FEBS Letters 587 (2013) 1571–1578

journal homepage: www.FEBSLetters .org

Hepatitis C virus targets the interferon-a JAK/STAT pathwayby promoting proteasomal degradation in immune cells and hepatocytes

0014-5793/$36.00 � 2013 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.febslet.2013.03.041

⇑ Corresponding author. Fax: +353 1 6772400.E-mail address: n.stevenson@tcd.ie (N.J. Stevenson).

1 These authors contributed equally to this work.

Nigel J. Stevenson a,⇑,1, Nollaig M. Bourke a,1, Elizabeth J. Ryan a, Marco Binder b, Liam Fanning c,James A. Johnston d, John E. Hegarty e, Aideen Long f,g, Cliona O’Farrelly a,g

a School of Biochemistry and Immunology, Trinity Biomedical Sciences Institute, Trinity College Dublin (TCD), Dublin, Irelandb Department of Infectious Diseases, Heidelberg University, Germanyc Molecular Virology Diagnostic & Research Laboratory, University College Cork, Cork, Irelandd Centre for Infection and Immunity, School of Medicine, Dentistry and Biomedical Sciences, Queen’s University Belfast, UKe National Liver Transplant Unit, St. Vincent’s University Hospital, Dublin, Irelandf Institute of Molecular Medicine, TCD, Irelandg School of Medicine, TCD, Ireland

a r t i c l e i n f o

Article history:Received 6 February 2013Revised 22 March 2013Accepted 28 March 2013Available online 12 April 2013

Edited by Hans-Dieter Klenk

Keywords:Hepatitis C VirusInterferonJAK/STATUbiquitinationProteasome

a b s t r a c t

JAK/STAT signalling is essential for anti-viral immunity, making IFN-a an obvious anti-viral thera-peutic. However, many HCV+ patients fail treatment, indicating that the virus blocks successfulIFN-a signalling. We found that STAT1 and STAT3 proteins, key components of the IFN-a signallingpathway were reduced in immune cells and hepatocytes from HCV infected patients, and upon HCVexpression in Huh7 hepatocytes. However, STAT1 and STAT3 mRNA levels were normal. Mechanisticanalysis revealed that in the presence of HCV, STAT3 protein was preferentially ubiquitinated, anddegradation was blocked by the proteasomal inhibitor MG132. These findings show that HCV inhib-its IFN-a responses in a broad spectrum of cells via proteasomal degradation of JAK/STAT pathwaycomponents.� 2013 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

1. Introduction

The anti-viral properties of interferon (IFN)-a have made it anobvious therapeutic agent to combat HCV infection, a major causeof chronic liver disease, hepatocellular carcinoma and liver trans-plantation [1]. Unfortunately, IFN-a therapy has a high failure rate,particularly in patients infected with genotype 1, with as few as30% responding to treatment, suggesting that HCV might targetthis pathway [2]. Several molecular mechanisms by which HCVevades the immune response have already been described. HCVblocks adaptive immunity, by stimulating specific T regulatorycells [3], as well as inhibiting normal T cell migration and IL-2secretion [4,5]. More recently, attention has turned to the innateimmune response and HCV has been shown to target NK cell activ-ity, dendritic cell function [6,7] and pathogen recognition path-ways, including inhibition of Toll-Like Receptor (TLR)-3 andretinoic acid-inducible gene (RIG)-I signalling [8]. However, the

reasons for non-response to endogenous or therapeutic IFN-a re-main poorly defined.

Anti-viral signalling commences upon IFN-a engagement withits receptor, thus activating tyrosine kinases, which phosphorylatereceptor docking sites for cytoplasmic signal transducer and acti-vator of transcription (STAT) proteins. STATs become phosphory-lated, form dimers and translocate to the nucleus, where theybind specific DNA-recognition motifs, promoting gene transcrip-tion. Disruption of STAT genes in mice has confirmed that theseproteins are essential in several biological processes, including cellgrowth and differentiation, as well as immune defence [9,10]. IFN-a signalling via STAT proteins leads to the upregulation of over 500IFN Sensitive Genes (ISGs), including key anti-viral and pro-inflam-matory mediators [11–14].

Signal transduction is often regulated by the homeostatic pro-cess of protein turnover via degradation of poly-ubiquitinated pro-teins through the 26S proteasome. Suppressor of cytokinesignalling (SOCS) proteins associate with elongin B, elongin C, cul-lin (cul) and the RING finger-containing protein, to form functionalElongin B/C-cul-SOCS box (ECS)-like E3 ligase complexes, that cata-lyse ubiquitin transfer to target proteins [15]. Viral proteins regu-larly act as E3 ligase proteins; in fact, we and others have shown

Table 1Primer sequences used for quantitative and qualitative RT-PCR.

qRT-PCR primersRPS15 F CGGACCAAAGCGATCTCTTCRPS15 R CGCACTGTACAGCTGCATCASTAT3 F GAGAAGGACATCAGCGGTAAGACSTAT3 R GCTCTCTGGCCGACAATACTTTSTAT1 F TGATGGCCCTAAAGGAACTGSTAT1 R AGGAAAACTGTCGCCAGAGART-PCR primersB-actin F GGACTTCGAGCAAGAGATGGB-actin R AGCACTGTGTTGGCGTACAGSTAT1 F CCGTTTTCATGACCTCCTGTSTAT1 R TGAATATTCCCCGACTGAGCSTAT3 F TTTCACTTGGGTGGAGAAGGSTAT3 R GCTACCTGGGTCAGCTTCAG

1572 N.J. Stevenson et al. / FEBS Letters 587 (2013) 1571–1578

that Paramyxoviruses, such as Respiratory Syncytial Virus (RSV),target STAT proteins for degradation via the proteosome [16–18].

Since many HCV-infected patients do not respond to IFN-a ther-apy we investigated the status of the JAK/STAT pathway in bothprimary immune cells and hepatocytes from HCV infected patients.We found that STAT1 and STAT3 proteins were reduced in all majorimmune cell populations and hepatocytes from HCV+ patients.Furthermore, we discovered that STAT3 was preferentially ubiqui-tinated and targeted for proteosomal degradation in the presenceof HCV. These findings demonstrate for the first time that essentialanti-viral components of the IFN-a pathway are reduced in bothimmune cells and hepatocytes, revealing a novel mechanism bywhich HCV blocks responsiveness to IFN-a therapy.

2. Patients

Blood from 17 patients (15 female, 2 male, ranging between 32and 71 years of age) who were PCR positive for HCV (Supplemen-tary Table 1) and had not undergone IFN-a/Ribavrin standardtreatment of care, were studied. All were informed and gave writ-ten consent. The study received ethical approval from the Researchand Ethics Committee at SVUH, in accordance with the guidelinesof the 1975 Declaration of Helsinki. Patients were tested for HCVantibodies using enzyme immunoassays (Abbott) and immunoblotassays (Chiron Corp., USA).

3. Materials and methods

3.1. PBMC isolation

PBMCs were isolated from 30 ml of blood by centrifugation(650g for 30 min) over Ficoll (Amersham, UK) and washed X2(290g for 10 min) with RPMI.

3.2. HCV RNA detection in PBMCs

PBMCs (resuspended in stabilising solution RNAlater (Ambion,UK)) were lysed according to the Magna Pure external lysis proto-col (Roche, UK) and cellular RNA from 1.5 � 106 cells was ex-tracted; real time qRT-PCR performed according to the COBASTaqman protocol HCV test v2.0 (Roche, UK). HCV titre recordedas international units (IU).

3.3. Cell culture

Huh7 cells were grown in DMEM supplemented with 6 lg/mlzeocin. PBMCs were cultured in RPMI containing 2% FCS, 250U/ml penicillin and 250 lg/ml streptomycin. Cells were treated with1000 U/ml human IFN-a (Roche, Switzerland) and 10 lM MG132(Calbiochem, USA).

3.4. Transfection

Huh7 cells expressing T7 polymerase were transfected for 16 husing Lipofectamine 2000 (Invitrogen, USA), with 0.75 lg (6 wellplates) or 5 lg (9 cm diameter plates) of pBRTM/HCV1-3011 DNAconstruct (a kind gift from Prof. Charles Rice, The Rockefeller Univer-sity) or EV (Invitrogen, USA) or with 0.1 lg of DNA in poly-D-lysineslides (Becton Dickson, USA). Huh7 cells were also transfected with5 lg HA-tagged-Ubiquitin or EV for 48 h using Lipofectamine 2000.

3.5. Immunoprecipitation

Huh7 cells, transfected with EV/HCV (5 lg) �/+MG132 treat-ment, or transfected with HA-Ubiquitin and HCV, were harvestedin mild lysis buffer (50 mM Hepes, 100 mM NaCl, 1 mM EDTA,

10% glycerol, 0.5% NP40) or RIPA buffer, substituted with aprotinin[5 lg/ml], leupeptin [5 lg/ml], PMSF [1 mM], Na3VO4 [1 mM]). Ly-sates were immunoprecipitated with protein-A agarose beads andSTAT3 antibody, before immunoblotting for HA (Sigma, USA), ubiq-uitin or STAT3 (Santa Cruz Biotechnologies, USA).

3.6. Immunoblotting

Cells were harvested in lysis buffer and analysed by immuno-blotting using pSTAT1, pSTAT3, STAT1 (New England Biolabs,USA), STAT3, ubiquitin (Santa Cruz Biotechnologies, USA), HCV-NS2, b-actin, c-tubulin, HA (Sigma, USA), HCV-E2 (Acris, USA),HCV-Core (Abcam, UK) and secondary antibodies labelled withinfrared dyes and the infrared Odyssey imager. Densitometry wascalculated using LI-COR software. (LI-COR, Germany).

3.7. RT-PCR

RNA was isolated from cells following the Trizol manufacturer’sprotocol (Invitrogen, USA). RT-PCR was performed using the One-step protocol (Qiagen, Germany). Cycling parameters: 30 min at50 �C, 95 �C for 15 min and 35 cycles of 94 �C, 58 �C and 72 �C for30 s, 72 �C for 10 min. qRT-PCR: as described before [19]. Humangene amplifications normalised to Ribosomal protein 15 (RPS15);Primers listed in Table 1.

3.8. HCV infection

Virus stock production as described previously [20]. In vitrotranscribed Jc1 RNA was electro-transfected into Huh7.5 cellsand supernatants collected 24, 48 and 72 h post transfection,pooled and infectious titer determined by TCID50. Huh7.5 cellswere inoculated with infectious supernatants (MOI = 3) for 24 h,washed with DMEM and treated with MG132 or DMSO. After 2 h,cells were harvested and lysed in 50 ll RIPA.

3.9. Immunohistochemistry

Liver sections (8 lm) were cut from paraffin-embedded, forma-lin-fixed samples and microwaved in a pressure cooker in sodiumcitrate buffer (27 ml 0.1 M citric acid, 123 ml 0.1 M tri sodiumcitrate in 1500 ml dH2O, pH 6). Sections were stained usingSTAT3(Santa Cruz Biotechnologies, USA) and STAT1 (New EnglandBiolabs) antibodies (1:80) for 1 h in antibody diluent (DAKO, Den-mark) and detected using a polymer detection kit (Leica Microsys-tems, Germany) and analysed by microscopy.

3.10. Immunofluorescence

As described previously [19], antibodies included anti-STAT3(Santa Cruz Biotechnologies, USA) labelled with Alexa 568 or 546

Fig. 1. STAT1 and STAT3 protein levels are reduced in PBMCs from HCV+ patients. (A) Lysates from PBMCs of individual HCV+ or healthy subjects, treated �/+ IFN-a for15 min, were analysed by immunoblotting for total or phosphorylated STAT1, STAT3 and b-actin (n = 3). STAT1 and STAT3 mRNA levels in PBMCs from HCV+ and healthyindividuals were measured by (B) qualitative RT-PCR (HCV n = 6; healthy n = 4) and (C) quantitative qRT-PCR (healthy n = 7; HCV n = 12). (D) Ratio of STAT3: house-keepingprotein optical density (OD) measured by densitometric analysis of immunoblots of STAT3 and b-actin or c-tubulin protein levels in PBMCs from HCV+ (n = 17) and healthy(n = 8) individuals. (E) Immunoblot of HCV-NS2 and b-actin from healthy (H) (n = 3) and HCV+ patient PBMC lysates (n = 3). (F) qRT-PCR detection of HCV RNA in PBMCs from5 HCV RNA+ individuals (patient numbers 1–5 inclusive), 2 HCV RNA negative patients (patient 6 spontaneously cleared HCV and patient 7 cleared HCV during IFN-a/ribavirin therapy) and 2 healthy controls (patient numbers 8 and 9).

N.J. Stevenson et al. / FEBS Letters 587 (2013) 1571–1578 1573

Fig. 2. STAT1 and STAT3 proteins are reduced in major immune cell lineages of HCV+ patients. PBMCs from healthy (n = 3) or HCV+ individuals (n = 7) were labelled withCD45, CD3, CD19, CD56, CD14, CD16 and HLA-DR antibodies. Cells were fixed, permeabilised and stained with (A) STAT1 and (B) STAT3 antibodies. Representative dot plots ofa healthy and a HCV+ individual shown. (C) Confocal micrographs of PBMCs from HCV+ individuals labelled with HCV-E2 along with FITC labelled CD19, HLA-DR and CD14 orCD56 and CD3 primary antibodies (n = 2). Nucleus labelled with DAPI. Bar, 10 lm.

1574 N.J. Stevenson et al. / FEBS Letters 587 (2013) 1571–1578

(Invitrogen, USA); HCV-Core (Abcam, USA) and FITC secondary(Sigma, USA) were used. DNA was stained using Prolong goldanti-fade reagent containing DAPI (Invitrogen, CA, USA). HCV-E2(Acris, USA) and Alexa 546 or 488 secondary; CD56, CD3 and Alexa647 secondary and FITC-conjugated antibodies to: CD19, CD14, andHLA-DR (BD biosciences, USA) were used. Microscopy was per-formed on an Olympus FV1000 laser scanning confocal microscope,using an UPlanSAPO 60X/1.35 NA oil objective at room tempera-ture. Cells were digitally sectioned by confocal laser scanning fluo-rescence microscopy at 0.3 lm per slice and analyzed using theOlympus FV1000 software.

3.11. Flow cytometry

PBMCs were surface labelled with FITC-conjugated antibodiesto: CD45, CD19, CD14, CD16 and HLA-DR; PerCP-conjugated anti-CD3 or Alexa-Fluor 488-conjugated anti-CD56 (BD Biosciences,USA). Cells were fixed with 4% paraformaldehyde for 30 min onice, washed X2 in PBS and resuspended in permeabilisation buffer(eBioscience). Cells were incubated with STAT1-PE and STAT3-APC(BD Biosciences, USA) for 30 min, washed in PBS and acquiredusing a Dako Cyan flow cytometer. Data was analysed using Flowjosoftware (Tree Star Inc, USA).

Fig. 3. Loss of STAT1 and STAT3 protein in hepatocytes and liver infiltratingleukocytes. (A) Immunoblot of STAT1, STAT3, HCV-NS2 and b-actin protein in Huh7cells transfected �/+ HCV DNA construct or EV control. (B) qRT-PCR of STAT1 andSTAT3 mRNA in Huh7 hepatocytes transfected with EV or HCV constructs (n = 3).(C) Immunohistochemistry of STAT1 and STAT3 in liver tissue from HCV+(representative of 3 individuals) and healthy individuals (representative of 3individuals). Bar, 50 lm.

N.J. Stevenson et al. / FEBS Letters 587 (2013) 1571–1578 1575

3.12. Statistical analysis

Statistical analysis was performed using GraphPad Prism ver-sion 5.01 for Windows (GraphPad Software, La Jolla, CA, USA).Groups were compared by paired Student’s T-test for parametricsamples with GAUSSIAN distribution, non-parametric Mann-Whitneyanalysis for patient samples and ANOVA with Bonferroni’s multiplecomparison post-test analysis as appropriate. ⁄P < 0.05, ⁄⁄P < 0.01,⁄⁄⁄P < 0.001.

4. Results

4.1. STAT1 and STAT3 proteins are reduced in PBMCs from HCV+individuals

To investigate the status of the JAK/STAT pathway in HCV pa-tients, we initially analysed STAT1 and STAT3 tyrosine phosphory-lation and protein levels in human PBMCs from HCV positive (+)and healthy individuals. We found that protein levels were lowerin PBMCs from HCV+ patients compared to healthy controls(Fig. 1A), and IFN-a-induced tyrosine phosphorylation of bothSTAT1 and STAT3 proteins were lower, although mRNA levels didnot differ significantly (Fig. 1B and C). Since this was the first dem-onstration of STAT3 protein reduction in HCV infected cells, proteinlevels were investigated in 16 additional treatment-naïve HCV+(genotype 1) patients and 8 healthy controls. Densitometric analy-sis revealed that STAT3 protein levels were significantly lower inPBMCs from HCV+ patients compared to healthy controls

(P < 0.001) Fig. 1D). Interestingly, we found HCV-NS2 protein byimmunoblotting (Fig. 1E) and HCV RNA by qRT-PCR (Fig. 1F) withinPBMCs from HCV+ patients, indicating that intracellular HCV maydirectly act to promote loss of STAT proteins.

4.2. STAT1 and STAT3 proteins are reduced in all major immune cellpopulations

To determine if STAT1 and STAT3 in PBMCs were depleted fromall major immune cell types, we analysed protein levels withinspecific immune cell populations using flow cytometry. Normallevels of STAT1 (Fig. 2A) and STAT3 (Fig. 2B) were observed in allthe major peripheral blood cell subsets from healthy controls,but were significantly reduced in all cell subsets from HCV+ pa-tients, including monocytes (CD14+), T cells (CD3+), B cells(CD19+), natural killer (NK) cells (CD56+), and HLA-DR+ antigenpresenting cells (APCs), demonstrating that loss of STAT proteinexpression was not limited to any particular cell type (Fig. 2Aand B). Furthermore, HCV was detected by confocal microscopyin all major immune cells types, demonstrating that its potentialto directly target intracellular STAT expression (Fig. 2C).

4.3. HCV-mediated loss of STAT1 and STAT3 in hepatocytes and liverinfiltrating leukocytes

To determine the effect of HCV upon STAT1 and STAT3 levels inliver cells, we next analysed their expression in Huh7 hepatocytestransfected with a HCV DNA construct. We found considerably re-duced levels of STAT1 and STAT3 protein (Fig. 3A) in HCV transfec-ted hepatocytes and as in PBMCs, mRNA levels were notsignificantly different (Fig. 3B). We further analysed STAT1 andSTAT3 in primary human liver tissue by immunohistochemistryand found STAT1 and STAT3 levels greatly reduced in hepatocytesand infiltrating leukocytes of HCV+ livers, compared to healthycontrols (Fig. 3C). Loss of STAT1 and STAT3 proteins in PBMCs,Huh7 hepatocytes, primary hepatocytes and liver infiltrating leu-kocytes demonstrates a conserved reduction across a spectrum ofcell types and reveals STAT3, for the first time, as a target for HCV’simmune evasion mechanisms.

4.4. HCV promotes STAT3 degradation via the proteasome

We and others have shown that the ubiquitin/proteasome sys-tem is used by viruses to degrade STAT proteins [17,18] and sinceSTAT1 had already been shown to be proteasomally degraded byHCV [21], we investigated the molecular mechanism responsiblefor the absence of STAT3 by blocking proteasomal activity withthe inhibitor, MG132. Using immunoblotting and confocal micros-copy, we found that absent STAT3 protein could be recovered inPBMCs from HCV+ patients following proteasomal inhibition(Fig. 4A and B). To determine if the loss of STAT3 in HCV transfec-ted hepatocytes was also proteasome-dependent, we transfectedHuh7 hepatocytes with a HCV construct and treated withMG132. We observed that HCV-mediated degradation of STAT3was blocked by proteasomal inhibition (Fig. 4C and D). WhileHCV infection with cell cultured HCV (HCVcc) had no effect onSTAT3 mRNA expression (Fig. 4E), it promoted STAT3 protein deg-radation, which was inhibited by MG132, further confirming aHCV-controlled mechanism of STAT3 degradation via the protea-some (Fig. 4F).

4.5. HCV enhances ubiquitination of STAT3

Since ubiquitinated proteins are targeted for proteasomal deg-radation, we analysed protein ubiquitination in HCV expressingHuh7s. We found that endogenous total protein ubiquitination in

Fig. 4. HCV promotes STAT3 ubiquitination and proteasomal degradation. (A) STAT3 immunoblot and (B) confocal micrograph of PBMCs from HCV+ and healthy individualstreated �/+ MG132 for 4 h (n = 3). STAT3 and the nucleus labelled with FITC and DAPI, respectively. Bar, 10 lm. (C) Immunoblot of STAT3, c-tubulin and HCV-NS2 proteinexpression, �/+ HCV–DNA transfection of Huh7 hepatocytes, �/+ MG132 (n = 3). (D) Confocal micrograph of STAT3, HCV-Core and the nucleus (N), labelled with FITC, Alexa568 and DAPI, respectively, �/+ HCV construct expression, alongside MG132 treatment (n = 3). Bar, 10 lm. (E) STAT3 mRNA from Huh7 cells mock or HCVcc infected (n = 6).(F) Immunoblot of STAT3, HCV-Core and c-tubulin proteins from Huh7 cells mock or HCVcc infected and treated �/+ MG132 for 4 h (n = 4). (G) Immunoblot of Huh7 cellsmock or HCVcc infected and treated �/+ MG132 for 4 h, whole cell lysates (WCLs) probed with ubiquitin and c-tubulin antibodies (n = 3). Values below the blot representdensitometric analysis. Ubiquitinated protein band intensity was normalised to c-tubulin band intensity and all values are presented relative to uninfected control cells (lane1), which was normalised to 1. (H) Huh7 cells mock or HCVcc infected treated with MG132 for 4 h were lysed and immunoprecipitated for STAT3 and immunoblotted forubiquitin. (I) Huh7 cells transfected �/+ HCV-DNA, Ubiquitin-HA or EV constructs and treated with MG132 for 4 h were lysed, immunoprecipitated for STAT3 andimmunoblotted for HA and STAT3. WCLs probed for HCV-E2 and b-actin (n = 3).

1576 N.J. Stevenson et al. / FEBS Letters 587 (2013) 1571–1578

N.J. Stevenson et al. / FEBS Letters 587 (2013) 1571–1578 1577

HCVcc infected cells was enhanced, compared to mock infectedcontrols and treatment of HCV infected cells with MG132 furtherincreased protein ubiquitination (Fig. 4G). To determine if STAT3ubiquitination was specifically enhanced by HCV, lysates fromHuh7 cells infected with HCVcc infected hepatocytes, �/+MG132, were immunoprecipitated for STAT3 and immunoblottedfor ubiquitin. While proteasomal inhibition with MG132 increasedlevels of ubiquitinated STAT3 (Fig. 4H, lane 4), HCV infection fur-ther enhanced levels of ubiquitinated STAT3, as seen by the pres-ence of multiple bands increasing in molecular weight,characteristic of polyubiquitination (Fig. 4H, lane 3). To furtheranalyse the specific ubiquitination of STAT3, we over-expressedhemagglutinin (HA)-tagged ubiquitin, �/+ transfection with HCVDNA construct, in Huh7 cells, before immunoprecipitating forSTAT3 and immunoblotting for ubiquitin. Once again, STAT3 ubiq-uitination was enhanced in HCV expressing cells, compared to con-trols (Fig. 4I). These results indicate that HCV promotes the directubiquitination of STAT3, revealing the novel mechanism by whichHCV promotes STAT3 protein degradation via the proteosome.

5. Discussion

This study describes the loss of STAT1 and STAT3 from HCV-in-fected immune cells and hepatocytes, and the mechanism bywhich HCV promotes STAT3 degradation through the proteosome,a process also used by other viruses, such as RSV, to evade immu-nity [17].

Since HCV is hepatotrophic, previous studies have focused onsignalling pathways within hepatocytes [22]. However, becausecirculating and liver infiltrating immune cells play such a vital rolein fighting HCV infection [23] and HCV has also been identified inimmune cells [24], we investigated the IFN-a JAK/STAT pathway inthe whole PBMC population and the major individual immune cellsubgroups. In fact, we found that STAT1 and STAT3 proteins wereabsent from circulating and liver infiltrating immune cells of HCVpatients, demonstrating widespread disruption of the anti-viralIFN-a pathway, which identifies the broad immune evasion strat-egy of HCV and widens our understanding of this virus’s spectrumof target cells.

Although STAT3 is known to be activated by IFN-a, its role inthis pathway remains poorly understood. Several studies, usingtransformed or primary hepatocytes, suggest that HCV either acti-vates STAT3 [25–29], has no impact [30–32] or blocks DNA binding[33]. However, this study indicates that HCV decreases STAT3 pro-tein in PBMCs, liver infiltrating leukocytes and hepatocytes frominfected patients, as well as Huh7 hepatocytes transfected or in-fected with HCV. These data support the results of studies showingdecreased STAT3 mRNA and protein in the HCV replicon system[34,35]. However, unlike these previous reports, we discoveredthat HCV does not target STAT3 mRNA. Rather, we found thatSTAT3 mRNA in Huh7 cells was elevated upon HCV expression,suggesting a homeostatic mechanism that attempts to replacethe STAT3 protein lost during HCV-mediated proteosomaldegradation.

STAT3 is well known to promote cell proliferation and thus tu-mour progression [36]; properties that are demonstrated in the li-ver, with increases in hepatic STAT3 activation linked to liverregeneration, fibrosis and hepatocellular carcinoma [25,37]. How-ever, even though STAT1 is well documented as an anti-viral medi-ator [38], as evidenced by STAT1 KO mice lacking IFN-aresponsiveness and exhibiting increased susceptibility to viralinfection [9,39], STAT3’s role in anti-viral signalling is disputed.Interestingly STAT3 was ruled out as an anti-viral mediator be-cause STAT3 homo-dimers remain inducible in STAT1�/� micethat succumb to viral infections. However, STAT1:STAT1 homo-

dimers and STAT1:STAT3 dimers also could not form in STAT1�/�mice [9], indicating a possible role for this hetero-dimer complexin anti-viral signalling. The functions of these STAT1:STAT3 hetero-dimers are not known, however the loss of STAT1 and STAT3 inHCV-infected cells may further indicate their important role duringanti-viral signalling.

In conclusion, we have shown that both STAT1 and STAT3 arereduced in primary immune cells and hepatocytes of HCV+ pa-tients and upon in vitro HCV expression or infection of hepato-cytes. We have also provided evidence of HCV-mediatedubiquitination and proteasomal degradation of STAT3, which out-lines a basic viral immune evasion mechanism used by HCV toblock the IFN-a JAK/STAT pathway in immune cells and hepato-cytes. This broad HCV-mediated immune evasion strategy may ex-plain poor responsiveness to therapeutic IFN-a and highlights apotenital target pathway for therapeutic restoration of anti-viralimmune responses.

Acknowledgements

We would like to thank Prof. Charles Rice (The Rockefeller Uni-versity) for the HCV construct and NS2 antibody; Prof. Ralf Bart-enschlarger (Heidelberg University) for Huh7 cells; Dr. OrlaHanrahan (Trinity College Dublin) for confocal microscopy assis-tance; Catherine Keogh and Dr. Aideen Collins for technical assis-tance and all patients and healthy volunteers who donated bloodand liver for this study and the nurses who drew blood and Dr.Niamh Nolan for liver analysis advice. This study was supportedby the Irish Health Research Board and Science Foundation Ireland.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.febslet.2013.03.041.

References

[1] Armstrong, G., Wasley, A., Simard, E., McQuillan, G., Kuhnert, W. and Alter, M.(2006) The prevalence of hepatitis C virus infection in the United States, 1999through 2002. Ann. Intern. Med. 144, 705–714.

[2] Health Protection Surveillance Centre, D., Ireland. (2007) National Hepatitis Cdatabase for infection acquired through blood and blood products.ed.^eds).

[3] Rowan, A.G., Fletcher, J.M., Ryan, E.J., Moran, B., Hegarty, J.E., O’Farrelly, C. andMills, K.H. (2008) Hepatitis C virus-specific Th17 cells are suppressed by virus-induced TGF-beta. J. Immunol. 181, 4485–4494.

[4] Petrovic, D. et al. (2011) Hepatitis C virus targets the T cell secretorymachinery as a mechanism of immune evasion. Hepatology 53, 1846–1853.

[5] Klenerman, P. and Thimme, R. (2012) T cell responses in hepatitis C: the good,the bad and the unconventional. Gut 61, 1226–1234.

[6] Ryan, E.J. and O’Farrelly, C. (2011) The affect of chronic hepatitis C infection ondendritic cell function: a summary of the experimental evidence. J. ViralHepat. 18, 601–607.

[7] Ryan, E.J., Stevenson, N.J., Hegarty, J.E. and O’Farrelly, C. (2011) Chronichepatitis C infection blocks the ability of dendritic cells to secrete IFN-a andstimulate T-cell proliferation. J. Viral Hepat. 18, 840–851.

[8] Horner, S. and Gale, M.J. (2009) Intracellular innate immune cascades andinterferon defenses that control hepatitis C virus. J. Interferon Cytokine Res.29, 489–498.

[9] Durbin, J., Hackenmiller, R., Simon, M. and Levy, D. (1996) Targeted disruptionof the mouse Stat1 gene results in compromised innate immunity to viraldisease. Cell 84, 443–450.

[10] Park, C., Li, S., Cha, E. and Schindler, C. (2000) Immune response in Stat2knockout mice. Immunity 13, 795–804.

[11] Der, S., Zhou, A., Williams, B. and Silverman, R. (1998) Identification of genesdifferentially regulated by interferon alpha, beta, or gamma using oligonucleotidearrays. Proc. Natl. Acad. Sci. USA 95, 15623–15628.

[12] Takeda, K., Noguchi, K., Shi, W., Tanaka, T., Matsumoto, M., Yoshida, N.,Kishimoto, T. and Akira, S. (1997) Targeted disruption of the mouse Stat3 geneleads to early embryonic lethality. Proc. Natl. Acad. Sci. USA 94, 3801–3804.

[13] Takeda, K., Kaisho, T., Yoshida, N., Takeda, J., Kishimoto, T. and Akira, S. (1998)Stat3 activation is responsible for IL-6-dependent T cell proliferation throughpreventing apoptosis: generation and characterization of T cell-specific Stat3-deficient mice. J. Immunol. 161, 4652–4660.

1578 N.J. Stevenson et al. / FEBS Letters 587 (2013) 1571–1578

[14] Radaeva, S. et al. (2002) Interferon-alpha activates multiple STAT signals anddown-regulates c-Met in primary human hepatocytes. Gastroenterology 122,1020–1034.

[15] Hershko, A. and Ciechanover, A. (1998) The ubiquitin system. Annu. Rev.Biochem. 67, 425–479.

[16] Banks, L., Pim, D. and Thomas, M. (2003) Viruses and the 26S proteasome:hacking into destruction. Trends Biochem. Sci. 28, 452–459.

[17] Elliott, J. et al. (2007) Respiratory syncytial virus NS1 protein degrades STAT2by using the Elongin-Cullin E3 ligase. J. Virol. 81, 3428–3436.

[18] Horvath, C. (2004) Weapons of STAT destruction. Interferon evasion byparamyxovirus V protein. Eur. J. Biochem. 271, 4621–4628.

[19] Stevenson, N.J., Murphy, A.G., Bourke, N.M., Keogh, C.A., Hegarty, J.E. and O’Farrelly,C. (2011) Ribavirin enhances IFN-a signalling and MxA expression: A novelimmune modulation mechanism during treatment of HCV. PLoS ONE 6, e27866.

[20] Pietschmann, T. et al. (2006) Construction and characterization of infectiousintragenotypic and intergenotypic hepatitis C virus chimeras. Proc. Natl. Acad.Sci. USA 103, 7408–7413.

[21] Lin, W., Choe, W., Hiasa, Y., Kamegaya, Y., Blackard, J., Schmidt, E. and Chung,R. (2005) Hepatitis C virus expression suppresses interferon signaling bydegrading STAT1. Gastroenterology 128, 1034–1041.

[22] Boonstra, A., van der Laan, L., Vanwolleghem, T. and Janssen, H. (2009)Experimental models for hepatitis C viral infection. Hepatology 50, 1646–1655.

[23] Rehermann, B. (2009) Hepatitis C virus versus innate and adaptive immuneresponses: a tale of coevolution and coexistence. J. Clin. Invest. 119, 1745–1754.

[24] Stamataki, Z. (2010) Hepatitis C infection of B lymphocytes: more tools toaddress pending questions. Expert Rev. Anti Infect Ther. 8, 977–980.

[25] Gao, B. (2005) Cytokines, STATs and liver disease. Cell. Mol. Immunol. 2, 92–100.

[26] Kawamura, H., Govindarajan, S., Aswad, F., Machida, K., Lai, M., Sung, V. andDennert, G. (2006) HCV core expression in hepatocytes protects againstautoimmune liver injury and promotes liver regeneration in mice. Hepatology44, 936–944.

[27] Yoshida, T., Hanada, T., Tokuhisa, T., Kosai, K., Sata, M., Kohara, M. andYoshimura, A. (2002) Activation of STAT3 by the hepatitis C virus core proteinleads to cellular transformation. J. Exp. Med. 196, 641–653.

[28] Sarcar, B., Ghosh, A., Steele, R., Ray, R. and Ray, R. (2004) Hepatitis C virus NS5Amediated STAT3 activation requires co-operation of Jak1 kinase. Virology 322,51–60.

[29] Zhu, H. et al. (2003) Gene expression associated with interferon alfa antiviralactivity in an HCV replicon cell line. Hepatology 37, 1180–1188.

[30] Blindenbacher, A. et al. (2003) Expression of hepatitis c virus proteins inhibitsinterferon alpha signaling in the liver of transgenic mice. Gastroenterology124, 1465–1475.

[31] Hazari, S., Taylor, L., Haque, S., Garry, R., Florman, S., Luftig, R., Regenstein, F.and Dash, S. (2007) Reduced expression of Jak-1 and Tyk-2 proteins leads tointerferon resistance in hepatitis C virus replicon. Virol. J. 4, 89.

[32] Heim, M., Moradpour, D. and Blum, H. (1999) Expression of hepatitis C virusproteins inhibits signal transduction through the Jak-STAT pathway. J. Virol.73, 8469–8475.

[33] Stärkel, P., De Saeger, C., Leclercq, I., Strain, A. and Horsmans, Y. (2005)Deficient Stat3 DNA-binding is associated with high Pias3 expression and apositive anti-apoptotic balance in human end-stage alcoholic and hepatitis Ccirrhosis. J. Hepatol. 43, 687–695.

[34] Larrea, E., Aldabe, R., Molano, E., Fernandez-Rodriguez, C., Ametzazurra, A.,Civeira, M. and Prieto, J. (2006) Altered expression and activation of signaltransducers and activators of transcription (STATs) in hepatitis C virusinfection: in vivo and in vitro studies. Gut 55, 1188–1196.

[35] Luquin, E., Larrea, E., Civeira, M., Prieto, J. and Aldabe, R. (2007) HCV structuralproteins interfere with interferon-alpha Jak/STAT signalling pathway.Antiviral Res. 76, 194–197.

[36] Kortylewski, M. and Yu, H. (2008) Role of Stat3 in suppressing anti-tumorimmunity. Curr. Opin. Immunol. 20, 228–233.

[37] Stärkel, P., Saeger, C., Leclercq, I. and Horsmans, Y. (2007) Role of signaltransducer and activator of transcription 3 in liver fibrosis progression inchronic hepatitis C-infected patients. Lab. Invest. 87, 173–181.

[38] Horvath, C. and Darnell, J.J. (1996) The antiviral state induced by alphainterferon and gamma interferon requires transcriptionally active Stat1protein. J. Virol. 70, 647–650.

[39] Meraz, M. et al. (1996) Targeted disruption of the Stat1 gene in mice revealsunexpected physiologic specificity in the JAK-STAT signaling pathway. Cell 84,431–442.