YIPF5 Is Essential for Innate Immunity to DNA Virus and Facilitates COPII-Dependent STING Trafficking

This information is current as Yong Ran, Mei-guang Xiong, Zhi-sheng Xu, Wei-wei Luo, of September 26, 2021. Su-yun Wang and Yan-Yi Wang J Immunol published online 7 August 2019 http://www.jimmunol.org/content/early/2019/08/06/jimmun ol.1900387 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2019/08/06/jimmunol.190038 Material 7.DCSupplemental http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 26, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts 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 © 2019 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published August 7, 2019, doi:10.4049/jimmunol.1900387 The Journal of Immunology

YIPF5 Is Essential for Innate Immunity to DNA Virus and Facilitates COPII-Dependent STING Trafficking

Yong Ran, Mei-guang Xiong, Zhi-sheng Xu, Wei-wei Luo, Su-yun Wang, and Yan-Yi Wang

STING plays central roles in the innate immune response to pathogens that contain DNA. Sensing cytoplasmic DNA by cyclic GMP-AMP synthase produces cyclic GMP-AMP, which binds to and activates STING and induces STING translocation from the to the perinuclear . However, this trafficking process has not been fully elucidated yet. In this study, we identified YIPF5 as a positive regulator of STING trafficking. YIPF5 is essential for DNA virus- or intracellular DNA- triggered production of type I IFNs. Consistently, knockdown of YIPF5 impairs cellular antiviral responses to DNA virus. Mech- anistically, YIPF5 interacts with both STING and components of COPII, facilitating STING recruitment to COPII in the presence

of cytoplasmic dsDNA. Furthermore, knockdown of components of COPII inhibits DNA virus-triggered production of type I IFNs, Downloaded from suggesting that COPII is involved in innate immune responses to DNA viruses. Collectively, our findings demonstrate that YIPF5 positively regulates STING-mediated innate immune responses by recruiting STING to COPII-coated vesicles and facilitating STING trafficking from the endoplasmic reticulum to Golgi, providing important insights into the molecular mechanisms of intracellular DNA-stimulated STING trafficking and activation. The Journal of Immunology, 2019, 203: 000–000.

ermline-encoded pattern recognition receptors can detect responses to pathogen-derived DNA and aberrantly accumulated http://www.jimmunol.org/ pathogen-associated molecular patterns, leading to a series cytoplasmic self-DNA (5–7). G of signaling cascades that subsequently result in inducing Upon sensing cytoplasmic DNA, cGAS catalyzes the synthesis type I IFNs and proinflammatory . These cytokines of the second messenger, cGAMP, which binds to endoplasmic subsequently induce wide-range of antiviral and reticulum (ER)–resident adaptor STING (also known as inflammatory , which mediate innate antiviral immune and in- MITA and ERIS) (6, 8–11) and triggers STING translocation. flammatory responses (1, 2). Acting as classic pathogen-associated STING moves from the ER via the to perinuclear molecular patterns, pathogen-derived nucleic acids are recognized by microsomal compartments, accompanied by the recruitment and a subset of pattern recognition receptors, such as viral RNA recog- activation of kinases TBK1 and IKKa/b, ultimately leading to in- nition by TLR3, RIG-I, and MDA5 and the recognition of viral DNA duction of IFNs and proinflammatory cytokines (6, 12, 13). Various by guest on September 26, 2021 by TLR9, cyclic GMP-AMP (cGAMP) synthase (cGAS), RNA po- studies suggest that STING trafficking is the critical process of lymerase III, and IFI16 (3–5). Many studies have established the innate immune responses to cytoplasmic DNA, and recent reports pivotal roles of the cGAS–STING pathway in innate immune show that several are involved in the regulation of STING trafficking, including the translocon-associated protein b (TRAPb) and the translocon adapter Sec61b, exocyst complex component Key Laboratory of Special Pathogens and Biosafety, Wuhan Institute of Virology, Sec5, iRhom2, SCAP, and SNX8 (6, 12, 14, 15). A thorough in- Chinese Academy of Sciences, Wuhan, Hubei 430071, China vestigation of STING trafficking regulation is important because ORCID: 0000-0002-7005-435X (Y.R.). disordered STING translocation contributes to autoinflammatory Received for publication April 3, 2019. Accepted for publication July 11, 2019. and autoimmune diseases. This work was supported by grants from the Strategic Pilot Project funded by the A recent study showed that STING transports to the ER–Golgi Chinese Academy of Sciences (XDB29010302), the Youth Innovation Promotion Association/Chinese Academy of Sciences (2017382), the National Science Fund intermediate compartment (ERGIC), recruits TBK1, and induces for Distinguished Young Scholars (31425010), and the National Natural Science IRF3 activation upon DNA stimulation (13), suggesting that ERGIC Foundation of China (31621061). is an important platform for STING activation. This indicates that Y.R. conceived and designed the study. Y.R., M.-g.X., Z.-s.X., and S.-y.W. performed STING trafficking is more complicated and well organized than the experiments. Y.R. and Y.-Y.W. analyzed the data. Y.R., Y.-Y.W., and W.-w.L. wrote the manuscript. All of the authors discussed the results and commented on the we had imagined. Exporting proteins from the ER is facilitated manuscript. via coat protein complex II (COPII)–coated vesicles, which Address correspondence and reprint requests to Dr. Yong Ran, Key Laboratory of are formed in specialized zones within the ER called the ER Special Pathogens and Biosafety, Wuhan Institute of Virology, Chinese Academy of exit sites (ERES) or transitional ER and located adjacent to the Sciences, Xiao Hong Shan Zhong Qu 44, Wuhan, Hubei 430071, China. E-mail address: [email protected] ERGIC (16, 17). However, the role of COPII-coated vesicles in The online version of this article contains supplemental material. STING trafficking and activation remains unclear. In this study, we found that COPII facilitated the translocation of STING Abbreviations used in this article: BFA, brefeldin A; BMDM, bone marrow–derived macrophage; cGAMP, cyclic GMP-AMP; cGAS, cGAMP synthase; COPII, coat protein from the ER to perinuclear puncta, which is essential for STING- complex II; ER, endoplasmic reticulum; ERES, ER exit site; ERGIC, ER–Golgi interme- mediated signaling. Furthermore, we found that YIPF5, a mem- diate compartment; HA, hemagglutinin; HFF, foreskin fibroblast; MEF, mouse embryonic fibroblast; MLF, mouse lung fibroblast; MOI, multiplicity of infection; ber of the Yip family (YIPF), is responsible for recruiting STING NP-40, Nonidet P-40; qPCR, quantitative real-time PCR; SeV, Sendai virus; shRNA, to COPII-coated vesicles, facilitating intracellular DNA-induced short-hairpin RNA; siRNA, small-interfering RNA; TRAPb, translocon-associated translocation of STING. Therefore, YIPF5 is essential for DNA protein b; VSV, vesicular stomatitis virus; WT, wild-type; YIPF, Yip family. virus- but not RNA virus-triggered innate immune responses. Copyright Ó 2019 by The American Association of Immunologists, Inc. 0022-1767/19/$37.50 Collectively, our findings provide important insights into the

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1900387 2 YIPF5 FACILITATES STING TRANSLOCATION AND ACTIVATION molecular mechanisms of intracellular DNA-stimulated STING Sec13 no. 1, 59-AGGAGGAGCAGAAGCTAGA-39, and no. 2, 59-GGCAA- trafficking and activation. TATGTGGTCACCTA-39. Transfection and reporter assays Materials and Methods 3 5 Constructs HEK293T cells (1 10 ) were seeded on 24-well plates and trans- fected the next day by a standard calcium phosphate precipitation ISRE, NF-kB, and IFN-b luciferase reporter plasmids and method. HeLa cells were transfected by Lipofectamine 2000. Empty mammalian expression plasmids for hemagglutinin (HA)-, Flag-tagged control plasmid was added to ensure that each transfection receives the cGAS, STING, TBK1, IRF3, Sec5, TRAPb, Sec61b, iRhom2, and STING same amount of total DNA. To normalize for transfection efficiency, were previously described (18, 19). Mammalian expression plasmids for 0.01 mgofpRL-TK(Renilla luciferase) reporter plasmid was added Flag-, HA-, RFP-tagged YIPF5 and their truncations and HA-tagged to each transfection. Luciferase assays were performed using a dual-specific SEC13, SEC23B, SEC24D, and SEC31A were constructed by standard luciferase assay kit. dsDNA oligonucleotide VACV70 (59-CCATCAGAA- molecular biology techniques. Expression plasmids for YIPF1–7 were AGAGGTTTAATATTTTTGTGAGACCATCGAAGAGAGAAAGAGAT- purchased from Origene. AAAACTTTTTTACGACT-39) was used to stimulate cells. Reagents, Abs, cells, and viruses Coimmunoprecipitation, immunoblot analysis, and native PAGE Poly(I:C) (InvivoGen), Lipofectamine 2000 (Invitrogen), GM-CSF (PeproTech), polybrene (Millipore), SYBR Green (Bio-Rad), dual-specific luciferase HEK293T cells, THP-1 cells, L929 cells, or MLFs were lysed in l ml of assay kit (Promega), anti-Flag (F3165), and anti–b-actin (A2228) Abs were Nonidet P-40 (NP-40) lysis buffer (20 mM Tris-HCl [pH 7.4], 150 mM from Sigma-Aldrich. Anti–phospho-IRF3 (Ser396, 4D4G), anti-STING (D2P2F), NaCl, 1 mM EDTA, 1% NP-40, and proteasome inhibitor mixture). Coim- and anti-SEC31A (D1G71) Abs were from Technology. Anti-HA munoprecipitation, immunoblot analysis, and native PAGE were performed

Ab (16B12) was from Covance. Anti-TBK1 (ab109735) and anti–phospho- as previously described (20). Downloaded from TBK1 (Ser172, EPITOMICS) Abs were from Abcam. Anti-SEC13 (A11613) and anti-GM130 (A5344) Abs were from ABclonal. Anti-IRF3 Ab (FL-425) STING dimerization and oligomerization assay was from Santa Cruz Biotechnology. Rabbit anti-YIPF5 (SK2473571) and Analysis of STING dimerization and oligomerization was performed as mouse anti-TGN38 (2F7.1) Abs were from Invitrogen. Mouse anti-GM130 described previously (23). HFFs were lysed in 0.5 ml of lysis buffer Ab (610822) was from BD Biosciences. The secondary Ab goat anti-mouse (10 mM PIPES-KOH buffer [pH 7], 50 mM NaCl, 5 mM MgCl2,5mM (31430) or goat anti-rabbit (31460) IgG1 conjugated to HRP was from EGTA, 10% glycerol, and a mixture of protease and phosphatase inhibi- Pierce. HEK293T cells were originally provided by G. Johnson (National 3

tors) containing 1% NP-40 and centrifuged at 20,000 g for 15 min. The http://www.jimmunol.org/ Jewish Health, Denver, CO). HeLa, THP-1, and L929 cells were pur- 3 2/2 cell lysates were mixed with 5 SDS buffer (200 mM Tris-HCl buffer chased from the American Type Culture Collection. Sting mouse [pH 6.8], 10% SDS, 25% glycerol, and 0.05% bromphenol blue) without lung fibroblasts (MLFs) and human foreskin fibroblasts (HFFs) were 2-ME and were analyzed by SDS-PAGE. previously described (20). Sendai virus (SeV), vesicular stomatitis virus (VSV; Indiana strain), HSV-1, and HSV-1–GFP/luciferase were prepared Quantitative real-time PCR as previously described (20, 21). Total RNAwas isolated for quantitative real-time PCR (qPCR) analysis Preparations of mouse embryonic fibroblasts and bone to measure mRNA abundance of the indicated genes. Data shown are marrow–derived macrophages the relative abundances of the indicated mRNA normalized to that of GAPDH/Gapdh. -specific primers used for qPCR were as fol- Mouse embryonic fibroblasts (MEFs) were prepared from day 13.5 C57BL/ lows: human GAPDH forward, 59-GAGTCAACGGATTTGGTCGT- by guest on September 26, 2021 6 mice embryos and cultured in DMEM supplemented with 10% FBS. Bone 39;humanGAPDH reverse, 59-GACAAGCTTCCCGTTCTCAG -39; marrow cells were isolated from the tibia and femur. For preparation of bone human IFNB1 forward, 59-TTGTTGAGA ACCTCCTGGCT-39; human marrow–derived macrophages (BMDMs), the bone marrow cells were IFNB1 reverse, 59-TGACTATGGTCCAGGCACAG-39;humanIFIT1 cultured in 10% M-CSF–containing conditional medium from L929 cells forward, 59-GCCTTGCTGAAGTGTG GAGGAA-39;humanIFIT1 for 3–5 d. All animals were maintained according to the guidelines of reverse, 59-ATCCAGGCGATAGGCAGAGATC-39; human IL6 forward, the animal facility at the Wuhan Institute of Virology, Chinese Academy 59-CCTCAGCCCCCTCTGG GGTC-39; human IL6 reverse, 59-AAGTC- of Sciences. All the experiments were conducted in accordance with the CGCCCTGTAGGTGAGGTT-39; human YIPF5 forward, 59-TGCCCATG- Guidelines for Animal Care and Use of the Wuhan Institute of Virology, ATCCTACTTTCCAG-39; human YIPF5 reverse, 59-AGTTGCTGTCCTT- Chinese Academy of Sciences. CCATGGCT-39; human SAR1A forward, 59-ATAATGCAGGCAAAAC- CACTCT-39; human SAR1A reverse, 59-TGATGTCGGATGTAGTGTTG- RNA interference GAA-39; human SAR1B forward, 59-TACAGTGGTTTCAGCAGTGTG-39; human SAR1B reverse, 59-AGTGGGATGTAATGTTGGGACA-39; human Both short-hairpin RNA (shRNA) and small-interfering RNA (siRNA) were SEC13 forward, 59-ACAACGCTCACACCATTGGCTG-39; human SEC13 used to knock down . pSuper.Retro-RNA interference 9 9 SEC23A plasmids were constructed according to the manufacturer’s instructions reverse, 5 -GTTGTCACAGCCACCTGAT-GCA-3 ; human forward, 9 9 SEC23A (Oligoengine). The following sequences were targeted for human YIPF5 5 -GGAGTCCGATTTAGTTGGAATGT-3 ; human reverse, 9 9 SEC23B 9 mRNA: 1) 59-GCGAACACTTACTTACATA-39 and 2) 59-GCGAACAC- 5 -AGGTCTCTCTTTCAGTGGTGT-3 ; human forward, 5 -ATGG- 9 SEC23B 9 TTACTTACATA-39. Establishment of shRNA-transduced stable cell lines CGACATACCTGGAGTTC-3 ; human reverse, 5 -GAAGGCCA- 9 SEC31A 9 was previously described (22). Chemically synthesized siRNA duplexes CACGTTCCAACTA-3 ; human forward, 5 -CAGGTGGACAAG- AACTCTTGGC-39; human SEC31A reverse, 59-GACTCAACAATCTCT- were obtained from Guangzhou RiboBio. Twenty-five nanomolar siRNA TTCCAGTTC-39; human SEC31B forward, 59-CTGCTGAAGCAAACA- was introduced into cells using GenMute (SignaGen) according to the 9 SEC31B 9 manufacturer’s instructions. After 6 h, the medium was replaced by fresh CAGGAGC-3 ; human reverse, 5 -CTCTCCAGTTCTTCAGGC- 9 Gapdh 9 medium, and cells were further incubated for 42 h. For siRNA pools, a TACAG-3 ; mouse forward, 5 -ACGGCCGCATCTTCTTGTGCA- 9 Gapdh 9 9 total of 50 nM siRNA was introduced into cells. The sequences of each 3 ;mouse reverse, 5 -ACGGCCAAATCCGTTCACACC-3 ;mouse Ifnb1 9 9 Ifnb1 siRNA oligonucleotide used in this study are as follows: control, 59-TT- forward, 5 -TCCTGCTGTGCTTCTCCACCACA-3 ; mouse reverse, 59-CCTCAGCCCCCTCTGGGGTC-39;mouseIfit1 forward, 59-CAG- CTCCGAACGUGUCACGT-39; human YIPF5 no. 1, 59-TGGCAAGTG- CAACCATGGGAGAGAATGCTG-39;mouseIfit1 reverse, 59-ACGTAGG- TCCTTGGATA-39,andno.2,59-CGGGATCAGTGCAATTGGA-39;murine 9 Il6 9 Yipf5 no. 1, 59-CCAACAACACCTCAGCCAT-39,andno.2,59-GGGCAG- CCAGGAGGTTGTGCAT-3 ;mouse forward, 5 -TATGAAGTTCCTCT- 9 Il6 9 9 ATTTACCAGCCAA-39; human SAR1A no. 1, 59-GAACAGATGCAATCA- CTGCAAGAGA-3 ;mouse reverse, 5 -TAGGGAAGGCCGTGGTT-3 ; Yipf5 9 9 GTGA-39,andno.2,59-CCAGTATATTGACTGATGT-39; SAR1B no. 1, 59- mouse forward, 5 -GAAGTGGAGGACCCTACAGCAA-3 ; mouse Yipf5 9 9 Sec13 GCATAACTTGAATTCAATA-39,andno.2,59-CTACCTTCCTGCTATCA- reverse, 5 -TGCCCAGTGTATGTCTGCTGTG-3 ;mouse forward, 59-AGGAGCAGAAGCTAGAGGCACA-39; and mouse Sec13 AT-39; SEC13 no. 1, 59-AGGAGGAGCAGAAGCTAGA-39,andno.2,59- reverse, 59-ATTGCCTGAGGCATCGT-CACAG-39. CCTATTACCGGAAAGTCAT-39; SEC23A no. 1, 59-GTTATGCTGGTAT- ATCTGA-39, and no. 2, 59-GCATAATGCTCCAATTCCT-39; SEC23B no. 1, Virus manipulation 59-CACGTTACATCAACACGGA-39, and no. 2, 59-CACTATGAGATGCTT- GCTA-39; SEC31A no. 1, 59-CCATAGCAGGTGGACAAGAAC-39, and no. 2, Viral infection was performed when cells were 70% confluent. The culture 59-TGCCAAGAGTTCAAACTCA-39; SEC31B no. 1, 59-GCCTGTGTTGT- medium was replaced by serum-free medium, and then HSV-1 or VSV GCAAAAGA-39,andno.2,59-CAGCGTCTGGAGTATCTAT-39;andmurine were added into the medium at various multiplicities of infection (MOI) The Journal of Immunology 3 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 1. Identification of YIPF5 as a positive regulatorofcGAS–STING–mediatedsignaling.(A) Effects of YIPF members on cGAS–STING– mediated activation of the IFN-b promoter. HEK293T cells were transfected with the IFN-b promoter luciferase plasmid and plasmids of cGAS, STING, and the indicated YIPF members. Twenty hours after transfection, luciferase assays were performed. (B) Effects of YIPF5 and YIPF2 on cGAS–STING–mediated activation of the IFN-b promoter, ISRE, and NF-kB reporter. The experiments were performed as in (A). (C and D)Effects of YIPF5 and YIPF2 on SeV- and poly(I:C)-triggered activation of the IFN-b promoter (C) and HEK293–TLR3 cells (D). (E and F)qPCRanalysisof mRNA levels of the indicated genes in WT and Yipf52/2 L929 cells. Cells were transfected with VACV70 or infected with SeV for 8 h. Graphs show mean 6 SD. n =3.*p , 0.05, **p , 0.01. according to the specific experiments. After 1 h of incubation, the Statistical analysis medium was replaced with normal medium. The titers of HSV-1 and VSV in supernatants were calculatedbystandardplaqueassays(22). Unpaired Student t tests were used for statistical analysis with Graph Pad HSV-1 (GFP and luciferase) replication in HFF cells was visualized by Prism software. Fluorescence imaging analysis was performed in a blinded fluorescent microscopy or determined by luciferase assay. fashion.

Immunofluorescent staining and confocal microscopy Results HeLa cells were transfected with the indicated plasmids by Lipofectamine YIPF5 positively regulates DNA virus-induced signaling 2000 (Invitrogen). Twenty-four hours after transfection, the cells were fixed with 4% paraformaldehyde for 10 min at room temperature and were Because YIPF members play roles in ER–Golgi transport, we wondered permeabilized in 0.1% Triton X-100 for 15 min on ice. Fixed and per- whether YIPF members are involved in STING-mediated innate meabilized cells were preincubated with 1% BSA in PBS before incubation immune signaling by regulating STING trafficking from the ER to with primary Abs. Cells were then incubated with secondary Abs conju- gated with Alexa Fluor 488/555/647 (Thermo Fisher Scientific). Nuclei Golgi. To do this, we first examined the effects of YIPF members were stained with DAPI. The cells were observed with an Olympus confocal on cGAS–STING–induced IFN-b promoter activation by the re- microscope under a 60-fold oil objective. porter assays. The results showed that overexpression of YIPF5 4 YIPF5 FACILITATES STING TRANSLOCATION AND ACTIVATION Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 2. Knockdown of YIPF5 impairs DNA virus-induced innate immune response. (A) qPCR analysis of mRNA levels of the indicated genes in YIPF5 stable-knockdown THP-1 cells. Cells were infected with HSV-1 or transfected with VACV70 for 8 h. (B and C) qPCR analysis of mRNA levels of the indicated genes in L929 cells transfected with control or Yipf5-siRNA. Cells were infected with HSV-1, transfected with VACV70 (B), or infected with SeV (C) for 8 h. (A and B) Graphs show mean 6 SD. n =3.(D) Immunoblot analysis of the indicated proteins in YIPF5-knockdown L929 cells. L929 cells were transfected with siRNA for 48 h and then were treated with transfected VACV70 for the indicated times. Native PAGE was used to detect IRF3 dimerization. The asterisk (*) indicates nonspecific bands. (E) qPCR analysis of mRNA levels of the indicated genes in BMDMs and primary MEFs as in (B). (F) Effects of YIPF5 depletion on cGAMP-induced expression of IFN-b. L929 and primary MEF cells were transfected with control or Yipf5-siRNA for 48 h, and then the cells were treated with cGAMP (100 nM) in digitonin permeabilization solution for 30 min, followed by another 4-h incubation in regular media before qPCR was performed. (E and F) Graphs show mean 6 SD. n =3.(A and B)*p , 0.05, **p , 0.01. (E and F)**p , 0.01. potentiated, whereas YIPF2 inhibited, cGAS–STING–mediated knockdown of YIPF5 markedly diminished transfected DNA- activation of the IFN-b promoter (Fig. 1A). By siRNA-mediated induced IFN-b expression, with iRhom2–siRNA as a positive knockdown of endogenous YIPF members, we found that only control (Supplemental Fig. 1A). In addition, YIPF5 dose-dependently The Journal of Immunology 5 potentiated cGAS–STING–mediated IFN-b promoter activation IFIT1,andIL-6 by transfected dsDNA (Fig. 2A). In similar as well as ISRE reporter and NF-kB reporter activation, rep- experiments, knockdown of Yipf5 decreased the mRNA levels resentingIRF3andNF-kB activation, respectively, which work of Ifnb1, Ifit1, and Il6 induced by HSV-1 infection or transfected cooperatively to activate the IFN-b promoter (Fig. 1B). In the same dsDNA in L929 cells (Fig. 2B, Supplemental Fig. 1E) but not experiment, YIPF2 inhibited cGAS–STING–mediated signal- SeV-induced Ifnb1 transcription (Fig. 2C). TBK1 phosphory- ing. However, YIPF5 had no obvious effects on SeV- and lation and IRF3 dimerization induced by cytoplasmic dsDNA poly(I:C)-induced activation of the IFN-b promoter, which triggered were also markedly inhibited by YIPF5 depletion (Fig. 2D). In RIG-I– and TLR3-mediated signaling, respectively (Fig. 1C, 1D). addition, knockdown of YIPF5 also inhibited transfected dsDNA- Compared with the vector control, only cells that stably express induced of p65, indicating that YIPF5 was also YIPF5 but not YIPF2 increased the transcription of Ifnb1, Ifit1, required for STING-mediated NF-kB activation (Supplemental and Il6 stimulated by HSV-1 or transfected dsDNA (Supplemental Fig. 1F, 1G). Furthermore, we examined the function of YIPF5 Fig. 1B). The discrepancy between Fig. 1B and Supplemental in primary BMDMs and MEFs and found that the results in pri- Fig. 1B is probably due to artificial effects of YIPF2 on cGAS– mary cells are consistent with the observation in immortalized STING–mediated signaling when expressed at high levels. The THP-1 and L929 cells (Fig. 2E). YIPF5 knockdown also impaired transcription of Ifnb1, Ifit1,andIl6 stimulated by HSV-1 or cGAMP-stimulated induction of Ifnb1 (Fig. 2F). Collectively, transfected dsDNA was markedly impaired in Yipf52/2 cells these results suggest that YIPF5 is essential for innate immune compared with wild-type (WT) cells, which further confirmed response to DNA virus both in human and mouse immortalized its function (Fig. 1E, 1F, Supplemental Fig. 1C). Taken together, cell lines and primary cells. these results suggest that YIPF5 specifically regulates DNA Downloaded from YIPF5 is essential for cellular antiviral response to HSV-1 virus-induced type I IFN production. Because YIPF5 is important for viral infection–induced expression Knockdown of YIPF5 impairs DNA virus-induced innate of the type I IFNs and its downstream IFN-stimulated genes, immune responses we wondered whether YIPF5 contributes to cellular antiviral To extensively confirm the role of YIPF5 in DNA virus-induced responses. Consistent with decreased type I IFN production,

innate immune responses, we performed a series of experiments knockdown of YIPF5 significantly enhanced HSV-1 replica- http://www.jimmunol.org/ to explore the function of endogenous YIPF5 by RNA interference- tion in THP-1 and L929 cells (Fig. 3A). In contrast, knock- mediated knockdown. First, we found that knockdown of YIPF5 down of YIPF5 had no obvious effects on VSV replication in impaired IFNB1, IFIT1,andIL6 induction by HSV-1 infection THP-1 and L929 cells (Fig. 3B). In addition, using GFP- and in THP-1 cells (Fig. 2A, Supplemental Fig. 1D). In addition, luciferase-expressed HSV-1, we confirmed that knockdown of knockdown of YIPF5 also impaired the induction of IFNB1, YIPF5 impaired cellular antiviral response in HFFs (Fig. 3C, 3D). by guest on September 26, 2021

FIGURE 3. Knockdown of YIPF5 attenuates antiviral response. (A) Knockdown of YIPF5 potentiates HSV-1 replication. THP-1 stably expressing control or YIPF5-shRNA or L929 cells transfected with siRNA were infected with HSV-1 (MOI = 0.1), and the supernatants were harvested 24 h post- infection, followed by analysis of HSV-1 production with standard plaque assay. (B) Knockdown of YIPF5 has no obvious effects on VSV replication. The experiments were similarly performed as in (A) (VSV, MOI = 0.1). (C and D) L929 cells transfected with siRNA were infected with HSV-1 (GFP and luciferase; MOI = 0.1) for 24 h (C) or the indicated times (D) before virus replication was examined by flow cytometry (C) and luciferase assay (D). Numbers adjacent to the outlined areas indicate percentages of GFP+ cells (C). Graphs show mean 6 SD. n =3.*p , 0.05, **p , 0.01. 6 YIPF5 FACILITATES STING TRANSLOCATION AND ACTIVATION

These observations suggest that YIPF5 is important for cellular that their membrane localization is important for the YIPF5–STING antiviral responses against DNA but not RNA viruses. interaction. YIPF5 is associated with STING YIPF5 is essential for STING trafficking We next determined whether YIPF5 is associated with signaling To explore the mechanism by which YIPF5 regulates STING- components in DNA virus-triggered pathways. Transient trans- mediated signaling, we first examined the STING dimerization fection and coimmunoprecipitation experiments revealed that and oligomerization, hallmarks of its activation. The results showed YIPF5 was associated with STING but not with cGAS and TBK1 that YIPF5 had no effects on transfected dsDNA-induced STING (Fig. 4A). The results agree with those of Fig. 2F, in which knock- dimerization or oligomerization (Supplemental Fig. 1H). YIPF5 down of YIPF5 inhibited cGAMP-induced induction of IFN-b, displayed discrete spots, with the highest density at the peri- suggesting that YIPF5 probably functions at STING level in nuclear region, which fits the feature of ERES (Supplemental cGAS–STING–mediated IFN induction signaling. In coimmuno- Fig. 2A, 2B). A little moiety of YIPF5 was colocalized with precipitation experiments of endogenous proteins, we found that cis-Golgi marker GM130 but not with trans-Golgi marker TGN38 YIPF5 was weakly associated with STING in unstimulated cells, (Supplemental Fig. 2A). These observations are consistent with a and this association was enhanced by stimulation with transfected previous report that YIPF5 was localized around the Golgi stacks dsDNA (Fig. 4B). Confocal microscopy also revealed that a small and revealed a high degree of colocalization with SEC31A, a moiety of YIPF5 was colocalized with STING after stimulation protein known to be enriched at ERES (24). Reports also indicate with transfected dsDNA (Fig. 4C). Similar to STING, YIPF5 is also that YIPF5 regulates ER–Golgi traffic by COPII-mediated ER ex- a membrane protein and contains five transmembrane domains. To port at ERES (24), so we investigated whether YIPF5 is involved in Downloaded from investigate the domains responsible for YIPF5–STING interaction, STING translocation. YIPF5 depletion markedly inhibited STING we performed coimmunoprecipitation experiments with a series of translocation to perinuclear puncta induced by transfection of truncations of YIPF5 and STING. We found that the C-terminal dsDNA (Fig. 5A, Supplemental Fig. 2B, 2C). In the same experi- transmembrane domains of YIPF5 could interact with STING ments, STING translocation was almost blocked by brefeldin A (Fig. 4D), and STING containing the fourth transmembrane do- (BFA) treatment because of Golgi stack collapse. Notably, depletion

main could interact with YIPF5 (Fig. 4E). These results suggest of YIPF5 in these experiments did not obviously change ER/Golgi http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 4. YIPF5 is associated with STING. (A) HEK293T cells were transfected with the indicated plasmids before coimmunoprecipitation and immunoblot analysis were performed with the indicated Abs. (B) L929 cells were treated by transfection of VACV70 at the indicated times, and then coimmunoprecipitation and immunoblot analysis were performed with the indicated Abs. (C) Immortalized Sting2/2 MLFs reconstituted with 33 Flag-tagged murine STING were untreated or stimulated with transfected VACV70 for 1 h. The cells then were fixed, permeabilized, and stained for YIPF5 (anti-YIPF5 Ab, green) and STING (anti-Flag Ab, red). Nuclei were stained with DAPI (blue). Original magnification 360. Magnified images of the boxed areas are shown. The arrows indicate the colocalization spots. (D and E) Domain mapping of YIPF5–STING association. The upper panels showed schematic representations of YIPF5 (C) and STING (D) truncations, respectively. The experiments were performed as in (A). The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 5. YIPF5 is essential for the trafficking of STING. (A) Detection of the subcellular localization of STING. Immortalized Sting2/2 MLFs reconstituted with 33 Flag-tagged murine STING (mSTING-3F) were pretreated with or without BFA (0.3 mg/ml) for 30 min and stimulated with transfected VACV70 for 2 h. The cells then were fixed, permeabilized, and stained for STING (anti-Flag Ab, red) and GM130. Nuclei were stained with DAPI (blue). Magnified images of the boxed areas are shown in lower panels (original magnification 360). The per- centage of cells with obvious perinuclear puncta was calculated by analyzing ∼100 cells. The error bar shows the SD of two biological repeats. (B–D) Effects of YIPF5 on the association of STING with TBK1 (B), IRF3 (C), and STING itself (D). Coimmunoprecipitation and immunoblot analysis were performed with the indicated Abs. (E) Effects of YIPF5 on endogenous association of STING with TBK1 and SEC31A. Sting2/2 MLFs reconstituted with mSTING-3F were transfected with control or Yipf5-siRNA. Forty-eight hours later, cells were unstimulated or stim- ulated with transfected VACV70 for 2 h before coimmunoprecipitation and immunoblot analysis were performed. (F) YIPF5 is associated with SAVI variants of STING. HEK293T cells were transfected with the indicated plasmids before coimmunoprecipitation and immunoblot analysis were performed with the indicated Abs. (G) Effects of YIPF5 on STING mutant–mediated activation of the IFN-b promoter. HEK293T cells were transfected with the IFN-b promoter luciferase plasmid and the indicated plasmids. (H) Effects of YIPF5 knockdown on STING mutant–mediated activation of the IFN-b promoter. HEK293T cells were transfected with control or YIPF5-siRNA. Thirty hours later, cells were retransfected with the IFN-b promoter luciferase plasmid and the indicated plasmids for 24 h before luciferase assays were performed. Graphs show mean 6 SD. n =3. *p , 0.05, **p , 0.01. morphology (Supplemental Fig. 2C, 2D). STING trafficking is es- ER exit (13, 25). In this study, we found that YIPF5 also interacted sential for recruitment of downstream TBK1 and IRF3. We found that with SAVI mutants and potentiated SAVI mutant–mediated IFN-b overexpression of YIPF5 enhanced STING interaction with TBK1 production when overexpressed, whereas knockdown of YIPF5 and IRF3 but not STING itself (Fig. 5B–D). In contrast, knockdown impaired SAVI mutant–mediated IFN-b production (Fig. 5F–H), ofYIPF5impairedtherecruitmentofTBK1bySTINGinthe indicating that YIPF5 is also involved in constitutive activation of presence of dsDNA (Fig. 5E). Research indicates that STING– STING–SAVI mutants. These results suggestthatYIPF5isimportant SAVI mutants induce STING-mediated signaling by constitutive for STING trafficking and STING-mediated signal transduction. 8 YIPF5 FACILITATES STING TRANSLOCATION AND ACTIVATION

YIPF5 recruits STING to COPII-coated vesicles dsDNA (Fig. 6D). These results suggest that YIPF5 interacts with Previous studies demonstrated that several proteins are involved STING at COPII-coated vesicles. We next explored STING as- in regulating STING trafficking, including TRAPb, Sec61b, Sec5, sociation with the COPII complex components in immortalized 2/2 3 and iRhom2 (6, 12, 18). However, we found that YIPF5 neither Sting MLFs reconstituted with 3 Flag-tagged mouse STING, was associated with these proteins (Supplemental Fig. 3A) nor af- which produced comparable levels of STING protein to those fected the association of STING with these proteins (Supplemental of WT MLFs (Supplemental Fig. 3D). STING interacted with Fig. 3B). The COPII complex, initiated by the small GTPase SAR1 SEC31A after dsDNA stimulation, which was disturbed by deple- on the ER membrane, consists of an inner and an outer layer that tion of YIPF5 (Fig. 5E). BFA treatment, which causes Golgi col- are made up of SEC23–SEC24 heterodimers and SEC13–SEC31 lapse, also disrupted the interaction of STING with SEC31A and heterotetramers, respectively (17). Studies indicate that YIPF5 SEC13 (Fig. 6E). These results suggest that YIPF5 recruits STING is recruited to COPII-coated vesicles and potentially assists their to COPII-coated vesicles, facilitating STING trafficking from ER fusion with their target membrane (24, 26). Our results also con- to Golgi. firmed the colocalization of YIPF5 with the COPII complex com- ponent SEC31A by confocal microscopy (Supplemental Fig. 3C). COPII complex is involved in STING-mediated signaling Furthermore, transient transfection and coimmunoprecipitation ex- We next explored the roles of COPII-coated vesicles in STING- periments revealed that YIPF5 was associated with COPII complex mediated innate immune response to intracellular DNA. We components SEC23B and SEC31A but not SEC13 and SEC24D designed siRNAs to knock down key components of the COPII (Fig. 6A). Interestingly, we found that STING also interacted with complex. Depletion of SAR1 (simultaneous knockdown of SAR1A SEC23B and SEC31A, although their associations were relatively and SAR1B), SEC13, SEC23 (simultaneous knockdown of Downloaded from weaker (Fig. 6B). YIPF5 could also enhance the association of STING SEC23A and SEC23B), and SEC31 (simultaneous knockdown with SEC23B and SEC31A (Fig. 6C). In addition, confocal micros- of SEC31A and SEC31B) inhibited HSV-1– and transfected dsDNA– copy showed that STING was partially colocalized with endoge- triggered induction of IFNB1 but did not inhibit SeV-triggered nous SEC31A and SEC13 only when stimulated with transfected induction of IFNB1 (Fig. 7A). Consistently, depletion of SEC13 http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 6. YIPF5 facilitates the association of STING with COPII. (A–C) HEK293T cells were transfected with the indicated plasmids before coimmunoprecipitation and immunoblot analysis were performed with the indicated Abs. (D) Immortalized Sting2/2 MLFs reconstituted with 33 Flag-tagged murine STING (mSTING-3F) were untreated or stimulated with transfected VACV70 for 1 h. The cells then were fixed, permeabilized, and stained for STING (anti-Flag Ab, green), Sec31A (anti-Sec31A Ab, red), and Sec13 (anti-Sec13 Ab, red). Nuclei were stained with DAPI (blue). Magnified images of the boxed areas are shown (original magnification 360). The arrows indicate the colocalization spots. (E) Immortalized Sting2/2 MLFs reconstituted with mSTING-3F were pretreated with or without BFA (0.3 mg/ml) for 30 min and stimulated with transfected VACV70 for 2 h as indicated. Coimmunoprecipitation and immunoblot analysis were performed with the indicated Abs. IgG-H, H chains of IgG. The Journal of Immunology 9 Downloaded from http://www.jimmunol.org/

FIGURE 7. COPII is essential for STING-mediated signaling. (A) siRNA pools of SAR1A plus SAR1B, SEC13, SEC23A plus SEC23B, and SEC31A plus SEC31B were transfected into HFFs to knock down the expression of SAR1, SEC13, SEC23, and SEC31. Cells were then stimulated by HSV-1 by guest on September 26, 2021 infection, VACV70 transfection, or SeV infection as indicated. The transcription of IFNB1 and knockdown efficiencies of each gene were analyzed by qPCR. Graphs show mean 6 SD. n =3.*p , 0.05, **p , 0.01. (B and C) Immunoblot analysis of the indicated proteins in SEC13-knockdown HFFs. HFFs were transfected with siRNA for 48 h and then were infected with HSV-1 (B) and SeV (C) for the indicated times. (D) Effects of SEC13-siRNA on the subcellular localization of STING. The experiments were performed as in Fig. 4J. Original magnification 360. The percentage of cells with obvious perinuclear puncta was calculated by analyzing ∼100 cells. The error bar shows the SD of two biological repeats. Original magnification 360. Graphs show mean 6 SD. n =2.*p , 0.05, **p , 0.01. (E) Effects of YIPF5 on STING-dependent autophagy. HFFs transfected with control or YIPF5-siRNA were stimulated with transfected VACV70 for the indicated times before immunoblot analysis was performed. inhibited HSV-1–induced phosphorylation of STING, TBK1, and and autoimmune diseases. A thorough investigation of STING IRF3 (Fig. 7B), which are key events of DNA virus-induced innate trafficking regulation is important for the development of treat- immune responses. However, depletion of SEC13 did not affect ment strategies for related diseases. In this study, we found that SeV-induced phosphorylation of TBK1 and IRF3 (Fig. 7C). Moreover, YIPF5 positively regulates STING-mediated innate immune re- knockdown of SEC13 markedly inhibited transfected DNA-induced sponses by recruiting STING to COPII-coated vesicles and facili- STING trafficking to perinuclear puncta (Fig. 7D). Recently, it tating STING trafficking from the ER to Golgi, which provides is reported that STING also activates autophagy during COPII- important insights into the molecular mechanisms of intracellular dependent STING translocation (27). Because YIPF5 is also essential DNA-stimulated STING trafficking and activation. for STING translocation, we wondered whether YIPF5 affects There are several lines of evidence to support the notion that YIPF5 STING-dependent autophagy. We found that knockdown of YIPF5 is essential for STING-mediated type I IFN production triggered by inhibited LC3 lipidation induced by transfected dsDNA, suggesting cytosolic DNA. First, knockdown of YIPF5 impaired IFN production that YIPF5 is involved in STING-dependent autophagy activa- induced by DNA virus and transfected dsDNA. Consistently, the tion (Fig. 7E). Taken together, these results suggest that the COPII antiviral responses to DNAvirus were impaired by YIPF5 depletion. complex is important for STING trafficking and STING-mediated Second, YIPF5 interacted with STING, which was enhanced by signaling. stimulation with dsDNA. Notably, in a large-scale study of innate immunity interactome for type I IFN production signaling by mass Discussion spectrometry, YIPF5 was identified as an interacting protein of STING trafficking is essential for STING-mediated signaling, STING, which supplies another support for our results (28). Third, which is involved in a variety of immune surveillances, such as YIPF5 enhanced TBK1 and IRF3 recruitment to STING, which is viral infection, DNA damage, mitochondria damage, senescence, necessary for STING-mediated signaling transduction. and tumor immunology. Clinical and genetic studies indicate that DNA stimulation–triggered STING trafficking and recruitment disordered STING translocation contributes to autoinflammatory of downstream signaling molecules are tightly regulated in a 10 YIPF5 FACILITATES STING TRANSLOCATION AND ACTIVATION spatiotemporal manner. STING is translocated across multiple needs to be clarified. Further exploring the details of these processes membrane compartments including the ER, ERGIC, Golgi, endo- by more sensitive and quantitative techniques would be of great some, and perinuclear microsome or punctate structure. The ER is interest, as it would provide new clues for understanding signaling the start site of STING translocation after its activation, but how transduction and membrane-associated trafficking and would be STING exits from the ER remains mysterious. In this study, we helpful for developing new strategies to modulate STING-related found that YIPF5 interacts with STING and facilitates STING diseases. trafficking by recruiting STING to COPII-coated vesicles, which are adjacent to ERES. First, our results and previous studies sug- Acknowledgments gest that YIPF5 is associated with components of COPII at ERES, We acknowledge Ding Gao of the Core Facility and Technical Support suggesting a role of YIPF5 in anterograde transport from the ER to Facility of the Wuhan Institute of Virology for technical assistance. Golgi at ERES (24). Indeed, it is reported that YIPF5 is responsible for generating the ER-derived at ERES during Brucella Disclosures infection (29). Second, YIPF5 was colocalized and interacted with The authors have no financial conflicts of interest. STING at COPII-coated vesicles, and their association was enhanced following DNA stimulation. Third, YIPF5 promoted the association of STING with the COPII complex. In addition, References knockdown of key components of COPII inhibited STING- 1. Takeuchi, O., and S. Akira. 2010. Pattern recognition receptors and inflamma- tion. Cell 140: 805–820. mediated innate immune response to DNA viruses. Consistent 2. Akira, S., S. Uematsu, and O. Takeuchi. 2006. Pathogen recognition and innate with our findings, recent studies also have shown that other factors immunity. Cell 124: 783–801. Downloaded from of the COPII complex are involved in STING trafficking and ac- 3. Burdette, D. L., and R. E. Vance. 2013. STING and the innate immune response to nucleic acids in the . Nat. Immunol. 14: 19–26. tivation (27, 30, 31). Together, these findings suggest that COPII 4. Ran, Y., H. B. Shu, and Y. Y. Wang. 2014. MITA/STING: a central and plays important roles in STING translocation. In this study, multifaceted mediator in innate immune response. Growth Factor Rev. 25: 631–639. we further identified YIFP5 as a crucial manipulator of COPII- 5. Wu, J., and Z. J. Chen. 2014. Innate immune sensing and signaling of cytosolic mediated STING trafficking, providing a new insight into the nucleic acids. Annu. Rev. Immunol. 32: 461–488. 6. Ishikawa, H., Z. Ma, and G. N. Barber. 2009. STING regulates intracellu- process of STING trafficking. http://www.jimmunol.org/ lar DNA-mediated, type I interferon-dependent innate immunity. Nature 461: Whether STING translocation is essential for STING-mediated 788–792. NF-kB activation is not clear yet. Reports indicate that ubiq- 7. Luo, W. W., and H. B. Shu. 2018. Delicate regulation of the cGAS-MITA- uitination on K224 of STING is required for STING translocation mediated innate immune response. Cell. Mol. Immunol. 15: 666–675. 8. Li, X. D., J. Wu, D. Gao, H. Wang, L. Sun, and Z. J. Chen. 2013. Pivotal roles and IRF3 activation but not NF-kB activation (32). STING-K224 of cGAS-cGAMP signaling in antiviral defense and immune adjuvant effects. ubiquitination is blocked by BFA treatment, suggesting that the Science 341: 1390–1394. 9. Gao, D., J. Wu, Y. T. Wu, F. Du, C. Aroh, N. Yan, L. Sun, and Z. J. Chen. 2013. STING ubiquitination on K224 probably occurs during the trans- Cyclic GMP-AMP synthase is an innate immune sensor of HIV and other portation process from the ER to the Golgi (32). In addition, we retroviruses. Science 341: 903–906. found that BFA treatment only inhibited IRF3 but not NF-kB ac- 10. Zhong, B., Y. Yang, S. Li, Y. Y. Wang, Y. Li, F. Diao, C. Lei, X. He, L. Zhang, P. Tien, and H. B. Shu. 2008. The adaptor protein MITA links virus-sensing by guest on September 26, 2021 tivation; however, knockdown of YIPF5 inhibited both IRF3 and receptors to IRF3 activation. Immunity 29: 538–550. NF-kB activation. The simplest explanation for these phenomena is 11. Sun, W., Y. Li, L. Chen, H. Chen, F. You, X. Zhou, Y. Zhou, Z. Zhai, D. Chen, that STING-mediated IRF3 and NF-kB activation branches after and Z. Jiang. 2009. ERIS, an endoplasmic reticulum IFN stimulator, activates innate immune signaling through dimerization. Proc. Natl. Acad. Sci. USA 106: the STING–ER exit and before the stage disturbed by BFA during 8653–8658. the transportation process from the ER to the Golgi because YIPF5 12. Ishikawa, H., and G. N. Barber. 2008. STING is an endoplasmic reticulum adaptor that facilitates innate immune signalling. [Published erratum appears in functions at the initial stage of STING translocation (the ER exit). 2008 Nature 456: 274.] Nature 455: 674–678. These findings provide clues to reveal the mechanism of STING- 13. Dobbs, N., N. Burnaevskiy, D. Chen, V. K. Gonugunta, N. M. Alto, and N. Yan. mediated NF-kB activation. 2015. STING activation by translocation from the ER is associated with infec- tion and autoinflammatory disease. Cell Host Microbe 18: 157–168. It is reported that YIPF5 plays roles in maintaining the integrity 14. Saitoh, T., N. Fujita, T. Hayashi, K. Takahara, T. Satoh, H. Lee, K. Matsunaga, of Golgi and ER (26, 33). Knockdown of YIPF5 caused partial S. Kageyama, H. Omori, T. Noda, et al. 2009. Atg9a controls dsDNA-driven disassembly of the Golgi apparatus and led to restructuring of the dynamic translocation of STING and the innate immune response. Proc. Natl. Acad. Sci. USA 106: 20842–20846. ER network into micrometer-sized concentric whorls (26, 33). 15. Chen, W., S. Li, H. Yu, X. Liu, L. Huang, Q. Wang, H. Liu, Y. Cui, Y. Tang, However, it is also reported that knockdown of YIPF5 did not P. Zhang, and C. Wang. 2016. ER adaptor SCAP translocates and recruits IRF3 to perinuclear microsome induced by cytosolic microbial DNAs. PLoS Pathog. induce morphological changes in the Golgi, ER, or ERGIC (34). 12: e1005462. Dykstra et al. (33) explained that the extent of YIPF5 depletion 16. Kuehn, M. J., J. M. Herrmann, and R. Schekman. 1998. COPII-cargo interac- by distinct siRNAs probably determines its effects on ER/Golgi tions direct protein sorting into ER-derived transport vesicles. Nature 391: 187– 190. morphology, and they found that low levels of YIPF5 are able to 17. Zanetti, G., K. B. Pahuja, S. Studer, S. Shim, and R. Schekman. 2011. COPII and maintain a relatively normal ER network. In our experiments, we the regulation of protein sorting in . [Published erratum appears in found that knockdown of YIPF5 did not cause obvious changes 2012 Nat. Cell Biol. 14: 221.] Nat. Cell Biol. 14: 20–28. 18. Luo, W. W., S. Li, C. Li, H. Lian, Q. Yang, B. Zhong, and H. B. Shu. 2016. to ER/Golgi morphology, which was probably due to small amounts iRhom2 is essential for innate immunity to DNA viruses by mediating trafficking of YIPF5 existing in siRNA-transfected cells. However, STING- and stability of the adaptor STING. Nat. Immunol. 17: 1057–1066. 19. Lian, H., J. Wei, R. Zang, W. Ye, Q. Yang, X. N. Zhang, Y. D. Chen, Y. Z. Fu, mediated signaling transduction was markedly impaired simulta- M. M. Hu, C. Q. Lei, et al. 2018. ZCCHC3 is a co-sensor of cGAS for dsDNA neously. These observations suggest that the effects of YIPF5 on recognition in innate immune response. Nat. Commun. 9: 3349. STING trafficking are unlikely to be due to morphological changes 20. Fu, Y. Z., S. Su, Y. Q. Gao, P. P. Wang, Z. F. Huang, M. M. Hu, W. W. Luo, S. Li, M. H. Luo, Y. Y. Wang, and H. B. Shu. 2017. Human cytomegalovirus tegument in ER/Golgi caused by YIPF5 depletion. protein UL82 inhibits STING-mediated signaling to evade antiviral immunity. Interestingly, YIPF5 is also important for activating disease- Cell Host Microbe 21: 231–243. associated variants of STING presented in patients exhibiting 21. Huang, Z. F., H. M. Zou, B. W. Liao, H. Y. Zhang, Y. Yang, Y. Z. Fu, S. Y. Wang, M. H. Luo, and Y. Y. Wang. 2018. Human cytomegalovirus protein UL31 autoinflammatory vasculopathy and autoimmunity (13, 25, 35), inhibits DNA sensing of cGAS to mediate immune evasion. Cell Host raising a question of whether YIPF5 is involved in these auto- Microbe 24: 69–80.e4. 22. Ran, Y., J. Zhang, L. L. Liu, Z. Y. Pan, Y. Nie, H. Y. Zhang, and Y. Y. Wang. inflammatory or autoimmune diseases. In addition, the relationship 2016. Autoubiquitination of TRIM26 links TBK1 to NEMO in RLR-mediated between YIPF5 and Sec61, TRAPb, and Sec5 in STING trafficking innate antiviral immune response. J. Mol. Cell Biol. 8: 31–43. The Journal of Immunology 11

23. Motani, K., S. Ito, and S. Nagata. 2015. DNA-mediated cyclic GMP-AMP activation of the IRE1 pathway of the unfolded protein response during Brucella synthase-dependent and -independent regulation of innate immune responses. infection. PLoS Pathog. 11: e1004747. J. Immunol. 194: 4914–4923. 30. Ogawa, E., K. Mukai, K. Saito, H. Arai, and T. Taguchi. 2018. The binding of 24. Tang, B. L., Y. S. Ong, B. Huang, S. Wei, E. T. Wong, R. Qi, H. Horstmann, and TBK1 to STING requires exocytic membrane traffic from the ER. Biochem. W. Hong. 2001. A membrane protein enriched in endoplasmic reticulum exit Biophys. Res. Commun. 503: 138–145. sites interacts with COPII. J. Biol. Chem. 276: 40008–40017. 31. Sun, M. S., J. Zhang, L. Q. Jiang, Y. X. Pan, J. Y. Tan, F. Yu, L. Guo, L. Yin, 25. Liu, Y., A. A. Jesus, B. Marrero, D. Yang, S. E. Ramsey, G. A. M. Sanchez, C. Shen, H. B. Shu, and Y. Liu. 2018. TMED2 potentiates cellular IFN responses K. Tenbrock, H. Wittkowski, O. Y. Jones, H. S. Kuehn, et al. 2014. Activated to DNA viruses by reinforcing MITA dimerization and facilitating its trafficking. STING in a vascular and pulmonary syndrome. N. Engl. J. Med. 371: 507–518. Cell Rep. 25: 3086–3098.e3. 26. Yoshida, Y., K. Suzuki, A. Yamamoto, N. Sakai, M. Bando, K. Tanimoto, 32. Ni, G., H. Konno, and G. N. Barber. 2017. Ubiquitination of STING at lysine Y. Yamaguchi, T. Sakaguchi, H. Akhter, G. Fujii, et al. 2008. YIPF5 and YIF1A 224 controls IRF3 activation. Sci. Immunol. 2: eaah7119. recycle between the ER and the Golgi apparatus and are involved in the main- 33. Dykstra, K. M., J. E. Pokusa, J. Suhan, and T. H. Lee. 2010. Yip1A structures the tenance of the Golgi structure. Exp. Cell Res. 314: 3427–3443. mammalian endoplasmic reticulum. Mol. Biol. Cell 21: 1556–1568. 27. Gui, X., H. Yang, T. Li, X. Tan, P. Shi, M. Li, F. Du, and Z. J. Chen. 2019. 34. Kano, F., S. Yamauchi, Y. Yoshida, M. Watanabe-Takahashi, K. Nishikawa, Autophagy induction via STING trafficking is a primordial function of the cGAS N. Nakamura, and M. Murata. 2009. Yip1A regulates the COPI-independent pathway. Nature 567: 262–266. retrograde transport from the Golgi complex to the ER. J. Cell Sci. 122: 2218– 28. Li, S., L. Wang, M. Berman, Y. Y. Kong, and M. E. Dorf. 2011. Mapping a 2227. dynamic innate immunity protein interaction network regulating type I interferon 35. Jeremiah, N., B. Neven, M. Gentili, I. Callebaut, S. Maschalidi, M. C. Stolzenberg, production. Immunity 35: 426–440. N. Goudin, M. L. Fre´mond, P. Nitschke, T. J. Molina, et al. 2014. Inherited STING- 29. Taguchi, Y., K. Imaoka, M. Kataoka, A. Uda, D. Nakatsu, S. Horii-Okazaki, activating mutation underlies a familial inflammatory syndrome with lupus-like R. Kunishige, F. Kano, and M. Murata. 2015. Yip1A, a novel host factor for the manifestations. J. Clin. Invest. 124: 5516–5520. Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

Supplementary figure 1. YIPF5 is involved in cytoplasmic DNA- and HSV-1- induced signaling. (A) L929 were transfected with pools of siRNA against the indicated genes. Forty-eight hours later, cells were transfected with VACV70 for 8 h. The transcription of Ifnb1 and knockdown efficiencies of each genes were analyzed by qPCR. (B) qPCR analysis of mRNA levels of the indicated genes in murine Yipf5- or Yipf2-stable-expressing L929 cells infected with HSV-1 or transfected with VACV70 for 8 h. (C-E) WB analysis of protein level of YIPF5 in Yipf5 knockout L929 cells generated by CRISPR/Cas9 (C), YIPF5 stable knockdown THP-1 cells (D), and L929 cells transfected with control or Yipf5-siRNA (E). (F) Knockdown of YIPF5 inhibits both IRF3 and NF-B activation. L929 cells were transfected with control or YIPF5- siRNA in L929 cells. Forty-eight hours later, cells were transfected with VACV70 for the indicated times. (G) BFA treatment inhibits IRF3 but not NF-B activation. L929 cells were pretreated with DMSO or BFA for 30 min before stimulating with transfected VACV70 for the indicated times. (H) The effects of YIPF5 on STING dimerization or oligomerization. HFFs stably-expressing YIPF5 or transfected with siRNA were stimulation with transfected VACV70 for the indicated times. STING dimerization or oligomerization was analyzed as described in Materials and Methods. *indicates non- specific bands. Graphs show mean ± SD. n=3.

Supplementary figure 2. The subcellular localization of YIPF5 and its effects on ER/Golgi morphology. All of experiments were performed in immortalized Sting-/- MLFs reconstituted with mSTING-3F. (A) Cells were untreated or stimulated with transfected VACV70 for 2 h, and then were fixed, permeabilized, and stained with the indicated antibodies. (B-D) Knockdown of YIPF5 did not obviously change Golgi/ER morphology. Cells were transfected with control or Yipf5-siRNA for 40 h. And then cells were untreated or stimulated with transfected VACV70 for 2 h. Cells were fixed, permeabilized, and stained with the indicated antibodies. The ER was visualized by staining with ER-tracker Green, and nuclei were stained with DAPI (blue). (B-D) are related to figure 5.

Supplementary figure 3. The association of YIPF5 with molecules in STING- mediated signaling. (A) YIPF5 did not interact with Sec5, TRAP, Sec61, and iRhom2. HEK293T cells were transfected with the indicated plasmids before co- immunoprecipitation and immunoblot analysis were performed with the indicated antibodies. (B) YIPF5 had no effects on the association of STING with Sec5, TRAP, Sec61, and iRhom2. HEK293T cells were transfected with the indicated plasmids before co-immunoprecipitation and immunoblot analysis were performed with the indicated antibodies. (C) YIPF5 is co-localized with SEC31A. Confocal microscopy of HeLa cells transfected with RFP-YIPF5 and HA-SEC31A. SEC31A was visualized by staining with anti-HA antibody. (D) Sting-/- MLFs reconstituted with 3×Flag-taged mouse STING produced comparable level of STING protein with wild-type MLFs. WB analysis of STING expression in wild-type MLF or Sting-/- MLF reconstituted with 3×Flag-tagged murine STING (mSTING-3F).