Receptors in Macrophages Γ Functional Association of CD9 With

Functional Association of CD9 with the Fcγ Receptors in Macrophages Keisuke Kaji, Sunao Takeshita, Kensuke Miyake, Toshiyuki Takai and Akira Kudo This information is current as of September 29, 2021. J Immunol 2001; 166:3256-3265; ; doi: 10.4049/jimmunol.166.5.3256 http://www.jimmunol.org/content/166/5/3256 Downloaded from

References This article cites 40 articles, 22 of which you can access for free at: http://www.jimmunol.org/content/166/5/3256.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

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

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 29, 2021

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 © 2001 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Functional Association of CD9 with the Fc␥ Receptors in Macrophages1

Keisuke Kaji,* Sunao Takeshita,* Kensuke Miyake,† Toshiyuki Takai,‡ and Akira Kudo2*

CD9, a member of the tetraspan family of proteins, is highly expressed on macrophages. Although a clear function for the molecule has yet to be described, we have found that the anti-CD9 mAb activates mouse macrophages. The rat anti-CD9 mAb, KMC8.8, ؅ but not the F(ab )2, induced tyrosine phosphorylation of proteins including syk and cbl and induced cell aggregation in the mouse macrophage cell line, J774, suggesting that co-cross-linking of CD9 and Fc␥R was required for the signal. Co-cross-linking of ,CD9-Fc␥R with KMC8.8 on macrophages from three different FcR-deficient mice, FcR ␥-chain؊/؊,Fc␥RIIB؊/؊, and Fc␥RIII؊/؊ revealed that Fc␥RIII is specific and crucial for syk phosphorylation. Although both KMC8.8 and the anti-Fc␥RIIB/III mAb, 2.4G2, evoked similar phosphorylation patterns, only KMC8.8 induced cell aggregation. Additionally, KMC8.8 treatment led to ␣ reduce levels of TNF- production and p42/44 extracellular signal-related kinase phosphorylation relative to 2.4G2 stimulation. Downloaded from Immunofluorescence staining showed that co-cross-linking of CD9-Fc␥R with KMC8.8 induced filopodium extension before cell aggregation, which was followed by simultaneous colocalization of CD9, Fc␥RIIB/III, Mac-1, ICAM-1, and F-actin at the cell-cell adhesion site. Moreover, KMC8.8 treatment of Fc␥R-deficient macrophages revealed that the colocalization of CD9, Fc␥RIII, Mac-1, and F-actin requires co-cross-linking of CD9-Fc␥RIII, whereas co-cross-linking of CD9-Fc␥RIIB induced the colocalization of only CD9 and Fc␥RIIB. Our results demonstrate that co-cross-linking of CD9 and Fc␥Rs activates macrophages; therefore,

CD9 may collaborate with FcRs functioning in infection and inﬂammation on macrophages. The Journal of Immunology, 2001, http://www.jimmunol.org/ 166: 3256–3265.

he most important accessory cells in immune responses Fc␥RI and Fc␥RIII receptors, nonreceptor tyrosine kinases includ- are phagocytes of the monocytic and myelocytic lineage, ing members of the Src and Syk/ZAP70 families are activated, T represented by the macrophage. In phagocytosis by mac- resulting in the phosphorylation of the ␥-chain ITAM. This is fol- rophages, Fc␥Rs (Fc receptor for IgG) are one of the most impor- lowed by tyrosine phosphorylation of downstream effectors, such tant receptor families. Fc␥Rs recognize IgGs, which are produced as phospholipase C␥, phosphatidylinositol 3-kinase, mitogen-acti- in response to pathogen invasions, and mediate phagocytosis of the vated protein kinase (MAPK), and cbl (2, 5, 6). Although both IgG-opsonized pathogens. During this process, cross-linking of Fc␥RI and Fc␥RIII function as phagocytotic receptors and trans- by guest on September 29, 2021 Fc␥Rs by the immune complex induces a wide variety of immune duce similar signals, the binding activity to IgGs is different; responses: Ab-dependent cellular cytotoxity, release of numerous Fc␥RI specifically binds to IgG2a with a high affinity, whereas inflammatory mediators, and expression changes of cell surface Fc␥RIII binds to IgG1, IgG2a, and IgG2b with a low affinity (2, 3). proteins involved in cell-cell adhesion and Ag presentation (1–3). In contrast, Fc␥RIIB is a monomeric receptor containing the im- Three classes of Fc␥R (Fc␥RI, Fc␥IIB, and Fc␥III) are ex- munoreceptor tyrosine-based inhibitory motif (ITIM), which re- pressed on mouse macrophages and share a highly homologous cruits the phosphatases Src homology 2 protein-1 and Src homol- extracellular portion for the IgG binding domain (1, 2). However, ogy 2 domain-containing inositol phosphatase, and it is unknown there are structural and functional differences in the various recep- how Fc␥RIIB contributes to phagocytosis in macrophages (2, 7). tor family members. Fc␥RI and Fc␥RIII exist as oligomeric com- CD9, which is highly expressed in macrophages, is a cell sur- plexes in which the ␣-chain, bearing the IgG binding domain, as- face glycoprotein belonging to the transmembrane 4 superfamily sociates with ␥-chain dimers that bear an immunoreceptor (TM4SF). The TM4SF is a group of cell surface proteins, includ- tyrosine-based activation motif (ITAM)3; these receptors do not ing at least 16 members such as CD37, CD53, CD63, CD81, and contain intrinsic tyrosine kinase activity (2, 4). On cross-linking of CD82. The structure of these proteins is typified by four hydro- phobic domains spanning the cell membrane and short N- and

*Department of Life Science, Tokyo Institute of Technology, Yokohama, Japan; †De- C-terminal cytoplasmic domains (around 5–14 aa in length) (8, 9). partment of Immunology, Saga Medical School, Saga, Japan; and ‡Department of CD9 was reported to be associated with various integrin family Experimental Immunology, Institute of Development, Aging and Cancer, Tohoku molecules, CD5, CD19, and other TM4SF proteins on the cell University, Sendai, Japan surface and has been postulated to participate in the regulation of Received for publication July 18, 2000. Accepted for publication December 19, 2000. cell growth, motility, and signaling (8–16). Besides monocytes/ The costs of publication of this article were defrayed in part by the payment of page macrophages, CD9 is expressed on certain hematopoietic lineage charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. cells such as platelets, subpopulations of lymphocytes, eosino- 1 This work was supported by a grant-in-aid for scientiﬁc research from the Japan phils, and basophils and on some other cell lineages such as en- Ministry of Education, Science, Sports and Culture. dothelial cells, myoblasts, and neuroblasts (8, 15, 16). Addition- 2 Address correspondence and reprint requests to Dr. Akira Kudo, Department of Life ally, CD9 is expressed in oocytes, and CD9-deﬁcient female mice Science, Tokyo Institute of Technology, 4259 Nagatsuta, Midoriku, Yokohama 226- showed serious sterility caused by a defect in the gamete fusion 8501, Japan. 3 Abbreviations used in this paper: ITAM, immunoreceptor tyrosine-based activation motif; ITIM, immunoreceptor tyrosine-based inhibition motif; TM4SF, transmem- toxin; MAPK, mitogen-activated protein kinase; M␤CD, methyl-␤-cyclodextrin; brane 4 superfamily; DIGs, detergent-insoluble glycolipid membranes; CTx, cholera BSS, buffer saline solution.

FIGURE 1. Tyrosine phosphorylation and cell aggregation induced by anti-CD9 mAb. A, J774 cells were incubated with 10 ␮g/ml of a control Ab, ␥ ␤ anti-CD9, anti-Fc RIIB/III, and anti- 1 integrin mAb, respectively, for 5 min at 37°C. Tyrosine-phosphorylated proteins were detected by anti- by guest on September 29, 2021 phosphotyrosine mAb PY99 in Western blotting. B, Before or after stimulation with anti-CD9 mAb (10 ␮g/ml, 37°C, 5 min), the lysates of J774 cells were precipitated with either anti-syk or anti-cbl Ab. Phosphotyrosine of immunoprecipitates was detected by PY99 in Western blotting. Blots were stripped and reprobed with anti-syk or anti-cbl Ab to reveal equal loading. Data are representative of three similar experiments. C, J774 cells were incubated with a

control mAb or anti-CD9 mAb for 10, 30, 60, 120, and 180 min in 1 mM CaCl2/HCMF at 37°C under constant agitation. The ratio of aggregation was represented by the index Nt/N0, where N0 is the initial cell number before incubation and Nt is the total particle number at incubation time t. Results are presented as mean Ϯ SD of four experiments. D, J774 cells were transferred to Counter chamber after incubation for 120 min with a control mAb (left) or anti-CD9 mAb (right), and aggregated cells were demonstrated by phase contrast photographs (ϫ200).

process (17–20). However, no abnormalities were detected in any duced much less TNF-␣ production and p42/44 MAPK phosphor- other tissues in CD9-deficient mice (18–20), suggesting that the ylation than Fc␥RIIB/III cross-linking. Finally, our results suggest function(s) of CD9 may be compensated by the presence of other that CD9 on mouse macrophages functionally associates with TM4SF proteins. Fc␥Rs and may modify signals for phagocytosis and inflammatory In human platelets, CD9 is thought to functionally associate responses. with Fc␥RIIA, an isoform of Fc␥RII bearing the ITAM in the cytoplasmic tail, which is not found in mice, and co-cross-linking of CD9 and Fc␥RIIA induces cell aggregation and activation. Materials and Methods Moreover, in various types of cells, CD9 is proposed to be in- Reagents volved in signal transduction coupled with cell activation, proliferation, and adhesion, as mAbs against CD9 can induce the acti- The monoclonal rat anti-mouse CD9 Ab, KMC8.8, was reported previously ␬ vation of T cells, the homotypic aggregation of pre-B cells, the (24), the isotype of which is IgG2a . KMC8.8 and anti-IgM mAb (MA6), as an isotype (IgG2a␬)-matched control mAb, were purified from ascitic adhesion and proliferation of Schwann cells, and can inhibit the fluids as described (24), and KMC8.8 was biotinylated by the manufactur- migration of leukemia cells (13, 21–23). However, in mouse mac- Ј er’s instruction (Pierce, Rockford, IL). F(ab )2 of KMC8.8 were generated ␥ ␤ rophages, the function of CD9 has not yet been investigated, and by Takara Shuzo (Tokyo, Japan). Anti-Fc RIIB/III mAb (2.4G2), anti- 1 the association of molecules with CD9 on macrophages are not integrin mAb, biotinylated anti-ICAM-1 mAb, and FITC-conjugated anti- known, despite their high level of CD9 expression. Mac-1 mAb were obtained from BD PharMingen (San Diego, CA). Anti- In the present study, we found that an anti-CD9 mAb, KMC8.8, syk mAb, anti-cbl-b mAb, anti-phosphorylated p42/44 MAPK mAb, anti-p42 MAPK mAb, and HRP-conjugated anti-phosphotyrosine mAb induced protein tyrosine phosphorylation, filopodium extension, (PY99) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). and cell aggregation caused by Fc␥RIIB/III-CD9 co-cross-linking Mouse anti-rat Fab polyclonal Abs, HRP-conjugated anti-rabbit IgG, anti- on mouse macrophages. The Fc␥RIIB/III-CD9 co-cross-linking in- mouse IgG Ab, and FITC-conjugated anti-rat IgG Ab were purchased from 3258 CD9 IN MACROPHAGES

Detection of protein tyrosine phosphorylation by Western blottimg J774 cells and bone marrow macrophages were washed and resuspended at 1 ϫ 107 cells/ml in DMEM or ␣-MEM containing 10 ␮g/ml indicated mAb. In some experiments, J774 cells were pretreated with 10 ␮g/ml anti- Fc␥RIIB/III mAb, 2.4G2 on ice for 20 min before stimulation with anti-rat IgG Ab or anti-CD9 mAb, KMC8.8. After 1–20 min incubation at 37°C, cells were washed with cold PBS and lysed with TNE buffer (10 mM Tris-HCl, pH 7.8, 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, 1 mM ␮ ␮ Na3VO4, 2 mM PMSF, 1 g/ml aprotinin, 100 g/ml leupeptin) on ice for 30 min. The cell lysate (30 ␮g/lane) was separated by 7.5% or 10% SDS- PAGE in reducing conditions, electrophoretically transferred to a nitrocel- lulose membrane, and then reacted with Abs. Bound Abs were visualized by chemiluminescence by using an ECL immunoblotting kit (Amersham, Buckinghamshire, U.K.). To strip bound Abs, the membranes were incubated in stripping buffer (100 mM 2-ME, 2% SDS, 62.5 mM Tris-HCl, pH 6.8) at 50°C for 30 min.

Immunoprecipitation

J774 cells were lysed with TNE buffer before or after stimulation with Downloaded from KMC8.8 for 5 min at 37°C. Four micrograms of anti-syk or cbl-b mAb plus protein G-Sepharose beads (Pharmacia Biotech, Uppsala, Sweden) were added to lysates, and they were incubated for2hat4°Cunder constant agitation. After washing beads with TNE buffer, the immunoprecipitated proteins were subjected to SDS-PAGE and Western blotting.

Aggregation assay http://www.jimmunol.org/ J774 cells were incubated with Abs at 37°C for the indicated time under 2ϩ constant agitation in 1 mM CaCl2/HCMF (10 mM HEPES-buffered Ca , Mg2ϩ-free Hank’s solution) containing 1% BSA. After incubation for t minutes, the total number of aggregated particles and cells in the cell sus-

pension (Nt) was counted with a counter chamber. The ratio of aggregation was represented by the index Nt/N0, where N0 is the initial cell number before aggregation (27).

Ј by guest on September 29, 2021 FIGURE 2. No effects of F(ab )2 of KMC8.8 in tyrosine phosphoryla- TNF-␣ measurement tion and cell aggregation. A, J774 cells were incubated with 10 ␮g/ml of ϫ 4 ␮ F(abЈ) or the whole of anti-CD9 mAb KMC8.8 for 1 or 5 min at 37°C. J774 cells (5 10 ) were cultured with 10 g/ml of KMC8.8 or 2.4G2 plus 2 anti-rat IgG Ab in 100 ␮l DMEM containing 10% FBS for various times After lysis of cells, tyrosine-phosphorylated proteins were detected by anti- (0–24 h) at 37°C. The amount of TNF-␣ in the collected supernatant was phosphotyrosine mAb PY99 in Western blotting. To reveal equal loading, measured with an ELISA kit (Amersham) according to the manufacturer’s blots were stripped and reprobed with anti-syk Ab. Data are representative instructions. of three similar experiments. B, J774 cells were incubated with a control Ј mAb, F(ab )2 of KMC8.8 and KMC8.8, respectively, for 60 min in 1 mM CaCl2/HCMF at 37°C under constant agitation. The ratio of aggregation, Cholesterol extraction N /N , was calculated. Results are presented as mean Ϯ SD of three 60 0 To remove cholesterol, suspended J774 cells (5 ϫ 106 cells/ml) were in- experiments. cubated for 15 or 30 min at 37°C in the presence of 20 mM M␤CD in BSA-containing buffer saline solution (BSA/BSS: 20 mM HEPES, pH 7.4,

135 mM NaCl, 5 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 5.6 mM glucose, and 1 mg/ml BSA; Ref. 28) and then washed with BSA/BSS before stimulation or analysis by flow cytometry. The cells stained with anti-CD9 Jackson ImmunoResearch (West Grove, PA), Southern Biotechnology As- mAb or anti-Fc␥RIIB/III mAb followed by FITC-conjugated anti-rat IgG sociates (Birmingham, AL), Bio-Rad (Hercules, CA), and ICN Pharma- were analyzed by using FACSCalibur (Becton Dickinson, San Jose, CA). ceuticals-Cappel Products (Costa Mesa, CA), respectively. FITC-conjugated cholera toxin (CTx) and methyl-␤-cyclodextrin (M␤CD) were obtained from Sigma (St. Louis, MO). Rhodamine-conjugated phalloidin Immunohistochemistry and laser-scanning confocal microscopy and FITC-conjugated streptavidin were purchased from Molecular Probes (Eugene, OR). J774 cells or bone marrow macrophages cultured on a coverslip were stimulated with KMC8.8 at 37°C for 0, 5 and 60 min. After washing with PBS, cells were fixed in 4% paraformaldehyde/PBS for 30 min at 25°C and then Cells permeabilized and blocked with 0.1% Triton X-100 and 1% skim-milk in PBS. Immunofluorescence staining was conducted by using Abs to CD9, The mouse macrophage cell line J774 was cultured in DMEM with 10% ␥ ␤ Fc RIIB/III, and 1 integrin, followed by a second Ab, FITC-conjugated FBS. M-CSF-dependent bone marrow macrophages were prepared from goat anti-rat IgG. ICAM-1 and Mac-1 were visualized with biotinylated wild-type mice and mice deficient in the FcR ␥-chain, Fc␥RIIB, and anti-ICAM-1 mAb, followed by FITC-conjugated streptavidin and FITC- Fc␥RIII as described previously (25, 26). In brief, the bone marrow cells conjugated anti-Mac1 mAb, respectively. F-actin and ganglioside GM1, a from femurs and tibias of adult mice were cultured in ␣-MEM containing detergent-insoluble glycolipid membranes (DIGs) marker (29), were 10% FBS and 1/10 vol of culture supernatant from the M-CSF-producing stained with rhodamine-conjugated phalloidin and FITC-conjugated CTx ␤ cell line CMG14-12 (final ϳ35,000 U/ml of M-CSF) for 3 days. After subunit, respectively. Microscopy was performed with a laser-scanning removing cells in suspension, adherent M-CSF-dependent bone marrow confocal microscope (CSU10; Yokogawa Electric, Tokyo, Japan), and cap- macrophages were harvested with 0.02% EDTA/PBS treatment and refed tured images were processed and superimposed by using the IPLab soft- or used for experiments. ware package (Scanalytics, Billerica, MA). The Journal of Immunology 3259 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 3. Protein tyrosine phosphorylation induced by co-cross-linking of Fc␥RIII and CD9. A, J774 cells were incubated in the presence or absence of anti-Fc␥RIIB/III mAb, 2.4G2, for 20 min at 4°C. After washing, cells were incubated with a control Ab, anti-rat IgG Ab, or anti-CD9 mAb, KMC8.8, for 5 min at 37°C. Phosphotyrosine was detected by PY99 in Western blotting. Reblotting with anti-syk mAb was done after stripping to reveal equal loading. B and C, M-CSF-dependent bone marrow macrophages from wild-type, FcR ␥-chainϪ/Ϫ,Fc␥RIIBϪ/Ϫ, and Fc␥RIIIϪ/Ϫ mice were stimulated with Ј KMC8.8 for 5 min at 37°C (B), or with KMC8.8-F(ab )2 for 5 min at 37°C followed by mouse anti-rat Fab Ab treatment for 15 min at 4°C (C). Western blotting was conducted as described in A. Data are representative of three (A) or two (B and C) similar experiments.

Results cipitation followed by Western blot analysis confirmed that the Protein tyrosine phosphorylation and cell aggregation induced phosphorylated proteins of 72 and 120 kDa were a nonreceptor by anti-CD9 mAb in the mouse macrophage cell line J774 tyrosine kinase, syk, and an adapter protein, cbl, respectively (Fig. 1B). Moreover, KMC8.8 induced time-dependent cell aggregation In human platelets, anti-CD9 mAbs induce protein tyrosine phos- in the presence of 1 mM CaCl (Fig. 1, C and D, right), and the phorylation and cell aggregation (10–12). To investigate the func- 2 same result was obtained in the absence of CaCl (data not shown). tion of CD9 on macrophages, the effect of an anti-CD9 mAb, 2 In the previous report, the activation of human platelets induced KMC8.8, on the mouse macrophage cell line, J774 was examined. by anti-CD9 mAbs required Fc receptor expression (10–12). After incubation with the mAb for 5 min at 37°C, J774 cells were Therefore, to avoid Fc␥R engagement in signal transduction, lysed, and tyrosine-phosphorylated proteins were detected by an cross-linking by F(abЈ) of KMC8.8 on J774 cells was examined. anti-phosphotyrosine mAb, PY99, in Western blot analysis. As 2 The incubation with F(abЈ) of KMC8.8 for 1 or 5 min did not shown in Fig. 1A, KMC8.8 induced the tyrosine phosphorylation 2 induce protein tyrosine phosphorylation (Fig. 2A) and did not of 55-, 60-, 72-, and 120-kDa proteins, and the same signals were cause cell aggregation after 60 min (Fig. 2B). The results suggest obtained by cross-linking with an anti-Fc␥RIIB/III mAb, 2.4G2. that the protein tyrosine phosphorylation and cell aggregation in- Neither a control mAb nor a mAb against ␤ integrin, which as- 1 duced by KMC8.8 require co-cross-linking of CD9 and Fc␥R with sociates with CD9 in B cells and endothelial cells (9), induced a whole molecule of the anti-CD9 Ab on J774 cells. protein tyrosine phosphorylation, despite being the same isotype as KMC8.8. CD43 and CD44 are expressed on J774 cells at a similar level to CD9, and mAbs against both surface molecules did not Specific induction of protein tyrosine phosphorylation by co- ␥ induced tyrosine phosphorylation. Furthermore, incubation with cross-linking of Fc RIII and CD9 with KMC8.8 KMC8.8 did not induce tyrosine phosphorylation in macrophages On mouse macrophages, three types of Fc␥R, Fc␥RI, Fc␥RIIB, derived from CD9-deficient mice (data not shown). Immunopre- and Fc␥RIII, are expressed. Fc␥RI and Fc␥RIII, but not Fc␥RIIB, 3260 CD9 IN MACROPHAGES Downloaded from http://www.jimmunol.org/

FIGURE 4. Different effect of anti-CD9 mAb from that of anti- Fc␥RIIB/III mAb in cell aggregation. A, J774 cells were incubated with or FIGURE 5. TNF-␣ production or p42/44 MAPK phosphorylation by by guest on September 29, 2021 without 10 ␮g/ml anti-Fc␥RIIB/III mAb 2.4G2 for 20 min at 4°C. After cross-linking by anti-CD9 mAb or anti-Fc␥RIIB/III mAb. A, J774 cells washing, either none or10 ␮g/ml anti-rat IgG Ab or anti-CD9 mAb (5 ϫ 104) were cultured with 10 ␮g/ml of anti-Fc␥RIIB/III mAb 2.4G2 KMC8.8 were reacted to cells for 1–20 min at 37°C. Phosphotyrosine was plus anti-rat IgG Ab (solid line) or anti-CD9 mAb KMC8.8 (dotted line) for ␣ detected by PY99 in Western blotting. To reveal equal loading, blots were 0, 12, 18, and 24 h (37°C, 5% CO2). The amount of TNF- in the collected stripped and reprobed with anti-syk Ab. Data are representative of three supernatant was measured with an ELISA kit. Results are presented as similar experiments. B, J774 cells were incubated with a control mAb, mean Ϯ SD of two experiments. When error bars were not seen, their sizes anti-Fc␥RIIB/III mAb 2.4G2, plus anti-rat IgG Ab, anti-CD9 mAb were smaller than the symbols. B, J774 cells were incubated with 10 ␮g/ml

KMC8.8, for 60 min in 1 mM CaCl2/HCMF at 37°C under constant agi- of a control mAb, 2.4G2, plus anti-rat IgG Ab or KMC8.8 for 5 min at tation. The ratio of aggregation, N60/N0, was calculated. Results are pre- 37°C. Phosphotyrosine was detected by PY99 in Western blotting. Phos- sented as mean Ϯ SD of three experiments. phorylated p42/44 MAPK and p42 MAPK were detected with the corre- sponding Ab on the same membrane after stripping. Data are representative of three similar experiments. C, After western blotting described above, transduce tyrosine phosphorylation signals through the ITAM of levels of phosphorylated p42 MAPK and p42 MAPK were quantified by the ␥ subunit (2), whereas Fc␥RIIB transduces dephosphorylation densitometric analysis, and the specific phosphorylation was calculated by signals through the ITIM. We tried to identify which Fc␥Rs can dividing the densitometric values for phosphorylated p42 MAPK by those Ϯ transduce tyrosine phosphorylation signals by cross-linking to- for p42 MAPK. Data are the mean ( SD) of values relative to the specific gether with CD9. We pretreated J774 cells with 2.4G2, used as the phosphorylation in cells treated with control mAb from three independent experiments. Fc␥RIIB/III blocker, at 4°C for 20 min before treatment of KMC8.8 so as to block co-cross-linking between Fc␥RIIB/III and CD9 in the treatment of KMC8.8. In J774 cells pretreated with 2.4G2, slight tyrosine phosphorylation of proteins, including syk, FcR ␥-chainϪ/Ϫ,Fc␥RIIBϪ/Ϫ, and Fc␥RIIIϪ/Ϫ mice. These mac- was observed after incubation at 37°C for 5 min (Fig. 3A, lane 4), rophages express functional Fc␥RI, Fc␥RIIB, and Fc␥RIII and the addition of anti-rat IgG Ab enhanced the level of phos- (wild-type), Fc␥RIIB (FcR ␥-chainϪ/Ϫ), Fc␥RI and Fc␥RIII phorylation (Fig. 3A, lane 5). This indicated that 2.4G2 did bind to (Fc␥RIIBϪ/Ϫ), or Fc␥RI and Fc␥RIIB (Fc␥RIIIϪ/Ϫ), as the FcR Fc␥RIIB/III at 4°C. The addition of KMC8.8 to J774 cells pre- ␥-chain is critical in facilitating either surface expression or ligand treated with 2.4G2 did not enhance the level of phosphorylation binding of the Fc␥RI and Fc␥RIII (30). Stimulation with KMC8.8 (Fig. 3A, lane 6), although KMC8.8 induced strong phosphoryla- for 5 min at 37°C induced tyrosine phosphorylation of proteins, tion in untreated J774 cells (Fig. 3A, lane 3). These results sug- including syk, in bone marrow macrophages from wild-type and gested that Fc␥RIII is more important than Fc␥RI for phosphory- Fc␥RIIBϪ/Ϫ mice (Fig. 3B, lanes 2 and 6) but not from FcR lation signals induced by KMC8.8. To confirm this, we prepared ␥-chainϪ/Ϫ and Fc␥RIIIϪ/Ϫ mice (Fig. 3B, lanes 4 and 8). These M-CSF-dependent bone marrow macrophages from wild-type, results demonstrate that co-cross-linking of Fc␥RIII and CD9 by The Journal of Immunology 3261

KMC8.8 induces tyrosine phosphorylation signals. Mouse Fc␥RI damine-conjugated phalloidin. Time-dependent filopodium exten- shows 10-fold higher binding affinity to Fc region of mouse sion was observed by staining of F-actin in J774 cells (Fig. 6). IgG2a␬ than Fc␥RIIB and Fc␥RIII (2). The isotype of rat anti-CD9 Additionally, localization of CD9 at filopodia was also observed mAb, KMC8.8, is IgG2a␬; however, the binding affinity of mouse (Fig. 6). Moreover, after Ab treatment, many cells adhered to Fc␥Rs is to the Fc region of rat IgG2a␬ is not well known. Thus, neighboring cells as if cells had moved toward the cells attached we examined whether co-cross-linking of Fc␥Rs and CD9 with on the extended filopodia, and CD9 and F-actin were densely co- Ј KMC8.8-F(ab )2 followed by mouse anti-rat Fab polyclonal Abs localized at the cell-cell adhesion sites (Fig. 6, arrow). CD9 has induces tyrosine phosphorylation signals. Treatment of macro- not been reported to be associated with actin filaments or to func- ␥ Ϫ/Ϫ ␥ Ϫ/Ϫ phages from wild-type, FcR -chain ,FcRIIB , and tion in cell-cell or cell-matrix adhesion directly, but CD9 can make Fc␥RIIIϪ/Ϫ mice with KMC8.8-F(abЈ) followed by the addition 2 complexes with CD5, CD19, heparin-binding-epidermal growth of mouse anti-rat Fab Abs induced tyrosine phosphorylation ex- factor, or ␤ integrin on the plasma membrane (8, 9, 13–16, 33). cept for macrophages from FcR ␥- chainϪ/Ϫ (Fig. 3C). Moreover, 1 Therefore, we examined whether other molecules are colocalized in macrophages lacking Fc␥RIIB containing the ITIM, the phos- at the cell-cell adhesion site after stimulation with KMC8.8. As phorylation level of syk was stronger than that in wild-type mac- shown in Fig. 7, the localization of Fc␥RIIB/III, Mac-1 (␣ ␤ rophages (Fig. 3, B and C). These results may suggest that co- M 2 cross-linking of any combination of Fc␥R and CD9 transduces integrin), and ICAM-1 at the cell-cell adhesion site was observed some signals or enhances FcR signals. after stimulation, and these molecules were colocalized with CD9 Because no specific Ab to each Fc␥R is available, it is difficult or F-actin at this site as detected by double staining (Fig. 7A, ␤ to compare the effects of co-cross-linking between CD9 and each arrow). In contrast, 1 integrin did not change localization after Downloaded from Fc␥R. Therefore, we compared the effects of Fc␥R-CD9 cross- stimulation, while F-actin did concentrate at the cell-cell adhesion linking by KMC8.8 with that of Fc␥RIIB/III cross-linking by the site (Fig. 7B, arrow). anti-Fc␥RIIB/III Ab, 2.4G2. In addition, localization of the ganglioside GM1, used as a marker of DIGs, was also observed at the cell-cell adhesion site ␥ Different effects of anti-CD9 and anti-Fc RIIB/III mAbs on with FITC-conjugated CTx ␤ subunit, suggesting that CD9 existed macrophage cell aggregation and TNF-␣ production at DIGs (Fig. 7A). http://www.jimmunol.org/ J774 cells were stimulated with KMC8.8, 2.4G2 alone, or 2.4G2 plus anti-rat IgG Ab, and then tyrosine-phosphorylated proteins were detected by Western blot analysis. An anti-Fc␥RIIB/III mAb, 2.4G2, induced time-dependent protein tyrosine phosphorylation, and stronger phosphorylation was induced by super cross-linking of 2.4G2 with a secondary Ab or with KMC8.8 alone (Fig. 4A). Moreover, in an aggregation assay, unlike incubation with KMC8.8, super cross-linking of 2.4G2 with anti-rat IgG Ab did not induce cell aggregation after 60 min (Fig. 4B), despite the similar by guest on September 29, 2021 high level of tyrosine phosphorylation (Fig. 4A). The cross-linking of Fc␥Rs by IgG immune complexes triggers a wide variety of effector functions including phagocytosis, Ab-dependent cellular cytotoxity, and release of inflammatory mediators like IL-1, IL-10, and TNF-␣ (6, 31, 32). We stimulated J774 cells with KMC8.8 or 2.4G2 plus anti-rat IgG Ab for 0, 12, 18, and 24 h and measured the TNF-␣ production in the supernatant with an ELISA kit. Fc␥RIIB/III cross-linking by 2.4G2 plus anti-rat IgG Ab induced the production of TNF-␣ from J774 cells rapidly, whereas Fc␥R- CD9 co-cross-linking by KMC8.8 induced less TNF-␣ (Fig. 5A). An increase in TNF-␣ mRNA expression was observed in J774 cells cultured with 2.4G2 plus anti-rat IgG Ab for 12 h but not with KMC8.8 (data not shown). It has been reported that the activation of p42 MAPK is necessary for TNF-␣ synthesis induced by Fc␥R cross-linking (6); therefore, the phosphorylation of p42/44 MAPK was examined. As shown in Fig. 5B, 2.4G2 plus anti-rat IgG Ab induced the phosphorylation of p42/44 MAPK after stimulation for 5 min. However, KMC8.8 did not induce but rather reduced it, although other major tyrosine-phosphorylated proteins were much the same in both treatments (Fig. 5, B and C). These results suggest that cross-linking of Fc␥R and CD9 with KMC8.8 induces differ- FIGURE 6. Filopodia extension and colocalization of CD9 and F-actin ent signals from that of Fc␥RIIB/III with 2.4G2. at the cell-cell adhesion site induced by anti-CD9 mAb. J774 cells were stimulated with anti-CD9 mAb KMC8.8 at 37°C for 0, 5, and 60 min. After Induction of filopodium extension and colocalization of CD9, fixation, permeabilization, and blocking, CD9 (left) or F-actin (center) was Fc␥R, Mac-1, ICAM-1, and F-actin after CD9 cross-linking stained with KMC8.8 followed by FITC-conjugated anti-rat IgG Ab or rhodamine-conjugated phalloidin. Right panels represent superimposed KMC8.8 induced aggregation of J774 cells in a suspension culture. images. Upper three panels indicate the adhesive sections to a coverslip To observe the process of cell aggregation, J774 cells adhered on observed by using a confocal microscopy. The lowest panels show images a coverslip were stimulated with KMC8.8. After fixing the cells, of the middle section of the cells. After stimulation with the Ab, CD9 and localization of CD9 or F-actin was examined by staining with F-actin were colocalized at the cell-cell adhesion site (arrows). Data are KMC8.8, followed by FITC-conjugated anti-rat IgG Ab or rho- representative of four similar experiments. 3262 CD9 IN MACROPHAGES

FIGURE 7. Localization of Fc␥RIIB/ III, Mac-1, ICAM-1, and GM1 at the cell- cell adhesion site. A and B, Before or after stimulation with anti-CD9 mAb KMC8.8 for 60 min, Fc␥RIIB/III, Mac-1, ICAM-1, ␤ 1 integrin, and F-actin were stained with Abs and reagents as described in Materials and Methods.InA, ganglioside GM1 was stained with FITC-conjugated CTx ␤ subunit as a DIGs marker. Right panels present the superimposed images of the two left panels, respectively. After stimulation with the Ab, CD9, F-actin, Fc␥RIIB/ Downloaded from III, Mac-1, ICAM-1, and GM1 were localized at the cell-cell adhesion site (A, arrows). In B, after stimulation with the ␤ Ab, 1 integrin was not localized at the cell-cell adhesion site, although F-actin was localized at the site (arrow). Data are

representative of three similar experiments. http://www.jimmunol.org/ by guest on September 29, 2021

Diminution of the signal from Fc␥RIIB/III or Fc␥R-CD9 cross- lation (Fig. 9,A and C, arrow). However, in FcR ␥-chainϪ/Ϫ and linking after extraction of cholesterol from DIGs Fc␥RIIIϪ/Ϫ bone marrow macrophages, CD9 and Fc␥RIIB were DIGs are postulated to represent plasma membrane domains that colocalized at the cell-cell adhesion site, but F-actin and Mac-1 may function as centers for signal transduction and membrane traf- were not (Fig. 9, B and D, arrow). These results suggest that ␥ ficking (28, 29). However, Fc␥R has not been reported to be in- Fc RIII is crucial for the localization of F-actin and Mac-1 at the volved in DIGs, whereas Src family kinases, which have been cell-cell adhesion site after stimulation with KMC8.8. In addition, ␥ implicated in the Fc␥R signals, exist in DIGs (28, 29). Therefore, it is possible that Fc RIIB can transduce some signal by cross- ␥ we tested the effects of disrupting DIGs on Fc␥R-CD9 cross-link- linking with CD9, because Fc RIIB and CD9 are colocalized at ␥ ing by extracting cholesterol with M␤CD (28). The M␤CD treat- the cell-cell adhesion site in Fc RIII-deficient bone marrow ment (20 mM for 30 min at 37°C) did not change the amount of macrophages. Fc␥RIIB/III and CD9 on the cell surface (Fig. 8A). However, tyrosine phosphorylation signals induced by both 2.4G2 plus anti-rat IgG or KMC8.8 were diminished in a time-dependent manner (Fig. Discussion 8B). This result suggests that DIGs may function in Fc␥R signaling The interaction of FcR with immune complexes at the cell surface in J774 cells. The result that KMC8.8 induces strong signals of induces a wide variety of immune responses. Our studies demon- tyrosine phosphorylation, filopodium extension and cell aggrega- strate that CD9, a member of the tetraspan family, is functionally tion without super cross-linking by a second Ab may be explained correlated with Fc␥Rs on the macrophage. In a previous report, an by localization of CD9 to a specific membrane structure like DIGs Ab against another tetraspan molecule, CD82, activated the human as reported in T cell activation induced by anti-CD9 mAb (13). monocyte cell line U937 and induced an increase in intracellular calcium after FcR-CD82 co-cross-linking (34). In addition, an in- The localization of Mac-1 and F-actin at the cell-cell adhesion teraction between FcR and CD9 has been reported on human plate- ␥ site induced by co-cross-linking of Fc RIII and CD9 lets; a monoclonal anti-CD9 Ab treatment resulted in phosphory- Ј To investigate the contribution of Fc␥R to the colocalization of lation of syk, and an intact anti-CD9 Ab, but not F(ab )2, induced CD9, F-actin, Fc␥RIIB/III, and Mac-1 at the cell-cell adhesion site rapid platelet aggregation (35). Moreover, recombinant Fc␥RIIa2 (Figs. 6 and 7), we examined the localization of these proteins by inhibited Fc-dependent anti-CD9-induced platelet aggregation in a immunostaining before or after stimulation with KMC8.8. In both dose-dependent manner (10). These results have not been observed wild-type and Fc␥RIIBϪ/Ϫ bone marrow macrophages, CD9, F- in mouse cells, and it may result from the fact that human platelets actin, Fc␥RIIB/III (in wild type), or Fc␥RIII (in Fc␥RIIBϪ/Ϫ) and express ITAM-containing Fc␥RIIA but mouse platelets do not. We Mac-1 were colocalized at the cell-cell adhesion site after stimu- demonstrated that treatment of an anti-CD9 mAb, KMC8.8, but The Journal of Immunology 3263

Ј cross-linking with an anti-CD9 F(ab )2 followed by mouse anti-rat Fab Abs induced phosphorylation of syk in macrophages from Fc␥RIIIϪ/Ϫ but not from FcR ␥-chainϪ/Ϫ mice. These results suggest that CD9 functionally associates with all Fc␥Rs and modifies the signals. Our data suggest that CD9 may be involved in Fc␥R-mediated phagocytosis and enhance or modify it like Mac-1/CR-3, a major receptor in the phagocytosis of complement-opsonized targets (36, 37). Moreover, in our preliminary experiments, colocalization of FcR and CD9 was observed in the phagosome when IgG-opsonized particles were engulfed (data not shown). Recently, it has been demonstrated that the small G proteins Cdc42/Rac and Rho are necessary for Fc␥R- and Mac-1-dependent phagocytosis, respectively (38), and the cross-talk of signal transduction among the small G proteins has been described (39). The excessive filopodium extension and concentration of F-actin at the cell-cell adhesion site induced by the CD9-Fc␥R co-cross-linking are thought to be a result of the activation of the small G proteins, which regu- lates the actin cytoskeleton. In fact, CD9 has been reported to be Downloaded from associated with small G proteins (40). The CD9 and Fc␥R co- cross-linking may specifically enhance the signals of small G proteins downstream of Fc␥R, and the machinery could be involved in the Fc␥R-Mac-1-dependent phagocytosis. We demonstrated the possibility that CD9 could work functionally in associating with Fc␥R in DIGs in J774 cells like CD9 in T cells behaving as a http://www.jimmunol.org/ coreceptor of TCR in DIGs (13). In some cells, co-cross-linking between surface molecules is constitutively present in DIGs and ITAM-containing immunoreceptors, which induces recruitment to lipid raft and activation of the receptors. However, these phenom- ena may include some artifacts and do not reflect the biological function of these molecules. Concerning CD9 on bone marrow macrophages, M␤CD does not cancel the activation with KMC8.8

out of accord with the experiment carried with J774 cells under the by guest on September 29, 2021 FIGURE 8. Diminution of tyrosine phosphorylation signal induced by cross-linking of Fc␥RIIB/III or Fc␥R-CD9 under cholesterol extraction. A, same condition (data not shown), so it is obscure whether DIGs Before (upper) or after (lower) incubation of J774 cells in the presence of contribute to the activation induced with KMC8.8. Therefore, fur- 20 mM M␤CD in BSA/BSS for 30 min at 37°C, surface expression of ther investigation is required. Fc␥RIIB/III (left) and CD9 (right) were analyzed by flow cytometry. The We and others have shown that CD9-deficient mice exhibit de- broken line shows the FACS profile from the negative control, streptavidin fective cell fusion between sperm and eggs, but no significant im- FITC for CD9, or FITC-conjugated anti-rat IgG for Fc␥RIIB/III. B, After mune system defects were observed (18–20). We revealed that the ␤ treatment with 20 mM M CD in BSA/BSS for 0, 15, and 30 min at 37°C, structure of microvilli on the egg plasma membrane seems to cap- J774 cells were incubated with a control Ab, anti-Fc␥RIIB/III mAb 2.4G2, ture the sperm before fertilization evokes phagocytosis and that plus anti-rat IgG Ab, anti-CD9 mAb KMC8.8, for 5 min at 37°C. Phos- photyrosine was detected by PY99 in Western blotting. Reblotting with CD9 is localized at the sperm-egg adhesion site (20). Sperm-egg anti-syk mAb was done after stripping to reveal equal loading. Data are fusion must require tight adherence of the egg plasma membrane representative of two (A) or three (B) similar experiments. to the sperm, likely being mediated by the reorganization of the actin cytoskeleton. We suggest that CD9 functions in this process. In vivo, the reorganization of the actin cytoskeleton after rec- ognition of IgG immune complexes by FcR is extremely impor- not the Fc fragment induces: 1) cell aggregation, 2) stronger phos- tant, not only for internalization of the complex, but also for ac- phorylation of syk, and 3) filopodium extension in macrophages, tivation or stabilization of adhesion molecules, which lead to resulting from co-cross-linking of Fc␥RIIB/III and CD9. More- chemotaxis and invasion of the inflammatory site by monocytes/ over, Fc␥RIII was critical for colocalization of Mac-1 and F-actin macrophages. CD9 has been reported to be involved in cell adhe- induced by the CD9 and Fc␥RIIB/III co-cross-linking with sion and motility in various types of cells, both of which require KMC8.8. the reorganization of the cytoskeleton. Our findings demonstrate Although syk was not phosphorylated after incubation with that co-cross-linking of CD9 and Fc␥R activates macrophages; KMC8.8 in Fc␥RIII-deficient macrophages, colocalization of therefore, CD9 may play a role when FcRs function in infection Fc␥RIIB and CD9 was detected in these cells, suggesting that and inflammation on macrophages. The detailed function of CD9 KMC8.8 can co-cross-link both Fc␥RIIB with CD9 as well as on macrophages in vivo will be investigated in Ag-stimulated Fc␥RIII. However, we could not detect the physical interaction of CD9-deficient mice. these molecules with CD9 in immunoprecipitation experiments using each detergent, Nonidet P-40, 3-[(3-cholamidopropyl)dimeth- ylammonio]-1-propanesulfonate, or Briji98 (data not shown). Acknowledgments Besides the association of Fc␥RIIB/III and CD9, it seems that We thank Drs. Christopher Paige and Heather Fleming for critical reading there is a functional association between Fc␥RI and CD9, as co- of the manuscript. 3264 CD9 IN MACROPHAGES Downloaded from

FIGURE 9. Localization of Mac-1 and F- http://www.jimmunol.org/ actin at the cell-cell adhesion site induced by co-cross-linking of Fc␥RIII and CD9. A–D, Bone marrow macrophages from wild-type (A), FcR ␥-chainϪ/Ϫ (B), Fc␥RIIBϪ/Ϫ (C), and Fc␥RIIIϪ/Ϫ (D) mice were stimulated with anti- CD9 mAb KMC8.8 for 5 min at 37°C. Before or after stimulation with the Ab, CD9 and F- actin (A--D), Fc␥RIIB/III and F-actin (A), Fc␥RIIB and F-actin (B and D), Fc␥RIII and

F-actin (C), Mac-1, and CD9 (A--D) were by guest on September 29, 2021 stained with Abs and reagents described as in Materials and Methods. Right panels present the superimposed images of the two left panels, respectively. Arrows indicate localization of each molecule detected with each Ab or the reagent at the cell-cell adhesion site. Data are representative of three similar experiments. The Journal of Immunology 3265

References 21. Masellis-Smith, A., G. S. Jensen, J. G. Seehafer, J. R. Slupsky, and A. R. Shaw. 1990. Anti-CD9 monoclonal antibodies induce homotypic adhesion of pre-B cell 1. Ravetch, J., and J.-P. Kinet. 1991. Fc receptors. Annu. Rev. Immunol. 9:457. lines by a novel mechanism. J. Immunol. 144:1607. 2. Gessner, J., H. Hiken, A. Tamm, and R. E. Schmidt. 1998. The IgG Fc receptor 22. Hadjiargyrou, M., and P. H. Patterson. 1995. An anti-CD9 monoclonal antibody family. Ann. Hematol. 76:231. promotes adhesion and induces proliferation of Schwann cells in vitro. J. Neu- 3. Deo, Y., M., F. Robert, R. F. Graziano, R. Repp, and Jan, G. J. van de Winkel. rosci. 15:574. 1997. Clinical significance of IgG Fc receptors and Fc␥R-directed immunother- 23. Ikeyama, S., M. Koyama, M. Yamaoko, R. Sasada, and M. Miyake. 1993. Sup- apies. Immunol. Today 18:127. pression of cell motility and metastasis by transfection with human motility- 4. Indik, Z., K., J. G. Park, S, Hunter, and A. D. Schreiber. 1995. The molecular related protein (MRP-1/CD9) DNA. J. Exp. Med. 177:1231. dissection of Fc␥ receptor mediated phagocytosis. Blood 86:4389. 5. Strzelecka, A., K. Kwiatkowska, and A. Sobota. 1997. Tyrosine phosphorylation 24. Oritani, K., X. Wu, K. Medina, J. Hudson, K. Miyake, J. M. Gimble, and Fc␥ receptor-mediated phagocytosis. FEBS Lett. 400:11. S. A. Burstein, and P. W. Kincade. 1996. Antibody ligation of CD9 modifies 6. Rose, D., M., B. W. Winston, E. D. Chan, D. W. Riches, P. Gerwins, production of myeloid cells in long-term cultures. Blood 87:2252. G. L. Johnson, and P. M. Henson. 1997. Fc␥ receptor cross-linking activates p42, 25. Azuma, Y., K. Kaji, R. Katogi, S. Takeshita, and A. Kudo. 2000. Tumor necrosis ␣ p38, and JNK/SAPK mitogen-activated protein kinases in murine macrophages: factor- induces differentiation of and bone resorption by osteoclasts. J. Biol. role for p42MAPK in Fc␥ receptor-stimulated TNF-␣ synthesis. J. Immunol. Chem. 275:4858. 158:3433. 26. Takeshita, S., K. Kaji, and A. Kudo. 2000. Identification and characterization of 7. Daeron, M. 1997. Fc receptor biology. Annu. Rev. Immunol. 15:203. the new osteoclast progenitor with macrophage phenotypes being able to differ- 8. Wright, M., D., and M. G. Tomlinson. 1994. The ins and outs of the transmem- entiate into mature osteoclasts. J. Bone Miner. Res. 15:1477. brane 4 superfamily. Immunol. Today 15:588. 27. Kawaguchi, J., S. Takeshita, T. Kashima, T. Imai, R.Machinami, A. Kudo. 1999. 9. Hemler, M., E., B. A. Mannion, and F. Berditchevski. 1996. Association of Expression and function of the splice variant of the human cadherin-11 gene in TM4SF proteins with integrins: relevance to cancer. Biochim. Biophys. Acta subordination to intact cadherin-11. J. Bone Miner. Res. 14:764. 1287:67. 28. Sheets, E., D., D. Holowka, B. Baird. 1999. Critical role for cholesterol in Lyn- 10. Gachet, C., A. Astier, H. de la Salle, C. de la Salle, W. H. Fridman, mediated tyrosine phosphorylation of Fc⑀RI and their association with detergent- J. P. Cazenave, D. Hanau, and J. L. Teillaud. 1995. Release of Fc␥RIIa2 by resistant membranes. J. Cell Biol. 17:877.

activated platelets and inhibition of anti-CD9-mediated platelet aggregation by 29. Viola, A., S. Schroeder, Y. Sakakibara, and A. Lanzavecchia. 1999. T lympho- Downloaded from recombinant Fc␥RIIa2. Blood 85:698. cyte costimulation mediated by reorganization of membrane microdomains. Sci- 11. Slupsky, J., R., J. C. Cawley, C. Kaplan, and M. Zuzel. 1997. Analysis of CD9, ence 283:680. CD32 and p67 signalling: use of degranulated platelets indicates direct involve- 30. Takai, T., M. Li, D. Sylvestre, R. Clynes, and J. V. Ravetch. 1994. FcR ␥ chain ment of CD9 and p67 in integrin activation. Br. J. Haematol. 96:275. deletion results in pleiotrophic effector cell defects. Cell 76:519. 12. Kuroda, K.,Y. Ozaki, R. Qi, N. Asazuma, Y. Yatomi, K. Satoh, S. Nomura, 31. Sutterwala, F., S., G. J. Noel, P. Salgame, and D. M. Mosser. 1998. Reversal of M. Suzuki, and S. Kume. 1995. Fc␥ II receptor-mediated platelet activation in- proinﬂammatory responses by ligating the macrophage Fc␥ receptor type I. duced by anti-CD9 monoclonal antibody opens Ca2ϩ channels which are distinct ϩ J. Exp. Med. 188:217. from those associated with Ca2 store depletion. J. Immunol. 155:4427. ␥

32. Sanchez-Mejorada, G., and C. Rosales. 1998. Fc receptor-mediated mitogen- http://www.jimmunol.org/ 13. Yashiro-Ohtani, Y., X. Y. Zhou, K. Toyo-Oka, X. G. Tai, C. S. Park, activated protein kinase activation in monocytes is independent of Ras. J. Biol. T. Hamaoka, R. Abe, K. Miyake, and H. Fujiwara. 2000. Non-CD28 costimula- Chem. 273:27610. tory molecules present in T cell rafts induce T cell costimulation by enhancing the 33. Sakuma, T., S. Higashiyama, S. Hosoe, S. Hayashi, and N. Taniguchi. 1997. CD9 association of TCR with rafts. J. Immunol. 164:1251. antigen interacts with heparin-binding EGF-like growth factor through its hepa- 14. Horvath, G., V. Serru, D. Clay, M. Billard, C. Boucheix, and E. Rubinstein. 1998. rin-binding domain. J. Biochem. 122:474. CD19 is linked to the integrin-associated tetraspans CD9, CD81, and CD82. 34. Lebel-Binay, S., C. Lagaudriere, D. Fradelizi, and H. Conjeaud. 1995. CD82, J. Biol. Chem. 273:30537. tetra-span-transmembrane protein, is a regulated transducing molecule on U937 15. Tachibana, I., and M. E. Hemler. 1999. Role of transmembrane 4 superfamily monocytic cell line. J. Leukocyte Biol. 57:956. (TM4SF) proteins CD9 and CD81 in muscle cell fusion and myotube mainte- nance. J. Cell Biol.146:893. 35. Ozaki, Y., K. Satoh, K. Kuroda, R. Qi, Y. Yatomi, S. Yanagi, K. Sada, 16. Banerjee, S., A., M. Hadjiargyrou, and P. H. Patterson. 1997. An antibody to the H. Yamamura, M. Yanabu, S. Nomura, et al. 1995. Anti-CD9 monoclonal anti- ␣ ␤ body activates p72syk in human platelets. J. Biol. Chem. 270:15119. tetraspan membrane protein CD9 promotes neurite formation in a partially 3 1 integrin-dependent manner. J. Neurosci. 17:2756. 36. Brown, E., J., J. F. Bohnsack, and H. D. Gresham. 1988. Mechanism of inhibition by guest on September 29, 2021 17. Chen, M. S., K. S. Tung, S. A. Coonrod, Y. Takahashi, D. Bigler, A. Chang, of immunoglobulin G-mediated phagocytosis by monoclonal antibodies that rec- Y. Yamashita, P. W. Kincade, J. C. Herr, and J. M. White. 1999. Role of the ognize the Mac-1 antigen. J. Clin. Invest. 81:365. integrin-associated protein CD9 in binding between sperm ADAM 2 and the egg 37. Graham, I. L., H. D. Gresham, and E. J. Brown. 1989. An immobile subset of ␣ ␤ plasma membrane CD11b/CD18 (Mac-1) is involved in phagocytosis of targets integrin 6 1: implications for murine fertilization. Proc. Natl. Acad. Sci. USA 96:11830. recognized by multiple receptors. J. Immunol. 142:2352. 18. Le Naour, F., E. Rubinstein, C. Jasmin, M. Prenant, and C. Boucheix. 2000. 38. Caron, E., and A. Hall. 1998. Identification of two distinct mechanisms of phago- Severely reduced female fertility in CD9-deficient mice. Science 287:319. cytosis controlled by different Rho GTPases. Science 282:1717. 19. Miyado, K., G. Yamada, S. Yamada, H. Hasuwa, Y. Nakamura, F. Ryu, 39. Nobes, C. D., and A. Hall. 1995. Rho, rac, and cdc42 GTPases regulate the K. Suzuki, K. Kosai, K. Inoue, A. Ogura, et al. 2000. Requirement of CD9 on the assembly of multimolecular focal complexes associated with actin stress fibers, egg plasma membrane for fertilization. Science 287:321. lamellipodia, and filopodia. Cell 81:53. 20. Kaji, K., S. Oda, T. Shikano, T. Ohnuki, Y. Uematsu, J. Sakagami, N. Tada, 40. Seehafer, J.,G., and A. R. Shaw. 1991. Evidence that the signal-initiating mem- S. Miyazaki, and A. Kudo. 2000. The gamete fusion process is defective in eggs brane protein CD9 is associated with small GTP-binding proteins. Biochem. Bio- of CD9-deficient mice. Nat. Genet. 24:279. phys. Res. Commun. 179:401.