Proc. Nati. Acad. Sci. USA Vol. 84, pp. 1384-1388, Mirch 1987 Immunology Crosslinking of surface antigens causes mobilization of intracellular ionized calcium in T (signal transduction/ activation/indo-1/flow cytometry) JEFFREY A. LEDBETTER*, CARL H. JUNEt, LAURA S. GROSMAIRE*, AND PETER S. RABINOVITCHt *Oncogen Corporation, 3005 First Avenue, Seattle, WA 98121; tNaval Medical Research Institute, Bethesda, MD 20814; and tDepartment of Pathology, University of Washington, Seattle, WA 98195 Communicated by George J. Todaro, November 17, 1986 (received for review August 15, 1986)

ABSTRACT Antibodies binding to a large subset of T-cell differentiation antigens, including CD2, CD4, CD5, CD6, CD7, C)8, Tp44, and CDw18, cause an increase in the cytoplasmic calcium concentration ([Ca2+k) after the antigens are crosslinked on the cell surface. Similar crosslinking- induced signals were seen for a subset of mouse differentiation antigens. The various antigens on human T cells differed in the extent ofcrosslinking required for generating the calcium signal, as evidenced by comparisons with monoclonal versus polyclonal second-step antibody. The [Ca2i] increase that occurs after crosslinking represents mobilization of cyto- plasmic calcium since the initial component of the signal is 8 resistant to depletion of extracellular calcium by chelation with EGTA. The [Ca2+] increase is completely inhibited by pre- treatment of cells with pertussis toxin, indicating that a substrate for pertussis toxin regulates the signal transduction. 5 Crosslinking of antigens other than the CD3/T-cell receptor 4~~~~~~~~~~~~~~. I complex did not result in T-cell proliferation. Crosslinking of CD2 and Tp44, but not other antigens, resulted in expression of functional interleukin 2 receptors. Comparisons of three different anti-CD3 antibodies showed that a second calcium FIG. 1. Measurement of [Ca2+], (violet/blue indo-1 ratio) vs. time signal was generated by crosslinking, even when the anti-CD3 in resting peripheral blood T cells after surface antigen crosslinking antibodies were used at optimal concentrations. with a polyclonal goat anti-mouse Ig second-step reagent added in excess at time 5 min. T cells were isolated from periphetal blood mononuclear cells by removal of adherent cells (m9nocytes and B Calcium plays an important role in regulating the responses cells) on a nylon wool column. (A) Antibody 10.2 (anti-CD5, 10 of excitable cells, and levels offree calcium in the cytoplasm ,ug/ml) added at time -30 sec. (B) Antibody G17-2 (anti-CD4, 10 are tightly controlled by calcium membrane channels, calci- ,ug/ml) added at -30 sec. (C) Antibody G3-7 (anti-CD7, 10 Agu/ml) um pumps, and release or uptake of calcium from membrane added at -30 sec. (D) Addition of the second step alone at 5 min. surfaces (for reviews, see refs. 1-3). Many hormones and Calibration of the violet/blue ratio, as described (20), shows that neurotransmitters exert their effects in part by raising the ratios of 1, 2, 3, and 4 correspond to [Ca2+], of 131 nM, 338 nM, 679 calcium concentration in the cytoplasm ([Ca2+]j), thus acti- nM, and 1346 nM, respectively. vating calcium-dependent enzymes that direct the cellular response (1-3). G3-7 (anti-CD7), G1-14 (anti-T200), G1-15 (anti-p220), G17-2 Calcium responses in human T lymphocytes have been (anti-CD4), G3-5 (anti-CD6), 64.1, 38.1, and G19-4 (anti- reported after antibody or antigen binding to the CD3/T-cell CD3), G10-1 (anti-CD8), 60.3 (anti-CDwl8), and 9.3 (anti- receptor complex (4-8) and after antibody binding to the CD2 Tp44) have been described and assigned to the indicated erythrocyte rosette receptor (9, 10). The "antigen-specific" cluster groups by the First and Second International Work- CD3/T-cell receptor and the "alternative pathway" CD2 shops on Leukocyte Differentiation Antigens (14, 15). Mono- systems lead to activation of T cells through the interleukin clonal rat antibodies to murine antigens anti-Thy-1.2 2 (IL-2) autocrine system (11-13). Here we examine the (30-H12), anti-LgplOO, Ly9 (30-C7), anti-Lyt-2 (53-6), anti- relation of surface antigen crosslinking to calcium regulation Lyl (53-7), anti-ThB (53-9), and anti-L3T4 (GK1.5) have been for these and other antigens on T cells. Crosslinking surface described (16, 17). antigens such as CD2, CD3, CD4, CD5, CD6, CD7, CD8, and Monoclonal rat anti-mouse K chain 187.1 (18) was purified Tp44 causes an increase in [Ca2+]i. In the case of CD2 and from culture fluids by affinity chromatography on mouse Tp44, the calcium response after antigen crosslinking is immunoglobulin-Sepharose. Polyclonal goat anti-rat Ig or accompanied by expression of functional IL-2 receptors. goat anti-mouse Ig was used as serum from hyperimmunized goats. AND METHODS EGTA and phorbol 12-myristate 13-acetate (PMA) were MATERIALS from Sigma. Pertussis toxin was from List Biological Labo- Monoclonal Antibodies and Reagents. Monoclonal antibod- ratories (Campbell, CA). Recombinant IL-2 (rIL-2) was from ies against human antigens 9.6 (anti-CD2), 10.2 (anti-CD5), Genzyme (Boston, MA).

The publication costs of this article were defrayed in part by page charge Abbreviations: [Ca2+]j, cytoplasmic calcium concentration; IL-2, payment. This article must therefore be hereby marked "advertisement" interleukin 2; rIL-2, recombinant IL-2; PMA, phorbol 12-myristate in accordance with 18 U.S.C. §1734 solely to indicate this fact. 13-acetate; G , guanine nucleotide-binding protein.

Downloaded by guest on October 1, 2021 1384 Immunology: Ledbetter et al. Proc. Natl. Acad. Sci. USA 84 (1987) 1385

Measurement of [Ca2"Ji. Cytoplasmic free calcium concen- radioactivity was measured in a liquid scintillation counter. tration was measured with the dye indo-1 (19) (Molecular Antibodies used in proliferation assays were dialyzed against Probes, Eugene, OR) and a model 50HH/2150 cell sorter phosphate-buffered saline and sterilized by filtration prior to (Ortho Diagnostic Systems, Westwood, MA) as we have use. Monoclonal antibodies were used at 1 ,g/ml, and described in detail (20). The histograms were analyzed by antibody for crosslinking (187.1) was used at 20 ,g/ml in programs that calculated the mean indo-1 violet/blue fluo- proliferation assays. rescence ratio vs. time. In addition, the percentage of responding cells vs. time was analyzed by programs that first RESULTS determined the value of the indo-1 ratio 2 SD above the mean ratio of control cells and then plotted the percentage of cells Crosslinking of surface antigens is thought to play a role in above this threshold value vs. time. There are 100 data points signal transduction through the membrane. Because cyto- on the x (time) axis on all flow cytometric data. Values above plasmic calcium is a central regulator of cellular responses, 1% in percentage of responding cells are significant. Human we measured the effects of antigen crosslinking on levels of lymphocytes were purified from peripheral blood by centrif- cytoplasmic free calcium using the dye indo-1 and a flow ugation on Ficoll/Hypaque. In many experiments, T cells cytometer. This dye, which is preloaded into the cells, were enriched by removal of adherent cells on nylon wool exhibits a shift in the emission spectrum from blue to violet columns (21). These cell populations contained 85-95% T upon the binding of Ca2+. Fig. 1 shows the effects on [Ca2+], cells and 5-15% natural killer (CD16+) null cells. Experi- after crosslinking of CD5, CD4, and CD7 antigens compared ments with murine T cells used from BALB/c to the second-step polyclonal antibody alone (Fig. 1D). The mice. antibodies were added just before time 0 on the display, Cell Culture. Measurement of proliferation was performed whereas the crosslinking antibody (see arrow) was added in by culturing cells in quadruplicate samples in flat-bottom, excess 5 min later. Crosslinking of these antigens caused 96-well microtiter plates at 5 x 104 cells per well in RPMI cytoplasmic calcium to quickly increase, as evidenced by the medium containing 15% human AB serum (Pel-Freez, Brown increase in the violet/blue ratio. The kinetics ofthe responses Deer, WI). Cell proliferation was measured by uptake of to crosslinking ofthese antigens are similar to that previously [3H]thymidine [6.7 Ci/mmol (1 Ci = 37 GBq); New England shown in response to antibodies CD3 and CD2. Table 1 Nuclear] after a 6-hr pulse. Uptake was measured by har- demonstrates that crosslinking of a variety of other antigens, vesting cells onto glass fiber filters with a cell harvester, and either with polyclonal or monoclonal second-step reagents Table 1. [Ca2+]i responses after antigen crosslinking on T cells Crosslinking % cells Polyclonal second step* Monoclonal second stept Antigen positive [Ca2+],, % responding [Ca2+]1, % responding (antibody) (range) nM cells nM cells Human peripheral blood T cellst None 131 0 131 0 CD2 (9.6) 90-98 610 94 145 15 (35.1) 478 92 138 2.8 CD3 (G19-4) 75-95 665 91 370 79 CD4 (G17-2) 40-60 201 39 141 2.0 CD5 (10.2) 75-90 158 24 172 45 CD6 (G3-5) 75-90 160 16 145 4.5 CD7 (G3-7) 85-95 145 17 141 3.0 CD8 (G10-1) 20-40 ND ND 170 20 Tp44 (9.3) 60-80 261 65 231 52 T200 (G1-14) 100 138 6.0 138 1.5 p220 (G1-15) 60-80 137 2.5 143 7.0 CDw18 (60.3) >95 144 11 133 0 BALB/c thymocytes§ None 131 0 Thy-1.2 (30-H12) 1126 98 Lyl (53-7) 170 22 Ly9 (LgplOO,30-C7) 214 40 ThB (53-9) 145 8.0 L3T4 (GK1.5) 137 1.5 Lyt-2 (53-6) 131 0 Prior to crosslinking, none ofthe antibodies caused a change in [Ca2+]j except for anti-CD3. The peak mean [Ca2+], ofthe total population and maximum % responding cells following the indicated treatment are shown. ND, not determined. *A goat anti-mouse Ig polyclonal second step was used in excess (25 Ml) after 5 min to induce the crosslinking response. tA rat anti-mouse K chain monoclonal antibody (187.1) was added in excess (50 ,g/ml) after 5 min to induce the crosslinking response. tT cells were isolated from peripheral blood by centrifugation on Ficoll followed by removal ofadherent cells on nylon wool columns. The range of positive cells for each antigen was not measured in the experiments shown but represents the expected normal range for expression of each antigen. §A goat anti-rat Ig polyclonal second step was used in excess (25 Al) after 5 min to induce the crosslinking response. Downloaded by guest on October 1, 2021 1386 Immunology: Ledbetter et al. Proc. Natl. Acad. Sci. USA 84 (1987) (as described below), produces a [Ca2+], signal. Of the pertussis toxin to alter the [Ca2+]i response. Pertussis toxin purified antibodies listed in Table 1, only anti-CD3 produces ribosylates and thereby inactivates Gi and other guanine a detectable increase in [Ca2+], when used without crosslink- nucleotide-binding (G proteins) that regulate signal ing. In at least one case, that of CD5, production of a transduction systems (22-29). Fig. 2 shows that the Tp44 "preformed complex" of antibody and a second-step cross- crosslinking signal (Fig. 2C) and CD8 crosslinking signal (Fig. linker before exposure to cells can also produce an increase 2F) are completely eliminated by a 1-hr pretreatment of the in [Ca2+]i (unpublished data). The data shown in Fig. 1 and cells with 10 ,ug of pertussis toxin per ml at 37°C. Thus, a G Table 1 are representative ofat least two or more experiments protein that is inactivated by pertussis toxin regulates the for each antigen. The [Ca2+]i responses after crosslinking signal transduction system stimulated by surface antigen were consistently seen. For several antigens such as CD7 and crosslinking. The data also suggest that the release ofinternal CD2 that are expressed on T cells and natural killer cells, we Ca2+ is a prerequisite for the entry of external Ca2", since do not yet know whether preferential calcium responses are pertussis toxin inhibits the Ca2+ response completely, where- occurring in lymphocyte subpopulations. as EGTA inhibits only a component of that response. This is Increase in [Ca2`1 from Mobilization of Cytoplasmic Sources. consistent with a recent model for cellular calcium entry To determine whether the Ca2+ that increases in the cytoplasm proposed by Putney (30). after crosslinking is ofan intracellular or an extracellular origin, Extent of Crosslinking Controls Magnitude of [Ca2+]J Re- experiments were performed in the presence of EGTA to sponse. For some surface antigens, the extent of crosslinking chelate extracellular Ca2". In the presence of EGTA, extrarel- determined the magnitude of the increase in [Ca2+]i. This is lular ionized calcium is =15 nM (20), whereas intracellular shown for the case of CD2 (erythrocyte rosette receptor) in ionized calcium in resting T cells is 131 nM. Therefore, an Fig. 3 by comparing the use of either a polyclonal goat increase in [Ca2+]i after stimulation cannot result from a calcium anti-mouse Ig second step (Fig. 3A) or a monoclonal rat influx when EGTA is present. Fig. 2 shows the response in anti-mouse K chain second step (Fig. 3B). For CD2, the of the surface or polyclonal second step was required to obtain a response in [Ca2+]i to crosslinking Tp44 antigen (Fig. 2A) all CD2+ cells. Similarly, Table 1 demonstrates that for of the CD8 antigen (Fig. 2D). In the presence of EGTA, there antigens such as CD4, CD6, CD7, the common leukocyte was still a significant increase in [Ca2]1i for both of these antigen T200, and CDw18, the polyclonal second step gen- antigens (Fig. 2 B and E). We found this same result (an erated a greater increase in [Ca2+]i among a larger fraction of EGTA-resistant increase in [Ca2+]j) for the crosslinking re- cells than did the monoclonal second step. In the case of sponse for all of the other antigens examined. Thus, although CDw18, the latter reagent produced no visible effect at all. the magnitude of the [Ca2+] response is decreased by EGTA, Only the CD5 antigen, which is highly mobile on the cell indicating a contribution from extracellular Ca2+, the initial surface, showed a larger increase in [Ca2+]i after crosslinking [Ca2]1i response is from a cytoplasmic source of Ca2'. with the monoclonal second step compared to the polyclonal To further characterize the crosslinking pathway for mo- second step (Table 1). The magnitude of the [Ca2 ]i response bilization of cytoplasmic Ca2+, we tested the ability of induced by crosslinking ofantibodies T200, p220, and CDw18 was much smaller than that of the other antibodies tested. Several of the antigens recognized by these antibodies are known to be relatively rigidly bound in the cell's membrane (31). Results of crosslinking antigens on murine thymocytes are also shown in Table 1. Crosslinking the Thy-1 antigen, Lyl antigen, and Ly9 antigen generated large increases in [Ca2+], whereas crosslinking of ThB, L3T4, and Lyt-2 generated smaller [Ca2+], signals or none at all. It is possible that factors such as antibody affinity and avidity, valency of epitopes on the antigens, and the intrinsic ability of antigens to transduce signals are all contributing to the magnitude of the response and the frequency of responding cells. T-Cell Activation by Crosslinking Surface Antigens. To investigate the effects of crosslinking on T-cell activation, we

FIG. 2. Sensitivity of crosslinking [Ca2+]i responses (violet/blue indo-1 ratio) to inhibition by EGTA and pertussis toxin. (A-C) Responses of peripheral blood T cells (prepared as described in the legend to Fig. 1) to crosslinking ofTp44 using antibody 9.3 (10 jug/ml) added at time -3 min. (D-F) Responses of peripheral blood T cells 0 5 to crosslinking ofCD8 using antibody G10-1 (10 ,ug/ml) added at time -3 min. In each panel, 50 ,g ofthe monoclonal second-step antibody 187.1 (rat anti-mouse K chain) per ml was added at time 2 min, indicated by the arrow. (A and D) [Ca2+]i response in the presence of extracellular Ca2 . (B and E) [Ca2 ]i response in the presence of FIG. 3. Crosslinking requirement for [Ca2+]i increases after 10 mM EGTA to chelate extracellular Ca2+ to levels below cyto- antibody binding to CD2 (erythrocyte rosette receptor) on peripheral plasmic [Ca2+]i. (C and F) [Ca2+]i response after a 1-hr pretreatment blood T cells (purified as described in the legend to Fig. 1). Anti-CD2 of the cells with 10 ,ug of pertussis toxin per ml at 37°C. The signals antibody 9.6 (10 ,ug/ml) was added at time -30 sec in A and B. The shown in D-F were gated to display only CD8+ cells using R- second-step antibody (25 ul) was added at 5 min, represented by the phycoerythrin-conjugated anti-CD8 antibody and flow cytometry arrow. (A) Polyclonal goat anti-mouse Ig second step added in with dual laser excitation as described (20) so that the response ofthis excess. (B) Monoclonal rat anti-mouse K antibody (187.1) added in subpopulation could be more easily visualized. excess (50 ,ug/ml). Downloaded by guest on October 1, 2021 Immunology: Ledbetter et al. Proc. Natl. Acad. Sci. USA 84 (1987) 1387

measured proliferation by uptake of [3H]thymidine in re- sponse to antigen crosslinking. By themselves, none of the 0 2._ Anti-lg A increases after antigen crosslinking resulted in direct "T 2.10 t [Ca2+]I Anti-CD3 300 c stimulation of proliferation. However, in the presence of U) rIL-2, crosslinking of either the CD2 or the Tp44 antigen, but Q .f j 5 *D 1.- 200 co not others, resulted in proliferation (Table 2). These results , . 0 0 - 0 ,:: were repeated in four other experiments. This implies that -, X. o crosslinking CD2 or Tp44 causes expression of high-affinity .0 IL-2 receptors. We also found IL-2 receptors expressed after 1100 CD2 or Tp44 crosslinking when measured by immuno- fluorescence using anti-CD25 (IL-2 receptor) antibodies and s 8o0 Anti-Ig B a flow cytometer (data not shown). For CD2, we found that a) -Anti-CD3 $ a polyclonal second step was more active in inducing IL-2 receptors than the monoclonal 187.1 antibody (data not 0 shown). Without crosslinking, anti-CD2 and anti-Tp44 did QUo k0o r-C I not induce expression of IL-2 receptors. In the presence of the phorbol ester PMA, the [Ca2+] signals from crosslinking 0 2 4 6 8 10 CD2 and Tp44 antigens also resulted in a synergistic prolif- Time, mmn eration response (Table 2). Thus, some of the crosslinking- induced [Ca2+]i responses, particularly in the case ofCD2 and FIG. 4. Responses in [Ca2+]i of peripheral blood mononuclear Tp44, may be involved in lymphocyte proliferation. Both of cells after direct binding of anti-CD3 monoclonal antibodies 64.1 these antigens have been implicated as receptors for func- (IgG2a) (---), G19-4 (IgG1) (-), or 38.1 (IgM) (....) at 20 ,ug/ml each, tional signals from other studies showing direct activation of added at time 2 min. At time 6 min, 100 ug of a second-step rat T cells using combinations of antibodies to CD2 (12, 13) and anti-mouse K chain (187.1) per ml was added to crosslink the CD3 synergistic proliferation responses to anti-Tp44 with PMA or antigen. (A) Mean [Ca2+], of the population. (B) Percentage of anti-Tp44 with anti-CD3 (32, 33). responding cells, determined as the proportion of cells with [Ca2+], Distinction Between [Ca2+]i Response After Anti-CD3 Bind- 2 SD above the mean [Ca2+], of the resting population (20). ing and CD3 Crosslinking. Antibody binding to the CD3/T- cell receptor complex can activate T cells. To explore antibodies can cause [Ca2+], to increase (5, 33), indicating whether CD3 crosslinking is involved in the CD3 activation that crosslinking of CD3 may not be required to mobilize pathway, we tested the effect of crosslinking on the [Ca2+] calcium in T cells and that CD3 crosslinking causes a second signal for three anti-CD3 antibodies. These three antibodies [Ca2+], signal through a different mechanism. differed in their ability to directly cause a [Ca2+]j increase (Fig. 4) even though they were each used at a saturating concentration that generated the maximal [Ca2+], signal. The DISCUSSION 64.1 antibody was least effective, whereas the 38.1 and G19-4 We have shown that cytoplasmic free calcium in T cells antibodies were more effective, although the G19-4 induced increases after crosslinking certain antigens on the cell a [Ca2+], signal that was more sustained. After crosslinking, surface. The effect was seen for a wide variety, but not all, however, each of the antibodies generated a second increase of surface antigens. The calcium increase appears to include in [Ca2+],, and the responses induced by crosslinking were significant mobilization of intracellular calcium stores. This almost equal in the percentage of responding cells (Fig. 4B). interpretation is based on the resistance of the initial [Ca2 ] Thus, the 64.1 and 38.1 anti-CD3 antibodies were very increase after crosslinking to depletion of extracellular Ca2+ different in direct stimulation of [Ca2+]i but were almost by addition of EGTA. In T lymphocytes and other cells, Ca2+ identical in the [Ca2+]i response after crosslinking (Fig. 4A). is released from cytoplasmic stores after formation ofinositol The differences between these three anti-CD3 antibodies in trisphosphate caused by receptor-mediated activation of their ability to directly cause increases in [Ca2+], may be due phospholipase C (3, 4, 7). Thus, it is possible that the to differences in affinity or differences in the epitopes mobilization of cytoplasmic Ca2+ after surface antigen cross- recognized. Thus, the [Ca2+]i signals delivered through CD3 linking that we have described is mediated by formation of by antibody binding before or after crosslinking may not inositol trisphosphate. Another potential mechanism for involve the identical signaling system. This is supported by mobilization of cytoplasmic Ca2+ regulated by guanine nu- previous reports that monovalent Fab fragments of anti-CD3 cleotides was recently described (34), although this latter

Table 2. Effect of antigen crosslinking on T-cell proliferation Additional signal* 187.1 PMA PMA PMA Antigen + + + + 187.1 (antibody) None 187.1 rIL-2 rIL-2 PMA rIL-2 187.1 + rIL-2 None 31 71 743 775 221 3,341 400 3,051 Anti-CD5 (10.2) 30 27 644 829 202 3,075 133 3,509 Anti-CD2 (9.6) 42 21 533 1929 98 3,099 16,627 18,543 Anti-CD4 (G17-2) 26 26 368 494 89 2,820 94 3,681 Anti-CD7 (G3-7) 64 21 678 717 688 3,277 92 3,804 Anti-CD8 (G10-1) 30 25 764 652 610 3,823 161 3,656 Anti-Tp44 (9.3) 26 240 1309 9208 9869 11,129 14,195 23,575 The effect of antigen crosslinking on T-cell proliferation was measured by uptake of [3H]thymidine for the last 6 hr of a 3-day experiment using peripheral blood mononuclear cells isolated by centrifugation on Ficoll. Uptake was measured in quadruplicate. Mean cpm are shown and the SEM were <15% of the mean at each point. Antibodies were used at 1 gg/ml. Data for one of four experiments are shown. *Monoclonal antibody 187.1 (anti-mouse K chain) was used at 20 gg/ml, PMA concentration was 1 ng/ml, and rIL-2 was 100 units/ml. Downloaded by guest on October 1, 2021 1388 Immunology: Ledbetter et al. Proc. Natl. Acad. Sci. USA 84 (1987)

mechanism has not yet been shown to be activated by We thank Mr. S. Chong Kim for expert technical assistance and receptor stimulation. Drs. Marilyn Parsons and Ed Clark for helpful discussions. This The increase in [Ca2+]i after surface antigen crosslinking is work was supported by Oncogen Corporation, by Public Health completely inhibited by pretreatment ofT cells with pertussis Service Grant AG01751, and by Naval Medical Research and toxin. In addition to Gi that regulates activation of adenylate Development Command, Research Task M0095.001.0045. cyclase (22, 23), pertussis toxin also ribosylates and thus 1. Gill, D. L. (1985) Adv. Cyclic Nucleotide Protein Phosphorylation Res. inactivates G proteins that regulate phospholipase C activa- 19, 195-212. tion in neutrophils (24-26). Other G proteins inactivated by 2. Exton, J. H. (1985) J. Clin. Invest. 75, 1753-1757. pertussis toxin include transducin that regulates a cGMP 3. Berridge, M. J. & Irvine, R. F. (1984) Nature (London) 312, 315-321. 4. Imboden, J. B. & Stobo, J. D. (1985) J. Exp. Med. 161, 446-456. phosphodiesterase (27, 28) and Go that couples signal trans- 5. Oettgen, H. C., Terhorst, C., Cantley, L. C. & Rosoff, P. M. (1985) Cell duction through brain cholinergic receptors (29). It has been 40, 583-590. reported that a G protein inactivated by ribosylation with 6. O'Flynn, D., Zander, E. D., Lamb, J. R., Beverley, P. C., Wallace, cholera toxin regulates inositol trisphosphate formation in a D. L., Tatham, P. E., Tax, W. J. & Linch, D. C. (1985) Eur. J. Immunol. 15, 7-11. T- cell line (35). Thus, in T cells it is likely that signal 7. Imboden, J. B., Weiss, A. & Stobo, J. D. (1985) J. Immunol. 134, transduction resulting in Ca2+ mobilization is regulated by G 663-665. proteins. 8. Weiss, A., Imboden, J., Shobach, D. & Stobo, J. (1984) Proc. Natl. The increase after crosslinking of CD2 and Tp44 Acad. Sci. USA 81, 4169-4173. [Ca2+], 9. Weiss, M. J., Daley, J. F., Hodgdon, J. C. & Reinherz, E. L. (1984) appears to be associated with expression of functional IL-2 Proc. Natl. Acad. Sci. USA 81, 6836-6840. receptors. Both of these antigens have been implicated as 10. June, C. H., Ledbetter, J. A., Rabinovitch, P. S., Martin, P. J., Beatty, receptors for signals in T-cell activation (12, 13, 32, 33). The P. G. & Hansen, J. A. (1986) J. Clin. Invest. 77, 1224-1232. signal after crosslinking other antigens, such as CD5, 11. Meuer, S. C., Hussey, R. E., Cantrell, D. A., Hodgdon, J. C., Schloss- [Ca2+], man, S. F., Smith, K. A. & Reinherz, E. L. (1984) Proc. Natl. Acad. CD8, CD4, etc., does not appear to be related in the same way Sci. USA 81, 1509-1513. to T-cell activation. This indicates that an increase in [Ca2+] 12. Meuer, S. C., Hussey, R. E., Falbi, M., Fox, D. A., Acuto, O., by itself is not sufficient for T-cell activation. It is possible, Fitzgerald, K. A., Hodgdon, R. C., Protentis, J. P., Schlossman, S. F. however, that the signals generated by crosslinking these & Reinherz, E. L. (1984) Cell 36, 897-906. 13. Brottier, P. L., Boumsell, L., Gelin, C. & Bernard, A. (1985) J. latter antigens are associated with other functional responses Immunol. 135, 1624-1631. of T cells. 14. Hansen, J. A., Martin, P. J., Beatty, P. G., Clark, E. A. & Ledbetter, Certain antibodies to the Thy-1 antigen on murine T cells J. A. (1984) in Leukocyte Typing, eds. Bernard, A., Boumsell, L., T-cell clones Dausset, J., Milstein, C. & Schlossman, S. (Springer, New York), pp. are able to stimulate lymphokine production by 195-212. (36). Recent studies with anti-Thy-1 antibodies showed that 15. Haynes, B. F. (1986) in Leukocyte Typing II, eds. Reinherz, E. L., crosslinking caused an increase in [Ca2+]j and led to activa- Haynes, B. F., Nadler, L. M. & Bernstein, I. D. (Springer, New York), tion of murine T cells in the presence of phorbol esters (37). Vol. 1, pp. 3-30. 16. Ledbetter, J. A. & Herzenberg, L. A. (1979) Immunol. Rev. 47, In these studies the [Ca2+]j response was thought to be unique 362-389. to Thy-i. Our results with crosslinking antigens on human 17. Dialynas, D. P., Quan, Z. S., Wall, K. A., Pierres, A., Quintans, J., and mouse cells, however, using the sensitive indo-1 system Loken, M. R., Pierres, M. & Fitch, F. W. (1983) J. Immunol. 131, 2445-2452. with a flow cytometer, indicate that the [Ca2+]i response to 18. Yelton, D. E., Desaymard, C. & Scharff, M. D. (1981) Hybridoma 1, crosslinking is not unique to the Thy-1 antigen. The phos- 5-11. phatidylinositol that is known to be associated with the 19. Grynkiewicz, G. M., Poenie, M. & Tsien, R. Y. (1985) J. Biol. Chem. 260, 3440-3450. carboxyl terminus of Thy-i (38) therefore may not be re- 20. Rabinovitch, P. S., June, C. H., Grossman, A. & Ledbetter, J. A. (1986) quired for generating the [Ca2+]i response. However, the J. Immunol. 137, 952-961. association of phosphatidylinositol with a surface antigen 21. Julius, H. H., Simpson, E. & Herzenberg, L. (1973) Eur. J. Immunol. 3, 645-649. may represent only one mechanism for altering phosphati- 22. Katada, T. & Ui, M. (1982) Proc. Natl. Acad. Sci. USA 79, 3129-3133. dylinositol metabolism. 23. Bokoch, G. M., Katada, T., Northrup, J. K., Hewlett, E. L. & Gilman, Crosslinking of many surface antigens transmits a signal A. G. (1983) J. Biol. Chem. 258, 2072-2075. that of and 24. Bokoch, G. M. & Gilman, A. G. (1984) Cell 39, 301-308. causes polymerization cytoskeletal components 25. Okajima, F. & Ui, M. (1984) J. Biol. Chem. 259, 13863-13871. their association with the membrane at the site of the antigen 26. Smith, C. D., Cox, C. C. & Snyderman, R. (1986) Science 232, 97-100. patches and caps (for review, see ref. 31). Of the antigens 27. Abood, M. E., Hurley, J. B., Pappone, M.-C., Bourne, H. R. & Stryer, studied this has been demonstrated for the L. (1982) J. Biol. Chem. 257, 10540-10543. here, directly only 28. Van Dop, C., Yamanaka, G., Steinberg, F., Sekura, R. D., Manclark, Thy-1 antigen (39, 40). However, this is a property of C. R., Stryer, L. & Bourne, H. R. (1984) J. Biol. Chem. 259, 23-26. crosslinking surface antigens such as H-2, Ig, Fc receptors, 29. Florio, V. A. & Sternweis, P. C. (1985) J. Biol. Chem. 260, 3477-3483. and other 30. Putney, J. W., Jr. (1986) Cell Calcium 7, 1-12. and even glycolipids in lymphocytes cell types (31). 31. Braun, J. & Unanue, E. R. (1983) Semin. Hematol. 20, 322-333. Calcium has been proposed as the transmembrane signal 32. Hara, T., Fu, S. M. & Hansen, J. A. (1985) J. Exp. Med. 161, generated by crosslinking for redistribution and polymeriza- 1513-1524. 33. Ledbetter, J. A., June, C. H., Martin, P. J., Spooner, C. E., Hansen, tion of cytoskeletal proteins (for review, see ref. 31). This J. A. & Meier, K. E. (1986) J. Immunol. 136, 3945-3952. concept is supported by the recent evidence that diacylglyc- 34. Gill, D. L., Ueda, T., Chueh, S.-H. & Noel, M. W. (1986) Nature erol, which is also produced during hydrolysis of phospha- (London) 320, 461-464. 35. Imboden, J. B., Shoback, D. M., Pattison, G. & Stobo, J. D. (1986) tidylinositol 4,5-bisphosphate to form inositol trisphosphate, Proc. Natl. Acad. Sci. USA 83, 5673-5677. is required for association of a-actinin with model mem- 36. MacDonald, H. R., Bron, C., Rousseaux, M., Horvath, C. & Cerottini, and forms of in the J. C. (1985) Eur. J. Immunol. 15, 495-501. branes part a-actinin/actin complexes 37. Kroczek, R. A., Gunter, K. C., Seligmann, B. & Shevach, E. M. (1986) cytoskeleton of activated platelets (41). Furthermore, the J. Immunol. 136, 4379-4384. unpolymerized form of actin interacts with phosphatidylino- 38. Tse, A. G. D., Barclay, A. N., Watts, A. & Williams, A. F. (1985) Science 230, 1003-1008. sitol 4,5-bisphosphate, suggesting a relationship between 39. Hoessli, D., Rungger-Brandle, E., Jockusch, B. M. & Gabbiani, G. microfilament formation and phosphatidylinositol metabo- (1980) J. Cell Biol. 84, 305-314. lism after receptor-ligand crosslinking (42). Thus, it is pos- 40. Laub, F., Kaplan, M. & Gitler, C. (1981) FEBS Lett. 124, 35-38. 41. Burn, P., Rotman, A., Meyer, R. K. & Burger, M. M. (1985) Nature sible that the [Ca2+]I increase after crosslinking certain T-cell (London) 314, 469-472. antigens is associated with cytoskeletal responses. 42. Lassing, I. & Lindberg, U. (1985) Nature (London) 314, 472-474. Downloaded by guest on October 1, 2021