Biomedical subjects
Publications and source records attributed to S Lindquist.
Heat-shock proteins help microorganisms cope with the toxic effects of a wide variety of stresses. Some help the organism grow under moderately stressful conditions, others help it to survive more extreme conditions. Surprisingly, the relative importance of individual proteins differs between organisms.
In Drosophila cells, regulatory mechanisms not only act to provide rapid induction of hsp70 during heat shock but also to prevent expression at normal temperatures. To determine whether expression of hsp70 is detrimental to cells growing at normal temperatures, we used heterologous promoters to force expression of the protein in tissue culture cells and in larval salivary glands. Initially, constitutive expression of hsp70 substantially reduces the rate of cell growth. With continued expression, however, growth rates recover. At the same time, the intracellular distribution of hsp70 changes. Immediately after induction, the protein is diffusely distributed throughout the cell, but as growth resumes it coalesces into discrete points of high concentration, which we term hsp70 granules. hsp70 granules are also observed both in wild-type Drosophila tissue culture cells and in salivary glands after extended periods of recovery from heat shock. The protein in these granules appears to be irreversibly inactivated. It cannot be dispersed with a second heat shock, and cells containing these granules do not show thermotolerance. Only partial overlap between hsp70 granules and lysosomes indicates that the granules form independently of lysosomes. We conclude that expression of hsp70 is detrimental to growth at normal temperatures. We suggest that the change in hsp70 distribution, from diffuse to granular, represents a mechanism for controlling the protein's activity by sequestration.
Most eukaryotic cells produce proteins with relative molecular masses in the range of 100,000 to 110,000 after exposure to high temperatures. These proteins have been studied only in yeast and mammalian cells. In Saccharomyces cerevisiae, heat-shock protein hsp104 is vital for tolerance to heat, ethanol and other stresses. The mammalian hsp110 protein is nucleolar and redistributes with growth state, nutritional conditions and heat shock. The relationships between hsp110, hsp104 and the high molecular mass heat-shock proteins of other organisms were unknown. We report here that hsp104 is a member of the highly conserved ClpA/ClpB protein family first identified in Escherichia coli and that additional heat-inducible members of this family are present in Schizosaccharomyces pombe and in mammals. Mutagenesis of two putative nucleotide-binding sites in hsp104 indicates that both are essential for function in thermotolerance.
The amp operon, which is located on the Escherichia coli chromosome, modulates the induction of plasmid-borne beta-lactamase genes by extracellular beta-lactam antibiotics. This suggests that the gene products AmpD and AmpE may function in the transduction of external signals. beta-Lactam antibiotics are analogs of cell wall components that can be released during cell wall morphogenesis of enterobacteria. The amp operon was studied to determine its importance in signal transduction during cell wall morphogenesis. The peptidoglycan compositions of amp mutants were determined by high-performance liquid chromatography and fast atom bombardment mass spectrometry. When a chromosomal or plasmid-borne copy of ampD was present, the amount of pentapeptide-containing muropeptides in the cell wall increased upon addition of the cell wall constituent diaminopimelic acid to the growth medium. These results suggest that beta-lactamase induction and modulation of the composition of the cell wall share elements of a regulatory circuit that involves AmpD. Escherichia coli requires AmpD to respond to extracellular signaling amino acids, such as diaminopimelic acid, and this signal transduction system may regulate peptidoglycan composition in response to cell wall turnover products.
In the yeast Saccharomyces cerevisiae, the splicing of mRNA precursors is disrupted by a severe heat shock. Mild heat treatments prior to severe heat shock protect splicing from disruption, as was previously reported for Drosophila melanogaster. In contrast to D. melanogaster, protein synthesis during the pretreatment is not required to protect splicing in yeast cells. However, protein synthesis is required for the rapid recovery of splicing once it has been disrupted by a sudden severe heat shock. Mutations in two classes of yeast hsp genes affect the pattern of RNA splicing during the heat shock response. First, certain hsp70 mutants, which overproduce other heat shock proteins at normal temperatures, show constitutive protection of splicing at high temperatures and do not require pretreatment. Second, in hsp104 mutants, the recovery of RNA splicing after a severe heat shock is delayed compared with wild-type cells. These results indicate a greater degree of specialization in the protective functions of hsps than has previously been suspected. Some of the proteins (e.g., members of the hsp70 and hsp82 gene families) help to maintain normal cellular processes at higher temperatures. The particular function of hsp104, at least in splicing, is to facilitate recovery of the process once it has been disrupted.
Prolonged exposure to high concentrations of human recombinant interleukin-1 beta (rIL-1 beta) has been reported to exert both suppressive and cytotoxic effects on pancreatic beta-cells during culture in vitro. In order to investigate the molecular mechanism(s) underlying the actions of rIL-1 beta on the beta-cell, we have exposed isolated rat pancreatic islets for 3 or for 24 h to 25 U/ml of rIL-1 beta. Subsequently the biosynthesis of heat shock proteins, as assessed by western blot analysis, and total protein biosynthesis patterns were studied, using one and two-dimensional gel electrophoresis of [35S]methionine labelled islet proteins from different subcellular compartments. It was found that rIL-1 beta exerted no specific effects on protein synthesis when added during a 3 h incubation period. However, after a 24 h incubation period, the synthesis of a group of acidic proteins with the approximate molecular weight of 35 kD was specifically inhibited in the rIL-1 beta treated islets. This alteration was predominantly associated with the endoplasmic reticulum fraction. The cytokine also inhibited the synthesis of four cytosolic proteins with the molecular weights 75, 85, 95 and 120 kD. In contrast, rIL-1 beta increased the expression of the heat shock protein hsp70 both in the microsomal and cytosolic fractions, in contrast to the islet nuclei in which no increase was found. These results show that exposure of pancreatic islets to rIL-1 beta is accompanied by specific alterations in the protein synthesis of the islet cells.(ABSTRACT TRUNCATED AT 250 WORDS)
To test the role of the heat shock protein hsp70 in induced thermotolerance and in the regulation of the heat-shock response, we established cell lines with altered expression of the Hsp70 gene. Underexpressing cells were created by transformation with antisense Hsp70 genes, and overexpressing cells by transformation with extra copies of the wild-type gene. Expression at normal temperatures was achieved by placing Hsp70 coding sequences under the control of the metallothionein promoter. Cells that expressed mutant hsp70s were created by transforming cells with deletion and frameshift mutations. The results indicate that hsp70 plays a major role in both thermotolerance and regulation. Surprisingly, they also indicate that these functions can be separated. Overexpression affected thermotolerance more than regulation; underexpression affected regulation more than thermotolerance. A carboxyl-terminal deletion of Hsp70 had a severe dominant-negative effect on thermotolerance but only a minor effect on regulation; an amino-terminal deletion strongly affected regulation but not thermotolerance. A model that explains these observations is presented.
Signalling by steroid hormones is mediated by receptor proteins that bind hormonal ligands and regulate the transcription of specific genes. The heat-shock protein hsp90 seems to associate selectively with unliganded receptors (aporeceptors), but it has not been determined whether this interaction affects receptor function in vivo. To address the role of hsp90, we have taken advantage of the capacity of mammalian steroid receptors to function in yeast. We constructed a strain of Saccharomyces cerevisiae in which hsp90 expression was regulatable and could be reduced more than 20-fold relative to wild type. At low levels of hsp90, aporeceptors seem to be mostly hsp90-free, yet fail to enhance transcription; on hormone addition, the receptors are activated but with markedly reduced efficiency. Thus hsp90 does not inhibit receptor function solely by steric interference; rather, hsp90 seems to facilitate the subsequent response of the aporeceptor to the hormonal signal. This is the first biological evidence that hsp90 acts in the signal transduction pathway for steroid receptors.
It has long been appreciated that selective transcription and translation play important roles in the heat shock response. More recently, regulatory strategies acting at the levels of RNA processing and message degradation have been shown to exert a profound effect on gene expression both during heat shock and during recovery from heat shock. In turn, as heat shock proteins accumulate, they affect those very processes that govern their expression.
hsp26, the small heat shock protein of Saccharomyces cerevisiae, accumulates in response to heat and other types of stress. It also accumulates during the normal course of development, as cells enter stationary phase growth or begin to sporulate (S. Kurtz, J. Rossi, L. Petko, and S. Lindquist, Science 231:1154-1157, 1986). Analysis of deletion and insertion mutations demonstrated that transcriptional control plays a critical role in regulating HSP26 expression. The HSP26 promoter was found to be complex and appears to contain repressing elements as well as activating elements. Several upstream deletion mutations resulted in strong constitutive expression of HSP26. Furthermore, upstream sequences from the HSP26 gene repressed the constitutive expression of a heterologous heat shock gene. We propose that basal repression and heat-induced depression of transcription play major roles in regulating the expression of HSP26. None of the recombinant constructs that we analyzed separated cis-regulatory sequences responsible for heat shock regulation from those responsible for developmental regulation of HSP26. Depression of HSP26 transcription may be the general mechanism of HSP26 induction in yeast cells. This regulatory scheme is very different from that described for the regulation of most other heat shock genes.
The rapid and dramatic induction of heat-shock proteins is accomplished by regulatory mechanisms acting at many different levels. Here we review current knowledge of two cytoplasmic mechanisms employed during the response in the fruit fly Drosophila melanogaster. (1) Heat-shock messages are translated with high efficiency during heat shock while most normal cellular messages are inactive. Sequences in the 5'-untranslated leader of heat shock mRNAs govern their preferential translation. (2) The messages for heat-shock proteins are unstable at normal temperatures. During heat shock, however, they are very stable and accumulate in large numbers. Sequences in their 3'-untranslated regions play a major role in determining their stability.
We have transferred the site-specific recombination system of the yeast 2 micron plasmid, the FLP recombinase and its recombination targets (FRTs), into the genome of Drosophila. Flies were transformed with an FLP gene under the control of hsp70 regulatory sequences and with a white gene flanked by FRTs. The heat-induced recombinase catalyzes recombination between FRTs, causing loss of white (seen somatically as white patches in the eye) and, less frequently, gain of white (seen as dark-red patches). Loss and gain frequencies vary with the severity of the heat shock, and patterns of mosaicism vary with the developmental stage at which the heat shock is applied. The recombinase is also active in the germline, producing white-eyed and dark-red-eyed progeny.
An antibody highly specific for heat-shock protein (hsp)26, the unique small hsp of yeast, and mutants carrying a deletion of the HSP26 gene were used to examine the physical properties of the protein and to determine its intracellular distribution. The protein was found in complexes with a molecular mass of greater than 500 kD. Thus, it has all of the characteristics, including sequence homology and induction patterns, of small hsps from other organisms. When log-phase cells growing in glucose were heat shocked, hsp26 concentrated in nuclei and continued to concentrate in nuclei when these cells were returned to normal temperatures for recovery. However, hsp26 did not concentrate in nuclei under a variety of other conditions. For example, in early stationary-phase cells hsp26 is induced at normal growth temperatures. This protein was generally distributed throughout the cells, even after heat shock. Similarly, in cells genetically engineered to synthesize hsp26 in the presence of galactose, hsp26 did not concentrate in nuclei, with or without a heat shock. To determine if the failure of hsp26 to concentrate in the nucleus of these cells was due to the fact that the protein had been produced at 25 degrees C or to a difference in the physiological state of the cell, we investigated the distribution of the heat-induced protein in cells grown under several different conditions. In wild-type cells grown in galactose or acetate and in mitochondrial mutants grown in glucose or galactose, hsp26 also failed to concentrate in nuclei with a heat shock. We conclude that the intracellular location of hsp26 in yeast depends upon the physiological state of the cell and not simply upon the presence or absence of heat stress. Our findings may explain why previous investigations of the intracellular localization of small hsps in a variety of organisms have yielded seemingly contradictory results.
When Drosophila cells are heat shocked, hsp70 messenger RNA (mRNA) is stable and is translated at high efficiencies. During recovery from heat shock, hsp70 synthesis is repressed and its messenger RNA (mRNA) is degraded in a highly regulated fashion. Dramatic differences in the timing of repression and degradation are observed after heat treatments of different severities. The 3' untranslated region (UTR) of the hsp70 mRNA was sufficient to transfer this regulated degradation to heterologous mRNAs. Altering the translational efficiency of the message or changing its natural translation-termination site did not alter its pattern of regulation, although in some cases it changed the absolute rate of degradation. We have previously shown that hsp70 mRNA is very unstable when it is expressed at normal growth temperatures (from a metallothionein promoter). We report here that the 3' untranslated region of the hsp70 mRNA is responsible for this instability as well. We postulate that a mechanism for degrading hsp70 mRNA pre-exists in Drosophila cells, that it is inactivated by heat shock and that it is the reactivation of this mechanism that is responsible for hsp70 repression during recovery. This degradation system may be the same as that used by other unstable mRNAs.
The cloned Citrobacter freundii ampC beta-lactamase is inducible in the presence of its regulatory gene ampR in Escherichia coli (Lindberg et al., 1985). The basal level of expression and inducibility are affected by two E. coli proteins encoded by the closely linked ampD and ampE genes. Deletion of both genes led to constitutive ampR-dependent overproduction of beta-lactamase, whereas an out-of-frame deletion in AmpD caused the basal expression to increase two-fold. This ampD1 mutant was inducible at lower beta-lactam concentrations than the wild type. An IS1 insertion in ampD was polar on ampE expression and increased basal beta-lactamase expression 30-fold while mediating a semi-constitutive phenotype. AmpE expressed from a recombinant plasmid in an ampD-ampE deletion mutant reduced basal beta-lactamase expression to wild-type levels but did not markedly reduce beta-lactam resistance since the cells became hyperinducible. In the absence of AmpD, increasing levels of AmpE therefore decrease the basal expression of AmpC beta-lactamase in an AmpR-dependent manner. AmpD modulated the response exerted on beta-lactamase expression by AmpE. The ampD gene encodes a 20.5kD cytoplasmic protein while the 32.1kD ampE gene product is an integral membrane protein with a likely ATP-binding site between the second and third putative transmembrane region. Since neither AmpD nor AmpE are needed for beta-lactam induction and since these proteins could not be covalently labelled by benzylpenicillin, they are not thought to act as beta-lactam-binding sensory transducers. Instead it is suggested that AmpD and AmpE sense the effect of beta-lactam action on peptidoglycan biosynthesis and relay this signal to AmpR.
Citrobacter freundii encodes an inducible chromosomal beta-lactamase. Induction requires the product of the ampR gene, which is transcribed in the opposite orientation from the ampC beta-lactamase gene. We show here that the AmpR protein acts as a transcriptional activator by binding to a DNA region immediately upstream of the ampC promoter. The DNase I footprint pattern was not affected by growth in the presence of beta-lactam inducer or by the use of extracts prepared from cells carrying the ampD2 allele leading to semiconstitutive production of beta-lactamase. It is suggested that activation of AmpR facilitates binding or open complex formation for RNA polymerase at the ampC promoter. The AmpR-binding site overlaps the ampR promoter, and beta-galactosidase activity was decreased from an ampR-lacZ transcriptional fusion when AmpR was expressed from a coresident plasmid, suggesting that ampR is autogenously controlled. The AmpR protein belongs to a family of highly homologous transcriptional activators that includes LysR, which regulates the E. coli lysine synthetase gene, and the NodD protein, which regulates expression of a number of genes involved in nodulation in Rhizobium. The lack of sequence homology to any known beta-lactam-binding protein suggests that AmpR does not bind directly to the beta-lactam inducer but interacts with a second messenger of unknown nature.
hsp82 is one of the most highly conserved and abundantly synthesized heat shock proteins of eucaryotic cells. The yeast Saccharomyces cerevisiae contains two closely related genes in the HSP82 gene family. HSC82 was expressed constitutively at a very high level and was moderately induced by high temperatures. HSP82 was expressed constitutively at a much lower level and was more strongly induced by heat. Site-directed disruption mutations were produced in both genes. Cells homozygous for both mutations did not grow at any temperature. Cells carrying other combinations of the HSP82 and HSC82 mutations grew well at 25 degrees C, but their ability to grow at higher temperatures varied with gene copy number. Thus, HSP82 and HSC82 constitute an essential gene family in yeast cells. Although the two proteins had different patterns of expression, they appeared to have equivalent functions; growth at higher temperatures required higher concentrations of either protein. Biochemical analysis of hsp82 from vertebrate cells suggests that the protein binds to a variety of other cellular proteins, keeping them inactive until they have reached their proper intracellular location or have received the proper activation signal. We speculate that the reason cells require higher concentrations of hsp82 or hsc82 for growth at higher temperatures is to maintain proper levels of complex formation with these other proteins.
Severe heat shocks block the splicing of intervening sequences from messenger RNA precursors. The RNA's that accumulate after a severe heat shock have normal transcription start sites and are uncut at both their 5' and 3' splice junctions. Some of these unspliced transcripts leave the nucleus and enter the pool of cytoplasmic messenger RNA. Translation of these RNA's proceeds into their intervening sequences, resulting in the production of abnormal proteins. Thus, the repression of normal transcription, which usually accompanies the heat shock response, may protect the cell from the large-scale synthesis of abnormal RNA's and aberrant proteins.