AS LONG AGO as 1962, a long time in biological research, a paper appeared in the scientific press describing the curious effect on fruit flies of a short exposure to high temperatures. The heat seemed to have ‘switched on’ new genes. On examining the giant ‘polytene’ chromosomes from the fly’s salivary glands, the researcher was surprised to observe the appearance of a new pattern of ‘puffs’ – areas that are actively transcribing the DNA into messenger RNA, to be translated ultimately into protein.
It took another 15 years to show that this heat-induced phenomenon is not unique to flies, as researchers discovered similar effects in birds and mammals. We now know that proteins induced by heat, or heat-shock proteins (hsps) as they are now often known, exist in almost all living organisms, including bacteria.
Advertisement
Of even greater significance was the discovery that several of these proteins are related; antibodies against a particular heat-shock protein from one species will usually cross-react with similar proteins from many other unrelated species. In our laboratory, for example, an antibody against a chicken heat-shock protein will recognise proteins from a shrimp.
Researchers have cloned several of the genes for these unusual proteins, and determined their sequence of chemical bases, providing important data on their degree of similarity. Two groups of proteins in particular, with medium-sized molecules of 70 and 90 kilodaltons (also known as hsp 70 and hsp 90), are among the most highly conserved proteins in nature. This suggests that they must do something which is essential to most, if not all forms of life. ¿ìè¶ÌÊÓƵs are now beginning to understand some of these essential functions.
At first, researchers thought that as these proteins appeared after heat shock, they most probably dealt with any harmful biochemical effects of such harsh treatment. This idea gained ground when biologists discovered that other stressful conditions such as recovery from oxygen starvation, contact with heavy metals and even alcohol could induce the proteins. Furthermore, cells expressing heat-shock proteins after exposure to chemicals or heat often became tolerant to normally lethal doses of these agents, suggesting that these proteins play some protective role .
Surprisingly, biologists soon discovered that most cells harbour considerable quantities of almost identical ‘heat-shock’ proteins that are only weakly inducible by heat, if at all. If these proteins have something important to do with controlling damage, why should cells go to the trouble of synthesising them after a heat shock when they are already present in abundance? One of the most widespread of these proteins in animal and plant cells, appropriately named ubiquitin, may provide an answer to this dilemma.
Ubiquitin is a relatively small protein that we know is involved in the ‘tagging’, or ubiquitination, of other proteins destined for breakdown by enzymes. It is also a heat-shock protein, which has naturally led to the suggestion that high temperature, as well as other stressful conditions, effectively denatures proteins or generally renders them ‘unserviceable’. Such damaged molecules are of no use to the cell and may even be harmful. Perhaps, researchers reasoned, a heat shock overloads the protein-destroying pathway directed by ubiquitin with junk proteins. The cell may then need to make more ubiquitin to meet the ‘tagging’ demand imposed by abnormally high levels of damaged proteins.
This theory is all well and good, but it does not explain how the cell switches on the production of heat-shock proteins, particularly hsp 70. Perhaps, researchers argued, ubiquitin itself controls the activation of the hsp 70 genes. Such control could work through interactions between ubiquitin and the recently discovered heat-shock transcription factor (HSTF). This protein binds to DNA and probably activates some of the hsp 70 genes.
We are still unsure about the precise role of hsp 70, but one clue comes from work on an enzyme that catalyses the removal of a particular protein, clathrin. Clathrin is a scaffolding protein that acts as a molecular cage to enclose and trap molecules on the cell surface. Vesicles form on the cell membrane and emerge into the cell surrounded by a cage of clathrin. The enzyme that removes this cage, which requires energy in the form of ATP, is closely related to hsp 70. Other observations show that in heat-shocked cells, hsp 70 binds to partially assembled ribosomes. By so binding, they may prevent the denaturation of these important structures, which are vital to protein synthesis.
So hsp 70 may be part of a salvaging or repair pathway. Hugh Pelham, of the Medical Research Council’s Laboratory of Molecular Biology at Cambridge, has suggested that hsp 70 may inhibit possibly catastrophic precipitation of denatured proteins by reversibly binding to them. Slightly damaged proteins could be salvaged by the removal of hsp 70, while irreversibly damaged proteins would be ubiquitinated and destroyed. The idea that damaged proteins can trigger the synthesis of heat-shock protein gains support from experiments where such proteins are injected into cells.
Yet even this idea falls short of providing an adequate explanation for the high levels of hsp 70 present in most unstressed cells. At least this was the case until recently, when biologists finally verified another predicted role for hsp 70. Researchers had long speculated that hsp 70 acts as a chaperone, binding to newly synthesised proteins that may be in a temporarily denatured state as they come off the construction site of the ribosomes. According to this idea, hsp 70 prevents the proteins from aggregating and precipitating in the cytoplasm before they are properly folded. Once free in the cytoplasm, hsp 70 would release itself from its embrace with these proteins. The energy used up in this process, in the form of hydrolysed ATP, would either contribute to the correct folding of the nascent protein, or simply alter the affinity of hsp 70 for the protein, or a combination of both. We now believe that something like this does happen.
Using an in vitro assay system, William Chiroco and his colleagues at the Rockefeller University, in New York, have shown that something dissolved in the cytoplasm stimulates a yeast protein (called prepro factor) to move across small membrane-bound ‘vesicles’ derived from yeast. The ‘pre’ of prepro signifies the presence of a disposable signal sequence, which labels a protein destined eventually to be secreted from the cell. The required soluble component turns out to be hsp 70.
If the yeast protein is denatured before it is introduced to the assay, it moves more rapidly across the membrane. This observation suggests that proteins are in a naturally denatured state as they move through the membrane. Given the idea that hsp 70 can bind to denatured or unfolded proteins, it follows that this hsp might be involved in chaperoning or shuttling unfolded proteins safely to and from specific sites in membranes. Unless protected by chaperones, such unfolded proteins would be insoluble and would tend to aggregate.
In another study, Raymond Deshaies and his colleagues, working at the University of California and the University of Wisconsin at Madison, used mutated strains of yeast to show that cells need hsp 70 to move proteins across membranes. In essence, the mutants do not make hsp 70, a condition which would be lethal were it not for the introduction into the yeast cells of another copy of the gene on a circular bit of DNA known as a plasmid. The hsp 70 gene on the plasmid is joined to a molecular ‘on-and-off switch’, a promoter that is sensitive to the sugar galactose. Adding galactose to the growth medium switches on the synthesis of hsp 70. Glucose however, switches the synthesis off. In the on position, the yeast translocates several proteins (including prepro factor) to their proper places inside the cells. In the off position, these proteins merely accumulate in the cytoplasm – clear evidence of a role for hsp in moving proteins.
Several other observations, particularly work by Sean Munro at Cambridge, are relevant to the story of hsp 70. Starving a cell of glucose induces it to produce a family of closely related proteins known as glucose regulated proteins, or grps for short. Unlike ‘mainstream’ hsp 70, these proteins seem to be fixed in membranous channels in the cell, the endoplasmic reticulum. One of these proteins, grp 78, associates with the heavy chain of antibody molecules, probably to inhibit aggregation of the chains prior to the production of mature antibodies. Splitting the grp/antibody complex also requires energy from the hydrolysis of ATP. These proteins probably function in much the same way as their cytoplasmic cousins, by stabilising other proteins destined for assembly into or across the membrane of the endoplasmic reticulum.
Cells make other equally important heat-shock proteins, including hsp 90 and the stress proteins of lower molecular weight. The hsp 90 seems to be a common component of certain hormone receptors , while the smaller hsps shows a tendency to form interesting aggregates, the precise function(s) of which are the subject of intensive investigations in their own right.
Researchers at Liverpool Polytechnic and the University of Liverpool are investigating the role of one such aggregate, isolated from the dormant cysts of a shrimp. The shrimp aggregate is a complex formed from multiple copies of a single (putative) heat-shock protein, associating with small RNA molecules in the cytoplasm, giving a native molecular mass of 600 kilodaltons. Other complexes are composed of different protein subunits. After heat shock, some are formed from small hsp precursors and all look similar under the electron microscope.
Aggregates from dormant cysts closely resemble similar complexes found in mammalian, avian and insect tissues. One of these, termed the prosome, probably regulates protein synthesis by selectively sequestering messenger RNA, the intermediary between gene and protein. The shrimp aggregate may also help to preserve messenger RNA during the dehydration and eventual desiccation of its cysts (which, incidentally, can survive for years under extreme environmental conditions). Desiccation is rather like an extreme form of heat shock.
Patricia Falkenburg and her colleagues at the Institute for Molecular Genetics, Heidelberg, and Andre-Patrick Arrigo and his colleagues at the Cold Spring Harbor Laboratory, New York, recently reported that the prosome was immunologically, structurally and biochemically related to a previously recognised complex of protein-destroying enzymes in several types of cell. This work vindicates the idea that structurally related complexes formed by smaller hsps are likely to be functionally related to them as well.
The unravelling of the mystery of how heat-shock proteins work in animal and plant cells is progressing well, at least for the large reservoir of these proteins in unstressed cells. You can be sure, however, that what started with overheated flies 20 years ago still has a long way to go.
* * *
HEAT-SHOCK PROTEINS HELP YOU TO ‘STAY IN THE KITCHEN’
ONE OF the most fascinating aspects of a cell’s response to a heat shock is the development of tolerance to normally lethal temperatures. For example, mammalian cells, which normally grow at 37 C, are rapidly killed at 45 C. If cells in culture are given a sub-lethal heat shock by exposure to 43 C for an hour followed by several hours’ recovery at 37 C, the surviving cells are tolerant to a subsequent heat shock at 45 C. Many will survive even higher temperatures.
Many scientists disagree about which heat-shock proteins are responsible for the development of thermotolerance. Some think that the hsp 70 family is responsible, others prefer to point the finger at the smaller heat-shock proteins. There are also those who maintain that heat-shock proteins are not required at all. This conclusion arose from the frequent observation that inhibitors of protein synthesis of hsp 70, for example, do not appear to block the development of thermotolerance.
But thermotolerance merely provides an additional protection from damage by heat. The heat-shock proteins already present in the cells probably dictate the baseline level of resistance to heat. The additional resistance seen after the first heat treatment must be due to some additional factor synthesised by the cells. William Welch and Lee Mizzen at the Cold Spring Harbor Laboratory, in New York, have produced evidence that a potent inhibitor of protein synthesis, cycloheximide, does not block the development of thermotolerance, because it somehow stabilises complexes called polysomes against heat-induced disaggregation. After a heat shock, these polysomes provide a head start in the resumption of normal protein synthesis.
There is also the possibility that thermotolerance results from a subtle redistribution of heat-shock proteins inside the cell (rather than their heat-induced expression per se). We already know that hsp 70 moves to the nucleus after a heat shock, but hsp 70 and 90 are also known to be associated with elements of the intracellular networks of filaments known as the cytoskeleton. These networks have important functions in almost all cellular events and, given the modulatory and protective roles postulated for heat-shock proteins, it is possible that the dynamic equilibrium of the cytoskeleton may be altered by heat shock. Its stabilisation is an important (and until recently, overlooked) factor in the acquisition of thermotolerance.
* * *
HSP 90, THE NEGLECTED HEAT-SHOCK PROTEIN?
MOST of the literature concerning the heat-shock response makes only a passing reference to hsp 90, even though it is also inducible by heat. Most unstressed cells contain considerable levels of this protein, and hsp 90 from different species appears to be immunologically related. Its function is only now becoming clear.
Monoclonal antibodies, raised against intact steroid hormone receptors, cross-react with hsp 90. Experiments showed that hsp 90 binds away from the hormone-binding site itself. This inevitably led to the conclusion that hsp 90 masks the DNA-binding site of the hormone receptors, until a hormone is positioned within the hormone-binding site (see Figure). When this happens, the hsp 90 is released, enabling the hormone/receptor complex to bind to the DNA. In contrast, receptors exposed to anti-hormones cannot bind DNA and remain associated with hsp 90. So this heat-shock protein definitely does regulate the activity of steroid hormones.
Hsp 90 also turns up in the machinery of protein synthesis. It is a component in a complex set of factors involved in the control of protein synthesis. The precise function of hsp 90 in this system is still unclear, but it appears to modulate the phosphorylation, and therefore the activity, of the alpha subunit of ‘initiation factor 2’ in animal and plant cells. Such modulatory activity is significant because the phosphorylation of proteins, or kinase activity as it is more commonly referred to, is of enormous importance in biochemical reactions. A cell’s response to heat shock may thus be mediated by an hsp.
Finally, as with hsp 70, hsp 90 has an immunologically related counterpart, grp 94, located within the endoplasmic reticulum. Its role in this compartment is still highly speculative, but it is thought to be involved in the assembly of other proteins.
Dr David Miller is lecturer in biochemistry and molecular biology at Liverpool Polytechnic