Eur. J. Immunol. 1991. 21: 979-983 LFA-1 and ICAM-1 in T help for B cells 979

Trevor Owens

Department of Neurology and

A role for adhesion molecules in contact-dependent T help for B cells*

Neurosurgery, McGill University, Montreal Neurological Institute, Montreal The role of cell contact in T-dependent B cell activation was examined. Small

resting B cells from C57BL/6 mice were cultured with CBA-derived, non- alloreactive cloned T helper cells in anti-T cell receptor Vg8-coated microwells. This induced polyclonal B cell activation to enter cell cycle (as measured by thymidine incorporation at 2 days) and to secrete immunoglobulin (as measured by an enzyme-linked immunoassay detecting high-rate Ig secretion at 5 days). The inclusion of monoclonal antibodies against LFA-1, ICAM-1 and CD4 in these cultures strongly inhibited antibody responses, although proliferative responses were only inhibited to about 50%. Inhibitory monoclonal antibodies did not significantly affect lipopolysaccharide-induced responses, T cell activation to interleukin (IL) 3 secretion, nor did they inhibit the formation of multicellular clusters containing Tand B cells. There was no correlation between the level of expression of adhesion molecules by T cells and their ability to induce B cell responses. Anti-LFA-1 abrogated T-dependent responses to IL 2 which were inducible after 2 days in culture, but did not inhibit the induction of this IL2 responsiveness. These results suggest that continued cell contact involving adhesion/accessory molecules induces B cells to proliferate and to respond to T cell lymphokines. A signaling role for cell interaction molecules on B cells is proposed, similar to the role of these and analogous molecules on Tcells.

1 Introduction 2 Materials and methods

The ability of B cells to process and present Ag for recognition by MHC class 11-restricted (CD4+) Th forms the basis for the cognate model of T help for B cell activation [ l , 21. Ag-specific, MHC class I1 restricted T h cells are activated to lymphokine production through recognition of Ag on B cells, and Ag-presenting B cells respond to these lymphokines by growth and differentia- tion to Ig production [ l , 21. Recent evidence suggests that T-B contact may itself be sufficient to induce B cell responses, in the absence of Ag-directed T lymphokine production or cognate interaction [3-61. Anti-TcR-acti- vated cloned T h cells induced small resting allogeneic B cells to both enter cell cycle and to high-rate Ig production [3]. Of particular interest was the fact that T h cells which did not produce direct-acting or classical B helper lymphokines could also mediate this form of help.

This report examines the participation of cell adhesion and accessory molecule ligand pairs in the induction of B cell activation through contact with activated T h cells using mAb to block interaction. Results indicate a role for LFA-l/ICAM-1 interactions in B cell responses, indepen- dent of direct effects on T o r B cells, or on their physical interaction to form cell clusters. A model for adhesion molecule participation in B cell signaling during interaction with T h is proposed on the basis of these observations.

[I 89011

2.1 mAb

F23.1 (mouse IgG2, anti-murineVg8; [7]), GK1.5 (rat IgG2h anti-L3T4 (CD4); [S]), 30H12 (rat IgG2b anti-Thy-1.2; [9]), 53-6.7.2 (rat IgG2, anti-Ly-2; [9]), 53-7.3 (rat IgG2, anti- Ly-1; [9]), F7D5 (mouse IgM anti-Thy-1.2; [lo]), FD441.8 (rat IgG2b anti-murine LFA-1; [l l]) , S4B6 (rat IgG2, anti-murine IL2; [12]), YNU1.7.4 (rat IgGZh anti-murine ICAM-1; [13]) and 11-1.23 (mouse IgG2, anti-H-2Kk; [14]) were affinity-purified from culture SN using either pro- tein A-Sepharose (Pharmacia, Montreal, Quebec, Ca- nada) or protein A-purified MAR-18.5 (mouse anti-rat x ; [15]) coupled to Sepharose 4B (Pharmacia). Optimal titers were determined by inhibition of MLR (FD441.8, GK1.5, YNU1.7.4) and/or by cell surface staining (GK1.5, 53- 6.7.2, F23.1, 30H12, F7D5).

2.2 B cells

Spleen cells from C57BL/6 female mice (Charles River Labs., St. Constant, Quebec, Canada) were depleted of RBC by hypotonic lysis, thenT-depleted by lysis with anti-T cell mAb (F7D5, GK1.5 and 53-6.7.2) + rabbit C (Cedar- lane, Hornby, Ontario, Canada). Cells were then fraction- ated on discontinuous Percoll (Pharmacia) gradients [3]. The cells which sedimented between e = 1.08 and 1.09 kg/l were used as B cells. These were > 99% Thy-1-, > 90% sIg+ , and homogeneously small.

* Supported by an operating grant and a Scholarship from MRC-Canada.

Correspondence: Trevor Owens, Montreal Neurological Institute, 3801 University Street, Montreal, Quebec, Canada H3A 2B4

Abbreviation: ABA: Azobenzenearsonate

2.3 Th cells

The hapten-specific T h clone E9.D4 (H-2k, F23.1+, Thy- 1.2+, CD4+) secretes IL2, IL3 and I F N y upon activation, but does not transcribe message for or secrete IL4 or IL5

0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991 00 14-298019 l/O4O4-O979$3.50 + .25/0

980 T. Owens Eur. J. Immunol. 1991. 21: 979-983

[16-181. Th cells were used following periods of 2-5 days resting culture, in the absence of Ag or growth factors, and were irradiated (1500 rad) before use. T h cells were myco- plasma free.Th cell activation was assessed by measurement of IL3, using the R6-X-E4.8.9 bioassay cell line (191, with the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) colorimetric readout (Sigma, St. Louis, MO; [20]). In some experiments haptenated Iak spleen cells were prepared by azobenzenearsonate (ABA)-coupling [16], and added at l@/culture.

T-B co-cultures were established by the addition of los B cells to T h cells in F23.1-coated microwelk [3]. LPS (S. abortus equi (Difco, Detroit, MI; 20 pg/ml) was added as a positive control for B cell activation. SN from F23.1-activated E9.D4 ( lo6 celllml) was added where appropriate. E9.D4-SN contained < 5 ng/ml soluble F23.1, insufficient to induce significant T h cell activation [3]. In some experiments Th cells were removed after time in culture, by depletion with (F7D5 + 11-1.23) + C.The purity of input and subsequently T-depleted B cells was assessed by FCM. Cultures were established in RPMI 1640 (Flow, Mississauga, Ontario, Canada), supplemented with 5% FCS (Flow), 2 mM L-glutamine (GibcoBRL, Burlington, Ontario, Canada), 5 x lo-' M 2-ME (Sigma) and penicillin (100 U/ml)-streptomycin (100 pg/ml) (Gibco).

B cells'entry into cell cycle was assessed by measurement of incorporation of [3H]dThd (ICN, Irvine, CA) following a 6-h pulse of 0.5 pCi = 18.5 kBq per microculture at 2 days of culture. Ig secretion was measured using an ELISA assay for high rate Ig production [3]. Briefly, cultures were washed, then cultured for 4 h in 100 pI total volume, after which the Ig content of 50 pI aliquots of SN was measured, using peroxidase-coupled sheep anti-mouse Ig (Bio/Can Scientific,Toronto, Ontario, Canada), and goat anti-mouse Ig (Antibodies Inc., Davis, CA) as a capture reagent.This ELISA was equivalently sensitive to IgM, IgG1, IgGz,,

Anli-ICAM

IgGzb and IgG3. Conversion of absorbance (405-492 nm) to Ig concentrations was carried out by an SLT 400AT microplate reader (Fisher Scientific, Montreal, Quebec, Canada) using a titration of affinity-purified F23.1 as reference.

2.5 Fluorescent labeling of Th cells and cluster analysis

Th cells were incubated for 5 min at 37°C with fluorescein diacetate (FDA; Sigma), at 10 pg/ml in culture medium, then washed extensively before addition to cultures. FDA labeling did not affect IL 3 secretion by Th cells in response to F23.1, nor their helper ability. Clusters containing fluorescent Th cells were analyzed after 16-24 h, resuspend- ing the contents of culture wells by reproducible pipetting (10 strokes at 30 pl displacement), then counting by fluorescence microscopy. Clusters were defined as aggre- gates of two or more cells. Two-cell conjugates were very rarely observed, and the frequency of clusters in unstimu- lated cultures was negligible.

2.6 FCM

Cells were stained with primary mAb (FD441.8, GK1.5, YNM.7.4) at 5 pg/ml for 20 min at 4 °C then washed once and stained with PE-coupled goat anti-rat Ig (Southern Biotechnology, Birmingham, AL) for a further 20 min. Fluorescence was analyzed using a FACScan (Becton Dickinson, Mountain View, CA).

3 Results

3.1 Induction of B cell responses by F23.1-activated E9.D4 T cells

Co-culture of small resting B cells with irradiated allogeneic E9.D4 Tcells in the presence of plastic-adsorbed anti-TcR antibody induced the entry of B cells into cell cycle (Fig. 1A

Control Anti-CD4

A n t i I F A - (

E9.04/F23.1 50% E9.04-SN

6 call8 alorm 0 F 2000 4000 6 0 0 0 8000 L 0 2 0 4 0 6 0 8 0 100

HIThyrnldlne incorporation (cprn) Ip secretlon (np/rnL)

O . 7 - 0.5 0 m 'p 0.4

v) 0.3 0 v) - Q. 0 .2

0 .1

0.0 Nil 2560 640 1 6 0 4 0

l l D i l u t i o n

0.0 ! - . Nil 6 4 0 1 6 0 4 0 1 0

l l D i l u t i o n

Figure 1. Inhibition of B cell responses by anti-adhesion mAb. Percoll-sepa- rated dense C57BL/6 B cells ( 1e5) were cultured with 4 x 104 (A, B) or 2 X 10" (C, D) irradiated (1500 rad) E9.D4 cells, in wells that had been coated with F23.1. Anti-LFA-1 and anti-CD4 mAb were added to a final concentration of 5 pglml, and anti-ICAM-1 was added as ascites fluid to a final dilution of 1/1OOO. Control cultures included T and B cells alone, co-cultures without F23.1. and B cells with E9.D4-SN (50%). (A, C) Entry into cell cycle was assessed at 48 h by measurement of ["IdThd incorpora- tion after a 6-h pulse. (B. D) Ig produc- tion was measured at 5 days by ELISA. Results show means k SD. ND: None detectable.The effect of mAb onTh cell responses was measured by assay of IL 3 in SN (As50-6Hn,), taken at 24 h from cultures containing 2 x 104 E9.D4 in

F23.1-coated wells (E). or cultured with lo6 haptenated Iak spleen cells (F). (0) Control (no antibody); (0) anti-eD4; (M) anti-LFA-1; (A) anti-LFA-1 plus anti-ICAM-1 (E only). Results show means of replicate bioassay titrations.

Eur. J. Immunol. 1991. 21: 979-983

Table 1. Effect of anti-adhesion mAb on Tand B cell activationa)

LFA-I and ICAM-1 in T hclp for B cells 981

I Control 67 Anti-ICAM-1 58 Anti-LFA- 1 54 Anti-CDJ NT Anti-LFA-1 + anti-ICAM-1 NT

EY.D-4 rcsponsc B cell responsc lo LPS F23.1-P ABAlIa' Entry into cycle Ig secretion

(Ulml IL3) (Ulml IL3) (cpm (3H]dThd) ( n g W

11

132 17.4 21800f 20 515 ? 10 NTh) NT 21 250 k 530 525 f 20 62 < 4 21 510 f 160 522 f 20 40 < 0.3 18 930 f 500 51x f 15 62 NT NT NT

a) Affinity-purified mAb GK1.5 or FD441.8 at 5 pg/ml, or YN111.7.4 ascites fluid at lll00 were added either to 100-yl cultures containing 2 x 104 E9.D4 in F23.1-coated wells (F23.1-P). or loh haptenated Iak spleen cells (ABA/Iak), or to 200-yl cultures containing 105 dense C57BL/6 B cells with 20 pglml LPS.Th cell activation was measured as IL3 titers in 24-h SN. B cell entry into cell cycle was assessed at 48 h by measurement of [3H]dThd incorporation following a 6-h pulse. Ig production was measured by ELISA at 5 days. Results arc presented as means * SD.

b) NT: not tested.

and C) and the activation of both high-rate Ig secretion and [3H]dThd incorporation at 5 days (Fig. 1B and D, and data not shown) The E9.D4 clone has no known alloreactivity [ 161, and there was no requirement for the addition of Ag or equivalent sIg ligands to cultures. SN from F23.1-activated E9.D4, at a concentration that corresponded to the activation of 3 x lo4 Tcells/culture, had no effect on resting B cells (Fig. 1C and D). RestingT cells did not induce B cell responses (Fig. 1C and D). Proliferative responses were

No addition

T cell SN

T cell SN + anti-IL-2

(A)

0 2 0 4 0 6 0 8 0 Ig secretion (nglmL)

(8)

No addition

T cell SN r T cell SN + anti-LFA-1

4 . . . . . . . i 0 1 0 2 0 3 0 4 0

Ig secretion (ng/mL)

Figure 2. Anti-LFA-1 inhibits IL2-dependent Ig secretion. Per- coll-separated dense C57BL/6 cells (2.5 x lo6) were cultured in 1 ml F23.1-coated wells with l@ irradiated (1500 rad) E9.D4 for 2 days. Cells were collected, and Tcells partially depleted by lysis with F7D5 (anti-Thy-1.2) plus 11-1.23 (anti-H-2Kk) plus C. The proportion of Tcells in culture was reduced by lysis from 10% to 1%-3%0. (A) The cells remaining were cultured at 5 X l@/well either alone, or with 50% E9.D4-SN f 5 pglml anti-IL2. Ig production was measured 4days later by ELISA. Results are presented as means+SD. (B) Cells were re-cultured at 105 cellslwell, either alone, or with 16.6% E9.D4-SN f 5 yglml anti- LFA-1. Ig production was measured by ELISA 4 days later. Results are shown as means f SD. ND: None detected.

55% (Fig. 1A) or 122% (Fig. 1C) of those induced by LPS and Ig responses were 16% (Fig. 2B) or 10% (Fig. 1D) of the LPS response. B cell activation was polyclonal, and was both abrogated by separation of T and B cells by a membrane, and enhanced by their clustering in tilted wells [3, 41.

3.2 Inhibition by anti-adhesion molecule mAb

3.2.1 Anti-adhesion mAb inhibit T-dependent B cell responses

Addition of affinity-purified anti-CD4 or anti-LFA-1, or anti-ICAM-1 ascites fluid to T-B co-cultures significantly inhibited T-dependent B cell responses (Fig. 1A-D). For proliferative responses, inhibition by anti-LFA-1 was between 24% and 60% (three experiments), between 35% and 54% by anti-CD4 (three experiments) and 27% and 73% by anti-ICAM-1. The inhibition of Ig responses was always stronger, being 65% to 97% by anti-LFA-1 (five experiments), 19% and 84% by anti-ICAM-1 and 83% to 96% by anti-CD4 (three experiments). Isotype-matched control mAb to other Tcell surface molecules (anti-Thy-1.2 and anti-Ly-1), or control ascites fluid at the same concen- tration had no effect (not shown). Anti-LFA-1 plus ICAM- 1 in combination did not increase the inhibition produced by either mAb alone (not shown).

3.2.2 Anti-adhesion mAb were not directly inhibitory for T or B cells

Anti-adhesion mAb reduced F23.1-stimulated IL 3 produc- tion by E9.D4 (in replicate cultures to those assayed for Fig. 1C and D) by at most 50% (Fig. 1E). InTable 1, these and data from another experiment are presented as units IL3/ml. Inhibitions of this magnitude are not sufficient to significantly affect these T-dependent B cell responses [3]. By contrast, the response to cellular Ag (ABA plus Iak) was abrogated by anti-CD4, and profoundly inhibited by anti- LFA-1 (Table 1, Fig. 1F). None of these mAb inhibited B cell responses to LPS (Table 1). These results indicate that

982 T. Owens Eur. J. Immunol. 1991. 21: 979-983

Table 2. % of fluorescent Tcells in clusters

Exp. Control Anti- Ant i-LFA- 1 Anti-CDJ no. ICAM-I

1 53 k 13.5 31 .8+5.5 N P ) NT 2 5.5+ 3.7 6.3? 3.3 8 . 2 f 5 . 4 NT 3 21.5kO.7 21.Ok5.7 22 .5k3 .5 23.5+0.5

a) FDA-labeled E9.D4 cells (2 X 104) were cultured in 200-1.11 microwells with 10s B cells, with or without 5 pg/ml affinity- purified mAb. After 16-24 h. cells were collected as described in Sect. 2.5, and the proportion of Tcells in multicell clusters analyzed by fluorescence microscopy. Results show means k SD.

b) NT: not tested.

the inhibition of T-dependent B cell responses, especially that of Ig secretion, did not result from direct effects of anti-adhesion mAb on Tor B cell activation.

3.2.3 Anti-adhesion mAb did not inhibit T-B clustering

Within 24 h of the initiation of culture, multicellular clusters can be observed. These contain Tand B cells, and although the formation of these clusters correlates with T-dependent B cell activation, their disruption (by pipet- ting) does not inhibit responses, and they re-form if disrupted. Neither the number of multicellular clusters, nor the number or frequency of T cells contained in such clusters, was affected by anti-adhesion mAb, although the proportion of T cells contained in clusters varied between experiments. Results from three experiments using fluor- escence-labeled Tcells are shown in Table 2.

3.3 No correlation between Th ability and adhesion molecule expression

If Tcells were pre-activated by culture with F23.1 and/or with IL2, there was no requirement for a Tcell activating stimulus during co-culture for B cell responses. In the experiments described in Table 3, T cells were cultured in IL2 for various times before addition to B cell cultures. Both their ability to induce B cell activation and their expression of adhesion molecules at the time of addition were assessed. There was no correlation between the level of expression of LFA-1 or of ICAM-1, and the ability of T cells to induce B cell activation.

3.4 LFA-1-mediated interactions required for progression to Ig secretion

Removal of T and B cells from F23.1-coated wells after 2 days, and partial Tcell depletion, abrogated further B cell response (Fig. 2A). The addition of SN from activated T cells restored the response, and this was inhibited by adding anti-IL2 mAb (Fig. 2A). rIL2 could replace SN from activated E9.D4 [21]. More rigorous depletion of T cells abrogated the response to IL2, and re-addition of irra- diated Tcells to rigorously T-depleted B cells restored B cell responses to IL2 [21]. This suggested a requirement for continued interaction with activated T cells for IL2- dependent progression to Ig secretion. Co-culture of Tand B cells for 2 days in the presence of anti-LFA-1 did not inhibit the subsequent response of these co-cultures to IL2 in the absence of F23.1, consistent with the weak effect of anti-LFA-1 on the entry of B cells into cycle. By contrast, inclusion of anti-LFA-1 with E9.D4-SN 2 days after remov- al of cells from F23.1-coated microwells, inhibited the Ig response (Fig. 2B). The concentrations of E9.D4-SN used in Fig. 2 were in the optimal or plateau range for reconsti- tution of the IL2-dependent B cells response ([21], and data not shown), and corresponded to the activation of 10-5 Tcells (Fig. 2A) or 3 x l@ Tcells (Fig. 2B).

4 Discussion

Extensive apposition of T and B cell membranes occurs during T-B interaction, together with redistribution of B and T cell surface molecules into the region of contact [22,23]. Non-cognate, Ag-independent contact events might therefore be important for B cell responses. Previous results showed a requirement for co-culture of Tand B cells on the same surface for non-cognate help by E9.D4 Tcells [3]. The inhibition by anti-LFA-1, anti-ICAM-1 and anti- CD4 shown here confirms that cell contact is required. Because anti-CD4 not only blocks CD4/Ia interaction, but can also inhibit T cell activation independently of such interaction [19,24], and since two otherwise inhibitory anti-MHC class I1 mAb did not inhibit B cell responses (Owens, unpublished), the inhibition by anti-CD4 is not as easily interpretable as that by anti-LFA-l/ICAM-1.

None of the mAb used strongly inhibited the T-dependent entry of B cells into cycle, as found by others (reviewed in [25]). By contrast, anti-cell adhesion molecule antibodies strongly inhibited the antibody response, and anti-LFA-1 appeared to inhibit at a relatively late stage of the response, after B cells had entered cell cycie. The contact event that induces cell cycle entry may involve other adhesion mole- cule ligand pairs, not targeted in this study.

Table 3. Expression of LFA-1 and ICAM-1 by resting and activated Th cells")

Tcells B cell Ig response Mean fluorescence (arbitrary units)

a) Tcells were either rested, as described in Sect. 2.3. or had been activated by overnight culture with F23.1.

F23. I N o F23.1 LFA-1 ICAM-1 Control followed by a 5-day culture in IL 2 (10 Ulml). Before addition to cultures (6 x 104 Tcells/lOs B cells). the

Resting 250+ 30 < 1 31.3 5.7 3.5 level of expression of LFA-1 and ICAM-1 on Tcells Activated 850 ? 110 720 + 50 23.1 5.2 3.7 was analyzed by FCM. Five days after initiation of

(ndml I&

culture, Ig secretion was measured by ELISA.

Eur. J. Immunol. 1991. 21: 979-983 LFA-1 and ICAM-1 in T help for B cells 983

The requirement that Tcells be activated for B cell response could reflect either a critical T cell surface density of adhesiodcell interaction molecules, andor a dependence on lymphokine production. There was no evidence for the former. The possibility that the requirement for T cell activation in this system reflects an LFA-1 conformational change, independent of levels of expression [26,27], remains to be examined.The induction of B cell entry into cycle is independent of lymphokine production by T cells [21, 25, 281, and the E9.D4 Th cells do not secrete “classical” B helper lymphokines (IL4, IL5), or any other lymphokine(s) that have direct effects on B cells. IL2 clearly plays a major role in driving B cells to Ig production, but it plays no role in cell cycle entry, and contact withTcells is also required for the induction of Ig secretion [21].

It is important that direct effects of mAb onTcell activation be excluded. It might be argued that the inhibition by mAb of IL 3 production (itself irrelevant to response) reflected the inhibition of other parameters of Th activation, them- selves involved in help for B cells. However, similar reductions in IL3 production in previous studies did not significantly affect helper activity in this system [3]. Fur- thermore, while the inhibition of proliferative responses may have been of similar magnitude to that of IL3 production, the inhibition of Ig responses was far more profound. This suggests that anti-adhesion mAb interfered with B cell signaling. In support of this, Hodgkin et al. who induced B cell activation by isolated T h cell membranes [29], could inhibit B cell responses with either of two anti-LFA-1 mAb (M. Kehry, personal communication). In that system, direct effects on T cells were not possible. Interestingly, neither anti-CD4 or anti-MHC class I1 mAb were inhibitory [29].

Signaling during T-dependent B cell activation may involve intermolecular aggregates on the B cell surface, by analogy with the role of accessory molecules in T cell activation [30-321. Adhesion molecules such as LFA-1 have been shown to transduce signals for T cell activation [26], and have been implicated in cognate help for B cells [26,33,34]. The results presented here show a role for LFA-1 and ICAM-1 distinct from either direct effects on T or B cell activation, and from gross effects on physical cell interac- tion, and are therefore consistent with the proposition that interactions involving these molecules contribute, during contact with Tcells, to B cell activation.

The expert technical assistance of Diane Heath is gratefully acknowledged. I would like to thank Michael Julius and Michael Ratcliffe (McGill University) for advice and discussions, and for comments on the manuscript. I am grateful to Fumio Takei (Univ. British Columbia) for YNIII. 7.4 ascites.

Received September 3, 1990; in revised form December 28, 1990.

5 References

1 Chesnut, R. W. and Grey, H. M., Adv. Irnrnunol. 1986. 39:

2 Tony, H.-P. and Parker, D. C., J. Exp. Med. 198.5. 161: 223. 3 Owens,T., Eur. J. Immunol. 1988. 18: 395. 4 Riedel, C., Owens, T. and Nossal, G. J. V., Eur. J. Irnmunol.

1988. 18: 403. 5 Julius, M. H. and Rammensee, H.-G.. Eur. J. Imrnunol. 1988.

18: 375. 6 Whalen, B. J.,Tony, H.-P. and Parker, D. C., J. Irnrnunol. 1988.

141: 2230. 7 Staerz, U.D., Rammensee, H.-G., Benedetto, J. D. and Bevan.

M. J, J. Irnrnunol. 1985. 134: 3994. 8 Dialynas, D. P.,Wilde, D. B., Marrack, P., Pierres. A.,Wall, K.

A., Havran, W., Otten, G., Loken, M. R., Pierres, M.. Kappler, J. and Fitch, F. W, Immunol. Rev. 1983. 74: 29.

9 Ledbetter, J. A. and Herzenberg, L. A., Imrnunol. Rev. 1979. 47: 63.

10 Lake, P., Clarke, E. A. , Korshidi, M. and Sunshine. G . H.. Eur. J. Irnrnunol. 1979. 9: 875.

11 Sarmiento, M., Dialynas, D. P., Lancki, D. W., Wall, K. A.. Lorber, M. I., Loken, M. R. and Fitch, F. W., Imrnunol. Rev. 1982. 68: 135.

12 Mosmann,T. R., Cherwinski, H., Bond, M.W., Giedlin, M. A. and Coffman, R. L., J. Immunol. 1986. 136: 2348.

13 Horley, K. J., Carpenito, C., Maker, B. and Takei. F., EMBO J. 1989. 8: 2889.

14 Oi,V.T., Jones, P. F!, Coding, J. W., Herzenberg, L. A. and Herzenberg, L. A., Curr. Top. Microbiol. Irnrnunol. 1978.81: 11.5.

15 Lanier, L. L., Gutman, G. A.,Lewis, D. E.. Griswold. S.T. and Warner, N. L., Hybridoma 1982. 1: 125.

16 Owens, T. and Miller, J. E A. P., J. Imrnunol. 1987. 138: 1687.

17 Kelso. A. and Owens,T., J. Immunol. 1988. 140: 11.59. 18 Williamson, D. J., Owens,T. and Pearse, M., J. Cell. Physiol.

19 Owens,T. and Fazekasde St. Groth, B., J. Irnmunol. 1987.138:

20 Mosmann, T. R., J. Irnmunol. Methods 1983. 65: 5.5. 21 Owens,T., Cell. Irnmunol. 1991. 133, in press. 22 Kupfer, A., Swain, S. L. and Singer, S. J., J. Exp. Med. 1987.

165: 1565. 23 Sanders,V. M., Snyder, J. M., Uhr, J. W. and Vitetta, E. S.. J.

Imrnunol. 1986. 137: 2395. 24 Bierer, B. E., Sleckman, B. P., Ratnofsky, S. E. and Burakoff. S.

J., Annu. Rev. Irnmunol. 1989. 7: 579. 25 Parker, D. C. (Ed.), Res. Imrnunol. 1990. 141: 405. 26 Springer, T. A., Nature 1990. 346: 425. 27 Figdor, C. G.,Van Kooyk,Y. and Keizer, G. D.. Imrnunol. Rev.

28 Krusimieir, M. and Snow, E. C., J. Imrnunol. 1988. 140:

29 Hodgkin, P. D.,Yamashita, L. C., Coffman, R. L. and Kehry,

30 Owens,T. and Zeine, R., Biochern. Cell Biol. 1989. 67: 481. 31 Owens,T., FazekasdeSt. Groth, B.andMiller,J. F. A. P., Proc.

32 Emmrich, F., Irnmunol. Today 1988. 9: 296. 33 Howard, D. R., Eaves, A. C. and Takai, F., J. Immunol. 1986.

34 Mazerolles, F., Lumbroso, C., Lecomte, 0.. Le Deist. F. and

51.

1989. 139: 245.

2402.

1990. 11: 277.

367.

M. R., J. Irnrnunol. 1990. 145: 2025.

Natl. Acad. Sci. USA 1987. 84: 9209.

136: 4013.

Fischer, A., Eur. J. Irnmunol. 1988. 18: 1229.