Proc. Natl. Acad. Sci. USA Vol. 83, pp. 5247-5251, July 1986

Antigen presentation by a B-cell line transfected with cloned immunoglobulin heavy- and light-chain genes specific for a defined hapten (helper T hybridomas//protoplast fusion) MOTOO WATANABE*, DALE R. WEGMANNt, ATSUO OCHI*, AND NOBUMICHI HozuMi* *Mount Sinai Hospital Research Institute, Department of Medical Genetics and Immunology, University of Toronto, 600 University Avenue, Toronto, ON, MSG 1X5, Canada; and tLilly Research Laboratories, 3252 Holiday Court, La Jolla, CA 92037 Communicated by Elvin A. Kabat, March 24, 1986

ABSTRACT The rearranged genes encoding immnunoglob- of presentation are not well-understood. To elucidate ulin heavy (mi) and light (K) chains specific for the hapten the biochemical and molecular events involved in this im- 2,4,6-trinitrophenyl (Tnp) were introduced into a B-lymphoma portant , it is necessary to study monoclo- line that bears surface IgG with an unknown specificity and nal B cells and T cells specific for defined . expresses surface Ia molecules. A transformant expressing Lanzavecchia (13) utilized an Epstein-Barr virus-trans- surface IgM specific for Tnp was obtained. The transformant formed human B-cell line specific for tetanus toxoid. These was found to present Tnp- to antigen ()-specific experiments elegantly demonstrated T-cell-B-cell collabora- T cells far more efficiently than the parental B-lymphoma line. tion; however, further genetic analysis would be difficult. This artificial-system, utilizing recombinant DNA technology The development of a mouse system would be more suitable and gene transfer, provides several approaches for the study of for fine biochemical genetic analysis due to the availability of T-cell-B-cell interactions. well-characterized recombinant mouse strains, T-cell clones, and monoclonal . Corley et al. (14) have used A T-cell-dependent immune response to a soluble antigen is hyperimmunization to establish a mouse monoclonal B-cell generated by the collaboration of three functionally distinct line specific for sheep erythrocytes. However, it is not clear immunocompetent cell populations, helper T (Th) cells, which are recognized by Ig and T-cell receptors in -producing cells (B cells), and antigen-presenting this system; therefore, it would be difficult to carry out cells (APCs) (1, 2). is a prerequisite for precise biochemical analysis of antigen processing and pre- the generation of an immune response. A variety of APCs sentation. For several reasons, a hapten-carrier system bearing Ia molecules on the membrane surface, including would be desirable for assessment of antigen processing and and dendritic cells, have been shown to activate presentation. Such a system would allow for the availability Th cells by antigen presentation (3-5). The antigen-presenting of defined antigenic specificities as well as flexibility in ability of macrophages has been well-documented (6). Fur- carrier choice. Further, much in vivo data on the immune ther, a number of experiments have established that antigen- response has been compiled by utilizing the hapten-carrier reactive Th cells recognize antigen presented in the context system. of Ia molecules encoded by the I region of the major Methods to construct T-cell lines and T hybridomas spe- histocompatibility complex (MHC). That is, Th cells and cific for defined antigens have been documented (15). In APCs must share the same haplotype of class II molecules contrast, there exists no established method in obtaining (MHC restriction) (7). antigen-specific B-cell lines. One of the goals of an immu- Recently, several investigators have turned their attention nologist is to establish monoclonal B-cell lines expressing sIg to B cells and their ability to present antigen to Th cells. with a defined hapten specificity. To this end, we have Several B-lymphoma lines bearing Ia molecules, as well as established a system in which recombinant DNA and gene- normal resting B cells, have been shown to have antigen- transfer techniques can be used to construct monoclonal presenting activity (8, 9). Each clone ofB cells carries surface B-cell lines expressing sIgM specific for a defined hapten. immunoglobulin (sIg) receptors specific for an antigen. These A previous report (16) described functional Ig production receptors may have an important role not only in the in myelomas and B-cell hybridomas after transfection of a activation of B cells but also in the efficient presentation of vector carrying the rearranged genes encoding A heavy chain antigen to T cells, as suggested by Chesnut and Grey (10). and K light chain (/krnp and KTnp) specific for the hapten Tnp. In a hapten-carrier system, B cells recognize haptenic More recently, the same vector was transferred into a determinants, whereas T cells recognize carrier determi- pre-B-like cell line (17). Although transformants expressing nants. By utilizing an enriched B-cell population specific for 1gM specific for Tnp (sIgMTnp) were obtained, these the hapten 2,4,6-trinitrophenyl (Tnp), Rock and coworkers transformants could not collaborate with Th cells in an (11, 12) showed that antigen (Tnp-modified protein) recog- antigen-specific manner, presumably due to the lack of Ia nized by the B-cell population was presented to Th cells molecules. In the present work, the same vector has been (termed specific antigen presentation) approximately 1000- transferred into a B-lymphoma line bearing Ia molecules and fold more efficiently than nonmodified protein by the same IgG with an unknown specificity. A resultant transformant B-cell population (nonspecific antigen presentation). expressing sIgMTnp was analyzed for specific antigen pre- Cellular properties of antigen presentation have been sentation mediated via sIg receptor molecules. well-studied. In contrast, biochemical and molecular aspects Abbreviations: APC, antigen-presenting cell; IL-2, interleukin 2; The publication costs of this article were defrayed in part by page charge KLH, keyhole limpet hemocyanin; MHC, major histocompatibility payment. This article must therefore be hereby marked "advertisement" complex; Ova, ; slg, surface immunoglobulin; Th cell, in accordance with 18 U.S.C. §1734 solely to indicate this fact. helper ; Tnp, 2,4,6-trinitrophenyl. 5247 5248 Immunology: Watanabe et al. Proc. Natl. Acad. Sci. USA 83 (1986) MATERIALS AND METHODS Sachs (National Institutes of Health, Bethesda, MD). The B-cell hybridoma (anti-Sp603) producing antibody against the Cell Lines. A B-lymphoma line, A20-2J (H-2d) (18), pro- of IgMTnP of Sp6 was a gift from G. Kohler (Max ducing sIgG of an unknown specificity and surface Ia antigen, Planck Institute, Freiburg, F.R.G.). was obtained from D. McKean (Mayo Clinic, Rochester). An Flow Microfluorimetric Analysis. Cells (5 x 105) were antigen-specific T-cell hybridoma (CAK1-22) specific for the incubated for 45 min with -conjugated goat anti- carrier keyhole limpet hemocyanin (KLH) was established mouse ,t antibody (Cappel Laboratories, Cochranville, PA) by the method of Harwell et al. (19). In brief, T cells were or with either biotinylated anti-I-Ad antibody (34-5-3S) or obtained by draining lymph nodes of (BALB/c x A/J)F1 biotinylated anti-idiotypic antibody (anti-Sp6O3). After wash- mice that had been injected through the foot pad with 100 ,g ing in Dulbecco's phosphate-buffered saline containing 2% of KLH. These T cells were then fused with BW5147 murine fetal bovine serum, the cells that had been treated with the lymphoma cells. A resultant line, CAK1-22, used in the biotinylated antibodies were incubated for another 45 min on following experiments was further characterized and found to ice with fluorescein-labeled avidin (Sigma). The cells were be Thy-l-positive, Lyt-2-negative, L3T4-positive, and I-Ed_ washed again and analyzed with an EPICS-C Coulter flow restricted. A T-cell hybridomna (3D0 18.3) that is ovalbumin microfluorimeter on a logarithmically amplified scale. (Ova)-specific and I-Ad-restricted was obtained from P. Assays for Antigen Presentation. The antigen-presentation Marrack (20). Interleukin 2 (IL-2)-dependent cell lines HT-2 assay was carried out by a slightly modified method of and CTL.L, utilized for assessing IL-2 production by the Kappler et al. (27). Unless otherwise indicated, 105 APCs and T-cell hybridomas after antigen presentation, were main- 5 x i104 T cells were cultured in 96-well plates in 0.2 ml of tained in RPMI 1640 medium containing 10% fetal bovine RPMI 1640 medium containing 10% fetal bovine serum, in the serum and 10% supernatant from concanavalin A (Con presence or absence of antigen. Prior to coculture, the APCs A)-stimulated rat spleen cells (21). were inactivated by treatment with mitomycin C (Sigma) at Vectors and Gene Transfer. The rearranged 9fnp- and 50 jig per ml at 370C for 30 min. After 24 hr of culture, KTnp-chain genes specific for Tnp were cloned from a hybrid- supernatants were collected for assay of IL-2 production. oma (Sp6) making IgMTnp (16, 22). The /UTnp and KTnp genes IL-2-dependent cell lines (CTL.L or HT-2) were used to were inserted into the vector pSV2neo, which confers resist- assay IL-2-dependent DNA synthesis. CTL.L (8 x 103) or ance to the antibiotic G418, a derivative of neomycin (Fig. 1) HT-2 (4 x 103) cells were incubated for 15-20 hr in 100 ,ul of (23). The vector (pR-HLTNP) was transfected into rj mj medium (RPMI 1640 with 10% fetal bovine serum) containing Escherichia coli strain K803. A protoplast fusion technique 10-20 ,ul of supernatant. DNA synthesis was estimated by was utilized to transform the A20-2J cells with pR-HLTNP incorporation of [methyl-3H]thymidine (1 ACi per well, 25 (24). One transformant, designated A20-HL, was used for Ci/mmol, Amersham; 1 Ci = 37 GBq) during an 8- to 12-hr further study. period. Data represent averages calculated from triplicate Antigens and Antibodies. Hapten (Tnp) conjugates of KLH cultures. and Ova were prepared as described (25). Tnp-Ova was calculated to carry =6 mol of Tnp per mol (43,500 g) of Ova. Tnp-KLH was calculated to bear =6 mol of Tnp per 100,000 RESULTS g of KLH (25). Culture supernatant from the B-cell hybrid- Transformation of a B-Lymphoma Line. In order to estab- oma line 14-4-4S was used as a source for anti-I-Ed antibody lish a cellular system in which specific antigen presentation (26). The B-cell hybridomas MK-D6, 28-16-8S, and 34-5-3S via sIg receptor molecules can be analyzed, a cell line bearing were used as sources ofanti-I-Ad antibodies (27-29). MK-D6, Ia molecules is required. The B-lymphoma line A20-2J, 14-4-14S, and 28-16-8S were purchased from the American expressing sIgG with an unknown specificity and surface Ia Type Culture Collection. 34-5-3S was obtained from D. H. molecules (H-2d), was chosen as the recipient. pR-HLTNp was introduced into A20-2J cells by a modified protoplast EcoRi B(3mHI fusion method (24). G418-resistant transformants were screened for sIgM production by using fluorescein isothio- cyanate-conjugated goat anti-mouse , antibody. One of the transformants (A20-HL) was chosen for further analysis. The parental line A20-2J secretes endogenous IgG to some extent. A Tnp-specific enzyme-linked immunosorbent assay (ELISA) revealed that the A20-HL cells secreted =200 ng of IgMTnP per ml of medium (24 hr), whereas the hybridoma line Sp603 (a derivative of Sp6) secreted =5 ug of IgMTnP per ml. To confirm the expression of Ig and Ia molecules, A20-HL and A20-2J cells were further characterized by flow microfluori- metric analysis (Fig. 2). The A20-HL cells gave a positive signal for sIgM expression, whereas the parental A20-2J cells did not show any significant fluorescein. The idiotype of the sIgM was analyzed by using an anti-idiotypic antibody (anti-Sp603). The A20-HL cells showed a much stronger fluorescein signal than the A20-2J cells, which exhibited EcoRI some background signal. This background may be due to nonspecific binding of the biotinylated anti-idiotypic anti- FIG. 1. Structure of the transducing vector pR-HLTNp. The body to the membrane surface of the A20-2J cells. These vector carries the rearranged pnp (16 kilobase pairs) and KTnp (9.6 results indicate that the A20-HL cells express sIgMTnP. These kilobase pairs) genes at the EcoRI and BamHI sites of pSV2neo, cells were also stained with a biotinylated anti-I-Ad antibody respectively (16). Black boxes indicate coding segments for the (34-5-3S). Expression of I-A molecules was not significantly Tnp-specific heavy-chain variable region (VHTNP), for the a heavy- different and chain constant (C,) and membrane (M) regions, and for the K between the A20-HL A20-2J cells. light-chain constant (CK) and Tnp-specific variable (VKTNP) regions. Presentation of Hapten-Modiflied Antigen by A20-HL Cells The directions of transcription of the Ig genes and the simian virus to a Th-Hybridoma Line. The ability of A20-HL to present 40 (SV40) early promoter are indicated by arrows. antigen (Tnp-KLH) to the KLH-specific, I-Ed-restricted Immunology: Watanabe et al. Proc. Natl. Acad. Sci. USA 83 (1986) 5249

d 0C-

UO It 8, C.) L-i(D X 1 10 *D a Tnp-KLH, Ag/ml E 'E 0 z Ic

Anti-l-Ad 5.

256 - IA

1 100 64 128 192 256 KLH, Ag/ml Channel number FIG. 3. Antigen presentation to KLH-specific Th cells (CAK1- 22). CAK1-22 cells (5 x 104) were cultured with 105 APCs in the FIG. 2. Flow microfluorimetric analysis of A20-2J (traces a) and presence ofthe antigen Tnp-KLH (A) or KLH (B) as indicated. HT-2 A20-HL (traces b) cells stained with fluoresceinated goat anti-mouse cells (4 x 103) were used for assay of IL-2-dependent [3H]thymidine ,u antibody (Top) or with either biotinylated anti-idiotypic antibody incorporation. Broken and solid lines indicate A20-2J and A20-HL, (anti-Sp603, Middle) or biotinylated anti-I-Ad antibody (34-5-3S, respectively. Bottom) followed by fluoresceinated avidin. Spleen cells from BALB/c (H-2d) mice were stained with 34-5-3S, but not spleen cells from CBA/J (H-2k) mice (data not shown). About 104 cells were If sIgMTnp receptors were involved in specific antigen analyzed for each tracing. Channel number is proportional to the presentation, anti-Sp603 antibody should block specific an- logarithm of the fluorescence. tigen presentation by A20-HL. As expected, the anti-idio-

Th-cell hybridoma line CAK1-22 was tested (Fig. 3A). As a Table 1. Inhibition of specific antigen (Tnp-KLH) presentation control, the original parental line A20-2J was used. Antigen- by competitor (Tnp-Ova) presentation activity ofthe A20-HL cells was saturated at 0.1 [3H]Thymidine ,ug/ml, whereas A20-2J required 200 of the same antigen ,ug incorporation per ml to reach a similar saturation level. When nonmodified Tnp-KLH, Tnp-Ova, APCs inhibition antigen (KLH) was utilized, both the A20-HL and A20-2J bug/ml ,ug/ml cpm % cells required higher concentrations of the antigen (250-500 Experiment I ,ug/ml) to reach saturation levels (Fig. 3B). The amount of A20-2J 0 0 5,254 KLH antigen required for maximal saturation showed no 200 0 58,689 0 significant difference between the A20-HL and A20-2J cells. 40 48,604 19 25 These results suggest that sIgMTnp receptors play an impor- 200 45,554 tant role for efficient specific-antigen presentation. A20-HL 0 0 11,123 0.5 0 0 Blocking of Specific Antigen Presentation.-If re- 81,659 sIgMTnp 8 10,549 100 were for the efficient of ceptors responsible presentation 40 13,123 97 Tnp-KLH, addition of other Tnp-proteins would be expected Experiment 2 to block antigen presentation of Tnp-KLH to the CAK1-22 A20-HL 0 0 1,297 cells. To study this point, Tnp-Ova was added to Tnp-KLH- 1.0 0 50,054 0 containing reaction media (Table 1). Antigen presentation by 0.2 12,103 78 the A20-HL cells was completely abrogated by the addition 0.6 3,211 96 of a low dose (8 ,g) of Tnp-Ova. Antigen presentation by x were cocultured with APCs and was same treatment. CAK1-22 cells (5 104) (105) A20-2J not significantly affected by Tnp-KLH in the presence or absence of the competitor Tnp-Ova. When the competitive inhibition by Tnp-Ova was more HT-2 cells (4 x 103) and CTL.L cells (8 x 103) were used in precisely titrated, it was found that 0.2 ,ug of Tnp-Ova experiments 1 and 2, respectively, for [3H]thymidine-incorporation inhibited presentation of 1 ,g of Tnp-KLH by "80%. assay for IL-2. 5250 Immunology: Watanabe et al. Proc. Natl. Acad. Sci. USA 83 (1986) typic antibody did inhibit specific antigen presentation (Table Table 3. Effect of monoclonal anti-Ia antibodies on 2). The same treatment with the anti-idiotypic antibody did Tnp-KLH presentation not affect antigen presentation by A20-2J. T-cell-B-cell interactions are strictly regulated by MHC Anti-Ia [3H]Thymidine restriction; therefore, the effect of anti-Ia antibodies on Tnp-KLH, antibody incorporation antigen presentation was tested (Table 3). CAK1-22 is I-Ed_ APCs Ag/ml (specificity) cpm % inhibition restricted. An anti-I-Ed antibody was shown to inhibit both A20-2J 0 - 4,654 specific and nonspecific antigen presentation, whereas anti- 200 - 53,300 0 I-Ad antibodies had no significant effect. Ova-specific Th cells 14-4-4S (I-Ed) 10,078 89 (3D0 18.3) are I-Ad-restricted (20). The anti-I-Ed antibody did MK-D6 (I-Ad) 43,714 20 not block the Ova-specific antigen presentation significantly, 28-16-8S (I-Ad) 44,259 19 but the anti-I-Ad antibodies abrogated the activity (data not 34-5-3S (I-Ad) 47,477 12 shown). These results confirm that specific antigen presen- A20-HL 0 - 7,154 tation is mediated via Ia molecules. 0.4 - 206,930 0 The Recombinant Vector as a Universal System for Specific 14-4-4S (I-Ed) 10,282 98 Antigen Presentation. We have shown that the antigen Tnp- MK-D6 (I-Ad) 210,564 0 KLH is presented to KLH-specific Th cells very efficiently. 28-16-8S (I-Ad) 167,957 20 With the appropriate choice of carrier, a wide variety of 34-5-3S (I-Ad) 157,437 25 antigen-specific Th cells can be utilized. Tnp-Ova was used for antigen presentation to Ova-specific cells CAK1-22 cells (5 x 104) were cocultured with APCs (105) and Th (Fig. 4). Tnp-KLH in the presence or absence of antibodies. Final dilution of A20-HL cells presented Tnp-Ova to Th cells -1000-fold more culture supernatants containing antibodies was 1:8. IL-2 production efficiently than did the A20-2J cells. was measured by [3H]thymidine incorporation of 4 x 103 HT-2 cells. DISCUSSION major role in antigen presentation in the immune response in The recombinant vector pR-HLTNP was transfected into the vivo. Generally, low amounts of antigen are required to elicit B-lymphoma line A20-2J. Several experiments indicated that a secondary immune response in vivo. the resultant transformant (A20-HL) expressed IgMTnp on the We found that competing antigen (Tnp-Ova) at 200 ng/ml membrane surface. No significant difference in expression of could block specific presentation of Tnp-KLH at 1 ,ug/ml by Ia molecules was found between the A20-HL and A20-2J "80%. The density of Tnp on the carriers was approximately cells. Since several investigators have suggested that in- 6 mol ofTnp per 43,500 g of Ova and 100,000 g of KLH. Thus creased expression of Ia molecules on APCs enhances 440 ng of Tnp-Ova and 1 ,ug of Tnp-KLH are equimolar with antigen-presenting activity (6, 30, 31) and since the experi- respect to Tnp molecules. These results indicate the two ments presented in this paper were designed to examine the hapten-carrier systems compete with each other on an ability of sIgMTnp to present specific antigen, this observation equimolar basis. The anti-idiotypic antibody against IgMTnp is critical. could abrogate specific antigen presentation. These results Antigen presentation of the hapten-carrier (Tnp-KLH) by clearly indicate that sIgMTflp molecules played an important A20-HL to the KLH-reactive, I-Ed-restricted Th hybridoma role in efficient antigen presentation. CAK1-22 was studied. The A20-HL cells were shown to At least two functions of sIg receptor molecules on B cells present Tnp-KLH, at a concentration as low as 100 ng/ml, to can be postulated. (i) slg receptors capture and focus antigen CAK1-22, resulting in the activation of T cells. This result is on the B-cell membrane surface. Bound antigen is efficiently comparable to the results obtained by utilizing enriched incorporated into B cells by endocytosis, processed inside normal Tnp-specific B cells (11, 12). The amount of Tnp- the B cells, and finally presented to T cells in association with KLH needed by the parental A20-2J cells to activate the Ia molecules (11-13). (ii) Crosslinking of sIg receptors by CAK1-22 cells was 2000-fold greater than that needed by the antigen leads to activation of B cells. This activation may A20-HL cells. Efficiency of nonspecific antigen presentation result in clonal expansion of B cells, production of growth by the A20-HL cells, however, did not exhibit any significant factors, increased expression of Ia molecules, or triggering of difference compared to that of the A20-2J cells. Because of B-cell differentiation (32, 33). the low dose of antigen required for specific antigen presen- Our gene-transfer system utilizing recombinant vectors tation by B cells, B cells bearing slg receptors may play a carrying rearranged Ig genes has obvious advantages over Table 2. Anti-idiotypic antibody (anti-Sp603) inhibits specific antigen (Tnp-KLH) presentation F 10- 0 [3H]Thymidine .0o Tnp-KLH, incorporation APCs ,ug/ml Anti-Sp603 cpm % inhibition 0. A20-2J 0 - 1,423 400 - 28,805 0 C E / 1:8* 24,778 15 EO 1:40* 24,975 14 A20-HL 0 - 1,565 I C- 1.0 - 43,683 0 O -~------0.001 0.01 0.1 1 10 100 1:8* 2,954 97 60 1000 1:40* 13,129 73 Tnp-Ova, ltg/ml CAK1-22 cells (5 x 104) were cocultured with APCs (105) and FIG. 4. Antigen presentation to Ova-specific Th cells (3D0 18.3). Tnp-KLH in the presence or absence of the anti-idiotypic antibody 3D0 18.3 cells (5 x 104) were cultured with 5 x 104 APCs in medium (culture supernatant from anti-Sp603). IL-2 production was mea- containing Tnp-Ova as indicated. CTL.L cells (8 x 103) were used for sured by [3H]thymidine incorporation of 8 x 103 CTL.L cells. assay of IL-2-dependent [3H]thymidine incorporation assay. Broken *Final dilution of anti-Sp603 culture supernatant. and solid lines indicate A20-2J and A20-HL, respectively. Immunology: Watanabe et al. Proc. Natl. Acad. Sci. USA 83 (1986) 5251

other systems described in the Introduction. (i) Virtually any 9. Ashwell, J. D., DeFranco, A. L., Paul, W. E. & Schwartz, B-cell line at different stages ofontogeny can be transformed. R. H. (1984) J. Exp. Med. 159, 881-905. Transformed B-cell lines at different differentiation stages 10. Chesnut, R. W. & Grey, H. M. (1981) J. Immunol. 126, 1075-1079. will be quite valuable for the study of B-cell activation and 11. Rock, K. L., Benacerraf, B. & Abbas, A. K. (1984) J. Exp. differentiation after antigenic stimulation or T-cell-B-cell Med. 160, 1102-1113. interaction. (ii) This system has flexibility with regard to 12. Abbas, A. K., Haber, S. & Rock, K. L. (1985) J. Immunol. carrier choice. The flexibility of our system was demonstrat- 135, 1661-1667. ed by changing the carrier protein from KLH to Ova. 13. Lanzavecchia, A. (1985) Nature (London) 314, 537-539. Tnp-Ova was presented to Ova-specific Th cells -1000-fold 14. Corley, R. B., LoCascio, N. J., Ovnic, M. & Haughton, G. more efficiently by A20-HL than by A20-2J. (iii) By manip- (1985) Proc. Natl. Acad. Sci. USA 82, 516-520. ulating the cloned Ig genes, we can construct Ig receptors 15. Fathaman, C. G. & Frelinger, J. G. (1983) Annu. Rev. Immu- differing in but encoding the same antigen specificity. nol. 1, 633-655. receptors, receptors, 16. Ochi, A., Hawley, R. G., Hawley, T., Shulman, M. J., The role of such Ig for instance IgD in Traunecker, A., Kohler, G. & Hozumi, N. (1983) Proc. Natl. antigen presentation and in B-cell activation and differenti- Acad. Sci. USA 80, 6351-6355. ation can be investigated. 17. Ochi, A. & Hozumi, N. (1986) J. Immunol. 136, 2705-2708. We have described use of recombinant DNA and gene- 18. Chesnut, R. W., Colon, S. M. & Grey, H. M. (1982) J. Immu- transfer techniques to establish a monoclonal B-cell line nol. 129, 2382-2388. bearing specific sIg receptor molecules. This system will be 19. Harwell, L., Skidmore, B., Marrack, P. & Kappler, J. W. useful in the study of several fundamental problems in (1980) J. Exp. Med. 152, 893-904. cellular immunology at the molecular level, such as antigen 20. Shimonkevitz, R., Kappler, J. W., Marrack, P. & Grey, H. presentation in relation to B-cell activation, the effect of (1983) J. Exp. Med. 158, 303-316. 21. Rock, K. L. & Benacerraf, B. (1983) Immunol. Rev. 76, 29-57. crosslinking of sIg receptors during antigen presentation, 22. Hawley, R. G., Shulman, M. J., Murialdo, H., Gibson, D. M. processing ofantigen, and biochemical changes in specific vs. & Hozumi, N. (1982) Proc. Natl. Acad. Sci. USA 79, nonspecific antigen presentation. 7425-7429. 23. Southern, P. J. & Berg, P. (1982) J. Mol. Appl. Genet. 1, We thank J. Pawling for expert technical assistance, Roland Tisch 327-341. and Joan Richardson for critical comments, and Nobukata Shinohara 24. Ochi, A., Hawley, R. G., Shulman, M. J. & Hozumi, N. (1983) for helpful suggestions and discussion. This work was supported by Nature (London) 302, 340-342. the National Cancer Institute and the Medical Research Council. 25. Good, A. H., Wofsy, L., Henry, C. & Kimura, J. (1980) in Selected Methods in Cellular Immunology, eds. Mischell, 1. Katz, D. H. & Unanue, E. R. (1973) J. Exp. Med. 137, B. B. & Shiigi, S. M. (Freeman, San Francisco), pp. 343-350. 967-990. 26. Ozato, K., Mayer, N. & Sachs, D. H. (1980) J. Immunol. 124, 2. Swierkosz, J. E., Rock, K., Marrack, P. & Kappler, J. W. 533-540. (1978) J. Exp. Med. 147, 554-570. 27. Kappler, J. W., Skidmore, B., White, J. & Marrack, P. (1981) 3. Rosenthal, A. S. & Shevach, E. M. (1973) J. Exp. Med. 138, J. Exp. Med. 153, 1198-1214. 1194-1212. 28. Ozato, K. & Sachs, D. H. (1981) J. Immunol. 126, 317-321. 4. Erb, P. & Feldman, M. (1975) J. Exp. Med. 142, 460-472. 29. Ozato, K., Mayer, N. M. & Sachs, D. H. (1982) Transplanta- 5. Sunshine, G. H., Katz, D. R. & Feldman, M. (1980) J. Exp. tion 34, 113-120. Med. 152, 1817-1822. 30. Beller, D. I. (1984) Eur. J. Immunol. 14, 138-143. 6. Unanue, E. R. (1984) Annu. Rev. Immunol. 2, 395-428. 31. Bekkhoucha, F., Naquet, P., Pierres, A., Marchetto, S. & 7. Haskins, K., Kappler, J. W. & Marrack, P. (1984) Annu. Rev. Pierres, M. (1984) Eur. J. Immunol. 14, 807-814. Immunol. 2, 51-66. 32. Monroe, J. G. & Cambier, J. C. (1983) J. Exp. Med. 158, 8. McKean, D. J., Infante, A. J., Nilson, A., Kimoto, M., Fath- 1589-1599. man, C. G., Walker, E. & Warner, N. (1981) J. Exp. Med. 154, 33. Kakiuchi, T., Chesnut, R. W. & Grey, H. M. (1983) J. Immu- 1419-1431. nol. 131, 109-114.