What the bacteriologists have learned about heat shock.

F. Neidhardt

DOI: https://doi.org/10.1101/GAD.1.2.109

IF: 12.89

1987-04-01

Genes & Development

Abstract:In the spring of 1982, word that something very like the heat shock response of Drosophila and other eukaryotes had been found in Escherichia coli excited great attention among the attendees of the first Cold Spring Harbor heat shock conference. The finding implied that this multigenic response of cells to high temperature must be nearly as ancient as cellular life itself. In addition, there was a general expectation that work with E. coli, as usual, would move very quickly, and before long would be generating information of value to the whole heat shock field. To what extent have these expectations been realized by the work of the ensuing 5 years? Progress on molecular mechanism has been satisfying, as can be judged from three closely related papers that have appeared within the past few months. These papers do not bring the analysis to an end, but the remaining questions about mechanism in the bacterial system can now be phrased with such precision that the complete solution seems not far off. As in many other cells, the heat shock response in bacteria consists of the transient induction of a set of some 20 proteins upon a shift from low to high temperature. The induction is brought about by stimulation of gene transcription by an active regulatory system, not by a direct effect of heat on the responding genes (references in Neidhardt et al. 1984). Work from several laboratories over the past 5 years has established that the heat shock genes of E. coli are found in unlinked operons and constitute a single regulon (regulatory unit) under the control of a positiveacting protein. This protein has been shown to function as an RNA polymerase sigma factor, and to have structural similarity to the major sigma factor, cr 7°. Originally called HtpR, as product of the htpR gene, and identified on gels as protein F33.4, it is now referred to as 03~; its gene bears the altemative designation rpoH. The halfdozen known sequences for promoters of heat shock genes share a family resemblance, distinguishable from those of other E. coli promoters. But, 032 appears to do more than simply determine the promoter specificity of RNA polymerase. Evidence both in vivo and in vitro suggests that the level of 03~ directly determines the activity of heat shock genes (references in Neidhardt et al. 1984 and Neidhardt and VanBogelen 1987). The paper by Grossman et al. (1987) in this issue of Genes & Development provides support for this simple model. First, these authors confirm and extend earlier findings that the level of 032 is rate limiting for synthesis of heat shock proteins. Increasing the rate of synthesis of 032 [by inducing expression from a Plac or a Ptac htpR {rpoH) fusion on a plasmid] increases the synthesis of heat shock proteins in the absence of a temperature shift. Second, they confirm the fact that 032, at least under conditions of massive oversynthesis, is an unstable protein (half-life of approximately 4 min) at 42°C. Finally, they show that when 032 is expressed from the phage h PL promoter, it is synthesized transiently at a high rate after temperature upshift, and then repressed by some post-transcriptional process. The repression of 032 synthesis is abolished in a strain carrying a mutant form (dnaK7561 of one of the heat shock genes previously implicated as a modulator of the heat shock response {references in Neidhardt et al. 1984 and Neidhardt and VanBogelen 1987). In sum, these findings support a relatively simple model in which the control of heat shock genes might be solely accomplished by changes in the cellular level of 032 brought about by increasing its rate of synthesis or decreasing its rate of degradation. Tilly et al. (1986) contribute another element to this picture by showing that cellular levels of transcripts of the htpR (rpoH) gene increase quickly {within minutes) and markedly {fivefoldl after temperature upshifts. The time course of increase correlates with the synthesis of heat shock proteins. Interestingly, htpR {rpoH) mRNA levels were observed to rise after temperature shifts even in htpR mutants which are unable to stimulate heat shock genes, so the increase must occur by a mechanism different in some way from that operating on the heat shock genes themselves. A third paper continues the theme of heat shock regulation by the level of o 32, but deals with the intriguing involvement of the heat shock response with viral infection. Bahl et al. {19871 have pursued the mechanism by which phage h induces heat shock proteins in E. coli shortly after infection. It had previously been demonstrated that this induction depended both on 032 and on early gene expression from the leftward PL transcription unit of the phage, but the specific inducing factor was not known {references in Neidhardt and VanBogelen 1987). These investigators report that overproduction from a plasmid system of one of the proteins encoded by this transcription unit, the h cIII protein, leads to induct ion of heat shock proteins at low temperature and prolongation of the response following a temperature upshift. Overproduction of another h protein, Eal0, does not alter the response after temperature upshift, so the cIII effect is presumed to be specific. The known properties of cIII include its ability to protect the h cII protein from proteolysis by a variety of cellular proteolytic systems, and the current work shows that cIII overpro-