Early Deletion and Late Positive Selection of T-Cells Expressing a Male-Specific Receptor in T-Cell Receptor Transgenic Mice

The ontogeny of T cells in T-cell receptor (TCR) transgenic mice, which express a transgenic αβ heterodimer, specific for the male (H-Y) antigen in association with H-2Db, was determined. The transgenic α chain was expressed on about 10% of the fetal thymocytes on day 14 of gestation. About 50% of day-15 fetal thymocytes expressed both α and β transchains and virtually all fetal thymocytes expressed the transgenicαβ heterodimer by day 17. The early expression of the transgenic TCR on CD4-8- thymocytes prevented the development of γδ cells, and led to accelerated growth of thymocytes and an earlier expression of CD4 and CD8 molecules. Up to day 17, no significant differences in T-cell development could be detected between female and male thymuses. By day 18 of gestation, the male transgenic thymus contained more CD4-8- thymocytes than the female transgenic thymus. The preponderance of CD4-8- thymocytes in the male transgenic thymus increased until birth and was a consequence of the deletion of the CD4+8+ thymocytes and their CD4-8+ precursors. By the time of birth, the male transgenic thymus contained half the number of cells as the female transgenic thymus. The deletion of autospecific precursor cells in the male transgenic mouse began only at day 18 of gestation, despite the fact that the ligand could already be detected by day 16. The preferential accumulation of CD4-8+ T cells, which expressed a high density of the transgenic TCR, occurred only after birth and was .obvious in 6-week-old female thymus. These data support the hypothesis that the positive selection of T cells expressing this transgenic heterodimer may involve two steps, i.e., the commitment of CD4+8+ thymocytes to the CD4-8+ lineage following the interaction of the transgenic TCR with restricting major histocompatibility molecules, followed by a slow conversion of CD4+8+ thymocytes into CD4-8+ T cells. In normal mice, the precursors of CD+4+8 and single positive thymocytes have the CD4-8- CD3-J11d+ (or M1/69 +) phenotype. Because of the early expression of the transgenic αβ heterodimer, this population was not detected in adult transgenic mice. All CD4-8- M1/ 69+ cells expressed the transgenic receptor associated with CD3 and could be readily grown in media containing T-cell lectins and interleukin 2.


INTRODUCTION
During ontogeny, T-cell receptor (TCR) gene rearrangement and expression follow an ordered sequence. In mice, full-size transcripts of y and c TCR *Corresponding author. genes are detected by day 14 of gestation (Brenner et al., 1988) and precede the appearance of full-size /l and cz TCR gene transcripts that are formed by day 15 and day 17 of gestation, respectively (Raulet et al, 1985;Snodgrass et al., 1985a;Snodgrass et al., 1985b). TCR yc receptors are detected by day 15 (Brenner et al., 1988), and c/J receptors by day 17 (Cristanti et. al., 1986). In /J TCR transgenic mice, we were unable to detect yc cells in the thymus. The TCR transgene prevented cell surface expression of all and rearrangement of certain y genes (Vy4/Cy4) . This may be one of several reasons why usually yc and c] TCRs are not expressed in the same cell. Other evidence supports the view that the lineages of yc and cz/J TCRexpressing cells are determined prior to rearrangement of the c locus: analysis of c-chain genes in c/J T cells showed that the c TCR locus exists in germline configuration in DNA circles that are excised during ( TCR-gene rearrangement (Winoto and Baltimore, 1989). Because in our studies of fl TCR transgenic mice, all CD4-8-thymocytes express the /J transgene , we conclude that the expression of a rearranged / TCR gene in cells of the yc lineage either prevents cell surface expression of yc genes or aborts further differentiation and expansion of this lineage. When we introduced rearranged TCR c and ] transgenes (Kisielow et al., 1988a), the same phenomenon was observed: yc cells were not detectable in these mice. In this report, we investigate the ontogeny of TCR-gene expression in the c/J TCR transgenic mice in order to study whether an abnormally early expression of the c and/J TCR genes interfered with the development of the yc lineage.
The transgenic TCR was specific for the HY antigen in the context of H-2D b MHC molecules. In female TCR transgenic mice of the H-2 b haplotype, T cells expressing the male-specific cz/ TCR were positively selected by the restricting H-2D b molecule in the absence of the male antigen and were of the CD4-8 + phenotype (Kisielow et al., 1988b;Teh et al., 1988). In male TCR transgenic mice of the H-2 b haplotype, CD4+8 + immature thymocytes expressing the transgenic TCR were deleted (Kisielow et al., 1988a;Teh et al., 1989). Here we describe the sequence of phenotypic changes associated with positive and negative selection as they occur during ontogeny. Our conclusion from these studies is that negative selection can occur at the same stage or even on an earlier stage of thymocyte development as positive selection. guish whether the T3.70 antibody binds to an idiotype defined by both the c and/J TCR chain or only the c TCR chain of the B6.2.16 clone from which the c and /J TCR genes were isolated . This was now determined by transfecting the c and /J TCR genes into the CD4/8 / thymona 110, which expresses its own a/J TCR but which does not express /J TCR chains staining with the F23.1 monoclonal antibody and which does not stain with the T3.70 antibody. It turned out that some of the transfected cells expressed the transfected c TCR gene without concommitant expression of the transfected /J TCR gene. As these cells stained with the T3.70 antibody but not the F23.1 antibody, we conclude that the T3.70 antibody defines a determinant present on the transgenic c chain only ( positive and negative CD4-8thymocytes could be induced to proliferate. This means that the TCR/ CD3 complex in the early M1/69 positive precursor T cells can transduce signals that lead to T-cell activation and proliferation (Table 1). This is in contrast to the in vivo CD4+8 + progeny of these cells, which cannot be induced to proliferate. This could mean that the signal transduction is changed when CD4-8-cells differentiate into CD4 /8 / thymocytes. CD4-8thymocytes contain the precursors for all other subsets (Kisielow et al., 1984;Fowlkes et al., 1985), whereas CD4 /8 + thymocytes contain the precursors of CD4/8and CD4-8 / thymocytes (Kisielow et al., 1988a;Nicolic-Zugic and Bevan, 1988 FIGURE 4. Ontogeny of T cells in nontransgenic and transgenic mice. Thymocytes of the indicated age were stained with phycoerythrin-labeled CD4 and fluorescein-labeled CD8 mAb and analyzed in the FACScan flow cytometer, as previously described (Kisielow et al., 1988a;Teh et al., 1988). thymuses was especially pronounced in newborn transgenic embryos contained two major popuanimals, where the majority of male thymocytes lations, namely, CD4-8-(T"iflT hi and CD4+8 / were CD4-8-. Although the signs of deletion were czT/JT . At birth, the female transgenic thymus clearly apparent at birth, the phenotypic changes contained, in addition to the above populations, a associated with positive selection were not yet evi-discrete population of CD4+8 cells that expressed dent. All thymuses contained a significant populahigh levels of the transgenic/J chain, but lacked the tion of CD4/8-thymocytes at this time, but CD4-8 + transgenic cz chain. This population was lacking in thymocytes were not yet present in significant num-the male transgenic thymus, which contained only bers even in the thymus from female transgenic cells expressing high levels of the receptor, some of newborns. The overrepresentation of CD4-8 / which stained only faintly with CD4 or CD8 antithymocytes, as noticed previously, was, however, bodies. clearly apparent 6 weeks after birth. In summary, these data indicate that 'the pheno- The expression of (z and 1 transgenic TCR chains typic changes associated with deletion (deletion of and CD4 and CD8 coreceptors are shown in Fig. 5 Because we did not find any sign of deletion of CD4+8 + thymocytes by day 17 of gestation, we wondered if this was due to a late appearance of the ligand in the embryonic thymus. This was tested by culturing the B6.2.1 clone from which the c and /J transgenes were obtained with X-irradiated thymocytes from 16-day-old embryos and interleukin 2. The individual embryos were sexed using a malespecific probe (Lamar and Palmer, 1984). As shown in Table 2, the ligand was expressed on thymocytes from 16-day-old embryos as the thymocytes from male embryos induced significant proliferation in the B6.1.16 clone. Thus, the ligand was present 2 days earlier before we could detect the first signs of deletion, which resulted in a decreased number of CD4 +8 + thymocytes.

DISCUSSION
The data on the ontogeny of the expression of the transgenic TCR indicate that both the /J and the c TCR chain expressed relatively early on CD4-8-M1/ 69 + thymocytes by day 14 to 15 of gestation. One reason for this is that already rearranged genes are expressed earlier than rearranging genes. In fact, it was noted by some investigators that even in normal mice, some "early" CD4-8 / precursors of CD4+8 / thymocytes express CD3 molecules presumably associated with the c/J TCR (Nicolic-Zugic et al., 1988). Another reason for the early expression of the TCR c chain might be that its expression is regulated by the /J TCR enhancer located 5kb downstream of C2: it is conceivable that the .random integration of both the /J TCR and c TCR DNA into the genome of the transgenic mice has occurred in such a way that the expression of both genes is regulated by the same /J TCR enhancer. Obviously, most of the CD4-8-M1/69 / cells can express the CD3 complex that is associated with the c TCR. It appears that a significant proportion of the CD4-8-M1/69 / cells expressing the c/J TCR can be induced by antibodies against the TCR complex to express lymphokine receptors and proliferate in lymphokine-containing media. This is, however, not so with CD4+8 / thymocytes expressing the c/J TCR associated with CD3. This could have two possible interpretations: one is that the TCR-signal transluction is altered when CD4/8 + thymocytes are formed from CD4-8-precursors. The other is that only a subset of CD4-8-J11d / thymocytes can be induced to proliferate. One could, for instance, speculate that these are cells of the yc lineage that express the TCR transgenes. This needs to be studied in more detail.
It is clear that the early expression of the a/ TCR prevents the development of the yc lineage. Because in mice markers other than the TCR cannot be used to identify the yc lineage, it is not certain whether this is due to an abortive differentiation of this lineage or to a block of expression and/or rearrangement of the yc TCR in that lineage.
The early expression of the c/J TCR leads to some acceleration in T-cell development, i.e., the earlier expression of CD8 molecule and an accelerated growth. This is well in line with observations in scid mice into which rearranged TCR genes were intro-8 H.S. TEH et al.
duced by breeding and that resulted in the expres-MATERIALS AND METHODS sion of CD8 and CD4 coreceptors as well as an enormous expansion of thymocytes in vivo (Scott et Mice al., 1989). The analysis of the thymus at various stages of C57BL/6 (H-2b) and C57L (H-2b) mice were purchased from the Jackson Laboratory, Bar Harbor, gestation clearly indicates that the phenotypic Maine. TCR transgenic mice expressing both the a changes associated with negative selection are and ] genes, which were isolated from a cytolytic apparent much earlier than those associated with T-cell clone with specificity for the male antigen in positive selection. Although there are no significant association with H-2Db, were produced as previdifferences between the male and female thymuses up to day 18, the deletion of CD4/8 / thymocytes ously described (Kisielow et al., 1988a; and their CD4-8 / precursors became apparent at al., 1988). The cz/J transgenic mice used in the that time such that at birth the thymus contained studies reported here had been backcrossed for five to six generations to C57L female mice. mostly CD4-8thymocytes and a much reduced lymphocyte population. Despite the fact that the ligand for the transgenic TCR could be detected Ontogeny Studies already by day 16 of gestation, we were unable to detect deletion before day 18. This could mean that Male c/J transgenic mice were mated with C57L females. The day of which a vaginal plug was the deletion requires a certain antigen-presenting observed was designated as day 0. At various times cell that develops relatively late during ontogeny or of gestation, fetal thymuses were collected and that the cells themselves develop the mechanism leading to deletion relatively late. We prefer the analyzed individually for expression of specific cell former possibility because in adult life even the surface molecules. Single cell suspensions of both earliest stages of newly developing CD4/8 / thymo-fetal thymus lobes from individual animals were cytes are deleted, i.e., these cells are equipped with prepared by rubbing the thymuses between two the program that leads to cell death, pieces of 1 cm nylon with 40-/ mesh size. Trans-In contrast to negative selection, the phenotypic genic and nontransgenic fetuses were distinguished on the basis of expression of the transgenic cz and/l changes associated with positive selection develop chains on thymocytes. quite late in that the preponderance of CD4-8 / thymocytes is only apparent well after birth. Although these findings are consistent with the MonoclonalAntibodies (mAb) view that negative selection can precede positive selection during T-cell development, they do not Phycoerythrin-labeled CD4 (antimouse L3T4)(Dialprovide firm evidence that tlis is the case. The ynas et al., 1983), biotinylated CD8 (antimouse Lyttransgenic mice, however, provide evidence that 2) (Ledbetter and Herzenberg, 1979), and fluoresnegative selection can occur at the same or earlier cein-labeled CD8 mAbs were purchased from Becton stages of T-cell development than positive selection: Dickinson, Mountain View, California. Fluoresceinmale transgenic mice lack or have much more labeled CD4 was a kind gift of Dr. Jeffrey Ledbetter, reduced numbers of CD4+8 thymocytes that are not Oncogen, Seattle. The F23.1 (Staerz et al., 1985) and male-specific and express endogenous c TCR chains T3.70  mAbs were purified by (Kisielow et al., 1988a;Teh et al., 1989;and Fig. 5). protein A chromatography and used either in the These cells are obviously absent in male' mice unconjugated or biotinylated form. The M1/69 because their precursors, namely, CD4+8 / thymo-hybridoma cells (Springer et al., 1978)  Cytometer (Becton Dickinson) as previously described (Kisielow et al., 1988a;Teh et al., 1988).
Transfection of ( and/i TCR Genes Into T-Cell Lines Ten /g each of cosmid DNA that contained the functional TCR c chain gene (COSHYa36) or the TCR/J chain gene (COSHY9.1.14) was linearized by PvuI digestion. After digestion, DNAs were dissolved in 400/1 of PBS and added to 400/zl of 2 xl0" 110TC CD4 /8 / thymoma in PBS. COSHYc36 and COSHY/J9.1.14 were cotransfected by electroporation using a Gene Pulser (Bio-Rad) under the condition of 960/F and 250 V. After transfection, cells were cultured in medium for 2 days, and then transformants were selected in media containing 2 mg/ml G418.
Isolation of CC4-CD8-Thymocytes CD4-CD8-cells were isolated by treating cells (at a concentration of 107 cells/ml in RPMI medium plus 5% FCS) with previously determined optimal concentrations of the anti-CD8 monoclonal antibody HO2.2 (Gottlieb et al., 1980) and the anti-CD4 monoclonal antibody RL172 (Ceredig et al., 1985) for 30min, on ice. Rabbit complement (Cedar Lane Laboratories, Canada) was added to a final concentration of 1/10 and the cells were incubated for 45 min at 37C. Remaining viable cells were subjected to a second depletion step to ensure the purity of the CD4-8-population. This process involved incubating the cells (5 xl0"/ml in PBS plus 5% FCS) with the anti-CD8 monoclonal antibody 53.6.7 (Ledbetter and Herzenberg, 1979) and the anti-CD4 monoclonal antibody GK1.5 (Dialynas et al., 1983) for 30 min on ice. The cells were washed twice, resuspended to 5 xl06 cells/ml and incubated with a predetermined optimal concentration of Dynabeads (magnetic beads, Dynal, Norway) conjugated to sheep antimouse Ig. The cells and beads were rotated for 30 min at 4C and then subjected to magnetic separation.