2367
Research Article
Overexpression of HSP70 inhibits cofilin
phosphorylation and promotes lymphocyte migration
in heat-stressed cells
Jodie P. Simard*, Danielle N. Reynolds, Alan P. Kraguljac, Graham S. T. Smith and Dick D. Mosser‡
Department of Molecular and Cellular Biology, University of Guelph, Guelph, ON N1G 2W1, Canada
*Present address: The Hospital for Sick Children, Program in Genetics and Genome Biology, University of Toronto, Toronto, ON M5G 1L7, Canada
‡
Author for correspondence ([email protected])
Journal of Cell Science
Accepted 28 March 2011
Journal of Cell Science 124, 2367-2374
© 2011. Published by The Company of Biologists Ltd
doi:10.1242/jcs.081745
Summary
Hyperthermia adversely affects cell structure and function, but also induces adaptive responses that allow cells to tolerate these
stressful conditions. For example, heat-induced expression of the molecular chaperone protein HSP70 can prevent stress-induced cell
death by inhibiting signaling pathways that lead to apoptosis. In this study, we used high-resolution two-dimensional gel electrophoresis
and phosphoprotein staining to identify signaling pathways that are altered by hyperthermia and modulated by HSP70 expression. We
found that in heat-shocked cells, the actin-severing protein cofilin acquires inhibitory Ser3 phosphorylation, which is associated with
an inhibition of chemokine-stimulated cell migration. Cofilin phosphorylation appeared to occur as a result of the heat-induced
insolubilization of the cofilin phosphatase slingshot (SSH1-L). Overexpression of HSP70 reduced the extent of SSH1-L insolubilization
and accelerated its resolubilization when cells were returned to 37°C after exposure to hyperthermia, resulting in a more rapid
dephosphorylation of cofilin. Cells overexpressing HSP70 also had an increased ability to undergo chemotaxis following exposure to
hyperthermia. These results identify a critical heat-sensitive target controlling cell migration that is regulated by HSP70 and point to
a role for HSP70 in immune cell functions that depend upon the proper control of actin dynamics.
Key words: Cofilin, HSP70, Heat shock, Slingshot phosphatase SSH1-L, LIM kinase, Cell migration
Introduction
A fundamental problem faced by cells of all organisms is the
ability to maintain vital functions when exposed to cellular stress.
Protein-damaging stresses have the capacity to disrupt cell structure,
metabolism and signaling pathways controlling cell proliferation
and cell death (Morimoto, 2008). Mild heat stress, which occurs in
humans during infection or heat stroke, enhances immunological
function through regulation of immune cell migration, proliferation,
maturation and cytokine secretion (Park et al., 2005). More extreme
exposures to elevated temperature can initiate apoptosis as a means
to remove irreversibly damaged cells. However, prior exposure to
mild heat stress induces an adaptive response that allows cells to
survive these normally lethal conditions. This adaptive response
occurs in part through the induced synthesis of the heat-shock
proteins, which include several families of molecular chaperones
(Young et al., 2004). The major stress-inducible heat-shock protein,
HSP70 is a potent anti-apoptotic factor (Mosser and Morimoto,
2004). For example, HSP70 can blunt the apoptotic signal by
preventing the heat-induced activation of the pro-apoptotic protein
Bax thereby blocking mitochondrial disruption and release of
caspase-activating proteins from the intermembrane space
(Stankiewicz et al., 2005). Stress kinases have been implicated in
the signaling of apoptotic pathways. Hyperthermia, and other
protein-damaging stresses activate JNK, and this activation is
inhibited in cells that overexpress HSP70 (Gabai et al., 1997;
Mosser et al., 1997). Inhibition by HSP70 expression is achieved
by preventing the heat-induced inactivation of JNK phosphatases
(Meriin et al., 1999). ERK is also activated in heat-stressed cells
by the aggregation and insolubilization of ERK phosphatases
(Yaglom et al., 2003). HSP70 prevents the activation of ERK by
protecting the ERK phosphatases MKP-1 and MKP-3 from heat
inactivation.
In this study, we attempted to identify additional signaling
pathways that are altered by hyperthermia and modulated by HSP70
expression. We used high-resolution two-dimensional gel
electrophoresis and phosphoprotein staining to identify proteins
that are differentially phosphorylated in cells exposed to
hyperthermia in the presence or absence of HSP70. From this
analysis, we identified cofilin as a heat-induced phosphoprotein
and found that its phosphorylation was diminished in cells
overexpressing HSP70. Cofilin is an actin-binding protein that has
an important role in the regulation of actin dynamics (Bernstein
and Bamburg, 2010; Van Troys et al., 2008). Cofilin regulates actin
severing, stabilization and nucleation in a concentration-dependent
fashion. Cofilin-mediated severing of actin filaments at the leading
edge of motile cells controls the formation of lamellipodia, which
is essential for T-cell migration and cancer cell metastasis
(Burkhardt et al., 2008; Wang et al., 2007). Cofilin activity is
spatially regulated within the cell through the activity of kinases
including LIM kinase (LIMK) that inactivates it by phosphorylation
of Ser3, resulting in the stabilization of actin filaments (Bernard,
2007; Nishita et al., 2005; Scott and Olson, 2007). Cofilin
reactivation by specific (slingshot, chronophin) and general
phosphatases (PP2A) cause F-actin severing and depolymerization
(Huang et al., 2006). We found that hyperthermia caused an
insolubilization of slingshot, which correlated with an increased
level of cofilin phosphorylation and a reduced capacity for
chemokine-stimulated cell migration. Cells expressing HSP70
2368
Journal of Cell Science 124 (14)
maintained higher levels of soluble slingshot, had reduced levels
of cofilin phosphorylation and retained the ability to migrate when
exposed to hyperthermia.
Results
Journal of Cell Science
Identification of cofilin as a phosphorylated protein in
heat-shocked cells
Given that several kinases and phosphatases are modulated during
heat shock, we set out to identify proteins that show altered
phosphorylation in response to hyperthermia and for which HSP70
exerted some control over this change in phosphorylation status.
PErTA70 cells, a human acute lymphoblastic T-cell line with
tetracycline-regulated expression of HSP70, were induced with
doxycycline to overexpress HSP70 (ON) and compared with noninduced cells following a 60 minute heat shock at 43°C and with
control cells incubated at 37°C. Proteins were separated by highresolution two-dimensional gel electrophoresis, stained with Pro-Q
diamond phosphoprotein stain and imaged (Fig. 1). Gels were
counterstained with SYPRO Ruby total protein stain to ensure that
any increased spot intensity in the Pro-Q-stained gels was due to
increased phosphorylation and not due to increased protein
abundance. A number of spots could be seen that were more intensely
stained in heat-shocked cells compared with control cells. These
spots appeared to be less intensely stained following heat shock
when HSP70 was expressed, suggesting that HSP70 could suppress
the heat-induced phosphorylation of these proteins. We focused our
attention on a protein with an approximate pI of 8.5 and a molecular
mass of 20 kDa. This spot was consistently found to be highly
phosphorylated after heat-shock exposure in the non-induced cells
with a substantially reduced level of increased phosphorylation in
the induced cells (Fig. 1A, circled spot in bottom right corner). This
spot was excised from Coomassie-Blue-stained gels and protein
identification was obtained by MALDI-TOF mass spectrometry
analysis and peptide mass fingerprinting. Protein Prospector MS-Fit
database search identified this spot to be non-muscle cofilin (sequence
coverage was 42% with a MOWSE score of 360). To verify that the
spot of interest was indeed cofilin and that heat shock affected its
level of phosphorylation, control and heat-shocked (43°C for 1 hour)
cells were subjected to 2D PAGE and immunoblotted using an anticofilin antibody (Fig. 1B). A spot corresponding to a pI value of ~8
was detected in extracts from control cells, whereas an acidic shift
of spots was present in extracts from heat-shocked cells (indicative
of phosphorylation).
We wished to know whether the heat-induced phosphorylation
of cofilin also occurs in other cell types besides the acute
lymphoblastic T-cell line PEER. Cofilin phosphorylation status
was examined by western blotting of extracts from PEER, HCT116
(colorectal carcinoma) and HeLa (cervical carcinoma) cells that
were exposed to 43°C for 30 minutes to 2 hours using antibodies
against cofilin and phosphorylated cofilin (cofilin-P) (Fig. 1C).
The anti-cofilin-P antibody detects inactive Ser3-phosphorylated
cofilin. Heat-induced cofilin phosphorylation was seen in all three
of the cell lines, indicating that it is a general phenomenon;
however, the PEER cells appeared to be the most sensitive.
Analysis of heat-induced cofilin phosphorylation and its
inhibition by HSP70
To determine the temperature profile of cofilin phosphorylation
and to characterize the ability of HSP70 to inhibit this heat-induced
phosphorylation, we exposed non-induced and HSP70overexpressing PErTA70 cells to temperatures ranging from 39°C
to 43°C for 1 hour and examined protein phosphorylation levels by
immunoblotting (Fig. 2A). This temperature range represents the
extremes of a mild febrile response to severe hyperthermia such as
might occur during heat stroke. The lowest temperature at which
increased cofilin phosphorylation could be detected was 40°C in
the non-induced cells, with 42°C and 43°C producing significantly
higher levels of phosphorylation. The level of cofilin
phosphorylation was much less at each temperature in the HSP70overexpressing cells compared with the non-induced cells. These
results indicate that exposure to hyperthermia causes inhibitory
phosphorylation of cofilin and that this is prevented in cells
Fig. 1. Global changes in protein phosphorylation in cells exposed to hyperthermia in the presence or absence of overexpressed HSP70. (A)Representative
Pro-Q-stained 2D gels containing extracts from equivalent cell numbers of non-induced (OFF) and HSP70-overexpressing (ON) PErTA70 cells cultured at 37°C
(control) or exposed to 43°C for 1 hour (heat shock). The circled spot represents a protein that shows increased levels of phosphorylation in heat-shocked cells and
which has reduced levels of heat-induced phosphorylation in cells overexpressing HSP70. The identity of this protein was determined by MALDI-TOF-based
peptide mass fingerprinting of duplicate gels stained with Coomassie Blue to be human non-muscle cofilin (Acc. no. P23528). (B)Immunoblot analysis of extracts
from control (C) and heat-shocked (HS) cells separated by 2DE and probed with an anti-cofilin antibody. (C)Immunoblot analysis of extracts from PEER, HCT116
and HeLa cells exposed to 43°C for 30, 60, 90 or 120 minutes probed with antibodies against cofilin and phosphorylated cofilin (p-cofilin).
Journal of Cell Science
Heat-induced cofilin phosphorylation
Fig. 2. Analysis of heat-induced cofilin phosphorylation and its inhibition
by HSP70. (A)Immunoblots showing levels of phosphorylated cofilin, cofilin,
HSP70 and actin in non-induced (OFF) and HSP70-overexpressing (ON)
PErTA70 cells incubated for 1 hour at the indicated temperatures (°C).
(B)Immunoblots from cells exposed to 43°C for 1 hour and collected either
immediately (0) or after incubation at 37°C for 2, 4 or 6 hours.
(C)Immunoblots from cells exposed to 39°C for 2, 4, 6 or 8 hours. For
comparison, the effects of a 1 hour exposure to 43°C is shown for the noninduced and induced cells. Bar graphs show levels of phosphorylated cofilin in
the non-induced and induced cells plotted relative to the maximum level +
s.e.m. (n3). *P<0.05 in comparison of OFF vs ON (One-tailed t-test).
expressing HSP70. When the non-induced cells were exposed to
43°C for 1 hour and then incubated at 37°C, the level of cofilin
phosphorylation decreased over time, but by 6 hours of recovery,
remained at a level that was still 30% of the maximum (Fig. 2B).
In the HSP70-overexpressing cells, the same heat treatment resulted
in reduced levels of heat-induced cofilin phosphorylation and the
rate of cofilin dephosphorylation was also significantly more rapid.
We next examined whether extended exposure to the moderately
elevated temperature of 39°C could affect cofilin phosphorylation
(Fig. 2C). Exposure to 39°C for 2 hours or longer caused cofilin
phosphorylation, which slowly diminished with increasing exposure
time. However, unlike acute exposure to temperatures above 41°C,
HSP70 overexpression did not prevent cofilin phosphorylation
during continuous exposure to 39°C.
2369
Fig. 3. Examination of cofilin localization. (A)Confocal microscopic images
of non-induced (OFF) and HSP70-overexpressing (ON) PErTA70 cells
cultured at 37°C (C) or exposed to 43°C for either 60 or 120 minutes (HS).
Cells were cytospun onto glass slides, fixed and incubated with an anti-cofilin
antibody followed by an Alexa-Fluor-594-conjugated secondary antibody.
Arrowheads indicate regions of intense localized cofilin staining. Scale bar:
20m. (B)Digital z-stack reconstruction of cells immunostained for cofilin
following 120 minute heat-shock treatment demonstrating either the loss of the
spatially concentrated localization of cofilin to a more diffuse granular
distribution in the non-induced cells (left), or maintenance of the spatially
concentrated localization in the HSP70-overexpressing cells (right). Image sets
were colored with higher intensity voxels represented as warmer colors, and
lower intensity voxels as cooler colors. Cross-section analyses (z-plane is
indicated by yellow line) for comparing stacks illustrate spatial concentration
of cofilin at distinct regions near the cell periphery (arrows) in a subpopulation
of cells. Digital reconstruction image series were acquired at a sample rate of
0.5m and were processed and analyzed using ImageJ software (National
Institutes of Health), and were compiled using Adobe Photoshop CS3.
(C)Quantification of cells containing brightly stained cofilin at one end of the
cell (polarized). Shown is the mean + s.e.m. (n4); P<0.005 for OFF-HS vs
ON-HS and for C vs HS in both OFF and ON cells (One-tailed t-test). There is
no significant difference between the control cells for OFF vs ON.
Effect of hyperthermia on cofilin distribution in heatshocked cells
We next examined whether the heat-induced phosphorylation of
cofilin was associated with a change in the intracellular distribution
of cofilin because this might provide insight into its effect on cell
function. In the non-stressed cells, cofilin was distributed in a
granular fashion throughout the cell with a prominent brightly
staining accumulation at one end of the cell (Fig. 3A). Exposure to
hyperthermia (43°C for 60–120 minutes) led to a loss of this polarized
cofilin localization. Most cells retained the diffuse granular
distribution, but the brightly staining accumulation at one end of the
cell was no longer evident. The extent of this loss was not as severe
in the cells overexpressing HSP70 that were exposed to hyperthermia.
Using image sets consisting of 0.5-m-thick CSLM serial sections,
we compiled a three-dimensional reconstruction of the cell that
demonstrates the spatially concentrated cofilin distribution (Fig. 3B).
2370
Journal of Cell Science 124 (14)
This analysis clearly shows that cofilin retains the polarized
distribution in cells exposed to hyperthermia when HSP70 is present,
but not when it is absent. A similar analysis in cells co-stained for
cofilin and DAPI demonstrates that the polarized accumulation of
cofilin is cytoplasmic (Fig. 3B). We quantified the percentage of
cells with bright-staining polarized cofilin (Fig. 3C) and found a
significant level of protection against the loss of these polarized
cofilin structures in the HSP70-overexpressing cells (ON, 44%
polarized vs OFF, 9% polarized).
Journal of Cell Science
Effect of hyperthermia on cell migration
Because exposure to hyperthermia resulted in cofilin
phosphorylation, which was associated with a loss of polarized
cofilin distribution, we next investigated whether heat-shock
treatment altered the ability of cells to migrate towards a chemokine
source. Chemotaxis was measured using Transwell chambers
containing 10–8 M stromal cell derived factor-1  (SDF-1) in the
lower chamber (Fig. 4A). Control and heat-shocked cells (43°C for
1 hour) were added to the upper wells of individual Transwell
chambers and incubated at 37°C for 2 hours. The migration index
(number of cells in the lower chamber as a percentage of the total
number of cells added to the upper chamber) was severely reduced
in the non-induced cells that were exposed to hyperthermia.
However, similarly to the results observed for cofilin distribution,
HSP70-overexpressing cells displayed a significant level of
protection against the heat-induced inhibition of cell migration
(ON, 19% vs OFF, 5.5%). These results suggest that HSP70 is able
to maintain a migration-competent state in cells exposed to
hyperthermia and that this competency is correlated with the
maintenance of a polarized distribution of cofilin.
We next examined the effect of exposure to 39°C on cell
migration because this treatment was found to increase the level of
cofilin phosphorylation in both control and HSP70-overexpressing
cells (Fig. 2C). In this experiment, the non-induced and induced
cells were mixed at a 1:1 ratio, exposed to hyperthermia and then
assayed for their ability to undergo chemotaxis. For the 39°C
treatment, cells were incubated at 39°C for 2 hours and then the
cells were allowed to migrate through Transwell chambers for 2
hours at 39°C. This was compared with cells that were incubated
at 43°C for 1 hour and assayed for migration at 37°C for 2 hours
and to control cells in which non-treated cells were assayed for
migration at 37°C for 2 hours (Fig. 4B). The results show that
lymphocyte migration is significantly impaired at 39°C relative to
the 37°C control cells. To determine whether HSP70 overexpression
protected the cells from this impairment, we next determined the
percentage of GFP-positive cells before and after migration (Fig.
4C). The induced cells express both HSP70 and GFP from an
IRES-containing transcript and were 96% GFP positive before
mixing with the non-induced GFP-negative (and HSP70-negative)
cells. Although HSP70-overexpressing cells had an increased ability
to migrate at 37°C after a 1 hour exposure to 43°C, these cells
showed no increased ability to migrate when exposed to 39°C.
This provides further evidence that heat-induced cofilin
phosphorylation impairs lymphocyte cell chemotaxis because this
was seen after exposure to either 39°C or 43°C. In addition,
resistance to this impairment occurs only under conditions in which
HSP70 is able to prevent heat-induced cofilin phosphorylation.
Insolubilization of slingshot in heat-shocked cells
To investigate the mechanism of heat-induced cofilin
phosphorylation, we initially examined whether levels of
Fig. 4. Effect of hyperthermia on cell migration. (A)Inhibition of cell
migration following exposure to 43°C for 1 hour. Cells were allowed to
migrate for 2 hours through 5m pore polycarbonate Transwell chambers into
medium containing 10–8 M SDF. The migration index was calculated by
counting the number of cells in the bottom chamber and dividing this by the
number of cells in the original suspension. Migration was examined at 37°C in
non-induced and HSP70-expressing cells that were either maintained at 37°C
(C) or exposed to 43°C for 1 hour (HS). Shown is the mean + s.e.m. (n5),
P<0.015 for OFF-HS vs ON-HS, P<0.002 for OFF C vs OFF HS, P<0.01 for
ON C vs ON HS (One-tailed t-test). There is no significant difference between
the control cells for OFF vs ON. (B)Non-induced and HSP70-expressing cells
were mixed 1:1 and assayed for cell migration at 37°C (37), at 39°C after a 2
hour incubation at 39°C (39) or at 37°C after a 1 hour incubation at 43°C (43).
In each case, the cells were allowed to migrate for 2 hours. Shown is the mean
+ s.e.m. (n3). (C)The percentage of GFP-positive cells (induced cells coexpress GFP) was measured by flow cytometry for the experiment shown in B.
Shown is the mean + s.e.m. (n3). P<0.05 for 37 vs 39 or 43 in B and P<0.005
for total vs migrated for cells exposed to 43°C in C (One-tailed t-test).
phosphorylated LIMK (LIMK-P) were altered in cells exposed to
43°C. Surprisingly, we found that heat shock did not affect the
level of LIMK-P in either the non-induced or HSP70-expressing
cells. Because HSP70 is known to prevent heat-induced protein
aggregation, we next considered whether hyperthermia altered the
solubility of LIMK or the cofilin phosphatases slingshot (SSH1L), chronophin or PP2A. Cells were exposed to 43°C for 30–120
minutes and were then divided into two aliquots. One sample was
lysed in 1% Triton X-100 and insoluble proteins were removed by
centrifugation (Fig. 5A, soluble). The pellets were resolubilized in
an equal volume of Laemmli buffer (Fig. 5B, pellet). The other
sample was lysed in buffer containing 2% SDS to recover total
cellular proteins (Fig. 5A, total). Exposure to hyperthermia caused
a depletion of soluble SSH1-L, which is indicative of aggregation
(Fig. 5A,B). LIMK was also subject to heat-induced
insolubilization; however, this was less extensive in comparison
with that of SSH1-L, and importantly, the amount of active LIMKP remained relatively constant in the soluble fraction. Whereas
Heat-induced cofilin phosphorylation
2371
Journal of Cell Science
occurred in the non-induced cells, reflecting the reduced ability of
these cells to dephosphorylate cofilin during recovery from
hyperthermia. Total and LIMK-P levels decreased during incubation
at 37°C in the non-induced cells. This loss of LIMK most likely
results from caspase-3-mediated cleavage of LIMK (Tomiyoshi et
al., 2004). LIMK levels remained intact in the HSP70overexpressing cells because HSP70 inhibits heat-induced caspase
activation (Mosser et al., 1997).
Fig. 5. Hyperthermia causes a loss of soluble LIMK and SSH1-L. Noninduced (OFF) and HSP70-overexpressing (ON) PErTA70 cells were exposed
to 43°C for 30, 60, 90 or 120 minutes and then lysed in either 1% Triton X-100
buffer or Laemmli buffer. Control cells ‘C’ remained at 37°C.
(A)Immunoblots showing levels of the indicated proteins in Triton-soluble
fractions and in total cell lysates (Laemmli extract). (B)Comparison of the
Triton-soluble and -insoluble fractions. (C)Quantification of the percentage of
LIMK and SSH1-L that remained soluble in cells exposed to hyperthermia
compared with control non-stressed cells. Shown is the mean + s.e.m. (n3).
*P<0.05 for comparison of OFF vs ON (One-tailed t-test).
SSH1-L was very sensitive to heat-induced insolubilization,
chronophin and PP2A were not affected. In HSP70-overexpressing
cells, the extent of SSH1-L (and LIMK) insolubilization was greatly
reduced compared with that in non-induced cells (Fig. 5C). These
results suggest that the heat-induced phosphorylation of cofilin
occurs as a consequence of the inactivation of SSH1-L and that
HSP70 prevents cofilin phosphorylation in heat-shocked cells by
protecting SSH1-L from insolubilization.
We also examined SSH1-L insolubilization in cells that were
exposed to 43°C for 1 hour and then incubated at 37°C for 2–6
hours in an attempt to correlate changes in cofilin phosphorylation
with the resolubilization of SSH1-L (Fig. 6A). These results show
a resolubilization of SSH1-L in the HSP70-overexpressing cells,
which correlates with the rapid dephosphorylation of cofilin.
However, only a minimal extent of SSH1-L resolubilization
Discussion
Heat shock can affect the phosphorylation status of many proteins,
although the significance of these changes to cell function at
elevated temperatures has been documented for only a few of them
(Kim et al., 2002). Our phosphoprotein analysis led to the finding
that the actin-binding protein cofilin becomes phosphorylated and
inactivated in cells exposed to hyperthermia. This inactivation of
cofilin resulted in an altered cellular distribution of cofilin and an
inhibition of chemokine-stimulated cell migration. The mechanism
responsible for heat-induced cofilin phosphorylation appears to be
a temperature-dependent insolubilization of the cofilin phosphatase
slingshot (Fig. 7). Overexpression of HSP70 prevented the heatinduced phosphorylation of cofilin by reducing the extent of
slingshot insolubilization. This was accompanied by improved
polarized cellular distribution of cofilin and hence increased
chemotaxis ability.
Exposure to hyperthermia has long been known to affect
cytoskeletal structures (Welch and Suhan, 1985). Many studies
have shown that heat-shock proteins participate in the synthesis
and assembly of cytoskeletal components and can protect from
heat-induced damage (Brown et al., 1996; Liang and MacRae,
1997; Mounier and Arrigo, 2002). Our findings show that HSP70
provides protection of the actin cytoskeleton by preventing the
heat-induced inactivation of cofilin and that this prevented the
inhibitory effect of hyperthermia on cell migration. Our
immunofluorescence studies showed that cofilin is spatially
concentrated within non-stressed cells and that exposure to
hyperthermia leads to a loss of this spatial concentration, which
was correlated with an impaired chemotactic response. Both the
polarized localization of cofilin and the ability to undergo
chemotaxis were protected from the effects of hyperthermia in
cells overexpressing HSP70. These results suggest that heat shock
affects cofilin localization and migration competency by disrupting
the localized control of cofilin phosphorylation. Interestingly, cofilin
phosphorylation and cell migration could be affected by both mild
(39°C) and extreme (43°C) exposures to hyperthermia; however,
HSP70 overexpression only provided protection to cells exposed
to 43°C. It should be noted that migration was measured at 37°C
after a 1 hour exposure to hyperthermia for the cells exposed to
43°C, whereas for the cells exposed to 39°C, cell migration was
measured at 39°C after a 2 hour pre-incubation at 39°C. Therefore,
continuous exposure to 39°C impairs additional components of the
cell migration apparatus in addition to affecting cofilin
phosphorylation. The inhibitory effect of mild hyperthermia on
cell migration has also been reported in human mononuclear
leukocytes exposed to 38.5°C (Roberts and Sandberg, 1979).
Exposure to mild or severe hyperthermia has been shown to
result in activation of multiple signaling pathways (Park et al.,
2005). In some cases, the activation occurs through the direct
stimulation of these pathways by upstream signaling events,
whereas in others, kinase activation occurs as a result of
phosphatase inactivation (Meriin et al., 1999; Yaglom et al., 2003).
2372
Journal of Cell Science 124 (14)
Journal of Cell Science
Fig. 6. Recovery of soluble LIMK and SSH1-L in
HSP70-expressing cells. Non-induced (OFF) and
HSP70-expressing (ON) PErTA70 cells were exposed to
43°C for 60 minutes and collected immediately ‘0’ or
incubated at 37°C for 2, 4 or 6 hours. Control cells (C)
remained at 37°C. Cells were lysed in either 1% Triton
X-100 buffer or Laemmli buffer. (A)Immunoblots
showing levels of the indicated proteins in Triton-soluble
and -insoluble (pellet) fractions and in the total cell lysate
(Laemmli extract) in the OFF and ON cells.
(B)Quantification of the percentage of LIMK and SSH1L that remained soluble in cells exposed to hyperthermia
compared to control non-stressed cells. Shown is the
mean + s.e.m. (n3). *P<0.05 for comparison of OFF vs
ON (One-tailed t-test).
For example, the heat-induced activation of ERK has been shown
to be dependent upon inactivation of the dual-specificity
phosphatases MKP1 and MKP3 (Yaglom et al., 2003), whereas
JNK activation results from inactivation of JNK phosphatase M3/6
(Palacios et al., 2001). Aggregation and inactivation of these
phosphatases resulted in the accumulation of the phosphorylated
substrate through the activity of the constitutively active upstream
kinases. We propose that a similar mechanism exists for cofilin
phosphorylation. We found that although there was no increase in
the level of LIMK phosphorylation in cells exposed to
hyperthermia, there was a loss in the total levels of both soluble
SSH1-L and soluble LIMK, although SSH1-L was clearly more
heat sensitive than LIMK. In addition, LIMK can be
dephosphorylated and inactivated by SSH1-L (Soosairajah et al.,
2005) and therefore the insoluble LIMK could potentially represent
the pool of LIMK that was associated with SSH1-L and subject to
insolubilization along with the aggregating SSH1-L (see Fig. 5).
Nevertheless, in cells exposed to hyperthermia, sufficient LIMKP remained available to phosphorylate cofilin and in conjunction
with depleting levels of SSH1-L, led to the overall accumulation
of cofilin-P. This effect appeared to be specific to SSH1-L because
chronophin and PP2A remained soluble.
The overexpression of HSP70 not only reduced the extent of
SSH1-L insolubilization in cells exposed to hyperthermia, it also
increased the rate of recovery of soluble SSH1-L when heatstressed cells were returned to 37°C. This enhanced recovery of
soluble SSH1-L is probably responsible for the elevated rate of
cofilin-P dephosphorylation in the HSP70-overexpressing cells.
During this recovery period, there was a loss of total LIMK, which
is subject to caspase-3-mediated cleavage (Tomiyoshi et al., 2004).
Inhibition of caspase-3 activity by HSP70 (Mosser et al., 1997)
prevented this loss of LIMK. Consequently, the combined loss of
SSH1-L activity and the eventual decline in LIMK levels led to an
inability to regulate actin dynamics in heat-stressed cells in the
absence of HSP70, whereas in its presence, both SSH1-L and
LIMK retained their ability to act on cofilin. HSP70 could also
have a direct role in stabilizing LIMK activity. Both HSP70 and
HSP90 have been shown to interact with LIMK1, and in the case
of HSP90, this interaction promotes LIMK1 homo-dimerization
leading to an increase in its activity (Li et al., 2006; Lim et al.,
2007).
The ability of cofilin to affect actin dynamics endows it with the
ability to control numerous cellular functions including not only
cell migration, but also the regulation of apoptosis, lipid metabolism
and gene expression (Bernstein and Bamburg, 2010). Not
surprisingly, interference with its normal function is linked to
diseases such as cancer (Bamburg and Wiggan, 2002). Our results
suggest that the ability of HSP70 to modulate cofilin activity could
have effects on cofilin-dependent processes, such as T-cell
activation and migration, that could contribute to disease. HSP70
is often overexpressed in tumor cells, and evidence suggests that
its overexpression contributes to tumorigenesis (Calderwood et al.,
2006; Mosser and Morimoto, 2004). In the case of bladder cancer,
overexpression of the testis-specific protein HSP70-2 was
Fig. 7. Model illustrating the effect of hyperthermia on cofilin activity.
Cofilin is inactivated by phosphorylation of Ser3 by LIMK and activated by
dephosphorylation of this residue by the cofilin phosphatase slingshot (SSH1L). LIMK is activated by kinases downstream from Rac, Cdc42 and Rho and
inactivated by slingshot. The spatially coordinated activity of LIMK and
SSH1-L is required for regulating actin dynamics and controlling actindependent cellular processes, including cell migration. Heat shock (HS) causes
the insolubilization of SSH1-L, leading to the accumulation of phosphorylated
cofilin and inhibition of cell migration.
Heat-induced cofilin phosphorylation
associated with increased malignancy because specific knockdown
resulted in reduced migration and invasion abilities in vitro and
suppressed tumor growth in a xenograft assay (Garg et al., 2010).
Clinically, hyperthermia increases cancer patient survival when
used in combination with radiation or chemotherapy. Our results
suggest that this benefit might be due in part to the suppressive
effects of hyperthermia on cancer cell migration. However, this
beneficial effect would be limited in tumor cells with elevated
levels of HSP70, and therefore therapies targeting HSP70 might
need to be used in combination with hyperthermia to be effective.
Materials and Methods
Cells and treatments
The effects of HSP70 (NM_005345) overexpression were examined using a human
acute lymphoblastic T-cell line (PEER) with tetracycline-regulated expression of
human HSP70 (PErTA70) (Mosser et al., 2000). HSP70 was induced by incubation
with 1.0 g/ml doxycycline for 24 hours before each experiment. Cells were
maintained at 37°C in a humidified 5% CO2 incubator in RPMI medium with 10%
fetal bovine serum (Invitrogen, Burlington, Canada). Cells were heat shocked in
medium supplemented with 20 mM HEPES buffer (pH 7.2) by immersion of logphase cells in a circulating water bath. A humidified 5% CO2 incubator was used for
long-term exposure to 39°C.
Journal of Cell Science
Two-dimensional gel electrophoresis, phosphoprotein staining and
identification
For two-dimensional gel analysis, 1.0⫻107 cells were collected, washed with PBS
and then with 250 mM sucrose in 10 mM Tris-HCl (pH 7.5). Cell pellets were
resuspended on ice in lysis solution (7 M Urea, 2 M Thiourea, 4% CHAPS, 2% IPG
buffer (pH 4–7), 60 mM DTT, 200 mM sodium orthovanadate, 200 mM sodium
fluoride and 200 mM aprotinin) and then sonicated. Cell debris was pelleted by
centrifugation (14,000 g, 5 minutes, 4°C) and the supernatants were clarified by
precipitation using a 2-D Clean-Up kit (GE Healthcare, Baie d’Urfe, Quebec,
Canada). Precipitated proteins were resuspended in rehydration buffer (7 M Urea, 2
M Thiourea, 2% CHAPS, 0.5% IPG buffer (pH 4–7), 40 mM DTT) and applied to
IPG strips (11 cm, pI 4–7) for rehydration and first-dimension focusing using an
ETTAN IPGPhor apparatus (GE Healthcare). The first dimension was performed at
20°C for 45,000 V hours with a maximum current of 50 A. Focused strips were
equilibrated in 10 mg/ml DTT and 25 mg/ml iodoacetamide, rinsed in SDS
electrophoresis buffer and loaded onto a precast 13 cm 10.5–14% Criterion gel (BioRad, Mississauga, Canada) and resolved at constant power. SDS-PAGE gels were
fixed overnight in 10% TCA, 50% methanol, rinsed with water and stained with ProQ diamond phosphoprotein stain (Invitrogen). Gels were destained in 50 mM sodium
acetate (pH 4.0), 4% acetonitrile, rinsed in water and scanned using a Typbon 9400
imager using 532 nm excitation and 580 nm emission at a resolution of 200 m and
a PMT of 580 V (GE Healthsciences). Images were visualized using ImageQuant
V1.4 software (GE Healthsciences). Pro-Q stained gels were subsequently poststained with SYPRO Ruby total protein stain (Invitrogen), destained with 10%
methanol, 7% acetic acid, rinsed with water and imaged using 532 nm excitation and
610 nm emission.
For protein identification, the two-dimensional gels were stained with 0.1%
Coomassie Blue G250, 40% methanol, 10% acetic acid and destained overnight in
5% methanol, 7% acetic acid. Protein spots of interest were selected based on
differential spot staining of a duplicate set of gels with Pro-Q Diamond and Sypro
Ruby. Spots were manually excised and washed in 50 mM ammonium bicarbonate.
The slices were dehydrated using 50% acetonitrile, 25 mM ammonium bicarbonate
and then reduced by treatment with 10 mM DTT for 30 minutes at 56°C followed
by alkylation with 100 mM iodoacetamide at room temperature. The gel slices were
again dehydrated and then rehydrated in 0.12 g/l trypsin (Trypsin Gold, Promega,
Madison, WI) in 50 mM ammonium bicarbonate. Digestion was carried out at 37°C
for 16 hours. Peptides were collected by first transferring the supernatants to new
tubes followed by removal of residual digestion mixture from gel slices with 5%
formic acid treatment and a wash with acetonitrile. The samples were evaporated
and then purified using C18 Ziptips (Millipore) immediately prior to MALDI-TOF
mass spectrometry analysis (performed by the Advanced Protein Technology Centre
at the Hospital for Sick Children, Toronto, Canada).
Immunoblotting
Cell lysis and immunoblotting were performed as described previously (Stankiewicz
et al., 2005). To examine protein solubility following exposure to hyperthermia, cells
were lysed in Triton buffer (10 mM HEPES pH 7.4, 100 mM NaCl, 5 mM MgCl2,
1 mM EGTA, 1% Triton X-100, 10 g/ml each of pepstatin A, leupeptin and
aprotinin, 10 mM sodium fluoride, 1 mM sodium vanadate, 20 mM sodium phosphate,
3 mM -glycerophosphate, 5 mM sodium pyrophosphate). Lysates were centrifuged
at 15,000 g for 10 minutes at 4°C. The supernatants were mixed with 2⫻ Laemmli
buffer (100 mM Tris-HCl pH 6.8, 20% glycerol, 4% SDS, 10% -mercaptoethanol)
2373
and heated to 95°C for 5 minutes before electrophoresis. The 1% Triton X-100
pellets were washed with lysis buffer and then resuspended in an equal volume of
1⫻ Laemmli buffer. The following antibodies against the following proteins were
used for immunoblotting: actin (ACTN05: NeoMarkers, Fremont, CA), Caspase-3
(SA-320-0100; BIOMOL International, Plymouth Meeting, PA), Chronophin/PDXP
C85E3 (4686, Cell Signaling Technology), Cofilin (C8736, Sigma), Ser3-P Cofilin
(3311S, Cell Signaling Technology), HSP70 (C92F3A-5: Stressgen/Assay Designs,
Ann Arbor, MI), LIMK (3842, Cell Signaling Technology), LIMK-1/2 Thr508/505P (sc-28409-R, Santa Cruz Biotechnology, Santa Cruz, CA), PP2A catalytic subunit (610555, Sigma), Slingshot SSH1 (ab46202, Abcam, Cambridge, MA).
Cell-migration analysis
Chemotaxis was measured essentially as described (Vicente-Manzanares et al.,
1999). Cells were resuspended in RPMI medium containing 0.1% human serum
albumin (HSA, Sigma) at a concentration of 5⫻106 cells/ml. One hundred microliters
of the cell suspension was added to the upper chamber of a Costar Transwell cell
culture chamber (6.5 mm diameter, 10 m thickness, 5-m-diameter pore size
polycarbonate membrane). The lower chamber contained 600 l RPMI, 0.1% HSA
medium supplemented with 10–8 M recombinant human stromal cell derived factor1 (SDF-1, PeproTech, Dollard des Ormeaux, Quebec, Canada). Cells were allowed
to migrate at 37°C for 2 hours in a humidified 5% CO2 incubator. The number of
cells in the lower chamber and in the original cell suspension was counted using a
hemocytometer and the migration index was calculated [(no. cells per ml in lower
well/no. cells per ml in original cell suspension) ⫻100]. Flow cytometry (Beckman
Coulter FC500) was used to measure the percentage of GFP-positive cells in
migration assays containing a mixture of non-induced and induced cells (PErTA70
cells co-express GFP and HSP70 from a dicistronic expression cassette).
Immunofluorescence
Cells (100 l of a 2⫻105 cells/ml suspension) were collected onto glass slides
(Shandon cytoslide blue mask, Fisher Scientific, Markam, Ontario, Canada) in a
cytocentrifuge (Cytospin 4, Shandon Products), air-dried, fixed in 4%
paraformaldehyde, permeabilized with 0.2% Triton X-100 in PBS and then blocked
with 3% bovine serum albumin in PBS. The slides were incubated at room
temperature for 1 hour with anti-cofilin antibody (1:1000 dilution, C8736, Sigma),
washed with PBS and then incubated with an Alexa-Fluor-594-conjugated rabbit
secondary antibody (Invitrogen, Burlington, Ontario, Canada). Fluorescent images
were acquired using a Leica multiphoton laser-scanning confocal microscope (DM
6000B microscope connected to a TCS SP5 system). Images were processed using
ImageJ software (National Institutes of Heath), and were compiled using Adobe
Photoshop CS3.
We thank Julian Kwan and Alicia Soltys for their contribution of
preliminary data and Mhairi Skinner for her comments on the
manuscript. This work was supported by grants from the Natural
Sciences and Engineering Research Council of Canada (NSERC) and
the Canadian Institutes of Health Research (CIHR). G.S.T.S. is a
recipient of a Doctoral Studentship from the Multiple Sclerosis Society
of Canada.
References
Bamburg, J. R. and Wiggan, O. P. (2002). ADF/cofilin and actin dynamics in disease.
Trends Cell Biol. 12, 598-605.
Bernard, O. (2007). Lim kinases, regulators of actin dynamics. Int. J. Biochem. Cell Biol.
39, 1071-1076.
Bernstein, B. W. and Bamburg, J. R. (2010). ADF/cofilin: a functional node in cell
biology. Trends Cell Biol. 20, 187-195.
Brown, C. R., Hong-Brown, L. Q., Doxsey, S. J. and Welch, W. J. (1996). Molecular
chaperones and the centrosome. A role for HSP 73 in centrosomal repair following heat
shock treatment. J. Biol. Chem. 271, 833-840.
Burkhardt, J. K., Carrizosa, E. and Shaffer, M. H. (2008). The actin cytoskeleton in T
cell activation. Annu. Rev. Immunol. 26, 233-259.
Calderwood, S. K., Khaleque, M. A., Sawyer, D. B. and Ciocca, D. R. (2006). Heat
shock proteins in cancer: chaperones of tumorigenesis. Trends Biochem. Sci. 31, 164172.
Gabai, V. L., Meriin, A. B., Mosser, D. D., Caron, A. W., Rits, S., Shifrin, V. I. and
Sherman, M. Y. (1997). Hsp70 prevents activation of stress kinases. A novel pathway
of cellular thermotolerance. J. Biol. Chem. 272, 18033-18037.
Garg, M., Kanojia, D., Seth, A., Kumar, R., Gupta, A., Surolia, A. and Suri, A. (2010).
Heat-shock protein 70-2 (HSP70-2) expression in bladder urothelial carcinoma is
associated with tumour progression and promotes migration and invasion. Eur. J.
Cancer 46, 207-215.
Huang, T. Y., DerMardirossian, C. and Bokoch, G. M. (2006). Cofilin phosphatases and
regulation of actin dynamics. Curr. Opin. Cell Biol. 18, 26-31.
Kim, H. J., Song, E. J. and Lee, K. J. (2002). Proteomic analysis of protein
phosphorylations in heat shock response and thermotolerance. J. Biol. Chem. 277,
23193-23207.
2374
Journal of Cell Science 124 (14)
Journal of Cell Science
Li, R., Soosairajah, J., Harari, D., Citri, A., Price, J., Ng, H. L., Morton, C. J., Parker,
M. W., Yarden, Y. and Bernard, O. (2006). Hsp90 increases LIM kinase activity by
promoting its homo-dimerization. FASEB J. 20, 1218-1220.
Liang, P. and MacRae, T. H. (1997). Molecular chaperones and the cytoskeleton. J. Cell
Sci. 110, 1431-1440.
Lim, M. K., Kawamura, T., Ohsawa, Y., Ohtsubo, M., Asakawa, S., Takayanagi, A.
and Shimizu, N. (2007). Parkin interacts with LIM Kinase 1 and reduces its cofilinphosphorylation activity via ubiquitination. Exp. Cell Res. 313, 2858-2874.
Meriin, A. B., Yaglom, J. A., Gabai, V. L., Zon, L., Ganiatsas, S., Mosser, D. D. and
Sherman, M. Y. (1999). Protein-damaging stresses activate c-Jun N-terminal kinase via
inhibition of its dephosphorylation: a novel pathway controlled by HSP72. Mol. Cell.
Biol. 19, 2547-2555.
Morimoto, R. I. (2008). Proteotoxic stress and inducible chaperone networks in
neurodegenerative disease and aging. Genes Dev. 22, 1427-1438.
Mosser, D. D. and Morimoto, R. I. (2004). Molecular chaperones and the stress of
oncogenesis. Oncogene 23, 2907-2918.
Mosser, D. D., Caron, A. W., Bourget, L., Denis-Larose, C. and Massie, B. (1997). Role
of the human heat shock protein hsp70 in protection against stress-induced apoptosis.
Mol. Cell. Biol. 17, 5317-5327.
Mosser, D. D., Caron, A. W., Bourget, L., Meriin, A. B., Sherman, M. Y., Morimoto,
R. I. and Massie, B. (2000). The chaperone function of hsp70 is required for protection
against stress-induced apoptosis. Mol. Cell. Biol. 20, 7146-7159.
Mounier, N. and Arrigo, A. P. (2002). Actin cytoskeleton and small heat shock proteins:
how do they interact? Cell Stress Chaperones 7, 167-176.
Nishita, M., Tomizawa, C., Yamamoto, M., Horita, Y., Ohashi, K. and Mizuno, K.
(2005). Spatial and temporal regulation of cofilin activity by LIM kinase and Slingshot
is critical for directional cell migration. J. Cell Biol. 171, 349-359.
Palacios, C., Collins, M. K. and Perkins, G. R. (2001). The JNK phosphatase M3/6 is
inhibited by protein-damaging stress. Curr. Biol. 11, 1439-1443.
Park, H. G., Han, S. I., Oh, S. Y. and Kang, H. S. (2005). Cellular responses to mild
heat stress. Cell. Mol. Life Sci. 62, 10-23.
Roberts, N. J., Jr and Sandberg, K. (1979). Hyperthermia and human leukocyte function.
II. Enhanced production of and response to leukocyte migration inhibition factor (LIF).
J. Immunol. 122, 1990-1993.
Scott, R. W. and Olson, M. F. (2007). LIM kinases: function, regulation and association
with human disease. J. Mol. Med. 85, 555-568.
Soosairajah, J., Maiti, S., Wiggan, O., Sarmiere, P., Moussi, N., Sarcevic, B., Sampath,
R., Bamburg, J. R. and Bernard, O. (2005). Interplay between components of a novel
LIM kinase-slingshot phosphatase complex regulates cofilin. EMBO J. 24, 473-486.
Stankiewicz, A. R., Lachapelle, G., Foo, C. P., Radicioni, S. M. and Mosser, D. D.
(2005). Hsp70 inhibits heat-induced apoptosis upstream of mitochondria by preventing
Bax translocation. J. Biol. Chem. 280, 38729-38739.
Tomiyoshi, G., Horita, Y., Nishita, M., Ohashi, K. and Mizuno, K. (2004). Caspasemediated cleavage and activation of LIM-kinase 1 and its role in apoptotic membrane
blebbing. Genes Cells 9, 591-600.
Van Troys, M., Huyck, L., Leyman, S., Dhaese, S., Vandekerkhove, J. and Ampe, C.
(2008). Ins and outs of ADF/cofilin activity and regulation. Eur. J. Cell Biol. 87, 649667.
Vicente-Manzanares, M., Rey, M., Jones, D. R., Sancho, D., Mellado, M., RodriguezFrade, J. M., del Pozo, M. A., Yanez-Mo, M., de Ana, A. M., Martinez, A. C. et al.
(1999). Involvement of phosphatidylinositol 3-kinase in stromal cell-derived factor-1
alpha-induced lymphocyte polarization and chemotaxis. J. Immunol. 163, 4001-4012.
Wang, W., Eddy, R. and Condeelis, J. (2007). The cofilin pathway in breast cancer
invasion and metastasis. Nat. Rev. Cancer 7, 429-440.
Welch, W. J. and Suhan, J. P. (1985). Morphological study of the mammalian stress
response: characterization of changes in cytoplasmic organelles, cytoskeleton, and
nucleoli, and appearance of intranuclear actin filaments in rat fibroblasts after heatshock treatment. J. Cell Biol. 101, 1198-1211.
Yaglom, J., O’Callaghan-Sunol, C., Gabai, V. and Sherman, M. Y. (2003). Inactivation
of dual-specificity phosphatases is involved in the regulation of extracellular signalregulated kinases by heat shock and hsp72. Mol. Cell. Biol. 23, 3813-3824.
Young, J. C., Agashe, V. R., Siegers, K. and Hartl, F. U. (2004). Pathways of chaperonemediated protein folding in the cytosol. Nat. Rev. Mol. Cell Biol. 5, 781-791.