Post-transcriptional regulation of MHC class II expression in human T cells

CELLULAR

IMMUNOLOGY

139,98-107 (1992)

Post-transcriptional Regulation of MHC Class II Expression in Human T Cells HOWARD S.CAPLEN,*

*Departments

SILVIA SALVADORI,* BERNDGANSBACHER,? ANDKARENS.ZIER*

of Medicine and Microbiology, Mt. Sinai School of Medicine, and tDepartment of Hematology/Lymphoma, Memorial-Sloan Cancer Center, New York, New York 10021 Received

May

21, 1991; accepted

September

New York, New York Kettering

10029;

9, 1991

Human T lymphocytes are among those cells which are cell surface class III in the resting state, but can be induced to express class II following treatment with appropriate stimulators. Although resting T cells do not express detectable surface class II, cell surface class II can be detected on purified T cells as early as 30 min following stimulation with PHA and PMA, well before the initiation of DNA synthesis, and the percentage of positive cells gradually increases with time. One hypothesis explaining this very rapid surface expression of class II is that the genes can be regulated post-transcriptionally in T cells. To test this, we used nuclear run-on assaysto measure the transcriptional rate of diverse class II genes in resting and activated T cells. Our results demonstrated that transcripts for DR, DP, and DQ could be detected in cells which were neither dividing nor transcribing mRNA for another marker of T cell activation, the IL-2 gene. Northern blot analysis demonstrated low to moderate steady-state levels of DRP mRNA in these cells. Moreover, treatment of activated T cells with cycloheximide resulted in superinduction of class II for DR, DQ, and DP. These results suggest that resting T cells can transcribe mRNA for class II genes, but that they do not express the protein product on the cell surface in a detectable way until following activation. In addition, they suggest that there may be a protein factor which negatively influences class II levels in T cells. Thus, the regulation of class II in T cells is complex and involves post-transcriptional regulation, at least in part. 0 1992 Academic Press, Inc.

INTRODUCTION The activation of normal T cells involves a series of events resulting in the acquisition of a number of functional properties not associated with resting T cells. During this process human T cells express a number of activation antigens, among them HLAclass II DR, DP, and DQ (l-7). Thus, resting T cells are among those cells which do not express surface class II, but can be induced following appropriate stimulation. As part of defining the functional role of class II determinants on T cells, it is informative to understand the molecular mechanisms underlying the regulation of their expression. Although class II molecules play a pivotal role in regulating the immune response, their role in T cell function is unknown. Two possibilities, analogous to their role on non-T cells, for which we and others have presented data is that they function in antigen presentation (8-12) and/or signal transduction (13-16). Despite the fact that resting T cells are surface class III, we previously reported that class II surface levels could be detected on a subset of purified T cells as early as 30 min following stimulation 98 0008-8749/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved

REGULATION

OF CLASS II IN T CELLS

99

with phytohemagglutinin (PHA) and phorbol myristic acid (PMA), well before the initiation of DNA synthesis, and that the percentage of positive cells increased with time (17). Northern and dot blot analysis revealed that class II mRNA accumulation could be detected 15 min after stimulation, the levels of which also increased with time. Immunoblot analysis of lysates using an anti-DR monoclonal antibody detected low levels, 17% of those in activated T cells, of a polypeptide of -28 kDa in the resting T cells. This suggested that, in addition to the synthesis of new class II molecules following activation, there might be some preformed polypeptides either within a cytoplasmic fraction or buried within the membrane in a cryptic form. Thus, the regulation of the expression of class II in T cells was complex and could be regulated via several pathways, including one which involved post-transcriptional mechanisms. To test this, we looked at transcription of class II in resting cells using nuclear runon assays. Transcripts for DR, DP, and DQ were detected in cells which were neither dividing nor transcribing the interleukin-2 (IL-2) gene, another marker for T cell activation. Next, we used Northern blot analysis to measure DR mRNA steady-state levels in resting T cells. We observed that in most cases low to moderate levels of DR/3 mRNA could be detected in resting cells. Moreover, treatment of class II+-activated T cells with cycloheximide resulted in superinduction of mRNA for DR, DQ, and DP. These results suggest that resting T cells can transcribe class II mRNA, but not express the protein product on the cell surface in a detectable way until following activation. In addition, they suggest that there may be a regulatory factor which negatively influences class II levels in T cells. MATERIALS

AND

METHODS

T Cell Isolation Purified resting T cells were prepared via negative selection with magnetic beads (Advanced Magnetics, Cambridge, MA). Briefly, peripheral blood lymphocytes (PBL) were resuspended in phosphate buffered saline (PBS) without Ca2+, Mg’+, 3% fetal bovine serum (FBS), called fluorescence-activated cell sorter (FACS) buffer, and an aliquot was removed for phenotypic analysis before separation. Cells to be purified were first allowed to adhere to plastic to deplete for adherent cells. The remaining cells then were centrifuged at 4°C at 1500 rpm for 10 min to obtain a pellet. One hundred microliters each/ lo6 cells of monoclonal antibodies directed against the nonT cell determinants CD1 1, CD 16, and CD 19 present on macrophages, NK cells, and B cells, respectively, in the form of culture supematants was added to the cell pellet. After gentle mixing, the cells were incubated for 30 min on ice in the dark, centrifuged as described above, and washed twice with FACS medium. Prewashed beads in FACS buffer were then added to the cell pellet (50 particles/106 cells). Following another 30 min incubation on ice in the dark, a magnet was used to draw the non-T cells which had bound the beads to one side of the flask. The T cells were then withdrawn using a pipet and transferred to a new flask. Following this procedure samples were tested in FACS analysis using an anti-CD3 monoclonal to assessthe purity of the separation. In 17 experiments we achieved a mean of 98.0% CD3+ cells + 1.8% (mean f SD). All cells were then allowed to “rest” overnight in RPM1 supplemented with 10% human serum at 37°C to allow any nonspecific activation events to dissipate before use. Activation of cells was as described in the text.

100

CAPLEN

ET AL.

Flow Cytometry Two hundred thousand lymphocytes for each antibody to be tested were resuspended in FACS buffer containing 0.02% sodium azide, and distributed into microtiter wells. The cells were centrifuged and the supernatant was poured off. One hundred microliters/ 1O6cells of appropriately diluted antibody was then added and the cells incubated on ice for 30 min. The cells were washed three times with 100 ~1 of cold FACS buffer and, after gently breaking up the cell pellet, 100 pl/ lo6 cells of fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse antibody was added to each sample. The cells were incubated on ice for another 30 min and washed as described above. They were then immediately analyzed for the percentage fluorescence with an Epics C equipped with logarithmic amplifiers for fluorescence measurements. The fluorescence originating from erythrocytes and dead cells was eliminated by threshold-gating of forward and 90” light-scatter signals. Isolation

of RNA, Northern Blotting,

Hybridization,

and Autoradiography

Total cellular RNA was isolated from T cells using RNAzol B and following the manufacturer’s instructions (Biotex). Briefly, cells were washed once in PBS and solubilized in RNAzol. One-tenth volume of chloroform was added to the homogenate, the tube vigorously shaken, and incubated on ice for 5 min. The suspension was centrifuged at 12,OOOg for 15 min and the aqueous phase containing the RNA was recovered. The RNA was precipitated by the addition of isopropanol followed by incubation on ice for 15 min. Following spectrophotometric determination of RNA yield, the sample was stored in TE at -70°C. The RNA was size-fractionated on a 1% agarose gel containing formaldehyde. Gels were stained with 0.5 pg/ml ethidium bromide to visualize the 18s and 28s ribosomal bands to assessthe integrity and also to confirm that approximately equal amounts of RNA were loaded per well. The RNA was transferred to a nylon filter (Gene Screen Plus, NEN-DuPont) using the technique of Northern blotting and then baked in a vacuum oven for 2 hr at 8O’C to fix the RNA. The fluid for prehybridization and for hybridization consisted of 0.2% bovine serum albumin, 0.2% polyvinyl pyrrolidone, 0.2% ficoll, 50 mM Tris-HCl (pH 7.5), 0.1% sodium pyrophosphate, 10% dextran sulphate, 1% sodium dodecyl sulphate (SDS), and 100 pg/ml salmon sperm DNA. After prehybridizing the filter in a plastic bag at 65°C for 4-18 hr, the bag was opened, and 1 X lo6 cpm/ml of a 32P-cDNA probe labeled by the random primer method was added (18). The bag was resealed and incubated overnight in a 65°C water bath. Subsequently, the filter was washed in a room temperature solution of 2X standard sodium citrate (SSC) and 0.5% SDS for 5 min, in a room temperature solution of 2X SSC and 0.1% SDS for 15 min, and then four times in a 55°C solution 0.1X SSC and 0.1% SDS for 30 min each time. The filter was blotted dry, wrapped in Saran Wrap, and exposed at -70°C to a piece of Kodak XAR X-ray film with an intensifying screen. The cDNA clones used for hybridization were DRPI (DR @), a 500-bp insert in the plasmid pBR 322 (19), kindly provided by Dr. Eric Long; p-II-j37 (DPP), a 600-bp insert in pBR322 (20); and p-II/31 (DQP), a 627-bp insert in pBR322 (2 l), kindly provided by Dr. Per Peterson; IL2, a 532-bp insert in p41, a gift of Dr. Gerald Crabtree (22); glyceraldehyde 3-phosphodehydrogenase (GAPD), a 1.2-kb insert in pBR322 obtained from the ATCC (23); and the epidermal growth factor receptor (EGFR), a 3.9-kb insert, a gift from Dr. J. Schlessinger (24).

REGULATION

OF CLASS II IN T CELLS

101

Nuclear Run-on Assays Nuclear run-ons were performed according to the procedure of Thompson and coworkers (25). Nuclei were isolated by disrupting the plasma membrane at 4°C with lysis buffer consisting of 10 rnA4 Tris (pH 7.4), 10 mM NaCl, 5 mM MgCl* , and 0.5% NP40. Nuclei were pelleted by centrifugation at 400g at 4°C. After removal of the supernatant and gentle vortexing, the pellet was resuspended in freezing buffer consisting of 50 mM Tris, (pH 8.3), 40% glycerol, 5 mM MgClz , and 0.1 mM EDTA and stored at -70”. To assesstranscription the nuclei were resuspended in a run on buffer composed of 25 mM Tris, (pH 8.0), 12.5 rnA4 MgC12, 750 mM KCl, 1.25 mM XTP (AGC), and 250 &i 32P-UTP and incubated at 30°C for 30 min. After incubation RNase-free DNase I was added with 10 mM CaC12, vortexed, and incubated 5 min at 30°C. One-tenth volume of 10X set buffer and 10 ~1 proteinase K (10 mg/ml) were added, the mixture transiently heated to 65°C and incubated at 42°C for 45 min. It was then phenol chloroform-extracted and precipitated with isopropanol. After centrifugation, the pellet was resuspended in Tris-EDTA (TE) and applied to a Sephadex G50 column. Radioactivity was checked and transcription was evaluated following prehybridization and hybridization for 24 hr of the labelled RNA with unlabelled cDNA bound to filter strips for 24 hr in a buffer made up of 10 mM TES, pH 7.4, 0.2% SDS, 10 mM EDTA, 250 /*g/ml Escherichia coli RNA, 0.3 M NaCl, and 1X Denhardt’s. When transcription in more than one preparation of cells was tested, equal numbers of counts per minute of labelled RNA were applied to replicate filters. After hybridization, the filters were washed two times in 0.1% SDS at room temperature and two times in 0.1X SSC and 0.1% SDS for 30 min at 65°C. Finally, they were treated with 10 pg/ml RNAse A for 30 min and exposed to XAR film (Kodak) at -70°C. RESULTS Class II Genes Are Transcribed in Both Resting and Activated T Cells Resting T cells in peripherd blood express no or very low levels of cell surface class II antigens, but become strongly positive following treatment with mitogens such as PHA or stimulation with specific antigen (results not shown). In previous studies we reported that increased steady-state levels of class II were detectable within 30 min of T cell activation and that resting T cells were positive on Western blots for DRP polypeptide chains (17). One possible explanation for these findings was that there was transcription of the class II genes in resting cells despite the lack of class II on the cell surface. To test this, purified T ceils were obtained from PBL using negative selection with magnetic beads as described under Materials and Methods. Subsequent testing with an anti-CD3 monoclonal antibody revealed that 99.4% of the cells were T cells (results not shown). Next, we performed run-on assaysusing nuclei from resting, class II surface negative purified T cells and activated, class II surface positive purified T cells. The positive control was GAPD, which is constitutively transcribed in T cells, and the negative control was EGFR, which is not transcribed in hematopoietic cells. Our results demonstrated that DR& DQP, and DPP were all transcribed in both resting and activated T cells (Figs. 1A and 1B). Moreover, DP had the highest transcriptional rate, followed by DR and DQ. As expected, neither the IL-2 nor the EGFR genes were expressed in the resting cells. The IL-2 was not expressed in the activated

102

CAPLEN

ET AL.

FIG. 1. Transcription of class II genes in resting (A) and activated (B) T cells. Nuclei were isolated from purified resting T cells or 7-day-old cultured T cells which had been activated with PHA and expanded in medium containing IL-2. Labeling of previously initiated mRNA transcripts was carried out as described under Materials and Methods and hybridized to filters on which the indicated unlabeled cDNA probes were bound. The shadow which appears on EGFR in A is due to nonspecific binding.

cells either, since these cells had been grown in exogenous IL-2 which results in downregulation of the endogenous gene (K. Zier, unpublished results). These findings suggested that class II expression in T cells might be regulated at least in part at the posttranscriptional level.

Expression of DR mRNA in Unstimulated, Resting T Cells Although we had evidence for transcription of class II genes in resting T cells, this did not imply that there would be detectable steady-state levels of class II mRNA in these cells. Evidence exists for post-transcriptional regulation of certain genes, such as that for c-myc, in T cells (26). To determine whether the mRNA transcribed in the nucleus was processed and transported to the cytoplasm we prepared purified resting T cells and isolated total cytoplasmic RNA. A portion of the cells was then activated with PHA and PMA and samples were taken for RNA isolation at 6, 12, 18, and 24 hr. Our positive control was GM 1500, a B cell line, and our negative control was HL60, a myelomonocytic cell line. Following electrophoresis of 3 pg of total cytoplasmic RNA per sample and transfer to a nylon filter, it was probed with a cDNA for the DRP chain. Staining of the gel with ethidium bromide indicated that slightly less material had been loaded at 12 and 18 hr compared to the other time points (results not shown). Figure 2 demonstrates that accumulation of DRP mRNA was detectable at all time points, including resting T cells, although levels for resting cells were lower

+-0

612 1824

FIG. 2. Northern blot analysis of steady-state levels of DR@ mRNA in resting T cells. RNA was isolated from purified resting T cells at indicated time points as described under Materials and Methods, electrophoresed on a 1% agarose gel, transferred to a nylon filter, and hybridized with cDNA for DRP. (+) represents the positive control, GM 1500, a B cell line. (-) represents the negative control, HL-60, a myelomonocytic cell line.

REGULATION

103

OF CLASS II IN T CELLS

than those observed for 24-hr activated T cells. Similar results have been observed in two additional experiments. Since the transcripts were of the expected size, it appeared that the primary transcripts had been appropriately processed. Taken together with earlier results demonstrating DR polypeptides in resting T cells using immunoblots (17), these data suggest that the lack of cell surface expression of class II on resting T cells might be due to a failure to transport or properly orient the molecule in the membrane. Cycloheximide

Treatment Increases Class II mRNA Levels in T Cells

Our results so far suggested that post-transcriptional mechanisms were involved in regulating class II expression in T cells. Studies on regulatory mechanisms affecting class II gene expression have demonstrated that there exist diverse regulatory proteins which bind to 5’ untranslated regions and which can exert either negative or positive influences (27-29). To determine whether there was evidence for a negative regulatory protein influencing the class II levels, T cells were stimulated with another phorbol ester, phorbol dibutyrate (PdB) and a calcium ionophore, ionomycin, or with 1% PHA for 72 hr and grown with and without cycloheximide (5 pg/ml) for the final 12 hr of culture. Vigorous proliferation was observed for T cells stimulated by PdB and ionomycin and low proliferation was seen for T cells stimulated with PHA (Table 1). RNA was isolated from these cells, electrophoresed, transferred to a filter, and probed for DRP. Staining with ethidium bromide demonstrated that similar amounts of material had been loaded per lane for cells stimulated with PdB and ionomycin and grown with or without cycloheximide and for cells stimulated with 1% PHA with or without cycloheximide, but that less material for both PHA samples had been loaded. Our results demonstrated that levels of accumulation of DR/3 mRNA following treatment with PdB and ionomycin were greatly increased in the presence of cycloheximide. The lower levels of DRP mRNA seen following stimulation with PHA were not affected by treatment with cycloheximide (Fig. 3). Similar results were observed for DQ@ and DPP (results not shown). Thus, the negative factor inhibited by cycloheximide appeared to be important in the PdB/ionomycin pathway, but not the PHA pathway.

TABLE 1 Proliferation of Purified T Cells to PdB/Ionomycin or PHA” Stimulus Medium PdB Ionomycin PdB/Ionomycin 1% PHA

Proliferation (cpm) 84 + 269 + 119k 45,450 f 1,677 f

44 24 40 895 565

a T cells were purified using magnetic beads as described under Materials and Methods. They were resuspended at a concentration of 0.05 X 106/well in complete medium. PdB was used at lo-* M, ionomycin at 1 PALM,and PHA at 1%. After 60 hr of culture they were labeled with 1 &i/well of ‘H-TdR for the final 12-15 hr of growth. Results represent the mean of triplicate samples from one representative experiment of9.

104

CAPLEN

ET AL.

FIG. 3. (A) Northern blot analysis of DFQ9steady-state levels in activated T cells with and without cycloheximide. (B) Ethidium bromide staining of gel.

DISCUSSION While it has been appreciated for about 10 years that human T cells convert from class II- to class II+ following activation, the underlying regulatory mechanisms have been little studied (l-7). The regulation of class II expression in general is complex and the control of these genes appears to differ depending upon the cell type studied. For example, studies by others have shown that class II regulation differs between B cells and macrophages (30). These studies are of importance because of the role that class II molecules play in antigen presentation and in signal transduction. Furthermore, it is believed that the inappropriate expression of class II molecules on certain cell types may be involved in the pathogenesis of autoimmune disease (31). We have previously observed that class II was detectable on the surface of purified T cells within 30 min of treatment with PHA and PMA (17). Furthermore, there were detectable steady-state levels on mRNA dot blots probed with DR cDNA by 15 min following stimulation and appropriately sized polypeptides were seen on immunoblots using an anti-DR monoclonal antibody. These findings could be accounted for by several mechanisms which are not necessarily mutually exclusive, among them increased transcription of the class II genes, effects on stability of the class II mRNA, alterations in the rate of translation and/or transport of the protein product, or the presence of preformed class II molecules within the T cell before stimulation. At the present time, we do not believe there is evidence to hypothesize differences in the intracellular traffic of class II between T cells and B cells. The results of the current studies provide new information on this subject. First, we have shown that class II genes can be transcribed in resting T cells which do not have detectable surface class II (Fig. 1A). Based upon these results, our earlier

REGULATION

OF CLASS II IN T CELLS

105

findings of very early detection of surface class II after activation could be explained in the context of the ongoing transcription. According to this reasoning, rapid surface expression is a consequence of the fact that less time is required to detect class II on the surface if the gene has already been transcribed, processed, and transported to the cytoplasm than if transcription must first be initiated. While other genes such as cmyc have been shown to be regulated at the post-transcription level (26), steady-state levels of mRNA are not detectable in unstimulated cells. Since with Western blotting we observed DR polypeptides within resting T cells, some translation of the protein occurred. It is important to emphasize that these results do not exclude a role for increased transcription in class II regulation in T cells, since our studies examined only 7-day-activated T cells when we knew class II levels were still increasing. If there is transcription of the gene, why can class II not be detected on the cell surface of resting cells? One possibility is that the protein product may not be transported to the cell surface, or that the monoclonal antibodies may fail to detect the product on the cell surface. This is consistent with results of Mittler et al. who have shown that cell surface class II antigen could be detected by direct immunofluorescence if the cells had first been reacted with another, irrelevant monoclonal antibody (32). These authors interpreted their findings as suggesting that class II was on the surface of the T cell in a cryptic form and that perturbation of the cell membrane by the first monoclonal antibody caused a reorientation of class II. The specificity of the results was demonstrated by the failure of other activation genes such as IL-2 to be expressed (Fig. IA). While the function of class II on T cells is not understood there is evidence consistent with its role in antigen presentation and/or signal transduction (8-16). We have previously hypothesized that the antigen presentation by T cells may be an amplification mechanism of activated cells (33). Recent evidence suggests that macrophages in the spleen are able to take up and process antigen which then can be transferred to B cells (34). These B cells then present this antigen to T cells, during which they receive T cell help in order to mature to antibody secreting cells. Our working hypothesis is that T cells can also present antigen and, while most studies suggest that they are inefficient at best at processing antigen, the above evidence that B cells can receive processed antigen passively from other accessory cells eliminates this as a block to T cells serving as antigen-presenting cells. According to this hypothesis, the presence of intracellular class II as either mRNA or polypeptides within the T cells would enable them to receive and present antigen readily. Furthermore, it has recently been reported that some “empty” class I molecules, devoid of peptides, expressed on the cell surface may be able to bind extracellular peptides (35). If such a mechanism were to be demonstrated for class II as well, it could impact on antigen presentation by T cells. Finally, our results suggest that there may be a protein which negatively regulates class II expression in normal T cells. Treatment of T cells activated with PdB and ionomycin with cycloheximide resulted in superinduction of DR, DQ, and DP (Fig. 3). However, treatment of T cells activated with PHA did not lead to similar results, consistent with the hypothesis that class II expression can be regulated by multiple distinct pathways. Thus, the mechanisms regulating class II in T cells are common to DR, DQ, and DP, but whether they are functional depends upon the activation pathway. This finding may account for the complexity underlying class II expression in T cells. One question of interest is the target of this negative regulatory protein and its identity. cDNAs for two regulatory proteins hXPB- 1 and YB- 1 which influence class

106

CAPLEN

ET AL.

II levels in other cell types have recently been cloned (36, 37). Evidence suggests that hXBP-1 may be a regulatory protein with a positive function, while YB-1 appears to be capable of exerting a negative effect in malignant T cell lines. Whether these proteins play a role in the regulation of class II in normal T cells or whether the negative factor we observed may in fact be YB-1 is currently under investigation in our laboratory. ACKNOWLEDGMENTS The authors appreciate the secretarial assistance of Ms. Letty Baez. We are grateful to Drs. Jeff Ledbctter and Craig Thompson for the anti-CD28 monoclonal antibody, to Dr. Craig Thompson for technical advice on the nuclear run-on assays,and to Andrew Pizzimenti for operation of the flow cytometer. This work was supported in part by NIH Grant A123519.

REFERENCES 1. Evans, R. L., Faldetaa, T. J., Humphrey, R. E., Pratt, D. M., Yunis, E. J., and Schlossman, S. F., J. Exp. Med. 148, 1440, 1978. 2. Greaves, M. F., Verbi, W., Festenstein, H., Papasteriadis, C., Jaraquemada, D., and Hayward, A., J. Immunol. 9, 356, 1979. 3. Ko, H. S., Fu, S. H., Winchester, R. J., Yu, D. T. Y., and Kunkel, H. G., 1. Exp. Med. 150,246, 1979. 4. Zier, K. S., Human Immunol. 4, 147, 1982. 5. Engleman, E. G., Charron, D. J., Benike, C. J., and Stewart, G. J., J. Exp. Med. 152, 99s, 1980. 6. Schendel, D. J., and Johnson, J. P., Eur. J. Immunol. 15, 1239, 1985. 7. Triebel, F., De Roquefeuil, S., Blanc, C., Charron, D. J., and Debre, P., Human Immunol. 15, 302, 1986. 8. Zier, K. S., and Pickard, A. R., J. Immunol. 135, 2274, 1985. 9. Ben-Nun, A., Strauss, W., Leeman, S. A., Cohn, L. E., Murre, C., Duby, A., Seidman, J. G., and Glimcher, L. H., Immunogenetics 22, 123, 1985. 10. Gerrard, T. L., Volkman, D. J., Jurgensen, C. H., and Fauci, A. S., Human Immunol. 17,416, 1986. Il. Lanzavecchia, A., Roosnek, E., Gregory, T., Berman, P., and Abrignai, S. Nature 334, 530, 1988. 12. Hewitt, C. R., and Feldmann, M. J., J. Immunol. 143,762, 1989. 13. Cambier, J. C., and Lehmann, K. R., J. Exp. Med. 170, 877, 1989. 14. Moretta, A., Accolla, R. S., and Cerottini, J-C., J. Exp. Med. 155, 599, 1982. 15. St-Pierre, Y., Nabavi, N., Ghogawala, Z., Glimcher, L. H., and Watts, T. H., J. Immunol. 143, 808, 1989. 16. Urban, N., and Zier, K. S., Human Immunol. 31, 170, 1991. 17. Zier, K. S., Gansbacher, B., Ikegaki, N., Kennett, R., and Polakova, K., Autoimmunity 5, 59, 1989. 18. Feinberg, A. P., and Vogelstein, B., Anal. Biochem. 132, 6, 1983. 19. Long, E. O., Wake, C. T., Gorski, J., and Mach, B., EMBO J. 2,389, 1983. 20. Gustafsson, K., Emmoth, E., Widmark, E., Boehme, J., Peterson, P. A., and Rask, L., Nature309, 76, 1984. 21. Gustafsson, K., Winman, K., Emmoth, E., Larhammar, D., Bohme, J., Hyldig-Nielsen, J. J., Ronne, H., Peterson, P. A., and Rask, L., EMBO J. 3, 1655, 1984. 22. Holbrook, N. J., Lieber, M., Fomace, A. J., Comeau, C. M., Wiskocil, R. L., and Crabtree, G. R., Proc. Natl. Acad. Sci. USA 81, 1634, 1984. 23. Tso, J. Y., Sun, S. H., Kao, T. H., Reece, K. S., and Wu, R., Nucleic Acids Res. 13, 2485, 1975. 24. Ullrich, A., Coussens, L., Haythck, S., Gray, A., Tam, A. W., Lu, J., Yarden, Y., Libermann, T. A., Schlessinger, J., Downward, J., Mayes, E. L. V., Whittle, N., Waterheld, M. O., and Seeburg, P. H., Nature 309,418, 1984. 25. Lindsten, T., June, C. H., and Thompson, C. B., EMBO J. 7, 787, 1988. 26. Dani, C., Blanchard, J. M., Piechaczyk, M., El Sabouty, S., Marty, L., and Jeanteur, P., Proc. Natl. Acad. Sci. USA 81, 7046, 1984. 27. Boss, J. M., and Strominger, J. L., Proc. Nat/. Acad. Sci. USA 83,9139, 1986.

REGULATION

OF CLASS II IN T CELLS

107

28. Liou, H-C., Boothby, M. R., Finn, P. W., Bavavi, N., Eddy, R., Shows, T., and Glimcher, L. H., J. Cell. Biochem. B 14, 210, 1988. 29. Darn, A., Durand, B., Marling, C., Le Meur, M., Benoist, C., and Mathis, D., Proc. NutI. Acad. Sci. USA 84,6249, 1987. 30. Maffei, A., Scarpellino, L., Bernard, M., Carra, G., Jotterand-Bellomo, M., Guardiola, J., and Accolla, R. S., J. Immunol. 139, 942, 1987. 31. Bottazzo, G. F., Pujol-Borrell, R., and Hanafusa, T., Lancer 2, 1115, 1983. 32. Mittler, R. S., Rao, P. E., Talle, M. A., Look, R., and Goldstein, G., J. Exp. Med. 158, 99, 1983. 33. Gansbacher, B., and Zier, K. S., Cell. Immunol. 117, 22, 1988. 34. van Rooijen, N., Claassen, E., Knurl, O., and Dijkstra, C. D., Prog. Histochem. Cytochem. 19, I, 1989. 35. Benjamin, R. J., Madrigal, J. A., and Parham, P., Nature 351, 74, 1991. 36. Didier, D. K., Schiffenbauer, H., Woulfe, S. L., Zacheis, M., and Schwartz, B. D., Proc. Nutl. Acad. Sci. USA 85,1322, 1988. 37. Liou, H. C., Boothby, M. R., and Glimcher, L. H., Science 242,69, 1988.