We have suggested that IDPs can carry out their function by six different molecular mechanisms. In one of these, when their function is realized by transient binding to the partner, they exert a protective effect, i.e. they function as chaperones. Traditionally, chaperones are thought of as highly sophisticated protein machines that assist the correct folding of partner molecules by a combination of mechanisms. Chaperones may prevent protein aggregation, actively unfold misfolded partners, solubilize aggregates, pull translocating proteins across membranes and remodel native proteins complexes.

Even in the case of structurally well-characterized chaperones, disordered segments missing from the X-ray structure are involved in function, as suggested in the case of Hsp90, GroEL and α-crystalline, for example. Similar observations have been made in the case of RNA-chaperones, such as heteronuclear ribonucleoprotein A1 (hnRNP A1), nucleocapsid proteins and the prion protein. In addition to such partial disorder in classical chaperones, there are several reports on fully disordered proteins displaying chaperone-like activity, such as α- and β-synuclein, α-casein, MAP2 and several plant stress proteins. In a detailed bioinformatic study on a collection of 27 RNA-chaperones and 20 protein-chaperones, we showed that protein chaperones have an elevated level of disorder (36.7% of their residues falling into disordered regions and 15% within IDRs ≥30 consecutive residues), whereas RNA chaperones are even more disordered (54.2% of their residues are predicted to be disordered regions and 40% fall within IDRs ≥30 consecutive residues). Thus, RNA chaperones are the most disordered functional class of proteins, exceeding in disorder even regulatory and signaling proteins. These observations motivated our quest into the details of chaperone activity of IDPs, with a particular focus on experimental by testing the “entropy transfer” model we suggested for the chaperone activity of IDPs.

Predicted disorder in chaperones and regulatory proteins. From Tompa and Csermely (2004) FASEB J. 18: 1169-75.

Expression of ERD10 and ERD14 in rapidly dividing tissues under non-stressed conditions. From Nylander et al. (2001) Plant Mol. Biol. 45 263

A key observation in the laboratory is the chaperone activity of plant stress protein termed late embryogenesis abundant (LEA) proteins. The primary line of defense of plants against stressful environmental changes causing dehydration, such as draught, low- or high temperature, freezing or increase in salinity, is the increased expression of proteins also expressed under normal conditions in seeds (why they appear abundant late in embryogenesis). LEA proteins are intrinsically disordered, and their anti-stress functions have been linked with a variety of molecular mechanisms, such as ion sequestration, membrane binding and stabilization, or acting as antioxidants. We have studied two LEA proteins that belong to the dehydrin subclass of LEA proteins, ERD10 (early responsive to dehydration) and ERD14. By a wide-line NMR relaxations approach it could be shown that they are highly hydrated, i.e. they bind much more water per protein mass than a globular protein. Thus, they probably shift the osmotic equilibrium of the cell under conditions of excessive water loss. We also demonstrated their chaperone activity in a series of in vitro assays, such as heat-induced inactivation of alcohol dehydrogenase and citrate synthase, thermal aggregation of firefly luciferase and chemically-induced inactivation of lysozyme. ERD10/14 are effective at a concentration equimolar to the substrate, which is commensurable with that of Hsp90. Our plan is to observe chaperone activity of these proteins in vivo, and also to carry out their in-depth structural analysis by multi-dimensional NMR, to understand the molecular details of chaperone function of IDPs and test the entropy-transfer model.

Chaperone activity of Hsp90 and ERD14 in ADH denaturation assay. From Kovacs et al. 2008) Plant. Physiol. 147, 381

The entropy transfer model incorporates two key features of the function of IDPs, recognition by short recognition elements and entropic effects of long disordered unbound regions. Implicitly, this model enables the disordered chaperone to act on partners of basically differing nature, at the extreme on both RNA and protein partners. We have provided evidence for this possibility by studying the protein chaperone activity of several proteins of the large ribosomal subunit (L15, L16, L18 and L19). These ribosomal proteins assist the assembly and increase the stability of ribosomal RNA. We found in three chaperone assays (thermal inactivation of ADH, chemical inactivation of lysozyme and refolding of denatured lysozyme) that these ribosomal proteins exhibit potent chaperone activity, commensurable with that of heat shock protein 90 kDa. Due to the generally high level of disorder in ribosomal proteins, this observation may be suggestive of the role of structural disorder in their dual activity, which also highlight possible novel aspects of the functioning of ribosomal proteins outside of the ribosome, noted several years ago. We have suggested the phenomenon of such promiscuous chaperone activity to be termed “Janus chaperones”.

Entropy transfer model of action of ID chaperones. From Tompa and Csermely (2004) FASEB J. 18: 1169-75.