Intrathymic selection of murine TCRa(3 + CD4 - CD8 - thymocytes

Transcription

1 International Immunology, Vol. 2, No Oxford University Press /90 $3.00 Intrathymic selection of murine TCRa(3 + CD4 - CD8 - thymocytes Mark Egerton and Roland Scollay The Waller and Eliza Hall Institute of Medical Research, Royal Parade, Parkville, Australia 3050 Key words clonal deletion, early thymocytes, positive selection, T cell receptor, thymocyte development Abstract The CD4CD8- thymocyte population contains the precursors of all other thymocytes. However, It also contains a significant proportion of cells which express surface TCRa/3, and have little or no precursor activity. Like peripheral T cells, but unlike most other thymocytes, these TCRa/3+CD4CD8- thymocytes do not express heat stable antigen. Both the origin and developmental status of these cells are unclear, and are the subject of this report. We have measured the proportion of V^8.1 + cells amongst TCR + HSA-CD4-CD8- thymocytes In Mls-1" versus Mls-1 b mice, In order to determine whether they have undergone negative selection. The proportions were similar In both strains, In contrast to mature T cells, indicating that neither they nor their precursors had undergone clonal deletion. We also measured the accumulation of these cells over the early life of the animal and found that It was extremely slow. Our data also show that although TCR-v\j8.1+ cells are reactive to Mls-1 1 In association with MHC class II, most mature TCR-V^8.1 + cells In Mls-1 b mice are CD8 +, suggesting an additional reactivity with MHC class I. We raise the possibility that TCR-V S 8.1+CD4CD8- thymocytes are derived from TCR-V^8.1 +CD4 + CD8+ thymocytes, and that the reactivity of TCR-V fl 8.1 with both MHC classes I and II has resulted In the down-regulation of both CD4 and CD8. Introduction Thymocytes lacking expression of the CD4 and CD8 surface molecules make up only 3-5% of the lymphoid cells in the murine thymus, but constitute a highly heterogeneous group of cells at varying stages of maturity (1-3). A useful marker that resolves two major fractions of CD4-CD8- thymocytes is heat stable antigen (HSA) (1,4,5). The major fraction contains cells that express high levels of HSA (HSA+ + ) are predominantly cycling (3,6,7) and account for nearly all CD4-CD8- thymocyte precursor activity, as judged by the ability to reconstitute an irradiated thymus (2,3,8). It is therefore believed these cells are engaged in the early events of mainstream T cell development in the thymus. The other major fraction of CD4-CD8- thymocytes is characterized by the complete lack of the HSA expression (HSA-). This group is absent from the embryonic thymus (9,10), and in the adult consists of non-cycling cells that turnover extremely slowly (8) and have little or no precursor activity in intrathymic reconstituton experiments (2,3,8). The majority of these cells express surface TCRa/3, with unusually high usage of Vp8 gene products, and account for nearly all TCRa/3 and Vg8 expression by CD4-CD8- thymocytes in some mouse strains (10-12). Minor fractions express TCR76 or no TCR at all (13,14). This evidence has led to the conclusion that TCRa/3+HSA-CD4-CD8- thymocytes are neither precursor cells nor part of the mainstream of T cell development (10), and their origin and function remains unclear. It is the question of the origin of the major TCRa/S+ group of HSA-CD4-CD8- thymocytes that we will address in this paper It has been assumed by some investigators that since TCRa/S+CD4-CD8- thymocytes do not express CD4 or CD8, they must be an unselected group of cells (15). However, the unusually high level of V^8 gene family usage in this population already suggests these cells have undergone some kind of selection Recent evidence from this laboratory that the CD8 gene is demethyiated in TCRa/3 + CD4-CD8- thymocytes (14) has suggested that these cells, or their precursors, may have once expressed CD8. Since positive and negative selection are believed to occur at the CD4+CD8+ stage (16-18), this raises the question of whether TCRa/3+CD4-CD8- thymocytes or their precursors have been subjected to these processes. We have tested this by taking advantage of two recent observations. Firstly, that TCRV^8.1 recognizes the Mls-1 stimulatory gene product (Mls-1 a ) in association with MHC class II (19). Secondly, that 50-60% of HSA-CD4-CD8- thymocytes express either TCRV08.I, 8.2or8.3inCBA/CaH(Mls-1b)mice(10-12), which Correspondence to' R. Scollay, as above Transmitting editor. T Hunig Received 11 September 1989, accepted 7 November 1989

2 158 Intrathymic selection of TCRa/3+CD4~CD8- thymocytes is a high enough proportion of this subset to allow an accurate comparison of TCRV^8.1 expression between Mls-1 disparate CBA mouse strains. Thus, in the thymus, Mls-1 a acts as a selfantigen when presented with MHC class II, with the result that both CD4+ and CD8+ TCRV/38.1+ cells are absent from the mature thymocyte and peripheral T cell populations, presumably as a result of clonal deletion at the CD4+CD8+ stage of development. We have asked whether there are less TCRj8.1 + cells among HSA-CD4-CD8- thymocytes in CBA/J (Mls-1 a ) mice than in CBA/CaH (Mls-1 b ) mice, and therefore whether Mls-1 b reactive cells have been deleted from this population HSA-CD4-C08" thymocytes Lymph node T cells Methods Mice Male CBA/CaH (Mls-1 b /2 b,h-2 k ) mice, used at 1-6 weeks of age, were bred under specific pathogen-free conditions at the Hall Institute animal facility Male CBA/J (Mls-1 a /2 a, H-2 k ) mice were used at 5-7 weeks of age, and were obtained from the CSIRO Animal Breeding Facility, North Ryde, Sydney. Lymphoid cell suspensions Thymus and lymph node cell suspensions were prepared as described elsewhere (1,20). Isolation of CD4-CD8- and HSA-CD4-CD8- thymocytes CD4-CD8- and HSA-CD4-CD8- thymocytes were isolated by a combination of cytotoxic and magnetic bead depletion. Full details of cytotoxic depletion procedures and mabs used for cytotoxic and magnetic bead depletion have been detailed elsewhere (1). In summary, for cytotoxic procedures they were anti-cd4, GK1.5; anti-cd8, ; and anti-hsa, B2A2 For magnetic bead depletion they were: anti-cd4, GK1 5; anti-cd8, ; and anti-hsa, M1/69. Briefly, thymus suspensions were treated with complement fixing anti-cd4 and anti-cd8 mab (and anti-hsa where applicable) and complement, followed by a density cut to retrieve viable cells The viable cells were then labelled with anti-cd4 and anti-cd8 mab (and anti-hsa where applicable) followed by addition of washed magnetic beads (Dynabeads, M-540, Dynal A.S., Oslo, Norway) at 5 mg/ml/10 7 cells. This suspension was incubated for 30 min on ice with frequent shaking, after which the free and cell-bound beads were removed using a strong magnet. A sample of each purified cell population was checked for purity using a second-stage fluorescein isothiocyanate (FITC)-conjugated anti-lg reagent, and routinely <3% of cells were of the depleted phenotype. Fluorescent reagents, mabs and flow cytometry For fluorescence procedures the mabs used were. anti-cd4, GK1.5; anti-cd8, , anti-hsa, M1/69; anti-cd3, 2C11; anti- TCRa/3, H (kindly provided by Dr Ralph Kubo); ant-tcr- Vp8, F23.2; anti-tcr-v p 8.2, F23.2; anti-tcr-vp8.1&8 2, KJ16. All antibodies were prepared in this laboratory and used as directly fluoresceinated reagents, as biotinylated reagents with phycoerythrin-avidin or texas-red avidin second stage, or as culture supernatants with a second-stage anti-lg coupled to either fluorescein or phycoerythrin fluorochromes. Detailed presentations of staining methods and two colour flow-cytometnc analysis have been described previously (1). Three-colour analyses were CD Log fluorescence Fig. 1. Fluorescence histograms illustrating F23 1, F23 2 and KJ16 mab stainmgof HSA-CD4"CD8' thymocytes from CBA/CaH (Mls-I 13 ) mice, and lymph node T cells from CBA/J (Mls-1 ) mice performed using a FACSTAR plus flow-cytometer (Becton- Dickenson, Sunnyvale, CA). FITC and phycoerythrin fluorochromes were excited using a single 488 nm argon laser, and individual FITC (peak 525 nm) and phycoerythrin (peak 570 nm) emissions were separated using a dichroic filter. Texas-red was excited by a separately aimed 600 nm dye-laser. Results Comparison of TCRV^8.1 expression by thymocytes of Mis disparate CBA mice HSA-CD4-CD8- The reactivity of TCRs containing particular /3 chain V regions has been shown in three separate cases. Cells expressing TCR- Vp8.1 are reactive to Mls-1 a together with MHC class II (19), TCR-V06+ cells are reactive to Mls-1 a together with I-E (21), and TCRV g 17a + cells are reactive to I-E alone (22). Hence in order to determine whether TCR-a/3+CD4-CD8- thymocytes had undergone negative selection, we measured the expression of TCR-Vp8.1 and TCR-V 3 6 by these cells in both CBA/J(Mls-1 a ) and CBA/CaH(Mls-1 b ) mice, and TCR-V fl 17a in I-E-(SWR) and I-E + (SWR x CBA/CaH)F, mice. Preliminary experiments indicated that both TCR-V^e and V^17a expression by HSA-CD4- CD8- thymocytes were too low (<2%) to accurately compare selected and non-selected strains (data not shown), hence we restricted our study to TCR-V^.I + expression. The proportion of cells expressing each of the three members of the V^8 gene family was determined by staining cells with F23.1 mab (anti-v08.1,2 and 3), KJ16 mab (anti-v s 8.1 and 2), and F23.2 mab (anti-vj8.2). The values for V^.I and Vg8.3 were obtained by subtraction. Examples of F23.1, F23.2, and KJ16 mab staining on CBA/CaH HSA-CD4-CD8- thymocytes and CBA/J lymph node T cells are presented in Fig. 1. The total proportion of cells expressing TCR-V^S, and the proportionate usage of Vg8.1, Vg8.2, and V S 8.3 gene products by TCR-V^+ cells are presented in Table 1.

3 Table 1. Proportionate usage of TCRV S 8 family members in Mis disparate mice Intrathymic selection of TCRa/3 + CD4~CD8- thymocytes 159 Lymph node cells % of T cells % of V^ i + cells HSA"CD4 % of cells "CD8" thymocytes % of V^} + cells CBA/CaH (Mls-1 b ) Total TCR-Vg8 + V/ V^8 2 + Vffl CBA/J (Mls-1") Total TCR-V^ V Vg8 2 + Vp Values were calculated from the average proportions of cells staining positively for F23.1, F23 2and KJ16 mab from two or three separate experiments The staining patterns were highly consistent between experiments Table 2. Kinetics of accumulation of HSA~CD4~CD8~ thymocytes Age (weeks) CD4 " CD8 "thymocytes %HSA" 8 ± 2 11 ± 2 14 ± 4 17 ± 2 27 ± 4 39 ± 5 HSA-CD4-CD8" Cells per thymus 1.2 x x x x x x 10 5 thymocytes %CD ± 1 70 ± ± 4 77 ± 7 %Vj ± 6 25 ± 6 34 ± 5 54 ± 2 60 ± 3 %TCR-a0 + Values expressed as a mean ± SD were obtained from three to eight separate experiments. All other values were obtained from two separate experiments. As has been previously established (19), there was almost complete absence of TCR-V^8.1 + peripheral T cells in Mls-1 a (CBA/J) mice but not in Mls-1 b (CBA/H) mice (1 % as compared with 7% of lymph node T cells, see Table 1). However, HSA-CD4-CD8- thymocytes in the two strains had similar proportions of TCR-Vp8.1 + cells, which were comparable to the proportion found amongst unseiected peripheral T cells in Mls-1 b mice. This indicated that they had not undergone clonal elimination with respect to Mls-1 a reactivity However, there was clearly a higher proportion of TCR-Vp8.2 + HSA-CD4-CD8- thymccytes than peripheral T cells in both strains, indicating that there may have been a selection process operating that favoured the accumulation of TCR-Vp8.2 + cells in the HSA-CD4-CD8- population, although the antigen reactivity of this receptor is unknown. 1 week 6 weeks 5 68 HSA CD3 TCR-»e Post-natal accumulation of HSA-CD4-CD8- thymocytes It has been established previously that HSA-CD4-CD8- cells are absent from the mouse embryo, and at birth (9,10), but account for nearly 40% of all CD4-CD8- thymocytes (in some mouse strains) by 6 weeks of age (10-12). More recently we studied the rate of turnover of these cells under semi-continuous tritiated thymidine ([ 3 H]TdR) labelling conditions, and found they turned-over extremely slowly at a rate of - 1 % (equivalent to 1.3 x 10 4 cells) per day (submitted). This suggested that although the entire HSA-CD4-CD8- group was generated sometime after birth, only a small proportion (1 % per day) of cells Log fluorescence TCR-V 88 Fig. 2. Fluorescence histograms showing HSA expression by CD4"CD8~ thymocytes, and CD3, TCRa0, and TCR-Vp8 expression by HSA-CD4-CD8" thymocytes from 1- and 6-week-dd CBA/CaH mee.

4 160 Intrathymic selection of TCRa/3+CD4-CD8- thymocytes FL1 Thymus CD4 Rg. 3. Representation of the gates used to select CD4 + CD8" and CD4"CD8 + subsets of CBA/CaH thymocytes and lymph-node cells for analysis of TCR expression. The gates used are indicated by straight black lines. were being generated in the 5- to 6-week-old mouse. Therefore it was unclear at which point during the early life of the animal the HSA-CD4-CD8- group was generated. To answer this question, we assessed the proportion of CD4-CD8- thymocytes that were HSA- in CBA/CaH mice ranging from 1 to 6 weeks of age (Table 2). We also calculated the number of HSA-CD4-CD8- cells present per thymus over this period (Table 2). CD3 and TCR-V^ expression by HSA-CD4-CD8- thymocytes and TCRa/3 expression by the CD3+ fraction were also measured (Table 2). Representative histograms at 1 and 6 weeks of age are presented in Fig. 2 The proportion of CD4-CD8- thymocytes that were HSAincreased steadily over the entire time period studied, as did the proportion of TCR-V08+ cells This increase was in the order of 2 x 10 s cells per week or 3 x 10 4 cells per day, which was consistent with previous data that showed a turnover rate for this population of 1.3 x 10 4 cells per day. The population increase was not due to the expansion of the few HSA-CD4- CD8- cells present at 1 week of age, since at no stage were there >3% of cells in cycle (determined by flow-cytometric DNA content analysis, data not shown). In 1-week-old mice a significant proportion of HSA-CD4-CD8- thymocytes were CD3 +, few expressed TCRa/3, and the majority of TCR expressed at this time was TCR-76. Thus, the post-natal accumulation of HSA-CD4-CD3- thymocytes was largely composed of TCRa/3+, and more particularly TCR-V P 8 + thymocytes, arising from an unknown source. E C o CD.0 E CD4 CD4+CD8- thymus CD4+C08- lymph node CD4-CD8+ thymus CD4"CD8 + lymph node TCR-Vfi. 1 + T cells preferentially express CD8 Given that the reactivity of developing thymocytes to Mls-1 a is dependent upon the co-expression of MHC class II by the Mis presenting cell and CD4 expression by the T cell (19,21,23), and that T cells can be positively selected according to their MHC reactivity alone (24-27), we wondered if in Mls-1 b mice, TCR- V08.I + cells would preferentially express CD4 due to positive selection by thymic MHC class II restricting elements. We therefore studied TCR-Vp8.1 expression by CD4+CD8- and CD4-CD8+ thymocytes and lymph node T cells using threecolour flow-cytometric analysis, and determining TCR-Vg8 gene Log fluorescence Fig. 4. Fluorescence histograms illustrating F23 1, F23.2 and KJ16 mab staining of CD4 + CD8" and CD4~CD8 + thymocytes and lymph node cells from CBA/CaH mice, gated as described in Figure 3

5 Intrathymic selection of TCRa0+CD4-CD8- thymocytes 161 Table 3. Proportionate usage of TCR-V^8 family members by mature thymocytes and lymph node T cells Thymocytes CD4+CD8" CD4-CD8+ Lymph node T cells CD4 + CD8" CD4-CD8+ Total TCR-V Vp8.1 + Vg8 2 + Vg The values for TCR-V^8 1, 2 and 3 usage are expressed as percentages of all TCR-Vg8 + cells from CBA/CaH mice Values were calculated using the average proportions of cells staining positively with F 23.1, F23 2 and KJ16 mab in two separate experiments. The staining patterns were almost identical in the two experiments family expression as previously described Fluorescence histograms of F23.1, F23 2, and KJ16 mab staining on CD4 + CD8- and CD4-CD8+ thymocytes and lymph node cells were obtained by gating the original three-colour analyses as shown in Fig. 3. Examples of these histograms are presented in Fig. 4. The proportions of cells expressing TCR-V fl 8 1, Vg8 2, and V^8 3 are presented in Table 3. Unexpectedly, not only did a higher proportion of CD4-CD8+ than CD4+CD8- cells express TCR-V P 8.1, but also the majority (65%) of TCR-V^.I cells were CD4-CD8 +. Conversely, TCR-Vp8.2 + cells were predominantly CD4 +. These data suggested that TCR-V^ 1 + cells were preferentially reactive to MHC class I in Mls-1 b mice, and had been positively selected on that basis. Discussion The data in this paper have shown that TCR-Vg8.1 + CD4-CD8- thymocytes exist in Mls-1 b mice in a similar proportion to those of Mls-1 mice Thus we have concluded that TCR-Vg8 1 + CD4- CD8- thymocytes represent a group of cells that have not undergone clonal deletion. Whether this applies only to TCR-V cells or to all TCRa/3-bearing HSA-CD4-CD8- cells is unclear, since we were only able to accurately measure TCR-Vg8 1 expression as an indicator of deletion. Given that clonal deletion is believed to occur at the CD4+CD8+ stage of intrathymic development (16-18), we conclude that TCR- V08 + HSA-CD4-CD8- cells cannot be the products of cells that have progressed beyond this stage. At face value, it may seem self-evident that thymocytes lacking CD4 and CD8 expression should represent an unselected population. In other words, our observation that were was no deletion of TCR-V S 8.1+CD4-CD8- thymocytes in Mls-1 a mice may be seen as evidence that these cells have never expressed CD4 and CD8 Yet recent studies in this laboratory have shown that the CD8 gene in TCRa/3 + HSA-CD4-CD8- thymocytes has been demethylated (14). The methytation pattern was consistent with previous expression of CD8, but was distinct from that of mature thymocytes or T cells, partial remethylation of the CD8 gene being required if these cells were the precursors of TCRa/3+HSA-CD4-CD8- thymocytes. Thus, although the precursors TCR-V^.I +CD4-CD8- thymocytes have not been subjected to clonal deletion, they may have expressed CD8, suggesting that these precursors were more likely to be CD4+CD8+ than CD4-CD8- thymocytes. If so, what is the mechanism whereby some TCR-V^.I +CD4+CD8+ thymocytes become CD4-CD8-? We have shown that TCR-V^.I + thymocytes, which are absent from the mature thymocyte and peripheral T cell populations of Mls-1 a mice, preferentially express CD8 in Mls-1 b mice in both the mature thymocyte and peripheral T cell populations. This preferential CD8 expression was presumably a result of positive selection by MHC class I elements expressed intrathymically, since several studies have shown that CD4+CD8- and CD4-CD8+ thymocytes are produced as a result of positive selection by MHC molecules (24-27). However, in Mls-1 a mice, TCR-V^8.1+CD4+CD8+ cells were deleted by an MHC class II restricted process (19). It seems then, that most TCR-Vg8.1 +CD4+CD8+ thymocytes are restricted to MHC class I in Mls-1 b mice, but in Mls-1 a mice these same cells are reactive to MHC class II associated with Mls-1 a determinants. Combined with evidence that thymocytes can be positively selected by MHC alone in the absence of nominal antigen (24-27) [in this case either the Mls-1 a antigen, or unknown antigen(s) in Mls-1 b mice], our results suggest that some TCR-V S 8.1 +CD4+CD8+ thymocytes may be capable of interacting with both MHC class I and class II in the thymus. If these interactions were to occur simultaneously, the result may be that both CD4 and CD8 are downregulated, resulting in the production of TCR-V^8.1 +CD4-CD8- thymocytes. Two obvious questions arise regarding this model (i) Why would only a small porportion of TCR-V^.I +CD4 + CD8 + thymocytes interact with both MHC classes, given that they may all be potentially reactive 7 (ii) Why is the proportion not less in Mls-1 a mice than in Mls-1 b mice, given that the presence of Mls-1 a should result in the deletion of all TCR-V^.I +CD4+CD8 + thymocytes 7 The answer to the first question may be that even if all TCR-V^8.1 + cells are potentially able to interact with both MHC classes, the limiting factor may be the chance of interacting simultaneously with class I and II, which is probably very low. Alternatively, cross-reactivity may be influenced by TCRor chain usage. The second question is more difficult to answer, except that even if Mls-1 a is presented together with class II, a chance simultaneous interaction with MHC class I may inhibit whatever mechanism normally results in the destruction of the TCR- Vg8.1 + thymocyte. This conclusion supports the hypothesis proposed by von Boehmer (28) that TCR bearing HSA-CD4- CD8- thymocytes were once CD4+CD8+ cells that interacted (by chance) with both class I and II MHC determinants simultaneously, resulting in down-regulation of both CD4 and CD8.

6 162 Intrathymic selection of TCRa/3+CD4 - CD8 - thymocytes Our results are in accordance with those of previous studies which have shown that although TCR-V? 6 is reactive to Mls-1 a (together with class II I-E), and TCR-V s 17a is reactive to I-E, in the absence of the relevant tolerizing molecules these TCRs can be expressed by either CD4+ or CD8+ mature T cells (25,26). However, the extremely low frequency of TCR-V^ and TCR- V ff 17a usage by HSA-CD4-CD8- thymocytes has prohibited us from extending our hypothesis to these cells. Consequently, at this stage we can only account for the minor, TCR-V^.I- 1 - fraction of HSA-CD4-CD8- thymocytes, although it may be that all TCRaj3 + CD4-CD8- thymocytes are the products of a 'cross-reaction' event, and as such be generated purely by accident. If so, the \/^2 + TCR must be particularly predisposed to MHC classes I and II cross-reactivity, since it is by far the most frequently expressed TCR on CD4-CD8- thymocytes. Alternatively, the over-expression of TCR-V^ 2 may be the result of a selection mechanism unrelated to that which we have proposed. This alternative mechanism may favour the retention of TCR-V08.2+ cells within the CD4-CD8- population to serve an as yet unknown intrathymic function. Our studies on the post-natal accumulation of TCR-bearing CD4-CD8- thymocytes showed that the production of these cells occurred exclusively after birth. The fact that this coincides with the initiation of TCR-V^- 1 - thymocyte deletion in Mls-1 a mice (29) is perhaps circumstantial evidence that the accumulation of these cells is related to the mtiation of tolerance ot Mls-1 a determinants. Given that the accumulation of HSA-CD4-CD8- cells proceeded at the rate of only 3 x 10 4 cells per day, these cells were only rarely produced when compared with the rate at which their proposed precursors turnover. The rate of CD4+CD8 + thymocyte turnover is ~3 x 10 7 cells per day, therefore the frequency at which TCRajS+HSA-CD4-CD8- cells arise would be roughly 1:1000, and the proposed MHC class I and II cross-reaction, even if responsible for the generation of all TCRa/S+HSA-CD4-CD8- thymocytes, must be a very rare event. Acknowledgements We wish to thank Ms Paula Gason for excellent technical assistance, and Dr Frank Battye, Mr Ralph Ross, and Ms Sheela Unnrthian of the WEHI Row Systems Laboratory for their help with the flow cytometry. We also thank Dr John O'Grady (CSIRO, Sydney) for supplying the CBA/J mice, Dr Ralph Kubo for providing the anti-tcr-a/3 mab, and Drs Ken Shortman, Grant Morahan, and Wendy Cook for helpful discussions regarding this paper This work was supported by Grant Al from the United States National Institutes of Health, by the C H Warman Research Fund, and by the National Health and Medical Research Council, Australia Abbreviations FITC HSA [ 3 H]TdR References fluorescein isothiocyanate heat stable antigen region tritiated thymidine 1 Wilson, A., D'Amico, A., Ewing, T.Scoday, R., and Shortman, K Subpopulations of early thymocytes- a cross-correlation flow-cytometnc analysis of adult mouse Ly 2" L3T4" (CD8~ CD4~) thymocytes using eight different surface markers. J. Immunol Cnspe, I. N, Moore, M W., Husmann, L. A, Smith, L, Bevan, M. J, and Shimonkevrtz, R P Differentiation potential of subsets of CD4~CD8~ thymocytes Nature 329: Pearse, M, Wu, L, Egerton.M, Wilson, A, Shortman, K, and Scollay, R 1989 An early thymocyte development sequence marked by transient expression of the IL-2 receptor. Proc Nati. Acad Sci. USA Scollay, R, Bartlett, P., and Shortman, K T cell development in the adult murine thymus. changes in the expression of the surface antigens Ly 2, L3T4 and B2A2 during development from early precusor cells to emigrants. Immunol, Rev Cnspe, I N. and Bevan, M J 1987 Expression and functional significance of the J11d marker on mouse thymocytes J Immunol 138: Howe, R C and MacDonald, H R Heterogeneity of immature (Lyt2~L3T4~) thymocytes Identification of four major phenotypically distinct subsets differing in cell cycle status and in in vitro activation requirements J. Immunol Ewing, T., Egerton.M, Wilson, A., Scollay, R, and Shortman, K Subpopulations of CD4~CD8~ murine thymocytes' differences in proliferation rate in vivo and proliferate responses in vitro Eur. J. Immunol Scollay, R, Wilson, A, D'Amico.A., Kelly, K, Egerton.M, Pearse, M., Wu, L, and Shortman K 1988 Developmental status and reconstitution potential of supopulations of munne thymocytes. Immunol Rev. 104:81. 9 Scollay, R Differences in Ly2" L3T4~ thymocytes from foetal and adult munne thymuses Immunol Lett Fowlkes, B. J, Kruisbeek, A. M., Ton-That, H, Weston, M A, Coligan, J E, Schwartz, R H, and Pardoll, D. M 1987 A novel population of T-cell receptor a/3-beanng thymocytes which predominantly expresses a single V^ gene family Nature Wilson, A, Ewing, T, Owens, T, Scollay, R, and Shortman, K 1988 T-cell antigen receptor expression by subsets of Ly2" L3T4~ (CD8~CD4~) thymocytes. J. Immunol Shortman, K., Wilson, A, Pearse, M., Gallagher, P, and SccJIay.R 1988 Mouse strain differences in subset distribution and T cell receptor expression among CD4~CD8~ thymocytes Immunol. Cell B10I Pearse, M., Gallagher, P., Wilson, A, Wu, L., Fisicaro, N, Miller, J, Scollay, R, and Shortman, K Molecular characterization of T-cell antigen receptor expression by subsets of CD4~CD8~ murine thymocytes. Proc Natl Acad. Sci. USA Wu, L, Pearse, M., Egerton, M, Petne, H., and Scollay, R 1990 CD4"CD8" thymocytes which express the T-cell receptor may have previously expressed CD8 Int Immunol Kotzkin, B. L., Babcock, S K, and Herron, L R 1988 Deletion of potentially self-reactive T cell specificities in L3T4~,Lyt-2" T cells of 1pr mice. J Exp. Med Hengartner, H., Odermatt, B, Schneider, R., Schreyer, M, Walle, G, MacDonald, H R., and Zinkernagel, R M Deletion of setfreactive T cells before entry into the thymus medulla Nature Sha, W. C, Nelson, C A, Newberry, R D, Kranz, D. M, Russell, J H, and Loh, D. Y Positive and negative selection of an antigen receptor on T cells in transgenic mice Nature Fowlkes, B. J., Schwartz, R. H, and Pardoll, D. M 1988 Deletion of self-reactive thymocytes occurs at a CD precursor stage. Nature Kappler, J. W., Staerz, U., White, J, and Marrack, P C Selftolerance eliminates T cells specific for Mis-modifies products of the major histocompatibility complex. Nature Shortman, K, Williams, N, and Adams, P The separation of different cell classes from lymphoid organs V. Simple procedures for the removal of cell debris, damaged cells and erythroid cells from lymphoid cell suspensions J. Immunol Methods MacDonald, H. R., Schneider, R, Lees, R. K, Howe, R. H, Acha- Orbea, H, Festenstein, H, Zinkernagel, R. M., and Hengartner, H T-cell receptor V/3 use predicts reactivity and tolerance to Mls-1 a- encoded antigens Nature Kappler, J. W., Roehm, N., and Marrack, P T cell tolerance by calonal elimination in the thymus Cell MacDonald, H R, Hengartner, H, and Pedrazzini, T Intrathymic deletion of self-reactive cells prevented by neonatal anti-

7 Intrathymic selection of TCRa0+CD4-CD8- thymocytes 163 CD4 antibody treatment Nature T cells Science Teh, H. S, Kisielow, P, Kishi, H, Uematsu, Y., Bluthmann, H, and 27 Scott, B, Bluthmann, H, Teh, H S, and von Boehmer, H The von Boehmer, H 1988 Thymic major histocompablity complex generation of mature T cells requires interaction of the a/3 T-cell antigens and the a/3 T-cell receptor determine the CD4/CD8 receptor with major histocompatibility antigens Nature 338:591 phenotype of T cells Nature von Boehmer, H 1988 The developmental biology of T lymphocytes 25 MacDonald, H R, Lees, R. K, Schneider, R, Zinkernagel, R. M, 1988 Annu Rev. Immunol and Hengartner, H Positive selection of CD4 + thymocytes 29 Schneider, R, Lees, R K, Pedrazzini, T., Zinkernage), R. M, controlled by MHC class II gene products Nature Hengartner, H, and MacDonald, H R 1989 Postnatal disappearance 26 Blackman, M A, Marrack, P, and Kappler, J 1989 Influence of the of self-reactive (V/36 + ) cells from the thymus of Mls-1 a mice. J Exp major histocompatibility complex on positive thymic selection of V/317a Med