CELLULAR

IMMUNOLOGY

142, 338-347 (1992)

Bispecific Monoclonal Antibody Regulation of FcyRIII-Directed Tumor Cytotoxicity by Large Granular Lymphocytes IRMA GARCIA DE PALAZZO, JOANNE KITSON, CICEK GERCEL-TAYLOR, STACY ADAMS, AND LOUIS M. WEINER Department of Medical Oncology, Fox Chase Cancer Center, 7701 Burholme Avenue, Philadelphia, Pennsylvania 191 I I Received December 23, 1991; accepted April 6, 1992 Bispecific monoclonal antibodies (BsMAbs) prepared by somatic cell fusion bind monovalently to their targets and yet are extremely potent enhancers of target cell lysis by relevant effector cells. The mechanisms underlying this efficiency are not known. To investigate this property, we studied the ability of selected antibodies to modulate potentiation of tumor lysis by a bispecific antibody (CL1 58) which targets FcyRlII-expressing cells, via the 3G8 epitope, to malignant cells expressing CAl9-9 antigen. Antibodies directed against the 3G8 and B73.1 epitopes of FcrRIII efficiently inhibited BsMAb-mediated SW948 tumor cell lysis by interleukin-2 (IL-2)-activated lymphocytes (PBLs). Unexpectedly, Leu 19 antibody reversed antibody-dependent but not antibody-independent lysis of 5’Cr-labeled SW948 cells by IL-2-activated PBLs in a concentration-dependent fashion. Leu 19 binds to CD56, a neural cell adhesion molecule (N-CAM) isoform expressed by large granular lymphocytes (LGLs). The effects of Leu 19 on bispecific antibody promotion of lysis were due to competition for binding to the 3G8 epitope of FcyRIII and led to inhibition of binding between LGLs and SW948 cells. Leu 19 did not inhibit antibody-dependent lysis by the monospecific, bivalent IgG,, variant of CAl9-9 antibody. These studies show that competition assayscan be useful in dissecting the relevant mechanisms underlying BsMAb-promoted lysis. Steric constraints between effector cell trigger molecules (i.e., FcyRIII) and CAM such as N-CAM may regulate the function of these molecules. Understanding the roles of diverse CAM in this phenomenon will facilitate effortsto expand and use defined effector cell populations with maximal lytic potential and to identify potentially responsive tumor phenotypes. 0 1992 Academic Press, Inc.

INTRODUCTION Large granular lymphocytes (LGLs) comprise the major cell population mediating MHC-unrestricted lysis of autologous and allogeneic malignant cells (1). The mechanisms by which LGLs recognize their targets remain unknown. However, several lines of evidence point to the roles of cell adhesion molecules (CAM) as accessory structures in the binding of LGLs to targets. LGLs are known to express CAM, such as CD2, LFA-1, and CD56, which has been identified as an isoform of neural cell adhesion molecule (N-CAM) (2). The function of N-CAM on LGLs is unknown. LGLs also express FcyRIII (CD16), the low affinity Fc receptor for aggregated immunoglobulin, and thus are potent mediators of antibody-dependent cellular cytotoxicity (3-5). We have prepared by somatic cell fusion a bispecific monoclonal antibody which targets tumor and LGLs monovalently. This antibody is an extremely potent augmenter 338 0008-8749/92 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

BISPECIFIC

ANTIBODY

REGULATION

OF

LGL

CYTOTOXICITY

339

of tumor lysis by LGLs when compared to its corresponding bivalent, monoclonal anti-tumor antibody (6). To understand the basis of this difference, we studied the roles of various cell surface molecules in regulating lysis of tumor by LGLs in concert with a bispecific antibody targeting FcyRIII and CA 19-9 antigen or with the corresponding bivalent, monospecific anti-tumor antibody. MATERIALS

AND

METHODS

Cells. The SW 948 adenocarcinoma cell line was obtained from American Type Culture Collection (Rockville, MD) and was maintained as described (7). Peripheral blood lymphocytes (PBL) were isolated from normal donor peripheral blood, following monocyte depletion on plastic by density gradient centrifugation on FicollIsopaque (8). PBL were activated by culturing at IO6 cells/ml in RPM1 1640 medium with 10% fetal cell serum containing 1000 units/ml recombinant interleukin-2 (Cetus Corp. Emeryville, CA) for 72 hr. Typical cultures contained 20-25% CD56+ and 15-20% CD16+ cells by flow cytometric analysis. Antibodies. Clone 158 (CL158) bispecific monoclonal antibody was prepared by fusion of hybridoma clones CA 19-9 (9) and 3G8 (10). The resulting antibody binds monovalently to CA19-9 antigen and FcyRIII and is a potent mediator of lysis of SW948 cells, which express CA 19-9 antigen, by LGLs expressing FcyRIII (6). Clonal supernatants diluted to approximately 5 rig/ml of bispecific antibody produced optimal lysis potentiation at effector:target (E:T) ratios of 5:l in the assays described below. These supernatants perform identically to bispecific antibody partially purified by ion exchange chromatography. Supernatant containing the CA19-9 IgG,, variant was a kind gift from Dr. Zenon Steplewski (Wistar Institute). Leu 19 (anti-CD56), Leu 4 (anti-CD3), anti-CD25, and Leu 1 lc (anti-CD 16) were purchased from Becton-Dickinson (Mountain View, CA). The 3G8 cells are routinely maintained in our laboratory as sources of antibody. The 3G8 antibody was purified from culture supernatants by ion exchange chromatography and labeled with fluorescein isothiocyanate using the QUICK Tag FITC Conjugation Kit (Boehringer-Mannheim, Indianapolis, IN). Cytotuxicity rtssays. Target cells were labeled with chromium-5 1 as described (1 l), and incubated at 5: 1 E:T ratios with IL-2-activated PBLs. In some experiments, the PBLs were enriched for LGLs by Percoll density gradient centrifugation ( 12). CL1 58 or CA 19-9 antibodies were added at the beginning of the assay to the cell suspension. Blocking antibodies were usually added to the suspension 30 min prior to addition of the anti-tumor antibodies, except where indicated. Binding assays. SW948 cells were grown as monolayers for 18 hr in 96-well flatbottomed plates (Costar, Cambridge, MA) at initial plating density of 5 X lo3 cells/ well. IL-2-activated PBLs were labeled with chromium-5 1 and lo4 labeled cells were added to washed SW948 monolayers at PBL:SW948 ratios of 2:l in the presence of various antibodies. The cells were coincubated for 1 hr at 37°C in L15 medium containing 15% fetal cell serum, and the wells were then washed and lysed with 100 ~1 1 N NAOH. The lOO+l samples were analyzed in a Beckman LS4000 gamma counter. The percentage of PBLs bound to SW948 was estimated using the formula %Bound =

EC - BC x 100. TC - BC

GARCIA

340

DE PALAZZO

ET AL.

EC refers to counts per experimental well, BC refers to background counts (wells without SW948 cells), and TC refers to total counts contained in the aliquoted PBLs. All samples were analyzed in triplicate. The standard deviations of the triplicates were always less than 5%. Results did not differ with coincubations of cells for 1 or 4 hr. Flow cytometry. Cells (2-5 X 105) were incubated in l-ml vol with antibodies at 4°C for 30 min. In some experiments, second antibodies (F(ab’)z fragments of goat anti-mouse immunoglobulin (Tago, Burlingame, CA)) were used if the primary antibodies were unlabeled. Between each antibody application, tubes were washed twice with phosphate-buffered saline, pH 7.4. Cell suspensions were analyzed using a FACScan flow cytometer (Becton-Dickinson), with gates set on cells using previously established size and light scatter criteria. In each experiment, isotype-matched negative control antibodies were used. The percentage of positive cells was determined by scoring as positive all cells within the gated population with fluorescence intensity exceeding that of 97% of the negative control population (13). In competition assays, competing, unlabeled antibodies were added 30 min prior to addition of conjugated antibodies. RESULTS Inhibition

of Bispecific Antibody Promoted Lysis

The ability of several antibodies to inhibit lysis of SW948 cells by LGLs and CL1 58 was evaluated using optimal concentrations of CL1 58 at E:T ratios of 5: 1 (Table 1). Blockade of FcyRIII by either B73.1 (Leu 1 lc) or 3G8 bivalent antibody abrogated lysis augmentation by CL158 Blockade by an isotype-matched anti-CD3 antibody (Leu 4) had no effect; similarly anti-CD25, anti-&-microglobulin and anti-Leu 7 affected neither antibody-independent nor CL158dependent lysis (data not shown). However, preincubation of the cell suspension with the same concentration (20 pg/ ml) of Leu 19 antibody reversed antibody-dependent but not -independent lysis. In repeated experiments, these results were confirmed (Fig. 1) with a highly significant inhibition of CL1 58 lysis augmentation by 20 &ml Leu 19 (P < 0.05 by t test). We confirmed that SW948 cells express N-CAM (Fig. 2) and that 20 pg/ml saturated Leu TABLE Effects

of Different

Bispecific None CL158’

MAb

Antibodies

I

on Lysis of SW948 Cells by Bispecific and IL-Z-Activated Lymphocytes Blocking

MAb”

None None Leu 19 Leu 4 873.1 3G8

Monoclonal

Antibody

% Lysis’ 21.5 59.0 25.5 55.0 25.0 11.5

f t f * + +

3.5 7.1 3.5 0 1.4 2.1

” Used at 20 &nl. b Mean of two separate, representative experiments. c Coincubation of ILZ-activated PBLs and 5’Cr-SW948 at E:T ratio of 5: 1 in presence of bispecific antibody CA 19-9 X 3G8 supematant (CL 158) containing 5 ng/mI bispecific antibody by quantitative ELISA.

BISPECIFIC

ANTIBODY

No Antibody

REGULATION

OF

LGL

CL158

CYTOTOXICITY

&specific

341

MAE

FIG. 1. Inhibition of bispecific antibody-mediated lysis by Leu 19 antibody. IL-2-activated PBLs were incubated with 5’Cr-labeled SW948 cells at 5: 1 E:T ratios. The pair of bars on the left depicts target lysis in the absence (open bar) and presence (closed bar) of 20 pg/mI Leu 19 antibody. The pair of bars on the right shows results of experiments in which CL1 58 supernatants containing 5 rig/ml bispecific antibody was added to the suspension of IL-2-activated PBLs and 5’Cr-SW948 targets in otherwise identical conditions. Open bars represent no added Leu 19: closed bars show results when Leu I9 was added to the cell suspension 30 min prior to addition of CL1 58. Numbers in parentheses show the number of experiments contributing to each experimental point.

19-dim and -bright lymphocytes by flow cytometric analysis (data not shown). The inhibition of CL158-promoted lysis by Leu 19 (Fig. 3) was concentration dependent. Thus, Leu 19 exposure interferes with augmentation of LGL-mediated tumor lysis by CL158.

I&ence

cf LGL or Tumor N-CAM Blockade on Lvsis Augmentation

Since N-CAM-mediated interactions occur via homotypic (i.e., N-CAM with NCAM) binding ( 14), blockade of effecters and targets separately should inhibit binding and, hence, lysis augmentation. Table 2 shows results when either effecters or targets were preincubated with 20 pg/ml Leu 19, washed, and added to labeled targets, compared with the addition of Leu 19 to cell suspensions already containing effecters and targets. The results clearly demonstrate that blockade of LGL CD56 by Leu 19 reverses CL158 lysis augmentation as completely as the addition of the antibody to a mixed cell population. Binding of this IgG, antibody to SW948 cells had no effect on CL 158 lysis potentiation until higher concentrations of Leu 19 were used. At these higher antibody concentrations, the blocking by Leu 19 of tumor binding may be explained by dissociation of antibody from tumor-associated N-CAM and subsequent binding to LGL CD56. The possibility that CD56 occupancy by Leu 19 affected accessibility of 3G8 antibody binding to FcyRIII was evaluated as shown in Table 3. Preincubation of IL-2-activated

342

GARCIA

FIG. 2. Expression of N-CAM by antibody (thick solid line), 20 &ml CA19-9 (dashed line); goat anti-mouse antibody and CA19-9 samples, while coerythrin. Fluorescent intensity was

DE PALAZZO

ET AL.

Log Fluorescence

Channel

SW948 adenocarcinoma cells. SW948 cells were incubated with no Leu 4 (thin solid line), 20 &ml Leu 19 (dotted line), or 20 &ml immunoglobulin conjugated with phycoerythrin was used in the no the Leu 19 and Leu 4 antibodies were directly conjugated to phyanalyzed by flow cytometry as described in the text.

PBLs with 20 yg/ml Leu 19 did not alter in any way the binding of fluoresceinated B73.1 (Leu 1 lc) to FcyRIII (not shown), but completely inhibited the binding of fluoresceinated 3G8 antibody. Similarly, unlabeled 3G8 inhibited the binding of phycoerythrin-labeled Leu 19 to IL-2-activated PBLs. The possibility that these antibodies recognize cross-reactive epitopes is excluded by the observation that 3G8, which does not bind to SW948 cells, also does not inhibit the binding of phycoerythrin-labeled Leu 19 to these cells. Thus, Leu 19 inhibits the binding of 3G8 antibody to FcyRIII.

50

Leu 19 @g/ml) FIG. 3. Concentration dependence of Leu 19 inhibition of bispecific antibody-mediated tumor lysis. Varying concentrations of Leu 19 antibody were added 30 min prior to addition of 5 rig/ml CL158 supernatants to a suspension of IL-2-activated PBLs and 5’Cr-labeled SW948 cells at E:T ratio of 5: I.

BISPECIFIC

ANTIBODY

REGULATION TABLE

Inhibition

of Antibody-Augmented

Experimental

OF

LGL

2

Lysis by Leu 19 via Interaction

conditions

LAK alone” LAK + CL158’ LAK + CL158 + Leu 19d Leu 19 added to Total cell suspension LAK only “Cr-SW948 only Leu 19 20 @g/ml Leu I9 100 fig/ml

with

7~ Specific “0-SW948

of Timing of FcyRIII

Lymphocytes lysis of cells

39.3 k 2.3” 81.3 k 10.7

35.3 + 32.3 k

7.5 5.5

75.7 + 11.9 38.0 f 8.5

’ “Cr-SW948 cells incubated with 5: I E:T ratios of IL-2 activated PBLs in assays. ’ Means k SEM for three representative experiments. ’ Clonal supernatant containing 5 rig/ml final assay concentration of CA 19-9 X antibody. ’ 20 pg/ml final assay concentration of Leu 19 antibody was added as described. with either effector or targets. the cells were washed and then incubated with the for 4 hr, followed by determination of label release.

&f&s

343

CYTOTOXICITY

4 hr chromium-5

3G8 bispecific

1 release

monoclonal

After Leu 19 was incubated unexposed cell population

Blockade on Lysis

Figure 4 shows the results of two representative experiments demonstrating that NCAM blockade only inhibits lysis if used relatively early following mixing of effector and target cells. In each experiment, substantial augmentation of lysis with CL 158 was completely inhibited if 20 pug/ml Leu 19 were added within the first 30 min of the start of the assay. If Leu 19 was added 60 or more min after the start, no effect on lysis was observed. These results were consistent with early, N-CAM-mediated LGL binding to SW948, despite the apparent lack of Leu 19 binding to the target cell. These experiments show that the effects of Leu 19 on CL158 lysis augmentation are directed against LGLs.

Binding of IL-2-Activated

PBLs and LGL to SW948 Cells

Two representative experiments are depicted in Table 4. In the first, IL-2-activated, “Cr-labeled PBLs binding to SW948 was increased from 22 to 43% with the addition of CL158. In this and all binding assays, visual confirmation of binding preceded detergent lysis of the well. Addition of Leu 19 decreased binding in a concentrationdependent manner. In the second experiment, IL-2-activated PBLs enriched for LGL content (60% CD56+ cells by flow cytometry) were used. Again, Leu 19 concentrationdependent inhibition of binding was observed; the parental IgG, antibodies CA 19-9 and 3G8 did not promote effector-target interactions.

Effticts of FcyRIII

3G8 Epitope Blockade on Conventional ADCC

To determine if the influence of Leu 19 on lysis is unique to BsMAbs with specificity for the 3G8 epitope of FcyRIII, the effects of Leu 19 on CL1 58 and CA19-9 IgG,,-

344

GARCIA DE PALAZZO

ET AL.

TABLE 3 Competitive Inhibition by Leu 19 of 3G8 Antibody Binding to Human LAK Cells Percentage of control binding” Cells Human LAKs

SW948

Unlabeled antibody Ox/ml)

3G8-FITC

Leu 19-PE

None Leu 19 (1) (10) (100) (1000) 3G8 (1) (10) (100) ( 1000) MOPC-2 1 (1) (10) (100) ( 1000) Leu 19 (10,000) 3G8 (10,000) MOPC-2 1 ( 10,000)

100 75 55 21 0 4.5 2.3 2.3 0 15 100 100 93 N.D.’ N.D. N.D.

100 54 0 0.1 0 100 0 0 0 83 81 92 12 0.2’ 114 100

a Cells were incubated for 30 min in varied concentrations of unconjugated antibodies, followed by 10 pg/ml labeled antibodies, as described in text. Representative experiment from series of three. Typically, 15% of human LAKs and 60% of SW948 cells bound Leu 19-PE and 5-10% of human LAKs bound 3G8FITC. Binding was determined by the percentage of cells with fluorescence greater than 97% of negative control population. ’ N.D., not determined. ’ 50% inhibition of Leu 19-PE binding at 100 rig/ml concentration of unconjugated Leu 19 antibody.

promoted lysis were studied. Table 5 shows the results of one of three representative experiments. At antibody concentrations promoting equivalent lysis augmentation, Leu 19 inhibited only the CL1 58 effects. The lack of Leu 19 influence on CA 19-9 IgG,,-promoted ADCC was apparent even at lower CA 19-9 concentrations that promote suboptimal ADCC. Thus, there is no evidence that N-CAM interactions become important in the setting of low anti-tumor antibody concentrations that engage a suboptional number of Fcr receptors on the effector cell surface. Instead, N-CAM molecule interactions are apparently required only for bispecific antibody-mediated lysis when the effecters and targets express N-CAM. The effects are not limited to SW948 cells since Leu 19 inhibition of CLl58-promoted lysis is also seen with the colonic adenocarcinoma cell line SW 1116 (data not shown). DISCUSSION Bispecific monoclonal antibodies which promote tumor lysis by LGLs via interactions with tumor antigens and FcyRIII have been shown previously to mediate these effects in the presence of competing human immunoglobulin. Thus, a role for conventional ADCC by these antibodies is unlikely (6). The present study shows that FcyRIII engagement is required for these bispecific antibodies to exert their effects

BISPECIFIC

ANTIBODY

REGULATION

Time of Leu 19 Addition

OF

LGL

CYTOTOXICITY

345

(minutes)

FIG. 4. Timing dependence of Leu 19 inhibition of SW948 cell lysis by CL1 58-directed lymphocytes. The results of two separate representative experiments evaluating the influence of timing of Leu I9 addition to antibody-directed tumor lysis are shown. IL-2-activated PBLs (solid circles) or LGLs (80% purity, open squares) were incubated with 5’Cr-labeled SW948 cells at 5: I E:T ratios in the presence or absence of 5 ng/ ml CL158 supernatant. When Leu 19 was not added, antibody-independent lysis was I6 and 32% and antibody-dependent lysis was 75 and 95% in the respective assays. Leu I9 antibody was added at the times indicated following the initiation of coincubation of effecters and targets and the percentage of SW948 cell lysis was calculated at 4 hr as described in the text.

since lysis potentiation by CL1 58 is competitively inhibited by its bivalent anti-FcyRIII antibody parent 3G8. While these studies usually employed bulk IL-2-activated lymphocytes, both the binding (Table 4) and lysis (Fig. 4) experiments demonstrated that these inhibitory effects are directed against LGLs. The competition between 3G8 and Leu 19 antibodies suggests the possibility of steric hindrance and that the Leu 19 epitope of N-CAM and 3G8 epitope of FcyRIII may be spatially juxtaposed on the cell membrane of LGLs. The binding assay data (Table 4) are of interest as they demonstrate engagement of additional LGLs when bispecific antibody is added to a culture containing IL-2activated LGLs and tumor. Thus, the bispecific antibody does not merely increase the avidity of binding of LGLs to already bound tumor. Additional experiments will be required to determine if the LGLs recruited to bind to tumor by bispecific antibody differ phenotypically from those which bind nonspecifically. It is possible that expression of FcyRIII by LGLs will prove to be a key determinant in this process and that separate LGL populations, discriminated by FcyRIII expression, mediate the LAK and antibody-dependent mechanisms of tumor lysis. Although FcyRIII-expressing cells may express CD3 antigen, blockade of CD3 did not alter CL158-promoted lysis. This does not support a role for T cells in the lysis promoted by CL 158 bispecific antibody. Other cytotoxic trigger molecules, such as CD2 ( I5), require evaluation as coordinate-activating structures for lysis triggered via FcyRIII. The results shown in Table 2 could be interpreted as showing that blockade of LGL-associated and tumor-associated N-CAM leads to inhibition of antibody-de-

GARCIA DE PALAZZO

346

ET AL.

TABLE 4 Anti-N-CAM

Antibody Effects on Binding of Large Granular Lymphocytes to Tumor

Experiment 1

% 5’Cr label bound to SW948 cells

Conditions

22 43

PBL” PBL + BsMAb PBL + BsMAb + Leu 19 5 be/ml 50 fig/ml LGLb LGL + BsMAb LGL + BsMAb + Leu 19 5 a/ml 25 &c/ml 50 fig/ml LGL + CAl9-9 LGL + 3G8

36 21 16 71 73 20 16 16 13

n IL-2-activated PBLs containing 15-20% LGLs by flow cytometry. b IL-2 activated LGLs (60% CD56+) enriched by Percoll gradient density centrifugation.

pendent lysis; these data also are consistent with current evidence that N-CAM interactions involve homotypic binding ( 14). However, the very low Leu 19 concentrations required to inhibit binding of 3G8 antibody to FcyRIII might have been present even after washing Leu 19-prepared SW948 cells. Although N-CAM regulation of CL1 5% promoted lysis is based on shared reactivity of the 3G8 and Leu 19 epitopes, the roles of cell adhesion molecules in regulating the migratory and lytic functions of lymphocytes are receiving increasing attention. Much has been learned about CAM interactions and T-lymphocyte function ( 17-19), but relatively little is known about CAM distribution and function in LGLs. Aside from N-CAM (2), LGLs are known to express laminin receptors (16), and a subset of LGLs expresses CD2 (15). Since identification of a universal LGL antigen receptor or recognition complex has proven elusive, it is reasonable to speculate that these promiscuously lytic cells bind to diverse targets via TABLE 5 Effects of N-CAM Blockade on Lysis Promoted by CLI 58 and the IgG,, Variant of CA 19-9 % Lysis of 51Cr-SW948 Antibodv None CL158 CAl9-9

Concentration (&ml) 0 12.5 1.25 300 30 3.0

With Leu 19”

Without Leu 19

20 27 27 70 45 24

21 70 91 70 46 23

a 20 rg/ml was added prior to addition of anti-tumor antibody.

BISPECIFIC

ANTIBODY

REGULATION

OF

LGL

CYTOTOXICITY

341

multiple CAM, with lysis dependent on the cumulative strength of binding, engagement of specific target structures, activation status of the LGLs, or, more likely, a combination of these factors. FcyRIII interactions provide additional points of contact, as well as triggers to initiate lysis (20). These studies have not identified molecules other than FcyRIII which regulate CL 158 lysis, although the roles of additional molecules need to be determined. Further study of the roles of diverse CAMS in facilitating FcyRIIItriggered lysis will provide additional insights into these considerations. The present study illustrates the complexity of investigating the contributions of various molecules to cell adhesion and lysis. The novel and unexpected finding that Leu 19 antibody inhibits FcyRIII-mediated lysis by LGLs originally implicated NCAM-based regulation of LGL binding to N-CAM-expressing tumor. These results underscore the need to consider multiple mechanisms by which antibodies to cell adhesion molecules may affect cellular function. ACKNOWLEDGMENTS This work was supported by National Cancer Institute Awards ROl CA50633, CA06927, and support by an appropriation from the Commonwealth of Pennsylvania. The authors gratefully acknowledge the expert secretarial assistance of Catherine Thompson.

REFERENCES 1. Philips, J. H., and Lamer, L. L., J. Exp. Med. 15, 1823, 1982. 2. Lanier, L. L., Testi, R., Bindl, J., and Phillips, J. H., J. hp. Med. 169, 2233, 1989. 3. Ortaldo, J. R., Woodhouse, C., Morgan, A. C., Herberman, R. B., Cheresh, D. A., and Reisfeld, R., J. Immunol. 138, 3566, 1987. 4. Munn, D. H., and Cheung, N. V., Cancer Rex 47,6600, 1987. 5. Weiner, L. M., Zarou, C. M., O’Brien, J., and Ring, D., J. Biol. Response Modif: 8, 221, 237. 1989. 6. Garcia de Palazzo, I. E., Gercel-Taylor, C., Kitson. J., and Weiner, L. M., Cancer Res. 50, 7 123, 1990. 7. Weiner, L. M., Moldofsky, P. J., Gatenby, R. A., O’Dwyer. J., O’Brien, J.. Litwin, S., and Comis, R. L.. Cancer Res. 48, 2568, 1988. 8. Boyum, A., Scund. J. Clin. Luh. Inve.vt. 21(Suppl. 97), 7, 1968. 9. Koprowski, H., Steplewski, Z., Mitchell, K.. Herlyn, M., Herlyn, D., and Fuhrer, J. P., Somutic Gel/ Genet. 5, 957, 1979. 10. Fleit, H. B., Wright, S. D., and Unkeless, J. C., Proc. Natl. Acad. Sci. USA 79, 3275, 1982. I 1. Cerottini, J-C., and Brunner, T., In “In Vitro Methods in Cell-Mediated Immunity” (B. R. Bloom, and P. R. Glade, Eds.). Academic Press, New York, 1971. 12. Timonen, T., Ortaldo, J. R., and Herberman, R. B., J. hp. Med. 663, 569, 198 I, 13. Weiner, L. M., O’Dwyer, J., Kitson, J., Comis, R. L.. Frankel, A. E., Bauer, R. J., Konrad, M. S., and Groves, E. S., Cancer Rex 49, 4062, 1989. 14. Cunningham, B. A., Hemperly, J. J., Murray, B. A., Prediger, E. A., Brackenbury, R., and Edelman, G. M.. Science 236, 799, 1987. 15. Ritz, J., Schmidt, R. E., Michon, J., Hercend, T., and Schlossman, S. F., In “Advances in Immunology” (F. Dixon, Ed.), Vol. 42, pp. 181-21 I. Academic Press, San Diego, 1988. 16. Hiserodt. J. C., Laybourn, K. A., and Varani, J., J. Immunol. 135, 1484, 1985. 17. Springer, T. A., Nature 346(6283), 425, 1990. 18. Patarroyo, M., Prieto, J., Rincon, J., Timonen, T., Lundberg, C., Linbom, L., Asjlo, B., and Gahmberg. C. G., Immtmol. Rev. 114, 67, 1990. 19. Singer, K. H., J. Leukocute Biol. 48(4), 367, 1990. 20. Unkeless, J. C., Scigliano, E., and Freedman, V. H., Annu. Rev. Immunol. 6, 251, 1988.