On the Role of the Pre T Cell Receptor in versus T Lineage Commitment

Transcription

1 Immunity, Vol. 9, , November, 1998, Copyright 1998 by Cell Press On the Role of the Pre T Cell Receptor in versus T Lineage Commitment Iannis Aifantis, Orly Azogui, Jacqueline Feinberg, Claude Saint-Ruf, Jan Buer, and Harald von Boehmer* Institut Necker INSERM U373 that address specifically the role of the pre-tcr in versus lineage commitment. Nevertheless, the study of function of the pre-tcr at this branchpoint appears to represent a worthwhile exercise, because unlike the TCR that is formed effectively only late in T cell Faculté de Médecine development (Raulet et al., 1995; Snodgrass et al., 1995), Necker-Enfants-Malades the pre-tcr is generated almost at the same time as 156 Rue de Vaugirard the TCR, possibly within the same CD25 44 cells Paris Cedex 15 of the CD4 8 population of thymocytes that undergo France,, and rearrangement (Godfrey et al., 1993; Petrie et al., 1995). The CD25 44 cells express the pre T cell receptor (pt ) gene that can associate covalently with Summary the TCR chain (Groettrup et al., 1993; Saint-Ruf et al., 1994) and noncovalently with signal-transducing CD3 The role of the pre T cell receptor (TCR) in lineage proteins (Groettrup et al., 1992) to form the pre-tcr. commitment to the versus lineage of T cells was Recent data are consistent with the notion that addressed by analyzing TCR chain rearrangements in and T cells arise from a common precursor and that T cells from wild-type and pre-tcr-deficient mice the TCR is involved in lineage commitment: in T by single cell polymerase chain reaction. Results show cells, and rearrangements could be clearly detected that the pre-tcr selects against T cells containing but with an apparent selection against in frame re- rearranged V genes and that T cell precursors but arrangements (Dudley et al., 1995; Livak et al., 1995; not T cells express the pre-tcr protein. Furthercells could still become T cells but left open the Kang et al., 1995). This could suggest that failed T more, pre-tcr-induced proliferation could not be dequestion of what initially determined versus lintected in T cells. We propose that the pre-tcr commits developing T cells to the lineage by an eage commitment. With regard to the role of the pre- instructive mechanism that has largely replaced an TCR, it is clear that this receptor plays a crucial role in evolutionary more ancient stochastic mechanism of the generation of lineage cells, while it has apparently lineage commitment. no essential function in T cell development (Fehling et al., 1995). In fact, the absolute number of T cells was Introduction increased 3- to 10-fold in pt -deficient mice (Fehling et al., 1995, 1997). This could suggest that the pre-tcr takes away cells from the lineage and commits them Mechanisms of lineage commitment are of general interto the lineage. This notion, however, became quesest in biology but conclusive experiments in this field tionable when TCR rearrangements were analyzed in are the exception rather than the rule. This also applies T cells: TCR rearrangements were reported to be to immunology, where lineage commitment has been frequent in T cells from peripheral lymphoid tissue studied extensively within the lineage of T lymphocytes with an apparent selection for in frame rearrangements that separates early into the sublineages of cells with (70%) (Dudley et al., 1995). A somewhat lower estimate or T cell receptors for antigen (Shortman and was obtained when Burtrum et al. (1996) analyzed thymic Wu, 1996). While T cells have a well defined role in T cells and found over 50% of TCR rearrangehumoral and cell-mediated adaptive immune responses, ments to be in frame. A further reduction in the estimate the role of the T cells is less clear but may include of TCR in frame rearrangements (42%) for thymic immunoregulatory as well as direct immunoprotective T cells was reported by Mertsching et al. (1997). Finally, function (Kyes et al., 1995). At a later point in developin fetal thymocytes as well as epithelial cells, there ment, the lineage splits into the CD4 and CD8 was apparently no sign of selection for or against in sublineages that contain functionally distinct cells exframe TCR rearrangements (Dudley et al., 1995; Mertpressing antigen receptors with different specificity for sching et al., 1997). In spite of these different results, it class II and class I MHC molecules, respectively (von was argued that T cells or their precursors underwent Boehmer, 1990). At both branchpoints, the same ques- TCR selection by the pre-tcr (Burtrum et al., 1996; tion has been asked: do distinct signals initated by dif- Mertsching et al., 1997), much like TCR selection ocferent receptors that are randomly generated instruct curs in the lineage (von Boehmer and Fehling, 1997). developing cells to follow either one or the other path- This concept would obviously be difficult to combine way or is lineage commitment initially made indepenwith the notion that the pre-tcr commits cells to the dently of receptor-specific signals (Robey et al., 1991). lineage. There are numerous experiments that relate to both Since the latter results were obtained by sampling the versus and the CD4 versus CD8 lineage separearrangements of a few V gene segments only in a ration. There are, however, relatively few experiments population of T cells that may be contaminated by DNA from T cells and since there was no consensus * To whom correspondence should be addressed ( vonboehm with regard to the extent of selection, it appeared able to repeat these studies by sampling a larger number These authors contributed equally to this work. of V gene segments and by analyzing single cells rather

2 Immunity 650 Figure 1. Flow Cytometric Analysis of Thymocytes from pt and pt / Mice Total thymocytes from pt and pt / mice were surface-stained with FITC-conjugated anti-tcr antibodies. Cells from pt mice were enriched for the CD4 8 subset by negative depletion of CD4/CD8 positive cells using Dynabeads (Dynal) prior to staining. TCR positive cells were sorted according to the indicated gates using a FACS Vantage (Becton Dickinson) cell sorter. than cell populations. TCR rearrangements of thymic 45% of cells from pt mice contained at least one T cells were therefore analyzed by single cell PCR V rearrangement while 73% of cells from pt / mice (Aifantis et al., 1997) using primers that can detect the exhibited V rearrangements. This difference in V rearrangement rearrangement of the most frequently used V gene segments. is highly significant and indicates that rerearrangement Direct sequencing of PCR products as well as arrangement of the TCR locus proceeds much further cytoplasmic staining with TCR antibodies was used to in developing T cells in the absence of the pre-tcr determine the proportion of in frame rearrangements. (Table 1). In order to evaluate the putative role of the pre-tcr The extent of V rearrangement could be underestiin selecting TCR rearrangements, this analysis was mated if the PCR method would selectively fail to detect carried out in wild-type as well as pt / mice. Our V rearrangements for whatever reason. If this were so, experiments lead to the conclusion that the pre-tcr then cells with two V rearrangements should be more has a role in committing cells to the lineage rather often scored as cells with only one allele than cells that than positively selecting T cells or their precursors have only one V rearrangement. Therefore, if this were with in frame TCR rearrangements. the case the ratio of VDJ/DJ over VDJ/VDJ rearranged mature T cells that is normally 60 to 40% (Casanova Results et al., 1991) should be much higher; for instance, if we failed to see 50% of all V rearrangements it should be More Extensive TCR Rearrangement in 75 to 25%. When this was tested with 204 mature T T Cells from pt / versus pt Mice cells exhibiting two distinct rearrangements, the ratio Thymocytes bearing TCRs on the cell surface were was found to be 58 to 42. Therefore, we can conclude sorted independently from two pt / and two pt that the PCR does not underestimate the frequency of mice (Figure 1) and subjected to single cell PCR in order V rearrangement and therefore the values in Table 1 to determine the rearrangement status of the TCR lo- represent correct estimates on the extent of V rearrangement cus utilizing primers as previously described (Aifantis in T cells. et al., 1997). These primers detect some 90% of V The difference in the extent of V rearrangement in rearrangements in mature T cells (Sarukhan et al., pt versus pt / mice was also evident by the higher 1994; Aifantis et al., 1997) and the frequency of amplified proportion of cells with two VDJ rearrangements in cells gene products of the different V families correspond from the latter (Table 1). well to the frequency obtained by staining with antibodies (Sarukhan et al., 1994) when mature T cells from Selection against In Frame TCR Rearrangements B6 mice were analyzed (data not shown). Since our primers in T Cells from pt but Not pt / Mice detect D as well as V rearrangements and since From the rearrangement analysis in single cells, one can D segments are known to rearrange prior to V segments, calculate the fraction of cells that should express a TCR we analyzed only these cells in greater detail that exhibited clearly two distinct rearrangements at the TCR locus presumably occurring on both alleles. Such Table 1. Efficacy of Amplification and Frequency of TCR cells represented 45% of T cells from pt / and Rearrangements 42% of T cells from pt mice (Table 1). Cells exhibiting only one rearrangement or no rearrangement at all Cells Exhibiting Type of Rearrangements b were not further studied since they may arise as the Mouse Two Rearrangements DJ/DJ VDJ/DJ VDJ/VDJ result of inefficient single cell distribution by the cell pt 61 (42%) a 34 (56%) 23 (38%) 4 (7%) sorter, inefficient DNA extraction, or occasionally by the pt / 40 (45%) 11 (28%) 22 (55%) 7 (18%) use of inappropriate V primers. The observed fre- a Percentages of cells with two rearrangements among all cells anaquency of cells with two detectable rearrangements was lyzed. similar to that found when mature T cells were anaall b Number and percentage of cells with distinct rearrangements of lyzed by the same technique (data not shown). When cells with two rearrangements. The significance of differences the quality of rearrangement was probed, it became between pt and pt / mice with regard to VDJ rearrangements was determined by Chi-square test ( p 0.005). clear that T cells from the various mice differed: only

3 Pre-TCR and versus Lineage Commitment 651 Table 2. Productive TCR Genes in T Cells Observed percentage of cells TCR Percentage of TCR cells Cytoplasmic Staining (expected if no selection) Thymus Lymphnode pt 16% 12,9,7 (9%) a 12,11 (11%) pt / 28% 25,21,25 (24%) 25,23 (24%) Percentage of V 12 cells among TCR cells Percentage of V 12 cells (expected from sequencing) Thymus Lymphnode pt 17% 6,7 (6%) 8,8 (8%) pt / 58% 12,9 (10%) 6,8 (7%) Sequencing of VDJ Alleles Percentage of in frame alleles (expected from staining) pt 20% 6/31 (19%) b pt / 33% 12/28 (43%) a Percentage of individual mice (mean in parentheses). b In frame alleles per total alleles (percentages in parentheses). Percentage of observed in frame alleles of T cells from pt and 24% of T cells from pt / mice contain cytoplasmic TCR chains. When these values are compared with the calculated values on the basis of V rearrangements in T cells, it is clear that there is some selection against in frame V rearrangements in pt mice but little or no selection in pt / mice. A more direct evaluation on the fre- quency of in frame rearrangements can of course be derived from the direct sequencing of VDJ alleles. This is a time and money consuming exercise and therefore one is tempted to limit this to a number of alleles that yields statistically significant results. In our case, we sequenced 31 and 28 TCR VDJ alleles from T cells from pt and pt / mice, respectively. The frequency of in frame alleles was 19% for cells from pt and 43% for cells from pta / mice (Table 2). These values indicate a significant difference in the frequency of in frame alleles in the two populations of cells with some selection against in frame rearrangements in the former and some selection for in frame alleles in the latter. The extent of selection, however, is difficult to extrapolate from these data because of the limited number of sequences. Here, the data from the combined analysis of rearrangement and the determination of cells with productive rearrangements by staining is more reliable, since they are based on the analysis of much larger numbers of cells. The data then indicate a significant difference in the frequency of productive TCR rearrangements in T cells from pt versus pt / mice and a selection against in frame rearrangement in the former and no selection in the latter mice. Clearly this result would not be expected if the pre-tcr was mediating TCR selection in T cells but indicates that the pre-tcr takes away cells from the lineage that harbor productive TCR genes, thereby reducing the proportion of cells with productive V rearrangements in that population. With respect to the frequency of particular V rearrangements (Table 3), a relatively high contribution of independent V 12 gene rearrangements to productive genes was noted in pt / but not pt mice. Because of the small sample size, this difference was not significant, but nevertheless we have analyzed the contribution of V 12 proteins in T cells by cytoplasmic staining protein in case there is no V selection assuming that a value of 33% of in frame rearrangements corresponds to a situation where there is selection neither for nor against in frame rearrangements. As shown in Table 2, this would amount to a value of 16% of cells from T cells from pt mice and 28% of cells from T cells from pt / mice carrying a productive TCR gene. Cells with productive V rearrangements can be di- rectly determined by cytoplasmic staining of T cells with a pantcr antibody. Figure 2 shows that this stain- ing works well, and Table 2 shows that on average 9% Figure 2. Assessment of TCR Protein Expression in T Cells by Cytoplasmic Staining in pt and pt / Mice (A) Thymus. (B) Lymph node. Cells were obtained from 6- to 8-week- old pt / mice and wild-type littermate controls. Cells were surface stained with FITC-conjugated anti-tcr ; cytoplasmic staining was performed with biotinylated anti-pantcr, followed by streptavidin- PE. Percentages were determined by FACScan as indicated. WT, wild-type mice; IC, intracellular.

4 Immunity 652 chains are formed after the cells have been committed Table 3. Characterization of TCR Rearrangements in Single Cells to the lineage, for instance by expressing a TCR. In that case, the formation of the pre-tcr could be pt Frame a pt / Frame a prevented by the rapid shutoff of the expression of the V6J2.6/D1J2 V16J2.2/D1J2 pt gene. If so, the pre-tcr protein should not be V5J1.4/D2J2 V12J2.2/D1J2 found in TCR expressing T cells. This was in fact V5J2.6/D2J2 V8J2.6/D2J2 found to be true at the level of mrna (Bruno et al., 1995) V5J2.1/D1J2 V5J2.5/D2J2 V14J2.6/D2J2 V12J2.4/D1J2 but needed to be established at the level of protein and V11J1.5/D1J2 V8J1.5/D1J1 therefore was tested by cytoplasmic staining of T V12J2.2/D2J2 V12J1.1/D1J2 cells with a newly generated monoclonal pt antibody V8J2.5/D1J2 V4J2.6/D1J1 (Figure 3). This demonstrated the complete absence of V12J2.6/D1J2 V14J2.1/D1J1 pt protein in T cells. On the other hand, this antibody V10J2.3/D1J2 V11J1.4/D2J2 clearly stained CD25 44 precursors of T cells. While V8J2.6/D2J2 V12J1.4/D2J2 V14J1.5/D1J2 V1J2.6/D2J2 these findings do not prove that TCR chain containing V1J2.6/D2J2 V12J2.6/D2J2 T cells arise by expression of the TCR chain after V1J2.4/D1J1 V10J1.5/D1J2 commitment to the lineage, they are consistent with V8J2.2/D1J2 V5J1.4/D1J1 the notion that the pre-tcr expression in T cells is V5J1.4/D2J2 V10J2/D2J2 ND V11J1.5/D2J2 V1J1/D1J1 ND halted by rapid termination of pt gene expression. V1J2.6/D2J2 V5J2/D1J2 ND V16J1.3/D2J2 V8J2.6/D2J2 V12J1.1/D1J2/D2J2 V16J2/D1J1 ND No Evidence for Pre-TCR-Induced Proliferation V15J2.6/D1J1/D2J2 V16J2/D1J1/D1J2 ND in TCR Expressing T Cells V15J2.6/D1J1/D2J2 V10J2.2/D2J2 In order to analyze whether TCR containing T cells V6J2.6/D1J1/D2J2 V10J1/V16J2.4 ND/ in the thymus were cycling, perhaps by a signal received V4J1.1/V2J2.5/D2J2 / V8J2.5/V12J2.3 / before being fully differentiated, we analyzed the num- V6J2.1/V8J2.5/D1J1 / V2J2.6/V5J2.6 / ber of TCR and TCR T cells in G2, S, and M V15J1.4/V11J2.1/D2J2 / V6J2.5/V15J2.6 / V5J2.1/V8J2.6 / V6J2.5/V10J2.5 / phase of the cell cycle in pt and pt / mice (Figure V5J2/V12J2.6 ND/ 4). For this purpose, T cells with or without cyto- V12J2.6/V8J1 /ND plasmic TCR chains were sorted and their DNA content Data shown represent VDJ or DJ rearrangements of each allele. was analyzed. In two independent experiments, the proa The frame of each VDJ rearrangement was determined by sequence portion of cells in these phases of cell cycle was consis- analysis as described in the Experimental Procedures. tently but only slightly higher in TCR versus and correspond to in frame and out of frame TCR re- TCR T cells. In order to test whether this was due to arrangements, respectively. ND, not determined. the pre-tcr, the same experiment was performed with cells from pt / mice. Again, more TCR T cells were in cycle even though the absolute number of cywith V 12 antibodies. Indeed, the proportion of T cling cells was reduced. This then indicates that the cells with V 12 proteins was up to 2-fold higher in the difference in DNA content of TCR and TCR cells is thymus but not the lymph nodes of pta / versus pt actually not due to the pre-tcr. Also, the higher propormice (Table 2). Nevertheless, this difference cannot tion of T cells in these phases of cell cycle in pt solely account for the higher proportion of T cells with versus pt / mice cannot be attributed to cell-autonocytoplasmic TCR chains in pt / mice. The higher mous signaling by the pre-tcr, since this was not only frequency of individual V 12 rearrangements in produc- true for TCR, T cells. These experiments obviously tive TCR alleles in pt / mice is of some interest do not rule out pre-tcr-induced proliferation at earlier because the V 12 segment lies more proximal to the J stages of development but they do make the point that cluster and therefore may rearrange more often at earlier in mature T cells pre-tcr-induced proliferation is not time points than more distal V gene segments. If so, detectable. the data would suggest that early in frame V rearrangements are preferentially counter selected in pt but not pt / mice. Discussion The single cell PCR approach has the advantage over the analysis of populations of cells that it can be used to determine precisely the fraction of cells with particular V rearrangements. In addition, by using a larger set of primers it is possible to reliably amplify two distinct alleles in single cells rather than amplifying many different genes in a population of cells. We believe that the single cell PCR approach as described by Aifantis et al. (1997) reliably estimates the cells with distinct re- arrangements since it does not underestimate V rearrangement: it determines correctly the ratio of VDJ/ DJ and VDJ/VDJ containing cells in mature T cells, Absence of Pre-TCR Expression in T Cells but Not in T Cell Precursors The above data collectively suggest that the pre-tcr takes away from the lineage cells with productive TCR genes by committing them to the lineage. If so, one needs to account for the 9% of T cells in pt mice that still express TCR chains. This could for instance be due to the fact that such TCR chains cannot pair with the pt protein as has been claimed for some IgH chains that cannot pair with the invariant pre-b cell receptor proteins (ten Boekel et al., 1998). Perhaps more likely is the possibility that these TCR

5 Pre-TCR and versus Lineage Commitment 653 Figure 3. Flow Cytometric Analysis of pt Protein Expression in CD25 T Cell Progenitors but Not in Thymocytes (A) Detection of pre-tcr chain expression by a new MAb on the surface of the TCR transfected SCID thymocyte cell line SCB29 (upper panel). Intracellular expression of pre- TCR protein and TCR protein (anti-pantcr ) in total thymocytes from RAG / and pt / mice (lower panel). (B) Cytoplasmic staining for pre-tcr protein in thymocytes and CD25 thymocytes. Cells were stained with the monoclonal antibody against the pre-tcr chain or with an IgG1 MAb control with irrelevant specificity (as an indicator of background fluorescence). IC, intracellular. TCR antibodies. The latter approach indicates that there is selection against in frame TCR rearrangements in T cells from pt mice and no selection in T cells from pt / mice. There are, however, still T cells with productive TCR genes in pt mice, which could be explained by assuming that these particular TCR proteins are formed after the cells have been committed to the lineage and expression of the pre-tcr gene has been terminated. The latter is consistent with but not proven by our finding that thymic T cells do not express the pre-tcr protein. Our findings are not easily compatible with the notion that the pre-tcr posi- tively selects T cells or their precursors with in frame TCR rearrangements (Burtrum et al., 1996; Mertsching et al., 1997) and we were unable to document any increased proliferation of T cells with in frame TCR which is 60 to 40 (Casanova et al., 1991). If V rearrangements were underestimated, this approach should yield different ratios, since in that case cells with two V rearrangements would be twice as often affected as cells with one V rearrangement, i.e., they would be more often scored as cells with less than two distinct alleles. Therefore, our results indicate that V rearrangement proceeds much further in T cells in the absence of the pre-tcr. If the pre-tcr were to take away cells from the lineage, it could only do so by affecting cells with productive TCR rearrangements, and therefore the frequency of in frame TCR rearrangements in T cells from pt and pt / mice should differ. This was clearly established by both direct sequencing of PCR products as well as by staining populations of T cells with known V rearrangements with Figure 4. DNA Content and Its Correlation with Cytoplasmic Expression of TCR Proteins in Cells from pt and pt / Mice Sorted thymocytes from pt and pt / mice were stained with PI and nuclei analyzed for DNA content as described in the Experimental Procedures. Percentages of cells in S G2/M phases of the cell cycle were determined by FACScan using a doublet discrimination unit.

6 Immunity 654 rearrangements that was dependent on the pre-tcr. min at room temperature, washed twice in PBS, and revealed for The latter results, however, do not rule out any pre-tcrnology, 20 min at room temperature by streptavidin-pe (Southern Biotechinduced Birmingham, AL) diluted in PBS plus 0.5%. Cytoplasmic proliferation at an earlier stage of T cell staining was followed by two washing steps in PBS and 2 15 min development. on a rocking platform in PBS plus 2% fetal calf serum (FCS) plus On the basis of these results, we favor a model by 0.5% saponin on ice. Finally, cells were washed in PBS plus 2% which commitment to the lineage is made largely FCS. Stained samples were analyzed and sorted twice on a FACS but not exclusively by an instructive mechanism that is Vantage (Beckton Dickinson, San Jose, CA). Sorted TCR, TCR triggered by pre-tcr-dependent signals. This would be populations were reanalyzed on a FASCscan (Beckton Dickinson) compatible with the notion that early expression of to check for purity. or transgenes does not divert cells from the lineage pt -Monoclonal Antibody (Kang et al., 1995; Bruno et al., 1996), whereas early A fragment of DNA encoding the extramembrane domain of the expression of TCR transgenes, i.e., early expression of murine pre-tcr protein was ligated into a pgex expression vector the pre-tcr, prevents cells from entering the lineage (Pharmacia Biotech, Piscataway, New Jersey) that contained a fu- (von Boehmer et al., 1998). This would indicate that the sion system with the GST (glutathione S-transferase) gene and a and the TCR generate similar signals but that the tac promoter inducible by IPTG. The vector was then introduced pre-tcr that assembles in a different way with CD3 into Escherichia coli by transformation. C57BL/6 mice were immu- molecules (Groettrup et al., 1992; Berger et al., 1997) nized with 100 g of purified fusion protein and MAbs were produced by fusion of lymph node cells with the murine myeloma cell line signals differently. This then would indicate that the pre- Ag8653 according to standard procedures. Antigen specificity of TCR protein is not merely a surrogate TCR chain but hybridoma supernatants was assessed by FACS analysis of the would endow the pre-tcr with functions that differ from TCR transfected SCID thymocyte cell line SCB29 (Figure 3). In the TCR. We would, however, argue that while comand addition, anti-pt specificity was confirmed by immunoprecipitation mitment to the lineage by the pre-tcr is an entirely immunoblotting of the 33 kda pt protein (data not shown). instructive event, lineage commitment to the and Cytoplasmic stainings with the biotinylated anti-pt MAb (2F5.3 clone) were performed with minor modifications of the above protolineage before the advent of the pre-tcr was detercol: cells were fixed with paraformaldehyde 0.1% in PBS and permemined by stochastic mechanisms and perhaps regu- abilized with 0.1% of saponin in PBS plus 3% FCS and all antibody lated by Notch-like receptors, which would commit cells incubation steps were carried out at 4 C for 30 min. to the lineage irrespective of the quality of rearrangements (Washburn et al., 1997). This would ex- Single Cell PCR plain why cells with productive and genes can still TCR single cells were sorted using a FACS Vantage (Beckton enter the lineage, i.e., express CD4 and CD8 mole- Dickinson) equipped with an automatic cell deposition unit (Beckton cules and silence TCR expression (Buer et al., 1997; Dickinson) (Figure 1). DNA from single cells was prepared as pre- viously described (Aifantis et al., 1997). TCR gene rearrangements Passoni et al., 1997) and why cells that harbor were amplified by a seminested two step-pcr protocol (Löffert et transgenes do not all enter the lineage. Thus, we are al., 1996; Aifantis et al., 1997). In the first step, both alleles were postulating an evolutionary ancient mechanism of amplified simultaneously by addition to each tube of 35 l ofa versus lineage commitment that worked stochas- mixture containing dntps, buffer, and Taq polymerase at 0.5 U per tically and independent of TCR-specific signals, while sample (Perkin Elmer), 15 5 primers (3 pmol of each) homologous with the advent of the pre-tcr, development of the to 13 V gene families, D 1 and D 2 genes (Aifantis et al., 1997), and two 3 primers (3 pmol of each) that primed downstream of the lineage was mostly mediated by a pre-tcr-dependent J 1 and J 2 cluster sequences, respectively (Aifantis et al., 1997). instructive mechanism, perhaps because this receptor The first round of amplification was done in a final volume of 60 l regulates Notch-like receptors that determine commit- for 5 cycles in which the annealing temperature decreased from ment to the lineage (Washburn et al., 1997). 68 C to60 C, followed by 25 cycles of amplification (30 sec at 94 C, 1 min at 58 C, and 1 min at 72 C) and finally 5 min at 72 C. For the Experimental Procedures second round of amplification, 1 l of the first PCR product was transferred into separate tubes, each containing a single 5 primer Mice in combination with the nested J 2 orj 1 3 primer (10 pmol of The pt / mice have been described elsewhere (Fehling et al., each), dntp, buffer reaction, and 1U of Taq polymerase in a final 1995). The genotype of mice with respect to endogenous pt was volume of 20 l. Amplification was then carried out for 35 cycles determined by PCR. C57BL/6 and RAG / mice were purchased following the procedure of the first PCR. V and J were identified from IFFA CREDO (France). Animals were analyzed at 6 8 weeks of by migration of the total PCR product on a 1.5% ethidium bromide- age. Animal care was in accordance with institutional guidelines. stained agarose gel and positives were purified using geneclean III (Bio 101, Vista, CA). Direct sequencing of the PCR products was performed using the Ready Reaction DyeDeoxy Terminator Cycle Cytoplasmic Staining for TCR and Cell Sorting sequencing kit (Applied Biosystems, Warrington, Great Britain) and Total thymocytes were derived from C57BL/6 mice and pt / mice. sequenced by automated sequencing (Applied Biosystem). Cells from C57BL/6 mice were enriched for the CD4 8 subset by negative depletion of CD4/CD8 positive cells using Dynabeads (Dynal, Oslo, Norway). For extracellular/intracellular double staining Cell Cycle Analysis (Buer et al., 1997), cells were first incubated with culture supernatant DNA content was determined by sorting the TCR and TCR of MAb 2.4G2 to block FC RII/III receptors. Cells were then stained T cells by cytoplasmic staining for TCR and surface staining for for TCR surface expression with FITC-conjugated anti-tcr TCR of CD4 8 thymocytes. Propidium iodide (PI) staining of (GL3, Pharmigen, San Diego, CA) at optimal concentration. After DNA was performed as described (Paramithiotis et al., 1995). The washing in PBS, cells were fixed in PBS plus 0.5% paraformaldehyde fixation step was done using 0.5% paraformaldehyde in PBS. Sorted for 15 min at room temperature, followed by two washing steps cells were treated with 50 g/ml RNAse (Sigma Chemicals Co., in PBS. Cells were then permeabilized in 0.5% saponin for 10 min Saint Louis, MO) and 0.03 g/ml of PI (Sigma Chemicals Co.) and at room temperature and washed in PBS. Intracellular staining with incubated for 5 min at 37 C before analysis. FACS analysis was biotinylated anti-pantcr (H57-597) or anti-v 12 (MR11-1, Pharmingen) diluted in PBS plus 0.5% saponin was performed for 20 performed on a FASCscan (Beckton Dickinson) using a doublet discrimination module.

7 Pre-TCR and versus Lineage Commitment 655 Statistics rearrangements of TCR gamma genes influence the commitment of Statistical analysis was performed using the StatView software (Abacells. progenitor cells to differentiate into alpha beta or gamma delta T cus, Berkley, CA). P values for significance of differences of distributions Eur. J. Immunol. 25, between groups (pt and pt / mice) were determined by Kyes, S., and Hayday, A. (1995). Disparate types of gamma delta T Chi-square analysis. cell. Res. Immunol. 141, Livak, F., Petrie, H.T., Crispe, N., and Schatz, D.G. (1995). In-frame Acknowledgments TCR gene rearrangements play a critical role in the / T cell lineage decision. Immunity 2, We would like to thank Corinne Garcia for expert technical support Löffert, D., Ehlich, A., Muller, W., and Rajewsky, K. (1996). Surrogate with cytofluorometric analysis and cell sorting and Renato Monteiro light chain expression is required to establish immunoglobulin heavy for the IgG1 clone. I. A. is a recipient of a Biotechnology grant from chain allelic exclusion during early B cell development. Immunity 4, the European Commission. J. B. is supported by a grant from the Deutsche Forschungsgemeinschaft. H. v. B is supported by the Mertsching, E., Wilson, A., MacDonald, H.R., and Ceredig, R.T. Institut Universitaire de France, by the Juvenile Diabetes Foundation (1997). T cell receptor gene rearrangement and transcription in (U.S.A.), and by the Körber Foundation (Germany). Supported in part adult thymic cells. Eur. J. Immunol. 27, by the Institut National de la Santé et de la Recherche Médicale, Paramithiotis, E., Jacobsen, K.A., and Ratcliffe, M.J. (1995). Loss Paris, and by the Faculté Necker Enfants Malades, Descartes Uniof surface immunoglobulin expression precedes B cell death by versité, Paris. apoptosis in the bursa of Fabricius. J. Exp. Med. 181, Received July 22, 1998; revised September 5, Passoni, L., Hoffman, E.S., Kim, S., Crompton, T., Pao, W., Dong, M.Q., Owen, M.J., and Hayday, A.C. (1997). Intrathymic selection events in cell development. Immunity 7, References Petrie, H.T., Livak, F., Burtrum, D., and Mazel, S. (1995). T cell receptor gene recombination patterns and mechanisms: cell death, res- Aifantis, I., Buer, J., von Boehmer, H., and Azogui, O. (1997) Essential cue, and T cell production. J. Exp. Med. 182, role of the pre-t cell receptor in allelic exclusion of the T cell receptor Raulet, D.H., Garman, R.D., Saito, H., and Tonegawa, S. (1985). beta locus. Immunity 5, Developmental regulation of T-cell receptor gene expression. Nature Berger, M.A., Dave, V., Rhodes, M.R., Bosma, G.C., Bosma, M.J., 314, Kappes, D.J., and Wiest, D.L. (1997). Subunit composition of pre-t Robey, E.A., Fowlkes, B.J., Gordon, J.W., Kioussis, D., von Boehmer, cell receptor complexes expressed by primary thymocytes: CD3 H., Ramsdell, F., and Axel, R. (1991). Thymic selection in CD8 delta is physically associated but not functionally required. J. Exp. transgenic mice supports an instructive model for commitment to Med. 186, a CD4 or CD8 lineage. Cell 64, Buer, J., Aifantis, I., DiSanto, J.P., Fehling, H.J., and von Boehmer, Rodewald, H.R., Kretzschmar, K., Swat, W., and Takeda, S. (1995). H. (1997). Role of different T cell receptors in the development of Intrathymically expressed c-kit ligand (stem cell factor) is a major pre-t cells. J. Exp. Med. 185, factor driving expansion of very immature thymocytes in vivo. Immunity Burtrum, D.B., Kim, S., Dudley, E.C., Hayday, A.C., and Petrie, H.T. 3, (1996). TCR gene recombination and alpha beta-gamma delta lin- Saint-Ruf, C., Ungewiss, K., Groettrup, M., Bruno, L., Fehling, H.J., eage divergence: productive TCR-beta rearrangement is neither ex- and von Boehmer, H. (1994). Analysis and expression of a cloned clusive nor preclusive of gamma delta cell development. J. Immunol. pre-t cell receptor gene. Science 266, , Sarukhan, A., Gombert, J.M., Olivi, M., Bach, J.F., Carnaud, C., and Bruno, L., Rocha, B., Rolink, A., von Boehmer, H., and Rodewald, Garchon, H.J. (1994). Anchored polymerase chain reaction based H.R. (1995). Intra- and extra-thymic expression of the pre-t cell analysis of the V repertoire in the non-obese diabetic (NOD) mouse. receptor alpha gene. Eur. J. Immunol. 7, Eur. J. Immunol. 24, Bruno, L., Fehling, H.J., and von Boehmer, H. (1996). The T cell Shortman, K., and Wu, L. (1996). Early T lymphocyte progenitors. receptor can replace the receptor in the development of gamma Annu. Rev. Immunol. 14, delta lineage cells. Immunity 5, Snodgrass, H.R., Dembic, Z., Steinmetz, M., and von Boehmer, H. Casanova, J.L., Romero, P., Widmann, C., Kourilsky, P., and Maryanvelopment (1985). Expression of T-cell antigen receptor genes during fetal deski, J.L. (1991). T cell receptor genes in a series of class I major in the thymus. Nature 315, histocompatibility complex-restricted cytotoxic T lymphocyte clones ten Boekel, E., Melchers, F., and Rolink, A.G. (1998). Precursor B specific for a Plasmodium berghei nonapeptide: implications for T cells showing H chain allelic inclusion display allelic exclusion at cell allelic exclusion and antigen-specific repertoire. J. Exp. Med. the level of pre-b cell receptor surface expression. Immunity 2, 174, Dudley, E.C., Girardi, M., Owen, M.J., and Hayday, A.C. (1995). Alpha von Boehmer, H. (1990). Developmental biology of T cells in T cellbeta and gamma delta T cells can share a late common precursor. receptor transgenic mice. Annu. Rev. Immunol. 8, Curr. Biol. 5, von Boehmer, H., and Fehling, H.J. (1997). Structure and function Fehling, H.J., Krotkova, A., Saint-Ruf, C., and von Boehmer, H. of the pre-t cell receptor. Annu. Rev. Immunol. 15, (1995). Crucial role of the pre-t-cell receptor alpha gene in develop- von Boehmer, H., Bonneville, M., Ishida, I., Ryser, S., Lincoln, G., ment of alpha beta but not gamma delta T cells. Nature 375, Smith, R.T., Kishi, H., Scott, B., Kisielow, P., and Tonegawa, S. Godfrey, D.I., Kennedy, J., Mombaerts, P., Tonegawa, S., and Zlotsuppresses rearrangement of the V 4 gene segment. Proc. Natl. (1988). Early expression of a T-cell receptor beta-chain transgene nik, A. (1994). Onset of TCR-beta gene rearrangement and role of Acad. Sci. USA 24, TCR-beta expression during CD3 CD4 CD8 thymocyte differenvon tiation. J. Immunol. 152, Boehmer, H., Aifantis, I., Azogui, O., Feinberg, J., Saint-Ruf, C., Zober, C., Garcia, C., and Buer, J. (1998). Crucial function of the Groettrup, M., Baron, A., Griffiths, G., Palacios, R., and von Boehmer, pre-t cell receptor in TCR selection, TCR allelic exclusion and H. (1992). T cell receptor (TCR) beta chain homodimers on the sur- versus lineage commitment. Immunol. Rev. 165, face of immature but not mature alpha, gamma, delta chain deficient lines. EMBO J., 11, Washburn, T., Schweighoffer, E., Gridley, T., Chang, D., Fowlkes, B.J., Cado, D., and Robey, E. (1997). Notch activity influences the Groettrup, M., Ungewiss, K., Azogui, O., Palacios, R., Owen, M.J., versus T cell lineage decision. Cell 88, Hayday, A.C., and von Boehmer, H. (1993). A novel disulfide-linked heterodimer on pre-t cells consists of the T cell receptor chain and a 33 kd glycoprotein. Cell 75, Kang, J., Baker, J., and Raulet, D.H. (1995). Evidence that productive