SFA-1/PETA-3 (CD151), a Member of the Transmembrane 4 ␣ ␤ Superfamily, Associates Preferentially with 5 1 Integrin and Regulates Adhesion of Human Virus Type 1-Infected T Cells to Fibronectin1

Hitoshi Hasegawa,2 Tetsuhiko Nomura, Kyoko Kishimoto, Kohsuke Yanagisawa, and Shigeru Fujita

In this study we have analyzed the adhesion molecules associated with and the biologic function of SFA-1/PETA-3 (CD151) in human T cell leukemia virus type 1 (HTLV-1)-infected T cells and in freshly isolated adult T cell leukemia (ATL) cells using an anti-CD151 ␣ ␤ ␣ mAb. The anti-CD151 mAb coprecipitated 5 1 integrin from HTLV-1-infected T cells. Conversely, an anti- 5 integrin mAb copre- cipitated CD151. The anti-CD151 mAb inhibited the adhesion of HTLV-1-infected T cells to fibronectin but did not have any effect on their adhesion to laminin, collagen type I, or collagen type IV. Moreover, antisense CD151 oligonucleotide-treated HTLV-1-infected T cells showed significant inhibition of adhesion to fibronectin. These findings showed that the CD151 molecule was associated with the ␣ ␤ ␣ ␤ 5 1 integrin molecule and that it enhanced 5 1 integrin-mediated adhesion to fibronectin. In addition, the expression levels of CD151, ␣ ␤ ␣ ␤ 4 1 integrin, and 5 1 integrin on ATL cells from lymph nodes of -type ATL patients were significantly higher than those on circulating ATL cells from leukemia-type ATL patients. This suggests that the increased expression of these integrins may contribute to lymphoma formation through the adhesion of ATL cells to the extracellular matrix and dendritic cells, rather than contributing to transmigration. The Journal of Immunology, 1998, 161: 3087–3095.

uman T cell leukemia virus type 1 (HTLV-1)3 is an ex- cells as well as probes obtained from normal CD4ϩ T cells and the ogenous human retrovirus closely linked with adult T MOLT-4 cell line (18). SFA-1 and PETA-3 were assigned CD151 at H cell leukemia (ATL). HTLV-1 has also been reported to the Sixth Human Leukocyte Differentiation Antigen Workshop. Hu- be associated with myelopathy, alveolitis, arthropathy, Sjo¨gren man SFA-1 (CD151) was found to be up-regulated upon transforma- syndrome, and uvenitis, which may result from immunologic al- tion by HTLV-1 and is trans-activated by Tax. The mRNA of the terations induced by HTLV-1 infection (1–5). The HTLV-1 ge- human SFA-1 (CD151) is comprised of approximately 1.6 kb nome encodes a 40-kDa , Tax, that functions as a transcrip- and encodes a protein of 253 amino acids. SFA-1 is a member of the tional trans-activator of viral and cellular . T cell transmembrane 4 superfamily (TM4SF). The human SFA-1/PETA-3 proliferation and immunologic alterations observed during (CD151) gene is a single gene located on 11p15.5 (19). HTLV-1 infection appear to be due to the effect of Tax on viral and Moreover, CD151 is conserved between human and mouse (20). cellular gene expression. Tax stimulates the expression of various PETA-3 was originally identified as a human surface glyco- cellular , including IL-2, IL-2R␣, granulocyte-macrophage protein, and it has been reported to regulate platelet aggregation and CSF, TNF-␤, TGF-␤,c-fos,c-jun, Krox-20, and Krox-24, and vi- mediator release (21–23). mentin and suppresses the expression of the genes, ␤-polymerase, The TM4SF is a family of membrane that are charac- p53, NF-1, and lck (5–17). However, the mechanism of HTLV-1- terized by the presence of four highly conserved transmembrane induced disease still remains to be elucidated. domains (reviewed by Refs. 24–26). This family currently has 19 Previously, to examine the changes in CD4ϩ T cells after HTLV-1 members that are found in species from Schistosoma to human: infection, we have cloned SFA-1 by differential hybridization of a CD9, CD37, CD53, CD63/ME491, CD81/TAPA-1, CD82/C33/ cDNA library, using probes obtained from an HTLV-1-infected T R2/KAI1, CO-029, A15, lbl, CD151/SFA-1/PETA-3, SAS, sm23, sj23, il-TMP, L6, peripherin, Rom-1, uroplakin Ia, and uroplakin

First Department of Internal Medicine, Ehime University School of Medicine, Shig- Ib. TM4SF members play roles in signal transduction pathways enobu, Ehime, Japan and regulate cell activation, development, proliferation, motility, Received for publication January 29, 1998. Accepted for publication May 11, 1998. and adhesion of a number of cell types. In addition, TM4SF mem- The costs of publication of this article were defrayed in part by the payment of page bers can form noncovalent associations with each other and with charges. This article must therefore be hereby marked advertisement in accordance other molecules, such as those involved in signal transduction and with 18 U.S.C. Section 1734 solely to indicate this fact. adhesion. In this report, we describe the adhesion molecules asso- 1 This work was supported in part by a grant-in-aid from the Osaka Cancer Research ciated with CD151 and its biologic function in HTLV-1-infected T Foundation and Scientific Research, and the Ministry of Education, Science, and Culture of Japan. cells and freshly isolated ATL cells. 2 Address correspondence and reprint requests to Dr. Hitoshi Hasegawa, First De- partment of Internal Medicine, Ehime University School of Medicine, Shigenobu, Materials and Methods Ehime 791-02, Japan. E-mail address: [email protected] Cells 3 Abbreviations used in this paper: HTLV-1, human T cell leukemia virus type 1; ATL, adult T cell leukemia; TM4SF, transmembrane 4 superfamily; PE, phyco- Human T cell lines, MOLT-4 and Jurkat; a human erythroleukemia cell erythrin; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate; line, K562; two human myelomonocytoid cell lines, HL60 and U937; and MFI, mean fluorescence intensity. NIH-3T3 cells were obtained from the American Type Culture Collection

Copyright © 1998 by The American Association of Immunologists 0022-1767/98/$02.00 3088 INTEGRIN ASSOCIATION WITH CD151 MOLECULE

(Rockville, MD). A human glioblastoma cell line, A172, and a human renal Preparation of mAbs carcinoma cell line, Caki-1, were obtained from the Japanese Cancer Re- search Resources Bank (Tokyo, Japan). These cell lines were maintained in mAbs were produced by hybridoma technology as described previously RPMI 1640 or DMEM supplemented with 10% heat-inactivated FCS (Life (34). In brief, hybridomas were produced through the fusion of Technologies, Gaithersburg, MD). Three HTLV-1-infected T cell lines, P3X63Ag8.653 cells with spleen cells from BALB/c mice immunized ␣ MT-2, MT-4, and HUT 102, were maintained in RPMI 1640 medium sup- against the NIH-3T3/pL2neoSR IIISFA-1 cells. The hybridoma culture ␣ plemented with 10% FCS. The HTLV-1-infected T cell line, SF-HT (27), supernatants that bound to NIH-3T3/pL2neoSR IIISFA-1 cells but not to ␣ was maintained in RPMI 1640 medium supplemented with 10% FCS and NIH-3T3/pL2neoSR III were screened. After screening, selected colonies 100 U/ml human rIL-2 (Takeda Chemical Industries, Osaka, Japan). Mono- were cloned twice or more by the limiting dilution method. After cloning, cytes were enriched as described previously (28). Normal T and B cells the Ab-producing hybridomas were inoculated into BALB/c mice treated were separated using the sheep erythrocyte-rosetting method (29). previously with pristane (Aldrich, Milwaukee, WI). Ascitic fluid contain- ing Ab was obtained after about 2 wk. The mAbs from the ascitic fluids were purified by affinity chromatography on a DEAE column. ATL patients Immunoprecipitation Mononuclear cells were isolated from peripheral samples from 14 patients with leukemia-type and from the lymph nodes of 10 patients with Cell surface proteins were labeled with 125I using carrier-free Na 125I and lymphoma-type by a Ficoll-Conray density gradient centrifugation. Con- lactoperoxidase (ICN, Costa Mesa, CA) following the instructions of the trol PBMCs were obtained from 10 normal healthy volunteers. The control manufacturer. Labeled cells were lysed in either Nonidet P-40 lysis buffer

lymph nodes samples were prepared from the reactive lymph nodes of (10 mM Tris-HCl (pH 8.0), 0.15 M NaCl, 3 mM MgCl2, 2 mM PMSF, eight HTLV-1-seronegative individuals who had undergone abdominal 0.5% Nonidet P-40), or CHAPS lysis buffer (10 mM Tris-HCl (pH 8.0), surgery. ATL was diagnosed according to the following clinical criteria: 0.15 M NaCl, 2 mM PMSF, 1 ␮g/ml antipain, 1 ␮g/ml pepstatin, 1 ␮g/ml serum Abs against HTLV-1-associated Ags; morphologic characteristics leupeptin, 1 ␮g/ml chymostatin, 10 mM iodoacetamine, and 1% CHAPS; showing highly convoluted nuclei; phenotypic analysis of ATL cells with all reagents were purchased from Sigma, St. Louis, MO). The lysates were anti-CD2, anti-CD4, and anti-CD25 mAbs; and monoclonal integration of centrifuged, and the supernatants were precleared overnight at 4°C with the HTLV-1 proviral genome in the cells. Using Shimoyama’s criteria (30), protein G-Sepharose (Pharmacia, Piscataway, NJ) precoated with normal 11 of 14 leukemia-type ATL patients were diagnosed with acute ATL, and rabbit serum. The lysates were then incubated with protein G-Sepharose 3 had chronic or smoldering ATL. Six of the leukemia-type ATL patients precoated with each mAb. After washing extensively with lysis buffer, had involvement of the skin, gut, or lymphoid organs such as the lymph samples were boiled for 2 min in Laemmli’s electrophoresis sample buffer nodes, liver, or spleen. The infiltration of ATL cells into the skin, gut, or and fractionated by SDS-PAGE. Radioactive bands were detected by lymph node was confirmed histochemically using biopsy samples or fol- autoradiography. lowing autopsy. Of the eight patients with noninfiltrating ATL, five had acute ATL, and three had chronic or smoldering ATL. In vitro adhesion assay ϩ Highly purified CD4 T cells were enriched by the negative immuno- Adhesion assays were performed as described by Sonnenberg et al. (35). selection from mononuclear cells by using a multiple mAb mixture and The 96-well plates were coated overnight at 4°C with fibronectin, collagen immunomagnetic beads, as described by Ishikawa et al. (31). Briefly, the type I, collagen type IV, or laminin at 100 ␮g/ml (all obtained from Sig- mononuclear cells prepared from peripheral blood and lymph nodes of ma). Plates were washed three times with PBS and blocked with 1% BSA ATL patients and control samples were incubated for 30 min at 4°C with in PBS for at least1hatroom temperature. The cell suspension was diluted a mixture of mAbs against CD8, CD11b, CD14, CD16, and CD20. After to contain 1 ϫ 104 cells/ml in serum-free DMEM and incubated with washing, Dynabeads (Japan Dynal, Tokyo, Japan) were added and incu- SFA1.4F11 mAb at the saturating concentration of 100 ␮g/ml for 30 min bated for1hat4°C, which were then removed with a Dynal magnetic at room temperature, and aliquots of 100 ␮l were added to each well. After particle concentrator. In all samples from ATL patients, the proportion of ϩ ϩ incubation for 1 h, the wells were washed, and the number of adherent cells ATL cells (CD4 and CD25 cells) in the negatively selected cells was was estimated using spectrophotometric absorbance following uptake of Ͼ90%. ATL cells from all samples expressed CD45RO. The ϩ ϩ 3-(4,5)-dimethylthiazol-2-yl-2,5 diphenyltetrazolium bromide (Sigma) by CD4 CD45RO T cells from control samples were purified by the same live cells (36). All experiments were conducted in triplicate at 37°C in a negative immunoselection technique using additional mAbs against CO incubator. HLA-DR and CD45RA. The percentage of CD4ϩCD45ROϩ T cells was 2 approximately 90%. Antisense CD151 oligonucleotides Oligonucleotides (32 mer) corresponding to the antisense sequence flank- ing the translation initiation region of the mRNA for human CD151 were Mouse mAbs to CD8 (OKT8), CD11b (OKM1), and HLA-DR (OKIa1) synthesized using phosphorothioate linkages because of their demonstrated Ј were obtained from the American Type Culture Collection (Rockville, resistance to nucleases. The sequences of the oligonucleotides were 5 - Ј MD). Mouse mAbs, anti-CD14 (322A-1), anti-CD20 (B1), anti-CD45RA GGCAAACGGTGCCACATGTTGTCTTCTTCTCG-3 (antisense) and Ј Ј (2H4), FITC-conjugated anti-CD4 (OKT4), phycoerythrin (PE)-conjugated 5 -CGGATAGGCTCCGAGAAGATCTGTACATGTGG-3 (nonsense). ␮ anti-CD25 (B1.49.9), and PE-anti-CD45RO (UCHL1) were purchased After the cells had been treated for 24 h with 5 M nonsense or antisense ␣ oligonucleotide, in vitro adhesion assays were performed. from Coulter (Hialeah, FL). Mouse mAbs to CD16 (3G8), integrin 2 ␣ ␣ (PIE6), 3 (PIB5), and v (VNR147) were purchased from Chemicon In- Flow cytometric analysis and immunoblotting ␣ ␣ ␣ ternational. Mouse mAbs to integrin 4 (HP2/1), 5 (SAM1), 6 (GoH3), ␤ and 1 (K20) were purchased from Immunotech (Marseille, France). Cells growing in monolayers were detached from the culture flasks by incubation at 37°C for 10 min with PBS containing 0.05% trypsin and 1 5 Cell transfection mM EDTA. The cells (2 ϫ 10 ) were washed once with 3% FCS-PBS and incubated with mAb, SFA1.4F11, or IgG1 mouse mAb (negative control) A recombinant plasmid, pL2neoSR␣IIISFA-1, was constructed by insert- at the saturating concentration (10 ␮g/ml) on for 30 min. After washing ing the 2.5-kb SalI fragment of pCDSR␣SFA-1 into the SalI site of the with 3% FCS-PBS, the cells were stained with FITC-conjugated goat anti- pL2neoSR␣III vector (32). The expression of the SFA-1 cDNA is regu- mouse IgG (Cappel, Westchester, PA) on ice for 30 min. After washing, lated by the SR␣ promoter. No transcription of the SFA-1 (CD151) gene the cells were resuspended in 3% FCS-PBS, and the fluorescence intensity was observed in NIH-3T3 cells (20). To transfect NIH-3T3 cells, cells at was analyzed on using a profile flow cytometer (EPICS, Coulter). about 50% confluence in a 10-cm dish were rinsed twice with Opti-MEM Flow cytometric analysis of freshly isolated ATL cells and control (Life Technologies) and were transfected with 10 ␮gof CD4ϩCD45ROϩ T cells was conducted as described by Ishikawa et al. pL2neoSR␣IIISFA-1 using lipofectin (Life Technologies) and following (31). Briefly, cells (2 ϫ 105) were incubated for 15 min at 4°C with human the instructions of the manufacturer. After incubation at 37°C for 6 h, the serum globulin (10 ␮g/ml; Green Cross, Osaka, Japan) to block Fc binding transfection mixture was removed. The cells were allowed to grow in fresh sites and then further incubated with FITC-anti-CD4 mAb for 30 min at RPMI 1640 medium supplemented with 10% FCS for 24 h and then were 4°C. After washing, the cells were further incubated with PE-conjugated ␮ ␣ ␣ added to G418 (Life Technologies) at the concentration of 400 g/ml. SFA1.4F11, PE-anti- 4 integrin mAb (HP2/1), PE-anti- 5 integrin mAb After 2 wk, several stable transformants were isolated and examined for the (SAM1), or PE-conjugated nonbinding control IgG1 mouse mAb at a sat- expression of SFA-1 by Northern blot analysis as described previously urating concentration (10 ␮g/ml) for 30 min at 4°C. After washing, the (33). Of these transformants, the clone with highest expression of SFA-1 is two-color immunostaining of cells was analyzed. The mean fluorescence ␣ ␣ ␣ designated NIH-3T3/pL2neoSR IIISFA-1 cells. intensity (MFI) values for staining of CD151, 4 integrin, and 5 integrin The Journal of Immunology 3089

FIGURE 1. Reactivity of SFA1.4F11 mAb to CD151-transfected NIH-3T3 cells. A, Flow cytometric analysis of the expression of SFA1.4F11 mAb on CD151-transfected NIH-3T3 cells. CD151-transfected (NIH3T3/pL2neoSR␣IIISFA-1; solid histogram) or control (NIH-3T3/pL2neoSR␣III; clear histo- gram) NIH-3T3 cells were incubated with SFA1.4F11 mAb and stained with FITC-conjugated goat anti-mouse IgG. The dotted line shows unstained NIH-3T3/pL2neoSR␣IIISFA-1 cells. Immunoprecipitation (B) and immunoblotting (C) of CD151-transfected NIH-3T3 cells using SFA1.4F11 mAb. NIH-3T3/pL2neoSR␣III (negative control; lane 1) and NIH-3T3/pL2neoSR␣IIISFA-1 (lane 2) cells were surface labeled with 125I, lysed in 0.5% Nonidet P-40, and then immunoprecipitated with SFA1.4F11. The precipitate was fractionated by SDS-PAGE and analyzed by radiography and immunoblotting using mouse serum immunized against the purified histidine-tagged CD151 protein.

ϩ were corrected by subtracting MFI obtained with the nonbinding control was expressed at low levels on CD4 T cells, T , B IgG1 mouse mAb. lymphocytes, granulocytes, and four other lymphoid cell lines, A recombinant plasmid, pQE30SFA-1, was constructed by inserting the Jurkat, MOLT-4, SF-EB, and HH-EB, whereas monocytes, U937 0.8-kb SacI fragment of pCDSR␣SFA-1 into the SacI site of the pQE30 vector. Histidine-tagged SFA-1 protein was purified using QIAexpress sys- cells, and all human nonhemopoietic cell lines used in our previous tem (Qiagen, Chatsworth, CA). Immunoblotting was performed by using study (18) showed significant expression of CD151. These results mouse serum immunized against the purified histidine-tagged SFA-1 pro- are in good agreement with those by obtained Northern blot tein as described previously (34). analysis (18).

Results Identification of CD151-associated proteins Isolation of mAb against CD151 To examine whether CD151 is associated on the cell surface with For use as a immunogen, we transfected the recombinant plasmid other molecules, 125I surface-labeled HUT 102 cells were ex- pL2neoSR␣IIISFA-1 into NIH-3T3 cells. Nontransfected NIH- tracted with the mild detergent CHAPS, and CD151 immunopre- 3T3 cells did not express any detectable levels of mRNA for cipitates were analyzed by SDS-PAGE. As shown in Figure 3A, CD151. After 2 wk, several stable transformants were isolated. anti-CD151 mAb SFA1.4F11 immunoprecipitated the CD151 Ag Among these transformants, one stable neomycin-resistant trans- itself (29 kDa). Additional protein bands of 160 and 130 kDa co- formant, designated NIH-3T3/pL2neoSR␣IIISFA-1, which has precipitated with CD151 molecule. These additional proteins were highest expression of CD151 mRNA, was used as an immunogen. also obtained from HUT 102 lysates prepared using the milder After injecting the NIH-3T3/pL2neoSR␣IIISFA-1 cells i.p. twice, detergents 1% Tween-20 and 1% Brij 58, but not from 0.5% Non- hybridomas were produced through the fusion of P3X63Ag8.653 idet P-40 lysate (data not shown). In addition, pretreatment of cells cells with spleen cells from immunized BALB/c mice. The anti- with EDTA did not alter the pattern of precipitation of these ad- CD151 mAb, designated SFA1.4F11, which bound to NIH-3T3/ ditional proteins (data not shown). These higher molecular mass pL2neoSR␣IIISFA-1 cells, but not to NIH-3T3/pL2neoSR␣III proteins (160 and 130 kDa) that coprecipitated from HUT 102 cells, was obtained (Fig. 1A). SFA1.4F11 was an IgG1 Ig with ␬ cells closely resembled the very late Ag integrins. Next, we repre- chain. We confirmed that this Ab reacted to CD151 by im- cipitated the anti-CD151 immunoprecipitate using anti-integrin ␣ ␣ ␣ ␣ ␣ ␣ ␤ munoprecipitation and immunoblotting using histidine-tagged ( 2, 3, 4, 5, 6, v, and 1) mAbs. As shown in Figure 3A, 160- ␣ ␤ SFA-1 protein (Fig. 1, B and C). and 130-kDa bands were reprecipitated by the anti- 5 and anti- 1 Abs but not by the other anti-integrin Abs. Conversely, the CD151 Expression of CD151 Ag on various cells ␣ molecule was also reprecipitated from an anti- 5 immunoprecipi- The surface expression of the CD151 Ag by various hemopoietic tate prepared from HUT 102 cells (Fig. 3B). However, CD151 was ␣ and nonhemopoietic cell lines was examined by flow cytometric not reprecipitated from an anti- 4 immunoprecipitate using a com- analysis using SFA1.4F11 (Fig. 2). Strong expression of CD151 bination of immunoprecipitation and immunoblotting (data not ϩ ␣ ␤ Ag was demonstrated on PHA-stimulated CD4 T cells and two shown). As shown in Figure 3C, 5 1 integrin was observed in HTLV-I-transformed T cell lines, SF-HT and HUT 102. The other anti-CD151 immunoprecipitates prepared from the other three two HTLV-1-transformed T cell lines, MT-2 and MT-4, also HTLV-1-infected T cell lines, MT-2, MT-4, and SF-HT. These showed strong expression of CD151 (data not shown). CD151 Ag findings show that the CD151 molecule associated preferentially 3090 INTEGRIN ASSOCIATION WITH CD151 MOLECULE

FIGURE 2. Flow cytometric analysis of the expression of CD151 Ag on various hemopoietic and nonhemopoietic cell lines. Cells were incubated with anti-CD151 mAb SFA1.4F11 (solid lines) or IgG1 mouse mAb (dotted lines) and stained with FITC-conjugated goat anti-mouse IgG.

␣ ␤ with 5 1 integrin in HTLV-1-infected T cell lines. In addition, from the MT-2 and SF-HT lysates. These additional proteins were unknown additional protein bands coprecipitated with the CD151 either absent or faintly present in anti-CD151 immunoprecipitates molecule: a 90-kDa band from the MT-2 lysate and 45-kDa bands prepared from HUT 102 and MT-4 cells and were obtained from

␣ ␤ 125 FIGURE 3. Coprecipitation of CD151 Ag and 5 1 integrin in HTLV-1-infected T cells. A, HUT 102 cells were surface labeled with I, lysed with 1% CHAPS, and then immunoprecipitated with anti-CD151 mAb SFA1.4F11. The anti-CD151 precipitate was eluted using 0.5% Nonidet P-40 and ␣ ␣ ␣ ␣ ␣ ␣ ␤ 125 ␣ reprecipitated with anti-integrin ( 2, 3, 4, 5, 6, v, and 1) mAbs. B, I surface-labeled HUT 102 cells were immunoprecipitated with anti- 5 mAb. ␣ The anti- 5 precipitate was reprecipitated with anti-CD151 mAb. C, The lysates prepared from four HTLV-1-infected T cell lines, HUT 102, MT-2, MT-4, and SF-HT, were immunoprecipitated with anti-CD151 mAb. The Journal of Immunology 3091

FIGURE 4. Inhibition of HTLV-1-infected T cell adhesion to extracellular matrix proteins by anti-CD151 mAb. Four kinds of extracellular matrix proteins, fibronectin, collagen type I, collagen type IV and laminin, were coated on 96-well plates as described in Materials and Methods. The adhesion of HTLV-1-infected T cells to extracellular matrix proteins was measured in the presence of anti-CD151 mAb SFA1.4F11 at the saturating concentration of 100 ␮g/ml. The percentage of the control value was compared with adhesion in the absence of Ab. All experiments were conducted in triplicate. Statistical analysis was performed using Student’s t test: * indicates p Ͻ 0.01. lysates prepared using the milder detergents, 1% CHAPS, 1% (by 85–90% adhesion) than that caused by SFA1.4F11 (data not Tween-20, and 1% Brij 58, but not from the 0.5% Nonidet P-40 shown). lysate (data not shown). Decreased adhesion of antisense CD151 oligonucleotide-treated Effect of anti-CD151 mAb on adhesion of HTLV-1- infected T HTLV-1-infected T cells to fibronectin cells to extracellular matrix proteins To demonstrate the physiologic significance of the interaction be- ␣ ␤ To examine whether anti-CD151 mAb affects the adhesion of tween CD151 and 5 1 integrin, we decided to decrease the ex- HTLV-1-infected T cells to extracellular matrix proteins, we per- pression level of CD151 to determine whether this would affect formed in vitro adhesion assays using fibronectin, laminin, colla- integrin function. The down-regulation of expression of CD151 gen type I, or collagen type IV. As shown in Figure 4, the anti- was achieved using an antisense oligonucleotide designed to in- CD151 mAb, SFA1.4F11, inhibited adhesion to fibronectin by 40 hibit the initiation of CD151 protein synthesis. Antisense oligonu- to 57% for all four HTLV-1-infected T cell lines, but did not have cleotide-treated, nonsense oligonucleotide-treated, and control un- any effect on adhesion to laminin, collagen type I, or collagen type treated cells showed no difference in cell viability, indicating that IV. The other anti-CD151 mAb, SFA1.2B4, which reacts with an the oligonucleotides were not toxic to the cells (data not shown). epitope different from that recognized by SFA1.4F11, was tested, As shown in Figure 5A, addition of the antisense CD151 oligonu- and inhibition of adhesion to fibronectin by SFA1.2B4 was lower cleotide to HUT 102 cell cultures resulted in significant inhibition

FIGURE 5. A, Flow cytometric analysis of the CD151 expression on antisense CD151 oligonucleotide-treated HTLV-1-infected T cells. After HUT 102 cells were treated for 24 h with 5 ␮M nonsense (clear histogram) or antisense (solid histogram) oligonucleotides, cells were incubated with anti-CD151 mAb SFA1.4F11 and stained with FITC-conjugated goat anti-mouse IgG. The dotted line shows unstained HUT 102 cells. B, Decreased adhesion of the antisense CD151 oligonucleotide-treated HTLV-1-infected T cells to fibronectin. After HUT 102 cells were treated for 24 h with 5 ␮M nonsense or antisense oligonucleotides, in vitro adhesion assays were performed. All experiments were conducted in triplicate. Statistical analysis was performed using Student’s t test: * indicates p Ͻ 0.01. 3092 INTEGRIN ASSOCIATION WITH CD151 MOLECULE

(Ͼ60%) of CD151 expression. Treatment with nonsense oligonu- cleotide at the same concentration did not result in down-regula- tion. Moreover, addition of antisense or nonsense CD151 oligo- ␣ ␤ nucleotides did not have any effect on the expression level of 5 1 integrin on HUT 102 cells (data not shown). The effect of the antisense CD151 oligonucleotide on adhesion to fibronectin was examined. As shown in Figure 5B, antisense oligonucleotide-treated HUT 102 cells showed significant inhibi- tion to 75% of adhesion to fibronectin, whereas there was very little difference in adhesion between the nonsense oligonucleotide- treated and untreated control cells. The other three HTLV-1-in- fected T cell lines, MT-2, MT-4, and SF-HT, also inhibited the adhesion to fibronectin significantly after antisense-oligonucleo- tide treatment (data not shown). These findings indicate that the ␣ ␤ CD151 molecule was associated with the 5 1 integrin molecule ␣ ␤ and that it enhanced the 5 1 integrin-mediated adhesion to fibronectin.

␣ ␤ ␣ ␤ Differential expression of CD151, 4 1 integrin, and 5 1 integrin on ATL cells in patients of leukemia-type or lymphoma-type ATL ␣ ␤ ␣ ␤ We analyzed the expression of CD151, 4 1 integrin, and 5 1 integrin on circulating ATL cells isolated from the peripheral blood of 14 patients with leukemia-type ATL or on ATL cells obtained from the lymph nodes of 10 lymphoma-type ATL pa- tients. ATL cells were enriched by the negative immunoselection. Since ATL cells had the phenotype CD4ϩCD45ROϩ, the expres- ␣ ␤ ␣ ␤ sion of CD151, 4 1 integrin, and 5 1 integrin on ATL cells was compared with that on CD4ϩCD45ROϩ T cells purified from con- ␣ ␤ trol samples. The integrin 4 chain associates with either 1-or ␤ ␣ ␤ ␣ ␤ 7-chains to form a heterodimeric glycoproteins 4 1 and 4 7.In ␤ all samples, the proportion of 1 integrin-expressed cells in ATL ϩ ϩ ␣ and that of control CD4 CD45RO T cells that expressed 4 in- tegrin were Ͼ90% (data not shown). Therefore, the expression of ␣ ␤ ␣ ␤ CD151, 4 1 integrin, and 5 1 integrin on ATL cells was as- sessed by the MFI for staining of anti-CD151 mAb (SFA1.4F11), ␣ ␣ anti- 4 integrin mAb (HP2/1), and anti- 5 integrin mAb (SAM1). As shown in Figure 6, there were no significant differences in the expression of CD151, ␣ ␤ integrin, and ␣ ␤ integrin between 4 1 5 1 FIGURE 6. Differential expression of CD151, ␣ ␤ integrin, and ␣ ␤ circulating ATL cells from leukemia-type ATL patients and con- 4 1 5 1 ϩ ϩ integrin on ATL cells in leukemia-type or lymphoma-type ATL patients. ϭ Ϯ ϩ ϩ trol CD4 CD45RO T cells (MFI: leukemia (n 14), 3.69 Freshly isolated ATL cells and CD4 CD45RO T cells were purified in 1.70; control PBMC (n ϭ 10), 3.21 Ϯ 1.56; vs control lymph node the negative immunoselection using a multiple mAb mixture as described (n ϭ 8), 3.38 Ϯ 1.30; 14.11 Ϯ 7.40 vs 10.72 Ϯ 4.47 vs 12.11 Ϯ in Materials and Methods. Flow cytometric analysis of CD4ϩCD45ROϩ T 5.34; 11.09 Ϯ 5.90 vs 9.36 Ϯ 4.20 vs 11.01 Ϯ 5.27, respectively). cells and ATL cells obtained from the peripheral blood of 10 normal vol- ␣ ␤ unteers (A, closed circles), the lymph nodes of eight HTLV-1-seronegative On the other hand, the expression levels of CD151, 4 1 integrin, ␣ ␤ individuals (B, closed squares), the peripheral blood of 14 leukemia-type and 5 1 integrin on ATL cells from the lymph nodes of lympho- ma-type ATL patients were significantly higher than those on cir- ATL patients (C, open circles, patients 1–14), and the lymph nodes of 10 culating ATL cells from leukemia-type ATL patients (MFI: lym- lymphoma-type ATL patients (D, open circles, patients 15–24) was per- formed with FITC-conjugated anti-CD4, PE-anti-CD151, PE-anti-␣ , PE- phoma (n ϭ 10), 9.55 Ϯ 5.96; vs leukemia (n ϭ 14), 3.69 Ϯ 1.70 4 anti-␣ , or PE-nonbinding control IgG1 mouse mAbs. Cells were electron- Ͻ Ϯ Ϯ Ͻ Ϯ 5 ( p 0.05); 25.47 11.42 vs 14.11 7.40 ( p 0.05); 20.16 ically gated based upon CD4 expression. Each point shows the MFI for Ϯ Ͻ 11.62 vs 11.09 5.90 ( p 0.05), respectively). This suggests that CD151, ␣ ,or␣ staining, which was corrected by subtracting MFI ob- ␣ ␤ ␣ ␤ 4 5 increased expression of CD151, 4 1 integrin, and 5 1 integrin tained with the nonbinding control IgG1 mouse mAb. The bars indicate the may act to retain ATL cells in the lymph nodes in some lympho- mean Ϯ SD of each group; the statistical analysis was performed using ma-type ATL patients. Student’s t test.

Discussion ␣ ␣ was observed in anti- 5 immunoprecipitates but not in anti- 4 im- In this report, we demonstrate for the first time a specific associ- munoprecipitates; 3) the anti-CD151 mAb inhibited the adhesion ␣ ␤ ation between CD151 and 5 1 integrin in HTLV-1-infected T to fibronectin of HTLV-1-infected T cell lines, but had no effect on ␣ ␤ cells, and we show that the CD151 molecule enhanced 5 1 inte- their adhesion to laminin, collagen type I, or collagen type IV; and grin-mediated adhesion to fibronectin. This interaction is likely to 4) antisense CD151 oligonucleotide-treated HTLV-1-infected T ␣ ␤ be meaningful for the following reasons: 1) 5 1 integrin was cells showed significantly inhibited of adhesion to fibronectin. reprecipitated from anti-CD151 immunoprecipitates prepared from Moreover, we also found that the expression levels of CD151, ␣ ␤ ␣ ␤ HTLV-1-infected T cell lines; 2) conversely, the CD151 molecule 4 1 integrin, and 5 1 integrin were higher on ATL cells from The Journal of Immunology 3093 the lymph nodes of lymphoma-type ATL patients than those on lation of CD151 expression on HUT 102 cells caused by antisense circulating ATL cells from leukemia-type ATL patients, suggest- oligonucleotide resulted in significant inhibition of adhesion to fi- ing that the expression levels of these molecules may affect the bronectin. Some members of the TM4SF function as adaptor pro- retention of ATL cells in lymph nodes. teins that organize the relative positions of other cell surface mol- The TM4SF has 19 members, each containing four highly con- ecules and modulate their function. For example, CD81 forms a served transmembrane domains, a number of cysteine residues, complex consisting of CD19, CD21, and Leu 13 on B cells, and and a major extracellular region between the third and fourth trans- this complex lowers the threshold for signal transduction through membrane domains. Although the biologic functions of the mem- the Ag receptor complex (46, 47). This result indicates that bers of the TM4SF are poorly understood, several studies of their CD81 facilitates interactions with integrins, resulting in cell adhe- functions using specific mAbs have been undertaken, and the re- sion, and that CD81 stabilizes the CD21/CD19 interaction. CD151 ␣ ␤ sults obtained suggest a role for this superfamily in signal trans- may also affect the accumulation of 5 1 integrin at focal contacts ␣ ␤ duction pathways and the regulation of cell activation, develop- or the maintenance of the appropriate conformation of 5 1 inte- ment, proliferation, motility, and adhesion (reviewed in Refs. 24– grin for optimal adhesion to fibronectin. ␣ ␤ 26). Moreover, TM4SF members have been shown to form It has been reported that the expression of 4 1 integrin and ␣ ␤ ␣ ␤ noncovalent associations with each other, integrins, and coreceptor 5 1 integrin on HTLV-1-infected lymphocytes and of 5 1 in- molecules, such as those involved in adhesion and signal trans- tegrin on T lymphocytes of patients with HIV-1 infection is in- duction. CD81, CD9, CD53, CD63, and CD82 have all been found creased and that alteration of expression of these integrins might in association with certain integrins in various types of human contribute to the abnormal proliferation of T cells in ATL or to the ␤ ␣ ␤ cells. All these molecules associate with the 1 integrins 3 1, abnormal localization of activated or infected T cells (59, 60). ␣ ␤ ␣ ␤ 4 1, and 6 1 (37–45). These associations appear to be impor- Fibronectin is widely distributed in plasma and connective tissue tant for cell-cell adhesion and migration. CD81 forms part of a subendothelial matrexes and stroma and is produced by various signaling complex with CD21, CD19, and Leu 13 on B cells (46, cells, such as endothelial, mesenchymal, and epithelial cells. In 47). CD81 also associates with CD82 as well as with CD4 and lymph nodes, fibronectin is mainly produced by stromal cells and CD8 coreceptors on T cells (48). The interaction of CD81/CD82 is present in all areas except the mantle zone (61). Fibronectin is with CD4 is strongly inhibited by p56lck binding to CD4. In the especially abundant in the interfollicular compartments, which are present study, human SFA-1, which was up-regulated by HTLV-I rich in T cells. The follicular dendritic cells in the germinal center ␣ ␤ ␣ ␤ infection and trans-activated by Tax, was associated with 5 1 express VCAM-1. 4 1 integrin expressed at increased levels on B ␣ ␤ integrin in HTLV-1-infected T cells, and it enhanced the 5 1 lymphoma cells has been reported to play a role in the formation integrin-mediated adhesion to fibronectin. PETA-3, which was of follicular non-Hodgkin’s lymphoma by binding to VCAM-1 originally identified as a human platelet surface glycoprotein, was expressed on follicular dendritic cells (62, 63). On the other hand, detected using a mAb, 14A2.H1. Although PETA-3 was present in adhesion to high endothelium has been reported to be ␣ ␤ ␣ ␤ low abundance on the platelet surface, 14A2.H1 stimulated platelet mediated by L-, 4 1 integrin, and 5 1 integrin, resulting aggregation and mediator release. In addition, this Ab showed syn- in contributing to the transmigration of lymphocytes (64, 65). The ␣ ␤ ␣ ␤ ergy with subthreshold concentrations of other agonists, ADP, expression levels of CD151, 4 1 integrin, and 5 1 integrin on adrenaline, collagen, and serotonin, in mediating platelet activa- ATL cells from lymph nodes of lymphoma-type patients were tion. Sincock et al. (49) examined the distribution of PETA-3 higher than those on circulating ATL cells from leukemia-type ␣ ␤ ␤ (CD151), CD9, CD63, 5 1, and the integrin 1-chain in normal patients. In ATL, the increased expression of these integrins may human tissues by the indirect immunoperoxidase and alkaline contribute to the formation of a lymphoma through adhesion of phosphatase-anti-alkaline phosphatase techniques. CD151 showed extracellular matrix and dendritic cells, rather than contributing to a broad distribution and was expressed by endothelium, epithe- transmigration. The fibronectin-rich microenvironment in the lium, Schwann cells, dendritic cells, and skeletal, smooth, and car- lymph node can also serve as a site of ATL cell adhesion and diac muscle. They showed colocalization of CD151 with CD9, support for ATL cell proliferation. Adhesion studies using ATL ␣ ␤ ␤ CD63, 5 1, and 1 in particular tissues, demonstrating that cells and HUVECs revealed that E-selectin-mediated adhesion was CD151/integrin complexes may occur. CD151 formed a complex a major pathway for the adherence of ATL cells to HUVECs, ␣ ␤ ␣ ␤ with 5 1 integrin, but not with CD9 or CD63, on HTLV-1-in- whereas the 4 1 integrin/VCAM-1 pathway was partly involved fected T cells (H. Hasegawa, et al., unpublished observations). in adhesion in some cases (31). Three pairs of adhesion molecules, Their finding and our results suggest that CD151 associates L-selectin/peripheral lymph node addressin, cutaneous lympho- ␣ ␤ ␣ ␤ strongly with 5 1 integrin in particular cells and tissues. cyte-associated Ag/E-selectin, and 4 7 integrin/mucosal ␣ ␤ The 5 1 integrin has been identified as the classical receptor VCAM-1 have been reported to be associated with preferential for fibronectin. At least eight integrins have been reported to bind recirculation of ATL cells to peripheral lymph nodes, skin, and ␣ ␤ ␣ ␤ ␣ ␤ ␣ ␤ ␣ ␤ ␣ ␤ ␣ ␤ to fibronectin: 3 1, 4 1, 5 1, v 1, IIb 3, v 3, v 6, and gastrointestinal mucosa, respectively (66). The circulating ATL ␣ ␤ 4 7 (50). Of these integrins, lymphocytes interact with fibronec- cells from patients with lymph node, skin, and gut involvement ␣ ␤ ␣ ␤ tin via two main receptors, 5 1 and 4 1 integrins. The interac- had higher expression levels of L-selectin, cutaneous lymphocyte- ␣ ␤ ␣ ␤ tion of 5 1 integrin with fibronectin results in augmented cell associated Ag, and 4 7 integrin, respectively, compared with adhesion, migration, differentiation, and signal transduction and those from patients without involvement of these organs. These has been implicated in important events in early embryogenesis, findings together with our results suggest that these adhesion mol- ␣ ␤ ␣ ␤ ␣ ␤ such as gastrulation (51–55). The 5 1 integrin-mediated adhesion ecules, rather than 4 1 integrin and 5 1 integrin, play an im- to fibronectin causes localization of the receptor to focal contacts portant role in the transmigration and organ infiltration of ATL and results in stable cell-substratum adhesion (56). The capacity of cells. ␣ ␤ ␣ ␤ 5 1 integrin to bind fibronectin is regulated by divalent cations In this report, we showed that CD151 associated with 5 1 in- ␣ ␤ and GM3 ganglioside. The recognition of fibronectin by 5 1 in- tegrin and regulated adhesion of HTLV-1-infected T cells and tegrin is enhanced by Mn2ϩ and is inhibited by Ca2ϩ (57). GM3 ATL cells to fibronectin. Since CD151 is widely expressed on ganglioside, within an optimal concentration range, enhances the various cells and tissues, however, CD151 may have other un- ␣ ␤ ability of 5 1 integrin to adhere to fibronectin (58). Down-regu- known biologic functions in other cells and tissues. It is necessary 3094 INTEGRIN ASSOCIATION WITH CD151 MOLECULE to study the biologic functions of CD151 further in other cells and leukemia virus type 1, is highly expressed in mature lymphocytes. Biochem. tissues. Biophys. Res. Commun. 222:164. 28. Hasegawa, H., Y. Satake, and Y. Kobayashi. 1990. Effect of cytokines on Jap- anese encephalitis virus production by human monocytes. Microbiol. Immunol. References 34:459. 29. Hasegawa, H., Y. Utsunomiya, M. Yasukawa, K. Yanagisawa, and S. Fujita. 1. Osame, M., K. Usuku, S. Izumo, N. Ijichi, H. Amatani, A. Igata, M. Matsumoto, 1994. Induction of G protein-coupled peptide receptor EBI 1 by human herpes- and M. Tara. 1986. HTLV-1 associated myelopathy, a new clinical entity. Lancet ϩ virus 6 and 7 infection in CD4 T cells. J. Virol. 68:5326. i:1031. 2. Vernant, J. C., G. Buisson, J. Magdeleine, J. De Thore, A. Jouannelle, 30. Shimoyama, M., and members of The Lymphoma Study Group (1984–87). 1991. C. Neisson-Vernant, and N. Monplaisir. 1988. T-lymphocyte alveolitis, tropical Diagnostic criteria and classification of clinical subtypes of adult T-cell leukemia- spastic paresis, and Sjogren syndrome. Lancet i:177. lymphoma. Br. J. Haematol. 79:428. 3. Nishioka, N., I. Maruyama, K. Sato, I. Kitajima, Y. Nakajima, and M. Osame. 31. Ishikawa, T., A. Imura, K. Tanaka, H. Shirane, M. Okuma, and T. Uchiyama. 1989. Chronic inflammatory arthropathy associated with HTLV-1. Lancet 1:441. 1993. E-selectin and vascular cell adhesion molecule-1 mediate adult T-cell leu- 4. Mochizuki, M., K. Yamaguchi, K. Takatsuki, T. Watanabe, S. Mori, and kemia cell adhesion to endothelial cells. Blood 82:1590. K. Tajima. 1992. HTLV-1 and uveitis. Lancet 339:1110. 32. Takebe, Y., M. Seiki, J. Fujisawa, P. Hoy, K. Yokota, K. Arai, M. Yoshida, and 5. Fujii, M., P. Sassone-Corsi, and I. M. Verma. 1988. c-fos promoter transactiva- N. Arai. 1988. The SR␣ promoter: an efficient and versatile mammalian cDNA tion by the tax1 protein of human T-cell leukemia virus type 1. Proc. Natl. Acad. expression system composed of the simian virus 40 early promoter and the R-U5 Sci. USA 85:8526. segment of human T-cell leukemia virus type 1 long terminal repeat. Mol. Cell. 6. Inoue, J.-K., M. Seiki, T. Taniguchi, S. Tsuru, and M. Yoshida. 1986. Induction Biol. 8:466. of interleukin-2 receptor gene expression by p40 encoded by human T-cell leu- 33. Hasegawa, H., M. Yoshida, T. Shiosaka, S. Fujita, and Y. Kobayashi. 1992. kemia virus type 1. EMBO J. 5:2883. Mutations in the envelope protein of Japanese encephalitis virus affect entry into 7. Sjekevitz, M., M. B. Feinberg, N. Holbrook, F. Wong-Staal, and W. C. Greene. cultured cells and virulence in mice. Virology 191:158. 1987. Activation of interleukin-2 and interleukin-2 receptor (tac) promoter ex- 34. Hasegawa, H., M. Yoshida, Y. Kobayashi, and S. Fujita. 1995. Antigenic analysis pression by the trans-activator (tat) gene product of human T cell leukemia virus of Japanese encephalitis viruses in Asia by using monoclonal antibodies. Vaccine type 1. Proc. Natl. Acad. Sci. USA 84:5389. 13:1713. 8. Miyatake, S., M. Seiki, M. Yoshida, and K.-I. Arai. 1988. T-cell activation sig- 35. Sonnenberg, A., C. J. T. Linders, P. W. Modderman, C. H. Damsky, nals and human T-cell leukemia virus type 1-encoded p40 protein activate the M. Aumailley, and R. Timpl. 1990. Integrin recognition of different cell binding mouse granulocyte-macrophage colony-stimulating factor gene through a com- fragments of laminin (P1, E3, E8) and evidence that ␣6␤1 but not ␣6␤4 functions mon DNA element. Mol. Cell. Biol. 8:5581. as a major laminin receptor for fragment E8. J. Cell. Biol. 110:2145. 9. Tschachler, E., M. Robert-Guroff, R. C. Gallo, and M. S. Reitz. 1989. Human 36. Camichael, J., W. G. Degraff, A. F. Gazdar, J. D. Minna, and J. B. Mitchell. 1987. T-lymphotropic-virus-1-infected T cells constitutively express lymphotoxin in Evaluation of a tetrazolium-based semi-automated colorimetric assay: assessment vitro. Blood 73:194. of radiosensitivity. Cancer Res. 47:943. 10. Kim, S.-J., J. H. Kehrl, J. Burton, C. L.Tendler, K.-T. Jeang, D. Danielpour, 37. Berditchevski, F., M. H. Zutter, and M. E. Hemler. 1996. Characterization of C. Thevenin, K. Y. Kim, M. B. Sporn, and A. B. Roberts. 1990. Transactivation novel complexes on the cell surface between integrins and proteins with 4 trans- ␤ ␤ of the transforming growth factor 1 (TGF- 1) gene by human T lymphotropic membrane domains (TM4 proteins). Mol. Biol. Cell 7:193. ␤ virus type 1 Tax: a potential mechanism for the increased production of TGF- 1 38. Radford, K. J., R. F. Thorne, and P. Hersey. 1996. CD63 associates with trans- in adult T cell leukemia. J. Exp. Med. 172:121. membrane 4 superfamily members, CD9 and CD81, and with ␤1 integrins in 11. Alexandre, C., and B. Verrier. 1991. Four regulatory elements in the human c-fos human melanoma. Biochem. Biophys. Res. Commun. 222:13. promoter mediate transactivation by HTLV-1 Tax protein. Oncogene 6:543. 39. Berditchevski, F., G. Bazzoni, and M. E. Hemler. 1995. Specific association of 12. Alexandre, C., P. Charnay, and B. Verrier. 1991. Transactivation of Krox-20 and CD63 with the VLA-3 and VLA-6 integrins. J. Biol. Chem. 270:17784. Krox-24 promoters by the HTLV-1 Tax protein through common regulatory el- ements. Oncogene 6:1852. 40. Rubinstein, E., F. Le Naour, M. Billard, M. Prenant, and C. Boucheix. 1994. CD9 13. Liliebaum, A., M. Duc Dodon, C. Alexandre, L. Gazzolo, and D. Paulin. 1990. antigen is an accessory subunit of the VLA integrin complexes. Eur. J. Immunol. Effect of human T-cell leukemia virus type 1 Tax protein on activation of the 24:3005. human vimentin gene. J. Virol. 64:256. 41. Nakamura, K., R. Iwamoto, and E. Mekata. 1995. Membrane-anchored heparin- 14. Jeang, K.-T., S. G. Widen, O. J. Semmes IV, and S. H. Wilson. 1990. HTLV-1 binding EGF-like growth factor (HB-ECF) and diphtheria toxin receptor-associ- ␣ ␤ transactivator protein, Tax, is a transrepressor of the human ␤ polymerase gene. ated protein (DRAP27)/CD9 form a complex with integrin 3 1 at cell-cell Science 247:1082. contact sites. J. Cell Biol. 129:1691. 15. Uittenbogaard, M. N., H. A. Giebler, D. Reisman, and J. K. Nyborg. 1995. Tran- 42. Rubinstein, E., F. Le Naour, C. Lagaudriere-Gesbert, M. Billard, H. Conjeaud, scriptional repression of p53 by human T-cell leukemia virus type 1 Tax protein. and C. Boucheix. 1996. CD9, CD63, CD81, and CD82 are components of a J. Biol. Chem. 270:28503. surface tetraspan network connected to HLA-DR and VLA integrins. Eur. J. Im- 16. Feigenbaum, L., K. Fujita, F. S. Collins, and G. Jay. 1996. Repression of the NF1 munol. 26:2657. gene by Tax may explain the development of neurofibromas in human T-lym- 43. Behr, S., and F. Schriever. 1995. Engaging CD19 or target of an antiproliferative photropic virus type 1 transgenic mice. J. Virol. 70:3280. 1 on human B lymphocytes induces binding of B cells to the interfol- ␣ ␤ 17. Lemasson, I., V. Robert-Hebmann, S. Hamaia, M. D. Dodon, L. Gazzolo, and licular stroma of human tonsils via integrin 4/ 1 and fibronectin. J. Exp. Med. C. Devaux. 1997. Transrepression of lck gene expression of human T-cell leu- 182:1191. kemia virus type 1-encoded p40tax. J. Virol. 71:1975. 44. Hadjiargyrou, M., Z. Kaprielian, N. Kato, and P. H. Patterson. 1996. Association 18. Hasegawa, H., Y. Utsunomiya, K. Kishimoto, K. Yanagisawa, and S. Fujita. of the tetraspan protein CD9 with integrins on the surface of S-16 Schwann cells. 1996. SFA-1, a novel cellular gene induced by human T-cell leukemia virus type J. Neurochem. 67:2505. 1, is a member of the transmembrane 4 superfamily. J. Virol. 70:3258. 45. Mannion, B. A., F. Berditchevski, S.-K. Kraeft, L. B. Chen, and M. E. Hemler. 19. Hasegawa, H., K. Kishimoto, K. Yanagisawa, H. Terasaki, M. Shimadzu, and 1996. Transmembrane-4 superfamily proteins CD81 (TAPA-1), CD82, CD63, ␣ ␤ S. Fujita. 1997. Assignment of SFA-1 (PETA-3), a member of the transmembrane and CD53 specifically associate with integrin 4 1 (CD49d/CD29). J. Immunol. 4 superfamily, to human chromosome 11p15.5 by fluorescence in situ hybridiza- 157:2039. tion. Genomics 40:193. 46. Matsumoto, A. K., D. R. Martin, R. H. Carter, L. B. Klickstein, J. M. Ahearn, and 20. Hasegawa, H., H. Watanabe, T. Nomura, Y. Utsunomiya, K. Yanagisawa, and D. T. Fearon. 1993. Functional dissection of the CD21/CD19/TAPA-1/Leu-13 S. Fujita. 1997. Molecular cloning and expression of mouse homologue of SFA- complex of B lymphocytes. J. Exp. Med. 178:1407. 1/PETA-3 (CD151), a member of the transmembrane 4 superfamily. Biochim. 47. Bradbury, L. E., G. S. Kansas, S. Levy, R. L. Evans, and T. F. Tedder. 1992. The Biophys. Acta 1353:125. CD19/CD21 signal transducing complex of human B lymphocytes includes the 21. Fitter, S., T. J. Tetaz, M. C. Berndt, and L. K. Ashman. 1995. Molecular cloning target of antiproliferative antibody-1 and Leu-13 molecules. J. Immunol. 149: of cDNA encoding a novel platelet-endothelial cell tetra-span antigen, PETA-3. 2841. Blood 86:1348. 48. Imai, T., M. Kakizaki, M. Nishimura, and O. Yoshie. 1995. Molecular analysis 22. Ashman, L. K., G. W. Aylett, P. A. Mehrabani, L. J. Bendall, S. Niutta, of the association of CD4 with two members of the transmembrane 4 superfam- A. C. Cambareri, S. R. Cole, and M. C. Berndt. 1991. The murine monoclonal ily, CD81 and CD82. J. Immunol. 155:1229. antibody, 14A2.H1, identifies a novel platelet surface antigen. Br. J. Haematol. 79:263. 49. Sincock, P. M., G. Mayrhofer, and L. K. Ashman. 1997. Localization of the transmembrane 4 superfamily (TM4SF) member PETA-3 (CD151) in normal 23. Roberts, J. J., S. E. Rodgers, J. Drury, L. K. Ashman, and J. V. Lloyd. 1995. ␣ ␤ Platelet activation induced by a murine directed against a human tissues: comparison with CD9, CD63, and 5 1 integrin. J. Histochem. novel tetra-span antigen. Br. J. Haematol. 89:853. Cytochem. 45:515. 24. Maecker, H. T., S. C. Todd, and S. Levy. 1997. The superfamily: 50. Akiyama, S. K., K. Olden, and K. M. Yamada. 1995. Fibronectin and integrins molecular facilitators. FASEB J. 11:428. in invasion and metastasis. Cancer Metastasis Rev. 14:173. 25. Hemler, M. E., A. B. Mannion, and F. Berditchevski. 1996. Association of 51. Plantefaber, L. C., and R. O. Hynes. 1989. Changes in integrin receptors on TM4SF proteins with integrins: relevance to cancer. Biochim. Biophys. Acta oncogenically transformed cells. Cell 56:281. 1287:67. 52. Schreiner, C. L., J. S. Bauer, Y. N. Danilov, S. Hussein, M. M. Sczekan, and 26. Wright, M. D., and M. G. Tomlinson. 1994. The ins and outs of the transmem- R. L. Juliano. 1989. Isolation and characterization of Chinese hamster ovary cell brane 4 superfamily. Immunol. Today 15:588. variants deficient in the expression of fibronectin receptor. J. Cell Biol. 109:3157. ␣ ␤ 27. Hasegawa, H., Y. Utsunomiya, K. Kishimoto, Y. Tange, M. Yasukawa, and 53. Giancotti, F., and E. Ruoslahti. 1990. Elevated levels of 5/ 1 fibronectin recep- S. Fujita. 1996. SFA-2, a novel bZIP transcription factor induced by human T-cell tor suppress the transformed phenotype in CHO cells. Cell 60:849. The Journal of Immunology 3095

54. Zutter, M. M., G. Mazoujian, and S. A. Santoro. 1990. Decreased expression of 61. Castanos-Velez, E., P. Biberfeld, and M. Patarroyo. 1995. Extracellular matrix integrin adhesive protein receptors in adenocarcinoma of the breast. proteins and integrin receptors in reactive and non-reactive lymph nodes. Immu- Am. J. Pathol. 137:863. nology 86:270. 55. Gui, G. P. H., J. R. Puddefoot, G. P. Vinson, C. A. Wells, and R. Carpenter. 1997. 62. Koopman, G., H. K. Parmentier, H. J. Schuurman, W. Newman, C. J. Meijer, and Altered cell-matrix contact: a prerequisite for breast cancer metastasis? Br. J. S. T. Pals. 1991. Adhesion of human B cells to follicular dendritic cells involves Cancer 75:623. both the lymphocyte function-associated antigen 1/intercellular adhesion mole- 56. Tawil, N., P. Wilson, and S. Carbonetto. 1993. Integrins in point contacts mediate cule 1 and very late antigen 4/vascular cell adhesion molecule 1 pathway. J. Exp. cell spreading: factors that regulate integrin accumulation in point contacts vs. Med. 173:1297. focal contacts. J. Cell Biol. 120:261. 63. Freedman, A. S., J. M. Munro, C. Morimoto, B. W. McIntyre, K. Rhynhart, 57. Gailit, J., and E. Rouslahti. 1988. Regulation of the fibronectin receptor affinity N. Lee, and L. M. Nadler. 1992. Follicular non-Hodgkin’s lymphoma cell adhe- by divalent cations. J. Biol. Chem. 263:12927. sion to normal germinal centers and neoplastic follicles involves very late anti- 58. Zheng, M., T. Tsuruoka, T. Tsuji, and S. Hakomori. 1992. Regulatory role of gen-4 and vascular cell adhesion molecule-1. Blood 79:206. GM3 ganglioside in integrin function, as evidenced by its effect on function of 64. Lasky, L. A., M. S. Singer, D. Dowbenko, Y. Imai, W. J. Henzel, C. Grimley, ␣5␤1-liposomes: a preliminary note. Biochem. Biophys. Res. Commun. C. Fennie, N. Gillett, S. R. Watson, and S. D. Rosen. 1992. An endothelial 186:1397. for L-selectin is a novel mucin-like molecule. Cell 69:927. 59. Dhawan, S., B. S. Weeks, F. Abbasi, H. R. Gralnick, A. L. Natkins, 65. Szekanecz, Z., M. J. Humphries, and A. Ager. 1992. Lymphocyte adhesion to M. E. Klotmon, K. M. Yamada, and P. E. Klotman. 1993. Increased expression high endothelium is mediated by two ␤1 integrin receptors for fibronectin, ␣4␤1 of ␣4␤1 and ␣5␤1 integrins on HTLV-I-infected lymphocytes. Virology 197:778. and ␣5␤1. J. Cell Sci. 101:885. 60. Torre, D., G. Ferrario, M. Issi, A. Pugliese, and F. Speranza. 1996. Expression of 66. Tanaka, Y., A. Wake, K. J. Horgan, S. Murakami, M. Aso, K. Saito, S. Oda, the ␣5␤1 fibronectin receptor on T lymphocytes of patients with HIV-1 infection. I. Morimoto, H. Uno, H. Kikuchi, Y. Izumi, and S. Eto. 1997. Distinct phenotype J. Clin. Pathol. 49:733. of leukemic T cells with various tissue tropisms. J. Immunol. 158:3822.