Functional analysis of monkey lymphocyte subsets defined by OKT4 and OKT8 monoclonal antibodies

CELLULAR

IMMUNOLOGY

77, 338-347 (1983)

Functional Analysis of Monkey Lymphocyte Subsets Defined by OKT4 and OKT8 Monoclonal Antibodies LOUIS N. MARTIN,

BOBBY J. GORMUS,

AND BRIAN

E. BOZELKA

Microbiology Department, Delta Regional Primate Research Center, Tulane University, Covington, Louisiana 70433, and Clinical Immunology Section, Tulane University School of Medicine, New Orleans, Louisiana 701 I2 Received January 7, 1983; accepted January 24, I983 The percentages of rhesus monkey blood lymphocytes (PBL) reactive with OKT4 and OKT8 antibodies and the OKT4/OKT8 ratio showed significant correlations with the log of the immunoglobulin plaque-forming cell (PFC) response after stimulation with pokeweed mitogen (PWM). These correlations suggested that monkey OKT4+ cells function as “helper” cells and OKT8+ cells function as “suppressor” cells for the PFC response. This was confirmed by separation and study of enriched T- and B-cell subpopulations. OKTIdepleted (OKT4+) and OKT4depleted (OKT8+) cells were obtained by treatment of purified T cells with antibody and complement. OKT4+ cells augmented the PWM-induced B-cell differentiation into PFC but OKT8+ cells did not. OKT8+ cells suppressed the PFC response by mixtures of B cells and OKT4+ cells. OKT8 antibodies also detected a suppressive cell subset in African green monkeys since the percentage of OKT8+ cells showed a negative correlation with the log PFC response. OKT4 antibodies failed to bind to African green monkey PBL. INTRODUCTION

Immune activities are regulated in part by T-lymphocyte subsets with opposing functions. Monoclonal antibodies are capable of differentiating between lymphocyte subsets which include regulatory cells that mediate help or suppression of immune functions. The OKT4 monoclonal antibodies react with 50 to 60% of human T cells, including cells which mediate help for B-cell and cytotoxic T-cell differentiation (l-3). The OKTS monoclonal antibodies react with 30 to 40% of T cells including cells with cytotoxic effector function and cells capable of suppressing B-cell differentiation (3, 4). OKT4+ helper T cells are required for pokeweed mitogen (PWM)‘stimulated B cells to differentiate into plaque-forming cells (PFC) producing immunoglobulin (3, 5). OKTV T cells cannot provide help for PWM stimulation of PFC differentiation but can suppress the production of PFC in PWM-stimulated cultures of B cells plus OKT4+ T cells (3, 5). ’ Abbreviations used: AET, S-(2-aminoethyl)isothiouronium bromide hydrobromide; BAy, BAr, BC, Salmonella typhimurium bacteria (B) coated respectively with IgG antibody, IgM antibody, or complement (C); BSA, bovine serum albumin; FCS, fetal calf serum; MNC, blood mononuclear cells isolated on FicollHypaque; PBL, adherent cell-depleted MNC; PFC, plaque-forming cells; PWM, pokeweed mitogen; R, cell membrane receptor; SIg, surface immunoglobulin; SRBC, sheep erythrocytes. 338 0008-8749/83 $3.00 Copyright 0 1983 by Academw Press. Inc. All rights of reproduction in any form reserved

MONKEY

LYMPHOCYTE

SUBSET

FUNCTION

339

Certain monoclonal antibodies that react with human lymphocyte subsets also react with lymphocytes from some nonhuman primate species (6-9). In the present study, we provide evidence that OKT4 and OKT8 antibodies react with rhesus monkey T-cell subsets that respectively have helper or suppressor functions similar to their human counterparts in the PWM-induced PFC assay. In addition, evidence is presented that African green monkey cells reactive with OKT8 antibodies function as suppressor cells in that species; OKT4 antibodies do not appear to recognize African green monkey T cells. MATERIALS

AND

METHODS

Monkeys. Twenty-one young rhesus monkeys (Macaca mulatta) aged 1 to 3 years, 7 adult rhesus (>4 years old) and 16 adult African green monkeys (Cercopithecus aethiops) were studied. Blood samples were obtained by femoral venipuncture from animals anesthetized with ketamine hydrochloride (Bristol Laboratories, Syracuse, N.Y .). Lymphocyte preparation. Blood samples were anticoagulated with sterile heparin (10 U/ml), and mononuclear cell (MNC) preparations enriched for peripheral blood lymphocytes were separated by a modification of the Ficoll-Hypaque procedure (10, 11). The MNC preparations which were obtained were greater than 98% viable by trypan blue exclusion and contained greater than 95% lymphocytes in differentially stained preparations. The MNC were suspended to 5 X 106/ml in RPM1 1640 culture medium (Grand Island Biological Co., Grand Island, N.Y.) supplemented with 10% heat-inactivated fetal calf serum (FCS). Adherent cells were removed by overnight incubation in 60-mm plastic dishes (Falcon 3002, Oxnard, Calif.) at 37°C in a 5% CO* humidified incubator. Analysis of 0KT-P and OKTF cells. The adherent cell-depleted MNC (PBL) were resuspended to 1O’/ml in RPMI- 10% FCS containing 0.1% NaN3, and 0.1 -ml aliquots were reacted with 5 ~1 OKT4 or OKT8 antibodies (Ortho Diagnostic Systems, Raritan, N.J.) at 4°C for 30 min. The cells were washed twice in 2 ml RPM1 1640 containing 5% FCS and 0.1% NaN3. The cell pellets were resuspended in 0.1 ml RPM1 and stained at 4°C for 30 min with 0.1 ml of a 1:lO dilution of fluoresceinconjugated goat anti-mouse IgG (Cappel Laboratories, Cochranville, Pa.). This dilution was determined to be optimal in preliminary experiments. The stained cells were washed twice in 2 ml RPM1 1640 containing 5% FCS and 0.1% NaN3, resuspended in 0.5 ml RPM1 1640 with no serum additive, and fixed by adding 0.5 ml 2% paraformaldehyde in phosphate-buffered saline. Control cells were treated with normal mouse serum diluted 1: 100 in place of the OKT4 or OKT8 antibodies. The fixed cells were stored at 4°C for no more than 2 days before analysis by flow cytometry on an FACS III (Becton-Dickenson, Sunnyvale, Calif.). The percentage of stained cells in the lymphocyte scatter peak was determined by analyzing at least 10,000 cells. Analysis of lymphocyte surface markers. (a) T-cell percentages were determined by rosette formation with sheep erythrocytes (SRBC) modified by treatment with S-(2aminoethyl)isothiouronium bromide hydrobromide (AET) by the method of Kaplan and Clark (12) as we have previously described (13). (b) Surface immunoglobulin (SIg) bearing PBL were quantified by direct immunofluorescent staining at 4°C with fluorochrome-conjugated F(ab’)z polyvalent antiserum (Cappel Laboratories, Coch-

340

MARTIN

ET

AL.

ranville, Pa., Kallestadt Laboratories, Chaska, Minn.) using cells incubated and washed at 37°C prior to staining (14). (c) Fey and Fcp receptors (R) were determined by rosette formation with heat-killed fluorochrome-labeled Salmonella typhimurium bacteria (B) after sensitization with rabbit anti-S. typhimurium IgG or IgM antibody (BAT or BAp, respectively). Details of this procedure will appear elsewhere. Briefly, B were sensitized by incubation at 37°C for 45 min with subagglutinating amounts of purified rabbit anti-B IgG or IgM antibody. After washing, these complexes were resuspended to lo8 per ml in cold RPM1 1640- 1% BSA; 0.1 ml of BAT or BAp was then mixed with 0.1 ml of cold PBL at 10 X lo6 PBL/ml of RPM1 1640-l % BSA. After 30 min of incubation in ice, unbound BAT or BAp were removed by two centrifugation washes at 125g. Pelleted PBL were resuspended and fixed in 2.5% glutaraldehyde. Cells were then cytocentrifuged, fixed, coverslipped, and viewed on a fluorescence microscope. Rosetted cells were those binding three or more BAT or BAp. (d) PBL C3R were detected by rosette formation with killed, fluorochromelabeled B sensitized with complement (C) by incubation with a 1:5 dilution of fresh mouse serum. The resultant BC complexes were washed, resuspended to lo8 BC/ml, and used to rosette PBL by methods analogous to those described above for BAy and BAp. The BC procedure represents a modification of our previously reported BAC method ( 15) and a modification of a procedure reported by Gelfand et al. ( 16) for the detection of C3R. PWM stimulation. MNC were cultured at a density of 106/ml in RPM1 1640 supplemented with 15% heat-inactivated FCS, penicillin ( 100 pg/ml), streptomycin (100 U/ml), and L-glutamine (2 mM). One-milliliter cultures in 12 X 75-mm plastic tubes (Falcon 2054) were incubated at 37°C in a humidified atmosphere of 5% CO2 in air. Experimental cultures received PWM (Grand Island Biological Co.) at the optimal final dilution of 1: 125; control cultures received no mitogen. PFC assay. The plaque assay with protein A-coated SRBC was modified from the procedure of Lanzavecchia et al. ( 18). SRBC were treated with CrC13 and were coated with staphylococcal protein A (17). The protein A was prepared from the Cowan I strain (American Type Culture Collection, Rockville, Md.) as described (19). Cultured MNC were washed three times in RPM1 1640 containing 1% bovine serum albumin (RPMI-BSA) and serial dilutions were prepared. Diluted cell aliquots (0.2 ml) were warmed to 40°C and mixed with 0.25 ml of 1.5% agarose (Sea-Plaque, Marine Colloids, Rockland, Maine) in RPM1 1640 and with 0.1 ml of 6% protein A-coated SRBC in RPMI-BSA. The mixtures were spread on glass slides (0.25 ml on each slide) previously treated with 0.1% agarose, and the agarose-cell mixtures were allowed to harden for 2 min. The slides were inverted over a 2-mm-deep chamber filled with RPMI-BSA containing a final dilution of 1: 14 guinea pig complement (Grand Island Biological Co.), previously absorbed at 4°C with 10% packed SRBC, and 1:50 diluted rabbit anti-rhesus monkey IgG (20) previously heat inactivated at 56°C and absorbed with SRBC. The anti-rhesus IgG reacted in immunoelectrophoresis with Fc and Fab fragments of IgG. The slides were incubated for 3 hr at 37°C in a humidified 5% COZ in air atmosphere. Plaques were counted in indirect light. The results were corrected for the dilution factor and expressed as the number of PFC/ lo6 cells cultured. Preliminary experiments with the PWM-induced PFC assay showed that stimulation with PWM at a final dilution of 1: 125 for 6 days produced optimal PFC responses for most individuals. A final anti-IgG dilution of I:50 was optimal for the detection of PFC.

MONKEY

LYMPHOCYTE

SUBSET

F-UNCTION

341

Separation of T and B cells. T and B cells were separately enriched based on the ability of T cells to form rosettes with SRBC ( 14). MNC at 1 X 107/ml in RPM1 1640 containing 20% heat-inactivated, SRBC-absorbed FCS were mixed with an equal volume of 1% AET-SRBC in the same medium, and the mixture was incubated at 37°C for 15 min. The cells were centrifuged at 2008 for 10 min and were held in pellets overnight in ice. One-half of the supernatant was discarded and the pellets were gently but thoroughly resuspended. The cells were layered over 12 ml FicollHypaque (Pharmacia) in 50-ml polycarbonate (Oak Ridge) centrifuge tubes, adding no more than 4 X lo7 resuspended MNC to each tube. The tubes were centrifuged at 400s for 30 min at 22°C. The B-cell-enriched interface cells were washed three times, resuspended at 107/ml, and held at 4°C until use in PFC cultures. The pellets, containing SRBC-rosetted T cells, were washed twice and the SRBC were lysed by resuspending the pellet in fresh autologous serum at 37°C (fresh serum from most adult monkeys contains enough natural hemolysin and complement to lyse the SRBC within 2 min). The enriched T cells were washed three times and were resuspended at 5 X lo6 cells/ml. Enrichment of OKT4+ and OKT8+ T cells. Enriched T cells were depleted of OKT8+ cells by incubating 5 X lo6 cells/ml for 30 min at 25°C with a final dilution of 1:lOO OKT8 antibody, followed by a 75-min incubation with sufficient rabbit complement added to make a final dilution of 1: 10 (3, 5). The OKT8-depleted cells were termed OKT4+. OKT8+ cells were similarly prepared by depletion of enriched T cells with OKT4 antibody and complement. In three experiments using cells from adult rhesus monkeys, the average percentage of PBL killed by OKT4 and complement was 5 1% and the average percentage killed by OKT8 and complement was 35%. The OKT4+ cells, OKT8+ cells, and the enriched B cells were analyzed for reactivity with monoclonal antibodies and for other lymphocyte surface markers. Functional assessment of subset-enriched lymphocyte preparations. The B cells, OKT4+ cells, and OKT8+ cells or various combinations were stimulated with PWM for 6 days. Thereafter, these cultures were examined by the PFC assay to determine the functional capabilities of the OKT4+ and OKT8+ T-cell subsets. Statistical analyses. Regression lines by the method of least squares, correlation coefficients (r), Student’s t tests for correlations and group means, and standard errors (SE) were calculated using the Wang 600 programmable calculator. RESULTS Correlations between T-lymphocyte subsets and PWM-induced PFC responses. The percentages of OKT4+ and OKT8+ cells were determined on adherent cell-depleted MNC (PBL) samples from 2 1 normal rhesus monkeys aged 1 to 3 years. The PWMinduced PFC responses were determined for each MNC preparation. The means + SE were: percentage OKT4+ cells, 28.8 + 1.2; percentage OKT8+ cells, 39.9 + 1.5; and PFC per lo6 cells cultured with PWM, 452 + 99. The mean + SE for the ratio of OKT4+ cells to OKT8+ cells was 0.75 f 0.05. There was a significant positive correlation between the percentage of OKT4+ cells and the log PFC response (Fig. 1) and a significant negative correlation between the percentage of OKT8+ cells and the log PFC response (Fig. 2). There was also a significant positive correlation between the OKT4/OKT8 ratio and the log PFC response (Fig. 3). These correlations indicated that the OKT4+ cells may function as

342

MARTIN

00

Log

PFC

ET AL.

per

Culture

FIG. 1. Positive correlation between the percentage of rhesus monkey OKT4* cells and the log PFC response (r = 0.4761; P < 0.05).

“helper” cells and the OKTS+ cells may function as “suppressor” cells for the PWMinduced differentiation of rhesus monkey B cells into functional immunoglobulinsecreting PFC. ’ Enrichment of lymphocyte subsets. Preparations enriched for the different lymphocyte subsets were separated in order to study the functional capability of rhesus monkey OKT4+ and OKT8+ cells. The surface markers expressed by unfractionated MNC are compared to the surface markers expressed by the lymphocyte subset-

OI

1.0 Log

PFC

per

2.0 Culture

3.0

FIG. 2. Negative correlation between the percentage of rhesus monkey OKT8+ cells and the log PFC response (r = 0.6697; P -c 0.005).

MONKEY

LYMPHOCYTE

1.0 Log

PFC

343

SUBSET FUNCTION

per

2.0 Culture

FIG. 3. Positive correlation between the OKT4/OKT8 < 0.00 I) in rhesus monkeys.

3.0

ratio and the log PFC response (r = 0.6862; P

enriched cell preparations in Table 1. The OKT4+ preparation was enriched for cells: reactive with OKT4 (75%) forming rosettes with AET-SRBC (84%), and bearing receptors for BAp ( 13%). The OKT4+ preparation was depleted of cells: reactive with OKT8 (16%), bearing SIg (2%), and bearing receptors for BAT (6%). In contrast, the OKT8+ preparation was enriched for cells: reactive with OKT8 (77%), forming rosettes with AET-SRBC (84%), and bearing receptors for BAy ( 15%); and was depleted of cells reactive with OKT4 (3%). The percentage of cells bearing receptors for BAp was low in the OKT8+ preparation (7%) relative to the percentage in the OKT4+ subset (13%). The B-cell preparation was enriched for cells with SIg (47%), receptors for BA-y (28%), and receptors for BC (32%), but was depleted for cells reactive with OKT4 (3%) or OKT8 (13%) and for cells forming rosettes with AET-SRBC (9%). TABLE 1 Surface Markers of Rhesus Monkey MNC and of Enriched Lymphocyte Subsets Percentage of cells with markerb Cell preparation’ MNC OKT4+ OKT8+ B cells

OKT4

OKT8

AET-SRBC rosettes

Sk

B&J rosettes

BAY rosettes

BC rosettes

34 75 3 3

45 16 77 13

70 84 84 9

14 2 ND’ 47

7 13 7 8

12 6 15 28

10 6 11 32

’ Ficoll-Hypaque-isolated mononuclear cells (MNC) were depleted of adherent cells, B cells were enriched by depletion of T cells forming rosettes with AET-SRBC, and T cells were enrichedCfor OKT4’ or OKT8+ cells by treatment with monoclonal antibody and complement. b Data represent the averages of two experiments. ‘ND, not determined.

344

MARTIN

ET AL.

These results indicated that the cells reactive with OKT4 for the most part comprised a different and separable T-cell subset from the cells reactive with OKT8. Functional d@erences between OKT4’ and OKTS’ cells in the PWM-induced PFC response. The culturing of 2 X lo5 B cells with PWM for 6 days did not result in a PFC response above unstimulated controls in two experiments (average, 10 PFC/ culture). The addition of OKT4+ cells to the B cells in these cultures resulted in a dose-dependent increase in the PFC response (Fig. 4a), although OKT4+ cells alone 300.

a.

200. 0 ; I ; % 0 L

OKT 8+

”

T Cells

OL

(X

lo-‘)

2.0 OKT .Y+ Cells

0.5 Added lo 2 X 10’

(X

4.0 164 Added

_

1.0 B Cells

6.0 to Cultures

6.0

FIG. 4. Helper and Suppressor activity of rhesus monkey T-lymphocyte fractions. Each point represents the average of two experiments. (a) Capacity of OKT4+ or OKT8+ cells to enhance PFC production by 2 X IO5 B cells. (b) Suppression by OKTS’ cells of PFC production by 2 X IO5 B cells plus 1 X 10’ OKT4+ cells.

MONKEY

LYMPHOCYTE

SUBSET FUNCTION

345

were unresponsive ( 10 PFC/culture). The addition of OKT8+ cells to B cells resulted in only a minimal increase in the PFC response (30 PFC/culture). These results indicated that OKT4+ but not OKT8+ cells were capable of providing help for the PWM-induced differentiation of rhesus monkey B cells. When OKT8+ cells were added to a responsive combination of 2 X lo5 B cells plus 1 X lo5 OKT4+ cells, a marked suppression of the PFC response resulted (Fig. 4b). In contrast, additional increases in the number of OKT4+ cells cultured with 2 X 10’ B cells resulted in increased PFC responses (not shown). These results indicated that OKT8+ cells were capable of suppressing the PWM-induced PFC response generated by rhesus B cells combined with OKT4+ T cells. Correlation between the percentage of OKT8’ African green monkey cells and the PWM-induced PFC response. It was of interest to determine whether subset percentages were correlated with the PWM-induced PFC response in other nonhuman primates. Sixteen adult African green monkeys were studied. Cells from this species did not react with OKT4 antibodies but did react with OKT8. The mean + SE for the percentage of OKT8+ cells in adherent cell-depleted MNC from the 16 monkeys was 49.4 + 3.3, and the mean ? SE for the PFC response by PWM-stimulated MNC was 1422.8 + 365.0. The percentage of OKT8+ cells was negatively correlated with the log PFC response (Fig. 5). These results indicated that OKT8+ cells function as suppressor cells in African green monkeys. DISCUSSION Previous studies have shown that certain monoclonal antibodies reactive with human T-cell subsets also react with nonhuman primate lymphocytes (6-8) and that nonhuman primates may be useful for determining the in vivo effects of the administration of monoclonal antibodies (7, 2 1). The present results indicate that rhesus monkey lymphocytes reactive with OKT4 or OKT8 monoclonal antibodies represent helper- and suppressor-T-cell subsets, respectively. The percentage of OKT4+ cells and the OKT4/OKT8 ratio showed sig-

FIG. 5. Negative correlation between the percentage of OKT8+ cells and the log PFC response by African green monkey cells (r = 0.5554, P c 0.05).

346

MARTIN

ET

AL.

nificant positive correlations with the log of the PWM-induced PFC response, while the percentage of OKT8+ cells showed a significant negative correlation with the log PFC response. OKT4+ T cells augmented the PFC response by B cells, but OKT8+ T cells suppressed the PFC response. Although no cells reactive with OKT4 were detectable in African green monkey PBL, the percentage of OKT8+ cells was negatively correlated with the log PFC response. These results, which show that the functions of cells with antigenic determinants reactive with OKT4 or OKTS are conserved among nonhuman primates, support the view that those molecules may confer a selective evolutionary advantage (8). Other studies with rhesus monkey cells have shown that T cells selected on the basis of a surface Fc receptor for IgM were almost completely depleted of OKT8+ cells (1.8%) and had decreased OKT4+ cells (29.7%) compared to percentages obtained for unfractionated T cells (OKT8+ = 33.3% and OKT4+ = 50.4%). T cells selected for the surface Fc receptor for IgG had a lower frequency of OKT4+ cells (9.1 W) and were almost completely depleted for OKT8+ cells (1.8%) (6). These results indicated that the distributions of receptors for IgM or IgG did not completely correspond to the distributions of the OKT4 or OKT8 markers. On the other hand, depletion of T cells which had Fc receptors for IgG resulted in a cell preparation with high numbers of OKT4+ cells (5 1.5%) and low numbers of OKT8+ cells (5.5%). The results of the present studies indicate that the enrichment of rhesus monkey OKT4+ T cells resulted in a concomitant enrichment for BAp rosette-forming T cells, and the enrichment for OKT8+ T cells led to enrichment for BAy rosette-forming T cells. In humans, T cells with receptors for the IgM-Fc fragment have been associated with helper function, and T cells with receptors for the IgG-Fc fragment have been associated with suppressor function (22, 23). The present results in rhesus monkeys are consistent with those observations in humans. The formation of rosettes with BAT, BAp, or BC provides a convenient method for the detection of cells with surface Fc receptors for different immunoglobulin classes or complement (B. J. Gormus, unpublished). In the present studies, our assay for the PWM-induced PFC response indicated a mean PFC response of 452 f 99 for rhesus monkeys 1 to 3 years old. Using a similar assay in humans, others have indicated a substantially higher PFC response (18, 24, 25). We have recently observed significantly higher (P < 0.05) PFC responses (mean f SE = 2407 2 861) among rhesus monkeys 4 years or older. These older monkeys had significantly higher (P < 0.025) percentages of OKT4 reactive cells (mean f SE = 37.0 t 2.9) compared to monkeys aged 1 to 3 years (mean 1 SE = 28.8 + 1.2). The older monkeys also tended to have higher OKT4/OKT8 ratios (mean f SE = 0.99 4 0.10) compared to the younger monkeys (mean f SE = 0.75 + 0.05). Other studies have also indicated higher OKT4/OKT8 ratios for adult rhesus monkeys (6, 8). Recent investigations with human cells have shown that PWM-induced PFC responses are quite low in human newborns, and that the responses increase slowly over a period of more than 12 years before adult-level responses occur (26). Our results suggest that rhesus monkeys also have a prolonged maturation period for the PFC response. ACKNOWLEDGMENTS This work Lia Pedroza

was supported of the Harvey

in part by USPHS Grants RR00 164, AI 19302, and RR05444. We thank Cell Sorter Facility for essential help in using the fluorescence-activated

Mrs. cell

MONKEY

LYMPHOCYTE

SUBSET FUNCTION

347

sorter, Mrs. Cynthia Trygg, Mrs. Patricia McNease, and Miss Renee Bruno for expert technical assistance, Mrs. Mary Ann Quiroz for secretarial assistance, and Mr. Murphy Dowouis for graphics.

REFERENCES 1. Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman, S. F., Proc. Nat. Acad. Sci. USA 76, 4061, 1979. 2. Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman, S. F., J. Immunol. 123, 2894, 1979. 3. Thomas, Y., Sosman, J., Irigoyen, O., Friedman, S. M., Kung, P. C., Goldstein, G., and Chess, L., J. Immunol.

125, 2402,

1980.

4. Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman, S. F., J. Immunol. 124, 130, 1980. 5. Miyawaki, T., Nagaoki, T., Yokoi, T., Yachie, A., Uwadana, N., and Taniguchi, N., J. Immunol. 128, 899, 1982. 6. Ellingsworth, L. R., and Osbum, B. I., Vet. Immunol. Immunopathol. 2, 541, 1981. 7. Cosimi, A. B,, Burton, R. C., Kung, P. C., Colvin, R., Goldstein, G., Lifter, J., Rhodes, W., and Russell, P. S., Transplant. Proc. 13, 499, 1981. 8. Haynes, B. F., Dowell, B. L., Hensley, L. L., Gore, I., and Metzgar, R. S., Science 215, 298, 1982. 9. Neubauer, R. H., Levy, R., Stmad, B. C., and Rabin, H., J. Immunogenet. 8, 433, 1981. 10. Thorsby, E., and Bratlie, A., “Histocompatibility Testing.” pp. 655-656. Williams & Wilkins, Baltimore, 1970. 11. Martin, L. N., Leslie, G. A., and Hindes, R., Int. Archs. Allergy Appl. Immunol. 51, 320, 1976. 12. Kaplan, M. E., and Clark, C., J. Immunol. Methods 5, 131, 1974. 13. Vessella, R. L., Gormus, B. J., Lange, P. H., and Kaplan, M. E., Znt. J. Cancer 21, 594, 1978. 14. Nomdedeu, B., Gormus, B. J., Banisadre, M., Rinehart, J. J., Kaplan, M. E., and Zanjani, E. D., Exp. Hematol.

8, 845, 1980.

15. Gormus, B. J., Basara, M. L., Cossman, J., Ameson, M. A., and Kaplan, M. E., Cell. Immunol. 55, 94, 1980. 16. Gelfand, J. A., Fauci, A. S., Green, T., and Frank, M. M., J. Immunol. 116, 595, 1976. 17. Gronowicz, E., Coutinho, A., and Melchers, F., Eur. J. Immunol. 6, 588, 1976. 18. Lanzavecchia, A., Sacchi, G., Maccario, R., Vitiello, A., Nespoli, L., and Ugazio, A. G., &and. J. Immunol.

10, 297, 1979.

19. Sjiiquist, J., Meloun, B., and Hjelm, H., Eur. J. Biochem. 29, 572, 1972. 20. Martin, L. N., J. Immunol. Meth. 50, 319, 1982. 21. Vecchione, R., Chatterjee, S. N., Billings, R., Bemoco, D., and Terasaki, P., Amer. Coil. Surg. Surgical Forum 32, 562, 1981. 22. Moretta, L., Ferrarini, M., Mingari, M. C., Moretta, A., and Webb, S. R., J. Immunol. 117, 217 1, 1976. 23. Moretta, L., Webb, S., Grossi, C., Lydyard, P., and Cooper, M., J. Exp. Med. 146, 184, 1977. 24. Pryjma, J., Munoz, J., Virella, G., and Fudenberg, H. H., Cell. Immunol. 50, 115, 1980. 25. Hammarstrom, L., Bird, A. G., B&ton, S., and Smith, C. I. E., Immunology 38, 181, 1979. 26. Miyawaki, T., Moriya, N., Nagaoki, T., and Taniguchi, N., Immunolog. Rev. 57, 61, 1981.