Developmental regulation of Vb genes in adult thymocytes
Eur. J. Immunol. 1990. 20: 41-45
Guido C. Miescher, Nan Shih Liaoo, Rosemary K. Lees, H. Robson MacDonald and David H. RauletO Ludwig Institute for Cancer Research, Lausanne Branch, Epalinges and Department Biology and Center for Cancer Research, MITO, Cambridge
Selective expression of V66 genes by B2A2CD4- CDS- T cell receptor y/6 thymocytes Most CD4-CD8- adult murine thymocytes characterized by absence of the B2A2 (Jlld) antigen express Tcell receptors (TcR) a / p and utilize preferentially Vp8.2 segments. To a much lesser extent, B2A2-CD4-CD8- thymocytes also express TcR y/6, as evidenced by biochemical and Northern blot analysis. We have now been able t o exclude the possibility that these cells might co-express both types of TcR: Vp8+ B2A2- CD4-CD8- thymocytes expressed no TcR 6 mRNA whereas the corresponding Vg8- subset transcribed full-length TcR y as well as 6 mRNA. Furthermore, the TcR y/6 expressing B2A2- thymocytes were found to use preferentially V66 genes. Conversely, they did not express V65 genes which are most frequently used by other TcR $&bearing populations such as B2A2+ CD4CD8- thymocytes or CD4-CD8- peripheral lymph node cells. RNase protection experiments showed that two particular V66 transcripts predominate in these y/6 populations, the most prominent V66 sequence being highly homologous if not identical to V,7.1. Our observations extend previous information on overlapping V, and V6 gene repertoires, particularly in the cross-hybridizing Va7Ns6 gene family. Moreover, our data suggest that B2A2- CD4-CD8- thymocytes represent a developmentally unique subset in which both V6 and Vg segments are nonrandomly expressed.
1 Introduction The ligands of the CD3-associated heterodimeric TcR a l p have been identified as the polymorphic MHC class I and class I1 proteins, generally in association with a peptide antigen [ l l r ] . MHC class I and class 11-reactive T cells represent functionally distinct Tcell populations [ 5 ] . However, it is controversial to what extent individual V segments may confer preferential reactivity to particular MHC products, rather than to the MHC-bound peptides [6]. In contrast,TcR y/6 composed of characteristicvy and V6 combinations are preferentially used by y/6 T cell populations in different tissues [7]. Thus, Thy-l+ dendritic epidermal cells have mostly TcR y/S with V,3 and V6l segments [8]*, intraepithelial lymphocytes of the gut use VY5JC.&genes [lo], splenic T cells use VY2JC,,1and to a large extent V, genes that have rearranged to JC,,4 [ l l , 121, whileVY2-V65TcR predominate in the adult thymus [9,13]. At present, the functional characteristics of these populations are not established and it is unclear to what extent these differences in V gene expression relate to differences in specificity for the still unknown ligands of y/6 T cells. We have shown previously that a subset of adult doublenegative (DN) CD4-CD8- thymocytes lacking the B2A2 ( J l l d ) antigen expresses TcR a/p utilizing predominantly theVg8.2 gene [14, 151. However, a subset of B2A2- DN thymocytes also express TcR $6 as evidenced by biochemical and Northern blot analysis [16]. A previous study of V, [I 77011 Correspondence: H. Robson MacDonald, Ludwig Institute for Cancer Research, Lausanne Branch, CH-1066 Epalinges, Switzerland Abbreviations: DN: Double negative
*
41
We have used throughout previously described nomenclatures for y genes [7] and for 6 genes [9].
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
usage in these cells demonstrated that V,2 was most frequently utilized [16], as is the case for peripheral y/6 T cells [ll]. We show here that this y/6 population expresses predominantly V6 genes that cross-hybridize with V,7 sequences. We therefore suggest that these cells are phenotypically and developmentally distinct from the majority of adult y/6 thymocytes which express the B2A2 protein.
2 Materials and methods 2.1 Cells Thymocyte and LN populations from adult female C57BL/6 and BALB/c mice 5 to 6 weeks of age, were prepared by negative selection using appropriate mAb as described [15]. Briefly, thymocytes (50 x 106/ml) were, treated twice by incubation at 37°C for 45 min with a mixture of rat IgM mAb directed against CD8 (3.168.8.1) and CD4 (LJCR.LAU. RL172.4), with or without rat IgM mAb B2A2, all at a 1/10 final dilution of hybridoma culture supernatant and a 1/20 final dilution of rabbit C SN (Cedarlane Laboratories, Hornsby, Ontario, Canada). After each treatment cells were passed over FicollHypaque (Pharmacia, Uppsala, Sweden) and washed three times. Two-color fluorescence staining with anti-CD4 or -CD8 reagents indicated < 1% contaminants. In some experiments B2A2- DN thymocytes were further subdivided into F ~ , J(i.e.Vg8+) + and F23.1- populations by sterile sorting prior to expansion in v i m .Total and DN LN, as well as B2A2- DN thymocyte populations, were stimulated with PMA (5 ng/ml) and ionomycin (250 ng/ml). To obtain TcR y/6 B2A2+ DN thymocytes we took advantage of our recent observation [16, 171 that these cells can be induced to proliferate selectively by stimulation with Con A (2 pg/ml). In each case, cells were expanded for 7 days in medium supplemented with 200 U/ml of human rIL 2 (provided by Biogen SA, Geneva, Switzerland). T-T hybrids were generated by fusing cultured B2A2- DN thymocytes with a 0014-2980/90/0101-0041$02.50/0
42
G. C. Miescher, N.S. Liao, R.K. Lees et al.
thioguanine-resistant variant of the BW5147 thymoma, as described [18]. HAT-resistant clones were further selected by fluorescence staining with mAb 145-2C11 (anti-CD3) and F23-1 (anti-Vp8). Immunoprecipitation [16] and Southern blot procedures [19] were done as described previously. 2.2 RNA analysis Total cellular RNA was extracted using the acid phenol method [20]. For Northern blot analysis, 10 pg of RNA was formaldehyde treated, fractionated on 1.5% agarose gels and transferred to Hybond nylon filters prior to hybridization with Ca, Cg, C,2 or Va7.2 antisense, 32P-labeledin vitro SP6 transcripts as described [19]. To verify that comparable amounts of RNA were loaded, the gels were stained with ethidium bromide and photographed prior to further processing.The 900-bp Eco RI fragment of ph12 [21],which corresponds to the constant and 3' untranslated regions of a TcR 6 cDNA and a Vp5-specific 300-bp Eco RI-S/pe I fragment of 244 [9], were 32Plabeled by random hexamer priming. Blots were stripped by boiling for 2 min in 5 mM Tris pH 8,0.1 mM EDTA and 0.1% SDS using a microwave oven, and then test exposed for 24 h prior to rehybridization. A 350 bp Eco RI - Not I (blunted) fragment of V,7.2 [22] was subcloned into the Eco RI and Hind I1 sites of pGEM I.The RNase protection assay was done with 3 pg of total RNA as described [19]. For mapping experiments, a Va7.2 template truncated with Hind I1 was used to obtain 32P-labeled transcripts of 175 bases that extend to the homologous Hind I1 site on Va7.1 (see Sect. 3.4, Fig. 4).
3 Results and discussion 3.1 TcR mRNA expression in Vp8+ and Vp8- B2A2DN thymocytes
In contrast with most a / p T cells, B2A2- DN thymocytes (population D, Fig. 1) express high levels of full-length TcRy transcripts and we have suggested that these cells might represent an early stage after divergence of $6 and a/pTcell lineages [16].We have now addressed the question of whether such cells might co-express TcR a@ and $6. Fresh cells of population D were therefore sorted into 60% F23.1' (i.e. V&V) and 40% F23.1- cells before in vitro
Eur. J. Immunol. 1990.20: 41-45 expansion and subsequent RNA analysis. We have shown previously that our culturing conditions, i.e. stimulation with PMA and ionomycin followed by culturing for 7 days in the presence of IL 2, did not change significantly either the proportion of cells expressing Vg8 determinants nor the TcR a or y mRNA levels of population D [16]. Both F ~ 3 . 1 ~ and F23.1- subsets clearly express full-lengthTcR a,p and y mRNA (Fig. 1A-C), implying that a sizeable portion of population D expressTcR a@ using Vp genes other than the Vg8 family, while confirming the relatively high TcR y mRNA in these unusual a / p Tcells.TcR 6 mRNA, however, was only expressed in the F23.1- subset (Fig. 1D) demonstrating that TCR y/6 could not be expressed on Vg8' cells of population D. TcR 6 mRNA levels in different preparations of these cells (cf. D, D F23.1+ and D F23.1- in Fig. lD, and data not shown) varied considerably less thanTcR a,p, and y mRNA. The low levels of TcR 6 mRNA in the Vp8subset are consistent with our previous estimate [16] that about 10% of population D are likely t o be yl6 Tcells. 3.2 V gene expression by B2A2- DN thymocytes
Preliminary screening by RNase protection analysis established that transcripts belonging to the V,7 gene family were expressed at high levels in B2A2- DN thymocytes (not shown). Northern blot analysis using aVa7 probe revealed surprisingly that this gene family is utilized mostly by the F23.1- subset and that the size of the transcripts (1.8 kb) would be appropriate for TcR 6 rather thanTcR a mRNA (Fig. 2A). Indeed,VG sequences that cross-hybridize with V,7 have been identified and designated as the V86 gene family [9]. This interpretation was corroborated by analyzing two T cell hybridomas. They were derived by fusing cultured cells from population D with the BW5147 thymoma and selecting for Va7 mRNA expression. D1 is CD3+ and F23.1-, transcribes theTcR 6 gene (Fig. 3C) and expresses TcR y/6 with CD3-associated heterodimeric 45and 35-kDa proteins (Fig. 3A); D2.1 has F23.1' TcR a/p and has deleted the TcR 6 gene (Fig. 3B). Northern blot analysis using aVb7 probe reveals a 1.8-kb band for D1 and a 1.6-kb band for D2.1 (Fig. 3D), consistent with the interpretation that the 1.8-kb mRNA represents V86 transcripts and that the 1.6-kb band corresponds to Va7 transcripts. It is also noteworthy that V86 mRNA levels are similar in the F23.1- subset of population D as in B2A2+ DN y/6
Figure 1. Northern blot analysis of TcR C, (A), Cb(B), C,(C) and Cs(D)mRNA levels in two independentpreparationsof cultured CD3+ B2A2+ DN thymocytes (C), B2A2- DN thymocytes (D). These are compared to BW5147 thymoma (BW), cultured total LN, cultured DN (4-8-) L N and to subsets of population D that were sorted into F23.,+ and F23.y cells prior to culturing.The same blots were re-hybridized in A-D. Populations D F23+ and D M3- were derived from a separate experiment than population D.
Eur. J. Immunol. 1990. 20: 41-45
Figure 2. Northern blot analysis of TcR V,7 (= v66; A) and V65 (B) mRNA 1evels.The same populations (and blots) were used as in Fig. 1.
thymocytes (population C), while Cg mRNA levels are over 6-fold lower in population D than in C (cf. Fig. 2A vs. Fig. lD).This observation suggests that V66 genes are used much more frequently by y/6 cells in the B2A2- F23.1- DN subset than in population C. Similar results as in C57BL16 were obtained using BALB/c mice (data not shown). As population C expresses high V65 mRNA levels (Fig. 2B), the absence of such transcripts in the F23.~- subset of D argues against the possibility of contaminants from population C. Our data, therefore, make it likely that there are y/6 thymocytes that have a restricted V6 usage and a distinct phenotype, i.e. B2A2-, that distinguish them from the majority of B2A2+ y/6 adult thymocytes.With respect toV, usage, the two populations do not differ significantly: both expressV,2 mRNA and 35-kDa CD3-associated protein, as well asV,1.2 mRNA; both have undetectable levels of V,3 and V,4 ([16] and data not shown).
3.3 Comparison of y/6 T cells in thymus and LN Our preliminary studies indicated that DN LN cells are B2A2- and F23.1- and that after expansion in vitro with PMA, ionomycin and IL2, TcR $6 could be demonstrated on these cells by immunoprecipitation (data not shown).We have now examined by RNA analysis whether or not this population resembled the B2A2- TcR y16 thymocytes from population D. A preparation of similarily cultured unselected LN cells expressed full-length TcR a and fi mRNA but no y or 6 mRNA (Fig. 1). As expected,V,7 transcripts had the same size as TcR a mRNA, i.e. 1.6 kb (Fig. 2A). DN LN cells, however, expressed essentially no TcR a mRNA, mostly short 1.1kbTcR fi mRNA and full-length y and 6 transcripts. V,7 cross-hybridizing transcripts of the size of Cg mRNA, i. e. 1.8 kb V66 transcripts, as well as Vg5 mRNA (Fig. 2B) could be demonstrated. DN LN yl6 T cells, therefore, resemble mostly the B2A2+ thymic population C rather than theTcR y/6 subset of D which does not express V65 mRNA. Analysis of the relative V66-usage of these populations also supports the conclusion that DN LN y/6 T cells are similar to the B2A2+ TcR y6 expressing thymocytes. Densitometric quantitation of the Northern blots (cf. Fig. 1 D and 2A) and standardization of the V66 mRNA levels with respect to the corresponding levels of TcR C6 mRNA indicated that population D had an over 6-fold higher usage of V66 transcripts than population C. According to the same calculation, the independent preparation of F23.1- cells of population D showed a 9-fold
Developmental regulation of Vb genes in adult thymocytes
43
Figure 3. (A) Immunopreciptiation of CD3-associated TcR molecules from Fu.l- (D 1)and F B . ~ + (D2.1) CD3+ T-T hybridomas as well as CD3+ B2A2+ DN thymocytes (C). 1251 surface-labeled cells (lo’) were solubilized using digitonin and were analyzed on 10% polyacrylamide gels under reducing conditions. While optimizing resolution of TcR polypeptides, the 21-kDa CD3 subunit was run off the gel.(B) Southern blot analysis for the presence or deletion of the Ca gene in T-T hybridomas (D1 and D2.1), the parental BW5147 thymoma (BW), and C57BL16 liver (GL): Ten micrograms of Eco RI-digested D N A was used in each lane. (C) Northern blot analysis of TcR C6 mRNA, and (D) TcR V,7.2 mRNA levels in theT-T hybridomas, BW thymoma and population P
L
increased V66 usage, while the relative mRNA expression levels of Vg6 were only 1.7-fold higher in DN LN cells than in the thymic population C. These observations could imply that population C may give rise to y/6 emigrants that, in analogy to a/@T cells, lose the B2A2 determinant upon maturation. The abundance of V,7 mRNA we find in a / B peripheral T cells is in opposition to the findings of Bill et al. [23] who note a marked underexpression of V,7 genes, when compared to their approximately 10% representation in the germ-line V, pool. Conceivably, differences of mouse strains, or selection of Tcells by culturing for 2 days in the presence of Con A and in the absence of exogenous IL 2, as well as preferential fusion with the BW5147 thymoma may account for the discrepancy.
3.4 RNase protection analysis of Va6Na7 transcripts The pattern of RNase-resistant fragments shows a considerable difference between a / P and y/6 T cell populations (Fig. 4). A 150-base, and often a 185-base fragment are prominent in the y/6 populations while they are only barely detectable in mRNA from total LN. This observation is in apparent contradiction with the Northern blot analysis showing similar levels of V,7 cross-hybridizing mRNA in these populations. A likely explanation is that the RNase conditions used allow quantitative detection only of sequences closely related to theV,7.2 probe (such a ~ V ~ 7 . 1 ) but not less homologous V,7 transcripts such as Vz53,Vz49 and V ~ 1 2which have only 78% to 80% homology with V,7.2 791. The interpretation of the RNase protection patterns was facilitated by the study of the two afore-mentioned T cell hybridomas. RNase protection assays with the a/fi clone D2.1 result in a band of the same size, i.e. 300 bases, as the cDNA insert (Fig. 4). This band is at best very faint in polyclonal a / P populations and represents transcripts of the
44
Eur. J. Immunol. 1990. 20: 41-45
G. C. Miescher, N.S. Liao, R.K. Lees et al.
D1
mRNA
a+
b+
*+
V a 7.2
& V 6 6 M23
** +
V 6 6 253
Figure 5. RNase mapping of V66 transcripts from the hybridoma D1 compared to knownV66 sequences. A Hind I1 -truncated Va7.2 template (23base pairs polylinker plus 150base pairs of Va7.2)was used to transcribe in vitro 32P-labeledantisense RNA (lStlane). This Va7.2 probe was hybridized to: D1, 3 pg cellular RNA from the TcR y/S hybridoma D1; M23, 10 ng of a cDNA restriction Figure 4. RNase protection analysis of TcR Va7 (= V66) mRNA in thymic and LN populations (see legend Fig. 1) as well as in T-T fragment containingthe completeVsequence [9];TA27,10ng of a 315 base pairs Hind I1 - Sac I cDNA fragment [24]; and 253, sense hybridomas between B2A2- DN thymocytes and BW5147 thymoRNA transcribed in vitro using T3 RNA polymerase (91. The ma. D 1is CD3+F23.1.-and expressesTcR y/6, while D2.1 is FZ3.,+. Va7.2 indicates the protected fragment of 300 bases corresponding observed RNase-resistantfragmentsare designated a, b, and c, and to the probe used. Protected fragments of 185 bases (*) and 150 are represented by solid bars in the accompanying scheme. The bases (**) correspond to V66 (=V,7) transcripts from separate cDNA restriction fragments (straight lines), the cellular RNA (wavy line) and the in vitro transcribed sense and antisense RNA genes. (wavy line plus box) are lined up to indicate the homologous sequences. Va7.2 locus. In the case of the y/6 clone D1, the RNaseresistant fragment of 150 bases is identical to the main fragment found in polyclonal y/6 populations (Fig. 4). This transcripts inTcR y/6 cells as well as theVg6 gene expressed fragment therefore identifies a different Vg6 gene transcript by D1 are highly homologous, if not identical, to V,7.1. Such a candidateVg6 gene,TCD61, has been reported [25] than either the 300- or 185- base bands. and has only two nucleotide mismatches with respect to RNase mapping experiments with available cDNA of Va7.1 in a sequence 3' to the restriction fragment used in various members of the V,7/Vg6 gene family were per- this study.Three further Vg6 sequences are known that have formed in an attempt t o identify the Vg6 transcript in the between 75% and 80% homology to Va7.1 i.e. 249, 253 TcR y/6 hybridoma D1. The main protected fragment seen and pA12 [9]. RNase protection experiments with 253 after hybridization of a V,7.2 RNA probe with a Va7.1 resulted in protected fragments less than half the size of cDNA restriction fragment [24] is identical to that seen those observed with Va7.1 (Fig. 5). after hybridization of the probe to mRNA from the D1 hybridoma (Fig. 5). Surprisingly, using V ~ 2 3 , a Va6 Bill et al. [23] have suggested that there may be about six sequence with 93% homology toVa7. 1[9], the main RNase different V66 genes. If the 300- base fragment seen in the protected fragment has the same size as the Va7.2 cDNA RNase protection experiments with cellular RNA samples corresponds to Va7.2NsM23 and the 150 base fragment restriction fragment used to generate the RNA probe. However, the sequence of Va7.2 [22], which was published corresponds to Va7. 1, then the 185 base band seen in most after completion of these experiments, revealed that this of the mRNA samples from polyclonal y/6 populations restriction fragment is identical to the homologous Vm3 (Fig. 4) may indicate the existence of an as yet unreported sequence. Our results, therefore, are consistent with the Vg6 gene with intermediate homology between V,7.1 and interpretation that it is possible, under the conditions of Va7.2. Alternatively, the observation that the size of this partial RNase digestion used, t o compare directly band of 185 nucleotides corresponds to the portion of the R N A R N A homoduplexes and RNA/DNA heterodu- Va7.2 probe up to the 5'end of the V segment could imply plexes. Accordingly, we suggest that the predominant Vg6 that the 5' untranslated and leader sequences of this
Eur. J. Immunol. 1990. 20: 41-45
transcript may be derived from a different gene by alternative splicing as has been noted for Vg8 [26].
4 Concluding remarks Our experiments suggest that B2A2- DN thymocytes can be separated into two mutually exclusive subsets: a small TcR y/6 population utilizing preferentially Va6 genes, and a majority of unusual TcR a / p thymocytes utilizing mostly the Vg8.2 gene and expressing high TcR y mRNA [161. Recent experiments indicate that y / 6 and a/P T cells may not be lineally related and may be derived from pre-commited stem cells [27]. The observation of developmental regulation of V gene expression in two phenotypically similar y/6 and a / p B2A2- thymocyte subsets is, therefore, intriguing. It is tempting to speculate that these thymocytes might have been subjected to selection processes. The expression of the CD5 molecule [16] and the absence of the B2A2 antigen are characteristic of mature T cells. Indeed, these cells are small resting lymphocytes [ 171 without precursor activity as assessed by in vivo reconstitution experiments [28]. Our recent results [29] indicate that F23.1+ B2A2- DN thymocytes have functional TcR mediating cellular proliferation, production of IL2 and IL3, as well as cytolysis. Moreover, the observation that in vivo treatment of neonatal mice with F23.1 mAb resulted in a marked reduction of this subset [29] would also be consistent with the notion that TcR a/p B2A2- DN thymocytes are terminally differentiated cells that might have been shaped by selecting or inducing processes. Whether any of these characteristics hold true for the TcR+ $6 B2A2- DN thymocytes remains conjectural at present. Vb and V, genes are interspersed in the same chromosomal region [21]. Alternative usage as V6 or V, genes has been described for V83N,6 [9], members of the Vs6N,7 gene family [22,25] and for a member of theV,4 gene family [13] in the mouse, as well as for DS6N,6 in man [30]. It is therefore of considerable interest to identify the particular ligands of y/6 TcR utilizing Vp6-V,2 genes. Conceivably, such ligands might also be involved in the selection process of a / P T cells carrying V,7 TcR. It remains to be seen whether the reported presence of an in-frame V86N,7-Ja1 rearranged gene on a DNA excision circle [25] might be a result of negative selection.
We thank Dr. !I Chienfor providing V, cDNA, Dr. E. Palmerfor the Va7.2cDNA and Dr. B. Arden for providing a TA27plasmid as well as an unpublished cDNA corresponding to V,+,23. Received June 1, 1989; in revised form August 21, 1989.
Developmental regulation of Vb genes in adult thymocytes
45
5 References 1 Zinkernagel, R. M. and Doherty, l? C., Nature 1974. 248: 701. 2 Bjorkman, P. J., Saper, M. A., Samraoui, B.. Bennett,W. S., Strominger, J. L. and Wiley, D. C., Nature 1987. 329: 506. 3 Townsend, A. R. M., Rothbard, J., Gotch, F. M., Bahadur. G., Wraith, D. and McMichael, A. J. Cell 1986. 44: 959. 4 Shimonkevitz, R., Kappler, J.W., Marrack, P. and Grey, H. M., J. Exp. Med. 1983. 158: 303. 5 Reinherz, E. L. and Schlossman, S., Cell 1980. 19: 821. 6 Davis, M. M. and Bjorkman, l? J., Nature 1988. 334: 395. 7 Raulet, D. H., Annu. Rev. Immunol. 1989. 7: 175. 8 Havran,W. and Allison, J. P., Nature 1988. 335: 443. 9 Elliott, J.F., Rock, E. P., Patten, P.A., Davis, M. M. and Chien, Y., Nature 1988. 331: 627. 10 Bonneville, M., Janeway, C. A., Ito, K., Haser,W., Ishida, I.. Nakanishi, N. and Tonegawa, S., Nature 1988. 336: 479. 11 Cron, R. Q., Koning, F., Maloy,W. L., Pardoll, D., Coligan, J. E. and Bluestone, J. A., J. Immunol. 1988. 141: 1074. 12 Pelkonen, J., Traunecker, A. and Karjalainen, K., EMBO J. 1987. 6: 1941. 13 Korman, A. J., Marusic-Galesic, S., Spencer, D., Kruisbeek, A. M. and Raulet, D. H., J. Exp. Med. 1988. 168: 1021. 14 Miescher, G. C., Howe, R. C., Budd, R. C. and Mac Donald, H. R., Ann. N . Y Acad. Sci. 1988. 532: 8 . 15 Budd, R. C., Miescher, G. C., Howe, R. C., Lees, R. K., Bron, C. and MacDonald, H. R., J. Exp. Med. 1987. 166: 577. 16 Miescher, G. C., Howe, R. C., Lees, R. K. and MacDonald, H. R., J. Immunol. 1988. 140: 1779. 17 Howe, R. C. and klxDonald, H. R., J. Immunol. 1988. 140: 1047. 18 Nabholz, M., Conzelmann, A., Acuto, 0..North, M., Haas, W., Pohlit, H., v. Boehmer, H., Hengartner, H., Mach, J.-l?. Engers, H . and Johnson, J. P., Immunol. Rev. 1980. 51: 125. 19 Miescher, G. C., Budd, R. C., Lees, R. K. and MacDonald, H. R., J. Immunol. 1987. 138: 1959. 20 Chomczynski, P. and Sacchi, N., Anal. Biochem. 1987. 162: 156. 21 Chien,Y., Iwashima, M., Kaplan, K. B., Elliott, J. F. and Davis, M. M., Nature 1987. 327: 677. 22 Yague, J., Blackman, M., Born,W., Marrack, P., Kappler, J. and Palmer, E., Nucleic Acids. Res. 1988. 16: (23): 11355. 23 Bill, J., Appe1,V. B. and Palmer, E., Proc. Natl. Acad. Sci. USA 1988. 85: 9184. 24 Arden, B., Klotz, J. L., Siu, G. and Hood, L. D.. Nature 1985. 316: 783. 25 Okazaki, K. and Sakano, H., EMBO J. 1988. 7 (6): 1669. 26 Chou, H. S., Anderson, S. J., Louie, M. C., Godambe, S. A., Pozzi, M. R.,Behlke, M. A., Huppi, K. and Loh, D.Y.. Proc. Natl. Acad. Sci. USA 1987. 84: 1992. 27 Winoto, A, and Baltimore, D., Nature 1989. 338: 430. 28 Crispe, I. N., Moore, M. W., Husmann, L. A.. Smith, L., Bevan, M. J. and Shimonkevitz, R. P., Nature 1987. 329: 336. 29 Howe, R. C., Pedrazzini,T. and MacDonald, H. R.. Eur. J. Immunol. 1989. 19: 25. 30 Guglielmi, P., Davi, F., D’Auriol, L.. Bones, J.-C., Dausset, J. and Bensussan, A., Proc. Natl. Acad. Sci. USA 1988. 85: 5634.
Eur. J. Immunol. 1990. 20: 41-45
Guido C. Miescher, Nan Shih Liaoo, Rosemary K. Lees, H. Robson MacDonald and David H. RauletO Ludwig Institute for Cancer Research, Lausanne Branch, Epalinges and Department Biology and Center for Cancer Research, MITO, Cambridge
Selective expression of V66 genes by B2A2CD4- CDS- T cell receptor y/6 thymocytes Most CD4-CD8- adult murine thymocytes characterized by absence of the B2A2 (Jlld) antigen express Tcell receptors (TcR) a / p and utilize preferentially Vp8.2 segments. To a much lesser extent, B2A2-CD4-CD8- thymocytes also express TcR y/6, as evidenced by biochemical and Northern blot analysis. We have now been able t o exclude the possibility that these cells might co-express both types of TcR: Vp8+ B2A2- CD4-CD8- thymocytes expressed no TcR 6 mRNA whereas the corresponding Vg8- subset transcribed full-length TcR y as well as 6 mRNA. Furthermore, the TcR y/6 expressing B2A2- thymocytes were found to use preferentially V66 genes. Conversely, they did not express V65 genes which are most frequently used by other TcR $&bearing populations such as B2A2+ CD4CD8- thymocytes or CD4-CD8- peripheral lymph node cells. RNase protection experiments showed that two particular V66 transcripts predominate in these y/6 populations, the most prominent V66 sequence being highly homologous if not identical to V,7.1. Our observations extend previous information on overlapping V, and V6 gene repertoires, particularly in the cross-hybridizing Va7Ns6 gene family. Moreover, our data suggest that B2A2- CD4-CD8- thymocytes represent a developmentally unique subset in which both V6 and Vg segments are nonrandomly expressed.
1 Introduction The ligands of the CD3-associated heterodimeric TcR a l p have been identified as the polymorphic MHC class I and class I1 proteins, generally in association with a peptide antigen [ l l r ] . MHC class I and class 11-reactive T cells represent functionally distinct Tcell populations [ 5 ] . However, it is controversial to what extent individual V segments may confer preferential reactivity to particular MHC products, rather than to the MHC-bound peptides [6]. In contrast,TcR y/6 composed of characteristicvy and V6 combinations are preferentially used by y/6 T cell populations in different tissues [7]. Thus, Thy-l+ dendritic epidermal cells have mostly TcR y/S with V,3 and V6l segments [8]*, intraepithelial lymphocytes of the gut use VY5JC.&genes [lo], splenic T cells use VY2JC,,1and to a large extent V, genes that have rearranged to JC,,4 [ l l , 121, whileVY2-V65TcR predominate in the adult thymus [9,13]. At present, the functional characteristics of these populations are not established and it is unclear to what extent these differences in V gene expression relate to differences in specificity for the still unknown ligands of y/6 T cells. We have shown previously that a subset of adult doublenegative (DN) CD4-CD8- thymocytes lacking the B2A2 ( J l l d ) antigen expresses TcR a/p utilizing predominantly theVg8.2 gene [14, 151. However, a subset of B2A2- DN thymocytes also express TcR $6 as evidenced by biochemical and Northern blot analysis [16]. A previous study of V, [I 77011 Correspondence: H. Robson MacDonald, Ludwig Institute for Cancer Research, Lausanne Branch, CH-1066 Epalinges, Switzerland Abbreviations: DN: Double negative
*
41
We have used throughout previously described nomenclatures for y genes [7] and for 6 genes [9].
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
usage in these cells demonstrated that V,2 was most frequently utilized [16], as is the case for peripheral y/6 T cells [ll]. We show here that this y/6 population expresses predominantly V6 genes that cross-hybridize with V,7 sequences. We therefore suggest that these cells are phenotypically and developmentally distinct from the majority of adult y/6 thymocytes which express the B2A2 protein.
2 Materials and methods 2.1 Cells Thymocyte and LN populations from adult female C57BL/6 and BALB/c mice 5 to 6 weeks of age, were prepared by negative selection using appropriate mAb as described [15]. Briefly, thymocytes (50 x 106/ml) were, treated twice by incubation at 37°C for 45 min with a mixture of rat IgM mAb directed against CD8 (3.168.8.1) and CD4 (LJCR.LAU. RL172.4), with or without rat IgM mAb B2A2, all at a 1/10 final dilution of hybridoma culture supernatant and a 1/20 final dilution of rabbit C SN (Cedarlane Laboratories, Hornsby, Ontario, Canada). After each treatment cells were passed over FicollHypaque (Pharmacia, Uppsala, Sweden) and washed three times. Two-color fluorescence staining with anti-CD4 or -CD8 reagents indicated < 1% contaminants. In some experiments B2A2- DN thymocytes were further subdivided into F ~ , J(i.e.Vg8+) + and F23.1- populations by sterile sorting prior to expansion in v i m .Total and DN LN, as well as B2A2- DN thymocyte populations, were stimulated with PMA (5 ng/ml) and ionomycin (250 ng/ml). To obtain TcR y/6 B2A2+ DN thymocytes we took advantage of our recent observation [16, 171 that these cells can be induced to proliferate selectively by stimulation with Con A (2 pg/ml). In each case, cells were expanded for 7 days in medium supplemented with 200 U/ml of human rIL 2 (provided by Biogen SA, Geneva, Switzerland). T-T hybrids were generated by fusing cultured B2A2- DN thymocytes with a 0014-2980/90/0101-0041$02.50/0
42
G. C. Miescher, N.S. Liao, R.K. Lees et al.
thioguanine-resistant variant of the BW5147 thymoma, as described [18]. HAT-resistant clones were further selected by fluorescence staining with mAb 145-2C11 (anti-CD3) and F23-1 (anti-Vp8). Immunoprecipitation [16] and Southern blot procedures [19] were done as described previously. 2.2 RNA analysis Total cellular RNA was extracted using the acid phenol method [20]. For Northern blot analysis, 10 pg of RNA was formaldehyde treated, fractionated on 1.5% agarose gels and transferred to Hybond nylon filters prior to hybridization with Ca, Cg, C,2 or Va7.2 antisense, 32P-labeledin vitro SP6 transcripts as described [19]. To verify that comparable amounts of RNA were loaded, the gels were stained with ethidium bromide and photographed prior to further processing.The 900-bp Eco RI fragment of ph12 [21],which corresponds to the constant and 3' untranslated regions of a TcR 6 cDNA and a Vp5-specific 300-bp Eco RI-S/pe I fragment of 244 [9], were 32Plabeled by random hexamer priming. Blots were stripped by boiling for 2 min in 5 mM Tris pH 8,0.1 mM EDTA and 0.1% SDS using a microwave oven, and then test exposed for 24 h prior to rehybridization. A 350 bp Eco RI - Not I (blunted) fragment of V,7.2 [22] was subcloned into the Eco RI and Hind I1 sites of pGEM I.The RNase protection assay was done with 3 pg of total RNA as described [19]. For mapping experiments, a Va7.2 template truncated with Hind I1 was used to obtain 32P-labeled transcripts of 175 bases that extend to the homologous Hind I1 site on Va7.1 (see Sect. 3.4, Fig. 4).
3 Results and discussion 3.1 TcR mRNA expression in Vp8+ and Vp8- B2A2DN thymocytes
In contrast with most a / p T cells, B2A2- DN thymocytes (population D, Fig. 1) express high levels of full-length TcRy transcripts and we have suggested that these cells might represent an early stage after divergence of $6 and a/pTcell lineages [16].We have now addressed the question of whether such cells might co-express TcR a@ and $6. Fresh cells of population D were therefore sorted into 60% F23.1' (i.e. V&V) and 40% F23.1- cells before in vitro
Eur. J. Immunol. 1990.20: 41-45 expansion and subsequent RNA analysis. We have shown previously that our culturing conditions, i.e. stimulation with PMA and ionomycin followed by culturing for 7 days in the presence of IL 2, did not change significantly either the proportion of cells expressing Vg8 determinants nor the TcR a or y mRNA levels of population D [16]. Both F ~ 3 . 1 ~ and F23.1- subsets clearly express full-lengthTcR a,p and y mRNA (Fig. 1A-C), implying that a sizeable portion of population D expressTcR a@ using Vp genes other than the Vg8 family, while confirming the relatively high TcR y mRNA in these unusual a / p Tcells.TcR 6 mRNA, however, was only expressed in the F23.1- subset (Fig. 1D) demonstrating that TCR y/6 could not be expressed on Vg8' cells of population D. TcR 6 mRNA levels in different preparations of these cells (cf. D, D F23.1+ and D F23.1- in Fig. lD, and data not shown) varied considerably less thanTcR a,p, and y mRNA. The low levels of TcR 6 mRNA in the Vp8subset are consistent with our previous estimate [16] that about 10% of population D are likely t o be yl6 Tcells. 3.2 V gene expression by B2A2- DN thymocytes
Preliminary screening by RNase protection analysis established that transcripts belonging to the V,7 gene family were expressed at high levels in B2A2- DN thymocytes (not shown). Northern blot analysis using aVa7 probe revealed surprisingly that this gene family is utilized mostly by the F23.1- subset and that the size of the transcripts (1.8 kb) would be appropriate for TcR 6 rather thanTcR a mRNA (Fig. 2A). Indeed,VG sequences that cross-hybridize with V,7 have been identified and designated as the V86 gene family [9]. This interpretation was corroborated by analyzing two T cell hybridomas. They were derived by fusing cultured cells from population D with the BW5147 thymoma and selecting for Va7 mRNA expression. D1 is CD3+ and F23.1-, transcribes theTcR 6 gene (Fig. 3C) and expresses TcR y/6 with CD3-associated heterodimeric 45and 35-kDa proteins (Fig. 3A); D2.1 has F23.1' TcR a/p and has deleted the TcR 6 gene (Fig. 3B). Northern blot analysis using aVb7 probe reveals a 1.8-kb band for D1 and a 1.6-kb band for D2.1 (Fig. 3D), consistent with the interpretation that the 1.8-kb mRNA represents V86 transcripts and that the 1.6-kb band corresponds to Va7 transcripts. It is also noteworthy that V86 mRNA levels are similar in the F23.1- subset of population D as in B2A2+ DN y/6
Figure 1. Northern blot analysis of TcR C, (A), Cb(B), C,(C) and Cs(D)mRNA levels in two independentpreparationsof cultured CD3+ B2A2+ DN thymocytes (C), B2A2- DN thymocytes (D). These are compared to BW5147 thymoma (BW), cultured total LN, cultured DN (4-8-) L N and to subsets of population D that were sorted into F23.,+ and F23.y cells prior to culturing.The same blots were re-hybridized in A-D. Populations D F23+ and D M3- were derived from a separate experiment than population D.
Eur. J. Immunol. 1990. 20: 41-45
Figure 2. Northern blot analysis of TcR V,7 (= v66; A) and V65 (B) mRNA 1evels.The same populations (and blots) were used as in Fig. 1.
thymocytes (population C), while Cg mRNA levels are over 6-fold lower in population D than in C (cf. Fig. 2A vs. Fig. lD).This observation suggests that V66 genes are used much more frequently by y/6 cells in the B2A2- F23.1- DN subset than in population C. Similar results as in C57BL16 were obtained using BALB/c mice (data not shown). As population C expresses high V65 mRNA levels (Fig. 2B), the absence of such transcripts in the F23.~- subset of D argues against the possibility of contaminants from population C. Our data, therefore, make it likely that there are y/6 thymocytes that have a restricted V6 usage and a distinct phenotype, i.e. B2A2-, that distinguish them from the majority of B2A2+ y/6 adult thymocytes.With respect toV, usage, the two populations do not differ significantly: both expressV,2 mRNA and 35-kDa CD3-associated protein, as well asV,1.2 mRNA; both have undetectable levels of V,3 and V,4 ([16] and data not shown).
3.3 Comparison of y/6 T cells in thymus and LN Our preliminary studies indicated that DN LN cells are B2A2- and F23.1- and that after expansion in vitro with PMA, ionomycin and IL2, TcR $6 could be demonstrated on these cells by immunoprecipitation (data not shown).We have now examined by RNA analysis whether or not this population resembled the B2A2- TcR y16 thymocytes from population D. A preparation of similarily cultured unselected LN cells expressed full-length TcR a and fi mRNA but no y or 6 mRNA (Fig. 1). As expected,V,7 transcripts had the same size as TcR a mRNA, i.e. 1.6 kb (Fig. 2A). DN LN cells, however, expressed essentially no TcR a mRNA, mostly short 1.1kbTcR fi mRNA and full-length y and 6 transcripts. V,7 cross-hybridizing transcripts of the size of Cg mRNA, i. e. 1.8 kb V66 transcripts, as well as Vg5 mRNA (Fig. 2B) could be demonstrated. DN LN yl6 T cells, therefore, resemble mostly the B2A2+ thymic population C rather than theTcR y/6 subset of D which does not express V65 mRNA. Analysis of the relative V66-usage of these populations also supports the conclusion that DN LN y/6 T cells are similar to the B2A2+ TcR y6 expressing thymocytes. Densitometric quantitation of the Northern blots (cf. Fig. 1 D and 2A) and standardization of the V66 mRNA levels with respect to the corresponding levels of TcR C6 mRNA indicated that population D had an over 6-fold higher usage of V66 transcripts than population C. According to the same calculation, the independent preparation of F23.1- cells of population D showed a 9-fold
Developmental regulation of Vb genes in adult thymocytes
43
Figure 3. (A) Immunopreciptiation of CD3-associated TcR molecules from Fu.l- (D 1)and F B . ~ + (D2.1) CD3+ T-T hybridomas as well as CD3+ B2A2+ DN thymocytes (C). 1251 surface-labeled cells (lo’) were solubilized using digitonin and were analyzed on 10% polyacrylamide gels under reducing conditions. While optimizing resolution of TcR polypeptides, the 21-kDa CD3 subunit was run off the gel.(B) Southern blot analysis for the presence or deletion of the Ca gene in T-T hybridomas (D1 and D2.1), the parental BW5147 thymoma (BW), and C57BL16 liver (GL): Ten micrograms of Eco RI-digested D N A was used in each lane. (C) Northern blot analysis of TcR C6 mRNA, and (D) TcR V,7.2 mRNA levels in theT-T hybridomas, BW thymoma and population P
L
increased V66 usage, while the relative mRNA expression levels of Vg6 were only 1.7-fold higher in DN LN cells than in the thymic population C. These observations could imply that population C may give rise to y/6 emigrants that, in analogy to a/@T cells, lose the B2A2 determinant upon maturation. The abundance of V,7 mRNA we find in a / B peripheral T cells is in opposition to the findings of Bill et al. [23] who note a marked underexpression of V,7 genes, when compared to their approximately 10% representation in the germ-line V, pool. Conceivably, differences of mouse strains, or selection of Tcells by culturing for 2 days in the presence of Con A and in the absence of exogenous IL 2, as well as preferential fusion with the BW5147 thymoma may account for the discrepancy.
3.4 RNase protection analysis of Va6Na7 transcripts The pattern of RNase-resistant fragments shows a considerable difference between a / P and y/6 T cell populations (Fig. 4). A 150-base, and often a 185-base fragment are prominent in the y/6 populations while they are only barely detectable in mRNA from total LN. This observation is in apparent contradiction with the Northern blot analysis showing similar levels of V,7 cross-hybridizing mRNA in these populations. A likely explanation is that the RNase conditions used allow quantitative detection only of sequences closely related to theV,7.2 probe (such a ~ V ~ 7 . 1 ) but not less homologous V,7 transcripts such as Vz53,Vz49 and V ~ 1 2which have only 78% to 80% homology with V,7.2 791. The interpretation of the RNase protection patterns was facilitated by the study of the two afore-mentioned T cell hybridomas. RNase protection assays with the a/fi clone D2.1 result in a band of the same size, i.e. 300 bases, as the cDNA insert (Fig. 4). This band is at best very faint in polyclonal a / P populations and represents transcripts of the
44
Eur. J. Immunol. 1990. 20: 41-45
G. C. Miescher, N.S. Liao, R.K. Lees et al.
D1
mRNA
a+
b+
*+
V a 7.2
& V 6 6 M23
** +
V 6 6 253
Figure 5. RNase mapping of V66 transcripts from the hybridoma D1 compared to knownV66 sequences. A Hind I1 -truncated Va7.2 template (23base pairs polylinker plus 150base pairs of Va7.2)was used to transcribe in vitro 32P-labeledantisense RNA (lStlane). This Va7.2 probe was hybridized to: D1, 3 pg cellular RNA from the TcR y/S hybridoma D1; M23, 10 ng of a cDNA restriction Figure 4. RNase protection analysis of TcR Va7 (= V66) mRNA in thymic and LN populations (see legend Fig. 1) as well as in T-T fragment containingthe completeVsequence [9];TA27,10ng of a 315 base pairs Hind I1 - Sac I cDNA fragment [24]; and 253, sense hybridomas between B2A2- DN thymocytes and BW5147 thymoRNA transcribed in vitro using T3 RNA polymerase (91. The ma. D 1is CD3+F23.1.-and expressesTcR y/6, while D2.1 is FZ3.,+. Va7.2 indicates the protected fragment of 300 bases corresponding observed RNase-resistantfragmentsare designated a, b, and c, and to the probe used. Protected fragments of 185 bases (*) and 150 are represented by solid bars in the accompanying scheme. The bases (**) correspond to V66 (=V,7) transcripts from separate cDNA restriction fragments (straight lines), the cellular RNA (wavy line) and the in vitro transcribed sense and antisense RNA genes. (wavy line plus box) are lined up to indicate the homologous sequences. Va7.2 locus. In the case of the y/6 clone D1, the RNaseresistant fragment of 150 bases is identical to the main fragment found in polyclonal y/6 populations (Fig. 4). This transcripts inTcR y/6 cells as well as theVg6 gene expressed fragment therefore identifies a different Vg6 gene transcript by D1 are highly homologous, if not identical, to V,7.1. Such a candidateVg6 gene,TCD61, has been reported [25] than either the 300- or 185- base bands. and has only two nucleotide mismatches with respect to RNase mapping experiments with available cDNA of Va7.1 in a sequence 3' to the restriction fragment used in various members of the V,7/Vg6 gene family were per- this study.Three further Vg6 sequences are known that have formed in an attempt t o identify the Vg6 transcript in the between 75% and 80% homology to Va7.1 i.e. 249, 253 TcR y/6 hybridoma D1. The main protected fragment seen and pA12 [9]. RNase protection experiments with 253 after hybridization of a V,7.2 RNA probe with a Va7.1 resulted in protected fragments less than half the size of cDNA restriction fragment [24] is identical to that seen those observed with Va7.1 (Fig. 5). after hybridization of the probe to mRNA from the D1 hybridoma (Fig. 5). Surprisingly, using V ~ 2 3 , a Va6 Bill et al. [23] have suggested that there may be about six sequence with 93% homology toVa7. 1[9], the main RNase different V66 genes. If the 300- base fragment seen in the protected fragment has the same size as the Va7.2 cDNA RNase protection experiments with cellular RNA samples corresponds to Va7.2NsM23 and the 150 base fragment restriction fragment used to generate the RNA probe. However, the sequence of Va7.2 [22], which was published corresponds to Va7. 1, then the 185 base band seen in most after completion of these experiments, revealed that this of the mRNA samples from polyclonal y/6 populations restriction fragment is identical to the homologous Vm3 (Fig. 4) may indicate the existence of an as yet unreported sequence. Our results, therefore, are consistent with the Vg6 gene with intermediate homology between V,7.1 and interpretation that it is possible, under the conditions of Va7.2. Alternatively, the observation that the size of this partial RNase digestion used, t o compare directly band of 185 nucleotides corresponds to the portion of the R N A R N A homoduplexes and RNA/DNA heterodu- Va7.2 probe up to the 5'end of the V segment could imply plexes. Accordingly, we suggest that the predominant Vg6 that the 5' untranslated and leader sequences of this
Eur. J. Immunol. 1990. 20: 41-45
transcript may be derived from a different gene by alternative splicing as has been noted for Vg8 [26].
4 Concluding remarks Our experiments suggest that B2A2- DN thymocytes can be separated into two mutually exclusive subsets: a small TcR y/6 population utilizing preferentially Va6 genes, and a majority of unusual TcR a / p thymocytes utilizing mostly the Vg8.2 gene and expressing high TcR y mRNA [161. Recent experiments indicate that y / 6 and a/P T cells may not be lineally related and may be derived from pre-commited stem cells [27]. The observation of developmental regulation of V gene expression in two phenotypically similar y/6 and a / p B2A2- thymocyte subsets is, therefore, intriguing. It is tempting to speculate that these thymocytes might have been subjected to selection processes. The expression of the CD5 molecule [16] and the absence of the B2A2 antigen are characteristic of mature T cells. Indeed, these cells are small resting lymphocytes [ 171 without precursor activity as assessed by in vivo reconstitution experiments [28]. Our recent results [29] indicate that F23.1+ B2A2- DN thymocytes have functional TcR mediating cellular proliferation, production of IL2 and IL3, as well as cytolysis. Moreover, the observation that in vivo treatment of neonatal mice with F23.1 mAb resulted in a marked reduction of this subset [29] would also be consistent with the notion that TcR a/p B2A2- DN thymocytes are terminally differentiated cells that might have been shaped by selecting or inducing processes. Whether any of these characteristics hold true for the TcR+ $6 B2A2- DN thymocytes remains conjectural at present. Vb and V, genes are interspersed in the same chromosomal region [21]. Alternative usage as V6 or V, genes has been described for V83N,6 [9], members of the Vs6N,7 gene family [22,25] and for a member of theV,4 gene family [13] in the mouse, as well as for DS6N,6 in man [30]. It is therefore of considerable interest to identify the particular ligands of y/6 TcR utilizing Vp6-V,2 genes. Conceivably, such ligands might also be involved in the selection process of a / P T cells carrying V,7 TcR. It remains to be seen whether the reported presence of an in-frame V86N,7-Ja1 rearranged gene on a DNA excision circle [25] might be a result of negative selection.
We thank Dr. !I Chienfor providing V, cDNA, Dr. E. Palmerfor the Va7.2cDNA and Dr. B. Arden for providing a TA27plasmid as well as an unpublished cDNA corresponding to V,+,23. Received June 1, 1989; in revised form August 21, 1989.
Developmental regulation of Vb genes in adult thymocytes
45
5 References 1 Zinkernagel, R. M. and Doherty, l? C., Nature 1974. 248: 701. 2 Bjorkman, P. J., Saper, M. A., Samraoui, B.. Bennett,W. S., Strominger, J. L. and Wiley, D. C., Nature 1987. 329: 506. 3 Townsend, A. R. M., Rothbard, J., Gotch, F. M., Bahadur. G., Wraith, D. and McMichael, A. J. Cell 1986. 44: 959. 4 Shimonkevitz, R., Kappler, J.W., Marrack, P. and Grey, H. M., J. Exp. Med. 1983. 158: 303. 5 Reinherz, E. L. and Schlossman, S., Cell 1980. 19: 821. 6 Davis, M. M. and Bjorkman, l? J., Nature 1988. 334: 395. 7 Raulet, D. H., Annu. Rev. Immunol. 1989. 7: 175. 8 Havran,W. and Allison, J. P., Nature 1988. 335: 443. 9 Elliott, J.F., Rock, E. P., Patten, P.A., Davis, M. M. and Chien, Y., Nature 1988. 331: 627. 10 Bonneville, M., Janeway, C. A., Ito, K., Haser,W., Ishida, I.. Nakanishi, N. and Tonegawa, S., Nature 1988. 336: 479. 11 Cron, R. Q., Koning, F., Maloy,W. L., Pardoll, D., Coligan, J. E. and Bluestone, J. A., J. Immunol. 1988. 141: 1074. 12 Pelkonen, J., Traunecker, A. and Karjalainen, K., EMBO J. 1987. 6: 1941. 13 Korman, A. J., Marusic-Galesic, S., Spencer, D., Kruisbeek, A. M. and Raulet, D. H., J. Exp. Med. 1988. 168: 1021. 14 Miescher, G. C., Howe, R. C., Budd, R. C. and Mac Donald, H. R., Ann. N . Y Acad. Sci. 1988. 532: 8 . 15 Budd, R. C., Miescher, G. C., Howe, R. C., Lees, R. K., Bron, C. and MacDonald, H. R., J. Exp. Med. 1987. 166: 577. 16 Miescher, G. C., Howe, R. C., Lees, R. K. and MacDonald, H. R., J. Immunol. 1988. 140: 1779. 17 Howe, R. C. and klxDonald, H. R., J. Immunol. 1988. 140: 1047. 18 Nabholz, M., Conzelmann, A., Acuto, 0..North, M., Haas, W., Pohlit, H., v. Boehmer, H., Hengartner, H., Mach, J.-l?. Engers, H . and Johnson, J. P., Immunol. Rev. 1980. 51: 125. 19 Miescher, G. C., Budd, R. C., Lees, R. K. and MacDonald, H. R., J. Immunol. 1987. 138: 1959. 20 Chomczynski, P. and Sacchi, N., Anal. Biochem. 1987. 162: 156. 21 Chien,Y., Iwashima, M., Kaplan, K. B., Elliott, J. F. and Davis, M. M., Nature 1987. 327: 677. 22 Yague, J., Blackman, M., Born,W., Marrack, P., Kappler, J. and Palmer, E., Nucleic Acids. Res. 1988. 16: (23): 11355. 23 Bill, J., Appe1,V. B. and Palmer, E., Proc. Natl. Acad. Sci. USA 1988. 85: 9184. 24 Arden, B., Klotz, J. L., Siu, G. and Hood, L. D.. Nature 1985. 316: 783. 25 Okazaki, K. and Sakano, H., EMBO J. 1988. 7 (6): 1669. 26 Chou, H. S., Anderson, S. J., Louie, M. C., Godambe, S. A., Pozzi, M. R.,Behlke, M. A., Huppi, K. and Loh, D.Y.. Proc. Natl. Acad. Sci. USA 1987. 84: 1992. 27 Winoto, A, and Baltimore, D., Nature 1989. 338: 430. 28 Crispe, I. N., Moore, M. W., Husmann, L. A.. Smith, L., Bevan, M. J. and Shimonkevitz, R. P., Nature 1987. 329: 336. 29 Howe, R. C., Pedrazzini,T. and MacDonald, H. R.. Eur. J. Immunol. 1989. 19: 25. 30 Guglielmi, P., Davi, F., D’Auriol, L.. Bones, J.-C., Dausset, J. and Bensussan, A., Proc. Natl. Acad. Sci. USA 1988. 85: 5634.
