- Email: [email protected]
Defective B-cell tolerance in New Zealand black mice
CELLULARIMM~UNOLOCY
113,183-191
(1988)
Defective B-Cell Tolerance in New Zealand Black Mice Fc Receptor-Independence
of Resistance to Low-Epitope-Density ELIOT
Department oflnternal
Tolerogens’
A. GOLDINGS
Medicine and the Harold C. Simmons Arthritis Research Center, University of Texas Health Science Center at Dallas, Dallas, Texas 75235 Received October 27, 1987; accepted December 6, 1987
Trinitrophenyl (TNP) human y-globulin with low-epitope-density tolerizes B cells from normal BDF, mice in a Fey receptor-dependent manner but does not tolerize B cells from preautoimmune NZB mice. In order to investigate the relationships between tolerance induction and epitope density independently of Fey receptor function in these two strains, TNP conjugates of two additional thymic-independent tolerogenic carriers, D-ghrtamic acid-o-lysine (D-CL) and carboxymethyl cellulose (CMC), were tested. A brief pulse with low-epitope-density conjugates such as TNP,.,-D-CL rendered unfractionated or T-cell-depleted spleen cells from BDF, but not NZB mice tolerant in a hapten-specific manner. Spleen cells from NZB mice, however, were susceptible to tolerization with TNPr,.+-CL. NZB mice were also resistant to tolerance induction in vivo with TNP,.+-CL, TNPx-CMC, and TNP,-CMC, all of which tolerize BDF, mice in vivo. Both strains were tolerized with TNP13.5-~GL and TNP,,-CMC in vivo. NZB mice were also significantly lesssusceptible to tolerance induction with TNP&MC when TNPFicoll was substituted for TNP Brucella abortus as the challenge antigen. These findings militate against the possibility that an Fcr receptor defect is the principal mechanism of resistance of NZB IB cells to tolerance induction with-low-epitope density conjugates. o 1988 Academic Press. Inc.
INTRODUCTION New Zealand black (NZB)2 mice which develop spontaneous autoimmunity are resistant to the induction of immunologic tolerance relative to other autoimmune and control mice (1). NZB splenic B cells required a relatively higher epitope density of the trinitrophenyl (TNP) hapten on the protein carrier human y-globulin (HGG) to induce unresponsiveness in vitro. Furthermore, NZB T cells and macrophages ’ Supported by a National Institutes of Health Grant AM-09989. ’ Abbreviations used: BDF, (C57BL/6 X DBA/Z) F,; CMC, carboxymethyl cellulose; D-CL, o-glutamic acid-D-lysine; FcyR, receptor for the Fc region of immunoglobulin G; FL-BA, fluorescein Brucella abortus, NZB (NZB/BtNJ), New Zealand black mice; PFC, plaque-forming cell; SLE, systemic lupus erythematosus; sIg, surface immunoglobulin; SRBC, sheep erythrocytes; TNP, trinitrophenyl; TNP-BA, trinitrophenyl Brucella abortus; TNP-CMC, trinitrophenyl carboxymethyl cellulose; TNP-D-CL, WinitrophenyJ-Dglutamic a&-D-lysine; TNP-HGG, trinitrophenyl human y-globulin; TNBS, 2,4,6-trinitrobenzene sulfonic acid, HBSS, Hanks’ balanced salt solution. 183 0008-8749/88$3.00 Copyright 0 1988 by Academic Press, Inc. All n&s of reproduction in any form reserved.
184
ELIOT
A. GOLDINGS
did not account for the observed abnormal state of resistance to B-cell tolerance induction. The mechanism of tolerance induction in normal B cells by low-epitope-density conjugates of TNP-HGG has been partially elucidated (2). Although the participation of T cells and macrophages is not necessary(1,2), a requirement for the Fc region of the HGG carrier has been documented (2). The latter presumably engagesthe Fey receptor (FcyR) of susceptible B cells either to promote their avidity for the tolerogen in a passive fashion or to transmit an additional signal to that delivered via the surface immunoglobulin (sIg) receptors specific for the TNP hapten. The latter possibility is suggestedby the demonstration that engagement of Fc receptors can interfere with polyphosphoinositide hydrolysis in a model of Fc-dependent B-cell inhibition by anti-immunoglobulin antibodies (3). Resistance in NZB B cells to TNP-HGG with low epitope density, therefore, might involve a defect in the expression or function of either the sIg receptor or the FcrR, disorders of both types of receptors, or alternatively a postreceptor defect. The possibility of a defect in B-cell FcyR function is suggestedby the previous demonstration of such defects in FcyR-mediated clearance by mononuclear phagocytes in mice (4) and patients (5) who develop systemic lupus erythematosus (SLE). The possibility that the tolerance defect might relate to alterations in surface immunoglobulin expression is suggestedby the high surface IgM to IgD ratio (6) and overall lower density of surface immunoglobulin (7) in NZB B cells. In order to determine whether the mechanism of tolerance resistance in NZB mice might involve a possible FcyR defect, B-cell unresponsiveness was examined in NZB and control mice with nonimmunoglobulin low-epitope-density tolerogens. Two previously characterized Fc-independent tolerogenic carriers capable of promoting hapten-specific B-cell tolerance, D-glutamic acid-D-lysine (D-GL) (8, 9) and carboxymethyl cellulose (CMC) (lo), were selectedfor study. The data indicate that resistance of NZB B cells to tolerogens with low epitope density is independent of the carrier used and suggestthat the basis of tolerance resistance in this strain relates to a defect of either surface immunoglobulin expression and function or alternatively a postreceptor defect. MATERIALS AND METHODS Animals. Female NZB/BINJ, (C57BL/6 X DBA/2)F,, and DBA/2J mice (The Jackson Laboratory, Bar Harbor, ME) were used at 6 to 10 weeks of age. Antigens. Carboxymethyl cellulose (sodium salt, medium viscosity grade, Sigma Chemical Co., St. Louis, MO) was haptenated with 2,4,6-trinitrobenzene sulfonic acid (TNBS, Sigma) at substitution ratios of 3, 6, 11, and 13 mol TNP per 100,000 Da of CMC asdescribed (8, 10). Copolymer D-glutamic acid-D-lysine (mol wt 49,000, ratio 60:40) (D-GL) (Miles-Yeda, Rehovot, Israel) was haptenated with TNBS at substitution ratios of 4.4, 5.5, 13.5, and 30.5 per mole of D-GL as described (9). TNP Brucellu abortus (BA) was prepared as described (11). Fluorescein (FL) BA was also prepared as described (12). TNPe5aecm Ficoll and FL aecm Ficoll were purchased from Biosearch (San Rafael, CA). Axsay of antibody-forming cells. Plaque-forming cells (PFC) were determined by a slide modification of the hemolysis in gel technique ( 13, 14). TNP-BA-stimulated cultures or spleen cells from TNP-BA- or TNP-Ficoll-injected mice were assayed against TNP-sheep erythrocytes (SRBC) ( 15). FL-BA-stimulated cultures of spleen cells from FL-BA- or FL-Ficoll-injected mice were assayedagainst FL-SRBC ( 16).
DEFECTIVE
B-CELL
TOLERANCE
IN NZB MICE
185
Induction and assessmentof unresponsiveness in vitro. Pooled spleen cells of several syngeneic mice were suspended in complete medium consisting of RPM1 1640 (Microbiologic Associates,Walkersville, MD) supplemented with 5% fetal calf serum (Microbiologic Associates), penicillin G, 150 U/ml, streptomycin, 0.25 mg/ml, Lglutamine, 0.3 mg/ml (Microbiologic Associates), gentamicin sulfate (Schering Corp., Kenilworth, NJ), 10 pg/ml, and 2-mercaptoethanol, 5 X 1O-5M. In several experiments, spleen cells were depleted of T cells by treatment with monoclonal antiThy 1.2 antibody (HO 13.4), a gift from Dr. Ellen S. Vitetta, and guinea pig complement (Pel-Freeze Biologicals, Inc., Rogers, AR) as described (11). After a single treatment, the proliferative responsesof the cells to T-cell mitogens (concanavalin A and phytohemagglutinin) were abolished entirely, whereas responsivenessto lipopolysaccharide remained intact. Cells were pulsed for 1 hr at 37°C at a cell density of 107/ml with various concentrations of TNP-D-GL, washed three times with Hanks’ balanced salt solution (HBSS) (Microbiological Associates),and immunized in microtiter wells (Costar, Data Packaging, Cambridge, MA) at a density of 5 X 106/ml with 0.001% TNP-BA or 0.00 1%FL-BA (volume packed organisms/culture volume). After 3 days of incubation with antigen, direct antihapten PFC responseswere assayed.The results of four replicate cultures are expressedasthe arithmetic mean PFC responsesper lo6 input cells ? SEM or as the percentage of the control PFC response. Induction1 and assessmentof unresponsiveness in vivo. Groups of five mice were injected iv Twitheither 1 mg TNP-D-GL on 2 successivedays or once with 250 pg TNP-CMC or HBSS as control and challenged iv 1 day later with both TNP-BA and FL-BA, 100 ~1 of a 0.5% of suspension each, or TNP-Ficoll and FL-Ficoll, 100 pg each. The mice were sacrificed 4-5 days later and the direct anti-hapten PFC responsesof individual spleen cell suspensions were assessed.The results are expressed as either the geometric mean PFC responsesper spleen or as the percentage of the control response. RESULTS Resistance
186
ELIOT A. GOLDINGS
”
TNP,3,5-D-GL
TNP,++GL
TOLEROGEN
FIG. 1. Selective resistance of NZB spleen cells to tolerance induction in vitro with low-epitope-density TNP-DGL. Splenocytes from NZB or BDF, mice were pulsed for 1 hr with 100 &ml of tolerogen or HBSS as control and challenged with TNP-BA. Direct anti-TNP PFC were measured 3 days later. Results pooled from 3 to 8 experiments for each point.
Resistance of NZB Mice to Tolerance Induction in Vivo with Low-Epitope-Density TNP-D-GL Intact NZB and BDF, mice were pretreated in vivo with either 2 mg of TNP,.,-DGL or TNP,3.5-~-GL and challenged with antigen. The anti-TNP response of BDFI 2000 A. BDF, 1600 1200 % *
600
“0 c
400
ii
1200
:
1000
!
I
8. NZB
i E’
t
600 600 400
0
100 jog TNPea-p-
10’ GL/ml
102
FIG. 2. Thymic independence of susceptibility of BDF, and resistanceof NZB mouse B cells to tolerance induction in vitro with low-epitope-density TNP-D-GL. T-Cell-depleted splenocytes were pulsed with TNP4..,-D-GL or HBSS as control for 1 hr challenged with TNP-BA (0) or FL-BA (0), and PFC assayed3 days later. Representative of 3 experiments.
DEFECTIVE B-CELL TOLERANCE IN NZB MICE
187
NZB BDF,
rNP5,5-rj-GL TOLEROGEN
FIG. 3. Selective resistance of NZB mice to tolerance induction in vivo with low-epitope-density TNPD-CL. Mice were injected iv with 1 mg oftolerogen or HBSS as control on 2 successivedays and challenged with TNP-BA. Four days later the mice were sacrificed and direct splenic PFC were assayed.
mice given TNPs, ,-D-GL was reduced by approximately 50% whereas the response of NZB mice was completely unaffected (Fig. 3). Mice of both strains, however, were tolerized by TNP,3.5-~-GL to the same degree, indicating a selective resistance of NZB mice to tolerance induction in vivo by TNP-D-GL with low epitope density. Resistance of NZB Mice to Tolerance Induction in Vivo with Low-Epitope-Density TNP-CMC A second1Fc-independent tolerogen was selected to study the relative epitope density requirements for tolerance induction in NZB and BDF, mice. A plot of the absolute anti-TNP responseper spleen to TNP-BA is given asa function of prior treatment in vivo with either HBSS, TNP,-CMC, or TNP13-CMC (Fig. 4). The responseof BDF, mice is reduced from 108,000 to 34,000 PFC/spleen after treatment with TNP3CMC. The response was further diminished to 5% of control by pretreatment with TNP,,-CMC. In contrast, the anti-TNP response of NZB mice was entirely unaffected by pretreatment with TNP,-CMC. Similarly, NZB mice were resistant to tolerance induction with TNP6-CMC (data not shown). Although the anti-TNP response of NZB mice after pretreatment with TNP,,-CMC was significantly reduced, the degree Ioftolerance achieved is lessthan that observed in BDF, mice. Resistance of the LybS B-Cell Subset in NZB Mice to in Vivo Tolerization with TNP,-C’MC Whereas the PFC response to the type 1 antigen, TNP-BA, includes contributions by both LybS- and LybS+ B-cell subsets( 17), the responseto the type 2 antigen, TNPFicoll, is exclusively generated by the more mature LybS+ subset (18). In order to examine the tolerance susceptibility of the Lyb5’ B-cell subpopulation, TNP-Ficoll was substituted for TNP-BA as the challenge antigen. After pretreatment with TNP3-
188
ELIOT A. GOLDINGS
q
z
CONTROL
•TNP~c~~C
q
TNP,+C
NZB
BDF, MOUSE STRAIN
FIG. 4. Resistance of NZB mice to tolerance induction in vivo with TNP,-CMC. Five mice per group of both BDF, and NZB mice were injected with either 250 pg IV of TNP,-CMC, TNP&MC, or HBSS, 24 hr later challenged with TNP-BA, and 4 days later assayedfor direct PFC responses.
CMC, the mice were challenged with both TNP-Ficoll and FL-Ficoll. The anti-TNP responses in normal BDF, and DBA/2 mice were approximately 20% of control, significantly more reduced than the anti-TNP response of NZB mice (P < 0.005, Fig. 5). The anti-FL responsesof all strains were not affected, indicating that the observed tolerance was hapten-specific. DISCUSSION The present study has confirmed our previous in vitro finding that tolerance induction at the B-cell level is defective in NZB mice when the epitope density of the toleroMOUSE
0
TNP-FIc0LL
q
FL-FICOLL
STRAIN NZB
DBA/E
BDF,
I 0
I 20
I 40 PERCENT
I I I 100 60 90 OF CONTROL RESPONSE
I 120
J 140
FIG. 5. Resistanceof the TNP-Ficoll-responsive B-cell subset in NZB mice to tolerance induction in vivo with TNP3-CMC. Five mice per group of BDF, , DBA/Z, and NZB mice were injected iv with either 250 pg of TNP,-CMC or HBSS, 24 hr later challenged with both TNP-Ficoll and FL-Ficoll, and 4 days later assayedfor direct PFC responses.
DEFECTIVE
B-CELL
TOLERANCE
IN NZB MICE
189
gen is low (1). Furthermore, it excludes the possibility that the mechanism of tolerance resistance in NZB mice resides solely with a defect in Fc receptor function. Whereas a defect in Fc receptor function has not been excluded by the present study, an additional defect must be invoked either at the level of surface immunoglobulin receptors for the hapten or a postreceptor defect preventing adequate transduction of the negative signal. The present study was stimulated by the findings of Waldschmidt et al. (2) that demonstrated a requirement for the Fc portion of the carrier for TNP-HGG when the hapten dens:ity was low. Defects in Fc-dependent clearance of opsonized erythrocytes presumably at the level of macrophages have been demonstrated in autoimmune New Zealanld mice (4). In addition, the number of Fc-receptor-positive B cells in such strains is diminished with time (19). On the other hand, defects in the density and isotype distribution of surface immunoglobulin receptors have also been reported in the NZB mouse (6,7). In order to explore the mechanism for the resistance of B cells to tolerance induction with TNPiO-HGG (l), the present study was conducted with tolerogens that circumvent the requirement for Fc receptor interaction to induce tolerance. Because splenic B cells in vitro and intact NZB mice have both proven to be resistant to Fc-independent tolerogens, defects of either surface immunoglobulin receptors or postreceptor abnormalities are suggestedby this report. The tolerogens selected for the present study have both been determined to exert their inhibitory effectsdirectly at the B cell level (8-10). The possibility that suppressor-T-cell induction might be necessaryfor tolerogenic action by the lightly haptenated conjugates used in the present study could not be excluded a priori. Therefore, experiments were performed to rule out a requirement for suppressor T-cell generation. Normal BDF, spleen cells depleted of T cells were susceptible to TNP,.,-D-GL in vitro (Fig. 2A). Similarly, suppressor cell generation in TNP3-CMC-injected mice was not detectable in in vitro mixing experiments (data not shown). The thymic independence of tolerance induction in normal mice by haptenated D-GL (8,9) and lack of suppressor-T-cell generation by haptenated CMC are in agreement with previous reports (10)~.In addition, the possibility that excessive helper-T-cell activity in NZB splenocytes might actively interfere with tolerance induction by TNP,.,-D-GL was excluded by the observation of resistance to tolerance induction by T-cell-depleted NZB spleen cells. Finally, the impaired responsesobserved in normal mice after the administration of low-epitope-density TNP-D-GL and TNP-CMC cannot be attributed to tolerization of helper T cells since the antibody responsesto TNP-BA both in vitro and in vivo (20) and TNP-Ficoll in vivo (2 1) do not depend upon helper-T-cell participation. The present study has not excluded a possible macrophage defect in determining the resistance of NZB mice to tolerance induction with Fc-independent tolerogens. This possibility is unlikely, however, in view of our previous experiments with TNP,,-HGG ( 1) that demonstrated tolerance resistancein macrophage-depleted NZB splenocytes and failure of NZB cells in mixing experiments to interfere with tolerance induction in normal DBA/2 splenocytes. Resistance to B-cell tolerance induction in NZB mice has previously been shown for TNP-BA-responsive B cells in NZB mice (1) and has been confirmed in the present study. The latter are comprised of both LybS- and Lyb5+ B cells (17). In order to assessthe susceptibility to tolerance induction of the LybS+ subset in NZB mice, the type 2 antigen TNP-Ficoll capable of stimulating only Lyb5+ cells was utilized as the antigenic challenge. Although the susceptibility of the LybS+ subset in NZB mice to
190
ELIOT
A. GOLDINGS
tolerance induction appearsto be greater (Fig. 5) than that of the TNP-BA responsive subset (Figs. 3, 4), the former was still more resistant than in the two normal strains tested. The Lyb5+ subpopulation in NZB mice has previously been shown to be responsible for the production of both polyclonal IgM and certain IgM autoantibodies (22,23). Another seriesof studies has detected an abnormal B-cell subset in this strain which bears the Ly 1 antigen and is responsible both for polyclonal hypersecretion of IgM and for IgM autoantibody formation (24-26). These subsetsappear to be largely but not entirely overlapping (27). Whether the Ly If cells irrespective of Lyb5 type exclusively account for the resistance to low-epitope tolerogens observed in NZB mice remains unresolved. A previous study has examined the susceptibility of NZB mice to one of the tolerogens used in the current study (28). The administration of 4.5 mg of TNP-CMC with an epitope density of 9-l 1 to NZB mice reduced the IgM antihapten response to TNP-Ficoll and TNP-lipopolysaccharide to the samedegreeas in normal strains. The secondary IgG anti-hapten responseto a T-dependent antigen, however, was resistant to this regimen. Since this degree of haptenation of the CMC carrier was nearly that of the conjugate capable of tolerizing NZB B cells in the present study (TNPi,-CMC (Fig. 4) and TNP,,-CMC (data not shown) and the dose of tolerogen was much greater, the studies are in accord. Defective tolerization of the primary IgM antihapten response has more recently been documented by Brooks and Aldo-Benson using B-cell lines from (NZB X NZW)F, mice (29) and by Cowdery et al. in surface immunoglobulin-negative marrow cells from NZB mice using the splenic focus assay (30). These two reports are in agreement with our previous ( 1, 12) and current reports. Recent studies examining the early events associated with B-cell activation have documented a series of biochemical changes occurring at both the membrane and intracellular levels after crosslinking of surface immunoglobulin receptors (3 1). It is possible that hapten-specific tolerance induction might involve an alteration in the pathways associated with activation. Since NZB B cells are refractory to tolerance induction under defined conditions, they may serve as a negative control in studies designed to examine biochemical pathways associated with tolerance induction in normal mouse B cells. Such an analysis would be facilitated by the use of enriched populations of hapten-specific cells. Recent reports have indicated that the predominant DNA speciesin circulation in patients with SLE is a relatively low-molecular-weight speciesconsisting of as little as 30 to 50 base pairs and usually s200 base pairs (32-33). In view of the work by Papalian and colleagues (34), which has demonstrated that the minimal size required for divalent attachment of anti-DNA antibody to its ligand is 20 to 30 basepairs, it is conceivable that circulating DNA might in fact function as a low-epitope ligand. Such an autoantigen might ordinarily feed-back on anti-DNA-producing B cells induced either by polyclonal activators that bypass surface immunoglobulin receptors or by environmental, crossreactive, multivalent immunogens (35). The resistance to tolerance induction in New Zealand mice by low-epitope-density tolerogens irrespective of the chemical nature of the carrier demonstrated by the present and previous studies (1, 12) suggeststhat such hypothetical regulatory mechanisms might fail to operate in NZB mice and thereby contribute to autoantibody production.
DEFECTIVE B-CELL TOLERANCE IN NZB MICE
191
ACKNOWLEDGMENTS I thank Ms. Mary Waldschmidt and Ms. Lilli Hansen for excellent technical assistance; Mrs. Theresa Stencer for skillful secretarial assistance; Dr. Ellen S. Vitetta for purified monoclonal anti-Thy 1.2 antibody; and Dr. Peter E. Lipsky for critical review of the manuscript.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
Goldings, E. A., J. Immunol. 131,2630, 1983. Waldschmidt, T. J., Borel, Y., and Vitetta, E. S., J. Immunol. 131,2204, 1983. Bijsterbosch, M. K., and Klaus, G. G. B., J. Exp. Med. 162, 1825, 1985. Magilavy, D. B., Rifai, A., and Plotz, P. H., J. Immunol. 126,770, 198 1. Frank, M. M., Lawley, T. J., Hamburger, M. I., and Brown, E. J., Ann. Intern. Med. 98,206, 1983. Cohen, P., Ziff, M., and Vitetta, E. S., J. Immunol. 121,973, 1978. Chused, T. M., Moutsopoulos, H. H., Sharrow, S. O., Hansen, C. T., and Morse, H. C., III, In “Genetic Control of Autoimmune Diseases” (N. R. Rose, P. E. Bigazzi, and N. L. Warner, Eds.), pp. 177190. Elsevier/North Holland, New York, 1978. 8. Nossal, G. J. V., Pike, B. L., and Katz, D. H., J. Exp. Med. 138,3 12, 1973. 9. Goldings, E. A., and Mosier, D. E., Eur. J. Immunol. 9,76, 1979. 10. Diner, I-1.E., Kunimoto, D., and Diener, E., J. Immunol. 122, 1886, 1979. 11. Cambier, J. C., Vitetta, E. S., Uhr, J. W., and Kettman, J. R., J. Exp. Med. 145,778, 1977. 12. Geldings, E. A., Cohen, P. L., McFadden, S. F., Ziff, M., and Vitetta, E. S., J. Exp. Med. 152, 730, 1980. 13. Mishell, R., and Dutton, R. W., Science 153, 1004, 1966. 14. Jeme, N. D., and Nordin, A. A., Science 140,405, 1963. 15. Rittenbe:rg, M. R., and Pratt, K. L., Proc. Sot. Exp. Biol. Med. 132,575, 1969. 16. Jondal, M., Holm, G., and Wigzell, H., J. Exp. Med. 136,307, 1972. 17. Mosier, D. E., and Subbarao, B., Immunol. Today 3,2 17, 1982. 18. Mosier, D. E., Mond, J. J., andGoldings, E. A., J. Immunol. 119, 1877, 1977. 19. Theofilopoulos, A. N., Eisenberg, R. A., Bourdon, M., Crowell, J. S., and Dixon, F. J., J. Exp. Med. 149,516, 1979. 20. Mond, J. J., Scher, I., Mosier, D. E., Blaese,M., and Paul, W. E., Eur. J. Immunol. 8,459, 1978. 2 1. Sharon, R., McMaster, P. R. B., Kask, A. M., Owens, J. D., and Paul, W. E., J. Immunol. 114, 1585, 1975. 22. Nakajima, P. B., Datta, S. K., Schwartz, R. S., and Huber, B. T., Proc. Natl. Acad. Sci. USA 76,46 13, 1979. 23. Taurog, J. D., Raveche, E. S., Smathers, T. A., Glimcher, L. H., Huston, D. P., Hansen, C. T., and Steinberg, A. D., J. Exp. Med. 153,221, 1981. 24. Manoha.r, V., Brown, E., Leiserson, W. M., and Chused, T. M., J. Immunol. 129,532, 1982. 25. Hayakawa, K., Hardy, R. R., Parks, D. R., and Herzenberg, L. A., J. Exp. Med. 157,202, 1983. 26. Hayakawa K., Hardy, R. R., Honda, M., Herzenberg, L. A., and Steinberg, A. D., Proc. Natl. Acad. Sci. USA 81,2494, 1984. 27. Smith, II. R., Yaffe, L. J., Chused, T. M., Raveche, E. S., Klinman, D. M., and Steinberg, A. D., Cell. Immunol. 92, 190, 1985. 28. Tsunetsugu, Y., and Fujiwara, M., Cell. Immunol. 701,303, 1982. 29. Brooks, M. S., and Aldo-Benson, M., J. Clin. Invest. 78,784, 1986. 30. Cowdery, J. S., Jacobi, S. M., Pitts, A. K., and Tyler, T. L., J. Immunol. 138,760, 1987. 3 1. Cambier, J. C., Monroe, J. G., Coggeshall, K. M., and Ramsom, J. T., Immunol. Toduy6,2 18, 1985. 32. Morimoto, C., Sano, H., Abe, T., Homma, M., and Steinberg, A. D., J. Immunol. 139, 1960, 1982. 33. McCoubrey-Hoyer, A., Okarma, T. B., and Holman, H. R., Amer. J. Med. 77,23, 1984. 34. Papalian, M., Lafer, E., Wong, R., and Stollar, B. D., J. Clin. Invest. 65,469, 1980. 35. Schwartz, R. S., and Stollar, B. D., J. Clin. Invest. 75,321, 1985.
113,183-191
(1988)
Defective B-Cell Tolerance in New Zealand Black Mice Fc Receptor-Independence
of Resistance to Low-Epitope-Density ELIOT
Department oflnternal
Tolerogens’
A. GOLDINGS
Medicine and the Harold C. Simmons Arthritis Research Center, University of Texas Health Science Center at Dallas, Dallas, Texas 75235 Received October 27, 1987; accepted December 6, 1987
Trinitrophenyl (TNP) human y-globulin with low-epitope-density tolerizes B cells from normal BDF, mice in a Fey receptor-dependent manner but does not tolerize B cells from preautoimmune NZB mice. In order to investigate the relationships between tolerance induction and epitope density independently of Fey receptor function in these two strains, TNP conjugates of two additional thymic-independent tolerogenic carriers, D-ghrtamic acid-o-lysine (D-CL) and carboxymethyl cellulose (CMC), were tested. A brief pulse with low-epitope-density conjugates such as TNP,.,-D-CL rendered unfractionated or T-cell-depleted spleen cells from BDF, but not NZB mice tolerant in a hapten-specific manner. Spleen cells from NZB mice, however, were susceptible to tolerization with TNPr,.+-CL. NZB mice were also resistant to tolerance induction in vivo with TNP,.+-CL, TNPx-CMC, and TNP,-CMC, all of which tolerize BDF, mice in vivo. Both strains were tolerized with TNP13.5-~GL and TNP,,-CMC in vivo. NZB mice were also significantly lesssusceptible to tolerance induction with TNP&MC when TNPFicoll was substituted for TNP Brucella abortus as the challenge antigen. These findings militate against the possibility that an Fcr receptor defect is the principal mechanism of resistance of NZB IB cells to tolerance induction with-low-epitope density conjugates. o 1988 Academic Press. Inc.
INTRODUCTION New Zealand black (NZB)2 mice which develop spontaneous autoimmunity are resistant to the induction of immunologic tolerance relative to other autoimmune and control mice (1). NZB splenic B cells required a relatively higher epitope density of the trinitrophenyl (TNP) hapten on the protein carrier human y-globulin (HGG) to induce unresponsiveness in vitro. Furthermore, NZB T cells and macrophages ’ Supported by a National Institutes of Health Grant AM-09989. ’ Abbreviations used: BDF, (C57BL/6 X DBA/Z) F,; CMC, carboxymethyl cellulose; D-CL, o-glutamic acid-D-lysine; FcyR, receptor for the Fc region of immunoglobulin G; FL-BA, fluorescein Brucella abortus, NZB (NZB/BtNJ), New Zealand black mice; PFC, plaque-forming cell; SLE, systemic lupus erythematosus; sIg, surface immunoglobulin; SRBC, sheep erythrocytes; TNP, trinitrophenyl; TNP-BA, trinitrophenyl Brucella abortus; TNP-CMC, trinitrophenyl carboxymethyl cellulose; TNP-D-CL, WinitrophenyJ-Dglutamic a&-D-lysine; TNP-HGG, trinitrophenyl human y-globulin; TNBS, 2,4,6-trinitrobenzene sulfonic acid, HBSS, Hanks’ balanced salt solution. 183 0008-8749/88$3.00 Copyright 0 1988 by Academic Press, Inc. All n&s of reproduction in any form reserved.
184
ELIOT
A. GOLDINGS
did not account for the observed abnormal state of resistance to B-cell tolerance induction. The mechanism of tolerance induction in normal B cells by low-epitope-density conjugates of TNP-HGG has been partially elucidated (2). Although the participation of T cells and macrophages is not necessary(1,2), a requirement for the Fc region of the HGG carrier has been documented (2). The latter presumably engagesthe Fey receptor (FcyR) of susceptible B cells either to promote their avidity for the tolerogen in a passive fashion or to transmit an additional signal to that delivered via the surface immunoglobulin (sIg) receptors specific for the TNP hapten. The latter possibility is suggestedby the demonstration that engagement of Fc receptors can interfere with polyphosphoinositide hydrolysis in a model of Fc-dependent B-cell inhibition by anti-immunoglobulin antibodies (3). Resistance in NZB B cells to TNP-HGG with low epitope density, therefore, might involve a defect in the expression or function of either the sIg receptor or the FcrR, disorders of both types of receptors, or alternatively a postreceptor defect. The possibility of a defect in B-cell FcyR function is suggestedby the previous demonstration of such defects in FcyR-mediated clearance by mononuclear phagocytes in mice (4) and patients (5) who develop systemic lupus erythematosus (SLE). The possibility that the tolerance defect might relate to alterations in surface immunoglobulin expression is suggestedby the high surface IgM to IgD ratio (6) and overall lower density of surface immunoglobulin (7) in NZB B cells. In order to determine whether the mechanism of tolerance resistance in NZB mice might involve a possible FcyR defect, B-cell unresponsiveness was examined in NZB and control mice with nonimmunoglobulin low-epitope-density tolerogens. Two previously characterized Fc-independent tolerogenic carriers capable of promoting hapten-specific B-cell tolerance, D-glutamic acid-D-lysine (D-GL) (8, 9) and carboxymethyl cellulose (CMC) (lo), were selectedfor study. The data indicate that resistance of NZB B cells to tolerogens with low epitope density is independent of the carrier used and suggestthat the basis of tolerance resistance in this strain relates to a defect of either surface immunoglobulin expression and function or alternatively a postreceptor defect. MATERIALS AND METHODS Animals. Female NZB/BINJ, (C57BL/6 X DBA/2)F,, and DBA/2J mice (The Jackson Laboratory, Bar Harbor, ME) were used at 6 to 10 weeks of age. Antigens. Carboxymethyl cellulose (sodium salt, medium viscosity grade, Sigma Chemical Co., St. Louis, MO) was haptenated with 2,4,6-trinitrobenzene sulfonic acid (TNBS, Sigma) at substitution ratios of 3, 6, 11, and 13 mol TNP per 100,000 Da of CMC asdescribed (8, 10). Copolymer D-glutamic acid-D-lysine (mol wt 49,000, ratio 60:40) (D-GL) (Miles-Yeda, Rehovot, Israel) was haptenated with TNBS at substitution ratios of 4.4, 5.5, 13.5, and 30.5 per mole of D-GL as described (9). TNP Brucellu abortus (BA) was prepared as described (11). Fluorescein (FL) BA was also prepared as described (12). TNPe5aecm Ficoll and FL aecm Ficoll were purchased from Biosearch (San Rafael, CA). Axsay of antibody-forming cells. Plaque-forming cells (PFC) were determined by a slide modification of the hemolysis in gel technique ( 13, 14). TNP-BA-stimulated cultures or spleen cells from TNP-BA- or TNP-Ficoll-injected mice were assayed against TNP-sheep erythrocytes (SRBC) ( 15). FL-BA-stimulated cultures of spleen cells from FL-BA- or FL-Ficoll-injected mice were assayedagainst FL-SRBC ( 16).
DEFECTIVE
B-CELL
TOLERANCE
IN NZB MICE
185
Induction and assessmentof unresponsiveness in vitro. Pooled spleen cells of several syngeneic mice were suspended in complete medium consisting of RPM1 1640 (Microbiologic Associates,Walkersville, MD) supplemented with 5% fetal calf serum (Microbiologic Associates), penicillin G, 150 U/ml, streptomycin, 0.25 mg/ml, Lglutamine, 0.3 mg/ml (Microbiologic Associates), gentamicin sulfate (Schering Corp., Kenilworth, NJ), 10 pg/ml, and 2-mercaptoethanol, 5 X 1O-5M. In several experiments, spleen cells were depleted of T cells by treatment with monoclonal antiThy 1.2 antibody (HO 13.4), a gift from Dr. Ellen S. Vitetta, and guinea pig complement (Pel-Freeze Biologicals, Inc., Rogers, AR) as described (11). After a single treatment, the proliferative responsesof the cells to T-cell mitogens (concanavalin A and phytohemagglutinin) were abolished entirely, whereas responsivenessto lipopolysaccharide remained intact. Cells were pulsed for 1 hr at 37°C at a cell density of 107/ml with various concentrations of TNP-D-GL, washed three times with Hanks’ balanced salt solution (HBSS) (Microbiological Associates),and immunized in microtiter wells (Costar, Data Packaging, Cambridge, MA) at a density of 5 X 106/ml with 0.001% TNP-BA or 0.00 1%FL-BA (volume packed organisms/culture volume). After 3 days of incubation with antigen, direct antihapten PFC responseswere assayed.The results of four replicate cultures are expressedasthe arithmetic mean PFC responsesper lo6 input cells ? SEM or as the percentage of the control PFC response. Induction1 and assessmentof unresponsiveness in vivo. Groups of five mice were injected iv Twitheither 1 mg TNP-D-GL on 2 successivedays or once with 250 pg TNP-CMC or HBSS as control and challenged iv 1 day later with both TNP-BA and FL-BA, 100 ~1 of a 0.5% of suspension each, or TNP-Ficoll and FL-Ficoll, 100 pg each. The mice were sacrificed 4-5 days later and the direct anti-hapten PFC responsesof individual spleen cell suspensions were assessed.The results are expressed as either the geometric mean PFC responsesper spleen or as the percentage of the control response. RESULTS Resistance
186
ELIOT A. GOLDINGS
”
TNP,3,5-D-GL
TNP,++GL
TOLEROGEN
FIG. 1. Selective resistance of NZB spleen cells to tolerance induction in vitro with low-epitope-density TNP-DGL. Splenocytes from NZB or BDF, mice were pulsed for 1 hr with 100 &ml of tolerogen or HBSS as control and challenged with TNP-BA. Direct anti-TNP PFC were measured 3 days later. Results pooled from 3 to 8 experiments for each point.
Resistance of NZB Mice to Tolerance Induction in Vivo with Low-Epitope-Density TNP-D-GL Intact NZB and BDF, mice were pretreated in vivo with either 2 mg of TNP,.,-DGL or TNP,3.5-~-GL and challenged with antigen. The anti-TNP response of BDFI 2000 A. BDF, 1600 1200 % *
600
“0 c
400
ii
1200
:
1000
!
I
8. NZB
i E’
t
600 600 400
0
100 jog TNPea-p-
10’ GL/ml
102
FIG. 2. Thymic independence of susceptibility of BDF, and resistanceof NZB mouse B cells to tolerance induction in vitro with low-epitope-density TNP-D-GL. T-Cell-depleted splenocytes were pulsed with TNP4..,-D-GL or HBSS as control for 1 hr challenged with TNP-BA (0) or FL-BA (0), and PFC assayed3 days later. Representative of 3 experiments.
DEFECTIVE B-CELL TOLERANCE IN NZB MICE
187
NZB BDF,
rNP5,5-rj-GL TOLEROGEN
FIG. 3. Selective resistance of NZB mice to tolerance induction in vivo with low-epitope-density TNPD-CL. Mice were injected iv with 1 mg oftolerogen or HBSS as control on 2 successivedays and challenged with TNP-BA. Four days later the mice were sacrificed and direct splenic PFC were assayed.
mice given TNPs, ,-D-GL was reduced by approximately 50% whereas the response of NZB mice was completely unaffected (Fig. 3). Mice of both strains, however, were tolerized by TNP,3.5-~-GL to the same degree, indicating a selective resistance of NZB mice to tolerance induction in vivo by TNP-D-GL with low epitope density. Resistance of NZB Mice to Tolerance Induction in Vivo with Low-Epitope-Density TNP-CMC A second1Fc-independent tolerogen was selected to study the relative epitope density requirements for tolerance induction in NZB and BDF, mice. A plot of the absolute anti-TNP responseper spleen to TNP-BA is given asa function of prior treatment in vivo with either HBSS, TNP,-CMC, or TNP13-CMC (Fig. 4). The responseof BDF, mice is reduced from 108,000 to 34,000 PFC/spleen after treatment with TNP3CMC. The response was further diminished to 5% of control by pretreatment with TNP,,-CMC. In contrast, the anti-TNP response of NZB mice was entirely unaffected by pretreatment with TNP,-CMC. Similarly, NZB mice were resistant to tolerance induction with TNP6-CMC (data not shown). Although the anti-TNP response of NZB mice after pretreatment with TNP,,-CMC was significantly reduced, the degree Ioftolerance achieved is lessthan that observed in BDF, mice. Resistance of the LybS B-Cell Subset in NZB Mice to in Vivo Tolerization with TNP,-C’MC Whereas the PFC response to the type 1 antigen, TNP-BA, includes contributions by both LybS- and LybS+ B-cell subsets( 17), the responseto the type 2 antigen, TNPFicoll, is exclusively generated by the more mature LybS+ subset (18). In order to examine the tolerance susceptibility of the Lyb5’ B-cell subpopulation, TNP-Ficoll was substituted for TNP-BA as the challenge antigen. After pretreatment with TNP3-
188
ELIOT A. GOLDINGS
q
z
CONTROL
•TNP~c~~C
q
TNP,+C
NZB
BDF, MOUSE STRAIN
FIG. 4. Resistance of NZB mice to tolerance induction in vivo with TNP,-CMC. Five mice per group of both BDF, and NZB mice were injected with either 250 pg IV of TNP,-CMC, TNP&MC, or HBSS, 24 hr later challenged with TNP-BA, and 4 days later assayedfor direct PFC responses.
CMC, the mice were challenged with both TNP-Ficoll and FL-Ficoll. The anti-TNP responses in normal BDF, and DBA/2 mice were approximately 20% of control, significantly more reduced than the anti-TNP response of NZB mice (P < 0.005, Fig. 5). The anti-FL responsesof all strains were not affected, indicating that the observed tolerance was hapten-specific. DISCUSSION The present study has confirmed our previous in vitro finding that tolerance induction at the B-cell level is defective in NZB mice when the epitope density of the toleroMOUSE
0
TNP-FIc0LL
q
FL-FICOLL
STRAIN NZB
DBA/E
BDF,
I 0
I 20
I 40 PERCENT
I I I 100 60 90 OF CONTROL RESPONSE
I 120
J 140
FIG. 5. Resistanceof the TNP-Ficoll-responsive B-cell subset in NZB mice to tolerance induction in vivo with TNP3-CMC. Five mice per group of BDF, , DBA/Z, and NZB mice were injected iv with either 250 pg of TNP,-CMC or HBSS, 24 hr later challenged with both TNP-Ficoll and FL-Ficoll, and 4 days later assayedfor direct PFC responses.
DEFECTIVE
B-CELL
TOLERANCE
IN NZB MICE
189
gen is low (1). Furthermore, it excludes the possibility that the mechanism of tolerance resistance in NZB mice resides solely with a defect in Fc receptor function. Whereas a defect in Fc receptor function has not been excluded by the present study, an additional defect must be invoked either at the level of surface immunoglobulin receptors for the hapten or a postreceptor defect preventing adequate transduction of the negative signal. The present study was stimulated by the findings of Waldschmidt et al. (2) that demonstrated a requirement for the Fc portion of the carrier for TNP-HGG when the hapten dens:ity was low. Defects in Fc-dependent clearance of opsonized erythrocytes presumably at the level of macrophages have been demonstrated in autoimmune New Zealanld mice (4). In addition, the number of Fc-receptor-positive B cells in such strains is diminished with time (19). On the other hand, defects in the density and isotype distribution of surface immunoglobulin receptors have also been reported in the NZB mouse (6,7). In order to explore the mechanism for the resistance of B cells to tolerance induction with TNPiO-HGG (l), the present study was conducted with tolerogens that circumvent the requirement for Fc receptor interaction to induce tolerance. Because splenic B cells in vitro and intact NZB mice have both proven to be resistant to Fc-independent tolerogens, defects of either surface immunoglobulin receptors or postreceptor abnormalities are suggestedby this report. The tolerogens selected for the present study have both been determined to exert their inhibitory effectsdirectly at the B cell level (8-10). The possibility that suppressor-T-cell induction might be necessaryfor tolerogenic action by the lightly haptenated conjugates used in the present study could not be excluded a priori. Therefore, experiments were performed to rule out a requirement for suppressor T-cell generation. Normal BDF, spleen cells depleted of T cells were susceptible to TNP,.,-D-GL in vitro (Fig. 2A). Similarly, suppressor cell generation in TNP3-CMC-injected mice was not detectable in in vitro mixing experiments (data not shown). The thymic independence of tolerance induction in normal mice by haptenated D-GL (8,9) and lack of suppressor-T-cell generation by haptenated CMC are in agreement with previous reports (10)~.In addition, the possibility that excessive helper-T-cell activity in NZB splenocytes might actively interfere with tolerance induction by TNP,.,-D-GL was excluded by the observation of resistance to tolerance induction by T-cell-depleted NZB spleen cells. Finally, the impaired responsesobserved in normal mice after the administration of low-epitope-density TNP-D-GL and TNP-CMC cannot be attributed to tolerization of helper T cells since the antibody responsesto TNP-BA both in vitro and in vivo (20) and TNP-Ficoll in vivo (2 1) do not depend upon helper-T-cell participation. The present study has not excluded a possible macrophage defect in determining the resistance of NZB mice to tolerance induction with Fc-independent tolerogens. This possibility is unlikely, however, in view of our previous experiments with TNP,,-HGG ( 1) that demonstrated tolerance resistancein macrophage-depleted NZB splenocytes and failure of NZB cells in mixing experiments to interfere with tolerance induction in normal DBA/2 splenocytes. Resistance to B-cell tolerance induction in NZB mice has previously been shown for TNP-BA-responsive B cells in NZB mice (1) and has been confirmed in the present study. The latter are comprised of both LybS- and Lyb5+ B cells (17). In order to assessthe susceptibility to tolerance induction of the LybS+ subset in NZB mice, the type 2 antigen TNP-Ficoll capable of stimulating only Lyb5+ cells was utilized as the antigenic challenge. Although the susceptibility of the LybS+ subset in NZB mice to
190
ELIOT
A. GOLDINGS
tolerance induction appearsto be greater (Fig. 5) than that of the TNP-BA responsive subset (Figs. 3, 4), the former was still more resistant than in the two normal strains tested. The Lyb5+ subpopulation in NZB mice has previously been shown to be responsible for the production of both polyclonal IgM and certain IgM autoantibodies (22,23). Another seriesof studies has detected an abnormal B-cell subset in this strain which bears the Ly 1 antigen and is responsible both for polyclonal hypersecretion of IgM and for IgM autoantibody formation (24-26). These subsetsappear to be largely but not entirely overlapping (27). Whether the Ly If cells irrespective of Lyb5 type exclusively account for the resistance to low-epitope tolerogens observed in NZB mice remains unresolved. A previous study has examined the susceptibility of NZB mice to one of the tolerogens used in the current study (28). The administration of 4.5 mg of TNP-CMC with an epitope density of 9-l 1 to NZB mice reduced the IgM antihapten response to TNP-Ficoll and TNP-lipopolysaccharide to the samedegreeas in normal strains. The secondary IgG anti-hapten responseto a T-dependent antigen, however, was resistant to this regimen. Since this degree of haptenation of the CMC carrier was nearly that of the conjugate capable of tolerizing NZB B cells in the present study (TNPi,-CMC (Fig. 4) and TNP,,-CMC (data not shown) and the dose of tolerogen was much greater, the studies are in accord. Defective tolerization of the primary IgM antihapten response has more recently been documented by Brooks and Aldo-Benson using B-cell lines from (NZB X NZW)F, mice (29) and by Cowdery et al. in surface immunoglobulin-negative marrow cells from NZB mice using the splenic focus assay (30). These two reports are in agreement with our previous ( 1, 12) and current reports. Recent studies examining the early events associated with B-cell activation have documented a series of biochemical changes occurring at both the membrane and intracellular levels after crosslinking of surface immunoglobulin receptors (3 1). It is possible that hapten-specific tolerance induction might involve an alteration in the pathways associated with activation. Since NZB B cells are refractory to tolerance induction under defined conditions, they may serve as a negative control in studies designed to examine biochemical pathways associated with tolerance induction in normal mouse B cells. Such an analysis would be facilitated by the use of enriched populations of hapten-specific cells. Recent reports have indicated that the predominant DNA speciesin circulation in patients with SLE is a relatively low-molecular-weight speciesconsisting of as little as 30 to 50 base pairs and usually s200 base pairs (32-33). In view of the work by Papalian and colleagues (34), which has demonstrated that the minimal size required for divalent attachment of anti-DNA antibody to its ligand is 20 to 30 basepairs, it is conceivable that circulating DNA might in fact function as a low-epitope ligand. Such an autoantigen might ordinarily feed-back on anti-DNA-producing B cells induced either by polyclonal activators that bypass surface immunoglobulin receptors or by environmental, crossreactive, multivalent immunogens (35). The resistance to tolerance induction in New Zealand mice by low-epitope-density tolerogens irrespective of the chemical nature of the carrier demonstrated by the present and previous studies (1, 12) suggeststhat such hypothetical regulatory mechanisms might fail to operate in NZB mice and thereby contribute to autoantibody production.
DEFECTIVE B-CELL TOLERANCE IN NZB MICE
191
ACKNOWLEDGMENTS I thank Ms. Mary Waldschmidt and Ms. Lilli Hansen for excellent technical assistance; Mrs. Theresa Stencer for skillful secretarial assistance; Dr. Ellen S. Vitetta for purified monoclonal anti-Thy 1.2 antibody; and Dr. Peter E. Lipsky for critical review of the manuscript.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
Goldings, E. A., J. Immunol. 131,2630, 1983. Waldschmidt, T. J., Borel, Y., and Vitetta, E. S., J. Immunol. 131,2204, 1983. Bijsterbosch, M. K., and Klaus, G. G. B., J. Exp. Med. 162, 1825, 1985. Magilavy, D. B., Rifai, A., and Plotz, P. H., J. Immunol. 126,770, 198 1. Frank, M. M., Lawley, T. J., Hamburger, M. I., and Brown, E. J., Ann. Intern. Med. 98,206, 1983. Cohen, P., Ziff, M., and Vitetta, E. S., J. Immunol. 121,973, 1978. Chused, T. M., Moutsopoulos, H. H., Sharrow, S. O., Hansen, C. T., and Morse, H. C., III, In “Genetic Control of Autoimmune Diseases” (N. R. Rose, P. E. Bigazzi, and N. L. Warner, Eds.), pp. 177190. Elsevier/North Holland, New York, 1978. 8. Nossal, G. J. V., Pike, B. L., and Katz, D. H., J. Exp. Med. 138,3 12, 1973. 9. Goldings, E. A., and Mosier, D. E., Eur. J. Immunol. 9,76, 1979. 10. Diner, I-1.E., Kunimoto, D., and Diener, E., J. Immunol. 122, 1886, 1979. 11. Cambier, J. C., Vitetta, E. S., Uhr, J. W., and Kettman, J. R., J. Exp. Med. 145,778, 1977. 12. Geldings, E. A., Cohen, P. L., McFadden, S. F., Ziff, M., and Vitetta, E. S., J. Exp. Med. 152, 730, 1980. 13. Mishell, R., and Dutton, R. W., Science 153, 1004, 1966. 14. Jeme, N. D., and Nordin, A. A., Science 140,405, 1963. 15. Rittenbe:rg, M. R., and Pratt, K. L., Proc. Sot. Exp. Biol. Med. 132,575, 1969. 16. Jondal, M., Holm, G., and Wigzell, H., J. Exp. Med. 136,307, 1972. 17. Mosier, D. E., and Subbarao, B., Immunol. Today 3,2 17, 1982. 18. Mosier, D. E., Mond, J. J., andGoldings, E. A., J. Immunol. 119, 1877, 1977. 19. Theofilopoulos, A. N., Eisenberg, R. A., Bourdon, M., Crowell, J. S., and Dixon, F. J., J. Exp. Med. 149,516, 1979. 20. Mond, J. J., Scher, I., Mosier, D. E., Blaese,M., and Paul, W. E., Eur. J. Immunol. 8,459, 1978. 2 1. Sharon, R., McMaster, P. R. B., Kask, A. M., Owens, J. D., and Paul, W. E., J. Immunol. 114, 1585, 1975. 22. Nakajima, P. B., Datta, S. K., Schwartz, R. S., and Huber, B. T., Proc. Natl. Acad. Sci. USA 76,46 13, 1979. 23. Taurog, J. D., Raveche, E. S., Smathers, T. A., Glimcher, L. H., Huston, D. P., Hansen, C. T., and Steinberg, A. D., J. Exp. Med. 153,221, 1981. 24. Manoha.r, V., Brown, E., Leiserson, W. M., and Chused, T. M., J. Immunol. 129,532, 1982. 25. Hayakawa, K., Hardy, R. R., Parks, D. R., and Herzenberg, L. A., J. Exp. Med. 157,202, 1983. 26. Hayakawa K., Hardy, R. R., Honda, M., Herzenberg, L. A., and Steinberg, A. D., Proc. Natl. Acad. Sci. USA 81,2494, 1984. 27. Smith, II. R., Yaffe, L. J., Chused, T. M., Raveche, E. S., Klinman, D. M., and Steinberg, A. D., Cell. Immunol. 92, 190, 1985. 28. Tsunetsugu, Y., and Fujiwara, M., Cell. Immunol. 701,303, 1982. 29. Brooks, M. S., and Aldo-Benson, M., J. Clin. Invest. 78,784, 1986. 30. Cowdery, J. S., Jacobi, S. M., Pitts, A. K., and Tyler, T. L., J. Immunol. 138,760, 1987. 3 1. Cambier, J. C., Monroe, J. G., Coggeshall, K. M., and Ramsom, J. T., Immunol. Toduy6,2 18, 1985. 32. Morimoto, C., Sano, H., Abe, T., Homma, M., and Steinberg, A. D., J. Immunol. 139, 1960, 1982. 33. McCoubrey-Hoyer, A., Okarma, T. B., and Holman, H. R., Amer. J. Med. 77,23, 1984. 34. Papalian, M., Lafer, E., Wong, R., and Stollar, B. D., J. Clin. Invest. 65,469, 1980. 35. Schwartz, R. S., and Stollar, B. D., J. Clin. Invest. 75,321, 1985.