The chaperone-like activity of a small heat shock protein is lost after sulfoxidation of conserved methionines in a surface-exposed amphipathic alpha-helix.

The small heat shock proteins (sHsps) possess a chaperone-like activity which prevents aggregation of other proteins during transient heat or oxidative stress. The sHsps bind, onto their surface, molten globule forms of other proteins, thereby keeping them in a refolding competent state. In Hsp21, a chloroplast-located sHsp in all higher plants, there is a highly conserved region forming an amphipathic alpha-helix with several methionines on the hydrophobic side according to secondary structure prediction. This paper describes how sulfoxidation of the methionines in this amphipathic alpha-helix caused conformational changes and a reduction in the Hsp21 oligomer size, and a complete loss of the chaperone-like activity. Concomitantly, there was a loss of an outer-surface located alpha-helix as determined by limited proteolysis and circular dichroism spectroscopy. The present data indicate that the methionine-rich amphipathic alpha-helix, a motif of unknown physiological significance which evolved during the land plant evolution, is crucial for binding of substrate proteins and has rendered the chaperone-like activity of Hsp21 very dependent on the chloroplast redox state.

Substitution of conserved methionines by leucines in chloroplast small heat shock protein results in loss of redox-response but retained chaperone-like activity.

During evolution of land plants, a specific motif occurred in the N-terminal domain of the chloroplast-localized small heat shock protein, Hsp21: a sequence with highly conserved methionines, which is predicted to form an amphipathic alpha-helix with the methionines situated along one side. The functional role of these conserved methionines is not understood. We have found previously that treatment, which causes methionine sulfoxidation in Hsp21, also leads to structural changes and loss of chaperone-like activity. Here, mutants of Arabidopsis thaliana Hsp21 protein were created by site-directed mutagenesis, whereby conserved methionines were substituted by oxidation-resistant leucines. Mutants lacking the only cysteine in Hsp21 were also created. Protein analyses by nondenaturing electrophoresis, size exclusion chromatography, and circular dichroism proved that sulfoxidation of the four highly conserved methionines (M49, M52, M55, and M59) is responsible for the oxidation-induced conformational changes in the Hsp21 oligomer. In contrast, the chaperone-like activity was not ultimately dependent on the methionines, because it was retained after methionine-to-leucine substitution. The functional role of the conserved methionines in Hsp21 may be to offer a possibility for redox control of chaperone-like activity and oligomeric structure dynamics.

Caenorhabditis elegans small heat-shock proteins Hsp12.2 and Hsp12.3 form tetramers and have no chaperone-like activity.

Four 12.2-12.6 kDa small heat-shock proteins (sHSPs) of Caenorhabditis elegans are the smallest known members of the sHSP family. They essentially comprise the characteristic C-terminal 'alpha-crystallin domain' of the sHSPs, having a very short N-terminal region, and lacking a C-terminal tail. Recombinant Hsp12.2 and 12.3 are characterized here. Far-UV CD spectra reveal, as for other sHSPs, predominantly a beta-sheet structure. By gel permeation and crosslinking, they are the first sHSPs shown to occur as tetramers, rather than forming the usual large multimeric complexes. Exceptionally, too, both appear devoid of in vitro chaperone-like abilities. This supports the notion that tetramers are the building blocks of sHSP complexes, and that higher multimer formation, mediated through the N-terminal domains, is a prerequisite for chaperone-like activity.

The lack of chaperonelike activity of Caenorhabditis elegans Hsp12.2 cannot be restored by domain swapping with human alphaB-crystallin.

The small heat shock proteins Hsp12.2 and alphaB-crystallin differ in that the former occurs as tetramers, without chaperonelike activity, whereas the latter forms multimers and is a good chaperone. To investigate whether the lack of chaperone activity of Hsp12.2 is primarily due to its tetrameric structure or rather to intrinsic sequence features, we engineered chimeric proteins by swapping the N-terminal, C-terminal, and tail regions of Hsp12.2 and alphaB-crystallin, designated as n-c-t and N-C-T, respectively. Three of the chimeric sHsps, namely N-c-T, n-c-T, and N-C-t, showed nativelike secondary and quaternary structures as measured by circular dichroism and gel permeation chromatography. Combining the conserved alpha-crystallin domain of Hsp12.2 with the N-terminal and tail regions of alphaB-crystallin (N-c-T) resulted in multimeric complexes, but did not restore chaperonelike activity. Replacing the tail region of Hsp12.2 with that of alphaB-crystallin (n-c-T) did not alter the tetrameric structure and lack of chaperone activity. Similarly, providing alphaB-crystallin with the tail of Hsp12.2 (N-C-t) did not substantially influence the multimeric complex size, but it reduced the chaperoning ability, especially for small substrates. These results suggest that the conserved alpha-crystallin domain of Hsp12.2 is intrinsically unsuitable to confer chaperonelike activity and confirms that the tail region in alphaB-crystallin modulates chaperonelike capacity in a substrate-dependent manner.

The rodent alphaA-crystallin gene: mutagenesis of a non-consensus 5'-splice site to study alternative splicing in vivo.

alphaA-Crystallin is a member of the small heat shock protein family that is abundantly expressed as a structural component in the vertebrate eye lens. In lenses of rodents and some other mammals, there occurs a minor variant of alphaA-crystallin, which has an insertion of 23 amino acid residues. This variant, alphaA(ins)-crystallin, results from differential integration of an optional exon into a small fraction of the mRNA. We have studied whether this alternative splicing is caused by a non-consensus cytosine in the 5' splice site adjacent to the optional exon. After replacement of the aberrant cytosine in the hamster alphaA-crystallin gene by a consensus thymine, and transient transfection of this gene in Chinese Hamster Ovary cells, the optional exon is indeed almost completely spliced into the mature mRNA. In contrast, replacement of the cytosine by adenine or guanine completely abolishes the splicing of the optional exon. Our results confirm that alternative splicing of the alphaA-crystallin primary transcript is mainly due to a non-consensus 5' splice site nucleotide. However, we conclude that the small size of the optional exon is probably an additional contributing factor and therefore it seems that the splicing mechanism is based on recognition of exons rather than introns.