RICERCA ITALIANA     

Research program

Theoretical and experimental approach to non-native states of proteins: formation of amyloid fibrils, unstructured and unfolded proteins.

University Co-ordinator

Università degli Studi di ROMA "La Sapienza" - SCIENZE BIOCHIMICHE - ROMA(RM)

Research Unit Leader

Carlo TRAVAGLINI ALLOCATELLI

Description

It is our aim to obtain a complete quantitative description of the folding pathways of different proteins. In particular, we will be focussed on developing innovative experimental techniques in order to describe the folding mechanism from microseconds to seconds of some representative model systems. Furthermore, the experimental work will be employed to benchmark molecular dynamics as well as Monte Carlo simulations in collaboration with the UR of Padova. The synergy between experiment and simulation will be centre stage in defining the basic rules of protein folding. We aim to dissect the folding problem in few directed questions: i) at which stage a given polypeptide chain selects its folding pathway? ii) are there common structural features in protein folding transition states? iii) can we apply the rules of protein folding for a productive de novo protein design? Finally, in collaboration with the UR of Firenze, we aim to understand the close relationships between productive folding and misfolding. In this context, it will be crucial to apply the novel approach of "the pre-sculpted energy landscape" introduced by the UR of Padova (13).
Model systems
i) The cytochrome c family (cyt c551 from Pseudomonas aeruginosa and the cyt c552 from Hydrogenobacter thermophylus and from Thermus thermophylus) represents a paradigmatic system for protein folding studies. The cyt c family consists of a typical globular all a-helical fold. Furthermore, the presence of a covalently bound heme represents a unique feature to monitor structure formation via different spectroscopic probes. Previous studies from this and others labs have already identified a small number of key interactions that drive the folding mechanism of different c-type cytochromes. Overall, it appears that the topological positions of such key residues is conserved within the protein family and similar folding intermediates and transition states may be identified across the whole family (14-16). On the basis of these observation we recently proposed the presence of a consensus folding mechanism for the cytochrome c family (17). Finally, we already developed expression vectors which allow heterologous expression of recombinant c-type cytochromes and their site-directed mutants in E.coli.
ii) the second PDZ (post-synaptic density-95/discs large/zonula occludens-1) domain from murine tyrosine phosphates (PDZ2 from PTP-BL). PDZ domains represent a large family of protein-interaction modules associated with a variety of unrelated proteins with different functions. The structure of PDZ domains corresponds to a compact globular fold, composed of six b strands and two a helices. We already developed expression vectors which allow heterologous expression of recombinant murine PDZ2 and site-directed mutants in E.coli.
Main goals of the project are:
1) Detailed description of the early events in protein folding.
Refolding kinetics is generally followed in a stopped-flow apparatus (dead time > 1 ms) by rapid dilution of a protein that has been dissolved in high denaturant. In this type of experiment, the kinetically resolved amplitude is often different than expected from the total signal change between the fully unfolded and the native state. This phenomenon, known as ‘burst phase collapse', implies that there is an early, very rapid, event typically involving the formation of a compact collapsed species (Dphys). Various proteins, including members of the cytochrome c family, show this behaviour (9, 18-19). A fundamental issue is whether the collapsed state Dphys represents an intermediate with at least some native-like structural features or an ensemble of structures with no defined secondary and tertiary structural elements. Early theoretical studies predicted that the burst phase collapse might be modelled by an un-cooperative transition (20). Following this view, the denatured chain would be characterised by a peculiar plasticity and would assume a different structure depending on experimental conditions. In this scenario, defined "downhill", the burst phase effect would reflect a non-specific response of the denatured polypeptide chain to the change in solvent composition without a barrier separating a continuum ensemble of states (second-order transition). Up to date, the limited experimental characterizations of the burst phase collapse do not clarify whether the fast transition between U and Dphys follows a second-order transition or a classical barrier limited first-order process. We aim to clarify this initial event, and particular attention will be given to the structural and mechanistic properties of the early collapsed species Dphys.
Different spectroscopic techniques have been used in combination with super-rapid mixing to show that, in the case of eukaryotic cytochrome c, the collapsed species is characterized by some structural properties (a-helical content – ref. 9; radius of gyration – ref. 18) typical of a compact and partially structured state. However, the presence and/or the role of specific short and long-range contacts it still not clear since, up to date, no f-value analysis has been carried out on this collapsed species. Our aim is to identify such interactions and to define their role in the stabilization of Dphys.
First year. We will build and set up a prototype of a continuous-flow instrument on the basis of the apparatus used by A. Fersht (Cambridge, UK). The mixing properties of the instruments should guarantee reaction dead-times of about 50-100 microseconds and it will be equipped with a high resolution, UV-enhanced, CCD camera. Particular attention will be given to the development of an instrument with the possibility to study chemical reactions monitoring both the fluorescence (conformational changes) and the absorbance properties (useful to study heme-containing proteins as the c-type cytochromes).
Second year. Design, expression and purification of site-directed mutants of different bacterial cytochromes c (cyt c551 from P. aeruginosa; cyt c552 from H. thermophilus) mapping in homologous regions of the two proteins; extensive kinetic characterization of the mutants obtained both in the microseconds and milliseconds time range. Some site directed mutants of P. aeruginosa cyt c551, mapping at the interface between the N- and C-terminal alfa helices, are already available in the laboratory. Structural description, obtained by f-value analysis, of the collapsed species formed in the microsec time-range for the two cytochromes and definition of its mechanistic role.
2) Transition and intermediate states and the selection of the folding mechanism.
A complete understanding of the folding process requires the description of all the transient species populated during the reaction. Although the structure of the native state is readily achieved at atomic resolution (X-ray diffraction, solution NMR…), the structural heterogeneity of the denatured state and the transient character of transition and intermediate states preclude their direct characterization. This part of the project is therefore focused on the description of intermediate and transition states populated in the two model systems described above employing both an experimental (phi-value analysis; this laboratory), and a theoretical approach (in collaboration with the U.R. Padova).
PDZ Domains. We have recently produced two site-directed mutants in which we introduced Trp residues necessary to carry out fluorescence-monitored equilibrium and kinetic experiments (pseudo-wt Y43W and F14W). The results have already demonstrated the presence of two transition states (TS1 and TS2) in the unfolding pathway of this small globular domain; ref. 21). In this part of the project we aim to depict the structure of TS1 and TS2 through phi-value analysis. Achievement of this goal will allow to discriminate between a nucleation-condensation and a diffusion-collision mechanism; these alternative folding mechanism are in fact probably selected on the basis of the structural properties of the transition states (see Introduction).
A second goal of this sub-project concerns the interesting hypothesis that residues playing key roles in the folding process are also important for function (22). It should be pointed out indeed that PDZ domains are present in a variety of proteins, where they are responsible of recognition events among proteins involved in different cellular processes (23-24).
First year. Design, expression and purification of site-directed mutants starting from the two pseudo-wt mapping throughout the structure of this small beta-domain. Structural description of the transition state TS1 and TS2 (through phi-value analysis) and definition of the folding mechanism.
Second year. We will study the binding reaction of the PDZ domain with small synthetic peptides (6-8 residues) representing the consensus recognition sequence. The availability of the two Trp-containing variants should allow to follow the bimolecular binding reaction by monitoring variations of the intrinsic fluorescence. If this were not the case, we will make use of dansylated peptides in order to follow variations of the extrinsic fluorescence. Rapid-mixing fluorescence spectroscopy will allow to determine the dissociation constants (Kd) for both the pseudo-wt and for the mutants employed in the characterization of the folding mechanism. Description of the transition state for the binding reaction and comparison with the transition states for folding.
Cytochromes c. We have recently hypothesized that in this protein family the folding mechanism is dictated by the common topological properties, independently of the phylogenetic distance, while sequence specific physical-chemical properties are responsible of the modulation of intermediate and transition state stability. This hypothesis will be further evaluated using a theoretical approach by molecular dynamics and/or Monte Carlo experiments (in collaboration with U.R. Padova). The specific goals which we aim to obtain are: i) structural characterization of transient states through phi-value analysis, ii) modification of the folding mechanism on rational backgrounds, iii) definition of the role of the heme cofactor in the folding process.
First year. Structural characterization of transient states by phi-value analysis. We aim to test the hypothesis which postulates that the structural region involving the two terminal a-helices plays a prominent role in the folding mechanism of this protein family. To this end we plan to design and characterize the folding kinetics of different site-directed mutants mapping throughout the structure of the protein. Over-and-above the deletions of specific tertiary contacts, we plan to alter, by site-directed mutagenesis, the stability and/or the propensity of the N- and C-terminal a-helices. Such a strategy will allow us to verify the hypothesis that accumulation of the obligatory intermediate identified in the millisecond time range (16) depends on the stability of these elements of secondary structure. However, an alternative hypothesis postulates that the degree of internal packing of the protein may control not only the stability of the native state, but also the stability of the partially folded intermediate state. Characterization of the kinetic folding mechanism of variants of P. aeruginosa c551 designed on the basis of H. thermophilus cyt c552, which presents a reduced number of internal packing defects, will help to test this hypothesis. Furthermore, such a strategy will allow us to study the possibility to alter the folding mechanism at will, opening the interesting perspective to correct potentially harmful misfolding processes (see the research sub-project of the U.R. Firenze).
A different aspect, up to date poorly investigated, concerns the unusually high number of variants with phi-value > 1 in the case of cyt c551 from P. aeruginosa. We will explore, employing also advanced computational methods (in collaboration with the U.R. Padova), if these residues do make interactions important in the selection of alternative folding pathways. Interestingly, the existence of parallel folding pathways, suggested by others employing computational methods (25), has been also proposed by our group in the case of cyt c551 from P. aeruginosa (26).
Second year. Definition of the role of the heme prosthetic group in the folding mechanism of c-type cytochromes. Thermodynamic analysis and folding kinetics characterization of the apo form (devoid of the heme group) of cyt c552 from H. thermophilus: because of its unusual stability it is indeed possible that the apo form of this cytochrome may retain some structural organization, contrary to what observed in the case of eukaryotic c-type cytochromes. The experimental protocol which allow to break the thioether bonds linking the heme group to the two Cys residues of the conserved motif Cys-Xaa-Xaa-Cys-His has been already developed in our laboratory. It will be interesting to test the hypothesis which postulates that the structure of the intermediate identified in the folding mechanism of this protein may reflect the structural organization of the apo-cyt c552. The description of the folding mechanism of apo-cyt c will shed new light in the folding of this protein family, bridging in vitro folding studies with the biogenesis and maturation processes of c-type cytochromes in the cell.