Spliced Form of B7-1 Characterization of a Soluble, Alternatively Primary

Primary Structure and Functional Characterization of a Soluble, Alternatively Spliced Form of B7-1

This information is current as Susan J. Faas, Michelle A. Giannoni, Angela P. Mickle, of September 29, 2021. Cheri L. Kiesecker, Deborah J. Reed, Dayang Wu, William L. Fodor, John P. Mueller, Louis A. Matis and Russell P. Rother J Immunol 2000; 164:6340-6348; ;

doi: 10.4049/jimmunol.164.12.6340 Downloaded from http://www.jimmunol.org/content/164/12/6340

References This article cites 51 articles, 22 of which you can access for free at: http://www.jimmunol.org/content/164/12/6340.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 by guest on September 29, 2021 • Fast Publication! 4 weeks from acceptance to publication

*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 © 2000 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Primary Structure and Functional Characterization of a Soluble, Alternatively Spliced Form of B7-11

Susan J. Faas, Michelle A. Giannoni, Angela P. Mickle, Cheri L. Kiesecker, Deborah J. Reed, Dayang Wu, William L. Fodor, John P. Mueller,2 Louis A. Matis, and Russell P. Rother3

Recent studies have suggested that soluble forms of B7-1 and B7-2 may exist, but transcripts that code for these molecules have not been previously described. In this study, we report the cloning and characterization of an alternatively spliced soluble form of porcine B7-1 (sB7-1) that lacks exons coding for both the transmembrane and cytoplasmic domains. Northern blot analysis of RNA from alveolar macrophages revealed an approximate 3:1 ratio of the transmembrane form of B7-1 mRNA relative to sB7-1 mRNA. Porcine B7-1 was present on the surface of both B and T cells following stimulation with PMA/ionomycin. A histidine- tagged form of porcine sB7-1 (sB7-1-His) interacted with both CD28 and CTLA-4, and effectively blocked IL-2 production from human responder cells stimulated with PHA and either porcine or human stimulator cells. In addition, sB7–1-His inhibited human Downloaded from T cell proliferation in response to porcine or human peripheral blood leukocytes. This study is the ﬁrst report of an alternatively spliced form of B7 that codes for a soluble protein. Furthermore, these data demonstrate that porcine B7-1 interacts with the human receptors CD28 and CTLA-4, suggesting a potential role for this molecule in pig to human xenotransplantation. Possible physiological functions for the soluble form of B7-1 are discussed. The Journal of Immunology, 2000, 164: 6340–6348.

n addition to TCR-mediated Ag recognition, T cell activation natively spliced messages or derived from full-length molecules on http://www.jimmunol.org/ is dependent on signals generated by interactions between T the cell surface through enzymatic cleavage or cell death. I cell costimulatory molecules and their ligands expressed on Several alternate transcripts exist for B7-1 and B7-2. For exam- APCs (1). The related proteins CD28 and CTLA-4 have been ple, multiple murine B7-2 alternatively spliced products have been shown to play critical roles in the regulation of T cell activation via identified (13). Additionally, an alternatively spliced form of mu- interactions with B7 ligands expressed on various APCs (2, 3). The rine B7-1 lacking the Ig C region-like extracellular domain has B7 molecules, B7-1 and B7-2, are type I transmembrane glyco- been described (14). Resolution of the B7-1 genomic organization proteins containing an Ig V region-like domain, an Ig C region-like in mouse and human indicates that both the transmembrane and domain, and a short cytoplasmic tail (3, 4). CD28-B7 interactions cytoplasmic domains are encoded by separate exons, making it amplify T cell activation signals, and TCR Ag recognition in the possible to exclude these regions without disrupting the extracel- by guest on September 29, 2021 absence of CD28 engagement may result in anergy (2, 3, 5, 6). lular domains (15, 16). Further analysis of the murine B7-1 Conversely, B7 signaling through CTLA-4 down-regulates T cell genomic organization identified an additional exon that encodes an activation, as evidenced by the finding that CTLA-4-deficient mice alternative cytoplasmic domain, confirming the presence of at least experience profound peripheral lymphoid expansion (7–10). one downstream splice site (17). Taken together, these studies sup- It has been suggested that naturally occurring soluble forms of port the possibility that soluble B7 molecules could be generated both B7-1 and B7-2 may exist. For example, one study identified through alternative splicing. a soluble factor in the medium from primary porcine endothelial In this study, we describe the cloning of the porcine B7-1 ho- cells that mediates T cell activation (11). It was hypothesized that mologue. Furthermore, this is the first study of an alternatively this soluble factor is porcine B7-2 based on the observations that spliced form of B7-1 that encodes a soluble protein lacking both its activity is CD28 dependent and that B7-1 mRNA is not detect- the transmembrane and cytoplasmic domains. Soluble porcine able in these cells. An additional report demonstrated detectable B7-1 (sB7-1)4 maintains its ability to bind both CTLA-4 and CD28 levels of soluble human B7-1 in the synovial fluid of arthritic pa- and can function to abrogate human T cell activation. tients (12). However, it has not been determined whether these putative soluble forms of B7-1 and B7-2 are encoded from alter- Materials and Methods Abs and recombinant proteins Various mAbs reactive with porcine cell surface markers were purchased Alexion Pharmaceuticals, New Haven, CT 06511 from VMRD (Pullman, WA), including anti-CD3 (clone 8E6), anti-IgM Received for publication September 20, 1999. Accepted for publication March (clone PG145A), anti-class I (clone PT85A), and anti-class II (clone 28, 2000. MSA3). The anti-porcine B7-2 mAb was generated at Alexion Pharma- ceuticals (New Haven, CT). The functionally blocking anti-CD28 mAb 9.3 The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance was a kind gift of Dr. Carl June (Department of Molecular and Cellular with 18 U.S.C. Section 1734 solely to indicate this fact. Engineering, University of Pennsylvania, Philadelphia, PA), and the anti- 1 CD59 mAb MEM43 was obtained from Biodesign International (Kenne- This work was supported in part by a National Institutes of Standards and Tech- bunkport, ME). Rabbit antiserum directed against porcine B7-1 was gen- nology-Advanced Technology Program grant to W.L.F. erated by repeated s.c. immunization with sB7-1 containing six histidine 2 Current address: Immunology, Cancer and Infectious Disease, Central Research Division, Pfizer Inc., Eastern Point Road, Groton, CT 06340. 3 Address correspondence and reprint requests to Dr. Russell P. Rother, Alexion Phar- 4 Abbreviations used in this paper: sB7-1, soluble B7-1; hB7-1, human B7-1; PAEC, maceuticals, Molecular Development, 25 Science Park, Box 15, New Haven, CT porcine aortic endothelial cell; tmB7-1, transmembrane B7-1; UTR, untranslated 06511. region.

Copyright © 2000 by The American Association of Immunologists 0022-1767/00/$02.00 The Journal of Immunology 6341 residues (sB7-1-His), according to methods routinely performed at Co- The transmembrane form of porcine B7-1 (tmB7-1) was isolated by calico (Reamstown, PA). The IgG fractions from preimmune and immune RT-PCR of freshly isolated porcine lung RNA using an oligonucleotide sera were purified by passage over a protein A-Sepharose column (Phar- from the 3Ј end of the sB7-1 coding region as the 5Ј primer (GCTAC Ј macia, Piscataway, NJ). The rabbit anti-histidine polyclonal IgG was pur- CAACACGATGCTTTCC) and oligo(dT16) as the 3 primer. Conditions chased from Santa Cruz Biotechnology (Santa Cruz, CA). FITC-conju- for RNA isolation and RT-PCR were otherwise identical to those described gated goat anti-rabbit and goat anti-mouse, and PE-labeled goat anti-rabbit above. The two major products resulting from the RT-PCR were cloned secondary reagents were obtained from Zymed (South San Francisco, CA). into pCR2.1-TOPO, and inserts were sequenced for identification. Human CTLA-4Ig was purchased from Ancell (Bayport, MN), while recombinant porcine P-selectin-His was cloned and purified at Alexion Pharmaceuticals. Generation of His-tagged sB7-1 sB7-1 tagged with a carboxyl-terminal histidine hexapeptide (sB7-1-His) Cells and cell lines was generated in the mammalian expression vector Apex3P (19) by PCR amplification of B7-1 cDNA. The 5Ј primer (CCGGGGATCCCTTCTGTTT To purify PBL, heparinized porcine peripheral blood was obtained from TCATCCTCATCAAGC) was derived from the 5Ј untranslated region (UTR) adult swine (Cocalico). Human peripheral blood was collected from of B7-1 and contained a BamHI site for subcloning. The 3Ј primer (GGCCT healthy adult volunteers by venipuncture. Blood samples from either spe- GCAGGTCATCAATGGTGATGGTGATGGTGGCATTTTTGCCAGTTG cies were diluted 1/2 with HBSS (Life Technologies, Grand Island, NY), AAGGTCTGTGAC) inserted the histidine tag just upstream of the stop codon and the mononuclear fraction containing PBL was obtained by centrifuga- and an Sse83371 subcloning site. For stable expression of sB7-1-His, 293- tion over a Ficoll density gradient (Lymphocyte Separation Medium; Cell- EBNA embryonic kidney cells (Invitrogen, Carlsbad, CA) were transfected gro, Herndon, VA). Low density cells were collected from the interface and with sB7-1-His/Apex3P, as previously described (19). Cells were grown in washed repeatedly in PBS containing 5% heat-inactivated bovine FCS DMEM containing 5% FCS, 2 mM L-glutamine, 100 IU/ml penicillin, and (HyClone, Logan, UT). Viable cells were enumerated by trypan blue ex- 100 ␮g/ml streptomycin (D10 medium) with puromycin at a final concen- clusion. The human T cell line, Jurkat, and the human B cell line, Raji, Downloaded from tration of 1 ␮g/ml. Selected cells were cloned by limiting dilution, and were obtained from American Type Culture Collection (Manassas, VA). those producing high levels of protein were chosen by Western blot anal- Porcine aortic endothelial cells (PAEC) were purchased from Cell Systems ysis of cell supernatants using rabbit anti-histidine polyclonal IgG. The (Kirkland, WA). sB7-1-His protein was purified by affinity chromatography using a nickel- charged nitrilotriacetic acid resin (Qiagen, Chatsworth, CA), as previously Cloning of porcine B7-1 described (20). For transient transfections, the full-length untagged version of sB7-1 in pCR2.1-TOPO was subcloned into the mammalian expression Total RNA was prepared from freshly isolated porcine PBL using the acid/ vector Apex3P. sB7-1/Apex3P and sB7-1-His/Apex3P were then trans- http://www.jimmunol.org/ guanidinium thiocyanate technique (18). Ten micrograms of total RNA fected into 293 cells, as described above. Cells were grown in D10 medium were heated at 65°C for 3 min, quenched on ice, and subjected to first for 12 h and subsequently transferred into the serum-free medium HB PRO strand cDNA synthesis for1hat37°C in the following 100 ␮l reaction (Irvine Scientific, Santa Ana, CA) for 36 h. Cell supernatants were then mixture: 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2,10mM ␮ collected for Western blot analysis. dithiothreitol, 0.20 mM of each dNTP, 0.5 g oligo(dT16), and 20 U of avian myeloblastosis virus reverse transcriptase (Seikagaku, Rockville, MD). The following oligonucleotide primers synthesized at Oligos Etc. Northern blot analysis (Wilsonville, OR) were generated from regions of high homology between human and mouse B7-1 sequence: 1) 5Ј-TGGCCCGAGTATAAGAAC Total RNA from porcine alveolar macrophages was kindly provided by Dr. CGGAC-3Ј and 2) 5Ј-TCAGTTTCAGGATCTTGGGAAA-3Ј.A5-␮l al- Michael Murtaugh. Northern blot analysis was performed using the North- iquot of the cDNA pool was used as a template in a 100 ␮l PCR reaction ernMax Kit based on the manufacturer’s protocol (Ambion, Arlington by guest on September 29, 2021 under the following reaction conditions: 50 mM KCl, 10 mM Tris-HCl (pH Heights, IL) and a total of 10 ␮g of RNA per lane. Blots were hybridized ␮ ␣ 32 9), 1.5 mM MgCl2, 0.1% (w/v) gelatin, 1% Triton X-100, 200 M each with various [ - P]UTP-labeled RNA probes that were generated using dNTP, 2.5 U Taq DNA polymerase (Perkin-Elmer Cetus, Norwalk, CT), the MAXIscript In Vitro Transcription Kit (Ambion). RNA transcripts and 25 pmol of each primer. PCR amplification was performed for 30 were synthesized from B7-1 DNA fragments contained in pCR2.1-TOPO. cycles (94°C for 1 min, 50°C for 1 min, 72°C for 1 min), followed by a 1 These fragments consisted of either the complete 3Ј UTR of sB7-1 (soluble cycle extension step at 72°C for 10 min. The resulting 338-bp fragment was probe), the complete transmembrane and cytoplasmic regions of tmB7-1 cloned into the pCR2.1 vector using the T/A cloning system, as described (transmembrane probe), or the extracellular domain, which is common to by the manufacturer (Invitrogen, San Diego, CA) and identified by se- both the soluble and transmembrane forms of B7-1 (common probe). Lev- quence analysis as a B7-1 homologue. Two gene-specific oligonucleotides els of each transcript were determined by densitometry using a Gel Doc were derived from the porcine B7-1 sequence, and a 250-bp fragment was 1000 (Bio-Rad, Hercules, CA) and NIH Image 1.61 software (downloaded generated by PCR. This DNA fragment was used to screen a ␭gt10 porcine from http://rsb.info.nih.gov/nih-image/download.html). macrophage library (a generous gift from Dr. Michael Murtaugh, Depart- ment of Veterinary PathoBiology, University of Minnesota, St. Paul, MN). To screen the ␭gt10 porcine macrophage library, approximately 1 ϫ 106 PMA/ionomycin activation of porcine PBL phage were isolated on nitrocellulose filters (Schleicher & Schuell, Keene, Porcine PBL intended for use as stimulator cells in mixed allogeneic or NH). Filters were denatured for 1.5 min (1.5 M NaCl and 0.5 N NaOH), xenogeneic lymphocyte cultures (see below) were resuspended in R10 me- neutralized for 5 min (1.5 M NaCl and .05 M Tris, pH 8.5), rinsed in 3ϫ Ϫ5 dium (RPMI containing 5 ϫ 10 M 2-ME, 10 mM L-glutamine, 100 SSC, air dried, and UV cross-linked in a UV Stratalinker 2400 (Stratagene, IU/ml penicillin, 100 ␮g/ml streptomycin, and 10% FCS) supplemented La Jolla, CA). Filters were then prehybridized in BSA/SDS buffer (1% with 1 ng/ml PMA (Sigma, St. Louis, MO) and 400 ng/ml ionomycin BSA, 7% SDS, 0.5 M sodium phosphate buffer, pH 6.8, and 1 mM EDTA) (Sigma). The cells were seeded in 24-well plates at 2.5 ϫ 106 cells/ml in for2hat65°C before addition of the porcine B7-1 fragment, previously a final volume of 2 ml, and incubated for 48–72 h at 37°C in 5% CO in labeled with ␣-32P (NEN, Pittsburgh, PA) using the Prime-It II random 2 air. Mitomycin C (50 ␮g/ml; Sigma) was added to the cells during the last primer kit (Stratagene) to a sp. act. of 1 ϫ 109 cpm/␮g of DNA. Mem- 30 min of culture. Activated cells were then harvested and washed exten- branes were hybridized at 60°C overnight and subsequently washed using sively to remove residual PMA/ionomycin and/or mitomycin C. Porcine the following conditions: two 30-min washes with 2ϫ SSC/0.1% SDS at PBL used in FACS analysis were stimulated with 20 ng/ml PMA and 200 room temperature, one 30-min wash with 0.5ϫ SSC/0.1% SDS at 50°C, ng/ml ionomycin for 48 h. and one 30-min wash with 0.2ϫ SSC/0.1% SDS at 60°C. Positive plaques present on duplicate lifts were purified and the B7-1 DNA was rescued by PCR using primers that flanked the insertion site of the ␭gt10 vector (Clon- Immunofluorescence and flow cytometry tech, Palo Alto, CA). After cloning the PCR fragment into pCR2.1-TOPO, both strands of the putative full-length clone were sequenced using the Activated cell populations were preincubated in PBS containing 5% goat chain termination method. Clones derived from different PCR reactions serum. Cells were then incubated with purified IgG from hyperimmune were also sequenced to rule out potential errors induced during amplifica- serum from a rabbit immunized with sB7-1-His (rabbit anti-porcine-B7-1) tion. The DNA templates were primed with vector sequence primers flank- or from preimmune rabbit serum, in combination with murine mAbs di- ing the multiple cloning site, or primers constructed from internal cDNA rected against porcine CD3, IgM, class II, and B7-2 cell surface Ags, in sequence. All clones isolated from the ␭gt10 porcine macrophage library PBS containing 2% goat serum. Cells were washed in the same and then were identified as soluble porcine B7-1 (sB7-1). reacted with FITC-labeled goat anti-mouse Ig and with PE-labeled goat 6342 IDENTIFICATION AND ANALYSIS OF A SOLUBLE FORM OF B7-1 anti-rabbit Ig. The cells were again washed and analyzed for surface im- mediating experimental allograft rejection (23, 24). In addition, an munofluorescence using a FACSort flow cytometer and CellQuest Soft- important role for these molecules in cellular xenograft rejection is ware (Becton Dickinson, Mountain View, CA). Further analysis was per- suggested by data showing enhanced graft survival when formed using WinMDI version 2.7 software (provided by Dr. Joseph Trotter, University of San Diego, CA). CD28-B7 interactions are inhibited (25). The demonstration of In experiments performed to assess the binding of sB7-1 to the surface functional CD28-B7 interactions across the species barrier in the of Jurkat cells, sB7-1-His (2.5 ␮g/ml) was preincubated in HBSS contain- potentially clinically relevant porcine to human xenotransplant ing either anti-CD28 mAb (25 ␮g/ml), anti-CD59 mAb (20 ␮g/ml), anti- model could have significant therapeutic implications, given the CD3 mAb (20 ␮g/ml), human CTLA-4Ig (20 ␮g/ml), or buffer alone before addition to Jurkat cells (2.5 ϫ 105 cells/reaction) for an additional demonstrated potency of the human anti-pig cellular immune incubation. Cells were then incubated with purified rabbit anti-porcine- response. B7-1 IgG (10 ␮g/ml), washed in HBSS, and finally incubated in the FITC- A recent study indicated that porcine B7-2 is recognized by conjugated goat anti-rabbit secondary reagent (1/100 dilution). All incu- human CD28, and that this interaction promotes activation of hu- bations were performed for 30 min at 4°C. Jurkat/sB7-1-His binding was man T cells by porcine APCs (26). To further investigate the im- detected by cell surface immunofluorescence and flow cytometry, as described above. portance of porcine B7 molecules in the human anti-pig immune response, porcine B7-1 was cloned from a ␭gt10 library generated Costimulation assays from porcine lung macrophages. In the initial cloning attempts, The Jurkat T cell costimulatory assay has been described previously (21, only cDNA encoding sB7-1 was obtained from the macrophage 22). Briefly, PAEC were seeded in wells of 96-well microtiter plates at 5 ϫ library, at a frequency of approximately one clone per 1 ϫ 105 104 cells/well in R10 medium, and cells were allowed to adhere overnight phage. This full-length clone lacked both the transmembrane and at 37°C. Jurkat T cells (1 ϫ 106 cells/well) were then added to the wells in ␮ cytoplasmic domains. sB7-1 cDNA contained 1206 bp comprised Downloaded from the presence or absence of 10 g/ml PHA (Sigma) and either serial dilu- Ј Ј tions of sB7-1-His, 50 ␮g/ml of human CTLA-4Ig, or 100 ␮g/ml of por- of a 304-bp 5 UTR, a 215-bp 3 UTR including a processing/ cine P-selectin-His. In some experiments, Raji cells (1 ϫ 106 cells/well) polyadenylation signal, and an open reading frame that encoded were substituted for PAEC, but were added to the Jurkat cells at the ini- 229 aa (data not shown; GenBank accession number AF203442). tiation of the experiment. The cultures were maintained at 37°C in 5% CO2 The abnormally long 5Ј UTR observed for sB7-1 corresponded to for 24 h, at which time the culture supernatants were harvested. IL-2 was measured in supernatants using an ELISA kit (Quantikine Human IL-2 that of B7-1 reported for other species (4, 27). To obtain tmB7-1,

Immunoassay; R&D Systems, Minneapolis, MN), according to the manu- RT-PCR was performed on porcine lung RNA using a 5Ј primer http://www.jimmunol.org/ facturer’s protocol. Briefly, serial dilutions of a human rIL-2 standard generated from the end of the sB7-1 coding region and oli- (R&D Systems) or culture supernatants were added, in duplicate, to ELISA go(dT16). A B7-1-specific band of approximately 340 bp was gen- wells that had been previously coated with a mAb specific for human IL-2, erated that encoded a putative transmembrane domain and a partial and plates were incubated overnight at 4°C. Unbound cytokine was re- moved by repeated washes. Bound IL-2 was detected using a second IL- cytoplasmic domain, but lacked sequence encoding the transla- 2-specific Ab conjugated with HRP, followed by addition of a hydrogen tional stop site and the 3Ј UTR (data not shown; GenBank acces- peroxide/chromogen substrate. The reaction was stopped by the addition of sion number AF203443). The truncation of tmB7-1 cDNA derived 2 N sulfuric acid. The OD of each well was determined using a microtiter from reverse-transcribed porcine lung RNA and the lack of detec- plate reader (Bio-Rad model 3550, Hercules, CA) set to 450 nm, with values corrected by subtraction of readings taken at 595 nm. IL-2 concen- tion of tmB7-1 in the oligo(dT)-primed porcine macrophage li- trations were calculated using Microplate Manager software (Bio-Rad). brary suggest strong 3Ј UTR secondary structure in this transcript. by guest on September 29, 2021 The predicted amino acid sequences for sB7-1, tmB7-1, and Mixed lymphocyte reactions human B7-1 (hB7-1) were compared (Fig. 1A). Sequences were Stimulator cells (mitomycin C-treated allogeneic human PBL or PMA/ segregated into domains based on exon boundaries identified for ionomycin-activated porcine PBL; 5 ϫ 105 cells/well) and responder lym- hB7-1 (15). Considering that sB7-1 and tmB7-1 amino acid se- ϫ 5 phocytes (human PBL; 5 10 cells/well) were combined in wells of a quences were identical before the transmembrane domain, only 96-well microtiter plate in the presence or absence of serially diluted sB7- 1-His or of the indicated concentrations of CTLA-4Ig or porcine P-selec- sB7-1 is depicted before this region. Excluding the transmembrane tin-His (final volume, 0.2 ml/well). Cells were maintained in R10 medium and cytoplasmic domains, which are highly divergent between spe- 3 ␮ for 4–5 days at 37°C in 5% CO2 in air. [ H]thymidine (1 Ci/well; NEN cies, porcine B7-1 and hB7-1 shared 65% sequence identity and an DuPont, Boston, MA) was added to the cell cultures during the last 16–18 overall conservation of the Ig V-like and Ig C-like structural do- h of the incubation. The cells were harvested onto glass fiber filters with an mains characteristic of other B7 molecules (28). The signal peptide automated sample harvester (Wallac, Turku, Finland) and the filters were counted in a beta liquid scintillation counter (Wallac). for sB7-1 was 29 aa in length, as determined by amino-terminal sequencing of purified protein. Virtually all amino acid residues Western blot analysis that have been shown to be critical for the binding of B7-1 to both Supernatants from 293 cells transiently transfected with sB7-1/Apex3P, CD28 and CTLA-4 (excluding methionine 47 and isoleucine 49) sB7-1-His/Apex3P, or Apex3P alone were subjected to SDS-PAGE (4– were highly conserved (29–33). A clone containing the complete 12% gradient gels) under reducing or nonreducing conditions. Proteins coding region for tmB7-1 was not found, but based on sequence were transferred to nitrocellulose and the membrane was blocked for1hin comparison with various other species, the terminal amino acid is blocking solution (Tris-buffered saline containing 5% dry milk). Blots were incubated for1hinfresh blocking solution containing either rabbit anti- expected to be very close to the translational stop site. histidine polyclonal IgG (0.2 mg/ml) or purified anti-porcine B7-1 poly- Various splice variants have been reported for the B7 molecules, clonal IgG (2 mg/ml). Blots were then washed three times with Ab wash none of which encode a soluble product (13, 14, 17). To examine solution (500 mM NaCl, 35 mM Tris, pH 7.4, 0.5 mM CaCl 2, 0.1% SDS, the splicing mechanism that generated sB7-1, the nucleic acid and 1% Nonidet P-40, and 0.5% deoxycholic acid) before the addition of fresh blocking solution containing HRP-conjugated goat anti-rabbit secondary amino acid sequences corresponding to the junction of exons 4 and Ab (1:5000; Zymed) for 15 min. Finally, blots were washed three times in 5 of hB7-1 were compared with porcine sB7-1 and tmB7-1 (Fig. Ab wash solution, incubated for 1 min in ECL Western blot reagent, and 1B). Sequence identity between the alternative forms of porcine exposed to ECL Hyperfilm (both from Amersham, Arlington Heights, IL). B7-1 were identical in exon 4, but showed no homology beginning with exon 5. A stop codon was generated in the beginning of exon Results 5 for sB7-1. These data suggest that sB7-1 is a splice variant lack- Molecular cloning and sequence analysis of porcine B7-1 ing exons coding for both the transmembrane and cytoplasmic do- Signals generated by the interactions of CD28 and CTLA-4 with mains and thus represent the first report of a naturally occurring the B7 molecules have been shown to be critically involved in soluble form of B7-1. The Journal of Immunology 6343 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 1. Sequence comparisons between porcine and human B7-1. A, Amino acid sequences of sB7-1, tmB7-1, and hB7-1 were aligned based on amino acid identity and structural similarity. Identical amino acids are denoted by asterisks, and gaps in the sequences are indicated by dashes. Assignment of structural domains is based on exon boundaries published for hB7-1 and are identified by the following abbreviations: signal peptide, SP; Ig variable-like domain, Ig V-like; Ig constant-like domain, Ig C-like; transmembrane domain, TM; cytoplasmic domain, CYT. The signal peptide is depicted by single underline. The transmembrane domain was determined using the PSORT II program (http://psort.nibb.ac.jp) and is indicated by dashed underline. Sites known to be critical for B7-1 binding to CD28 and/or CTLA-4 are shown by double underline. Translational stop sites are indicated by closed diamonds, while the closed circles at the end of the tmB7-1 sequence indicate that the stop codon was not identified in this molecule. The transmembrane and cytoplasmic domains are absent in sB7-1. B, Partial nucleic acid sequence flanking the junction site of hB7-1 exons 4 and 5 was aligned with sB7-1 and tmB7-1 based on sequence similarities. Encoded amino acid sequences for each are also shown. Protein domains corresponding to exons 4 and 5 are depicted as extracellular or transmembrane. The translational stop site for sB7-1 is indicated by an asterisk.

Northern analysis of porcine B7-1 splice variants rophage library. The tmB7-1-specific probe hybridized with a transcript of approximately 3 kb, which represents the membrane form To determine the relative levels of expression of sB7-1 and of porcine B7-1. As expected, the tmB7-1-specific probe did not tmB7-1 transcripts, Northern blot analysis was performed on total recognize the 1.3-kb species. Finally, the common probe hybrid- RNA isolated from porcine alveolar macrophages. Reactive RNA ized with both the 1.3- and 3-kb species. The relative densities of species were detected with either a sB7-1-specific probe, a tmB7- the two species reactive with the common probe were determined 1-specific probe, or a probe common to both sB7-1 and tmB7-1. using National Institutes of Health Image 1.61 software. The The sB7-1-specific probe hybridized with a single species of ap- tmB7-1 mRNA was present at a 3-fold excess relative to sB7-1. proximately 1.3 kb (Fig. 2). The size of this message approximates These data indicate that both sB7-1 and tmB7-1 transcripts are the size of the sB7-1 cDNA clone isolated from the porcine mac- well represented in porcine alveolar macrophages. 6344 IDENTIFICATION AND ANALYSIS OF A SOLUBLE FORM OF B7-1

FIGURE 2. Northern blot analysis of soluble and transmembrane forms of B7-1. Total RNA from porcine alveolar macrophages was electropho- resed, transferred to nitrocellulose, and hybridized with 32P-labeled RNA transcripts. Probes were generated from sequence restricted to sB7-1 (sol- Downloaded from uble) or tmB7-1 (transmembrane) or sequence common to both forms of the molecule (common). RNA size in kilobases is depicted on the left of the ﬁgure, while mRNA species that speciﬁcally hybridized with sB7-1 or tmB7-1 are indicated by arrows on the right.

Cell surface expression of porcine B7-1 http://www.jimmunol.org/ Most studies have indicated that the B7-1 costimulatory molecule is either undetectable or present on only a small subset of resting human B and T cells, but that it is dramatically up-regulated in these cell types upon stimulation by various means (34). To assess the cell surface expression of porcine B7-1 on various subsets of stimulated cells, porcine PBL were evaluated following stimulation with PMA/ionomycin using two-color immunoﬂuorescence

(B7-1 was detected by PE-, while other cell surface markers were by guest on September 29, 2021 detected by FITC-conjugated secondary reagents). Following stim- FIGURE 3. Expression of cell surface B7-1 on subpopulations of PMA/ ulation with PMA/ionomycin, the majority of CD3-, class II-, and ionomycin-stimulated porcine PBL. Porcine PBL were isolated from pe- B7-2-positive cells were also B7-1 positive (Fig. 3, B, C and E, ripheral blood samples and analyzed for the cell surface expression of B7-1 respectively). Virtually all IgM-positive cells were B7-1 positive following incubation with PMA/ionomycin for 48 h at 37°C. Cells were (Fig. 3D). These data suggest that a membrane-bound form of reacted with IgG purified from either preimmune rabbit serum (A) or polyclonal rabbit anti-porcine B7-1 (B–E). Lymphocyte subsets were identified porcine B7-1 is abundant on the surface of both peripheral T cells by reactivity with mAbs specific for porcine T cells (CD3; B), APC (class and APC following cell stimulation. II; C), B cells (IgM; D), or B7-2-expressing cells (E). Cells were then reacted with a combination of PE-conjugated goat anti-rabbit and FITC- sB7-1 binds human CD28 and CTLA-4 conjugated goat anti-mouse reagents (A–E). The major lymphoid popula- The ligands for several porcine adhesion and costimulatory mol- tions were identified on the basis of forward and side scatter characteristics, ecules have been shown to be conserved across species, including and two-color flow-cytometric analysis was performed on 10,000 gated humans. These include the ligands for porcine E-selectin, VCAM, cells. The percentage of positive cells enumerated is indicated. and B7-2 (20, 26, 35). The fact that amino acids generally shown to be critical for the binding of B7-1 to both CD28 and CTLA-4 are conserved in porcine B7-1 (Fig. 1A) suggests that this molecule cells (Fig. 4B). By contrast, an isotype-matched irrelevant mAb could interact with human CD28 and CTLA-4. To confirm this had no effect on sB7-1-His/CD28 binding. The interaction between prediction, purified sB7-1-His was incubated with human Jurkat porcine B7-1 and human CTLA-4 was also confirmed by ELISA, cells. Jurkat cells constitutively express CD28, but do not express as plate-coated sB7-1-His bound to CTLA-4Ig in a titratable fash- CTLA-4 under any culture conditions (36). sB7-1-His specifically ion (data not shown). These data demonstrate that a naturally oc- bound to Jurkat cells (Fig. 4A), but failed to bind a Jurkat deriv- curring soluble form of porcine B7-1 has the ability to bind both ative cell line that does not express CD28 (TIB 153, data not human CD28 and CTLA-4. Furthermore, the binding of porcine shown). The binding of sB7-1-His to CD28 was effectively B7-1 to human receptors suggests that this interaction could play blocked by an anti-CD28 mAb, further establishing the specificity a role in T cell activation during pig to human xenotransplantation. of this interaction. An isotype-matched irrelevant mAb did not interfere with sB7-1-His/CD28 binding. sB7-1 inhibits IL-2 production following costimulation of human It has been demonstrated that both CD28 and CTLA-4 bind to Jurkat T cells very similar sites on B7-1 (29). To analyze the ability of porcine To examine whether sB7-1 could interact functionally with human B7-1 to bind to CTLA-4, sB7-1-His was preincubated with human T cells, sB7-1-His was titrated into the Jurkat costimulation assay CTLA-4Ig before its addition to Jurkat cells. Human CTLA-4Ig and its effect on T cell activation was evaluated by detection of effectively inhibited the binding of sB7-1-His to CD28 on these IL-2 in the culture supernatants. Jurkat cells do not constitutively The Journal of Immunology 6345

FIGURE 4. Porcine B7-1 binding to the human receptors CD28 and Downloaded from CTLA-4. Human Jurkat cells that are known to express CD28, but not CTLA-4, were incubated with sB7-1-His, and binding was determined by FACS analysis (see Materials and Methods). A, Depicts the binding of sB7-1-His to Jurkat cells that have been previously incubated with buffer alone (No Ab), with a functionally blocking anti-CD28 mAb, or with an FIGURE 5. Soluble B7-1 blocks IL-2 production following costimula- irrelevant isotype-matched mAb (Anti-CD59). B, Shows the binding of tion of human T cells. Cells of the Jurkat line were incubated with PHA and sB7-1-His following preincubation with buffer alone (No Ab), with human either PAEC (A) or Raji cells (B) under the following costimulatory con- http://www.jimmunol.org/ CTLA-4Ig, or with an irrelevant isotype-matched mAb (Anti-CD3). The ditions: PAEC or Raji cells alone (F), or with the indicated concentrations binding of the FITC-conjugated secondary reagent in the absence of sB7- of CTLA-4Ig (Ⅺ), P-selectin-His (‚), or sB7-1-His (E). The cell cultures 1-His is shown in both panels (Control). Histograms represent data from a were maintained at 37°C for 24 h, at which time the supernatants were single experiment, one of three so performed. harvested and assayed by indirect sandwich ELISA to determine the concentration of hIL-2, as described in Materials and Methods. Results are expressed as the mean of duplicate ELISA determinations, converted to elaborate IL-2, even in the presence of PHA, which provides a pg/ml IL-2 using Microplate Manager software. primary signal through the TCR. However, PHA signaling in the presence of APCs or a stimulatory anti-CD28 mAb results in sig- by guest on September 29, 2021 nificant IL-2 production (37). of increasing amounts of sB7-1-His, an irrelevant histidine-tagged Jurkat cells were incubated alone or were cocultured with PAEC protein, or CTLA-4Ig, and responder cell proliferation was as- or Raji cells in the presence or absence of sB7-1-His, an irrelevant sessed on day 4–5 of culture. The addition of sB7-1-His effectively histidine-tagged protein or human CTLA-4Ig. As expected, in the inhibited human T cell proliferation in a dose-dependent fashion in absence of PAEC or Raji cells, Jurkat cells did not generate de- response to both allogeneic and xenogeneic stimulator cells (Fig. tectable IL-2 under any of the culture conditions tested (data not 6, A and B, respectively). Addition of human CTLA-4Ig at con- shown). By contrast, Jurkat cells generated high levels of IL-2 centrations of 100 ␮g/ml also effectively inhibited cell prolifera- when stimulated with PHA in the presence of either PAEC (Fig. tion in both assays, while addition of porcine P-selectin-His at 100 5A) or Raji cells (Fig. 5B). Addition of sB7-1-His inhibited the ␮g/ml had no effect. These results indicate that binding of porcine production of IL-2 in a dose-dependent manner, with maximal sB7-1-His to CD28 and/or CTLA-4 ligands on human T cells in- inhibition equivalent to that observed by the addition of human hibits their activation by allogeneic or xenogeneic stimulation in a CTLA-4Ig. Inhibition of IL-2 production was virtually complete at concentration-dependent manner. high doses of sB7-1-His, regardless of the source of APCs (porcine or human) used to provide the secondary signal. Addition of re- sB7-1 exists as both a monomer and dimer combinant porcine P-selectin-His did not significantly influence The generation of sB7-1 through alternative splicing creates a cys- IL-2 production in these assays. Nonspecific toxicity of the sB7- teine residue before a translational termination signal (Fig. 1). Al- 1-His preparation due to excipient effects was excluded by evalu- though this cysteine residue is included in the recombinantly pro- ating similar volumes of dialysis buffer, collected during the final duced protein, the addition of a histidine tag on the carboxyl dialysis of the protein (data not shown). Since Jurkat cells do not terminus of sB7-1 could affect potential disulfide bond formation, express CTLA-4, these data suggest that sB7-1-His inhibits IL-2 and thus dimerization of the molecule. To investigate the ability of production by binding to CD28 and blocking signaling through the native molecule to dimerize, sB7-1 was transiently expressed this molecule. in 293 cells and compared with sB7-1-His under reducing and nonreducing gel electrophoresis, followed by Western blot analy- sB7-1 inhibits human allogeneic and xenogeneic MLRs sis. Under reducing conditions, both proteins migrated as a doublet To determine the potential effect of sB7-1-His on T cell activation at approximately 40 kDa, with the histidine-tagged version of the under conditions in which both CD28 and CTLA-4 molecules protein running just above the untagged version, most likely due to were present, MLRs were performed. Short-term primary human the hexapeptide tag (Fig. 7B, lanes 2 and 3, respectively). Con- allogeneic and xenogeneic MLRs were established by coculturing versely, while sB7-1-His ran identically under nonreducing con- human PBL with mitomycin C-treated human (Fig. 6A) or porcine ditions, the untagged version of sB7-1 showed an additional major (Fig. 6B) PBL, respectively. MLRs were performed in the presence species at approximately 80 kDa, presumably due to dimerization 6346 IDENTIFICATION AND ANALYSIS OF A SOLUBLE FORM OF B7-1

FIGURE 7. Multimericity of native and histidine-tagged forms of sB7-1. sB7-1-His and an untagged version of sB7-1 were transiently expressed in 293 cells, subjected to SDS-PAGE under reducing (A and B)or nonreducing (C) conditions, and transferred to nitrocellulose for Western blot analysis. Reactivity of cell supernatant proteins from 293 cells transfected with vector alone (lane 1), sB7-1-His cDNA (lane 2), or cDNA from an untagged version of sB7-1 (lane 3) is shown. Membranes were reacted with either rabbit anti-histidine polyclonal IgG (A) or puriﬁed anti-porcine B7-1 polyclonal IgG (B and C). Protein size in kDa is depicted on the right

of the ﬁgure. Downloaded from

soluble molecules maintain the ability to bind their ligands (38). Furthermore, an anti-inﬂammatory role for soluble P-selectin has been suggested by evidence that it can inhibit both CD18-depen-

dent neutrophil adhesion to endothelium and neutrophil superoxide http://www.jimmunol.org/ release (39, 40). Potential immune modulation by a soluble cytokine receptor has also been suggested by the observation that mutations in the extracellular domain of the 55-kDa TNFR-1 in in- dividuals with autoinflammatory syndrome result in a decrease in the shedding of the functionally antagonistic soluble form of the FIGURE 6. Soluble B7-1 inhibits lymphocyte proliferation in an allo- receptor (41). geneic and xenogeneic MLR. Porcine PBL were stimulated with PMA/ The naturally occurring soluble form of B7-1 described in the ionomycin for 48–72 h, then treated with mitomycin C before their use as present study represents a potential regulatory component of the stimulator cells in MLRs, as described in Materials and Methods. Human by guest on September 29, 2021 PBL were isolated and seeded at 5 ϫ 105 cells/well in triplicate wells of immune system that heretofore has not been described. A histi- 96-well plates in the presence or absence of 5 ϫ 105 similarly prepared and dine-tagged version of sB7-1 effectively bound CD28 and CTLA-4 mitomycin C-treated human allogeneic stimulator cells (A) or porcine PBL molecules and blocked T cell activation in both allogeneic and (B). Cells were maintained at 37°C for 4–5 days, with [3H]thymidine xenogeneic MLRs. B7-1 has previously been expressed as a re- added to the cell cultures during the last 16–18 h of incubation. Cells were combinant soluble molecule by replacing the transmembrane and harvested onto glass fiber filters and assayed for the incorporation of cytoplasmic domains with the Fc region of an Ab or with an oligo- [3H]thymidine, as described in Materials and Methods. Human PBL pro- histidine tag (2, 42, 43). These recombinant soluble B7-1 mole- E liferation in response to stimulator cells was evaluated in the absence ( ) cules (rsB7-1) also maintain their ability to bind both CD28 and F f or presence of the indicated amounts of sB7-1-His ( ), CTLA-4Ig ( ), or CTLA-4 (2, 42–44). In addition, plate-immobilized rsB7-1 effec- P-selectin-His (Ⅺ). Proliferation of human responder cells alone (छ)or tively stimulates T cells in conjunction with a primary signal (anti- human or porcine stimulator cells alone (‚) is indicated. Results are expressed as the mean cpm of triplicate wells. CD3 mAb) (2, 43). The ability of these recombinant molecules to block T cell activation was not tested. Inhibition of T cell activation by sB7-1-His most likely occurs through the inhibition of CD28 binding to B7-1 and/or B7-2 on the of the molecule through cysteine bonding (Fig. 7C, lanes 2 and 3, APCs. Since it has been shown that Ag presentation in the absence respectively). As expected, Ab reactive against the histidine tag of costimulation through CD28 results in T cell anergy (5, 6), it is recognized sB7-1-His, but not the untagged version of the protein interesting to speculate that endogenous sB7-1 may exert this same (Fig. 7A, lanes 2 and 3, respectively). No reactivity was observed effect in vivo by blocking the binding of CD28 to its ligands. in samples prepared from 293 cells transfected with vector alone Blockade of CD28 interactions with the B7 proteins may also re- (Fig. 7, all panels, lane 1). These results suggest that native sB7-1 duce the production of Bcl-x , a molecule important in preventing may exist as both a monomer and dimer. Furthermore, preliminary L apoptosis, as engagement of CD28 has been shown to promote the studies with sB7-1 preparations containing primarily homodimer production of this survival factor (45). indicate that this multimer also functions to block T cell activation In the present study, the inclusion of a carboxyl-terminal histi- in an MLR (data not shown). dine hexapeptide prevented B7-1 disulfide-linked dimer formation, as an untagged version of the molecule migrated primarily as a Discussion dimer under nonreducing electrophoretic conditions, while the his- The physiological significance of soluble forms of otherwise trans- tidine-tagged version ran entirely as a monomer. Apparently, the membrane-linked proteins is not known. However, the levels of highly charged histidine tag restricts disulfide bond formation be- soluble adhesion molecules and cytokine receptors, whether prod- tween cysteine residues found at the carboxyl end of the molecule. ucts of alternative splicing or enzymatic cleavage, increase during This finding raises the question of whether sB7-1-His may be func- inflammation, infection, or malignancies, and the majority of these tionally different from the untagged version of the molecule. Our The Journal of Immunology 6347 preliminary studies with sB7-1 preparations containing primarily curring at less than 14 copies per cell) with an average occurrence homodimer indicate that the homodimer also blocks T cell activa- in an oligo(dT) reverse-transcribed library (49). tion in an MLR (data not shown). In addition, the recent resolution Recently, a soluble form of hB7-1 has been identified in the of the crystal structure of human rsB7-1 indicates that this mole- synovial fluid of arthritic patients (12). It is not known whether this cule undergoes a rapid monomer-dimer exchange that favors non- soluble molecule is also encoded by an alternatively spliced mes- disulfide-bonded homodimer formation (46). These data suggest sage or rather, represents a result of enzymatic cleavage from the the possibility that sB7-1-His utilized in the present study may also cell surface. Northern blot analysis has identified at least four dif- exist as a homodimer. ferent hB7-1 transcripts that may represent alternatively spliced The ability of sB7-1 to form a stable homodimer may have other molecules (4). Using RT-PCR, we have also cloned a putative important functional consequences. Based on crystal structure alternatively spliced form of hB7-1 that encodes a protein lacking data, it has been predicted that the increase in avidity between the transmembrane and cytoplasmic domains (data not shown). B7-1 and CTLA-4 homodimers may serve to stabilize the B7-1/ However, this species was not detectable by Northern blot analy- CTLA-4 signaling complex, which would facilitate the termination sis, suggesting either that it was not present in the source of RNA of T cell activation (46). It is interesting to speculate that cysteine analyzed or it represented a rare transcript. The relevance of this bond formation between sB7-1 molecules would further favor the alternatively spliced human soluble B7-1 is currently under generation of stable homodimers. However, it is not known investigation. whether a sB7-1 homodimer would activally signal through Based on anatomical and physiological considerations, the pig is CTLA-4 or alternatively, block CTLA-4 signaling by competing considered the most likely candidate as a xenogeneic organ donor with B7-1 on the surface of the APC. Definitive studies addressing for human transplants. However, cellular mechanisms of rejection Downloaded from the physiologic function(s) of sB7-1 are currently underway. of pig organs mediated by recognition of porcine adhesion and/or Alternative splicing of a heteronuclear transcript can result in costimulatory molecules are likely to represent at least one major the generation of a soluble molecule from an otherwise membrane- barrier to the success of such transplants (50). For example, por- linked protein. This generally occurs by removal of an exon coding cine E-selectin, VCAM, and B7-2 have all been shown to interact for the transmembrane domain by use of an alternative down- with the human homologues of their receptors and to mediate hu-

stream splice site. Examples of soluble proteins that are generated man leukocyte adhesion or T cell costimulation (20, 26, 35). In the http://www.jimmunol.org/ through this mechanism include LFA-3, P-selectin, and many of present study, we show that porcine B7-1 is recognized by the the cytokine receptors (38, 47). Alternatively spliced products human receptors CD28 and CTLA-4 and that it can also serve as have also been described for both B7-1 and B7-2 (13, 14), but to a ligand for human T cell costimulation by porcine APCs. These our knowledge, this is the first report of an abundant endogenous data suggest that porcine B7-1 may represent a good candidate for mRNA that lacks the transmembrane and cytoplasmic domains blockade or targeted gene knockout in the xenotransplantation set- and encodes a functional, soluble B7 protein. ting. However, the functional redundancy described for B7-1 and Evidence that sB7-1 reported in the present study is indeed an B7-2 suggests that blockade of B7-1 alone may not be sufficient to alternatively spliced product and not an incompletely processed adequately inhibit T cell activation. Indeed, the combination of mRNA with the stop codon generated from unspliced intron se- functionally blocking anti-murine B7-1 and B7-2 mAbs is required by guest on September 29, 2021 quence includes the following: 1) the 1.2-kb full-length cDNA for to effectively inhibit primary MLRs and allogeneic transplant re- sB7-1 corresponds in size to one of two species identified by jection, while inhibition with anti-B7-1 mAbs alone does not (24, Northern blot analysis of RNA derived from porcine macrophages, 51, 52). Thus, we are currently investigating the effectiveness of and a probe generated from sequence specific to sB7-1 (3Ј UTR functionally blocking mAbs to porcine B7-1 and B7-2 either in- sequence) recognized this transcript; 2) a known processing/poly- dependently or in combination during xenotransplantation. adenylation signal is found at the end of the 3Ј UTR sequence of sB7-1; 3) the sB7-1 message is well represented in the porcine Acknowledgments macrophage cDNA library; and 4) there is an absence of a splice We thank Aaron Bourret for technical assistance, and Dr. Scott Rollins for donor site at the point that the sB7-1 cDNA sequence diverges helpful discussions. from that of tmB7-1. In addition, using PCR analysis of porcine genomic DNA, we demonstrated that the end of the sB7-1 coding References region is not juxtaposed to the downstream untranslated region, 1. Mueller, D. L., M. K. Jenkins, and R. H. Schwartz. 1989. Clonal expansion providing further support that this sequence is not unspliced intron versus functional inactivation: a costimulatory signaling pathway determines the (data not shown). outcome of T cell antigen receptor occupancy. Annu. Rev. Immunol. 7:445. A tmB7-1-specific probe recognized a mRNA species of ap- 2. Linsley, P. S., W. Brady, L. Grosmaire, A. Aruffo, N. K. Damle, and J. A. Ledbetter. 1991. Binding of the B cell activation antigen B7 to CD28 proximately 3 kb by Northern blot analysis, which corresponds to costimulates T cell proliferation and interleukin 2 mRNA accumulation. J. Exp. that reported previously for stimulated porcine lymph node cells Med. 173:721. (11). However, this study did not demonstrate the presence of the 3. Freeman, G. J., J. G. Gribben, V. A. Boussiotis, J. W. Ng, V. A. Restivo, L. A. Lombard, G. S. Gray, and L. M. Nadler. 1993. Cloning of B7-2: a CTLA-4 sB7-1 transcript in lymph node cell RNA, which may indicate its counter-receptor that costimulates human T cell proliferation. Science 262:909. absence or low level of expression in this tissue source. Differen- 4. Freeman, G. J., A. S. Freedman, J. M. Segil, G. Lee, J. F. Whitman, and L. M. Nadler. 1989. B7, a new member of the Ig superfamily with unique ex- tial expression of alternatively spliced B7 products in various cell pression on activated and neoplastic B cells. J. Immunol. 143:2714. types and in response to different stimuli has been previously re- 5. Harding, F. A., J. G. McArthur, J. A. Gross, D. H. Raulet, and J. P. Allison. 1992. ported (13, 48). CD28-mediated signaling co-stimulates murine T cells and prevents induction of anergy in T cell clones. Nature 356:607. Although mRNA for sB7-1 was not as abundant as the tmB7-1 6. Gimmi, C. D., G. J. Freeman, J. G. Gribben, G. S. Gray, and L. M. Nadler. 1993. transcript (3-fold less), it was well represented in the alveolar mac- Human T-cell clonal anergy is induced by antigen presentation in the absence of rophage library. The complete absence of tmB7-1 cDNAs allowed B7 costimulation. Proc. Natl. Acad. Sci. USA 90:6586. 7. Kearney, E. R., T. L. Walunas, R. W. Karr, P. A. Morton, D. Y. Loh, the total number of clones representing sB7-1 to be determined. Of J. A. Bluestone, and M. K. Jenkins. 1995. Antigen-dependent clonal expansion of a total of approximately 3 ϫ 106 plated phage, approximately 30 a trace population of antigen-specific CD4ϩ T cells in vivo is dependent on CD28 ϫ 5 costimulation and inhibited by CTLA-4. J. Immunol. 155:1032. sB7-1 clones were isolated (1 clone per 1 10 phage). This 8. Krummel, M. F., and J. P. Allison. 1995. CD28 and CTLA-4 have opposing frequency is indicative of a low-abundance mRNA (mRNAs oc- effects on the response of T cells to stimulation. J. Exp. Med. 182:459. 6348 IDENTIFICATION AND ANALYSIS OF A SOLUBLE FORM OF B7-1

9. Walunas, T. L., D. J. Lenschow, C. Y. Bakker, P. S. Linsley, G. J. Freeman, 32. Peach, R. J., J. Bajorath, J. Naemura, G. Leytze, A. Aruffo, and P. S. Linsley. J. M. Green, C. B. Thompson, and J. A. Bluestone. 1994. CTLA-4 can function 1995. Both extracellular immunoglobulin-like domains of CD80 contain residues as a negative regulator of T cell activation. Immunity 1:405. critical for binding T cell surface receptors CTLA-4 and CD28. J. Biol. Chem. 10. Tivol, E. A., F. Borriello, A. N. Schweitzer, W. P. Lynch, J. A. Bluestone, and 270:21181. A. H. Sharpe. 1995. Loss of CTLA-4 leads to massive lymphoproliferation and 33. Guo, Y., Y. Wu, X. Kong, and M. Liu. 1998. Identification of conserved amino fatal multiorgan tissue destruction, revealing a critical negative regulatory role of acids in murine B7-1 IgV domain critical for CTLA4/CD28:B7 interaction by CTLA-4. Immunity 3:541. site-directed mutagenesis: a novel structural model of the binding site. Mol. Im- 11. Davis, T. A., N. Craighead, A. J. Williams, A. Scadron, C. H. June, and K. P. Lee. munol. 35:215. 1996. Primary porcine endothelial cells express membrane-bound B7-2 (CD86) 34. Lenschow, D. J., T. L. Walunas, and J. A. Bluestone. 1996. CD28/B7 system of and a soluble factor that co-stimulate cyclosporin A-resistant and CD28-depen- T cell costimulation. Clin. Immunol. Immunopathol. 14:233. dent human T cell proliferation. Int. Immunol. 8:1099. 12. McHugh, R. S., W. D. Ratnoff, R. Gilmartin, K. W. Sell, and P. Selvaraj. 1998. 35. Rollins, S. A., M. J. Evans, K. K. Johnson, E. A. Elliot, S. P. Squinto, L. A. Matis, Detection of a soluble form of B7-1 (CD80) in synovial fluid from patients with and R. P. Rother. 1994. Molecular and functional analysis of porcine E-selectin arthritis using monoclonal antibodies against distinct epitopes of human B7-1. reveals a potential role in xenograft rejection. Biochem. Biophys. Res. Commun. Clin. Immunol. Immunopathol. 87:50. 204:763. 13. Borriello, F., J. Oliveros, G. J. Freeman, L. M. Nadler, and A. H. Sharpe. 1995. 36. Freeman, G. J., D. B. Lombard, C. D. Gimmi, S. A. Brod, K. Lee, J. C. Laning, Differential expression of alternate mB7-2 transcripts. J. Immunol. 155:5490. D. A. Hafler, M. E. Dorf, G. S. Gray, H. Reiser, et al. 1992. CTLA-4 and CD28 14. Inobe, M., P. S. Linsley, J. A. Ledbetter, Y. Nagai, M. Tamakoshi, and T. Uede. mRNA are coexpressed in most T cells after activation: expression of CTLA-4 1994. Identification of an alternatively spliced form of the murine homologue of and CD28 mRNA does not correlate with the pattern of lymphokine production. B7. Biochem. Biophys. Res. Commun. 200:443. J. Immunol. 149:3795. 15. Selvakumar, A., B. K. Mohanraj, R. L. Eddy, T. B. Shows, P. C. White, and 37. Williams, T. M., D. M. Moolten, H. Makni, H. W. Kim, J. A. Kant, and B. Dupont. 1992. Genomic organization and chromosomal location of the human M. Kamoun. 1992. CD28-stimulated IL-2 gene expression in Jurkat T cells oc- gene encoding the B-lymphocyte activation antigen B7. Immunogenetics 36:175. curs in part transcriptionally and is cyclosporine-A sensitive. J. Exp. Med. 148: 16. Selvakumar, A., P. C. White, and B. Dupont. 1993. Genomic organization of the 2609. mouse B-lymphocyte activation antigen B7. Immunogenetics 38:292. 38. Meager, A., C. Bird, and A. Mire-Sluis. 1996. Assays for measuring soluble 17. Borriello, F., G. J. Freeman, S. Edelhoff, C. M. Disteche, L. M. Nadler, and cellular adhesion molecules and soluble cytokine receptors. J. Immunol. Methods Downloaded from A. H. Sharpe. 1994. Characterization of the murine B7-1 genomic locus reveals 191:97. an additional exon encoding an alternative cytoplasmic domain and a chromo- 39. Gamble, J. R., M. P. Skinner, M. C. Berndt, and M. A. Vadas. 1990. Prevention somal location of chromosome 16, band B5. J. Immunol. 153:5038. of activated neutrophil adhesion to endothelium by soluble adhesion protein 18. Chomczynski, P., and N. Sacchi. 1987. Single-step method of RNA isolation by GMP-140. Science 249:414. acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162: 156. 40. Wong, C. S., J. R. Gamble, M. P. Skinner, C. M. Lucas, M. C. Berndt, and 19. Evans, M. J., S. L. Hartman, D. W. Wolff, S. A. Rollins, and S. P. Squinto. 1995. M. A. Vadas. 1991. Adhesion protein GMP-140 inhibits superoxide release by human neutrophils. Proc. Natl. Acad. Sci. USA 88:2397.

Rapid expression of an anti-human C5 chimeric Fab utilizing a vector that rep- http://www.jimmunol.org/ licates in COS and 293 cells. J. Immunol. Methods 184:123. 41. McDermott, M. F., I. Aksentijevich, J. Galon, E. M. McDermott, 20. Mueller, J. P., M. J. Evans, R. Cofiell, R. P. Rother, L. A. Matis, and E. A. Elliot. B. W. Ogunkolade, M. Centola, E. Mansfield, M. Gadina, L. Karenko, 1995. Porcine vascular cell adhesion molecule (VCAM) mediates endothelial cell T. Pettersson, et al. 1999. Germline mutations in the extracellular domains of the adhesion to human T cells. Transplantation 60:1299. 55 kDa TNF receptor, TNFR1, define a family of dominantly inherited autoin- 21. Murray, A. G., M. M. Khodadoust, J. S. Pober, and A. L. M. Bothwell. 1994. flammatory syndromes. Cell 97:133. Porcine aortic endothelial cells activate human T cells: direct presentation of 42. Van der Merwe, P. A., D. L. Bodian, S. Dainke, P. Linsley, and S. J. Davis. 1997. MHC antigens and costimulation by ligands for human CD2 and CD28. Immunity CD80 (B7-1) binds both CD28 and CTLA-4 with low affinity and very fast 1:57. kinetics. J. Exp. Med. 185:393. 22. Rollins, S. A., S. P. Kennedy, A. J. Chodera, E. A. Elliott, G. B. Zavoico, and 43. Gerstmayer, B., U. Pessara, and W. Wels. 1997. Construction and expression in L. A. Matis. 1994. Evidence that activation of human T cells by porcine endo- the yeast Pichia pastoris of functionally active soluble forms of the human co- thelium involves direct recognition of porcine SLA and costimulation by porcine stimulatory molecules B7-1 and B7-2 and the B7 counter-receptor CTLA-4.

ligands for LFA-1 and CD2. Transplantation 57:1709. FEBS Lett. 407:63. by guest on September 29, 2021 23. Turka, L. A., P. S. Linsley, H. Lin, W. Brady, J. M. Leiden, R.-Q. Wei, 44. Linsley, P. S., W. Brady, M. Urnes, L. S. Grosmaire, N. K. Damle, and M. L. Gibson, X.-G. Zheng, S. Myrdal, D. Gordon, et al. 1992. T-cell activation J. A. Ledbetter. 1991. CTLA-4 is a second receptor for the B cell activation by the CD28 ligand B7 is required for cardiac allograft rejection in vivo. Proc. antigen B7. J. Exp. Med. 174:561. Natl. Acad. Sci. USA 89:11102. 24. Lenschow, D. J., Y. Zeng, K. S. Hathcock, L. A. Zuckerman, G. J. Freeman, 45. Boise, L. H., A. J. Minn, P. J. Noel, C. H. June, M. A. Accavitti, T. Lindsten, and J. R. Thistlethwaite, G. S. Gray, R. J. Hodes, and J. A. Bluestone. 1995. Inhibition C. B. Thompson. 1995. CD28 costimulation can promote T cell survival by of transplant rejection following treatment with anti-B7-2 and anti-B7-1 antibod- enhancing the expression of Bcl-XL. Immunity 3:87. ies. Transplantation 60:1171. 46. Ikemizu, S., R. J. C. Gilbert, J. A. Fennelly, A. V. Collins, K. Harlos, E. Y. Jones, 25. Lenschow, D. J., Y. Zeng, J. R. Thistlethwaite, A. Montag, W. Brady, D. I. Stuart, and S. J. Davis. 2000. Structure and dimerization of a soluble form M. L. Gibson, P. S. Linsley, and J. A. Bluestone. 1992. Long-term survival of of B7-1. Immunity 12:51. xenogeneic pancreatic islet grafts induced by CTLA-4Ig. Science 257:789. 47. Yamashita, K., C. R. Parish, H. S. Warren, and L. C. Harrison. 1997. A multi- 26. Maher, S. E., K. Karmann, W. Min, C. C. W. Hughes, J. S. Pober, and meric form of soluble recombinant sheep LFA-3 (CD58) inhibits human T-cell A. L. M. Bothwell. 1996. Porcine endothelial CD86 is a major costimulator of proliferation. Immunology 92:39. xenogeneic human T cells. J. Immunol. 157:3838. 48. Inobe, M., N. Aoki, P. Linsley, J. A. Ledbetter, R. Abe, M. Murakami, and 27. Freeman, G. J., G. S. Gray, C. D. Gimmi, D. B. Lombard, L. J. Zhou, M. White, T. Uede. 1996. The role of the B7-1a molecule, an alternatively spliced form of J. D. Fingeroth, J. G. Gribben, and L. M. Nadler. 1991. Structure, expression, and murine B7-1 (CD80), on T cell activation. J. Immunol. 157:582. T-cell costimulatory activity of murine homologue of human B lymphocyte ac- 49. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Construction and analysis of tivation antigen B7. J. Exp. Med. 174:625. cDNA libraries. Molecular Cloning: A Laboratory Manual, 2nd Ed. Cold Spring 28. Judge, T. A., M. Liu, P. J. Christensen, J. J. Jak, and L. A. Turka. 1995. Cloning Harbor Laboratory Press, Cold Spring Harbor, p. 8.1. the rat homolog of the CD28/CTLA-4-ligand B7-1: structural and functional analysis. Int. Immunol. 7:171. 50. Matis, L. A., E. A. Elliott, M. J. Evans, J. P. Mueller, S. A. Rollins, R. P. Rother, 29. Guo, Y., Y. Wu, M. Zhao, X.-P. Kong, and Y. Liu. 1995. Mutational analysis and and M.-C. Shanafelt. 1995. The molecular basis of human anti-porcine cellular an alternatively spliced product of B7 defines its CD28/CTLA-4-binding site on interactions. Xeno 3:1. immunoglobulin C-like domain. J. Exp. Med. 181:1345. 51. Lenschow, D. J., G. H.-T. Su, L. A. Zuckerman, N. Nabavi, C. L. Jellis, 30. Ellis, J. H., M. N. Burden, D. V. Vinogradov, C. Linge, and J. S. Crowe. 1996. G. S. Gray, J. Miller, and J. A. Bluestone. 1993. Expression and functional sig- Interactions of CD80 and CD86 with CD28 and CTLA4. J. Immunol. 56:2700. nificance of an additional ligand for CTLA-4. Proc. Natl. Acad. Sci. USA 90: 31. Fargeas, C. A., A. Truneh, M. Reddy, M. Hurle, R. Sweet, and R.-P. Sekaly. 11054. 1995. Identification of residues in the V domain of CD80 (B7-1) implicated in 52. Chen, C., and N. Nabavi. 1994. In vitro induction of T cell anergy by blocking functional interactions with CD28 and CTLA4. J. Exp. Med. 182:667. B7 and early T cell costimulatory molecule ETC-1/B7-2. Immunity 1:147.