RICERCA ITALIANA     

Identification of folding and misfolding determinants by site-directed mutagenesis.

Università degli Studi di Firenze

Abstract

Dialysis-related amyloidosis represents an inevitable and severe complication of long-term hemodialysis and is characterised by the deposition in essential tissues, such as the skeletal muscle, of fibrillar aggregates, termed amyloid aggregates, formed by beta-2-microglobulin (b2-m). The research project presented here aims at extending the present knowledge on the involvement of both specific residues of b2-m sequence, and its different structural regions in the folding and misfolding events. Thus, this project will focus on several topics that require the involvement and the cooperation of different research units (RU). The dealed topics are:
1) Production of b2-m variants. We will design a number of mutations that are useful to study both the processes of folding and aggregation.
2) Characterisation of the solution structure of the variants by standard 1D and heteronuclear 2D and 3D NMR spectroscopy. Gross structural comparisons between different mutants will be obtained by analyisis of the paramagnetic attenuation pattern measured in the presence and absence of a spin label or even from the differential water accessibilty pattern. NMR spectroscopy will be also used to obtain the profiles of segmental mobilities between different mutants and information on the stability of the different secondary structure elements. Real time NMR determinations will then enable us to measure the kinetics of the slow refolding intermediates of several b2-m variants and infer structural conclusions about the nature of the slow-evolving intermediates.
3) Determination of the thermodynamic and kinetic parameters of the folding and unfolding reaction of b2-m. The conformational stabilities of the mutants will be determined by acquiring equilibrium GdnHCl-induced unfolding curves using intrinsic fluorescence and/or circular dichroism as spectroscopic probes. The folding and unfolding rate constants will be determined for each protein variant at different Gdn-HCl concentrations using a sopped-flow device. To gain insight into the b2-m folding mechanism, we will determine the phi-values of each mutant. This methodology will provide information on the level of structure that the mutated residue forms in the folding transition state and in the major folding intermediates in order to elucidate the folding mechanism at a molecular level.
4) Determination of the solvent exposure and structural flexibility of b2-m. Integrated strategies such as hydrogen/deuterium exchange or limited proteolysis coupled to mass spectrometry, will be applied to the study of b2-m variants to determine the flexible and/or solvent exposed regions of the transient folding and misfolding intermediates.
5) Fibrillogenesis. We will monitor the aggregation of the b2-m mutants both spectroscopically, using specific assays, and by atomic force microscopy (AFM). We will also evaluate the effect of surface charge and hydrophobicity on the aggregation process by zeta potential measurements and by the ability to interact with model membranes such as liposomes, planar membranes and phospholipid monolayers. Another task of this research project is the determination of the aggregation propensity of the whole b2-m sequence by comparisons between the aggregation rate of the variants and those of wild-type b2-m.
6) Study of the process of fibrillogenesis in the presence of fibrillar collagen. A recent finding of some of the applicant RUs shows that collagen fibrils can trigger b2-m amyloid aggregation and explain the tissue specificity of the amyloid deposits. We will develop this topic studying the process of fibril growth in the precence of type I and type II collagen fibrils and we will also investigate the effect of heparin on b2-m amyloid aggregation in presence of collagen fibrils. <<<

Principal Investigator

Fabrizio Chiti Università degli Studi di FIRENZE

Research Objectives

Dialysis-related amyloidosis represents an inevitable and severe complication of long-term haemodialysis. Under this pathological condition, protein aggregates known as amyloid fibrils, accumulate in essential tissues, such as the skeletal muscle, interfering with their normal functions. A major constituent of the amyloid fibrils related to this pathological condition is beta-2-microglobulin (b2-m). The research program described here aims at characterising the underlying events of folding and misfolding of such protein at a molecular level.
The first objective of the research program is to gain insight, at a residue level, of the process by which b2-m folds, i.e. convert from its unfolded state to its fully folded, functional state. The proposed research will use previous models of folding, previously characterised by the research units participating to this PRIN as well as other internationally recognised groups, to investigate the folding process with deep molecular insight. We propose to identify the residues that participate to the formation and stabilisation of the native state and of the various partially folded states that accumulate during folding.
The second objective of the program is to determine, with similar molecular insight, the residues or regions of the sequence that promote the process of amyloid aggregation of b2-m. The process of aggregation will be studied not just for free b2-m, but also for the protein in the presence of collagen. Indeed, collagen has been found to promote amyloid formation of b2-m in vitro, to be associated with amyloid deposits in vivo and is therefore suggested to play a major role in the amyloidogenesis process occurring naturally by b2-m and to determine its tissue selectivity.
With a full characterisation of both the folding and aggregation processes the network of the interconversions of the various conformational states accessible to b2-m will be characterised not just in their thermodynamic and kinetic terms, but also by elucidating at a molecular level the mechanisms of these conformational steps. This will constitute a fundamental step in the elucidation of the pathogenesis of Dialysis-related amyloidosis and will provide the key targets for therapeutic design. <<<

Timescale

24 months

National and international background

Amyloidosis is an heterogeneous disease category in which peculiar proteins undergo a pathological self aggregation into insoluble fibrils (1). These fibrils accumulate in target tissues causing irreversible damage. In the last few years the RU of Pavia has gathered a significant collection of natural amyloidogenic proteins isolated from different biological sources, in particular the different isoforms of b2-microglobulin (b2-m), present in amyloid deposits of different patients affected by dialysis–related amyloidosis (DRA), have been characterised. Native b2-m is 99 residues in length and has a seven stranded beta-sandwich fold typical of the immunoglobulin superfamily. The protein contains two beta-sheets (one comprising beta-strands A, B, D, E and the other beta-strands C,F and G) that are held togheter by a single disulphide bond linking Cys25 and Cys80. In vivo, b2-m is present as the non-polymorphic light chain of the class I major histocompatibility complex (MHCI). As part of its normal catabolic cycle, b2-m dissociates from the MHCI complex and is trasported in the serum to the kidneys where the majority is degraded (2). In haemodialised patients (&gt;20000 people in Italy, and &gt;250000 in the EU) the b2-m concentration increases by 20–50 fold and by a mechanism that is currently not well understood, b2-m self aggregate into amyloid fibrils that tipically accumulate in the musculoskeletal system (3). b2-m is mainly present in amyloid fibrils as full length protein without amino acid mutations although significant amounts (~30%) of truncated form lacking the first 6 residues (DN6b2-m) is also present (4,5). Radford’s group has shown that full length b2-m can create fibrils at low pH and that fibrillar aggregation is most likely primed by a partially unfolded state presenting significant similarities with the molten globule state (6). A few years ago it has been shown that DN6b2-m can create fibrils at neutral pH and most likely this process does not require any relevant 3D modification for the assumption of the “amyloidogenic” conformation (7). Recently the Radford’s group has confirmed that DN6b2-m forms fibrils at pH 7.0 in the absence of seeds, suggesting that this species could initiate the fibrillogenesis in vivo (8).
By limited proteolysis it has been shown that surface protein topology of fibrillar b2-m is similar to that of native truncated b2-m but differs in many cleavage sites in respect to the full-length protein (9). These findings should support the investigation of conformational similarities between the truncated protein in the native like state and intermediates of the folding/unfolding pathway. The existence of these partially folded species prompt us to investigate putative functional characteristic of these conformers such as the binding to molecular target in tissue (i.e. collagen and glycosaminoglycans) (10). With regard to the folding process it was discovered that b2-m has an unexpected non two-state folding pathway and after two fast phases a final slow phase is present that allows the recovery of the native structure in approximately 20 min at 20°C in physiologic solution (11,12). During the structural conversion associated with this slow phase, b2-m can elongate preformed fibrils even at neutral pH (13). Other conditions, compatible with the formation of fibrils at physiologic pH were discovered by Goto's group (14).
All the studies on structure and folding dynamics carried out on b2-m in the last five years were finalised to the discovery of conditions suitable for the formation of amyloid fibrils.
This particular issue has been deeply debated in december 2004 in a meeting where all the expert in this field were gathered in a workshop sponsored by the japanese JSPS and italian CNR (the work is summarised in a monographic issue published by BBA (Biochim Biophys Acta. 1753, 2005, pp 1-153). We can distinguish the methods of fibrillogenesis in vitro in three categories:
A) methods requiring a partial unfolding of the protein in drastic denaturing conditions (i.e. pH 2.5);
B) methods in which an organic solvent such TFE or a detergent such SDS is introduced with the function of enhancing intermolecular interactions and introduce local structural transitions;
C) methods implying the addition of cofactors that could mimic the microenvironment in which amyloidogenesis occurs in vivo (i.e. including collagen and heparin).
These three types of methods can utilize the wild-type protein or its truncated species ubiquitarious of natural b2-m amyloid fibrils and the truncated species facilitate the fibrils formation.
The third approach (C) has been recently undertaken by many groups including Radford’s group which has reported the role of different biological factors in promoting b2-m fibrillogenesis at neutral pH in vitro (7). Such factors include components commonly found in amyloid deposits of b2-m and other proteins that are known to affect fibrillogenesis in vitro and components of the joint enviroment. These researchers have demonstrated that fibril seeds incubated with SAP, apoE, collagen, synovial fluid or uremic serum are able to seed fibril formation of b2-m in the absence of organic solvents or detergents. These factors may act synergistically to promote fibril formation. The same authors also demonstrated that these factors may act in concert with heparin, generating fibrils in vitro. Recently it was demonstrated that heparin efficiently accelerates the formation of gelsolin amyloid by enabling intermolecular beta-sheet formation (15) and it well known that the transition of monomeric amyloid-beta peptides to a beta-sheet conformation is accelerated by heparin (16). In our opinion it is exteremely interesting to discover “biocompatible” ways of making fibrils in vitro and the RU of Genoa found that in vitro, in temperature and pH conditions similar to those occurring in peri-articular tissues in the presence of flogistic processes, b2-m amyloid fibrils can spring up on the surface of type I fibrillar collagen (17). This result indicates that collagen plays a crucial role in b2-m amyloid deposition under physiopathological conditions and suggests an explanation for the strict specificity of DRA toward tissues of the skeletal system. We hypothesize that the presence of positively charged regions along the collagen fibre could play a direct role in b2-m fibrillogenesis. This hypothesis is strengthen by the discovery that a positively charged polymer such as polylysine has an effect similar to collagen on b2-m fibrillogenesis (17).

A detailed understanding of the mechanism(s) by which b2-m folds from its fully unfolded state to its fully native state and by which the protein converts from its soluble state to fibrillar aggregates is of fundamental importance to elucidate the pathogenesis of DRA and for designing therapeutic approaches for this pathological condition. One of the most informative approaches generally used to gain insight into the mechanisms of folding of a globular protein is the phi-value analysis, introduced by Alan Fersht (18) and fully exploited by several groups worldwide over the past 15 years. The analysis consists in the production of a number (typically 25-50) single-point mutants of the protein of interest. The mutations are selected to cover different structural regions of the native protein and to cause only small deletions of the side chains. The typical mutation for a phi-value analysis is the substitution of an isoleucine residue with a valine, or of a threonine residue with a serine. Conservative mutations of this type do not change substantially the chemical nature of the mutated residue, do not induce steric hindrance of newly inserted chemical groups and are therefore aimed at minimising the structural distortion of the compact native fold of the protein of interest. All the mutants are then analysed to determine (a) the free energy change of unfolding extrapolated in the absence of chemical denaturant, (b) the folding rate and (c) the unfolding rate, again in the absence of denaturant. The comparison between these three parameters for the mutant and the same parameters for the wild-type protein provides valuable information on the level of structure that the mutated residue forms in the folding transition state and in the major folding intermediate (if any). In other words, by the comparative analysis of all the mutants produced and the wild-type protein it is possible to trace the involvement of the mutated residues, one by one, in the process of folding of the studied protein.

The same mutants can be used to investigate the involvement of the mutated residues in the aggregation process of the studied protein. Indeed, the rate of the aggregation process and the stability of the resulting fibrils can be evaluated for all the mutants and compared with those of the wild-type protein (19). These systematic measurements can be carried out under conditions in which the native protein is populated or under conditions in which the protein populates a partially folded state that is thought to be the direct precursor of amyloid fibrils. This investigation, along with the comparison between the obtained results with those expected on the basis of the most modern algorithm and with kinetic and thermodynamic data of the folding reaction can provide important information on a number of issues. First they can shed light on which residues or regions of the sequence can promote aggregation (19); second, they can tell whether a process of full or partiall unfolding is required to trigger aggregation or whether aggregation involves the assembly of native or native-like molecules (20); third, when the analysis is coupled to a structural investigation of the amyloidogenic conformational state it can provide information on the relationship between aggregation-promoting regions (or residues) and the structure of the amyloidogenic state. <<<