NF90 Exerts Antiviral Activity through
Regulation of PKR Phosphorylation and
Stress Granules in Infected Cells
This information is current as
of June 18, 2017.
Xi Wen, Xiaofeng Huang, Bobo Wing-Yee Mok, Yixin
Chen, Min Zheng, Siu-Ying Lau, Pui Wang, Wenjun Song,
Dong-Yan Jin, Kwok-Yung Yuen and Honglin Chen
J Immunol published online 12 March 2014
http://www.jimmunol.org/content/early/2014/03/12/jimmun
ol.1302813
Permissions
Email Alerts
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2014 by The American Association of
Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Subscription
Published March 12, 2014, doi:10.4049/jimmunol.1302813
The Journal of Immunology
NF90 Exerts Antiviral Activity through Regulation of PKR
Phosphorylation and Stress Granules in Infected Cells
Xi Wen,*,† Xiaofeng Huang,*,† Bobo Wing-Yee Mok,*,† Yixin Chen,‡ Min Zheng,*,†
Siu-Ying Lau,*,† Pui Wang,*,† Wenjun Song,*,† Dong-Yan Jin,x Kwok-Yung Yuen,*,† and
Honglin Chen*,†
N
F90 (also known as NFAR1 or DRBP76) was first identified as an IL-2 promoter-binding protein in activated
T cells and was found to regulate IL-2 gene expression via
stabilization of IL-2 mRNA (1–5). A variety of independent studies
subsequently found that NF90 has dsRNA-binding properties. NF90
is predominantly localized within the nucleus (6, 7), although it also
has been found in the cytoplasm (8). It interacts with PKR and is a
substrate for phosphorylation by PKR (7–11). Further studies (12,
13) revealed that phosphorylation of NF90 by PKR is necessary
for association of the NF90/NF45 complex, shuttling of NF90 between the nucleus and cytoplasm, and NF90 function in translational regulation and host antiviral defense.
*State Key Laboratory for Emerging Infectious Diseases, Department of Microbiology,
The University of Hong Kong, Hong Kong Special Administrative Region, People’s
Republic of China; †Research Centre of Infection and Immunology, The University
of Hong Kong, Hong Kong Special Administrative Region, People’s Republic of
China; ‡National Institute of Diagnostics and Vaccine Development for Infectious
Diseases, School of Life Sciences, Xiamen University, Xiamen 361005, China; and
x
Department of Biochemistry, The University of Hong Kong, Hong Kong Special
Administrative Region, People’s Republic of China
Received for publication October 18, 2013. Accepted for publication February 18,
2014.
This work was supported in part by the Research Grants Council of the Hong Kong
Special Administrative Region (7620/10M and 7629/13M) and the Areas of Excellence Scheme of the University Grants Committee (Grant AoE/M-12/06).
Address correspondence and reprint requests to Dr. Honglin Chen, State Key Laboratory for Emerging Infectious Diseases, Department of Microbiology, The University
of Hong Kong, 21 Sassoon Road, Laboratory Block, Pokfulam, Hong Kong Special
Administrative Region, People’s Republic of China. E-mail address: [email protected]
Abbreviations used in this article: co-IP, coimmunoprecipitation; eIF2a, eukaryotic
initiating factor 2 a; G3BP, Ras GAP SH3-domain–binding protein; MEF, mouse
embryonic fibroblast; MOI, multiplicity of infection; NP, nucleoprotein; NS1, nonstructural protein 1; PKR, protein kinase; poly(I:C), polyinosinic-polycytidylic acid;
RBD, RNA-binding domain; RBDm, RNA-binding domain mutant; siRNA, small
interfering RNA; WT, wild-type.
Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00
www.jimmunol.org/cgi/doi/10.4049/jimmunol.1302813
NF90 was found to be involved in host antiviral mechanisms
targeting various viruses. It suppresses the function of Ebola virus
polymerases through interaction with VP35 (14), inhibits HIV
replication through interaction with HIV-1 TAR RNA (15, 16),
represses internal ribosome entry sites in rhinoviruses (17, 18) and
negatively regulates influenza virus replication through interaction
with viral nucleoprotein (NP) (19). However, other studies (20–23)
found that NF90 is required for the replication of some positivestranded RNA viruses and is important for expression of E6 protein
in human papillomavirus–infected cells. It is postulated that, although members of the NF90 family generally serve as components of host antiviral responses, some viruses may have adapted
a mechanism to hijack NF90, retasking it for viral replication and
weakening host defenses (20). A study in which NF90 (NFAR) was
depleted in mouse embryonic fibroblast (MEF) cells suggested that
NF90 exerts its antiviral activity through modulation of translation
in host cells in a PKR-dependent manner (12, 13).
In addition to its involvement in the regulation of the IL-2
promoter in T cells, dsRNA-dependent protein kinase (PKR) is
the most well-defined host factor interacting with NF90. PKR is
an IFN-inducible gene that plays a critical role in host antiviral
responses (24, 25). It is natural to hypothesize that the antiviral
function exerted by NF90 may be signaled through PKR-related
pathways. The primary role of PKR in the antiviral response is its
inhibition of translation of viral mRNAs through phosphorylation
of eukaryotic initiation factor 2 a (eIF2a) (26). PKR is also recognized for its role in regulating cellular inflammatory signals (27,
28). A previous study (29) found that NF90 is able to inhibit yeast
growth when coexpressed with PKR, but not on its own, implying
that NF90 may activate PKR to cause translational inhibition.
However, this same study (29) found no evidence of NF90 activation of PKR in an in vitro assay. Conversely, it was found that NF90
inhibited PKR phosphorylation, presumably through competition for
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
NF90 was shown to exhibit broad antiviral activity against several viruses, but detailed mechanisms remain unclear. In this study, we
examined the molecular basis for the inhibitory effect of NF90 on virus replication mediated through protein kinase (PKR)associated translational regulation. We first verified the interaction between NF90 and PKR in mammalian cells and showed that
NF90 interacts with PKR through its C-terminal and that the interaction is independent of NF90 RNA-binding properties. We
further showed that knockdown of NF90 resulted in significantly lower levels of PKR phosphorylation in response to dsRNA induction and influenza virus infection. We also showed that high concentrations of NF90 exhibit negative regulatory effects on PKR
phosphorylation, presumably through competition for dsRNA via the C-terminal RNA-binding domain. PKR activation is essential
for the formation of stress granules in response to dsRNA induction. Our results showed that NF90 is a component of stress granules.
In NF90-knockdown cells, dsRNA treatment induced significantly lower levels of stress granules than in control cells. Further evidence for an NF90–PKR antiviral pathway was obtained using an NS1 mutated influenza A virus specifically attenuated in its
ability to inhibit PKR activation. This mutant virus replicated indistinguishably from wild-type virus in NF90-knockdown cells,
but not in scrambled control cells or Vero cells, indicating that NF90’s antiviral function occurs through interaction with PKR.
Taken together, these results reveal a yet-to-be defined host antiviral mechanism in which NF90 upregulation of PKR phosphorylation restricts virus infection. The Journal of Immunology, 2014, 192: 000–000.
2
Materials and Methods
Cells and viruses
HEK 293T cells, MEF cells, HeLa cells, and Vero cells were maintained in
Dulbecco’s minimal essential medium, whereas MDCK cells were cultured in Eagle’s MEM. Culture media were supplemented with 10% FBS,
100 IU penicillin G/ml, and 100 ml streptomycin sulfate/ml, and cells were
incubated at 37˚C in a 5% CO2 atmosphere. NF90-knockdown or scrambled control 293T cells were kindly provided by Dr. Christopher Basler
(Mount Sinai School of Medicine, New York, NY) (14). PKR-knockout
(PKR2/2) and wild-type (WT) (PKR+/+) primary MEF cells were a generous
gift from Drs. John C. Bell (Ottawa Health Research Institute, Ottawa, ON,
Canada) and Craig McCormick (Dalhousie University, Halifax, NS, Canada)
(34).
WT and NS1-mutated influenza viruses (A/WSN/1933 [H1N1]) were
propagated in MDCK cells cultured in MEM containing antibiotics and 1
mg TPCK-trypsin/ml at 37˚C for 48 h. Supernatants were harvested at
designated time points, and viruses were titrated in MDCK cells by plaque
assay. Briefly, virus culture supernatant samples were serially diluted in
PBS and adsorbed onto confluent MDCK cells for 1 h at 37˚C. The inoculum was removed, and the cells were washed with PBS and covered
with 2 ml an agar medium (1% agarose, 1 mg TPCK-trypsin/ml in MEM).
After 2 d of incubation, plaques were counted, and the virus titer (PFU/ml)
was calculated.
Plasmid construction
The NP, PB1, PB2, and PA genes, derived from the H5N1 strain A/VNM/
1194/2004 and the H1N1 strain A/WSN/1933, were cloned into pCDNA3.1Flag/-V5 and pCMV-Flag vectors, respectively. The NS1 gene was cloned
into a Flag-tagged vector (33). For construction of recombinant virus, the
viral genomes were cloned into the pHW2000 vector (37). The QuikChange
II Site-Directed Mutagenesis Kit (Stratagene) was used to introduce R38A
and R35A (19, 33) and I123A/M124A/K126A/N127A (33, 38) point mutations into NS1, as well as to construct an NS1-deletion mutant (delNS1)
with stop codons introduced into residues C13 and L15. Plasmids pcDNA3.1NF90-V5 (NF90-V5) and pcDNA3.1-NF90-Flag were generated by cloning
the NF90 gene (GenBank accession number NM_004516; https://www.ncbi.
nlm.nih.gov/genbank/) into pcDNA3.1-V5 and pcDNA3.1-Flag vectors,
respectively. NF90-385-388A mutant was generated using the primer, 59GAGGAGAAGTCGCCCAGCGCAGCGGCGGCGAAGATTCAGAAGAAAGAG-39, and a complementary reverse primer. NF90-432/555A mutant
was made using primers 59-ACGACAAGCGCGCCGTCATGGAGGT-39
and 59-ATGCCCCCATCGCTACCATGTCTGT-39 and their complementary
reverse primers. NF90-T188/T315A mutant was made using primers 59-GTATTAGCTGGAGAAGCGCTATCAGTCAACG-39 and 59-CAACGGGAAGATATCGCACAGAGTGCGCAGC-39 and their complementary reverse
primers. NF90 (1–369 aa) and NF90 (370 aa-c) truncation mutants were
constructed by inserting NF90 1–369 aa or 370 aa-c fragments into pCDNA3.
1-Flag vectors. PKR (GenBank accession number NM_002759) was cloned
into pCDNA3.1-V5. Truncated mutants of PKR, including PKR (1–257) and
(258–551), also were constructed in the same vector.
Small interfering RNA knockdown of NF90
Transient knockdown of NF90 was achieved by transfection of NF90-specific
small interfering RNA (siRNA) (NM_004516_stealth_1155 59-AAAGAAGCCACUGAUGCUAUUGGGC-39 and its reverse complementary sequence) or 39-untranslated region siRNA (59-AATAGGACCAAGTTCAAAGGC-39 and its reverse complementary sequence), which were obtained
from Life Technologies and Sigma, respectively. HEK 293T cells were
transfected with 50 nM siRNA using Lipofectamine RNAiMAX transfection reagent (Invitrogen). After 24 h of incubation, NF90-silenced cells
were transfected with 100 ng the different NF90 constructs and incubated
for an additional 24 h. Cells were subsequently treated with 150 ng poly(I:C)
for 6 h. The effect of the various NF90 constructs on stress granule formation
was visualized using an LSM700 confocal microscope in the Core facility at
Li Ka Shing Faculty of Medicine, University of Hong Kong.
Transfection and coimmunoprecipitation
For transfection experiments, subconfluent (70%) monolayer cells were
transiently transfected or cotransfected with plasmids using TransIT-LT1
transfection reagent (Mirus). A total of 10–500 ng plasmid was added to
each well in a 24-well plate, in accordance with the user manual. For dsRNA
treatment, 100 ng poly(I:C) (Invitrogen) and 400 ng pUC19 carrier DNA
were combined to make a mix containing 500 ng DNA, which was then
transfected into cells (34).
Coimmunoprecipitation (co-IP) was performed using Dyna Beads
(Invitrogen). Specific Abs were added to the beads and incubated at room
temperature for $20 min. Cell lysates were obtained by adding lysis buffer
(50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% Triton X-100, and protease
inhibitor mixture; Roche) to the cells, shaking for 15 min on ice, and then
centrifuging. Cell lysate supernatants were added to the Ab-coupled beads,
and the mixture was incubated at room temperature for 2 h or overnight at
4˚C. Finally, beads were washed three times with lysis buffer, and Western
blotting was performed.
Immunoprecipitation products or diluted cell lysate supernatants were
mixed with SDS loading buffer and heated at 95˚C for 10 min. Samples were
fractionated by 12% SDS-PAGE and then blotted onto nitrocellulose membranes (Bio-Rad). After blocking with 3% skim milk, membranes were
incubated with the primary Abs anti–Flag M2 (Sigma) at 1:2000 dilution,
anti-V5 (Invitrogen) at 1:2000 dilution, anti-PKR (Santa Cruz) at 1:1000
dilution, anti–p-PKR (T446; Abcam) at 1:1000 dilution, anti-NP (19) at
1:2000 dilution, anti-DRBP76 (NF90; BD) at 1:500 dilution, anti-tubulin
(Sigma) at 1:5000 dilution, or anti-NS1 (33) at 1:2000 dilution. IRDye
680– or IRDye 780–labeled donkey anti-mouse or anti-rabbit secondary
Abs (Li-Cor Biosciences) were used at a dilution of 1:5000. Membrane
blots were scanned to visualize Ab–protein bands using the Odyssey imaging system (Li-Cor Biosciences). Relative levels of protein present in
bands were analyzed by estimation of band density using ImageJ software.
Immunofluorescence assay and microscopy
Cells were grown on Millicell EZ slides (Millipore) and transiently
transfected with plasmids or infected with virus. For indirect immunofluorescence assay, cells were fixed for 15 min using 4% paraformaldehyde in
PBS, followed by permeabilization with 0.2% Triton X-100 in PBS for
5 min. The cells were washed with PBS and blocked with 5% normal
donkey serum in PBS at 37˚C for 30 min, followed by incubation for 1 h at
37˚C with primary Abs diluted in PBS as follows: NP 1:500, DRBP76
(NF90) 1:50, G3BP 1:200, PABP1 1:200, V5 1:200, and Flag 1:200. The
slides were washed with PBS and incubated with anti-mouse, anti-rabbit,
or anti-chicken conjugates (FITC, rhodamine, or aminomethylcoumarin
acetate, respectively), as appropriate, for 30 min at room temperature. For
costaining procedures in which primary Abs raised from the same species
were used, the primary Abs were coupled with Zenon mouse IgG labeling
reagents (Invitrogen) prior to application onto the cells, and the use of
secondary Abs was omitted (33). After staining, the cells were washed
several times and mounted in mounting buffer, with or without DAPI
(VECTASHIELD). Signals were examined using an LSM 700 confocal
microscope (Zeiss). Images were taken using a 633 or 403 oil-immersion
lens, and captured images were processed using ZEN software (Carl Zeiss).
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
dsRNA binding, because the C-terminal of NF90 possesses RNAbinding ability (6).
Activation of PKR leads to the phosphorylation of eIF2a and
results in the stalling of mRNA translation in cells (30). Activation
of eIF2a triggers formation of stress granules that are composed
of multiple RNA-binding proteins, including TIA-1 and Ras GAP
SH3-domain–binding protein (G3BP)-1 (31, 32). Interaction between influenza A virus and host stress granules was demonstrated
recently (33, 34). Activation of PKR to induce the formation of
stress granules is regarded as a hallmark of the cellular response
to virus infection (34–36). This study explored the possibility of
NF90 involvement in PKR activation and the formation of stress
granules in response to dsRNA treatment or virus infection of
cells. We found that NF90 interacts with PKR via its C-terminal
domain and further demonstrated that NF90 is a previously uncharacterized component of cellular stress granules formed when
cells are treated with stress-inducing agents, such as polyinosinicpolycytidylic acid [poly(I:C)] or arsenite. NF90 may inhibit virus
replication by regulating levels of PKR phosphorylation. Compared with control cells, stable NF90-knockdown 293T cells exhibit
lower levels of activated PKR and form significantly fewer stress
granules following treatment with poly(I:C). Using an NS1 mutant
influenza A virus, which is specifically attenuated in its ability to
antagonize PKR activation but unaltered with respect to other
functions, this study confirms that NF90 is required for PKR activation in response to virus infection. NF90 may serve as a regulator for PKR activation in response to dsRNA in cells.
NF90 REGULATES ACTIVATION OF PKR
The Journal of Immunology
3
Quantitative real-time PCR
NF90 and V5-tagged full-length and C-terminal– and N-terminal–
truncated versions of PKR (Fig. 1A). NF90 and PKR plasmids were
coexpressed in 293T cells, and co-IP with anti-V5 Ab showed that
both N- and C-terminals of PKR interact with NF90 (Fig. 1B).
NF90 contains a DZF motif in the N-terminal and an RNA-binding
domain (RBD) in the C-terminal (Fig. 1A). We then investigated the
PKR-interacting domain on the NF90 molecule by constructing 1–
369 N-terminal and 370–702 C-terminal truncates and RBD mutant
(RBDm) of NF90. These Flag-tagged truncated or mutated forms
of NF90 were coexpressed with V5-tagged full-length PKR in
293T cells and coprecipitated with anti-Flag. Our result showed that
only the C-terminal of NF90 interacts with PKR and that the interaction is independent of NF90 RNA-binding properties (Fig. 1C).
NF90-knockdown or scrambled 293T cells were infected with WSN WT
or mutant viruses at a multiplicity of infection (MOI) of 0.1. Total RNAs
were extracted at 24 h postinfection using RNAiso (Takara). cDNAs were
synthesized by reverse transcription using SuperScript II Reverse Transcriptase (Invitrogen) and random primers. A SYBR Green–based real-time
PCR method (Roche) was used to detect target mRNAs with the LightCycler system (Roche). Primers for detecting IFN-b were 59-GCCGCATTGACCATCT-39 and 59-CACAGTGACTGTACTCCT-39. Actin mRNA
was quantified, and the information was used to normalize the total RNA
concentration between different samples, as described previously (19). A
reaction mix of 20 ml was composed of 10 pmol each gene-specific primer,
10 ml SYBR Green Master Mix, and 2 ml cDNA. The amplification program was as follows: 95˚C for 10 min, followed by 40 cycles of 95˚C for
10 s, 60˚C for 10 s, and 72˚C for 15 s. The specificity of the assay was
confirmed by melting-curve analysis at the end of the amplification program.
Results
NF90 interacts with PKR
FIGURE 1. Interaction between NF90 and
PKR. (A) Illustration of NF90 domain required
for interaction with PKR. (B) Co-IP analysis of
PKR and NF90. V5-tagged full length (FL) and
N-terminal (N) and C-terminal (C) truncates of
PKR were coexpressed with Flag-tagged NF90
by transfection into 293T cells. At 48 h posttransfection, cell lysates were prepared and
precipitated with anti-V5 Ab and then blotted
with anti-Flag Ab. (C) Reciprocal co-IP analysis
of NF90 and PKR. Flag-tagged full-length (FL),
N-terminal (N), and C-terminal (C) truncates
and RBD mutated NF90 were coexpressed with
V5-tagged PKR by transfection in 293T cells
for 48 h. Cell lysates were subjected to co-IP
with anti-Flag and blotted with anti-V5 Abs.
Right panels in (B) and (C) show co-IP input
material: 1/20 diluted lysates. The strong low
m.w. bands seen in (B) and (C) are M2-V5 (M2
protein of influenza A virus (A/Vietnam/1194/04)
and GFP-Flag (full-length GFP), respectively,
which were included as negative controls.
Previous reports (6, 7, 9) demonstrated that NF90 interacts with
PKR and is one of the substrates of PKR. The phosphorylated form
of NF90, MPP4, also was identified as a phosphoprotein detected
during the M phase of the cell cycle (9, 39). Other studies indicated
that NF90 may activate PKR and increase eIF2a phosphorylation
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
NF90 was shown previously to interact with PKR using yeast twohybrid and GST pull-down assays (6, 29). To verify their interaction
in mammalian cells and map the interacting domains on both proteins, we first tested interactions between Flag-tagged full-length
NF90 enhances phosphorylation of PKR upon induction by
dsRNA
4
NF90 into NF90-knockdown cells to reveal the effect of NF90 on
PKR phosphorylation in response to dsRNA. We found that PKR
phosphorylation was enhanced in response to low amounts of
transfected NF90, but the positive effect disappeared when higher
concentrations (500 ng) of NF90 were introduced into the NF90knockdown cells (Fig. 2C). To test whether PKR and NF90 may
be competing for dsRNA binding in poly(I:C)-treated cells, we
used NF90 truncated at either the N-terminal or RBD (C-terminal)
or containing mutations abolishing RNA-binding ability (RBDm)
(Fig. 1A) (41). We first confirmed that the full-length and N-terminal
forms of NF90 bind poly(I:C) and that NF90 RBDm does not bind
dsRNA by performing poly(I:C) bead pull-down assays (data not
shown). Interestingly, it was found that, at the NF90 concentration
that exhibited inhibitory effects (500 ng), WT (full-length) NF90
and the mutant form containing only the C-terminal significantly
inhibited PKR activation following poly(I:C) treatment of 293T
cells, whereas N-terminal and RBDm NF90 exerted little or no
inhibitory effect, respectively (Fig. 2D) (41). The different p-PKR
expression levels associated with the NF90 mutants also may be
due to the positive or negative effect of these clones on PKR activation. It appears that NF90 has dual and conflicting functions, in
FIGURE 2. NF90 is required for phosphorylation of PKR in response to dsRNA activation. (A) Levels of PKR phosphorylation were determined by Western
blotting of lysates of 293T cells in which knockdown of NF90 with siRNA for 56 h was followed by treatment with poly(I:C) (100 ng) or viral RNA for 6 h.
Different cell treatments are indicated in the figure. Efficiency of NF90 knockdown (right panel). A scrambled siRNA (sc) was used as a control. (B) Levels of PKR
phosphorylation in NF90 stably knocked-down cells (kd) and scrambled control cells (sc) were determined as in (A). Levels of NF90 in both NF90-knockdown and
scrambled cells also were examined (right panel). b-tubulin was detected to normalize cell numbers in (A) and (B). (C) Dose-dependent NF90 regulation of PKR
phosphorylation in cells treated with poly(I:C) dsRNA. NF90-knockdown cells were transfected with different amounts of NF90-encoding plasmid 18 h prior to
dsRNA treatment. Cells were treated with poly(I:C), collected, and lysed for Western blot analysis. The amounts of NF90 transfected are indicated; empty vector
was used to bring the total DNA in each well to 500 ng. (D) 293T cells were transfected with 500 ng of NF90 WT, truncation mutants 1–369 aa (N) or 370 aa-c (C),
site-mutated mutant 432/555A (RBDm) (Fig. 1A), or control pCDNA3.1 vector (vec) before treatment with poly(I:C). Sample preparation and Western blotting
were carried out as described above. Levels of NF90 produced from transfected plasmids were analyzed using anti-Flag Ab. Relative levels of protein present in
bands were analyzed by estimation of band density using ImageJ software in Western blots.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
in yeast, but these positive effects on PKR phosphorylation could
not be seen when purified proteins were used in vitro (29, 40),
suggesting that additional factors may be missing in the in vitro
assay. However, in the in vitro study, it was demonstrated that high
concentrations of NF90 inhibited PKR activation, possibly through
competitive binding to dsRNA (40). These observations from yeast
and in vitro studies prompted us to investigate the effect of NF90 on
PKR phosphorylation in mammalian cells. In this study, we examined the levels of the phosphorylated form of PKR in both NF90knockdown and scrambled control 293T cells following PKR activation induced by transfection of either dsRNA [poly(I:C)] or viral
RNA. We found that, although expression of PKR protein is not
changed, the level of the phosphorylated form of PKR is significantly lower (∼3–4–fold) in NF90-knockdown cells compared with
that in scrambled siRNA-treated cells (Fig. 2A). The observation
that NF90 is required for PKR phosphorylation in response to
dsRNA treatment was confirmed in a pair of 293T cell lines: one in
which NF90 was stably knocked down with specific short hairpin
RNA and the other a control created using a scrambled short hairpin
RNA (14) (Fig. 2B). To delineate the mechanism for the regulation
of PKR activation by NF90, we transfected increasing amounts of
NF90 REGULATES ACTIVATION OF PKR
The Journal of Immunology
that it can support PKR phosphorylation (Fig. 2A) while inhibiting
the level of PKR phosphorylation through sequestration of dsRNA
and consequent prevention of PKR activation via the C-terminal
RBD (Fig. 2C, 2D). A study (13) found that PKR phosphorylated
T188 and T315 of NF90 and regulated NF90 cellular shuttling.
The T188A/T315A (attenuated) and the T188D/T315D (constitutively
active) form of NF90 did not exhibit any altered activity with effect
on PKR phosphorylation in response to dsRNA treatment compared
with WT NF90 (data not shown). These results establish a mechanism in which NF90 is involved in a complex process of host
regulation of gene expression exerted through the PKR pathway.
NF90 is a component of stress granules
NF90 was shown to inhibit replication of a broad range of viruses
and to form complexes with NF45 and shuttle between the nucleus
and cytoplasm (13–19). However, a detailed mechanism describing
how NF90 negatively regulates virus replication remains unclear.
One of the well-defined pathways set in motion following PKR
activation was found to associate with the regulation of mRNA
translation through induction of stress granules in virus-infected
cells (25, 34, 35). Taken together with the observed interaction
between NF90 and PKR, a mechanism for regulation of protein
synthesis involving these two proteins seems likely. This study
tested the possibility that NF90 is an undefined component of
PKR-induced stress granules. We examined stress granules induced
by treatment with arsenite and poly(I:C) for the presence of NF90.
NF90 was found to colocalize with the stress granule markers
PABP1 and G3BP (Fig. 3A), as well as TIA-1 (data not shown), in
both arsenite- and poly(I:C)-induced granules. In untreated control
cells, NF90 predominantly localized in the nucleus and did not
colocalize with PABP1 in the cytoplasm. To investigate whether
NF90 is involved in the formation of stress granules, we used the
pair of NF90 stable knockdown 293T cell lines described above
(14) (Fig. 2B). It is notable that significantly more stress granules
were observed in the scrambled control cells compared with the
NF90-knockdown cells following induction with poly(I:C) (Fig.
3B, p , 0.05). Interestingly, there was no significant difference in
granule formation between NF90-knockdown and scrambled cells
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 3. NF90 colocalizes with
markers of stress granules. (A) 293T
cells were transfected with poly(I:C)
for 6 h, treated with sodium arsenite
for 30 min, or left untreated, and immunofluorescence assays were performed as described previously (33).
Stress granules were examined using
Abs specific for G3BP or PABP1 and
NF90 (anti-DRBP76) and visualized
using a confocal microscope. DAPI
is shown in blue in merged images
(original magnification 363). (B)
Scrambled or NF90 stable knockdown
(kd) cells were treated with poly(I:C)
or arsenite and examined for the
formation of stress granules (SG)
by staining with anti-G3BP and
DAPI. Stress granules (G3BP) present in the cytoplasm are shown in
green. (C) 293T cells were treated with
poly(I:C) or mock treated for 4 h;
treatment medium was removed and
replaced with medium containing
2-AP (with final working concentration of 10 mM) in vehicle (acetic
acid) or vehicle only. For statistical
analysis of cells containing stress
granules (SG), ∼100–200 cells were
examined for the presence of SG in
each of three independent experiments, and the average percentage
of SG+ cells was calculated. Error bars
represent SD. (D) Enlarged images
of stress granules in scrambled and
NF90-knockdown (kd) cells after
treatment with poly(I:C) (original
magnification 340). *p , 0.05,
**p , 0.01.
5
6
FIGURE 4. NF90 RNA binding and phosphorylation by PKR are required for localization to stress granules in response to poly(I:C)
treatment. (A) Plasmids containing V5-tagged
NF90 (NF90-WT) and NF90 nuclear localization (NF90-NLSm) and RNA binding (NF90RBDm) mutants, as well as Flag-tagged phosphorylation mutant (NF90-A), as shown in Fig.
1A, were transfected into 293T cells. Cells were
treated with poly(I:C) for 6 h, fixed, and stained
with anti-V5 or Flag, along with anti-PABP1 or
anti-TIA Abs, to assess colocalization of NF90
and stress granules (original magnification 363).
(B) Western blot analysis of NF90 expression in
293T cells treated with NF90-specific siRNA or
scrambled siRNA or mock treated. b-tubulin was
used as a loading control. (C) NF90 was knocked
down with siRNA, as described above, and cells
were transfected with WT NF90 or RNA-binding
(NF90-RBDm) or phosphorylation (NF90-A)
mutants. Formation of stress granules (SG) in
response to poly(I:C) treatment was estimated
in cells (left panel). Values represent the means
of 10 randomly selected microscopic field of views
for each sample (∼100 cells/field of view); error
bars show SD. Expression of NF90 constructs was
confirmed by Western blotting (lower panel). *p ,
0.05, **p , 0.005, ****p , 0.0001. (D) Representative images of stress granules (red) in cells
expressing WT or different mutants of NF90
(original magnification 340).
was not able to localize to stress granules (Fig. 4A). PKR was
reported to phosphorylate T188 and T315 of NF90, thereby regulating NF90 cellular shuttling (13). We found that the phosphorylation-deficient T188A/T315A mutant (NF90-A) (Fig. 1A) exhibited
a deficiency in localization to stress granules in response to dsRNA
treatment (Fig. 4A). The control, 293T cells transfected with vector,
showed no stress granules (data not shown). By using siRNA
targeting the 39-untranslated region of NF90, we were able to knock
down NF90 expression to a minimal level in 293T cells (Fig. 4B).
We then examined different NF90 mutants for the ability to restore
formation of stress granules in NF90-knockdown cells. Although
both RBDm (NF90-RBDm) and phosphorylation mutant (NF90-A)
possess low-level stress granule formation activity, the ability of
these mutants to restore stress granule formation in poly(I:C)treated NF90 knockdown cells was considerably weaker than that
of WT NF90 (Fig. 4C, 4D). Taken together, these results suggest
that RNA binding and phosphorylation of NF90 are required for
induction of stress granules via dsRNA activation of PKR.
Mutant of influenza A virus associated with PKR activation
NF90 was shown to be involved in the host antiviral response and
to limit influenza virus replication (12, 19). We attempted to determine whether there is an NF90–PKR antiviral pathway that
enhances activation of PKR, using influenza virus as a model. The
nonstructural protein 1 (NS1) of influenza A virus is able to inhibit
PKR activation and stress granule formation induced by PKR
activation, thereby supporting virus replication (33, 34). We previously described a panel of NS1 mutant viruses that is not able to
inhibit formation of stress granules (33) (Fig. 5A). We confirm in
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
treated with arsenite (Fig. 3B, lower panels). Because poly(I:C),
but not arsenite, activates PKR (42), these results suggest that
NF90’s role in the formation of stress granules requires PKR. To
further verify that NF90’s facilitation of stress granule induction
in cells in response to poly(I:C) treatment is mediated by PKR, we
tested whether treatment with the PKR inhibitor, 2-AP (43), affected stress granule induction. Our results showed that 2-AP, but
not the vehicle-only control, reduced stress granule formation in
NF90 WT cells treated with poly(I:C) (Fig. 3C, p , 0.01). These
results clearly demonstrate that NF90-regulated stress granule
formation in response to dsRNA induction is mediated by PKR.
In contrast to the scrambled 293T cells, in which distinct stress
granules were observed, PABP1 exhibited a diffuse pattern of expression in NF90-knockdown cells (Fig. 3D), suggesting that NF90
may be a core nucleating component for the formation of stress
granules in response to dsRNA induction. Just as we established
that the C-terminal of NF90 is required for interaction with PKR
(Fig. 1C), it was important to identify which domain of NF90
is associated with stress granules. Because NF90 predominantly
localizes to the nucleus in normal conditions, we created a mutant
NF90 protein that does not have the ability to localize to the nucleus
to better understand NF90’s association with stress granules. When
four substitutions were introduced into the putative nuclearlocalization signal of NF90 (residues 385–388) (Fig. 1A), this mutated NF90 (NF90-NLSm) was found exclusively in the cytoplasm
of poly(I:C)-treated cells, where it localized to stress granules (Fig.
4A). This seems to suggest that the nuclear-localization ability of
NF90 is not associated with its interaction with stress granules. In
contrast, an RBDm (NF90-RBDm) that lacked RNA-binding ability
NF90 REGULATES ACTIVATION OF PKR
The Journal of Immunology
of PB1, PB2, PA, and NP) were coexpressed with NF90, only NP
was targeted to stress granules (Fig. 6A, lower panel). Using NP
as a marker for influenza A virus–induced stress granules, we found
that, although WT virus suppresses formation of stress granules, NP
was clearly detectable in stress granules in 293T, MEF, and HeLa
cells infected with 123-127A mutant influenza virus at 24 h postinfection. In addition, the nonfunctional R35A NS1 and del-NS1
mutants also failed to inhibit formation of NP-associated stress
granules in infected cells (Fig. 6B–D). Quantitation of stress
granule–forming 293T cells infected with WT or mutant viruses
further confirmed that inhibition of PKR blocks induction of stress
granules upon virus infection (Fig. 6E). The characteristics of
these NS1 mutants of influenza A virus indicated their suitability
for further use in studying the PKR-related antiviral pathway.
NF90 antiviral activity is not associated with expression of
IFN-b
We next performed a series of infections, using WT and mutant
influenza A viruses, to demonstrate the role of NF90 in the PKRinduced antiviral pathway and to confirm that it is independent of
IFN expression. We previously showed that there is a significant
increase in the replication of PR8 (H1N1) and 1194 (H5N1) influenza viruses in cells in which NF90 has been transiently knocked
down using siRNA targeting NF90 compared with cells treated
with control siRNA (19). Using stable NF90-knockdown cells, we
observed a similar significant negative effect (p = 0.0001) of NF90
on virus replication in cells infected with another H1N1 virus
strain, A/WSN/1933 (Fig. 7A). To verify that the altered virusreplication efficiency is associated with the NF90–PKR pathway,
we compared the growth kinetics of WT WSN virus with those of
mutant WSN viruses containing different NS1 mutations, which
were attenuated in inhibition of PKR activation (123-127A) or
FIGURE 5. Characterization of an NS1 mutant of influenza A virus that is attenuated in the ability to inhibit PKR phosphorylation in response to dsRNA
activation. (A) Illustration of the NS1 protein of influenza A virus (A/WSN1933), showing locations of mutations used in this study. (B) 293T cells were
transfected with pCMV expression vectors containing WT or R35A, R38A, or 123-127A quadruple NS1 mutants or with vector expressing GFP, or they
were left untransfected prior to treatment with poly(I:C). Cells were harvested and lysed for Western blot analysis of NS1 and PKR expression and PKR
phosphorylation. Expression of b-tubulin was analyzed for normalization of protein loading. (C) 293T cells were infected with WT or R38A or 123-127A
NS1 mutant viruses for 24 h, and PKR activation levels were examined. Virus infection is shown by viral NP expression. (D) 293T cells were treated with
siRNA targeting NF90 or scrambled siRNA for 56 h and then infected with WT or 123-127A NS1 mutant virus or were mock infected. Cell lysates were
examined for levels of PKR phosphorylation at 24 h postinfection. Levels of total PKR also were estimated as a control. Relative levels of protein present in
bands were analyzed by estimation of band density using ImageJ software in Western blots, as described above.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
this study that the NS1 mutants that cannot inhibit stress granule
formation also have lost the ability to inhibit PKR activation induced by poly(I:C) or virus infection (Fig. 5B, 5C). Consistent with
previous observations, the R38A mutant form of NS1, which is
attenuated in its ability to suppress host IFN expression, retains
the ability to inhibit PKR phosphorylation similarly to WT NS1
(33, 44). The R35A mutant, which has lost both RNA-binding and
dimerization ability (45), was used as a nonfunctional NS1 control
and showed no activity in inhibiting PKR activation. Although
RNA-binding activity is not affected, as described in a previous
study (33), NS1 mutants that combine I123A/M124A/K126A/N127A
(123-127A) quadruple mutations exhibited a reduced ability to
inhibit PKR activation induced by poly(I:C) or virus infection
(Fig. 5B, 5C). Using this PKR-specific NS1 mutant, we further
demonstrated that knockdown of NF90 produced a negative
effect on PKR phosphorylation upon virus infection (Fig. 5D).
We then examined whether the 123-127A quadruple mutant had
lost the ability to suppress formation of stress granules, in addition
to being unable to inhibit PKR activation. We first examined
whether any influenza A virus components were targeted to stress
granules. We showed previously that NP interacts with RAP55 and
colocalizes to P-bodies and stress granules (33). Moreover, NP also
interacts with NF90 (19). Because NF90 is associated with stress
granules, as shown above, it seems possible that it negatively
regulates influenza A virus replication by targeting viral NP to
stress granules. To test this idea, individual PB1, PB2, PA, and NP
influenza virus proteins were expressed in cells, and their colocalization with stress granules was examined following treatment
with poly(I:C). Only NP colocalized with the stress granule marker
G3BP following poly(I:C) induction (Fig. 6A, upper panel). A
colocalization assay with another stress granule marker, PABP1,
showed a similar result (data not shown). When RNP (composed
7
8
NF90 REGULATES ACTIVATION OF PKR
IFN expression (R38A) (as described in Fig. 5A), in both NF90
stable-knockdown and scrambled 293T cells. In the scrambled
293T cells, antiviral mechanisms involving expression of IFN-b
and activation of PKR are intact, with both 123-127 and R38A
mutant viruses replicating to a significantly lower level (p = 0.0001)
compared with the WT virus (Fig. 7B). In the NF90-knockdown
cells, it is notable that 123-127A quadruple mutant virus showed
no sign of attenuation and replicated to a level similar to the WT
virus, whereas replication of R38A mutant virus was severely
attenuated, presumably due to the antiviral effects of IFN (Fig. 7C).
The R38A mutant is well recognized for its reduced ability to inhibit
the host IFN antiviral response (46). To differentiate the effects
caused by the IFN antiviral response from those associated with
the NF90–PKR pathway, we compared virus-replication efficiency
in Vero cells, which are deficient in IFN expression. In contrast to
the result obtained in NF90-knockdown cells, R38A mutant virus
is able to replicate to a level close to that of WT virus, whereas the
123-127A quadruple mutant is significantly attenuated (p = 0.001)
(Fig. 7D). PKR-knockout (PKR2/2) and WT (PKR+/+) primary
MEF cells were used to further verify the linkage of NF90 and PKR.
Western blot showed NF90 expression in both PKR-knockout
(PKR2/2) and WT (PKR+/+) primary MEF cells. Using this influenza A virus–infection model, we further demonstrated that PKR is
a downstream effector for NF90 antiviral activity, because infection
with mutant viruses that are unable to inhibit PKR activation (NS1
R35A or 123-127A) induces stress granules in PKR WT (PKR+/+)
cells, but not in PKR knockout (PKR2/2) cells with normal NF90
(Fig. 7F, 7G), and expression of PKR by transfection restores
stress granule–forming ability in PKR2/2 cells (Fig. 7G, lower
panel). PKR is an IFN-inducible gene, and it is not known whether
the differences in stress granule induction may be due to variations
in inhibition of IFN expression by WT and mutant NS1 proteins.
However, examination of inhibition of IFN-b expression in WT
and mutant virus-infected cells revealed that both the WT and 123127A quadruple mutant viruses are able to inhibit IFN-b expression, but R38A virus cannot (Fig. 7E). Taken together, these results support the idea of an underlying mechanism whereby NF90
is associated with PKR activation in the host antiviral response,
and the related antiviral signaling is independent of induction of
IFN-b expression (Fig. 8).
Discussion
NF90 was found to play an important role in host innate immunity
against various virus infections (12–15). We (19) previously found
that NF90 negatively regulates influenza A virus replication through
interaction with viral NP. Although the significance of NF90 in
restricting virus replication is evident from the magnitude of virus
replication increase in NF90-knockdown cells (12, 19), the molecular
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 6. The influenza A virus mutant with attenuated ability to inhibit PKR
phosphorylation also lost the ability to
suppress formation of stress granules upon
virus infection. (A) Plasmids expressing
NP and individual subunits of the RNP
complex derived from the A/Vietnam/
1194/2004 (H5N1) strain of influenza A
virus were transfected into 293T cells
for 16 h. Cells were then treated with
poly(I:C) for 6 h before fixation. Colocalization of NP, PB1, PB2, and PA with
G3BP was examined as described above.
V5-tagged NP and Flag-tagged PB1, PB2,
and PA were coexpressed in 293T cells
(upper panel, original magnification 363),
and cells were examined for colocalization
with G3BP (lower panel, original magnification 363). Cells were stained with
Abs specific for G3BP, V5 (NP), and Flag
(PB1, PB2, and PA). NP-associated stress
granules were examined in 293T (B),
MEF (C), and HeLa (D) cells infected
with WT or 123-127A or R35A mutant
viruses by examining cells for colocalization of G3BP and NP in the cytoplasm
(original magnification 363). (E) Formation of stress granules (SG) in 293T cells
infected with WSN WT or 123-127A,
R35A, or R38A NS1 mutant viruses.
Stress granules were examined by staining
cells with G3BP-specific Ab at 24 h postinfection. Virus-infected cells were visualized by staining for the presence of NP
protein. For statistical analysis, 60–300
cells were examined, and the percentage
of infected cells containing stress granules was determined. The average values
from two independent experiments were
calculated and graphed; error bars represent SD.
The Journal of Immunology
9
details of the NF90-associated antiviral mechanism remain largely
unknown. NF90 is one of the evolutionarily conserved members of
the dsRNA-binding protein family and is expressed abundantly in
various human cells. Previous studies (6, 7, 9, 10, 13, 40) identified
and confirmed PKR as the key interacting partner for NF90 in cells,
in addition to its function in regulating IL-2 gene expression in
T cells (4). PKR plays an important role in mediating innate immunity to viral infection and is required for NF90 antiviral activity,
as demonstrated in a vesicular stomatitis virus infection experiment
using PKR-knockout (PKR2/2) and control (PKR+/+) MEFs (13,
25). Thus, it is natural to reason that the antiviral activity exerted
by NF90 may lie within the PKR-signaling pathway.
Phosphorylation of PKR is essential for activation of its function,
and it can be regulated by host RNA-binding proteins other than
dsRNA (47–49). Two models currently describe PKR activation:
the autoinhibition and dimerization models (50). It was suggested
that binding of either dsRNA or PACT to PKR may interrupt
intramolecular interaction between the kinase and dsRNA-binding
domains, leading to an open conformation of the PKR molecule
and allowing autophosphorylation to occur (49). NF90 is phosphorylated by PKR in its RBD (6). Previous studies (7, 9, 29)
showed that NF90 interacts with PKR in the yeast two-hybrid and
GST-pull down assays. We confirmed interaction between NF90
and PKR in 293T cells using a co-IP assay and showed that it is
independent of RNA binding of NF90, because the RBDm could
interact with PKR (Fig. 1C). On testing the effect of NF90 on PKR
activation, we found that low concentrations of NF90 restore some
PKR activity in response to poly(I:C) induction in NF90 stably
knocked-down cells. However, when high concentrations of NF90
were introduced by transfection, PKR phosphorylation was inhibited in these cells. Furthermore, the results showed that the inhibitory effect is exerted by the dsRNA-binding domain, presumably
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 7. NF90 inhibition of influenza A virus is mediated via the PKR pathway. (A) Growth kinetics in NF90 stable-knockdown and scrambled
293T cells infected with the influenza A virus A/WSN/1933 at an MOI of 0.005. Growth kinetics of influenza viruses containing specific mutations that
attenuate their ability to inhibit PKR activation were determined in different cell lines. Scrambled (B) and NF90-knockdown (C) 293T cells and Vero cells
(D) were infected with WT A/WSN/1933 (H1N1) or mutant viruses containing NS1 R38A or 123–127A quadruple mutations at an MOI of 0.05.
Supernatants were collected at 8, 24, and 48 h postinfection (h.p.i.) for titration by plaque assay. (E) NF90 inhibition of influenza A virus is not caused by
expression of IFN-b. NF90-knockdown (kd) cells and scrambled control cells were infected with WT or NS1 R38A or 123-127A mutant viruses at an MOI
of 0.1 or were mock infected (PBS). At 24 h postinfection, cellular RNAs were extracted, and quantitative real-time PCR was performed. IFN-b mRNA
levels were determined after normalization. (F) Western blot showing NF90 and PKR expression in both PKR-knockout (PKR2/2) and WT (PKR+/+)
primary MEF cells. Expression of b-tubulin was assessed as a loading control. (G) PKR-knockout (PKR2/2) and WT (PKR+/+) primary MEF cells were
infected with R35A or 123–127A virus for 24 h before being subjected to immunofluorescence assay for visualization of stress granules (upper panel,
original magnification 363). G3BP, NP, and DAPI are shown in green, red, and blue, respectively. PKR expression in PKR-knockout MEF cells was
restored by transfection of a V5-tagged PKR construct (lower panel, original magnification 363). MEF cells with restored PKR expression were infected
with 123–127A mutated influenza A virus. Formation of antiviral stress granules was examined by immunostaining using anti-V5 (PKR), anti-NP, and antiG3BP Abs. Error bars represent SD. *p , 0.05, **p , 0.01, ***p , 0.001, t test.
10
NF90 REGULATES ACTIVATION OF PKR
through competition with PKR for dsRNA ligand in an RNAbinding–dependent manner, as was suggested previously (6). This
hypothesis was confirmed in this study using an NF90 mutant
containing mutations in the RBD, which shows no inhibition of
PKR phosphorylation when transfected into 293T cells at high
concentrations (Fig. 2D). NF90 predominantly localizes to the nucleus but may be exported to the cytoplasm to perform its functions
(5, 8). NF90 is also phosphorylated by PKR, and activated NF90
does not inhibit PKR activation (6). It was proposed that shuttling
of NF90 between nucleus and cytoplasm is regulated through phosphorylation of NF90 by PKR (13). However, these studies did not
investigate whether NF90 exerted any effect on PKR. Our results
suggest that a physiological concentration of NF90 is required for
optimal activation of PKR in response to stimuli, such as cellular
dsRNA or intermediate dsRNA products generated during virus
replication. Because PKR phosphorylation can activate downstream antiviral cascades, including NF-kB and inflammasomes
(28, 51), it is postulated that NF90 may serve as a regulator for
PKR activation in response to cellular dsRNA stimuli or virus
infection. Our findings suggest a mechanism in which NF90 may
act to regulate PKR activity in cells through balancing PKR phosphorylation in response to dsRNA via its N- and C-terminal functions.
The PKR-activation function of NF90 lies in the N-terminal region.
In contrast, interaction between NF90 and PKR also may modulate
the activation of PKR through competition with dsRNA via its
C-terminal RBD; this also was reported in an in vitro study (6).
Hosts use multiple antiviral mechanisms to restrict virus infection; formation of stress granules is an important antiviral response
that limits virus replication by stalling viral protein synthesis (52,
53). Activation of PKR is required for the formation of stress granules in response to virus infection (34–36, 54). This study attempts to
delineate the mechanism of NF90 antiviral activity through examination of NF90 involvement in PKR activation and formation of
antiviral stress granules using the influenza A virus–infection model.
We found that NF90 is regularly associated with stress granules
formed in response to treatments with oxidative substances (arsenite) or dsRNA mimics [poly(I:C)]. It was further demonstrated
that NF90 may be required for the formation of stress granules in
the dsRNA-induced host response, because there is a significant
reduction in the levels of stress granules formed in response to
poly(I:C) treatment in NF90 stably knocked-down cells. However,
there was no significant difference in granule formation between
NF90 knockdown and scrambled cells treated with arsenite, which
does not involve PKR activation (42) (Fig. 3B), whereas treatment
with a PKR phosphorylation inhibitor, 2-AP (43), inhibited stress
granule induction (Fig. 3C). It seems likely that NF90’s association with stress granules is coupled with its ability to interact with
PKR and regulate its activation. Stable knockdown of NF90 renders 293T cells ineffective in induction of phosphorylation of PKR
in response to poly(I:C) treatment.
To further verify the linkage between NF90 and PKR signaling
in the host antiviral response, we generated an influenza A virus
specifically mutated in the NS1 gene that is attenuated in its ability
to inhibit PKR activation but otherwise functions normally. We
found that NP of this mutant is targeted to stress granules, presumably
through interaction with NF90 and/or other NP-interacting proteins.
Using this PKR-specific virus model, this study clearly demonstrated
that this mutant virus is able to replicate to a similar level as WT virus
in NF90-knockdown cells but not in scrambled control cells or in Vero
cells, which are deficient in expression of IFN-b. Our experiments also
showed that this specific mutant virus is able to inhibit IFN-b expression just as WT virus does but that the R38A mutant, which has
lost RNA-binding ability, is attenuated in terms of suppression
of IFN. These lines of evidence strongly support the hypothesis
that NF90 antiviral function is signaled through the PKR pathway.
Although further studies are needed to reveal the details of how
NF90 enhances PKR phosphorylation and how NF90 is recycled
back to the nucleus after engaging in PKR activation in the
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 8. Working model of PKR-mediated NF90 antiviral pathway. In normal conditions, NF90 migrates to the cytoplasm to facilitate PKR phosphorylation in response to virus infection or the presence of dsRNA during virus replication. Activation of PKR leads to the phosphorylation of eIF2a and
subsequently triggers formation of stress granules that involve NF90, PKR, and other stress granule components. However, viruses adapt different
mechanisms to antagonize NF90–PKR antiviral activity; for instance, influenza A viruses express NS1 to inhibit phosphorylation of PKR, which also may
interact with NF90, to enhance virus replication. PKR also may be activated by IFN to exert antiviral activity in an NF90-independent manner.
The Journal of Immunology
Acknowledgments
We thank Dr. Christopher Basler for the gift of NF90-knockdown and
scrambled 293T cells and Drs. Craig McCormick and John C. Bell for
kindly providing PKR knockout (PKR2/2) and WT (PKR+/+) primary MEF
cells. This work was done by X.W. in partial fulfillment of the requirements for a Ph.D. degree at the University of Hong Kong.
Disclosures
The authors have no financial conflicts of interest.
References
1. Durand, D. B., J. P. Shaw, M. R. Bush, R. E. Replogle, R. Belagaje, and
G. R. Crabtree. 1988. Characterization of antigen receptor response elements
within the interleukin-2 enhancer. Mol. Cell. Biol. 8: 1715–1724.
2. Corthésy, B., and P. N. Kao. 1994. Purification by DNA affinity chromatography
of two polypeptides that contact the NF-AT DNA binding site in the interleukin 2
promoter. J. Biol. Chem. 269: 20682–20690.
3. Kao, P. N., L. Chen, G. Brock, J. Ng, J. Kenny, A. J. Smith, and B. Corthésy.
1994. Cloning and expression of cyclosporin A- and FK506-sensitive nuclear
factor of activated T-cells: NF45 and NF90. J. Biol. Chem. 269: 20691–20699.
4. Shi, L., W. R. Godfrey, J. Lin, G. Zhao, and P. N. Kao. 2007. NF90 regulates
inducible IL-2 gene expression in T cells. J. Exp. Med. 204: 971–977.
5. Shim, J., H. Lim, J. R Yates, and M. Karin. 2002. Nuclear export of NF90 is
required for interleukin-2 mRNA stabilization. Mol. Cell 10: 1331–1344.
6. Parker, L. M., I. Fierro-Monti, and M. B. Mathews. 2001. Nuclear factor 90 is
a substrate and regulator of the eukaryotic initiation factor 2 kinase doublestranded RNA-activated protein kinase. J. Biol. Chem. 276: 32522–32530.
7. Saunders, L. R., D. J. Perkins, S. Balachandran, R. Michaels, R. Ford,
A. Mayeda, and G. N. Barber. 2001. Characterization of two evolutionarily
conserved, alternatively spliced nuclear phosphoproteins, NFAR-1 and -2,
that function in mRNA processing and interact with the double-stranded
RNA-dependent protein kinase, PKR. J. Biol. Chem. 276: 32300–32312.
8. Liao, H. J., R. Kobayashi, and M. B. Mathews. 1998. Activities of adenovirus
virus-associated RNAs: purification and characterization of RNA binding proteins. Proc. Natl. Acad. Sci. USA 95: 8514–8519.
9. Patel, R. C., D. J. Vestal, Z. Xu, S. Bandyopadhyay, W. Guo, S. M. Erme,
B. R. Williams, and G. C. Sen. 1999. DRBP76, a double-stranded RNA-binding
nuclear protein, is phosphorylated by the interferon-induced protein kinase,
PKR. J. Biol. Chem. 274: 20432–20437.
10. Langland, J. O., P. N. Kao, and B. L. Jacobs. 1999. Nuclear factor-90 of activated
T-cells: A double-stranded RNA-binding protein and substrate for the doublestranded RNA-dependent protein kinase, PKR. Biochemistry 38: 6361–6368.
11. Satoh, M., V. M. Shaheen, P. N. Kao, T. Okano, M. Shaw, H. Yoshida,
H. B. Richards, and W. H. Reeves. 1999. Autoantibodies define a family of proteins
with conserved double-stranded RNA-binding domains as well as DNA binding
activity. J. Biol. Chem. 274: 34598–34604.
12. Pfeifer, I., R. Elsby, M. Fernandez, P. A. Faria, D. R. Nussenzveig, I. S. Lossos,
B. M. Fontoura, W. D. Martin, and G. N. Barber. 2008. NFAR-1 and -2 modulate
translation and are required for efficient host defense. Proc. Natl. Acad. Sci. USA
105: 4173–4178.
13. Harashima, A., T. Guettouche, and G. N. Barber. 2010. Phosphorylation of the
NFAR proteins by the dsRNA-dependent protein kinase PKR constitutes a novel
mechanism of translational regulation and cellular defense. Genes Dev. 24:
2640–2653.
14. Shabman, R. S., D. W. Leung, J. Johnson, N. Glennon, E. E. Gulcicek,
K. L. Stone, L. Leung, L. Hensley, G. K. Amarasinghe, and C. F. Basler. 2011.
DRBP76 associates with Ebola virus VP35 and suppresses viral polymerase
function. J. Infect. Dis. 204(Suppl. 3): S911–S918.
15. Agbottah, E. T., C. Traviss, J. McArdle, S. Karki, G. C. St Laurent, III, and
A. Kumar. 2007. Nuclear Factor 90(NF90) targeted to TAR RNA inhibits transcriptional activation of HIV-1. Retrovirology 4: 41.
16. Hoque, M., R. A. Shamanna, D. Guan, T. Pe’ery, and M. B. Mathews. 2011.
HIV-1 replication and latency are regulated by translational control of cyclin T1.
J. Mol. Biol. 410: 917–932.
17. Merrill, M. K., E. Y. Dobrikova, and M. Gromeier. 2006. Cell-type-specific repression of internal ribosome entry site activity by double-stranded RNA-binding
protein 76. J. Virol. 80: 3147–3156.
18. Merrill, M. K., and M. Gromeier. 2006. The double-stranded RNA binding
protein 76:NF45 heterodimer inhibits translation initiation at the rhinovirus type
2 internal ribosome entry site. J. Virol. 80: 6936–6942.
19. Wang, P., W. Song, B. W. Mok, P. Zhao, K. Qin, A. Lai, G. J. Smith, J. Zhang,
T. Lin, Y. Guan, and H. Chen. 2009. Nuclear factor 90 negatively regulates influenza virus replication by interacting with viral nucleoprotein. J. Virol. 83:
7850–7861.
20. Isken, O., C. W. Grassmann, R. T. Sarisky, M. Kann, S. Zhang, F. Grosse,
P. N. Kao, and S. E. Behrens. 2003. Members of the NF90/NFAR protein group
are involved in the life cycle of a positive-strand RNA virus. EMBO J. 22: 5655–
5665.
21. Isken, O., M. Baroth, C. W. Grassmann, S. Weinlich, D. H. Ostareck,
A. Ostareck-Lederer, and S. E. Behrens. 2007. Nuclear factors are involved in
hepatitis C virus RNA replication. RNA 13: 1675–1692.
22. Gomila, R. C., G. W. Martin, and L. Gehrke. 2011. NF90 binds the dengue virus
RNA 39 terminus and is a positive regulator of dengue virus replication. PLoS
ONE 6: e16687.
23. Shamanna, R. A., M. Hoque, T. Pe’ery, and M. B. Mathews. 2013. Induction of
p53, p21 and apoptosis by silencing the NF90/NF45 complex in human papilloma virus-transformed cervical carcinoma cells. Oncogene 32: 5176–5185.
24. Meurs, E., K. Chong, J. Galabru, N. S. Thomas, I. M. Kerr, B. R. Williams, and
A. G. Hovanessian. 1990. Molecular cloning and characterization of the human
double-stranded RNA-activated protein kinase induced by interferon. Cell 62:
379–390.
25. Balachandran, S., P. C. Roberts, L. E. Brown, H. Truong, A. K. Pattnaik,
D. R. Archer, and G. N. Barber. 2000. Essential role for the dsRNA-dependent
protein kinase PKR in innate immunity to viral infection. Immunity 13: 129–141.
26. Levin, D., and I. M. London. 1978. Regulation of protein synthesis: activation by
double-stranded RNA of a protein kinase that phosphorylates eukaryotic initiation factor 2. Proc. Natl. Acad. Sci. USA 75: 1121–1125.
27. Pindel, A., and A. Sadler. 2011. The role of protein kinase R in the interferon
response. J. Interferon Cytokine Res. 31: 59–70.
28. Lu, B., T. Nakamura, K. Inouye, J. Li, Y. Tang, P. Lundbäck, S. I. Valdes-Ferrer,
P. S. Olofsson, T. Kalb, J. Roth, et al. 2012. Novel role of PKR in inflammasome
activation and HMGB1 release. Nature 488: 670–674.
29. Parker, L. M., I. Fierro-Monti, T. W. Reichman, S. Gunnery, and M. B. Mathews.
2001. Double-stranded RNA-binding proteins and the control of protein synthesis and cell growth. Cold Spring Harb. Symp. Quant. Biol. 66: 485–497.
30. Hershey, J. W. 1991. Translational control in mammalian cells. Annu. Rev.
Biochem. 60: 717–755.
31. Buchan, J. R., and R. Parker. 2009. Eukaryotic stress granules: the ins and outs of
translation. Mol. Cell 36: 932–941.
32. White, J. P., and R. E. Lloyd. 2012. Regulation of stress granules in virus systems. Trends Microbiol. 20: 175–183.
33. Mok, B. W., W. Song, P. Wang, H. Tai, Y. Chen, M. Zheng, X. Wen, S. Y. Lau,
W. L. Wu, K. Matsumoto, et al. 2012. The NS1 protein of influenza A virus
interacts with cellular processing bodies and stress granules through RNAassociated protein 55 (RAP55) during virus infection. J. Virol. 86: 12695–12707.
34. Khaperskyy, D. A., T. F. Hatchette, and C. McCormick. 2012. Influenza A virus
inhibits cytoplasmic stress granule formation. FASEB J. 26: 1629–1639.
35. Okonski, K. M., and C. E. Samuel. 2013. Stress granule formation induced by
measles virus is protein kinase PKR dependent and impaired by RNA adenosine
deaminase ADAR1. J. Virol. 87: 756–766.
36. Onomoto, K., M. Jogi, J. S. Yoo, R. Narita, S. Morimoto, A. Takemura,
S. Sambhara, A. Kawaguchi, S. Osari, K. Nagata, et al. 2012. Critical role of an
antiviral stress granule containing RIG-I and PKR in viral detection and innate
immunity. PLoS ONE 7: e43031.
37. Hoffmann, E., G. Neumann, Y. Kawaoka, G. Hobom, and R. G. Webster. 2000. A
DNA transfection system for generation of influenza A virus from eight plasmids. Proc. Natl. Acad. Sci. USA 97: 6108–6113.
38. Min, J. Y., S. Li, G. C. Sen, and R. M. Krug. 2007. A site on the influenza A virus
NS1 protein mediates both inhibition of PKR activation and temporal regulation
of viral RNA synthesis. Virology 363: 236–243.
39. Matsumoto-Taniura, N., F. Pirollet, R. Monroe, L. Gerace, and J. M. Westendorf.
1996. Identification of novel M phase phosphoproteins by expression cloning.
Mol. Biol. Cell 7: 1455–1469.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
cytoplasm, it is tempting to postulate that, in normal conditions, NF90
migrates to the cytoplasm to facilitate PKR phosphorylation in response to virus infection or the presence of dsRNA during virus
replication. Activation of PKR leads to the phosphorylation of eIF2a
and subsequently triggers formation of stress granules that involve
both NF90 and PKR (Fig. 8). We also found that phosphorylation of
NF90 by PKR is required for NF90–PKR signaling to induce antiviral stress granules (Fig. 4). Based on a previous study (13) that
reported that NF90 is phosphorylated by PKR and retained in the
cytoplasm during the interaction, it seems reasonable to suggest that
NF90 and PKR are mutually regulated. It may be speculated that an
accumulation of NF90 in the cytoplasm may subsequently compete
with PKR for dsRNA binding and, thereby, form a feedback loop to
restore PKR back to its latent form; similarly, NF90 would be
recycled back to the nucleus after the stimuli dissipates. Further
studies are needed to unveil the regulatory loop of the NF90–PKR
pathway. Although we showed that IFN-b is not involved in NF90–
PKR activation, it would be interesting to study how NF90 and IFN
coordinately activate PKR in the antiviral state induced during virus
infection. Host innate immunity relies on pattern recognition
receptors, which reside primarily in the cytosol or cellular membranes, to recognize pathogen-associated molecule patterns that
are usually intermediate products generated from the viralreplication process. Because NF90 predominantly resides in the
nucleus, it would be important to examine whether it is a yet-tobe-defined nuclear pattern recognition receptor, like IFI16 (55).
11
12
40. Coolidge, C. J., and J. G. Patton. 2000. A new double-stranded RNA-binding
protein that interacts with PKR. Nucleic Acids Res. 28: 1407–1417.
41. Parrott, A. M., M. R. Walsh, T. W. Reichman, and M. B. Mathews. 2005. RNA
binding and phosphorylation determine the intracellular distribution of nuclear
factors 90 and 110. J. Mol. Biol. 348: 281–293.
42. McEwen, E., N. Kedersha, B. Song, D. Scheuner, N. Gilks, A. Han, J. J. Chen,
P. Anderson, and R. J. Kaufman. 2005. Heme-regulated inhibitor kinasemediated phosphorylation of eukaryotic translation initiation factor 2 inhibits
translation, induces stress granule formation, and mediates survival upon arsenite
exposure. J. Biol. Chem. 280: 16925–16933.
43. Lindquist, M. E., B. A. Mainou, T. S. Dermody, and J. E. Crowe, Jr. 2011.
Activation of protein kinase R is required for induction of stress granules by
respiratory syncytial virus but dispensable for viral replication. Virology 413:
103–110.
44. 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.
45. Wang, W., K. Riedel, P. Lynch, C. Y. Chien, G. T. Montelione, and R. M. Krug.
1999. RNA binding by the novel helical domain of the influenza virus NS1
protein requires its dimer structure and a small number of specific basic amino
acids. RNA 5: 195–205.
46. Min, J. Y., and R. M. Krug. 2006. The primary function of RNA binding by the
influenza A virus NS1 protein in infected cells: Inhibiting the 29-59 oligo (A)
synthetase/RNase L pathway. Proc. Natl. Acad. Sci. USA 103: 7100–7105.
NF90 REGULATES ACTIVATION OF PKR
47. Pfaller, C. K., Z. Li, C. X. George, and C. E. Samuel. 2011. Protein kinase PKR
and RNA adenosine deaminase ADAR1: new roles for old players as modulators
of the interferon response. Curr. Opin. Immunol. 23: 573–582.
48. Patel, R. C., and G. C. Sen. 1998. PACT, a protein activator of the interferoninduced protein kinase, PKR. EMBO J. 17: 4379–4390.
49. Li, S., G. A. Peters, K. Ding, X. Zhang, J. Qin, and G. C. Sen. 2006. Molecular
basis for PKR activation by PACT or dsRNA. Proc. Natl. Acad. Sci. USA 103:
10005–10010.
50. Cole, J. L. 2007. Activation of PKR: an open and shut case? Trends Biochem.
Sci. 32: 57–62.
51. Kumar, A., J. Haque, J. Lacoste, J. Hiscott, and B. R. Williams. 1994. Doublestranded RNA-dependent protein kinase activates transcription factor NF-kappa
B by phosphorylating I kappa B. Proc. Natl. Acad. Sci. USA 91: 6288–6292.
52. Beckham, C. J., and R. Parker. 2008. P bodies, stress granules, and viral life
cycles. Cell Host Microbe 3: 206–212.
53. White, J. P., A. M. Cardenas, W. E. Marissen, and R. E. Lloyd. 2007. Inhibition
of cytoplasmic mRNA stress granule formation by a viral proteinase. Cell Host
Microbe 2: 295–305.
54. Simpson-Holley, M., N. Kedersha, K. Dower, K. H. Rubins, P. Anderson,
L. E. Hensley, and J. H. Connor. 2011. Formation of antiviral cytoplasmic
granules during orthopoxvirus infection. J. Virol. 85: 1581–1593.
55. Veeranki, S., and D. Choubey. 2012. Interferon-inducible p200-family protein
IFI16, an innate immune sensor for cytosolic and nuclear double-stranded DNA:
regulation of subcellular localization. Mol. Immunol. 49: 567–571.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017