Specific inhibition of the PKR-mediated antiviral response by the murine cytomegalovirus proteins m142 and m143.

JOURNAL OF VIROLOGY, Feb. 2009, p. 1260–1270 Vol. 83, No. 3
0022-538X/09/$08.00�0 doi:10.1128/JVI.01558-08
Copyright © 2009, American Society for Microbiology. All Rights Reserved.
Specific Inhibition of the PKR-Mediated Antiviral Response by the
Murine Cytomegalovirus Proteins m142 and m143�
Matthias Budt,1 Lars Niederstadt,1 Ralitsa S. Valchanova,1† Stipan Jonjic´,2 and Wolfram Brune1*
Division of Viral Infections, Robert Koch Institute, Nordufer 20, 13353 Berlin, Germany,1 and Department of Histology and
Embryology, Faculty of Medicine, University of Rijeka, Brace Branchetta 20, 51000 Rijeka, Croatia2
Received 23 July 2008/Accepted 12 November 2008
Double-stranded RNA (dsRNA) produced during viral infection activates several cellular antiviral responses.
Among the best characterized is the shutoff of protein synthesis mediated by the dsRNA-dependent protein kinase
(PKR) and the oligoadenylate synthetase (OAS)/RNase L system. As viral replication depends on protein synthesis,
many viruses have evolved mechanisms for counteracting the PKR and OAS/RNase L pathways. The murine
cytomegalovirus (MCMV) proteins m142 andm143 have been characterized as dsRNA binding proteins that inhibit
PKR activation, phosphorylation of the translation initiation factor eIF2�, and a subsequent protein synthesis
shutoff. In the present study we analyzed the contribution of the PKR- and the OAS-dependent pathways to the
control of MCMV replication in the absence or presence of m142 and m143. We show that the induction of eIF2�
phosphorylation during infection with an m142- and m143-deficient MCMV is specifically mediated by PKR, not by
the related eIF2� kinases PERK or GCN2. PKR antagonists of vaccinia virus (E3L) or herpes simplex virus (�34.5)
rescued the replication defect of anMCMV strain with deletions of bothm142 andm143.Moreover, m142 andm143
bound to each other and interacted with PKR. By contrast, an activation of the OAS/RNase L pathway by MCMV
was not detected in the presence or absence of m142 and m143, suggesting that these viral proteins have little or no
influence on this pathway. Consistently, an m142- and m143-deficient MCMV strain replicated to high titers in
fibroblasts lacking PKR but did not replicate in cells lacking RNase L. Hence, the PKR-mediated antiviral response
is responsible for the essentiality of m142 and m143.
The intrusion of an infectious agent is noticed by target cells
through specific receptors that recognize pathogen-associated
molecular patterns (32). These sensors trigger the induction of
an antimicrobial response aimed at elimination of the patho-
gen. Many different structural features of microbes can activate
such a response, among them virus-associated nucleic acids
such as long double-stranded RNA (dsRNA), which is absent
from uninfected cells (51). dsRNA not only constitutes the
genetic material of dsRNA viruses but is also produced in
infected cells by positive-strand RNA viruses and some DNA
viruses, especially those with large genomes and genes ar-
ranged on both strands of the viral DNA genome (63).
Toll-like receptor 3 (TLR3) and the RNA helicases RIG-I
and MDA5 serve as sensors for dsRNA. Upon activation, they
induce signaling cascades culminating in the expression of type
I interferons (IFNs) (58). These IFNs induce the expression of
a plethora of antiviral genes, which can interfere with the viral
replication cycle (54). The IFN-inducible gene products com-
prise the dsRNA-dependent protein kinase (PKR) and oligo-
adenylate synthetases (OAS). Both PKR and OAS are directly
activated by dsRNA. Hence, dsRNA induces the expression of
these antiviral effector proteins and is also necessary for their
activation.
Upon binding to dsRNA, PKR dimerizes and undergoes
autophosphorylation to gain full catalytic activity (30, 46, 59).
Once activated, PKR phosphorylates the eukaryotic transla-
tion initiation factor eIF2� (10, 39). In its phosphorylated
state, eIF2� forms a stable complex with the nucleotide ex-
change factor eIF2B, which is then no longer recycled for
initiation of protein translation by GDP/GTP exchange (57).
Consequently, PKR activation leads to a global block to pro-
tein synthesis in the infected cell, which can hamper the pro-
duction of virus progeny. However, it is important to note that
while eIF2� is inactivated by PKR, it also constitutes an im-
portant cellular stress checkpoint utilized by three additional
signaling pathways. These are activated by different cellular
malfunctions, all of which require a temporary halt of protein
synthesis to overcome the cause of stress. The PKR-related
endoplasmic reticulum kinase (PERK) responds to protein
overload in the endoplasmic reticulum (18), while the kinase
GCN2 (general control non-derepressible 2) reacts to dysregu-
lation of amino acid metabolism (64) or to UV light (12).
Finally, the heme-regulated inhibitor HRI functions as a
checkpoint for hemoglobin biosynthesis in reticulocytes (34).
So far, antiviral functions have been reported only for the
eIF2� kinases PKR, PERK, and GCN2 (3, 15, 26), and only
these will be considered in the present study.
The members of the OAS protein family are encoded by
several cellular genes and possess a very specific catalytic ac-
tivity in being able to condense ATP molecules via unusual
2�-5�-phosphodiester linkages (23, 49). The resulting oligomers
of variable length bind and activate the latent RNase L en-
zyme, which then cleaves different RNA species, among them
viral RNAs, mRNAs, and rRNAs (14, 55, 65). As a result,
synthesis of viral proteins is inhibited, and viral RNA genomes
are directly destroyed.
* Corresponding author. Mailing address: Division of Viral Infec-
tions, Robert Koch Institute, Nordufer 20, 13353 Berlin, Germany.
Phone: 49 30 187542502. Fax: 49 30 18107542502. E-mail: BruneW
@rki.de.
† Present address: Institute of Physiology, Charite´-Universita¨tsmedizin
Berlin, Arnimallee 22, 14195 Berlin, Germany.
� Published ahead of print on 19 November 2008.
1260
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

The prominent antiviral role of PKR is underscored by the
fact that many different viruses have evolved proteins that in
one way or another prevent the activation of the PKR signaling
cascade. For instance, the vaccinia virus (VV) E3L, the herpes
simplex virus type 1 (HSV-1) US11, and influenza virus NS1
proteins bind dsRNA and PKR and inhibit its activation (7, 35,
42). The �34.5 protein, also from HSV-1, functions as a regu-
latory subunit of the cellular PP1 phosphatase, directing it to
dephosphorylate eIF2�, thereby terminating the PKR-induced
signal (20).
Murine cytomegalovirus (MCMV) is a representative of the
Herpesviridae and possesses several properties that should
make it vulnerable to the antiviral host response: a large DNA
genome with genes encoded on both strands, harboring the
potential for extensive dsRNA formation; a prolonged repli-
cation cycle allowing enough time for the cell to mount an
effective defense; and finally the establishment of a latent state
in an organism with an activated immune response (56).
Therefore, it is not surprising that MCMV encodes several
different proteins that interfere with various host defense
mechanisms.
We have recently shown that the MCMV proteins m142 and
m143 are essential for MCMV replication and are both re-
quired to inhibit PKR activation, eIF2� phosphorylation, and
a shutdown of protein synthesis (62). Child and colleagues
showed that the two proteins can be coimmunoprecipitated
from lysates of infected cells and function together to bind
dsRNA (9). Thus, MCMV’s system differs from other viral
systems (such as E3L, NS1, or US11) in that a synergism of two
different proteins is required to prevent PKR activation. How-
ever, whether these viral proteins also inhibit other dsRNA-
dependent host defenses and whether other eIF2� kinases
contribute to the protein synthesis shutoff have not been de-
termined.
In the present study, we show that the function of m142 and
m143 is specifically directed against the PKR pathway. When
cells were infected by an MCMV strain lacking these genes,
eIF2� was phosphorylated specifically by PKR, not by the
related eIF2� kinases PERK or GCN2. The OAS/RNase L
pathway, by contrast, was not detectably activated during
MCMV infection. Consistent with these findings, an m142- and
m143-deficient MCMV replicated to titers equivalent to the
wild-type (wt) virus in PKR-deficient cells but did not grow in
RNase L-deficient cells. These results suggested that PKR is
the predominant (if not the only) target of m142 and m143.
MATERIALS AND METHODS
Virus mutagenesis. All mutant viruses were constructed on the basis of an
MCMV variant expressing the green fluorescent protein (MCMV-GFP), which
has been used in previous studies (4, 27). The MCMV-GFP genome, cloned as
a bacterial artificial chromosome (BAC), was modified in the Escherichia coli
strain DY380 as described previously (4, 5). For generation of an MCMV mutant
with a deletion of both m142 and m143 (MCMV��m142/m143), a zeocin resis-
tance gene (zeo) was PCR amplified with primers that contained 50-nucleotide
sequences homologous to positions 199621 to 199670 and 202594 to 202643,
respectively, and used for homologous recombination. For generation of a re-
vertant virus, a fragment spanning the genomic region of m142 and m143 (nu-
cleotides 195201 to 200853 of the MCMV genome [MCMV Rev m142/m143])
was excised with EcoRI and SspI from a pUC19 plasmid containing the cloned
MCMV strain Smith HindIII/I region (kindly provided by M. Messerle, Han-
nover Medical School, Hannover, Germany). NIH 3T3 cells were transfected
with the purified fragment and superinfected with MCMV��m142/m143 2 h
posttransfection. Since wt NIH 3T3 cells do not support replication of the
MCMV��m142/m143 mutant virus, only revertant virus was recovered.
The US11 and �34.5 genes from HSV-1 and E3L from VV were introduced into
the region of m02 to m06 (m02-m06) of MCMV��m142/m143 by homologous
recombination by means of the pReplacer plasmid, as previously described (27). The
VV E3L gene was purchased as a synthetic, codon-optimized (for expression in
mouse cells) version from Geneart (Regensburg, Germany). The synthetic E3L
sequence is as follows: 5�-ATGAGCAAGATCTACATCGACGAGAGAAGCGA
CGCCGAGATCGTGTGCGCCGCCATCAAGAACATCGGCATCGAGGGC
GCCACAGCCGCCCAGCTGACCAGGCAGCTGAACATGGAGAAGCGGG
AGGTGAACAAGGCCCTGTACGACCTGCAGAGAAGCGCCATGGTGTA
CAGCAGCGACGACATCCCCCCCAGATGGTTCATGACCACCGAGGCCG
ACAAGCCCGACGCCGACGCCATGGCCGACGTGATCATCGACGACGTG
AGCCGGGAGAAGAGCATGAGAGAGGACCACAAGAGCTTCGACGATG
TGATCCCCGCCAAGAAGATCATCGACTGGAAGGACGCCAACCCCGTG
ACCATCATCAACGAGTACTGCCAGATCACCAAGAGGGACTGGTCCTT
CAGGATCGAGAGCGTGGGACCCAGCAACAGCCCCACCTTCTACGCCT
GCGTGGACATCGACGGCAGGGTGTTCGACAAGGCCGACGGCAAGAG
CAAGAGGGACGCCAAGAACAACGCCGCCAAGCTGGCCGTGGACAAG
CTGCTGGGCTACGTGATCATCAGGTTCTGA-3�. The HSV-1 US11 gene was
amplified by PCR from pCIneo-US11 (kindly provided by I. Mohr, New York
University, NY), introducing a hemagglutinin (HA) epitope tag at the 5� end, and
the �34.5 gene was excised from pRC-CMV-�34.5. The three genes were inserted
into the multiple cloning site of pReplacer (27). A fragment containing a kan
selection marker, the viral gene driven by a phosphoglycerate kinase promoter, and
homology arms for homologous recombination was excised from the pReplacer
plasmid with KpnI and SacI and used for inserting the sequence into the m02-m06
region of the MCMV��m142/m143 BAC as described previously (27). The resulting
BACs were purified with Nucleobond PC100 columns (Macherey and Nagel, Du¨ren,
Germany) and transfected into immortalized PKR�/� fibroblasts, which support full
replication of m142- and m143-deficient viruses.
Viral DNA from MCMV Rev m142/m143 was isolated from supernatants of
infected cells essentially as described previously (48). Genomic integrity was
confirmed by EcoRI digestion and separation on ethidium bromide-stained 0.6%
agarose gels and by sequencing the mutated sites.
Cells and viruses. NIH 3T3 fibroblasts were cultured in Dulbecco’s modified
Eagle’s medium (DMEM) with 10% newborn calf serum. Murine 10.1 fibroblasts
(19) and human 293A cells (Invitrogen) were grown in DMEM with 5% fetal calf
serum. Primary murine embryonic fibroblasts (MEFs) from BALB/c mice were
used at passage 3 for experiments. Primary PKR�/� MEFs (66) were provided by
U. Kalinke (Paul Ehrlich Institute, Langen, Germany). Immortalized 3T3-like
fibroblasts were generated from these cells by continuous passaging, as described
previously (66). Immortalized RNase L�/� fibroblasts and corresponding wt
fibroblasts (67) were provided by R. H. Silverman (Cleveland Clinic, Cleveland,
OH). PERK�/� GCN2�/� double mutant cells (24) were obtained from D. Ron
(New York University, New York, NY). All primary and immortalized MEFs
were cultured in DMEM with 10% fetal calf serum, 100 �M �-mercaptoethanol,
and nonessential amino acids (PAA Laboratories, Pasching, Austria).
For generation of E3L-expressing cells, the E3L gene was cloned in the
pMSCV-puro plasmid (Clontech). Production of retroviral vectors using the
Phoenix packaging cell line and transduction of NIH 3T3 cells were done as
described previously (4). Transduced cells were selected with 4 �g/ml puromycin
and grown as bulk cultures without clonal selection.
All virus stocks of viruses deficient in both m142 and m143 were produced on
PKR�/� cells. Virus was harvested from the supernatant by ultracentrifugation,
resuspended in complete medium, and titrated on the same cells according to the
50% tissue culture infective dose (TCID50) method (37). For analysis of virus
replication, cells were seeded in six-well dishes and infected with different
MCMV strains. The supernatants were harvested daily, and fresh medium was
added. Viral titers were determined on PKR�/� MEFs as described above. All
growth kinetics experiments were done in triplicate, and the means and standard
deviations were used for the diagrams.
Immunoprecipitation and Western blotting. In the expression plasmid
pcDNA-m142HA (62), the HA tag was replaced by a FLAG tag by PCR with the
forward primer 5�-AAA GAA TTC CAC CAT GGA CGC CCT GTG CGC
GGC-3� and the reverse primer 5�-AAA CTC GAG TCA CTT ATC ATC GTC
ATC CTT GTA GTC GTC GTC ATC GTC GGC GTC CGC-3�. The C-
terminal FLAG tag introduced by the reverse primer is underlined, and the
restriction sites are in italics. The PCR product was cloned in the pcDNA3 vector
with EcoRI and XhoI. pcDNAm143FLAG was generated by the same method
from pcDNA-m143HA (62) with the forward primer 5�-AAA GGA TCC ACC
ATG TCT TGG GTG ACC GGA GAT-3�, the reverse primer 5�-AAA GAA
TTC TCA CTT ATC ATC GTC ATC CTT GTA GTC CGC GTC GGT CGC
VOL. 83, 2009 PKR BUT NOT OAS INHIBITION BY m142 AND m143 1261
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

TCT CTC GTC-3�, and cloning with BamHI and EcoRI. pcDNA-HA-US11 was
generated from pReplacer-HA-US11 by HindIII and EcoRI cloning.
For analysis of the interaction of m142, m143, and PKR, 293A or NIH 3T3
cells grown in 10-cm dishes were transfected with HA- or FLAG-tagged m142
and/or m143 expression plasmids for 48 h. Cells were then lysed on ice in cold
immunoprecipitation lysis buffer (140 mM NaCl, 5 mM MgCl2, 20 mM Tris, pH
7.6, and 1% [vol/vol] Triton X-100 with complete protease inhibitor cocktail
[Roche]). Insoluble material was removed by centrifugation. Equal amounts of
protein were subjected to immunoprecipitation using either rabbit anti-HA or
anti-FLAG antiserum (Sigma) and protein G-Sepharose (Amersham). Precipi-
tates were washed twice with washing buffer 1 (0.15 M NaCl, 10 mM Tris, pH 7.6,
2 mM EDTA, 0.2% Triton X-100), twice with washing buffer 2 (0.5 M NaCl, 10
mM Tris, pH 7.6, 2 mM EDTA, 0.2% Triton X-100), and once with washing
buffer 3 (20 mM Tris, pH 8.0), and precipitated proteins were released by boiling
in reducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-
PAGE) sample buffer. Proteins were separated by SDS-PAGE, blotted on ni-
trocellulose membranes, and detected by immunoblotting as described previ-
ously (62). RNase III digestion of immunoprecipitates was done by resuspending
the precipitated complexes in 100 �l of RNase III reaction buffer (Ambion) and
incubation with or without 2 U of dsRNA-specific RNase III (Ambion) for 30
min at 37°C. Samples were then washed and analyzed as above.
For analysis of eIF2� phosphorylation, cells were treated with 1 �M thapsi-
gargin ([TG] Calbiochem) for 1 h or were infected with 3 TCID50/cell of the
MCMV strains indicated in the figures, with centrifugal enhancement (two
rounds of centrifugation at 2,000 g for 15 min, with 180° inversion of the plates
in between). After 16 h, cells were lysed on ice in cold radioimmunoprecipitation
assay buffer (50 mM Tris, pH 7.2, 150 mM NaCl, 1% [vol/vol] Triton X-100, 0.1%
[wt/vol] SDS, and 1% [wt/vol] sodium deoxycholate with complete protease
inhibitor cocktail), lysates were cleared by centrifugation, and equal amounts of
protein were separated by SDS-PAGE and blotted as above. The following
antibodies were used for detection in immunoblotting analysis: rabbit anti-HA
(mouse monoclonal antibody [MAb] 3F10; Sigma), anti-FLAG (mouse MAb
M2; Sigma), anti-phospho-eIF2�, anti-total eIF2� (both from Cell Signaling
Technology), anti-MCMV immediate-early antigen 1 ([IE1] MAb CROMA
101), anti-MCMV early antigen 1 ([E1] MAb CROMA103), anti-MCMV glyco-
protein B ([gB] MAb SN 1.07), and anti-PKR (mouse MAb B10; Santa Cruz
Biotechnology).
RNA analysis. Total RNA was isolated from six-well dishes using an RNeasy
kit (Qiagen) according to the manufacturer’s protocol. RNA samples (1.5 �g
each) were stained with ethidium bromide, heat denatured in 25 �l of loading
buffer (50% formamide, 6% [vol/vol] formaldehyde, 5% [vol/vol] glycerol in
MOPS buffer [20 mM morpholinepropanesulfonic acid, 5 mM sodium acetate, 1
mM EDTA, pH 7.0]) for 15 min at 75°C, and separated on 1% agarose form-
aldehyde gels in MOPS buffer at 50 V for 3 h.
OAS activation assay. Detection of ATP oligomerization to 2�,5�-linked oligo-
adenylates (2-5A) by activated OAS enzyme was performed as described previously
(52). Briefly, cell lysates were incubated with ATP and [�32P]ATP, and the resulting
oligomers were separated on 20% acrylamide–7 M urea gels in a sequencing gel
apparatus (Bethesda Research Laboratories) and subjected to autoradiography.
Immunofluorescence. Murine 10.1 cells were seeded on gelatin-coated glass
coverslips and infected 24 h later with intact or UV-inactivated MCMV or with
encephalomyocarditis virus ([EMCV] strain Maus Elberfeld, kindly provided by
A. Nitsche, Robert Koch Institute) for 16 h. Cells were fixed with 3% parafor-
maldehyde in phosphate-buffered saline, permeabilized with 0.5% Triton X-100,
and stained with either anti-dsRNA MAb J2 (English and Scientific Consulting,
Szira´k, Hungary) or an isotype-matched control antibody and with an anti-
mouse-Alexa594 conjugate (Invitrogen). Nuclei were counterstained with DAPI
(4�,6�-diamidino-2-phenylindole). Images were acquired using an epifluores-
cence microscope.
RESULTS
Accumulation of dsRNA during MCMV infection. Although
CMVs are DNA viruses whose replication cycle does not in-
clude a dsRNA intermediate, their genomic organization, with
gene expression occurring from both strands, harbors the po-
tential for dsRNA formation inside the infected cell. To test
whether dsRNA is detectable in MCMV-infected cells, we
stained MCMV-infected and control cells with a MAb specif-
ically recognizing dsRNA (63) or an isotype-matched control
antibody. By immunofluorescence analysis (Fig. 1), dsRNA
was detected in MCMV-infected cells although the signal was
not as strong as in cells infected with EMCV. By contrast,
mock-infected cells and cells infected with UV-inactivated
MCMV were negative, indicating that viral gene expression
and/or viral replication are required for the production of
dsRNA. MCMV-encoded expression of GFP was used as an
infection control. Cells transfected with polyinosinic:poly(C)
[poly(I:C)] showed a coarsely speckled pattern after staining
with the J2 MAb, similar to what has was seen in a previous
study (63). The detection of dsRNA in MCMV-infected cells
suggested that the virus needs to respond to dsRNA-induced
host antiviral defenses.
Viral PKR antagonists rescue an m142 and m143 double
mutant MCMV. A previous study has shown that the MCMV
proteins m142 and m143 bind dsRNA as a heterooligomeric
complex (9). Single-gene deletion mutants failed to prevent
PKR activation and a global shutdown of protein synthesis
upon infection (62). To analyze the properties of MCMV in
the absence of both proteins, we generated a double knockout
mutant, MCMV��m142/m143, lacking both genes. Using
FIG. 1. Accumulation of dsRNA in MCMV-infected cells. Murine 10.1 fibroblasts were mock infected or infected with the picornavirus EMCV
(MOI of 0.3), MCMV (MOI of 3), or UV-inactivated MCMV for 16 h. As an additional positive control, cells were transfected with the synthetic
dsRNA poly(I:C). Cells were stained with the dsRNA-recognizing MAb J2 or an isotype-matched control immunoglobulin G and an Alexa594-
conjugated secondary antibody. Nuclei were counterstained with DAPI. MCMV-encoded GFP is shown in green.
1262 BUDT ET AL. J. VIROL.
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

BAC technology, the region of the m142 and m143 genes was
replaced with a zeocin selection cassette by homologous re-
combination in E. coli carrying the parental MCMV BAC. To
exclude effects of unwanted second-site mutations, we also
generated a revertant virus, MCMV Rev m142/m143. To this
end, a fragment of the MCMV genome spanning the deleted
region was recombined with the MCMV��m142/m143 ge-
nome in NIH 3T3 fibroblasts. Since both genes, m142 and
m143, are essential for MCMV replication (62), only repaired
virus could grow on NIH 3T3 cells.
We also wanted to analyze whether the replication defect of
the ��m142/m143 mutant and, hence, the function of m142
and m143 can be compensated by well-characterized viral an-
tagonists of the PKR signaling pathway. Therefore, we inserted
genes encoding the PKR inhibitors E3L from VV and US11
from HSV-1, as well as the eIF2� phosphatase regulator �34.5
from HSV-1 into the genome of the ��m142/m143 mutant
virus. The genes were inserted by homologous recombination
into the m02-m06 region of the MCMV genome as described
previously (27, 62). Schematic representations of the recombi-
nant MCMV genomes and the EcoRI restriction patterns are
shown in Fig. 2A and B, respectively. The mutated sites in the
genome were also verified by BAC sequencing (Fig. 2C).
When we analyzed the course of infection of the virus mu-
tants in normal fibroblasts (NIH 3T3 cells) with an intact PKR,
parental and revertant viruses replicated to the same titers,
while the double deletion mutant MCMV��m142/m143 was
unable to replicate (Fig. 3A). The PKR antagonists E3L and
�34.5 and, to a lesser extent, also US11 partially rescued the
replication of the ��m142/m143 mutant (Fig. 3A). Thus, ex-
pression of foreign viral PKR antagonists can overcome the
replication block induced by the deletion of m142 and m143,
further confirming a role of m142 and m143 in the PKR anti-
viral pathway. A similar partial rescue was observed when E3L
was expressed in trans after retroviral transduction of NIH 3T3
cells. E3L-expressing but not parental NIH 3T3 cells could
support replication of the ��m142/m143 mutant (Fig. 3B).
Activation of PKR by dsRNA leads to phosphorylation of
the translation initiation factor eIF2� and subsequent protein
synthesis shutoff (43). After infection at a high multiplicity of
infection ([MOI] 5 TCID50/cell), an increased amount of phos-
phorylated eIF2� was detected in cells infected with
MCMV��m142/m143 but not in cells infected with the paren-
tal or the revertant virus. As expected, eIF2� phosphorylation
was antagonized by the expression of E3L, �34.5, or US11 (Fig.
3C). Treatment of cells with TG, an activator of the eIF2�
kinase PERK (61), was used as a positive control for eIF2�
phosphorylation. An apparent discrepancy between the almost
complete suppression of eIF2� phosphorylation compared to
the only moderate rescue of virus growth by the �34.5 and
US11 proteins emerged. We therefore studied the course of
the infection in more detail by analyzing the expression of the
viral IE1, E1, and gB proteins. At 16 h postinfection when
eIF2� phosphorylation was measured, viral gene expression
was similar in all virus mutants. By contrast, the mutants
MCMV��m142/m143 and MCMV��m142/m143 expressing
US11 showed reduced IE1 and E1 levels and completely failed
to express gB at 48 h postinfection. From this we conclude that
protein expression during the early phase of infection is sus-
tained in all complementing mutants, but US11 fails to prevent
protein synthesis shutoff during the late phase.
MCMV��m142/m143 induces eIF2� phosphorylation spe-
cifically through PKR, not via the related kinases PERK or
GCN2. Phosphorylation of eIF2� is a converging point of several
cellular signaling pathways, which respond to different types of
cellular stress. Therefore, we wanted to analyze the contribution
of the eIF2� kinases PKR, PERK, and GCN2 to eIF2� phos-
phorylation observed after infection with MCMV��m142/m143.
wt, PKR�/�, and PERK�/� GCN2�/� fibroblasts were infected
with MCMV or treated with the PERK activator TG. As shown
in Fig. 4, the MCMV��m142/m143 virus induced eIF2� phos-
phorylation in wt and PERK- and GCN2-deficient cells but not in
PKR-deficient cells. Conversely, TG induced eIF2� phosphory-
lation in wt and PKR-deficient cells but not in PERK- and GCN2-
deficient cells. From these results we conclude that MCMV in-
fection in the absence of m142 and m143 selectively activates
eIF2� phosphorylation by PKR and not by the related kinases
PERK or GCN2.
m142 and m143 interact with each other and with PKR.
Since m142 and m143 have been shown to bind dsRNA jointly
but not individually, two possible mechanisms for the inhibi-
tion of PKR are conceivable. First, the binding of m142 and
m143 to dsRNA might occupy all attachment sites for PKR
and sequester the activating substrate away from the kinase. In
this case, a physical interaction between m142 and m143 and
PKR would not be necessary. Alternatively, m142 and m143
could interact with PKR, thereby inhibiting its activation. To
distinguish between the two mechanisms, we checked for an
interaction of m142 and m143 with PKR in coimmunoprecipi-
tation assays. We expressed m142 and m143 with a C-terminal
HA tag (m142HA and m143HA, respectively) or a FLAG tag
(m142FL and m143FL, respectively) in murine NIH 3T3 or
human 293 cells by transient transfection. Cells were lysed, and
equal amounts of protein were subjected to immunoprecipita-
tion with anti-HA, anti-FLAG, or isotype-matched control an-
tibodies coupled to protein G-Sepharose. The samples were
then analyzed by immunoblotting for HA, FLAG, and endog-
enous PKR. After isolated expression of m142 or m143 in NIH
3T3 cells, the protein could be precipitated, but an association
with PKR was not detected. In contrast, when m142HA and
m143FL were cotransfected, m143FL and the endogenous cel-
lular PKR coprecipitated with m142HA (Fig. 5A). Similarly,
the m142-m143-PKR complex could be precipitated via
m142FL with an anti-FLAG antibody (Fig. 5B). The same
complex was also detected in lysates of transfected 293 cells.
Here, immunoprecipitation with anti-HA (Fig. 5C) but not
with control antibodies yielded all three components in the
precipitate. Interestingly, the expression levels of m142 and
m143 were increased when both proteins were coexpressed,
suggesting that the m142-m143 complex is more stable than the
individual proteins. In some immunoprecipitation experi-
ments, we observed small amounts of PKR coprecipitating
with the individual viral proteins, m142HA or m143FL (data
not shown). However, since PKR coprecipitated much more
efficiently in the presence of both viral proteins, the interac-
tions of the individual proteins with PKR are probably weaker
and of lesser relevance than the interaction of the m142-m143
complex with PKR. Combined with our previous observation
that deletion of one of the two viral genes from the viral
VOL. 83, 2009 PKR BUT NOT OAS INHIBITION BY m142 AND m143 1263
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

genome completely abolishes the virus’s ability to inhibit PKR
activation (62), we concluded that both m142 and m143 are
required for a functional interaction with PKR. The results
argue for an inhibition by physical interaction rather than by
sequestering dsRNA away from PKR.
Interestingly, the m142-m143 complex coprecipitated endoge-
nous PKR from human 293 cells and from murine NIH 3T3 cells
even though the murine and human PKRs are only about 60%
identical on the amino acid level (31). This suggested that the viral
proteins interact with a highly conserved domain of PKR.
To analyze whether the complex between m142, m143, and
PKR is mediated by protein interaction or bridged by dsRNA,
we expressed m142HA and m143FL proteins in 293 cells and
precipitated the proteins with anti-HA antibodies. The com-
plexes were resuspended in reaction buffer for the dsRNA-
specific RNase III, incubated with or without enzyme, washed
extensively, and analyzed by immunoblotting (Fig. 5D). Diges-
tion of dsRNA did not reduce the amount of PKR in the
complex, indicating that the m142-m143-PKR complex does
not depend on dsRNA. By contrast, RNase III digestion
FIG. 2. Construction of MCMV deletion and replacement mutants. (A) A GFP-expressing MCMV served as the parental wt virus for all mutants.
Open reading frames m142 and m143 were deleted and replaced by a bacterial zeocin resistance gene (zeo), yielding MCMV��m142/m143. The
revertant virus, Rev m142/m143 (Rev), was constructed as described in Materials and Methods. The genes encoding the viral PKR antagonists E3L, �34.5,
or US11 were inserted at an ectopic location into the MCMV��m142/m143 BAC, replacing open reading frames m02 to m06 and yielding
MCMV��m142/m143 expressing E3L, �34.5, or US11 (��/E3L, ��/�34.5 or ��/US11, respectively). EcoRI (E) restriction sites and the expected
fragment sizes are indicated. (B) The EcoRI restriction pattern of the parental and mutant genomes in an ethidium bromide-stained agarose gel is shown.
Asterisks indicate altered fragments (shown in panel A) where they are clearly visible. Fragments of 4.8, 9, and 10 kb are difficult to identify as they
comigrate with other genome fragments. Note that virion DNA was used for the revertant virus, as this virus was constructed by homologous
recombination in fibroblasts. All other MCMV genomes are shown as BAC DNA. Consequently, Rev m142/m143 lacks two bands of approximately 2
and 6 kb containing the BAC cassette, which is removed during virus reconstitution. The 22-kb fragment containing the joined termini of the MCMV
genome (in the wt and the ��m142/m143 BAC) is also absent in Rev m142/m143 due to the linear structure of the virion DNA. The lower portion of
the gel is shown as an overexposed image to visualize the smaller fragments. (C) The mutated sites of all five mutant viruses were verified by sequencing.
The 3� end of the inserted genes (underlined; stop codon in italics) and the transitions between MCMV and inserted sequences are shown. Numbers
indicate nucleotide positions within the MCMV genome. EcoRI restriction sites are shown in bold. p, phosphoglycerate kinase promoter.
1264 BUDT ET AL. J. VIROL.
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

strongly reduced the amount of PKR coprecipitating with
US11, confirming the previously described RNA dependence
of the US11-PKR complex (6).
MCMV infection does not activate the OAS/RNase L path-
way. The previous experiments have shown that MCMV infec-
tion produces dsRNA and that activation of the dsRNA-de-
pendent PKR is prevented by m142 and m143. This raised the
question of whether the dsRNA-activated antiviral OAS/
RNase L pathway is also turned on by MCMV in the presence
or absence of m142 and m143. To detect oligoadenylate syn-
thetase activity, we applied an enzymatic assay, in which the
substrate [�32P]ATP is metabolized and incorporated into
2-5A (52). The reaction mixtures were then separated on poly-
acrylamide-urea gels and subjected to autoradiography (Fig.
6A). We were unable to detect 2-5A in MCMV-infected cells,
regardless of whether the cells had been pretreated with IFN-�
to induce OAS expression. Only after the addition of poly(I:C)
were slower-migrating ATP derivatives (indicative of 2-5A for-
mation) synthesized (Fig. 6A).
To date, the only known function of 2-5A is activation of the
latent RNase L enzyme, which cleaves different RNA species
FIG. 3. Rescue of MCMV��m142/m143 by heterologous PKR antagonists. (A) Murine fibroblasts were infected at an MOI of 0.01 TCID50/cell
with wt MCMV, MCMV��m142/m143 (��), MCMV Rev m142/m143 (Rev), or deletion mutants expressing E3L (��/E3L), �34.5 (��/�34.5) or
US11 (��/US11) proteins. Virus released into the supernatant was titrated on PKR�/� cells. DL, detection limit. (B) Parental NIH 3T3 cells and
NIH 3T3 cells transduced with an E3L-expressing retrovirus were infected with MCMV��m142/m143 or Rev m142/m143, and virus replication
was analyzed as described above. (C) Murine fibroblasts were infected at an MOI of 5 TCID50/cell. Cell lysates were harvested 16 h later and
probed by immunoblotting for phospho-eIF2� (P-eIF2�). Total eIF2� and the MCMV IE1 protein were detected in the lysates as loading and
infection controls, respectively. (D) The expression of viral IE1, E1, and gB proteins was analyzed by immunoblotting with lysates from NIH 3T3
cells infected with 5 TCID50/cell of wt MCMV (wt), MCMV��m142/m143 (��), or the mutant expressing E3L (E) or US11 (U), respectively, for
the indicated times. Actin is shown as a loading control.
FIG. 4. eIF2� phosphorylation by PKR in MCMV��m142/
m143-infected cells. Normal, PKR�/�, or PERK�/� GCN2�/� fi-
broblasts were mock treated or stimulated with 1 �M TG for 1 h to
induce PERK-mediated eIF2� phosphorylation (p-eIF2�). Infec-
tion of cells with wt MCMV or Rev m142/m143 (Rev) for 16 h (MOI
of 5) did not induce eIF2� phosphorylation in the three cell types.
Infection with MCMV��m142/m143 (��) induced eIF2� phosphor-
ylation in normal and PERK�/� GCN2�/� cells but not in PKR-
deficient fibroblasts. Total eIF2� and the MCMV IE1 protein were
detected in the lysates as loading and infection controls, respec-
tively.
VOL. 83, 2009 PKR BUT NOT OAS INHIBITION BY m142 AND m143 1265
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

including the abundant rRNAs. Therefore, we also assayed the
activity of RNase L in response to MCMV infection. Cells
were pretreated with IFN-� and infected as indicated in Fig.
6B. PKR�/� cells were used for this assay in order to exclude
any contribution of PKR. Total RNA was isolated, stained with
ethidium bromide, and separated on denaturing formalde-
hyde-MOPS agarose gels. Consistent with the results of the
OAS assay (Fig. 6A), we did not observe degradation of cel-
lular rRNA during MCMV infection. Only when poly(I:C) was
transfected into IFN-�-treated cells was rRNA degradation
detected as a ladder of degradation products (Fig. 6B). Under
the same conditions, rRNA degradation was not detected in
FIG. 5. Interaction of m142 and m143 with PKR. Murine NIH 3T3 cells (A and B) or human 293 cells (C) were transfected with plasmids
encoding HA- or FLAG-tagged m142 and/or m143. Polyclonal anti-HA (HA), anti-FLAG (FL), or control (ctrl) antibodies were used for
immunoprecipitation (IP). Precipitates were analyzed by immunoblotting (IB) for the presence of m142, m143, and PKR. PKR was coprecipitated
only in the presence of both the m142 and m143 proteins. (D) To analyze the dsRNA dependence of this interaction, 293 cells were either
transfected with m142HA and m143FL or with HA-US11. The immunoprecipitates were incubated with or without dsRNA-specific RNase III,
washed, and analyzed by immunoblotting. RNase III reduces the coprecipitation of PKR with US11 but not with the m142-m143 complex.
FIG. 6. MCMV neither activates nor inhibits the OAS/RNase L pathway. (A) To analyze OAS activation during MCMV infection, PKR�/�
fibroblasts were preincubated for 24 h with (�) or without (�) 500 U/ml IFN-�. Cells were then infected at an MOI of 5 TCID50/cell with viruses,
as indicated in the figure. Cell lysates were harvested 16 h later and used in an OAS activity assay using [�32P]ATP. The 2-5A enzyme was separated
on a 20% polyacrylamide-urea gel and detected by autoradiography. As a positive control, OAS activity was stimulated by adding 5 �g/ml poly(I:C).
(B) Poly(I:C) transfection but not MCMV infection induced cleavage of ribosomal 28S and 18S RNAs. PKR�/� cells were pretreated with IFN
and infected with MCMV, as described in panel A, or transfected with poly(I:C). Total RNA was isolated and separated on denaturing
formaldehyde-MOPS agarose gels. (C) rRNA was not cleaved in RNase L�/� cells, indicating that the rRNA degradation shown in panel B is
RNase L specific. (D) IFN-pretreated NIH 3T3 or PKR�/� cells were infected with the indicated viruses for 16 h and transfected with different
amounts of poly(I:C) (pIC). rRNA cleavage was analyzed 6.5 h after transfection. MCMV infection did not inhibit poly(I:C)-induced RNase L
activation. ��, MCMV��m142/m143; Rev, MCMV Rev m142/m143.
1266 BUDT ET AL. J. VIROL.
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

RNase L�/� cells, confirming that this effect is mediated spe-
cifically by RNase L (Fig. 6C).
The failure to observe RNase L activity during MCMV in-
fection could be attributed to a lack of a strong activating
stimulus (as suggested by the results shown in Fig. 6A) or to an
active inhibition of this pathway by MCMV. To test for the
latter, we analyzed MCMV’s ability to inhibit RNase L activa-
tion induced by an exogenous stimulus. To this end, cells were
pretreated with IFN, infected with MCMV, and subsequently
transfected with poly(I:C). As shown in Fig. 6D, MCMV in-
fection did not significantly affect poly(I:C)-induced rRNA
degradation, regardless of the presence or absence of m142
and m143. Taken together, these results suggested that the
OAS/RNase L pathway is neither activated nor actively inhib-
ited by MCMV and that the dsRNA-binding proteins m142
and m143 do not play an important role in controlling this
dsRNA-activated pathway.
PKR, but not RNase L, controls replication of MCMV��
m142/m143. The results of the previous experiments showed
that MCMV infection in the absence of m142 and m143 acti-
vates PKR, whereas activation of the OAS/RNase L pathway
was not detectable. This suggested that the PKR pathway, but
not the OAS/RNase L pathway, inhibits replication of
MCMV��m142/m143. We tested this hypothesis by analyzing
the replication kinetics of this virus in wt fibroblasts and in
fibroblasts derived from PKR or RNase L knockout mice. As
expected, MCMV��m142/m143 did not replicate in wt and
RNase L-deficient cells (Fig. 7), nor did MCMV mutants lack-
ing only one of the two genes (data not shown). However, the
MCMV��m142/m143 mutant grew to similar titers as the pa-
rental virus and the revertant virus in PKR�/� cells (Fig. 7).
Hence, we concluded that PKR, but not OAS/RNase L, con-
trols MCMV replication in the absence of m142 and m143 and
that PKR is responsible for the essentiality of m142 and m143.
DISCUSSION
Cells respond to viral infection by secreting type I IFNs,
which stimulate specific receptors on infected and neighboring
cells. This activates the expression of IFN-stimulated genes,
many of which encode proteins with antiviral functions (54).
PKR and OAS are two important representatives of these
antiviral effector proteins (50). Both are activated by dsRNA
and hamper viral replication by repressing protein synthesis.
The importance of PKR and OAS is underscored by the fact
that several viruses express gene products that inhibit the PKR
or the OAS pathway (43, 55).
In MCMV, the m142 and m143 proteins play a crucial role
in blocking the host antiviral response. In the absence of one or
both proteins, MCMV infection leads to PKR activation,
eIF2� phosphorylation, and attenuation of protein synthesis
(62). The two proteins form a dsRNA-binding complex and are
together able to rescue the replication of an E3L-deficient VV
(9). However, m142 and m143 differ from previously charac-
terized viral PKR inhibitors in that they are absolutely essential
for virus replication in cell culture (38, 62). Two possible rea-
sons for this essential nature are conceivable: either the PKR-
mediated antiviral response is so effective against MCMV that
the virus depends on an efficient inhibitory mechanism, or the
two viral proteins possess an additional function beyond PKR
inhibition. Such additional functions have been identified in a
number of viral PKR inhibitors: human CMV (HCMV) TRS1
appears to have a role in viral DNA replication and assembly
of DNA-containing capsids (1, 45), VV E3L binds Z-DNA
(29), influenza virus NS1 inhibits RIG-I (47), and HSV-1 US11
prevents OAS activation (52). We also assayed the replication
of the PKR antagonist-expressing virus mutants in PKR�/�
cells in order to clarify whether the incomplete complementa-
tion of these proteins might be due to a dominant-negative
effect of the viral PKR antagonist on MCMV replication. How-
ever, all virus mutants grew to similar titers in PKR�/� cells,
arguing against such an effect (Fig. 8).
Our first approach to delineate the functional spectrum of
m142 and m143 was to complement an MCMV strain deficient
in both m142 and m143 with PKR antagonists of HSV-1 and
VV. These proteins reduced eIF2� phosphorylation when ex-
pressed from the genome of MCMV��m142/m143 and par-
tially restored virus replication (Fig. 3). Similar results have
been obtained previously with the HCMV TRS1 gene product
(62). By contrast, a knockout of the host effector protein PKR
completely restored replication of the MCMV��m142/m143
virus to titers indistinguishable from the parental virus (Fig. 7).
This suggested that the PKR antagonists of HCMV, HSV-1,
and VV inhibit murine PKR less efficiently than the MCMV
proteins or that the expression levels and kinetics in our system
were not adequate for a complete inhibition of PKR. Another
possible explanation could be that the heterologous viral genes
were inserted into the MCMV genome alone even though they
synergize with a second viral protein in their natural viral
context: US11 synergizes with �34.5 in HSV-1 (44), TRS1
FIG. 7. PKR, but not RNase L, controls replication of m142- and m143-deficient MCMV. PKR�/�, RNase L�/�, and wt fibroblasts were
infected with wt MCMV, MCMV��m142/m143 (��), or MCMV Rev m142/m143 (Rev) at an MOI of 0.01 TCID50/cell. Virus titers in the
supernatants were determined by titration on PKR�/� cells. DL, detection limit; dpi, days postinfection.
VOL. 83, 2009 PKR BUT NOT OAS INHIBITION BY m142 AND m143 1267
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

synergizes with IRS1 in HCMV (8), and E3L synergizes with
the PKR pseudosubstrate K3L in VV (11).
We show in this study that the m142-m143 complex interacts
with PKR and that the viral proteins do not bind to PKR
individually or do so with lower affinity. Interestingly, the com-
plex was not sensitive to digestion with RNase III, suggesting a
protein-protein interaction of the components. It is unlikely
that the short 12- to 30-bp dsRNA fragments produced by
RNase III would be sufficient to maintain such a complex.
However, it cannot be completely excluded that bridging
dsRNA molecules are inaccessible for digestion in the m142-
m143-PKR complex. It seems likely that the interaction of
m142 and m143 with PKR is necessary for inhibition of PKR
activation because an interaction with PKR is essential for the
function of the influenza virus NS1 and the HCMV TRS1 and
IRS1 proteins (17, 33). However, this hypothesis can be tested
only after the dsRNA- and PKR-binding domains of m142 and
m143 have been identified and mutated separately.
Another antiviral effector system that shares many similari-
ties with the PKR-dependent pathway is the OAS/RNase L
pathway. It is upregulated by type I IFNs, activated by dsRNA,
and attenuates protein expression (55). Surprisingly, we found
no indication that the OAS/RNase L pathway is activated dur-
ing MCMV infection, regardless of whether m142 and m143
were expressed. Moreover, RNase L knockout cells did not
support replication of m142- and/or m143-deficient viruses,
whereas PKR knockout cells allowed replication of these vi-
ruses to titers equivalent to the parental virus (Fig. 7). Two
possible explanations are conceivable. On the one hand,
dsRNA levels in MCMV-infected cells could be insufficient for
a robust activation of OAS. On the other hand, MCMV might
express other proteins that inhibit the OAS/RNase L pathway.
Candidates for such a function would be, for instance, the gene
products of M23 and M24, which were shown to bind dsRNA.
(9). The inability of MCMV to block rRNA degradation upon
poly(I:C) transfection argues in favor of the first possibility.
However, we cannot completely exclude the possibility that
MCMV encodes inhibitors of the OAS/RNase L pathway that
are sufficient to inhibit OAS/RNase L activities induced by
endogenously produced dsRNA but insufficient to block activ-
ities triggered by transfected poly(I:C).
The lack of detectable RNase L activation by MCMV is in
accordance with the observation that at least some DNA vi-
ruses are less sensitive to the OAS/RNase L pathway than
RNA viruses (55). For instance, no significant activation of the
OAS/RNase L pathway was detected after infection of cells
with varicella zoster virus or simian virus 40 (13, 21), and the
impact of this pathway on HSV-1 and VV was strongly depen-
dent on the system investigated (55).
Besides PKR and OAS, at least two other classes of pattern
recognition receptors are activated by dsRNA. TLR3 senses
the presence of dsRNA in endosomes, and the RNA helicases
MDA5 and (to a lesser extent) RIG-I detect dsRNA in the
cytosol (53). It is unclear how the cytosolic proteins m142 and
m143 could gain access to endosomes to inhibit TLR3 activa-
tion. On the other hand, it might be worth investigating the
influence of m142 and m143 on RIG-I or MDA5 activation.
While both helicases can be activated by poly(I:C), it has re-
cently become clear that RIG-I is preferentially activated by
5�-triphosphorylated single-stranded RNA (22) and that
MDA5 is activated predominantly by long dsRNA (28, 53).
Whether such RNA species are present in MCMV-infected
cells is unknown. It has been shown, however, that MDA5 is
not required for the induction of IFN-� by MCMV in vivo (16).
Viral interference with dsRNA-activated pathways does not
necessarily require that the viral antagonist itself bind dsRNA.
Viral proteins can interfere with signal transduction at various
steps of the signaling cascades. In the case of MCMV, it has
recently been shown that the M45 protein inhibits the adaptor
protein RIP1 (36, 60), a protein that mediates NF-
B activa-
tion in response to TLR3 and RIG-I/MDA5 stimulation (2, 40,
41). Interestingly, a study identified a role of PKR in the
TLR3-dependent pathway (25), hinting at a possible indirect
role for m142 and m143 in TLR3 signaling. Although the fact
that MCMV��m142/m143 replicated to wt levels in PKR�/�
fibroblasts indicated that PKR is the primary target for m142
and m143, an influence on other signaling pathways cannot be
entirely excluded. Further investigations will be necessary to
reveal the full spectrum of the activities of m142 and m143 and
the role of other viral proteins in subverting antiviral pathways
of the host cell.
ACKNOWLEDGMENTS
We thank Katrin Berger for excellent technical assistance, R. H.
Silverman for RNase L�/� cells, D. Ron for PERK�/� GCN2�/� cells,
U. Kalinke and J. Pavlovic for PKR�/� cells, I. Mohr for �34.5 and
US11 expression plasmids, A. Nitsche for EMCV, C. Adlhoch for help
with the sequencing gel apparatus, and S. Voigt for a critical reading of
the manuscript.
This work was supported by grants from the Deutsche Forschungs-
gemeinschaft (Bu2323/1-1 to M.B. and SFB421 TP B14 to W.B.).
REFERENCES
1. Adamo, J. E., J. Schroer, and T. Shenk. 2004. Human cytomegalovirus TRS1
protein is required for efficient assembly of DNA-containing capsids. J. Vi-
rol. 78:10221–10229.
2. Balachandran, S., E. Thomas, and G. N. Barber. 2004. A FADD-dependent
innate immune mechanism in mammalian cells. Nature 432:401–405.
3. Berlanga, J. J., I. Ventoso, H. P. Harding, J. Deng, D. Ron, N. Sonenberg, L.
Carrasco, and C. de Haro. 2006. Antiviral effect of the mammalian transla-
tion initiation factor 2� kinase GCN2 against RNA viruses. EMBO J. 25:
1730–1740.
4. Brune, W., M. Nevels, and T. Shenk. 2003. Murine cytomegalovirus m41
FIG. 8. Normal replication of PKR antagonist-expressing MCMV
in PKR�/� cells. Cells were infected with wt MCMV Rev m142/m143
(Rev) or MCMV��m142/m143 expressing E3L (��/E3L), �34.5 (��/
�34.5), or US11 (��/US11) at an MOI of 0.1 TCID50/cell. Virus titers
in the supernatants were determined by titration on PKR�/� cells. DL,
detection limit; dpi, days postinfection.
1268 BUDT ET AL. J. VIROL.
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

open reading frame encodes a Golgi-localized antiapoptotic protein. J. Virol.
77:11633–11643.
5. Brune, W., M. Wagner, and M. Messerle. 2006. Manipulating cytomegalo-
virus genomes by BAC mutagenesis: strategies and applications, p. 61–89. In
M. J. Reddehase (ed.), Cytomegaloviruses: molecular biology and immunol-
ogy. Caister Academic Press, Wymondham, United Kingdom.
6. Cassady, K. A., and M. Gross. 2002. The herpes simplex virus type 1 U(S)11
protein interacts with protein kinase R in infected cells and requires a
30-amino-acid sequence adjacent to a kinase substrate domain. J. Virol.
76:2029–2035.
7. Chang, H. W., J. C. Watson, and B. L. Jacobs. 1992. The E3L gene of
vaccinia virus encodes an inhibitor of the interferon-induced, double-
stranded RNA-dependent protein kinase. Proc. Natl. Acad. Sci. USA 89:
4825–4829.
8. Child, S. J., M. Hakki, K. L. De Niro, and A. P. Geballe. 2004. Evasion of
cellular antiviral responses by human cytomegalovirus TRS1 and IRS1. J. Vi-
rol. 78:197–205.
9. Child, S. J., L. K. Hanson, C. E. Brown, D. M. Janzen, and A. P. Geballe.
2006. Double-stranded RNA binding by a heterodimeric complex of murine
cytomegalovirus m142 and m143 proteins. J. Virol. 80:10173–10180.
10. Clemens, M. J., and A. Elia. 1997. The double-stranded RNA-dependent
protein kinase PKR: structure and function. J. Interferon Cytokine Res.
17:503–524.
11. Davies, M. V., H. W. Chang, B. L. Jacobs, and R. J. Kaufman. 1993. The E3L
and K3L vaccinia virus gene products stimulate translation through inhibi-
tion of the double-stranded RNA-dependent protein kinase by different
mechanisms. J. Virol. 67:1688–1692.
12. Deng, J., H. P. Harding, B. Raught, A. C. Gingras, J. J. Berlanga, D.
Scheuner, R. J. Kaufman, D. Ron, and N. Sonenberg. 2002. Activation of
GCN2 in UV-irradiated cells inhibits translation. Curr. Biol. 12:1279–1286.
13. Desloges, N., M. Rahaus, and M. H. Wolff. 2005. Varicella-zoster virus does
not significantly induce cell defence mechanism mediated by the 2-5A/
RNase L pathway during its replication cycle. Med. Microbiol. Immunol.
194:25–31.
14. Floyd-Smith, G., E. Slattery, and P. Lengyel. 1981. Interferon action: RNA
cleavage pattern of a (2�-5�)oligoadenylate-dependent endonuclease. Sci-
ence 212:1030–1032.
15. Garcia, M. A., E. F. Meurs, and M. Esteban. 2007. The dsRNA protein
kinase PKR: virus and cell control. Biochimie 89:799–811.
16. Gitlin, L., W. Barchet, S. Gilfillan, M. Cella, B. Beutler, R. A. Flavell, M. S.
Diamond, and M. Colonna. 2006. Essential role of mda-5 in type I IFN
responses to polyriboinosinic:polyribocytidylic acid and encephalomyocardi-
tis picornavirus. Proc. Natl. Acad. Sci. USA 103:8459–8464.
17. Hakki, M., E. E. Marshall, K. L. De Niro, and A. P. Geballe. 2006. Binding
and nuclear relocalization of protein kinase R by human cytomegalovirus
TRS1. J. Virol. 80:11817–11826.
18. Harding, H. P., Y. Zhang, and D. Ron. 1999. Protein translation and folding
are coupled by an endoplasmic-reticulum-resident kinase. Nature 397:271–
274.
19. Harvey, D. M., and A. J. Levine. 1991. p53 alteration is a common event in
the spontaneous immortalization of primary BALB/c murine embryo fibro-
blasts. Genes Dev. 5:2375–2385.
20. He, B., M. Gross, and B. Roizman. 1997. The �134.5 protein of herpes
simplex virus 1 complexes with protein phosphatase 1� to dephosphorylate
the � subunit of the eukaryotic translation initiation factor 2 and preclude
the shutoff of protein synthesis by double-stranded RNA-activated protein
kinase. Proc. Natl. Acad. Sci. USA 94:843–848.
21. Hersh, C. L., R. E. Brown, W. K. Roberts, E. A. Swyryd, I. M. Kerr, and G. R.
Stark. 1984. Simian virus 40-infected, interferon-treated cells contain 2�,5�-
oligoadenylates which do not activate cleavage of RNA. J. Biol. Chem.
259:1731–1737.
22. Hornung, V., J. Ellegast, S. Kim, K. Brzozka, A. Jung, H. Kato, H. Poeck, S.
Akira, K. K. Conzelmann, M. Schlee, S. Endres, and G. Hartmann. 2006.
5�-Triphosphate RNA is the ligand for RIG-I. Science 314:994–997.
23. Hovanessian, A. G., and J. Justesen. 2007. The human 2�-5�oligoadenylate
synthetase family: unique interferon-inducible enzymes catalyzing 2�-5� in-
stead of 3�-5� phosphodiester bond formation. Biochimie 89:779–788.
24. Jiang, H.-Y., S. A. Wek, B. C. McGrath, D. Lu, T. Hai, H. P. Harding, X.
Wang, D. Ron, D. R. Cavener, and R. C. Wek. 2004. Activating transcription
factor 3 is integral to the eukaryotic initiation factor 2 kinase stress response.
Mol. Cell. Biol. 24:1365–1377.
25. Jiang, Z., M. Zamanian-Daryoush, H. Nie, A. M. Silva, B. R. Williams, and
X. Li. 2003. Poly(I-C)-induced Toll-like receptor 3 (TLR3)-mediated acti-
vation of NF
B and MAP kinase is through an interleukin-1 receptor-
associated kinase (IRAK)-independent pathway employing the signaling
components TLR3-TRAF6-TAK1-TAB2-PKR. J. Biol. Chem. 278:16713–
16719.
26. Jordan, R., L. Wang, T. M. Graczyk, T. M. Block, and P. R. Romano. 2002.
Replication of a cytopathic strain of bovine viral diarrhea virus activates
PERK and induces endoplasmic reticulum stress-mediated apoptosis of
MDBK cells. J. Virol. 76:9588–9599.
27. Jurak, I., and W. Brune. 2006. Induction of apoptosis limits cytomegalovirus
cross-species infection. EMBO J. 25:2634–2642.
28. Kato, H., O. Takeuchi, E. Mikamo-Satoh, R. Hirai, T. Kawai, K. Matsushita,
A. Hiiragi, T. S. Dermody, T. Fujita, and S. Akira. 2008. Length-dependent
recognition of double-stranded ribonucleic acids by retinoic acid-inducible
gene-I and melanoma differentiation-associated gene 5. J. Exp. Med. 205:
1601–1610.
29. Kim, Y. G., M. Muralinath, T. Brandt, M. Pearcy, K. Hauns, K. Lowen-
haupt, B. L. Jacobs, and A. Rich. 2003. A role for Z-DNA binding in vaccinia
virus pathogenesis. Proc. Natl. Acad. Sci. USA 100:6974–6979.
30. Kostura, M., and M. B. Mathews. 1989. Purification and activation of the
double-stranded RNA-dependent eIF-2 kinase DAI. Mol. Cell. Biol. 9:1576–
1586.
31. Kuhen, K. L., X. Shen, E. R. Carlisle, A. L. Richardson, H. U. Weier, H.
Tanaka, and C. E. Samuel. 1996. Structural organization of the human gene
(PKR) encoding an interferon-inducible RNA-dependent protein kinase
(PKR) and differences from its mouse homolog. Genomics 36:197–201.
32. Lee, M. S., and Y. J. Kim. 2007. Pattern-recognition receptor signaling
initiated from extracellular, membrane, and cytoplasmic space. Mol. Cells
23:1–10.
33. Li, S., J. Y. Min, R. M. Krug, and G. C. Sen. 2006. Binding of the influenza
A virus NS1 protein to PKR mediates the inhibition of its activation by either
PACT or double-stranded RNA. Virology 349:13–21.
34. Lu, L., A. P. Han, and J. J. Chen. 2001. Translation initiation control by
heme-regulated eukaryotic initiation factor 2� kinase in erythroid cells under
cytoplasmic stresses. Mol. Cell. Biol. 21:7971–7980.
35. Lu, Y., M. Wambach, M. G. Katze, and R. M. Krug. 1995. Binding of the
influenza virus NS1 protein to double-stranded RNA inhibits the activation
of the protein kinase that phosphorylates the elF-2 translation initiation
factor. Virology 214:222–228.
36. Mack, C., A. Sickmann, D. Lembo, and W. Brune. 2008. Inhibition of proin-
flammatory and innate immune signaling pathways by a cytomegalovirus
RIP1-interacting protein. Proc. Natl. Acad. Sci. USA 105:3094–3099.
37. Mahy, B. W. J., and H. O. Kangro. 1996. Virology methods manual. Aca-
demic Press, San Diego, CA.
38. Menard, C., M. Wagner, Z. Ruzsics, K. Holak, W. Brune, A. E. Campbell,
and U. H. Koszinowski. 2003. Role of murine cytomegalovirus US22 gene
family members in replication in macrophages. J. Virol. 77:5557–5570.
39. Meurs, E. F., Y. Watanabe, S. Kadereit, G. N. Barber, M. G. Katze, K.
Chong, B. R. Williams, and A. G. Hovanessian. 1992. Constitutive expression
of human double-stranded RNA-activated p68 kinase in murine cells medi-
ates phosphorylation of eukaryotic initiation factor 2 and partial resistance to
encephalomyocarditis virus growth. J. Virol. 66:5805–5814.
40. Meylan, E., K. Burns, K. Hofmann, V. Blancheteau, F. Martinon, M. Kelli-
her, and J. Tschopp. 2004. RIP1 is an essential mediator of Toll-like receptor
3-induced NF-
B activation. Nat. Immunol. 5:503–507.
41. Michallet, M. C., E. Meylan, M. A. Ermolaeva, J. Vazquez, M. Rebsamen, J.
Curran, H. Poeck, M. Bscheider, G. Hartmann, M. Konig, U. Kalinke, M.
Pasparakis, and J. Tschopp. 2008. TRADD protein is an essential compo-
nent of the RIG-like helicase antiviral pathway. Immunity 28:651–661.
42. Mohr, I., and Y. Gluzman. 1996. A herpesvirus genetic element which affects
translation in the absence of the viral GADD34 function. EMBO J. 15:4759–
4766.
43. Mohr, I. J., T. Pe’ery, and M. B. Mathews. 2007. Protein synthesis and
translational control during viral infection, p. 545–599. In M. B. Mathews, N.
Sonenberg, and J. W. B. Hershey (ed.), Translational control in biology and
medicine. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
44. Mulvey, M., J. Poppers, D. Sternberg, and I. Mohr. 2003. Regulation of
eIF2� phosphorylation by different functions that act during discrete phases
in the herpes simplex virus type 1 life cycle. J. Virol. 77:10917–10928.
45. Pari, G. S., and D. G. Anders. 1993. Eleven loci encoding trans-acting factors
are required for transient complementation of human cytomegalovirus ori-
Lyt-dependent DNA replication. J. Virol. 67:6979–6988.
46. Patel, R. C., P. Stanton, N. M. McMillan, B. R. Williams, and G. C. Sen.
1995. The interferon-inducible double-stranded RNA-activated protein ki-
nase self-associates in vitro and in vivo. Proc. Natl. Acad. Sci. USA 92:8283–
8287.
47. Pichlmair, A., O. Schulz, C. P. Tan, T. I. Naslund, P. Liljestrom, F. Weber,
and C. Reis e Sousa. 2006. RIG-I-mediated antiviral responses to single-
stranded RNA bearing 5�-phosphates. Science 314:997–1001.
48. Redwood, A. J., M. Messerle, N. L. Harvey, C. M. Hardy, U. H. Koszinowski,
M. A. Lawson, and G. R. Shellam. 2005. Use of a murine cytomegalovirus
K181-derived bacterial artificial chromosome as a vaccine vector for immu-
nocontraception. J. Virol. 79:2998–3008.
49. Roberts, W. K., A. Hovanessian, R. E. Brown, M. J. Clemens, and I. M. Kerr.
1976. Interferon-mediated protein kinase and low-molecular-weight inhibi-
tor of protein synthesis. Nature 264:477–480.
50. Sadler, A. J., and B. R. Williams. 2008. Interferon-inducible antiviral effec-
tors. Nat. Rev. Immunol. 8:559–568.
51. Saito, T., and M. Gale, Jr. 2007. Principles of intracellular viral recognition.
Curr. Opin. Immunol. 19:17–23.
52. Sanchez, R., and I. Mohr. 2007. Inhibition of cellular 2�-5� oligoadenylate
VOL. 83, 2009 PKR BUT NOT OAS INHIBITION BY m142 AND m143 1269
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from

synthetase by the herpes simplex virus type 1 Us11 protein. J. Virol. 81:3455–
3464.
53. Schlee, M., W. Barchet, V. Hornung, and G. Hartmann. 2007. Beyond
double-stranded RNA-type I IFN induction by 3pRNA and other viral nu-
cleic acids. Curr. Top. Microbiol. Immunol. 316:207–230.
54. Sen, G. C., and S. N. Sarkar. 2007. The interferon-stimulated genes: targets
of direct signaling by interferons, double-stranded RNA, and viruses. Curr.
Top. Microbiol. Immunol. 316:233–250.
55. Silverman, R. H. 2007. Viral encounters with 2�,5�-oligoadenylate synthetase
and RNase L during the interferon antiviral response. J. Virol. 81:12720–
12729.
56. Simon, C. O., C. K. Seckert, N. K. A. Grizmek, and M. J. Reddehase. 2006.
Murine model of cytomegalovirus latency and reactivation: the silencing/
desilencing and immune sensing hypothesis, p. 483–500. In M. J. Reddehase
(ed.), Cytomegaloviruses: molecular biology and immunology. Caister Aca-
demic Press, Wymondham, United Kingdom.
57. Sudhakar, A., A. Ramachandran, S. Ghosh, S. E. Hasnain, R. J. Kaufman,
and K. V. Ramaiah. 2000. Phosphorylation of serine 51 in initiation factor 2�
(eIF2�) promotes complex formation between eIF2 alpha(P) and eIF2B and
causes inhibition in the guanine nucleotide exchange activity of eIF2B. Bio-
chemistry 39:12929–12938.
58. Takeuchi, O., and S. Akira. 2007. Recognition of viruses by innate immunity.
Immunol. Rev. 220:214–224.
59. Thomis, D. C., and C. E. Samuel. 1995. Mechanism of interferon action:
characterization of the intermolecular autophosphorylation of PKR, the
interferon-inducible, RNA-dependent protein kinase. J. Virol. 69:5195–
5198.
60. Upton, J. W., W. J. Kaiser, and E. S. Mocarski. 2008. Cytomegalovirus M45
cell death suppression requires receptor-interacting protein (RIP) homo-
typic interaction motif (RHIM)-dependent interaction with RIP1. J. Biol.
Chem. 283:16966–16970.
61. Urano, F., X. Wang, A. Bertolotti, Y. Zhang, P. Chung, H. P. Harding, and
D. Ron. 2000. Coupling of stress in the ER to activation of JNK protein
kinases by transmembrane protein kinase IRE1. Science 287:664–666.
62. Valchanova, R. S., M. Picard-Maureau, M. Budt, and W. Brune. 2006.
Murine cytomegalovirus m142 and m143 are both required to block protein
kinase R-mediated shutdown of protein synthesis. J. Virol. 80:10181–10190.
63. Weber, F., V. Wagner, S. B. Rasmussen, R. Hartmann, and S. R. Paludan.
2006. Double-stranded RNA is produced by positive-strand RNA viruses
and DNA viruses but not in detectable amounts by negative-strand RNA
viruses. J. Virol. 80:5059–5064.
64. Wek, S. A., S. Zhu, and R. C. Wek. 1995. The histidyl-tRNA synthetase-
related sequence in the eIF-2� protein kinase GCN2 interacts with tRNA
and is required for activation in response to starvation for different amino
acids. Mol. Cell. Biol. 15:4497–4506.
65. Wreschner, D. H., J. W. McCauley, J. J. Skehel, and I. M. Kerr. 1981.
Interferon action—sequence specificity of the ppp(A2�p)nA-dependent ri-
bonuclease. Nature 289:414–417.
66. Yang, Y. L., L. F. Reis, J. Pavlovic, A. Aguzzi, R. Schafer, A. Kumar, B. R.
Williams, M. Aguet, and C. Weissmann. 1995. Deficient signaling in mice
devoid of double-stranded RNA-dependent protein kinase. EMBO J. 14:
6095–6106.
67. Zhou, A., J. Paranjape, T. L. Brown, H. Nie, S. Naik, B. Dong, A. Chang, B.
Trapp, R. Fairchild, C. Colmenares, and R. H. Silverman. 1997. Interferon
action and apoptosis are defective in mice devoid of 2�,5�-oligoadenylate-
dependent RNase L. EMBO J. 16:6355–6363.
1270 BUDT ET AL. J. VIROL.
by on January 10, 2009
jvi.asm.org
D
ow
nloaded from