course of about six hours depending on the cavitein. Many of the amide protons exchanged before the first scan could be acquired (~5min) and are not likely involved in buried hydrogen bonds or then are 293 easily accessible to the solvent. However, a few amides were visible for a few hours, and their calculated protection factors are outlined in Table 4.6. Table 4.6. Tabulated Data from the Amide H/D Exchange Experiments of the Capping Caviteins in a 50 mM pD 5.02 CD3COOD/CD 3COO"Na +Buffer at 20 °C. Cavitein Name Amide Proton Chemical Shift0 (ppm) First-Order Rate Constant (h"1) Half-Life (h) Protection Factord LG2 a 8.5 3.37 x IO"2 20 (6.3 ± 0.5) x 103 LG3 a 8.5 2.97 x 10"2 23 (7.3 ± 0.5) x 103 LG2C b 8.5 1.39 0.50 (1.6 ± 0.2) x 102 LG2_nocap 8.6 8.32 x 10"1 0.80 (2.6 ± 0.3) x l'O2 LG3_nocap 8.6 1.67 x 10"1 4.2 (1.3 ± 0 . 3 ) x 103 LG2C_nocape 8.5 2.08 0.33 (1.0 ± 0.2) x 102 a data taken from Chapter 2. b data taken from Chapter 3. c only the data on the most protected proton is included. d these values are based on the half-life of an unprotected proton at 20 °C at pD 5.02 to be 3.18 x 10"3 h. e the exchange data was collected for LG2C_nocap after which the sample was lost during purification. Examining Table 4.6 it is clear that the protons at -8.5 ppm for LG2_nocap, LG3_nocap, and LG2C_nocap exhibit less protection from exchange, than the protons at -8.5 ppm for LG2, LG3 and LG2C, respectively. From the N-H/D exchange experiments it is clear that by removing the glycine cap at the N- or C-termini of the peptides results in a change in the 294 dynamics of the cavitein. It is likely that removing the glycine cap may leave the hydrophobic core slightly more exposed due to the unraveling of the helices at their termini, and hence more accessible to the solvent. The N-H/D exchange data further support the necessity of including capping residues in our helix designs. 4.3.5 ANS B i n d i n g As was explained in Chapter 2, ANS binding can be used as a diagnostic to probe the tertiary structures of proteins, with molten globule proteins generally binding ANS with a high affinity.14 ANS binding was studied by fluorescence spectroscopy for the capping caviteins, and their emission spectra are shown in Figure 4.7. Under standard experimental conditions15 negligible binding was observed for any of the caviteins. The studies were completed using three different concentrations per cavitein of 50, 100 and 150 uM and the curves were indistinguishable (note Figure 4.7 only show data for the 50 uM cavitein concentration for clarity's sake). 295 Figure 4.7. Fluorescence Emission Spectra of 2 uM ANS in the Presence of 95 % Ethanol, 100 % Methanol, 50 uM Capping Caviteins at 20 °C in pH 7.0 50 mM Sodium Phosphate Buffer. •LG2 •LG3 •LG2C • LG2_nocap •LG3_nocap •LG2C_nocap •EtOH MeOH Wavelength (nm) No binding of ANS was observed for any of the caviteins outlined in Figure 4.7. The amide regions of the *H NMR spectra of LG2, LG3, LG2_nocap, and LG3_nocap suggest native-like tertiary structures, which also correlate well to no binding of ANS observed for these caviteins. From the N-H/D exchange data it was apparent that removing a glycine cap does indeed alter the dynamics of the cavitein, however, it does not necessarily result in the formation of a hydrophobic pocket to which ANS has been suggested to preferentially bind. It is not surprising that no ANS was found to bind to any of the caviteins studied in this thesis, since all other caviteins designed and studied in our group have not bound ANS either. 296 4.3 Summary and Conclusions This Chapter focused on studying the effects of different C- and N-capping moieties. Numerous caviteins were synthesized with peptides linked either via their N o r C-termini including caviteins with and without capping residues at the ends of the peptide helices. The first two caviteins of interest, LG2_nocap and LG3_nocap, contained peptides linked to the cavitand template via their N-termini and were synthesized without caps at their C-termini, respectively. LG2_nocap and LG3_nocap served as uncapped counterparts of reference caviteins, LG2 and LG3, respectively. Many of the caviteins in this thesis were designed in order to further evaluate the native-like characteristics of LG2 and LG3. The LG2C cavitein, on the other hand, was made up of peptides linked via their C-termini to the cavitand template via disulfide bonds. LG2C was the reference compound for LG2C_nocap, which lacked glycine residues at the N-cap positions. It was hypothesized that the caviteins without capping residues would be less stable than their capped reference caviteins, and that the glycine residue would serve as a better C-cap than N-cap. From the experimental data, in general, it appears as though the removal of the glycine capping residue, from either termini, had some effects on the cavitein structure or stability. It is important to assess the oligomeric states of the caviteins, which were found to exist as monomers in solution when studied by CD spectroscopy and A U C . Furthermore, from the CD spectra it was evident that removing the glycine C-caps from LG2 and LG3, or the glycine N-cap from LG2C resulted in an observable loss in the oc-helicities of the caviteins, although all of the uncapped caviteins did remain a-helical. The observed loss in a-helicity when removing the glycine N-cap from LG2C was the most significant compared to the other caviteins suggesting that glycine indeed serves as a significant N-cap. It can be concluded that the glycine •  .' • • 297 '. '• cap is a useful N- or C-cap in the design of our caviteins, and can potentially not only satisfy the hydrogen bonding requirements of the terminal residues at the C- or N-termini, but also plays a crucial role in maintaining the a-helicity of the peptide helices. The GuHCl denaturation experiments show similar cavitein stability toward the chemical denaturant for both LG2_nocap and LG3_nocap, and their capped counterparts, respectively. It Was predicted that the uncapped caviteins would be less stable than their capped counterparts, although this was not clearly observed in the GuHCl denaturation experiments. The tertiary structures of the caviteins were assessed by their amide dispersion in their ] H NMR spectra, by their protection factors determined by N-H/D exchange, and by ANS binding studies. The ! H NMR spectra of LG2, LG2_nocap, LG3, and LG3_nocap each show -13 distinguishable dispersed amide signals indicative of a well-defined amide backbone with a high content of tertiary structure. The spectrum for LG2C was characteristic of a molten globule-like protein and showed no distinct signals in the amide region. Previously studied disulfide linked cavitiens in our group have shown molten globule-like characteristics as well. l b The ! H NMR spectra of LG2 and LG2_nocap and of LG3 and LG3_nocap were very similar, respectively. In this way, it can be concluded that the tertiary structures of the caviteins with and without caps are most likely similar, supported by their virtually indistinguishable ! H NMR spectra. The experimentally determined protection factors for the amide proton at -8.5 ppm in LG2_nocap was less by a factor of about ten than in LG2, and for LG3_nocap the protection factor for a similar proton was less by a factor of about six. The protection factors for LG2C and LG2C_nocap were similar to each other and were lower than the protection factors for any of the other caviteins. Therefore, it is reasonable to conclude that caviteins having peptides linked via their C-termini, e.g. LG2C and LG2C_nocap, have more mobile hydrophobic cores resulting in 298 lower protection factors. Lastly, the dynamics of the caviteins were also assessed by A N S binding studies, and none of the caviteins were found to bind A N S in any considerable amount. Overall this Chapter presented the design and synthesis of four-helix bundles containing peptides with and without N - and C-caps. It was predicted that the caviteins lacking glycine capping residues would be less stable than their capped counterparts, and would exhibit more molten globule-like characteristics, if the glycine caps were indeed responsible for holding the tertiary structures intact. The experimental evidence did support that there is a change in the structure and dynamics of the cavitein when a glycine cap is removed from the termini of the peptides. However, the evidence is not convincing in concluding that these uncapped caviteins are less stable than their capped counterparts, as both the capped and uncapped caviteins have similar A G ° H 2 O values of unfolding. Furthermore, it can be concluded that caviteins, which contained peptides linked to the cavitand template via their C-termini have more mobility within their hydrophobic cores resulting in molten globule-like characteristics, than those caviteins containing peptides linked via their N-termini. A possible explanation for the molten globule-like characteristics of LG2C and LG2C_nocap could be that the disulfide bonds linking the peptides to the template inhibit the proper packing of the hydrophobic core. From all of the experimental data, it is very important to include capping residues in our designs, since the caps not only prevent the helices from unraveling at the ends, but also help hold the tertiary contacts in place. In the future, other amino acid residues could be studied to determine whether a more suitable capping residue exists other than glycine. Furthermore, other caviteins could be designed with peptides linked via their C-termini with different linkages, other than the disulfide bond, between the peptides and the template. 299 4.4 Experimental 4.4.1 General All reagents used for the synthesis of the caviteins were purchased from Aldrich. Short descriptions of the general experimental details and modifications to the studies performed in Chapter 2 will be noted, but for a thorough explanation of reversed-phase HPLC purification, MALDI-MS procedures, CD and ANS binding studies, GuHCl denaturation and NMR experiments, and sedimentation equilibrium analyses see the experimental of Chapter 2. 4.4.2 Peptide Synthesis All of the peptides were synthesized by the solid-phase synthesis using the automated peptide synthesizer. Peptides lg3 and lg2 had been synthesized previously in Chapter 2, and the synthesis of lg2_nocap, and lg3_nocap followed similar experimental procedures as those described for lg2 and lg3. In the case of lg2c_nocap, the synthesis was identical to the synthesis of lg2c outlined in Chapter 3 except that lg2c_nocap lacked an N-terminal glycine residue. The peptides were characterized for purity by the inspection of a single peak by analytical HPLC (>95 % pure), and using MALDI-mass spectrometry. 300 Table 4.7. % Yields and MALDI-MS Characterization of the "Activated" Peptides. Peptide % Yield Mass (Da.) lg2a 85 1860 lg3a 82 1916 lg2cb 68 2037 lg2_nocap 78 1802 lg3_nocap 80 1860 lg2c_nocap 61 1980 a peptide sequences from Chapter 2. b peptide sequence from Chapter 3. 4.4.3 Cavitein Synthesis The LG2_nocap and LG3_nocap caviteins were synthesized as described in Chapter 2, while the synthesis of LG2C_nocap was equivalent to the synthesis of LG2C described in Chapter 3. The masses of the caviteins after purification were confirmed by MALDI-mass spectrometry (Table 4.8), and were characterized by physical studies. 301 Table 4.8. % Yields and MALDI-MS Characterization of the Caviteins. Cavitein % Yield Mass (Da.) LG2 a 29 8016 LG3 a 34 8240 L G 2 C b 32 8424 LG2_nocap 33 7784 LG3_nocap 31 8016 LG2C_nocap 25 8200 a cavitein sequences from Chapter 2. b cavitein sequence from Chapter 3. 4.4.4 CD Studies All CD spectra were acquired as described in Chapter 2. The GuHCl denaturation experiments were also performed as described in Chapter2. 4.4.5 Analytical Ultracentrifugation (AUC) Experiments Sedimentation equilibrium experiments were performed as described in Chapter 2. No partial specific volume or sedimentation velocity studies were carried out for the caviteins in this Chapter. Each sample was run at 20 °C at concentrations of 10, 50 and 80 uM and at rotor ' • 302 speeds of 27000, 35000 and 40000 rpm. Table 4.9 lists specific experimental details for each experiment. 303 Table 4.9. Experimental Parameters for Sedimentation Equilibrium Experiments for the Caviteins. Cavitein Cavitein Concentration (uM) Calculated Partial Specific Volume (mL/g) Predominant Species LG2 a 10 0.7814 Monomer LG2 a 50 0.7814 Monomer LG2 a 80 0.7814 Monomer LG3 a 10 0.7770 Monomer LG3 a 50 0.7770 Monomer L G 3 a 80 0.7770 Monomer L G 2 C b 10 0.7740 Monomer L G 2 C b 50 0.7740 Monomer L G 2 C b 80 0.7740 Monomer LG2_nocap 10 0.7816 Monomer LG2_nocap 50 0.7816 Monomer LG2_nocap 80 0.7816 Monomer LG3_nocap 10 0.7814 Monomer LG3_nocap 50 0.7814 Monomer LG3_nocap 80 0.7814 Monomer LG2C_nocap 10 0.7792 Monomer LG2C_nocap 50 0.7792 Monomer LG2C_nocap 80 0.7792 Monomer a data from Chapter 2. b data from Chapter 3. 304 Figures 4.8 and 4.9 show the plots of the sedimentation equilibrium data for LG3_nocap, and LG2C_nocap, respectively. Note that only the data for one concentration at one rotor speed is shown for each of the caviteins as a representative plot since all of the plots at the remaining concentrations and rotor speeds were similar. The plots for LG2 and LG3 can be found in Chapter 2, and that of LG2C can be found in Chapter 3. 305 Figure 4.8. Sedimentation Equilibrium Concentration Distributions of LG3_nocap at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 p:M. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. CO o I < o o •p P 2.500 L <C 0.000 u -2.500 17.625 o 0 o © O O o 0 0 o ° * o O o o o o o o ° o o o o o 17.875 10.125 18.375 Radius Squared/2 (cmA2) 5.0001 1 1 1 1 1 1 — 17.625 17.875 18.125 18.375 Radius Squared/2 (cmA2) 306 Figure 4.9. Sedimentation Equilibrium Concentration Distributions of LG2C_nocap at a Rotor Speed of 27000 rpm in 50 mM Sodium Phosphate Buffer, pH 7.0, 20 °C at 10 uM. In the Lower Panel The Solid Line Represents a Theoretical Fit to a Monomer Equilibrium. The Upper Panel Represents the Residuals for the Fit. o • < o ti o ••s <L> Q 5.000 g < 0.000 -5.000 17.250 as O © <?a8 o O o O " © 0 o o X© © O 0 ©a o O OO O O^ O O 0 O O o o° ° O o © o o o° © 0 , ° 0 * o° 17.750 18.250 Radius Squared/2 (cmA2) 18.750 0.150 I ti o o C o U 0.000 17.250 17.750 18.250 Radius Squared/2 (cmA2) 10.750 307 4.4.6 Nuclear Magnetic Resonance (NMR) Experiments The ID ! H NMR (including N-H/D exchange) spectra of the caviteins were recorded as described in Chapter 2. No variable temperature or 2D NMR experiments were performed on the caviteins in this Chapter. 4.4.7 ANS Binding ANS fluorescence experiments were performed as described in Chapter 2, although no fluorescence measurements were made in the presence of GuHCl. 308 4.5 References 1. Pauling, L. ; Corey, R.B.; Branson, H R Proc. Natl Acad. Sci. U.S.A. 1951,37, 205 -210. 2. Doig, A.J.; Sternberg, M.J.E. Prot. Sci. 1995, 4, 2247-2251. 3. Marqusee S.; Baldwin R.L. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 8898-8902. 4. (a)Padmanabhan S. et al Nature 1990, 344, 268-270. (b) Merutka G.; Stellwagen E. ; Biochemistry, 1990, 29, 894-898. 5. Presta, L .G. ; Rose, G.D. Science 1988, 240, 1632-1641. 6. Gibb, B.C., Mezo, A.R.; Sherman, J.C. Tetrahedron Lett. 1995,36, 7587-7590. 7. (a) Mezo, A. R.; Sherman, J.C. J. Am. Chem. Soc, 1999, 121, 8983-8994. (b) Causton, A.S.; Sherman, J.C. Bioorg. Med. Chem. 1999, 7, 23-27. 8. (a)Johnson Jr., W . C Proteins Struc. Func. Genet. 1990, 7, 205-214. (b) Woody, R W . Methods Enzymol 1995, 246, 34-70. 9. Hansen, J . C ; Lebowitz, J.; Demeler, B. Biochemistry 1994,33,13155-13163, 10. Johnson, M.L.; Cbrreia, J.J.; Yphantis, D.A.; Halvorson, H.R. Biophys. J. 1981, 36, 575-588. 11. Wuthrich, K. NMR of Proteins and Nucleic Acids; Wiley: New York, 1986. 12. (a) Wiltscheck, R.; Kammerer, R.A.; Dames, S.A.; Schulthess, T.; Blommers, M.J.; Engel, J.; Alexandrescu, A.T. Prot. Sci. 1997, 6, 1734-1745. (b) Greenfield, N.J.; Montelione, G.T.; Farid, R.S.; Hitchcock.DeGregori, S.E. Biochemistry 1998, 37, 7834-7843. 13. (a) Englander, S.W.; Downer, N.W.; Teitelbaum, H. Annu. Rev. Biochem. 1972, 41, 903-905. (b) Raschke, T.M.; Marqusee, S. Curr. Opin. Biotech. 1998, 9, 80-86. (c) Englander, S.W.; Mayne, L . ; Bai, Y.; Sosnick, T.R. Protein Sci. 1997,6, 1101-1109. (d) Jeng, M.-F.; Englander, S.W.; Elove, G.A.; Wand, A.J.; Roder, H. Biochemistry 1990, 29, 10433-10437. (e) Hughson, F.M.; Wright, P.E.; Baldwin, R.L Science 1990, 249, 1544-1548. 14. Kuwajima, K. Proteins Struc. Fund. Genet. 1989, 6, 87-103. 15. (a) Roy, S.; Ratnaswamy, G.; Boice, J.A.; Fairman, R.; McLendon, G.; Hecht, M.H. J. Am. Chem. Soc. 1997, 119, 5302-5306. (b) Raleigh, D.P.; Betz, S.F.; DeGrado, W.F. J. Am. Chem. Soc. 1995,117, 7558-7559. 309 CHAPTER 5: Summary, Conclusions, and Future Work 5.0 Thesis Summary This thesis presented the efforts towards the design and synthesis of native-like de novo proteins using a cavitand macrocycle as template. Chapter 1 introduced the basics of protein folding and structure with an emphasis on a-helices and a-helical motifs. It described the protein-folding problem, which is the uncertain relationship between the primary amino acid sequence and the folded native state, and provided examples of current research groups who use de novo protein design to study the relationship between primary sequence and tertiary structure. The ability to design, synthesize and characterize tie novo proteins can help not only to understand how individual amino acids contribute to the stability and structure of a protein, but has also emerged as a useful tool in the design of proteins with specific functions. One main obstacle encountered in the design and synthesis of native-like de novo proteins is the lack of information available to diagnose the respective native-like properties. Chapter 2 focused on diagnosing native-like structures, in addition to utilizing the standard experimental techniques used to characterize protein tertiary structures. Our approach, which could be used to identify native-like structure, was to design a corresponding protein that would exhibit non native-like properties. Chapter 3 explained the synthetic methodology which was developed to link more than one kind of peptide onto the cavitand template. The caviteins synthesized in our lab to date had resulted from the simultaneous ligation of four peptides strands to the four-thiol moieties on the cavitand template. In this way, the caviteins had been limited to having only one type of peptide 310 sequence attached within one bundle (i.e. four identical helices). Two different synthetic approaches were explored to link more than one kind of peptide on to the cavitand template. Due to synthetic simplicity, approach one was used which involved using ideal experimentally determined peptide equivalents. Having the ability to synthesize hetero-TASPs opened the opportunity to create a variety of de novo proteins, including an anti-parallel four-helix bundle. The hetero-TASP families were each designed with specific goals in mind, for example, attempting to model "knobs-into-holes" hydrophobic packing, to enhance native-like properties, and to study the changing of hydrophobicity within the bundle. Lastly, Chapter 4 investigated the use of the glycine residue and cavitand template as suitable N- or C-caps. Caviteins containing peptides linked to the cavitand template via meir Isl-and C-termini with glycine capping residues at the helix termini were synthesized and characterized. Similar caviteins lacking the glycine caps at the helix termini were synthesized, characterized, and compared in terms of structure and stability with their respective capped counterparts. 5.1 Experimental Conclusions De novo protein design includes the study of protein folding and structure by predicting a '' protein sequence that will adopt a particular fold, and then experimentally testing the prediction. One of the ultimate goals of de novo protein design involves the synthesis of native-like structures, a challenging task. All of the projects in this thesis focused on designing native-like proteins, diagnosing native-like structures, and studying the effects of amino acid or helix substitutions on the Overall tertiary structure. 311 Previously, LG2 and LG3 had been synthesized and studied by CD, GuHCl denaturation and 'H NMR. 1 They were re-synthesized and characterized further in order to investigate their native-like characteristics. The only difference between the LG2 and LG3 caviteins was the linker length between the peptide helices and the cavitand template, containing two and three glycines, respectively. The physical characterization of the two caviteins supported that a single glycine residue within the linker Can have a tremendous effect on the secondary and tertiary structure of the cavitein. The CD spectra of LG2 and LG3 showed a high extent of a-helical structure, with LG2 being slightly more helical than LG3. The intrinsic kinetic stabilities and thermodynamic stabilities of LG2 and LG3 were found to be similar; however, the two caviteins exhibited dissimilar amounts of dispersion in the amide regions of their l H NMR spectra. The proton signals in the amide region of the lH NMR spectrum of LG2 were more dispersed, than respective signals in LG3. Yet the proton signals in the amide region LG3 were considerably sharper than the signals observed for LG2. Mezo's original conclusions suggested that LG2 was more native-like than LG3, because LG2 exhibited greater dispersion in the amide region than LG3. However, from my interpretation, LG3 is more native-like supported by the presence of sharp signals in the amide region. Furthermore, the slightly lower dispersion in the amide region of LG3 could be explained by including the possibility of a coiled coil structure, which generally exhibits less dispersion in the amide region than does a square bundle.2 In this way, it is clear that similar experimental data can be interpreted quite differently. Regardless, both LG2 and LG3 exhibit a high degree of native-like character. From the studies on LG2 and LG3, it can be concluded that the linker between the cavitand template and the peptide helices strongly dictates the overall tertiary structure. The insertion or deletion of a single glycine residue within the linker can 312 entirely change the topology of the tertiary structure. Mezo also synthesized and studied LG1, and found it to exist as a stable dimer.1 Therefore, it is essential that the linker is designed to be short enough to take advantage of the stabilizing effects of the cavitand template, yet be flexible enough to allow the helices to pack into an ideal hydrophobic bundle. The leucine residues of LG2 and LG3 were then replaced with norleucine units in order to investigate the putative native-like structures of LG2 and LG3. It appears that the substitution of leucine residues in LG2 with norleucine units did not greatly affect the tertiary structure of the cavitein (supported by comparable signals in the amide region of their • ' H NMR spectra), suggesting that perhaps LG2 was not as native-like as previously thought, or that NG2 is actually quite native-like. The amide region in the ! H NMR spectrum for LG3 shows -14 relatively sharp and disperse amide signals. Comparing the amide region of LG3 to that of NG3, the amide region of LG3 is sharper than NG3 which has only -10 distinguishable amide signals. An interpretation is that the side chains in LG3 are well packed, characteristic of native-like structure, while the hydrophobic core of NG3 may have a little more motion leading to average signals. There are a few types of mobility which could be considered as a source of slight molten globule-like characteristics of our caviteins. Envision that the helices of the four-helix bundle are rigid, maintaining a high extent of secondary structure, with the side chains within the hydrophobic core moving around. Another possibility could be that the helices are slowly moving, although still keeping the hydrophobic core and tertiary structure intact. One final possibility would be that the helices remain highly a-helical and rigid, but they splay open to expose the hydrophobic core slightly, and then re-close. It is difficult to exclude any of the 313 above mentioned mobilities, although from the results from Chapter 2, it appears as though all four caviteins are quite well-packed. The similar amounts a-helicility between LG3 and NG3 would support that the helices of NG3 remain almost entirely unchanged. However, the slightly more broad signals in the amide region of NG3 would support that the side chains are moving a little more than in LG3. However, since the 'H NMR spectra of LG3 and NG3 are so similar, the hydrobobic packing of these caviteins is also likely similar. In this way, the entropic cost of restricting the norleucine side chains into a similar number of conformations as the leucine side chain is reflected in NG3 being less stable. In this way, the substitution of the leucine residues to norleucine units did to some degree reduce the native-like character of their corresponding caviteins, mainly in LG3. Norleucine amino acids due to their unbranched side chains relative to those of leucine are prone to more conformational freedom, and resulted in slightly more molten globule-like structures, as hypothesized. The CD, NMR, and fluorescence data were consistent with the leucine-based caviteins exhibiting an appreciable extent of native-like structure. The LG2 and LG3 caviteins are typified as some of the most native-like de novo proteins synthesized to date. In Chapter 3, a methodology was developed for the synthesis of four-helix bundles containing different peptides attached within one bundle. This method opened an opportunity to create a variety of hetero-TASPs with specific goals in mind. Three different hetero-TASP families were designed and synthesized, and are listed are follows: LG3/LG2 family, LG3/LG2C family, and the LG3/AG3 family. It was the a,b and a,c substituted hetero-TASPs that were of main interest. The four-helix bundles of the LG3/LG2 caviteins contained two different peptide sequences of differing linker lengths between the cavitand template and the peptide helices. 314 Synthesizing a hetero-TASP having peptides with linkers of differing lengths could provide a means to having the side chains of adjacent helices "out of register". This arrangement was intended to enable the side chains to arrange themselves in such a way to afford the preferred "knobs-into-holes" packing arrangement. If our previously designed caviteins were unable to pack efficiently due to an overpacked hydrophobic core, for example by side chains clashing in the same plane, then creating an LG3/LG2 substituted hetero-TASP could potentially overcome the packing limitations of the initial design. Characterization of the caviteins of this family confirmed that they were all a-helical by their characteristic CD curves, and exhibited cooperative two-state unfolding curves indicative of a native-like tertiary structure. However, the amide regions of the ! H NMR spectra of the LG3/LG2 a,b and a,c substituted caviteins were similar in appearance to the amide region of LG2. The broad signals in the amide regions of 2LG3«2LG2_ab and 2LG3«2LG2_ac reflected a molten globule-like structure. It was perhaps the influence of the lg2 peptide helices that are more closely linked to the cavitand template as compared to lg3 peptide helices, which largely dictate the overall tertiary structure. The design of creating a four-helix bundle with peptides having linkers of different lengths to improve the hydrophobic packing was not as ideal as predicted. The studies on the LG3/LG2 family further confirmed that the linker between the cavitand template and the peptide helices is responsible for the formation of a well-packed hydrophobic core, and hence for creating a native-like cavitein. The second family of hetero-TASPs to be designed was the LG3/LG2C family, which included the first anti-parallel TASP to be synthesized in our laboratory. The four-helix bundles of the anti-parallel caviteins were made up of peptides having different linkers. Two peptides were attached via their N-termini with a thioether linkage and the remaining two peptides were linked via a disulfide bond through their C-termini. Having linkers of differing length was again 315 hypothesized to enable the side chains to be "out of register" within the four-helix bundle, and to pack like "knobs-into-holes". Furthermore, the anti-parallel design was the first time we could study stability of a cavitein with a designed reduced helix macrodipole. From the experimental data, it was determined that all of the caviteins of this family were a-helical in structure, and exhibited cooperative unfolding curves characteristic of a well-defined structure. However, the LG2C cavitein appeared to be molten globule-like in structure from the broad undefined proton signals in the amide region of its ! H NMR spectrum, even though it was slightly more stable toward GuHCl than any of the other caviteins in this family. It seems intuitive that native-like proteins should be more stable than molten globule-like proteins, although clearly this is not always the case. The LG3/LG2C a,b and a,c substituted caviteins were of similar stability to each other and to LG3, although their amide regions of their ! H NMR spectra were not as defined as that of LG3. It can be concluded that the anti-parallel four-helix bundles did not exhibit more native-like behaviour than LG3. The reduced native-like structure of 2LG3»2LG2C_ab and 2LG3»2LG2C_ac could be due to the formation of an inefficient hydrophobic core resulting from the influence of the restrictive disulfide bond linkages between the two lg2c peptides and the cavitand template. The experimental evidence from this family of hetero-TASPs further supports the conclusion that the linkers between the cavitand template and the peptide helices are fundamental in generating an optimal hydrophobic core, and hence native-like structure. DeGrado and coworkers have synthesized numerous de novo anti-parallel four helix bundles which have demonstrated strong native-like characteristics.4 DeGrado and coworkers, however, do not include templates into their designs, and thus they cannot be directly compared to our systems. 316 Lastly, the LG3/AG3 family was created to examine the effects of decreasing the hydrophobicity of an entire peptide helix on the secondary and tertiary structures of the resulting four-helix bundle. The all leucine helices of the LG3 cavitein were one by one replaced with an all alanine helix in the creation of the LG3/AG3 cavitein variants. This design was created to study whether our previous all leucine-based hydrophobic cores were overpacked, and whether we could improve the native-like characteristics of our proteins by reducing the hydrophobicity of the bundle. From the GuHCl denaturation experiments it was shown that the replacement of a single lg3 peptide helix with an ag3 helix resulted in almost a complete loss in thermodynamic stability. Furthermore, substituting additional lg3 peptide helices with ag3 helices resulted in increasingly non-cooperative unfolding curves in the presence of GuHCl. This loss in structural stability was in accordance with a substantial loss in a-helicity deduced by the far-UV CD spectra. The lH NMR spectra of 2LG3»2AG3_ab and 2LG3«2AG3_ac exhibited broad signals in their amide regions, distinctive of molten globule-like structures. Looking at the experimental evidence as a whole, perhaps the resulting molten globule-like characteristics for the ag3-subsituted bundles resulted from the inherently too small alanine side chain inefficiently filling the hydrophobic pocket. The arylthiol cavitand template is approximately 10 A in diameter, and because of the rigidity of the cavitand, it likely controls the range of possible interhelical distances of our peptide helices. This would only be true under the assumption that the linker between the helices and the template is short enough to take advantage of the directing effects of the cavitand. Therefore, using the arylthiol cavitand as a template in our de novo proteins, may in fact limit the choice of hydrophobic residue in our peptide helix designs, in order to create native-like proteins. 317 Unfortunately, the synthesis of the hetero-TASPs described in this thesis did not result in more native-like structures than exhibited by LG3 or LG2. However, the experimental data obtained from the hetero-TASPs did confirm the importance of the protein design features on the construction of native-like tertiary structures. Chapter 4 focused on studying the suitability of glycine as an N- and C-capping residue, respectively. Appropriately designed caviteins including ones with peptides linked via both their N- and C-termini to the cavitand template, and with and without glycine caps, respectively, were synthesized and characterized. No caviteins had previously been synthesized in our lab without caps, and therefore, it was of interest to determine whether the helix caps actually played a role in the overall cavitein structure and stability. All of the experimental evidence acquired from these "capping Caviteins" reflected on the importance of including capping residues in our designs. Specifically the low protection factors for the uncapped caviteins suggested that the caps do aid in holding the ends of the helices from fraying. The glycine cap appeared to be vital in holding the upper part of the helices intact most likely via hydrogen bonding to the unsatisfied amide protons on the adjacent amino acid residues. Since only the glycine residue was examined, it would be interesting to test other commonly occurring N- and C-capping residues to generate some more general conclusions about the suitability and efficiency of capping residues for our caviteins. De novo protein design is a diverse field which includes the study of protein structure and the forces involved in protein folding and stability, the design and synthesis of novel compounds, and the creation of functional proteins. One of the main challenges in the area of protein research is solving the protein-folding problem. De novo design has emerged as a useful tool to study the folding process. From the experimental data acquired from the array of de novo caviteins that were described in this thesis, we have increased our knowledge on the factors that 318 govern protein secondary and tertiary structure, as well as, protein stability. Creating native-like de novo proteins is not an elementary task, and this thesis suggested some design features which are important for obtaining a native-like structure. Specific to our cavitein design is the necessity of an ideal linker between the template and the peptide helices. The effect of removing one glycine from the linker has shown to alter the tertiary structures of our caviteins quite drastically. Therefore, designing and testing different linkers in order to find an optimum one, will always be important in the synthesis of our caviteins. A more general outcome from the research in this thesis is that it appears that it is the formation of the hydrophobic core, which dictates the overall tertiary structure of the cavitein. Many noncovalent forces have been shown to play an important role in protein folding, but it is essentially the creation of an optimal hydrophobic core that is fundamental in generating a native-like protein.5 In this way, a crucial understanding the noncovalent forces of protein folding, and how they can be incorporated into the protein design are vital for the creation of native-like structures. This thesis has unveiled and confirmed many significant aspects of de novo protein design, but has also opened the door to other avenues of research. 5.2 Future Work De novo design has already helped unravel some of the questions underlying protein folding. The ability to design, synthesize and characterize tie novo proteins facilitates the understanding of how individual amino acids contribute to the stability and structure of a protein, and is an efficient approach to designing proteins with specific functions.6 The comprehension 319 of the factors influencing and controlling the tertiary structures and stabilities of our caviteins is well underway, although there remains room for further research. The physical studies that have been described in this thesis provide experimental support for the native-like tertiary structures of our caviteins, however, a crystal structure would inevitably be useful to elucidate the arrangement of the hydrophobic core and of the cavitein as a whole. Very few crystal structures have been obtained for de novo proteins,7 and no crystal structure has ever been acquired for a TASP. Several attempts were made at growing crystals of LG3 and LG2, although only precipitates prevailed. Molecular modelling is another valuable tool for attaining a three dimensional "picture" of a protein, and some modelling has been performed on our caviteins,8 but even more information could be garnered by expanding the scope of such studies. Furthermore, from the projects described in this thesis it would be interesting to continue studying the design and synthesis of the hetero-TASPs. Although the hetero-TASPs synthesized in this thesis did not result in highly native-like caviteins, it could be that a suitable four-helix bundle made up of different peptide helices has not been yet designed. It was predicted that by changing the length of the linkers for two of the peptide sequences within the four-helix bundle (i.e. 2LG3«2LG2_ac) would enhance the packing of the hydrophobic core, and hence increase the native-like properties of the resulting cavitein. Perhaps designing a 2LG3»2LGl_ac hetero-TASP with a larger difference in the linker lengths of the attached peptides would result in a hydrophobic core which is well-packed. It was difficult to conclude what was inefficient in the design of the hetero-TASPs in this thesis, but additional hetero-TASPs should be studied before this concept is dropped. In order to design a more native-like hetero-TASP, molecular dynamics simulations could be performed on a few different hetero-TASP designs, in order to assess the design with 320 the most potential. The molecular dynamics simulations may provide a clearer picture of how the side chains within the hydrophobic core are arranging themselves, and provide insights for a better design. Since it was difficult to determine the precise reasons for the molten globule-like characteristics of the hetero-TASPs analyzed in this thesis, it would be useful to also perform molecular dynamics simulations on these caviteins, in the hopes of obtaining possible explanations for the non-ideal packing of the hydrophobic core. Over the past few years several research projects in our group have focused on the design and synthesis of de novo cavitand-based four-helix bundles,1 and much less time has been devoted to three- or five-helix bundles.9 Research in the area of denovo proteins has also largely focused on four-helix bundles,10 and again less research has been dedicated to larger helical bundle proteins.11 It would be interesting to focus on designing specific peptide sequences suitable for five-helix bundles, and then exploring the effects of amino acid substitutions on the overall structures and stabilities of the resulting caviteins. Furthermore, the strength and specificity of the noncovalent interactions involved in holding a five-helix bundle intact could be evaluated by synthesizing a tetrasubstituted cavitein on a 5-cavitarid. The peptide sequence should be designed specifically for a five-helix bundle, and then by using physical techniques such as A U C to probe whether the binding of a free peptide could be detected to create the five-helix bundle. A U C has been used to study protein-protein interactions and protein-peptide interactions,12 which are essential components of cell cycle control, signal transduction, and intermediary metabolism among other biological functions. A large amount of information has been gathered in our laboratory on the effects of amino acid substitutions on the noncovalent forces involved in contributing to the stability and native-like structure of our caviteins. Since many of the factors controlling four-helix bundle stability and structure are well understood, it would be interesting to move into designing 321 caviteins with specific applications. Other research groups have introduced functionality into their protein models by creating de novo protein ion-Channels,13 proteins binding cofactors,14 peptide receptors,15 and catalysts to name a few.16 With examples of de novo proteins designed with specific functions, it would be interesting to extend this knowledge to our caviteins. 322 5.3 References 1. Mezo, A. R.; Sherman, J C . J. Am. Chem. Soc, 1999,121, 8983-8994. 2. (a) Wiltscheck, R.; Kammerer, R.A.; Dames, S.A.; Schulthess, T.; Blommers, M.J.; Engel, J.; Alexandrescu, A.T. Prot. Sci. 1997, 6, 1734-1745. (b) Greenfield, N.J.; Montelione, G.T.; Farid, R.S.; Hitchcock-DeGregori, S.E. Biochemistry 1998, 37, 7834-'.7843. 3. Crick, F.H.C. Acta Crystallogr. 1953, 6, 689-697. 4. (a) Hill, R.B.; DeGrado, W.F. J. Am. Chem. Soc. 1998, 1201 1138-1145. (b) Betz, S.F.; DeGrado, W.F. Biochemistry 1996, 35, 6955-6962. (c) Betz, S.F.; Liebman, P.A.; DeGrado, W.F. Biochemistry 1997, 36, 2450-2458. 5. DeGrado, W.F.; Lear, J.D. J. Am. Chem. Soc. 1985,107, 7684-7689. 6. (a) Kaplan, J.; DeGrado, W.F. Proc. Natl. Acad. Sci. USA 2004,101, 11566-11570. (b) Mason, J.M.; Arndt, K . M . Chem. Biochem.2004, 5, 170-176. (c) Pinto, Dieckmann, G.R.; Ghandi, C.S.; Papworth, C.G.; Braman, J.; Shaughnessy, M.A.; Lear, J.D.; Lamb, R.A.; DeGrado, W.F. Proc. Natl. Acad. Sci. USA 1997, 94, 11301-11306. (d) Lear, J. D.; Gratkowski, H.; DeGrado, W.F. Biochem. Soc. Trans. 2001, 29, 559-564. (e) Rossi, P.; Tecilla, L. ; Baltzer, P. Chem. Eur. J. 2004,10, 4163-4170. 7. (a) O'Shea, E.K.; Kim, P.S.; Alber, T. Science 1991,254, 539-544. (b) Harbury, P.B.; Zhang, T.; Kim, P.S.; Alber, T. Science 1993, 2t52, 1401-1407. (c) Harbury, P.B.; Kim, P.S.; Alber, T. Nature 1994, 371, 80-83. (d) DeGrado, W.F.; Eisenberg, D. Science 1990, 249, 543-546. (e) DeGrado, W.F.; Eisenberg, D. Science 1993, 259, 1288-1293. 8. Scott, W.R.P.; Seo, E.S.; Huttunen, H.; Wallhorn, D.; Sherman, J . C ; Straus, S.K. Proteins: Struc. Func. Bioinfor. 2006 (in press). 9. Causton, A.S.; Sherman, J.C. Bioorg. Med. Chem. 1999, 7, 23-27. 10. (a) Ho, S.P.; DeGrado, W.F. J. Am. Chem. Soc. 1987, 109, 6751-6758. (b) Tripet, B.; Wagschal, K.; Lavigne, P.; Mant, C.T.; Hodges, R.S. J. Mol. Biol. 2000, 300, 377-402. (c) Wagschal, K.; Tripet, B.; Hodges, R.S. J. Mol. Biol. 1999, 285, 785-803. (d) Kwok, S.C.; Hodges, R.S. J. Biol. Chem. 2003, 278, 35248-35254. (e) Beasley, J.R.; Hecht, M.H. J. Biol. Chem. 1997, 272, 2031-2034. 11. (a) Chin, T . -M.; Berndt, B.D.; Yang, N.C. J. Am. Chem. Soc. 1992,114, 2279-2280. (b) Lutgring, R.; Chmielewski, J. J. Am. Chem. Soc. 1994,116, 6451-6452. 12. Rivas, G.; Stafford, W.; M i n t o n , A . P . M^/ioJ5 1999,19, 194-212. 323 13. (a) Akerfeldt, K.S.; Kim, R.M.; Camac, D.; Groves, J.T.; Lear, J.D.; DeGrado, W.F. J. Am. Chem. Soc. 1992,114, 9656-9657. (b) Pinto, L .H. ; Dieckmann, G.R.; Ghandi, C.S.; Papworth, C.G.; Braman, J.; Shaughnessy, M.A.; Lear, J.D. Proc. Natl. Acad. Sci. USA 1991,94, 11301-11306. (c) DeGrado, W.F. J.Am. Chem. Soc. 1997,119, 3212-3217. (d) Senes, A., Engel, D. E. & DeGrado, W. F. Current Opinion in Structural Biology 2004,14,465-479. 14. (a) DeGrado, W.F. Proc. Natl. Acad. Sci. USA 2000, 97, 6298-6305. (b) Cochran, F. V.; Wu, S. P.; Wang, W.; Nanda, V ; Saven, J. G.; Therien, M . J ; DeGrado, W. F. J. Am. Chem. Soc. 2005 127, 1346-1347. 15. Lombardi, A.; Bryson, J.W.; DeGrado, W.F. J. Am. Chem. Soc. 1997,119, 12378-12379. 16. (a) Severin, K.; Lee, H.; Kennan, A.J.; Ghadiri, M.R. Nature 1997, 389, 706-709. (b) Gibney, B.R.; Rabanal, F.; Reddy, K.S.; Dutton, P. L. Biochemistry 1998, 37, 4635-4643. 324 Appendix A. The one- and three-letter codes for the amino acids mentioned in this thesis, including the twenty commonly occurring amino acids, and their side chains. H O Amino Acid One-Letter Code Three-Letter Code Side Chain: R = Glycine G Gly H Alanine A. ' Ala C H 3 Valine V Val CH(CH 3 ) 2 Leucine L Leu CH 2 CH(CH 3 ) 2 Isoleucine I He isobutyl Serine • s. Ser C H 2 O H Threonine T Thr CH(OH)CH 3 Cysteine c . Cys C H 2 S H Methionine M Met C H 2 C H 2 S C H 3 Proline P Pro see below Aspartic Acid D Asp C H 2 C O O H Asparagine N Asn C H 2 C O N H 2 Glutamic Acid E Glu C H 2 C H 2 C O O H Glutamine Q Gin C H 2 C H 2 C O N H 2 Lysine K Lys C H 2 ( C H 2 ) 3 N H 2 Arginine R Arg (CH 2) 3NHC(NH)(NH 2) Histidine H His see below Phenylalanine F Phe CH 2 Ph Tyrosine ' Y Tyr CH 2 Ph- p a r a OH Tryptophan W Trp see below Norleucine8 Nle C H 2 C H 2 C H 2 C H 3 norleucine does not have a one letter code but is referred to as N in this thesis. R = Proline Histidine Tryptophan 325 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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.831.1-0059769/manifest

Comment

Related Items