Cellular & Molecular Immunology (2016) 13, 94–102 ß 2015 CSI and USTC. All rights reserved 1672-7681/15 $32.00 www.nature.com/cmi

RESEARCH ARTICLE

Interferon alpha (IFNa)-induced TRIM22 interrupts HCV replication by ubiquitinating NS5A

Chen Yang1,6, Xinhao Zhao2,6, Dakang Sun5,6, Leilei Yang2, Chang Chong2, Yu Pan3, Xiumei Chi3, Yanhang Gao3, Moli Wang4, Xiaodong Shi3, Haibo Sun3, Juan Lv3, Yuanda Gao3, Jin Zhong2, Junqi Niu3 and Bing Sun1,2

TRIM22, a tripartite-motif (TRIM) protein, is upregulated upon interferon alpha (IFNa) administration to hepatitis C virus (HCV)-infected patients. However, the physiological role of TRIM22 upregulation remains unclear. Here, we describe a potential antiviral function of TRIM22’s targeting of the HCV NS5A protein. NS5A is important for HCV replication and for resistance to IFNa therapy. During the first 24 h following the initiation of IFNa treatment, upregulation of TRIM22 in the peripheral blood mononuclear cells (PBMCs) of HCV patients correlated with a decrease in viral titer. This phenomenon was confirmed in the hepatocyte-derived cell line Huh-7, which is highly permissive for HCV infection. TRIM22 over-expression inhibited HCV replication, and Small interfering RNA (siRNA)-mediated knockdown of TRIM22 diminished IFNa-induced anti-HCV function. Furthermore, we determined that TRIM22 ubiquitinates NS5A in a concentration-dependent manner. In summary, our results suggest that TRIM22 upregulation is associated with HCV decline during IFNa treatment and plays an important role in controlling HCV replication in vitro. Cellular & Molecular Immunology (2016) 13, 94–102; doi:10.1038/cmi.2014.131; published online 16 February 2015

Keywords: HCV; IFNa; NS5A; TRIM22; ubiquitin

INTRODUCTION Interferon alpha (IFNa) plus ribavirin is the standard therapy Hepatitis C virus (HCV) is a global problem, and according to for HCV patients in China.5 IFNa is a broad-spectrum anti-viral the World Health Organization, 150 million people are cur- molecule and plays an important role in the anti-viral innate rently infected.1 HCV is an enveloped, positive-sense single- immune response.6 Endogenous IFNa is induced by viruses stranded RNA virus of the family Flaviviridae.2 The viral RNA through pattern recognition receptors, such as Toll-like recep- is translated as a precursor polyprotein, which is then spliced tors, RIG-I and MDA5.7,8 IFNa elicits anti-viral activity by bind- into 10 separate proteins.3 The three structural proteins are E1, ing to Heterodimer of interferon alpha receptor1(IFNAR1) and E2 and core, while p7, NS2, NS3, NS4A, NS4B, NS5A and NS5B interferon alpha receptor 2 (IFNAR2) and inducing Janus-acti- are the non-structural proteins.3 HCV replication mainly vated kinase 1 and tyrosine kinase 2, which in turn phosphorylate occurs in hepatocytes. HCV infection is frequently not detected and thereby activate signal transducer and activator of transcrip- during the acute phase of infection, and most patients become tionproteins1and2(STAT1andSTAT2).9 IFN-regulatory chronically infected.4 Some chronic HCV patients may further factor 9 forms a complex with the STAT1/2 complex in the cell develop liver cirrhosis or even liver cancer.4 nucleus, leading to the upregulation of interferon-stimulated

1State Key Laboratory of Cell Biology, Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China; 2Key Laboratory of Molecular Virology and Immunology, Institut Pasteur of Shanghai, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China; 3Hepatology Section, First Hospital, University of Jilin, Changchun, China; 4Infectious Diseases Department, Fourth Hospital, University of Jilin, Changchun, China and 5Experiment Center of Clinical Medicine, Affiliated Hospital of Binzhou Medical University, Binzhou, China 6 These authors contributed equally to this work. Correspondence: Dr B Sun, State Key Laboratory of Cell Biology, Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031, China; New Biological Building, No. 320 Yueyang Road, Shanghai 200031, China. E-mail: [email protected] Dr JQ Niu, Hepatology Section, First Hospital, University of Jilin, No. 71 Xinmin Street, Changchun 130021, China. E-mail: [email protected] Dr J Zhong, Key Laboratory of Molecular Virology and Immunology, Institut Pasteur of Shanghai, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200025, China; New Biological Building, No. 320 Yueyang Road, Shanghai 200031, China. E-mail: [email protected] Received: 21 July 2014; Revised: 5 December 2014; Accepted: 5 December 2014 IFN-a-induced TRIM22 interrupts HCV replication C Yang et al 95 genes (ISGs), which are involved in biological processes such as (Product Code: 17-0891-01) following the manufacturer’s lipid metabolism, apoptosis, inflammation and protein degra- instructions. dation.9 Some ISGs affect the translation of viral proteins, such as protein kinase R.9 IFNa may also decrease viral RNA Reagents and antibodies stability by inducing 29,59-oligoadenylate synthetase during Recombinant IFNa was purchased from eBioscience, San HCV treatment.9 Diego, CA, USA. Anti-FLAG monoclonal antibody (M2) and The tripartite motif (TRIM) family of proteins is defined by anti-actin antibody were purchased from Sigma-Aldrich, St. the highly conserved RBCC signature domain, which contains a Louis, MO, USA. Anti-HA monoclonal antibody was pur- RING finger domain, one or two B-Box domains, and a coiled- chased from Covance. Rabbit anti-TRIM22 polyclonal anti- coil domain.10,11 The RING finger domain contains a specia- body was purchased from Sigma (Cat. No. HPA003307). lized zinc finger and functions as a ubiquitin protein ligase, either alone or as a part of a multi-subunit E3 protein complex. Reverse transcription and real-time RT-PCR The associated B-Box and coiled-coil domains are involved in Total RNA was obtained using TRIzol reagent (Life Techno- protein interactions and homo/heterodimerization. TRIM pro- logies, Grand Island, NY, USA) according to the manufacturer’s teins are involved in a variety of cellular functions, including instructions. Total cellular RNA (2 mg) was reverse-transcribed apoptosis, transcription, differentiation and regulation of cell with SuperScript II reverse transcriptase (Life Technologies), cycle progression.10,11 In addition, some TRIM proteins exhibit and 2 ml of the reverse-transcribed cDNA was used in real-time anti-viral activity and may be involved in innate immunity.10,11 RT-PCR with the 7900HTFast Real-Time PCR System (Applied Many TRIM proteins have recently emerged as IFN-induced Biosystems, Foster City, CA, USA) and a traditional agarose gel proteins involved in various cellular processes, including cell assay. The primers used for TRIM22 detection were 59-GGTT- proliferation and antiviral activities. TRIM22, also known as GAGGGGATCGTCAGTA-39 (L) and 59-TTGGAAACAGATT- Staf-50, is a natural antiviral effector of HIV-1,12 HBV13 and TTGGCTTC-39 (R). influenza.14 The dependence of the anti-viral function of TRIM22 on its ubiquitination function varies depending on Plasmids the virus.10,11 For example, TRIM22 inhibits influenza by tar- Sequences encoding TRIM22 were amplified by PCR from geting its nucleoprotein, but also inhibits the activity of the cDNA from Huh-7 cells challenged with IFNa for 12 h. HBV core promoter not through ubiquitination.13 TRIM22 and its truncation mutants were cloned into the We previously determined that TRIM22 is highly upregu- pcDNA3.1-59-FLAG vector. PCR amplification was performed lated in peripheral blood mononuclear cell (PBMC) samples in a Mastercycler (Eppendorf, Hamburg, Germany), and PCR from HCV patients treated with IFNa, as well as in Huh-7 cells products were separated on 1% agarose gels and visualized by treated with IFNa.13 The function of TRIM22 was further veri- ethidium bromide staining. fied by examining its effects on the HCV Huh-7-Con-1 repli- The sequences of the PCR primers were as follows: con system or JFH1 virus. Here, the mechanism of TRIM22 is TRIM22 forward: 59-ATGGATTTCTCAGTAAAGGTA-39 explored. TRIM22 reverse: 59-TCAGGAGCTCGGTGGGCACAC-39 TRIM22A reverse: 59-CCGCTCGAGtcaaatgtttggtgaccttgg- MATERIALS AND METHODS tgttcctgagac-39 Subjects and study protocol TRIM22B reverse: 59-CCGCTCGAGtcacatcccactcagatctgg- The study was approved by the ethics committee of the 1st tactcggaa-39 Hospital of Jilin University and was registered with Clini- TRIM22C forward: 59-CCATCGATatggcaagttcttaaagagctg- calTrials.gov, ID NCT01434212. All enrolled subjects were acagatgtccag-39 hepatitis B surface antigen-negative, hepatitis B surface antigen TRIM22D forward: 59- CCATCGATatgcataaacgaggtggtcaa- antibody-negative and HIV antibody-negative, with no sero- ggaatgtcagg-39 logical or histological evidence of a cause of liver disease NS5A forward: 59-AATGAATTCatgtccggatcctggctccgcg-39 other than HCV. Treatment consisted of standard IFNa NS5A reverse: 59-ccgggtaccatgcagcacacggtggtatcgtcctc-39 (Recombinant Human IFNa-2b; Kaiyinyisheng Inc., Beijing, NS5A-D1 reverse: 59-ccgggtaccattgatccccgtgccaagcgc-39 China, FDA Approval number: S20030032) plus ribavirin NS5A-D12 reverse: 59-ccgggtaccatttctgatatggtgctctcgctcag- (Weilake, FDA Approval number: H10940157) for 48 weeks. accc-39 Five million units of IFNa-2b were administered subcuta- NS5A-D23 forward: 59- AATGAATTCatgcctccatctgaggcga- neously every other day at the clinic throughout treatment. gctcctca-39 Ribavirin was dosed at 15 mg/kg/day in two divided doses. Serum HCV RNA levels were measured using the COBAS Virus, cell culture and transfection AmpliPrep/COBAS TaqMan assay (Roche Molecular Diag- HCV JFH1 virus was kindly provided by Takaji Wakita, nostics, Pleasanton, CA, USA) with a lower limit of quantifica- National Institute of Infectious Diseases, Japan.15,16 CON1 repli- tion of 15 IU/ml. Measurements were obtained at the following con cells were provided by Ralf Bartenschlager, Department of time points after IFNa treatment initiation: 0 h, 12 h, 24 h, 37 h Infectious Diseases, Molecular Virology, Heidelberg University, and 43 h. PBMCs were separated using GE Percoll reagent Heidelberg, Germany.17 JFH1 full genome Huh-7 cells were

Cellular & Molecular Immunology IFN-a-induced TRIM22 interrupts HCV replication C Yang et al 96

generated as previously described.18 HEK293T, Huh-7 or HCV UUCUUAATT-39. HEK293T cells or Huh-7 cells were trans- replicon cells were cultured in DMEM (Life Technologies) sup- fected with the siRNA targeting TRIM22 or control siRNA plemented with heat-inactivated fetal calf serum (10%), penicil- (20 nM final concentration) using Lipofectamine 2000 (Life lin (100 U/ml) and streptomycin (100 mg/ml). The cells were Technologies) or Lipofectamine LTX (Life Technologies) using grown at 37 uC in a humidified incubator with a 5% CO2 the standard method described in the manufacturer’s protocol. atmosphere. Empty vector was added to normalize the final plasmid amount. HEK293T cells were seeded into 6-well plates or 24-well The degree of gene silencing was confirmed by RT-PCR or plates for 24 h. The cells were transiently transfected using immunoblotting 24 h after transfection. Lipofectamine (Life Technologies) according to the manufac- turer’s instructions. The plasmid amounts were normalized by Statistical analysis the addition of empty plasmid. Normally distributed continuous variables were compared using t-tests. The Mann–Whitney test was used when a normal Immunoblotting and immunoprecipitation distribution was rejected (Shapiro test, P,0.05). In all cases, a P For immunoblotting, cells were washed with ice-cold phos- value #0.05 was considered significant. IBM SPSS V.19 soft- phate-buffered saline and lysed with ice-cold lysis buffer con- ware was used for statistical analysis. taining 1.0% (v/v) Nonidet P40, 20 mM Tris-HCl pH 8.0, 10% (v/v) glycerol, 150 mM NaCl, 1 mM NaF, 2 mM sodium ortho- RESULTS vanadate, 1 mM sodium pyrophosphate and a protease inhib- TRIM22 is induced in PBMCs from HCV patients by Type I itor cocktail (Roche, Basel, Switzerland). After incubation on IFNa treatment and is associated with a decrease in HCV ice for 15 min, the supernatant was obtained by centrifuging titers the crude lysates at 12 000 g for 15 min to remove cell debris. TRIM22 is upregulated by IFNa treatment. In this study, we Total protein was separated on a 10% SDS–polyacrylamide gel evaluated whether this association occurred in HCV patients and transferred to a nitrocellulose membrane (Bio-Rad, who were treated with IFNa. Patient PBMCs were collected Berkeley, CA, USA). For immune detection, membranes were after IFNa treatment, and TRIM22 expression was analyzed washed with TBS-T (20 mM Tris, 500 mM NaCl, 0.1% Tween by both real-time PCR and immunoblotting. A significant 20) and blocked in a 5% powdered-milk solution in TBS-T for increase in TRIM22 expression was observed in a real-time 1 h. After washing with TBS-T, the membranes were separately PCR assay 12 h after the initiation of IFNa treatment probed with anti-HA (HA1.1) (Covance, Princeton, NJ, USA) (Figure 1a). TRIM22 induction was also observed at the protein (1 : 1000), anti-FLAG M2 (1 : 1000), anti-TRIM22 (1 : 500) level by western blotting (Figure 1c). Concomitantly, the HCV (Sigma) and anti-beta-actin (1 : 1000) antibodies for 1 h. virus titer in the blood decreased rapidly (Figure 1b). The Secondary peroxidase-labeled anti-rabbit or anti-mouse anti- change in early virus kinetics after IFNa administration suggests bodies were incubated for 1 h at room temperature. Anti-NS5A that TRIM22 is involved in IFNa-induced antiviral effects. was raised in Dr Jin Zhong’s lab. Protein detection was visua- lized by ECL according to the manufacturer’s directions TRIM22 is upregulated in Huh-7 cells treated with IFNa (Pierce, Waltham, MA, USA). To determine whether IFNa is capable of inducing TRIM22, we For immunoprecipitation, total cell lysates were subjected to examined TRIM22 expression in the Huh-7 human hepatocyte immunoprecipitation with 0.5 mg of mouse anti-HA (HA1.1) cell line after IFNa treatment. When Huh-7 cells were treat- (Covance) monoclonal antibody or rabbit anti-TRIM22 poly- ed with IFNa for 6 h, TRIM22 mRNA expression increased, clonal antibodies in 500 ml of 1% (v/v) lysis buffer. Binding reaching a maximum at 18 h post-stimulation (Figure 2a). between the antibody and antigen was allowed to occur at 4 uC Accordingly, western blotting with a TRIM22-specific anti- for 2 h in suspension under constant rotation. Then, the pro- body revealed that TRIM22 protein levels increased signifi- tein A/G agarose suspension was added, and the mixture was cantly after 12 h of stimulation (Figure 2b). HCV infection incubated for 2 h at 4 uC with constant agitation. The immu- triggered endogenous TRIM22 expression (Figure 2c), albeit noprecipitated complexes were washed three times with 1% IP at lower levels than that induced by direct IFNa stimulation. buffer, and the proteins were eluted by adding 30 ml of 2% SDS– Taken together, these results demonstrate that TRIM22 is PAGE sample buffer, followed by boiling for 5 min. Sepharose induced following IFNa stimulation of Huh-7 cells. beads were pelleted by centrifugation in a microfuge for 5 min. The supernatant containing proteins was separated by SDS- TRIM22 is an anti-HCV molecule PAGE, followed by staining with mouse anti-FLAG M2 mono- We next determined whether IFNa-induced TRIM22 can sup- clonal antibody (Sigma). press HCV replication. The Huh-7 cell line was transfected with the Con1 replicon genome as described previously by Lohmann siRNA et al.17 After selection with G418, we obtained a cell line that TRIM22 Stealth Select RNAi (Catalog# 1299003) was purchased expressed the deleted Con1 genome stably, Huh-7-Con1- from Life Technologies: short interfering negative control rep. TRIM22 was over-expressed in Huh-7-Con1-rep cells sequence 59-UUCUCCGAACGUGUCACGUTT-39,si159-CA- (Figure 3a), leading to a decrease in HCV levels of approximately CCAAACAUUCCGCAUAATT-39,si259-GGAUGCUGCAAG- 50% compared to control cells 48 h following transfection and

Cellular & Molecular Immunology IFN-a-induced TRIM22 interrupts HCV replication C Yang et al 97

a b

5 7 *** *** 4 *** 6 *** *** 3 *** *** *** 5 2

1 4 HCV RNA log10 IU/ml HCV RNA 0 3 0 1 3 0 1 2 3 Trim22 RNA Relative Fold Change RNA Trim22 0.5 1.5 Time after treatment (days) Time after treatment (days) c

TRIM22

actin

Days 0 0.513 1.5

Figure 1 TRIM22 is induced in PBMCs by Type I IFN treatment of HCV patients and is associated with a decrease in HCV levels. (a) The TRIM22 mRNA level was significantly increased in the PBMCs of HCV patients approximately 12 h after initiation of IFNa treatment, as measured by real- time PCR. Comparison of the HCV mRNA levels with the baseline HCV mRNA level at the remaining time points revealed significant differences. ***P,0.001 as determined by Student’s t-test. (b) HCV viral titers (shown as the logarithm) in patients’ blood decreased markedly, as measured by COBAS (as described in the section on ‘Materials and methods’). Comparison of the HCV titer with the baseline virus titration at other time points revealed significant differences. ***P,0.001, as determined by Student’s t-test. (c) TRIM22 is induced in patient PBMC samples, as determined by western blotting (a representative patient result is shown). HCV, hepatitis C virus; IFN, interferon; PBMC, peripheral blood mononuclear cell; TRIM, tripartite motif. approximately 72% at 96 h (Figure 3c). Inhibition of HCV was TRIM22 interacts with HCV NS5A also observed at the protein level by western blot assay Because TRIM22 is an anti-HCV ISG, we determined whether (Figure 3b). To suppress TRIM22 expression, two siRNAs were TRIM22 directly interacts with the proteins encoded by HCV. designed, and their knockdown efficiency was evaluated by west- First, we screened all HCV proteins by co-immunoprecipita- ern blotting (Figure 3d). Cells were infected with the HCV JFH- tion (co-IP) (Supplementary Figure 1), which revealed that 1 virus (multiplicity of infection (MOI) of 0.1). When the TRIM22 specifically binds the NS5A protein (Figure 4a). infected cells were treated with control siRNA and IFNa in Second, the endogenous TRIM22-NS5A interaction was combination, HCV RNA was significantly decreased. By con- observed (Figure 4b). We next identified the domain of trast, under the same conditions, si1 and si2 against TRIM22 at TRIM22 that is essential for the interaction between TRIM22 the same concentration as the control siRNA suppressed and NS5A. Four truncations of TRIM22 lacking the SPRY and TRIM22 expression and subsequently inhibited IFNa treat- the coiled coil domains (A), the SPRY domain (B), the RING ment-mediated suppression of HCV at both the protein and and B-box domains (C) or only the RING domain (D) were RNA levels (Figure 3e and f). Thus, TRIM22 is an anti-HCV constructed to identify the domains required for the inter- molecule that can negatively regulate HCV replication. action between TRIM22 and NS5A (Figure 4c). TRIM22

a b IFNa IFNa Time (hours) 01862412 Time (hours) 0 12 2436 48

TRIM22 TRIM22

actin actin

c IFN JFH1 PBS TRIM22 Huh-7 actin

Figure 2 TRIM22 is upregulated in Huh-7 cells treated with IFNa.(a, b) TRIM22 mRNA is upregulated in Huh-7 cells treated with IFNa (50 IU/ml). Time-series (time points are shown in the figure) samples were collected and prepared for RT-PCR or western blotting. The primers are described in the section on ‘Materials and methods’. (c) TRIM22 is induced by JFH-1 (MOI50.1) 24 h after JFH-1 infection but at a lower level than induced by IFNa (50 IU/ml). HCV, hepatitis C virus; IFN, interferon; MOI, multiplicity of infection.

Cellular & Molecular Immunology IFN-a-induced TRIM22 interrupts HCV replication C Yang et al 98

a d SiNC Si1 Si2

NC TRIM22 IFNa TRIM22 FLAG-TRIM22 TRIM22 actin endogeneous actin e b SiNC Si1 Si2 SiNC+ IFNSi1+IFNSi2+ IFN GFP TRIM22IRF1 GFP TRIM22IRF1 TRIM22 48h NS5A 72h actin 96h

NS5A actin f 15 c * NS NS *** 48h 10 72h 3 *** * 96h *** *** *** 2 5 ***

1 Relative Fold Change 0 Relative Fold Change 0 Si1 Si2 SiNC GFP IRF1 GFP IRF1 GFP IRF1 Si1+IFNSi2+ IFN SiNC+ IFN TRIM22 TRIM22 TRIM22

Figure 3 TRIM22 has anti-viral activity against HCV. (a) Western blot analysis of TRIM22 expression with anti-TRIM22 antibody after transfection of cells with FLAG-TRIM22 (0.5 mg/well, 24-well plate). Endogenous TRIM22 is shown as a positive control in the lane labeled ‘IFNa’ in Huh-7 cells that were treated with IFNa (50 IU/ml) for 24 h and collected for western blotting. (b, c) A marked decrease in the virus replicon levels was observed in TRIM22-over-expressing Huh-7-con1-replicon cells 48–96 h after infection. Huh-7-con1-replicon cells were transfected with GFP, TRIM22 and the anti-viral ISG and IRF1 plasmids. Samples were collected at 48 h, 72 h and 96 h after transfection. Real-time RT-PCR and western blotting were performed as described in the section on ‘Materials and methods’. (d) The knockdown efficiency of the siRNAs was determined by western blotting. Huh-7 cells were transfected with siRNA or TRIM22 (siRNA: 50 nmol per well/24-well plate, TRIM22: 100 ng per well/24-well plate), and cells were collected for western blotting 24 h after transfection. (e, f) Knockdown of TRIM22 diminished the anti-viral effects of IFNa. Huh-7 cells were transfected with siRNA (50 nmol per well/24-well plate). After 24 h, the transfected cells were infected with JFH1 (MOI50.1). At 24 h after infection, the cells were treated with IFNa (50 IU/ml) for an additional 48 h. Samples were collected for real-time PCR assays and western blotting. *P,0.05 as determined by Student’s t-test. ***P,0.001 as determined by Student’s t-test. HCV, hepatitis C virus; IFN, interferon; IRF, IFN-regulatory factor; ISG, interferon-stimulated gene; PBMC, peripheral blood mononuclear cell; TRIM, tripartite motif.

mutants A and B (which lack the SPRY domain) were unable to (Figure 5b). MG132 blocked the TRIM22-mediated degrada- interact with NS5A (Figure 4d), indicating that the SPRY tion of NS5A (Figure 5d). TRIM22 does not degrade other domain is important for the interaction between TRIM22 HCV proteins, indicating that the degradation is specific for and NS5A. This interaction suggests that the NS5A protein NS5A (Figure 5a). Based on the importance of NS5A in HCV plays a role in the inhibition of HCV function by TRIM22. replication and the results of our study, we conclude that We also determined which domain of NS5A is essential for TRIM22 is induced by IFNa and plays a role in negatively the interaction between TRIM22 and NS5A. NS5A contains regulating HCV replication by ligating ubiquitin to NS5A. three protein domains: Domain 1, Domain 2 and Domain 3. We created three truncations of NS5A, NS5A-D1 (Domain 1), DISCUSSION NS5A-D12 (Domain 1 and 2) and NS5A-D23 (Domain 2 and In the present study, upregulation of TRIM22 by IFNa was 3) (Figure 4e). Only NS5A-D23 did not interact with TRIM22, associated with a decrease in HCV titer in the blood of patients indicating that domain 1 is essential for the interaction between on the first day after IFNa administration. Because liver biopsy NS5A and TRIM22 (Figure 4f). In addition, TRIM22 and NS5A samples were not available, it is unclear whether TRIM22 was colocalized in the cytoplasm of Huh-7 cells around lipid drop- also induced in hepatocytes. Previous research13 has demon- lets (Figure 4j and Supplementary Figure 2). strated that TRIM22 is upregulated in liver cell lines, consistent with its up-regulation in PBMCs. In the present study, upre- TRIM22 is an ubiquitin ligase for NS5A gulation of TRIM22 was also observed in Huh-7 cells after TRIM family members feature a common E3 ubiquitin ligase IFNa treatment. domain; thus, TRIM22 may be a ubiquitin ligase for NS5A. There is still no consensus on the localization of TRIM22.19 TRIM22 efficiently ligated ubiquitin to NS5A (Figure 5c) and When 293T or COS7 cells were transfected with human TRIM22 degraded NS5A in a concentration-dependent manner tagged with GFP, TRIM22 was localized in the cytoplasm.19–21

Cellular & Molecular Immunology IFN-a-induced TRIM22 interrupts HCV replication C Yang et al 99

a df 123 Flag-TRIM22 TRIM22-ATRIM22-BTRIM22-CTRIM22-DTRIM22 NS5A-D1NS5A-D12NS5A-D23NS5A1NC HA-NS5A

actin Flag-TRIM22 HA-NS5A IP:HA WB:Flag

IP:Flag WB:HA HA-NS5A Flag-TRIM22 b IFN + – TRIM22 IP:TRIM22 IP:HA IP:Flag IB:NS5A WB:Flag WB:HA IP:NS5A IB:TRIM22 IP:Flag NS5A IP:HA WB:HA WB:Flag actin

c A e D1 B D12 C D D23 TRIM22 NS5A RING B-Box Coiled Coil SPRY Domain1 Domain2 Domain3

j MERGE NS5A TRIM22 DAPI Co-local

+IFN

MERGE NS5A TRIM22 DAPI LD

+IFN

–IFN

Figure 4 TRIM22 interacts with HCV NS5A. (a) Both FLAG-TRIM22 (1.5 mg/well, six-well plate) and HA-NS5A (1.5 mg/well, six-well plate) were expressed in 293T cells. For lane 1, cells were transfected with only HA-NS5A. For lane 2, cells were transfected with only FLAG-TRIM22. For lane 3, cells were transfected with both HA-NS5A and FLAG-TRIM22. Co-IP experiments were performed with anti-FLAG and anti-HA antibodies 24 h after transfection of cells with the plasmids. The results of the co-IP experiments revealed that TRIM22 interacts with HCV NS5A. (b)The endogenous interaction of TRIM22 and NS5A was evaluated. Twenty-four hours after adding IFNa (50 IU/ml) to the HCV Huh-7-Con-1 replicon, the cells were harvested and subjected to endogenous IP with anti-TRIM22 and anti-NS5A antibodies. (c) Schematic diagram of the four truncated forms of TRIM22: TRIM22A (lacks coiled-coil and SPRY domains), B (lacks SPRY domain), C (lacks RING and B-Box domains) and D (lacks RING domain). (d) The SPRY and coiled-coil domains are important for the interaction of TRIM22 with NS5A because only TRIM22 C and TRIM22 D interact with NS5A. All TRIM22 truncations (1.5 mg/well, six-well plate) and HA-NS5A (1.5 mg/well, six-well plate) were expressed in 293T cells. Co- IP experiments were performed using anti-FLAG and anti-HA antibodies 24 h after transfection of cells with the plasmids. (e) Schematic diagram of the three truncated forms of NS5A: NS5A-D1 (contains only domain 1), NS5A-D12 (contains domains 1 and 2) and NS5A-D23 (contains domains 2 and 3). (f) Domain 1 of NS5A is important for the interaction of TRIM22 with NS5A because only NS5A-D23 does not interact with NS5A. All HA- NS5A truncations (1.5 mg/well, six-well plate) and FLAG-TRIM22 (1.5 mg/well, six-well plate) were expressed in 293T cells. Co-IP experiments were performed using anti-FLAG and anti-HA antibodies 24 h after transfection of cell with the plasmids. (j) IFNa (50 IU/ml) treatment was initiated 24 h after Huh-7 cells were infected with the JFH-1 virus (MOI of 0.1). TRIM22 was colocalized with NS5A in the cytoplasm. Immunofluorescence was performed IFNa using anti-TRIM22 and anti-NS5A 12 h after IFNa treatment. LD staining was conducted with bodipy dye (Life Technologies D2228). The region of co-localization is indicated in white in the figure labeled ‘Co-local’. TRIM22 and NS5A were colocalized near the lipid droplet, which is reported to be important for HCV assembly. co-IP, co-immunoprecipitation; HCV, hepatitis C virus; IFN, interferon; MOI, multiplicity of infection; TRIM, tripartite motif.

Cellular & Molecular Immunology IFN-a-induced TRIM22 interrupts HCV replication C Yang et al 100

ab TRIM22+Ubiquitin Exogenous NA5A –+ –+ –+ Exogenous NS5A TRIM22(ng) 0200 1000 IFN(IU/ml) 050 100 CORE E1 FLAG-TRIM22 TRIM22 HA-NS5A HA-NS5A E2 Actin Actin P7 Endogenous NS5A Endogenous NA5A NS2 TRIM22(ng) 0200 1000 IFN(IU/ml) 050 100 NS3 FLAG-TRIM22 NS4A TRIM22 HA-NS5A NS4B NS5A Actin Actin NS5A NS5B cd IFN –++ TRIM22 –+ MG132 ––+ Ubiquitin + + IFN –+

IP:HA IP:NS5A Ubi WB:His WB:Ubi

NS5A NS5A NS5A

TRIM22 TRIM22 TRIM22

actin actin actin

Figure 5 TRIM22 is an ubiquitin ligase of NS5A. (a) TRIM22 specifically degrades NS5A. 293T cells are transfected. HCV proteins are transfected together with TRIM22 and ubiquitin. Western blotting was performed 48 h after plasmid transfection. (b) Exogenous NS5A: Huh-7 cells transfected with HA-NS5A plasmid (100 ng/well in a 24-well plate) were transfected with FLAG-TRIM22 (upper left: 0, 200 and 1000 ng/well in a 24-well plate) and HIS-Ubiquitin (40 ng/well in a 24-well plate) plasmids or treated with IFNa (upper right: IFNa 0, 50 and 100 IU/ml triggered 24 h after HA-NS5A plasmid transfection). After 48 h, samples were collected for western blotting. TRIM22 degraded NS5A (100 ng/well in a 24-well plate) in a dose- dependent manner. Endogenous NS5A: Huh-7-Con-1 replicon cells were transfected with FLAG-TRIM22 (lower left: 0, 200 and 1000 ng/well in a 24-well plate) and HIS-Ubiquitin (40 ng/well in a 24-well plate) or treated with IFNa (lower right: 50 IU/ml). Samples were detected by western blotting 48 h after plasmid transfection or IFNa treatment. TRIM22 degraded endogenous NS5A (100 ng/well in a 24-well plate) in a dose- dependent manner. (c) Upper panel: FLAG-TRIM22 (1.5 mg/well, six-well plate), HA-NS5A (1.5 mg/well, six-well plate) and HIS-Ubiquitin (0.3 mg/ well, six-well plate) were expressed in 293T cells. Immunoprecipitation of NS5A with anti-HA antibody and western blotting of ubiquitin with an anti- HIS antibody were performed 24 h after plasmid transfection. TRIM22 efficiently ligated ubiquitin to NS5A. Lower panel: endogenous ubiquitination of NS5A was detected. Huh-7-Con-1 replicon cells were treated with IFNa (50 IU/ml). Twenty-four hours after IFNa treatment, cells were harvested and immunoprecipitated with an anti-NS5A antibody. Then, western blotting was performed with an anti-ubiquitin antibody. (d) MG132 (20 nM) successfully blocked the endogenous ubiquitination of NS5A and NS5A degradation. HCV, hepatitis C virus; IFN, interferon; TRIM, tripartite motif.

By contrast, when HepG2 cells were transfected with myc-tagged with the final outcome of IFNa therapy. Furthermore, experi- TRIM22, TRIM22 was localized in the nucleus.19,22,23 Moreover, mental evidence suggests that the function of NS5A includes in the serum-starved U2OS cell line, TRIM22 was localized in abrogation of STAT-1 translocation,29 suppression of STAT-1 both the nucleus and cytoplasm.19,24,25 Gao and colleagues phosphorylation 30 and repression of the protein kinase R pro- observed that TRIM22 was mainly localized in the nucleus of tein kinase.31–33 Because of its important function in control- HepG2 cells when it suppresses HBV survival.13 The discrepancy ling HCV replication, many molecules have been designed to in localization compared to our present study may be due to the target NS5A, some of which have exhibited promising treat- different cell lines we used (Huh-7 cell lines). ment outcomes in clinical research.34,35 Other potential The mechanism by which TRIM22 suppresses HCV replica- mechanisms of the anti-HCV effects of TRIM22 should be tion is of substantial interest. Studies of changes in HCV kine- evaluated in future research. tics during IFNa therapy have demonstrated that the dramatic We performed experiments demonstrating that a single decrease in viral titer on the first day of IFNa therapy is mainly mutation (K to R) of NS5A can all be ubiquitinated, which caused by the inhibition of HCV replication.26 All HCV proteins implies that multiple ubiquitination sites might appear in were screened for interaction with TRIM22, resulting in the sole NS5A (Supplementary Figure 3). At present, it is unclear which identification of NS5A. NS5A is a viral factor that helps HCV lysine in NS5A is ubiquitinated by TRIM22. The sequence of establish a niche in the host cell by turning off the IFNa signaling the HCV virus is constantly changing in the host.36 Therefore, a pathway.27 The sequence of a certain NS5A region, called the large HCV mutation repertoire exists in every HCV patient.37 interferon sensitivity-determining region,28 is highly associated To uncover the clinical association between the ubiquitinated

Cellular & Molecular Immunology IFN-a-induced TRIM22 interrupts HCV replication C Yang et al 101 sites and the resistance to IFNa, next generation sequencing 4 Levrero M. Viral hepatitis and liver cancer: the case of hepatitis C. should be used to determine the mutation status of NS5A at a Oncogene 2006; 25: 3834–3847. 5 Lai CL. Antiviral therapy for hepatitis B and C in Asians. J population level because HCV has a highly dynamic muta- Gastroenterol Hepatol 1999; 14 Suppl: S19–S21. 37 tional profile. Different time points should also be taken into 6 Schoggins JW, Wilson SJ, Panis M, Murphy MY, Jones CT, Bieniasz P consideration, and the association between the mutational pro- et al. A diverse range of gene products are effectors of the type I file and therapy should then be assessed. We will address those interferon antiviral response. Nature 2011; 472: 481–485. 7 Prens EP, Kant M, van Dijk G, van der Wel LI, Mourits S, van der Fits L. questions in a future study. IFN-alpha enhances poly-IC responses in human keratinocytes by In this study, we observed that TRIM22 is induced after inducing expression of cytosolic innate RNA receptors: relevance for IFNa treatment and functions to degrade NS5A and, in turn, psoriasis. J Invest Dermatol 2008; 128: 932–938. suppress HCV replication. Overall, our study indicates that 8 Xagorari A, Chlichlia K. Toll-like receptors and viruses: induction of innate antiviral immune responses. Open Microbiol J 2008; 2: 49– TRIM22 has a novel anti-HCV effect during IFNa therapy in 59. chronically HCV-infected patients and that this effect is 9 Feld JJ, Hoofnagle JH. Mechanism of action of interferon and ribavirin mediated by the ubiquitination of NS5A. in treatment of hepatitis C. Nature 2005; 436: 967–972. 10 Ozato K, Shin DM, Chang TH, Morse HC 3rd. TRIM family proteins and their emerging roles in innate immunity. Nat Rev Immunol 2008; 8: AUTHORS’ CONTRIBUTIONS 849–860. BS, JN and JZ proposed and managed the project. CY designed 11 Nisole S, Stoye JP, Saib A. TRIM family proteins: retroviral restriction the research project, performed most of the experiments and and antiviral defence. Nat Rev Microbiol 2005; 3: 799–808. wrote the manuscript. XZ performed most of the experiments 12 Barr SD, Smiley JR, Bushman FD. The interferon response inhibits HIV particle production by induction of TRIM22. PLoS Pathog 2008; involving the HCV virus. DS performed the molecular cloning. 4: e1000007. LY performed the microarray and analyzed the results. SX and 13 Gao B, Duan Z, Xu W, Xiong S. Tripartite motif-containing 22 inhibits CC assisted with molecular cloning and immunoprecipitation the activity of hepatitis B virus core promoter, which is dependent experiments. YP and YG recruited the cohort and led the clini- on nuclear-located RING domain. Hepatology 2009; 50: 424– cal research. MW, XS, HS and JL collected the biological samples 433. 14 Di Pietro A, Kajaste-Rudnitski A, Oteiza A, Nicora L, Towers GJ, and follow-up information from the patients. YG performed Mechti N et al. TRIM22 inhibits influenza A virus infection by the HCV RNA titration assay. XC was responsible for ethical targeting the viral nucleoprotein for degradation. J Virol 2013; 87: approval. 4523–4533. 15 Wakita T, Pietschmann T, Kato T, Date T, Miyamoto M, Zhao Z, et al. Production of infectious hepatitis C virus in tissue culture from a CONFLICTS OF INTEREST cloned viral genome. Nat Med 2005; 11: 791–796. The authors of this study have no conflicts to disclose. 16 Zhong J, Gastaminza P, Cheng G, Kapadia S, Kato T, Burton DR et al. Robust hepatitis C virus infection in vitro. Proc Natl Acad Sci USA ACKNOWLEDGEMENTS 2005; 102: 9294–9299. 17 Lohmann V, Korner F, Koch J, Herian U, Theilmann L, Bartenschlager We thank R Bartenschlager for providing the CON1 replicon and T R. Replication of subgenomic hepatitis C virus RNAs in a hepatoma Wakita for providing the JFH1 virus. We thank Wanyin Tao, Qiang cell line. Science 1999; 285: 110–113. Ding, Yu Xiang and Yongfen Xu for experimental support. We thank 18 He Y, Weng L, Li R, Li L, Toyoda T, Zhong J. The N-terminal helix Dahui Zhao, Shuai Yang and all of the other nurses who helped collect alpha0 of hepatitis C virus NS3 protein dictates the subcellular patient blood. We thank Shijie Sun and Jizheng He from Roche localization and stability of NS3/NS4A complex. Virology 2012; Diagnostics, China, for support with the COBAS technique. We thank 422: 214–223. Li Li and Wenjing Xuan for ordering and preparing the experimental 19 Hattlmann CJ, Kelly JN, Barr SD. TRIM22: a diverse and dynamic reagents. antiviral protein. Mol Biol Int 2012; 2012: 153415. 20 Herr AM, Dressel R, Walter L. Different subcellular localisations of This work was supported by a grant from the National 973 Key Project TRIM22 suggest species-specific function. Immunogenetics 2009; (2013CB530504), the National 863 Key Project (2012AA020103), 61: 271–280. National Science and Technology Major Projects (2013ZX10004-101- 21 Reymond A, Meroni G, Fantozzi A, Merla G, Cairo S, Luzi L et al. The 005, 2012ZX10002-007-003 and 2013ZX10004-003-003), grants from tripartite motif family identifies cell compartments. EMBO J 2001; the National Natural Science Foundation of China (31030029, 20: 2140–2151. 31230024 and 81361120409) and by the CAS/SAFEA International 22 Duan Z, Gao B, Xu W, Xiong S. Identification of TRIM22 as a RING Partnership Program for Creative Research Teams. finger E3 ubiquitin ligase. Biochem Biophys Res Commun 2008; 374: 502–506. Supplementary Information accompanies the paper on Cellular & 23 Yu S, Gao B, Duan Z, Xu W, Xiong S. Identification of tripartite motif- containing 22 (TRIM22) as a novel NF-kappaB activator. Biochem Molecular Immunology’s website. (http://www.nature.com/cmi). Biophys Res Commun 2011; 410: 247–251. 24 Petersson J, Lonnbro P, Herr AM, Morgelin M, Gullberg U, Drott K. The human IFN-inducible p53 target gene TRIM22 colocalizes with the centrosome independently of cell cycle phase. Exp Cell Res 2010; 1 Slomski A. WHO issues guidelines on HCV amid drug cost controversy. 316: 568–579. JAMA 2014; 311: 2262–2263. 25 Sivaramakrishnan G, Sun Y, Tan SK, Lin VC. Dynamic localization of 2 Chevaliez S, Pawlotsky JM. HCV genome and life cycle. In: Tan SL, tripartite motif-containing 22 in nuclear and nucleolar bodies. Exp editor. Hepatitis C Viruses: Genomes and Molecular Biology. Norfolk: Cell Res 2009; 315: 1521–1532. Horizon Bioscience, 2006. 26 Neumann AU, Lam NP, Dahari H, Gretch DR, Wiley TE, Layden TJ 3 Suzuki T, Aizaki H, Murakami K, Shoji I, Wakita T. Molecular biology et al. Hepatitis C viral dynamics in vivo and the antiviral efficacy of of hepatitis C virus. J Gastroenterol 2007; 42: 411–423. interferon-alpha therapy. Science 1998; 282: 103–107.

Cellular & Molecular Immunology IFN-a-induced TRIM22 interrupts HCV replication C Yang et al 102

27 Qashqari H, Al-Mars A, Chaudhary A, Abuzenadah A, Damanhouri G, 32 Gale M Jr, Blakely CM, Kwieciszewski B, Tan SL, Dossett M, Tang NM Alqahtani M et al. Understanding the molecular mechanism(s) of et al. Control of PKR protein kinase by hepatitis C virus nonstructural hepatitis C virus (HCV) induced interferon resistance. Infect Genet 5A protein: molecular mechanisms of kinase regulation. Mol Cell Biol Evol 2013; 19: 113–119. 1998; 18: 5208–5218. 28 Enomoto N, Sakuma I, Asahina Y, Kurosaki M, Murakami T, 33 Gale MJ Jr, Korth MJ, Tang NM, Tan SL, Hopkins DA, Dever TE et al. Yamamoto C et al. Mutations in the nonstructural protein 5A gene Evidence that hepatitis C virus resistance to interferon is mediated and response to interferon in patients with chronic hepatitis C virus through repression of the PKR protein kinase by the nonstructural 5A 1b infection. N Engl J Med 1996; 334: 77–81. protein. Virology 1997; 230: 217–227. 29 Cao J, Zhou Y, Gong GZ. Effect of HCV NS5A on STAT1 34 Pereira AA, Jacobson IM. New and experimental therapies for HCV. phosphorylation and nuclear translocation induced by IFN alpha- Nat Rev Gastroenterol Hepatol 2009; 6: 403–411. 2b. Zhonghua Gan Zang Bing Za Zhi 2006; 14: 894–897. Chinese. 35 de Clercq E. The design of drugs for HIV and HCV. Nat Rev Drug Discov 30 Gong GZ, Cao J, Jiang YF, Zhou Y, Liu B. Hepatitis C virus non- 2007; 6: 1001–1018. structural 5A abrogates signal transducer and activator of 36 Campo DS, Dimitrova Z, Yamasaki L, Skums P, Lau DT, Vaughan G transcription-1 nuclear translocation induced by IFN-alpha through et al. Next-generation sequencing reveals large connected networks of dephosphorylation. World J Gastroenterol 2007; 13: 4080–4084. intra-host HCV variants. BMC Genomics 2014; 15 Suppl 5: S4. 31 Gale MJ Jr, Korth MJ, Katze MG. Repression of the PKR protein kinase 37 Beerenwinkel N, Gunthard HF, Roth V, Metzner KJ. Challenges and by the hepatitis C virus NS5A protein: a potential mechanism of opportunities in estimating viral genetic diversity from next- interferon resistance. Clin Diagn Virol 1998; 10: 157–162. generation sequencing data. Front Microbiol 2012; 3: 329.

Cellular & Molecular Immunology