Expression of poly(A)-binding protein is upregulated during
recovery from heat shock in HeLa cells
Shuhua Ma*, Rumpa B. Bhattacharjee* and Jnanankur Bag
Department of Molecular and Cellular Biology, University of Guelph, Canada
In response to extracellular signals, cells regulate pro-
tein synthesis, and adapt to environmental changes
such as the level of growth factors, temperature, and
availability of nutrients [1]. The regulation of protein
synthesis is a complex process, as many factors are
required for mRNA translation. Translation initiation
is believed to be the rate-limiting step in protein syn-
thesis, and is usually regulated by several important
initiation factors, including poly(A)-binding protein
(PABP) 1 and eukaryotic initiation factors (eIFs) 4E,
4G, 2a, and 4A. eIF4E binds to the 5¢-cap of mRNA to
facilitate initiation of translation. eIF4A is a helicase
enzyme that removes the secondary structure from the
5¢-UTR of mRNA, and eIF4G acts as a bridge protein
to interact with eIF4E, PABP1, eIF4A and eukaryotic
release factor 3 [1,2]. At the beginning of the transla-
tion initiation cycle, eIF2a forms a ternary complex
with met-tRNAi and GTP [3]. Fine tuning of these
steps is achieved by regulating the interaction of these
initiation factors with each other and the mRNA [1].
PABP1 is an RNA-binding protein with a high
affinity for the poly(A) tail of mRNA [4]. It is mainly
distributed in the cytoplasm, and may shuttle between
the cytoplasm and the nucleus [5]. By interacting with
Keywords
eIF4G; HSP27; HSP70; mRNA translation;
poly(A)-binding protein

during recovery from heat shock. Therefore, our results raise the possibility
that the association of HSP27 with eIF4G may not be sufficient to suppress
cap-dependent translation during heat shock. In addition, we provide
evidence that the terminal oligopyrimidine cis-element of PABP1 mRNA
is responsible for the preferential increase of PABP1 mRNA translation
in cells undergoing recovery from heat shock.
Abbreviations
ARS, autoregulatory sequence; eIF, eukaryotic initiation factor; FITC, fluorescein isothiocyanate; b-gal, b-galactosidase; GFP, green
fluorescent protein; HSP, heat shock protein; PABP, cytoplasmic poly(A)-binding protein; TOP, terminal oligopyrimidine tract.
552 FEBS Journal 276 (2009) 552–570 ª 2008 The Authors Journal compilation ª 2008 FEBS
eIF4G and the poly(A) tail simultaneously, PABP1 is
believed to bring the 5¢-end and 3¢-end of mRNA into
close proximity and stimulate the translation of
mRNA by forming a closed-loop structure. Circular-
ization of mRNA is believed to promote the transla-
tion reinitiation by recycling 40S ribosomal subunit for
the next round of translation [1]. Interactions with
additional regulatory proteins allow PABP1 to regulate
the translation initiation step [6–11]. For example,
PAIP1 interacts with PABP1 through its ‘PABP1-inter-
acting motif 1’, containing a stretch of acidic amino
acids [8], and acts as a translational enhancer. Over-
expression of PAIP1 increases the rate of mRNA
translation in COS cells [9]. Another regulatory pro-
tein, PAIP2, disrupts the interaction of PABP1 with
PAIP1, and suppresses PABP1-dependent mRNA
translation [8,10]. Both PAIP1 and PAIP2 have similar
PABP1-binding motifs, and compete for binding with
PABP1 [10]. As such, PAIP2 works as an inhibitor of
mRNA translation by competing with PAIP1. Eukary-

growth-dependent manner [14]. As such, the TOP may
allow translation of PABP1 mRNA to be coordinately
regulated with a number of other TOP-containing
mRNAs encoding polypeptides involved in mRNA
translation, including elongation factors 1 and 2 and
ribosomal protein S6 kinases [15].
The PABP1 gene behaves like an early response
gene, as its expression level responds quickly to a
change in the cellular demand for protein synthesis
[16]. Under a variety of cellular stress conditions,
such as heat shock, mRNA translation undergoes
rapid changes. In response to heat shock stress, cells
induce the expression of a unique set of proteins
called heat shock proteins (HSPs) [17,18]. HSPs are
also produced at a basal level in cells under normal
conditions, and have important functions such as
facilitating appropriate protein folding as molecular
chaperones, and aiding in the assembly of protein
complexes; they also participate in translocation of
proteins across cellular membranes, as well as provid-
ing protection against cellular stresses [18]. HSPs are
also termed ‘stress proteins’, because they are overex-
pressed to improve cell survival under a variety of
additional cell-damaging conditions, such as exposure
to heavy metal ions, alcohol, and hypoxia, and glu-
cose deprivation [3]. There are several HSP families,
including HSP90, HSP70, HSP60, HSP110, and the
low molecular weight HSPs such as HSP10, HSP20,
HSP27, HSP32, and HSP40 [18]. The predominant
HSPs, HSP70 and HSP27, have been well studied.

GTP complex, which is hydrolyzed to an inactive
eIF2a–GDP complex at the end of the initiation
cycle. The GDP is then released from the eIF2a–
GDP complex, and reactivates eIF2a for the next
round of translation [3]. When eIF2a is phosphory-
lated, the GDP cannot be released from it, and the
global mRNA translation rate decreases. It has been
shown that eIF2a is increasingly phosphorylated from
mild to severe heat shock conditions as the cells are
subjected to an increasing temperature, and protein
synthesis is repressed by the presence of hyperphosph-
orylated eIF2a. However, HeLa cells subjected to a
mild heat shock (42 °C for 30 min) did not show
increased eIF2a phosphorylation, even though protein
synthesis was inhibited [3,20]. Thus, it has been sug-
gested that phosphorylation of eIF2a may not be the
dominant mechanism of inhibition of translation of
normal cellular mRNAs under heat shock, and there-
fore other regulatory mechanisms must exist.
The reduction in elF4F activity is probably also
responsible for suppression of translation of most
mRNAs during heat shock [22]. During heat shock
(44 °C, 2 h) in 293T cells, eIF4G was shown to be
trapped within the insoluble heat shock granules by
HSP27, and dissociated from PABP1. It has been sug-
gested that eIF4F-dependent mRNA translation is
inhibited by heat-induced HSP27 during heat shock
[22].
In addition to eIF4G, the PABP1 gene as an early
response gene may play a very important role in the

PABP1 expression during recovery from
heat shock
As PABP1 is important for regulation of mRNA
translation, we examined whether its cellular level is
adjusted during heat shock and subsequent recovery
periods. The results of western blot analyses (Fig. 1)
show that there was a small reduction in the cellular
abundance of PABP1 and its polypeptide partner
eIF4G immediately after the heat shock treatment.
However, during the period of recovery from heat
shock, when translation of normal cellular mRNA
resumes, there was an approximately 2.5-fold increase
in the cellular PABP1 level. In contrast, the cellular
level of eIF4G did not show a similar increase. The
increase of PABP1 abundance also correlated very well
with the increased expression of HSP27 and HSP70. In
order to assess whether the increased PABP1 level was
due to an increase in the cognate mRNA level, we
measured the PABP1 mRNA level by real time
RT-PCR. The results (Fig. 2A) show that the PABP1
mRNA levels were not altered by exposure to heat
shock or following the subsequent recovery phase. As
these results suggest translational control of PABP1
expression, we examined the distribution of PABP1
mRNA between the translationally active polysomes
and repressed subpolysomal fractions by sucrose gradi-
ent fractionation, as previously described [23]. The
results (Fig. 2B–D) show that in exponentially growing
HeLa cells, approximately 30–40% of cytoplasmic
PABP1 mRNA was present in the translated polyso-

phase of cells, using both coimmunoprecipitation and
immunofluorescence microscopy. The results of coim-
munoprecipitation studies (Fig. 3A,C) show that
although the cellular abundance of eIF4G was not
affected significantly during heat shock and recovery,
the level of its association with PABP1 was reduced.
As compared to untreated control cells, after a 2 h
heat shock treatment, approximately 40–50% less
PABP1 coimmunoprecipitated with eIF4G. This reduc-
tion, however, was a little higher than what was
expected from the reduced abundance of both polypep-
tides in heat-shocked cells. During the recovery period
from heat shock, as compared to the non-heat-shocked
control cells, almost a three-fold increase in the associ-
ation of PABP1 with eIF4G was observed. This
increase occurred without a concomitant increase in
the cellular level of eIF4G during recovery from the
thermal stress. As the cellular abundance of PABP1 in
cells recovered from heat shock also increased approxi-
mately three-fold, our results suggest that the majority
of excess PABP1 was associated with eIF4G during
the recovery phase. We also examined the association
between PABP1 and eIF4G in RNase-treated cell
extracts, and found that the coimmunoprecipitation of
PABP1 by eIF4G antibody was independent of the
presence of intact mRNA. We used b-actin as a nega-
tive control, and did not detect this polypeptide in our
immunoprecipitated samples with the eIF4G antibody.
In addition, as a loading control, we measured the
level of b-actin in total cell extracts before they were

ing controls. The values below each lane represent the relative abun-
dance of each polypeptide using an arbitrary scale where the level in
control cells was considered to be 1.00. (B) The experiment
described in (A) was repeated three times, and the averages are
shown here as mean ± standard error (SE). The linear response of
the western blotting experiments is shown in Fig. S1.
S. Ma et al. PABP expression during heat shock recovery
FEBS Journal 276 (2009) 552–570 ª 2008 The Authors Journal compilation ª 2008 FEBS 555
non-heat-shocked control cells, approximately 50% of
total cellular PABP1 was associated with eIF4G. In
heat-shocked cells, the level of eIF4G-associated
PABP1 was reduced to approximately 25% of total
cellular PABP1. However, following 12 or 24 h of
recovery, the level was increased to approximately
75% of total cellular PABP1. In addition, as expected,
all samples showed 90–100% of total eIF4G in the
immunoprecipitates.
PABP1 is a phosphoprotein, and its interaction with
poly(A) and eIF4G is enhanced by phosphorylation
[24]. Therefore, we investigated PABP1 phosphoryla-
tion by two-dimensional gel electrophoresis, followed
by western blotting as previously described [25]. In
exponentially growing cells (Fig. 3B), PABP1 was pres-
ent in multiple phosphorylated states, as shown by its
different isoelectric points. Upon inhibition of phos-
phorylation by U0126, an inhibitor of the MKK2
kinase pathway [25,26], PABP1 migrated as a more
basic protein with a higher isoelectric point, and
appeared as a single spot in two-dimensional gels. Fol-
lowing 24 h of recovery from heat shock, the abun-

β-actin
80
90
100
PABP
30
40
50
60
70
0
10
20
% in polysome
Fraction Number (6-11)
-HS
+HS
Re24
1.0
1.2
1.4
con HS
Re12 h Re24 h
0.2
0.4
0.6
0.8
Relative mRNA level
A
0.0

cytoplasmic b-actin also showed some aggregated gran-
ular structures and perinuclear localization in approxi-
mately 30–40% of cells. In contrast, the perinuclear
granular localization of eIF4G and PABP1 was more
pronounced and noticeable in a much larger percentage
(> 85%) of cells. Following 2 h of heat shock at 44 °C,
almost complete translocation of both eIF4G and
PABP1 within the cell nucleus as granules was observed.
During the same treatment, b-actin remained mostly
cytoplasmic, with some granular aggregates. Similar
nuclear translocation was also observed within a shorter
time period of 1–1.5 h when the cells were heat shocked
at 46 °C (results not shown). Following nuclear translo-
cation, we observed detectable dissociation of some
eIF4G granules from PABP1 (Fig. 4, enlarged inset).
Furthermore, neither polypeptide was present within the
nucleolus. Interestingly, during the recovery period,
PABP (-RNase)
eIF4G
PABP (+ RNase)
eIF4E
β-actin (Co-IP)
PABP
eIF4G
3
3.5
4
4.5
con
HS

Control
Re24 h
+
20um U0126
PABP
Re24 h
β-actin
Control
pH 116
E
B
Re12 h Re24 h
1.0 0.5 3.4 4.0
1.0 1.0 1.1 1.0
1.1 0.9 0.9 1.0
1.0 0.9 1.0 0.9
HS
Fig. 3. Coimmunoprecipitation of polypeptides with antibody to eIF4G and analyses of PABP1 phosphorylation. (A) Approximately
5 · 10
5
HeLa cells following heat shock and recovery were lysed in a lysis buffer and reacted with the eIF4G antibody as described in Exper-
imental procedures. The antigen–antibody complex was captured with protein A–Sepharose beads and eluted with SDS containing gel load-
ing buffer. The eluted samples were subjected to SDS ⁄ PAGE and western blotting using antibodies against PABP1, eIF4G, eIF4E and
b-actin (Santa Cruz) to detect coimmunoprecipitated polypeptides, according to the method described in Experimental procedures. Some
samples were treated with 1 lgÆlL
)1
RNase A and T1 for 10 min at 20 °C to degrade RNA before addition of the antibody. The effective-
ness of this treatment was determined by examining the absence of intact rRNA by gel electrophoresis of RNA samples prepared from the
treated cell extracts. (B) Mock immunoprecipitation was performed using 1.5 lg of preimmunized rabbit serum, and the eluted fractions
from protein A–Sepharose beads were examined for the presence of PABP1 and eIF4G as described above. (C) Western blots of three sepa-

eIF4G and PABP1 were able to relocate to the cyto-
plasm and were mostly colocalized. This suggests that
the pre-existing nuclear PABP1 and eIF4G in heat-
shocked cells could travel to the cytoplasm and reassoci-
ate during the recovery phase. However, the cellular
abundance of both polypeptides was reduced as a result
of inhibition of new protein synthesis. As mRNA trans-
lation was inhibited by cycloheximide treatment, we
could not determine whether the relocated PABP1 and
eIF4G were capable of participating in mRNA transla-
tion. eIF4G and PABP1 appeared to be completely
colocalized in cells that had recovered for 12 and 24 h.
As all of the PABP1 in exponentially growing cells was
not colocalized with eIF4G, these observations support
the coimmunoprecipitation results, which suggest an
increased association between the two polypeptides dur-
ing the recovery period. It is possible that post-transla-
tional modifications of polypeptides during the recovery
period might allow PABP1 to interact with eIF4G at a
stoichiometry of more than 1 : 1. It should be noted
here that, in addition to the C-terminal domain, PABP1
has another eIF4G-binding domain within its N-termi-
nal RNA-binding domain 1 [27]. As a negative control,
we also examined the colocalization of b-actin with
eIF4G in exponentially growing cells. As is evident from
the confocal images, b-actin was not colocalized with
eIF4G in exponentially growing cells. Furthermore, fol-
lowing recovery from heat shock, the cytoplasmic distri-
bution of b-actin returned to what was observed in
exponentially growing cells.

aggregates, especially at a very high expression level
[28]. Together, our results suggest that, following heat
shock treatment, both eIF4G and PABP1 translocate
to the cell nucleus as detergent-insoluble granular
aggregates.
Cis-acting translational control element involved
in regulating PABP1 mRNA translation during
recovery from heat shock
The mRNA encoding PABP1 consists of two well-
known translational control elements, the TOP and the
ARS in the 5¢-UTR [13]. To examine which of the two
cis-elements are involved in upregulating translation of
the PABP1 mRNA during recovery from heat shock,
we introduced both cis-elements either individually or
in combination at the 5¢-UTR of the reporter b-galac-
tosidase (b-gal) mRNA. Following transfection with
different constructs, cells were subjected to 2 h of heat
shock and recovery for 12 and 24 h, and the level of
b-gal polypeptide was measured by western blotting.
The results (Fig. 6) show that the presence of the ARS
had no stimulatory effect on the b-gal expression level
during recovery from heat shock. In contrast, the pres-
ence of the TOP element in the 5¢-UTR of the reporter
construct resulted in an increase in b-gal abundance
during recovery from heat shock. Furthermore, the
presence of the ARS with the TOP did not prevent the
increase in b-gal level from occurring during recovery
from heat shock. However, as the ARS has an inhibi-
tory effect on mRNA translation [23], its presence
together with the TOP element in the reporter con-

shown.
PABP expression during heat shock recovery S. Ma et al.
560 FEBS Journal 276 (2009) 552–570 ª 2008 The Authors Journal compilation ª 2008 FEBS
from heat shock was due to enhancement of TOP-con-
taining reporter mRNA translation.
Association of HSPs with eIF4G and PABP1
The association of HSP27 with eIF4G following heat
shock has been previously described as a possible
mechanism for repression of translation of normal cel-
lular mRNA [22]. Therefore, to extend our knowledge
further, we expanded our investigation to examine the
association of both eIF4G and PABP1 with the two
predominant HSPs, such as HSP27 and HSP70,
following heat shock, as well as during the recovery
period. We used both coimmunoprecipitation and
immunofluorescence confocal microscopy to study the
interaction between the polypeptides. The results of
coimmunoprecipitation studies (Fig. 7A,B) show that
the antibodies to both eIF4G and PABP1 immunopre-
cipitated HSP27 from extracts of heat-shocked and
recovered cells. A small amount of HSP27 was also
detected in the immunoprecipitated samples of expo-
nentially growing cells. By analyzing the percentage of
total cellular HSP27 found in the coimmunoprecipitat-
ed samples, we found that approximately 25–30% of
cellular HSP27 was associated with eIF4G in both
control and heat-shocked cells. However, owing to the
induction of HSP expression following heat shock, the
total amount of eIF4G-bound HSP27 was increased
almost two-fold (Fig. 7A). As there was very little

Transfection
ARS-β-gal
TOP-β-gal
Top- ARS-β-
gal
1.00 1.00 0.95 1.00 1.02
1.00
1.00 1.21 1.10 1.15
1.
00
1.
0
1 1.
0
1 1.
0
7
1.07 1.03 1.02 1.00
1.00 0.74
0
.
68

1.
98
2.
0
4
2.25 2.10
1.

Fig. 6. Characterization of the translational control element of PABP1 mRNA responsible for upregulation of its translation during recovery
from heat shock. (A) Reporter b-gal constructs with different PABP1 mRNA cis-elements were prepared as described in Experimental proce-
dures. The cytomegalovirus (CMV) promoter was used to drive b-gal expression; TOP, ARS and TOP + ARS sequences were placed within
the 5¢-UTR of the b-gal gene, and the parental construct pCMV-SPORT–b-gal (Invitrogen) was used as a control. The nucleotide sequences
of the TOP and ARS cis control elements of the PABP1 mRNA are shown. (B) Approximately 2 · 10
5
HeLa cells were transfected with dif-
ferent CMV– b-gal expression vectors, and 48 h after transfection, cells were heat shocked at 44 °C for 2 h and allowed to recover at 37 °C
for 0, 12 and 24 h as described in Experimental procedures. The control cells were not heat shocked, and were harvested after 48 h of
transfection. For mock transfection, cells were treated with the transfection reagent without any plasmid DNA. The cellular level of b-gal
was measured by western blotting and quantified by scanning images. The abundance of b-actin was measured as a loading control. The
numbers at the bottom of each lane represent cellular abundance relative to the transfected control cells maintained at 37 °C, after adjusting
for the loading control. (C) The values of mean ± standard error (SE) of three independent experiments are shown here. (D) Transfected cells
were treated as described in (B), total cellular RNA was isolated, and the abundance of different mRNAs was measured by real-time RT-PCR
as outlined in Experimental procedures. The average of three independent experiments is shown.
S. Ma et al. PABP expression during heat shock recovery
FEBS Journal 276 (2009) 552–570 ª 2008 The Authors Journal compilation ª 2008 FEBS 561
increase in the association of HSP27 with eIF4G in
these cells. During recovery from heat shock, the
percentage of input HSP27 present in the immuno-
precipitated samples remained almost unchanged
(Fig. 7D). The association of PABP1 with HSP27
appeared to change more significantly in heat-shocked
and recovered cells than what was observed for eIF4G.
Approximately 7% of cellular HSP27 was coimmuno-
precipitated by PABP1 antibody from extracts of heat-
shocked cells. When cells were allowed to recover for 12
and 24 h, approximately 25–30% of cellular HSP27 was
found in the immunoprecipitated samples. Therefore, it
appears that the association of PABP1 with HSP27

As a fraction of the nuclear PABP1 was still associated
with eIF4G (Fig. 3), this observation can be explained
if PABP1 does not bind HSP27 directly but associates
with it through eIF4G. During recovery from heat
shock, within 12 h, eIF4G, PABP1 and most of the
HSP27 were found in the cytoplasm. In addition, both
eIF4G and PABP1 remained colocalized with HSP27.
-HS HS Re12h Re24h
HSP 27
10 17 25 20
HSP70
1.0 1. 2.5 2.0
1.0 1.7 0.7
eIF4G
HSP 27
1.0 0.9 0.9 0.8
HSP70
1.0 1.3
(0.6) (2.2) (2.5)
PABP
1.0 0.7 1.9 1.8
HSP 27
HSP70
50
AD
EB
C
con HS Re12 h Re24 h
20
30

with the HSP27, a significant fraction of the HSP27
was also localized independently. This was expected, as
HSP27 also functions as a chaperone to properly fold
cellular polypeptides [18].
A
B
Fig. 8. Cellular colocalization of HSP27 with eIF4G and PABP1 following heat shock and recovery. Cells were immunostained with appropri-
ate antibodies and viewed by laser scanning confocal microscopy as described in the legend to Fig. 4. HSP27 was labeled using Texas Red-
conjugated secondary antibody, and eIF4G and PABP1 were labeled with an appropriate FITC-labeled secondary antibody.
S. Ma et al. PABP expression during heat shock recovery
FEBS Journal 276 (2009) 552–570 ª 2008 The Authors Journal compilation ª 2008 FEBS 563
Our results suggest that eIF4G could associate with
HSP27 in the cell nucleus and translocate to the cyto-
plasm as a complex with HSP27 during recovery from
heat shock. This process perhaps facilitates refolding
of eIF4G into its native form during recovery from
heat shock. We propose that PABP1 associates with
HSP27 by virtue of its interaction with eIF4G. As
capped mRNAs are translated during recovery from
heat shock, it is likely that the association of eIF4G
with HSP27 may not be the sole basis for the reduced
translation of capped mRNA in heat-shocked HeLa
cells, and therefore these results differ from what was
reported previously for H293T cells [22].
Discussion
Translational control of PABP1 expression during
recovery from heat shock
PABP1 is considered to be a genuine translation initia-
tion factor [29], and its activity and expression level
are controlled by a multitude of regulatory networks

the PABP1 mRNA is translated less efficiently than
the control b-actin mRNA, which was present predom-
inantly (90%) in the polysomal fraction. However,
during recovery from heat shock, the efficiency of
PABP1 mRNA translation reached the same level as
that of the b-actin mRNA. Similar preferential
enhancement of PABP1 mRNA translation was also
observed during liver regeneration and growth stimula-
tion of cells following serum starvation [14,30]. Thus,
it appears that a preferential increase of PABP1
mRNA translation occurs whenever there is a demand
for an increase in global mRNA translation. The
increase in the cellular PABP1 level may act as a cue
for cells to stimulate global mRNA translation. The
PABP1 mRNA is generally inefficiently translated
under normal growth conditions through feedback
inhibition mediated by binding of PABP1 to its own
ARS cis-element [23]. It is not known how the TOP
overrides the inhibition by the ARS during recovery
from heat shock. However, a small reduction of
PABP1 abundance after heat shock could conceivably
relieve the feedback control, and enhance PABP1
mRNA translation through its TOP cis-element.
Association between PABP1 and eIF4G during
heat shock and recovery
The function of PABP1 in mRNA translation depends
on its interaction with eIF4G. Earlier, it was suggested
that this interaction is inhibited by sequestration of
eIF4G into insoluble granules in a complex with
HSP27; as a result, cap-dependent mRNA translation

previous study, where eIF4G was not found in the
nucleus of H293T cells following heat shock at 44 °C
for 2 h. However, even in this previous report, most of
the eIF4G accumulated as granules near the perinucle-
ar site, which we also observed during a somewhat
shorter heat shock period. The observed difference in
the nuclear localization of eIF4G distribution between
the two studies could be simply due to the use of dif-
ferent cell lines. Cultured immortal cell lines may differ
with respect to the time and temperature for the opti-
mum stress response. In our studies, we found that the
nuclear localization of eIF4G and PABP1 was depen-
dent on the length of exposure of cells as well as the
temperature. Therefore, under more stringent condi-
tions, similar nuclear localization of those two poly-
peptides might also occur in H293T cells. Under
normal conditions, both PABP1 and eIF4G are
predominantly localized in the cell cytoplasm; never-
theless, both polypeptides contain weak nuclear locali-
zation signals to facilitate nuclear entry under some
circumstances [32,33]. A small fraction of eIF4G has
been previously shown to be present in the nuclear
fraction, and to bind the cap-binding complex involved
in splicing [34]. It is believed that the nuclear eIF4G
may associate with pre-mRNA through its interaction
with the polypeptide partners of the cap-binding com-
plex, such as CBP20 and CBP80, and accompany the
mRNA to the cytoplasm, where partner switching
occurs to form the eIF4F complex. This process could
conceivably couple mRNA translation with mRNA

the cytoplasm during the recovery period. During the
recovery period, we have also shown an increase in
the association of PABP1 with eIF4G, and a concom-
itant increase in its phosphorylation during the recov-
ery period. Thus, it appears that the excess PABP1
produced during recovery is mostly utilized for mRNA
translation.
Translational control elements of PABP1 mRNA
The main function of the ARS element of PABP1
mRNA is to regulate the cellular PABP1 level under
all circumstances [13]. The TOP cis-element is involved
in regulating PABP1 mRNA translation during growth
stimulation [14]. A number of mRNAs coding for fac-
tors involved in mRNA translation, such as ribosomal
proteins and elongation factor eEF1a, contain a simi-
lar TOP element [14,15]. It is believed that the presence
of this common cis-element allows coordinated regula-
tion of translation of mRNAs that participate in the
same biochemical step, such as mRNA translation. We
have shown here that the TOP cis-element of PABP1
mRNA is involved in the upregulation of PABP1
mRNA translation during recovery from heat shock.
Previous studies have shown that translation of TOP-
containing mRNAs is stimulated during growth by
changing the size of the polysomes [14]. This could
occur either by moving the 40S ribosomal subunit at a
faster rate along the 5¢-UTR, or by increasing the rate
of elongation. However, these two mechanisms are not
necessarily mutually exclusive. We have previously
shown that the presence of the ARS in the 5¢-UTR of

studies, the stimulatory effect of the TOP on reporter
mRNA translation was maintained even when the
ARS was also present. However, the overall cellular
abundance of the reporter b-gal was less in cells trans-
fected with both ARS- and TOP-containing construct
than what was observed with the construct containing
TOP alone. This is in good agreement with previous
observations that the ARS slows the rate of translation
initiation of the endogenous PABP1 mRNA in the
presence of the TOP control element [23].
Association of HSPs with PABP1 and eIF4G
during heat shock and recovery
The HSPs function as molecular chaperones, and inter-
act with many cellular proteins. Previous studies have
linked HSP27 with the eIF4G subunit of the eIF4F
complex [22]. Here we showed colocalization of HSP27
with both eIF4G and PABP1 in heat-shocked cells. Fol-
lowing 24 h of recovery from heat shock, most of the
HSP27 accumulated in the cytoplasm with eIF4G and
PABP1, and possibly remained colocalized within a
complex. However, it is not known whether or not the
HSP27-associated eIF4G participates in translation.
Thus, the previous observation regarding the sequestra-
tion of eIF4G by HSP27 may still explain, at least in
part, the reduced cap-dependent translation of mRNA
in heat-shocked cells [22]. In our studies, we have shown
that PABP1 also associates with HSP27, albeit at a
much reduced level in heat-shocked cells than what was
observed with eIF4G. Perhaps HSP27 does not directly
interact with PABP1 but was coimmunoprecipitated

recovery is accompanied by a preferential upregula-
tion of PABP1 expression.
Experimental procedures
Cell culture and heat shock treatment
HeLa cells were grown at 37 °C in the DMEM supple-
mented with 10% fetal bovine serum for 2–3 days, until
cells were fully spread and the desired degree of confluence
was obtained. Cells were subjected to heat shock at 44 °C
for 2 h, and incubated at 37 °C for 0, 12 and 24 h to allow
them to recover. The cells were washed and fixed for
immunofluorescence analysis, or harvested for biochemical
studies.
Transfection of cells
Approximately 2–5 · 10
5
subconfluent HeLa cells grown on
35 mm dishes were transfected with 2 lg of plasmid DNA
using 10 lg of lipofectamine (Invitrogen, Burlington,
Canada), according to protocols supplied by the manufac-
turer. In short, the DNA was mixed with lipofectamine and
PABP expression during heat shock recovery S. Ma et al.
566 FEBS Journal 276 (2009) 552–570 ª 2008 The Authors Journal compilation ª 2008 FEBS
100 lL of Opti-MEM (Invitrogen) for 30 min before being
added to the culture dish and incubated for 6 h with an
additional 0.9 mL of Opti-MEM. After 6 h, 1 mL of
growth medium containing 20% fetal bovine serum was
added. Following a total of 12 h of incubation, the medium
was replaced with fresh DMEM containing 10% fetal
bovine serum. Cells were usually harvested after an addi-
tional 24–36 h of incubation at 37 °C.

)1
aprotinin, 0.1 mm phen-
ylmethanesulfonyl fluoride, and 10 lgÆmL
)1
leupeptin]. The
cells were lysed by passing them through a 27G syringe.
The cell lysate was centrifuged at 5000 g for 5 min, and the
supernatant was used for analysis. Immunoprecipitation
was carried out by adding 1.5 l g of goat monoclonal anti-
eIF4G or rabbit polyclonal PABP1 antibody (Santa Cruz)
to the cell extract. After overnight incubation at 4 °Cona
rotary mixer, 50 lL of protein A–Sepharose bead slurry
(Amersham Biosciences, Piscataway, NJ, USA) was added,
and the incubation was continued for 3 h at 4 °C. After
washing six times with 1 mL of lysis buffer, resin-bound
proteins were eluted from the beads by adding 30 lLof2·
Laemmli sample buffer (4% SDS, 20% glycerol, 120 mm
Tris ⁄ HCl, pH 6.8, 200 mm dithiothreitol, and 0.1% brom-
ophenol blue) and heating the mixture for 5 min at 95 °C.
The proteins were then resolved by 12% SDS ⁄ PAGE, and
the presence of coimmunoprecipitated polypeptides was
examined in the eluted samples by western blotting using
the appropriate antibody.
Immunofluorescence confocal microscopy
HeLa cells plated on glass coverslips were used for exam-
ining the cellular localization of polypeptides by using
immunofluorescence confocal microscopy. Cells were
washed with NaCl ⁄ P
i
and fixed with methanol at )20 °C

objectives was used for all cells. Images were recorded with
a resolution of 512 · 512 pixels. A series of optical slices
that spanned the depth of the cell was obtained (normally
eight slices, 0.5 lm apart). Other parameters were used with
default settings. In order to avoid cross-talk between FITC
and Texas Red channels, the signals were scanned individu-
ally with one channel on and the other off. The images for
the same cells from two channels were merged with the
overlap function. To differentiate the autofluorescence, con-
trol cells without immunostaining were examined to set up
the settings for the stained cells.
Measurement of mRNA levels
Total cellular RNA was isolated using the High Pure
RNA Isolation Kit (Roche Biochemical, Indianapolis, IN,
USA). The quality and quantity of the RNA were deter-
mined by 1.5% agarose gel electrophoresis and spectropho-
tometry, respectively. The levels of specific mRNAs,
including PABP1, b-actin and b-gal, in the samples were
determined either by quantitative real-time RT-PCR using
S. Ma et al. PABP expression during heat shock recovery
FEBS Journal 276 (2009) 552–570 ª 2008 The Authors Journal compilation ª 2008 FEBS 567
a Rotor-gene 3000 (Corbett Research, NSW, Australia) or
by RT-PCR. An aliquot of total RNA (1 lg) was reverse
transcribed at 42 °C for 1 h in a total reaction volume of
25 lL using SuperScript II reverse transcriptase (Invitrogen,
Burlington, Canada) and 150 ng of random primers. After
the reaction, 5 lL of the cDNA sample was amplified by
PCR in a total reaction volume of 25 lL, using the plati-
num SYBR green qPCR supermix-UDG kit (Invitrogen,
Burlington, Canada) and 100 ng of primers specific for

debris by centrifugation at 12 000 g for 10 min, the super-
natant fraction was centrifuged in a 12 mL 10–50% sucrose
gradient, containing 25 mm Hepes (pH 7.0), 50 mm KCl,
2mm MgOAc, 50 lgÆmL
)1
cycloheximide and 15 mm
2-mercaptoethanol at 100 000 g in a Beckman SW 41Ti
rotor for 3 h as previously described [23]. Gradient frac-
tions of approximately 1 mL each were collected using an
Auto Densi-Flow IIC apparatus (Buchler Instruments, Fort
Lee, NJ, USA). Total RNA from each fraction was isolated
using the Triazole RNA isolation kit as described by the
manufacturer (Roche), and precipitated with ethanol using
5 lg of yeast tRNA as a carrier.
Plasmid construction
A reporter b-gal construct containing the ARS region
(nucleotides 71–131) of human PABP1 (GeneBank ID:
Y00345) was generated by ligating double-stranded oligode-
oxynucleotides with BamHI and NheI sticky ends into the
site between BamHI and NheI of the pCMV-Sport–b-gal
plasmid vector. To generate the TOP–b-gal construct, the
TOP region (nucleotides 1–31) of human PABP1 was
inserted into the transcription start site of pCMV-Sport–
b-gal by using a megaprimer-based method [40]. A mega-
primer was produced by PCR with primer 1 containing an
EcoRI site and primer 2 (primer sequences are listed in
Table 2). The PCR product was used as the megaprimer,
and treated with exonuclease I to digest the remaining sin-
gle-stranded primers. The megaprimer was used as the
upstream primer to amplify another region of the vector by

b-gal (AS) CAGCACCGCATCAGCAAG (2560–2577)
PABP expression during heat shock recovery S. Ma et al.
568 FEBS Journal 276 (2009) 552–570 ª 2008 The Authors Journal compilation ª 2008 FEBS
Acknowledgements
This work was supported by funds from a discovery
grant from the Natural Science and engineering
Research council (NSERC) of Canada.
References
1 Gebauer F & Hentze MW (2004) Molecular mecha-
nisms of translational control. Nat Rev Mol Cell Biol 5,
827–835.
2 Mendez R & Richter JD (2001) Translational control
by CPEB: a means to the end. Nat Rev Mol Cell Biol 2,
521–529.
3 Panniers R (1994) Translational control during heat
shock. Biochimie 76, 737–747.
4 Gorlach M, Burd CG & Dreyfuss G (1994) The mRNA
poly (A)-binding protein: localization, abundance and
RNA-binding specificity. Exp Cell Res 211 , 400–407.
5 Afonina E, Stauber R & Pavlakis GN (1998) The human
poly (A)-binding protein 1 shuttles between the nucleus
and the cytoplasm. J Biol Chem 271, 13015–13024.
6 Uchida N, Hoshino S, Imataka H, Sonenberg N &
Katada T (2002) A novel role of the mammalian
GSPT ⁄ eRF3 associating with poly(A)-binding protein
in cap ⁄ poly(A)-dependent translation. J Biol Chem 277,
50286–50292.
7 Sagliocco F, Laloo B & Cosson B (2006) The
ARE-associated factor AUF1 binds poly (A) in vitro in
competition with PABP1. Biochem J 400, 337–347.

ratus is regulated at the translational level. Eur J Bio-
chem 267, 6321–6330.
16 Mangus DA, Evans MC & Jacobson A (2003) Poly(A)-
binding proteins: multifunctional scaffolds for the post-
transcriptional control of gene expression. Genome Biol
24, 223, doi:10.1186/gb-2003-4-7-223.
17 Santoro MG (2002) Heat shock factors and the control
of stress response in inflammation. Biochem Pharmacol
59, 55–63.
18 Nollen E & Morimoto RI (2002) Chaperoning signaling
pathways: molecular chaperones as stress-sensing ‘heat
shock’ proteins. J Cell Sci 115, 2809–2816.
19 Trinklein ND, Murray JI, Hartman SJ, Botstein D &
Myers RM (2004) The role of heat shock transcription
factor 1 in the genome-wide regulation of the mamma-
lian heat shock response. Mol Cell Biol 15, 254–1261.
20 Sierra JM & Zapata JM (1994) Translational regulation
of the heat shock response. Mol Biol Rep 19, 211–220.
Table 2. Primers and sequences for cloning TOP–b-gal and TOP–ARS–b-gal. Upper and lower case letters represent vector and PABP
sequences respectively. The Restriction enzymes’ recognition sequences are shown in italic.
Primer name Primer sequence (5¢-to3¢)
Primer 1 with EcoRI site GACCCGGGAATTCCGGACCGG (nucleotides 4222–4242 of PCMV-Sport–b-gal plasmid)
Primer 2 containing TOP AAGCAGAGCTCGTTTAGTGAACCGCcttctccccggcggttagtgctgagagtgcTCAGATCG
CCTGGAGACGCC (nucleotides 4403–4380 of PCMV-Sport–b-gal + TOP + nucleotides
4379–4360 of PCMV-Sport–b-gal plasmid)
Primer 3 with NcoI site CGCTATTACCATGGTGATGC (nucleotides 4588–4608 of PCMV-Sport–b-gal plasmid)
Sense ARS with PstI and SalI CGGTTCACTAAACGAGCTCTGCTGCAGaaaaaatccaaaaaaaatctaaaaaaatcttttaaaa
aaccccaaaaaaatttacaaaaaaGTCGACaatgc
Antisense ARS with PstI and SalI gcattGTCGACttttttgtaaatttttttggggttttttaaaagatttttttagattttttttg
gattttttCTGCAGCAGAGCTCGTTTAGTGAACCG

eIF4B bind the poly(A)-binding protein through over-
lapping sites within the RNA recognition motif
domains. J Biol Chem 282, 25247–25258.
28 Wang Q, Mosser DD & Bag J (2005) Induction of
HSP70 expression and recruitment of HSC70 and
HSP70 in the nucleus reduce aggregation of a polyala-
nine expansion mutant of PABP1N1 in HeLa cells.
Hum Mol Genet 14, 3673–3684.
29 Kahvejian A, Svitkin YV, Sukarieh R, M’Boutchou
MN & Sonenberg N (2005) Mammalian poly (A)-bind-
ing protein is a eukaryotic translation initiation factor,
which acts via multiple mechanisms. Genes Dev 19,
104–113.
30 Berger L, Bag J & Sells BH (1992) Translation of
poly (A)-binding protein mRNA is regulated by growth
conditions. Biochem Cell Biol 70, 770–778.
31 Stolovich M, Tang H, Hornstein E, Levy G, Cohen R,
Bae SS, Birnbaum MJ & Meyuhas O (2002) Transduc-
tion of growth or mitogenic signals into translational
activation of TOP mRNAs is fully reliant on the phos-
phatidylinositol 3-kinase-mediated pathway but requires
neither S6K1 nor rpS6 phosphorylation. Mol Cell Biol
22, 8101–8113.
32 Brune C, Munchel SE, Fischer N, Podtelejnikov AV &
Weis K (2005) Yeast poly (A)-binding protein
Pab1 shuttles between the nucleus and the
cytoplasm and functions in mRNA Export. RNA 11,
517–531.
33 Coldwell MJ, Hashemzadeh-Bonehi L, Hinton TM,
Morley SJ & Pain VM (2004) Expression of

40 Tyagi R, Lai R & Duggleby RG (2004) A new
approach to ‘megaprimer’ polymerase chain reaction
mutagenesis without an intermediate gel purification
step. BMC Biotechnol 4, 2, doi:10.1186/1472-6750-4-2.
Supporting information
The following supplementary material is available:
Fig. S1. Dose response of western blotting. HeLa cells
were lysed in gel sample buffer as described in Experi-
mental procedures and the indicated volume of cell
lysate was used for western blotting using PABP,
HSP70 and beta-actin antibodies to examine the linear
response of the detection technique.
This supplementary material can be found in the
online version of this article.
Please note: Wiley-Blackwell is not responsible for
the content or functionality of any supplementary
materials supplied by the authors. Any queries (other
than missing material) should be directed to the corre-
sponding author for the article.
PABP expression during heat shock recovery S. Ma et al.
570 FEBS Journal 276 (2009) 552–570 ª 2008 The Authors Journal compilation ª 2008 FEBS