BTNL2, a Butyrophilin/B7-Like Molecule, Is a Negative Costimulatory Molecule Modulated in Intestinal Inflammation

This information is current as Heather A. Arnett, Sabine S. Escobar, Eva Gonzalez-Suarez, of September 28, 2021. Alison L. Budelsky, Lori A. Steffen, Norman Boiani, Ming Zhang, Gerald Siu, Avery W. Brewer and Joanne L. Viney J Immunol 2007; 178:1523-1533; ; doi: 10.4049/jimmunol.178.3.1523

http://www.jimmunol.org/content/178/3/1523 Downloaded from

References This article cites 40 articles, 12 of which you can access for free at: http://www.jimmunol.org/content/178/3/1523.full#ref-list-1 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 28, 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 © 2007 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

BTNL2, a Butyrophilin/B7-Like Molecule, Is a Negative Costimulatory Molecule Modulated in Intestinal Inflammation

Heather A. Arnett,1* Sabine S. Escobar,1* Eva Gonzalez-Suarez,* Alison L. Budelsky,* Lori A. Steffen,* Norman Boiani,† Ming Zhang,§ Gerald Siu,§ Avery W. Brewer,‡ and Joanne L. Viney2*

Butyrophilin-like 2 (BTNL2) is a butyrophilin family member with homology to the B7 costimulatory molecules, polymorphisms of which have been recently associated through genetic analyses to sporadic inclusion body myositis and sarcoidosis. We have characterized the full structure, expression, and function of BTNL2. Structural analysis of BTNL2 shows a molecule with an extracellular region containing two sets of two Ig domains, a transmembrane region, and a previously unreported cytoplasmic tail. Unlike most other butyrophilin members, BTNL2 lacks the prototypical B30.2 ring domain. TaqMan and Northern blot analysis indicate BTNL2 is predominantly expressed in digestive tract tissues, in particular small intestine and Peyer’s patches. Immunohisto- Downloaded from chemistry with BTNL2-specific Abs further localizes BTNL2 to epithelial and dendritic cells within these tissues. Despite its homology to the B7 family, BTNL2 does not bind any of the known B7 family receptors such as CD28, CTLA-4, PD-1, ICOS, or B and T lymphocyte attenuator. Because of its localization in the gut and potential role in the immune system, BTNL2 expression was analyzed in a mouse model of inflammatory bowel disease. BTNL2 is overexpressed during both the asymptomatic and symptomatic phase of the Mdr1a knockout model of spontaneous colitis. In functional assays, soluble BTNL2-Fc protein inhibits the proliferation of murine CD4؉ http://www.jimmunol.org/ T cells from the spleen, mesenteric lymph node, and Peyer’s patch. In addition, BTNL2-Fc reduces proliferation and cytokine produc- tion from T cells activated by anti-CD3 and B7-related protein 1. These data suggest a role for BTNL2 as a negative costimulatory molecule with implications for inflammatory disease. The Journal of Immunology, 2007, 178: 1523–1533.

he regulation of T cell responses is achieved through Ag- (4–8) is expressed in both lymphoid and nonlymphoid cells (9). A specific TCRs acting together with coreceptors that can second inhibitory receptor PD-1 (10, 11) binds to either of two T modulate activation. One of the principal costimulatory B7-related ligands PD-L1 (also known as B7-H1) (12, 13) and pathways for naive T cells is initiated when CTLA-4 and CD28 PD-L2 (or B7-DC) (14, 15). Another inhibitory receptor, B and T

engage B7 class molecules during TCR engagement by a primary lymphocyte attenuator (BTLA), binds to HVEM, previously iden- by guest on September 28, 2021 signal (1). These secondary costimulatory signals can either inhibit tified as a binding partner for LIGHT and lymphotoxin ␣. B7-H3, or promote T cell responses, depending on the receptor they en- another B7 homolog (16, 17), binds an unidentified receptor on gage. For example, B7-1 and B7-2 can bind either CD28, resulting activated T cells and serves as a negative regulator of T cell acti- in T cell stimulation, or CTLA-4, resulting in T cell inhibition (2). vation (18). A more recently described B7 family member, B7-H4 Negative signals can contribute to the attenuation of T cell re- (also called B7S1 or B7x) (19–21), acts as a novel costimulator sponses and so are important in regulating T cell tolerance. The and regulates the threshold of T cell activation. balance between stimulatory and inhibitory signals is critical to an Butyrophilin-like 2 (BTNL2) has been identified as a B7-like effective immune response. molecule within the butyrophilin gene family, located in the MHC In the past several years, new members of the B7 ligand and the locus in mice and humans (22). Recently, several studies have corresponding CD28/CTLA-4 receptor families have been identi- linked the gene for BTNL2 to two inflammatory diseases, sporadic fied, increasing our understanding of immune cell function. A sec- inclusion body myositis and sarcoidosis (23–25); however, no ond activating receptor in T cells is ICOS (3); its ligand B7-related function has been demonstrated for BTNL2. Although the buty- protein 1 (B7RP1)3 (also known as LICOS, GL50, B7-H2, or B7h) rophilin family of proteins share homology with the B7 family, no immune responses have been associated with BTNL1, the first and most broadly expressed butyrophilin. Because of its close struc- *Inflammation, †Protein Sciences, and ‡Pathology, Amgen, Seattle, WA 98119; and §Inflammation, Amgen, Thousand Oaks, CA 91320 tural homology to B7-1, BTNL2 has been hypothesized to have T cell costimulatory properties (25). Received for publication December 7, 2005. Accepted for publication October 17, 2006. In this study we characterize the expression of BTNL2 in normal The costs of publication of this article were defrayed in part by the payment of page and diseased tissue, and show evidence that BTNL2 down-regulates charges. This article must therefore be hereby marked advertisement in accordance T cell activation and may play a role in maintaining tolerance and with 18 U.S.C. Section 1734 solely to indicate this fact. resetting the balance of the immune response in the gut. 1 H.A.A. and S.S.E. contributed equally to this work. 2 Address correspondence and reprint requests to Dr. Joanne L. Viney, Department of Inflammation, Amgen, 1201 Amgen Court West, Seattle, WA 98119. E-mail address: Materials and Methods [email protected] Cloning 3 Abbreviations used in this paper: B7RP1, B7-related protein 1; BTLA, B and T lymphocyte attenuator; BTNL, butyrophilin-like; MLN, mesenteric lymph node; PP, Murine and human BTNL2 (accession nos. NM_023145 and NM_019602, Peyer’s patch; CHO, Chinese hamster ovary. respectively) were originally identified from the National Center for Bio- technology Information (NCBI) database based on homology to other B7 Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00 family molecules. RACE PCR was performed to confirm 5Ј and 3Ј ends of www.jimmunol.org 1524 BTNL2 DOWN-REGULATES T CELL ACTIVATION the human and mouse BTNL2 genes based on the original deposits using 1% Triton X-100, with a protease inhibitor mixture (Boehringer Mann- RACE-Ready Marathon cDNA from human colon and mouse spleen (BD heim). Proteins were separated on 4–20% reducing Tris-glycine gel, trans- Clontech). In addition, cDNAs from human T84 (colon carcinoma) cells, ferred to nitrocellulose, and blocked overnight at 4°C in TBST plus 3% Caco2 (colon carcinoma) cells, and mouse colon were also used to confirm nonfat milk. After washing in TBST, the membranes were incubated with the final cDNA sequences. The final full-length coding region cDNAs of the murine BTNL2–77p polyclonal antisera or control sera for1hat4°C, human and mouse BTNL2 were then amplified as single cDNA contigs, followed by goat anti-rabbit HRP (1/5000; Amersham Biosciences) and with the new additional 3Ј sequence, and then subcloned into pCR4-Topo developed by chemiluminescence (Amersham Biosciences). (Invitrogen Life Technologies) and appropriate mammalian expression vectors. Independent PCR analyses were performed to confirm each final Immunohistochemistry murine and human BTNL2 sequence. The sequence results were compared with both the predicted sequence published by Stammers et al. (22) and the Frozen sections were fixed in acetone at 4°C for 10 min, washed in TBS NCBI database. Transmembrane and signal peptide predictions were predicted (pH 8.0; Sigma-Aldrich) three times, then blocked for endogenous perox- ϩ with TMHMM and SignalP (www.cbs.dtu.dk/services/TMHMM-2.0 and idase in glucose oxidase solution (b-D glucose (G-5250; Sigma-Aldrich), www.cbs.dtu.dk/services/SignalP). glucose oxidase (Sigma-Aldrich), sodium azide (Sigma-Aldrich), and TBS (pH 8.0)) for 30 min at 37°C. Sections were then treated with avidin and Quantitative real-time PCR biotin blocking solutions (Vector Laboratories) for 15 min each, rinsed in TBST and blocked in Tris/NaCl/blocking solution (TNB; PerkinElmer) Expression of murine BTNL2 was measured by quantitative real-time RT- for 30 min. Primary Ab was applied using anti-BTNL2 at 19.5 ␮g/ PCR using the ABI PRISM 7900HT sequence detection system (Applied Bio- ml (xmuRIBD006, rat IgG2a, clone M517; Amgen) or 20 ␮g/ml systems) and normalized to the expression of a housekeeping gene (HPRT). (xmuRIBD006, rat IgG2a, clone M518; Amgen), anti-B220 at 0.025 ␮g/ml Ј Three PCR primer sets spanning the 1/2 exon junction (forward) 5 -GCTGA (01121D, rat IgG2a; BD Pharmingen), or using goat anti-human Fc (Jack- Ј Ј CTATAAAGCACCCAGATGAC-3 (nucleotides 60–84), (reverse) 5 -GAA son ImmunoResearch Laboratories), and incubated overnight at 4°C. Slides GATGCCCTGCTAACGTGTC-3Ј (130–151), (probe) 5Ј-AGTGGTCGGTC were washed in TBST followed by a 30-min incubation in the appropriate Downloaded from Ј CTAACCTCCCAATCTTG-3 (90–117); spanning the 5/6 exon junction secondary Ab (biotinylated rabbit anti-rat IgG or anti-goat IgG; Vector Ј Ј Ј (forward) 5 -ACCCTGCAGTCCACGTGTATG-3 (863–883), (reverse) 5 - Laboratories) at 2.5 ␮g/ml. Slides were washed in TBST then incubated for Ј Ј GGCATCAGTCACCAATGAAGTC-3 (933–954), (probe) 5 -CACGTG 30 min in streptavidin HRP (NEL750; PerkinElmer), followed by chromo- Ј GCTGGAGAGCAGATGGTAGAATAC-3 (895–924); and spanning the gen development for 5 min (liquid diaminobenzidine; BioGenex). Slides Ј Ј 7/8 exon junction (forward) 5 -GGACCTGATCAAGGTGAAACG-3 were counterstained with hematoxylin solution (DakoCytomation), dehy- Ј Ј (1422–1442), (reverse) 5 -ACACACAGCAGCAATCAGGAAA-3 drated, cleared, and mounted in synthetic resin (PolyMount; Poly Ј Ј (1465–1486), (probe) 5 -ACGGACCAATGAACA-3 (1449–1463) were Scientific). used to determine relative expression of BTNL2 cDNA and confirm ex- http://www.jimmunol.org/ pression patterns. All RNA was DNase-treated (DNA-free; Ambion) be- Naked DNA studies fore generation of cDNA using a TaqMan Reverse Transcription kit (Ap- plied Biosystems). Using 20 ng of total RNA equivalent cDNA, samples Naked DNA injections were used to induce expression of murine were subject to quantitative PCR using 2ϫ TaqMan Universal buffer (Ap- BTNL2-Fc in vivo, as previously described (30, 31). The 10 ␮gofDNA plied Biosystems) with 200 nM forward and reverse primers and a 900 nM (huFc.pEF100G or IgK-BTNL2-huFc.pEF100G) in sterile saline was de- probe with the following PCR conditions 50°C for 2 min; 95°C for 10 min; livered in a volume equal to 10% of the mouse body weight by tail vein 95°C for 15 s; 60°C for 1 min (40 times). Each PCR analysis was run in injection to female BALB/c mice (The Jackson Laboratory). The construct triplicate for each biological sample included in the study. containing BTNL2-Fc in this experiment used an IgK leader in place of the native leader to improve expression. The vector, pEF100G is Gateway Preparation of murine BTNL2-Fc adapted and uses the human elongation factor 1␣ promoter. Endotoxin levels were kept below 2 UEq/injection. Blood samples were collected via by guest on September 28, 2021 A cDNA was constructed encoding the extracellular domain of murine the tail vein from mice on days 1, 3, 7, 10, 14, and 20 (mice were injected BTNL2 in-frame with the human IgG -Fc mutein (BTNL2-Fc). The human 1 on day 0), and serum was analyzed by ELISA for Fc expression. Tissue IgG -Fc mutein encodes 98–330 aa of SWISSProt P01857 with the fol- 1 was collected from a subset of mice on day 7 for analysis by immunohis- lowing mutations: L117A, L118E, and G120A. These amino acids have tochemistry for the presence of Fc protein. All tissues were frozen in OCT been shown to be critical for high affinity binding to Fc receptors (26, 27). and stored at Ϫ70°C. All studies using mice were performed in accordance Mutation of these amino acids reduces affinity to Fc receptors Ͼ100-fold, with the guidelines and approval of the Amgen Institutional Animal Care described in U.S. Patent Nos. 5,457,035 and WO 93/10151 (28). COS and Use Committee. protein kinase B cells (E5) were transfected with the BTNL2-Fc mamma- lian expression construct cDNA with Lipofectamine 2000 (Invitrogen Life In vitro analysis of CD4ϩ T cell proliferation Technologies) and cultured in complete DMEM with 0.5% low Ig serum, as described by Ettehadieh et al. (29). Seven days posttransfection, super- Mouse. Single-cell suspensions were prepared from murine splenocytes, natants were harvested and purified by protein A-Sepharose Fast Flow MLN, and PP cells. Tissues were harvested from at least five female (Amersham Biosciences) column chromatography. C57BL/6 mice per experiment, and the CD4ϩ cells were purified using the Spin Sep CD4ϩ negative selection kit (Stem Cell Sciences). Purity of Generation of BTNL2 polyclonal Abs CD4ϩ cells was Ͼ90% as assessed by FACS analysis. A total of 1–2 ϫ 105 ϩ Three synthetic peptides representing the extracellular murine BTNL2 pro- CD4 cells/well were added to precoated flat-bottom 96-well plates. Plates were coated with variable concentrations of anti-CD3 mAb (clone tein sequence were independently KLH-conjugated, pooled, and then in- ␮ jected into New Zealand White rabbits for antisera production. Before im- 2C11; BD Pharmingen) and 10 g/ml goat anti-human Fc (Jackson ImmunoResearch Laboratories) in PBS at 4°C overnight. Wells were then munization, preimmune serum was collected as a negative control. ␮ Antisera were collected from the immunized rabbits and reactivity verified washed with PBS and coated with 10 g/ml of the specified amount of the by ELISA against the original peptides and the murine BTNL2-Fc protein. indicated fusion protein for4hatroom temperature. In the human B7RP1-Fc costimulation assay, two fusion proteins were plated simulta- Generation of BTNL2 mAbs neously to the same well, to a final concentration of 10 ␮g/ml (unless indicated otherwise) in PBS. After a final PBS wash, purified CD4ϩ Two Lewis rats were immunized with 20 ␮g of murine BTNL2-Fc in equal splenocytes, or MLN or PP cells were added to wells in a final volume of volume with Titermax (Cytogen), in a total volume of 100 ␮l. Three weeks 200 ␮l of complete RPMI 1640 medium containing 10% FBS and 55 pM later, animals were boosted with 20 ␮g of the same material emulsified in 2-ME. Proliferation of CD4ϩ cells was determined by incorporation of 1 IFA. Sixteen days later, one rat was i.v. boosted with 8 ␮g of murine ␮Ci/well of [3H]thymidine during the last 6 h (splenocytes) or 16 h (MLN BTNL2-Fc. Three days later, the rat was sacrificed and spleen removed. and PP cells) of the 72-h culture. Spleen cells were fused to NS1 myelomas using PEG1500. Supernatants Human. Human T cells were purified from human PBMC using a CD4ϩ were screened by ELISA against the murine BTNL2-Fc protein and an T cell isolation kit from Miltenyi Biotec, resulting in a population of cells irrelevant Fc protein (p7.5-Fc). Abs were purified over protein A-agarose containing Ͼ90% CD4ϩ cells. Plates were precoated by adding 100 ␮lof (Amersham Biosciences). PBS containing varying concentrations of anti-CD3 (clone HIT3a; BD Western blot analysis Pharmingen) with or without BTNL2-Fc (10 ␮g/ml) or negative control protein Chinese hamster ovary (CHO)-Fc (10 ␮g/ml). Plates were incu- Cell lysates (50 ϫ 106 cells/ml) from splenocytes, mesenteric lymph node bated at 4°C overnight and then washed twice with PBS. Purified T cells (MLN) cells, and Peyer’s patch (PP) cells were prepared by lysis in PBS (1 ϫ 105) were added to precoated 96-well flat-bottom plates in a volume The Journal of Immunology 1525 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 1. Structural analysis of BTNL2. A, Comparison of BTNL2 to the B7 and butyrophilin families. B, Comparison of full-length human and murine BTNL2 at the amino acid level. C, Alternative splice variants for BTNL2 in mice and humans. 1526 BTNL2 DOWN-REGULATES T CELL ACTIVATION Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021 The Journal of Immunology 1527 Downloaded from

FIGURE 3. Expression of BTNL2 is up-regulated in a mouse model of inflammatory bowel disease. A, Expression levels of BTNL2 in the colonic RNA from Mdr1a knockout mice that spontaneously develop colitis were compared with expression levels in the parental strain, FVB wild-type mice. Relative http://www.jimmunol.org/ expression was established for BTNL2 by quantitative RT-PCR, normalized to expression levels of housekeeping gene HPRT, using primers spanning the exon 7/8 junction in BTNL2. Each point represents an independent mouse/RNA sample. Statistical significance was determined comparing FVB wild-type p Ͻ 0.0001. B, Tissue from the distal colons of normal or Mdr1a knockout mice were analyzed by H&E ,ء .mice with either asymptomatic or colitic mice (left panels) and immunohistochemistry for BTNL2 (middle panels). The colon from the colitic Mdr1a knockout mouse had a large inflammatory infiltrate, and higher numbers of BTNL2ϩ cells than the noninflamed colon of an FVB wild-type mouse. Magnifications are shown for each image, and an isotype control image (right panels) from a serial section is included.

of 200 ␮l of complete RPMI 1640 medium containing 10% FBS. Prolif- of T cell produced cytokines. The plate was read with a Luminex 100 by guest on September 28, 2021 eration of CD4ϩ cells was determined by incorporation of 1 ␮Ci/well system. Each experiment was performed with triplicate biological repli- [3H]thymidine (Amersham Biosciences) during the last 6–8 h of a 72-h cates. The data shown are representative of at least two independent ex- assay. periments. All results were expressed as the mean Ϯ SEM of the biological replicates in a single experiment. Values for probability were calculated by Proteinase K treatment of BTNL2-Fc unpaired two-tailed t test using Prism software version 4.01 (GraphPad Software). Those cytokines that demonstrated a statistically significant In a final volume of 50 ␮l, BTNL2-Fc protein (15 ␮g) was incubated with value of p Ͻ 0.05 (two-tailed t test) in at least two independent experiments a 25% slurry of Glycine-Affi-10 (control) or Proteinase K-Affi-10 coupled and those with cytokine values above the reported least-detectable dose for resin overnight at 4°C. The resin was removed by spinning and supernatant the assay are included. collected. The “degraded” BTNL2-Fc protein was diluted to a final con- centration of 10 ␮g/ml for plate coating based on pretreatment equivalent and used in an in vitro analysis of mouse CD4ϩ T cell proliferation as described. Transfections and FACS Cytokine measurements 293-MSR cells were transfected with the full-length murine cDNAs for B7RP1, CD80, PD-L2, BTLA, BTNL2, or vector only with Lipofectamine In a standard anti-CD3 proliferation assay as earlier described, murine 2000 (Invitrogen Life Technologies) per the manufacturer’s instructions. Two CD4ϩ cells were isolated from spleens and costimulated with anti-CD3, days posttransfection, cells were harvested with nonenzymatic cell dissociation human B7RP1-Fc (5 ␮g/ml) and/or inhibited with murine BTNL2-Fc (10 buffer (Invitrogen Life Technologies). One million cells were then stained on ␮g/ml). After 64 h of stimulation, 100 ␮l of supernatant was harvested ice for 30 min with each murine BTNL2-Fc, HVEM-Fc, ICOS-Fc, PD-1-Fc, from each condition. The supernatants (25 of 100 ␮l) were then assayed for CTLA-4-Fc, and CD28-Fc at 10 ␮g/ml. ICOS-Fc, PD-1-Fc, and CD28-Fc cytokine levels using the Linco Research mouse 22-Plex (no. MCYTO- were purchased from R&D Systems; all others were developed in-house. Fol- Ј 70K-PMX22) kit containing premixed beads capable of detecting a variety lowing a wash, bound Fc protein was detected with PE-conjugated F(ab )2

FIGURE 2. Expression of BTNL2 in normal tissues. A, BTNL2 is expressed in the gastrointestinal tissues, as well as the spleen and lymph node, of mice. Quantitative RT-PCR with an exon 7/8 PCR primer set was used to define the expression pattern of BTNL2 in normal C57BL6/J mouse tissue. Additional primer sets spanning exon 1/2 and exon 5/6 showed similar results (data not shown). Expression values are shown relative to that of the housekeeping gene, HPRT. B, BTNL2 protein is expressed in PP. Protein lysates were prepared from cells isolated from MLN, PP, and spleen and separated on 4–20% reducing, Tris-glycine gel followed by Western blot analysis with preimmune sera (left) or polyclonal BTNL2 antisera (right). C, Full-length murine BTNL2-transfected 293-MSR cells express a protein at a similar m.w. as the endogenous protein detected in PP tissues. 293-MSR cells were transfected with vector only or murine BTNL2-TM (full-length version) mammalian expression cDNAs. Two days posttransfection, cell lysates were prepared (50 ϫ 106/ml) and separated on 4–20% reducing Tris-glycine gel and transferred to nitrocellulose. Western blot analysis was performed with polyclonal BTNL2 sera, or preimmune sera. D, Immunohistochemistry with anti-BTNL2 mAb was performed on a variety of tissues shown by quantitative PCR to express BTNL2, including small intestine, large intestine (proximal), inguinal lymph node, spleen, and PP. Immunohistochemistry on adjacent sections withan isotype control Ab is shown for each of these panels. 1528 BTNL2 DOWN-REGULATES T CELL ACTIVATION

FIGURE 4. BTNL2-Fc does not bind any known B7 family member re- ceptors. 293-MSR cells were trans- fected with the full-length murine trans- membrane cDNAs of B7RP1, CD80, PD-L2, BTLA, BTNL2 or empty vec- tor only using Lipofectamine 2000. Two days posttransfection, cells were harvested and stained with 10 ␮g/ml of the indicated Fc protein and detected Downloaded from Ј with PE-conjugated F(ab )2 goat anti- human Fc␥ specific secondary. Sam- ples were fixed and analyzed on a FACSCalibur. A, Histographs of FACS analysis are shown. B, The summary of FACS analysis is also indicated. http://www.jimmunol.org/ by guest on September 28, 2021

goat anti-human Fc (Jackson ImmunoResearch Laboratories). Samples were coding a signal peptide, two Ig-like domains, heptad linker, two fixed and analyzed using a FACSCalibur (BD Immunocytometry Systems). additional Ig-like domains, a transmembrane domain, and a cyto- plasmic tail (Fig. 1, B and C). A splice variant missing the second Results Ig-like domain in the extracellular region (exon 3) was also iden- Cloning of BTNL2 tified. The significance of this finding is unknown but a similar The gene encoding BTNL2 is contained in the MHC locus in both human variant was also identified. Cloning of the human counter- mouse and human. BTNL2 was originally described in a genomic part yielded two full-length transmembrane alternative splice vari- analysis as a putative soluble butyrophilin protein with an un- ants with either exon 2 or exons 2 and 3 deleted. A single cDNA known function (22). The butyrophilin and B7 family members are representing a full-length transmembrane version of human ancestrally related, and past genomic analysis has identified other BTNL2 with all four Ig-like domains was not identified. PCR of butyrophilin-like molecules later determined to be new B7 family various exon combinations (i.e., exons 1–3, 2–6, 4–6) yielded members with immunoregulatory activity. Overall, amino acid se- products that suggest a full-length version does exist in human, quence similarity of B7 and butyrophilin is low, but domain struc- although it may not be the dominant form, as it appears to be in the ture is conserved (Fig. 1A). All members contain a signal peptide, mouse. two Ig-like domains (IgV and IgC), a transmembrane domain, and a cytoplasmic domain. Two distinguishing features of most buty- rophilins from B7 family members is the inclusion of a 170 aa Expression analysis of BTNL2 in normal tissue B30.2 domain in the cytoplasmic domain (32). Unlike most buty- Quantitative RT-PCR was used to define the expression pattern of rophilins, this domain is not found in BTNL2, but like butyrophi- BTNL2 in normal C57BL/6J mouse tissue (Fig. 2A). As previously linsa7aaheptad sequence linker separating the two sets of IgC/ reported (22, 25), expression of BTNL2 mRNA is detected in a IgV domains is present. number of tissue types including spleen, lymph node, stomach, Full-length cloning and RACE PCR analysis was performed to MLN, bone marrow, small intestine, cecum, lung, large intestine, confirm the 5Ј and 3Ј end of the putative human and mouse BTNL2 PP, and thymus. The highest levels of expression are detected in genes. Human and mouse BTNL2 are encoded by eight exons, and small intestine, PP, and cecum tissue. Northern blot analysis of have a 64% identify score (Fig. 1B). The original published Ge- murine tissue detects BTNL2 predominantly in digestive tract tis- neScan analysis theorized only a soluble protein containing the sue especially in the colon and small intestine (data not shown). signal peptide, Ig-like domains and heptad linker. We have iden- Analysis by immunoblot with an anti-BTNL2 polyclonal Ab con- tified a full-length transmembrane cDNA of murine BTNL2 en- firms expression of BTNL2 protein in PP (Fig. 2B). Murine The Journal of Immunology 1529

BTNL2 runs at a higher m.w. than predicted by amino acid se- quence, likely due to further glycosylation or other posttransla- tional modifications. Western blot analysis of cell lysates of 293- EBNA cells transfected with full-length transmembrane BTNL2 detected BTNL2 protein with a m.w. consistent with the BTNL2 band detected in PP cells using the same BTNL2 polyclonal anti- sera (Fig. 2C). Together, these data suggest that the additional mass is likely due to posttranslational modifications. To further localize BTNL2 expression within tissue, an immu- nohistochemical survey across an array of mouse tissues was per- formed with a mAb to BTNL2. BTNL2ϩ cells are found in the tissues indicated by the RNA analysis, including small intestine, large intestine, PP, inguinal lymph node, and spleen (Fig. 2D). In the small intestine, BTNL2 is predominantly expressed in the ep- ithelial cells of the villi but not the crypts. Interestingly, colonic epithelial cells have been previously implicated in the suppression of T cells in the gut, and it is possible that the same may be true for epithelial cells from the small intestine (33). In other areas of the intestine and other tissues analyzed, BTNL2 is localized to Downloaded from cells with the morphological appearance of dendritic cells. Expression analysis of BTNL2 in diseased tissue Numerous mouse models of inflammatory bowel disease have been developed to mimic acute and chronic inflammatory bowel

diseases in human. In the Mdr1a knockout model of inflammatory http://www.jimmunol.org/ bowel disease, mice spontaneously develop colitis in a specific pathogen-free environment (34, 35). It is thought that an aberrant FIGURE 5. The cognate for BTNL2 is expressed in PP and liver. Mice immune response to commensal bacterial Ags leads to the intes- were injected with DNA expression vectors containing mouse BTNL2-Fc tinal inflammation, characterized by dysregulated epithelial cell cDNA or human Fc cDNA only. A, Serum Fc levels in individual mice growth and massive leukocyte infiltration. Before the onset of clin- were measured by an ELISA at varying time points following the injection ical signs, changes in mRNA levels of inflammatory mediators can of DNA containing BTNL2-Fc or Fc only. Mice injected with BTNL2-Fc be detected (35). Using quantitative RT-PCR to compare the total DNA were followed for 20 days while mice injected with a control vector RNA from the colons of Mdr1a knockout mice to the parental containing huFc were followed for 10 days. B, Tissues were harvested from strain, FVB wild-type mice, BTNL2 is found to be up-regulated in mice at day 7 following injection of the naked DNA expression vectors and by guest on September 28, 2021 both asymptomatic and colitic Mdr1a knockout mice (Fig. 3A). immunohistochemistry was performed on frozen tissue sections using an An increase in BTNL2 expression in the inflamed colon was anti-human Fc Ab. Binding along the sinusoidal endothelium of the liver further investigated at the protein level by immunohistochemistry. and the vascular endothelium of PP is observed in mice injected with the ϩ vector containing BTNL2-Fc, but not in mice injected with a vector con- Numbers of BTNL2 cells with the morphological appearance of taining Fc alone. dendritic cells were increased in the lamina propria of the inflamed colon of an Mdr1a knockout mouse compared with the expression in FVB wild-type mice (Fig. 3B). Affymetrix array analysis of human sion of BTNL2-Fc, or a control vector for expression of Fc only. BTNL2 expression levels in RNA isolated from patients with Crohn’s At various time points following injection of BTNL2-Fc or control disease and ulcerative colitis is inconclusive due to low expression Fc DNA, serum levels in individual mice were measured by an levels of BTNL2 in these samples (data not shown). ELISA for Fc (Fig. 5A). Equivalent serum concentrations of Fc protein were achieved in mice injected with either the BTNL2-Fc BTNL2 does not bind to known B7 family receptors construct or the Fc control construct. By analogy with other family members that have immune function, Immunohistochemistry was performed for the Fc tag in mice we predicted that BTNL2 might bind to cell surface receptors on injected with BTNL2-Fc or control Fc expression vectors 7 days T cells and/or other cells at sites of inflammation and Ag presen- after DNA injection. Staining is evident along the sinusoidal en- tation. To confirm that known B7 family ligands and receptors are dothelium of the liver and the vascular endothelium of PP (Fig. not a binding partner for BTNL2, cells were transfected with the 5B); this staining was not present in control mice expressing Fc full-length transmembrane version of murine BTNL2 and tested alone, although similar expression levels of Fc were detected in the for binding to many known recombinant soluble B7 family mem- serum (Fig. 5A). Immunohistochemistry was also performed using ber receptors by FACS, including CTLA-4, CD28, ICOS, PD-1, a mAb against BTNL2. Similar patterns of staining are observed, and HVEM or membrane-bound BTLA. No binding to BTNL2 although it is important to note that this Ab detects endogenous was detected for any of these molecules, although each receptor BTNL2, as well as the tissue-bound BTNL2-Fc. Although we have bound to its known ligands (Fig. 4). not identified the binding partner for BTNL2, it appears to be expressed on sinusoidal endothelium in the liver and in the vas- BTNL2 binds to a counterstructure present in PP and the liver cular endothelium within the PP. To identify tissues expressing the binding partners for BTNL2, we induced in vivo expression of Fc-tagged BTNL2, and then used BTNL2 is a negative regulator of T cell proliferation and immunohistochemistry to identify which tissues contained bound cytokine production BTNL2-Fc. Mice were injected with either naked DNA encoding B7 family proteins have been shown to have activity in T cell mouse BTNL2-Fc, with an IgK leader sequence to force expres- costimulatory assays that involve activating T cells through their T 1530 BTNL2 DOWN-REGULATES T CELL ACTIVATION Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 6. BTNL2 is a negative regulator of T cell proliferation. A, BTNL2-Fc inhibits anti-CD3-stimulated proliferation of murine CD4ϩ cells isolated from spleen, MLN, and PP. The splenocytes and MLN and PP cells were stimulated with plate-bound anti-CD3 (clone 2C11; 1, 2, and 4 ␮g/ml, respectively) and either immobilized BTNL2-Fc (10 ␮g/ml), murine B7-2 (2 ␮g/ml), or a protein control CHO-Fc (10 ␮g/ml). B, BTNL2-Fc inhibits varying concentrations of anti-CD3-stimulated proliferation. CD4ϩ splenocytes were incubated with decreasing amounts of plate-bound anti-CD3 as indicated, and BTNL2-Fc or CHO-Fc at 10 ␮g/ml. C, BTNL2-Fc inhibits anti-CD3-induced proliferation of CD4ϩ cells in a dose-dependent manner. CD4ϩ cells were stimulated with 4 ␮g/ml plate-bound anti-CD3 and decreasing amounts of BTNL2-Fc as indicated. CHO-Fc was included at 10 ␮g/ml. D, Proteinase K (PK) treatment of BTNL2-Fc protein restores anti-CD3-induced proliferation. BTNL2-Fc was treated with proteinase K-coupled Affi-10 or control Glycine-Affi-10 coupled resin overnight. Resin was removed and pretreatment equivalent of proteinase K-treated BTNL2-Fc was used to precoat The Journal of Immunology 1531 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 7. BTNL2-Fc inhibits the costimulatory signal of human B7RP1-Fc. A, BTNL2-Fc inhibits anti-CD3 and human B7RP1-Fc induced prolif- eration. CD4ϩ cells were incubated with plate-bound anti-CD3 (clone 2C11) at 1 ␮g/ml and BTNL2-Fc at 10 ␮g/ml and human B7RP1-Fc at 5 ␮g/ml. CHO-Fc at 10 or 5 ␮g/ml was used as a negative control. Supernatants (100 ␮l) from each well were then collected at 64 h. The supernatant was replenished with fresh culture medium containing [3H]thymidine (1 ␮Ci) followed by an additional 6-h incubation. B, BTNL2-Fc inhibits the costimulatory B7RP1-Fc induced enhanced cytokine production. Cytokine production was assayed with the Linco mouse cytokine/chemokine multiplex plate (22 analytes), and results interpolated using the standard curve included with the kit. All results were expressed as the mean Ϯ SEM of three biological replicates assayed. p ϭ 0.05 was considered significant, as calculated by unpaired two-tailed t test, comparing anti-CD3/B7RP1-Fc costimulation with and without ,ء BTNL2-Fc. Data are representative of at least two independent experiments. cell receptors with varying doses of Ag. Using costimulation with performed on the BTNL2-Fc protein. Digestion of the protein with immobilized anti-CD3 mAb, the effect of murine BTNL2-Fc pro- proteinase K abolishes the inhibitory activity of BTNL2-Fc, dem- tein was assessed on murine CD4ϩ T cell responses. Proliferation onstrating that the observed inhibitory activity is protein specific of freshly purified CD4ϩ T cells from spleen, MLN, or PP by (Fig. 6D). Viability studies demonstrate that BTNL2-Fc does not anti-CD3 mAb is inhibited by the addition of BTNL2-Fc, but not promote death of the T cells (data not shown). BTNL2-Fc can an irrelevant Fc protein, CHO-Fc (Fig. 6A). BTNL2-Fc can inhibit similarly inhibit human T cell proliferation (Fig. 6E). Human and T cell proliferation across a range of anti-CD3 doses (Fig. 6B), and mouse B lymphocyte proliferation is not inhibited by BTNL2-Fc the inhibition by BTNL2-Fc is dose-dependent (Fig. 6C). To en- (data not shown). sure the effects of T cell proliferation were not due to copurifica- B7RP1, typically expressed by APCs, behaves as a potent co- tion of contaminants or toxic products, proteinase K digestion was stimulatory factor in the presence of anti-CD3, through binding to

plates for a standard anti-CD3 induced proliferation assay of murine CD4ϩ cells. E, Mouse BTNL2-Fc inhibits anti-CD3-induced proliferation of human T cells. Purified human T cells were incubated with decreasing amounts of plate-bound anti-CD3 (clone HIT3a) and immobilized BTNL2-Fc or CHO-Fc at 10 ␮g/ml. Proliferation was measured by [3H]thymidine uptake during the last6hinsplenocytes assays or 16 h in MLN/PP and human PBMC assays. Data are representative of three or more experiments. All results are expressed as the mean Ϯ SEM. 1532 BTNL2 DOWN-REGULATES T CELL ACTIVATION

ICOS, expressed by activated T cells. Addition of B7RP1-Fc and of epithelia and dendritic cells in the colons of mice with signs of anti-CD3 to murine CD4ϩ T cell cultures results in enhanced pro- colitis. Overexpression of negative costimulatory molecules likely liferation over anti-CD3 alone. This costimulation can be proves valuable in controlling inflammation where down-modula- effectively blocked by the addition of BTNL2-Fc (Fig. 7A) but not tion of the immune response is desirable following the initial in- a control protein, CHO-Fc. flammatory event. The inhibitory effects of BTNL2-Fc on T cell We further explored the functional effect of BTNL2-Fc in this activation in vitro suggest that the up-regulation of BTNL2 in this assay by analyzing cytokine production in the different culture mouse model of colitis may contribute to the attempted amelioration conditions. Cytokine profiling of murine splenocyte CD4ϩ T cell of the inflammatory response. Of note, other anti-inflammatory mol- supernatants by multiplexed Luminex-based analysis demonstrates ecules, such as IDO and IL-10, are also up-regulated in concor- statistically significant increases in the expression of TNF-␣, GM- dance with onset of colitis, highlighting the complex balance of CSF, IL-10, IL-4, IL-6, IL-17, and IFN-␥ by anti-CD3/B7RP1 proinflammatory and anti-inflammatory signals that accompany costimulation of CD4ϩ T cells. The addition of BTNL2-Fc, when disease processes. compared with the inclusion of a control protein, CHO-Fc, effec- Mice lacking either BTLA or PD-1 are predisposed to develop tively blocks these increases ( p Ͻ 0.05) (Fig. 7B). The presence of autoimmune phenotypes with signs of progressive arthritis and a human Fc tag alone (CHO-Fc) does appear to have some incon- glomerulonephritis, suggesting their roles in the inhibition of in- sistent inhibitory activities in these assays, although no statistically flammatory responses and in maintaining peripheral tolerance (11, significant differences are observed. Together, the inhibitory ef- 41), therefore it is possible that mice lacking BTNL2 might be fects on both proliferation and cytokine expression demonstrate a more susceptible to developing spontaneous autoimmune disease. functional role for BTNL2-Fc in decreasing the activation state of This hypothesis remains to be tested. Downloaded from CD4ϩ T cells in the presence of costimulatory molecules. In summary, the characterization of a new member of the B7 family provides new insights into regulation of effector T cells, Discussion improving our understanding of organ-specific autoimmune dis- Recently, BTNL2 has been linked in separate genetic studies to eases and immune evasive strategies used by tumor cells and par- sarcoidosis and sporadic inclusion body myositis (23, 25). The asites. The expression patterns of BTNL2 in normal and diseased

BTNL2 SNP that results in a premature exon-splice site is associ- gut tissue, together with its ability to down-regulate T cell activa- http://www.jimmunol.org/ ated with an elevated risk of sarcoidosis, a multisystem disorder tion in vitro, makes it an attractive protein to be implicated in characterized by granulomas. Originally established in a white promoting intestinal tolerance, and point to a role for BTNL2 in German population (25), this association has also been indepen- regulating T cell-mediated responses in the gut by dampening im- dently demonstrated in an African American population (24). Spo- mune responses. radic inclusion body myositis is an acquired autoimmune inflam- matory muscle disease diagnosed by chronic muscular weakness. Acknowledgments Both diseases are characterized by inappropriate T cell activation We thank John Delaney, Melissa Lee, Irene Corpuz, and Jane Carter for the (36–38). Although the functional effects of the polymorphisms in production of BTNL2 protein, Peter Baum for initial contributions to the BTNL2 have yet to be established, these genetic linkages com- cloning, Li-ya Huang for immunohistochemistry, Duke Virca for protein- by guest on September 28, 2021 bined with the homology to the B7 family would predict a role for ase K digestion of BTNL2-Fc, Dean Thompson for screening polyclonal BTNL2 in the activation of T cells. Abs, Fergus Byrne, Jennitte Stevens, D. J. Caughey, and Todd Juan for Despite the genetic linkages, no function has been previously HVEM and BTLA reagents, Deanna Hill for assistance in necropsies, and demonstrated for BTNL2. Using a soluble Fc-tagged BTNL2 (ex- the expertise contributed by Jacque Wilk, Afsi Mozaffarian, Ryan Swanson, and the Amgen Washington Sequencing, FACS, and Array tracellular domain only), we have shown that BTNL2 is capable of groups. diminishing the proliferative responses of CD4ϩ lymphocytes from spleen, MLN, and PP in response to anti-CD3. In addition, Disclosures BTNL2 effectively down-regulates the proliferative and cytokine H. A. Arnett, S. S. Escobar, E. Gonzalez-Suarez, A. L. Budelsky, expression responses to B7RP1 costimulation. As the ligand for L. A. Steffen, N. Boiani, M. Zhang, G. Siu, A. W. Brewer, and J. L. Viney ICOS, B7RP1 acts as a costimulatory trigger in combination with all hold stock interest in Amgen. anti-CD3 to generate a high state of activation in T cells. Disrup- tion of this pathway has been beneficial in regulating autoimmu- References nity in a number of mouse models of disease, including colitis, 1. Greenwald, R. J., G. J. Freeman, and A. H. Sharpe. 2005. The B7 family revis- multiple sclerosis, and arthritis (1, 39). The observed ability of ited. Annu. Rev. Immunol. 23: 515–548. BTNL2 to prevent B7RP1-ICOS costimulation (both proliferation 2. Sansom, D. M., C. N. Manzotti, and Y. Zheng. 2003. What’s the difference between CD80 and CD86? Trends Immunol. 24: 314–319. and cytokine production) suggests that increased BTNL2 presents 3. Hutloff, A., A. M. Dittrich, K. C. Beier, B. Eljaschewitsch, R. Kraft, a novel avenue to modulate T cell activation in inflammatory dis- I. Anagnostopoulos, and R. A. Kroczek. 1999. ICOS is an inducible T-cell co- stimulator structurally and functionally related to CD28. Nature 397: 263–266. ease. While under review, a complementary report was published 4. Swallow, M. M., J. J. Wallin, and W. C. Sha. 1999. B7h, a novel costimulatory confirming an inhibitory role for BTNL2 in proliferation of mouse homolog of B7.1 and B7.2, is induced by TNF␣. Immunity 11: 423–432. T cells exposed to anti-CD3, and demonstrates binding of 5. Yoshinaga, S. K., J. S. Whoriskey, S. D. Khare, U. Sarmiento, J. Guo, T. Horan, G. Shih, M. Zhang, M. A. Coccia, T. Kohno, et al. 1999. T-cell co-stimulation BTNL2-Fc to activated B and T cells, suggesting that its cognate through B7RP-1 and ICOS. Nature 402: 827–832. is present on these cells (40). Nguyen et al. (40) also report a 6. Ling, V., P. W. Wu, H. F. Finnerty, K. M. Bean, V. Spaulding, L. A. Fouser, decrease in anti-CD3/CD28-induced IL-2 production by T cells in J. P. Leonard, S. E. Hunter, R. Zollner, J. L. Thomas, et al. 2000. Cutting edge: identification of GL50, a novel B7-like protein that functionally binds to ICOS the presence BTNL2-Fc; this is consistent with our observed receptor. J. Immunol. 164: 1653–1657. blockade of IL-2 production in anti-CD3/B7RP1 stimulated cells. 7. Wang, S., G. Zhu, A. I. Chapoval, H. Dong, K. Tamada, J. Ni, and L. Chen. 2000. Costimulation of T cells by B7-H2, a B7-like molecule that binds ICOS. Blood Inappropriate or inadequate expression of a negative costimu- 96: 2808–2813. latory molecule could contribute to the abnormal inflammation 8. Brodie, D., A. V. Collins, A. Iaboni, J. A. Fennelly, L. M. Sparks, X. N. Xu, and/or immune responses observed in a number of diseases. We P. A. van der Merwe, and S. J. Davis. 2000. LICOS, a primordial costimulatory ligand? Curr. Biol. 10: 333–336. have shown that in the Mdr1a knockout model of colitis, BTNL2 9. Liang, L., E. M. Porter, and W. C. Sha. 2002. Constitutive expression of the B7h is up-regulated and widely expressed in cells with the morphology ligand for inducible costimulator on naive B cells is extinguished after activation The Journal of Immunology 1533

by distinct B cell receptor and interleukin 4 receptor-mediated pathways and can 26. Canfield, S. M., and S. L. Morrison. 1991. The binding affinity of human IgG for be rescued by CD40 signaling. J. Exp. Med. 196: 97–108. its high affinity Fc receptor is determined by multiple amino acids in the CH2 10. Ishida, Y., Y. Agata, K. Shibahara, and T. Honjo. 1992. Induced expression of domain and is modulated by the hinge region. J. Exp. Med. 173: 1483–1491. PD-1, a novel member of the immunoglobulin gene superfamily, upon pro- 27. Jefferis, R., J. Lund, and J. Pound. 1990. Molecular definition of interaction sites grammed cell death. EMBO J. 11: 3887–3895. on human IgG for Fc receptors (huFc␥R). Mol. Immunol. 27: 1237–1240. 11. Nishimura, H., N. Minato, T. Nakano, and T. Honjo. 1998. Immunological stud- 28. Baum, P. R., R. B. Gayle III, F. Ramsdell, S. Srinivasan, R. A. Sorensen, ies on PD-1 deficient mice: implication of PD-1 as a negative regulator for B cell M. L. Watson, M. F. Seldin, E. Baker, G. R. Sutherland, K. N. Clifford, et al. responses. Int. Immunol. 10: 1563–1572. 1994. Molecular characterization of murine and human OX40/OX40 ligand sys- 12. Freeman, G. J., A. J. Long, Y. Iwai, K. Bourque, T. Chernova, H. Nishimura, tems: identification of a human OX40 ligand as the HTLV-1-regulated protein L. J. Fitz, N. Malenkovich, T. Okazaki, M. C. Byrne, et al. 2000. Engagement of gp34. EMBO J. 13: 3992–4001. the PD-1 immunoinhibitory receptor by a novel B7 family member leads to 29. Ettehadieh, E., S. Wong-Madden, T. Aldrich, K. Lane, and A. E. Morris. 2001. negative regulation of lymphocyte activation. J. Exp. Med. 192: 1027–1034. Over-expression of protein kinase B␣ enhances recombinant protein expression 13. Dong, H., G. Zhu, K. Tamada, and L. Chen. 1999. B7-H1, a third member of the in transient systems. In Animal Cell Technology: From Target to Market. B7 family, co-stimulates T-cell proliferation and interleukin-10 secretion. Nat. E. Lindner-Olsson, eds. Kluwer Academic. Dordrecht, The Netherlands, pp. 31– Med. 5: 1365–1369. 35. 14. Latchman, Y., C. R. Wood, T. Chernova, D. Chaudhary, M. Borde, I. Chernova, 30. Liu, F., Y. Song, and D. Liu. 1999. Hydrodynamics-based transfection in animals Y. Iwai, A. J. Long, J. A. Brown, R. Nunes, et al. 2001. PD-L2 is a second ligand by systemic administration of plasmid DNA. Gene Ther. 6: 1258–1266. for PD-1 and inhibits T cell activation. Nat. Immunol. 2: 261–268. 15. Tseng, S. Y., M. Otsuji, K. Gorski, X. Huang, J. E. Slansky, S. I. Pai, A. Shalabi, 31. Zhang, G., V. Budker, and J. A. Wolff. 1999. High levels of foreign gene ex- T. Shin, D. M. Pardoll, and H. Tsuchiya. 2001. B7-DC, a new dendritic cell pression in hepatocytes after tail vein injections of naked plasmid DNA. Hum. molecule with potent costimulatory properties for T cells. J. Exp. Med. 193: Gene Ther. 10: 1735–1737. 839–846. 32. Henry, J., I. H. Mather, M. F. McDermott, and P. Pontarotti. 1998. B30.2-like 16. Chapoval, A. I., J. Ni, J. S. Lau, R. A. Wilcox, D. B. Flies, D. Liu, H. Dong, domain proteins: update and new insights into a rapidly expanding family of G. L. Sica, G. Zhu, K. Tamada, and L. Chen. 2001. B7-H3: a costimulatory proteins. Mol. Biol. Evol. 15: 1696–1705. molecule for T cell activation and IFN-␥ production. Nat. Immunol. 2: 269–274. 33. Cruickshank, S. M., L. D. McVay, D. C. Baumgart, P. J. Felsburg, and 17. Sun, M., S. Richards, D. V. Prasad, X. M. Mai, A. Rudensky, and C. Dong. 2002. S. R. Carding. 2004. Colonic epithelial cell mediated suppression of CD4 T cell Downloaded from Characterization of mouse and human B7-H3 genes. J. Immunol. 168: activation. Gut 53: 678–684. 6294–6297. 34. Panwala, C. M., J. C. Jones, and J. L. Viney. 1998. A novel model of inflam- 18. Prasad, D. V., T. Nguyen, Z. Li, Y. Yang, J. Duong, Y. Wang, and C. Dong. matory bowel disease: mice deficient for the multiple drug resistance gene, 2004. Murine B7-H3 is a negative regulator of T cells. J. Immunol. 173: mdr1a, spontaneously develop colitis. J. Immunol. 161: 5733–5744. 2500–2506. 35. Wilk, J. N., J. Bilsborough, and J. L. Viney. 2005. The mdr1aϪ/Ϫ mouse model 19. Sica, G. L., I. H. Choi, G. Zhu, K. Tamada, S. D. Wang, H. Tamura, A. I. of spontaneous colitis: a relevant and appropriate animal model to study inflam- Chapoval, D. B. Flies, J. Bajorath, and L. Chen. 2003. B7-H4, a molecule of the matory bowel disease. Immunol. Res. 31: 151–160.

B7 family, negatively regulates T cell immunity. Immunity 18: 849–861. 36. Co, D. O., L. H. Hogan, S. Il-Kim, and M. Sandor. 2004. T cell contributions to http://www.jimmunol.org/ 20. Prasad, D. V., S. Richards, X. M. Mai, and C. Dong. 2003. B7S1, a novel B7 the different phases of granuloma formation. Immunol. Lett. 92: 135–142. family member that negatively regulates T cell activation. Immunity 18: 863–873. 37. Dalakas, M. C. 2004. Inflammatory disorders of muscle: progress in polymyo- 21. Zang, X., P. Loke, J. Kim, K. Murphy, R. Waitz, and J. P. Allison. 2003. B7x: sitis, dermatomyositis and inclusion body myositis. Curr. Opin. Neurol. 17: a widely expressed B7 family member that inhibits T cell activation. Proc. Natl. 561–567. Acad. Sci. USA 100: 10388–10392. 22. Stammers, M., L. Rowen, D. Rhodes, J. Trowsdale, and S. Beck. 2000. BTL-II: 38. Megens-de Letter, M. A., L. H. Visser, P. A. van Doorn, and H. F. Savelkoul. a polymorphic locus with homology to the butyrophilin gene family, located at 1999. Cytokines in the muscle tissue of idiopathic inflammatory myopathies: the border of the major histocompatibility complex class II and class III regions implications for immunopathogenesis and therapy. Eur. Cytokine Netw. 10: in human and mouse. Immunogenetics 51: 373–382. 471–478. 23. Price, P., L. Santoso, F. Mastaglia, M. Garlepp, C. C. Kok, R. Allcock, and 39. Totsuka, T., T. Kanai, R. Iiyama, K. Uraushihara, M. Yamazaki, R. Okamoto, N. Laing. 2004. Two major histocompatibility complex haplotypes influence sus- T. Hibi, K. Tezuka, M. Azuma, H. Akiba, et al. 2003. Ameliorating effect of

ceptibility to sporadic inclusion body myositis: critical evaluation of an associ- anti-inducible costimulator monoclonal antibody in a murine model of chronic by guest on September 28, 2021 ation with HLA-DR3. Tissue Antigens 64: 575–580. colitis. Gastroenterology 124: 410–421. 24. Rybicki, B. A., J. L. Walewski, M. J. Maliarik, H. Kian, and M. C. Iannuzzi. 40. Nguyen, T., X. K. Liu, Y. Zhang, and C. Dong. 2006. BTNL2, a butyrophilin-like 2005. The BTNL2 gene and sarcoidosis susceptibility in African Americans and molecule that functions to inhibit T cell activation. J. Immunol. 176: 7354–7360. Whites. Am. J. Hum. Genet. 77: 491–499. 41. Watanabe, N., M. Gavrieli, J. R. Sedy, J. Yang, F. Fallarino, S. K. Loftin, 25. Valentonyte, R., J. Hampe, K. Huse, P. Rosenstiel, M. Albrecht, A. Stenzel, M. A. Hurchla, N. Zimmerman, J. Sim, X. Zang, et al. 2003. BTLA is a lym- M. Nagy, K. I. Gaede, A. Franke, R. Haesler, et al. 2005. Sarcoidosis is associated phocyte inhibitory receptor with similarities to CTLA-4 and PD-1. Nat. Immunol. with a truncating splice site mutation in BTNL2. Nat. Genet. 37: 357–364. 4: 670–679.