The regulation of specific transcription factor activity is essential
to the normal development and maintenance of all organisms. A
large number of transcription factors preexist in various cells in
a latent inactive form and are ‘activated’ to promote transcription
of specific genes once the cell has received certain stimuli. There
are a wide variety of mechanisms which regulate the activity of
preexisting transcription factors including: binding of a specific
ligand to the factor to activate the factor’s activity (e.g. Cu2+ to
the ACE1 factor, steroid hormones to the steroid hormone
receptors); removal of an inhibitory protein from the
transcription factor (e.g. hsp90 from the glucocorticoid receptor,
inhibitor κB (IκB) from nuclear factor κB (NFκB));
transcription factor modification (e.g. phosphorylation of cyclic
AMP responsive element binding protein, CREB); and
subcellular localization (e.g. SV40 large T antigen) (for reviews
see Whiteside and Goodbourn, 1993; Latchman, 1995). Some
transcription factors use various combinations of the above
mechanisms to regulate their activity.
One well-studied transcription factor, which preexists in all
eukaryotic cells, is a protein called heat shock factor (HSF).
HSF mediates the transcription of the heat shock genes. All
organisms respond to elevated temperatures and other forms of
‘stress’ such as inhibitors of oxidative respiration, sulfydryl
reagents, certain heavy metals, and the generation of abnormal
proteins or oxygen radicals within cells, by inducing the
transcription of a family of genes known as the heat shock (hs)
genes. The products of these genes, the hs proteins (hsps), help
cells recover from the effects of the stress and protect them
from further trauma (Morimoto et al., 1994; Parsell and
Lindquist, 1994). For reviews of HSF function see Morimoto
(1998), Mager and De Kruijff (1995), Wu (1995) and Voellmy
All eukaryotes appear to possess HSF1, the major stress-
inducible form of HSF. However, many species possess other
HSFs in addition to HSF1. For example, mammals and
chickens both probably have at least three different HSF genes
each (Sarge et al., 1991; Nakai and Morimoto, 1993; Nakai et
al., 1997). The other HSFs appear either to respond to more
Journal of Cell Science 112, 2765-2774 (1999)
Printed in Great Britain © The Company of Biologists Limited 1999
The induction of the heat shock genes in eukaryotes by heat
and other forms of stress is mediated by a transcription
factor known as heat shock factor 1 (HSF1). HSF1 is
present in unstressed metazoan cells as a monomer with
low affinity for DNA, and upon exposure to stress it is
converted to an ‘active’ homotrimer that binds the
promoters of heat shock genes with high affinity and
induces their transcription. The conversion of HSF1 to its
active form is hypothesized to be a multistep process
involving physical changes in the HSF1 molecule and the
possible translocation of HSF1 from the cytoplasm to the
nucleus. While all studies to date have found active HSF1
to be a nuclear protein, there have been conflicting reports
on whether the inactive form of HSF is predominantly a
cytoplasmic or nuclear protein. In this study, we have made
antibodies against human HSF1 and have reexamined its
localization in unstressed and heat-shocked human HeLa
and A549 cells, and in green monkey Vero cells.
Biochemical fractionation of heat-shocked HeLa cells
followed by western blot analysis showed that HSF1 was
mostly found in the nuclear fraction. In extracts made from
unshocked cells, HSF1 was predominantly found in the
cytoplasmic fraction using one fractionation procedure, but
was distributed approximately equally between the
cytoplasmic and nuclear fractions when a different
procedure was used. Immunofluorescence microscopy
revealed that HSF1 was predominantly a nuclear protein
in both heat shocked and unstressed cells. Quantification of
HSF1 staining showed that approximately 80% of HSF1
was present in the nucleus both before and after heat stress.
These results suggest that HSF1 is predominantly a nuclear
protein prior to being exposed to stress, but has low affinity
for the nucleus and is easily extracted using most
biochemical fractionation procedures. These results also
imply that HSF1 translocation is probably not part of the
multistep process in HSF1 activation for many cell types.
Key words: Heat Shock, Heat Shock Factor, Nuclear localization,
Human heat shock factor 1 is predominantly a nuclear protein before and
after heat stress
Philippe A. Mercier*, Neil A. Winegarden and J. Timothy Westwood
Department of Zoology, University of Toronto, Mississauga, Ontario, Canada L5L 1C6
*Present address: Department of Anatomy and Cell Biology, University of Saskatchewan, Saskatoon, SK, CANADA S7N 5E5
Author for correspondence (e-mail: email@example.com)
Accepted 28 May; published on WWW 21 July 1999
severe forms of heat stress (Tanabe et al., 1997) or to other
types of stimuli. For example, HSF2 is active during
embryogenesis, spermatogenesis and erythroid differentiation
(Sistonen et al., 1992; Mezger et al., 1994; Sarge et al., 1994),
but its activity is not stimulated during heat shock.
The transformation of HSF1 from its inactive to active form
is believed to be a multistep process (Sarge et al., 1993; Cotto
et al., 1996; Wu, 1995). These steps have been proposed to
include: (1) conversion of HSF1 from monomer to a
homotrimer (Westwood et al., 1991; Westwood and Wu, 1993;
Baler et al., 1993; Sarge et al., 1993); (2) secondary changes
in HSF conformation such that certain domains believed to be
involved in the transcription promoting activities of HSF1
become exposed (Chen et al., 1993; Westwood and Wu, 1993;
Wu, 1995); and (3) stress-induced hyperphosphorylation of
HSF1 (Sorger and Pelham, 1988; Larson et al., 1988; Sarge et
al., 1993; Cotto et al., 1996; Winegarden et al., 1996). It has
also been proposed that for mammalian (Sarge et al., 1993;
Baler et al., 1993), chicken (Nakai et al., 1995) and Drosophila
cells (Zandi et al., 1997), translocation of HSF1 from the
cytoplasm to the nucleus is a step that regulates HSF activity.
The homotrimerization of HSF1 has been shown to be essential
for HSF to bind the heat shock element(s) (HSEs) found
upstream of all hs genes with high affinity (Sorger and Pelham,
1988; Clos et al., 1990; Fernandes et al., 1994; Wu, 1995). The
exact roles of other changes in HSF1, including stress-induced
hyperphosphorylation, are yet to be determined for HSF1 in
All studies to date have been in agreement that the trimeric
‘active’ form of HSF1 is a nuclear protein. However, there has
been some controversy as to the location of HSF1 prior to
stress in metazoans. In Drosophila, a number of reports have
suggested that HSF1 is a nuclear protein before and after stress
(Westwood et al., 1991; Wu et al., 1994; Shopland and Lis,
1996; Orosz et al., 1996), but one report has suggested that it
is predominantly cytoplasmic prior to heat shock and nuclear
after heat stress (Zandi et al., 1997). Similarly, while some
studies have suggested that mammalian HSF1 is always a
nuclear protein (Wu et al., 1994; Martinez-Balbas et al., 1995),
others have suggested that mammalian HSF1 is predominantly
a cytoplasmic protein under nonshock conditions and is
translocated into the nucleus upon heat stress (Sarge et al.,
1993; Baler et al., 1993; Sistonen et al., 1994). Chicken HSF1
has been reported to be a cytoplasmic protein before heat stress
and a nuclear protein after heat stress (Nakai et al., 1995).
Xenopus HSF1 has been reported to be a nuclear protein before
and after heat stress (Mercier et al., 1997). In tomatoes, the
localization of two different HSFs has been studied (Lyck et
al., 1997). The constitutive, stress-activated HSF called HSFA1
(formerly Lp-HSF8) appears to be distributed between the
nucleus and cytoplasm before heat shock and is exclusively
nuclear after heat shock.
In this study we have reexamined the localization of
mammalian HSF1 prior to stress. Using an affinity-purified
preparation of a polyclonal antibody made against a
bacterially produced maltose binding protein + Homo sapiens
HSF1 (MBP+HsHSF1) fusion protein, we confirm that
HsHSF1 is predominantly a nuclear protein before and after
stress. This finding implies that for many cell types, the
translocation of HsHSF1 from the cytoplasm to the nucleus
is not a step involved in regulating HSF1 transcriptional
activity and that any cellular signals that are involved in the
activation/deactivation of HSF1 are probably realized in the
MATERIALS AND METHODS
Human HeLa and A549 cells, and green monkey Vero cells were
grown in 45% Dulbecco’s Modified Eagles medium (Sigma), 45% F-
12 Ham nutrient mixture, Hepes modification (Sigma) supplemented
with 25 mM NaHCO3, 2% Fetal clone II (Hyclone), 8% Cool Calf 2
(Sigma), 20 µg/ml gentamycin (Sigma), pH 7.4, at 37°C in T-75 tissue
culture flasks (Sarstedt). A549 and Vero cells were generously
provided by W. Furyaya and S. Grinstein (Hospital for Sick Children,
Toronto, Ontario). Heat-shocked cells were prepared by immersion of
T-flasks in a circulating water bath at the temperatures indicated in
the Fig. legends. HeLa cells used for experiments in Fig. 3A,B were
grown in suspension Eagles Minimum Essential Medium, Joklik
modification (Sigma), supplemented with 10 mM Hepes, pH 7.4, 25
mM NaHCO3, 10% Cool Calf 2 (Sigma), 20 µg/ml gentamycin in
spinner flasks (Bellco Biotechnology). Prior to these experiments, 5
ml samples of cell suspension at a concentration of 4-5×105cells/ml
were transferred to 50 ml polypropylene tubes and aerated by shaking
at 180 rpm, 37°C, for 5 hours.
Preparation of protein extracts
Whole cell extracts
Cells were scraped into the medium and pelleted; the medium was
removed, the pellet resuspended in 3 volumes of 1× SDS sample
buffer (Ausubel et al., 1995) and then sonicated for 10 seconds with
a Heat Systems Ultrasonic sonicator.
NP-40 based fractionation
This procedure is a modification of the procedure described by Baler
et al. (1993). Briefly, 2.5×106cells were pelleted, washed once in 1×
PBS and resuspended in 150 µl Nonidet P-40 (NP-40) low salt buffer
containing 50 mM Tris-Cl, pH 8.0, 50 mM NaCl, 1% NP-40, 1 mM
dithiothreitol (DTT), 0.5 mM phenylmethylsulfonyl fluoride (PMSF).
The nuclei were pelleted by centrifugation at 7,000 g for 1 minute at
4°C. The supernatant was removed and saved as the cytoplasmic
fraction. 150 µl NP-40 high salt buffer (50 mM Tris-Cl, pH 8.0, 450
mM NaCl, 1% NP-40, 1 mM dithiothreitol, 0.5 mM PMSF) was then
added to the nuclei. After incubation for 10 minutes at 4°C the sample
was sonicated for 5 seconds and the resulting extract was the nuclear
The procedure used was modified from one described by Lee et al.
(1994). Briefly, 2.5×106cells were pelleted, washed once in 1× PBS,
resuspended in 1 packed cell volume of Buffer A (10 mM Hepes,
pH 8.0, 1.5 mM MgCl2, 10 mM KCl, 1 mM DTT, 0.5 mM PMSF)
and incubated on ice for 15 minutes. The cells were then lysed by
passing them through a 28.5-gauge needle attached to a 0.5 ml
syringe 5 times and the nuclei pelleted by centrifugation for 20
seconds at 12,000 g. The supernatant was removed and saved as the
cytoplasmic fraction. The nuclei were then resuspended in one
packed cell volume of Buffer C (20 mM Hepes, pH 8.0, 1.5 mM
MgCl2, 25% glycerol, 420 mM NaCl, 0.2 mM EDTA, 1 mM DTT,
0.5 mM PMSF) and incubated on ice for 30 minutes with occasional
shaking. The sample was then centrifuged for 5 minutes at 12,000
g at 4°C to separate the nuclear soluble (supernatant), and nuclear
insoluble (pellet) fractions. To prepare cytoplasmic and nuclear
protein extracts for SDS-PAGE, 1/6 volume of 6× SDS sample
buffer (Ausubel et al., 1995) was added, and for the nuclear
insoluble fraction, two packed cell volumes of 1× SDS sample buffer
P. A. Mercier and others
2767HSF1 is a nuclear protein in unstressed cells
(Ausubel et al., 1995) was added and the mixture sonicated for 10
Electrophoretic mobility-shift assays (EMSAs)
HSF1 DNA binding activity was analyzed by electrophoretic
mobility-shift assay. Binding reactions contained 3 µl of protein
extract, 1 µl 10× buffer mix (100 mM Hepes, pH 7.9, 30% glycerol
w/v) and 1 µl 10× bovine serum albumin/nucleotide mix (20 mg/ml
bovine serum albumin fraction V, 0.5 mg/ml E. coli DNA, 2 mg/ml
tRNA, 0.2 mg/ml poly d(N)6, 0.5 mg/ml poly(dI-dC)/poly(dI-dC),
NaCl to 200 mM final concentration, and double distilled water to 9
µl. This mixture was incubated on ice for 10 minutes. 1 µl of a 32P-
labelled oligonucleotide containing a heat shock element (HSE) (0.2
pmol) (Winegarden et al., 1996) was then added to the above mixture.
Binding reactions were then allowed to proceed for 10 minutes at
room temperature (18-21°C), after which 2 µl of 6× loading dye
(0.25% Bromophenol Blue, 30% glycerol, 3× Tris-borate-EDTA
buffer) was added. Samples were electrophoresed, dried and exposed
to film as previously described (Winegarden et al., 1996).
Production of anti-human HSF1 antibodies
The cDNA clone of human HSF1, pBS10811 (Rabindran et al.,
1991) was digested with EcoR1 and religated. Clones were selected
that had HSF1 in the opposite orientation from the original
pBS10811. The HSF1 coding sequences were removed with Xba1
and HindIII and cloned into pMalc2 (New England Biolabs)
digested with Xba1 and HindIII. This yielded a construct which
produced an in-frame fusion protein between maltose binding
protein (MBP) and full-length HsHSF1. E. coli containing either
MBP or MBP+HsHSF1 constructs were induced to overexpress, and
induced proteins were purified essentially as per the manufacturer’s
instructions (New England Biolabs). Antibodies were produced to
MBP+HsHSF1 in rabbits by an initial injection of 1.5 mg of purified
antigen mixed 1:1 in complete Freund’s adjuvant followed by four
boosts, at 3-week intervals, with 0.75 mg of antigen mixed 1:1 with
incomplete Freund’s adjuvant. Anti-HsHSF1 antisera (PM95-1) was
affinity purified following a procedure described by Harlow and
Lane (1988). Briefly, crude antiserum was passed over MBP coupled
to Affigel 10 beads (BioRad) until all of the anti-MBP antibodies
had been removed (as tested by western blot analysis). The anti-
MBP-depleted serum was passed over MBP+HsHSF1 coupled to
Affigel-10. Bound anti-HsHSF1 antibodies were eluted with low pH
buffer (Harlow and Lane, 1988).
Western blot analysis
Proteins fractionated by SDS-PAGE (8% polyacrylamide for HSF,
12% for histone H1 and IκB) were electroblotted onto nitrocellulose
(Biotrace NT, Gelman) and the blots blocked with 5% powdered
milk in TBST (20 mM Tris, pH 7.6, 137 mM NaCl, 0.1% Tween
20). Antibodies were diluted in 2% gelatin (BioRad) in TBST. For
Fig. 1, the antibody dilutions were 1:2000 for preimmune; 1:2000
for crude anti-MBP-HsHSF1 (PM95-1); 1:100 for affinity-purified
anti-HsHSF1 (PM95-1); and 1:1000 for SR191 (a gift from S.
Rabindran and C. Wu, National Institutes for Health, Bethesda,
MD). For Fig. 2, the antibody dilutions were 1:2000 for PM95-1
antibody, the anti-histone antibody (Chemicon International Inc.,
cat. no. MAB052), and 1:1000 for the anti-IκB antibody (Santa Cruz
Biotechnology, cat. no. SC847). Blots were incubated in the primary
antibody for 1 hour at room temperature, washed with TBST, and
then incubated for 45 minutes with secondary antibody, which was
either alkaline phosphatase-conjugated goat anti-rabbit IgG
(BioRad) diluted 1:2000 in 2% gelatin/TBST for the anti-HSF
antibodies, or horseradish peroxidase-conjugated goat anti-mouse
IgG (Sigma) diluted 1:40,000 in 5% powdered milk/TBST for the
anti-histone and anti-IκB antibodies. Blots were then washed in
TBST and developed for alkaline phosphatase activity using 5-
Bromo-4-chloro-3-indoylphosphate p-Toluidine Salt (BCIP) and
Nitroblue Tetrazolium Chloride (NBT) reagents (BRL Life
Technologies) or for horseradish peroxidase activity using a
chemiluminescence reagent kit (NEN Life Science) as per the
Cells, grown in chamber slides (NUNC), were either kept in non-
shock conditions (37°C) or heat shocked (42°C) by placing slides in
a humidified container floating in a water bath. The medium was
removed and the cells were fixed in 4% paraformaldehyde in 1× PBS.
The slides were either placed in 95% ethanol or stained immediately.
After blocking with 0.5% BSA, 0.5% Tween-20 in 1× PBS for 20
minutes at room temperature, slides were incubated in 1:10 affinity-
purified anti-MBP+HsHSF1 (PM95-1), 1:250 anti-HsHSF1 antibody
(SR191), or 1:200 anti-hsp70 (Stressgen, cat. no. SPA-810) in 0.5%
BSA, 0.5% Tween 20 in 1× PBS for 1 hour. Slides were then washed
3 times in 1× PBS, followed by incubation for 30 minutes with FITC-
conjugated goat anti-rabbit antibody (1:200; Cappel, cat. no. 55655)
or rhodamine-conjugated goat anti-mouse antibody (1:500; Jackson
Immunoresearch Laboratories Inc.) diluted in 0.5% BSA, 0.5%
Tween-20 in 1× PBS. The slide was washed 3 times in 1× PBS and
then the DNA was stained for fluorescence with Hoechst 33342
(Sigma) diluted to 1 µg/ml in 1× PBS for 5 minutes. The slide was
then rinsed once in 0.05% Tween 20 in 1× PBS. The cells were
mounted in 1 drop of antifade (0.1% phenylenediamine (Sigma) in
70% glycerol) before a coverslip was applied. Photography was
performed using a Nikon Microphot fluorescence microscope, a
Nikon Plan 40× objective and Fujichrome Sensia 400 ISO film.
Exposure times were 8 seconds for Hoechst-stained preparations, and
16 seconds (with neutral density filters) for FITC-stained samples. 35
mm transparencies and autoradiograms were digitized using an Agfa
Arcus II scanner. When necessary, digitized images were adjusted for
contrast and brightness using Adobe Photoshop.
Hoffman modulation contrast microscopy
Modulation contrast images of the cells were taken using a Zeiss
Axiovert 100 using a Hoffman HMC 20× LWD objective and
corresponding condenser (Modulation Optics Inc., Greenvale NY).
Images were captured using Northern Exposure software (v2.9d)
(Empix Imaging; Mississauga, Ontario), a Sony black and white
CCD video camera (Model XC-75) and an ATI Optimus frame
Quantification of immunofluorescence
To determine the relative fluorescence (HSF1 staining) in the nucleus
and cytoplasm, HeLa cells were incubated with SR191 antiserum
(1:250 dilution) and processed as described above. Fluorescent
images of cells were captured using Northern Exposure software and
a Sony CCD video camera as described above, except that the camera
was attached to a Nikon Microphot fluorescence microscope (Nikon
Plan 40× objective). For nuclear images (Hoechst stain), two neutral
density filters (total value 6) were used and 0.5 second integrated
exposures were taken. FITC (anti-HsHSF1) images were captured in
the absence of neutral density filters, using 2.5 second integrated
exposures. Quantification analysis of the images was performed
using Northern Exposure. In order to quantify nuclear area, the
Hoechst captures were thresholded such that only nuclei were
selected. Nuclei were manually traced, and the software was asked
to calculate perimeter, density and area of the traced section. FITC
fluorescence (anti-HSF1) was quantified similarly. For each
condition in a single experiment, three different slides were used, and
10 to 20 cells from each slide were quantified. Each set of conditions
was performed in three separate experiments. Cells stained with
secondary antibody alone were also thresholded to select the
fluorescent sections of the image. Since the fluorescence across the
cells was uniform, small sections in the center of the image were
selected for density and area quantification.
Production of polyclonal antibodies specific to
We produced polyclonal antibodies to a maltose binding
protein+human HSF1 (MBP+HsHSF1) fusion protein. The
cDNA for HsHSF1 was cloned behind an MBP expression
construct to facilitate rapid large-scale preparation. Using this
construct under control of the lac repressor, MBP+HsHSF1
was overexpressed and purified (Fig. 1A). During
MBP+HsHSF1 protein degraded to yield a product that is
presumed to contain, on the basis of its apparent molecular
half of the full-length
size, 15 kDa of HsHSF1 fused to the MBP moiety as well as
other degradation products. The eluate from the purification
was used to raise polyclonal antibodies by injection into
rabbits. Crude antiserum raised against MBP+HsHSF1 was
affinity purified to yield antibodies specific to HsHSF1. The
ability of the crude and affinity-purified antibodies (PM95-1)
to recognize HsHSF1 in HeLa cell extracts was tested by
western blot analysis (Fig. 1B). A different polyclonal
antiserum made against HsHSF1 produced in bacteria
(SR191) (provided by S. Rabindran and C. Wu; Rabindran et
al., 1991) was also tested for comparison (Fig. 1B). Both the
affinity-purified PM95-1 antibodies and SR191 recognized an
82 kDa band corresponding to HsHSF1 in nonshocked cells
(37°C) and 82 kDa to 95 kDa band(s) in heat-shocked cells
(42°C, 15 minutes), corresponding to differentially
phosphorylated forms of HsHSF1. The relative mobility of
hypophosphorylated and hyperphosphorylated HsHSF1
reported here is in good agreement with the previously
described relative mobility of HsHSF1 (Rabindran et al.,
1991, 1994). Both affinity-purified PM95-1 antibodies and
SR191 also recognized a 71 kDa protein, which could be a
degradation product of HsHSF1 (Rabindran et al., 1994).
Localization of HSF1 by biochemical fractionation of
Two groups have judged HsHSF1 to be a cytoplasmic protein
under non-shock conditions based on the biochemical
fractionation of unshocked and heat-shocked cells (Baler et
al., 1993; Sarge et al., 1993; Sistonen et al., 1994). In this
study we used two different biochemical fractionation
techniques to produce cytoplasmic and nuclear extracts; a
Dignam-type fractionation scheme (Lee et al., 1994) and a
fractionation scheme based on the nonionic detergent NP-40
(Baler et al., 1993). The nuclei of unshocked and heat
shocked HeLa cells were biochemically fractionated from
the cytosol and the presence of HSF1 was tested by western
blot analysis (Fig. 2A,B) and electrophoretic mobility-shift
assay (EMSA) (Fig. 3A,B). Purity of the fractions was
assessed by examining each fraction for the presence or
absence of IκBα, a cytoplasmic marker, and histone H1, a
nuclear marker (Fig. 2A,B). As expected, IκBα was only
detected in the cytoplasmic fractions. Histone H1 was only
found in the nuclear fractions for the NP-40 fractionation
scheme (Fig. 2A) and almost all of it was found in the
insoluble nuclear fractions for the Dignam fractionation
scheme (Fig. 2B). The HsHSF1 fractionation results are in
agreement with previous studies, i.e. that HSF1 in unshocked
cells fractionates to the cytosol. However, the proportion of
HSF1 in the cytoplasmic fraction differs, depending on the
fractionation scheme used. That is, with the NP-40 based
fractionation procedure, almost all of the unshocked HSF1
was found in the cytoplasmic fraction (Fig. 2A). Using the
modified Dignam procedure, a large portion of the unshocked
HSF1 was found in the nuclear fraction and some in the
insoluble fraction (IN) (Fig. 2B). In cells heat shocked for
either 15 minutes (Fig. 2) or 180 minutes (not shown) at
42°C, most of the HSF1 fractionated to the nucleus but there
was a small portion of HSF1 that was still found in the
cytoplasm (Fig. 2A,B).
The protein extracts made using the two different
extraction procedures were also examined for HSF1 binding
P. A. Mercier and others
Fig. 1. Production of antibodies that specifically recognize HsHSF1.
(A) E. coli containing the MBP+HsHSF1 fusion construct were
induced to overexpress MBP+HsHSF1 by the addition of isopropyl-
β-D-galactoside (IPTG) for 0 or 3 hours. A protein extract (soluble
extract) was made from the induced bacteria, passed over amylose
resin, and the flow-through and eluate were examined for the
presence of MBP+HsHSF1. The proteins in the various samples were
separated by SDS-PAGE and stained with Coomassie Blue. The
eluate was injected into rabbits to produce polyclonal antisera. (B)
Whole cell extracts of unshocked (37°C) and heat-shocked (42°C, 15
minutes) HeLa cells were used to test the specificity of antibodies to
HsHSF1 by western blot analysis. Blots were incubated with
preimmune serum (1:2000), crude anti-MBP-HsHSF1 (PM95-1;
1:2000), affinity-purified PM95-1 (1:100), and a different crude anti-
HsHSF1 antiserum (SR191; 1:1000). ns, nonspecific.
2769HSF1 is a nuclear protein in unstressed cells
activity to HSEs in an EMSA. Unshocked cells showed
virtually no HSF1 binding activity in either the cytoplasmic
or nuclear fractions (Fig. 3). In heat-shocked extracts,
virtually all of the HSF1 binding activity was found in the
nuclear fractions for both fractionation schemes, although a
small amount of activity is seen in the 15 minute heat-
shocked NP-40 cytoplasmic fraction (Fig. 3B). These results
are consistent with those obtained previously by others (Sarge
et al., 1993; Baler et al., 1993), with the exception that we
did not observe as much attenuation of HSF1 binding activity
or reversion to cytoplasmic localization during prolonged
heat shock (Sarge et al., 1993).
Localization of HSF1 by indirect
The distribution of HSF1 within cells was examined by indirect
immunofluorescence in three different cell types: human HeLa
cells (Fig. 4), human kidney A549 cells (Fig. 5) and green
monkey kidney Vero cells (Fig. 6). HSF1 staining was
examined in both unshocked (37°C) and heat-shocked (42°C,
15 minutes) cells using two different HSF1 antibodies: SR191,
a polyclonal antiserum made against bacterially produced
human HSF1; and affinity-purified PM95-1 containing
polyclonal antibodies made against a bacterially produced
maltose binding protein-human HSF1 fusion protein. Both
antisera gave essentially the same results in all three cell types.
In both unshocked and heat-shocked cells, HSF1 is seen mainly
in the nucleus but not in the nucleolus (Figs 4F,I,L,O; 5C,F,I,L;
6C,F,I,L). Some fluorescence is observed in the cytoplasm.
However, a proportion of this fluorescence can be attributed to
the fluorescence seen using the secondary antibody alone,
which stains the entire cell faintly (Fig. 4C). The distribution
of HSF1 we observed is entirely consistent with previous
studies that used the SR191 antisera (Rabindran et al., 1991;
Wu et al., 1994; Martinez-Balbas et al., 1995). They are also
consistent with results obtained with a polyclonal anti-mouse
HSF1 antiserum used to stain HeLa and other mammalian cell
lines (M. Vujanac, O. Bensaude and E. Zimarino, manuscript
in preparation). We also examined the distribution of hsc/hsp
70 before and after heat shock (Fig. 4Q,S). Hsc/hsp70 was
predominantly cytoplasmic prior to heat shock (Fig. 4Q) and
predominantly nuclear after heat shock (Fig. 4S). These results
are consistent with previous findings using this antibody
(Welch and Feramisco, 1984) and confirm that the cells were
Fig. 2. The location of HsHSF1, as determined by biochemical
fractionation, is dependent on both the state of stress in the cell and
the fractionation procedure used. Unshocked (37°C) and heat-
shocked (42°C, 15 minutes) HeLa cells were fractionated either by
(A) an NP-40 based, or (B) a modified Dignam fractionation
procedure. Fractions were examined for the presence of HSF1 by
western blot analysis using crude anti-MBP+HsHSF1 (PM95-1;
1:2000). The same samples were also examined for the presence of a
known nuclear (histone H1) and cytoplasmic (IκBα) protein. C,
cytoplasmic fraction; N, nuclear fraction; IN, nuclear insoluble
Fig. 3. Active HSF1 fractionates to the nuclear compartment.
Unshocked (37°C) and heat-shocked (42°C, 15 or 180 minutes)
HeLa cells were fractionated either by (A) an NP-40 based, or (B)
the modified Dignam fractionation procedure. Fractions were
examined for the presence of active HSF1 using gel mobility-shift
2770 P. A. Mercier and others
Fig. 4. Human HSF1 is a nuclear protein before and after stress in HeLa cells. Cells were either kept in non-shock conditions (37°C) or were
heat shocked (42°C, 15 minutes), fixed, then were incubated with one of two different anti-human HSF1 antibodies: affinity-purified PM95-1
(F,I) or SR191 (L,O), followed by staining with a fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit secondary antibody.
Fluorescence staining due to the secondary antibody alone is shown in (C). DNA was visualized by staining with Hoechst 33342 (B,E,H,K,N).
Whole cells were visualized using Hoffman modulation contrast microscopy (A,D,G,J,M). Other cells were incubated with an anti-hsp70
antibody (SPA-10) followed by staining with a rhodamine-conjugated secondary antibody (Q,S).
Fig. 5. HSF1 is a nuclear protein before and after stress in human A549 cells. Human A549 cells were kept in non-shock conditions (37°C) or
were heat shocked (42°C, 15 minutes), fixed, then stained with one of two different anti-human HSF1 antibodies, as described in the legend of
2771HSF1 is a nuclear protein in unstressed cells
indeed heat shocked and that our fixation technique worked for
both nuclear and cytoplasmically localized proteins.
We examined the distribution of HSF1 before and after
heat shock using image analysis software (Northern
Exposure) on digitally captured images. This analysis was
performed on cells stained in three separate experiments, all
giving essentially the same results. The results of one
such analysis are shown in Table 1. Several different
measurements and calculations were made to try and address
three major points. First, what proportion of the fluorescence
observed is due to HSF1? Second, what is the amount of
HSF1 staining in the nucleus relative to the cytoplasm? Third,
does either the total amount or the distribution of HSF1
staining change when comparing unshocked to heat shocked
cells? To determine the amount of fluorescence that can be
attributed to HSF1, the total amount of fluorescence for each
nucleus and each cytoplasm was determined and the average
of 40 to 50 cells was calculated. From this value, the
fluorescence due to the FITC secondary antibody alone (no
anti-HSF1 antibodies) was subtracted. Secondary antibody
staining alone produced a uniform staining pattern over the
entire cell in both unshocked (Fig. 4C) and heat-shocked cells
(not shown). Approximately 29% of the total nuclear
fluorescence could be attributed to the secondary antibody in
both unshocked and heat-shocked cells (see Table 1).
Approximately 63% of the total cytoplasmic fluorescence
could be attributed to the secondary antibody in both
unshocked and heat-shocked cells.
Unshocked cells appeared to have slightly more HSF1
staining than heat-shocked cells (186 versus 167 units), but this
difference is well within the standard deviation of each value.
The large standard deviations observed in the fluorescence in
both the nucleus and cytoplasm appears to be mostly due to
the large variation in nuclear and cytoplasmic size from cell to
cell. The average amount of HSF1 staining per unit area of
nucleus and cytoplasm was almost identical between control
and heat-shocked cells.
The proportion of total HSF1 in the nucleus was essentially
the same in unshocked and heat-shocked cells (78% and 79%,
respectively). These results indicate that most (approximately
80%) of the HSF1 appears to be in the nucleus before and after
heat shock. The remaining 20% of the staining attributed to
HSF1 appears to be localized throughout the cytoplasm and
does not appear to be translocated into the nucleus upon heat
Fig. 6. HSF1 is a nuclear protein before and after stress in monkey Vero cells. Vero cells were kept in non-shock conditions (37°C) or were heat
shocked (42°C, 15 minutes), fixed, then stained with one of two different anti-human HSF1 antibodies, as described in the legend of Fig. 4.
Table 1. Subcellular distribution of HSF1 in unshocked
and heat-shocked HeLa cells
(42°C, 15 minutes)
Nuclear HSF1 staining
Total average fluorescence per nucleus1
Fluorescence due to secondary antibody
Average HSF1 staining per nucleus
(Total fluorescence − secondary
Average nuclear area (arbitrary units)
Average HSF1 staining/unit nuclear area3
Cytoplasmic HSF1 staining
Total average fluorescence per cytoplasm1
Fluorescence due to secondary antibody
Average HSF1 staining per cytoplasm
Average cytoplasmic area (arbitrary units)
Average HSF1 staining/ unit area of
Percentage of total HSF1 in nucleus4
1Total average fluorescence per nucleus (or cytoplasm) = sum of the
fluorescence measured in nucleus/total number of cells measured. The
fluorescence for each cell cytoplasm was measured by subtracting nuclear
fluorescence from total fluorescence for that cell. Values are ± s.d.
(unshocked, n=42 cells; heat shocked, n=50 cells).
2Fluorescence due to secondary antibody staining = average staining of
secondary antibody alone/unit cell area × average size of nucleus (or
cytoplasm). Values are ± s.d. (unshocked, n=7; heat shocked, n=5).
3Average HSF1 staining/unit area of nucleus (or cytoplasm) = average
HSF1 staining per nucleus (or cytoplasm) / average size of nucleus (or
4Percentage of total HSF1 in nucleus = average HSF1 staining per nucleus
× 100 / (average HSF1 staining per nucleus + average HSF1 staining per
Regulation of mammalian hs genes by HSF1 has been
postulated to involve translocation of HSF1 from the cytoplasm
to the nucleus (Sarge et al., 1993; Baler et al., 1993; Sistonen
et al., 1994; Morimoto, 1998). Some of the early studies
determining the subcellular localization of HSF1 involved
biochemical fractionation of mammalian cells followed by
western blot analysis (Baler et al., 1993; Sistonen et al., 1994).
These studies suggested that HSF1 was a cytoplasmic protein
that translocated from the cytoplasm to the nucleus with heat
stress. Using a similar western blot analysis of biochemically
fractionated mammalian cells, we obtained essentially the
same results. That is, after biochemical fractionation, HsHSF1
from unshocked cells was found in the cytoplasmic fraction,
while after heat shock, HsHSF1 was found predominantly in
the nuclear fraction. However, the difference in the amount of
HsHSF1 found in the cytoplasm of unshocked cells differs
depending on which fractionation technique is used. Using a
modified Dignam extraction procedure, a large proportion of
the HSF1 was found in the nuclear fraction prior to heat shock.
Using an NP-40 based extraction procedure, there was almost
no detectable HSF1 in the nuclear fraction prior to heat shock.
Therefore, the distribution of HSF1 prior to heat shock is
highly dependent on the fractionation procedure used and
raises questions as to whether biochemical fractionation can
accurately determine the location of HSF1.
A few studies have examined the localization of mammalian
HSF1 by immunofluorescence microscopy. One study, using
an anti-mouse HSF1 polyclonal antisera, has shown that in
mouse 3T3 and in human HeLa cells, HSF1 is predominantly
cytoplasmic before heat shock and predominantly nuclear after
heat shock (Sarge et al., 1993). Other studies, which have used
an anti-human HSF1 antibody (SR191), have shown that in
HeLa cells, HSF1 is found almost exclusively in the nucleus
before and after heat shock (Wu et al., 1994; Martinez-Balbas
et al., 1995). An exception to this localization pattern was
observed in mitotic cells where HSF1 was found to be
distributed throughout the entire cell and excluded from
condensed chromosomes (Martinez-Balbas et al., 1995). One
of the cell types examined in all the studies was human HeLa
cells, indicating that the observed differences are likely due to
differences in the two antibodies used and not to different cells
being examined. Using a different affinity-purified antibodies
against HsHSF1, we obtained results that were vitually
identical to those obtained by the latter studies. That is, HSF1
was found to be predominantly a nuclear protein in unshocked
and heat-shocked interphase cells.
A factor that might contribute to the different results
obtained with the different anti-HSF1 antibodies is the
preparation of the HSF1 protein used as antigen. Some studies
(e.g. Sarge et al., 1993) used bacterially expressed mouse HSF1
extracted from a denaturing SDS gel as antigen, while others
(Wu et al., 1994; Martinez-Balbas et al., 1995) used human
HSF1 that was purified by chromatography and was not
denatured prior to introduction into rabbits. The antigen
produced in our study was not denatured prior to injection. This
raises the possibility that antibodies raised against
nondenatured HSF1 may not recognize all forms, and in
particular, the inactive monomeric form of HSF1. We do not
believe this to be the case because the amount of nuclear HSF1
staining, as measured by fluorescence quantification of
immunostained cells, was the same before and after heat shock,
indicating that the anti-HSF1 antiserum recognized unshocked
HSF1 with the same efficiency as shocked HSF1.
Studies by Cotto et al. (1997) and Jolly et al. (1997) have
examined the subnuclear localization of mammalian HSF1 by
fluorescence microscopy before and after heat and other
stresses. Using a green fluorescent protein-HSF1 fusion protein
as well as monoclonal antibodies that recognize either HSF1
directly or an HSF1-Flag tagged fusion protein, it was found
that HSF1 stained the entire nucleus during heat shock (except
for the nucleolus) (Cotto et al., 1997). In addition, HSF1
formed bright irregularly shaped granules whose appearance
and disappearance roughly
transcriptional activity. In agreement with our current findings,
both natural HSF1 and the transfected HSF1 fusion proteins
appeared to be predominantly nuclear in unstressed cells (Cotto
et al., 1997; Jolly et al., 1997).
The behaviour we observe for human HSF1 in this study is
identical to what has been previously observed for Drosophila
HSF (Westwood et al., 1991; Wu et al., 1994; Wisniewski et
al., 1996; Orosz et al., 1996). That is, biochemical fractionation
of Drosophila SL2 tissue culture cells showed that HSF
appeared to be a cytoplasmic protein before heat shock and a
nuclear protein after heat shock (Wu et al., 1994).
Immunofluorescence microscopy revealed that Drosophila
HSF was a nuclear protein before and after heat shock
(Westwood et al., 1991). This conclusion was based on
experiments using three different polyclonal antisera made
against Drosophila HSF1 and two types of Drosophila cells:
Schneider line 2 (SL2) and the salivary gland cells of
Drosophila melanogaster third instar larvae. Two of the
polyclonal antisera were made against bacterially produced
and purified Drosophila HSF (Westwood et al., 1991; Wu et
al., 1994; Wisniewski et al., 1996; J. T. Westwood and C. Wu,
unpublished) while the third was made against a peptide
sequence found within HSF (Orosz et al., 1996). A fourth
polyclonal antisera against Drosophila HSF immunostained
polytene chromosomes of salivary glands in a similar fashion
(Shopland et al., 1995; Shopland and Lis, 1996). In Drosophila
embryos, it has been found that HSF is excluded from the
nucleus in both unshocked and heat-shocked embryos until
stage 13, at which time it becomes a nuclear protein in both
unshocked and heat-shocked embryos (Wang and Lindquist,
1998). The redistribution of HSF correlates with the embryo’s
ability to induce heat-shock gene transcription, i.e. hsp70
transcription only occurs when HSF is a nuclear protein and
when embryos are heat shocked.
When biochemically fractionated nuclei from heat-shocked
Drosophila SL2 cells were examined by immunofluorescence
microscopy, they stained brightly with anti-HSF1 antibody,
whereas nuclei from unshocked cells had virtually no HSF (Wu
et al., 1994). Consistent with these results, it has previously
been shown, using immunofluorescence microscopy, that
unshocked HeLa cells extracted with a buffer containing NP-
40 no longer showed HSF1 staining in the nucleus (or any other
location), whereas HSF1 in heat-shocked cells was resistant to
NP-40 extraction and was still present in the nucleus
(Martinez-Balbas et al., 1995).
A study by Orosz et al. (1996) has identified the amino acid
sequence responsible for HSF nuclear localization in
correlated with HSF1
P. A. Mercier and others
2773HSF1 is a nuclear protein in unstressed cells
Drosophila, and this sequence conformed to a nuclear
localization motif found in other nuclear proteins. Mutant
HSF1 molecules lacking this sequence localize to the
cytoplasm rather than the nucleus in both unshocked and heat-
shocked cells when examined by immunofluorescence
microscopy, while wild-type HSF was nuclear before and after
heat shock (Orosz et al., 1996).
Not all studies examining the localization of Drosophila
HSF have given the same results. Using immunofluorescence
microscopy, Zandi et al. (1997) concluded that Drosophila
HSF is a cytoplasmic protein prior to heat shock and a nuclear
protein after heat shock. It is not clear why there is such a
discrepancy in the localization results but it is most likely due
to the different antibodies being used, since the cell line that
was examined (SL2) was apparently the same in both studies.
Further experimentation with more cell types and
quantification of the staining should help clarify the issue.
It is intriguing that one of the heat shock proteins, hsp27,
has localization and biochemical fractionation properties
similar to HSF. Immunofluorescence microscopy suggested
that, like HSF, hsp27 was predominantly a nuclear protein in
Drosophila S3 cells that were untreated, heat shocked, heat
shocked and allowed to recover at control temperatures, or
treated with ecdysone (Beaulieu et al., 1989). When cells were
biochemically fractionated using an extraction buffer
containing NP-40, hsp27 was found in the detergent insoluble
fraction of heat-shocked cells whereas in recovering and
ecdysone-treated cells, hsp27 was found in the detergent-
soluble fraction. Other proteins have also been shown to have
different solubilities after heat shock. Michels et al. (1995)
demonstrated that both cytoplasmic and nuclear localized
luciferase expressed in mammalian cells were detergent (Triton
X-100)-soluble in unstressed cells but were detergent-insoluble
after heat shock.
Apparent discrepancies between biochemical fractionation
and immunomicroscopy results have also been observed for
other transcription factors. For example, it was believed for
many years that unliganded (non-DNA binding) steroid
receptors were cytoplasmic while the liganded DNA binding
form of the receptor was nuclear. These conclusions were
based primarily on the location of the receptor after standard
biochemical fractionation of cells (Perrot-Applanat et al.,
1992). The availability of specific monoclonal antibodies as
well as different fractionation procedures has shown that the
estrogen receptor, and in fact all steroid receptors except the
glucocorticoid receptor, are predominantly nuclear proteins at
all times (King and Greene, 1984; Welshons et al., 1984;
Perrot-Applanat et al., 1992).
We have demonstrated
microscopy that for several cell types, mammalian HSF1 is
predominantly a nuclear protein in unshocked and heat-shocked
cells. This observation differs from results obtained using
biochemical fractionation procedures where HSF1 is found
mostly in the cytoplasmic fraction of unshocked cells. However,
because the amount of HSF1 fractionating to the ‘cytoplasm’ is
dependent on the make-up of the buffer being used, we argue
that most, if not all, of the HSF1 seen in the cytoplasmic fraction
is probably an artifact of the fractionation procedure. Thus,
monomeric HSF1 ‘leaches out’ or is extracted from the nucleus
during fractionation, whereas trimeric HSF1 remains tightly
associated with the nucleus (DNA) during extraction. We have
previously demonstrated that Xenopus HSF1 is a nuclear protein
in unshocked (and heat-shocked) mature oocytes (Mercier et al.,
1997). In that study, nuclei were physically dissected from the
oocyte under oil which prevented any possible ‘leaching’ of
HSF1. The results of this and other studies (Westwood et al.,
1991; Wu et al., 1994; Martinez-Balbas et al., 1995; Orosz et al.,
1996; Mercier et al., 1997) demonstrate that Drosophila,
Xenopus, and mammalian HSF1 appear to be predominantly
nuclear proteins before and after heat shock. This observation
implies that the evolutionarily conserved mechanism designed
to detect heat and other environmental responses may be a
nuclear response with HSF1 responding to stress-induced
changes occurring within the nucleus.
We thank Gail Peskleway and Mary Sopta for their contributions
to the production of the human HSF1 antibodies and S. Rabindran and
C. Wu for the SR191 antibody. We also thank N. Ovsenek for
comments on the manuscript. This work was supported by a grant
from the Medical Research Council of Canada to J.T.W.
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G.,
Smith, J. A. and Struhl, K. (1995). In Current Protocols in Molecular
Biology (ed. K. Janssen), p. 10.2.30. John Wiley and Sons, USA.
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,
Beaulieu, J. F., Arrigo, A. P. and Tanguay, R. M. (1989). Interaction of
Drosophila 27,000 Mrheat-shock protein with the nucleus of heat-shocked
and ecdysone-stimulated culture cells. J. Cell Sci. 92, 29-36.
Chen, Y., Barlev, N. A., Westergaard, O. and Jakobsen, B. K. (1993).
Identification of the C-terminal activator domain in yeast heat shock factor:
independent control of transient and sustained transcriptional activity.
EMBO J. 12, 5007-5018.
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.
Cotto, J. J., Kline, M. and Morimoto, R. I. (1996). Activation of Heat Shock
Factor 1 binding precedes stress-induced serine phosphorylation. J. Biol.
Chem. 271, 3355-3358.
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-
Fernandes, M., O’Brien, T. and Lis, J. T. (1994). Structure and regulation
of heat shock gene promoters. In The Biology of Heat Shock Proteins
and Molecular Chaperones (ed. R. I. Morimoto, A. Tissières and C.
Georgopoulos), pp. 375-393. Cold Spring Harbor Press, Cold Spring
Harlow, E. and Lane, D. (1988). Antibodies, a Laboratory Manual. Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Jolly, C., Morimoto, R. I., Robert-Nicoud, M. and Vourc’h, C. (1997).
HSF1 transcription factor concentrates in nucler foci during heat shock:
relationship with transcription sites. J. Cell Sci. 110, 2935-2941.
King, W. J. and Greene, G. L. (1984). Monoclonal antibodies localize
oestrogen receptor in the nuclei of target cells. Nature 307, 745-747.
Larson, J. S., Schuetz, T. J. and Kingston, R. E. (1988). Activation in vitro
of sequence-specific DNA binding by a human regulatory factor. Nature
Latchman, D. S. (1995). Eukaryotic Transcription Factors, 2nd edn., pp. 292-
308. Academic Press, San Diego, CA.
Lee, K. A., Zerivitz, K. and Akusjarvi, G. (1994). Small-scale preparation
of nuclear extracts from mammalian cells. In Cell Biology, A Laboratory
Handbook (ed. J. E. Celis), pp. 668-673. Academic Press, San Diego, CA.
Lyck, R., Harmening, U., Hohfeld, I., Treuter, E., Scharf, K.-D. and Nover,
L. (1997). Intracellular distribution and identification of the nuclear
localization signals of two plant heat-stress transcription factors. Planta 202,
2774 Download full-text
Mager, W. H. and De Kruijff, A. J. J. (1995). Stress-Induced Transcriptional
Activation. Microbiol. Rev. 59, 506-531.
Martinez-Balbas, M. A., Dey, A., Rabindran, S. K., Ozato, K. and Wu, C.
(1995). Displacement of sequence specific transcription factors from mitotic
chromatin. Cell 83, 29-38.
Mercier, P. A., Foksa, J., Ovsenek, N. and Westwood, J. T. (1997). Xenopus
heat shock factor 1 is a nuclear protein before heat stress. J. Biol. Chem.
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.
Michels, A. A., Nguyen, V. T., Konings, A. W., Kampinga, H. H. and
Bensaude, O. (1995). Thermostability of a nuclear-targeted luciferase
expressed in mammalian cells. Destabilizing influence of the intranuclear
microenvironment. Eur. J. Biochem. 234, 382-389.
Morimoto, R. I., Tissières, A. and Georgopoulos, C. (1994). Progress and
perspectives on the biology of heat shock proteins and molecular
chaperones. In The Biology of Heat Shock Proteins and Molecular
Chaperones (ed. R. I. Morimoto, A. Tissières and C. Georgopoulos), pp. 1-
30. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
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.
Nakai, A. and Morimoto, R. I. (1993). Characterization of a novel chicken
heat shock transcription factor, heat shock factor 3, suggests a new
regulatory pathway. Mol. Cell. Biol. 13, 1983-1997.
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.
Nakai, A., Tanabe, M., Kawazoe, Y., Inazawa, J., Morimoto, R. I. and
Nagata, K. (1997). HSF4, a new member of the human heat shock factor
family which lacks properties of a transcriptional activator. Mol. Cell. Biol.
Orosz, A., Wisniewski, J. and Wu, C. (1996). Regulation of Drosophila heat
shock factor trimerization: global sequence requirements and independence
of nuclear localization. Mol. Cell. Biol. 16, 7018-7030.
Parsell, D. A. and Lindquist, S. (1994). Heat shock proteins and stress
tolerance. In The Biology of Heat Shock Proteins and Molecular Chaperones
(ed. R. I. Morimoto, A. Tissières and C. Georgopoulos), pp. 457-494. Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Perrot-Applanat, M., Guiochon-Mantel, A. and Milgrom, E. (1992).
Immunolocalization of steroid hormone receptors in normal and tumour
cells: mechanisms of their cellular traffic. Cancer Surv. 14, 5-30.
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.
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,
Sarge, K. D., Zimarino, V., Holm, K., Wu, C. and Morimoto, R. I. (1991).
Cloning and characterization of two mouse heat shock factors with
distinct inducible and constitutive DNA-binding ability. Genes Dev. 5,
Sarge, K. D., Murphy, S. P. and Morimoto, R. I. (1993). Activation of heat
shock gene transcription by heat shock factor 1 involves oligomerization,
acquisition of DNA-binding activity, and nuclear localization and can occur
in the absence of stress. Mol. Cell. Biol. 13, 1392-1407.
Sarge, K. D., Park, S. O., 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.
Shopland, L. S., Hirayoshi, K., Fernandes, M. and Lis, J. T. (1995). HSF
access to heat shock elements in vivo depends critically on promoter
architecture defined by GAGA factor, TFIID, and RNA polymerase II
binding sites. Genes Dev. 9, 2756-2769.
Shopland, L. S. and Lis, J. T. (1996). HSF recruitment and loss at most
Drosophila heat shock loci is coordinated and depends on proximal
promoter sequences. Chromosoma 105, 158-171.
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-
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.
Sorger, P. K. and Pelham, H. R. (1988). Yeast heat shock factor is an essential
DNA-binding protein that exhibits temperature-dependent phosphorylation.
Cell 54, 855-864.
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.
Voellmy, R. (1994). Transduction of the stress signal and mechanisms of
transcriptional regulation of heat shock/stress protein gene expression in
higher eukaryotes. Crit. Rev. Eukaryotic Gene Expression 4, 357-401.
Wang, Z. and Lindquist, S. (1998). Developmentally regulated nuclear
transport of transcription factors in Drosophila embryos enable the heat
shock response. Development 125, 4841-4850.
Welch, W. J. and Feramisco, J. R. (1984). Nuclear and nucleolar localization
of the 72,000 dalton heat shock protein in heat shocked mammalian cells.
J. Biol. Chem. 259, 4501-4513.
Welshons, W. V., Lieberman, M. E. and Gorski, J. (1984). Nuclear
localization of unoccupied oestrogen receptors. Nature 307, 747-749.
Westwood, J. T., Clos, J. and Wu, C. (1991). Stress-induced oligomerization
and chromosomal relocalization of heat-shock factor. Nature 353, 822-827.
Westwood, J. T. and Wu, C. (1993). Activation of Drosophila heat shock
factor: conformational change associated with a monomer-to-trimer
transition. Mol. Cell. Biol. 13, 3481-3486.
Whiteside, S. T. and Goodbourn, S. (1993). Signal transduction and nuclear
targeting: regulation of transcription factor activity by subcellular
localisation. J.Cell Sci. 104, 949-955.
Winegarden, N. A., Wong, K., Sopta, M. and Westwood, J. T. (1996).
Sodium salicylate decreases intracellular ATP, induces both heat shock
factor binding and chromosomal puffing, but does not induce hsp 70 gene
transcription in Drosophila. J. Biol. Chem. 271, 26971-26980.
Wisniewski, J., Orosz, A., Allada, R. and Wu, C. (1996). The C-terminal
region of Drosophila heat shock factor (HSF) contains a constitutively
functional transactivation domain. Nucl. Acids Res. 24, 367-374.
Wu, C. (1995). Heat Shock Transcription Factors: Structure and Regulation.
Annu. Rev. Cell Dev. Biol. 11, 441-469.
Wu, C., Clos, J., Giorgi, G., Haroun, R. I., Kim, S.-J., Rabindran, S. K.,
Westwood, J. T., Wisniewski, J. and Yim, G. (1994). Structure and
Regulation of Heat Shock Transcription Factor. In The Biology of Heat
Shock Proteins and Molecular Chaperones (ed. R. I. Morimoto, A. Tissières
and C. Georgopoulos), pp. 395-416. Cold Spring Harbor Laboratory Press,
Cold Spring Harbor, NY.
Zandi, E., Tran, T.-N. T., Chamberlain, W. and Parker, C. S. (1997).
Nuclear entry, oligomerization, and DNA binding of the Drosophila heat
shock transcription factor are regulated by a unique nuclear localization
sequence. Genes Dev. 11, 1299-1314.
Note added in proof
Brown and Rush (Brain Res. 821, 333-340 (1999)) have
recently demonstrated using immunocytochemistry that HSF1
is predominantly localized in the nucleus of unstressed and
heat stressed neuronal and glial cell types in postnatal rat
P. A. Mercier and others