Signaling from LFA-1 contributes signal transduction through CD2 alternative pathway in T cell activation

CELLULAR

142, 145-158 (1992)

IMMUNOLOGY

Signaling from LFA-1 Contributes Signal Transduction Alternative Pathway in T Cell Activation’

through CD2

AKIRA YAMADA,**~ TAKAKO KANEYUKI,* YOSHIHIROTORIMOTO,~ JOHNF.DALEY,~CATHERINE M. PRADO,? ANDMITCHEL M. YOKOYAMA* *Department of Immunology, Kurume University School of Medicine, Kurume 830, Japan; and the TDivision of Tumor Immunology, Dana-Farber Cancer Institute. Harvard Medical School, Boston, Massachusetts 02115 Received January 28, 1992; accepted February 29, 1992

LFA- 1, a member of the integrin family of molecules, is involved in mediating cellular adhesion in all phases of the immune response, playing a role in the interaction of helper T cells as well as in killing of target cells by both cytotoxic T cells and natural killer cells. We have developed a monoclonal antibody, anti-HVS6B6, which recognizes a functionally unique epitope of the LFAI molecule. Although this mAb itself was not mitogenic against T cells, it induced a strong proliferative response when added to T cells with submitogenic concentrations of anti-CD2 (antiT 112 and anti-T I 13) mAbs. In contrast, other anti-LFA- 1 mAbs (CD 11a and CD1 8) suppressed this anti-CD2 mAb-induced T cell proliferation. Kinetic studies showed that anti-HVS6B6 acts on an early event in CD2-mediated T cell activation. Although Tl I,-epitope expression induced by anti-T1 l2 mAb was not affected by treatment of cells with anti-HVS6B6, both Ca2+ influx and phosphatidylinositol turnover induced by anti-CD2 mAbs were markedly enhanced by the pretreatment of T cells with anti-HVS6B6 mAb. These results indicate that the LFA-1 mediating signal contributes to a very early phase of signal transduction during CDZ-mediated T cell activation. 0 1992 Academic

Press. Inc.

INTRODUCTION The lymphocyte function-associated antigen- 1 (LFA- 1) (CD 11a/CD 18) is a member of the integrin family of molecules, which includes Mac-l, pl50/90, the VLA/extracellular matrix protein receptors, as well as platelet glycoprotein IIb/IIIa (1). LFA-1 is found on most leukocytes and plays an important role in the interaction of helper T cells, killing of target cells by both cytotoxic T lymphocytes and natural killer cells, and in adhesion of leukocytes to high endothelial venules (HEV) (2-4). Both intercellular adhesion molecule 1 (ICAM-1) and ICAM- have been identified as natural ligands for LFA- 1 (5, 6). Recently, it has been suggestedthat LFA-I may participate in T cell activation via the CD3/T cell receptor pathway as well asthe CD2 alternative pathway (7-l 1). Thus, I This work was supported in part of a grant from the Ministry of Science, Education, and Culture of Japan and a grant from the Fukuoka Cancer Society. ’ To whom correspondence and reprint requests should be addressed. 145 0008-8749/92 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

146

YAMADA

ET AL.

anti-LFA- 1, when immobilized in solid-phase, has been shown to induce a mitogenic signal in conjunction with submitogenic dosesof either anti-CD3 or PMA (8). Noesel et al. (9) showed that mAbs against the LFA-I fi subunit generally inhibited, whereas mAbs against the LFA-1 a subunit enhanced proliferation of T cells following stimulation with immobilized anti-CD3. Anti-LFA-1 mAbs have also been reported to inhibit or enhance CD2-mediated T cell proliferation ( 11). Thus, the LFA- 1 molecule appearsto play a role in both the CD3 and CD2 pathways of T lymphocyte activation. Although the role of LFA-1 on T cell activation has been reported extensively, the mechanisms of LFA- 1 mediating up- or down-regulation of T cell activation are still unclear. In the present study, we describe the mAb anti-HVS6B6 which recognizes a functionally unique epitope of the LFA- 1 molecule. This antibody has a comitogenic effect on T cells when used in combination with anti-CD2 mAbs (anti-T 1l2 and anti-T 113). Moreover, we have demonstrated that anti-HVS6B6 affects transmembrane signal transduction in T cells via the CD2 pathway as measured by Ca2+ influx and phosphatidylinositol turnover. MATERIALS AND METHODS Preparation of cells. Human PBMC were isolated from heparinized blood of healthy volunteer donors by Ficoll-Paque (Pharmacia Fine Chemicals, Piscatway, NJ) density gradient centrifugation. PBMC were further separatedinto E-rosette positive (E+) and negative (E-) populations with SRBC as previously described ( 12). E+ cells were used as T cells. These purified T cells contained ~97% CD2-positive cells and ~2% CD1 lbpositive cells. Plastic adherent monocytes were prepared from E- cells. In some experiments, we used highly purified T cells which contained ~0.5% of nonspecific esterase-positivemonocytes. Preparation of the highly purified T cells is as previously described (35). Antibodies. The mAb anti-HVS6B6 was made in the following manner. Spleen cells from a BALB/c mouse immunized with Herpes virus simiri-transformed Callithrix jacchus (common marmoset) T cells were fused with the murine myeloma cell line NS- 1. The mAb anti-HVS6B6 was screenedfor by utilizing assaysof T cell proliferation following stimulation of T lymphocytes by the mitogenic combination of two antiCD2 mAbs (anti-T1 l2 and anti-T 113). Other mAbs, anti-T 112 (CD2), anti-T 113 (CD2), anti-T4 (CD4), anti-8F2 (VLA4, CD49d), anti-lF7 (CD26), anti-S6Fl (LFA-l), anti2F12 (LFA- 1 (Ychain) and anti- 1OF12 (LFA- 1 p chain) were also used in the present study. Their production and characterization has been described elsewhere (2, 1318). FITC-conjugated F(ab’)2 goat anti-mouse Ig was obtained from Tago (Burlingame, CA). Flow cytometric analysis of lymphocyte populations. Cells were stained by indirect immunofluorescence with FITC-conjugated F(ab’)2-specific goat anti-mouse Ig and analyzed using the EPICS V cell sorter (Coulter, Hialeah, FL). Background fluorescence reactivity was determined with control ascites obtained from mice injected with a nonsecreting hybridoma. For staining, all mAbs were used in Ab excessat dilutions of 1:400 to 1:500. Prior to staining, all diluted ascites were centrifuged at 100,OOOg for 40 min to remove aggregatedmaterials. Immunoprecipitation assays.Cells (1 X 107)were washed twice with PBS and then labeled with 37 MBq (1 mCi) of [1251]NaI(Amersham, Arlington Heights, IL) by

ROLE

OF LFA-I

ON CD2 SIGNAL

TRANSDUCTION

147

lactoperoxidase-catalyzed iodination (19). After iodination, cells were washed three times and then lysed in 1 ml of lysis buffer (50 mM Tris-HCl (pH 7.4) 150 mM NaCl, 1%NP-40, 1 mA4 EDTA, 1 mM PMSF, 10 wg/ml aprotinin (Sigma, St. Louis, MO)) on ice for 30 min. The cell lysate was centrifuged at 10,000 rpm for 25 min at 4°C and further precleared overnight at 4°C with formalin-fixed S. aureus Cowan I organisms (Sigma). In sequential immunoprecipitation experiments, the cell lysate was further precleared four times with mAb-coupled protein A-Affi-Gel (Bio-Rad, Richmond, CA). The precleared lysate was incubated with mAb at 4°C for 6 hr and then precipitated with Protein A-coupled Affi-Gel. Precipitates were washed, once with lysis buffer containing 0.5 it4 NaCl, once with lysis buffer containing 0.5% deoxycholate, and once with lysis buffer alone. Immunoprecipitates were solubilized with SDS-PAGE sample buffer and analyzed by 6% SDS-PAGE under reducing conditions. Cell culture andproliferation assay. T cells were suspendedin RPM1 1640 medium (GIBCO, Gland Island, NY) supplemented with 10%human serum (from AB-positive donors, heat inactivated), 25 mA4 Hepes, and 50 pug/mlgentamicin, and cultured in 96-well U-bottom culture plates at a concentration of 1 X lo5 cells/O.2 ml/well in 5% COz-humidified atmosphere at 37°C. To assessthe anti-CD2-induced proliferative response, T cells were cultured with a 1:200- 1:1000 dilution of anti-T 112 and antiTI l3 mAb containing ascites for 3 days. Before cell harvest, cells were pulse-labeled with 37 KBq/well of [3H]methylthymidine for 12 hr. 13H]TdR incorporation was measured in a scintillation counter. Measurement of cytoplasmic Ca2+ion level. Calcium ion sensitive dye (Indo- 1 AM) loaded cells were used for CaZf influx studies. Measurement of Indo-l fluorescence by flow cytometry is described elsewhere (20). Briefly, T cells were incubated with Indo-l AM (Calbiochem, San Diego, CA) at a concentration of 10 pg/ml at 37°C for 20 min. After having been washed, cells were suspendedin MEM supplemented with 2.5% newborn calf serum. The measurement of Indo-l fluorescence of individual cells was performed on the EPICS V flow cytometer using an argon UV laser. Ca2+levels are shown as the ratio of 515 nm/405 nm wave length emission. Phosphatidylinositol turnover. T cells were suspended in vitamin-free RPM1 1640 medium (GIBCO) at a concentration of 20 X lo6 cells/ml and cultured at 37°C overnight with 1.85 MBeq/ml (50 pCi/ml) of [3H]2-myo-inositol (370-740 GBq/mmol, NEN DuPont, Boston, MA). After having been labeled, cells were washed three times with Hanks’ balanced salt solution (HBSS) and suspended in 10 mM LiCl-10 mM Hepes containing HBSS at a concentration of 1 X lo7 cells/ml. After a IO-min preincubation at 37°C l-ml aliquots were further incubated with mAbs at 37°C for 15 min. Inositol phosphates (IPn) were extracted by ice-cold TCA and ethylether and further purified using the Dowex 1 X 8 column method (21). RESULTS Development and Characterization of mAb Anti-HVS6B6 The mAb anti-HVS6B6 was raised in a mouse inoculated with a Herpes virus saimiritransformed common marmoset T cell line as described under Materials and Methods. This mAb was selectedbecauseof its ability to enhance anti-CD2 (anti-T 112 and antiT 113)-induced T cell proliferation. Peripheral blood T cells from healthy donors were cultured with anti-CD2 mAbs for 3 days, and the proliferative responseswere measured

148

YAMADA

ET AL.

TABLE 1 HVS6B6 mAb Enhances Anti-CD2 mAb (Anti-T1 l2 and Anti-T1 1,)~InducedProliferation of T Cells” [3H]TdR uptake (cpm) b Stimulated with

mAb

Expt 1

Expt 2

Expt 3

Anti-T I 1zf3

None HVS6B6

99,339 190,754 96,480 5,628

63,769 96,180 66,542 7,784

39,152 51,390 41,490 12,678

CD4 2F12

n T cells were cultured with anti-T 1I2 and anti-T 113 mAbs ( I:200 dilution) in the presence of indicated mAb ( 1:100 dilution of each ascites) for 3 days, and pulse labeled with [‘H]TdR. b Values are mean cpm of triplicate samples. SD was always less than 15%of the mean.

by [3H]thymidine uptake (Table 1). Anti-CD2 mAbs-induced T cell proliferation was enhanced in the presenceof anti-HVS6B6 mAb, whereas the isotype-matched (IgGl) control mAbs anti-2Fl2 (which recognizesLFA- 1) and anti-CD4, respectively, inhibited or had no effect on the CDZmediated T cell response. Immunoprecipitation experiments were then performed to determine the nature of the antigen defmed by anti-HVS6B6. Anti-HVS6B6 precipitated a 1SO-kDaband and coprecipitated a 95-kDa molecule as a minor band from ‘251-surfacelabeled PBMC (Fig. 1A, lane c). The isotype control mAb, 2F12, which recognizes the LFA- 1 a-chain (CD1 1a), similarly preincubated a 180-kDa band as well as a 95-kDa leukocyte integrin common fi2-chain (Fig. 1A, lane b). 10F12, a mAb specific for the common P-chain

B

A abcde

abcdefghi .*” 0

200-

11697-

FIG. 1. Molecular structure of the HVS6B6 antigen. (A) Comparison of immunoprecipitation by antiHVS6B6, anti-CD1 la, anti-CD18, and anti-CD1 lb mAbs. Lysates from “‘I-surface labeled PHA-activated T cells were immunoprecipitated by anti-T6 (CDl, lane a), anti-2F12 (CD1 la, lane b), anti-HVS6B6 (lane c), anti-lOFl2 (CD18, lane d), or anti-MO1 (CD1 lb, lane e). Immunoprecipitated samples were analyzed by SDS-PAGE (6% acrylamide) under reducing condition. (B) Sequential immunoprecipitation analysis of the HVS6B6 antigen. Cell lysate from the ‘ZSI-surfacelabeled H-9 cell line was precleared with control mAb anti-T6 (lanes a-c), anti-2F12 (lanes d-f), or anti-HVS6B6 (lanes g-i). Precleared ceI1lysates were then immunoprecipitated by anti-lF7 (CD26, lanes a, d, g), anti-2Fl2 (CD1 la, lanes b, e, h), or anti-HVS6B6 (lanes c, f, i), and then analyzed by 6% SDS-PAGE under reducing condition.

ROLE OF LFA-1

ON CD2 SIGNAL

TRANSDUCTION

149

(CD18) of LFA-1, precipitated a 95-kDa common P-chain and a 150-kDa ~150 achain (CD1 lc), 170-kDa MO- 1 (CR3) a-chain (CD 11b), and 1SO-kDaLFA- 1 a-chain (Fig. 1A, lane d). Anti-MO-1 also coprecipitated the 95-kDa P-chain (Fig. lA, lane e). These results suggestthat anti-HVS6B6 may recognize the a-chain of the LFA- 1 molecule. To further clarify this possibility, sequential immunoprecipitations were performed (Fig. 1B). When a lysate from surface-labeled PBMC was precleared using anti-LFA- 1 CY(2F12), the bands precipitated by anti-HVS6B6 were lost (lane f). Similarly, when the lysate was precleared with anti-HVS6B6, the bands precipitated by anti-LFA- 1 CYwere markedly diminished (lane h). However, preclearing with anti-T6 (CDl) affected neither anti-HVS6B6 nor anti-LFA-1 CYimmunoprecipitation (lanes b, c). The control anti- 1F7 immunoprecipitation also was unchanged by preclearing with anti-T6 (lanes a, d, and g). These results indicate that anti-HVS6B6 recognizes the a-chain of the LFA-1 molecule. Next, we examined the cytofluorographic profile of HVS6B6 antigen expression on various types of leukocytes. As shown in Fig. 2, the HVS6B6 antigen is expressedon virtually all leukocytes. Its staining profile is almost the same as that seen with the conventional anti-LFA-1 mAb, 2F12. However, the density of expression and percentage of cells expressing HVS6B6 antigen in PBMC and peripheral blood T cells was always lower than that detected by staining using conventional anti-LFA- 1 mAbs. The average percentage of HVS6B6-positive T cells from 10 different donors was 65 -t 20% (mean + SD). The distribution of HVS6B6 antigen on various T and B leukemic cell lines is also the same as that detected by anti-2F12, although its density on the cells appearslower than that of conventional LFA- I antigens (data not shown). Staining with FITC-conjugated anti-2F12 was completely blocked by a pretreatment with anti2F12 itself or anti-YH-384, both of which reacted with ICAM- interacting epitope on the LFA-1 a-chain, whereas anti-HVS6B6 pretreatment did not block the anti2F12 staining (data not shown). Thus, the anti-HVS6B6 recognizing epitope differed from the ICAM- 1 binding site. Indeed, conventional anti-LFA- 1 mAbs (anti-2F12 or anti-YH384) strongly inhibited LFA-l/ICAM-l-dependent T cell responses,such as

NEG

2F12

HVS6B6

pBLlala

E+LIYIIlil,a

E-;Imm

4lzlam

pMNlliLlil Fluorescense

Intensity

FIG. 2. Reactivity and cytofluorographic profile of anti-HVS6B6 and anti-2F12 mAbs on peripheral blood leukocyte populations. PBL; peripheral blood mononuclear leukocytes, MO; monocyte enriched population, PMN; polymorphonuclear leukocytes.

150

YAMADA

ET AL.

PHA-induced T cell proliferation or allo antigen-specific MHC classI-restricted CD8+ cytotoxic T cell function. In contrast, the inhibitory effect of anti-HVS6B6 on those responseswas very weak. Our preliminary experiments also suggestedthat the HVS6B6 epitope differed from cation-dependent epitopes such as NKI-L16 (22), since EDTA or EGTA treatment of cells could not diminish the staining with anti-HVS6B6. E#ect of Various Anti-LFA-1 mAbs on Anti-CD2-Induced Proliferation of T Cells Further experiments were performed to determine whether other anti-LFA- 1 mAbs also enhance T cell activation through the CD2 pathway. As shown in Table 2, antiHVS6B6 itself was not mitogenic, and the combination of anti-HVS6B6 with either anti-T 112 or anti-T 113did not induce any T cell proliferation. Anti-HVS6B6 enhanced T cell proliferation only to the combination of anti-T 112 plus anti-T 113. On the other hand, two other anti-LFA-1 mAbs, lOF12 (CD1 8) and 2F12 (CD1 la), inhibited antiCD2-induced T cell proliferation. Another mAb, anti-S6F1, which recognizesa unique epitope of the LFA- 1 molecule that can distinguish between suppressor and cytotoxic effector populations of CD8+ T cells (16), had no effect on the anti-CD2-induced T cell response.Although we further examined more than 20 clones of anti-LFA- 1 mAbs, which were obtained from the 4th International Workshop on Human Leukocyte Differentiation Antigens, none of the anti-LFA- 1 mAbs enhanced anti-CD2-induced proliferation (data not shown). The induction of T cell proliferation by anti-HVS6B6 was most dramatic when suboptimal concentrations of anti-CD2 mAbs were used (Table 2, experiment 3). It should be noted that the enhancement of anti-CD2-induced T cell proliferation by anti-HVS6B6 was observed during the entire time course we examined (between Day 3 and Day 6), suggestingthat this is not simply due to a change in the kinetics of the anti-CD2-induced proliferative response (data not shown). TABLE 2 Effect of Various Anti-LFA-1

mAbs on Anti-CD2

mAb-Induced

[‘H]TdR Expt 1

2 3

Stimulated with” None Anti-T1 l2 Anti-T 1 I r Anti-T 112+3 None Anti-T1 12+3 None Anti-T 11Z+3 1:200 1:lOOO

Nil

HVS6B6 ’

203’ 221 259 96,165 143 127,069 189 186,267 9,962

Proliferation

of T Cells

uptake (cpm)

2Fl2’

lOFl2’

S6Fl b

8F2’

267 235 331 150,902 226 176,108 142

181 164 183 17,943 346 6,948 138

271 274 263 17,081 260 21,110 191

196 221 258 95,684 225 127,332 ND

ND ND ND ND 263 129,150 ND

221,369 40,116

24,903 1,351

18,945 1,645

ND ND

ND ND

a In experiments 1 and 2, 1:200 dilution of anti-T 112 and/or anti-T 113 m Ab ascites were used. b Purified anti-LFA-I or anti-CD49d (8F2) mAbs were added to the T cell culture at a concentration 5 a/ml. ’ Values are mean cpm of triplicate samples. SD was always less than 15% of the mean.

of

ROLE

OF LFA-I

ON CD2 SIGNAL

151

TRANSDUCTION

Dose Responseof Anti-LFA-1 mAbs on Anti-CD2-Induced Proliferation of T Cells Of four anti-LFA- 1 mAbs (three anti-CD 11a and one anti-CD 18 mAbs) tested in our culture system, only anti-HVS6B6-enhanced anti-CD2 induced T cell proliferation. Since HVS6B6 antigen expression by T lymphocytes is lower than conventional LFA1 antigen expression, it is conceivable that the relative dosage of the conventional mAbs used in our experiments was simply higher. Thus, it is possible that at lower doses, even conventional anti-LFA- 1 mAbs might also have enhancing effects on antiCD2-induced T cell responses. To test this possibility, the mAb anti-2F 12 was assessed at concentrations ranging from 50 pg/ml to 0.5 rig/ml. As shown in Fig. 3B, no enhancing effects were observed. In contrast, anti-HVS6B6 mAb strongly enhanced the anti-CD2-induced T cell responses at concentrations ranging from 1 through 20 pg/ ml (Fig. 3A). Thus, using higher concentrations of anti-HVS6B6 mAb did not result in suppression of anti-CD2-induced T cell activation.

Kinetics of the Effect ofAnti-HVShB6 on Anti-CD2 mAbs Induced T Cell Proliferation Next we attempted to determine which stage of anti-CD2-induced T cell activation was affected by this mAb. As shown in Table 3, anti-HVS6B6 exhibited maximal enhancement of T cell proliferation when the antibody was added prior to or at the same time as the stimulator-y anti-CD2 mAbs. When anti-HVS6B6 was added 30 min after the anti-CD2 stimulation, only a moderate level of enhancement was observed.

80

0.5 1 2 5 10 20 50 mAb Concentration (rrg/ml)

B bs

mAb Concentration

(x5ug/ml)

FIG. 3. Dose response of anti-LFA-1 mAbs on anti-CD2 mAb-induced proliferation of T cells. T cells were cultured with anti-T 112 and anti-T 113 mAbs (I: 1000 dilution) in the presence of various concentrations of anti-HVS6B6 (A) or anti-2F12 mAb (B). Gray horizontal bars show the mean f SD of control (anti-CD2 mAbs alone) response.

152

YAMADA

ET AL.

TABLE 3 HVS6B6 mAb Affects Early Events in T Cell Activation mAb added” None Anti-T1 12+salone Anti-T1 12+, + anti-HVS6B6 Anti-T1 12+, + anti-HVS6B6 Anti-T1 12+3+ anti-HVS6B6 Anti-T 11zf3 + anti-HVS6B6 Anti-T1 l2+3 + anti-HVS6B6 Anti-T1 lZ+s + anti-HVS6B6 Anti-T1 lZ+s + anti-HVS6B6

Induced by the CD2 Pathway [‘H]TdR

(-0.5 hr) (0 hr) (0.5 hr) (1 hr) (6 hr) (24 hr) (48 hr)

uptake (cpm) b

80 6,896 42,725 40,628 24,256 7,741 6,891 6,366 6,703

f 6 f 1527 + 5370 f 3962 + 1954 -c 567 f 326 + 456 f 467

a Anti-T I I2 and anti-T 113 mAbs ( 1: 1000 diluted ascites) were added to the T cell culture at 0 hr. Purified anti-HVS6B6 mAb (10 pg/ml) was added at the indicated time. * Values show the mean + SD of triplicate samples.

Addition of anti-HVS6B6 at 1 hr or later resulted in no enhancement. These results suggestthat anti-HVS6B6 affectsearly events in T cell activation by the CD2 pathway. Effect of Anti-HVS6B6 on Tl 13-Epitope Expression Since the Tl l,-epitope expression is required for cell activation by anti-T1 l2 plus anti-T1 13 mAbs, we sought to determine whether anti-HVS6B6 could effect anti-

Tl l,-expression FIG.4. Effect of anti-HVS6B6 on T 1ls-epitope expression induced by anti-T1 with 1:lOOO diluted anti-T1 l2 at 37°C for 1 hr in the presence of 10 &ml (d), or anti-T1 l2 alone (b). After the incubation period, cells were washed and anti-T1 13, and analyzed by flow cytometry. (a) The T 11,-epitope expression

l2 mAb. T cells were incubated of anti-HVS6B6 (c), anti-8F2 stained with FITC-conjugated on unstimulated T cells.

ROLE OF LFA-I

ON CD2 SIGNAL

TRANSDUCTION

153

Tl 1,-induced Tl lrepitope expression on T cells. Tl 13-epitopeexpression was measured by immunofluorescence using FITC-conjugated anti-T 113 mAb and flow cytometric analysis. As shown in Fig. 4a, only a small number of resting T cells expressed the T 113-epitope(3%). Addition of a suboptimal concentration ( 1:1000 diluted ascites) of anti-T 112 induced T 113-epitopeexpression (Fig. 4b). Under these conditions, about 65% of T cells expressedthe T 113-epitope at 1 hr. Figure 4c shows the T 113-epitope expression on T cells incubated with this concentration of anti-T1 l2 mAb and antiHVS6B6 mAb. Both the density of Tl 13-epitope expression and the percentage of cells expressing T 113were similar to that of cells incubated with anti-T 112 mAb alone. Likewise, the control mAb anti-SF2 had no effect on the 7’113-epitope expression (Fig. 4d). Theseresults suggestthat the enhancement of anti-CD2-induced T cell proliferation by anti-HVS6B6 was not due to the enhanced expression of the Tl l3 epitope. Effect of Anti-HVS6B6 on Anti-CD2 mAb Induced Ca2+In&x It has been shown that stimulation of T cells with the combination of anti-T 1l2 plus anti-T1 13 mAbs rapidly increasescytoplasmic Ca*+ ion levels (23). To determine whether treatment of T cells with anti-HVS6B6 affects a change in intracellular Ca*’ levels, T cells loaded with the Ca2+-sensitivedye Indo- 1 were preincubated with antiHVS6B6 for 15 min and then stimulated with anti-CD2 mAbs and analyzed by flow cytometry with argon uv laser. As shown in Fig. 5a, after treatment of T cells with ionomycin, a marked Ca*+ influx was observed. Anti-HVS6B6 itself did not induce

.* -1 100

zoo-

s 400- iI@ E BOO-

.A.

a

t

I

0

5

1-T

10

T

15

20

I I b. . ,I! . .C . .u . d. * .J

0

5

1015200

5

10

15

20

Time (minutes) FIG. 5. Effect of anti-HVS6B6 mAb on anti-CD2 mAb-induced Ca2+ influx. Indo-l AM loaded T cells were preincubated with (d, f-h) or without (a-c, e) anti-HVS6B6 mAb (IO &ml) for 15 min at room temperature and then subjected to flow cytometry (EPICS V) to measure the basal level of cytoplasmic Ca’+. At the arrow, stimulants were added and measurement was continued for up to 1000 sec. (a) stimulated by 1 rg/ml of ionomycin; (b) stimulated by anti-HVS6B6 mAb; (c) stimulated by I:200 diluted anti-T1 I2 and anti-T 113 mAbs; (d) preincubated with anti-HVS6B6 and then stimulated by 1:200 diluted anti-T 1 I z and anti-T 113 mAbs; (e) stimulated by 1: 1000 diluted anti-T 11z and anti-T1 13; (f) preincubated with antiHVS6B6 and then stimulated by 1: 1000 diluted anti-T1 I2 and anti-T1 1s mAbs; (g) preincubated with antiHVS6B6 and then stimulated by 1:200 diluted anti-T1 I2 mAb; (h) preincubated with anti-HVS6B6 mAb and then stimulated by 1:200 diluted anti-T1 I, mAb.

154

YAMADA

ET AL.

a change in cytoplasmic free-Ca‘+ levels (Fig. 5b). Figure 5c illustrates the maximal Ca2+influx seen after incubation of cells with an optimal concentration of anti-CD2 mAbs. Under the sameconditions, the addition of anti-HVS6B6 mAb enhanced maximal Cazf levels, but the magnitude of this enhancement was small (Fig. 5d). When a suboptimal concentration of anti-CD2 mAbs was usedto stimulate theselymphocytes, only a small number of T cells responded (Fig. 5e). However, in the presence of antiHVS6B6, a significant Ca2+influx was induced (Fig. 5f). In contrast, a combination of anti-HVS6B6 and either anti-T 112 or anti-T 113 mAbs was unable to induce a Ca*+ influx (Figs. 5g and 5h). Under the same conditions, anti-2F12 did not affect the antiCD2-induced Ca2+influx (data not shown). Efect ofAnti-HVS6B6 mAb on Anti-CD2 mAbs Induced Phosphatidylinositol Turnover Products of phosphatidylinositol turnover, such as 1,4,5-inositol trisphosphate (IP3) cause an elevation of cytoplasmic-free Ca2+ion levels (24-28). Therefore, to further clarify the mechanism by which anti-HVS6B6 enhanced anti-CD2-induced elevation of cytoplasmic Ca2+ levels and activation, we evaluated the effect of anti-HVS6B6 mAb on anti-CD2 mAb-induced hydrolysis of phosphatidylinositol-bis-phosphate (PIP2). For this purpose, T cells were labeled with [3H]2-myo-inositol and then stimulated with anti-CD2 mAbs and/or anti-HVS6B6 mAb in the presence of lithium chloride to inhibit the activity of IP-phosphatase. After 15 min, inositol phosphates (IPn) were extracted and separatedby the Dowex ion exchange column method (21). As shown in Table 4, the anti-CD2 mAb-induced hydrolysis of PIP2 causesIP3, IP2, and IP production. IP3 and IP2 production induced by anti-CD2 mAb stimulation was enhanced by the pretreatment of T cells with anti-HVS6B6. However anti-HVS6B6 alone did not induce any change in IPn levels. These results suggestthat the pretreatment of T cells with anti-HVS6B6 mAb affectsa very early phaseof signal transduction in the CD2 pathway of T lymphocyte activation. TABLE 4 HVS6B6 mAb Augments the Hydrolysis of Phosphatidylinositol-bis-phosphate (PlP2) of T Cells Induced by Anti-CD2 mAbs” [‘Hlmyo-inositol (cpm) Expt

Treated with

IP

IPZ

Ip3

None Anti-HVS6B6 Anti-T 112+3 Anti-T 11rt3 + anti-HVS6B6 None Anti-HVS6B6 Anti-T1 lz+3 Anti-T1 lz+s + anti-HVS6B6

952 941 8,475 9,114 1,229 1,226 19,952 19,020

48 51 742 1182 219 220 4601 5232

56 54 388 1088 295 313 1627 2961

a [3H]2-myo-inositol-labeled T cells were preincubated in 10 mM LiCl-10 mM Hepes-HBSS at 37°C for 10 min. After preincubation, cells were stimulated with a I:200 dilution of anti-T1 l2 and anti-T1 1s mAb ascites with or without 10 pg/ml of anti-HVS6B6 mAb and then incubated for a further 15 min at 37°C. Termination of the reaction and the purification protocol for the IPn are described under Materials and Methods.

ROLE OF LFA-I

ON CD2 SIGNAL

TRANSDUCTION

155

DISCUSSION In the present study, we have described a monoclonal antibody, anti-HVS6B6, which has costimulatory effectson T cells when added in combination with anti-CD2 mAbs, anti-T 1l2 and anti-T1 13. The cell-surface structure defined by this mAb is comprised of two glycoproteins with relative molecular massesof 180 and 95 kDa. Sequential immunoprecipitation studies indicated that anti-HVS6B6 recognizes an epitope of the a-chain of LFA- 1 molecule. Anti-HVS6B6 antibody, which by itself is not mitogenic, has a unique capacity to stimulate T cell activation via the CD2 pathway. Other anti-LFA-1 antibodies tested in our study had either no effect or an inhibitory effect on anti-CD2-induced T cell activation. Since HVS6B6 antigen expression appears to be lower than conventional LFA-1 antigen expression, it remains possible that such an enhancing effect on antiCD2-induced T cell activation may be observed if lower concentrations of other antiLFA- 1 antibodies are used in concert with anti-CD2 to induce T cell responses.However, this possibility appearsunlikely since lower concentrations of conventional antiLFA-1 mAbs tested did not result in an enhancement of anti-CD2-induced T cell responses,nor did much higher concentrations of anti-HVS6B6 result in an inhibition of this response. Kinetics studies suggestthat this mAb affects an early phase of the CD2-mediated response. Expression of the T 113 epitope is required for activation of T cells via the CD2 (anti-T 112 and anti-T 11J pathway (23). This epitope, hidden on resting T cells, is induced by anti-T 112. Thus, the possibility existed that anti-HVS6B6 enhanced the T 113 epitope expression induced by anti-T 112, resulting in further activation of T cells following the in vitro addition of anti-T 113. However, as shown in Fig. 4, this possibility is unlikely. After stimulation of T cells with the combination of anti-T 11z and anti-T 1l3 mAbs or with anti-CD3 mAb, cytoplasmic free Ca2+ levels rapidly increase. This elevation of cytoplasmic Ca2+levels is mainly caused by the products of phosphatidylinositol4,5-bisphosphate (PIP2) hydrolysis, such as inositol- 1,4,5-trisphosphate ( 1,4,5-IP3) or inositol- 1,3,4,5-tetrakisphosphate (1,3,4,5-IP4) (24-28). Treatment of T cells with anti-HVS6B6 mAb enhances Ca2’ influx induced by antiCD2 (T 112and T 113) mAbs, although anti-HVS6B6 by itself did not induce alteration of cytoplasmic Ca2+levels (Fig. 3). Moreover, IP3 production induced by anti-CD2 stimulation was also enhanced by the pretreatment of T cells with this mAb (Table 4). Taken together the above results strongly suggestthat anti-HVS6B6 affects the very early phase of signal transduction through the CD2 pathway. The literature contains conflicting findings regarding the effect of anti-LFA- 1 mAb on Ca2+ influx. Ledbetter et al. (29) reported that the cross-linking of the LFA-1 molecules by murine anti-LFA-1 mAbs and goat anti-mouse Ig antibody induced weak Ca2+ influx, whereas the cross-linking of the LFA-1 molecules with rat antimouse Ig antibody as a second antibody did not induce Ca2+ influx. Furthermore, others have shown that the cross-linking of LFA-1 molecules does not induce Ca2+ influx (30). Anti-HVS6B6 mAb wasalso unable to induce any alteration of cytoplasmic Ca2+levels (either with or without cross-linking by anti-mouse Ig antibody). Only in combination with anti-T1 l2 and anti-T1 13 mAbs could this particular anti-LFA- 1 mAb enhance Ca2+influx. An enhancing effect of soluble anti-LFA- 1 on solid phase anti-CD3-induced T cell activation was reported by Noesel et al. (9). They indicated that a panel of anti-LFA- 1 a-chain (CD1 la) mAbs had an enhancing effect on CD3-

156

YAMADA

ET AL.

induced T cell activation, whereas anti-LFA- 1 P-chain (CD 18) mAbs had either an inhibitory or no effect on this T cell response. Anti-HVS6B6 had at most a weak enhancing effect in this solid-phase anti-CD3 system (data not shown). More recently, Wacholtz et al. (10) demonstrated that the cross-linking of LFA-1 and CD3 by goat anti-mouse Ig prolonged Ca” influx over that observedwith anti-CD3 alone. However, in this study, the initial Ca2+influx (both the cytoplasmic Ca2+level and percentage of responding cells) observed following LFA- l/CD3 cross-linking was the sameas that observed following CD3 cross-linking alone. In contrast, in our anti-CD2 activation system,binding of anti-HVS6B6 to the cell surfaceprovided sufficient signal to enhance the Ca2+influx induced by a pair of anti-CD2 mAbs, resulting in the enhancement of IL-2 receptor expression, IL-2 production (data not shown), and T cell proliferation. Huet et al. (11) demonstrated that anti-LFA-1 mAbs have two different effects on the activation of T cells by mitogenic combinations of anti-CD2 mAbs. They showed that anti-LFA-1 mAbs have an inhibitory effect on T cell activation induced by the combination of anti-GT2 and 9.6/T111 mAbs, which can induce T cell proliferation only in the presence of macrophages. These findings underscore the role of direct contact betweenthe accessorycells and T cells in this mode of CD2-mediated activation. Interestingly, anti-LFA- 1 mAbs enhanced the T cell activation induced by a different combination of anti-CD2 mAbs (anti-D66 and 9.6/T 11, mAbs), which can induce T cell activation in the absence of macrophages. These investigators have speculated that anti-LFA-1 mAbs may block a late occurring inhibitory contact between cell populations via the LFA- 1 molecule in this anti-CD2 mAb-mediated T cell activation. Cerdan et al. (3 1) also showed a similar result using a combination of anti-CD2 (antiT 11, and anti-T1 12)stimulation with PMA. Although the combination of anti-T1 l2 and anti-T 113 mAbs used in our study can induce T cell activation in the absenceof macrophages(32) the addition of macrophagesresults in even more potent activation. However, the addition of conventional anti-LFA- 1 mAb strikingly inhibited the CD2mediated T cell activation both in the presence and absence of macrophages. These data suggestthat the inhibition of anti-LFA- 1 in this CDZmediated T cell activation is not due solely to the inhibition of macrophage-T cell interaction. The enhancing effect of anti-HVS6B6 on anti-CD2-induced T cell activation was also observed when highly purified T cells were used instead of crude T (E+) cells (data not shown). Moreover, F(ab’)2 fragment of anti-HVS6B6 also expressedsimilar enhancing effect (data not shown). Therefore, it is very difficult to believe that the monocyte contributes to the enhancing effect of anti-HVS6B6. While the present data are unable to delineate the precise physiologic role of the LFA-1 molecule in T cell activation, we speculate that the augmentation of activation by anti-HVS6B6 could be due to the delivery of a specific positive signal, perhaps mimicking T cell interaction with macrophages. Alternatively, conventional anti-LFA- 1 mAbs might either block these costimulatory signals or interactions between different cell populations or they might direct a negative signal through the LFA-1 molecule. Further studies will be required to clarify these points. A costimulatory effect of fibronectin through VLA-5 (CD49e/CD29) on CD3-induced T cell activation has been reported along with a description of its mechanism (15, 33-36). Namely, signaling from the VLA-5 induces AP-1 transcription factor which regulates IL-2 gene expression. However, VLA-5-mediated signaling does not modify CD3-induced early events, such as Ca2+influx or protein kinase C activation (36). Thus, even though both the LFA- 1 and the VLA-5 belong to the integrin family

ROLE

OF LFA-1

ON CD2 SIGNAL

TRANSDUCTION

157

and both of them provide costimulatory signals, enhancing mechanisms of those two integrin molecules on T cell activation are completely different. Since the LFA- 1 and the VLA-5 belong to different subfamilies of integrin, each subfamily might have a different signal transduction mechanism. In the present paper, we have demonstrated that a functionally unique anti-LFA1 mAb, anti-HVS6B6, enhances the proliferation of T cells induced by the mitogenic combination of anti-CD2 mAbs, anti-T1 l2 and anti-T1 13, through its effect on an early step in the CD2-mediated signal transduction pathway. In addition to those of previous studies, our results strongly suggest that the LFA-1 molecule is not only involved in cellular adhesion, but is also involved in CD2-mediated signal transduction during T cell activation. ACKNOWLEDGMENTS We thank Drs. Takami Matsuyama and David M. Rothstein for their helpful discussions. We also thank Drs. S. F. Schlossman and C. Morimoto for their generous gift of mAbs and for their criticisms.

REFERENCES 1. Hynes, R. 0.. Cell 48, 549, 1987. 2. Tedder, T. F., Schmidt, R. E., Rudd, C. E., Kormacki, M. M., Ritz, J., and Schlossman, S. F., Eur. J. Immunol. 16, 1539, 1986. 3. Springer, T. A., Dustin, M. L., Kishimoto, T. K., and Marlin, S. D., Annu. Rev. Immunol. 5,223, 1987. 4. Pals, S. T., Otter, A. D., Miedema, F., Kabel, P., Keizer, G. D., Scheper, R. J., and Meijer, C. J. L. M., J. Immunol. 140, 1851, 1988. 5. Makgoba, M. W., Sanders, M. E., Lute, G. E. G., Dustin, M. L., Springer, T. A., Clark, E. A., Mannoni, P., and Shaw, S., Nature 331, 86, 1988. 6. Stuanton, D. E., Dustin, M. L., and Springer, T. A., Nature 339,61, 1989. 7. Walker, C., Bettens, F., and Pichler, W. J., Eur. J. Immunol. 17, 1611, 1987. 8. Carrera, A. C., Rincom, M., Sanchez-Madrid, F., Lopez-Botet, M., and Landazuri, M. 0. D., J. Immunol. 141, 1919, 1988. 9. Noesel, C. V., Miedema, F., Brouwer, M., Rie, M. A. D., Aarden, L. A., and Lier, R. A. W. V., Nature 333, 850, 1988. 10. Wacholtz, M. C., Patel, S. S., and Lipsky, P. E., J. Exp. Med. 170,431, 1989. 11. Huet, S., Wakasugi, H.. Sterkers, G., Gilmour, J., Tursz, T., Bounsell, L., and Bernard, A., J. Immunol. 137, 1420, 1986. 12. Morimoto, C., Letvin, N. L., Boyd, A. W., Hagan, M., Brown, H. M., Kornacki, M. A., and Schlossman, S. F., J. Immunol. 134, 3762, 1985. 13. Meuer, S. C., Hussey, R. E., Fabbi, M., Fox, D., Acute, O., Fitzgerald, K. A., Hodgdon, J. C., Protentis, J. P., Schlossman, S. F., and Reinherz, E. L., Cell 36, 897, 1984. 14. Reinherz, E. L., and Schlossman, S. F., Cell 19, 821, 1980. 15. Matsuyama, T., Yamada, A., Kay, J., Yamada, K. M.. Akiyama, S. K., Schlossman. S. F., and Morimoto. C., J. Exp. Med. 170, 1133, 1989. 16. Morimoto, C., Rudd, C. E., Letvin, N. L., and Schlossman, S. F., Nature 330, 479, 1987. 17. Morimoto, C., Torimoto, Y., Levinson, G., Rudd, C. E., Schrieber, M., Dang, N. H., Letvin, N. L., and Schlossman, S. F., J. Immunol. 143, 3430, 1989. 18. Schriever, F., Freedman, A. S., Freeman, G., Messner, E., Lee, G., Daley, J., and Nadler. L., J. Exp. Med. 169,2043, 1989. 19. Hubbard, A. L., and Cohn, Z. A., In Biochemical Analysis of Membranes. (A. H. Maddy, Ed.), p. 427, Champman and Hall, London, 1976. 20. Breitmeyer, J. B., Oppenheim, S. O., Daley, J. F., Levine, H. B., and Schlossman, S. F., J. Immunol. 138, 726, 1987. 21. Benidge, M. J., Dawson, R. M. C., Dawson, C. P., Heslop, J. P., and Irvine, R. F., Biochem. J. 212, 473, 1983.

158

YAMADA

ET AL.

22. van Kooyk, Y., Weder, P., Hogervorst, F., Verhoeven, A. J., van Seventer, G., te Velde, A. A., Borst, J., Keizer, G. D., and Figdor, C. G. J. Cell Biol. 112, 345, 199 1. 23. Alcover, A., Ramarli, D., Richardson, N. E., Chang, H. C., and Reinherz, E. L., Immunol. Rev. 95, 5, 1987. 24. Berridge, M. J., and Irvine, R. F., Nature 312, 315, 1984. 25. Hokin, L. E., Annu. Rev. Biochem. 54, 205, 1985. 26. Majerus, P. W., Connolly, T. M., Dedimym, H., Ross, T. S., Bross, T. E., Ishi, H., Bamsal, V. S., and Wilson, D. B., Science 234, 15 19, 1986. 27. Irvine, R. F., and Moore, R. M., Biochem. J. 240, 917, 1986. 28. Irvine, R. F., Letcher, A. J., Heslop, J. P., and Benidge, M. J., Nature 320, 631, 1986. 29. Ledbetter, J. A., June, C. H., Grosmaire, L. S., and Rabinovitch, P. S., Proc. Natl. Ad. Sci. USA 83, 1384, 1987. 30. OFlynn, K., Krensky, A. M., Beverley, P. C. L., Burakof, S. J., and Linch, D. C., Nature 313, 686, 1985. 31. Cerdan, C., Lipcey, C., Lopez, M., Nunes, J., Pierres, A., Mawas, C., and Olive, D., Cell. Immunol. 123,344, 1989. 32. Matsuyama, T., Anderson, P., Daley, J. F., Schlossman, S. F., and Morimoto, C., Eur. J. Immunol. 18, 1473, 1988. 33. Shimizu, Y., van Seventer, G. A., Horgan, K. J., and Shaw, S., J. Immunol. 145, 59, 1990. 34. Davis, L. S., Oppenheimer-Marks, N., Bednarczyk, J. L., McIntyre, B. W., and Lipsky, P. E., J. Immunol. 145,785, 1990. 35. Yamada, A., Nojima, Y., Sugita, K., Dang, N. H., Schlossman, S. F., and Morimoto, C., Eur. J. Immunol. 21, 319, 1991. 36. Yamada, A., Nikaido, T., Nojima, Y., Schlossman, S. F., and Morimoto, C., J. Zmmunol. 146,53, 1991.