H. Jane Dyson Home Page

Primary research techniques include NMR spectroscopy for study of structure and dynamics, mass spectrometry, and equilibrium and kinetic CD and fluorescence spectroscopy. Molecular cloning techniques have been an extremely powerful addition to the arsenal, allowing us to prepare labeled proteins in the amounts necessary for structural studies by NMR. (Bracketed numbers refer to references in the publications list).

How does the amino acid sequence code for the three-dimensional structure of the protein?

Initiation of the Protein Folding Process

The initial steps in protein folding are the most relevant for the exploration of the link between sequence and folded structure. They are also the most difficult to study experimentally. We tackled this problem by using short peptide fragments of proteins as a model system for the earliest events in the folding process. The initial focus of this work was to determine the sequence specificity for residual structure in short peptides in solution, as a model system for secondary structure propensities in unfolded proteins that might act as folding initiation sites. Reverse turns were found to be highly sequence-dependent (Refs 9, 47, 87 from the Publication List), and in some cases contained a sufficiently high population of turn conformations that the structure of the turn could be calculated (48). In order to determine which parts of proteins potentially contained folding initiation sites, three large sets of peptides were studied, corresponding to the complete sequences (in pieces) of Themiste zostericola myohemerythrin (31, 75), French bean plastocyanin (32) and sperm whale myoglobin (35-37, 72). These studies remain the definitive work in this field. More recently the emphasis has been on the use of new NMR techniques to probe residual structure in denatured forms of the full-length protein.

NMR Characterization of Unfolded Proteins and Folding Pathways

NMR is the method of choice for the examination of the solution conformational ensembles of equilibrium folding intermediates and unfolded states of proteins. This work represents a major thrust of research in my group at the present time and is conducted in an equal collaboration with Dr. Peter E. Wright. We have succeeded in assigning the backbone resonances for acid-unfolded apomyoglobin (127) and for low-salt unfolded plastocyanin (132). The chemical shift indices reveal differences in the backbone conformational ensemble between the two proteins, which indicate that apomyoglobin populates a backbone conformations, while plastocyanin populates b. This result is entirely consistent with the previously-published peptide studies. In addition, the extremely exciting observation of anomalies in the dynamics of the acid-unfolded form of apomyoglobin has led to the hypothesis that these indicate transient contacts between the termini of the protein. Further work using spin-labeled apomyoglobin (manuscript in preparation) has verified these observations and supported the hypothesis: these results provide the first experimental observation of the initial events in the collapse of a protein during folding. The most exciting aspect is that the collapse appears to be specific – only those areas of the molecule that are involved in the folding intermediate and present in the final folded structure are found to be in transient contact in the acid-unfolded form.

There are no published studies of the structure of a folding intermediate. X-ray crystallography is impossible due to the fluxional nature of such intermediates, and, while NMR is obviously the method of choice, none of the recognized intermediates of other proteins such as lactalbumin give good NMR spectra. The apomyoglobin intermediate appears to be an exception, with excellent NMR spectra. This work has been published (82,112).

Folding experiments on apomyoglobin using quench-flow methods combined with mass spectrometry (94) provide definitive evidence that this protein folds by a single major pathway, and that the observed folding intermediate is obligatory. The folding pathway has been further characterized by the study of several mutants, including a double mutant where the H helix is destabilized (95) and a mutant where the E helix is stabilized by the replacement of the functional distal histidine with a phenylalanine (117). Recent mutant studies have focused on the B helix region, where changes have been made at a number of positions, resulting in the stabilization of the protein in some cases and destabilization in others. Kinetic and thermodynamic studies on these mutants reveal a fascinating variety of folding mechanisms, even though the mutant proteins differ in only one or two residues. Two manuscripts are in preparation on this work.

The question as to whether the folding pathway of a protein family is conserved if the structure is preserved, or whether the folding pathway depends more critically on details of the amino acid sequence has been addressed by comparing two proteins, leghemoglobin and myoglobin, which are very similar in structure but have very low sequence homology. In this case, it is found that, while both proteins fold via early formation of a helical burst phase intermediate, there are major differences in the constitution of this intermediate according to details of the amino acid sequence differences between them (115).

Structure of Chaperone Domains

The Escherichia coli chaperone DnaJ contains a number of domains whose functions have been mapped, but for which structural information is sketchy. The cysteine-rich (CR) domain of DnaJ was thought to contain zinc binding sites, and the spacing and number of the cysteines was thought to indicate that it would fold as a classic zinc-finger domain. We cloned and expressed the CR domain of DnaJ, and determined the solution structure by NMR (116). It proved to form a novel fold incorporating two zinc ions and 6 short b-strands in an overall V-shaped b-hairpin structure. The difference between this structure, which is associated with chaperones which bind polypeptides, and those of the DNA-binding zinc fingers is probably related to the difference in the function.

Unfolded but Functional Proteins

It has recently been shown that a number of protein systems are unfolded or only partly folded until they bind ligand or substrate. While this has been known for many years to apply to peptide hormones, we now realize that certain transcription factors, cyclin-dependent kinase inhibitors and other proteins are unfolded in the absence of their natural receptors. In addition, folding processes have been implicated in a number of disease states, such as Alzheimer’s Disease and the prion diseases such as kuru and bovine spongiform encephalopathy (“mad cow” disease). We have been involved in the structure elucidation of some of the fragments of the prion protein (a collaboration with Drs. Fred Cohen and Stan Prusiner at UCSF). This promising new field has already given us hints that the state of folding of the fragments will be crucial in the conversion between cellular and infectious forms of the prion protein. A paper has been published on the structure and dynamics of the long fragment (PrP29-231) of the hamster prion protein (79). A recent new direction has been the exploration of the binding of Cu(II) to repeated sequences (termed “octarepeats”) in this fragment, which appear to bind this metal exclusively and highly specifically (96). A detailed examination of the dynamics of two fragments of the Syrian hamster prion protein (PrP90-231 and PrP29-231) (125) revealed a number of anomalous features that are related to the presence of two domains, one folded and one unfolded, in the same molecule. This work will have wide applicability in the study of other systems with significant proportions of unfolded structure. More recently, we have been engaged in the structure determination of a prion protein homologue present in tissues other than the brain. This protein, termed Doppel, has a similar overall fold to that of the prion protein, but differences in local structure suggest that its function may differ significantly (126).

An invited review synthesizing our recent thoughts on unfolded, functional proteins has appeared (100). We have also made a number of studies of baseline NMR parameters that can be used particularly in the studies of unfolded and partly folded proteins, but which are applicable also to the rapid elucidation of secondary structure in folded proteins. These include proton (56) and backbone ¹³C and ¹⁵N (119) random coil chemical shifts, and methodology and tabulations for the correction of backbone random coil chemical shifts for the effects of local sequence (130). Sequence correction of random coil chemical shifts has been an important missing factor in the more accurate prediction of secondary structure from chemical shifts, and the method developed in our laboratory appears to be widely applicable and particularly useful for ¹³CO shifts. These methods have been integrated into a widely used NMR analysis program, and are referred to on the Web site of the NMR data depository BioMagResBank.

Can we understand function in terms of protein structure and dynamics?

Thioredoxin

A number of protein systems have been studied to elucidate the link between chemistry, structure, dynamics and function. The system that has been studied the longest in my laboratory is E. coli thioredoxin, a long-standing collaboration with Professor Arne Holmgren of the Karolinska Institute, Sweden. This small (108-residue) thiol-disulfide oxidoreductase has a multitude of functions in the cell, including the vital process of ribonucleotide reduction to form deoxyribonucleotides for DNA synthesis. Thioredoxins occur in all living organisms and in viruses; mammalian thioredoxins have been shown to have a vital role in cellular control mechanisms and have recently been implicated in human disease processes: thioredoxin is found at elevated levels in the serum of AIDS patients. Thioredoxins are related structurally to protein disulfide isomerases, but differences in redox potential mean that there is a clear functional difference between the two. Indeed, one of the primary functions of thioredoxin in the cell is as a protein disulfide reductase, a function vital for the prevention of misfolded proteins in vivo. An extensive series of NMR experiments culminated in the complete resonance assignment for both oxidized (disulfide) and reduced (dithiol) thioredoxins (13, 26, 42) and the calculation of high-resolution structures of the two forms of the protein (46). Backbone dynamics were measured (34) and this information, together with a study of amide proton hydrogen exchange (54) allowed us to determine that functional differences in phage systems between oxidized and reduced thioredoxin were due to differences in the flexibility of the molecules, rather than to structural differences. The contribution from my laboratory has been particularly significant in the delineation of the mechanism of E. coli thioredoxin. The reduction reaction of thioredoxin depends critically on the movement of protons, during the two-electron-two-proton transfer reaction as a substrate disulfide is reduced. An extensive series of experiments on the pH-dependence of the NMR spectrum has given important insights into the complex relationship between the mechanism of this enzyme and local structure at the active site, including the presence of conserved, buried charged residues (24, 60, 62, 73, 92). During the course of this work, we advanced a revolutionary hypothesis that the mechanism included a shared proton between the two active-site cysteines (60). A great deal of attention has been paid to this in the recent literature. In a collaboration with Dr. Don Bashford at TSRI, we were able to show that such a structure would be consistent with a recent high-resolution X-ray crystal structure and with the electrostatic environment of the active site (84).

Glutaredoxin-2

A major recent effort has gone into the determination of the structure of another member of the thioredoxin-glutaredoxin family, E. coli glutaredoxin-2 (Grx-2). This protein is present in large quantities in the bacterial cell, but was not detected in earlier assays for glutaredoxin since it has no activity in the ribonucleotide reductase assay, the original screen for glutaredoxins. Until recently its primary function has been a mystery, but our collaborator in Sweden, Dr. Holmgren, was able to show that it is part of a pathway to detoxify arsenic compounds. For us, the intriguing aspect of Grx-2 was its lack of sequence homology with other glutaredoxins and thioredoxins and its significantly greater size (215 amino acids, compared to 90-110 residues for other members of the family). Resonance assignments are complete for reduced Grx-2 (103) and the structure determination is complete (133). Interestingly, the molecule appears to contain a mini-domain that is similar in topology to the smaller glutaredoxins, with the rest of the molecule consisting of long helices. According to our collaborator, the small glutaredoxin-like domain does not appear to be stable by itself. The closest structural similarity of Grx-2 is with glutathione-S-transferases, but no detectable GST activity is found for the molecule. These studies illustrate the economical use of similar protein scaffoldings to perform diverse tasks in the cell.

Rusticyanin

A clear relationship between structure and function was seen for another system, a metalloprotein obtained from an acidophilic bacterium. Thiobacillus ferrooxidans rusticyanin has the astonishing property that it contains a highly stable Type I or “blue” copper site at pHs well below 4. In fact the protein is stable and functional in sulfuric acid at pH 0.5, in spite of a copper coordination site that includes two histidines, which would normally be protonated at such a low pH. The solution structure of reduced rusticyanin (69), obtained utilizing a novel method of gene synthesis (58,77) clearly shows that the acid-stability of rusticyanin resides in the composition of the environment of the copper site, which contains a high proportion of aromatic groups, causing steric hindrance to the dissociation of the ligand histidines. Further collaborative work with Dr. Chris Bender has resulted in new approaches to the pulsed EPR of copper-containing proteins (78).

Dynamics in Enzyme Action

The relationship between dynamics of the polypeptide chain and enzyme catalysis is being explored. The success of structure-based drug design has so far been quite modest. This may be because the design process incorporates only static structural information. Our working hypothesis is that a requirement for efficient enzyme catalysis may well be that the active site be flexible, and that enzymes have therefore evolved to incorporate this flexibility. To test this hypothesis we will use mutagenesis coupled with full characterization of changes in enzymatic function, structure and dynamics. As an initial approach to the subject, two very different enzyme systems are being studied, in collaboration with Dr. Peter Wright: dihydrofolate reductase, for which we already have clear evidence that dynamics play a role in catalysis, and a metallo-b-lactamase for which there is evidence from the crystal structure of flexibility in the active site region. Both of these enzymes are clinically important potential targets for anti-cancer drugs in the case of DHFR and to combat antibiotic resistance in the case of the metallo-b-lactamase. Our approach should be exceptionally powerful in elucidating the role of dynamics in catalysis. A number of papers have been published on the preliminary NMR characterization of the metallo-b-lactamase system (89, 90,104) and a paper describing the dynamics of the system in the presence and absence of a tight-binding inhibitor has been published (120). Further work on this system will include the characterization of mutant proteins designed to explore the role of a “flexible flap” over the active site cleft in the catalytic mechanism and in allowing a broad range of substrate specificity.

Protein-Protein Interactions: Immunogenic Peptides, Antibodies and Cell Adhesion Domains

The function of a number of clinically important systems requires interactions between proteins. NMR provides a particularly sensitive tool for exploration of such interactions. We have been concerned for many years with the question of the immunogenicity of peptides (in collaboration with Drs. Richard Lerner and Ian Wilson). A hypothesis that immunogenic peptides may have a higher-than-usual propensity for folded conformations in solution (8) has been explored by studying the solution conformational preferences of an extensive series of such peptides (3, 8, 9, 10, 20, 25, 28, 29, 40, 67,70). These studies have added greatly to our understanding of the nature of peptide immunogenicity.

Related to the issue of design for catalysis is a long-standing project on the characterization and redesign of the Fv fragment of a catalytic antibody (in collaboration with Drs. Richard Lerner and Stephen Benkovic). This work is extremely promising for the understanding of the influence of structure on function for enzymes. Since enzymes have evolved over (in most cases) millions of years, they are exquisitely tuned to the reactions they catalyze, and may also be tolerant of mutations. By contrast, catalytic antibodies have much lower efficiency and specificity: a knowledge of the local structure and dynamics of the catalytic site should allow us to experiment with changing residues so as to change or enhance specificity and efficiency. We are in the process of studying the Fv fragment of the catalytic antibody NPN43C9 by NMR, and have succeeded in demonstrating the presence of covalently bound enzyme-substrate complexes by mass spectrometry (45,55), and the NMR characterization of the Fv fragment both free and in complex with a product inhibitor is now complete (105). A paper describing the backbone dynamics of these species is in preparation, and the work on the 43C9 antibody fragment will continue in studies of the hapten complex. New directions in this project include the commencement of studies of a family of aldolase catalytic antibodies in an effort to understand differences in substrate specificity. These studies should have far-reaching effects on our thinking about enzyme catalysis as well as for design of specific catalysts.

A project initiated in collaboration with scientists from Novartis obtained a number of exceptionally exciting results. Resonance assignments and a solution structure for LFA-1 have been completed (107, 111); these results were used directly by Dr. Uli Hommel of Novartis Basel to screen compounds for binding. Several promising new leads have already arisen from this work. Our interest in this system is in the understanding of the interactions between LFA-1 and its physiological partner ICAM-1, in parallel with a collaborative effort with Dr. Bruce Cunningham to probe the interactions between the immunoglobulin-like domains of NCAM (99,135). These projects illustrate the excellent sensitivity of the NMR method as a tool to probe protein-protein interactions: the NMR spectrum is used directly to map the sites of interaction.

Other Collaborative Projects

I have recently become involved in the collaborative efforts in Dr. Peter Wright’s lab, investigating the mechanisms of transcriptional activation. This work ties in well with the protein-interaction theme of my research. It is also highly relevant to the work on unfolded proteins, since a number of transcriptional activation domains are unfolded or molten-globular in the absence of their physiological ligands (100). I have been particularly involved in the resonance assignment and structure determination of a number of interaction domains; a number of publications have resulted from these collaborative projects (81, 86, 88, 93, 98, 102, 108, 113, 114, 118, 121, 122, 123, 124).

Biographical Information

Education

University of Sydney, B.Sc.(Hons, Class 1), 1973 (Biochemistry)

University of Sydney, Ph.D., 1977 (Inorganic Chemistry)

Professional Record

1973-1976 Tutor, Department of Inorganic Chemistry, University of Sydney,

New South Wales, Australia (supervisor James Beattie)

1977-1978 Postdoctoral Fellow, Department of Biology, Massachusetts Institute

of Technology, Cambridge, MA (supervisor Paul Schimmel).

1979-1984 UNESCO Lecturer, School of Chemistry, University of New South Wales

Australia.

1984-1988 Research Associate, Department of Molecular Biology, The Scripps

Research Institute.

1988-1992 Assistant Member, Department of Molecular Biology, The Scripps

Research Institute.

1992-1997 Associate Professor, Department of Molecular Biology, The Scripps

Research Institute.

1997-2001 Associate Professor with Tenure, Department of Molecular Biology,

The Scripps Research Institute

2001-Present Professor, Department of Molecular Biology,

The Scripps Research Institute

Last revised: March 13, 2007

H. Jane Dyson Home Page

Contents

Information

Position

Location

Department

Mailing Address

Favorite Links

Contact Information

E-mail address

Web address

Office phone

Administrative Assistant

Fax

Research Themes

How does the amino acid sequence code for the three-dimensional structure of the protein?

Initiation of the Protein Folding Process

NMR Characterization of Unfolded Proteins and Folding Pathways

Structure of Chaperone Domains

Unfolded but Functional Proteins

Can we understand function in terms of protein structure and dynamics?

Thioredoxin

Glutaredoxin-2

Rusticyanin

Dynamics in Enzyme Action

Protein-Protein Interactions: Immunogenic Peptides, Antibodies and Cell Adhesion Domains

Other Collaborative Projects

Biographical Information

Education

Professional Record