Heat shock factor

EVOLUTIONARY HOMOLOGS (part 1/2)

Yeast Heat shock factor

Heat shock elements (HSEs) are promoter sites that bind heat shock factor. An in vivo footprinting analysis of the S. cerevisiae HSP82 promoter has shown that yeast heat shock factor binds to strong HSEs in non-heat-shocked cells. Upon heat shock, however, additional binding of HSF becomes apparent at weak heat shock elements of the promoter as well. Recovery from heat shock results in a dramatic reduction in HSF binding at both strong and weak HSEs, consistent with a model in which HSF binding is subject to a negative feedback regulation by heat shock proteins. In vivo KMnO4 footprinting reveals that the interaction of the TATA-binding protein (TBP) with this promoter is also modulated: heat shock slightly increases TBP binding to the promoter and this binding is reduced upon recovery from heat shock. KMnO4 footprinting does not reveal a high density of polymerase at the promoter prior to heat shock, but a large open complex between the transcriptional start site and the TATA box is formed rapidly upon activation, similar to that observed in other yeast genes (Giardina, 1995a).

The anti-inflammatory drug sodium salicylate (aspirin) modulates the activity of specific transcription factors in humans. Salicylate and sorbate, another organic acid, stimulate DNA binding by yeast HSF in vivo. Surprisingly, salicylate inhibits heat shock gene transcription even in cells induced by a prior heat shock. This inhibition of transcription occurs at a step after HSF and transcription factor IID binding but before promoter melting by RNA polymerase. Salicylate appears to generate a tight binding but activation-impotent HSF by means of cytoplasmic acidification, since inhibiting proton efflux from cells triggers this same DNA binding and inhibition of heat shock gene expression (Giardina, 1995b).

In vitro DNA-binding assays demonstrate that the heat shock transcription factor (HSF) from the yeast Saccharomyces cerevisiae can adopt an altered conformation when stressed. This conformation, reflected in a change in electrophoretic mobility, requires that two HSF trimers be bound to DNA. Single trimers do not show this change, which appears to represent an alteration in the cooperative interactions between trimers. HSF isolated from stressed cells displays a higher propensity to adopt this altered conformation. Purified HSF can be stimulated in vitro to undergo the conformational change by elevating the temperature or by exposing HSF to superoxide anion. Mutational analysis maps a region critical for this conformational change to the flexible loop between the minimal DNA-binding domain and the flexible linker that joins the DNA-binding domain to the trimerization domain. The significance of these findings is discussed in the context of the induction of the heat shock response by ischemic stroke, hypoxia, and recovery from anoxia, all known to stimulate the production of superoxide (Lee, 2000).

The mutation M232V affects a highly conserved position within the DNA-binding domain of HSF. To assess whether the mutant might differ from wild-type HSF in DNA-binding ability, HSF binding to a synthetic HSE that accommodates only a single trimer was examined. There is no detectable difference in complex formation between wild-type and mutant HSF. Both form a single complex under nonshocked conditions. Although the inherent DNA-binding ability of HSFM232V appears normal, it might be possible that the lesion affects the ability of two bound trimers to interact. To test this possibility, the DNA-binding assays were repeated using a synthetic HSE that binds two HSF trimers (HSE6T). On this HSE, HSF forms three distinct complexes (I, II, and III), whose abundances vary depending on the source of HSF. Complex I has the identical mobility as a complex that forms on an HSE that can accommodate a single HSF trimer and therefore represents the binding of a single trimer to this HSE. Complex II has a mobility identical to that of a complex comprised of six HSF monomers and thus represents the binding of two trimers to HSE6T. Wild-type HSF from nonshocked cells forms almost exclusively complexes I and II, whereas HSF from heat-shocked cells forms a modest amount of complex III in addition to complexes I and II. Complex III becomes very abundant when HSFM232V is examined. The amount of complex III increases with HSF activity, showing quite clearly that the DNA-binding pattern of HSF is affected by its in vivo activity state. These observations suggest that complex III might represent an active species of HSF (Lee, 2000).

It has long been thought that the activation of HSF upon heat shock must involve a conformational change that 'unmasks' the transcriptional activation domains of HSF. There have been a number of analyses of HSF conformational changes, but these have generally examined metazoan HSF free in solution and have provided information primarily on the monomer-trimer transition. The observation that salicylate can induce human HSF to trimerize and bind DNA without concomitant activation of hsp70 expression suggests that there is likely to be a second aspect of HSF activation beyond trimerization. The observations presented here provide information about this second step, at least for Saccharomyces HSF (Lee, 2000).

The argument that complex III represents an active form of HSF is based on three kinds of evidence. The first is correlative, as it must be for an assay based on electrophoretic mobility. The propensity of wild-type HSF to form complex III in vitro depends, at least in part, on the physiological state of the cells from which the HSF is extracted. Heat shock favors the formation of complex III; recovery from heat shock disfavors it. Chemical stresses that affect the heat shock response also favor the formation of complex III. These correlations are imperfect, however. In part, this may reflect the fact that HSF can be induced to form complex III by exposure to superoxide, which is generated spontaneously in the DNA-binding reactions that are exposed to atmospheric oxygen. The imperfect correlation may also reflect the fact that reduced glutathione, present in whole-cell extracts, counteracts the induction of complex III in vitro (Lee, 2000).

The second line of evidence is that HSF can be induced to form complex III in vitro in response to agents that are known to induce HSF activity in vivo: heat and oxidation. If the mechanism of HSF activation were a direct effect of these stress agents, rather than a complex signal transduction cascade, this is the very result that would be predicted. The third line of evidence that complex III represents an active conformation of HSF is genetic. Several mutations that elevate HSF activity in vivo also increase the formation of complex III in vitro. Together, these three lines of evidence suggest that complex III may, indeed, represent a conformation that HSF can adopt upon activation (Lee, 2000).

The data do not reveal the nature of the conformational change that gives rise to complex III. However, they do provide two important pieces of information. First, the change in conformation is detectable only when two HSF trimers are bound cooperatively to DNA; singly bound trimers show no detectable differences in gel mobility shift assay. This suggests that the conformational change represents a change in the nature of the cooperative interaction between trimers. This is intriguing, because it helps explain how differences in the architecture of HSEs can exhibit differences in biological activity that are not easily explained simply by differences in binding affinity. For example, the HSEs of the yeast CUP1 gene and of the human IL1beta gene do not provide for cooperative interactions between trimers and display activities that are distinct from traditional HSEs. In contrast, traditional HSEs of natural heat shock genes typically have binding sites for two or more HSF trimers. In both yeast and Drosophila, these cooperative interactions have been shown to be either important (yeast) or critical (Drosophila) for normal heat shock-inducible transcription (Lee, 2000).

These data also map the regions that are important for complex III formation. The N-terminal "masking" domain, the flexible loop joining the DNA-binding domain to the linker domain, and residue Met₂₃₂ all seem to play roles in the transition between complexes II and III. Interestingly, these all share a certain physical proximity. The structural data for the DNA-binding domain indicate that the N-terminal domain joins the DNA-binding domain near the flexible loop, which in turn is capable of closely approaching Met₂₃₂. It is imagined that the flexible loop adopts one conformation in complexes I and II and a different conformation in complex III (Lee, 2000).

The data indicate that the transition to complex III can be induced in vitro by stimulation of purified HSF by heat or by the superoxide anion O₂^-. This indicates that HSF is directly responsive to the two stresses that are best known as inducers of the heat shock response: heat and oxidative stress. The difference that yeast HSF responds to O₂^- whereas Drosophila HSF responds to H₂O₂ may be caused by methodological differences or by differences in the physiology or in the sequences of HSF, but the similarity is striking. The production of O₂^- is coupled in vivo to the production of H₂O₂ via the dismutation of superoxide into H₂O₂. The major source of cellular O₂^- is mitochondrial metabolism, via which, it has been estimated, 1%-2% of electrons are 'spilled' into the generation of superoxide. The mitochondrial Mn-dependent superoxide dismutase rapidly converts much of this to H₂O₂. The generation of superoxide will increase during stresses that affect mitochondrial activity. Thus, the same basic physiological stress creates both species of reactive oxygen (Lee, 2000).

The finding that HSF is directly responsive to superoxide (or, in the case of Drosophila HSF, its byproduct H₂O₂) provides a mechanistic explanation for the finding that the heat shock response is induced by such stresses as recovery from anoxia, hypoxia, and ischemic stroke. For aerobic organisms, all of these stresses have the same result: a burst of superoxide production. Via the dismutation of O₂^- to H₂O₂ and the subsequent conversion to the hydroxyl radical (·OH), these treatments lead to significant cellular damage -- the classic 'reperfusion injury'. It has long been thought that the induction of a heat shock response was the result of the ·OH-induced protein damage, stimulating an unfolded protein response. These findings indicate that HSF is activated directly by the oxidative stress and may therefore occur before there is extensive protein damage. Thus, the induction of the heat shock response by anoxic stresses is immediate and potentially protective, rather than a mere reaction to damage previously done (Lee, 2000).

In analyses of protein damage by reactive oxygen species, it is usually the ·OH that causes damage after production of O₂^-. Hydroxylation of aromatic amino acid residues can usually be prevented by ·OH scavengers, but not by superoxide dismutase. It is therefore surprising that it is O₂^- itself that activates HSF. Superoxide seems to work directly, and probably reversibly, on yeast HSF. It is unlikely that O₂^- simply binds to HSF, because at least some HSF remains in its 'complex III-competent' form after extraction from stressed cells. More likely, O₂^- modifies one or more amino acid residues, in which case conversion of activated HSF back to the inactive form must entail the reversal of this modification (Lee, 2000).

It is possible that the inactivation of HSF during recovery from stress requires two activities: hsp70 proteins to help refold HSF and one or more glutathione-dependent enzymatic activities. It is striking that high concentrations of GSH can completely prevent nonshocked HSF from binding to DNA, as judged by gel mobility shift assay. This suggests that there may be an 'inactive' form of HSF that has a very low affinity for DNA, as is the case in metazoans. On the basis of these findings, it is suspected that this may occur via the GSH-dependent deactivation of HSF (Lee, 2000).

C. elegans Heat shock factor

In a screen for suppressors of activated GOA-1 (Galpha_o) under the control of the hsp-16.2 heat-shock promoter, three genetic loci were identified that affect heat-shock-induced GOA-1 expression. The cyl-1 (coding for an RS splice factor, Drosophila homolog: CG16903; vertebrate homolog Cyclin L) mutants are essentially wild type in appearance, while hsf-1 and sup-45 mutants have egg-laying defects. The hsf-1 mutation also causes a temperature-sensitive developmental arrest, and hsf-1 mutants have decreased life span. Western analysis indicates that mutations in all three loci suppress the activated GOA-1 transgene by decreasing its expression. Heat-shock-induced expression of hsp-16.2 mRNA is reduced in cyl-1 mutants and virtually eliminated in hsf-1 and sup-45 mutants, as compared to wild-type expression. The mutations could also suppress other transgenes under heat-shock control. cyl-1 and sup-45, but not hsf-1, mutations suppressed a defect caused by a transgene not under heat-shock control, suggesting a role in general transcription or a post-transcriptional aspect of gene expression. hsf-1 encodes the C. elegans homolog of the human heat-shock factor HSF1, and cyl-1 encodes a cyclin most similar to cyclin L. It is believed that HSF-1 acts in heat-shock-inducible transcription and CYL-1 acts more generally in gene expression (Hajdu-Cronin, 2004).

Xenopus Heat shock factor

There are distinct stress-inducible and developmentally regulated heat shock response element (HSE)-binding activities in Xenopus. The stress-induced HSE-binding complex is recognized by antiserum against mammalian HSF1, but not by HSF2 antiserum, suggesting that a Xenopus homolog of HSF1 is the major component of this activity. There is in addition an unusual constitutive HSE-binding complex in unstressed stage I and II oocytes, but not in later stage oocytes, eggs, or developing embryos. This constitutive complex is unaffected by heat or chemical treatments and is not recognized by either HSF1 or HSF2 antiserum. Appearance of the constitutive HSE-binding activity during oogenesis corresponds closely with peak levels of hsp70 mRNA. It is thought that the constitutive HSE-binding activity in early oocytes is formed by a unique developmentally regulated heat shock factor that may play a role in the expression of heat shock proteins during early stages of oogenesis (Gordon, 1997).

Stress-induced expression of the heat shock (hs) genes in eukaryotes is mediated by a transcription factor known as heat shock factor 1 (HSF1). HSF1 is present in a latent, monomeric form in unstressed metazoan cells and upon exposure to heat or other forms of stress is converted to an "active" trimeric form, which binds the promoters of hs genes and induces their transcription. The conversion of HSF1 to its active form is hypothesized to be a multistep process involving (1) oligomerization of HSF1, plus (2) additional changes in its physical conformation, (3) changes in its phosphorylation state, and for some species (4) translocation from the cytoplasm to the nucleus. Oligomerization of HSF appears to be essential for high affinity DNA binding, but it remains unclear what their mechanistic roles may be, or whether the other steps occur in all organisms. Germinal vesicles (nuclei) that are physically dissected from unshocked Xenopus laevis oocytes contain no HSF1 binding activity. Interestingly, in vitro heat shock treatments of isolated nuclei from unshocked oocytes activate HSF1 binding, indicating that HSF1 must have been present in the unshocked nuclei prior to isolation. Induction of HSF1 binding is not observed in enucleated oocytes. Western blot analysis using an affinity-purified polyclonal antibody made against X. laevis HSF1 shows that HSF1 is present in equal amounts in unshocked and shocked oocytes and isolated nuclei. HSF1 is not detected in enucleated oocytes. These results clearly demonstrate that HSF1 is a nuclear protein in oocytes prior to exposure to stress. In Xenopus oocytes, therefore, HSF1 translocation from the cytoplasm to the nucleus is not part of the multistep process of HSF1 activation. These results also imply that the signals and/or factors involved in HSF1 activation must have their effect in the nuclear compartment (Mercier, 1997).

Mammalian Heat shock factor

Cell and tissue injury activate the inflammatory response through the action(s) of arachidonic acid and its metabolites, leading to the expression of acute-phase proteins and inflammatory cytokines. At the molecular level, little is known how arachidonic acid regulates the inflammatory response. As inflammation is also associated with local increase in tissue temperatures, arachidonic acid has been examined to determine whether it is directly involved in the heat shock response. Extracellular exposure to arachidonic acid induces heat shock gene transcription in a dose-dependent manner via acquisition of DNA-binding activity and phosphorylation of heat shock factor 1 (HSF1). Exposure of cells to low concentrations of arachidonic acid, which by themselves does not induce HSF1 DNA-binding activity, reduces the temperature threshold for HSF1 activation from elevated temperatures that are not physiologically relevant (> 42 degrees C) to temperatures that can be attained during the febrile response (39-40 degrees C). These results indicate that elevated heat shock gene expression is a direct consequence of an arachidonic acid-mediated cellular response (Jurivich, 1994).

Promoter-proximal pausing during transcriptional elongation is an important way of regulating many diverse loci, including the human hsp70 gene. Pausing of RNA polymerase can be enhanced by chromatin structure. Activation of hsp70 leads to disruption of transcribed chromatin in front of RNA polymerase. In vivo, disruption of chromatin in the first 400 bp of the transcribed region of hsp70 following heat shock is resistant to the transcriptional inhibitor alpha-amanitin. Disruption of chromatin farther downstream also occurs following activation but is sensitive to alpha-amanitin, suggesting that polymerase movement is needed to disrupt distal portions of the hsp70 gene. In vitro, disruption of transcribed chromatin is dependent on the presence of the human heat shock factor 1 (HSF1) activation domains. These experiments demonstrate that HSF1 can direct disruption of chromatin in transcribed regions. It is suggested that this is one of the mechanisms used by HSF1 to facilitate transcriptional elongation. Previous studies have shown that SWI/SNF-containing fractions augment the readthrough of pausing on hsp70 in an activator-dependent fashion (Brown, 1997).

A preparation of mixed male germ cell types from mouse testis exhibit a lower temperature threshold for activation of HSF1 DNA binding relative to other mouse cell types. Is the phenomenon of reduced HSF1 activation temperature common to all testis cell types, both somatic and germ cell types, or is it a special property of male germ cells? A purified population of pachytene spermatocytes, one of the male germ cell types, exhibits a profile of reduced HSF1 activation temperature identical to that observed for the mixed germ cell preparation, with a threshold HSF1 activation temperature of 35 degrees C. Activation of HSF1 DNA binding in male germ cells by incubation at 38 degrees C is accompanied by the classic cellular stress response parameters of heat-induced HSF1 phosphorylation and increased expression of the hsp72 stress protein. In contrast, a preparation of somatic testis cell types exhibits HSF1 activation only at temperatures of 42 degrees C and above, a profile identical to that observed for mouse liver cells and mammalian cell lines. These results demonstrate that the phenomenon of reduced HSF1 activation temperature is a unique property of male germ cell types within the mammalian testis and demonstrate that HSF1 activated at this lower temperature threshold is fully capable of mediating a productive cellular stress response in these cell types (Sarge, 1995).

Heat shock factor 1 inhibits the expression of c-fos (See Drosophila Fos), an immediate early gene that controls responses to extracellular stimuli for growth and differentiation. Heat shock factor 1 inhibits the transcription of the c-fos gene and antagonizes the activating effects of the signal transducing protein Ras on the c-fos promoter and on the promoter of another Ras responsive gene, uPA. This property is specific for heat shock factor 1; c-fos repression is not seen with the structurally related protein heat shock factor 2. Repression involves different molecular mechanisms when compared with those involved in transcriptional activation by heat shock factor 1, and specifically does not require binding to the c-fos promoter. Thus, in addition to its known role as a transcriptional activator of the cellular heat shock response, heat shock factor 1 also antagonizes the expression of Fos, a key component of the ubiquitous AP-1 transcription factor complex; as such, it could influence multiple aspects of cell regulation (Chen, 1997).

Heat shock transcription factor 1 (HSF-1) activates the transcription of heat shock genes in eukaryotes. Under normal physiological growth conditions, HSF-1 is a monomer. Its transcriptional activity is repressed by constitutive phosphorylation. Upon activation, HSF-1 forms trimers, acquires DNA binding activity, increases transcriptional activity, and appears as punctate granules in the nucleus. In this study, using bromouridine incorporation and confocal laser microscopy, it has been demonstrated that newly synthesized pre-mRNAs colocalize to the HSF-1 punctate granules after heat shock, suggesting that these granules are sites of transcription. Evidence that glycogen synthase kinase 3beta (GSK-3beta) and extracellular signal-regulated kinase mitogen-activated protein kinase (ERK MAPK) participate in the down regulation of HSF-1 transcriptional activity. Transient increases in the expression of GSK-3beta facilitate the disappearance of HSF-1 punctate granules and reduce hsp-70 transcription after heat shock. ERK is shown to be the priming kinase for GSK-3beta. Taken together, these results indicate that GSK-3beta and ERK MAPK facilitate the inactivation of activated HSF-1 after heat shock by dispersing HSF-1 from the sites of transcription (He, 1998).

GSK-3 consists of two isoforms: GSK-3alpha (51 kDa) and GSK-3beta (46 kDa). It was first identified as an enzymatic activity that phosphorylates and inactivates glycogen synthase. A second role of GSK-3 was found when studies showed that inhibition of phosphatase type I activity is relieved when GSK-3 phosphorylates phosphatase inhibitor 2. At least 15 other substrates have been reported to be phosphorylated by GSK-3, including the transcription factors c-Jun, JunD, c-myb, c-myc, L-myc, CREB, and NF-AT, most of which become inactivated when phosphorylated by GSK-3. GSK-3 tends to phosphorylate serine/threonine residues located next to a proline which, in turn, is near another serine residue that has been prephosphorylated by some other protein kinase (referred to as priming kinase). GSK-3 is constitutively active and, as a result, suppresses many of its substrates under normal physiological growth conditions. It appears that the activity of HSF-1 can be down regulated by protein kinases that are activated by diverse signal transduction pathways. The ERK MAPK pathway is activated during cell growth and development by multiple signaling pathways, which in turn are activated by growth factor receptors, G protein-coupled receptors, ceramide production, and a protein kinase C-dependent pathway. Recent evidence has suggested that ERK MAPK activation by heat shock may be through ceramide activation of protein kinase Raf-1. The pathways leading to GSK-3beta regulation are complex. GSK-3beta activity is down regulated by PKB/Akt, p70S6K, or p90rsk as a result of phosphorylation on serine residues. Activation of PKB/Akt leads to increased cell survival, as is the case with activation of ERK. The ability of ERK to mediate cell survival is dependent on the activation of transcription factors such as Elk1 and repair of damaged proteins. The ability of PKB/Akt to mediate cell survival is likely to be dependent on downstream effectors such as p70S6K and protein translation, activation of FRAP/TOR, and inhibition of GSK-3beta. Interestingly, heat shock stimulates the activity of GSK-3beta and ERK MAPK. The increase in GSK-3beta activity may occur through its phosphorylation on a tyrosine residue by an unknown tyrosine kinase. Thus, it appears that when activated, HSF-1 reduces the expression of most other genes and must be inactivated in a timely manner for cell proliferation to continue. The cell has developed an elegant mechanism for doing this, since some of the enzymes that control cell proliferation are capable of inactivating HSF-1 (He, 1998 and references).

Heat-shock factor-1 (HSF1) is a major transactivator of stress-inducible genes in response to environmental changes, but it is also implicated in extra-embryonic development and female fertility in mice. Mouse embryos whose mothers lack this protein are unable to develop properly beyond the zygotic stage, although oocytes are ovulated and fertilized normally. Wild-type spermatozoa do not save zygotes from lethality, indicating that the reproductive failure of females deficient in this factor is caused by a 'maternal effect' mutation, and that HSF1 from the mother normally controls early post-fertilization development (Christians, 2000).

Female mice lacking the gene encoding HSF1 (Hsf1^-/- females) have normal ovaries and reproductive tracts, indicating that folliculogenesis and oogenesis do not require HSF1 expression, by contrast with the Drosophila Hsf mutation and others affecting fertility in mammals. Ovulated eggs of Hsf1^-/- females remain properly arrested at metaphase II until fertilization, after which zygotes form with two pronuclei and a second polar body. Although mutant embryos produced by Hsf1 ^-/- females can initiate early development, they do not survive after transplantation into the oviduct of Hsf1 wild-type mice, indicating that the causes of Hsf1^-/- infertility are intrinsic rather than extrinsic (Christians, 2000).

To determine at which stage development breaks down, an analysis was made of the cleavage efficiency of preimplantation embryos produced by Hsf1^-/- females, using Hsf1^+/- females as controls, mated with either wild-type or homozygous males. The percentage of blastocysts at 3.5 days of development was the same and independent of the genotype of the father. In contrast, the embryos of homozygous females mated with either Hsf1 ^+/+ or Hsf1^-/- males are blocked mainly at the 1-cell stage, and development at 3.5 days is poor . Unlike the embryos produced by heterozygous females, none of these mutant embryos reach the blastocyst stage at the appropriate time, and 18% embryos from the Hsf1 ^-/- x Hsf1^-/- intercross reached the 2-cell stage. A significantly higher fraction of those sired by Hsf1^+/+ males progress beyond the 1-cell division, but the difference does not persist beyond this stage. These findings show that the survival of the offspring is determined mainly by the maternal genotype, and that Hsf1^-/- is therefore a maternal-effect mutation (Christians, 2000).

Embryos might die at the 1-2-cell stage because of a failure to initiate the zygotic transcriptional activity required to complete the transition from maternal to embryonic control of development. Since spontaneous expression of Hsp70.1 indicates the onset of transcriptional activity at the late 1-cell stage, Hsp70.1-Luc transgenic lines, in which a luciferase reporter gene is driven by the murine Hsp70.1 promoter, were used to test this idea directly. At 1.5 days of development, embryos from Hsf1 ^-/- and Hsf1^-/- females that had been mated with Hsf1^+/+ transgenic males both showed luciferase activity significantly above the background level. The Hsf1-mutant zygotes therefore apparently express heat-shock proteins, showing that zygotic transcriptional activity can begin without HSF1 (Christians, 2000).

These results do not exclude the possibility that the level of transcriptional activity, or the control of the target genes to be transcribed, might have been altered in mutant embryos. Furthermore, ultrastructural abnormalities affect mainly the nuclei of embryos from null females at the 2-cell stage, indicating that, even with transcriptional activity, reduced survival is in part related to the decrease in structural integrity of the early embryonic nucleus. The key role for control by maternal-origin HSF1 in early development raises the possibility that in defective form it may be a cause of post-fertilization abnormalities associated with infertility in mammals, including humans (Christians, 2000).

Heat shock factor oligomerization and DNA-binding specificity>

Heat shock transcription factors (HSFs) are multi-zipper proteins with high-affinity binding to DNA that is regulated by heat shock-induced trimerization. Formation of HSF trimers is dependent on hydrophobic heptad repeats located in the amino-terminal region of the protein. Two subregions at the carboxyl-terminal end of human HSF1 have been identified that maintain the monomeric form of the protein under normal conditions. One of these contains a leucine zipper motif that is conserved between vertebrate and insect HSFs. These results suggest that the carboxyl-terminal zipper may suppress formation of trimers by the amino-terminal HSF zipper elements by means of intramolecular coiled-coil interactions that are sensitive to heat shock (Rabindran, 1994).

In unstressed mammalian cells, HSF1 is present mainly in complexes with an apparent molecular mass of about 200 kDa, unable to bind to DNA. Heat treatment induces a shift in the apparent molecular mass of HSF1 to about 700 kDa, concomitant with the acquisition of DNA-binding ability. Cross-linking experiments suggest that this change in complex size may reflect the trimerization of monomeric HSF1. Human HSF1 expressed in Xenopus oocytes does not bind DNA, but derepression of DNA-binding activity, as well as oligomerization of HSF1, occurs during heat treatment at the same temperature at which hsp gene expression is induced in this organism, suggesting that a conserved Xenopus protein(s) plays a role in this regulation. Inactive HSF1 resides in the cytoplasm of human cells; on activation it rapidly translocates to a soluble nuclear fraction, and shortly thereafter it becomes associated with the nuclear pellet. On heat shock, activatable HSF1, which might already have been posttranslationally modified in the unstressed cell, undergoes further modification (Baler, 1993).

Heat stress regulation of human heat shock genes is mediated by human heat shock transcription factor hHSF1, which contains three 4-3 hydrophobic repeats (LZ1 to LZ3). In unstressed human cells (37 degrees C), hHSF1 appears to be in an inactive, monomeric state that may be maintained through intramolecular interactions stabilized by transient interaction with hsp70. Heat stress (39 to 42 degrees C) disrupts these interactions, and hHSF1 homotrimerizes and acquires heat shock element DNA-binding ability. hHSF1 expressed in Xenopus oocytes also assumes a monomeric, non-DNA-binding state and is converted to a trimeric, DNA-binding form upon exposure of the oocytes to heat shock (35 to 37 degrees C in this organism). Because endogenous HSF DNA-binding activity is low, and anti-hHSF1 antibody does not recognize Xenopus HSF, this system was used for mapping regions in hHSF1 that are required for the maintenance of the monomeric state. The results of mutagenesis analyses strongly suggest that the inactive hHSF1 monomer is stabilized by hydrophobic interactions involving all three leucine zippers that may form a triple-stranded coiled coil. Trimerization may enable the DNA-binding function of hHSF1 by facilitating cooperative binding of monomeric DNA-binding domains to the heat shock element motif. This view is supported by observations that several different LexA DNA-binding domain-hHSF1 chimeras bind to a LexA-binding site in a heat-regulated fashion; that single amino acid replacements disrupting the integrity of hydrophobic repeats render these chimeras both constitutively trimeric and DNA binding, and that in these assays LexA itself binds stably to DNA only as a dimer but not as a monomer (Zuo, 1994).

Comparison of wild-type and mutant hHSF1 expressed in Xenopus oocytes and human HeLa cells suggests that retention of hHSF1 in the monomeric form depends on hydrophobic repeats (LZ1 to LZ3) and a carboxy-terminal sequence element in hHSF1 as well as on the presence of a titratable factor in the cell. Oligomerization of hHSF1 appears to induce DNA-binding activity as well as to uncover an amino-terminally located nuclear localization signal. A mechanism distinct from that controlling oligomerization regulates the transcriptional competence of hHSF1. Components of this mechanism have been mapped to a region, including LZ2 and nearby sequences downstream from LZ2, that is clearly separated from the carboxy-terminally located transcription activation domain(s). A fold-back structure is thought to mask the transcription activation domain in the unstressed cell but it is opened up by modification of hHSF1 and/or binding of a factor facilitating hHSF1 unfolding in the stressed cell. Activation of hHSF1 appears to involve at least two independently regulated structural transitions (Zuo, 1995).

HSF1 exists in unstressed cells in an inactive form, which is converted to the DNA binding from upon exposure of cells to elevated temperature. A protocol has been developed for isolation of the non-DNA binding form of recombinant mouse HSF1, involving expression and affinity purification of HSF1 as a fusion with the glutathione S-transferase protein in Escherichia coli, followed by specific protease cleavage to release pure HSF1 protein. The purified inactive HSF1 can be converted to the DNA binding form by heat treatment in vitro. This conversion is accompanied by oligomerization of HSF1 from a monomeric to a trimeric native structure, similar to that observed for HSF1 in heat-shocked cells. These results indicate that elements residing in the HSF1 polypeptide are sufficient both for maintenance of this factor in the non-DNA binding form and for its heat-induced conversion to the DNA binding form, and support a role for HSF1 as the "molecular thermostat" in eukaryotic cells, which senses adverse environmental conditions and activates the cellular stress response (Goodson, 1995a).

One step in the pathway leading to transcriptional activation of heat shock genes involves heat shock factor 1 (HSF1) trimerization, required for high-affinity binding of this activator protein to heat shock elements (HSEs) in the promoters. Previous studies have shown that in vivo the trimerization is negatively regulated at physiological temperatures by a mechanism that requires multiple hydrophobic heptad repeats (HRs) which may form a coiled coil in the monomer. To investigate the minimal requirements for negative regulation, mouse HSF1 translated in both rabbit reticulocyte lysate or extracted from Escherichia coli were examined after limited expression. Under these conditions HSF1 behaves as a monomer that can be induced by increases in temperature to form active HSE-binding trimers. Mutations of either HR region cause activation in both systems. Temperature elevations and acidic buffers activate purified HSF1, and mild proteolysis excises fragments that form HSE-binding oligomers. These results suggest that oligomerization can be repressed in the monomer and that repression can be relieved in the apparent absence of regulatory proteins. An intramolecular mechanism may be central for the regulation of this transcription factor in mammalian cells, although not necessarily sufficient (Farkas, 1998).

Eukaryotic heat shock transcription factors (HSF) regulate an evolutionarily conserved stress-response pathway essential for survival against a variety of environmental and developmental stresses. Although the highly similar HSF family members have distinct roles in responding to stress and activating target gene expression, the mechanisms that govern these roles are unknown. A loop within the HSF1 DNA-binding domain has been identified that dictates HSF isoform specific DNA binding in vitro and preferential target gene activation by HSF family members in both a yeast transcription assay and in mammalian cells. These characteristics of the HSF1 loop region are transposable to HSF2 and sufficient to confer DNA-binding specificity, heat shock inducible HSP gene expression and protection from heat-induced apoptosis in vivo. In addition, the loop suppresses formation of the HSF1 trimer under basal conditions and is required for heat-inducible trimerization in a purified system in vitro, suggesting that this domain is a critical part of the HSF1 heat-stress-sensing mechanism. It is proposed that this domain defines a signature for HSF1 that constitutes an important determinant for how cells utilize a family of transcription factors to respond to distinct stresses (Ahn, 2001).

In mammals, three nonredundant HSF family members have been identified. The physiological and environmental signals that activate HSF2 and HSF4, as well as the respective roles of all three HSFs to developmental pathways, remain to be elucidated; however, detailed biochemical analyses of these proteins have revealed interesting and significant differences with respect to their activation and DNA binding characteristics. Previous genetic analyses of hsf1^-/- cells have established that HSF1 is the only heat-inducible HSF in mice, and in vitro biochemical studies have shown differences between HSF1 and HSF2 in the degree of cooperative DNA binding on the hsp70 promoter. A multifunctional region of the HSF1 DNA-binding domain, the loop, has been mapped that allows discrimination between different HSE-HSF interactions and acts as a part of the sensing and regulatory mechanism for responding to thermal stress. The HSF1-specific DNA-binding property can be transposed to HSF2 by creating a hybrid of HSF2 with the loop of HSF1, and significantly, the HSF2Loop1 chimera is capable of activating transcription of an HSP70-lacZ reporter in yeast and the endogenous inducible hsp70 gene in mouse cells. Based on previous data that HSF1 binds to HSEs with greater cooperativity than HSF2 does, it is proposed that the HSF1 loop facilitates interactions between HSF1 trimers, thereby promoting its binding to extended HSEs. Such cooperative DNA binding has also been established for lower eukaryotic HSFs from yeast and fly. The HSF2 loop seems much less capable of promoting such cooperativity between trimers, as evidenced by its smaller footprint over the hsp70 HSE and weaker activation of the yeast HSP70-lacZ reporter. Therefore the loop domain may be a critical determinant in allowing HSF1 and HSF2 to activate target genes selectively based on the arrangement of the HSE (Ahn, 2001).

The idea that different arrangements of response elements for a transciption factor may influence the pattern of gene regulation was proposed to explain how the glucocorticoid receptor could function as both a transcriptional activator and repressor depending upon the promoter context. The allosteric control hypothesis posited that distinct DNA recognition sequences influence the activity of a transcriptional activator by dictating the conformation of the DNA-binding protein bound to the site, which in turn specifies the protein surfaces available for interactions with other transcription factors. Similarly, differences in the orientation of the HSE have been reported to promote distinct binding by yeast HSF. An arrangement of two repeats of the tail-tail sequence (nTTCnnGAAn) was found to bind two HSF trimers much more efficiently than two repeats of the head-head orientation (nGAAnnTTCn). Furthermore, a comparison of yeast HSF binding to the gapped CUP1 promoter and a continuous HSE by proteolytic cleavage assays revealed differences in the sensitivity of HSF to the protease. This difference in conformation was suggested to be one factor in determining the temperature threshold of HSF target gene activation. Although these studies show that yeast HSF is capable of adapting to different HSEs, this analysis of HSF1 and HSF2 suggests that the mammalian proteins exhibit more defined preferences for DNA binding and target gene activation. Although both proteins bind to the same core HSE sequences, differential binding and transactivation from shorter versus longer arrays of the pentameric HSE repeats by HSF2 and HSF1, respectively, would provide a basis for differential target gene activation in mammals. Interactions with other cis- and trans-acting factors may influence the protein-DNA interactions at some promoters. Although these data have addressed the molecular basis for discrimination for HSEs by HSF1 and HSF2, there may be overlapping targets for HSF1 and HSF2. For example, hsp70 can be activated in an HSF2-dependent manner during hemin-induced differentiation of K562 cells. However, it was observed that HSF1 and HSF2 differentially occupy the hsp70 HSE and activate the gene to significantly different levels in response to the heat shock and hemin, respectively. Thus, it is proposed that distinctions in the arrangement of the HSEs will provide an additional layer of selectivity among HSF-dependent target gene expression, especially under limiting levels of HSF1 and HSF2 protein concentrations, and may also modulate transcriptional potency. A full characterization of HSF2-preferred HSEs in vivo must await the identification of bona fide HSF2 target genes (Ahn, 2001).

A question that is raised is how might the loop facilitate cooperative interactions. A potential answer may be found from the cocrystal structure of two K. lactis HSF DBDs bound to two inverted HSE repeats. Although disordered in that structural model, the loop was predicted to lie in the interface between the two DNA-bound HSF DBDs. The loop of one DBD was thought to be juxatoposed against the turn of the helix-turn-helix motif of an adjacent DBD. Although not encompassed by the structure, the loop-and-turn interface was also predicted to bridge interactions between adjacent trimers bound to an extended HSE. Thus the loop may comprise part of a critical protein-protein interaction surface that is responsible for stabilizing contacts between individual DBDs within a DNA-bound trimer and between adjacent trimers (Ahn, 2001).

Even more striking, it has been found that the HSF1 loop is sufficient for conferring heat-inducible trimerization to purified HSF2 in vitro when transposed into this isoform, suggesting that the loop is part of the critical heat-sensing mechanism that distinguishes HSF1. A physiological significance for the HSF1 loop in cells is shown because the HSF2Loop1 protein, like HSF1 but not HSF2, is capable of rescuing transfected hsf1^-/- MEFs from heat-induced apoptosis. Therefore, it is proposed that the HSF1 loop modulates the monomer-to-trimer transition in response to heat stress in vitro and in cells in addition to facilitating cooperative binding on extended HSEs. A potential model is that the loop undergoes a conformational change upon heat stress, which relieves intramolecular repression of the monomer and exposes the trimerization domain (Ahn, 2001).

Heat shock factor alternative splicing

Continue: Evolutionary Homologs part 2/2