Roles of the heat shock transcription factors in
regulation of the heat shock response and beyond
LILA PIRKKALA,*,1 PÄIVI NYKÄNEN,* AND LEA SISTONEN*,†,2
*Turku Centre for Biotechnology, University of Turku and Åbo Akademi University; and
†
Department of Biology, Åbo Akademi University, Turku, Finland
The heat shock response, characterized
by increased expression of heat shock proteins (Hsps)
is induced by exposure of cells and tissues to extreme
conditions that cause acute or chronic stress. Hsps
function as molecular chaperones in regulating cellular
homeostasis and promoting survival. If the stress is too
severe, a signal that leads to programmed cell death,
apoptosis, is activated, thereby providing a finely tuned
balance between survival and death. In addition to
extracellular stimuli, several nonstressful conditions
induce Hsps during normal cellular growth and development. The enhanced heat shock gene expression in
response to various stimuli is regulated by heat shock
transcription factors (HSFs). After the discovery of the
family of HSFs (i.e., murine and human HSF1, 2, and 4
and a unique avian HSF3), the functional relevance of
distinct HSFs is now emerging. HSF1, an HSF prototype, and HSF3 are responsible for heat-induced Hsp
expression, whereas HSF2 is refractory to classical
stressors. HSF4 is expressed in a tissue-specific manner; similar to HSF1 and HSF2, alternatively spliced
isoforms add further complexity to its regulation. Recently developed powerful genetic models have provided evidence for both cooperative and specific functions of HSFs that expand beyond the heat shock
response. Certain specialized functions of HSFs may
even include regulation of novel target genes in response to distinct stimuli.—Pirkkala, L., Nykänen, P,
Sistonen, L. Roles of the heat shock transcription
factors in regulation of the heat shock response and
beyond. FASEB J. 15, 1118 –1131 (2001)
ABSTRACT
Key Words: heat shock element 䡠 hsps 䡠 HSF family 䡠 knockout
and transgenic HSF models
BACKGROUND
The heat shock response was discovered in 1962 by F.
Ritossa, who detected a new puffing pattern upon heat
shock in the polytene chromosomes of the fruit fly
Drosophila buschii (1). Today it is well known that all
organisms share a common molecular stress response
that includes a dramatic change in the pattern of gene
expression and the elevated synthesis of a family of
stress-induced proteins called heat shock proteins
(Hsps) (for a review, see ref 2). Expression of Hsps
usually results in repair of damaged proteins and
1118
survival of the cell, mainly through the chaperone
function of Hsps (reviewed in ref 3). The common
signal generated by various stress stimuli is likely to be
protein damage, but how the stress sensing mechanism
in the cell exactly operates is not known.
The heat shock response can also induce a death
signal that leads to apoptosis or rapid necrosis. The
expression of small Hsps, especially Hsp27, and the
inducible Hsp70 has been shown to enhance the survival of mammalian cells exposed to numerous types of
stimuli that induce stress and apoptosis (for reviews, see
refs 4, 5). The antiapoptotic Hsp27 and Hsp70 are
abundantly expressed in many malignant human tumors (reviewed in ref 5). With regard to the cytoprotective functions of Hsps, Hsp70 has been shown to
contribute to protection of myocardium from ischemic
injuries (6 – 8). The basis for the cardioprotective activity of Hsp70 is likely to be related to its ability to prevent
protein aggregation during ischemic stress. Considering the key role of Hsps in protection against stressinduced damage, it is of utmost importance to elucidate the regulatory mechanisms responsible for Hsp
expression.
The inducible Hsp expression is regulated by the
heat shock transcription factors (HSFs). In response to
various inducers such as elevated temperatures, oxidants, heavy metals, and bacterial and viral infections,
most HSFs acquire DNA binding activity to the heat
shock element (HSE), thereby mediating transcription
of the heat shock genes, which results in accumulation
of Hsps (for reviews, see refs 9 –11). Since the isolation
of a single HSF gene from Saccharomyces cerevisiae (12,
13) and Drosophila melanogaster (14), several members of
the HSF family have been found in vertebrates
(HSF1– 4) and plants (Fig. 1) (for reviews, see refs 9,
11, 26, 27). The existence of multiple HSFs in vertebrates and plants suggests that different HSFs mediate
the responses to various forms of physiological and
environmental stimuli (for a review, see ref 28).
On the basis of early studies, it has been a general
assumption that HSF1, the functional vertebrate homo1
Present address: The Ruttenberg Cancer Center, Mount
Sinai School of Medicine, New York, NY USA
2
Correspondence: Turku Centre for Biotechnology, Tykistokaty 6B, 20521 Turku, Finland. E-mail lea.sistonen@btk.
utu.fi
0892-6638/01/0015-1118 © FASEB
Figure 1. The HSF family: comparison of
structural domains. The conserved domains
of the distinct HSFs—i.e., the DNA binding
domain (DBD), the oligomerization domain
(HR-A/B), the carboxyl-terminal heptad repeat (HR-C), and the isoform-specific regions—are indicated. The mammalian
HSF1, 2, and 4 have been isolated from
human (h) and mouse (m) (15–21; P. Nykänen et al., unpublished data). The human
HSF1-␤ isoform has not been characterized;
the domain structure is illustrated based on
the predicted homology to the mouse counterpart. Lp-HsfA1, A2, and B1 are cloned
from tomato (22) and are presented as an
example of the existence of the interspecies
conserved domains in plant HSFs. HSFs isolated from chicken (23), zebrafish (24), X.
laevis (25), D. melanogaster (14), and S. cerevisiae (12, 13) are termed cHSF1–3, zHSF1a, b,
XHSF, DrHSF, and ScHSF, respectively.
logue of the HSF found in yeast and the fly, is activated
by diverse forms of stress. Subsequently, HSF3, a unique
avian HSF, has been shown to function as a heatresponsive transcription factor. In contrast to HSF1 and
HSF3, HSF2 is not activated in response to classical
stress stimuli, but under developmentally related conditions. In support of the original findings, neither
HSF2 nor any other HSF is able to functionally substitute for HSF1 or to rescue the heat shock response in
HSF1 knockout mice or in cells derived from these
animals (29 –31). However, the roles of distinct HSFs
have been proposed to overlap depending on stimulatory signals. For example, human HSF2 and HSF4, but
not HSF1, are capable of complementing the viability
defect and conferring thermotolerance in S. cerevisiae
cells carrying a lethal HSF deletion (21, 32). The
differential activities of HSFs do not exclude the possiDIFFERENTIAL FUNCTIONS OF HSFs
bility that different members of the HSF family could
also cooperate in order to regulate expression of their
target genes, as has been demonstrated for avian HSF1
and HSF3 (33). This review focuses on the current
knowledge on transcriptional regulation of the heat
shock response, with the emphasis on novel discoveries
on the differential functions of the HSF family members.
STRUCTURAL AND FUNCTIONAL FEATURES
OF HSFs
The importance of HSFs as regulators of the heat shock
response is reflected by their high cross-species conservation in evolution. The HSFs are composed of func1119
Figure 2. A) An alignment of predicted amino acid sequences of the mammalian and chicken
HSFs. The boxed green region corresponds to the DBD, the light blue to the HR-A/B, and the
dark blue to the HR-C, which is absent in HSF4. The isoform-specific regions are marked in orange.
The alignment was carried out using MALIGN (34, 35).
(continued on next page)
tional modules that are indicative of the complex
regulation of these transcription factors, both at the
intra- and intermolecular levels. The amino-terminal
1120
Vol. 15
May 2001
helix-turn-helix DNA binding domain (DBD) is the
most conserved functional domain of HSFs (Figs. 1 and
2). In fact, DBD is the only domain where comparative
The FASEB Journal
PIRKKALA ET AL.
Figure 2. (continued ) B) Species-specific relationship of HSFs
among higher eukaryotes. The amino acid sequences were
aligned and analyzed as in panel A. The values next to the
branches indicate the percent identities between different
HSFs. C) The DBD is the most conserved region of mammalian and avian HSFs. Total and domain-specific percentage
amino acid identities of human, mouse, and chicken HSFs are
presented in comparison with hHSF1.
structural data is available, since the crystal structure of
Kluyveromyces lactis and the solution structures of K.
lactis, D. melanogaster, and tomato HSF DBDs have been
determined (36 –39). Activation-induced trimerization
of HSFs is mediated by three arrays of hydrophobic
heptad repeats (HR-A/B) characteristic for helical
coiled-coil structures, i.e., leucine zippers (Fig. 1 and
Fig. 3) (41). The trimeric assembly of HSFs is unusual,
as leucine zipper proteins typically are known to asso-
Figure 3. Functional domains of HSF1 and HSF2. The DNA
binding domain (DBD) and the heptad repeats (HR-A/B, C)
are marked as in Fig. 1. The regulatory domain (RD) of HSF1,
containing sites for both constitutive and inducible serine
phosphorylation, regulates the carboxyl-terminally located
transcriptional activation domains (AD1, 2). HSF2 possesses
one relatively strong activation domain at the carboxyl terminus, which is negatively regulated by an adjacent domain
(40). Additional weak regulatory and activation domains have
been found in the internal region of HSF2. See text for more
details.
DIFFERENTIAL FUNCTIONS OF HSFs
ciate as homo- or heterodimers. The suppression of
HSF trimerization is likely to be mediated by another
region of hydrophobic heptad repeats (HR-C) adjacent
to the carboxyl terminus of the protein (42). The HR-C
is well conserved among the vertebrate HSFs but poorly
conserved in plant and S. cerevisiae HSFs, which may be
a reason underlying constitutive trimerization of HSF
in S. cerevisiae (43, 44; reviewed in ref 9).
Alternative splicing leads to the generation of functionally distinct HSF isoforms, thereby providing an
additional regulatory level to control gene expression.
As illustrated in Fig. 1, two isoforms of mammalian
HSF1 and HSF2, generated by alternative splicing of
exon 11 adjacent to the HR-C, have been found (18, 19,
45, 46). Alternative splicing appears to be a common
feature of several HSFs, as HSF4 also exists as two
differentially generated isoforms (21). Although the
regulatory mechanisms of formation of the isoforms are
not known, tissue-specific expression has been reported
(18, 19, 45). The most recent example of regulation via
alternative splicing is zebrafish HSF1 (zHSF1), which is
expressed as two isoforms in a tissue- and temperaturedependent manner (24).
HSFs AND THE HEAT SHOCK RESPONSE
HSF1 as a prototype of heat shock transcription
factors
In vertebrates, HSF1 has been identified as the HSF
that mediates stress-induced heat shock gene expression on the basis of the ability of HSF1 to display
1121
inducible DNA binding activity, oligomerization, and
nuclear localization in response to environmental stressors, such as elevated temperatures and cadmium sulfate, or exposures to the amino acid analog L-azetidine2-carboxylic acid or proteasome inhibitors (23, 31,
47–50). HSF1 is also responsible for mediating the
expression of Hsp70 in response to whole body stress in
rodents, highlighting the central role for HSF1 even
during physiological stress (reviewed in ref 51).
Oligomeric status of HSF1
In contrast to HSF in budding yeast, D. melanogaster HSF
and HSF1 of higher eukaryotes exist as monomers in
unstressed cells. It has been of great interest to study
how the monomeric status of HSF1 is retained under
nonstressful conditions. Studies showing that HSF1
produced in bacteria and overexpressed in mammalian
cells acquires DNA binding activity spontaneously in
the absence of heat shock (15, 48, 52, 53) led to the
conclusion that the intrinsic activity of endogenous
HSF1 is negatively regulated by a titratable cellular
factor in mammalian cells. However, when recombinant or overexpressed mammalian HSF1 was expressed
at lower concentrations, the DNA binding activity did
not occur spontaneously but was shown to be regulated
by heat (42, 45, 54, 55). Therefore, oligomerization of
the factor could be repressed in the monomer, possibly
by intramolecular interactions between the HR domains (Fig. 3) (42, 54). This concept is further supported by studies showing direct activation of purified
fruit fly HSF by heat and oxidative stress (56). Other
inducers of the heat shock response have no effect on
activation of D. melanogaster HSF in vitro, implying that
except for heat and oxidation, heat shock inducers
exert their effects indirectly (56).
On the basis of the experiments on negative regulation of HSF1 activity, studies intended to identify HSF1
interacting proteins were initiated. In fact, evidence for
the regulatory role for Hsp70 in HSF1 deactivation had
already been provided in earlier studies using a variety
of cell models and experimental strategies (for a review
and original references, see ref 10). In addition to
Hsp70, other molecular chaperones have been indicated to regulate the activity of HSF1. For example,
reduced levels of Hsp90 have been demonstrated to
lead to HSF1 activation in vitro (57). In vivo, geldanamycin that specifically binds and inhibits Hsp90, activates HSF1, indicating that Hsp90-containing HSF1
complex is present in the unstressed cells, but dissociates on stress (57). This provides evidence that Hsp90
by itself or in context with multichaperone complexes
would be an important repressor of HSF1. Similar kinds
of results have been obtained in the oocytes of Xenopus
laevis, where formation of a heterocomplex between
components of the Hsp90 chaperone machinery and
HSF1 has been suggested to play a key role in modulating different steps of the HSF1 activation– deactivation pathway (58, 59). Repression of HSF1 activity by
interaction with various Hsps is essential, especially
1122
Vol. 15
May 2001
considering the need for strictly regulated chaperone
levels in maintaining the cellular homeostasis.
Besides the feedback regulation of HSF1 by Hsps,
other proteins have been found to bind HSF1. For
example, in a yeast two-hybrid screen, a novel 8.5 kDa
nuclear protein termed heat shock factor binding
protein 1 (HSBP1) was found to interact with the
oligomerization domain of an active HSF1, thereby
negatively affecting HSF1 DNA binding activity (60).
During inactivation of HSF1 to an inert monomer,
HSBP1 is associated with Hsp70. In Caenorhabditis elegans, overexpression of the homologous CeHSB-1 results in reduced ability of the organism to cope with
thermal and chemical stresses, whereas loss of CeHSB-1
leads to similar or slightly better survival rates than in
wild-type animals (60). The association of HSBP1 with
HSF1 is interesting with regard to the mechanisms
allowing constitutive DNA binding activity of S. cerevisiae
HSF, since HSBP1 has been found in every organism
studied so far except in S. cerevisiae (60; L.-J. Tai, S.
McFall, and R. I. Morimoto, personal communication).
All of the HSF1 interacting proteins reported to date
act as negative regulators of HSF1 activity, suggesting
that a multicomplex might be required to keep HSF1 in
a state that can readily be activated. It is of interest to
learn whether there are any positive regulators of HSF1
or whether fast activation of HSF1 is an intrinsic
property that has been evolutionary conserved to allow
a rapid autoactivation of HSF1 upon stress stimuli.
HSF1 regulation is linked to cellular signaling pathways
The strict regulation of HSF1 is further emphasized by
the fact that its DNA binding and transactivating capacities are uncoupled. This is exemplified in S. cerevisiae,
where HSF binds DNA constitutively (12, 43, 61) but
fails to activate transcription without further stimuli.
Treatment of mammalian cells with sodium salicylate
and other nonsteroidal anti-inflammatory drugs induces formation of the HSF1 DNA binding trimer
without transcriptional activity (62, 63), demonstrating
that in mammalian cells the DNA binding and transcriptional activities of HSF1 are also uncoupled. In
mammalian model systems, phosphorylation is an important determinant of the transactivating potency of
HSF1. The salicylate-induced HSF1 is constitutively but
not inducibly phosphorylated on serine residues,
whereas the heat-induced HSF1 is both constitutively
and inducibly phosphorylated (64). Further, the druginduced, transcriptionally inert intermediate can be
converted to a transcriptionally active state by a subsequent exposure to heat shock, during which the only
detectable change in HSF1 is its hyperphosphorylation
(64). These results have formed a basis for further
studies to determine the phosphorylation sites and to
understand their roles in regulation of HSF1.
As illustrated in Fig. 3, HSF1 contains two distinct
carboxyl-terminal activation domains, AD1 and AD2,
which are under the control of a centrally located,
heat-responsive regulatory domain (RD) (65). Since
The FASEB Journal
PIRKKALA ET AL.
AD1 does not appear to be heat-regulated by itself, the
regulatory domain of HSF1 has been proposed to play
a key role in sensing heat stress in humans (66).
Constitutive phosphorylation of two specific serine/
proline motifs, S303 and S307, is important for the
function of the RD and may be critical for negative
regulation of HSF1 transcriptional activity at normal
temperatures, since substitution of serine to alanine
causes constitutive transcriptional competence (67–
69). Constitutive phosphorylation of S363 has also been
found to negatively regulate HSF1 under normal
growth conditions (70). The positive role of phosphorylation in regulation of HSF1 transcriptional activity is
only emerging. So far, one in vivo phosphorylated site
(S230), which has a positive effect on HSF1 transactivating capacity, has been characterized (C. I. Holmberg, personal communication). By phosphopeptide
mapping of in vivo 32P-labeled HSF1 and by using
phosphopeptide-specific antibodies that recognize the
phosphorylated form of S230, it has been demonstrated
that phosphorylation of S230 is enhanced upon heat
shock, and, consequently, mutation of S230 to alanine
correlates with decreased transcriptional activity of the
human HSF1.
To elucidate the mechanisms by which the cell senses
stress, it is important to analyze the cellular signaling
pathways leading to the stress response. On the basis of
in vitro analyses and overexpression studies, S303 and
S307 have been suggested to be targeted by the glycogen synthase kinase 3␤ (GSK-3␤) and the extracellular
signal-regulated kinase, whereas S363 might be a site
for phosphorylation by the c-Jun N-terminal kinase
(JNK) (67, 68, 70 –73). The S230, in turn, seems to be
a good substrate for calcium/calmodulin-dependent
protein kinase II (CaMK II) (C. I. Holmberg, personal
communication). The majority of the critical phosphorylation sites and kinases/phosphatases responsible for
HSF1 phosphorylation still need to be elucidated.
Stress induces formation of nuclear HSF1 granules
Much attention has recently been paid to specific
structures localized in the nucleus, collectively called
nuclear bodies, which most likely serve as organizational centers to control the activities of various transcriptional regulators. It had already been observed by
1993 that during heat shock, HSF1 localizes into the
nucleus, forming granule-like structures in human but
not in murine cells (48). HSF1 granules are also
formed upon other stresses, such as proteasome inhibition, and exposure to heavy metals and the amino
acid analog azetidine, but not by treatment with the
anti-inflammatory agent sodium salicylate, which does
not induce hyperphosphorylation or transcriptional
activity of HSF1 (74 –77). This suggests that depending
on the stress stimuli, the various events associated with
HSF1 activation are differentially affected.
The formation of granules seems to be an intrinsic
property of human cells and not that of HSF1, since
nuclear HSF1 granules can be detected upon expresDIFFERENTIAL FUNCTIONS OF HSFs
sion of exogenous mouse HSF1 in human cells (74).
The appearance of HSF1 granules parallels the activation of HSF1 and the transient induction of heat shock
gene transcription (74, 77). During recovery from heat
shock, HSF1 granules are no longer detected, but HSF1
rapidly relocalizes to the same structures upon subsequent reexposure to heat (76). A target for the HSF1
foci has recently been shown to be a centromeric
region on human chromosome 9 rich in repetitive
sequences (C. Jolly and R. I. Morimoto, personal communication). Although the structure and function of
granules are still unclear, as they do not colocalize with
previously described nuclear structures (74), it has
been proposed that granules could be sites where
activated HSF1 is stored and recycled, thereby coordinating the regulation of heat shock gene expression.
Alternatively, granules might represent sites where
HSF1 has activities distinct from transcriptional regulation, such as a structural role in protecting hypersensitive sites of the genome (76).
HSF3, an avian-specific HSF?
In avian cells, a novel heat shock factor, HSF3, has been
cloned in addition to the homologs of mammalian
HSF1 and HSF2 (23). Like HSF1, HSF3 is a stressresponsive transcription factor. Avian HSF2 and HSF3
are expressed at comparable levels among various
tissues, whereas the amount of HSF1 varies greatly in
distinct tissues (78; reviewed in ref 27). Upon activation, a nuclear localization signal is required for oligomerization of an inactive HSF3 dimer to a nuclear
trimer that has properties of a transactivator (79, 80).
The threshold temperatures required to activate HSF1
and HSF3 are different, as only HSF1 is activated upon
mild heat shock, in contrast to coactivation of HSF1
and HSF3 after severe heat shock (78, 81). Further, the
amount of HSF3 increases after a severe heat shock,
whereas HSF1 protein levels diminish (81), suggesting
that HSF3 has a role during severe and persistent stress
in avian cells. Despite efforts, no HSF3 homologue has
been found in other than avian cells, raising the
possibility that the mechanisms used to cope with stress
might be organism specific.
A genetic approach has unraveled an interesting
interdependent relationship between chicken HSF1
and HSF3: disruption of the HSF3 gene results in
impaired heat shock response and loss of thermotolerance in chicken B lymphocyte DT40 cells expressing
normal levels of HSF1 (33). In the absence of HSF3,
HSF1 does not trimerize upon heat shock, even at
milder temperatures, indicating that HSF3 directly
influences HSF1 activity and has a dominant role in the
regulation of heat shock response (33). In DT40 cells
harboring a double knockout of HSF1 and HSF3, the
basal expression of hsp90␣ mRNA is markedly decreased, in contrast to the unaffected basal expression
levels of other hsp mRNAs (A. Nakai, personal communication). Therefore, the constitutive expression of
1123
Hsp90␣ has been proposed to determine the capacity
of stress resistance of vertebrate HSFs.
HSF4: a novel family member to be characterized
Expression of the most recently discovered mammalian
HSF, HSF4, is restricted to certain tissues, as shown by
RT-PCR analysis (20, 21). Similar to HSF1 and HSF2,
HSF4 has two isoforms: HSF4a and HSF4b (Fig. 1). The
splicing of HSF4 mRNA is complex, as two alternative
5⬘ splice sites have been found in exons 8 and 9 with a
frameshift in the transcript (21). The original cDNA
clone encodes HSF4a, which has been reported to be
transcriptionally inactive and to function as a repressor
of heat shock gene expression (20). In contrast, the
more recently reported HSF4b isoform has the potential for transactivating heat shock genes and substituting for the yeast HSF (21). HSF4 has one unique
structural feature: both isoforms lack the HR-C necessary for suppression of HSF trimer formation (Figs. 1
and 2), leading to constitutive DNA binding activity of
HSF4 in vitro (21). However, no constitutive binding of
HSF4 to the HSEs in the human hsp70 promoter has
been detected by in vivo genomic footprinting of some
human cell lines (82, 83). To establish whether HSF4 is
a stress-responsive factor will require a more elaborate
analysis of HSF4 functions in various cell types.
FUNCTIONS OF HSFs EXPAND BEYOND THE
HEAT SHOCK RESPONSE
Heat shock gene expression is crucial for survival of
cells exposed to extracellular stress stimuli, but also
during normal cellular physiology. Since the discovery
of a family of HSFs, there has been speculation on their
functions unrelated to the stress response. The nonstress functions include, for example, roles during cell
cycle, embryonic development, cellular differentiation,
and spermatogenesis. Understanding of the functions
of the distinct HSFs during developmental and differentiation-related states is only emerging. Here, we have
summarized the current knowledge on HSF functions
beyond the classical stress response.
HSF2, a nonstress-responsive member of the HSF
family
The expression of Hsp70, induced during hemin-mediated erythroid maturation of the human K562 erythroleukemia cells (84), was shown in 1989 to be mediated through the HSE, suggesting that HSF-regulated
transcription may also play a role during nonstress
conditions (85). In 1991, the cDNAs for human and
murine HSF2 were isolated simultaneously with the
HSF1 counterparts (15–17). Subsequently, it has been
well established that the hsp70 transcription induced by
hemin in K562 cells is mediated predominantly by
HSF2 (reviewed in refs 10, 28).
The incapability of HSF2 to respond to classical stress
1124
Vol. 15
May 2001
stimuli and the importance of HSF2 in the heminmediated erythroid differentiation pathway of K562
cells suggest a role for HSF2 in controlling development and differentiation-specific gene expression. In
addition to the up-regulation of HSF2 expression upon
activation (46, 86), its diminished expression appears
to be an important control mechanism during cellular
differentiation. This has been exemplified by the rapid
down-regulation of HSF2 during TPA-mediated megakaryocytic differentiation (87). The opposite HSF2
expression patterns in K562 cells differentiating along
either the erythroid or megakaryocytic lineages are
intriguing, since constitutive HSF2 expression has been
observed in various cell types and tissues (16, 18).
However, the possible role of HSF2 in the distinct
differentiation pathways of progenitor cells remains
remote until alternative model systems to K562 cells
become available.
HSF2 has also been proposed to be responsible for
the specific Hsp expression observed during developmental processes, and the constitutive HSF2 HSE binding activity during early embryogenesis as well as in
mouse ES and carcinoma cells has been studied extensively (88 –91). HSF2 binds to DNA and is abundantly
expressed in the developing mouse heart at E11.5–12.5
(92). In the central nervous system, HSF2 HSE binding
activity has been shown to remain high until E15.5 (93).
In addition, high levels of HSF2 have been detected in
mouse testis, and the expression of HSF2 mRNA has
been shown to be regulated with respect to the developmental stages of mouse and rat spermatogenesis (94,
95). During mouse embryogenesis and rat spermatogenesis, the pattern of Hsp expression does not correlate with HSF2 DNA binding activity (92, 93, 95),
whereas opposite results have been obtained during
mouse spermatogenesis (94). Some of the discrepancies could be explained by the finding that the ratio of
the alternatively spliced HSF2 isoforms (Fig. 1) varies
significantly between different tissues and cell types
(18, 19). In adult mouse testis, the transcriptionally
more active HSF2-␣ isoform is predominantly expressed, whereas the HSF2-␤ isoform is more abundant
during mouse embryonic development and rat spermatogenesis (93–95). It is plausible that HSF2-␤ may
not be transcriptionally active enough to regulate heat
shock gene expression. This isoform might also activate
transcription of genes different from heat shock genes
or interact with other transcription factors to either
antagonize or stimulate their transcriptional properties
(93). The relevance of HSF2 in development and
spermatogenesis is likely to be elucidated when HSF2
null mice will be available.
Regulation of HSF2 expression is complex and involves
several regulatory levels
HSF2 displays only a 35% overall identity to HSF1, but
the DNA binding and oligomerization domains are
highly conserved (Fig. 2). Cloning of the genomic
structure of mouse HSF2 has revealed that the gene is
The FASEB Journal
PIRKKALA ET AL.
located on chromosome 10 and is composed of 13
exons spanning at least 43 kbp in the genome (96). The
functional domains of mouse HSF2 appear to be related to defined groups of exons: the DNA binding
domain is comprised of exons 1, 2, and 3, the oligomerization domain (HR-A/B) of exons 4, 5, and 6, and the
HR-C of exons 10 and 11 (96). The human HSF2 gene
shows similar overall structure and is located on chromosome 6 (P. Nykänen et al., unpublished results). The
cytoplasmic localization of inactive HSF2 has been
implicated to require interaction between the HR-A/B
and HR-C domains (97). This interaction could mask
the nuclear localization signals present on both sides of
the HR-A/B domain, which upon induction would be
exposed by disruption of the intramolecular interaction (97).
Regulation of HSF2 expression uses both transcriptional and post-transcriptional control mechanisms, as
increased HSF2 transcription accompanied by mRNA
stabilization has been shown to precede the hemininduced HSF2 protein accumulation in K562 cells
undergoing erythroid differentiation (86, 87). There is
some evidence that HSF2 does not, at least at the
transcriptional level, regulate its own expression, as no
HSE has been found in the mouse HSF2 promoter
(96). However, this does not exclude the existence of
regulatory elements in the context of a more complete
promoter, since only ⬃400 bp of the promoter sequence has been analyzed. By using the protein translation inhibitor cycloheximide, it has been found that
the stabilization of HSF2 mRNA by hemin does not
involve novel synthesis of an HSF2 mRNA binding
protein (L. Pirkkala and L. Sistonen, unpublished
results). Closer characterization of the HSF2 promoter
and regulatory factors as well as identification of the
HSF2 mRNA stability determinants are obvious subjects
for future studies.
HSF2 itself is a labile protein, which is targeted for
degradation by the 26S proteasome by a covalent
attachment of multiubiquitin (L. Pirkkala et al., unpublished results). (98). This suggests that the heminmediated increase in HSF2 protein could be due to a
stabilizing effect of hemin on some yet unidentified
HSF2 interacting protein(s). Several candidates for
HSF2 interacting proteins have been found using yeast
two-hybrid screens, including nucleoporin p62 (99)
and the testis-specific protein HSF2BP (100). Yoshima
and co-workers (99) have proposed that HSF2 would
compete with importin-␤ for binding to nucleoporin
p62; consequently, the lack of p62 in testis would
facilitate the nuclear localization of HSF2 and transactivation of heat shock genes observed in mouse testis
(94). HSF2 also interacts with the PR65 subunit of
protein phosphatase 2A (PP2A) (101). A region in
PR65 previously shown to be involved in interacting
with the catalytic subunit of PP2A is required for
binding to HSF2 (102). This suggests that HSF2 would
directly compete with the catalytic subunit for binding
to PR65, thereby modulating PP2A activity. Regulation
of PP2A activity by HSF2 could thus provide a mechaDIFFERENTIAL FUNCTIONS OF HSFs
nism for cross talk between heat shock gene expression
and PP2A-regulated pathways in the cell (101, 102),
suggesting a nontranscriptional function for HSF2. The
proposed specialized functions of HSF2 might depend
on HSF2 interacting proteins, but so far no compelling
functional evidence for the biological roles of HSF2
partner proteins has been obtained.
The 18 amino acid sequence present in the HSF2-␣
isoform, but missing from the HSF2-␤ isoform, has
been suggested to confer an increase in transcriptional
activity of HSF2. Using luciferase as a reporter gene
under control of the hsp70 promoter has demonstrated
that the mouse HSF2-␣ isoform is a more potent
transcriptional activator than the HSF2-␤ (19). Stable
overexpression of either mouse HSF2-␣ or HSF2-␤
isoform in human K562 cells has revealed that the two
isoforms indeed have distinct functions in vivo: overexpression of HSF2-␤ inhibits both hemin-induced hsp70
and hsp90 transcription and erythroid differentiation
(46). Therefore, HSF2-␣ acts as a potential activator
and HSF2-␤ as a suppressor of the hemin-induced gene
expression, and the isoform ratio might be critical for
the human hematopoietic differentiation process.
Moreover, HSF2-␤ negatively controls the amount of
HSF2 protein accumulated in a cell (46). This type of
regulation could be a relevant mechanism to attenuate
the erythroid differentiation process and induction of
heat shock gene expression. However, the mechanisms
behind the differential regulation and function of the
two isoforms are still obscure.
Differential effects of HSF2 and HSF1 on hsp70 were
ultimately suggested to be due to distinct DNA binding
patterns at the human hsp70 promoter in vivo on
hemin and heat treatments, respectively (83), and to
their slightly different binding site specificities (103,
104). This has been further supported by the observation that synergistic activation of HSF1 and HSF2
causes an enhanced hsp70 transcription, possibly by
forming heterotrimers, which correlates with an alteration in the DNA binding pattern in vivo (86). These
findings also suggested that HSF1 and HSF2 might
regulate the expression of distinct genes. For example,
during erythroid differentiation of K562 cells, HSF2
might be a strong candidate, either alone or in combination with other transcription factors, to regulate the
expression of erythroid-specific genes (87). One potential HSF2 target gene is thioredoxin (TRX), the first
non-heat shock gene the expression of which is regulated in an HSF2-dependent manner (105). More studies are required to determine the role of HSF2 in
regulation of TRX expression.
Knockout and transgenic organisms reveal
unexpected functions for HSF1
The yeast homologue of vertebrate HSF1 is essential for
cell growth and viability (12, 13, 106), and this requirement has been suggested to involve a role in the
regulation of basal heat shock gene expression. Similar
to the budding yeast HSF, fruit fly HSF is essential for
1125
the heat shock response in vivo but, unlike yeast HSF, is
dispensable for growth and viability under nonstressful
conditions in adult flies (107). The fruit fly HSF is
required during oogenesis and early larval development, as shown by analyzing phenotypes harboring
mutant alleles of the D. melanogaster HSF gene. However, these two genetically independent functions are
not mediated through Hsp induction (107).
The Hsps have vital functions during embryogenesis
and in protecting embryos against the effects of various
toxic agents. Spontaneous and inducible Hsp expression has been demonstrated in mouse early embryonic
cells and embryonal carcinoma cells (108 –110). HSF1
and HSF2 are specifically expressed at different stages
of embryonic development, as HSF1 is already abundant in oocytes, whereas HSF2 is undetectable in oocytes and its expression increases during the preimplantation period (111). At morula and blastocyst stages,
the amounts of HSF1 and HSF2 become comparable.
Until the recently reported HSF1 knockout mouse
(30), no evidence of the in vivo functions of mammalian HSFs during development had been provided.
HSF1-deficient mice display severe defects in the chorioallantoic placenta, resulting in increased prenatal
lethality. Still, a small number of mice lacking HSF1
survive, demonstrating that HSF1 is not obligatory for
development (30). After birth, growth of the HSF1
knockout mice is retarded and the females are sterile
(30). The requirement of HSF1 as a maternal factor
during early cleavage of mammalian development has
recently been demonstrated elegantly in HSF1 knockout embryos (112). The HSF1 null oocytes exhibit
normal morphological appearance, but the development of fertilized embryos derived from the knockout
oocytes is arrested during preimplantation before the
blastocyst stage. Fertilization of HSF1-deficient oocytes
with paternal wild-type HSF1 fails to release the blockage, indicating the essential requirement for maternal
HSF1 in early postfertilization development (112).
However, the HSF1 knockout zygotes spontaneously
express the Hsps, showing that zygotic transcriptional
activity can begin without HSF1 (112).
HSF1 null mice generated using a heterozygous
mother and a homozygous HSF1-deficient father neither exhibit classical heat shock response nor acquire
thermotolerance, and the HSF1-deficient embryonic
fibroblasts exposed to heat stress die via apoptosis (29,
30). In the absence of HSF1, exaggerated tumor necrosis factor ␣ (TNF-␣) production and increased mortality after endotoxin and inflammatory challenge have
been observed (30). In an independent study, inhibition of TNF-␣ transcription by febrile range temperatures was shown to result from partial activation and
binding of HSF1 to the proximal promoter or 5⬘untranslated region of TNF-␣ (113). This finding raises
the interesting possibility of HSF1 acting as a repressor
of certain genes, already proposed for prointerleukin
1␤ and c-fos genes (114, 115).
Generation of transgenic mice expressing constitutively active HSF1 in the testis has revealed an unex1126
Vol. 15
May 2001
pected phenotype: the constitutive expression of an
active form of HSF1 results in a blockage of spermatogenesis at the pachytene stage, accompanied by an
increased number of apoptotic spermatocytes (116).
These results are the opposite of previous studies in
mouse and fruit fly that indicate the requirement of
HSF1 for acquisition of thermotolerance by inducing
Hsps and supporting cell survival against thermal stress
(29, 107). The mechanism responsible for active HSF1
being sufficient to induce apoptosis of late pachytene
spermatocytes is not known, but it has been hypothesized that HSF1 may induce or repress target genes
involved in germ cell development or apoptosis (116).
Hsp70 –2 is normally abundantly expressed at the
pachytene stage, but in mice deficient in Hsp70 –2,
spermatogenesis is blocked at this stage and the males
are infertile (117). Therefore, it is tempting to speculate that HSF1 might repress hsp70 –2 expression in the
pachytene spermatocytes. The spectrum of HSF1 functions has thus become more diverse, including a requirement for developmental processes and postnatal
growth in addition to regulation of the stress response
under adverse conditions.
HSF1 plays an essential role in the ubiquitin
proteolytic pathway
Ubiquitin–proteasome-mediated proteolysis is an essential pathway for protein degradation. The selective
destruction of many short-lived regulatory proteins as
well as damaged proteins is mediated by the 26S
proteasome (for a review, see ref 118). When the
ubiquitin–proteasome network is down-regulated, certain heat shock proteins such as Hsp70, among other
molecular chaperones, are induced (49, 50, 98, 119,
120; reviewed in ref 118). It is therefore of interest to
understand the functions of the distinct HSFs and their
roles in regulating Hsp expression during ubiquitin–
proteasome-mediated degradation. To resolve the contradictory results from various laboratories concerning
activation of the different members of the HSF family
(HSF1–3) on proteasome inhibition (49, 50, 98), the
specific roles of HSF1 and HSF2 in the regulation of the
ubiquitin–proteasome network have recently been investigated. By using different strategies to abrogate
HSF1 and HSF2 activities, the key regulatory role of
HSF1 in the ubiquitin proteolytic pathway was established in HSF1 null cells by the capability of exogenous
HSF1 to restore the HSF DNA binding activity and
inducible Hsp70 expression upon proteasome inhibition (31). Despite a prominent increase in HSF2 protein, the transactivating capacity of this factor could not
be detected.
It has been speculated that the distinct HSFs could
have cooperative functions. The finding that other
HSFs cannot functionally substitute for HSF1 when the
ubiquitin–proteasome-mediated degradation is inhibited indicates that not all functions of the distinct HSFs
are overlapping. This observation is further supported
by studies of HSF1 knockout models (29, 30). The
The FASEB Journal
PIRKKALA ET AL.
activation of HSF1 upon disturbances in the degradation pathway leading to enhanced expression of Hsp70
and other molecular chaperones is an interesting subject for future studies as to the importance of proteasome-mediated degradation of a multitude of proteins
involved in progression of cell cycle and tumorigenesis.
Stress-independent functions of HSF3 in the cell cycle
An important role for HSFs unrelated to the heat shock
response originates from studies demonstrating HSF3
activation in unstressed proliferating cells by direct
binding to c-Myb, a transcription factor involved in
hematopoiesis and cell proliferation, through their
respective DNA binding domains (121). Since the
c-Myb protein is highly expressed and required for the
G1-to-S transition of the cell cycle simultaneously with
expression of Hsp70, it is plausible that the c-Mybinduced activation of HSF3 may contribute to the cell
cycle-dependent expression of heat shock genes (121).
Recently, the c-Myb-HSF3 interaction has been found
to be disrupted by direct binding of the p53 tumor
suppressor to HSF3, leading to proteasome-dependent
degradation of c-Myb and down-regulation of hsp70
and hsp90␣ expression (122). Mutated forms of p53
found in certain tumors are not able to inhibit c-Mybinduced HSF3 activation, suggesting an HSF3-mediated
interplay between the c-Myb proto-oncogene and the
p53 tumor suppressor in regulation of the cell cycle and
apoptosis (122).
CONCLUSIONS AND PERSPECTIVES FOR
FUTURE RESEARCH
In only 10 years since cloning of the first vertebrate
HSFs, it has become apparent how wide a variety of
cellular responses the distinct HSFs can mediate, despite their striking structural similarities. The current
knowledge on the differential functions of HSFs presented in this review is summarized and illustrated in a
simplified form in Fig. 4. According to the model, the
various exogenous and endogenous stimuli activate
distinct members of the HSF family, in most cases
leading to induction of heat shock gene expression. In
contrast to the generally held notion, HSF1 appears not
only to be a stress factor, but is also involved in certain
developmental and differentiation-related processes.
The functions of HSF1 not associated with stress have
only been established after the generation of HSF1deficient cell and animal models. Therefore, to elucidate the specific role of HSF2, gene disruption experiments should be most revealing. Moreover, generation
of transgenic and double knockout animals and the cell
lines derived from them should further clarify the
potential overlapping functions of HSFs. Functional
cooperativity between the avian HSF1 and HSF3 has
already been described. Unraveling the mechanistic
basis of this cooperation will undoubtedly facilitate
DIFFERENTIAL FUNCTIONS OF HSFs
Figure 4. Model of differential HSF functions. Environmental
stressors and dysfunctions in the ubiquitin–proteasome pathway induce accumulation of aberrant and short-lived proteins, which in turn leads to dissociation of HSF1/Hsp
complexes and activation of HSF1. Activation of the heat
shock response ultimately leads to synthesis of Hsps, thereby
forming a negative feedback loop for attenuation of the stress
response. Note that the details of the heat-induced pathway,
which culminates in activation of HSF3, are still unclear, as
are the mechanisms by which avian HSF1 and HSF3 cooperate under severe stress. Beyond the heat shock response,
HSF3 cooperates with c-Myb to ensure an adequate supply for
Hsps, an enhanced need for which is observed during the cell
cycle progression. Binding of p53 to HSF3 disrupts the
HSF3-c-Myb interaction leading to ubiquitin-mediated degradation of c-Myb and down-regulation of Hsp expression.
HSF2 is activated under certain developmental conditions
and upon erythroid differentiation of K562 cells. During
megakaryocytic differentiation, HSF2 expression is downregulated in K562 cells. The HSF2-␤ isoform is able to inhibit
erythroid differentiation and accumulation of HSF2, as well
as transcriptional activation of the HSF2 targets, i.e., the heat
shock genes. Moreover, both HSF1 and HSF2 are likely to
have additional target genes. The stimuli responsible for
activation of HSF4 are uncharacterized, but the in vitro
evidence points to opposite roles for the two HSF4 isoforms in
regulation of the heat shock genes.
understanding of the interdependent relationship between distinct HSFs. This is particularly important,
since the incapability of HSFs to functionally substitute
for each other might be unique for mammalian cells: in
mutated yeast cells carrying a lethal HSF deletion, both
human HSF2 and HSF4, but not HSF1, can function as
a stress-responsive factor and replace the yeast HSF,
thereby rescuing the viability defect.
1127
Considering the multiple HSF isoforms yielding increased complexity in their regulatory functions, characterization of the mechanisms responsible for expression and regulation of the distinct isoforms is of great
importance. For example, existence of HSF1 isoforms
has been reported, but whether they possess differential features, similar to HSF2 and HSF4, has remained
unexplored. The avian HSF3 might be a unique member of the HSF family, because only a single mRNA
instead of multiple alternatively spliced isoforms, has
been found to date. Chromosomal localization of HSFs
is likely to shed light on the HSF locus and the
neighboring genes. By the aid of large-scale gene
analysis using microarray techniques, there remains the
exciting possibility of finding new, unexpected HSF
target genes.
With regard to the conservation of the heat shock
response in evolution, it is of interest to note that at
least two species lacking an inducible heat shock response have been found. The absence of heat shock
response in a highly cold-adapted, stenothermal Antarctic teleost fish (Trematomus bernacchii) and a freshwater cnidarian (Hydra oligactis) might be due to evolutionary adaptation to stable subzero temperatures
without a need to cope with high temperatures (123,
124). The lack of inducible stress response could be
due to elimination of the heat shock genes or some
components in the gene regulatory pathway during
evolution. In the case of T. bernacchii, the mechanism
remains to be elucidated, but in H. oligactis, the hsp70
mRNA has been shown to be highly unstable during
heat shock (125).
From the evolutionary point of view, phylogenetic
analysis has unexpectedly revealed that the zebrafish
HSF1 is more homologous to other vertebrate HSF1s
than to the blue gill sunfish HSF (24). This indicates
that fish may have several distinct HSFs or that different
fish species have unusually divergent HSFs. As the
multiple HSFs in higher organisms are supposed to
differentially regulate gene expression in response to
distinct signals, it is plausible that fishes have several
HSFs, considering the need of many fish species to
tolerate large fluctuations of temperature and other
physicochemical parameters in their natural environment. In further support of this hypothesis, the complexity of HSFs in plants also appears to be higher than
in other eukaryotic organisms; in the tomato, for
example, HSFs are encoded by a gene family with up to
five members (for a review, see ref 26). The organismspecific numbers of HSF family members most likely
will soon be established by analyzing the complete
genome sequences.
The ongoing efforts to identify HSF interacting proteins will certainly continue. Although transient protein–protein interactions have traditionally been studied in vitro, it will be most challenging to demonstrate
the in vivo significance of these interactions. Understanding the structural basis of phosphorylation-mediated regulation of HSF1 and other HSFs will provide
1128
Vol. 15
May 2001
insight into regulation of the stress-sensing mechanisms. Obviously, the ultimate challenge will be to
elaborate our detailed knowledge of the HSFs to understand the significance of this transcription factor
family in the adaptation to the diverse biological environments to which organisms are exposed.
We thank our colleagues for generously providing results
prior to publication. We regret that due to space limitations,
many original important studies have not been cited directly
but rather through the relevant reviews. We are grateful to
Yvonne Nymalm, Tiina Salminen, and Stefanie Tran for
excellent computing assistance. John E. Eriksson and members of the Sistonen laboratory, particularly Tero-Pekka
Alastalo, Carina Holmberg, and Marko Kallio, are acknowledged for critical reading and valuable comments on the
manuscript. Work in the authors’ laboratory is supported by
the Academy of Finland, the Sigrid Jusélius Foundation, the
Wihuri Foundation, and the Finnish Cancer Organizations.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Ritossa, F. (1962) A new puffing pattern induced by temperature shock and DNP in Drosophila. Experientia 18, 571–573
Lindquist, S., and Craig, E. A. (1988) The heat shock response.
Annu. Rev. Genet. 22, 631– 677
Hartl, F. U. (1996) Molecular chaperones in cellular protein
folding. Nature (London) 381, 571–580
Arrigo, A.-P. (1998) Small stress proteins: chaperones that act
as regulators of intracellular redox state and programmed cell
death. Biol. Chem. 379, 19 –26
Jäättelä, M. (1999) Escaping cell death: survival proteins in
cancer. Exp. Cell Res. 248, 30 – 43
Marber, M. S., Mestril, R., Chi, S.-H., Sayen, M. R., Yellon,
D. M., and Dillmann, W. H. (1995) Overexpression of the rat
inducible 70-kD heat stress protein in a transgenic mouse
increases the resistance of the heart to ischemic injury. J. Clin.
Invest. 95, 1446 –1456
Plumier, J.-C. L., Ross, B. M., Currie, R. W., Angelidis, C. E.,
Kazlaris, H., Kollias, G., and Pagoulatos, G. N. (1995) Transgenic mice expressing the human heat shock protein 70 have
improved post-ischemic myocardial recovery. J. Clin. Invest. 95,
1854 –1860
Radford, N. B., Fina, M., Benjamin, I. J., Moreadith, R. W.,
Graves, K. H., Zhao, P., Gavva, S., Wiethoff, A., Sherry, A. D.,
Malloy, C. R., and Williams, R. S. (1996) Cardioprotective
effects of 70-kDa heat shock protein in transgenic mice. Proc.
Natl. Acad. Sci. USA 93, 2339 –2342
Wu, C. (1995) Heat shock transcription factors: structure and
regulation. Annu. Rev. Cell Dev. Biol. 11, 441– 469
Morimoto, R. I. (1998) Regulation of the heat shock transcriptional response: cross talk between a family of heat shock
factors, molecular chaperones, and negative regulators. Genes
Dev. 12, 3788 –3796
Morano, K. A., and Thiele, D. J. (1999) Heat shock factor
function and regulation in response to cellular stress, growth,
and differentiation signals. Gene Exp. 7, 271–282
Sorger, P. K., and Pelham, H. R. (1988) Yeast heat shock factor
is an essential DNA-binding protein that exhibits temperaturedependent phosphorylation. Cell 54, 855– 864
Wiederrecht, G., Seto, D., and Parker, C. S. (1988) Isolation of
the gene encoding the S. cerevisiae heat shock transcription
factor. Cell 54, 841– 853
Clos, J., Westwood, J. T., Becker, P. B., Wilson, S., Lambert, K.,
and Wu, C. (1990) Molecular cloning and expression of a
hexameric Drosophila heat shock factor subject to negative
regulation. Cell 63, 1085–1097
Rabindran, S. K., Giorgi, G., Clos, J., and Wu, C. (1991)
Molecular cloning and expression of a human heat shock
factor, HSF1. Proc. Natl. Acad. Sci. USA 88, 6906 – 6910
Sarge, K. D., Zimarino, V., Holm, K., Wu, C., and Morimoto,
R. I. (1991) Cloning and characterization of two mouse heat
The FASEB Journal
PIRKKALA ET AL.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
shock factors with distinct inducible and constitutive DNAbinding ability. Genes Dev. 5, 1902–1911
Schuetz, T. J., Gallo, G. J., Sheldon, L., Tempst, P., and
Kingston, R. E. (1991) Isolation of a cDNA for HSF2: evidence
for two heat shock factor genes in humans. Proc. Natl. Acad. Sci.
USA 88, 6911– 6915
Fiorenza, M. T., Farkas, T., Dissing, M., Kolding, D., and
Zimarino, V. (1995) Complex expression of murine heat shock
transcription factors. Nucleic Acids Res. 23, 467– 474
Goodson, M. L., Park-Sarge, O.-K., and Sarge, K. D. (1995)
Tissue-dependent expression of heat shock factor 2 isoforms
with distinct transcriptional activities. Mol. Cell. Biol. 15, 5288 –
5293
Nakai, A., Tanabe, M., Kawazoe, Y., Inazawa, J., Morimoto, R. I.,
and Nagata, K. (1997) HSF4, a new member of the human heat
shock family which lacks properties of a transcriptional activator. Mol. Cell. Biol. 17, 469 – 481
Tanabe, M., Sasai, N., Nagata, K., Liu, X.-D., Liu, P. C. C.,
Thiele, D. J., and Nakai, A. (1999) The mammalian HSF4 gene
generates both an activator and a repressor of heat shock genes
by alternative splicing. J. Biol. Chem. 274, 27845–27856
Scharf, K.-D., Rose, S., Zott, W., Schöffl, F., and Nover, L.
(1990) Three tomato genes code for heat stress transcription
factors with a region of remarkable homology to the DNAbinding domain of the yeast HSF. EMBO J. 9, 4495– 4501
Nakai, A., and Morimoto, R. I. (1993) Characterization of a
novel chicken heat shock transcription factor, HSF3, suggests a
new regulatory pathway. Mol. Cell. Biol. 13, 1983–1997
Råbergh, C. M. I., Airaksinen, S., Soitamo, A., Björklund, H. V.,
Johansson, T., Nikinmaa, M., and Sistonen, L. (2000) Tissuespecific expression of zebrafish (Danio rerio) heat shock factor
1 mRNA in response to heat stress. J. Exp. Biol. 203, 1817–1824
Stump, D. G., Landsberger, N., and Wolffe, A. P. (1995) The
cDNA encoding Xenopus laevis heat-shock factor 1 (XHSF1):
nucleotide and deduced amino-acid sequences, and properties
of the encoded protein. Gene 160, 207–211
Nover, L., Scharf, K.-D., Gagliardi, D., Vergne, P., CzarneckaVerner, E., and Gurley, W. B. (1996) The Hsf world: classification and properties of plant heat stress transcription factors.
Cell Stress Chap. 1, 215–223
Nakai, A. (1999) New aspects in the vertebrate heat shock
factor system: Hsf3 and Hsf4. Cell Stress Chap. 4, 86 –93
Leppä, S., and Sistonen, L. (1997) Heat shock response—
pathophysiological implications. Ann. Med. 29, 73–78
McMillan, D. R., Xiao, X., Shao, L., Graves, K., and Benjamin,
I. J. (1998) Targeted disruption of heat shock transcription
factor 1 abolishes thermotolerance and protection against
heat-inducible apoptosis. J. Biol. Chem. 273, 7523–7528
Xiao, X., Zuo, X., Davis, A. A., McMillan, D. R., Curry, B. B.,
Richardson, J. A., and Benjamin, I. J. (1999) HSF1 is required
for extra-embryonic development, postnatal growth and protection during inflammatory responses in mice. EMBO J. 18,
5943–5952
Pirkkala, L., Alastalo, T.-P., Zuo, X., Benjamin, I. J., and
Sistonen, L. (2000) Disruption of heat shock factor 1 reveals an
essential role in the ubiquitin proteolytic pathway. Mol. Cell.
Biol. 20, 2670 –2675
Liu, X.-D., Liu, P. C. C., Santoro, N., and Thiele, D. J. (1997)
Conservation of a stress response: human heat shock transcription factors functionally substitute for yeast HSF. EMBO J. 16,
6466 – 6477
Tanabe, M., Kawazoe, Y., Takeda, S., Morimoto, R. I., Nagata,
K., and Nakai, A. (1998) Disruption of the HSF3 gene results in
the severe reduction of heat shock gene expression and loss of
thermotolerance. EMBO J. 17, 1750 –1758
Johnson, M. S., and Overington, J. P. (1993) A structural basis
for sequence comparisons. An evaluation of scoring methodologies. J. Mol. Biol. 233, 716 –738
Johnson, M. S., May, A. C. W., Rodionov, M. A., and Overington, J. P. (1996) Discrimination of common protein folds:
application of protein structure to sequence/structure comparisons. Methods Enzymol. 266, 575–598
Damberger, F. F., Pelton, J. G., Harrison, C. J., Nelson,
H. C. M., and Wemmer, D. E. (1994) Solution structure of the
DNA-binding domain of the heat shock transcription factor
determined by multidimensional heteronuclear magnetic resonance spectroscopy. Protein Sci. 3, 1806 –1821
DIFFERENTIAL FUNCTIONS OF HSFs
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
Harrison, C. J., Bohm, A. A., and Nelson, H. C. M. (1994)
Crystal structure of the DNA binding domain of the heat shock
transcription factor. Science 263, 224 –227
Vuister, G. W., Kim, S. J., Orosz, A., Wu, C., Bax, A. (1994)
Solution structure of the DNA-binding domain of Drosophila
heat shock transcription factor. Nature Struct. Biol. 1, 605– 614
Schultheiss, J., Kunert, O., Gase, U., Scharf, K.-D., Nover, L.,
and Rüterjans, H. (1996) Solution structure of the DNAbinding domain of the tomato heat stress transcription factor
HSF24. Eur. J. Biochem. 236, 911–921
Yoshima, T., Yura, T., and Yanagi, H. (1998) Function of the
C-terminal transactivation domain of human heat shock factor
2 is modulated by the adjacent negative regulatory segment.
Nucleic Acids Res. 26, 2580 –2585
Sorger, P. K., and Nelson, H. C. (1989) Trimerization of a yeast
transcriptional activator via a coiled-coil motif. Cell 59, 807– 813
Rabindran, S. K., Haroun, R. I., Clos, J., Wisniewski, J., and Wu,
C. (1993) Regulation of heat shock factor trimer formation:
role of a conserved leucine zipper. Science 259, 230 –234
Jakobsen, B. K., and Pelham, H. R. (1988) Constitutive binding
of yeast heat shock factor to DNA in vivo. Mol. Cell. Biol. 8,
5040 –5042
Jakobsen, B. K., and Pelham, H. R. (1991) A conserved
heptapeptide restrains the activity of the yeast heat shock
transcription factor. EMBO J. 10, 369 –375
Goodson, M. L., and Sarge, K. D. (1995) Heat-inducible DNA
binding of purified heat shock transcription factor 1. J. Biol.
Chem. 270, 2447–2450
Leppä, S., Pirkkala, L., Saarento, H., Sarge, K. D., and Sistonen,
L. (1997) Overexpression of HSF2-␤ inhibits hemin-induced
heat shock gene expression and erythroid differentiation in
K562 cells. J. Biol. Chem. 272, 15293–15298
Baler, R., Dahl, G., and Voellmy, R. (1993) Activation of
human heat shock genes is accompanied by oligomerization,
modification, and rapid translocation of heat shock transcription factor HSF1. Mol. Cell. Biol. 13, 2486 –2496
Sarge, K. D., Murphy, S. P., and Morimoto, R. I. (1993)
Activation of heat shock gene transcription by HSF1 involves
oligomerization, acquisition of DNA binding activity, and
nuclear localization and can occur in the absence of stress. Mol.
Cell. Biol. 13, 1392–1407
Kawazoe, Y., Nakai, A., Tanabe, M., and Nagata, K. (1998)
Proteasome inhibition leads to the activation of all members of
the heat-shock-factor family. Eur. J. Biochem. 255, 356 –362
Kim, D., Kim, S.-H., and Li, G. C. (1999) Proteasome inhibitors
MG132 and lactacystin hyperphosphorylate HSF1 and induce
hsp70 and hsp27 expression. Biochem. Biophys. Res. Commun.
254, 264 –268
Holbrook, N. J., and Udelsman, R. (1994) Heat shock protein
gene expression in response to physiologic stress and aging. In
The Biology of Heat Shock Proteins and Molecular Chaperones
(Morimoto, R. I., Tissières, A., and Georgopoulos, C., eds) pp.
577–593, Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY
Rabindran, S. K., Wisniewski, J., Li, L., Li, G. C., and Wu, C.
(1994) Interaction between heat shock factor and hsp70 is
insufficient to suppress induction of DNA-binding activity in
vivo. Mol. Cell. Biol. 14, 6552– 6560
Zuo, J., Rungger, D., and Voellmy, R. (1995) Multiple layers of
regulation of human heat shock transcription factor 1. Mol.
Cell. Biol. 15, 4319 – 4330
Zuo, J., Baler, R., Dahl, G., and Voellmy, R. (1994) Activation
of the DNA-binding ability of human heat shock transcription
factor 1 may involve the transition from an intramolecular to
an intermolecular triple-stranded coiled-coil structure. Mol.
Cell. Biol. 14, 7557–7568
Farkas, T., Kutskova, Y. A., and Zimarino, V. (1998) Intramolecular repression of mouse heat shock factor 1. Mol. Cell. Biol.
18, 906 –918
Zhong, M., Orosz, A., and Wu, C. (1998) Direct sensing of heat
and oxidation by Drosophila heat shock transcription factor.
Mol. Cell 2, 101–108
Zou, J., Guo, Y., Guettouche, T., Smith, D. F., and Voellmy, R.
(1998) Repression of heat shock transcription factor HSF1
activation by HSP90 (HSP90 complex) that forms a stresssensitive complex with HSF1. Cell 94, 471– 480
1129
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
1130
Ali, A., Bharadwaj, S., O’Carroll, R., and Ovsenek, N. (1998)
HSP90 interacts with and regulates the activity of heat shock
factor 1 in Xenopus oocytes. Mol. Cell. Biol. 18, 4949 – 4960
Bharadwaj, S., Ali, A., and Ovsenek, N. (1999) Multiple components of the HSP90 chaperone complex function in regulation of heat shock factor 1 in vivo. Mol. Cell. Biol. 19, 8033– 8041
Satyal, S. H., Chen, D., Fox, S. G., Kramer, J. M., and Morimoto, R. I. (1998) Negative regulation of the heat shock
transcriptional response by HSBP1. Genes Dev. 12, 1962–1974
Sorger, P. K., Lewis, M. J., and Pelham, H. R. (1987) Heat
shock factor is regulated differently in yeast and HeLa cells.
Nature (London) 329, 81– 84
Jurivich, D. A., Sistonen, L., Kroes, R. A., and Morimoto, R. I.
(1992) Effect of sodium salicylate on the human heat shock
response. Science 255, 1243–1245
Lee, B. S., Chen, J., Angelidis, C., Jurivich, D. A., and Morimoto, R. I. (1995) Pharmacological modulation of heat shock
factor 1 by anti-inflammatory drugs results in protection
against stress-induced cellular damage. Proc. Natl. Acad. Sci.
USA 92, 7207–7211
Cotto, J. J., Kline, M., and Morimoto, R. I. (1996) Activation of
heat shock factor 1 DNA binding precedes stress-induced
serine phosphorylation. J. Biol. Chem. 271, 3355–3358
Green, M., Schuetz, T. J., Sullivan, E. K., and Kingston, R. E.
(1995) A heat shock-responsive domain of human HSF1 that
regulates transcription activation domain function. Mol. Cell.
Biol. 15, 3354 –3362
Newton, E. M., Knauf, U., Green, M., and Kingston, R. E.
(1996) The regulatory domain of human heat shock factor 1 is
sufficient to sense heat stress. Mol. Cell. Biol. 16, 839 – 846
Chu, B., Soncin, F., Price, B. D., Stevenson, M. A., and
Calderwood, S. (1996) Sequential phosphorylation by mitogen-activated protein kinase and glycogen synthase kinase 3
represses transcriptional activation by heat shock factor 1.
J. Biol. Chem. 271, 30847–30857
Knauf, U., Newton, E. M., Kyriakis, J., and Kingston, R. E.
(1996) Repression of human heat shock factor 1 activity at
control temperature by phosphorylation. Genes Dev. 10, 2782–
2793
Kline, M. P., and Morimoto, R. I. (1997) Repression of the heat
shock factor 1 transcriptional activation domain is modulated
by constitutive phosphorylation. Mol. Cell. Biol. 17, 2107–2115
Chu, B., Zhong, R., Soncin, F., Stevenson, M. A., and Calderwood, S. K. (1998) Transcriptional activity of heat shock factor
1 at 37°C is repressed through phosphorylation on two distinct
serine residues by glycogen synthase kinase 3␣ and protein
kinases C␣ and C␰. J. Biol. Chem. 273, 18640 –18646
Kim, J., Nueda, L., Meng, Y.-H., Dynan, W. S., and Mivechi,
N. F. (1997) Analysis of the phosphorylation of human heat
shock transcription factor by MAP kinase family members.
J. Cell. Biochem. 67, 43–54
He, B., Meng, Y.-H., and Mivechi, N. F. (1998) Glycogen
synthase kinase 3␤ and extracellular signal-regulated kinase
inactivate heat shock transcription factor 1 by facilitating the
disappearance of transcriptionally active granules after heat
shock. Mol. Cell. Biol. 18, 6624 – 6633
Dai, R., Frejtag, W., He, B., Zhang, Y., and Mivechi, N. F.
(2000) JNK targeting and phosphorylation of heat shock
factor-1 suppress its transcriptional activity. J. Biol. Chem. 275,
18210 –18218
Cotto, J. J., Fox, S. G., and Morimoto, R. I. (1997) HSF1
granules: a novel stress-induced nuclear compartment of human cells. J. Cell Sci. 110, 2925–2934
Jolly, C., Morimoto, R. I., Robert-Nicoud, M., and Vourc’h, C.
(1997) HSF1 transcription factor concentrates in nuclear foci
during heat shock: relationship with transcription sites. J. Cell
Sci. 110, 2935–2941
Jolly, C., Usson, Y., and Morimoto, R. I. (1999) Rapid and
reversible relocalization of heat shock factor 1 within seconds
to nuclear stress granules. Proc. Natl. Acad. Sci. USA 96, 6769 –
6774
Holmberg, C. I., Illman, S. A., Kallio, M., Mikhailov, A., and
Sistonen, L. (2000) Formation of nuclear HSF1 granules varies
depending on stress stimuli. Cell Stress Chap. 5, 219 –228
Kawazoe, Y., Tanabe, M., Sasai, N., Nagata, K., and Nakai, A.
(1999) HSF3 is a major heat shock responsive factor during
chicken embryonic development. Eur. J. Biochem. 265, 688 – 697
Vol. 15
May 2001
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
Nakai, A., Kawazoe, Y., Tanabe, M., Nagata, K., and Morimoto,
R. I. (1995) The DNA-binding properties of two heat shock
factors, HSF1 and HSF3, are induced in the avian erythroblast
cell line HD6. Mol. Cell. Biol. 15, 5268 –5278
Nakai, A., and Ishikawa, T. (2000) A nuclear localization signal
is essential for stress-induced dimer-to-trimer transition of heat
shock transcription factor 3. J. Biol. Chem. 275, 34665–34671.
Tanabe, M., Nakai, A., Kawazoe, Y., and Nagata, K. (1997)
Different thresholds in the responses of two heat shock transcription factors, HSF1 and HSF3. J. Biol. Chem. 272, 15389 –
15395
Abravaya, K., Phillips, B., and Morimoto, R. I. (1991) Heat
shock-induced interactions of heat shock transcription factor
and the human hsp70 promoter examined by in vivo footprinting. Mol. Cell. Biol. 11, 586 –592
Sistonen, L., Sarge, K. D., Phillips, B., Abravaya, K., and
Morimoto, R. I. (1992) Activation of heat shock factor 2 during
hemin-induced differentiation of human erythroleukemia
cells. Mol. Cell. Biol. 12, 4104 – 4111
Singh, M. K., and Yu, J. (1984) Accumulation of a heat
shock-like protein during differentiation of human erythroid
cell line K562. Nature (London) 309, 631– 633
Theodorakis, N. G., Zand, D. J., Kotzbauer, P. T., Williams,
G. T., and Morimoto, R. I. (1989) Hemin-induced transcriptional activation of the hsp70 gene during erythroid maturation in K562 cells is due to a heat shock factor-mediated stress
response. Mol. Cell. Biol. 9, 3166 –3173
Sistonen, L., Sarge, K. D., and Morimoto, R. I. (1994) Human
heat shock factors 1 and 2 are differentially activated and can
synergistically induce hsp70 gene transcription. Mol. Cell. Biol.
14, 2087–2099
Pirkkala, L., Alastalo, T.-P., Nykänen, P., Seppä, L., and Sistonen, L. (1999) Differentiation lineage-specific expression of
human heat shock transcription factor 2. FASEB J. 13, 1089 –
1098
Mezger, V., Bensaude, O., and Morange, M. (1989) Ununsual
levels of heat shock element-binding activity in embryonal
carcinoma cells. Mol. Cell. Biol. 9, 3888 –3896
Mezger, V., Renard, J.-P., Christians, E., and Morange, M.
(1994) Detection of heat shock element-binding activities by
gel shift assay during mouse preimplantation development.
Dev. Biol. 165, 627– 638
Mezger, V., Rallu, M., Morimoto, R. I., Morange, M., and
Renard, J.-P. (1994) Heat shock factor 2-like activity in mouse
blastocysts. Dev. Biol. 166, 819 – 822
Murphy, S. P., Gorzowski, J. J., Sarge, K. D., and Phillips, B.
(1994) Characterization of constitutive HSF2 DNA-binding
activity in mouse embryonal carcinoma cells. Mol. Cell. Biol. 14,
5309 –5317
Eriksson, M., Jokinen, E., Sistonen, L., and Leppä, S. (2000)
Heat shock factor 2 is activated during mouse heart development. Int. J. Dev. Biol. 44, 471– 477
Rallu, M., Loones, M. T., Lallemand, Y., Morimoto, R. I.,
Morange, M., and Mezger, V. (1997) Function and regulation
of heat shock factor 2 during mouse embryogenesis. Proc. Natl.
Acad. Sci. USA 94, 2392–2397
Sarge, K. D., Park-Sarge, O.-K., Kirby, J. D., Mayo, K. E., and
Morimoto, R. I. (1994) Expression of heat shock factor 2 in
mouse testis: potential role as a regulator of heat-shock protein
gene expression during spermatogenesis. Biol. Reprod. 50, 1334 –
1343
Alastalo, T.-P., Lönnström, M., Leppä, S., Kaarniranta, K.,
Pelto-Huikko, M., Sistonen, L., and Parvinen, M. (1998) Stagespecific expression and cellular localization of the heat shock
factor 2 isoforms in the rat seminiferous epithelium. Exp. Cell
Res. 240, 16 –27
Manuel, M., Sage, J., Mattéi, M.-G., Morange, M., and Mezger,
V. (1999) Genomic structure and chromosomal localization of
the mouse Hsf2 gene and promoter sequences. Gene 17,
115–124
Sheldon, L. A., and Kingston, R. E. (1993) Hydrophobic
coiled-coil domains regulate the subcellular localization of
human heat shock factor 2. Genes Dev. 7, 1549 –1558
Mathew, A., Mathur, S. K., and Morimoto, R. I. (1998) The
heat shock response and protein degradation: regulation of
HSF2 by the ubiquitin-proteasome pathway. Mol. Cell. Biol. 18,
5091–5098
The FASEB Journal
PIRKKALA ET AL.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
Yoshima, T., Yura, T., and Yanagi, H. (1997) The trimerization
domain of human heat shock factor 2 is able to interact with
nucleoporin p62. Biochem. Biophys. Res. Commun. 240, 228 –233
Yoshima, T., Yura, T., and Yanagi, H. (1998) Novel testisspecific protein that interacts with heat shock factor 2. Gene
214, 139 –146
Hong, Y., and Sarge, K. D. (1999) Regulation of protein
phosphatase 2A activity by heat shock transcription factor 2.
J. Biol. Chem. 274, 12967–12970
Hong, Y., Lubert, E. J., Rodgers, D. W., and Sarge, K. D. (2000)
Molecular basis of competition between HSF2 and catalytic
subunit for binding to the PR65/A subunit of PP2A. Biochem.
Biophys. Res. Commun. 272, 84 – 89
Kroeger, P. E., Sarge, K. D., and Morimoto, R. I. (1993) Mouse
heat shock transcription factors 1 and 2 prefer a trimeric
binding site but interact differently with the HSP70 heat shock
element. Mol. Cell. Biol. 13, 3370 –3383
Kroeger, P. E., and Morimoto, R. I. (1994) Selection of new
HSF1 and HSF2 DNA-binding sites reveals differences in
trimer cooperativity. Mol. Cell. Biol. 14, 7592–7603
Leppä, S., Pirkkala, L., Chow, S. C., Eriksson, J. E., and
Sistonen, L. (1997) Thioredoxin is transcriptionally induced
upon activation of heat shock factor 2. J. Biol. Chem. 272,
30400 –30404
Gallo, G. J., Prentice, H., and Kingston, R. E. (1993) Heat
shock factor is required for growth at normal temperatures in
the fission yeast Schizosaccharomyces pombe. Mol. Cell. Biol. 13,
749 –761
Jedlicka, P., Mortin, M. A., and Wu, C. (1997) Multiple
functions of Drosophila heat shock transcription factor in vivo.
EMBO J. 16, 2452–2462
Bensaude, O., and Morange, M. (1983) Spontaneous high expression of heat-shock proteins in mouse embryonal carcinoma
cells and ectoderm of day 8 mouse embryo. EMBO J. 2, 173–177
Bensaude, O., Babinet, C., Morange, M., and Jacob, F. (1983)
Heat shock proteins, first major products of zygotic gene
activity in mouse embryo. Nature (London) 305, 331–333
Morange, M., Diu, A., Bensaude, O., and Babinet, C. (1984)
Altered expression of heat shock proteins in embryonal carcinoma and mouse early embryonic cells. Mol. Cell. Biol. 4,
730 –735
Christians, E., Michel, E., Adenot, P., Mezger, V., Rallu, M.,
Morange, M., and Renard, J.-P. (1997) Evidence for the
involvement of mouse heat shock factor 1 in the atypical
expression of the HSP70.1 heat shock gene during mouse
zygotic genome activation. Mol. Cell. Biol. 17, 778 –788
DIFFERENTIAL FUNCTIONS OF HSFs
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
Christians, E., Davis, A. A., Thomas, S. D., and Benjamin, I. J.
(2000) Maternal effect of Hsf1 on reproductive success. Nature
(London) 407, 693– 694
Singh, I. S., Viscardi, R. M., Kalvakolanu, I., Calderwood, S.,
and Hasday, J. D. (2000) Inhibition of tumor necrosis factor-␣
transcription in macrophages exposed to febrile range temperature. J. Biol. Chem. 275, 9841–9848
Cahill, C. M., Waterman, W. R., Xie, Y., Auron, P. E., and
Calderwood, S. K. (1996) Transcriptional repression of the
prointerleukin 1␤ gene by heat shock factor 1. J. Biol. Chem.
271, 24874 –24879
Chen, C., Xie, Y., Stevenson, M. A., Auron, P. E., and Calderwood, S. K. (1997) Heat shock factor 1 represses Ras-induced
transcriptional activation of the c-fos gene. J. Biol. Chem. 272,
26803–26806
Nakai, A., Suzuki, M., and Tanabe, M. (2000) Arrest of
spermatogenesis in mice expressing an active heat shock
transcription factor 1. EMBO J. 19, 1545–1554
Eddy, E. M. (1998) HSP70 –2 heat-shock protein of mouse
spermatogenic cells. J. Exp. Zool. 282, 261–271
Ciechanover, A. (1998) The ubiquitin–proteasome pathway:
on protein death and cell life. EMBO J. 17, 7151–7160
Bush, K. T., Goldberg, A. L., and Nigam, S. K. (1997) Proteasome inhibition leads to a heat-shock response, induction of
endoplasmic reticulum chaperones, and thermotolerance.
J. Biol. Chem. 272, 9086 –9092
Meriin, A. B., Gabai, V. L., Yaglom, J., Shifrin, V. I., and
Sherman, M. Y. (1998) Proteasome inhibitors activate stress
kinases and induce Hsp72. J. Biol. Chem. 273, 6373– 6379
Kanei-Ishii, C., Tanikawa, J., Nakai, A., Morimoto, R. I., and
Ishii, S. (1997) Activation of heat shock transcription factor 3
by c-Myb in the absence of cellular stress. Science 277, 246 –248
Tanikawa, J., Ichikawa-Iwata, E., Kanei-Ishii, C., Nakai, A.,
Matsuzawa, S., Reed, J. C., and Ishii, S. (2000) p53 suppresses
the c-Myb-induced activation of heat shock transcription factor
3. J. Biol. Chem. 275, 15578 –15585
Bosch, T. C. G., Krylow, S. M., Bode, H. R., and Steele, R. E.
(1988) Thermotolerance and synthesis of heat shock proteins:
These responses are present in Hydra attenuata but absent in
Hydra oligactis. Proc. Natl. Acad. Sci. USA 85, 7927–7931
Hofmann, G. E., Buckley, B. A., Airaksinen, S., Keen, J. E., and
Somero, G. N. (2000) Heat-shock protein expression is absent
in the Antarctic fish Trematomus bernacchii (family Nototheniidae). J. Exp. Biol. 203, 2331–2339
Brennecke, T., Gellner, K., and Bosch, T. C. (1998) The lack of
a stress response in Hydra oligactis is due to reduced hsp70
mRNA stability. Eur. J. Biochem. 255, 703–709
1131