C-Src Suppresses Dendritic Cell Antitumor Activity Via T Cell Ig and Mucin Protein-3 Receptor

c-Src Suppresses Dendritic Cell Antitumor Activity via T Cell Ig and Mucin Protein-3 Receptor

This information is current as Ravindra Gujar, Neeraj Maurya, Vinod Yadav, Mamta of September 25, 2021. Gupta, Saurabh Arora, Neeraj Khatri and Pradip Sen J Immunol published online 20 July 2016 http://www.jimmunol.org/content/early/2016/07/19/jimmun ol.1600104 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2016/07/19/jimmunol.160010 Material 4.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 25, 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 © 2016 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published July 20, 2016, doi:10.4049/jimmunol.1600104 The Journal of Immunology

c-Src Suppresses Dendritic Cell Antitumor Activity via T Cell Ig and Mucin Protein-3 Receptor

Ravindra Gujar,* Neeraj Maurya,* Vinod Yadav,* Mamta Gupta,* Saurabh Arora,* Neeraj Khatri,† and Pradip Sen*

The enhanced expression of T cell Ig and mucin protein-3 (TIM-3) on tumor-associated dendritic cells (DCs) attenuates antitumor effects of DNA vaccines. To identify a potential target (or targets) for reducing TIM-3 expression on tumor-associated DCs, we explored the molecular mechanisms regulating TIM-3 expression. In this study, we have identiﬁed a novel signaling pathway (c-Src→Bruton’s tyrosine kinase→transcription factors Ets1, Ets2, USF1, and USF2) necessary for TIM-3 upregulation on DCs. Both IL-10 and TGF-b, which are produced in the tumor microenvironment, upregulated TIM-3 expression on DCs via this pathway. Suppressed expression of c-Src or downstream Bruton’s tyrosine kinase, Ets1, Ets2, USF1, or USF2 blocked

IL-10– and TGF-b–induced TIM-3 upregulation on DCs. Notably, in vivo knockdown of c-Src in mice reduced TIM-3 expression Downloaded from on tumor-associated DCs. Furthermore, adoptive transfer of c-Src–silenced DCs in mouse tumors enhanced the in vivo antitumor effects of immunostimulatory CpG DNA; however, TIM-3 overexpression in c-Src–silenced DCs blocked this effect. Collectively, our data reveal the molecular mechanism regulating TIM-3 expression in DCs and identify c-Src as a target for improving the efﬁcacy of nucleic acid–mediated anticancer therapy. The Journal of Immunology, 2016, 197: 000–000.

n view of the initial discovery of T cell Ig and mucin protein-3 HAVCR2 promoter in DCs is not known. Accordingly, the events http://www.jimmunol.org/ (TIM-3) as a Th1-specific cell surface protein (1), earlier regulating HAVCR2 expression in general and specifically in DCs I studies primarily focused on how TIM-3 regulates Th1 re- are ill defined. Moreover, because TIM-3 impacts DC-mediated sponses (2, 3). However, deciphering the role of TIM-3 in innate antitumor immunity, the identification of transcription factors immunity has recently gained considerable interest due to the fact regulating HAVCR2 expression in DCs may reveal potential new that TIM-3 is also expressed on innate effectors, including den- targets for cancer treatment. dritic cells (DCs) (4, 5). In DCs, TIM-3 acts as a critical regula- In the field of cancer immunotherapy, DNA vaccination has tor of antitumor immunity (6, 7), but the molecular mechanisms emerged as a promising therapeutic approach (10). A recent study that regulate TIM-3 expression remain obscure. Although the showed that the antitumor efficacy of DNA vaccines is limited by transcription factors T-bet in T cells and SMAD2 and SMAD4 elevated TIM-3 expression on tumor-associated DCs (TADCs) (6). by guest on September 25, 2021 in a mast cell line have been proposed to regulate HAVCR2 However, the potential molecular target for reducing TIM-3 ex- (which encodes TIM-3) expression (8, 9), only T-bet has been pression on TADCs remains unidentified. Accordingly, we set out shown (in T cells) to bind to the HAVCR2 promoter (8). In the to unravel the molecular mechanism regulating TIM-3 expression case of DCs, reports on transcription factors regulating HAVCR2 in DCs. In this study, we provide evidence for a novel pathway in expression are much more limited. As yet, only T-bet has been which the immunosuppressive cytokines IL-10 and TGF-b, found implicated in the regulation of HAVCR2 mRNA expression; in the tumor milieu (11), upregulate TIM-3 expression on DCs. however, T-bet deficiency does not affect TIM-3 protein ex- Specifically, we demonstrated that both IL-10 and TGF-b upreg- pression in DCs (8). Furthermore, whether T-bet binds to the ulated TIM-3 surface expression on DCs via a common signaling pathway that involved sequential activation of c-Src and Bruton’s tyrosine kinase (Btk; a member of the Tec nonreceptor tyrosine *Division of Cell Biology and Immunology, Council of Scientific and Industrial Research– Institute of Microbial Technology, Chandigarh 160036, India; and †Division of Animal kinase family), leading to the recruitment of Ets and USF tran- Facility, Council of Scientific and Industrial Research–Institute of Microbial Technology, scription factors to the HAVCR2 promoter. We also showed that Chandigarh 160036, India silencing of the proximal mediator c-Src reduced TIM-3 expres- ORCID: 0000-0003-0038-4617 (V.Y.). sion on TADCs and significantly improved the in vivo antitumor Received for publication January 20, 2016. Accepted for publication June 17, 2016. effects of immunostimulatory CpG DNA in animal models. These This work was supported by grants from the Council of Scientific and Industrial findings provide a valuable insight for improving the efficacy of Research and the Department of Biotechnology, Government of India. nucleic acid–mediated antitumor therapy. Address correspondence and reprint requests to Dr. Pradip Sen, Division of Cell Biology and Immunology, Council of Scientific and Industrial Research–Institute of Microbial Technology, Sector 39A, Chandigarh 160036, India. E-mail address: Materials and Methods [email protected] Reagents The online version of this article contains supplemental material. The following Abs and reagents were purchased from Santa Cruz Bio- Abbreviations used in this article: B6, C57/BL6; BMDC, bone marrow–derived technology: anti-Ets1, anti-Ets2, anti-USF1, anti-USF2, anti–TGF-b type I dendritic cell; Btk, Bruton’s tyrosine kinase; ChIP, chromatin immunoprecipitation; receptor (TGF-bRI), anti–IL-10 receptor 1 (IL-10R1), anti–b-actin, HRP- Ctrl, control; DC, dendritic cell; DMS, dimethylsulfate; HuMoDC, human monocyte- conjugated anti-rabbit IgG, and protein A/G PLUS-Agarose beads. Abs to derived DC; IL-10R1, IL-10 receptor 1; ODN, oligodeoxynucleotide; sDC, splenic 416 223 DC; SEAP, secreted alkaline phosphatase; siRNA, small interfering RNA; TADC, c-Src, Btk, p-c-Src (Tyr ), and p-Btk (Tyr ) were obtained from Cell tumor-associated DC; TGF-bRI, TGF-b type I receptor; TIM-3, T cell Ig and mucin Signaling Technology. For immunoprecipitation and chromatin immuno- protein-3; vivo-MO, vivo-morpholino. precipitation (ChIP) assays, rabbit IgG was purchased from R&D Systems. Neutralizing anti–IL-10 and anti–TGF-b Abs were from eBioscience and Copyright Ó 2016 by The American Association of Immunologists, Inc. 0022-1767/16/$30.00 R&D Systems, respectively. PE-conjugated anti-mouse TIM-3 and anti-

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1600104 2 c-Src–TIM-3 AXIS IN DC IMPEDES ANTITUMOR RESPONSES human TIM-3 and allophycocyanin-conjugated anti-mouse CD11c were IL-10 (20 ng/ml) or TGF-b (40 ng/ml) in complete RPMI 1640 medium from BioLegend. The following isotype-matched control Abs were used: (10% FBS, penicillin/streptomycin, L-glutamine, and 2-ME). In some ex- mouse IgG1 (R&D Systems); rat IgG1 (BioLegend); and PE-conjugated periments, DCs were incubated with MC-38 or CT26-wt tumor cell su- rat IgG1, rat IgG2, and mouse IgG1 (all from BioLegend). Recombinant pernatants (50% of total medium) in the presence or absence of 10 mg/ml mouse GM-CSF and IL-4 were from PeproTech; mouse IL-10, human neutralizing anti–IL-10 or anti–TGF-b Ab or respective isotype controls IL-10, and human TGF-b were from R&D Systems; and human GM-CSF (rat IgG1 for anti–IL-10 and mouse IgG1 for anti–TGF-b). and IL-4 were from Miltenyi Biotec. Anti-mouse CD11c MicroBeads and anti-human CD14 MicroBeads were obtained from Miltenyi Biotec. The Primer extension analysis SMARTpool small interfering RNAs (siRNAs) targeting mouse and human Primer extension analysis was done as described (14) using mouse or c-Src, Btk, Ets1, Ets2, USF1, or USF2 were obtained from Dharmacon. 32 human HAVCR2-speciﬁc [ P]-end–labeled antisense primers, the se- Nontargeting scrambled control siRNA was purchased from Santa Cruz quences and locations (relative to the translation initiation sites) of which Biotechnology. were as follows: mouse HAVCR2, primer A (+9 to 221) 59-TGAAAA- Study approval CATGAGTACTTGGCAGGGGAAATC-39 and primer B (211 to 237) 59- CAGGGGAAATCCAAGGACAGCTCTGTG-39;humanHAVCR2,primerC All animal experiments were done with the approval of the Institutional Animal (257 to 276) 59-CAAATGGACTGGGTACTTCT-39, and primer D (272 Ethics Committee of Institute of Microbial Technology. For human monocyte- to 292) 59-CTTCTTCCAACTGTCTACTCC-39. The extracted primer ex- derived DC (HuMoDC) preparation, the buffy coats of healthy donors were tension products were run on a 10% sequencing gel along with the appropriate obtained from Fortis Hospital (Mohali, India), with approval of the Biosafety sequencing ladders. Committee of Institute of Microbial Technology and the Ethics Committee of Fortis Hospital. Informed consents were obtained from all blood donors. Quantitative RT-PCR Mice The cDNA synthesis and quantitative RT-PCR analyses were performed

using the SuperScriptIII Platinum SYBR Green one-step qRT-PCR kit Downloaded from BALB/c and C57/BL6 (B6) mice were maintained at the Institute of Mi- (Invitrogen) and gene-speciﬁc primers (Supplemental Table I). Expression crobial Technology animal facility and used at 8–12 wk of age. was quantiﬁed by the change in threshold method (DDCT) and normalized to the Actb mRNA (encoding b-actin) expression. Tumor cell culture and ELISA In vivo footprinting MC-38 (gifted by G. Shurin; University of Pittsburgh) and CT26-wt (American Type Culture Collection) tumor cells (1 3 106 cells/ml) were Treatment of DCs and the naked DNA with dimethylsulfate (DMS) and

cultured as described (12). After 48 h, supernatants were assayed for IL-10 ligation-mediated PCR were done as described (15). Details of primers used http://www.jimmunol.org/ and TGF-b using mouse IL-10 and TGF-b ELISA kits (eBioscience) and for footprint analyses are mentioned in Supplemental Table I. subsequently used for DC treatment. EMSA DC preparation and treatment Nuclear extracts were prepared as described (16). EMSA was performed Mouse bone marrow–derived DCs (BMDCs), splenic DCs (sDCs), and using various 32P-labeled DNA probes (Supplemental Table I) specific for HuMoDCs were prepared as described (13). BMDCs and sDCs (5 3 106 mouse and human HAVCR2 promoters. An OCT-1 probe, 59-TGTCG- cells/well) were treated with mouse IL-10 (25 ng/ml) or human TGF-b AATGCAAATCACTAGAA-39, was used as control. Bands were visual- (10 ng/ml), and HuMoDCs (5 3 106 cells/well) were treated with human ized using a phosphoimager (Fujifilm FLA-9000; Fujifilm). by guest on September 25, 2021

FIGURE 1. IL-10 and TGF-b secreted from tumor cells upregulate TIM-3 expression on DCs. (A) ELISA of IL-10 and TGF-b in supernatants (Sup) of MC-38 and CT26-wt tumor cells at 48 h after culture. Error bars represent SD. BMDCs from B6 (B) or BALB/c (C) mice were left untreated (UT) or treated for 48 h with supernatants (50% of total medium) of MC-38 (B) or CT26-wt (C) tumor cells in the presence of no Ab, isotype-matched control Ab (isotype) or neutralizing Ab to IL-10 [anti–IL-10 (B)] or TGF-b [anti–TGF-b (B and C)]. TIM-3 expression on DCs was measured by FACS. For this and other ﬁgures, numbers in histograms show mean ﬂuorescence intensities of TIM-3 after subtracting isotype background. Rat IgG1 in (B) and mouse IgG1 in (B) and (C) represent isotype-matched control Abs for anti–IL-10 and anti–TGF-b, respectively. Data are representative of three independent experiments. *p , 0.001. Med, medium. The Journal of Immunology 3

ChIP Densitometry analysis ChIP assays were performed with rabbit IgG or Abs to Ets1, Ets2, USF1, or Densitometry analysis was performed using Scion Image software (Scion USF2 using the ChIP-IT kit (Active Motif). Enrichment of speciﬁc DNA Corporation). fragments was measured by quantitative PCR using primers mentioned in Supplemental Table I. Results were normalized to ChIP with rabbit IgG Reporter assay (control) and input DNA. Custom GLuc-ON dual-reporter constructs containing wild-type mouse Immunoprecipitation, DNA pulldown assay, and immunoblot HAVCR2 promoter fragment [21298 to +43 region; HAVCR2(Wt)pro] or analysis similar promoter fragment containing mutated Ets- or USF-binding site [HAVCR2(MutEts)pro and HAVCR2(MutUSF)pro, respectively] were ob- DCs were lysed with the cell lysis buffer (Cell Signaling Technology). tained from GeneCopoeia. These reporter plasmids encoded the secreted Immunoprecipitation and immunoblot analysis were performed as de- Gaussia luciferase under control of the HAVCR2 promoter and the secreted scribed (13). DNA pulldown assay of nuclear extracts of DCs using bio- alkaline phosphatase (SEAP) under control of the CMV promoter. tinylated oligonucleotides (Supplemental Table I), followed by immunoblot RAW264.7 mouse macrophage cells (1 3 106) were transfected in tripli- analysis was done as described (17). cates with either of the above-mentioned reporter constructs (800 ng) along Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 2. rIL-10 and TGF-b upregulate TIM-3 expression in DCs. HAVCR2 mRNA (encoding TIM-3) expression in BALB/c BMDCs, BALB/c sDCs, B6 BMDCs, and HuMoDCs treated with IL-10 (A) or TGF-b (B) for 12 h or left untreated was assessed by quantitative RT-PCR. Results were normalized to the expression of Actb mRNA (encoding b-actin) and are presented as fold change relative to untreated DCs. Error bars represent SD. TIM-3 expression on BALB/c BMDCs, BALB/c sDCs, B6 BMDCs, and HuMoDCs treated with IL-10 (C) or TGF-b (D) for 48 h or left untreated was measured by FACS. Isotype represents immunostaining with an isotype-matched control Ab. Numbers in histograms indicate mean fluorescence intensity of TIM-3 after subtracting isotype background. Data are representative of three independent experiments. *p , 0.001. 4 c-Src–TIM-3 AXIS IN DC IMPEDES ANTITUMOR RESPONSES with 800 ng mammalian expression vectors encoding Ets1 (pcDNA3.1- cells were enriched using a dead cell removal kit (Miltenyi Biotec). TIM-3 Ets1), Ets2 (pcDNA3.1-Ets2), USF1 (pCMV-USF1), or USF2 (pCMV-USF2) expression on sDCs and TADCs was assessed via flow cytometry after or respective empty vector using the TransIT-2020 transfection reagent gating on CD11c+ population. (Mirus). At 36 h after transfection, supernatants were assayed for the activities of Gaussia luciferase and SEAP using the Secrete-Pair Dual Lumines- Immunohistochemistry Gaussia cence Assay kit (GeneCopoeia). The luciferase activity was normalized Tumor tissues were excised and then fixed overnight in 4% paraformal- to the activity of SEAP. The Ets1 and Ets2 constructs were provided by dehyde (Sigma-Aldrich). Sectioning of paraffin-embedded tissues and T.M. Nowling (Medical University of South Carolina) and USF1 and USF2 immunohistochemical staining were done at IMGENEX India (Bhuba- ` ´ constructs by M.D. Galibert (Universite Rennes, Rennes, France). neswar, India). Expression of c-Src in tumor tissue sections was detected by RNA-mediated interference immunohistochemical staining with anti–c-Src Ab (1:50; Sigma-Aldrich) using MACH 1 detection kit (Biocare Medical) following the manufac- DCs were transfected with 60 nmol siRNA using Lipofectamine RNAiMAX turer’s protocol. Sections were counterstained with hematoxylin (Vector reagent (Invitrogen). Laboratories).

Delivery of vivo-morpholinos and assessment of TIM-3 DC transfer experiments expression on TADCs B6 or BALB/c BMDCs transfected with control siRNA or c-Src–speciﬁc B6 and BALB/c mice were injected s.c. in right ﬂank with MC-38 and siRNA were injected intratumorally (1 3 106 DCs/mouse) into MC-38 CT26-wt tumor cells (1 3 106 cells/mouse), respectively. Mice were then or CT26-wt tumor-bearing syngeneic mice (on days 12, 13, 15, and 16 injected i.v. with nontargeting control vivo-morpholino (Ctrl vivo-MO, 59- posttumor cell inoculation). Mice were then injected intratumorally with CCTCTTACCTCAGTTACAATTTATA-39) or c-Src–targeting vivo-morpholino control oligodeoxynucleotide (Ctrl-ODN) or CpG-ODN (30 mg/mouse; (c-Src vivo-MO, 59-CTTGCTCTTGTTGCTGCCCAT-39; Gene Tools) at 12.5 Invivogen) on days 13, 14, and 16 after tumor inoculation. In some ex- mg/kg body weight or with PBS for 5 consecutive d (day 21–25 posttumor periments, prior to intratumoral administration, c-Src siRNA-treated BMDCs Downloaded from inoculation). As a control, tumor-free mice were injected with PBS at (1 3 106) were transfected with 2 mg empty vector or TIM-3–expressing corresponding time points. After 26 d of tumor inoculation, single-cell vector (B6 TIM-3/pMKITneo or BALB/c TIM-3/pMKITneo; gifted by suspensions were prepared from spleens and established tumors. Viable H. Akiba, Juntendo University, Tokyo, Japan) using the TransIT-2020 http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 3. Identiﬁcation of transcription start site of the HAVCR2.(A) The locations of antisense primers used for primer extension analysis of the mouse (top) and human (bottom) HAVCR2. The transcription start sites (+1) of the mouse (B) and human (C) HAVCR2 were mapped by primer extension analysis using primers indicated in (A). DNA sequencing ladders were generated using respective primers. (D) The nucleotide sequences around the transcription start sites (arrow) of the mouse (top) and human (bottom) HAVCR2. Data are representative of three independent experiments. CS, coding strand; NCS, noncoding strand. The Journal of Immunology 5 Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 4. Identification of transcription factors mediating IL-10– and TGF-b–induced TIM-3 upregulation on mouse DCs. (A) Top, In vivo footprinting of mouse HAVCR2 promoter. Open circles indicate DMS-protected (weaker) bands in IL-10– or TGF-b–treated but not in untreated (UT) BALB/c BMDCs (band intensities are depicted graphically below). Bottom, Summary of the analysis. Base positions are relative to the transcription start site. (B) Details of EMSA probes. Mouse HAVCR2 promoter–specific Pr1 and Pr2 probes contain putative wild-type Ets- and USF-binding sites, and MutEts-Pr1 and MutUSF-Pr2 probes contain mutations (in italics) at Ets- and USF-binding sites, respectively. EMSA of nuclear extracts of BALB/c BMDCs (C and F), BALB/c sDCs (D), and B6 BMDCs (E) treated with IL-10 or TGF-b for indicated times, assayed with indicated probes. Numbers below lanes (Figure legend continues) 6 c-Src–TIM-3 AXIS IN DC IMPEDES ANTITUMOR RESPONSES transfection reagent (Mirus). Tumor volume was measured as described top). Examination of the sequences encompassing these DMS- (18). Expression of IL-12 in tumor lysates was assessed using mouse IL-12 protected G residues with the TFSEARCH program indicated ELISA kit (eBioscience) on day 28 after tumor inoculation. putative binding sites for the Ets and USF transcription factors 2 2 2 2 Flow cytometry ( 106CTGGAGG 100 and 119AACGTG 114, respectively; Fig. 4A, bottom). EMSA using mouse HAVCR2 promoter-specific Pr1 and Flow cytometry was performed with a C6 Accuri flow cytometer (BD Biosciences). Data were analyzed with FlowJo software (Tree Star). Pr2 probes, which contained putative Ets- and USF-binding sites, respectively (Fig. 4B), demonstrated an increased binding of nu- Statistical analyses clear proteins to each of these probes after IL-10 or TGF-b One-way ANOVA (SigmaPlot 11.0 program) was used for all statistical treatment of mouse BMDCs and sDCs (Fig. 4C–E). In contrast, analyses. A p value ,0.05 was considered significant. mutation of the Ets (MutEts-Pr1)- and USF (MutUSF-Pr2)– binding probes blocked this nuclear protein binding (Fig. 4B, 4F). Results Furthermore, Pr1 and Pr2, but not MutEts-Pr1 and MutUSF-Pr2, Tumor cell–secreted IL-10 and TGF-b upregulate TIM-3 successfully precipitated Ets and USF proteins, respectively, expression on DCs from nuclear extracts of BMDCs treated with IL-10 or TGF-b Because IL-10 and TGF-b are generally expressed in the tumor (Fig. 4G). Correspondingly, ChIP assays showed an increased microenvironment (11), we first assessed whether tumor cell–de- recruitment of Ets1, Ets2, USF1, and USF2 to the HAVCR2 pro- b rived IL-10 and TGF-b influenced TIM-3 expression by DCs. For moter after IL-10 or TGF- treatment (Fig. 4H). We further observed that the human HAVCR2 promoter also has Ets- and these experiments, we used the colon carcinoma cell lines MC-38 2 2 2 2

129 123 244 239 Downloaded from and CT26-wt (19, 20), which secrete high levels of IL-10 and USF-binding sequences ( CAGGATG and CATCTG , TGF-b (Fig. 1A). We treated BMDCs, established from B6 mice, respectively) that in turn were bound by these transcription factors b with culture supernatants from MC-38 tumor cells (H2Kb)and upon IL-10 or TGF- treatment of HuMoDCs (Supplemental assessed TIM-3 expression. Whereas treatment with MC-38 culture Fig. 1A–E). These results demonstrate that both IL-10 and b supernatants upregulated TIM-3 expression on B6 BMDCs, the TGF- induce Ets and USF binding to the HAVCR2 promoter in addition of IL-10– or TGF-b–neutralizing Ab diminished TIM-3 mouse and human DCs. induction (Fig. 1B). TIM-3 expression was also upregulated on Next, we tested whether Ets and USF transactivate the http://www.jimmunol.org/ BALB/c BMDCs in a TGF-b–dependent manner by CT26-wt cell mouse HAVCR2 promoter in a reporter assay. Overexpression of (H2Kd) supernatants (Fig. 1C). Furthermore, DC stimulation with Ets1, Ets2, USF1, or USF2 in RAW264.7 mouse macrophages rIL-10 or TGF-b was sufﬁcient to increase TIM-3 expression at strongly enhanced the activity of wild-type HAVCR2 promoter both mRNA and protein levels (Fig. 2). Cotreatment with IL-10 and (Supplemental Fig. 2). However, HAVCR2 promoter activity TGF-b had no additive or synergistic effect on TIM-3 expression was drastically reduced when the Ets- or USF-binding site in (data not shown). Because MC-38 and CT26-wt tumor cell super- the HAVCR2 promoter was mutated (Supplemental Fig. 2). Fur- natants contain several other factors as well as IL-10 and TGF-b thermore, downregulating the expression of Ets1, Ets2, USF1, or b (6, 21, 22), we used rIL-10 and TGF-b in subsequent experiments. USF2 by siRNA inhibited IL-10– and TGF- –induced TIM-3 up- by guest on September 25, 2021 These results demonstrate that TIM-3 expression by DCs is upregulation by BMDCs (Fig. 4I, 4J) and HuMoDCs (Supplemental regulated by tumor cell–derived IL-10 and TGF-b. Fig. 1F, 1G). Collectively, these data demonstrate that Ets and USF positively regulate HAVCR2 promoter activity, leading to TIM-3 upregulation on DCs requires Ets and USF increased TIM-3 expression by DCs upon IL-10 and TGF-b As an initial step to determine the transcription factors regulating stimulation. HAVCR2 expression, we sought to deﬁne the transcriptional start sites for mouse and human HAVCR2. Primer extension analysis c-Src–Btk signaling drives TIM-3 upregulation on DCs using two different primers mapped the transcription start site of We then examined the signaling events upstream of Ets and USF the mouse HAVCR2 to a C residue and of the human HAVCR2 to a recruitment to the HAVCR2 promoter upon IL-10 or TGF-b G residue, which were located 52 and 115 nt upstream of the stimulation. Computational analysis using the Scansite 3 program translational start codon (ATG), respectively (Fig. 3). Hence, we (http://scansite3.mit.edu) predicted c-Src (a nonreceptor tyrosine designated the respective C and G residues of the mouse and kinase) as a potential common interaction partner for the intra- human HAVCR2 as +1 nt. cellular domains of both mouse and human IL-10 and TGF-b Given that both cytokines upregulated TIM-3 expression by receptors. Therefore, we determined whether c-Src activation, as DCs, we investigated whether IL-10 and TGF-b induced binding of measured by tyrosine 416 (Tyr416) phosphorylation of c-Src (23), any common transcription factor (or factors) to the proximal re- was induced by both IL-10 and TGF-b. Within 2.5 min after IL-10 gion of the HAVCR2 promoter. In vivo footprint analysis revealed or TGF-b treatment, phosphorylation of c-Src was enhanced in DMS-protected G residues at positions 2103 and 2104 on the BALB/c BMDCs and sustained up to 6 h (Fig. 5A). Phosphory- coding strand and at positions 2106 and 2117 on the noncoding lation of c-Src was also induced in BALB/c sDCs and HuMoDCs strand of the mouse HAVCR2 promoter in both IL-10– and after IL-10 or TGF-b treatment (Fig. 5B, 5C). Thus, both IL-10 TGF-b–treated BMDCs but not in untreated BMDCs (Fig. 4A, and TGF-b induce c-Src activation in mouse and human DCs.

in (C) represent densitometry (normalized to OCT-1 binding [control]), relative to that of untreated BMDCs (0 h). (G) Nuclear proteins from BALB/c BMDCs treated with IL-10 or TGF-b for 0.5 h or left untreated were subjected to pulldown assay using streptavidin (SA)-coupled Dynabeads and specified biotin- labeled oligonucleotides and immunoblotted (IB) for indicated proteins. Input, nuclear extracts (5%) before pulldown. (H) ChIP quantitative PCR analysis of the recruitment of Ets1, Ets2, USF1, and USF2 to the HAVCR2 promoter after IL-10 or TGF-b treatment of BALB/c BMDCs for 0.5 h. Results are presented as fold enrichment relative to those of untreated BMDCs. Error bars represent SD. (I) Immunoblot analysis confirming siRNA-mediated silencing of Ets1, Ets2, USF1, and USF2 in BALB/c BMDCs. b-Actin serves as a loading control. (J) Analyzing (by FACS) the effect of Ets1, Ets2, USF1, or USF2 silencing on TIM-3 expression by BALB/c BMDCs treated with IL-10 (left panels) or TGF-b (right panels) for 48 h or left untreated. Numbers in histograms show mean fluorescence intensity of TIM-3. Data are representative of three independent experiments. *p , 0.001. - siRNA, no siRNA. The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 5. c-Src and Btk are necessary for IL-10– and TGF-b–induced TIM-3 upregulation on DCs. Immunoblot analysis of total and p-c-Src (A–C) or Btk (D–F)inBALB/cBMDCs(A and D), BALB/c sDCs (B and E), and HuMoDCs (C and F) after IL-10 or TGF-b treatment for indicated times. Numbers below lanes in (A) and (D) represent densitometry, normalized to total c-Src (A)orBtk(D) and presented relative to that of untreated DCs (0 min). Immunoprecipitation (IP) and immunoblot (IB) analysis of the interaction of c-Src and Btk with IL-10R1 (G)andTGF-bRI (H) in BALB/c BMDCs treated with IL-10 (G)orTGF-b (H) (time, above lanes). IgG represents IgG (IP control), and WCL is whole-cell lysate (no IP). (I)BALB/c BMDCs and HuMoDCs were left untransfected (- siRNA) or transfected with control siRNA or c-Src– or Btk-speciﬁc siRNA. Cell lysates were immunoblotted for c-Src, Btk, or b-actin. BALB/c BMDCs (J) and HuMoDCs (K) were transfected as in (I) and then cultured with (+) or without (2) IL-10 or TGF-b for 0.5 h. EMSA was performed with indicated probes (for probe details, see Fig. 4 and Supplemental Fig. 1). FACS analysis of TIM-3 expression on BALB/c BMDCs (L) and HuMoDCs (M) transfected with indicated siRNAs (left margin) and then treated with IL-10 (left panels) or TGF-b (right panels) for 48 h or left untreated. Numbers in histograms show mean ﬂuorescence intensity of TIM-3. Data are representative of three independent experiments. 8 c-Src–TIM-3 AXIS IN DC IMPEDES ANTITUMOR RESPONSES

It is reported that Btk can function either as an upstream reg- TGF-b stimulation. Together, these data suggest that both IL-10 and ulator or downstream effector of c-Src (13, 24). Accordingly, we TGF-b induce c-Src–Btk signaling, which triggers Ets and USF verified whether IL-10 and TGF-b induced Btk activation in DCs. binding to the HAVCR2 promoter and upregulation of TIM-3 ex- We assessed Btk activation by measuring phosphorylation of Btk pression by DCs. at Tyr223 (13). Phosphorylation of Btk was enhanced in BALB/c BMDCs at 15 min after IL-10 or TGF-b treatment and persisted Silencing of c-Src reduces TIM-3 expression on TADCs and up to 6 h (Fig. 5D). IL-10 and TGF-b similarly induced Btk enhances antitumor effects of CpG DNA phosphorylation in BALB/c sDCs and HuMoDCs (Fig. 5E, 5F). Having shown that c-Src activation is an early event required for Thus, Btk, like c-Src, is activated by both IL-10 and TGF-b in IL-10– and TGF-b–induced TIM-3 upregulation on DCs (Fig. 5, DCs. Furthermore, both c-Src and Btk were immunoprecipitated Supplemental Fig. 3), we determined whether c-Src silencing af- with IL-10R1 and TGF-bRI after IL-10 and TGF-b treatment, fected TIM-3 expression on TADCs, which often encounter IL-10 respectively (Fig. 5G, 5H). Because the kinetic analyses showed and TGF-b in the tumor microenvironment (11). Notably, IL-10 an early activation of c-Src relative to Btk (Fig. 5A, 5D), c-Src and TGF-b secreted in MC-38 and CT26-wt tumor cell superna- could be an upstream regulator of Btk in IL-10 and TGF-b sig- tants induced c-Src phosphorylation in BMDCs (Fig. 6A, 6B). naling. To test this possibility, we silenced c-Src and Btk ex- Moreover, silencing of c-Src by siRNA blocked the upregulation pression by siRNA (Supplemental Fig. 3A, 3B) and analyzed the of TIM-3 on BMDCs mediated by the MC-38 or CT26-wt cell effect of these silencing on IL-10– and TGF-b–stimulated acti- supernatants (Fig. 6C). These data indicate that tumor cell–se- vation of Btk and c-Src. Whereas c-Src silencing blocked Btk creted IL-10 and TGF-b upregulate TIM-3 expression on DCs in a phosphorylation after IL-10 and TGF-b stimulation, Btk silencing c-Src–dependent manner. Downloaded from had no effect on c-Src phosphorylation (Supplemental Fig. 3C, Next, we assessed the in vivo role for c-Src in the regulation of 3D). Moreover, c-Src silencing prevented Btk to interact with TIM-3 expression by TADCs. For this, we delivered c-Src–specific IL-10R1 or TGF-bRI despite IL-10 or TGF-b stimulation, re- vivo-MO i.v. into MC-38 or CT26-wt tumor-bearing mice, which spectively (Supplemental Fig. 3E). These results confirmed that in turn reduced c-Src expression in the tumors (Fig. 7A, 7B). c-Src acts upstream of Btk in both IL-10 and TGF-b signaling Compared with sDCs from tumor-bearing or tumor-free mice,

pathways. TADCs from MC-38 or CT26-wt tumors expressed much higher http://www.jimmunol.org/ Next, to verify whether c-Src and Btk were required for IL-10– and levels of TIM-3 (Fig. 7C, 7D, PBS-treated panels). However, TGF-b–induced binding of Ets and USF to the HAVCR2 promoter TIM-3 expression on TADCs was greatly reduced after c-Src si- and upregulation of TIM-3 expression, we silenced c-Src and Btk lencing through vivo-MO–mediated interference (Fig. 7C, 7D). expression in BALB/c BMDCs and HuMoDCs by siRNA (Fig. 5I). These results revealed c-Src as a critical mediator of enhanced Silencing of c-Src or Btk inhibited Ets and USF binding to the TIM-3 expression on TADCs in vivo. HAVCR2 promoter and prevented TIM-3 upregulation on BALB/c Because high expression of TIM-3 by TADCs suppresses nucleic BMDCs (Fig. 5J, 5L) and HuMoDCs (Fig. 5K, 5M) upon IL-10 and acid–mediated antitumor responses (6), it seemed likely that the by guest on September 25, 2021

FIGURE 6. c-Src silencing inhibits TIM-3 upregulation on DCs incubated with tumor cell supernatants. Immunoblot analysis of total and phosphorylated c-Src in B6 (A) and BALB/c (B) BMDCs that were left untreated (2) or treated (+) for 2.5 min with supernatants (Sup) of MC-38 (A) or CT26-wt (B) tumor cells in the presence of Abs as indicated (above lanes). Numbers below lanes represent densitometry results (as in Fig. 5A), presented relative to expression in control DCs (i.e., untreated DCs plus no Ab). Rat IgG1 in (A)andmouseIgG1in(A)and(B) represent isotype control Abs for anti–IL-10 and anti–TGF-b, respectively. (C) Tumor cell supernatants fail to upregulate TIM-3 expression on c-Src–silenced DCs. B6 (left panels) and BALB/c (right panels) BMDCs were transfected with indicated siRNAs (left margin) and then left untreated (UT) or treated for 48 h with supernatants of MC-38 (left panels) or CT26-wt (right panels) cells. TIM-3 expression was measured by FACS. Numbers in histograms indicate mean ﬂuorescence intensity values of TIM-3. Data are representative of three independent experiments. The Journal of Immunology 9 Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 7. c-Src silencing reduces TIM-3 expression on TADCs. B6 and BALB/c mice bearing MC-38 (A) and CT26-wt (B) tumors, respectively, were injected i.v. with PBS, nontargeting Ctrl vivo-MO, or c-Src–targeting vivo-MO (c-Src vivo-MO). Expression of c-Src in s.c. tumors was determined by immunoblot analysis of tumor lysates (with b-actin as a loading control; top) and by immunohistochemical staining of tumor tissue sections (original magnification 3100; bottom). (C and D) Top, FACS analysis of TIM-3 expression (right panels showing histograms) by DCs (CD11c+ gated cells outlined in the left panels) from established tumors (Tumor) or spleens of MC-38 (C) or CT26-wt (D) tumor-bearing mice that had been injected with PBS or vivo-MOs as above or from spleens of tumor-free mice (control) injected with PBS. Bottom, Graphs show cumulative data of TIM-3 expression (presented as mean fluorescence intensity [MFI]) on DCs from two independent experiments (n = 3 mice/group in each experiment). Each symbol represents an individual mouse; horizontal bars represent the mean. *p , 0.01. FSC, forward light scatter. 10 c-Src–TIM-3 AXIS IN DC IMPEDES ANTITUMOR RESPONSES downregulation of TIM-3 on DCs upon c-Src silencing would of control DCs (Fig. 8B, 8C). However, the ability of c-Src–silenced improve the antitumor effects of nucleic acids. To confirm this DCs to enhance CpG-ODN–induced IL-12 production was blocked scenario, we transferred c-Src–silenced DCs and control (control upon forced expression of TIM-3 in these cells (Fig. 8B, 8C). To- siRNA-transfected) DCs into syngeneic mice bearing MC-38 or gether, these results suggest that c-Src plays a key role in sup- CT26-wt tumor. We then injected these mice with a nucleic acid– pression of the antitumor effects of CpG-ODN and that c-Src based adjuvant CpG-ODN, a potential therapeutic agent for cancer mediates this inhibitory effect by increasing the expression of TIM-3 treatment (25). Although treatment with CpG-ODN led to inhi- on TADCs (Fig. 9). bition of tumor growth in the presence of control DCs, the efficacy of CpG-ODN was enhanced by transfer of c-Src–silenced DCs Discussion (Fig. 8A). In contrast, forced expression of TIM-3 in c-Src–silenced The level of TIM-3 expression on TADCs influences the efficacy of DCs blocked the tumor growth inhibitory effect of CpG-ODN antitumor therapies (6). However, the regulatory mechanism of (Fig. 8A). Consistent with these findings, CpG-ODN–induced intra- TIM-3 expression is ill defined. In this study, we have delineated tumoral expression of the potent antitumor cytokine IL-12 (26) was the signaling pathway responsible for TIM-3 upregulation on higher in the presence of c-Src–silenced DCs than in the presence DCs and identified c-Src as a molecular target to reduce TIM-3 Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 8. c-Src–driven TIM-3 expression on DCs impedes in vivo antitumor activities of CpG DNA. (A) Tumor growth kinetics in BALB/c and B6 mice given s.c. inoculation of CT26-wt (left panel) and MC-38 (right panel) tumor cells, respectively, assessed after transfer (intratumorally) of no cells (PBS) or of syngeneic BMDCs that had been transfected with control siRNA or c-Src–specific siRNA alone or together with an empty vector (Vec) or TIM-3–expressing vector (TIM-3) (key), followed by intratumoral injection of Ctrl ODN or CpG-ODN (as described in Materials and Methods). Data are compilation of two separate experiments (n = 3 mice/group in each experiment). Error bars represent SD. ELISA of IL-12p70 in lysates of CT26-wt (B) and MC-38 (C) tumors isolated from mice treated as in (A), assessed 28 d after tumor inoculation. Each symbol represents an individual mouse. Data are compilation of two separate experiments (n = 3 mice/group in each experiment). The horizontal bars represent the mean. *p , 0.001. The Journal of Immunology 11 expression on TADCs and enhance the antitumor effects of immu- ology. For instance, Btk has been shown to negatively regulate nostimulatory DNA. LPS-induced DC maturation (27). Also, our previous studies have In the tumor microenvironment, TIM-3 expression is upregu- implicated Btk in c-MET (hepatocyte growth factor receptor) and lated on DCs via various tumor cell–secreted immunosuppressive TIM-3 signaling that leads to DC suppression (5, 13). However, factors, including IL-10 (6). Our findings demonstrate that TIM-3 the role of Btk in IL-10 and TGF-b signaling remains unknown. In expression is also upregulated by TGF-b. Furthermore, we pro- this study, we describe Btk as an important mediator of both IL-10 vide evidence that the transcription factors Ets and USF play a and TGF-b signaling pathways. In addition, to the best of our critical role in IL-10– and TGF-b–induced upregulation of TIM-3 knowledge, our findings are the first to demonstrate the regulation expression by DCs. This differs from T cells in which constitutive of Ets and USF activity by Btk. TIM-3 expression is dependent on T-bet (8). Until now, the role of Another key observation made in this study is that c-Src acti- Ets and USF in the regulation of TIM-3 expression has remained vation is an early event induced by IL-10 and TGF-b that is needed unaddressed. Our findings therefore demonstrate a previously for downstream Btk activation and subsequent recruitment of Ets unrecognized function for Ets and USF. and USF to the HAVCR2 promoter. Although activation of c-Src Our second finding that Btk is necessary for upregulation of by TGF-b has been reported in other cell types (28), a similar role TIM-3 expression on DCs establishes a new role for Btk in DC for IL-10 is not known. Our findings also suggest that c-Src plays immunoregulation. Indeed, both IL-10 and TGF-b required Btk to a key role in regulating TIM-3 expression by TADCs in vivo. induce Ets and USF binding to the HAVCR2 promoter. So far, only For instance, silencing of c-Src expression blocked upregulation a few studies have addressed the role for Btk in DC immunobi- of TIM-3 by TADCs. In addition, we found that blocking c-Src Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 9. Crosstalk between c-Src and TIM-3 in DCs suppresses the antitumor effects of CpG DNA. c-Src impedes CpG DNA-induced antitumor responses in two mutually nonexclusive ways (demarcated by dotted line). First (left side of the dotted line), in response to IL-10 and TGF-b secreted from tumor cells, c-Src activates the downstream signaling pathway (Btk→transcription factors Ets1/2 and USF1/2) to upregulate TIM-3 expression on DCs (data reported in this study). The upregulated TIM-3 then prevents the binding of CpG DNA to TLR9 and thereby attenuates the antitumor effects of CpG DNA (for detailed mechanism, see Ref. 6). Second (right side of the dotted line), c-Src mediates the inhibitory effect of TIM-3 on DC activation and maturation by blocking the NF-kB pathway [our previous findings (5)], thus suppressing antitumor immunity mediated by CpG DNA (6). The gray-shaded area represents the events, which include NF-kB–driven DC activation and maturation, blocked by TIM-3 signaling. 12 c-Src–TIM-3 AXIS IN DC IMPEDES ANTITUMOR RESPONSES expression and thereby preventing increased expression of TIM-3 cell surface protein Tim-3 regulates macrophage activation and severity of an autoimmune disease. Nature 415: 536–541. by TADCs enhanced the antitumor effects of CpG DNA in vivo. This 2. Sabatos, C. A., S. Chakravarti, E. Cha, A. Schubart, A. Sańchez-Fueyo, later finding demonstrates the therapeutic potential of targeting X. X. Zheng, A. J. Coyle, T. B. Strom, G. J. Freeman, and V. K. Kuchroo. 2003. c-Src. Studies have reported that the antitumor effects of CpG DNA Interaction of Tim-3 and Tim-3 ligand regulates T helper type 1 responses and induction of peripheral tolerance. Nat. Immunol. 4: 1102–1110. are largely dependent on its ability to induce IL-12 secretion from 3. Sańchez-Fueyo, A., J. Tian, D. Picarella, C. Domenig, X. X. Zheng, TADCs (6, 29). However, the elevated TIM-3 expression by TADCs C. A. Sabatos, N. Manlongat, O. Bender, T. Kamradt, V. K. Kuchroo, et al. 2003. greatly impairs CpG DNA-induced IL-12 secretion (6). TIM-3 can Tim-3 inhibits T helper type 1-mediated auto- and alloimmune responses and promotes immunological tolerance. Nat. Immunol. 4: 1093–1101. mediate this inhibitory effect in two possible ways. First, elevated 4. Anderson, A. C., D. E. Anderson, L. Bregoli, W. D. Hastings, N. Kassam, C. Lei, TIM-3 expression by TADCs may interfere with endosomal local- R. Chandwaskar, J. Karman, E. W. Su, M. Hirashima, et al. 2007. Promotion of ization of CpG DNA, thus preventing the interaction between CpG tissue inflammation by the immune receptor Tim-3 expressed on innate immune cells. Science 318: 1141–1143. DNA and TLR9 (6). Consequently, CpG DNA cannot trigger 5. Maurya, N., R. Gujar, M. Gupta, V. Yadav, S. Verma, and P. Sen. 2014. NF-kB signaling (via TLR9), which is necessary for subsequent Immunoregulation of dendritic cells by the receptor T cell Ig and mucin protein- 3 via Bruton’s tyrosine kinase and c-Src. J. Immunol. 193: 3417–3425. IL-12 production (6, 30). Second, TIM-3 may transduce inhibitory 6. Chiba, S., M. Baghdadi, H. Akiba, H. Yoshiyama, I. Kinoshita, H. Dosaka-Akita, signals that block CpG DNA-induced NF-kB activation and IL-12 Y. Fujioka, Y. Ohba, J. V. Gorman, J. D. Colgan, et al. 2012. Tumor-infiltrating production. This possibility has been spurred by our recent obser- DCs suppress nucleic acid-mediated innate immune responses through interactions between the receptor TIM-3 and the alarmin HMGB1. Nat. Immunol. 13: vations that triggering of TIM-3 signaling by anti–TIM-3 Ab blocks 832–842. LPS-induced NF-kB activation and IL-12 secretion (5). In the same 7. Nagahara, K., T. Arikawa, S. Oomizu, K. Kontani, A. Nobumoto, H. Tateno, report, we have demonstrated c-Src as a critical component of the K. Watanabe, T. Niki, S. Katoh, M. Miyake, et al. 2008. Galectin-9 increases Tim-3+ dendritic cells and CD8+ T cells and enhances antitumor immunity via k Downloaded from TIM-3 signaling pathway that inhibits NF- B activation and sub- galectin-9-Tim-3 interactions. J. Immunol. 181: 7660–7669. sequent IL-12 secretion. In addition, we have shown in this study 8. Anderson, A. C., G. M. Lord, V. Dardalhon, D. H. Lee, C. A. Sabatos-Peyton, that c-Src was essential for enhanced TIM-3 expression by TADCs. L. H. Glimcher, and V. K. Kuchroo. 2010. T-bet, a Th1 transcription factor regulates the expression of Tim-3. Eur. J. Immunol. 40: 859–866. Therefore, downregulation of c-Src expression not only blocks the 9. Kim, J. S., D. C. Shin, M. Y. Woo, M. H. Kwon, K. Kim, and S. Park. 2012. inhibitory effects of TIM-3 signaling but also reduces the TIM-3 T cell immunoglobulin mucin domain (TIM)-3 promoter activity in a human levels on TADCs (Fig. 9). In the absence of TIM-3, CpG DNA can mast cell line. Immune Netw. 12: 207–212. 10. Yang, B., J. Jeang, A. Yang, T. C. Wu, and C. F. Hung. 2014. DNA vaccine for readily trigger TLR9-mediated NF-kB signaling and elicit IL-12 cancer immunotherapy. Hum. Vaccin. Immunother. 10: 3153–3164. http://www.jimmunol.org/ secretion from TADCs. In fact, our data have shown that CpG 11. Fricke, I., and D. I. Gabrilovich. 2006. Dendritic cells and tumor microenvironment: a dangerous liaison. Immunol. Invest. 35: 459–483. DNA-induced IL-12 production in the tumor milieu was signifi- 12. Parker, K. H., P. Sinha, L. A. Horn, V. K. Clements, H. Yang, J. Li, K. J. Tracey, cantly increased after c-Src silencing that reduced TIM-3 expres- and S. Ostrand-Rosenberg. 2014. HMGB1 enhances immune suppression by sion on TADCs. In this way, suppression of c-Src in DCs can facilitating the differentiation and suppressive activity of myeloid-derived suppressor cells. Cancer Res. 74: 5723–5733. enhance the antitumor efficacy of CpG DNA. 13. Singhal, E., P. Kumar, and P. Sen. 2011. A novel role for Bruton’s tyrosine ki- In summary, our study has established a pivotal role for c-Src in nase in hepatocyte growth factor-mediated immunoregulation of dendritic cells. the regulation of TIM-3 expression in DCs, and identifies the J. Biol. Chem. 286: 32054–32063. 14. Zhu, J. D., Q. Fei, P. Wang, F. Lan, D. Q. Mao, H. Y. Zhang, and X. B. Yao. pathway of c-Src→Btk→Ets1/2, USF1/2 as controlling this pro- 2006. Transcription of the putative tumor suppressor gene HCCS1 requires cess. In addition, we have demonstrated that c-Src downregulation binding of ETS-2 to its consensus near the transcription start site. Cell Res. 16: by guest on September 25, 2021 in TADCs potentiates the antitumor effects of CpG DNA. Earlier 780–796. 15. Lefevre, P., C. Lacroix, H. Tagoh, M. Hoogenkamp, S. Melnik, R. Ingram, and we had shown that c-Src is also required for TIM-3–induced in- C. Bonifer. 2005. Differentiation-dependent alterations in histone methylation hibition of DC activation and maturation (Fig. 9) (5). Notably, and chromatin architecture at the inducible chicken lysozyme gene. J. Biol. Chem. 280: 27552–27560. inhibition of TADCs by TIM-3 has been linked to an impaired 16. Beg, A. A., T. S. Finco, P. V. Nantermet, and A. S. Baldwin, Jr. 1993. Tumor nucleic acid–mediated antitumor response (6). These reports to- necrosis factor and interleukin-1 lead to phosphorylation and loss of I kappa gether with our present study suggest that the suppression of c-Src B alpha: a mechanism for NF-kappa B activation. Mol. Cell. Biol. 13: 3301– 3310. in DCs augments the antitumor effects of nucleic acids in two 17. Chan, Y. H., M. F. Chiang, Y. C. Tsai, S. T. Su, M. H. Chen, M. S. Hou, and mutually nonexclusive ways: 1) by downregulating TIM-3 ex- K. I. Lin. 2009. Absence of the transcriptional repressor Blimp-1 in hemato- pression on DCs; and 2) by blocking the inhibitory effects of poietic lineages reveals its role in dendritic cell homeostatic development and function. J. Immunol. 183: 7039–7046. TIM-3 signaling on DCs (Fig. 9). Collectively, our results provide 18. Saji, H., W. Song, K. Furumoto, H. Kato, and E. G. Engleman. 2006. Systemic a molecular blueprint for targeting c-Src to improve the antitumor antitumor effect of intratumoral injection of dendritic cells in combination with efficacy of DNA vaccines. local photodynamic therapy. Clin. Cancer Res. 12: 2568–2574. 19. Shurin, M. R., Z. R. Yurkovetsky, I. L. Tourkova, L. Balkir, and G. V. Shurin. 2002. Inhibition of CD40 expression and CD40-mediated dendritic cell function Acknowledgments by tumor-derived IL-10. Int. J. Cancer 101: 61–68. 20. Zhang, B., S. K. Halder, S. Zhang, and P. K. Datta. 2009. Targeting transforming We thank Drs. Ashwani Kumar (Institute of Microbial Technology) and growth factor-beta signaling in liver metastasis of colon cancer. Cancer Lett. Prafullakumar B. Tailor (National Institute of Immunology) for reading the 277: 114–120. 21. Liu, Q., C. Zhang, A. Sun, Y. Zheng, L. Wang, and X. Cao. 2009. Tumor- manuscript, Dr. Galina V. Shurin (University of Pittsburgh) for providing high low the MC-38 cell line, Drs. Tamara Nowling (Medical University of South educated CD11b Ia regulatory dendritic cells suppress T cell response through arginase I. J. Immunol. 182: 6207–6216. Carolina), Marie-Dominique Galibert (Unive`rsite´ Rennes), and Hisaya 22. Schmitz, V., H. Vilanueva, E. Raskopf, T. Hilbert, M. Barajas, C. Dzienisowicz, Akiba (Juntendo University) for sharing plasmid constructs, Fortis Hospi- M. Gorschluter, J. Strehl, C. Rabe, T. Sauerbruch, et al. 2006. Increased VEGF tal (Mohali, India) for providing human buffy coat, and the Institute of levels induced by anti-VEGF treatment are independent of tumor burden in Microbial Technology Animal House Facility for providing mice for ex- colorectal carcinomas in mice. Gene Ther. 13: 1198–1205. 23. Singhal, E., and P. Sen. 2011. Hepatocyte growth factor-induced c-Src- perimentation. phosphatidylinositol 3-kinase-AKT-mammalian target of rapamycin pathway inhibits dendritic cell activation by blocking IkB kinase activity. Int. J. Biochem. Cell Biol. 43: 1134–1146. Disclosures 24. Li, Z., M. I. Wahl, A. Eguinoa, L. R. Stephens, P. T. Hawkins, and O. N. Witte. The authors have no financial conflicts of interest. 1997. Phosphatidylinositol 3-kinase-gamma activates Bruton’s tyrosine kinase in concert with Src family kinases. Proc. Natl. Acad. Sci. USA 94: 13820–13825. 25. Krieg, A. M. 2007. Development of TLR9 agonists for cancer therapy. J. Clin. Invest. 117: 1184–1194. References 26. Tugues, S., S. H. Burkhard, I. Ohs, M. Vrohlings, K. Nussbaum, J. Vom Berg, 1. Monney, L., C. A. Sabatos, J. L. Gaglia, A. Ryu, H. Waldner, T. Chernova, P. Kulig, and B. Becher. 2015. New insights into IL-12-mediated tumor sup- S. Manning, E. A. Greenfield, A. J. Coyle, R. A. Sobel, et al. 2002. Th1-specific pression. Cell Death Differ. 22: 237–246. The Journal of Immunology 13

27. Kawakami, Y., N. Inagaki, S. Salek-Ardakani, J. Kitaura, H. Tanaka, K. Nagao, 29. Krepler, C., V. Wacheck, S. Strommer, G. Hartmann, P. Polterauer, K. Wolff, W. Xiao, H. Nagai, M. Croft, and T. Kawakami. 2006. Regulation of dendritic H. Pehamberger, and B. Jansen. 2004. CpG oligonucleotides elicit antitumor cell maturation and function by Bruton’s tyrosine kinase via IL-10 and Stat3. responses in a human melanoma NOD/SCID xenotransplantation model. J. In- Proc. Natl. Acad. Sci. USA 103: 153–158. vest. Dermatol. 122: 387–391. 28. Kim, J. T., and C. K. Joo. 2002. Involvement of cell-cell interactions in the rapid 30. Bode, K. A., F. Schmitz, L. Vargas, K. Heeg, and A. H. Dalpke. 2009. Kinetic of stimulation of Cas tyrosine phosphorylation and Src kinase activity by trans- RelA activation controls magnitude of TLR-mediated IL-12p40 induction. J. forming growth factor-beta 1. J. Biol. Chem. 277: 31938–31948. Immunol. 182: 2176–2184. Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021 Supplemental Table 1. Details of primers and oligonucleotides used in this study Primer sequences for in vivo footprint analyses of the mouse HAVCR2 promoter DNA strand Primers Primer positions Primer sequence (5'→3') Coding strand Antisense Primer 1 +52 to +37 gagtacttggcagggg (Biotinylated at 5' end) Antisense Primer 2 +42 to +27 caggggaaatccaagg Antisense Primer 3 +35 to +14 aatccaaggacagctctgtgta

Non-coding strand Sense Primer 1 -290 to -270 gatcattggaatactctctcc (Biotinylated at 5' end) Sense Primer 2 -277 to -258 ctctctcctcttgccatgac Sense Primer 3 -271 to -246 cctcttgccatgactaatgcttttct All position numbers are relative to the transcription start sites.

Primer sequences for quantitative RT-PCR Gene Forward primer sequence (5'→3') Reverse primer sequence (5'→3') Mouse HAVCR2 ttggagtgggagtctctgct aatcctgactgctcctgcat Mouse Actb tggaatcctgtggcatcca taacagtccgcctagaagca Human HAVCR2 tccaaggatgcttaccaccag gccaatgtggatatttgtgttagatt Human Actb aggcaccagggcgtgat gcccacataggaatccttctgac

Primer sequences for ChIP-qPCR Promoter Regions Forward primer sequence Reverse primer sequence amplified (5'→3') (5'→3') Mouse HAVCR2 -157 to +42 actggaacattctctgatgtg caggggaaatccaagg Human HAVCR2 -308 to -75 catctgtactttgtgttccccgc cagtcacattaaaggaaaccaaag All position numbers are relative to the transcription start sites.

Probes used for EMSA and DNA pull-down assays Probes Positions Sequence (5'→3') Mouse HAVCR2 promoter Pr1 Probe (carrying wild-type Ets site) -109 to -85 TCACTGGAGGTCAGACATCCTGGGG MutEts-Pr1 probe (carrying mutant Ets site) -109 to -85 TCAATAAAAATCAGACATCCTGGGG Pr2 probe (carrying wild-type USF site) -125 to -108 AGTACTAACGTGGTAATC MutUSF-Pr2 probe (carrying mutant USF site) -125 to -108 AGTACTAATTAAGTAATC

Human HAVCR2 promoter Pr3 Probe (carrying wild-type Ets site) -138 to -115 GAACACTTACAGGATGTGTGTAGT MutEts-Pr3 Probe (carrying mutant Ets site) -138 to -115 GAACACTTACAAACTGTGTGTAGT Pr4 Probe (carrying wild-type USF site) -251 to -229 TGTATCTCATCTGAGACATTATT MutUSF-Pr4 Probe (carrying mutant USF site) -251 to -229 TGTATCTTCTCCCAGACATTATT All position numbers are relative to the transcription start sites; binding sites for the transcription factors are underlined and mutated bases are italicized.

1 SUPPLEMENTAL FIGURES

Supplemental Figure 1. Both IL-10 and TGF- require Ets and USF to upregulate TIM-3 expression on human DCs. (A) Details of human HAVCR2 promoter-specific probes used for EMSA. Pr3 and Pr4 probes contain putative wild-type Ets- and USF-binding sites, and MutEts- Pr3 and MutUSF-Pr4 probes contain mutations (italics) at Ets- and USF-binding sites, respectively. Base positions are relative to the transcription start site. (B and C) EMSA of nuclear extracts of HuMoDCs treated with IL-10 or TGF- for indicated times; assayed with indicated probes. OCT-1 binding serves as an internal control. Numbers below lanes in (B): densitometry readings (as in Fig. 4C). (D) DNA pull-down assay, followed by immunoblot analysis reveals that IL-10 or TGF- treatment (for 0.5 h) of HuMoDCs induces the binding of nuclear Ets and USF proteins to their putative target sequences in the human HAVCR2 promoter. The biotinylated oligonucleotides used in this assay are indicated above lanes. Input, nuclear extracts (5%) before pull-down. (E) ChIP-qPCR analysis to determine the occupancy of Ets1, Ets2, USF1 and USF2 on the HAVCR2 promoter in HuMoDCs left untreated (UT) or treated with IL-10 or TGF- for 0.5 h. Results are presented as in Fig. 4H. Error bars represent SD. *p < 0.001; **p = 0.007. (F) Immunoblot analysis depicting siRNA-mediated silencing of Ets1, Ets2, USF1 and USF2 in HuMoDCs.-actin serves as a loading control. (G) Treatment with IL-10 (left panels) or TGF- (right panels) for 48 h fails to upregulate TIM-3 expression (measured by FACS) on HuMoDCs after silencing of Ets1, Ets2, USF1 or USF2 by siRNA. Numbers in histograms show MFI values of TIM-3. Data are representative of three separate experiments.

Supplemental Figure 2. Ets1, Ets2, USF1 and USF2 transactivate the HAVCR2 promoter. (A) Schematic of the HAVCR2 promoter-Gaussia luciferase (GLuc) constructs used in reporter assays. Expression of Gaussia luciferase is controlled by the mouse HAVCR2 promoter fragment (-1298 to +43), either wild-type (HAVCR2(Wt)pro) or mutated at Ets- (HAVCR2(MutEts)pro) or USF- (HAVCR2(MutUSF)pro) binding site. (B) Activation of the wild-type HAVCR2 promoter or its mutant variants (illustrated in A) at 36 h after transfection of RAW264.7 cells with plasmids encoding Ets1, Ets2, USF1 or USF2; assessed by measuring the activity of secreted Gaussia luciferase in the supernatants. Results were normalized to the activity of secreted alkaline phosphatase (encoded by same reporter plasmids; internal control) and are presented relative to those of cells cotransfected with wild-type HAVCR2 reporter and an empty vector (pcDNA3.1 for Ets1 and Ets2, or pCMV for USF1 and USF2). Error bars represent SD. *p < 0.001. Data are representative of three independent experiments.

Supplemental Figure 3. c-Src acts upstream of Btk in IL-10 and TGF- signaling. (A and B) Immunoblot analysis showing siRNA-mediated silencing of c-Src (A) or Btk (B) in BALB/c BMDCs. -actin serves as a loading control. (C and D) Immunoblot analysis illustrating the effect of c-Src silencing on Btk phosphorylation (C) and of Btk silencing on c-Src phosphorylation (D) in BALB/c BMDCs treated with IL-10 or TGF- for indicated time points. Total Btk (C) and c-Src (D) serve as loading controls. (E) Co-IP assessing the impact of c-Src silencing on the interaction of Btk with IL-10 and TGF- receptors. BALB/c BMDCs were treated with IL-10 (left panel) or TGF- (right panel) for indicated times after c-Src silencing by siRNA. Cell lysates were immunoprecipitated with anti-IL-10R1 (left panel), anti-TGF-RI (right panel) or control IgG, and immunoblotted for indicated proteins. Data are representative of three independent experiments.