Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Probing the stability of a model jellyroll [beta]-sandwich protein Everts, Sarah 2001

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

Item Metadata

Download

Media
831-ubc_2001-0189.pdf [ 7.74MB ]
Metadata
JSON: 831-1.0061491.json
JSON-LD: 831-1.0061491-ld.json
RDF/XML (Pretty): 831-1.0061491-rdf.xml
RDF/JSON: 831-1.0061491-rdf.json
Turtle: 831-1.0061491-turtle.txt
N-Triples: 831-1.0061491-rdf-ntriples.txt
Original Record: 831-1.0061491-source.json
Full Text
831-1.0061491-fulltext.txt
Citation
831-1.0061491.ris

Full Text

P R O B I N G T H E S T A B I L I T Y O F A M O D E L J E L L Y R O L L j3- S A N D W I C H P R O T E I N by S A R A H E V E R T S B.Sc, The University of Guelph, 1998 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Chemistry) We (this thesis as conforming ie re/qtlmjd standard THE UNIVERSITY OF BRITISH COLUMBIA March 2001 ©SarahEverts, 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date DE-6 (2/88) MSc Thesis - Sarah Everts ii A B S T R A C T A major challenge in the field of protein folding involves dissecting the thermodynamic and kinetic pathways by which a polypeptide acquires its unique, minimal free-energy structure. While much research has focussed on a-helical containing proteins, the folding pathways of their all-(5-sheet counterparts remains less well defined. To this end, thermodynamic investigations of the second N-terminal cellulose binding domain of Cellulomonas fimi endoglucanase C (CBDN2) were carried out. This 153 residue domain is composed of 11 P-strands that assemble into two anti-parallel p-sheets with the topology of a jellyroll p-sandwich. Using circular dichroism (CD) spectroscopy, CBD N 2 is shown to undergo fully reversible unfolding induced by both heating and the addition of denaturants. Unfolding of CBD N 2 occurs in a largely pH independent manner. This thesis describes the results of "native state hydrogen exchange (HX)" experiments in non-denaturing concentrations of GuanidineDCl (GuDCl). Amide hydrogen to deuteron exchange provides a means of obtaining residue specific stability, since these rates reflect the structural fluctuations and hydrogen bond disruptions required for solvent exposure. Many of these rates increase when measured in the presence of low GuDCl concentration, since the free-energy barrier against structural fluctuations is decreased, although the protein remains globally folded. The core residues of CBD N 2 exhibit a high degree of protection, exchanging only as a result of global unfolding fluctuations. The stability measured for these unfolding fluctuations corresponds to the global stability measured by CD denaturation. Amide groups in loop regions and near the termini exhibited much less protection, exchanging via low energy fluctuations. There appeared to be no difference in exchange patterns or residue specific stability between residues in sheet-A and sheet-B, suggesting a two-state folding mechanism and the absence of MSc Thesis - Sarah Everts iii any partially unfolded intermediates. Residues most protected from HX were highly conserved among the 8 most sequence homologous proteins. MSc Thesis - Sarah Everts iv T A B L E O F C O N T E N T S Abstract ii Table of Contents iv List of Tables viii List of Figures x Abbreviations xiv Acknowledgements xv Chapter 1 - Introduction 1 1.0 General Considerations 1 1.1 Native State Hydrogen Exchange 3 1.2 The Intrinsic Hydrogen Exchange Rate for Unstructured Peptides5 1.2.1 Theoretical Framework 6 1.2.2 Determination of Unstructured (Intrinsic) Peptide Exchange Rate, k,t 6 1.3 Measurement of HX in Native Proteins 8 1.4 Distinguishing EX1 and EX2 11 1.5 Native State HX in Low Concentrations of Denaturant 12 1.6 Native State HX of Stability Mutants 15 1.7 Survey of the Results of Native State HX Experiments 17 1.7.1 Identification of "Stability Domains"/ Partially Unfolded Forms from Denaturant Native State HX Experiments 17 1.7.2 Native State HX as a Function of Pressure 20 1.7.3 Native State HX as a Function of Temperature 20 1.7.4 Other Mutation Work 20 1.7.5 Native State HX on a Thermophilic Homologue of RNase 21 1.7.6 Tryptophan Indole Hydrogen Exchange 22 1.7.7 Transition Between Conformational Isomers 22 1.7.8 Dimer Formation 23 1.7.9 Exchange of Peptide/Micelles Mixtures 23 MSc Thesis - Sarah Everts v 1.7.10 Native State HX in the Presence of Low Concentrations of TMAO 23 1.8 Native State HX of (3-sheets in Proteins 24 1.9 The Second N-Terminal Cellulose-Binding Domain from Cellulomonas fimi 25 Chapter 2 - Thermodynamic and Kinetic Parameters for the Unfolding Transition of CBDN2 32 2.0 Overview 32 2.1 Materials and Methods 33 2.1.1 Sample Conditions 33 2.1.2 CD Spectroscopy 33 2.1.3 Stopped-Flow Fluorescence Spectroscopy 34 2.1.4 Differential Scanning Calorimetry 34 2.1.5 Temperature Induced Protein Unfolding 37 2.1.6 Analysis of GuDCl and Temperature Denaturation Profiles 39 2.2 Thermodynamic Results 41 2.2.1 Unfolding Reversibility of CBDN2 after Thermal and Chemical Denaturation 41 2.2.2 GuDCl Denaturation as a Function of pD 43 2.2.3 Thermal Denaturation as a Function of pD 45 2.2.4 DSC of CBDN2 as a Function of pD 47 2.2.5 Thermal Denaturation as a Function of GuDCl to obtain ACp 47 2.2.6 Comparison of ACp and &.GV Results 50 2.2.7 Thermodynamic Conclusions 52 2.3 Kinetic Results 53 2.3.0 Refolding of CBDN2 from GuDCl Denaturation 53 MSc Thesis - Sarah Everts vi Chapter 3 - NMR Experiments 56 3.0 Overview of the Native State HX Experiment 56 3.1 Material and Methods 57 3.1.1 Experimental Conditions 57 3.1.2 NMR Acquisition 58 3.1.3 NMR ID Data Processing and Analysis 59 3.1.4 NMR 2D Data Processing and Analysis 59 3.1.5 Spectral Assignment 60 3.1.6 Native State HX Protocol 62 3.1.7 Analysis of the HX Time Series 64 3.2 Native State HX NMR Results 65 3.2.1 General Hydrogen Exchange Results 65 3.2.2 HX trends in p-strands 65 3.2.3 Highly Protected Residues 69 3.2.4 Determination of AG f^DjO) and m^ 72 3.2.5 Low Denaturant Dependent Exchange 78 3.2.6 Large Denaturant Dependent Exchange 80 3.2.7 Comparison of AGHX(D20) and UGV 82 3.2.8 Stability in p-hairpins 82 3.2.9 Hydrogen Bond Breakage 84 3.2.10 The Disulphide Bond 86 3.2:11 Cellooligosaccharide Binding Cleft 88 3.2.12 EX2 and pD Dependent HX.. 88 3.3 Conclusions 91 Chapter 4 - Concluding Remarks 92 MSc Thesis - Sarah Everts vii Appendix A - Derivation of the general hydrogen exchange behavior of native proteins 96 Appendix B - Protein Production and Purification 99 B.l Unlabelled Protein Production 99 B.2 15N Labeled Protein Production 100 Appendix C - Derivation of AGU as a function of AHm, Tm, and ACp 101 Appendix D - Intrinsic peptide exchange rate for CBDN2 103 D.l Code for the C-program that calculates intrinsic peptide exchange rates 104 Appendix E - Normalization Factor for the HX Time Series from the 3 H 5 2 resonance of Leul33 and the H p 2 resonance of Asn47 107 Bibliography 109 MSc Thesis - Sarah Everts viii L I S T O F T A B L E S Table 1.1 A Survey of Protein Native State Experiments HX using NMR 18 Table 2.1 Thermodynamic Results for GuDCl denaturation experiments at 25 ° C . . . 43 Table 2.2 Thermodynamic Parameters from Thermal Denaturation of CBD N 2 as a function of pD 45 Table 2.3 DSC Data for CBD N 2 as a Function of pD 47 Table 2.4 Thermodynamic Parameters from Temperature Denaturation of CBD N 2 as a function of [GuDCl] at pD=5.4 48 Table 2.5 Tabulation of ACp and &GV results for CBD N 2 50 Table 2.6 Refolding Rates for CBD N 2 in Non-Denaturing Concentrations of GuDCl.. 53 Table 3.1 A comparison of sequence conservation in CBD N 2 69 Table 3.2 Native State AGHX(D20) and m values as fit to (3.4-3.7). Those values with no associated fitting error where obtained from a linear fit to two points 75 Table 3.3 Amide hydrogen and amino acid solvent exposure for residues that exchange in a denaturant independent way. The solvent exposure values listed here are averaged over all the CBD N 2 NMR-based ensemble of structures. Highlighted in bold are those residues that exchange with low denaturant dependence, but which have less MSc Thesis - Sarah Everts solvent accessible area exposed than a highly protected residue. Solvent exposure was measured using MOLMOL (Koradi, 2000) 79 Table 3.4 AG^GDp) and m^ for the p-hairpin between B1 and B2 83 Table 3.5 AG^CDp) and m for the p-hairpin between A3 and A4 84 Table D . l Intrinsic H ^ D Amide Exchange Rates for CBD N 2 at 25°C andpD=6.0 103 Ft? Table E . l Normalization Factor for the HX Time Series from the 3 H resonance of Leul33 and the H p 2 resonance of Asn47 108 MSc Thesis - Sarah Everts L I S T O F F I G U R E S Figure 1.1 Examples of (a) rough and (b) smooth conformational or free-energy landscapes (Dill, 1997) 2 Figure 1.2 (a) Base Catalyzed Reaction Equation with (b) mechanism (Englander, 1983) 7 Figure 1.3 1 5N-'H HSQC spectrum of CBDN 2. As exchange proceeds, peak intensities decrease and are fit to a single exponential decay function (insert) 10 Figure 1.4 Boltzman distribution of a folded protein with (a) no denaturant present, and (b) in a solution of denaturant of sufficiently low concentration that the protein remains folded. While all structural conformations are possible, at any given time the native state is the most populated state. When less than half of the protein molecules are in the native state, the protein is considered denatured 12 Figure 1.5 Typical profiles of HX in the presence of denaturant (GuDCl) 13 Figure 1.6 Hypothetical protein with three "stability domains" that sequentially acquire structure with a representation of the effect of mutation in one domain on all domain stabilities as measured by native state HX 16 Figure 1.7 Proteins that exhibit "stability domains" as determined from native state HX in low concentrations of denaturant. These regions of similar HX stability are identified by a uniform color of the backbone, or of probes located on the backbone 19 Figure 1.8 (a) The Greek key motif occurs when four adjacent anti-parallel P-strands are arranged in a pattern similar to the repeating unit of one MSc Thesis - Sarah Everts of the ornamental patterns used in ancient Greece, (b) The jellyroll motif (Branden and Tooze, 1991). (c) Schematic diagram showing the jellyroll P-sandwich topology of CBDN 2. Strands A2-A5 and B2-B5 comprise the jellyroll motif and are colored accordingly. The position of the disulphide between Cys38 and Cysl42 is indicated. The tertiary structure can be envisioned by folding the figure such that sheet B lies below sheet A. The lengths of the strands and loops are not drawn to scale 26 Figure 1.9 (a) Side and (b) Top views of the tertiary structure of CBDN2. The disulphide bond comprised of Cys38 and Cysl42 is colored in yellow and those residues involved in the binding are colored in pink and purple (Tyr91 and Trp49) 28 Figure 1.10 Sequence alignment among the 8 most homologous proteins to CBDN, Completely conserved residues are shaded in black, residues with >75% identity are in purple, and those with > 50% identity in turquoise. Residues involved in p-strands are indicated by a thick line above the sequence. Arrows indicate the Cysteine residues involved in the disulphide bond required for the native structure. The sequence alignment was done using GeneDoc from sequences obtained from the SwissProt database 29 Figure 2.1 (a) Native and GuDCl unfolded CD spectra of CBDN2.(b) Plotdata fits at multiple wavelengths to the GuDCl denaturation profiles, (c) Native and temperature unfolded spectra of CBDN2.(d) Plotdata fits at multiple wavelengths to the temperature denaturation profiles 40 Figure 2.2 Refolding of CBDN2 after (a) temperature and (b) GuDCl denaturation at pD=6.0 42 Figure 2.3 (a) The fraction unfolded of CBDN2 in the presence of GuDCl as a MSc Thesis - Sarah Everts xii Figure 2.4 Figure 2.5 Figure 2.6 Figure 2.7 Figure 3.1 Figure 3.2 Figure 3.3 function of pH at 25 °C. (b) GuDCl unfolding midpoint as a function of pD. (c) AGufD.O) as a function of pD 44 (a) The fraction unfolded during a temperature melt of CBD N 2 as a function of pD. The dependence of (b) T m and (c) AF£m on pD for CBD 46 (a) The fraction unfolded during temperature denaturation of CBD N 2 as a function of GuDCl at pD=5.4. The (b) T m and (c) AF£m as a function of GuDCl at pD=5.4 (c) 48 Comparison of ACp obtained from DSC and CD measurements 51 Refolding of CBD N 2 in non-denaturing concentrations of GuDCl. (a) Stopped-flow refolding traces and (b) residuals for the 3 exponential fit to this data 54 Histogram of the weighted chemical shift difference A5w between (a) 0-1.5M GuHCl and (b) pH 4-8 in the HSQC spectra of CBDN 2. Gaps are proline residues or residues with overlapping frequencies 61 The exchange profile of Ala74 as a function of [GuDCl] 66 P-strands B3 and B4. B4 is an edge strand of p-sheet B that displays measurable exchange for only alternating residues. P-strand B4 amide groups with measurable exchange are indicated by orange spheres, while the solvent exposed residues which exhibit fast exchange are indicated by red spheres. For reference, the orange side chains point in the direction of the hydrophobic core, while the red side chains point towards the protein exterior. This figure was created using the Biosym software, Insight 2000 67 MSc Thesis - Sarah Everts xiii Figure 3.4 Top view of CBDN2. (3-strands containing residues that remain unexchanged after 9 months in 1.6 M GuDCl are colored blue, while less protected strands are colored orange. This figure was created using the Biosym software, Insight 2000 71 Figure 3.5 The free energy of HX, AGHX, as a function of [GuDCl] for CBDN2 residues that exchange (a) with a simple linear denaturant dependence, (b) with a denaturant independence at low GuDCl and a denaturant dependence at high GuDCl, and (c) with small denaturant dependence at low GuDCl and a larger denaturant dependence at high GuDCl 74 Figure 3.6 Distribution of AGHX(D20 ) and m values for CBDN2 76 Figure 3.7 Side and top view of the (a) residue stability, AGHX(D20), of CBDN2 to HX and (b) their relative mHX values indicated by a color gradient. In the side view, the A-sheet is above the B-sheet. The A-sheet is in the foreground of the top view. This figure was created using the Biosym software, Insight 2000 77 Figure 3.8 AGHX(D20) and for (a) (5-Sheet A and (b) (3-sheet B and (c) coil regions mapped on the amino acid sequence 81 Figure 3.9 AGH X profiles of residues that form H-bonds to each other 85 Figure 3.10 (a) Stability profile and (b) structure of Cys38 and Cysl42, the residues comprising the cysteine bond in CBDN2 87 Figure 3.11 The linear fit of log(kex) vs pD at 25 °C mapped onto the CBDN2 sequence 90 MSc Thesis - Sarah Everts xiv A B B R E V I A T I O N S CD CBDN CBDN AG,, AG„ AGH X(D 2Q) DSC GuDCl GuHCl H-bond HSQC HX IPTG n nref NMR pD PUF wt circular dichroism second N-terminal cellulose-binding domain from Cellulomonas firm first N-terminal cellulose-binding domain from Cellulomonas fimi the apparent difference in free energy of a native protein (N) relative to it's unfolded state (U) the apparent difference in free energy for an amide group between states closed and open to exchange the apparent difference in free energy for an amide group between states open and closed to exchange, extrapolated to zero concentration of chemical denaturant in D 20 solution differential scanning calorimetry Guanidine deuterochloride Guanidine hydrochloride hydrogen bond heteronuclear single quantum correlation hydrogen exchange isopropyl P-D-thioglucopyranoside the index of refraction the index of refraction of a reference solution nuclear magnetic resonance pH of samples dissolved in D 20, corrected for the isotope effect partially unfolded form wild type MSc Thesis - Sarah Everts xv A C K N O W L E D G E M E N T S I wish to thank my supervisor Lawrence Mcintosh for his support and direction during my time in his lab. His enthusiasm and commitment to excellent science is inspiring. I appreciated the opportunity to work on a project that deviated from the usual focus of the laboratory. I would like to acknowledge Dr. Louise Creagh for making the D.S.C. measurements. I wish to thank all the members of the Mcintosh laboratory for their help and their humor over the past years. It was a real pleasure to work with people of excellent minds, lively hearts, and musically accommodating ears. To all the dear hearts who came to my rescue when the tunnel seemed too long and too dark, I love and appreciate you deeply. I hope that I may give as much as I have received. Finally, I thank my mother Nikki for being my most resilient supporter, and for encouraging me amidst all my endeavors. Chapter 1 - Introduction 1 C h a p t e r 1 I n t r o d u c t i o n 1 . 0 G e n e r a l C o n s i d e r a t i o n s With the maturation of the genome sequencing projects, and parallel burst of activity in the biotechnology industry, the field of protein science is experiencing not simply an expansion of knowledge, but an ability to expeditiously exploit this knowledge. Anfinsen's observation, that primary sequence dictates protein structure, is routinely exploited by protein engineering research (Creighton, 1993). Yet prediction of the processes by which a protein acquires and maintains this structure is still comparatively nascent and successful manipulation of the folding trajectory is virtually non-existent. In recent years, however, a theoretical framework for the interpretation of protein folding data has been carefully refined, and a host of experiments designed to probe both the kinetics and thermodynamics of the acquisition of protein structure. Protein folding is a kinetic process driven to the preeminent goal of minimizing free energy. While the final structure is generally determined by the lowest energy conformation, the pathway to this fold is not unique, and the rate of folding not necessarily indicative of the protein's global stability. The relationship between global stability and folding kinetics lies in the conformational or "free-energy" landscape of the protein. The conformational landscape is an abstraction illustrating the ensemble of available conformations of a protein as a function of free energy (Dill and Chan, 1997). A "rough" landscape describes a protein that has backbone arrangements wherein slight deviations in conformation lead to radically different free energy (Figure 1.1a). A "smooth" landscape is one wherein backbone conformations grade uniformly with energy (Fig 1.1b). Both types of landscapes may have relatively deeper energy wells leading to intermediates with observable lifetimes. Folding rate is slowed both by the number of local energy minima in the folding landscape, as well as the depth of these minima. Figure 1.1: Examples of (a) rough and (b) smooth conformational or free-energy landscapes (Dill and Chan, 1997). Chapter 1 - Introduction 3 Only recently has there been success in extending the abstraction of a conformational landscape to generalized rate-determining parameters. Baker (2000) has shown that overall protein topology seemingly determines the important features of a protein's free-energy landscape. In particular, contact order, defined as the average sequence separation of residues that are in physical contact, correlates inversely with folding rates of proteins. The salient feature in the relationship between contact and folding rate is that it appears general for proteins from both oc-helical and P-sheet structural subgroups (Baker, 2000). While Levinthal's paradox1 dictates that proteins do not randomly sample all possible conformations along the path(s) to tertiary structure, they may begin their collapse in any conformation. Furthermore, folding proceeds as an iterative sampling of conformational "ensembles", which are groups of partially folded structures with similar energy, en route to the lowest energy conformation. Given that conformational free-energy propels the acquisition of structure, there is much interest in measuring the thermodynamic stability of structural elements within a protein. By acquiring information about the substructure stability of a protein, insights on the contours of and navigation within its conformational landscape can be obtained. An experimental technique that probes structure at a residue specific level and relinquishes information about the structural contacts significant in the stability of the final structure is native-state hydrogen exchange. 1.1 N a t i v e S t a t e H y d r o g e n E x c h a n g e Over 40 years ago Linderstrom-Lang observed that the hydrogen of amide groups in native proteins exchanged with those of the solvent at rates that varied by orders of magnitude (Englander et al, 1997). While surface amide residues are expected to exchange readily with the 1 Levinthal argued that i f a protein sampled all possible structural conformations along the folding pathway then folding rates o f an average sized protein would be ~10 3 0 years (Creighton, 1993). Chapter 1 - Introduction 4 solvent, it is remarkable that amides completely buried in the protein core display similar, albeit slower, behavior. Buried amide residues require disruption of H-bonds, solvent penetration and/or global fluctuations of structure before exchange can occur. The rate of hydrogen exchange (HX) is often a reflection of the local stability of a H-bond, since H-bonds must be disrupted prior to exchange. Interpretation of this stability in terms of the flexibility/rigidity in the region of an amide group can help provide detailed structural information at the residue specific level. HX in a folded protein is typically described by the general two-step model: K kint closed ^ open —exchanged (1-1) In this model an equilibrium exists between the open and closed protein conformations. Exchange with the solvent can only occur from the open form, at the intrinsic rate of exchange of an unstructured peptide, knt. A general expression for the observed exchange rate kex can be derived using a standard eigenvalue method for solving a series of differential equations obtained from (1.1), and is done in Appendix 1. Under conditions where kcl > > kop, corresponding to the predominance of the native state, the observed exchange rate can be written: k„r = * d + * i n t (1.2) Two limiting cases emerge from this model (Englander and Kallenbach, 1983). When the intrinsic exchange rate is much faster than the closing rate, that is k t a t » kcl, (1.2) reduces to: In this so-called EX1 regime, the observed exchange rate is equal to the rate of opening. Those residues that exchange as a result of the same structural fluctuations will have measured exchange rates that are all equal to kop of that fluctuation (and independent of k J . The more advantageous EX2 regime occurs when the refolding rate is much greater than the intrinsic rate Chapter I - Introduction 5 of exchange, that is, kcl » kiM. In the EX2 regime, the observed rate of exchange can be written as: K o p is interpreted as an equilibrium constant for the structural opening during which HX occurs. It's inverse, l/K o p, is a residue specific protection factor, Pop=k x/kjnt. A protection factor of 100 implies that the structure of the protein slows exchange by 100-fold relative to a random-coil state. With an equilibrium constant K o p, the free-energy for structural opening leading to the HX of a given residue can be determined as follows: In order to obtain AG^, measurements of the observed exchange rate in a native protein (kex, section 1.3) and the intrinsic rate of exchange of the unstructured peptide (kint, section 1.2) are required. 1 . 2 T h e I n t r i n s i c H y d r o g e n E x c h a n g e R a t e f o r U n s t r u c t u r e d P e p t i d e s Over the past 40 years extensive descriptions of the factors influencing hydrogen exchange in unstructured peptides have emerged. (Molday et al, 1972; Englander and Mayne, 1983; Bai et al, 1993) k -^-k -Kex - i Kint -Kct (1.4) A G „y = -RT\nK„n = -RTln^ (1.5) int Chapter 1 - Introduction 6 1 .2 .1 T h e o r e t i c a l F r a m e w o r k In aqueous solutions, peptide HX is catalyzed by hydroxyl and hydronium ions, as well as water. k.mt=kA[H + ] + kB[OH} + kw ( 1 6 ) kw is the water catalyzed exchange rate constant, while kA and kB are the acid and base + catalyzed exchange rate constants. kim has a first order pH dependence for specific OH" and H 30 catalysis. Due to the relative magnitude of these effects, acid catalysis dominates for pH < 3, while base catalysis dominates at pH > 3. The constant rate of water catalysis is so small that it is generally negligible except near pH 3. Exchange requires the formation of a hydrogen-bonded encounter complex between catalyst and amide group followed by the partitioning of the hydrogen between the donor and acceptor according to their relative pKa values. This complex forms according to a second order diffusion limited rate constant. One can readily visualize the formation of an encounter complex with OH' ions H-bonded to an amide hydrogen (Figure 1.2). Conversely, the mechanism for acid catalyzed hydrogen exchange remains controversial given the possibility of protonating the oxygen or nitrogen in the amide. Most native state hydrogen exchange experiments are measured near physiological pH, and thus, base catalysis takes precedence. 1 . 2 . 2 D e t e r m i n a t i o n o f U n s t r u c t u r e d ( I n t r i n s i c ) P e p t i d e E x c h a n g e R a t e , k , n t HX rates of unstructured peptides are significantly influenced by the nature of the side chains flanking the amide residue in question. Molday et al first showed that charged groups and Chapter 1 - Introduction 7 polar side chains have additive inductive effects on acid and base catalyzed HX (1972). Later work revealed that the extent to which a side-chain exhibited p-sheet propensity scaled with its Figure 1.2: (a) Base Catalyzed Reaction Equation with (b) mechanism (Englander and Kallenbach, 1983). Chapter 1 - Introduction 8 peptide amide HX rate (with a correlation coefficient of 0.96) (Bai and Englander, 1994). This is a reflection of the side-chain blocking effect. Namely, in shielding the amide group from making H-bonds with solvent HX catalysts, the side-chains also shield the amide group from solvent H-bonds that would compete with secondary structure formation. HX blocking energy increases 5-fold more rapidly than the P-sheet H-bond blocking energy. This is because the solvent catalyst in the HX encounter complex (an ion, for all intents and purposes) has hydration shells that constitute a volume larger than the single water molecule involved in secondary structure competing H-bonds. Comprehensive studies of specific neighboring side-chain effects on amide HX residues have resulted in tabulation of acid and base catalysis, in FLp and D 20, for all amino acid pairs. Thus a modified version of (1.6) that includes the additive inductive electrostatic and steric effects of the flanking side chains, allows calculation of the intrinsic amide HX rates of an unstructured peptide, as a function of pH (or pD) and temperature (Bai et al, 1993). Experimental agreement is good under conditions of low ionic strength. 1 . 3 M e a s u r e m e n t o f H X i n N a t i v e P r o t e i n s Amide hydrogen exchange is most commonly and sensitively measured using Nuclear Magnetic Resonance (NMR) spectroscopy. In addition NH -> ND exchange has been detected using infrared mass spectroscopy and NH—»NT using gel filtration and scintillation counting (Tito et al, 2000; Loftus et al, 1986; Connelly et al, 1993; Arlington and Robertson, 2000a,b). 1 5N-'H Heteronuclear Single-Quantum Correlation (HSQC) NMR experiments produce 2-dimesional spectra with a single peak for every 1 5N atom directly bonded to a 'H atom. The col axis (direct detection) displays the 'HN chemical shift of an amide while the co2 axis (indirect detection) displays the l 5 N chemical shift (Figure 1.3). Since HSQC spectra exhibit resonances Chapter 1 - Introduction 9 for every amide group in a protein, it is a powerful residue specific detection tool for measuring HX. H —> D buffer exchange can be initiated by transferring a protonated protein into a deuterated solution. As exchange occurs, HSQC peak intensity decreases since 'H NMR probes are insensitive to deuterons. By fitting the peak intensity of HSQC spectra taken at incremental time points to a single exponential decay function, HX rates of residues in the protein can be measured (Figure 1.3 insert). Practically speaking, not all rates can be measured because many surface amides exchange too rapidly for accurate analysis, while others have overlapping resonances. The time between transfer into D 20 and acquisition the first spectra is approximately 15 minutes while the time to record an HSQC spectrum of satisfying signal to noise on a 500 MHz spectrometer is also 15 minutes. Acquiring at least 5 spectra to produce enough data points to precisely define a decay curve requires that the amide groups do not completely exchange before the "dead-time" of 1.5 hours. For most globular proteins, only the surface amides exchange within the experiments dead-time. Chapter 1 - Introduction 10 H Chemical shift (ppm) Figure 1.3: ,5N-H HSQC spectrum of CBDN2. As exchange proceeds, peak intensities decrease and are fit to a single exponential decay function (insert). Chapter 1 - Introduction 11 1 . 4 D i s t i n g u i s h i n g E X 1 a n d E X 2 Protection factors, and hence K o p and AG H X can be extracted from protein HX experiments provided that the exchange occurs in the EX2 regime. This requires that the refolding rate of the structural opening leading to exchange is much faster than the intrinsic rate of exchange for all the residues exposed in the opening. That is, kcl » kint. Comparison of global refolding rates of proteins to intrinsic rates calculated as in section 1.2.2 can confirm EX2 exchange. More direct proof of EX2 exchange is obtained by measuring the pH dependence of native protein exchange rates, k . ' ex In the EX2 regime kex oc kiM, and kjnt cc [OH]. Therefore an increase of pH by one unit should lead to an increase in kex by ten-fold. Conversely in the EX1 regime kex=kopen and pH-independent HX is expected. A transition from EX2 exchange to EX1 exchange is seen when the pH is raised sufficiently high to cause k t e to switch from being much less than the refolding rate kcl, to much greater. Determination of the pH dependence of amide exchange (kex) provides a residue level probe of the exchange regime. Since establishing the exchange regime is so integral to the interpretation of native state HX results, considerable effort is often extended to studying the pH dependence of exchange. Arlington and Robertson measured the pH dependence of native state exchange of turkey ovoid mucoid third domain (OMTKY3) to observe the transition from EX2 exchange to that of EX1 near pH ~ 8.5 (1997). It was possible to obtain kop in the EX1 regime and kex in the EX2 regime. They used this information to observe the residue specific refolding rate and found that this was highly correlated to the residue specific stability, while k had no correlation. Chapter 1 - Introduction 12 1 . 5 N a t i v e S t a t e H X i n L o w C o n c e n t r a t i o n s o f D e n a t u r a n t The success of determining residue specific stabilities using (1.6) under native conditions has lead to an extension of this experiment wherein HX rates are measured in solutions containing low concentrations of denaturant. Slowing protein refolding by making more favorable contacts with the hydrophobic core of a protein than would water, denaturants lower the unfolding free-energy, AG 0 . Present in a high enough concentration, denaturants reduce the free-energy of unfolding such that a protein is completely denatured. Present in low concentrations, below the midpoint of unfolding, denaturants spread the Boltzman distribution of protein conformations so that partially unfolded forms are transiently populated (Figure 1.4). In effect this lowers the energy barrier to structural fluctuations within a protein, without denaturing it. Figure 1.4: Boltzman distribution of a folded protein with (a) no denaturant present, and (b) in a solution of denaturant of sufficiently low concentration that the protein remains folded. While all structural conformations are possible, at any given time the native state is the most populated state. When less than half of the protein molecules are in the native state, the protein is considered denatured. Conformational Space The increased structural fluctuations occurring in the presence of low concentrations of denaturant causes overall protein HX rates to augment. Paralleling equilibrium studies of denaturant-induced protein unfolding, phenomenological estimates of the relationship between the free-energy associated with unfolding fluctuations leading to HX and denaturant concentration reveal a linear relationship (Bai et al, 1995). Chapter 1 - Introduction 13 AGHX - AGHX (D20) + mHX [denaturant ] (1.7) where AGH X(D 20) is the stability of a particular residue extrapolated to zero concentration of denaturant, and m is the denaturant dependence of the free energy of exchange. Since denaturant promotes the exposure of non-polar protein surface area, residues with a large denaturant dependence (large mHX) exchange during large ("global") unfolding fluctuations. Under conditions that favor the native state, AGH X(D 20) and m^ for those residues which have the highest protection factors should correspond to the global unfolding free energy, AGV, and mu measured by CD or fluorescence spectroscopy. When the extrapolated AG^CC^O) and values are less than the global unfolding AGV and n^, then exchange occurs as a result of so-called "sub-global" structural fluctuations. Figure 1.5: Typical profiles ofHX in the presence of denaturant (GuDCl). Chapter I - Introduction 14 Many residues in native state HX experiments exchange in with extremely low denaturant dependence (mHX ~ 0), throughout a large range of denaturant concentrations. Often referred to as "local" exchange, denaturant independent exchange is attributed to local structural fluctuations that do not expose new non-polar surface area and correspond to extrapolated AGm(D20) values which are far below the global unfolding free energy (Llinas and Marqusee, 1999). Since fluctuations leading to denaturant independent exchange are low in energy, they do not require stabilization of non-polar residues with denaturant molecules. Usage of the word "local" to describe denaturant independent exchange may be misleading because it implies little overall structural fluctuation is involved in this sort of exchange. Recent simulations on Staphylococcal nuclease suggest that denaturant independent exchange of many residues is consistent with significant surface area exposure of this protein (Wooll et al, 2000). The salient feature is that while Wooll's work indicates that fluctuations leading to denaturant independent exchange can expose a lot of surface area, this does not mean the surface area exposed is non-polar. If the surface area exposed were non-polar, a favorable interaction would lead to a significant HX dependence on denaturant. A given amide residue can exchange "locally" at low concentrations of denaturant, exchange with some denaturant dependence at intermediate concentrations of denaturant ("subglobal" exchange), and exchange "globally" near the unfolding mid point concentration of denaturant. Exchange profiles of residues within the same element of secondary structure, or structural domain, will occasionally superimpose at the "subglobal" and "global" level, leading to similar extrapolated AGHX(D20) and mHX-values. These regions of similar structural stability, called "partially unfolded forms" (PUF's) and "stability domains," have been interpreted as domains that fold synchronously. The formation of these putative structures, identified from their corresponding stabilities, is sometimes interpreted to represent the dominant kinetic folding Chapter 1 - Introduction 15 pathway (Xu et al, 1998). Several issues immediately arise as a result of extracting kinetic conclusions from an equilibrium study. Since native state HX studies probe the increasing population of unfolded forms, if any interpretations of kinetic pathways are to be made, they should be of both the unfolding pathway and the folding pathway. Or more generally, is the folding pathway the same as the unfolding pathway? In principle, if unfolding is reversible, then the protein must traverse the same conformational/energetic landscape regardless of it's initial state (native or denatured) (Creighton, 1992; Dill and Chan, 1997). To fully understand how a protein folds, a complete construction of the conformation landscape is required. The identification of stability domains of native state HX generates a coarse grain contour of this landscape. However, interpretation of the trajectory through the conformational landscape from equilibrium studies is a delicate extrapolation. Specifically, while suggesting that the most stable domains of a protein fold first, followed by the less stable domains, is intuitive, it is also an extrapolation. One route to validation2 of the structural folding pathway is achieved through "stability mutants." 1 . 6 N a t i v e S t a t e H X o f S t a b i l i t y M u t a n t s Consider a hypothetical protein (Figure 1.5) with 3 domains, A, B, and C, where A is the most stable domain, unfolding only when the protein globally denatures. B and C display subglobal stability dependence, where AGB(D20) > AGc(D20). One could suggest therefore that this proteins unfolds in the order of the C domain first (since it is least stable), then B, and finally A. One way to confirm step-wise unfolding is to introduce a mutation in B that increases the stability of this domain by 10 kJ/mol. If native state HX experiments are performed on the mutant, and folding proceeds in the order predicted, then the stability of the C domain should be Chapter 1 - Introduction 16 the same in both the wild type and mutant. If the C domain unfolds first, independently of B, then any stability increase in B will be inconsequential to C. Conversely, the stability of both A and B should increase by 10 kJ/mol, because B unfolds before A. Figure 1.6: Hypothetical protein with three "stability domains" that sequentially acquire structure with a representation of the effect of mutation in one domain on all domain stabilities as measured by native state HX. 2 Other techniques to establish the order of folding events are kinetics of folding/unfolding and pulsed label H X Chapter 1 - Introduction 17 The "stability labeling" method was used to establish that the three partially unfolded forms observed in the native state HX of cytochrome C, and the 2 partially unfolded forms in Rnase T 1 represented sequential intermediates in the folding pathway. (Xu et al, 1998; Chamberlin and Marqusee, 2000) The free-energy impact of a mutation is often evaluated as the difference in stability between mutant and wild type3, or the AAG^. k„y(mui) k;„t ,k„r(mut) AAGHX=AGHX(mut)-AGHX(wt) = -RT\n(-^ l—f^) = -RTln(°x\ J (1.8) *int keAwt) kex(wt) Direct comparison of HX rates eliminates the need to calculate the intrinsic rate of exchange of the unstructured peptide, which is an empirical approximation. 1 . 7 S u r v e y o f t h e R e s u l t s o f N a t i v e S t a t e H X E x p e r i m e n t s 1 .7 .1 I d e n t i f i c a t i o n o f " S t a b i l i t y D o m a i n s " / P a r t i a l l y U n f o l d e d F o r m s f r o m D e n a t u r a n t N a t i v e S t a t e H X E x p e r i m e n t s Native state HX experiments as a function of denaturant have been performed on 11 proteins with a broad distribution of ct-helical and P-sheet secondary structure. Subglobal domains have been identified in six proteins: Cytochrome C, T4 lysozyme (Llinas and Marqusee, 1999), LysN (Alexandrescu et al, 1999), Ribonuclease HI (Chamberlin et al, 1996), Ribonuclease A (Mayo et al, 1993), and Cytochrome b562 (Fuentes and Wand, 1998b) (Figure 1.7). In general, the stability domains constitute an a-helix, or group of a-helices, or a P-sheet. Proteins to display no "stability domains" are: peptostreptococcal protein L (Yi et al, 1997), Chymotrypsin Inhibitor A (Neira et al, 1997), Barnase (Chu et al, 1999), Barstar4 (Bhuyan and Udgaonkar, 1998), P-Lactoglobulin (Ragona et al, 1999), CD2, Chymotrypsin Inhibitor 2 (Itzhaki et al, 1997). In general, these proteins have only global denaturant dependent Chapter 1 - Introduction 18 exchange or denaturant independent exchange at low denaturant that switches to a global dependence at high denaturant. Table 1.1: A Survey of Protein Native State Experiments HX using NMR Protein Variable Partially Unfolded Forms 2° Structure First Author, Publication Date T4 Lysozyme G u D C l yes 7oc,3p Ll inas , 1999 L y s N G u D C l yes 3cc,5P Alexandrescu, 1999 Rnase H G u D C l yes 5a,5(3 Chamberlin, 1996 Rnase H Thermo G u D C l yes 5a,53 Hol l i en , 1999 Cytochrome c G u D C l yes 3 a X u , 1998; B a i , 1995 Cytochrome b562 G u D C l yes 4 a Fuentes, 1998b Barstar G u D C l 4 yes 4a,3(3 Bhuyan, 1998 CD2-mol ten globule G u D C l yes Parker, 2001 R N a s e A G u D C l no M a y o , 1993 peptostreptococcal protein L G u D C l no 4 3 , 1 a Yi ,1997 beta-Lactoglobulin Urea no 9 3 , 1 a Ragona, 1999 C o l d Shock Protein A Urea no 53 Jaravine, 2000 Barnase Urea, G u D C l no 2a ,43 Chu(1999) Chymotrypsin Inhibitor 2 G u D C l no Itzhaki, 1997 Chymotrypsin Inhibitor 2 Temperature no l a , 6 B Itzhaki, 1997 Rnase T l Temperature no la,73 Mul l in s , 1997 Cytochrome c Temperature yes 3 a X u , 1998; B a i , 1995 Cytochrome b562 Pressure yes 4 a Fuentes, 1998a Ovomucoid third domain P H 33 ,1a Arrington, 1997 Alamethicin in micelle a Yee , 1999 Rnase T l Mutation l a , 7 3 Huyghues-Despointes, 1999 C s p A T M A O 53 Jaravine, 2000 Pr ion Protein a,3 Hosszu, 1999 Stefin A Dimer 53,1a Jerala, 1999 Carbonic anhydrase Trp, G u D C l 4a ,73 Jonasson,1999 3 It is generally assumed that mutations affect the folded state of a protein. Chapter 1 - Introduction 19 (b) T4 Lysozyme (Llinas and Marqusee, 1999) (d) LysN (Alexandrescu et al, 1999) Figure 1.7: Proteins that exhibit "stability domains" as determined from native state HX in low concentrations of denaturant. These regions of similar HX stability are identifies by a uniform color of the backbone, or of probes located on the backbone. 4 Bhuyan et al claim PUF's exist in Barstar but the data is inconclusive on this point. Chapter 1 - Introduction 20 1 . 7 . 2 N a t i v e S t a t e H X a s a F u n c t i o n o f P r e s s u r e Hydrostatic pressure was used as a variable for native state HX experiments on apocytochrome b562. Application of hydrostatic pressure initiates the reversible unfolding of proteins in water because the volume of the protein-solvent system is smaller for the unfolded state (Fuentes, 1998a). Native state HX results between 1 bar and 1.1 kbar showed 3 regions of unique stability that corresponded to the stability domains observed in experiments as a function of denaturant. 1 .7 . 3 N a t i v e S t a t e H X a s a F u n c t i o n o f T e m p e r a t u r e Temperature can be used as a variable with which to study native state HX, since the Boltzmann distribution of partially unfolded states spreads out at temperatures close to, but not beyond the unfolding temperature midpoint, Tm. When temperature is used as the perturbant, values for AH and AS can be obtained at a residue specific level. Cytochrome c temperature dependent experiments revealed the same "stability domains" as the native state HX experiments as a function of denaturant (Bai et al, 1995). Stability measurements from temperature native state HX of Chymotrypsin Inhibitor 2 (CI2) revealed only global and denaturant independent exchange, also consistent with earlier denaturant studies (Itzhaki et al, 1997). That the general results of native state HX experiments from diverse denaturants (urea, GuDCl, pressure and temperature) agree ascertains that native state HX results are not an artifact of the experimental conditions. 1 . 7 . 4 O t h e r M u t a t i o n W o r k Aside from inferring the order in which domains of similar stability fold, native state HX on mutant proteins has been used to study the effects of a single mutation on the properties of the entire protein. Chapter 1 - Introduction 21 NSHX experiments on stability (reduction) mutants of CI2 showed that the HX rates of all residues exchanging through "local" fluctuations were unaffected. However, for a small number of residues in the mutant protein that exchanged through "global" fluctuations, the effect of mutation on the HX rate constant was the same as for the equilibrium constant for unfolding. These residues corresponded to the most slowly exchanging residues in the wild-type protein (Neira et al, 1997). Conversely, mutant studies (one located in the core, one on the surface) of Rnase T l show that while the residues in the wild type which exchange "globally" have a A A G ^ = AAGy and a A G ^ A G y , this is also true of many residues that exchange through "local" fluctuations (Huyghues-Despointes et al, 1999). Stability mutants of barnase show the same behaviour as Rnase T l (Perrett et al, 1995). The common result among the mutant studies of CI2, barnase and Rnase T l is that the stability of the global exchangers is affected by the mutation in the same way as the global stability (that is, A G ^ A G y and AAG H X =AAG U ) . One factor likely affecting the results of "local" exchange behavior among these proteins is in the placement of the mutations. While the CI2 mutations5 were embedded in elements of secondary structure, the Rnase T l mutations were placed on the surface an in only one small part of the interior. V 1 . 7 . 5 N a t i v e S t a t e H X o n a T h e r m o p h i l i c H o m o l o g u e o f R N a s e H Thermophilic protein homologues are nature's "stability mutants". Comparison of thermophilic and mesophilic homologues of RNase H have been made using native state HX. RNase H has stability domains that are also detected in the thermophilic protein. For all residues, 5 Mutations in CI2: one I—»V in the hydrophobic core, all others in either a P-strand or an a helix. Chapter 1 - Introduction 22 a proportional increase in HX stability (AG^CDjO)) is observed. In fact when the global stability of the thermophilic protein is decreased to that of the mesophilic protein in the presence of 0.88 M GuDCl, the residue specific stabilities of the homologues are virtually the same. Evolution has selected multiple site-mutations in the mesophilic protein to stabilize the thermophilic protein in a delocalized fashion (Hollien and Marqusee, 1999). 1 . 7 . 6 T r y p t o p h a n I n d o l e H y d r o g e n E x c h a n g e Native state H —» D exchange experiments at 7 tryptophan indole side-chains was performed in human carbonic anhydrase II (HCAII) as a function of low concentrations of denaturant. Indole HX shows similar dependence on denaturant as amide HX. Several of the Trp residues showed little protection to exchange, while others exchanged in a denaturant independent way before shifting to global exchange. Only Trp97, a member of the hydrophobic core, exchanged "globally" for all denaturant concentrations. HCAII native state HX protection factors were on par with those measured on a previously identified "molten globule6" of this protein (Jonasson et al, 1999). This work highlights the lack of clear distinction between a molten globule and the native state. 1 . 7 . 7 T r a n s i t i o n B e t w e e n C o n f o r m a t i o n a l I s o m e r s Human prion protein (PrP) has a propensity to adopt alternative folds, suggesting that several alternative conformations exist in equilibrium with the native state. Native state HX experiments identified a hyper-stable protected core of approximately 10 residues in the region of a disulphide bond in PrPc (Hosszu et al, 1999). The results suggest that the transition route 6 A molten globule refers to a population of partially folded protein, at p H extremes, that maintains indeterminate tertiary contacts. Chapter I - Introduction 23 from the PrPc conformer to the detrimental PrPSc conformer occurs through an unfolded state that retains only this small region of structure, instead of a highly organized intermediate. 1 . 7 . 8 D i m e r F o r m a t i o n The protection factors of stefin A as a monomer were compared to those under conditions where stefin A forms a dimer. Residues with increased protection in the dimer corresponded to those residues whose chemical shifts were affected by the monomer to dimer transition. As protein concentration was incrementally decreased to initiate dimer to monomer transitions rapid exchange of all protected residues in the dimer occurred, suggesting that the transition state is unfolded to a large extent (Jerala and Zerovnik, 1999). 1 . 7 . 9 E x c h a n g e o f P e p t i d e / M i c e l l e s M i x t u r e s Native state HX was performed on Alamethicin, a 20 residue ot-helical antibiotic peptide that inserts into bacterial membranes and self-associates to form ion channels, in the presence of micelles. Protection factor data supports the theory that one face of the helix interacts most strongly with the micelle (Yee et al, 1999). 1 . 7 . 1 0 N a t i v e S t a t e H X i n t h e P r e s e n c e o f L o w C o n c e n t r a t i o n s o f T M A O Trimethylamine N-oxide (TMAO) interactions with exposed peptide backbone are unfavorable, thereby stabilizing native proteins by predominantly increasing the free-energy of the denatured state, while only marginally increasing the free-energy of the native state. Thus increased protection as a result of TMAO reflects changes in the properties of the denatured state as opposed to the native state. Unlike denaturant, which stabilizes elements of a protein Chapter 1 - Introduction 24 (hydrophobic residues), TMAO uniformly affects the entire protein (since it interacts with the backbone), generating results that reflect free-energy properties of the entire protein. Native State HX on cold shock protein A (CspA) in the presence of TMAO lead to increased protection throughout the entire protein. These results suggest that the free-energy of partially and globally unfolded species are increased, which makes exchange competent conformations less accessible from the native state. No stability subdomains were observed as a function of TMAO for CspA (Jaravine et al, 2000). 1 . 8 N a t i v e S t a t e H X o f p - s h e e t s i n P r o t e i n s Many of the proteins studied using the native state FIX were comprised of both a-helices and p-sheets. In all proteins the P-sheets had uniform stabilities or protection factors. That is, no "sub-global" unfolding was detected among strands of the same sheet. P-strands are often among the most stable elements of secondary structure. As a counterpoint, the strands in these proteins often lacked enough probes in the P-strands to justify conclusive statements about specific stability trends of P-sheets in proteins. There are four proteins of majority P-sheet composition that have been studied using native state FIX. The Alexandrescu group has focussed on the OB-fold superfamily, studying both LysN7 and Cold Shock Protein A to determine whether conserved native state structures reflect conserved mechanisms of structure formation. Their results suggest that the most conserved elements of structure are the most stable to unfolding, and that this stability correlates with sequence hydrophobicity (Jaravine et al, 2000; Alexandrescu et al, 1999). P-Lactoglobulin is a 9-stranded P-barrel protein with a terminal a-helix. Native state FfX experiments indicate that it folds in a 2-state process. All the residues in this protein, with the 7 Ant icodon binding domain of the L y s S isoform of E.coli l y s y l - t R N A synthetase Chapter 1 - Introduction 25 exception of the last p-strand in the sequence and the helix, exchange solely as a result of global unfolding fluctuations, while all other probes exchange exclusively in a denaturant independent way (Ragona et al, 1999). Stopped-flow experiments on CD2.dl8, a 9-stranded P-barrel with one a-helix revealed folding of this protein included the transient accumulation of an intermediate. Native state HX in denaturant concentrations which traversed the native-intermediate transition revealed sub-global exchange profiles whose AGHX(D20) and m^ values agreed with those obtained from a kinetic burst phase transition (Parker and Marqusee, 2001). 1 . 9 T h e S e c o n d N - T e r m i n a l C e l l u l o s e - B i n d i n g D o m a i n f r o m C e l l u l o m o n a s f i m i Enzymes that degrade allomorphs of cellulose are comprised of one or more cellulose binding domains (CBDs) as well as a catalytic domain. Classified by sequence similarity into at least 13 families, over 200 CBDs have been identified in cellulolytic bacteria and fungi. CBDs act to associate cellulases with their substrate, and increase catalysis either passively by increasing the local concentration of substrate, or more actively by aiding in the disruption of the structure of cellulose. The Cellulomonas fimi 1,4-P-glucanase CenC has 2 homologous tandem N-terminal binding domains, CBDN 1 and CBDN 2. With 10 and 11 P-strands, CBD N 1 and CBD N 2 have a similar jellyroll P-sandwich topology as elucidated from NMR structure determination (Johnson et al, 1996, Brun et al, 2000). The jellyroll motif consists of four Greek key motifs formed across an imaginary barrel, which flattens to form a sandwich of 2 anti-parallel P-sheets (Figure 1.8). C D 2 . d 1: The N-terminal domain o f rat C D 2 Figure 1.8: (a) The Greek key motif occurs when four adjacent anti-parallel B-strands are arranged in a pattern similar to the repeating unit of one of the ornamental patterns used in ancient Greece, (b) The jellyroll motif (Branden and Tooze, 1991). Chapter 1 - Introduction 2 7 N C (c) A l A2 A5 A3 A4 Figure 1.8: (c) Schematic diagram showing the jellyroll B-sandwich topology of CBDN2. Strands A2-A5 and B2-B5 comprise the jellyroll motif and are colored accordingly. The position of the disulphide between Cys38 and Cysl42 is indicated. The tertiary structure can be envisioned by folding the figure such that sheet B lies below sheet A. The lengths of the strands and loops are not drawn to scale. Figure 1.9: (a) Side and (b) Top views of the tertiary structure of CBDm. The disulphide bond comprised of Cys38 and Cysl42 is colored in yellow and those residues involved in the binding are colored in pink and purple (Tyr91 and Trp49). Chapter 1 - Introduction 29 Figure 1.10: Sequence alignment among the 8 most sequence-similar proteins to CBDN2. (all are endoglucanses, with the exception of one laminase) Completely conserved residues are shaded in black, residues with >75% identity are in purple, and those with > 50% identity in turquoise. Residues involved in B-strands are indicated by a thick line above the sequence. Arrows indicate the Cysteine residues involved in the disulphide bond required for the native structure. The sequence alignment was done using GeneDoc from sequences obtained from the SwissProt database. Chapter 1 - Introduction 30 CBD N 2 has 8 strands forming the jelly-roll motif with three extra strands that lengthen the A and B sheets by 1 and 2 respectively. The cellulose binding site of CBDN 2, identified by cellooligosaccharide induced chemical shift deviations in HSQC spectra, is a groove across the side of one p-sheet which contains 13 residues directly involved in binding. (Figure 1.9) The 8 most similar sequences are from other endoglucanases, and identity among the sugar binding residues and within the P-strands is common. (Figure 1.10). The P-sandwich topology is common in sugar binding domains, as evidenced by prevalence of the OB9-folding motif, the FN310 module, and the IG11 module (Murkin, 1993; Bork 1996). These latter domains are however much smaller than CBD N 2 having on average only 5 P-strands, although a few have similar stabilities (Plaxco et al, 1997). While many kinetic folding studies had focussed on the P-sandwich topology, at the beginning of this study there was correspondingly little known about the residue specific stability, and conformational landscape of this folding topology. CBD N 2 is an ideal representative for the P-sandwich topology, and is a good candidate for the native state HX experiment. The large size of CBD N 2 ensured that multiple residues per secondary element could be probed for greater confidence on results. As well, the excellent stability (and reversibility) to chemical denaturant indicates that HX of CBD N 2 could be measured throughout a large range of non-denaturing [GuDCl]'s. The motivation for this work was to explore the thermodynamic stability of CBDN 2, primarily as a representative of the jellyroll class of P-sheet proteins, but also as a useful protein for industrial cellulose degradation. By identifying the factors that affect global stability of the 9 The (oligonucleotide/oligosaccharide binding)-fold is a 5 stranded P-sheet coiled to form a closed P-barrel. "Fibronectin type III module 1 1 Immunoglobulin "superfamily" Chapter I - Introduction 31 protein, as well as the distribution of stability within the structure, we hope to gain insight on the free-energy landscape of CBD N 2 , and identify the common characteristics (conservation, function) of stable residues. Chapter 2 - Thermodynamic and Kinetic Parameters of CBDI 32 C h a p t e r 2 T h e r m o d y n a m i c a n d K i n e t i c P a r a m e t e r s f o r t h e U n f o l d i n g T r a n s i t i o n o f C B D N 2 2 . 0 O v e r v i e w Measurements of thermodynamic parameters for the global denaturation of CBD N 2 were performed using circular dichroism (CD) spectroscopy. Reversible temperature and [GuDCl]-dependent unfolding transitions of CBD N 2 were measured as a function of pD. Results from parallel Differential Scanning Calorimetry (DSC) unfolding experiments are also included. Analysis of thermodynamic parameters requires that the folding/unfolding reaction is reversible, thus the reversibility of CBD N 2 from GuDCl and temperature denaturation is also shown. Stopped flow fluorescence experiments to obtain CBD N 2 refolding rates in non-denaturing concentrations of GuDCl were performed. These refolding rates are compared to intrinsic unstructured peptide amide hydrogen exchange rates in chapter 3 to confirm EX2 exchange in the native state HX experiments. Chapter 2 - Thermodynamic and Kinetic Parameters of CBDN2 33 2 . 1 M a t e r i a l s a n d M e t h o d s 2 . 1 . 1 S a m p l e C o n d i t i o n s CBD N 2 was produced and purified as described in Appendix B. D 20 was used instead of FLO in all experiments to be consistent with the NMR FIX data. All cited pD are those corrected for the isotope effect (pD = pHm e a s u r e d + 0.4). Potassium phosphate buffer at 5mM (CD and Fluorescence) or 50mM (DSC) was dissolved in D 20 and adjusted to the correct pD using DC1 or NaOD. 1-2 mg/ml of CBD N 2 were used for the CD and DSC experiments while 0.1 mg/ml was used for the fluorescence samples. For the GuDCl denaturation experiments, GuHCl (Sigma) was dissolved in 5 mM potassium phosphate-D20 prior to pD adjustment. GuDCl concentration was determined from the difference between the index of refraction of a reference solution and the solution containing GuDCl, according to (2.1). Measurements to determine [GuDCl] within ±0.05 M were made at room temperature on a Spectronic Instruments refractometer (model number 334610) from the difference between the index of refraction of a reference solution, nrep and a solution containing GuHCl, n (Pace et al, 1992). [GuHCl] = 57.147 (n - nj + 38.68 (n - nj - 91.6 (n - nj (2.1) Thermodynamic parameters are extracted from systems in equilibrium. Thus all samples made for GuDCl denaturation were left to equilibrate overnight, and samples in thermal denaturation experiments were allowed to equilibrate 5 minutes per 1 °C temperature increment. 2 . 1 . 2 C D S p e c t r o s c o p y All CD spectroscopy experiments were performed on a JASCO-720 spectrophotometer. The scan rate for all experiments was 50 nm/min with a response time of 2 seconds to yield a Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDN2 34 response wavelength of 1.71 nm, in keeping with manufacturer's suggestions. The bandwidth was chosen to be 2.0 nm with a resolution of 0.5 nm. Six spectra were acquired for every condition variable and averaged by the acquisition software. 2 . 1 . 3 S t o p p e d - F l o w F l u o r e s c e n c e S p e c t r o s c o p y Refolding kinetics of CBD N 2 were monitored using a stopped-flow apparatus from Bio-Logic (France). The fluorescence detector was a SFM-400/S and the four syringe mixer was a MOS-450/AF-CD. The refolding buffer was located in 3 syringes, while the denatured protein was placed in the fourth. The cuvette was TC-100/10 with a dead volume of 30.2 pL. The maximum dead-time for the volumes used in these experiments was 2.3 ms (corresponding to 3 refolding buffer samples of 150 pL, a denatured protein sample of 70 pL, and a mixing time of 40 ms). Two scans were averaged for every refolding concentration. 2 . 1 . 4 D i f f e r e n t i a l S c a n n i n g C a l o r i m e t r y DSC measurements were performed with a Calorimetry Sciences Corp. (Spanish Fork, UT) model 4207. Background excess thermal power scans were obtained with buffer in the sample cell and were subtracted from the scans for each 0.5 mL sample solution containing 1.5 to 2.0 mg/mL CBDN 2. For each pD, 4-5 thermograms were obtained. The scan rate used was 1 °C/min Reversibility of unfolding was determined by reheating the sample after cooling in the calorimeter. 2 . 1 . 5 D e n a t u r a n t I n d u c e d P r o t e i n U n f o l d i n g Many proteins exhibit two-state unfolding when monitored by conventional spectroscopic techniques (Pace et al, 1992). Indeed the extraction of thermodynamic unfolding parameters is Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDN2 35 most often performed using equations that assume two-state behavior, with validation of this assumption based on the efficacy of the fit. In a two-state analysis model, the sum of the fraction of folded (fF) and unfolded (fw) protein is normalized to 1: fu+fF = l (2-2) Thus any observed signal y during an unfolding reaction can be written as: y = yFfF+yufu (2.3) where yF and yv represent the signal characteristic to the folded and unfolded state, respectively. In general there is a linear dependence on denaturant concentration of the folded and unfolded signal and therefore: yv = bv + a ^ denaturant] yF =bF + aF [denaturant] (2.4) where bu / F is the extrapolated unfolded/folded signal at 0 M denaturant and a^ is the dependence of the unfolded/folded signal on denaturant. Combining (2.2) and (2.3) to solve for fv: yF-y (2.5) fu = y F - y u An equilibrium constant K for the unfolding transition is the ratio of unfolded versus folded protein and can be written: Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDN2 36 Since this expression neglects the presence of denaturant in the solution, K is not a full, but an apparent, equilibrium constant. The free energy of unfolding can be elicited from the thermodynamic expression: AG(denaturant) = -RT ln(K) f AG(denaturant) = -RTln(——) (2.7) For denaturant unfolding reactions, the free energy of unfolding is observed to vary with denaturant concentration in a linear fashion (Pace et al, 1992): AG (denaturant) = AGV (H20) + mv [denaturant] (2.8) AGV (FLO) or AGU (D20) is the free energy of unfolding in the absence of denaturant, and mv is the dependence of AG (denaturant) on [denaturant]. mv is often interpreted as a measure of the protein surface area exposed to the denaturant upon unfolding. Setting (2.7) equal to (2.8), and substituting in (2.4) and (2.5), one can derive that the observed signal for a denaturant induced unfolding of a protein is: y = bF + aF [den] + (bv + au [derive ( - A G ^ ^ O ) - * ! ) " 1 + e RT ( - A M " 2 0 ) - * 1 ) Kl (2.9) Non-linear least squares fitting of y versus [denaturant] allows for the extraction of AGu(F£20) (or in this case, AGu(D20)) and mu. Chapter 2 - Thermodynamic and Kinetic Parameters of CBDN2 37 2 . 1 . 5 T e m p e r a t u r e I n d u c e d P r o t e i n U n f o l d i n g Two state analysis of thermal denaturation profiles proceeds inititally in the same way as denaturant unfolding analysis. A full equilibrium constant is extracted from the spectroscopic signal changes (2.6), and is inserted into the equation for free-energy: AG(T) = -RT In K(J) = -RT l n ( ^ ) ( 2 1 Q ) In denaturant unfolding reactions, extraction of the unfolding free-energy in native conditions, AGU (I-LO), is made simple by the empirical observation that AG is linearly dependent on denaturant concentration (2.8). In thermal unfolding reactions, however, the free energy varies non-linearly with temperature because the heat capacity of a folded and unfolded protein is significantly different. To obtain the free energy of unfolding from a thermal denaturation profile, it is necessary to independently measure the change in heat capacity ACp. The free-energy of the temperature unfolding transition (derived in Appendix C) is then written: AG, (T) = AHm (1 -1-) - ACp (Tm -T + T ln(^)) < 2' 1 1 > m m where T m is the temperature at the midpoint of the unfolding transition, AHm is the enthalpy change at this midpoint, and ACp is the change in heat capacity of the folded and unfolded protein. The assumption in (2.11) is that the unfolding process is two-state. Unlike GuDCl denaturation experiments, however, the equilibrium constant is not an apparent, but a full one. Without foreknowledge of ACp, AGV cannot be extracted directly from a temperature denaturation experiment; however, the midpoint unfolding temperature Tm , and enthalpy change, AHm, can in the following manner. Free-energy is related to enthalpy according to (2.12): Chapter 2 - Thermodynamic and Kinetic Parameters of CBDl AG(T) - AH(T) - TAS 38 (2.12) At the midpoint temperature, AG(Tm)=0. Therefore from (2.12): A C A H M = — (2.13) (2.12) can be re-written as: A G ( 7 - ) . A H ( l . f ) 4 i Setting (2.14) equal to (2.10), and including the linear dependence of the folded and unfolded signal with temperature, the signal observed during a temperature unfolding reaction can be written as: = bF+aFT + (bu+auT)e y A H ) ) l + e R T T m (2.15) where bF=folded signal at T=0 aF=slope of folded signal on T bu=unfolded signal at T=0 au= slope of unfolded signal on T T=temperature of the sample Tm=melting temperature R=Gas constant By performing a non-linear least squares fit to (2.15), AH and T m can be extracted from spectroscopic data. Since a measure of the free-energy of unfolding at 25 °C is desired, determination of the change in heat capacity due to temperature denaturation is necessary for substitution into (2.11). Chapter 2 - Thermodynamic and Kinetic Parameters of CBDt N2 39 ACp, defined as the variation of enthalpy change with temperature, or: d(AH) dT (2.16) and can be measured from the slope of AH versus T m graph for temperature denaturation data taken at different pH values, or in the presence of GuDCl. The assumption is that while the different solution conditions affect Tm, ACp is constant. AC p is extracted from the strong temperature dependence of the enthalpy (2.16). 2 . 1 . 6 A n a l y s i s o f G u D C l a n d T e m p e r a t u r e D e n a t u r a t i o n P r o f i l e s An entire CD spectrum was obtained at every temperature or [GuDCl] increment. As a result, the unfolding transition of CBD N 2 could be monitored at multiple wavelengths. The greatest difference in CD signal during temperature denaturation occurred between 235-240 nm. In the GuDCl denaturation experiments, the greatest difference in CD signal of denatured and native protein occurred between 222-228 nm. Since GuDCl has significant CD signal, a corresponding buffer blank was subtracted from every GuDCl denaturation spectra. Non-linear least squares fitting of the unfolding traces to equations (2.9) or (2.15) was performed in Plotdata to extract AGV (D20) and m^  for GuDCl unfolding experiments or AHm and T m for the temperature denaturation experiments. Thermodynamic parameters from each wavelength fit were then averaged. The standard deviation of the set of standard errors from the fit parameters was taken to be the error in the averaged parameter (Taylor, 1982). Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDj 40 Figure 2.1: (a) Native and GuDCl unfolded CD spectra of CBDm. (b) Plotdata fits at multiple wavelengths to the GuDCl denaturation profiles, (c) Native and temperature unfolded spectra of CBDN2. (d) Plotdata fits at multiple wavelengths to the temperature denaturation profiles. Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDN2 41 2 . 2 T h e r m o d y n a m i c R e s u l t s 2 . 2 . 1 U n f o l d i n g R e v e r s i b i l i t y o f C B D N 2 a f t e r T h e r m a l a n d C h e m i c a l D e n a t u r a t i o n CBD N 2 unfolds reversibly upon thermal denaturation. Refolding from 80 °C at pD=6.0 is shown in Figure 2.2. The folded and refolded traces do not perfectly overlap, presumably because some protein aggregation occurs at high temperature. Nevertheless, the CD spectral profile of both folded and refolded CBD N 2 are the same and distinct from that of the unfolded spectra. Refolding from GuDCl denaturation was established by taking a spectra of CBD N 2 in 5 M GuDCl, then diluting/concentrating it multiple times until the GuDCl concentration was below 0.1 M, where CBD N 2 is folded. A new spectra of the refolded sample was then compared to native protein spectra. Since it is very difficult to concentrate the refolded protein to the exact concentration of the initial sample, the two spectra do not superimpose. However their respective morphology is the same, establishing that CBD N 2 refolds reversibly from GuDCl denaturation. Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDt ^•1 , , , , , r-210 220 230 240 250 260 270 Wavelength (nm) Figure 2.2: Refolding of CBDN2 after (a) GuDCl denaturation at pD=6.0 and (b) temperature. Chapter 2 - Thermodynamic and Kinetic Parameters of CBDN2 43 2 . 2 . 2 G u D C l D e n a t u r a t i o n a s a F u n c t i o n o f p D GuDCl denaturation of CBD N 2 was monitored at 3 pD values (5.0, 6.0, 7.0) using CD spectroscopy. This was done to examine any stability trends as a function of pD, and to obtain a AGy at pD=6.0, and T=25 °C, for comparison with NMR experiments (Figure 2.3). Table 2.1: Thermodynamic Results for GuDCl denaturation experiments at 25 °C [ G u D C l ] " 2 m A G „ ( D 2 0 ) p D (M) (kJ/mol«M) (kJ/mol) ± + ± 5.0 2.05 0.02 -21.0 4.3 43.2 8.9 6.0 2.16 0.02 -17.0 2.0 36.7 4.3 7.0 2.31 0.02 -13.8 1.6 31.8 3.7 Table 2.1 details the results of the unfolding of CBD N 2 in GuDCl. The most notable result was an increase in the unfolding midpoint of CBD N 2 by 0.13 M per unit increase in pD. This indicates that, under denaturing conditions the stability of CBD N 2 increases slightly with pD. However, the "m-value", a gauge of the protein surface area exposed in response to the presence of GuDCl (Creighton, 1992), decreases by 3.6 kJ/mol-[GuDCl] with each pD unit. Consistent with the common interpretation of m-values, this suggests that CBD N 2 surface area exposure, upon unfolding, becomes less sensitive to the presence of GuDCl. This result is surprising, and a physical interpretation is hard to envision. However one might propose that if CBD N 2 becomes less sensitive to the denaturing effects at higher pD, it is consistent that the unfolding midpoint would increase with pD. The extrapolated values of AGu(D20) indicated that the stability of CBD N 2 actually decreases with increasing pD. The reverse trend of AGU(D2Q) as compared to the midpoint value, Chapter 2 - Thermodynamic and Kinetic Parameters of CBD] 44 1.0-0.8 1 2 0.6-«§ D § 0.4-•a 0.2-0.0-o « o v • • o • o On • 8 c e o § 8 o T CO o 5.0 o 6.0 • 7.0 T" T" 235-] 2.30-2.25-2.20-2.15-2.10-2.05-2.00-0.0 0.5 1.0 1.5 20 2.5 3.0 3.5 4.0 4.5 5.0 [GuDCl] (M) O cf Slope 0.13 +/- 0.01 Intercept 1.39 +/- 0.08 5.0 — i — 6.0 - 1 — 6.5 —I— 7.0 Figure 2.3: (a) The fraction unfolded of CBDN2 in the presence of GuDCl as a function ofpH at 25 "C. (b) GuDCl unfolding midpoint as a function ofpD. (c) AGJD20) as a function ofpD. Chapter 2 — Thermodynamic and Kinetic Parameters of CBDm 45 [GuDCl]1/2, is due to the associated changes in n -^values. However the significance of the trend is unclear based on the large errors associated with the extrapolation. Indeed the fitting error is so large that the AG,j(D20) could well be equal for the different pD values (Figure 2.3(c)). 2 . 2 . 3 T h e r m a l D e n a t u r a t i o n a s a F u n c t i o n o f p D . Thermal denaturation experiments of CBD N 2 were measured using CD spectroscopy at pD = 5.4, 6.0, 7.0, and 8.9. Below pD = 5.4, these experiments could not be performed as the protein aggregates. Table 2.2 shows the results of the nonlinear least squares fits to the temperature denaturation data. The range of 3.5 pD units did not suffice to show a change in melting temperature larger that 2 degrees. Table 2.2: Thermodynamic Parameters from Thermal Denaturation of CBD N 2 as a function of pD p D m CC) A H (kJ/mol) ± + 5.4 67.4 0.1 600 29 6.0 67.0 0.1 480 22 7.0 68.0 0.5 350 35 8.9 65.8 0.2 590 58 The motivation for acquiring thermal denaturation data as a function of pD was to extract a value for the change in heat capacity during unfolding, A C p , as required for extrapolation of A G u to 25 ° C (equation 2.16). Since the spread in T m was so small and scattered, a linear fit of A H versus T was not viable for obtaining A C . m m O p Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDl 46 G 1.0 0.8 0.6-s § 0.4-•a 0.2-0.0-68.5-68.0-67.5 • 67.0 • 66.5 66.0 65.5 65.0. o 5.4 • 6.0 A 7.0 O 8.9 o 6 0 OA • O 06 ° o A —I II 1 i i r~ —j™ 1 i |— 50 55 60 65 70 Temperature (°Q 700-600-S 500-s 400-300-- 1 1 1 75 80 pD - 1 — i — ' — i — 1 — r -5.0 5.5 6.0 6.5 — I — 7.0 7.5 — I 1 1— 8.0 8.5 - i 1 r 1 1 1 1 1 1 1 1 1 1 1 1 1 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 — I — 9.0 pD Figure 2.4: (a) The fraction unfolded during a temperature melt of CBDN2 as a function ofpD. The dependence of(b)Tm and (c) AHm on pDfor CBDN2. Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDN2 Al 2 . 2 . 4 D S C o f C B D N 2 a s a F u n c t i o n o f p D In collaboration with Dr. Louise Creagh the thermodynamics governing the unfolding of CBD N 2 were measured as a function of pD using DSC. Thermograms were fit to a two-state model to extract AH m and Tm. However, as also seen by the CD spectroscopic analysis, the observed unfolding midpoint temperatures varied by only 4 °C between pD 5.6 and pD 12.0. Again, since the spread in T m was so small, there is little confidence in a linear regression in a plot of AHm and T m (Figure 2.6). Scatter in the data is attributed to ill-defined baselines. Table 2.3: DSC Data for CBD N 2 as a Function of pD p D T m CC) + A H m (kJ/mol) + 5.6 70.1 0.1 486 9 6 74.2 0.1 540 0 7 73.8 0.5 528 11 7.8 74.0 0.1 542 9 8.6 71.4 0.6 516 21 10 72.1 0.1 516 13 10.9 70.9 0.2 497 7 12 70.4 0.2 486 8 A C p 12.7 2 3 2 . 2 . 5 T h e r m a l D e n a t u r a t i o n a s a F u n c t i o n o f G u D C l t o o b t a i n A C P Given that obtaining AC p from thermal denaturation profiles as a function of pD was met with little success, thermal denaturation experiments at a constant pD (5.4) with low, non-denaturing concentrations of GuDCl were performed in anticipation of a larger spread of T m and AH values. To analyze this data it was necessary to assume that the ACp was independent of [GuDCl]. Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDj 48 LO-OK-'S 2 0.6-1 .2 0 4 ' 1 0.2-0.0-70 • v 0 a a o a o t D o A ° ° A ] A O V O A V , ° ^ ^ O V v oo V o o o 0 M V 0.4 M A 0.8 M O 1.2 M • 1.6 M -I— 30 65 H 60 G 40 50 60 70 Temperature ( °Q 80 90 1 _ e 55. 50. 45 • 650-600-550-500 H 450-400-350-300-250 —i— i—i— i—i— i—i— i—i— i—i— i—r—'—i—'—i—•—r 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 [GuDCl] (M) I —i—i—|—i—|—i—|—i—i—i—|—i—(—i—i—i—i—i—r 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 [GuDCl] (M) Figure 2.5: (a) The fraction unfolded during temperature denaturation ofCBDN2 as a function of GuDCl at pD-5.4. The temperature unfolding midpoint (b) as a function of GuDCl at pD=5.4. Chapter 2 - Thermodynamic and Kinetic Parameters of CBDN2 49 Table 2.4 presents the results of the temperature denaturation experiments as a function of pD. As would be expected, the temperature unfolding midpoint, Tm, decreased by ~ 5 °C as the [GuDCl] increased by 0.4M. While the T m values for this study were spread out and monotonically decreasing, the corresponding enthalpies did not have a corresponding decreasing trend, making linear regression analysis to obtain AC unproductive and statistically insignificant. Table 2.4: Thermodynamic Parameters from Temperature Denaturation of CBD N 2 as a function of [GuDCl] at pD=5.4 [ G u D C l ] m C Q (kJ/mol) (M) ± + 0 67.4 0.1 599 29 0.4 63.0 0.5 367 38 0.8 57.0 0.3 395 32 1.2 52.1 0.2 429 11 1.6 45.8 0.4 299 31 An alternative for obtaining ACp was elicited from (2.11). Instead of obtaining the AG^T) using ACp, the converse was tried. Prior to the temperature unfolding experiments as a function of denaturant, a simple GuDCl unfolding experiment was performed at pD=5.4. Thus AG([GuDCl]) was known for each [GuDCl] used in the temperature study. By substituting AHr a and Tm, from Table 2.4 and AG([GuDCl]) at each [GuDCl] into (2.11), 5 values for ACp were obtained and averaged. This value for ACp, as well as the DSC-ACp and a rule-of-thumb determination are listed in Table 2.5. Chapter 2 - Thermodynamic and Kinetic Parameters of CBDN2 50 2 . 2 . 6 C o m p a r i s o n o f A C p a n d A G y R e s u l t s A major goal of the G u D C l and temperature denaturation experiments was to obtain a A G 0 at 25°C and pD=6.0. The A C p values determined from different methods were substituted, in conjunction with A H m and T m from the temperature melt at pD=6.0, into (2.11) to give AGV detailed in table 2.5. Table 2.5: Tabulation of A C and AG, , results for C B D N , p U A G ^ D . O ) (kJ/mol) (kJ/mol-+ K ) at T=25°C and pD=6.0 + Rule of Thumb 7 . 7 ' N / A 38.9" N / A C D 8.5 2.0 36.7" 6.1 d D S C 12.7 1.4 2 7 . 6 ° 5.1 d C D ( G u D C l Unfolding) N / A 36.7 4.3 (a) A s a rule of thumb one can estimate A C p by multiplying the number of residues a protein has by 12 kcal/residue/mol/K. (b) (2.11) was used to fit this data, and the values for A H m and T m were taken from the temperature denaturation values as a function of p D at pD=6.0 (c) Obtained using equation (2.11) and the values for T m and A H m obtained from the D . S . C . experiments performed in conjunction with Dr . Louise Creagh. (d) bAGu = The "Rule of Thumb" and C D determination of A C are in better agreement than with the D S C results. While calorimetry (DSC) is generally thought to be the most cogent manner to obtain A C p , the scatter within the D S C data and the extracted A C p are both large. Figure 2.6 shows the slope of the line used to extract A C p from the D S C data, along with another line representing the A C p value obtained from C D experiments. The agreement between A G W obtained from G u D C l and temperature unfolding experiments using C D spectroscopy is surprising, given that the data used to obtain the respected Chapter 2 - Thermodynamic and Kinetic Parameters of CBDN2 51 values of AGV is completely independent. The AGU (D20) value obtained from DSC experiments is extremely low. 1 550-525-500-A C = 8 . 5 . - - -475 4 - ' . • • • ' A C =12.7 450-D S C Fit C D Result - 1 ' 1 ' r -68 69 70 —r-71 73 - 1 — 74 72 T m (°Q 75 Figure 2.6: Comparison of AC obtained from DSC and CD measurements Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDN2 52 2 . 2 . 7 T h e r m o d y n a m i c C o n c l u s i o n s The intrinsic assumption in all the analysis was that CBD N 2 unfolds in a 2-state transition. Given the general excellence of the fits of data to 2-state models, from both DSC and CD measurements, CBD N 2 does not appear to have any detectable folding intermediate. CBD N 2 exhibits a surprising lack of pD dependent stability, evidenced by the narrow range of the temperature unfolding midpoint, Tm, in both DSC and CD unfolding experiments. This is likely a consequence of few titratable side-chains in the structure of CBD N 2 (Brun, 2000). Given this robust characteristic, CBD N 2 makes an excellent candidate for degradation procedures in the pulp and paper industry. While CBD N 2 displayed pD independence for temperature denaturation experiments, GuDCl denaturation experiments revealed a increase in the unfolding midpoint concentration by 0.13 M with every unit pD increase. These results suggest that the presence of GuDCl must perturb the protein in such a way as to introduce a pD sensitivity. If this pD sensitivity is truly a reflection of GuDCl sensitivity, then the extrapolated value of AGW to zero denaturant (AGyODjO)) should be pD independent, so as to be consistent with the temperature denaturation experiments. Indeed, while there is a scatter in the AGV (D20) from the GuDCl unfolding experiments, all the values are equal within their associated error. Chapter 2 - Thermodynamic and Kinetic Parameters ofCBDN2 53 2 . 3 K i n e t i c R e s u l t s 2 . 3 . 0 R e f o l d i n g o f C B D N 2 f r o m G u D C l D e n a t u r a t i o n Refolding kinetics at pD=6.0 of CBD N 2 into 0.11, 0.21, 0.58, and 0.74 M GuDCl from 5 M GuDCl were fit with the best residuals to a 3 independent exponential functions using Plotdata non-linear regression software. y = Alek'' +A2ek^' +A3ek^ +yoo (2-17) Table 2.6: Refolding Rates for CBD N 2 in Non-Denaturing Concentrations of GuDCl [GuDCl ] (M) Refolding Rate (1/s) Fast Phase Med ium Phase S low Phase + + + 0.11 23.7 2.2 1.2 0.04 0.110 0.0019 0.21 5.4 0.5 0.48 0.018 0.032 0.0010 0.58 6.3 0.3 0.20 0.0058 0.0092 0.00035 0.74 7.9 0.4 0.14 0.0035 0.0063 0.00025 Despite the success of 2-state models in describing the thermal and chemical denaturation, three refolding phases were observed in the refolding kinetics of CBDN 2. The existence of multiple slower phases in the refolding kinetics of CBD N 2 is likely a result of cis I trans isomerization of this 13 proline protein. The native structure of CBD N 2 places Pro 76 in the higher energy cis conformation, which likely relaxes to the lower energy trans conformation in the denatured protein. The slowest phase in the refolding kinetics is likely the required trans -> cis conformation of Pro 76 to form native CBDN 2. In general, not all Pro residues relax to the trans conformation in the denatured state. Instead an equilibrium will be reached with 6-38 % of the Xaa-Pro bonds in the cis Chapter 2 - Thermodynamic and Kinetic Parameters of CBDl 54 Figure 2.7: Refolding ofCBDN2 in non-denaturing concentrations of GuDCl. (a) Stopped-flow refolding traces and (b) residuals for the 3 exponential fit to this data. Chapter 2 - Thermodynamic and Kinetic Parameters of CBDN2 55 conformation, depending on the identity of the preceeding amino acid, Xaa (Huyghues-Despointes et al, 1999a). Xaa for Pro 76 is Thr 75, and has a comparitively low probability of maintaining the cis conformation in the denatured state (9%; Huyghues-Despointes et al, 1999a). On the other hand, the Xaa residue of Prol20 is Phell9, which has a 23% probability of adopting the higher energy cis conformation in the unfolded state. Since refolding requires Prol20 to be in the trans conformation, it is possible that the medium refolding phase (Table 2.6) consists of the cis-^trans conformation of Prol20 (Huyghues-Despointes et al, 1999a). The salient feature of the proposed origin of the medium and slow refolding phases is that they are a consequence of allowing the prolines of CBD N 2 to equilibrate in the denatured state. If indeed these latter phases in the refolding kinetics of CBD N 2 are due to proline-isomerization, unfolding fluctuations in the native state likely refold according to the fast rate of Table 2.6, since by definition the protein is in equilibrium in the native conformation. Chapter 3 - NMR Experiments 56 C h a p t e r 3 N M R E x p e r i m e n t s 3 . 0 O v e r v i e w o f t h e N a t i v e S t a t e H X E x p e r i m e n t Amide hydrogen to deuterium exchange (HX) is measured quantitatively using 2D N M R HSQC spectroscopy. As a given 15N-labelled amide hydrogen exchanges for a deuteron, it's signal intensity in a 'H- 1 5 N HSQC spectrum decreases. Fitting HSQC peak intensity versus time to the equation that describes a single exponential decay yields the first-order rate constant for exchange. The exchange rate of a given amide hydrogen in a folded protein is dependent upon the local or global fluctuations that lead to solvent exposure and/or disruption of hydrogen bonds. The presence of small concentrations of denaturant such as GuDCl destabilizes the protein enough to enhance amide exchange via fluctuations, but not enough to globally unfold the protein. According to models of this "native state" H X , the equilibrium constant responsible for exposing a given amide group to solvent for exchange can be extracted from the protection factor, a ratio of the observed rate constant for exchange in the protein relative to that of a reference unstructured peptide. A site-specific free-energy for the fluctuations leading to exchange can consequently be obtained by extrapolating to conditions of no denaturant. Chapter 3 - NMR Experiments 57 3 . 1 M a t e r i a l a n d M e t h o d s 3 . 1 . 1 E x p e r i m e n t a l C o n d i t i o n s The native state HX experiment requires that a protein remains globally folded through a GuDCl concentration range large enough for trends in the denaturant-dependent exchange to be elucidated. Thus one chooses solution conditions that maintain the protein in a stable folded state in as high of a [GuDCl] as possible. However, several other factors must also be considered when selecting solution conditions and NMR acquisition parameters. These include maintaining a soluble sample, with HSQC spectra that have clearly isolated peaks, and choosing a pH such that amide HX is in the EX2 limit and slow enough for reliable measurement of a first order decay. On the other hand, if the HX is too slow, then months pass before complete amide exchange has occurred, and perturbations to peak intensity including spectrometer instability, solvent evaporation, and protein precipitation decrease the accuracy of the exponential decay profiles. One solution for these latter problems is the acquisition of a reference ID 'H spectrum along with each HSQC spectrum. One assumes that a resolved upfield resonance in the ID 'H spectrum is exchange-insensitive and thus the area of the corresponding peak can be used as a normalizing factor for the associated HSQC spectrum. In keeping with the goal of minimizing HX rates, the experimental temperature was chosen to be 25 °C rather than 35 °C (at which CBD N 2 was previously characterized) (Brun et al, 2000). A water bath was maintained at 25 °C for sample storage between data acquisition. An even lower temperature was not chosen since this would require use of a cooling bath over a period of months and since NMR spectral quality quickly decreases due to slower tumbling of CBD N 2 in solution. The experimental pD of 6.0 was chosen based on several considerations. Since every unit increase in pD raises the first-order base catalyzed HX reaction 10-fold, which can lead to EX1 instead of EX2 exchange, the pD was chosen to be as low as possible. However temperature Chapter 3 - NMR Experiments 58 denaturation of CBD N 2 at pD=5.0 (pH=4.6) led to aggregation at temperatures higher than 40°C, preventing an independent measure of the stability of the protein. This likely reflects the minimal solubility of CBD N 2 near its predicted isoelectric point (pi) of -4.1. Since both GuDCl and temperature denaturation proved reversible at pD 6.0, this condition was selected as a compromise for the NMR HX experiments. Thermodynamic parameters of CBD N 2 at this and other pD values are discussed in Chapter 2. 3 . 1 . 2 N M R A c q u i s i t i o n All spectra were acquired on a Varian Unity 500 MHz NMR spectrometer equipped with a single axis gradient triple resonance probe. Experiments with 'H N detection utilized the enhanced sensitivity pulsed-field gradient approach (Kay et al, 1992; Muhandiram and Kay, 1994). 'H was referenced with respect to an external DSS standard at 0.00 ppm. The I SN referencing was to an external sample of 2.9 M I5NH4C1 in 1 M HC1 at 24.93 ppm. The 'H spectral width was 6500 Hz with 1024 complex points. The I 5 N spectral width was 1600 Hz with 96 complex points. To observe the exchange of amides with limited protection, the first four HSQC spectra were acquired with 4 transients per free induction decay (FID) for an acquisition time of 15 minutes each. In order to improve signal-to-noise, all subsequent spectra comprised of 16 transients per FID, for a total experiment time of 1 hour. The peak intensities of the first four spectra were increased by a factor of 4 after processing. ID 'H spectra were acquired immediately after each HSQC spectrum so that the area under an exchange insensitive peak at -0.25 ppm could be used to normalize the amide intensity. 1024 complex points were acquired in 128 transients with a spectral width of 8000 Hz. Chapter 3 - NMR Experiments 59 3 . 1 . 3 N M R I D D a t a P r o c e s s i n g a n d A n a l y s i s 'H data was zero-filled to 2048 points prior to the Fourier transform. Only zero-order phase correction was performed, and a linear baseline subtracted from points in the region of a peak due to the 3 H 8 2 of Leul33 at - 0.23 ppm and the H p 2 of Asn 47 at - 0.27 ppm. A Felix macro was used to determine the area of this peak by summing the intensity of each data point contributing to the peak. This area was used to normalize the intensity of the crosspeaks in the corresponding HSQC spectrum for spectrometer shift, protein degradation and solvent evaporation. Generally the correction adjusted the intensities of the HX time points by a factor between 0.96-1.2, although the range was extended as low as 0.5 and as high as 1.8. Appendix E presents the multiplication factor for all HSQC spectra. 3 . 1 . 4 N M R 2 D D a t a P r o c e s s i n g a n d A n a l y s i s Approximately 30 HSQC spectra were taken for each of the 10 samples with concentrations of GuDCl ranging from 0 to 1.8 M at pD=6.0, and for the 5 samples at different pD values and [GuDCl]=1.6 M, to allow for the full time range of exchange. This amounted to an approximate total of 450 HSQC spectra. The HSQC data was processed using Felix95 software. The spectra were zero filled to 2048 points in the proton dimension and 1024 in the nitrogen dimension. Soft gaussian apodization (line broadening = -8 and Gaussian parameter = 0.1) in the 'H dimension and sinebell squared apodization (window size = 96 points and phase shift = 80°) in the 1 5N dimension were applied. Each spectrum was individually phase corrected in the *H dimension. 90 ° zero-order and 180° first-order phase corrections were performed in the 1 5N dimension. A zero* order polynomial for the 'H dimension was also used to correct baselines in each spectrum. Chapter 3 - NMR Experiments 60 3 . 1 . 5 S p e c t r a l A s s i g n m e n t NMR spectral assignments were determined previously for CBD N 2 at 35 °C, pH=6 (Brun, 2000). To extend these assignments to the conditions utilized for this study, HSQC spectra of CBD N 2 were recorded as a function of temperature, GuDCl and pH. i) T e m p e r a t u r e D e p e n d e n t S p e c t r a l C h a n g e s HSQC spectra of CBD N 2 were recorded at 35, 32, 30, 27, and 25 °C in order to obtain 'H and 1 5N chemical shift assignments for the protein at 25 °C. In general, some minor chemical shift variations are expected due to the temperature dependence of the D 20 lock. The average absolute value chemical shift in the 'H dimension was 0.065 ± 0.02 ppm, and 0.094 ± 0.006 ppm in the 1 5N dimension. i i) G u H C I D e p e n d e n t S p e c t r a l C h a n g e s HSQC spectra of fully protonated CBD N 2 were acquired with GuHCI at 0, 0.3, 0.6, 0.9, 1.2 and 1.5 M. CBD N 2 is folded under these conditions. Amide residues exposed to the solvent tended to show the largest chemical shift changes. Interestingly, above average changes in chemical shifts were observed for the amides of many solvent exposed leucines, as well as the residues near or hydrogen bonded to these leucines. For example, the relatively large chemical shift change resulting from the GuHCI titration occurs for L102 in P-strand A4 (see figure 3.1a) and its hydrogen-bonding partner, M78, a member of P-strand A3. It is generally thought that GuHCI acts as a protein denaturant by reducing the energetically unfavorable exposure of hydrophobic residues to the polar solvent. Association of GuHCI with the amino acid leucine appears to be particularly favorable, consequently producing relatively large chemical shift changes to those groups near in space to solvent exposed leucines. The average absolute value chemical shift in the 'H dimension was 0.080 ± 0.04 ppm, and 0.219 ± 0.019 ppm in the 1 5N dimension. Chapter 3 - NMR Experiments 61 P-strand L148 Figure 3.1: Histogram of the weighted chemical shift difference Adj between (a) 0-1.5M GuHCI and (b) pH 4-8 in the HSQC spectra ofCBDm. Gaps are proline residues or residues with overlapping frequencies. ' A 8 w = . J (Ao l f f ) 2 + ( ± A 6 U A , ) 2 Chapter 3 - NMR Experiments 62 i i i ) p H T i t r a t i o n Thirteen HSQC spectra were acquired between pH 4 and 8, using HC1 to initially reduce the pH from a value of 6, followed by the addition of NaOH to increase the pH. In general, amide groups of residues with or near solvent exposed titratible side chains (Asp, Glu, His) had chemical shifts showing the largest pH-dependence. The average absolute value chemical shift in the 'H dimension was 0.091 ± 0.017 ppm, and 0.46 ± 0.06 ppm in the 1 5N dimension (Fig. 3.1b). 3 . 1 . 6 N a t i v e S t a t e H X P r o t o c o l Native state HX of CBD N 2 was measured at pD=6.0 in the presence of 0, 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, 1.4, 1.6, and 1.8 M GuDCl. As well, exchange profiles of CBD N 2 at 1.6 M GuDCl were measured at pD = 5.0, 5.5, 6.0, 6.5, and 7.0. Generally 10 mg of 1 5N labeled CBD N 2 (Appendix B) in 500 pL of 50 mM potassium phosphate buffer (pH 6), 0.02 % sodium azide and 50 mM sodium chloride was lyophilized prior to being dissolved into and equivalent volume of D 20 and a specific GuDCl concentration. GuDCl was obtained from GuHCl through 8 D 20 dilution cycles wherein 1 g of GuHCl was dissolved in 20 ml of D 20 followed by lyophilization. The difference between the index of refraction of a reference solution,2 nref> and a solution containing GuDCl, n, was measured at room temperature on an Spectronic Instruments refractometer (model number 334610), and used to determine [GuDCl] within ~ 0.05M from (3.1) (Pace et al, 1992): [GuHCl] = 57.147 (n - nj + 38.68 (n - nj - 91.6 (n- nj (3.1) After resuspension of the lyophilized CBDN 2, the pD was quickly adjusted to 6.0 using DC1 or NaOD and the sample briefly centrifuged to remove any undissolved protein. 2 A reference sample was always taken prior to the addition of G u H C l in order to account for the extraneous refraction signal by H 2 0 , D 2 0 or buffer. Chapter 3 - NMR Experiments 63 The solution was then transferred to an NMR tube and placed in the spectrometer with the probe temperature set to 25 °C. 'H tuning and shimming was performed, followed by determination of the 90° pulse length for 'H. The 90° pulse increased from 9 to 14 ms between [GuDCTJ=0 and [GuDCl]=1.8 M. Generally the time between dilution into D 20 and the start of the first HSQC spectra was 20 minutes. Zero time for the exchange was set to be the point of protein rehydration in D 20. The time point for exchange was taken to be the midpoint of the acquisition period of a given HSQC spectrum. The final pD of all NMR samples, measured when exchange was complete or after 9 months, did not deviate by more than 0.1 unit from the initial value. The 90° pulse length was also re-measured for all NMR samples after exchange was complete. Deviations were restricted to 0.2 ps for all samples except for [GuDCl] = 0.4 M, where the deviation was 0.7 ps. The refractive index of HX samples, measured when exchange was complete, was slightly higher than at the time of sample preparation, due to the presence of buffer and protein. Accounting for these additional contributions, [GuDCl]fmal ~ [GuDCl]initial as determined by refractometry. 3 . 1 . 7 A n a l y s i s o f t h e H X T i m e S e r i e s Peak intensities were measured using the titration module of Felix95. The intensity of a given peak was measured at the center of an encompassing box defined by a width of ~ 0.04 ppm in the 'H dimension and ~ 0.03 ppm in the 1 5N dimension. The shape of the peaks did not change or move significantly during the time course of the HX measurements. This scenario allows determination of the extent of exchange from the center intensity in the box that encompasses a peak. Decay rates obtained from fits to maximum intensities within a box were similar, within the error of the fit, to rates obtained from center intensities. The maximum determination however, had a consistently higher associated fitting error due to the effects of low signal-to-noise associated with amides nearly completely exchanged. Chapter 3 - NMR Experiments 64 Before fitting was performed, the peak intensities of all amide resonances in a HSQC spectrum recorded at a given time point, were normalized to the most resolved upfield peak in the 'H NMR spectrum of CBDN 2. The peak intensity of the first four time points, which corresponded to HSQC spectra taken with 4 instead of 16 transients, were also multiplied by a factor of 4. A single exponential was fit to the peak intensities of each resolvable amide groups in C B D _ The constant baseline value loo represents the residual protons present in the sample, preventing complete NH -> ND exchange. Nonlinear.least squared fits were performed in Origin. These fits were done individual residues at a time at each [GuDCl] to ensure quality of the data. Exchange rates with fitting errors much more than 10% of the fitted value were discarded. N2' I(t) = I0e-k"'+Ic CO (3.2) Chapter 3 - NMR Experiments 65 3 . 2 N a t i v e S t a t e H X N M R R e s u l t s 3 . 2 . 1 G e n e r a l H y d r o g e n E x c h a n g e R e s u l t s A total of 71 amide groups in CBD N 2 had measurable hydrogen exchange at more than 2 different GuDCl concentrations. This is approximately half of the amide groups in the 153 residue protein, a typical fraction in native state hydrogen exchange experiments. A complete data set is precluded due to chemical shift overlap or exchange within the dead time of the experiment (-1.5 hours, since at least 5 time points are required for accurate fitting). Residues that exchanged faster than the dead time of the experiment were located in either the flexible loop regions, the N and C termini, or on the outer strands of the P-sheets with solvent-exposed amide groups. Exchange at [GuDCl] = 0 M occurred over 6 orders of magnitude with protection factors that ranged from-102 to-108. The residues for which FLX could be determined at 2 or more [GuDCl] are listed in Table 3.2. All elements of secondary structure had several probes, with the exception of p-strand BO. In the latter case, only 1 of the 3 amino acids that comprise this strand showed measurable protection. As [GuDCl] increased, individual amide HX rates likewise increased. This is illustrated in Figure 3.2 for Ala74, which is part of P-strand B3. 3 . 2 . 2 H X t r e n d s i n p - s t r a n d s A common trend of fast exchange, or limited protection, exists among alternating residues in P strands at the edge of sheets A and B (Figure 3.3). Amide groups exposed to the solvent were not detectable after transfer to D 20, while the intervening groups, intramolecularly H-bonded to the neighbor strand, exhibited significant protection from exchange. Chapter 3 - NMR Experiments 66 Time (minutes) Figure 3.2: The exchange profile of Ala74 as a function of [GuDCl] Chapter 3 - NMR Experiments 67 Figure 3.3 B-strands B3 and B4. B4 is an edge strand of B-sheet B that displays measurable exchange for only alternating residues. B-strand B4 amide groups with measurable exchange are indicated by orange spheres, while the solvent exposed residues which exhibit fast exchange are indicated by red spheres. For reference, the orange side chains point in the direction of the hydrophobic core, while the red side chains point towards the protein exterior. This figure was created using the Biosym software, Insight 2000. Chapter 3 - NMR Experiments 68 This feature is a subtle indication of a several issues governing the analysis of FIX in (3-sheet proteins. Unlike a helices, where H-bonding is a sequence-local interaction between the i"1 and ith+4 residues, (3 sheets have non-local H-bonds. As well, backbone conformation in an a-helix is constrained such that most residues involved are H-bonded. Thus similar protection is expected for residues in the same a helix, since any structural fluctuations uniformly disrupt all the a-helical H-bonds. In a P-strand, the additional conformational flexibility does not force all amides to orient such that they form H-bonds. As well, only alternating residues form H-bonds with neighboring strands. Thus the fast exchange of a solvent exposed amide group does not necessarily indicate low stability in that P-strand. Indeed p-strand B4 shows an alternating pattern of fast exchange for solvent exposed residues, while interior facing residues are among the most protected in CBDN 2. The previously reported secondary structure of CBD N 2 was based on patterns of secondary chemical shifts, 3JN H_H C X scalar couplings, NOE-derived distance restraints, and hydrogen bonds identified through a qualitative HX analysis. The exchange patterns scrutinized in this study for the most part agree well with the previously reported secondary structure of CBDN 2. However, a few modifications to the secondary structure boundaries are suggested. For example, Tyrl39 is among the most protected from HX, suggesting that it is the first residue in the p-strand B5, instead of the reported Glul40. Several residues located at the boundary of p-strands exchanged in a GuDCl-insensitive manner, namely Leu9 from BO, Ser65 from B3 and Thrl50 from B5. In contrast the interior residues of p-strands B3 and B5 contain some of the most protected residues in CBD N 2 (Figure 3.8). Chapter 3 - NMR Experiments 69 3 . 2 . 3 H i g h l y P r o t e c t e d R e s i d u e s A residue was deemed a "slow exchanger" when there was no detectable change in protection level after 9 months in [GuDCl]=0.6 M. All slow exchangers were members of (3-strands (including Tyrl39 in B5). The sole exception is Gly85, a deeply buried residue in the A3-A4 loop. Table 3.1: A comparison of sequence conservation in CBD, Percent Conservation 4 most homologous 8 most homologous A l l residues 52 33 Residues with H X information 56 36 Residues that are unexchanged after 9 months in [GuDCl]=1.6 M 80 55 Percent Conservation was calculated in the fol lowing manner: If the residue was 100%, 75%, or 50% conserved among the 4 (or 8) homologous sequences, it was assigned a conservation value of 100,75, or 50. Residues that were not conserved at a l l , or less that 50 % conserved were assigned a value of 0. Percent conservation was then calculated as the sum of the conservation value for a given set of residues divided the theoretical sum i f al l residues in the set were 100% conserved, multiplied by 100. Sum _of conservation _ values WcConservaWn = • 100 Theoretical _ sum _ for _ 100% _ conservation In some proteins, residues that exhibit a high level of protection are also those that are highly conserved (Alexandrecu et al, 1999). This is also true for CBDN 2. 20 out of the 23 most protected residues in CBD N 2 are conserved to some level among the 4 most homologous sequences (Figure 1.10), while 18 out of the 23 highly protected residues are conserved among the 8 most homologous sequences. Similar conservation was observed for the entire protein and among those residues that were probes in the native state HX experiment (Table 3.1). However significantly more percent conservation exists among those residues that are highly protected to exchange. Chapter 3 - NMR Experiments 70 |3-strands A2, B2, A3, B3, B4, A5, and B5, which contained these highly protected residues, were all members of the jellyroll structural component of CBDN 2. Boundary strands to the P sheet BO, A l , BI and A4 had no residues that remained unexchanged after 9 months in 1.6 M GuDCl, suggesting that these residues exchange via local fluctuation unraveling at the N-terminus (for BO) and fraying of strands at the edges of the P-sheets (for A l , BI, and A4). B4 the analogous edge strand to A4 is a member of the most slowly exchanging strands by virtue of a single "slow exchanger", Tyrl l l , which forms a H-bond with Leu68 another "slow exchanger." The C-terminal strand, B5 is both the longest strand and among the most protected to exchange. Figure 3.4 shows the strands containing the most protected amides colored in blue and the remainder colored in orange. Chapter 3 - NMR Experiments 7 1 Figure 3.4: Top view ofCBDN2. B-strands containing residues that remain unexchanged after 9 months in 1.6 M GuDCl are colored blue, while less protected strands are colored orange. This figure was created using the Biosym software, Insight 2000. Chapter 3 - NMR Experiments 72 3 . 2 . 4 D e t e r m i n a t i o n o f A G H X ( D 2 0 ) a n d m H X . AGHX=-RT\n(^-) kint ( 3 . 3 ) Free energy of FIX, AG H X , at a given [GuDCl], was determined according to (3.3), where the observed protein exchange rate kex was obtained for CBD N 2 by NMR spectroscopy as described in section 3.1, and the intrinsic rates of a polypeptide chain, ktat, was calculated according to the method of Bai et al as discussed in detail in Appendix D (1993). For almost all the residues measured, A G ^ had a simple linear dependence on [GuDCl], be it large or small, which could be fit to (3.4). AGHX = AGHX (D20) + mHX [GuDCl] (3-4) Figure 3.5(a) shows the exchange profiles of residues that have a mono-phasic dependence on [GuDCl]. Only 9 amide groups (Val31, Val58, Met78, Arg81, Phell9, Hisl32, Glul40, Glnl45, Leul48) in CBD N 2 displayed bi-phasic exchange behavior: that is, exchange with little or no denaturant dependence at low [GuDCl], and yet large denaturant dependence at higher [GuDCl]. The transition occurred near 1.2 M GuDCl (Figure 3.5(b,c)). For those residues that displayed biphasic exchange the data were fit to a modified version of 3.4 (Chamberlin et al, 1996): AGHX = -RTHKsubglobal +Kglobal) (3.5) where the large denaturant dependent phase is described by: ^Gglobal(D20) + mglobal[GuDCl] (3.6) Kglobal = e x P( : — ) and where K s u b g l o b a l is described by: _ n_,AGsub(D2Q) + msub[GuDCl]^ „ ~ K subglobal ~ e x P( — ) ^ •' > Chapter 3 - NMR Experiments 73 When there is no denaturant dependence in the subglobal phase, m s u b is set to zero in the fitting algorithm. The entire set of fitted values for A G ^ C D j O ) and m H X are listed in Table 3.2. Chapter 3 - NMR Experiments 74 • Alal6 A Leull7 X Phell9 o Tyrl39 [GuDCl] (M) • Met78 © Hisl32 X Glul40 [GuDCl] (M) 1 ' 1.8 —1 2.0 X Val31 • Glnl45 • Leul48 [GuDCl] (M) Figure 3.5: The free energy of HX, AGHX, as a function of [GuDCl] for CBDN2 residues that exchange (a) with a simple linear denaturant dependence, (b) with a denaturant independence at low GuDCl and a denaturant dependence at high GuDCl, and (c) with small denaturant dependence at low GuDCl and a larger denaturant dependence at high GuDCl. Chapter 3 - NMR Experiments 75 Table 3.2: Native State AGHX(D20) and m values as fit to (3.4-3.7). Those values with no associated fitting error were obtained from a linear fit to two points. AG^CD.O) m A G ^ C D p ) m (kJ/mol) (kJ/mol-M) (kj/mol) (kJ/mol-M) AA ± 2° AA ± + 2° L9 16.5 0.2 -1.3 0.3 B0 G85 43.3 1.1 -10.9 0.8 L10 21.8 0.3 -1.7 0.3 G87 24.4 0.6 -2.4 0.7 F15 20.2 0.1 -1.5 0.1 G88 17.1 0.8 -0.4 0.7 A16 20.3 0.2 -0.9 0.2 Y91 14.8 0.3 -2.8 0.4 S23 44.8 1.0 -11.1 0.8 A l R92 21.9 0.2 -2.6 0.3 L24 18.2 -21.5 A l A94 34.5 0.2 -5.8 0.2 A4 S28 17.1 -0.7 F95 35.0 0.3 -7.0 0.2 A4 V31 42.5 6.8 -13.1 3.6 BI Q97 20.4 0.3 -1.3 0.3 A4 28.3 1.9 -1.8 3.9 A100 47.6 1.6 -12.3 1.2 A4 A33 19.3 0.5 -1.2 0.4 BI L102 22.2 0.2 -3.7 0.3 A4 M37 42.3 0.9 -9.5 0.6 B2 T103 20.6 0.1 -3.1 0.2 A4 C38 45.1 0.3 -10.0 0.2 B2 A107 37.6 -9.5 B4 V39 42.0 0.4 -10.9 0.3 B2 R109 26.5 0.2 -2.2 0.2 B4 D40 42.3 0.2 -11.8 0.2 B2 Y l l l 40.9 0.8 -11.1 0.6 B4 G43 15.3 0.4 -1.2 0.3 F113 40.0 0.8 -10.1 0.6 B4 W49 14.8 0.3 -3.5 0.7 S115 32.0 0.3 -3.1 0.3 B4 A51 24.5 0.2 -2.8 0.1 A2 L117 18.3 0.2 -1.2 0.2 L53 40.0 0.8 -10.4 0.5 A2 F119 30.0 1.7 -4.4 1.0 V54 38.5 3.6 -10.7 2.3 A2 28.7 1.7 N/A Y55 40.3 1.0 -10.4 0.8 A2 G127 32.8 1.6 -5.7 1.2 A5 G57 28.0 0.2 -1.9 0.2 Q128 42.8 1.3 -10.2 1.0 A5 V58 28.7 0.3 -3.3 0.3 V129 41.8 0.9 -11.9 0.7 A5 G61 24.7 0.3 -2.6 0.3 A130 42.8 0.4 -11.3 0.3 A5 E64 21.6 0.2 -2.0 0.2 H132 .43.4 3.8 -9.2 2.4 A5 S65 15.6 0.9 -1.4 2.3 B3 34.0 0.7 N/A L68 39.9 1.1 -11.3 0.7 B3 G134 21.6 0.1 -3.9 0.2 S69 42.6 0.8 -10.6 0.6 B3 K135 18.7 0.4 -2.9 0.4 F70 39.9 0.5 -8.9 0.4 B3 Y139 39.5 0.1 -9.4 0.1 T71 40.4 0.4 -9.6 0.3 B3 E140 43.1 3.2 -12.3 1.9 B5 S73 42.1 0.3 -9.4 0.2 B3 30.1 1.0 N/A A74 36.3 0.5 -6.1 0.5 B3 F141 41.1 0.5 -10.8 0.4 B5 T75 39.2 0.3 -8.7 0.2 B3 C142 46.9 0.7 -10.9 0.5 B5 M78 43.9 2.1 -11.7 1.3 S144 22.5 0.2 -0.4 0.2 B5 28.7 0.1 N/A A3 Q145 34.1 0.4 -4.0 0.3 B5 R81 46.7 19 -12.5 10 A3 V146 33.7 0.4 -7.3 0.3 B5 27.2 0.4 -1.7 0.8 L148 42.4 9.2 -11.4 5.1 B5 V82 38.1 0.4 -9.5 0.3 A3 26.0 0.5 -0.7 1.0 V84 36.3 0.9 -9.7 0.6 A3 T149 41.0 0.6 -11.4 0.5 B5 T150 23.6 0.6 -1.8 0.5 B5 Chapter 3 - NMR Experiments 76 Figure 3.6 shows the distribution of AG^DjO ) and mHX-values for CJ3DN2. There are two distinct exchange regimes: those amide groups that exchange via "local" fluctuations which are characterized by low m values (that is, small denaturant dependence) and a range of small AGH X(D 20), and those amide groups that exchange due to global unfolding. The latter exhibit large mH X and AGH X(D20) values which have a strong correlation. For global unfolding exchange, this correlation with AGH X(D 20) indicates that when a large change in solvent exposure (large m) occurs, there is a corresponding large free energy cost for the structural reorganization that allows exchange. Furthermore, the clustering among those residues that exchange globally in both sheets is indicative of a 2 state folding mechanism for CBDN 2. 2-0--2-| -6-a -8. I <= -10--12--14-Local Fluctuations • m 9 * • • n " 1 • • • Mono- Bi-Phasic Phasic • o A-Sheet • • Loop • • B-sheet • * J o D n < 1 — 1 — i — 1 — i — < — i ' 1 — 1 r ~ 15 20 25 30 35 40 45 Global Fluctuations 50 AG™ (D.O) (kJ/mol) Figure 3.6: Distribution ofAGHJD20) and m values for CBDi Chapter 3 - NMR Experiments 77 (a) (b) 47.6 44.1 40.6 37.1 33.6 30.1 26.0 23.2 19.2 15.2 12.7 -12.0 -10.8 -9.6 -8.4 -7.2 -6.0 -4.8 -3.6 -2.4 -1.2 0 AGH X(D:0) (U/mol) m H X (kJ/mol-M) Figure 3.7: Side and top view of the (a) residue stability, AGH]/D20), of CBDN2 to HX and (b) their relative mm values indicated by a color gradient. In the side view, the A-sheet is above the B-sheet. The A-sheet is in the foreground of the top view. This figure was created using Biosym -Insight 2000 software. Chapter 3 - NMR Experiments 78 3 . 2 . 5 L o w D e n a t u r a n t D e p e n d e n t E x c h a n g e Residues in CBD N 2 displaying only low denaturant dependent exchange (small mHX and small AGHX(D20)), even at the highest concentration of GuDCl, are listed in Table 3.3. Many of these residues have at least partially solvent-exposed side chains, and most are present in coil regions of CBD N 2 (Figure 3.7 (a) green, (b) yellow) The average m^ value for these residues is -2.2± 0.2 kJ/mol-M and the average AGHX(D20) is 20.2±0.7 kJ/mol. Low denaturant dependent exchange is thought to take place during low-energy conformation fluctuations ("breathing"), which is common in native proteins. Dubbed "local fluctuation" exchange, it is traditionally expected to result from only small surface area reorganization of the protein. However, recent simulation work by Wooll et al suggest that an amide hydrogen which exchanges in a denaturant-independent manner does not necessarily do so as a sole result of small structural fluctuations but could exchange in this manner as a result of significant surface area exposure of the protein (-10%) (2000). The amount of amino acid surface area exposed to the solvent was evaluated for those residues in CBD N 2 that exchanged with a small denaturant dependence. Several of these residues were less solvent accessible than the average for highly protected residues. Furthermore, the percentage of solvent exposure was less than the percentage of amino acid surface area occupied by an amide hydrogen. Thus the exchange of LeulO, Phel5, Leu 102, and Glyl34 likely requires significant structural fluctuations. Chapter 3 - NMR Experiments 79 Table 3.3: Amide hydrogen and amino acid solvent exposure for residues that exchange in a denaturant independent way. The solvent exposure values listed here are averaged over all the CBD N 2 NMR-based ensemble of structures. Highlighted in bold are those residues that exchange with low denaturant dependence, but which have less solvent accessible area exposed than a highly protected residue. Solvent exposure was measured using MOLMOL (Koradi, 2000). %Surface Accessible Total Area of A A Surface Area (A2) Surface Area(A 2) N H relative to entire % Solvent Exposure 2° structure A A A A Residue Particular A A average Leu9 76 303.7 7.5 24.9 10.8 B 0 LeulO 11 293.3 7.5 3.8 10.8 Phel5 1 339.6 6.6 0.2 6.9 A l a l 6 26 231.3 9.6 11.2 16.5 Leu24 34 295.6 7.5 11.6 10.8 A l Ala33 56 229.8 9.6 24.3 16.5 B I Ser65 39 240.4 9.3 16.1 17.9 -B3 Gly87 34 200.6 11.0 16.9 17.5 Tyr91 154 354.2 6.3 43.6 18.6 Arg92 170 358.0 6.5 47.6 29.3 Gln97 54 306.4 7.2 17.5 17.4 A 4 Leul02 9 306.6 7.5 3.0 10.8 A4 L e u l l 7 74 304.8 7.5 24.2 10.8 Asp122 115 266.0 8.2 43.1 36.9 Glyl34 2 201.4 11.0 1.1 17.5 L y s l 3 5 67 336.5 6.5 20.0 20.0 Se r l44 18 240.1 9.3 7.4 17.9 B 5 T h r l 5 0 21 264.9 8.5 7.7 18.0 -B5 L o w denaturant dependence average A l l C B D N , A A average H ig h ly protected residues average 14.6 19.0 7.5 ' - ' indicates P-strand boundary Chapter 3 - NMR Experiments 80 3 . 2 . 6 L a r g e D e n a t u r a n t D e p e n d e n t E x c h a n g e . Amide hydrogens that exchange with large denaturant-dependence are referred to as "global exchangers" since the fluctuations that lead to exchange require global unfolding of the protein, or AG^fDjO) ~ AG,j. A structural representation of CBD N 2 (Figure 3.7) with the magnitude of AGHX(D20) and mHX displayed in a color gradient serves to illustrate that the most stable residues exist in the f3-sheets. The average AGHX(D20) for these residues is 41.9±0.4 kJ/mol while the average mHX is -10.7±0.2 kJ/mol-M. While several trends exist within the global folding data, overall the FfX behavior CBD N 2 is consistent with a 2-state model in which global fluctuations involve the entire protein, rather than sub-domains, such as sheets A or B. Figure 3.8 shows the values AGH X(D20 ) and m mapped along the sequence of CBDN 2. In regions of P-strand secondary structure, AG^CC^O) and m^ have similar large values. This excludes the possibility that the individual P sheets have different stabilities (either within sheets A or B, or in comparison between sheets). This is not unexpected as it is unlikely that a jellyroll P-sandwich protein would have sub-domains with distinct stability given the non-local interactions that make up its secondary and tertiary structure. Solvent exposure of residues within a strand requires large-scale fluctuations involving the entire sheet. This likely requires disruption of the hydrophobic core of the protein and hence fluctuations in the opposing sheet as well. Chapter 3 - NMR Experiments 8 1 Figure 3.8: AGHX(D20) and m for (a) fi-Sheet A and (b) fi-sheet B and (c) random coil regions mapped on the amino acid sequence. Chapter 3 - NMR Experiments 82 3 . 2 . 7 C o m p a r i s o n o f A G H X ( D 2 0 ) a n d A G u GuDCl and temperature unfolding experiments, as monitored by CD spectroscopy, gave a free energy of unfolding in D 20 for CBD N 2 of 36.7 ±4.3 kJ/mol. Given the extrapolation involved, this value agrees well with the average A G ^ (D20) of 41.9±0.4 kJ/mol calculated from those residues which exchange globally. Upon closer inspection, there is a small overestimation of stability determined by HX methods. This originates because the structural fluctuations that give rise to HX occur on a time scale much faster than cis-trans proline isomerization. Thus in HX experiments, one measures the energy input required to not only to unfold the protein, but to unfold the protein to a higher energy ds-X-Pro denatured conformation. In contrast, during equilibrium unfolding experiments, denatured samples can undergo isomerization from a higher energy ds-X-Pro to a lower energy trans-X-Pvo conformation. Thus in HX experiments of proteins with native c/s-prolines, there is an overestimation of the energy of unfolding by the additional energy of isomerization. The energy adjustment is determined from the equilibrium constant (KPR0) of isomerization of a given Xaa-Pro residue. K P R 0 varies from 1.65 when Xaa is a Trp to 15.67 when Xaa is another Pro. CBD N 2 has 13 prolines, of which one (Pro76) is in a cis conformation in the native state. The calculated adjustment is 9.4 kJ/mol, which results in a predicted unfolding A G u ~ 46 kJ/mol in the absence of isomerization. Although this is slightly higher than the average value of 41.9 kJ/mol for HX global fluctuations, the correction is qualitatively consistent, and in agreement with a AGHX(D20) of 47 kJ/mol for the most slowly exchanging residue, Cysl42. 3 . 2 . 8 S t a b i l i t y i n ( 3 - h a i r p i n s (3-hairpins are a structural motif consisting of two anti-parallel P-strands, typically joined by a loop of approximately 2-5 residues. The only motif among p sheets that are analogous to an a-helix, P-hairpins are secondary structure elements formed by sequence-local hydrogen bonds. Chapter 3 - NMR Experiments 83 In CBD N 2 there are 2 P-hairpins, while the rest of the P-sheets are comprised of non-sequence-local P-strand interaction. In both cases, the p-hairpins occur at the edge of the P-sheets of CBDN 2. The first P-hairpin is between BI and B2 (Table 3.4), while the second is between A3 and A4 (Table 3.5). Table 3.4: A G ^ D ^ ) and mHX for the P-hai rpin between BI and B2. A A 2 ° structure A G H X ( D 2 0 ) (kJ/mol) ± m H x ^kJ/mol-M) + V 3 1 B I 28.3 1.9 -1.9 2.9 42.5 6.5 -13.1 3.6 A 3 3 B I 19.3 0.5 -1.2 0.4 M 3 7 B 2 42.3 0.9 -9.5 0.6 C38 B 2 45.1 0.3 -10.0 0.2 V 3 9 B 2 42.0 0.4 -10.9 0.3 D 4 0 B 2 42.3 0.2 -11.8 0.2 AGHX(D20) of residues Met37, Val39, and Glu40 in B2 and the global unfolding phase AGHX(D20) of Val31 in BI seem to agree well, with and average AGHX(D20) of 42.3 kJ/mol. Ala33 of B1 is 24% solvent exposed (higher than the CBD N 2 average of 17%; Table 3.2), and so is not expected to show cooperative protection. Cys38 of B2, which forms a disulphide bond integral to the structure of CBDN 2, has a larger AGHX(D20) than the average for this hairpin. There is no consistent stability trend in the A3-A4 hairpin (Table 3.5). There is some agreement in the AGHX(D20) and m^ near the junction of the strands, but not before Val82 in A3 and not beyond F95 in A4. Chapter 3 - NMR Experiments 84 Table 3.5: AG^CE^O) and m for the P-hai rpin between A3 and A4 A A 2 ° A G H X ( D 2 0 ) (kJ/mol) m H X (kJ/mol-M) structure • ± . ± M 7 8 A 3 28.7 0.1 N / A 44.0 2.2 -11.7 1.3 N81 A 3 27.2 0.4 -1.7 0.8 46.6 19.1 -12.5 10.2 V 8 2 A 3 38.1 0.4 -9.5 0.3 V 8 4 A 3 36.3 0.9 -9 .7 0.6 A 9 4 A 4 34.5 0.2 -5.8 0.2 F95 A 4 35.0 0.3 -7.0 0.2 Q97 A 4 20.4 0.3 -1.3 0.3 A 1 0 0 A 4 47.6 1.6 -12.3 1.2 L102 A 4 22.2 0.2 -3.7 0.3 T103 A 4 20.6 0.1 -3.1 0.2 3 . 2 . 9 H y d r o g e n B o n d B r e a k a g e An amide group must be accessible to a catalyst for H X to occur. Since the encounter complex leading to exchange requires formation of a H-bond with a solvent, structural fluctuations that break internal H-bonds are required. Exchange profiles of residues that form cross-P-strand H-bonds often overlap, since structural fluctuations that break internal H-bonds are also those that promote exchange. Figure 3.9 highlights two such examples. Chapter 3 - NMR Experiments 34 32 30 6 S2<H i 24-22 20 I I a Phel41 • Val39 i 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 [GuDCl] (M) 34 32-30-6 28 26 4 24 22 A 20 o Tyr55 • Vall29 5 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 [GuDCl] (M) Figure 3.9: AGHX profiles of residues that form H-bonds to each other. Chapter 3 - NMR Experiments 86 3 . 2 . 1 0 T h e D i s u l p h i d e B o n d CBD N 2 has a single disulphide bond between Cys38 and Cysl42 that serves to cross-link the p-strands B2 and B5, both of which contain many residues exhibiting slow exchange via global fluctuations. Reduction of this disulphide bond leads to the unfolding of CBD N 2 under otherwise benign conditions. The amide groups of these two cysteine residues have exchange profiles that overlap (Figure 3.10). Their matching AGHX(D20) and mHX values are higher than any other residues in strand B2 and B5, and with the exception of AlalOO, the rest of the protein. In addition, the stability of these residues is larger than the global stability of CBDN 2, measured by CD spectroscopy and corrected for proline isomerization. "Super" stability in the region of a disulphide bond has been reported previously, most recently by Laszlo et al (1999) in their FIX study of human prion protein. This phenomenon is generally interpreted as an indicator of residual structure in the unfolded state. It is curious that the amide groups of these cysteines have the same stability pattern given that they are not involved in a mutual hydrogen-bond, but rather have amide groups that are oriented in opposite directions. Indeed, Cys38 forms a H-bond with Val31 of BI while Cysl42 forms a H-bond with Ser73 of B3; the profiles of Ser73 and Val31 do not overlap with the corresponding cysteines. Since the B2-B5 disulphide bond is close to the connection to the A2-A5 strands, any residual structure would likely contain hydrophobic contacts between the A and B sheet in this region. Chapter 3 - NMR Experiments 87 Figure 3.10: (a) Stability profile and (b) structure ofCys38 and Cysl42, the residues comprising the cysteine bond in CBDN2. Chapter 3 - NMR Experiments 88 3 . 2 . 1 1 C e l l o o l i g o s a c c h a r i d e B i n d i n g C l e f t Previous studies based on the measurement of chemical shift pertubation upon addition of cellooligosaccharide and intermolecular protein-sugar NOE interactions have identified the cleft across P-sheet A as the site of CBD N 2 sugar binding (Figure 1.9) (Brun, 2000). Since structure and function are so intimately related in proteins, there was interest in exploring whether these residues displayed above average HX stability. Since backbone fluctuations that result in HX would also disrupt the orientation of side chains involved in sugar binding, do those amino acids integral to sugar binding display high protection? All residues except Trp49 and Tyr91 displayed protection of at least 32 kJ/mol, though usually closer to 40 kJ/mol. Both Trp49 and Tyr 91 are in loop regions flanking the binding cleft, and therefore would be unlikely to exhibit such a high degree of protection. 3 . 2 . 1 2 E X 2 a n d p D D e p e n d e n t H X Extraction of residue specific stability from exchange profiles using equation 3.3 requires that exchange occur in the so-called EX2 regime. This necessitates that the rate of refolding (kd) be much greater that the intrinsic rate of exchange (kint) such that kM=K k^ ,. Except under conditions of moderately high pH, an EX2 mechanism is generally expected. All experiments occurred at pH = 5.6 or pD = 6.0. One method to establish that hydrogen exchange is occurring in the EX2 regime is to measure the overall amide exchange rate (kex) as a function of pD. Since the expression for kex in the EX2 regime has a linear dependence on kint, this rate is expected to increase -10 fold with every unit pH increment. In contrast, if exchange is occurring via the EX1 regime, then kex depends only on the rate of opening kop and will display pH independence. Since refolding rate is slowest at high [GuDCl], experiments as a function of pD (5.0, 5.5, 6.0, 6.5, and 7.0) were performed at 1.6 M GuDCl. If the rate of peptide exchange kint was still . Chapter 3 - NMR Experiments 89 significantly smaller than the refolding rate at this GuDCl concentration, then EX2 exchange could be extrapolated to all smaller [GuDCl]. Surprisingly, instead of finding a linear increase in log(kex) with pD (indicative of EX2) or a pD independence in the exchange rates (indicative of EX1), there was a decrease in kex for most residues in CBD N 2 (Figure 3.11). The only positive slope occurred for residues in loop regions, while all the residues involved in a P-strand displayed an unambiguous decrease in FLX rate as a function of pD. It is striking that the slope of log(kex) versus pD approaches -1 which is the inverse of the expected behavior (a slope of +1). Although surprising, this result can be explained by denaturation experiments (Chapter 2) that showed an increase of stability for CBD N 2 in the presence of GuDCl as a function of pD. Therefore under these conditions, any increase in kint due to an elevation in pD is offset by a larger increase in the stability of CBDN 2, culminating in a reduction in kex. Since GuDCl has such a perturbing effect, a similar experiment as a function of pD at 0 M GuDCl would ascertain EX2 exchange. Unfortunately, exchange occurs extremely slow (with time constants > 9 months for most residues) under non-denaturing conditions, making such an experiment extremely time consuming, and likely futile. Despite the unusual HX results, it is clear that EX1 is not the exchange regime. As well, stopped flow refolding experiments discussed in chapter 2 reveal refolding rates that are at least an order of magnitude larger than the internal peptide exchange rate (listed in Appendix D), the criteria for EX2 exchange. Chapter 3 - NMR Experiments (3-strand residues 0.8 -j 0.6-1 ^ T -r -1.2-J - L - - L Residue Figure 3.11: The linear fit oflog(kex) vs pD at 25 °C mapped onto the CBDN2 sequence. Chapter 3 - NMR Experiments 91 3 . 3 C o n c l u s i o n s Native state hydrogen exchange experiments in non-denaturing concentrations of GuDCl were performed on CBDN 2. Most residues exchanged with a mono-phasic dependence on denaturant, allowing simple extrapolation to AGHX(D20). There appeared to be no difference in exchange patterns between sheet-A and sheet-B, suggesting a two-state folding mechanism, or a free-energy landscape that has no significant crevasses. The most slowly exchanging residues were confined to the P-strands contributing to the hydrophobic core of the protein. Stabilities extracted from exchange rates showed linear dependence on [GuDCl]. The most protected residues had AGHX(D20) that agreed with global free-energy of unfolding measured by CD denaturation experiments when accounting for the perturbation from proline isomerizations. The most protected residues had AGHX(D20) ~ 42 kJ/mol. Among these were the cysteines involved in the disulphide bond required for stabilizing the native state of CBDN 2. Residues that were hydrogen bonded together showed similar exchange profiles, highlighting that H-bond breaking fluctuations must occur before the encounter complex can be formed, and base catalysis can proceed. Some of the residues that exchanged with low denaturant dependence were significantly buried, and exchange likely occurred though fluctuations requiring significant structural reorganization. No structural domains of dissimilar stability were observed. These results support a two state folding model for CBDN 2. Chapter 4 - Concluding Remarks 92 C h a p t e r 4 C o n c l u d i n g R e m a r k s The results from this work support a 2-state folding mechanism for CBDN 2. Global unfolding data described in detail in chapter 2, from both CD spectroscopy and DSC fit well to 2-state models. Proline isomerization, the probable basis for muliexponential refolding phases observed in stopped flow experiments, does not challenge the notion of 2-state folding. Instead it suggests the existence of small crevases in the conformational landscape, consistent with the "rough" landscape shown in Figurel.l. It is, however, the native state HX experiments as a function of low concentrations of denaturant that provide the conclusive proof of two-state unfolding. Several proteins that exhibited 2-state unfolding as monitored by fluorescence and CD spectroscopy (Cytochrome C, Ribonuclease H) revealed several "partially unfolded forms" (PUFs) when probed by native state HX. Identification of these PUF's relied on relating similar extrapolated free-energies of unfolding (AGHX(D20)) to residues in similar structural elements. Often a PUF consisted of a single or group of a-helices or an entire P-sheet. At the beginning of this project, no all-P-sheet protein had revealed PUF's. It was unclear whether this was a consequence of some intrinsic characteristic of P-sheet proteins, or whether the proteins studied were too small or too unstable to provide enough probes for detection of PUF's. CBDN 2 is exempt from both these latter considerations since it is very stable, and has at least 2 HX probes, although generally 4-5, for each of the 11 P-strands that comprise this protein. Results from these native state HX experiments revealed predominantly mono-phasic exchange that occurred solely through global unfolding fluctuations or via small-scale low energy fluctuations. No difference in exchange patterns existed between sheet-A and sheet-B. Chapter 4 - Concluding Remarks 93 It has recently been suggested that PUF's exist regardless of whether they can be detected (Rumbley, 2001). The results of this thesis, and a survey of native state HX literature suggest an alternative view. Namely that PUF's will only be observed in regions of sequence-local secondary structure. That is, independent fluctuations of structural sub-domains occur only if the backbone is not constrained by non-local hydrogen bonds, a-helices, which contain sequence-local H-bonds, are often found to be PUFs. Most P-sheet structures (with the exception of P-hairpins, and P-sheets composed of sequential p-hairpins) have non-local H-bonds. ct-P-proteins exhibit PUFs which include entire P-sheets, or P-strands that are sequence-local to a-helices (Rnase H). In CBDN 2, the p-hairpin between BI and B2 was the only structural element to display extrapolated stability within an extremely narrow range of values (although it was still in the global unfolding regime). While this "sequence-local" hypothesis does not negate the significance of PUF's, it does suggest that all proteins do not necessarily fold in a series of sub-native structural domains. Other folding motivators, like hydrophobic collapse may initially play a more important role than the formation of secondary structure H-bonds. Indeed, one might ask whether HX protection results from hydrophobic residue sequestering or maintenance of H-bonds? Evidence in this thesis supports the significance of H-bonds in HX protection. Several CBD N 2 residues involved in mutual H-bonds have completely overlapping HX profiles, with global AGHX(D20). While the overlapping HX profiles suggest that H-bonds must be broken before HX can occur, it is the global AGHX(D20) which indicates global unfolding fluctuations leading to exchange result from the H-bond breakage. If a single H-bond breakage was required for HX then measured rates would result in AGHX(D20) values that were on the order of H-bond formation (~2-3kJ/mol). Often several H-bonds must be broken before a fluctuation can lead to Chapter 4 - Concluding Remarks 94 exchange. Is the measured A G ^ the sum of broken H-bonds, or are there further factors significant in preventing fluctuations leading to exchange? An obvious additional contribution to HX protection could be the sequestering of hydrophobic residues, since fluctuations that expose the amide group to solvent catalysts, might also expose an energetically unfavorable hydrophobic side chain. In the case of a P-sandwich, residues along a strand have side chains whose orientation alternate between the hydrophobic interior and the solvent exposed exterior. If hydrophobic sequestering were indeed a significant factor in HX protection, one would expect to see an alternating pattern of stability along the sequence of a beta-strand. Fluctuations that disrupted only every second residue would require significant backbone strain, an improbable situation. Indeed Figure 3.8 shows no evidence of alternating patterns of stability along strands that form the P-sandwich, but also does not negate the importance of hydrophobic packing in stabilizing the tertiary contact between the 2 P-sheets of CBD N 2 . One possible way to establish the significance of hydrophobic interactions independent of backbone strain would be to do a native state HX experiment on a sequence-local P-sandwich -composed of 2 p-sheets that were both formed by consecutive p-hairpins, as opposed to CBD N 2 , whose P-sheets are not sequence-local due to its jelly-roll topology. If hydrophobic interactions were not significant in HX protection, one expects the separate sheets would represent individual PUFs. If hydrophobic interactions were significant, both sheets would exhibit similar stability. Native state HX data of stability mutants and stability homologues has yet to be potently exploited. Comparison of mesophilic and thermophilic homologues show a uniform increase in stability for all residues in the thermophilic protein. A point mutation that decreased the stability of CI2 showed effects only on the stability of residues that exchange through "global" fluctuations, while mutation that increased the stability of Rnase T l resulted in an increase in Chapter 4 - Concluding Remarks 95 extrapolated AGHX(D20) for all residues. With knowledge of residue specific stability of CBDN 2, it becomes simple to introduce expedient mutations. One can envision mutations aimed to stabilize residues that exchange through local fluctuations, such that they exchange through global fluctuations. How would this affect the stability of other residues in the protein? How would this perturb the conformational landscape? Would this alter the folding pathways or kinetics of this protein? Would it be useful in industrial applications? Not all these questions can be answered by native state exchange experiments. However, a better understanding of how mutations affect the type of data obtained from native state HX could lead to better interpretations of this data in the context of a proteins conformational landscape. It is, after all, the conformational landscape that provides insight to protein folding pathways, and the topology of this landscape that will act as the reference for strategies aimed at the manipulation of these protein folding pathways. Appendix A - Derivation of kex 96 A p p e n d i x A Der ivat ion of the genera l h y d r o g e n e x c h a n g e behav ior of nat ive p ro te ins Native state HX is often described by the following equation: Ko kml closed ^ open exchanged ( A . l ) k, cl where kop is the rate constant associated with the structural opening which can lead to exchange while kc, is the refolding rate to the native state. ktat is the intrinsic amide exchange rate of an unstructured peptide. The following chemical equations are obtained from the reaction described by (A.l). d[closed] dt = kcl [open] - kop [closed] (A.2) d [ ° ^ = koP[closed] -(kd + kiDt)[open] ( A 3 ) d[exchanged] , r = kmt[open] dt int F (AA) The method of eigenvalues can be used to solve a system of first order differential equations where there are n equations and n dependent variables. A matrix is set up where each row consists of one first order differential equation (ie (A.2), (A.3), or (A.4)), while the columns are filled by the coefficients of each variable. The eigenvalue, X, is subtracted from the diagonal elements of this matrix. If the determinant of the matrix is set to zero, then n solutions of X are obtained. In this case X represents the overall exchange rate for a native protein. Appendix A - Derivation ofkt 97 [closed] d[closed] dt op d[open] Kn dt op d[exchanged\ 0 dt The deterrninant of (A.5) is: [open] [exchanged] k„ -(kcl+kiat)-X k-.„, 0 0 -X = 0 (*_ + X)(kcl + tm + X)(-X) - {-X)kkd = 0 (A.5) (A.6) After some algebra, (A.6) reduces to: -X(X2 -X(kcl + kmt +k ) + k k-m) = 0 The solutions for X are: X — \),X — 2 op (A.7) (A.8) The first solution is trivial and can be ignored. It is useful to simplify the second 2 equations of (A.8). The first step is to factor out (kcl +kop +kint)2 from the root to obtain: (Ki+Kp+kiJ±(kcl+kop+kA 1 4kSMk int op X = (Ki+Kp+kint) (A.9) For a native protein, the closing rate is much faster than the opening rate, such that 4k k k « < k.. This means that —— is much smaller than 1. ( * d + * „ + * * ) a The binomial approximation states that ( l±s )* ~ l ± f e if 1 > » s . Thus the binomial approximation can be used to simplify the root in A.9. Appendix A - Derivation of k^ 98 *kiDlkop 2(kcl+kop+kint)2 4k k (*«< +K +kint)±(kcl +kop ( ^ d +^o p + i^nt) (A. 10) A simple equation for HX in native proteins can therefore be obtained from the negative root of (A. 10): KK (A.i i ) (k,+k +k.,) v cl op int' Appendix B - Protein Production and Purification 99 A p p e n d i x B P r o t e i n P r o d u c t i o n a n d P u r i f i c a t i o n JM101 electro-competent cells were electroporated with a pTUG plasmid containing the gene encoding Cellulomonas fimi wild type CBD N 2 gene, as well as the gene encoding for kanamycin resistance. Frozen stocks of electroporated cells were made from individual colonies on kanamycin agar plates and used within a 3 month period. B . I U n l a b e l l e d P r o t e i n P r o d u c t i o n i) E x p r e s s i n g t h e C B D s i n E . c o l i J M I O I Overnight cultures with 50 pg/ml of kanamycin were directly inoculated with frozen stock cells and incubated at 30 °C. Sterile TYP media (16 g of bacto-tryptone, 16 g yeast extract, 5 g NaCL, 2.5 g KjHPO,, (dibasic) per 1 litre of dFLO) was inoculated with overnight cultures of the JM101 cells. After approximately 2.5 hours, when the media OD^ was greater than 0.5, protein expression was induced with the addition of 1 mM IPTG. The media was left in the shaker at 30 °C for 36 to 48 hours for maximal protein yield. No significant increase in protein yield resulted from the use of a fermentor as compared to simple shaker Feinback flasks. i i) P r o t e i n P u r i f i c a t i o n CBD N 2 is produced in JM101 cells as a secreted protein. The cells were centrifuged at 5000 rpm for 20 minutes and the supernatant retained. Attempts to extract protein in th periplasmic by osmotic shock treatment did not prove appreciably fruitful. IM NaCl, 30 g of Avicel per llitre, and 3-5 mg of EDTA was added to the supernatant and the slurry stirred at 4 °C for at least 4 hours for complete Avicel binding. The suspension was then vacuum filtered through glass fiber filter and allowed to air dry. 30 g of Avicel per llitre was added to the flow-Appendix B - Protein Production and Purification 100 through and stirred for at least 8 hours to allow binding of any remaining protein, followed by vacuum filtering. The dried Avicel cakes were washed with with 500 mL 50 mM potassium phosphate buffer (pH 7), IM NaCl. CBD N 2 was then eluted with 500 mL of distilled FLO (4° C). The conductivity of this solution was measured and diluted with distilled FLO until the reading was less than 20 mS. This solution was filtered through 4 layers of glass fiber, and then through a 0.45 pm non-cellulose membrane to remove any residual Avicel. The supernatant was loaded on a Q-Sepharose column, equilibrated with 10 mM KPhos buffer at pFf=6.0, and then eluted via a salt gradient using a 10 mM KPhos and 1 M NaCl buffer at pH=6.0. CBD N 2 generally eluted from the column at about 55% NaCl. Purity was checked using SDS-PAGE and yields ranged between 60-120 mg/liter. Protein concentration was measured from the absorption at 280 nm using an extinction coefficient of 20504 1/cm • M (Brun et al, 2000). B . 2 1 5 N L a b e l e d P r o t e i n P r o d u c t i o n Labeled protein was produced in the same way as unlabeled with the following modifications. Overnight cultures in TYP media was centrifuged for 10 minutes at 6000 rpm and the supernatant disposed. Cells were re-suspended and then further inoculated in M9 minimal media (with lg per litre of 1 5N labeled Celltone). Typically the time taken to reach OD^O.5 was 3 hours. M9 media recipe consists of 1 litre of dH O and 6 g Na2HP04, 3 g KH P04, 0.5 g NaCl and 1 g 15[NH4]C1. After autoclaving, the following aliquots of filter-sterilized solutions were added: lOOul IM CaCl2, 2 ml IM MgS04, 1 ml 0.01M FeQ, 1 ml 1 mg/ml Thiamine, 20 ml 40% Glucose, and 1ml of 50 mg/ml kanamycin (Brun et al, 2000). Appendix C - Derivation ofAGu 101 A p p e n d i x C D e r i v a t i o n o f A G y a s a f u n c t i o n o f A H m , T m , a n d A C P We want to derive (C.l) from the van't Hoff equation. AG(T) = A H m ( i - l - ) - ACp (Tm-T-T ln(^-)) (C.l) The van't Hoff equation is written as follows: d i y , K ) = m ( C 2 ) MK R where T is Temperature, K is the equilibrium constant for unfolding, and AH is the change in enthalpy during unfolding, and R is the gas constant. We begin by rewriting d(-^) as --^-dT and substituting this into (C.2) to obtain: Rd(lnK) = ^-dT T2 (C.3) We want to integrate both sides from T to Tm, but must account for the fact that AH= AH(T). Making use of the integration rule: * ,6 b \udv = uv -\vdu dT T2 and setting dv = and u = AH we can integrate (C.3) as follows: T" AH R\dQnK) = — T 1 JjdAH ( C 4 ) T T Appendix C - Derivation of AGV 102 dT By multiplying the 2nd term on the right of (C.4) by — and by making use of the fact that dT AC„ = (C.4) can be written as: p dT T"> AH RldQnK) = — T 1 Tm T™dT - A C P J — p T T Integrating (C.5) one obtains: Add and subtract AHm from the right side of (C.7) to obtain: Making use of the fact that AHm = 7/AC and AH = TACn we can write (C.8) as: o m m p p x ' - AG(T) = - TACp - TACp ln(^) + Tm ACp - AHm Finally, grouping like terms together one obtains: (C.5) AG(Tm) AG(T)_AHm AH A C T (C.6) Making use of the fact that AG(Tm)=0, and multiplying both sides of (C.6) by T: TAH T (C.7) • AG(T) = -j^ -AH- TACp ln(^) r p i TT rri AG(T) = -AH-TACn ln(^) + AHm -AHm Tm p T m m (C.8) (C.9) AG(7> AHm(l-^-)-ACp(Tm-T-Tln(j-)) (C10) '.ix D - kinlfor CBDi A p p e n d i x D I n t r i n s i c p e p t i d e e x c h a n g e r a t e f o r C B D N 2 Table D . l : Intrinsic H^D Amide Exchange Rates for CBDN2 at 25°C and pD=6.0. AA K, (i/s) AA K, (i/s) AA K, (i/s) AA K, (i/s) S2 1.63E+01 D40 6.07E-02 V80 1.92E-02 P120 L3 8.77E-02 L41 2.91E-02 R81 1.46E-01 P121 D4 5.17E-02 P42 V82 5.54E-02 D122 4.82E-02 S5 2.59E-01 G43 1.79E-01 L83 3.19E-02 G123 2.06E-01 E6 1.03E-01 G44 4.60E-01 V84 2.06E-02 D124 1.24E-01 V7 2.36E-02 Q45 2.84E-01 G85 2.26E-01 A125 1.10E-01 E8 3.74E-02 G46 4.93E-01 E86 7.64E-02 P126 L9 3.11E-02 N47 7.64E-01 G87 2.20E-01 G127 1.79E-01 L10 2.71E-02 P48 G88 4.60E-01 0128 2.84E-01 Pll W49 3.74E-02 G89 4.60E-01 V129 5.29E-02 H12 6.07E-01 D50 6.50E-02 A90 2.47E-01 A130 1.21E-01 T13 9.62E-01 A51 1.10E-01 Y91 8.98E-02 F131 9.62E-02 S14 6.21E-01 G52 3.11E-01 R92 2.26E-01 H132 1.21E+00 F15 1.92E-01 L53 6.50E-02 T93 2.36E-01 L133 2.97E-01 A16 1.92E-01 V54 2.06E-02 A 94 2.65E-01 G134 1.92E-01 E17 5.17E-02 Y55 6.50E-02 F95 9.62E-02 K135 2.26E-01 S18 2.77E-01 N56 5.80E-01 E96 5.93E-02 A136 2.20E-01 L19 8.77E-02 G57 6.50E-01 Q97 1.36E-01 G137 3.11E-01 G20 1.92E-01 V58 4.93E-02 G98 4.93E-01 A138 2.47E-01 P21 P59 S99 5.80E-01 Y139 8.98E-02 W22 3.74E-02 V60 1.92E-02 A100 3.34E-01 E140 5.80E-02 S23 3.04E-01 G61 2.26E-01 P101 F141 6.81E-02 L24 8.77E-02 E62 7.64E-02 L102 2.53E-02 C142 6.81E-01 Y25 5.54E-02 G63 2.20E-01 T103 8.77E-02 1143 8.98E-02 G26 3.49E-01 E64 7.64E-02 G104 4.93E-01 S144 2.31E-01 T27 2.10E-01 S65 2.77E-01 E105 7.64E-02 Q145 3.83E-01 S28 6.21E-01 Y66 1.79E-01 P106 V146 5.29E-02 E29 1.03E-01 V67 3.74E-02 A107 9.62E-02 S147 2.84E-01 P30 L68 3.19E-02 T108 1.42E-01 L148 8.77E-02 V31 1.92E-02 S69 2.42E-01 R109 3.19E-01 T149 8.77E-02 F32 6.97E-02 F70 1.92E-01 E110 8.57E-02 T150 2.26E-01 A33 1.92E-01 T71 1.63E-01 Yl l l 6.36E-02 S151 6.21E-01 D34 8.38E-02 A 72 2.65E-01 A112 1.88E-01 A152 5.29E-03 G35 2.06E-01 S73 3.92E-01 F113 9.62E-02 R36 2.97E-01 A 74 3.34E-01 T114 1.63E-01 M37 2.71E-01 T75 1.42E-01 S115 6.21E-01 C38 7.64E-01 P76 N116 1.03E+00 V39 9.62E-02 D77 3.34E-01 L117 9.19E-02 M78 1.08E-01 T118 8.77E-02 P79 F119 1.52E-01 Appendix D - kjnlfor CBDI 104 D . I C o d e f o r t h e C - p r o g r a m t h a t c a l c u l a t e s i n t r i n s i c p e p t i d e e x c h a n g e r a t e s #include <stdio.h> #include <stdlib.h> #include <string.h> #include <math.h> #define R 1.98709 /* kcal/molK */ #definearNT-1.32 #definebrNT 1.62 #define a i d 0.96 #defineblCT-1.8 #define T 298 #define Tref 278 /* Activation Energies are in kcal/mol */ #define EaA 14 #define EaB 17 #defineEaW19 #definekA 1.19 #define kB 9.90 #define kW -2.5 #define pD 6.0 #define pOD (15.65-6.0) /* This program calculates the intrinsic exchange rate of a polypeptide according to the Protein Structure Function and Genetics paper by Bai et al vol 17 pp75-86 1993 It requires 2 input files: one with the aa seq in 3 letter code with modification to the 3 letter code as described below called aaseq_letters.txt and a file with the modification constants determine by Bai et al, both in the directory of compilation. One must also specify an output file for the krc constants. to compile: cc -lm krc.c -o krc.exe then to run it just type krc.exe at the prompt Some comments: It is important to properly define your chosen kA,kB,kW as well as pOD (check your dissociation constant) and pD above. The aa sequence file must be 3 letter coded and all cis PROLINES must have the modified 3 letter code PRC instead of PRO. Trans proline maintain the 3 letter PRO. Non Disulfide bonded Cysteines must have the 3 letter code CYN, whereas disulfide Cysteines maintain the CYS 3 letter code. All titratable side chains are assumed to be in charged form, and if the pH you interested in is below 4.4 or above 6.5 then the bai_aa.txt must have the uncharged data replaced, according to the Bai paper. NOte that calculating krc for proteins with non bonded cysteines at a pH above 8.5 is impossible since data for charged non-disulphide-cysteines is unavailable. Appendix D - kjnlfor CBDl Each amino acid should be on a seperate line in the sequence file. The aa sequence file must exist in the directory of the compiled program, and file name must be modified prior to compiling in the following code. */ main() { int flag=l; float ar,al,bl,br,brnext,arnext,krc; char AAL[4],AAR[4],BAI[4]; FILE *cfPtr,*cfPtrBAI,*cfPtrOUT; /* AA SEQUENCE FILE NAME */ if ((cfPtr = fopen("aaseq_letters.txt","r"))==NULL) printf("Sequence file can't be opened, baby. \n"); else { if ((cfPtrBAI = fopen("bai_aa.txt","r"))==NULL) printf("File bai_aa.txt can't be opened, baby. \n"); else { if ((cfPtrOUT = fopen("krc.txt","w"))==NULL) printf("Output file can't be opened, baby. \n"); else fputs(" AA BAI krc \n",cfPtrOUT); { do { rewind(cfPtrBAI); fscanf(cfPtr,"%s\n",AAL); if (feof(cfPtr)> flag=-l; switch (flag) { case 1: ar=arNT; br=brNT; flag+=l; break; case-1: al=alCT; /*printf("hi %f %f\n",pow(10,5),(10A5));*/ bl=blCT; krc= (pow(10,(kA+al+ar-pD))*(exp((-EaA/R)*((l/T)-(l/Tref)))) + pow(10,(kB+bl+br-pOD))*(exp((-EaB/R)*((l/T)-(l/TreO))) + pow(10,(kW+bl+br))* EaW/R)*((l/T)-(l/Tref))))); Appendix D - kjnlfor CBDj krc=krc/60; printf(" AA is %s and BAI is %s and krc is %f\n",AAL,BAI,krc); fprintf(cfPtrOUT,"%s %s %f\n",AAR,AAL,krc); break; default: fscanf(cfPtrBAi;'%s%f%f%f%f\n",BAI,&al,&arnext,&bl,&brnext); while (Ifeof(cfPtrBAI) && (strcmp(AAL,BAI) !=0)) fscanf(cfPtrBAi;,%s%f%f%f%i\n",BAI,&al,&arnext,&bl,&brnext); if (strcmp(AAL,BAI) == 0) { if ((strcmp(AAL,"PRO") == 0) || (strcmp(AAL,"PRC") == 0)) { printf("AAL is %s BAI is %s\n",AAL,BAI); fprintf(cfPtrOUT,"%s %s \n",AAR,AAL); } else { krc= (pow(10,(kA+al+ar-pD))*(exp((-EaA )^*((l/T)-(1/Tref)))) + pow(10,(kB+bl+br-pOD))*(exp((-EaB/R)*((l/T)-(l/Tref)))) + pow(10,(kW+bl+br))*(exp((-EaW/R)*((l/T)-(l/rref))))); krc=krc/60; printf(" AA is %s and BAI is %s and krc is %f\n",AAL,BAI,krc); fprintf(cfPtrOUT,"%s %s %f\n",AAR,AAL,krc); } } /*end of strcmp is *i else printf("Couldn't find the AA %s\n",AAL); ar=arnext; br=brnext; strcpy(AAR,AAL); flag+=l; break; } /*end of switch */ } while (Ifeof(cfPtr)); /*end of while Ifeof cfPtr*/ } /* end of file OUT else */ } /* end of file cfPtrBAI selection loop ELSE */ fclose(cfPtr); fclose(cfPtrBAI); } /* END OF ELSE */ return 0; } /*END OF MAIN*/ Appendix E - Normalization Factor for the HX Time Series A p p e n d i x E N o r m a l i z a t i o n F a c t o r f o r t h e H X T i m e S e r i e s f r o m t h e 3 H 5 2 r e s o n a n c e o f L e u 1 3 3 a n d t h e H r e s o n a n c e o f A s n 4 7 . Time [GuDCl ] Series [GuDCl] T ime ( M ) Data Peak Area N o r m (M) Series Peak Area N o r m Point Factor Data Point Factor 0.0 1-19 1.71E+06 1.00 0.4 1-13 7.77E+05 1.00 20 1.67E+06 0.98 14-17 8.08E+05 1.04 21 1.70E+06 0.99 18 7.95E+05 1.02 22 1.74E+06 1.02 19 7.40E+05 0.95 23 1.74E+06 1.02 20 8.28E+05 1.07 24 1.85E+06 . 1.08 21 8.48E+05 1.09 25 1.85E+06 1.08 22 8.50E+05 1.09 26 1.80E+06 1.05 23 8.06E+05 1.04 27 1.65E+06 0.96 24 8.99E+05 1.16 28 1.78E+06 1.04 25 8.76E+05 1.13 29 1.81E+06 1.06 26 8.08E+05 1.04 30 1.69E+06 0.99 27 8.27E+05 1.06 mean 1.75E+06 1.02 mean 8.22E+05 1.06 stdev 6.83E+04 0.04 stdev 4.32E+04 0.06 sterr 1.97E+04 0.01 sterr 1.25E+04 0.02 0.2 1-9 1.30E+06 1.00 0.6 1-9 8.56E+05 1.00 10-18 1.32E+06 1.02 10-15 9.39E+05 1.10 19 1.23E+06 0.95 16-17 9.30E+05 1.09 20 1.41E+06 1.08 18 8.43E+05 0.98 21 1.36E+06 1.05 19 1.40E+06 1.64 22 1.34E+06 1.03 21 9.59E+05 1.12 23 6.61E+05 0.51 22 9.15E+05 1.07 24 1.46E+06 1.12 23 1.51E+06 1.76 25 1.55E+06 1.19 24 1.44E+06 1.68 26 1.57E+06 1.21 25 1.44E+06 1.68 27 1.52E+06 1.17 26 1.32E+06 1.54 28 1.17E+06 0.90 mean 1.14E+06 1.33 29 1.31E+06 1.01 stdev 2.75E+05 0.32 mean 1.32E+06 1.02 sterr 8.28E+04 0.10 stdev 2.33E+05 0.18 sterr 6.46E+04 0.05 0.8 1-19 9.11E+05 1.00 20 9.55E+05 1.05 21 9.76E+05 1.07 22 9.54E+05 1.05 23 1.36E+06 1.49 24 1.45E+06 1.59 25 1.51E+06 1.66 26 1.39E+06 1.53 27 1.26E+06 1.38 28 1.12E+06 1.23 mean 1.19E+06 1.30 stdev 2.32E+05 0.25 sterr 7.33E+04 0.08 Appendix E - Normalization Factor for the HX Time Series [GuDCl ] Time Series Peak Area N o r m [GuDCl ] T ime Series Peak Area N o r m ( M ) Data Point Factor (M) Data Point Factor 1.0 1-17 7.63E+05 1.00 1.6 1-6 9.37E+05 1.00 18-20 8.05E+05 1.06 7-16 9.73E+05 1.04 21 8.27E+05 1.08 17-21 9.59E+05 1.02 22 8.41E+05 1.10 22 1.04E+06 1.11 23 8.66E+05 1.13 23 1.02E+06 1.09 24 8.06E+05 1.06 24 1.01E+06 1.08 25 8.72E+05 1.14 26 1.04E+06 1.11 26 8.65E+05 1.13 27 1.02E+06 1.09 27 8.61E+05 1.13 28 1.01E+06 1.08 28 1.37E+06 1.80 29 1.04E+06 1.11 29 1.43E+06 1.87 30 1.01E+06 1.08 30 1.01E+06 1.32 mean 1.01E+06 1.07 31 1.31E+06 1.72 stdev 3.46E+04 0.04 32 1.27E+06 1.66 sterr 1.04E+04 0.01 mean 9.93E+05 1.30 stdev 2.40E+05 0.31 1.8 1-17 7.97E+05 1.00 sterr 6.42E+04 0.08 18 7.85E+05 0.98 21-22 8.02E+05 1.01 1.2 1-15 3.73E+05 1.00 23 6.74E+05 0.85 16 4.89E+05 1.31 mean 7.65E+05 0.96 17 4.89E+05 1.31 stdev 6.08E+04 0.08 18-19 4.99E+05 1.34 sterr 3.04E+04 0.04 20 5.17E+05 1.39 21 5.50E+05 1.47 22 5.52E+05 1.48 23 5.64E+05 1.51 24 5.60E+05 1.50 25 5.04E+05 1.35 mean 5.10E+05 1.37 stdev 5.64E+04 0.15 sterr 1.78E+04 0.05 1.4 1-15 6.29E+05 1.00 16 7.90E+05 1.26 17 7.59E+05 1.21 18 7.55E+05 1.20 mean 7.33E+05 1.17 stdev 7.12E+04 0.11 sterr 3.56E+04 0.06 Bibliography 109 B I B L I O G R A P H Y Alexandrescu, A. T., Jaravine, V. A., Dames, S. A. & Lamour, F. P. (1999). NMR hydrogen exchange of the OB-fold protein LysN as a function of denaturant: the most conserved elements of structure are the most stable to unfolding. J Mol Biol 289(4), 1041-54. Arrington, C. B. & Robertson, A. D. (1997). Microsecond protein folding kinetics from native-state hydrogen exchange. Biochemistry 36(29), 8686-91. Arrington, C. B. & Robertson, A. D. (2000). Microsecond to minute dynamics revealed by EX1-type hydrogen exchange at nearly every backbone hydrogen bond in a native protein. J Mol Biol 296(5), 1307-17. Arrington, C. B. & Robertson, A. D. (2000). Correlated motions in native proteins from MS analysis of NH exchange: evidence for a manifold of unfolding reactions in ovomucoid third domain. J Mol Biol 300(1), 221-32. Bai, Y., Milne, J. S., Mayne, L. & Englander, S. W. (1993). Primary structure effects on peptide group hydrogen exchange. Proteins 17(1), 75-86. Bai, Y., Milne, J. S., Mayne, L. & Englander, S. W. (1994). Protein stability parameters measured by hydrogen exchange. Proteins 20(1), 4-14. Bai, Y. & Englander, S. W. (1994). Hydrogen bond strength and beta-sheet propensities: the role of a side chain blocking effect. Proteins 18(3), 262-6. Bai, Y., Sosnick, T. R., Mayne, L. & Englander, S. W. (1995). Protein folding intermediates: native-state hydrogen exchange. Science 269(5221), 192-7. Baker, D. (1998). Metastable states and folding free energy barriers [news]. Nat Struct Biol 5(12), 1021-4. Baker, D. (2000). A surprising simplicity to protein folding. Nature 405(6782), 39-42. Bibliography 110 Balbach, J., Forge, V., Lau, W. S., van Nuland, N. A., Brew, K. & Dobson, C. M. (1996). Protein folding monitored at individual residues during a two- dimensional NMR experiment. Science 274(5290), 1161-3. Bhuyan, A. K. & Udgaonkar, J. B. (1998). Two structural subdomains of barstar detected by rapid mixing NMR measurement of amide hydrogen exchange. Proteins 30(3), 295-308. Bork, P., Downing, A. K., Kieffer, B. & Campbell, A. D. (1996). Structure and distribution of modules in extracellular proteins. Quaterly Reviews pf Biophysics 29(2), 119-167. Branden, C. & Tooze, J. (1991). Introduction to Protein Structure. 1 edit, Garland Publishing Inc., New York. Brun, E., Johnson, P. E., Creagh, A. L., Tomme, P., Webster, P., Haynes, C. A. & Mcintosh, L.P. (2000). Structure and binding specificity of the second N-terminal cellulose- binding domain from Cellulomonas fimi endoglucanase C. Biochemistry 39(10), 2445-58. Chamberlain, A. K., Handel, T. M. & Marqusee, S. (1996). Detection of rare partially folded molecules in equilibrium with the native conformation of RNaseH. Nat Struct Biol 3(9), 782-7. Chamberlain, A. K. & Marqusee, S. (1997). Touring the landscapes: partially folded proteins examined by hydrogen exchange. Structure 5(7), 859-63. Chamberlain, A. K. & Marqusee, S. (1998). Molten globule unfolding monitored by hydrogen exchange in urea. Biochemistry 37(7), 1736-42. Chamberlain, A. K., Fischer, K. F., Reardon, D., Handel, T. M. & Marqusee, A. S. (1999). Folding of an isolated ribonuclease H core fragment. Protein Sci 8(11), 2251-7. Chamberlain, A. K. & Marqusee, S. (2000). Comparison of equilibrium and kinetic approaches for determining protein folding mechanisms. Adv Protein Chem 53, 283-328. Bibliography 111 Chu, R. A., Takei, J., Barchi, J. J., Jr. & Bai, Y. (1999). Relationship between the native-state hydrogen exchange and the folding pathways of barnase. Biochemistry 38(43), 14119-24. Clarke, J. & Fersht, A. R. (1996). An evaluation of the use of hydrogen exchange at equilibrium to probe intermediates on the protein folding pathway. Fold Des 1(4), 243-54. Clarke, J., Itzhaki, L. S. & Fersht, A. R. (1997). Hydrogen exchange at equilibrium: a short cut for analysing protein-folding pathways? Trends Biochem Sci 22(8), 284-7. Clarke, J. & Itzhaki, L. S. (1998). Hydrogen exchange and protein folding. Curr Opin Struct Biol 8(1), 112-8. Clarke, J., Cota, E., Fowler, S. B. & Hamill, S. J. (1999). Folding studies of immunoglobulin-like beta-sandwich proteins suggest that they share a common folding pathway. Structure Fold Des 7(9), 1145-53. Connelly, G. P., Bai, Y., Jeng, M. F. & Englander, S. W. (1993). Isotope effects in peptide group hydrogen exchange. Proteins 17(1), 87-92. Cota, E. & Clarke, J. (2000). Folding of beta-sandwich proteins: three-state transition of a fibronectin type III module. Protein Sci 9(1), 112-20. Coutinho, J. B., Gilkes, N. R., Warren, R. A., Kilburn, D. G. & Miller, R. C , Jr. (1992). The binding of Cellulomonas fimi endoglucanase C (CenC) to cellulose and Sephadex is mediated by the N-terminal repeats. Mol Microbiol 6(9), 1243-52. Creagh, A. L., Ong, E., Jervis, E., Kilburn, D. G. & Haynes, C. A. (1996). Binding of the cellulose-binding domain of exoglucanase Cex from Cellulomonas fimi to insoluble microcrystalline cellulose is entropically driven. Proc Natl Acad Sci USA 93(22), 12229-34. Creamer, T. P. & Rose, G. D. (1994). Alpha-helix-forming propensities in peptides and proteins. Bibliography 112 Proteins 19(2), 85-97. Creighton, T. E., Ed. (1992). Protein Folding. 1 edit. New York: W.H.Freeman and Company. Creighton, T. E. (1993). Proteins. Second edit, W. H. Freeman and Company, New York. Dabora, J. M. & Marqusee, S. (1994). Equilibrium unfolding of Escherichia coli ribonuclease H: characterization of a partially folded state. Protein Sci 3(9), 1401-8. Dabora, J. M., Pelton, J. G. & Marqusee, S. (1996). Structure of the acid state of Escherichia coli ribonuclease Ffl. Biochemistry 35(37), 11951-8. de Alba, E., Rico, M. & Jimenez, M. A. (1997). Cross-strand side-chain interactions versus turn conformation in beta- hairpins. Protein Sci 6(12), 2548-60. Dill, K. A. & Chan, H. S. (1997). From Levinthal to pathways to funnels. Nat Struct Biol 4(1), 10-9. Englander, S. W. & Kallenbach, N. R. (1983). Hydrogen exchange and structural dynamics of proteins and nucleic acids. Quarterly Reviews of Biophysics 16, 521-655. Englander, S. W. & Mayne, L. (1992). Protein folding studied using hydrogen-exchange labeling and two- dimensional NMR. Annu Rev Biophys Biomol Struct 21, 243-65. Englander, S. W., Sosnick, T. R., Englander, J. J. & Mayne, L. (1996). Mechanisms and uses of hydrogen exchange. Curr Opin Struct Biol 6(1), 18-23. Englander, S. W., Mayne, L., Bai, Y. & Sosnick, T. R. (1997). Hydrogen exchange: the modern legacy of Linderstrom-Lang. Protein Sci 6(5), 1101-9. Englander, S. W. (1998). Native-state HX. Trends Biochem Sci 23(10), 378; discussion 379-81. Fischer, K. F. & Marqusee, S. (2000). A rapid test for identification of autonomous folding units in proteins. J Mol Biol 302(3), 701-12. Bibliography 113 Fuentes, E. J. & Wand, A. J. (1998). Local stability and dynamics of apocytochrome b562 examined by the dependence of hydrogen exchange on hydrostatic pressure. Biochemistry 37(28), 9877-83. Fuentes, E. J. & Wand, A. J. (1998). Local dynamics and stability of apocytochrome b562 examined by hydrogen exchange. Biochemistry 37(11), 3687-98. Gerstein, M. & Levitt, M. (1997). A structural census of the current population of protein sequences. Proc Natl Acad Sci USA 94(22), 11911-6. Ghaemmaghami, S., Fitzgerald, M. C. & Oas, T. G. (2000). A quantitative, high-throughput screen for protein stability. Proc Natl Acad Sci USA 97(15), 8296-301. Gilkes, N. R., Jervis, E., Henrissat, B., Tekant, B., Miller, R. C., Jr., Warren, R. A. & Kilburn, D. G. (1992). The adsorption of a bacterial cellulase and its two isolated domains to crystalline cellulose. JBiol Chem 267(10), 6743-9. Goedken, E. R., Raschke, T. M. & Marqusee, S. (1997). Importance of the C-terminal helix to the stability and enzymatic activity of Escherichia coli ribonuclease H. Biochemistry 36(23), 7256-63. Grantcharova, V. P. & Baker, D. (1997). Folding dynamics of the src SH3 domain. Biochemistry 36(50), 15685-92. Hilser, V. J. & Freire, E. (1997). Predicting the equilibrium protein folding pathway: structure-based analysis of staphylococcal nuclease. Proteins 27(2), 171-83. Hollien, J. & Marqusee, S. (1999). Structural distribution of stability in a thermophilic enzyme. Proc Natl Acad Sci USA 96(24), 13674-8. Hollien, J. & Marqusee, S. (1999). A thermodynamic comparison of mesophilic and thermophilic ribonucleases H. Biochemistry 38(12), 3831-6. Bibliography 114 Hosszu, L. L., Baxter, N. J., Jackson, G. S., Power, A., Clarke, A. R., Waltho, J. P., Craven, C. J. & Collinge, J. (1999). Structural mobility of the human prion protein probed by backbone hydrogen exchange. Nat Struct Biol 6(8), 740-3. Huyghues-Despointes, B. M., Scholtz, J. M. & Pace, C. N. (1999). Protein conformational stabilities can be determined from hydrogen exchange rates. Nat Struct Biol 6(10), 910-2. Huyghues-Despointes, B. M., Langhorst, U., Steyaert, J., Pace, C. N. & Scholtz, J. M. (1999). Hydrogen-exchange stabilities of RNase Tl and variants with buried and solvent-exposed Ala -> Gly mutations in the helix. Biochemistry 38(50), 16481-90. Itzhaki, L. S., Neira, J. L. & Fersht, A. R. (1997). Hydrogen exchange in chymotrypsin inhibitor 2 probed by denaturants and temperature. J Mol Biol 270(1), 89-98. Jaravine, V. A., Rathgeb-Szabo, K. & Alexandrescu, A. T. (2000). Microscopic stability of cold shock protein A examined by NMR native state hydrogen exchange as a function of urea and trimethylamine N- oxide. Protein Sci 9(2), 290-301. Jerala, R. & Zerovnik, E. (1999). Accessing the global minimum conformation of stefin A dimer by annealing under partially denaturing conditions. J Mol Biol 291(5), 1079-89. Johnson, P. E., Creagh, A. L., Brun, E., Joe, K., Tomme, P., Haynes, C. A. & Mcintosh, L. P. (1998). Calcium binding by the N-terminal cellulose-binding domain from Cellulomonas fimi beta-l,4-glucanase CenC. Biochemistry 37(37), 12772-81. Johnson, P. E., Brun, E., MacKenzie, L. F., Withers, S. G. & Mcintosh, L. P. (1999). The cellulose-binding domains from Cellulomonas fimi beta-1, 4- glucanase CenC bind nitroxide spin-labeled cellooligosaccharides in multiple orientations. J Mol Biol 287(3), 609-25. Jonasson, P., Kjellsson, A., Sethson, I. & Jonsson, B. H. (1999). Denatured states of human carbonic anhydrase II: an NMR study of hydrogen/deuterium exchange at tryptophan-indole-H(N) sites. FEBS Lett 445(2-3), 361-5. Bibliography 115 Kay, L., Keifer, P. & Saarinen, T. (1992). J Am Chem Soc 114, 10663-10665. Keck, J. L. & Marqusee, S. (1996). The putative substrate recognition loop of Escherichia coli ribonuclease H is not essential for activity. J Biol Chem 271(33), 19883-7. Kern, G., Handel, T. & Marqusee, S. (1998). Characterization of a folding intermediate from HTV-1 ribonuclease H. Protein Sci 7(10), 2164-74. Koradi, R. (2000). MOLMOL 2.6 edit. Spectrospin, Faellanden. Lee, B. & Richards, F. M. (1971). The interpretation of protein structures: estimation of static accessibility. J Mol Biol 55(3), 379-400. Li, R. & Woodward, C. (1999). The hydrogen exchange core and protein folding. Protein Sci 8(8), 1571-90. Liu, H., Farr-Jones, S., Ulyanov, N. B., Llinas, M., Marqusee, S., Groth, D., Cohen, F. E., Prusiner, S. B. & James, T. L. (1999). Solution structure of Syrian hamster prion protein rPrP(90-231). Biochemistry 38(17), 5362-77. Llinas, M. & Marqusee, S. (1998). Subdomain interactions as a determinant in the folding and stability of T4 lysozyme. Protein Sci 7(1), 96-104. Llinas, M., Gillespie, B., Dahlquist, F. W. & Marqusee, S. (1999). The energetics of T4 lysozyme reveal a hierarchy of conformations. Nat Struct Biol 6(11), 1072-1078. Loftus, D., Gbenle, G. O., Kim, P. S. & Baldwin, R. L. (1986). Effects of denaturants on amide proton exchange rates: a test for structure in protein fragments and folding intermediates. Biochemistry 25(6), 1428-36. Marqusee, S. & Sauer, R. T. (1994). Contributions of a hydrogen bond/salt bridge network to the stability of secondary and tertiary structure in lambda repressor. Protein Sci 3(12), 2217-Bibliography 116 25. Mayo, S. L. & Baldwin, R. L. (1993). Guanidinium chloride induction of partial unfolding in amide proton exchange in RNase A [see comments]. Science 262(5135), 873-6. McCallister, E. L., Aim, E. & Baker, D. (2000). Critical role of beta-hairpin formation in protein G folding. Nat Struct Biol 7(8), 669-73. Merkel, J. S., Sturtevant, J. M. & Regan, L. (1999). Sidechain interactions in parallel beta sheets: the energetics of cross- strand pairings. Structure Fold Des 7(11), 1333-43. Minor, D. L. & Kim, P. S. (1994). Measurement of the beta-sheet-forming propensities of amino acids. Nature 367(6464), 660-3. Minor, D. L. & Kim, P. S. (1994). Context is a major determinant of beta-sheet propensity. Nature 371(6494), 264-7. Minor, D. L. & Kim, P. S. (1996). Context-dependent secondary structure formation of a designed protein sequence. Nature 380(6576), 730-4. Molday, R. S., Englander, S. W. & Kallen, R. G. (1972). Primary structure effects on peptide group hydrogen exchange. Biochemistry 11(2), 150-8. Moore, J. W. & Pearson, R. G. (1981). Kinetics and Mechanism. 3rd edit, John Wiley & Sons, Toronto. Muhandiram, D. & Kay, L. (1994). J Mag Res Series B 103, 203-216. Mullins, L. S., Pace, C. N. & Raushel, F. M. (1997). Conformational stability of ribonuclease Tl determined by hydrogen- deuterium exchange. Protein Sci 6(7), 1387-95. Myers, J. K. & Oas, T. G. (1999). Contribution of a buried hydrogen bond to lambda repressor folding kinetics. Biochemistry 38(21), 6761-8. Bibliography 117 Neira, J. L., Itzhaki, L. S., Otzen, D. E., Davis, B. & Fersht, A. R. (1997). Hydrogen exchange in chymotrypsin inhibitor 2 probed by mutagenesis. J Mol Biol 270(1), 99-110. Oberhauser, A. F., Marszalek, P. E., Carrion-Vazquez, M. & Fernandez, J. M. (1999). Single protein misfolding events captured by atomic force microscopy. Nat Struct Biol 6(11), 1025-1028. Pace, C. N., Shirley, B. A. & Thomson, J. A. (1992). Measuring the conformational stability of a protein. In Protein Folding (Creighton, T., ed.), pp. 309-329. W.H. Freeman, New York. Parker, M. J., Dempsey, C. E., LOrch, M. & Clarke, A. R. (1997). Acquisition of native beta-strand topology during the rapid collapse phase of protein folding. Biochemistry 36(43), 13396-405. Parker, M. J. & Marqusee, S. (1999). The cooperativity of burst phase reactions explored. J Mol Biol 293(5), 1195-210. Parker, M. J. & Marqusee, S. (2000). A statistical appraisal of native state hydrogen exchange data: evidence for a burst phase continuum? /Mol Biol 300(5), 1361-75. Parker, M. J. & Marqusee, S. (2001). A Kinetic Folding Intermediate Probed by Native State Hydrogen Exchange. J Mol Biol 305(3), 593-602. Perrett, S., Clarke, J., Hounslow, A. M. & Fersht, A. R. (1995). Relationship between equilibrium amide proton exchange behavior and the folding pathway of barnase. Biochemistry 34(29), 9288-98. Plaxco, K. W., Spitzfaden, C , Campbell, I. D. & Dobson, C. M. (1997). A comparison of the folding kinetics and thermodynamics of two homologous fibronectin type III modules. J Mol Biol 270(5), 763-70. Qian, H. & Chan, S. I. (1999). Hydrogen exchange kinetics of proteins in denaturants: a generalized two-process model. J Mol Biol 286(2), 607-16. Bibliography 118 Ragona, L., Fogolari, F., Romagnoli, S., Zetta, L., Maubois, J. L. & Molinari, H. (1999). Unfolding and refolding of bovine beta-lactoglobulin monitored by hydrogen exchange measurements. J Mol Biol 293(4), 953-69. Ragona, L., Confalonieri, L., Zetta, L., De Kruif, K. G., Mammi, S., Peggion, E., Longhi, R. & Molinari, H. (1999). Equilibrium unfolding CD studies of bovine beta-lactoglobulin and its 14-52 fragment at acidic pH. Biopolymers 49(6), 441-50. Raschke, T. M. & Marqusee, S. (1997). The kinetic folding intermediate of ribonuclease H resembles the acid molten globule and partially unfolded molecules detected under native conditions [published erratum appears in Nat Struct Biol 1997 Jun;4(6):505]. Nat Struct Biol 4(4), 298-304. Raschke, T. M. & Marqusee, S. (1998). Hydrogen exchange studies of protein structure. Curr Opin Biotechnol 9(1), 80-6. Raschke, T. M., Kho, J. & Marqusee, S. (1999). Confirmation of the hierarchical folding of RNase H: a protein engineering study [published erratum appears in Nat Struct Biol 2000 May;7(5):431]. Nat Struct Biol 6(9), 825-31. Richardson, J. S. (1977). beta-Sheet topology and the relatedness of proteins. Nature 268(5620), 495-500. Rumbley, J., Hoang, L., Mayne, L. & Englander, S. W. (2001). Inaugural Article: An amino acid code for protein folding. Proc Natl Acad Sci USA 98(1), 105-112. Sadqi, M., Casares, S., Abril, M. A., Lopez-Mayorga, O., Conejero-Lara, F. & Freire, E. (1999). The native state conformational ensemble of the SH3 domain from alpha- spectrin. Biochemistry 38(28), 8899-906. Shtilerman, M., Lorimer, G. H. & Englander, S. W. (1999). Chaperonin function: folding by forced unfolding. Science 284(5415), 822-5. Bibliography 119 Skelton, N. J., Kordel, J., Akke, M. & Chazin, W. J. (1992). Nuclear magnetic resonance studies of the internal dynamics in Apo, (Cd2+)1 and (Ca2+)2 calbindin D9k. The rates of amide proton exchange with solvent. J Mol Biol 227(4), 1100-17. Spudich, G. & Marqusee, S. (2000). A change in the apparent m value reveals a populated intermediate under equilibrium conditions in escherichia coli ribonuclease HI [In Process Citation]. Biochemistry 39(38), 11677-83. Stryer, L. (1995). Biochemistry. 4rth edit, W.H. Freeman and Company, New York. Taylor, J. R. (1982). An Introduction to Error Analysis. First edit, University Science Books, Mill Valley. Tito, P., Nettleton, E. J. & Robinson, C. V. (2000). Dissecting the hydrogen exchange properties of insulin under amyloid fibril forming conditions: a site-specific investigation by mass spectrometry. J Mol Biol 303(2), 267-78. van Holde, K. E., Johnson, W. C. & Ho, P. S. (1998). Physical Biochemistry, Prentice Hall, Upper Saddle River. Wildegger, G., Liemann, S. & Glockshuber, R. (1999). Extremely rapid folding of the C-terminal domain of the prion protein without kinetic intermediates. Nat Struct Biol 6(6), 550-3. Woodward, C. & Li, R. (1998). The slow-exchange core and protein folding. Trends Biochem Sci 23(10), 379-81. Wooll, J. O., Wrabl, J. O. & Hilser, V. J. (2000). Ensemble modulation as an origin of denaturant-independent hydrogen exchange in proteins. J Mol Biol 301(2), 247-56. Xu, Y., Mayne, L. & Englander, S. W. (1998). Evidence for an unfolding and refolding pathway in cytochrome c. Nat Struct Biol 5(9), 774-8. Bibliography 120 Yee, A., Szymczyna, B. & O'Neil, J. D. (1999). Backbone dynamics of detergent-solubilized alamethicin from amide hydrogen exchange measurements. Biochemistry 38(20), 6489-98. Yi, Q., Scalley, M. L., Simons, K. T., Gladwin, S. T. & Baker, D. (1997). Characterization of the free energy spectrum of peptostreptococcal protein L. Fold Des 2(5), 271-80. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061491/manifest

Comment

Related Items