Increased thymic B cells but maintenance of thymic structure, T cell differentiation and negative selection in lymphotoxin-alpha and TNF gene-targeted mice.

TNF, lymphotoxin (LT) and their receptors are expressed constitutively in the thymus. It remains unclear whether these cytokines play a role in normal thymic structure or function. We have investigated thymocyte differentiation, selection and thymic organogenesis in gene targeted mice lacking LTalpha, TNF, or both (TNF/LTalpha-/-). The thymus was normal in TNF/LTalpha-/- mice with regard to cell yields and stromal architecture. Detailed analysis of alphabeta and gammadelta T cell-lineage thymocyte subsets revealed no abnormalities, implying that neither TNF nor LT play an essential role in T cell differentiation or positive selection. The number and distribution of thymic CD11c+ dendritic cells was also normal in the absence of both TNF and LTalpha. A three-fold increase in B cell numbers was observed consistently in the TNF/LTalpha-/- thymus. This phenotype was due entirely to the LTalpha deficiency and associated with changes in the hemopoietic compartment, rather than the thymic stromal compartment of LTalpha-/- mice. Finally, specific Vbeta8+ T cell deletion within the thymus following intrathymic injection of staphylococcal enterotoxin B (SEB) was TNF/LT independent. Thus, despite the presence of these cytokines and their receptors in the normal thymus, there appears no essential role for either TNF or LT in development of organ structure or for those processes associated with T cell repertoire selection.


INTRODUCTION
The thymus provides the microenvironment necessary for the development of T cells from lymphoid progenitor cells as well as for selective elimination of T cells that are potentially autoreactive. This process, termed negative selection, inhibits maturation of those cells expressing receptors reactive to self antigens and hence promotes the maintenance of self tolerance (Kisielow and von Boehmer, 1995). Intrathymic negative selection involves active elimination of cells through the induction of apoptosis (programmed cell death) in negatively selected cells (Surh and Sprent, 1994). This appears to involve a signal through the TCR of the developing thymocyte in conjunction with second signal(s) which are as yet poorly defined (Page et al., 1993;Page et al., 1996).
Both TNF and LT have a well documented role in the induction of apoptosis in general, inducing programmed cell death through signalling via the p55 TNF receptor (TNFRI) and LTI3R (Satin et al., 1995;Ware et al., 1995;Browning et al., 1996;Nagata, 1997). Several studies have demonstrated constitutive expression of TNF in the thymus in vivo (Giroir et al., 1992;Murphy et al., 1992;Wolf and Cohen, 1992;Deman et al., 1996), moreover, the thymus was the only organ in which the TNF promoter was constitutively active (Giroir et al., 1992). Similarly, LTc and LT3 are constitutively expressed in the thymus (Wolf and Cohen, 1992;Pokholok et al., 1995), as are the TNF receptors TNFRI (p55) and TNFRII (p75) to which TNF and the LTc3 homotrimers bind (Ryffel et al., 1991Tartaglia et al., 1991Murphy et al., 1994).
The receptor for the membrane bound LTo/13 complex, LT3R, has been shown to be highly expressed in thymic epithelial tissue (Ware et al., 1995). TNF is produced by both immature and mature thymocyte subsets following stimulation in vitro (Fischer et al., 1991;Zlotnik et al., 1992). Several studies have implicated a role for TNF in T cell differentiation and/or proliferation, at least in vitro. TNF has been implicated as an important stimulatory factor at several points of T cell differentiation Zuniga-Pflucker et al., 1995). Conversely, TNF leads to rapid apoptosis of immature thymocytes in cell suspension (Hernandez-Caselles and Stutman, 1993) and mice overexpressing TNF in their T cells have reduced CD4+CD8 + double positive (DP) thymocyte numbers (Probert et al., 1993). TNF expression also coincides with the ability of fibroblast antigen presenting cells to induce negative selection in vitro (Page et al., 1993). Although many biological effects of TNF appear to occur via TNFRI, TNFRII appears to be both necessary and sufficient for at least some aspects of TNF mediated thymocyte stimulation (Tartaglia et al., 1991;Grell et al., 1998). LTt/3 and TNF play an essential role as mediators of lymphoid organ ontogeny and maintenance of lymphoid structure (De Togni et al., 1994;Pasparakis et al., 1996;Koni et al., 1997;Korner et al., 1997;Cook et al., 1998), at least in part through induction of B and T cell homing chemokines (Ngo et al.,1_999).
Despite this abundance of indirect evidence suggesting a role for TNF and LT in thymic T cell development, clear support for such a role in vivo has been elusive. Antibody blocking experiments (Sytwu et al., 1996) have failed to indicate a role for TNF in T cell development or negative selection in vivo. However, such studies may be complicated by the difficulty of completely saturating antigenic sites in the thymus, particularly with non-saturating doses of antibody (Gabor et al., 1997), due to the, at least partially effective, blood thymus barrier (Raviola and Karnovsky, 1972). Studies with TNFRI/RII deficient mice (Pfeffer et al., 1993;Page et al., 1998), or transgenic mice expressing soluble LTI3R and TNFRI (Ettinger et al., 1998), showed no major thymic abnormalities although in vitro, but not in vivo negative selection, was impaired in the absence of TNFR signalling (Page et al., 1998). Given the constitutive expression of TNF/LT molecules and their receptors in the thymus, it is surprising that no clear role for these factors in thymus/T cell development has been identified thus far. However, considering that TNF and LT share some functional characteristics, it remained possible that redundancy in the activity of these factors, and/or their receptors, had obscured their role in thymus/T cell development in these studies. Specifically, signalling of membrane LTt/3 via the LTR would be maintained in TNFRI/RII deficient mice (Page et al., 1998), while signalling via TNFRII may be maintained in TNFRI deficient mice (Pfeffer et al., 1993) and transgenic mice expressing soluble LT3R and TNFRI (Ettinger et al., 1998). We hypothesized that TNF and/or LT, through effects on lymphoid ontogeny or through their capacity to induce apoptosis, may have an as yet unidentified function in thymic physiology/T cell development. To test this, we used previously described C57BL/6 mice with targeted disruption of the TNF, TNF/LTc (Korner et al., 1997) or LTot  genes. Mice in which the gene for LTa is deleted lack both the secreted LTot3 and predominant membrane LTctl32 forms of the LT molecule, the latter by virtue of the fact that expression of the LT[3 molecule at the cell-surface fails or is non-functional in the absence of LTc (Browning et al., 1993). Thus, TNF/LTc -/mice are completely LT and TNF-deficient, and not only lack lymph nodes and Peyer's patches (Eugster et al., 1996) but exhibit profound changes to the microarchitecture of the spleen that are greater in magnitude to that seen in mice that lack either cytokine alone ( (Korner et al., 1997;Riminton et al., 1998) and unpublished observations). We have used these mice to perform a comprehensive analysis of thymocyte subsets and thymic structure. This included detailed examination of thymocyte differentiation, including positive and negative selection (mediated by the superantigen SEB), as well as non-T lineage cells (B cells, macrophages, dendritic cells and thymic stromal cells). The results indicated no major differences in T lineage development relative to wild-type (WT) mice. However, increased numbers of B cells were identified in the TNF/LTot -/thymus by flow cytometry and immunohistology and this was attributed to the absence of LT. In summary, this study indicates that TNF and LT at best play a dispensable or redundant role in most aspects of intrathymic T cell development.

RESULTS
Thymic Structure is Maintained in the Absence of TNF and LT To determine whether the absence of both TNF and LT affected the integrity of the thymus, TNF/LTct -/thymuses were examined for gross structural abnormalities and for microarchitectural changes. These thymuses were macroscopically normal, comparable to WT thymuses in terms of tissue mass and cellularity. However, TNF/LTc -/thymocyte viability, as assessed by Trypan blue dye exclusion, was slightly but reproducibly lower than that of WT mice (Table I). Thymic stromal compartments were examined by immunohistology with mAb that facilitated the identification of particular regions of the thymus. These included molecules such as MHC class II, expressed by the fine cortical epithelial network and more densely expressed in the thymic medulla, and MTS33, expressed by cortical thymocytes and isolated medullary epithelial cell clusters (Godfrey et al., 1990) both of which enabled clear distinction between cortex and medulla ( Figure 1). No differences were detected in the thymus from WT versus TNF/LTot -/mice. Lymphocyte markers CD4 (not shown) and CD8 (Figure 1), both densely label the cortex and are less frequent in the thymic medulla. These also revealed no differences between the TNF/LTc -/and WT thymus. MTS 12 (not shown) and MTS16 ( Figure 1) label thymic blood vessels and associated connective tissue lining the perivascular space, respectively (Godfrey et al., 1990). Again, the staining pattern with both mAb was identical in thymi from WT and TNF/LTc -/mice illustrating conserved cortical/medullary compartmentalisation and normal vascular integrity. Thymic dendritic cells, identified by CD 11 c expression (Figure 1), were confined to the thymic medulla in both WT and TNF/LTct -/mice.

Major Thymocyte Subsets are Maintained in the TNF/LTc "/ Thymus
To determine whether the absence of TNF and LTt affected proportions of thymocyte subsets, multi-color flow cytometric analysis of isolated cells was performed ( Figure 2). Plots of CD4 vs CD8 thymocytes from WT and TNF/LTt -/mice showed no difference in the proportions of CD4-CDS-, CD4+CD8+, CD4+CDS-and CD4-CD8 + populations. tTCR (Figure2) and ?TCR expression (not shown) of each of these subsets was also examined revealing marginally fewer etTCR+cells within the CD4-CDSpopulation in the TNF/LTt -/thymus (p<0.05, Mann-Whitney U-test). This was not reflected in a loss of NKI.I+tTCR + cells (NKT cells) that represent a significant proportion of cTCR+CD4-CD8 thymocytes (Levitsky et al., 1991), as these were present in normal numbers in TNF/LT -/thymuses (not shown). HSA expression was also tested as this molecule is down-regulated at a late stage in medullary thymocyte development, well after thymocytes reach the CD4 or CD8 single positive (SP) stage. Again, no differences were detected between WT and TNF/LTt -/mice (not shown).

Proportions of Non-T-Lineage Cells in the Thymus of TNF/LTct "/ Mice
The expression of a range of markers defining non-T lineage cells within the heterogeneous CD4-CD8thymocyte subpopulation was examined. Equivalent numbers of CDllc + cells (highly enriched for dendritic cells) were found in both WT and TNF/LTot -/mice, consistent with the data of Figure 1 (not shown). In contrast, a three-fold increase in B220 + cells and a 2-fold increase in Mac-1 + cells was detected in CD4-CD8thymocytes from TNF/LTot -/compared to WT mice ( Figure 2).
Most B220 + Cells in the TNF/LTt "/" Thymus Exhibit a Peripheral B2 Phenotype The majority (>95%) of mature B cells found in the blood and lymphoid tissues of mice are of the B2 type with a well defined phenotype (Hardy and Hayakawa, 1994). Most (~90%) B220 + cells in the spleen are IgD high while few express the CD5 antigen that is characteristic of B1 cells, found particularly in the peritoneal cavity (Kipps, 1989). In contrast, a high proportion of B cells in the thymus are CD5+Mac-1 + (Miyama-Inaba et al., 1988). These B cells are thought to develop within the thymus from a local progenitor population (Mori et al., 1997). A more detailed phenotypic analysis was performed to determine whether the increase in B220+cells within the CD4-CD8thymic population in TNF/LTot -/mice represented a selective expansion of thymic B cells ( Figure 3A). Of B220 + cells in the WT mouse, around 60% were CD5 +re and around 65% IgD-ve, confirming the over-representation of this atypical subset in the thymus. In contrast, of B220+cells in the TNF/LTct-/-, 25% were CD5 +ve and around 50% IgD-ve. Thus, B cells in the TNF/LTc -/thymus were enriched for normal peripheral type B2 cells rather than the thymic variety. This also suggested that the more numerous Mac-1 + cells in the TNF/LTc -/thymus ( Figure 2) were probably not CD5+IgD-vethymic B cells but macrophage-lineage cells. Perivascular lymphocytic infiltrates in the liver and lung of mice lacking LTc have been described previously (Banks et al., 1995). Thus, the TNF/LTc -/thymuses were examined histologically to determine whether the increased thymic B cells were present as perivascular infiltrates. This was tested by double staining thymuses for blood vessel-associated connective tissue using MTS 16, which clearly identifies the outer border of perivascular spaces, and B cells using anti-B220. The increased frequency of B cells had no association with vasculature, but rather were localized in the thymic parenchyma (Figure 3 B and C).

Identification of the Basis of the Increased B Cell
Numbers in TNF/LTc'/'Mice To determine the cytokine responsible for increased B cell numbers in the thymus of TNF/LTot -/mice, the proportion of B220 + thymocytes was examined from mice deficient for either TNF (Korner et al., 1997) or LTot alone  one TNF/LT(z--mouse. The proportion of positive cells within each region is given as the Flow cytometric data shown is from one WT and mean percent _+ SEM of analyses from between 3 and 8 individual mice where n 8 (etITCR), n 6 (B220) and n 3 (CD lb) CD4-CD8-B220 + cells in TNF -/thymuses was similar to that of WT thymuses whereas B220 + cells in LTc -/thymuses were increased ( Figure 4) and similar to that observed in the TNF/LTt -/thymus ( Figure 2). This analysis indicated that the lack of LTc rather than TNF was responsible for increased B cell numbers in the TNF/LT(z -/thymus. To determine whether the increased B cell numbers in the LT(z -/thymus was due to the bone marrow-derived lymphoid component, or to some defect in the thymic stromal elements in the absence of LTt, irradiation, bone marrow chimeras were established where either LTc -/or WT bone marrow cells were injected into irradiated RAG-1 -/recipient mice and left to enable reconstitution to occur (Figure 4). Despite the presence of WT thymic stroma in both cases, recipients of LTt -/bone marrow showed increased B cell numbers indicating that defects in the LT -/hematopoietic compartment underlay the deposition of increased B cells in the thymus.

The Influence of TNF/LTc on Negative Selection in vivo
The influence of TNF/LTt on intrathymic negative selection was tested by intrathymic injection of the superantigen SEB. SEB directly binds to members of the VI8 TCR family on T cells and MHC-II molecules on antigen presenting cells (Woodland and Blackman, 1993). Intrathymic injection was used because systemic treatment with SEB can lead to non-specific deletion of thymocytes (Lin et al., 1992), possibly due to the increased levels of TNF that follows activation of peripheral T cells. Similar, TNF-mediated, non-specific deletion of non-TCR transgenic DP thymocytes has been reported following peptide-specific stimulation of i.v. adoptively transferred TCR transgenic T cells (Martin and Bevan, 1997). Thus, intrathymic injection of SEB avoided the artefacts associated with systemic T cell stimulation and the generation of levels of TNF suffi-cient in themselves to cause effects on thymus that would be present in WT but reduced in TNF/LTc -/mice. The dose of SEB (5tg) was shown in titration studies to effectively delete V8 + cells in WT mice when administered into the thymus, whereas the same dose administered i.v. did not (not shown), indicating a local intrathymic effect. Moreover, the proportion of V8 + splenic T cells was not affected following intrathymic or i.v. injection of 5tg SEB (M.J. Gabor, manuscript in preparation). The thymus was removed 20 hours following injection and thymocyte subsets defined by CD4 and CD8 expression analysed for the proportion of V[8 expressing cells ( Figure 5). The extent of V[8 + thymocyte deletion within CD4+CD8 and CD4-CD8 + subsets in TNF/LTc -/mice following SEB injection was identical to that of WT mice. SEB treatment did not lead to a decrease in non-SEB reactive V[5 + CD4 or CD8 SP thymocytes in either WT or TNF/LTt -/mice, their proportion increasing marginally due to a marked drop in the number of V[8 + thymocytes.

DISCUSSION
Several studies have suggested a role for TNF and LT in thymocyte proliferation and differentiation, and in negative selection Zuniga-Pflucker et al., 1995). However, studies using a neutralising anti-TNF antibody, TNF-RI/RII deficient mice, and transgenic mice expressing soluble LTR or TNFRI have failed to support a role for these factors in thymus physiology (Sytwu et al., 1996;Ettinger et al., 1998;Page et al., 1998). It is important to point out that despite these studies, a detailed analysis of thymic structure and T cell development, including intrathymic positive and negative selection, has not been carried out in the absence of both TNF and LT, hence their potential influence remains unclear. In this study, we have thoroughly examined many aspects of thymic structure, thymocyte differentiation, positive and negative selection in mice rendered genetically deficient for TNF and LT, The thymus appeared grossly normal, with no difference in cell yield or thymic weight, and no difference in thymic architecture as defined by a range of antibodies against thymocytes and stromal cells. However, we observed consistently a minor albeit significant decrease in cell viability among thymocytes from TNF/LTt -/mice. It is difficult to be certain what this means, although it is possible that there is reduced macrophage activation due to the absence of TNF, which would slow down the clearance of dead or dying thymocytes that constantly takes place during T cell development (Surh and Sprent, 1994). The absence of any gross structural abnormalities in these thymuses contrasts with that of the spleen and lymph nodes of these mice, where as previously reported, the sPleens are devoid of B cell follicles and the lymph nodes are non-existent (Banks et al., 1995;Eugster et al., 1996;Korner et al., 1997), an effect substantially due to the absence of LT (De Togni et al., 1994;Banks et al., 1995;Ettinger et al., 1996;Mackay et al., 1997). Thus, factors regulating the generation and/or maintenance of the thymus are clearly distinct from those controlling the peripheral lymphoid organs.
Most thymocyte subsets were present in normal proportions. This suggests that the transition through early CD3/tTCR-CD4-CD8-triple negative (TN) stages, and from the TN to the DP stage does not absolutely depend on either TNF or LTt, despite earlier studies in cell suspension culture showing a positive influence for TNF on these immature cells Zuniga-Pflucker et al., 1995). It is possible that the functions of TNF/LT are redundant and replaced by other factors in these mice, but it should also be considered that the influence of TNF in the earlier studies (all based upon in vitro experiments) does not represent a physiological role for this factor at this stage of T cell development. It should be pointed out that although the total CD4-CD8population was normal in TNF/LTc -/mice, the increased proportion of B cells and macrophages within this compartment must mean that some other CD4-CD8cells are correspondingly diminished. Although a small decrease was detected in the WT TNF'I" LT'/" WT--RAG-l"/" LT'/'-'RAG-I"I" 23.5 B220 FIGURE 4 Increased Thymic B Cells are due to LTct Deficiency in the Lymphoid Compartment of the Thymus. Thymocytes from mice as indicated were triple-labelled for CD4, CD8 and B220 expression. Histogram profiles represent B220 expression on CD4-CDS-thymocytes from mice as indicated. The fourth and fifth histograms represent thymocytes from irradiated C57BL/6 RAG-14mice that were were reconstituted with bone marrow cells derived from either WT or LTct--C57BL/6 mice. Reconstituted thymuses were assessed 12 weeks later, when engraftment was complete. Histogram profiles of WT, TNF -/and LTcz -/thymocytes are representative of at least 3 different mice. The fourth histogram is derived from mouse and the fifth representative of 2 mice frequency of cI3TCR+CD4-CD8 cells, this is unlikely to account for the difference. As lineage marker (Mac-l, B220, Gr-1, TER119, CD3, CD4, CD8) negative cells, representing real thymocyte precursors (Godfrey and Zlotnik, 1993) were not directly investigated it is possible that these cells were also reduced. However, if this was the case it had no apparent downstream effects on subsequent T cell subsets. The increased presence of B cells in the thymus due to LT deficiency was an unexpected result. This may be due to dysregulated control over the point at which thymocytes lose their multilineage potential, believed to occur between the CD44+CD25 and CD44+CD25 + TN stages (Godfrey and Zlotnik, 1993), leading to an increase in B cell differentiation from early thymocyte precursors. However, thymic B cells are normally considered to be B1 type cells (CD5+), whereas the increased B cells detected in the present study more closely resembled conventional B2 type cells, as found in spleen and lymph nodes (Miyama-Inaba et al., 1988). An alternative possibility therefore is that LTt is important in controlling the trafficking or proliferation of peripheral B cells that have found their way to the thymus. A more mundane explanation is that the three-fold increase in circulating leukocytes seen in LTt -/mice (Banks et al., 1995;Riminton et al., 1998) may simply lead to a spill-over of B cells (and monocytes, Figure 2) into the thymus. Perivascular accumulations of lymphocytes are seen in some tissues in LTt -/mice (Banks et al., 1995). Notably however, the B cells were not localized to the perivascular space, but deeper within the tissue (Figure 3). In the absence of a clearer understanding of the ontogeny and regulation of B cells in the thymus, it is difficult to speculate any further on the basis of the B cell increase in LTt -/mice.
Our studies have also shown that positive selection is not associated with TNF/LTc signalling, as post selection CD4+CD8 and CD4-CD8 + thymocytes were present in normal proportions. This is not a contentious issue, as these factors have never been associated with positive selection. A more equivocal problem is the role of TNF and LT in intrathymic negative selection, with evidence for (Hernandez-Caselles and Stutman, 1993;Page et al., 1996;Page et al., 1998) and against (Sytwu et al., 1996;Page et al., 1998) a role for TNF in this process, and no direct evidence either way for LT. The results presented in this manuscript are the most definitive to date, as negative selection has been studied in an in vivo thymic microenvironment which is structurally normal, yet completely deficient in both TNF and membrane LTet or secreted LTt3. These results show that negative selection induced by the superantigen SEB, was clearly TCR-mediated and completely Groups of three WT and TNF/LTc -/mice were anesthetized, the thymus exposed surgically and injected with either PBS or 5gg SEB in PBS (10tl), then the incision closed with a single surgical staple. Twenty hours later, mice were killed and tissues obtained for analysis. Cells were triple-labeled for CD4, CD8 and either VI38 (mAb F23.1 8.1,8.2,8.3) or VI35 (mAb MR9-4) and analysed by flow cytometry. Bars represent the percentage (mean SEM, where appropriate) of VI38 or VI35 cells within the CD4+CD8 or CD8+CD4 thymocyte populations. The data shown here represents the outcome of a single experiment in which all mice were injected with the same SEB preparation on the same day and then all mice sacrificed and cell phenotype examined together en the following day. A second study in which groups of two mice were examined produced a qualitatively identical outcome but with a marginally reduced magnitude of SEB-induced deletion (PBS-injected thymus essentially as per this figure. SEB-injected thymus, CD4+CD8-VI38 mean of 13%. CD4-CD8+V38 mean of 7%). Representative spleen data derived from a single experiment is shown. N.D. not done independent of these factors. Although superantigens such as SEB do not bind to the same part of the T cell receptor as conventional antigen (Woodland and Blackman, 1993), they do require costimulatory signals such as CD28 for the activation of peripheral T cells, suggesting a similar interaction between developing thymocytes and thymic antigen presenting cells involved in deletion. Clearly, it is technically very difficult to measure negative selection in response to a conventional antigen in non-TCR transgenic mice, and TCR transgenic TNF/LTt -/-C57BL/6 mice were not available. Furthermore, analysis of negative selection in such mice would require at least two mouse lines including an MHC class I and MHC class II-restricted TCR transgene to study selection of both CD4 and CD8 T cells.
Our results clearly demonstrated the rapid deletion of many SEB-reactive thymocytes within 20 hours of exposure to this superantigen. This included both less mature DP thymocytes (not shown) but was more apparent in the CD4 and CD8 SP thymocyte subsets, due to the significantly higher levels of TCR expression in this compartment. Although early studies had suggested that most negative selection occurs at the DP stage rather than the SP stage of T cell development (Kisielow and von Boehmer, 1990), this is clearly not absolute (Surh and Sprent, 1994;Kishimoto and Sprent, 1997;Kishimoto et al., 1998). The reason why some SP thymocytes survived this deletion following SEB encounter is uncertain, although it may reflect variations in the level of maturity among these cells, such that the most mature ones are resist-ant to deletion, as previously reported following anti-CD3 and SEB challenge in vivo (Kishimoto and Sprent, 1997;Kishimoto et al., 1998).
Taken together, this study provides the most definitive results to date of the role of TNF and LT in thymic function, clearly showing that both cytokines are dispensable for most aspects of intrathymic T cell differentiation, including early thymocyte differentiation, positive and negative selection and maintenance of thymic structure. However, the results indicate minor roles for these factors, including maintenance of cell viability in the thymus, and possibly a role for LTc in the generation or maintenance of B cell numbers in the thymus.

Cell Suspensions
Tissues were removed and gently ground between the frosted ends of microscope slides in PBS containing 5% FCS and 0.02% Na-Azide. Cells were washed by pelleting and resuspending in the same buffer. Cell numbers and viability were determined using a hemocytometer and Trypan blue (Sigma Chemical Co. Castle Hill, NSW, Australia) dye exclusion.
Fluorescence data was obtained using a FACScan TM (Becton Dickinson, San Jose CA. USA) or FACStar Plus TM (Becton Dickinson) and analysed using CEL-LQuest 3.0 software TM (Becton Dickinson).
Intrathymic Injections with SEB Mice were anesthetized by i.p. injection of (0.75mg) ketamine hydrochloride (Ketapex, Apex Laboratories Pty. Ltd. St. Marys, NSW, Australia) and (0.35 mg) Xylazine (Rompun, Bayer Ltd. Pymble, NSW. Australia). Once anesthetized, the thoracic cavity was opened via a small midline incision, and 5 gg of SEB (Sigma) in 10 gl of PBS, or PBS alone, was injected intrathymically. The incision was closed using a single surgical staple and the mice injected with 0.02 mg Buprenorphine analgesic (Temgesic, Reckitt & Colman Products Ltd. Hu,ll, UK) and allowed to recover in a warm environment. Mice were killed 20 hours after treatment and tissues harvested. Generation of Radiation Bone Marrow Chimeras C57BL/6 RAG-1 -/mice were lethally irradiated (550 rad of y-radiation day -2 and day 0) and on day 0, injected i.v. with 2 x 107 bone marrow cells derived from either WT or LTt -/-C57BL/6 mice. Thymuses of mice were assessed 12 weeks later, when engraftment was complete .