The discovery of chaperonins

Both the editorial in this issue of Nature Structural Biology and the paper on page 1132 focus on protein folding research. Here we note that our understanding of the general properties of folding reactions in vivo took a major step forward in the late 1980s, when research from several disparate fields revealed the widespread function of the GroEL (hsp60) related proteins, members of the class of chaperones known as the chaperonins.

GroEL is required for the correct assembly of the oligomeric structure that connects the head to the tail of λ phage. It was identified in 1972–1973 by genetic approaches, in the search for factors involved in the replication of bacteriophages^1,2 and its name reflects this history: 'Gro' stands for phage growth; 'E' indicates that the growth defect can be overcome by a mutation in the phage head gene E; and 'L' stands for 'large subunit'.

In the next decade, several lines of research converged. Research on the assembly of the multisubunit ribulose-bisphosphate carboxylase-oxygenase (Rubisco) enzyme in chloroplasts led to the discovery of the Rubisco binding protein, which assists assembly but is not part of the final structure³. Sequencing of the gene for this protein⁴ revealed high homology (∼50% identity) with GroEL and, to distinguish these as a family of proteins, they were named the chaperonins. In addition, the study of mitochondrial protein import uncovered a temperature sensitive lethal mutation in the hsp60 gene of yeast (named for 'heat shock protein' and its approximate molecular weight). These mutants could transport the test protein ornithine transcarbamoylase into mitochondria but were unable to assemble the active trimer⁵. Sequencing of the hsp60 gene⁶ revealed homology to both GroEL and the Rubisco binding protein. Later, a distinct but similar family of chaperonin proteins were found in archaebacteria and eukaryotes⁷, thus demonstrating the ubiquitous requirement for in vivo chaperonin function.

Around the same time, in vivo studies of mitochondrial import and subsequent folding of dihydrofolate reductase (DHFR) showed a requirement for hsp60⁸. This was particularly interesting since DHFR is a monomeric protein that can fold spontaneously without chaperonin assistance in vitro. Thus, this work clearly indicated that protein folding reactions in vitro and in vivo can have different characteristics, a point that is of major interest today as researchers attempt to reconcile the large amount of in vitro and in vivo folding data, much of which has accumulated in the decade since the discovery of the chaperonins.