Translation of Viral mRNA without Active eIF2: The Case
Ewelina Welnowska., Miguel Angel Sanz*., Natalia Redondo, Luis Carrasco
Centro de Biologı ´a Molecular ‘‘Severo Ochoa’’ (CSIC-UAM), Universidad Auto ´noma de Madrid, Madrid, Spain
Previous work by several laboratories has established that translation of picornavirus RNA requires active eIF2a for
translation in cell free systems or after transfection in culture cells. Strikingly, we have found that encephalomyocarditis
virus protein synthesis at late infection times is resistant to inhibitors that induce the phosphorylation of eIF2a whereas
translation of encephalomyocarditis virus early during infection is blocked upon inactivation of eIF2a by phosphorylation
induced by arsenite. The presence of this compound during the first hour of infection leads to a delay in the appearance of
late protein synthesis in encephalomyocarditis virus-infected cells. Depletion of eIF2a also provokes a delay in the kinetics of
encephalomyocarditis virus protein synthesis, whereas at late times the levels of viral translation are similar in control or
eIF2a-depleted HeLa cells. Immunofluorescence analysis reveals that eIF2a, contrary to eIF4GI, does not colocalize with
ribosomes or with encephalomyocarditis virus 3D polymerase. Taken together, these findings support the novel idea that
eIF2 is not involved in the translation of encephalomyocarditis virus RNA during late infection. Moreover, other
picornaviruses such as foot-and-mouth disease virus, mengovirus and poliovirus do not require active eIF2a when maximal
viral translation is taking place. Therefore, translation of picornavirus RNA may exhibit a dual mechanism as regards the
participation of eIF2. This factor would be necessary to translate the input genomic RNA, but after viral RNA replication, the
mechanism of viral RNA translation switches to one independent of eIF2.
Citation: Welnowska E, Sanz MA, Redondo N, Carrasco L (2011) Translation of Viral mRNA without Active eIF2: The Case of Picornaviruses. PLoS ONE 6(7): e22230.
Editor: Eliane F. Meurs, Institut Pasteur, France
Received March 18, 2011; Accepted June 17, 2011; Published July 14, 2011
Copyright: ? 2011 Welnowska et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by a Grant from Direccion General de Investigacion Cientifica y Tecnica (DGICYT) (BFU2009-07352) and the Institutional Grant
awarded to the Centro de Biologı ´a Molecular "Severo Ochoa" by the Fundacio ´n Ramo ´n Areces. The funders had no role in study design, data collection and
analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
. These authors contributed equally to this work.
The genome of picornaviruses comprises a molecule of single-
stranded RNA of positive polarity that also acts as the only viral
mRNA that is translated in infected cells . Upon binding of the
virion to its receptor, the naked viral particles deliver the ssRNA
molecule to the cytoplasm, where it is recognized and translated by
the cellular protein synthesizing machinery . This early viral
translation is followed by RNA replication giving rise to large
amounts of RNA molecules of positive polarity, some of which
may serve as new mRNAs to direct the massive synthesis of viral
proteins during the late phase of infection [3,4,5]. This late viral
translation is accompanied by a profound inhibition of cellular
protein synthesis. The mechanism by which picornavirus mRNA
is translated has been analyzed from the early days of research on
eukaryotic protein synthesis. In fact, encephalomyocarditis virus
(EMCV) RNA was the first viral mRNA to be translated in a
mammalian cell free system . Shortly afterwards, the require-
ments for different eIFs were investigated, revealing that eIF2 was
necessary for EMCV mRNA translation . Since then, all
experiments with picornavirus mRNAs have provided overwhelm-
ing evidence for requirement of eIF2 for the initiation of
picornavirus protein synthesis in cell free systems and in culture
cells transfected with these mRNAs [8,9,10]. The elegant
experiments by Pestova et al.  using reconstituted translation
systems with all the purified components indicate that not all eIFs
are necessary for EMCV translation in vitro. These investigators
have observed that only a central domain of eIF4G was necessary
for EMCV RNA translation, while eIF4E and eIF4B were
dispensable . The exclusion of eIF2 from these systems
abolished protein synthesis directed by picornavirus mRNAs. The
presence of IRES elements in mRNAs was also initially found in
picornavirus mRNAs [6,13]. The structure and the eIF require-
ments for the translation of the different IRES-containing
picornavirus RNAs may vary among the different species
investigated. Based on these differences, at present four classes of
picornavirus IRESs can be considered , but all of them require
eIF2 for efficient translation in cell free systems.
The function of eIF2 is to bind Met-tRNAiand GTP to form the
ternary complex Met-tRNAi-eIF2-GTP, which interacts with the P
site on the 40S ribosomal subunit, establishing the interaction
between the initiator AUG codon with the anticodon present in
Met-tRNAi[15,16,17]. Binding of the 60S ribosomal subunit to the
pre-initiation complex promotes cleavage of GTP, displacing eIF2-
GDP from the ribosome. The eIF2-GDP complex is recycled to
eIF2-GTP by the activity of the recycling factor eIF2B. Factor eIF2
is composed of three subunits, known as a, b and c [15,16]. Subunit
eIF2a is a 36 kDa protein that contains a serine residue at position
51 (Ser-51), which can be phosphorylated by four different cellular
protein kinases. Nutrient deprivation or cellular stresses, such as
PLoS ONE | www.plosone.org1 July 2011 | Volume 6 | Issue 7 | e22230
heat-shock or viral infection, can activate some of these protein
kinases [18,19,20]. GCN2 is activated by amino acid starvation,
PKR phosphorylates eIF2 in response to double-stranded RNA,
PERK is activated by protein misfolding at the endoplasmic
reticulum and HRI phosphorylates eIF2 in the absence of HEME.
Phosphorylation of eIF2a impairs the GDP-GTP recycling
catalyzed by eIF2B. Therefore, the ternary complex Met-tRNAi-
eIF2-GTP is not generated and thus binding of this complex to the
40S ribosome is hampered. Even partial phosphorylation of eIF2
can lead to total abrogation of translation .
Study of eIF2 phosphorylation in picornavirus infected cells has
yielded varying results. Some reports suggested that this factor
[22,23], while other investigators found substantial eIF2 phos-
phorylation after PV infection, particularly at late times [24,25].
Of interest, PKR becomes highly activated, yet it is hydrolyzed in
PV-infected cells although this hydrolysis is not directly executed
by any of the PV proteases (2A or 3C) [24,26]. Mengovirus
infection of mouse L-cells provokes a substantial activation of
PKR, leading to eIF2 phosphorylation between 3–7 h after virus
absorption . The inactivation of eIF2 was coincident with the
global inhibition of cellular and viral translation. Interferon
treatment of culture cells stimulates, among others, PKR and
the 29-59 A system blocking EMCV translation . Direct
evidence that activation of PKR alone suffices to block EMCV
growth was provided by a cell line that stably synthesizes PKR
. All these findings pointed to the idea that active eIF2 was
necessary to sustain picornavirus translation. The partial phos-
phorylation of eIF2 arising in picornavirus-infected cells as
infection progresses might be partially responsible for the shut-
down of cellular translation and the arrest of viral protein
synthesis. Recent findings from our laboratory have provided
evidence that Sindbis virus subgenomic mRNA exhibits a dual
mechanism of translation. This mRNA follows a canonical
mechanism when it is directly electroporated in cells or is
translated in cell free systems, while it does not require eIF4G
nor eIF2 for efficient translation in the infected cells . A similar
mechanism may be used by other viruses, including the Cricket
paralysis virus . In view of these findings, we reappraised the
analysis for the participation of eIF2 during picornavirus RNA
translation. Our present results indicate that EMCV protein
synthesis does not require active eIF2 at late infection times, while
this factor is necessary at early times, suggesting that EMCV
mRNA translation can also follow a dual mechanism for the
synthesis of viral proteins.
Induction of eIF2a phosphorylation profoundly arrests
cellular translation, while EMCV protein synthesis is
Initially, we wished to test the effect of induction of eIF2a
phosphorylation on EMCV translation in infected cells. Previous
work has established that treatment of culture cells with
compounds such as dithiothreitol (DTT), thapsigargin (Tg) or
arsenite (Ars) causes phosphorylation of eIF2a leading to a
profound arrest of cellular translation [32,33]. Mouse embryo
fibroblasts (MEFs) were infected or not with EMCV and at 4 hpi
the test compounds were added to the medium and incubated for
1 h. Protein synthesis was estimated by addition of [35S]Met/Cys
15 min after the addition of the different compounds and
incubated for 45 min. Protein synthesis was then analyzed by
SDS PAGE followed by fluorography and phosphorylation of
eIF2a was tested by Western blot (Figures 1A and B). Treatment
with 400 mM DTT, 200 mM Ars and 1 mM Tg has no effect on
the total amount of eIF2a, while the phosphorylated form of this
factor clearly increases in the presence of any of these three
compounds in control cells (Figure 1B). As a consequence, cellular
protein synthesis strongly diminishes in the presence of these
inhibitors (Figure 1A). By contrast, synthesis of EMCV proteins is
almost unaffected by treatment with these agents, despite the fact
that strong eIF2a phosphorylation is found in the infected cells.
For instance, treatment of mock-infected cells with DTT induces
92% inhibition of cellular translation, as calculated by densito-
metric analysis, while EMCV protein synthesis only decreases by
24% (Figure 1A). Cellular translation was calculated by densi-
tometry of the most prominent band that corresponds to actin,
whereas viral translation was calculated by densitometry of all viral
proteins. Notably, phosphorylation of eIF2a is clearly apparent in
EMCV-infected cells at 5 hpi even in the absence of test
compounds. This suggests that EMCV infection induces the
phosphorylation of eIF2a.
It should be noted that Ars partially affects the proteolytic
cleavage of the EMCV polyprotein, leading to the accumulation of
viral precursors and the diminution of viral proteins of low MW.
Therefore, we wished to test in more detail the action of DTT on
cellular and viral translation. To this end, mock- or EMCV-
infected cells were treated at 5 hpi with different DTT
concentrations (125, 250 and 500 mM) and protein synthesis was
measured from 5.15–6 hpi (Figure 1C). Increasing concentrations
of DTT induce an almost total inhibition of cellular translation
while EMCV protein synthesis is much less affected under these
conditions (Figures 1C and 1D). These findings reveal that
substantial EMCV protein synthesis occurs at late times of EMCV
infection after induction of eIF2a phosphorylation by different
compounds. To estimate the percentage of eIF2a phosphorylated
by treatment of culture cells with DTT or Ars, isoelectric focusing
was carried out. In untreated cells, most of eIF2a (95%) remains
unphosphorylated, whereas in the presence of DTT or Ars almost
all eIF2a (90–100%) becomes phosphorylated (Figure S1). These
results agree well with our previous observations on the percentage
of eIF2a phosphorylated in BHK cells infected with Sindbis virus
and treated with Ars . Therefore, this potent phosphorylation
of eIF2a leads to the inhibition of cellular translation. The finding
that Ars has little effect on late EMCV protein synhtesis suggests
that this compound exhibits little toxicity on cellular processes that
may influence mRNA translation, such as ATP or GTP synthesis.
To further estimate the potential Ars toxicity on cellular protein
synthesis, we employed the mouse cell line that expresses a form of
eIF2a that cannot be phosphorylated. This cell line expresses an
eIF2a bearing a point mutation at serine 51 (S51A). Addition of
different Ars concentrations strongly inhibits cellular translation in
control MEFs, whereas under these conditions Ars has almost no
effect on protein synthesis in MEFs(S51A), demonstrating that the
major effect of Ars on translation is mediated by the induction of
eIF2a phosphorylation (Figure S2A). To further analyze the
differential action of Ars in MEFs and MEFs(S51A), EMC-luc
mRNA was transfected in these cells in presence or absence of Ars.
Notably, luc synthesis was blocked in MEFs by about 85% in
presence of Ars, whereas this compound had almost no effect in
MEFs(S51A) (Figure S2B). However, we have found that, for
unknown reasons, this variant cell line cannot be infected by
several animal viruses tested, including EMCV and Sindbis virus.
EMC-luc translation upon eIF2a phosphorylation in
culture cells and in cell free systems
In Sindbis virus-infected cells, we have demonstrated that
translation is coupled to transcription. Thus, late viral subgenomic
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org2 July 2011 | Volume 6 | Issue 7 | e22230
mRNA exhibits a different requirement for eIFs when they are
transcribed by the Sindbis virus replication machinery, as
compared to their requirements when electroporated into culture
cells [30,34]. Overwhelming evidence obtained over the years in
many laboratories has established that translation directed by
EMCV RNA requires the participation of eIF2 [11,35].
Therefore, our results described above indicating that eIF2 may
not participate in the initiation of EMCV RNA translation were
quite unexpected. In order to examine the requirement of eIF2 on
translation driven by EMCV IRES, we used an EMC-luc mRNA
synthesized by in vitro transcription, which contains the luc gene
immediately behind the IRES sequence of EMCV. BHK cells
were electroporated with EMC-luc and the action of Ars was
tested. For comparative purposes cells were also electroporated
with Cap-luc or CrPV IGR-luc mRNAs and then treated with
different concentrations of Ars (0, 50, 100 and 200 mM) for
75 min. After that time luc activity was measured and the amount
of phosphorylated eIF2a was analyzed (Figure 2A). At the highest
dose of Ars, Cap-luc mRNA was inhibited by about 80%, while
CrPV IGR-luc which is resistant to eIF2a phosphorylation was
inhibited by only 20% (Figure 2A). Notably, luc synthesis directed
by EMC-luc exhibited a high sensitivity to Ars, with 90%
inhibition at 50 mM Ars. Analysis of eIF2a indicated that this
factor was phosphorylated in Ars-treated cells (Figure 2A).
Next, in vitro translation of these different mRNAs was tested
and the effect of poly(I:C) analyzed. For this purpose, rabbit
reticulocyte lysates were programmed with EMC-luc, Cap-luc and
CrPV IGR-luc mRNAs, in the absence or presence of the
inhibitor. After incubation, luc activity was estimated. Poly(I:C)
rendered an inhibition of EMC-luc translation of about 90%,
similar to that found with Cap-luc, while CrPV IGR-luc was
almost unaffected by this compound (Figure 2B). These results
indicate that unphosphorylated eIF2a must be present in the cell
or in vitro for efficient initiation of translation of EMC-luc. In
addition, these findings contrast with those reported above
(Figure 1), illustrating that late viral protein synthesis takes place
when eIF2a is phosphorylated in EMCV-infected cells.
In EMCV-infected cells, preferential translation of viral mRNAs
synthesized by viral transcription is observed . Thus, EMC-luc
mRNAs transfected in these cells at late times of infection are
excluded from translation. Taking into account these consider-
ations, we wanted to assay the effect of Tg on the translation of
EMC-luc mRNA in EMCV-infected cells. To this end, EMCV-
infected MEFs were transfected with EMC-luc mRNA at different
hpi and the action of 1 mM Tg was tested (Figure 2C). Translation
of exogenous EMC-luc mRNA decreases when it is transfected at
late times of EMCV infection, in good agreement with our
previous results . Strikingly, Tg blocks EMC-luc mRNA
translation at all hpi tested, pointing to a different behavior of
EMCV RNA made from transcription or that transfected into
cells, as regards to the requirement for active eIF2. Similar findings
were obtained in BHK cells infected with EMCV and transfected
Figure 1. Effect of different inducers of eIF2a phosphorylation on cellular and EMCV translation. A) MEFs were mock- or EMCV-infected
at 10 pfu/cell. Cells were subsequently pre-treated with 200 mM arsenite (Ars), 400 mM DTT, or 1 mM thapsigargin (Tg) for 15 min and then labelled for
45 min with [35S]Met-Cys in presence of the same compounds. Samples were submitted to SDS-PAGE, fluorography and autoradiography. B) Western
blot analysis of eIF2a and phosphorylated eIF2a using the same samples as in panel A and antibodies anti-phospho-eIF2a (1:1000 dilution) and anti-
eIF2a (1:1000 dilution). C) MEFs mock- or EMCV-infected at 10 pfu/cell were pre-treated for 15 min with 0, 125, 250 or 500 mM DTT and then labelled
for 45 min with [35S]Met-Cys in presence of the same amounts of DTT. Samples were then collected and submitted to SDS-PAGE, fluorography and
autoradiography. D) Cellular and viral protein synthesis examined by densitometric analysis of the autoradiograph shown in panel C. The protein
bands analyzed are indicated by an asterisk.
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org3 July 2011 | Volume 6 | Issue 7 | e22230
Figure 2. Translation of in vitro made mRNAs: Action of eIF2a phosphorylation. A) Cap-luc, EMC-luc, or CrPV IGR-luc mRNAs synthesized in
vitro by T7 RNA polymerase were electroporated in BHK cells and seeded in DMEM (10% FCS). Different amounts of Ars (0, 50, 100 and 200 mM) were
added and cells were incubated for 75 min before harvesting to analyze luc activity. The values shown are percentages of the value of their
respective Ars untreated samples and are means 6 SD of three independent experiments (left panel). The phosphorylated form of eIF2a and total
eIF2a were determined in parallel by Western blot employing specific antibodies (right panel). B) Rabbit reticulocyte lysates were pre-treated or not
with 0.5 mg/ml poly(I:C) for 30 minutes. Subsequently, 100 ng Cap-luc, EMCV-luc, or CrPV IGR-luc mRNAs were added and incubated for 1 h at 30uC.
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org4 July 2011 | Volume 6 | Issue 7 | e22230
with EMC-luc mRNA (Figure S3). It should be noted that EMCV
translation becomes resistant to Tg inhibition as infection
progresses. Thus, there is more inhibition of viral translation by
Tg at 2–3 and 3–4 hpi than at 4–5 and 5–6 hpi. Once again EMC-
luctransfected inthese cellsis excluded from translation, butTg was
suggest that in EMCV infected cells there is not a transacting viral
factor that could confer eIF2-independence. Therefore, in the same
infected cells, EMCV RNAs that are synthesized by the viral
transcription machinery are more resistant to the phosphorylation
of eIF2a than transfected EMC-luc mRNAs.
Induction of eIF2a phosphorylation at the early stages of
During the early stages of EMCV infection, genomic RNA is
released to the cytoplasm for translation, whereas at late times of
infection the viral mRNAs that participate in protein synthesis are
produced by viral transcription. Our present results indicate that
phosphorylation of eIF2a has little effect on viral protein synthesis
in the late phase of EMCV infection. Therefore, we wanted to
analyze the requirement for active eIF2 during the early stages of
EMCV infection. To this end, MEFs were infected with EMCV
and next treated or not with 200 mM Ars for 1 h. The cells were
then washed and incubated with fresh medium and cell samples
were collected at 1, 2, 3, 4, 5 and 6 hpi. EMCV proteins are
evidenced by radioactive labelling at 4 hpi (Figure 3A, lane 4). In
addition, the appearance of EMCV 3D polymerase can be
evidenced by Western blot at 3 hpi (Figure 3B, lane 3). Strikingly,
as the infection progresses, an increase in the phosphorylated form
of eIF2a was observed (Figure 3B). When Ars is added at the
beginning of infection, inhibition of cellular protein synthesis
occurs (Figure 3C, lane 1) and this inhibition correlates with
phosphorylation of eIF2a (Figure 3D, lane 1). After removal of
Ars, the amount of phosphorylated eIF2a decreased to control
levels, while cellular translation recovered. Significantly, viral
protein synthesis is delayed, such that viral proteins start to be
detected at 6 hpi (Figure 3C, lane 6). When eIF2a is
phosphorylated at the beginning of infection, viral protein
synthesis is delayed by about 2 h as compared to control cells.
This finding suggests that for EMCV to begin translation, eIF2a
needs to be dephosphorylated. To further analyze the effect of
eIF2 phosphorylation on early translation of EMCV, cells were
infected with EMCV, and at 2 hpi Ars was added at various
concentrations (0, 50, 100, 200 and 400 mM). Then after 1 h of
incubation, cells were harvested and samples were analyzed by
Western blot using monoclonal anti-3D antibodies. Synthesis of
EMCV 3D was strongly inhibited by the presence of Ars at these
early times of infection, correlating with the phosphorylation of
eIF2a (Figure 3E). In summary, translation of EMCV RNA is
blocked at early times of infection if eIF2a is phosphorylated, while
during the late phase of EMCV infection, viral protein synthesis
can take place in the presence of phosphorylated eIF2a.
Synthesis of EMCV proteins in cells with eIF2a depletion
The use of siRNAs constitutes a useful tool to deplete eIFs in
culture cells in order to examine their functioning during viral
infection. A difficulty with this approach is that total depletion of
the protein to be investigated is rarely achieved, but this approach
maynevertheless indicateto what extenta givenfactorisinvolved in
viral protein synthesis. Another potential problem is that some viral
mRNAs mayexhibita dualmode of translation, requiring thefactor
early in the infection, but not at late times. In this case, a delay in
viral protein synthesis may occur in those cells withpartialdepletion
of the factor, while in strongly depleted cells, abrogation of viral
translation and replication will occur. To assess the involvement of
eIF2a in the translation of EMCV RNA, HeLa cells were depleted
with siRNAs. To achieve this, cells were transfected with a mixture
of four siRNAs designed to deplete eIF2a mRNA. At 36 h after
siRNA transfection, HeLa cells were infected with EMCV. Samples
were recovered at 3, 4, 5, 6 and 7 hpi and labeled proteins were
analyzed (Figure 4A). Western blot analysis against eIF2a indicates
that this factor is silenced by 90% (Figure 4B) and this depletion
blocks cellular protein synthesis by 72% as estimated by
densitometric analyses (Figure 4A). Notably, EMCV protein
synthesis is delayed by about 1–2 h and strongly decreases as
compared to undepleted cells infected with EMCV, although it can
be clearly detected at late times of infection (6 and 7 hpi). The delay
and decrease in EMCV translation in eIF2a-depleted cells are
consistent with the idea that this factor participates in viral
translation early during infection. To estimate the degree in which
EMCV RNA synthesis is affected in eIF2a-depleted HeLa cells,
[3H]uridine incorporation was estimated in presence of 5 mg/ml
actinomycin D (Figure S4). A very strong inhibition was found in
viral RNA replication consistent with the idea that early synthesis of
viral proteins is necessary for genome replication. Despite this
inhibition, once some viral RNA replication has taken place,
translation of EMCV RNA at late times of infection shows little
dependence on the presence of eIF2a (5–8 hpi).
Subcellular localization of eIFs and ribosomes in
Another way to analyze the participation of eIFs in viral protein
synthesis is to investigate their subcellular localization. Cytoplasmic
animal viruses synthesize their proteins in a focal manner,
particularly at late times of infection [30,36,37]. Ribosomes are
present at those foci together with the eIFs that participate in
translation, while those factors that are not involved in protein
synthesis or viral replication are excluded from these foci. To
investigate the subcellular localization of eIF2a after EMCV
infection, MEFs were seeded on glass slides and infected or not
with EMCV. At 5 hpi cells were fixed and incubated with the
corresponding antibodies as indicated in Figures 5 and 6 prior to
immunofluorescence analysis. In mock infected cells, eIF2a
colocalizes with ribosomal protein P in the cytoplasm (Figure 5A).
By contrast, those two proteins do not colocalize in EMCV infected
cells (Figure 5A). EMCV 3D polymerase is clearly observed in the
cytoplasm, indicating the viral replicative sites (Figure 5B). Both the
cytoplasmic sites for viral translation and RNA replication are
located in a perinuclear region and overlap, consistent with the idea
that transcription and translation are coupled processes .
Notably, EMCV 3D does not colocalize with eIF2a. Using the
ImageJ program with the Just Another Co-localization Plugin
(JaCoP), the colocalization rate was calculated . For eIF2a and
ribosomal protein P the Pearson’s Coefficient was 0.96 on the 0 to 1
Luc synthesis was estimated by measuring luc activity. The values shown are percentages of the value of their respective poly(I:C) and are means 6
SD of three independent experiments untreated samples (left panel). The phosphorylated form of eIF2a and total eIF2a were determined in parallel
by Western blot (right panel). C) MEFs cells were infected with EMCV (10 pfu/cell) and next transfected with in vitro made EMC-luc mRNA at different
times after infection. Cells were incubated for 75 min with the transcription mixture containing 5 mg EMC-luc mRNA for each L-24 well in presence or
absence of 1 mM Tg and then collected to measure luc activity. Luc activity values are means 6 SD of three independent experiments.
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org5 July 2011 | Volume 6 | Issue 7 | e22230
scale (0–0.5 indicates no colocalization and 0.5–1, colocalization),
indicating almost total colocalization between eIF2a and ribosomes
in mock-infected cells, while in EMCV infected cells this coefficient
was 0.28, suggesting that there was no colocalization for eIF2a and
ribosomes.Inthe case ofeIF2a andEMCV 3Dprotein,thePearson
Coefficient was 0.32, which further suggests that there is no
colocalization between those two proteins. Therefore, eIF2 is
excluded from viral replicative foci.
To compare the above results with other factors that are involved
in EMCV translation, eIF4G localization was investigated. This
initiation factor is present in the cytoplasm around the nucleus
colocalizing completely with ribosomal protein P in both mock and
EMCV infected cells (Figure 6A); the Pearson Coefficient for eIF4G
and ribosomal protein P was 0.89, and 0.92, respectively. In
addition, eIF4GI also colocalizes with EMCV 3D protein
(Figure 6B) (Pearson’s coefficient 0.9) suggesting that eIF4G
participates in EMCV translation. Therefore, the results obtained
on the subcellular localization of eIF2a further support the notion
that eIF2, contrary to eIF4GI, is not involved in the initiation of
EMCV protein synthesis.
Requirement of active eIF2 for RNA translation with other
After demonstrating that EMCV RNA exhibits a dual mode for
translation—i.e. this RNA requires the presence of active eIF2 in
Figure 3. Treatment with Ars at early times of EMCV infection. A) MEFs were infected with EMCV (10 pfu/cell) and then protein synthesis was
determined by labelling with [35S]Met-Cys every h from 1 to 6 hpi. B) Western blot analysis of the samples obtained in panel A using anti-3D, anti-
phospho-eIF2a or anti-eIF2a antibodies. C) MEFs were infected with EMCV as in panel A and then treated with 200 mM Ars during 1 h (0–1 hpi). Next,
Ars was washed and fresh medium was added. Protein synthesis was analyzed at the time of treatment with Ars and every h thereafter until 6 hpi. D)
Western blot performed with anti-phospho-eIF2a or anti-eIF2a antibodies using the same samples as in panel C. E) MEFs cells were infected with
EMCV (10 pfu/cell) and at 2 hpi treated with different amounts of Ars; one h later samples were harvested and the amount of polymerase 3D
produced was determined by Western blot. The amount of eIF2a phosphorylated and total eIF2a was also determined. The arrows indicate viral
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org6 July 2011 | Volume 6 | Issue 7 | e22230
infected cells early during infection, but not at late times—we
wished to examine the involvement of eIF2 in RNA translation of
other picornaviruses. For this purpose, BHK cells were infected
with FMDV, a member of the Aphtovirus genus, and at 3 hpi Ars
(50, 100 and 200 mM) was added to the medium and incubated for
1 h. Protein synthesis was estimated by incubation with [35S]Met/
Cys during 45 min, from 3.15–4 hpi in the presence or absence of
Ars. Cells were then collected and the synthesized proteins
analyzed (Figure 7A). Phosphorylation of eIF2a and cleavage of
eIF4GI were also analyzed (Figure 7B). As expected, addition of
Ars strongly induced eIF2a phosphorylation. No inhibition of
FMDV protein synthesis was observed by Ars under all the
concentrations tested. By contrast, cellular translation was almost
totally blocked at 100 mM Ars. These findings clearly indicate that
FMDV RNA translation takes place in the presence of
phosphorylated eIF2a during the late phase of infection.
To further analyze the effect of inducers of eIF2a phosphor-
ylation on early and late protein synthesis in picornavirus-infected
cells, two replicons, one from mengovirus and another from PV,
were analyzed. Both replicons contain the luc gene replacing the
coding region for viral structural proteins. Mengovirus is closely
related to EMCV and both belong to the Cardiovirus genus, while
PV is the prototype member of the Enterovirus genus. These
replicons have the advantage that early translation can be assayed
by estimating luc synthesis, whereas the synthesis of late proteins
can be examined by radioactive labelling when the shut-off of host
translation has occurred. After electroporation of these replicons
into BHK cells, 200 mM Ars were added to the culture medium
and luc activity was measured at 75 min. As controls Cap-luc and
CrPV IGR-luc were used. Remarkably, luc synthesis from each
replicon, as well as from Cap-luc mRNA, was drastically inhibited
by Ars, whereas luc synthesis directed by CrPV-luc was unaffected
in the presence of Ars (Figure 8A). The effect of this compound on
late viral translation was assayed using different Ars concentrations
(50, 100 and 200 mM). At late times of replication (5 h post-
electroporation), Ars has little effect on Mengovirus protein
synthesis (Figure 8C). Thus, only 30% inhibition was found in
the presence of 200 mM Ars, whereas cellular translation usually
diminished by over 90% under these conditions. The PV replicon
exhibited a similar behavior to the Mengovirus one as regards the
inhibitory action of Ars (Figure 8D). Altogether, these results
provide strong evidence that synthesis of picornavirus proteins
does not require eIF2a during the late phase of infection.
Participation of eIF2 in the formation of the ternary complex
Met-tRNAi–eIF2-GTP is a crucial event in the initiation of
translation of most mRNAs whether of cellular or viral origin
[16,17]. However, mRNA translation of some viruses, such as
HCV or CrPV, does not require this factor [39,40,41,42]. Our
present observations indicate that picornavirus RNA translation
Figure 4. EMCV infection of Hela cells depleted or not of eIF2a. A) Hela cells transfected with a mixture of siRNAs targeting eIF2a mRNA or
mock Hela cells were infected with EMCV (10 pfu/cell) at 36 h post-transfection. Next, protein synthesis was determined by [35S]Met-Cys labelling at
the hpi indicated. Samples were analyzed by SDS-PAGE, fluorography and autoradiography. B) Western blot analysis of samples from panel A using
anti-eIF2a or anti-phospho-eIF2a antibodies. As a control, the amount of eIF4GI in each sample was determined using specific antibodies against this
factor. The arrows indicate viral proteins.
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org7 July 2011 | Volume 6 | Issue 7 | e22230
takes place when eIF2a is phosphorylated, revealing that this
factor is not necessary to translate this RNA at late times of
infection. If so, the functioning of IRES elements from HCV,
CrPV IGR, and picornaviruses reflects more similarities than
previously suspected. Moreover, some cellular mRNAs bearing
IRES elements can also be translated when eIF2a becomes
Dual mechanism for EMCV translation
The concept that mRNA structure determines the mechanism by
which translation takes place has not been supported by our recent
findings demonstrating that a viral mRNA such as Sindbis virus
subgenomic (SV sg-mRNA) exhibits a dual mode for initiation of
translation . This mRNA requires active eIF2 and intact eIF4F
complex when it is translated in cell free systems or when
electroporated in culture cells. However, SV sg-mRNA efficiently
directs translation in the presence of phosphorylated eIF2a or upon
eIF4G cleavage in virus-infected cells [30,45]. Consistent with these
findings, our present results support the notion that picornavirus
RNA also exhibits a dual mode for its translation. Thus, EMCV
RNA is efficiently translated at late times of infection when eIF2a
has been phosphorylated. By contrast, as shown in this and other
studies, in vitro protein synthesis driven by EMCV IRES is
profoundly blocked upon phosphorylation of eIF2a [41,46]. A
similar situation is found when this IRES containing RNA is
transfected in cells or at early times of EMCV infection. Our
conclusion is that EMCV RNA can be translated following a
canonical mechanism as regards to the use of eIF2, early during
infection. As infection progresses, the cellular environment is
modified such that this RNA can now direct translation in the
absence of active eIF2. Therefore, EMCV RNA has a dual mode
for translation, despite the fact that this RNA possesses the same
structure at early and late times of infection. If true, the mechanism
by which picornavirus RNA is translated would depend on two
parameters: 1) the structure of this mRNA and 2) the environment
in which translation is examined. In addition, our present findings
provide an explanation for the partial resistance of EMCV in cells
that express PKR .
The dual mode of translation for viral mRNAs occurs not only
with SV sg- mRNA and picornavirus RNAs, but also with CrPV
Figure 5. Subcellular localization of eIF2a, ribosomal protein P and EMCV 3D protein in Mock and EMCV infected cells. Hela cells
were seeded on glass coverslips and mock infected or infected with EMCV (10 pfu/cell). At 5 hpi, cells were fixed and permeabilized. A) Ribosomal
protein P and eIF2awere detected by indirect immunofluorescence in mock- and EMCV-infected cells. ToPro 3 indicates the localization of the
nucleus. B) Localization of eIF2a and EMCV 3D proteins. The cell outline was defined by differential interference contrast microscopy (Nomarski).
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org8 July 2011 | Volume 6 | Issue 7 | e22230
mRNA . We have speculated that the presence of viral
proteins is responsible for the switch between these two modes to
initiate translation . In this regard, Hantavirus N protein is
able to replace the eIF4F complex, thus the mechanism of viral
translation in this instance is due to N protein . Also, PV 2Apro
can rescue the translatability of SV sg-mRNAs bearing a
picornavirus IRES .
Picornavirus translation in the presence of
It is puzzling to envisage how EMCV RNA initiation might
occur in the absence of eIF2. Several possibilities exist: one is that
a cellular protein or factor can act as a substitute for eIF2. This
may be the case for HCV RNA translation, where eIF5B acts as a
substitute for eIF2 . It has also been proposed that eIF2A
could act as a substitute for eIF2 in infections with Sindbis virus
. These data have been questioned recently as it has been
suggested that other cellular proteins such as ligatin or MCT-1
and DENR can replace eIF2 during the initiation of HCV protein
synthesis or during the translation of SV sg-mRNA . However,
the authors of that study did not find that ligatin can replace eIF2
for the initiation on EMCV RNA. In fact, ligatin has been
identified as eIF2D . This factor can replace eIF2 for the
translation of some cellular mRNAs. Binding of aminoacyl-tRNA
to the ribosomal P-site is promoted by eIF2D in a GTP-
independent fashion . In principle, it should be possible that
the function of eIF2 was replaced by eIF2A or eIF2D in
picornavirus infected cells. Another possibility is that the IRES
itself directly binds to the 40S, or even to the 80S ribosome, at the
P site late during infection, directly triggering the elongation
phase. If so, the activity of picornavirus IRESs may be more
similar to CrPV IGR IRES than previously thought [40,52].
Therefore, to know exactly which mechanism is acting during
initiation with the different IRES-containing mRNAs thus far
identified in virus species, the mechanism of translation has to be
examined in virus-infected cells. The results obtained in cell free
systems or even in culture cells transfected with these mRNAs may
be misleading and cannot be extrapolated to the physiological
situation. As illustrated in the present work, the mechanism for
initiation of translation on EMCV RNA requires active eIF2 in
Figure 6. Subcellular localization of eIF4G, ribosomal protein P and EMCV 3D protein in infected cells. Hela cells were seeded on glass
coverslips and mock infected or infected with EMCV (10 pfu/cell). At 5 hpi cells were fixed and permeabilized. A) Ribosomal protein P and eIF4G were
detected by indirect immunofluorescence in mock- and EMCV-infected cells. ToPro 3 indicates the localization of the nucleus. B) Localization of eIF4G
and EMCV 3D proteins. The cell outline was defined by differential interference contrast microscopy (Nomarski).
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org9 July 2011 | Volume 6 | Issue 7 | e22230
vitro, this being the mode of translation closer to the canonical
mechanism than that observed in the infected cells during the bulk
of viral translation.
A variety of mechanisms to initiate translation on viral
Two major mechanisms for the initiation of translation are
known in eukaryotic cells: m7G cap-dependent or m7G cap-
independent [16,17]. This division is mainly based on whether or
not a m7G cap structure is present at the 59 end of mRNAs and/
or the requirement for eIF4E during mRNA translation. However,
this simplistic classification may lead to some confusion because
there are capped mRNAs that do not require eIF4E. In some
instances, such as Adenoviruses, Influenza virus or Hantavirus, a
viral protein recognizes the m7G cap structure of viral mRNA and
replaces eIF4E or even eIF4F [47,53,54]. Thus, translation
depends on the presence of a m7G cap structure, but eIF4E is
dispensable. This is also the case of SV sg-mRNA, which does not
require intact eIF4G but still needs the m7G cap structure at the
59-end of this mRNA [45,48]. When defining the mechanism of
initiation it seems more adequate to refer to the eIFs that are
involved in translation . According to whether eIF2 is required
for translation, one of two different mechanisms of initiation is
defined. One is the canonical mechanism that uses the ternary
complex Met-tRNAi-eIF2-GTP while the other does not require
this factor. In this last case a variety of mechanisms can be
operative depending on the type of mRNA examined and the
conditions analyzed. The situation is that depending on the
cellular or viral mRNA considered and the type of assay employed
(in vitro, intact cells, stress situations, viral infections, etc.) the
requirements for eIFs can vary. This picture may be slightly more
complicated if we bear in mind that a given mRNA can exhibit
different mechanisms of initiation, reflecting the plasticity of some
RNAs in accommodating stress situations.
Materials and Methods
Cell Cultures and Viruses
The cell lines used in this work were: HeLa, BHK-21 and mouse
embryo fibroblasts (MEFs). The mouse cell line MEFs(S51A) that
contains an unphosphorylatable form of eIF2a was kindly provided
by D. Ron and R.J.Kaufman (Department of Biological Chemistry,
MI, USA). Cells were grown at 37uC in Dulbecco’s Modified
Eagle’s Medium (DMEM) supplemented with 5% fetal calf serum
(FCS) (HeLa and BHK) or 10% FCS (MEFs) and nonessential
amino acids. Infection with EMCV or with foot-and-mouth disease
virus (FMDV) was carried out at a multiplicity of 10 pfu/cell.
The constructs pKs luc and pTM1-luc have been already
described . These plasmids were used as DNA template to
obtain Cap-luc and EMC-luc mRNA by in vitro transcription using
the T7 RNA polymerase. Plasmid T7 Rluc DEMCV IGR-Fluc
 was employed to obtain CrPV IGR-luc mRNA. The
constructs encoding the PV replicon pRluc31  and the
mengovirus replicon RZ-pMluz  have been already described.
Protein metabolic labeling and Western blot analysis
Protein synthesis was analyzed by replacing DMEM growth
media with 0.2 ml methionine–cysteine free DMEM supplement-
ed with 2 ml EasyTagTMEXPRESS
[35S]Met-Cys (11 mCi/ml, 37.0 Tbq/mmol; Perkin Elmer) per
well of an L-24 plate. Cultures were pre-treated with the amounts
indicated in each case of dithiothreitol (DTT), thapsigargin (Tg) or
arsenite (Ars) for 15 min, before labeling for 45 min in the
presence of the tested compounds. The cells were then collected in
the appropriate gel loading buffer (62.5 mM Tris–HCl, pH 6.8,
2% SDS, 0.1 M DTT, 17% glycerol, and 0.024% bromophenol
blue) and analyzed by electrophoresis in SDS-polyacrylamide gels
(SDS-PAGE), followed by fluorography and autoradiography.
Specific rabbit polyclonal antibodies raised against phospho-eIF2a
(Ser 51) (Cell Signaling Technology) or total eIF2a (Santa Cruz
Biotechnology) were used in Western blot analysis at 1:1000
dilution antisera. Mouse monoclonal antibodies raised against
EMCV 3D protein (a generous gift from A. Palmenberg,
University of Wisconsin, Madison, USA) were used at 1:1000
dilution. Rabbit polyclonal antibodies raised against peptides
derived from the N- and C-terminal regions of human eIF4GI
were also used at a 1:1000 dilution . Anti-rabbit and anti-
mouse immunoglobulin G antibodies coupled to peroxidase
(Promega) were used at a 1:5000 dilution.
35S Protein Labeling mix,
In vitro transcription and transfection
Plasmids were used as templates for in vitro RNA transcription
with T7 RNA polymerase (Promega). To obtain Cap-luc mRNA,
an m7G(59)ppp(59)G cap analog was added to the transcription
mixture. For transfection, subconfluent BHK cells were harvested,
washed with ice-cold phosphate-buffered saline (PBS), and
Figure 7. Action of Ars on FMDV infection. A) BHK cells were mock
infected or infected with FMDV (10 pfu/cell). At 3 hpicultures were
treated with different amounts of Ars (0, 50, 100 and 200 mM) for
15 min and next labeled by [35S]Met-Cys labelling in presence of the
same concentrations of Ars from 3.15–4 hpi. Samples were collected
and analyzed by SDS-PAGE, fluorography and autoradiography. B) The
phosphorylated form of eIF2a and total eIF2a were determined in
parallel by Western blot with specific antibodies. The cleavage of eIF4GI
was also analyzed by Western blot using specific antibodies against this
factor. The arrows indicate viral proteins.
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org10 July 2011 | Volume 6 | Issue 7 | e22230
resuspended at a density of approximately 2.56106cells/ml in the
same buffer. Subsequently, 20 mg of in vitro transcribed RNA were
added to 0.4 ml cell suspension and the mixture was transferred to a
by generating two consecutive 1.5-kV, 25-mF pulses with a Gene
Pulser apparatus (Bio-Rad), as previously described .
In vitro translation
In vitro translation was carried out in rabbit reticulocyte lysates.
To induce phosphorylation of eIF2a, extracts were treated with
0.5 mg/ml poly(I:C) (Pharmacia Biotech) for 30 min. Subsequent-
ly, 100 ng of different mRNAs were added and incubated for 1 h
at 30uC. Protein synthesis was estimated by measuring luc activity.
Figure 8. Translation of Mengovirus and PV replicons. Effect of Ars on early and late viral protein synthesis. A) BHK cells were
electroporated with Mengo-luc, Polio-luc replicons, Cap-luc or CrPV IGR-luc mRNAs; all these RNAs were synthesized by in vitro transcription.
Electroporated cells were seeded in DMEM (10% FCS) in presence or absence of 200 mM Ars. 75 min later cells were collected and lysed to measure
luc activity. The values shown are percentages of the value of their respective Ars untreated samples and are means 6 SD of three independent
experiments. B) The phosphorylated form of eIF2a and total eIF2a were determined in parallel by Western blot with specific antibodies. C and D) BHK
cells were electroporated with Mengo-luc replicon (C) or polio-luc replicon (D) and at 5 h post-electroporation protein synthesis was determined by
[35S]Met-Cys labelling in presence of different concentrations of Ars (0, 50, 100 and 200 mM). The arrows indicate viral proteins.
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org 11 July 2011 | Volume 6 | Issue 7 | e22230
Transfection of HeLa cells
To transfect interference RNAs (siRNAs), HeLa cells were
grown in 24-well plates with antibiotic- and antimycotic free
DMEM supplemented with 5% FCS to 60–70% confluence. To
make up the transfection mixture, 2 ml of Lipofectamine 2000
(Invitrogen) were added to 50 ml of Opti-MEM I Reduced Serum
Medium (Opti-MEM I) (Invitrogen) and then incubated for 5 min
at room temperature. Simultaneously, the siRNA mixture was
prepared with 100 pmol of a mixture of four siRNAs targeting
eIF2a mRNA (Dharmacon; Thermo Scientific) in 50 ml of the
Opti-MEM I for each L-24 well and then incubated at room
temperature for 5 min. The final mixture was subsequently
prepared with 50 ml of Lipofectamine suspension and 50 ml of
the siRNA mixture by incubation for 30 minutes at room
temperature. To transfect HeLa cells with siRNAs, cell medium
was removed and 100 ml of Opti-MEM I followed by 100 ml of the
transfection/siRNAs mixture obtained were added to each well.
Cells were then incubated at 37uC for 4 h. After incubation, the
transfection medium was removed and the cultures continued in
fresh medium. At 36 h post-transfection HeLa cells were infected
with EMCV (10 pfu/cell) to determine viral protein synthesis.
Luciferase activity measurement
HeLa cells were harvested with buffer containing 25 mM
glycylglycine (pH 7.8), 0.5% of Triton X-100 and 1 mM dithio-
threitol. Luciferase (luc) activity was measured using Moonlight
2000 apparatus (Analytical Luminescence Laboratory) using the
Luciferase Assay System (Promega).
HeLa cells were seeded on glass cover slips prior to infection with
EMCV (10 pfu/cell). At 5 hpi, cells were fixed in 4% PFA for
15 min, washed twice withPBS, and then permeabilized for 10 min
with 0.2% Triton X-100 in PBS. Subsequent antibody incubations
were carried out for 2 h with specified primary antisera and
correspondingfluorescence-conjugated secondaryantibody at room
temperature. Cover slips were then mounted in ProLong Gold anti-
fade reagent (Invitrogen) and examined with a Zeiss LSM510
Inverted confocal laser-scanning microscope (Bio-Rad/Zeiss) with
Plan-Apochromat 63X/1.4 oil objective. Mouse monoclonal
antibodies raised against eukaryotic ribosomal P protein , or
EMCV 3D protein (a gift from A. Palmenberg, University of
Wisconsin, Madison, USA) were used for immunofluorescence at
1:10 and 1:200 dilutions, respectively. Rabbit polyclonal antibodies
raisedagainst eIF4GIoreIF2a wereused at1:100dilution.To-pro3
(Invitrogen) was employed at 1:500 dilution.
sphorylated eIF2a in culture cells. Effect of inhibitors.
HeLa cells were untreated or treated for 60 min with 200 mM Ars
or 400 mM DTT. Afterwards cell monolayers were collected and
proteins were separated by isoelectric focusing and transferred to a
nitrocellulose membrane as described before . The phosphor-
ylated and unphosphorylated forms of eIF2a were detected by
anti-eIF2a rabbit polyclonal antibodies and quantified by
densitometric scanning of the corresponding bands.
Analysis of phosphorylated and unpho-
synthesis was analyzed in MEFs or MEFs(S51A) treated with
different concentrations of Ars as indicated in the Figure. Culture
cells were pretreated for 15 min with Ars in DMEM without
methionine and cysteine. Then, 15 mCi of [35S]Met-Cys for each L-
24 well were added and incubation was continued for 1 h. Cells were
collected in sample buffer and proteins synthesized during this time
were analyzed bySDS-PAGE, fluorographyand autoradiography as
described in Materials and Methods. B) Luc synthesis in MEFs or
MEFs(S51A) transfected with EMC-luc mRNA in the presence of
different concentrations of Ars. Culture cells were transfected with 5
mg of EMC-luc mRNAper wellof an L-24 plate inthe presenceof 0,
200 or 400 mM Ars. 75 min later cell monlayers were collected and
lysed to measure luc activity. The percentage to the values of the
respective samples untreated with Ars is represented. Luc activity
values are means 6 SD of three independent experiments.
Effects of Ars on translation in MEFs. A) Protein
EMCV-infected cells. A) BHK cells were infected with EMCV
(10 pfu/cell) and next transfected with EMC-luc mRNA at
different times after infection. The cells were incubated for 75 min
with the transcription mixture containing 5 mg EMC-luc mRNA
per well of an L-24 plate in presence or absence of 1 mM Tg and
then collected to measure luc activity. Luc activity values are
means 6 SD of three measures of the same experiment. B) Protein
synthesis was analyzed in parallel. In this case the cultures were
treated or not with 1 mM Tg for 15 min before adding 15 mCi of
[35S]-Met, Cys per well of an L-24 plate, and continue the
incubation for 1 h. The arrows indicate viral proteins.
Translation of EMC-luc mRNA transfected in
cells. Hela cells transfected with a mixture of siRNAs targeting
eIF2amRNA or mock Hela cells were infected with EMCV (10
pfu/cell) at 36 h post-transfection. Viral RNA was subsequently
labeled with [3H]uridine (20 mCi/ml, final concentration) in the
presence of 5 mg/ml actinomycin D. At the indicated hpi
[3H]uridine incorporated was quantified in a liquid scintillation
spectrometer as described before . Cpm values are means 6
SD of three measures of the same experiment.
EMCV RNA synthesis in eIF2-depleted HeLa
We wish to thank J.P.G.-Ballesta (Centro de Biologia Molecular ‘‘Severo
Ochoa’’, Madrid) for kindly providing the anti-ribosome antibodies, A.
Palmenberg, (University of Wisconsin, Madison, USA) for the mouse
monoclonal antibodies against EMCV 3D protein and the mengovirus
replicon RZ-pMluz, E. Martı ´nez-Salas (Centro de Biologia Molecular
‘‘Severo Ochoa’’, Madrid) for providing a sample of FMDV, P. Sarnow
(Stanford University, California, USA) and Y. Kusov (University of
Lu ¨beck, Germany) for providing plasmids T7 Rluc DEMCV IGR-Fluc and
the PV replicon pRluc31 respectively.
Conceived and designed the experiments: EW MAS NR LC. Performed
the experiments: EW MAS NR. Analyzed the data: EW MAS NR LC.
Wrote the paper: LC MAS.
1. Agol V (2002) Picornavirus genome: an overview. In: BLSaE W, ed. Molecular
biology of picornavirus. Washington, D.C.: ASM Press. pp 127–148.
ed. Molecular biology of picornaviruses. WhasingtonD.C.: ASM Press. pp 61–70.
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org 12July 2011 | Volume 6 | Issue 7 | e22230
3. Hogle JM (2002) Poliovirus cell entry: common structural themes in viral cell Download full-text
entry pathways. Annu Rev Microbiol 56: 677–702.
4. Paul A (2002) Possible Unyfying Mechanism of picornavirus Genome
Replication. In: Semler BWE, ed. Molecular biology of picornaviruses.
Washington, D.C.: ASM Press. pp 227–246.
5. Belov GA, Ehrenfeld E (2007) Involvement of cellular membrane traffic proteins
in poliovirus replication. Cell Cycle 6: 36–38.
6. Kerr IM, Martin EM (1971) Virus protein synthesis in animal cell-free systems:
nature of the products synthesized in resonse to ribonucleic acid of
encephalomyocarditis virus. J Virol 7: 438–447.
7. Schreier MH, Staehelin T (1973) Initiation of eukaryotic protein synthesis: (Met-
tRNA f -40S ribosome) initiation complex catalysed by purified initiation factors
in the absence of mRNA. Nat New Biol 242: 35–38.
8. Scheper GC, Thomas AA, Voorma HO (1991) The 59 untranslated region of
encephalomyocarditis virus contains a sequence for very efficient binding of
eukaryotic initiation factor eIF-2/2B. Biochim Biophys Acta 1089: 220–226.
9. Svitkin YV, Meerovitch K, Lee HS, Dholakia JN, Kenan DJ, et al. (1994)
Internal translation initiation on poliovirus RNA: further characterization of La
function in poliovirus translation in vitro. J Virol 68: 1544–1550.
10. Thomas AA, Rijnbrand R, Voorma HO (1996) Recognition of the initiation
codon for protein synthesis in foot-and-mouth disease virus RNA. J Gen Virol
77(Pt 2): 265–272.
11. Pestova TV, Kolupaeva VG, Lomakin IB, Pilipenko EV, Shatsky IN, et al.
(2001) Molecular mechanisms of translation initiation in eukaryotes. Proc Natl
Acad Sci U S A 98: 7029–7036.
12. Kolupaeva VG, Lomakin IB, Pestova TV, Hellen CU (2003) Eukaryotic
initiation factors 4G and 4A mediate conformational changes downstream of the
initiation codon of the encephalomyocarditis virus internal ribosomal entry site.
Mol Cell Biol 23: 687–698.
13. Pelletier J, Sonenberg N (1988) Internal initiation of translation of eukaryotic
mRNA directed by a sequence derived from poliovirus RNA. Nature 334:
14. Belsham GJ (2009) Divergent picornavirus IRES elements. Virus Res 139:
15. Merrick WC (2004) Cap-dependent and cap-independent translation in
eukaryotic systems. Gene 332: 1–11.
16. Pestova TLJR, Hellen CU (2007) The mechanism of translation initiation in
eukaryotes. In: MB Matheus NS JH, ed. Translational Control in Biology and
Medicine. New York, USA: Cold Spring Harbor Laboratory Press. pp 87–128.
17. Sonenberg N, Hinnebusch AG (2009) Regulation of translation initiation in
eukaryotes: mechanisms and biological targets. Cell 136: 731–745.
18. Chen JJ, London IM (1995) Regulation of protein synthesis by heme-regulated
eIF-2 alpha kinase. Trends Biochem Sci 20: 105–108.
19. Fernandez J, Yaman I, Sarnow P, Snider MD, Hatzoglou M (2002) Regulation
of internal ribosomal entry site-mediated translation by phosphorylation of the
translation initiation factor eIF2alpha. J Biol Chem 277: 19198–19205.
20. Wek RC, Jiang HY, Anthony TG (2006) Coping with stress: eIF2 kinases and
translational control. Biochem Soc Trans 34: 7–11.
21. Hershey JW (1989) Protein phosphorylation controls translation rates. J Biol
Chem 264: 20823–20826.
22. Ransone LJ, Dasgupta A (1987) Activation of double-stranded RNA-activated
protein kinase in HeLa cells after poliovirus infection does not result in increased
phosphorylation of eucaryotic initiation factor-2. J Virol 61: 1781–1787.
23. Ransone LJ, Dasgupta A (1988) A heat-sensitive inhibitor in poliovirus-infected
cells which selectively blocks phosphorylation of the alpha subunit of eucaryotic
initiation factor 2 by the double-stranded RNA-activated protein kinase. J Virol
24. Black TL, Safer B, Hovanessian A, Katze MG (1989) The cellular 68,000-Mr
protein kinase is highly autophosphorylated and activated yet significantly
degraded during poliovirus infection: implications for translational regulation.
J Virol 63: 2244–2251.
25. O’Neill RE, Racaniello VR (1989) Inhibition of translation in cells infected with
a poliovirus 2Apro mutant correlates with phosphorylation of the alpha subunit
of eucaryotic initiation factor 2. J Virol 63: 5069–5075.
26. Black TL, Barber GN, Katze MG (1993) Degradation of the interferon-induced
68,000-M(r) protein kinase by poliovirus requires RNA. J Virol 67: 791–800.
27. DeStefano J, Olmsted E, Panniers R, Lucas-Lenard J (1990) The alpha subunit
of eucaryotic initiation factor 2 is phosphorylated in mengovirus-infected mouse
L cells. J Virol 64: 4445–4453.
28. Rice AP, Duncan R, Hershey JW, Kerr IM (1985) Double-stranded RNA-
dependent protein kinase and 2-5A system are both activated in interferon-
treated, encephalomyocarditis virus-infected HeLa cells. J Virol 54: 894–898.
29. Meurs EF, Watanabe Y, Kadereit S, Barber GN, Katze MG, et al. (1992)
Constitutive expression of human double-stranded RNA-activated p68 kinase in
murine cells mediates phosphorylation of eukaryotic initiation factor 2 and
partial resistance to encephalomyocarditis virus growth. J Virol 66: 5805–5814.
30. Sanz MA, Castello A, Ventoso I, Berlanga JJ, Carrasco L (2009) Dual
mechanism for the translation of subgenomic mRNA from Sindbis virus in
infected and uninfected cells. PLoS One 4: e4772.
31. Garrey JL, Lee YY, Au HH, Bushell M, Jan E (2010) Host and viral translational
mechanisms during cricket paralysis virus infection. J Virol 84: 1124–1138.
32. Harding HP, Zhang Y, Ron D (1999) Protein translation and folding are
coupled by an endoplasmic-reticulum-resident kinase. Nature 397: 271–274.
33. Brostrom CO, Brostrom MA (1998) Regulation of translational initiation during
cellular responses to stress. Prog Nucleic Acid Res Mol Biol 58: 79–125.
34. Sanz MA, Castello A, Carrasco L (2007) Viral translation is coupled to
transcription in Sindbis virus-infected cells. J Virol 81: 7061–7068.
35. Pestova TV, Hellen CU, Shatsky IN (1996) Canonical eukaryotic initiation
factors determine initiation of translation by internal ribosomal entry. Mol Cell
Biol 16: 6859–6869.
36. Katsafanas GC, Moss B (2007) Colocalization of transcription and translation
within cytoplasmic poxvirus factories coordinates viral expression and subjugates
host functions. Cell Host Microbe 2: 221–228.
37. Castello A, Quintas A, Sanchez EG, Sabina P, Nogal M, et al. (2009) Regulation
of host translational machinery by African swine fever virus. PLoS Pathog 5:
38. Bolte S, Cordelieres FP (2006) A guided tour into subcellular colocalization
analysis in light microscopy. J Microsc 224: 213–232.
39. Terenin IM, Dmitriev SE, Andreev DE, Shatsky IN (2008) Eukaryotic
translation initiation machinery can operate in a bacterial-like mode without
eIF2. Nat Struct Mol Biol 15: 836–841.
40. Deniz N, Lenarcic EM, Landry DM, Thompson SR (2009) Translation
initiation factors are not required for Dicistroviridae IRES function in vivo.
RNA 15: 932–946.
41. Wilson JE, Pestova TV, Hellen CU, Sarnow P (2000) Initiation of protein
synthesis from the A site of the ribosome. Cell 102: 511–520.
42. Lancaster AM, Jan E, Sarnow P (2006) Initiation factor-independent translation
mediated by the hepatitis C virus internal ribosome entry site. RNA 12:
43. Gerlitz G, Jagus R, Elroy-Stein O (2002) Phosphorylation of initiation factor-2
alpha is required for activation of internal translation initiation during cell
differentiation. Eur J Biochem 269: 2810–2819.
44. Fernandez J, Yaman I, Merrick WC, Koromilas A, Wek RC, et al. (2002)
Regulation of internal ribosome entry site-mediated translation by eukaryotic
initiation factor-2alpha phosphorylation and translation of a small upstream
open reading frame. J Biol Chem 277: 2050–2058.
45. Castello A, Sanz MA, Molina S, Carrasco L (2006) Translation of Sindbis virus
26S mRNA does not require intact eukariotic initiation factor 4G. J Mol Biol
46. Shimoike T, McKenna SA, Lindhout DA, Puglisi JD (2009) Translational
insensitivity to potent activation of PKR by HCV IRES RNA. Antiviral Res 83:
47. Mir MA, Panganiban AT (2008) A protein that replaces the entire cellular eIF4F
complex. EMBO J 27: 3129–3139.
48. Sanz MA, Welnowska E, Redondo N, Carrasco L (2010) Translation driven by
picornavirus IRES is hampered from Sindbis virus replicons: rescue by
poliovirus 2A protease. J Mol Biol 402: 101–117.
49. Ventoso I, Sanz MA, Molina S, Berlanga JJ, Carrasco L, et al. (2006)
Translational resistance of late alphavirus mRNA to eIF2alpha phosphorylation:
a strategy to overcome the antiviral effect of protein kinase PKR. Genes Dev 20:
50. Skabkin MA, Skabkina OV, Dhote V, Komar AA, Hellen CU, et al. (2010)
Activities of Ligatin and MCT-1/DENR in eukaryotic translation initiation and
ribosomal recycling. Genes Dev 24: 1787–1801.
51. Dmitriev SE, Terenin IM, Andreev DE, Ivanov PA, Dunaevsky JE, et al. (2010)
GTP-independent tRNA delivery to the ribosomal P-site by a novel eukaryotic
translation factor. J Biol Chem 285: 26779–26787.
52. Pestova TV, Lomakin IB, Hellen CU (2004) Position of the CrPV IRES on the
40S subunit and factor dependence of IRES/80S ribosome assembly. EMBO
Rep 5: 906–913.
53. Xi Q, Cuesta R, Schneider RJ (2004) Tethering of eIF4G to adenoviral mRNAs
by viral 100k protein drives ribosome shunting. Genes Dev 18: 1997–2009.
54. Burgui I, Yanguez E, Sonenberg N, Nieto A (2007) Influenza virus mRNA
translation revisited: is the eIF4E cap-binding factor required for viral mRNA
translation? J Virol 81: 12427–12438.
55. Shatsky IN, Dmitriev SE, Terenin IM, Andreev DE (2010) Cap- and IRES-
independent scanning mechanism of translation initiation as an alternative to the
concept of cellular IRESs. Mol Cells 30: 285–293.
56. Wilson JE, Powell MJ, Hoover SE, Sarnow P (2000) Naturally occurring
dicistronic cricket paralysis virus RNA is regulated by two internal ribosome
entry sites. Mol Cell Biol 20: 4990–4999.
57. Andino R, Rieckhof GE, Achacoso PL, Baltimore D (1993) Poliovirus RNA
synthesis utilizes an RNP complex formed around the 59-end of viral RNA.
EMBO J 12: 3587–3598.
58. Fata-Hartley CL, Palmenberg AC (2005) Dipyridamole reversibly inhibits
mengovirus RNA replication. J Virol 79: 11062–11070.
59. Aldabe R, Feduchi E, Novoa I, Carrasco L (1995) Efficient cleavage of p220 by
poliovirus 2Apro expression in mammalian cells: effects on vaccinia virus.
Biochem Biophys Res Commun 215: 928–936.
60. Vilella MD, Remacha M, Ortiz BL, Mendez E, Ballesta JP (1991)
Characterization of the yeast acidic ribosomal phosphoproteins using monoclo-
nal antibodies. Proteins L44/L45 and L449 have different functional roles.
Eur J Biochem 196: 407–414.
Picornavirus Translation without eIF2
PLoS ONE | www.plosone.org13 July 2011 | Volume 6 | Issue 7 | e22230