| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gene-Deleted Mouse1
,
* Institut National de la Santé et de la Recherche Médicale Unité 548, Commissariat à l’Energie Atomique-Grenoble, Grenoble, France;
Developmental and Molecular Immunology, Department of Clinical and Biological Sciences, Center for Biomedicine, University of Basel, Basel, Switzerland; and
Institut National de la Santé et de la Recherche Médicale Unité 645, Université Franche-Comté, Établissement Français du Sang, Bourgogne Franche-Comté, Institut Fédératif de Recherche 133, Besançon, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
(pT) gene-deleted mice, the positively selectable CD4+CD8+ double-positive thymocyte pool is only 1% that in wild-type mice. Consequently, their peripheral T cell compartment is severely lymphopenic with a concomitant increase in proportion of CD25+FoxP3+ regulatory T cells. Using mixed bone marrow chimeras, where thymic output was 1% normal, the pT–/– peripheral T cell phenotype could be reproduced with normal cells. In the pT–/– thymus and peripheral lymphoid organs, FoxP3+CD4+ cells were enriched. Parabiosis experiments showed that many pT–/–CD4+ single-positive thymocytes represented recirculating peripheral T cells. Therefore, the enrichment of FoxP3+CD4+ single-positive thymocytes was not solely due to increased thymic production. Thus, the pT–/– mouse serves as a model system with which to study the consequences of chronic decreased thymic T cell production on the physiology of the peripheral T cell compartment. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
TCR, an early critical stage in their development involves successful rearrangements of TCR genes and expression of a TCR protein on the cell surface in association with a pre-existing protein called pre-T (pT)4 (3). The TCR/pT heterodimer and associated CD3 components are collectively called the pre-TCR and expressed by a CD25+CD44lowCD117low subpopulation of CD4–CD8–, double-negative (DN) cells, so-called DN3 cells (4). Via associated CD3 components, the pre-TCR delivers signals resulting in Notch-dependent, IL-7-independent proliferation of DN3 cells and their further differentiation, via a CD25–CD44–CD117– DN4 intermediate, to CD4+CD8+, double-positive (DP) cells (5). Following rearrangement of TCR genes and successful pairing of TCR and TCR proteins, the TCR heterodimer is expressed on the surface of DP cells. Then, the specificity of the TCR heterodimer expressed by a DP cell is assessed by the processes collectively called positive and negative selection (1). By the process of positive selection, those DP cells with a TCR of intermediate affinity continue their differentiation into the appropriate CD4 or CD8 single-positive (SP) compartment before migrating to peripheral lymphoid organs.
The creation of pT gene-deleted (pT–/–) mice was instrumental in demonstrating the crucial role played by the pre-TCR in controlling thymocyte development (6). In pT–/– mice, there is an incomplete block of thymocyte development at the DN3 to DN4 transition characterized by an accumulation of DN3 cells, drastic reduction in DN4 and DP cell number and reduced thymic cellularity (6, 7). As recently reviewed (4), most pT–/– DP cells do not express a cytoplasmic TCR protein and some retain CD25 surface expression. CD5 is an important surface molecule associated with thymocyte selection (8) and is initially up-regulated at the DN3 to DN4 transition. CD5 expression by DP cells plays a significant role in modulating TCR signaling during positive and negative selection (9). As shown using TCR-transgenic mouse lines, there appears to be a correlation between a cell’s CD5 expression level and its TCR affinity; increasing a cell’s CD5 expression can protect DP cells from negative selection (9).
Some time ago, a subpopulation of CD4 peripheral T cells, characterized by bright expression of CD5 was described (10) which had potent immunoregulatory function in that they could protect mice from autoimmunity induced by neonatal thymectomy (11). Later, this CD5high T cell subpopulation was found to express CD25 and CD25 became a surrogate marker for these so-called regulatory T cells (TRegs) (12). Intriguingly, we now know that the broad TCR repertoire of TRegs (13, 14, 15) appears to have an autoreactive bias and therefore, their bright expression of CD5 may be a mechanism by which they escape negative selection in the thymus. However, CD25 expression is not restricted to TReg and does not identify all cells with immunoregulatory function. The discovery that expression of a transcription factor of the forkhead group, called FoxP3, is intimately associated with the TReg lineage has provided a powerful tool with which to study the development and biology of TRegs (16, 17, 18). As recently reviewed (19, 20), TRegs inhibit the proliferation of activated and naive T cells in vitro, and from in vivo experiments the presence of TReg prevents the development of autoimmunity mostly by limiting the expansion of activated T cells.
pT–/– mice have been extensively studied in the context of thymocyte development (21). Thus, pT–/– mice have a small thymus with corresponding drastic reduction in TCR but normal TCR
T cell development. Less attention has been paid to the consequences of this chronic reduction in TCR cell production on the composition and function of the peripheral T cell compartment. In this report, we show that the peripheral T cell compartment of pT–/– mice is persistently lymphopenic associated with a concomitant increase in proportion of CD25+ FoxP3+, functional TReg. Additional analysis of the pT–/– DP compartment shows that only about half the cytoplasmic TCR-positive cells express CD5 at a level comparable to wild-type (WT) controls. Although the proportion of FoxP3+ CD4 SP pT–/– thymocytes was dramatically increased, there was no enrichment of FoxP3+ DP thymocytes and parabiosis experiments showed that many pT–/– CD4 SP thymocytes represented recirculating peripheral T cells. By using mixed TCR–/–:WT bone marrow (BM) chimeras with a severe, sustained reduction in thymic output, we show that the pT–/– peripheral T cell defect could be reproduced using normal cells. A similar enrichment in FoxP3+ cells was also observed in both IL-7–/– and IL-7R–/– (22, 23) T lymphopenic mice. Therefore, the pT–/– mouse serves as a model system with which to study the consequences of a chronic decreased thymic T cell production on the physiology of the peripheral T cell compartment.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
All mice in these studies were on the C57BL/6 (B6, or WT) background. B6 mice deficient for the pT gene (pT–/–) (6) were a gift from Dr. H. J. Fehling (University of Ulm, Ulm, Germany); B6 mice deficient for the RAG-2 gene (RAG-2–/–) were obtained from Iffa-Credo; B6 Ly5.1 congenic (B6.Ly5.1) and B6 mice expressing GFP from the CDTA and B6 mice deficient for TCR gene expression (TCR–/–) from Dr. H. R. MacDonald (Ludwig Institute of Cancer Research, Lausanne, Switzerland) and Dr. J. Kirberg, (Max Planck Institute of Immunobiology, Freiburg, Germany). All mice were bred in the Commissariat à l’Energie Atomique animal facilities or Department of Clinical and Biological Sciences. Parabiosis was conducted as previously described (24). All animals were kept in specific pathogen-free conditions and 6- to 24-wk-old females were used for experiments which were conducted according to institutional guidelines.
BM chimeras
BM cells obtained by flushing femurs from donor mice with PBS were depleted of T cells as previously described (25). For mixed chimeras, BM from both donor mouse strains were T cell depleted and mixed at various ratios indicated in the text before injection into irradiated recipients. Recipient mice were lethally irradiated (900 rad) with a 137Cs source and received 5–15 x 106 BM cells i.v. 4 h after irradiation (24). Mice were sacrificed for analysis 7–8 wk after reconstitution.
Cell preparation
Single-cell suspensions from BM, thymus, pooled inguinal, axillary, brachial, and cervical lymph nodes (LN) or spleen were prepared by pressing through a 100-µm nylon mesh into PBS 1% FCS. For flow cytometry, cells were washed and resuspended in PBS containing 1% FCS and 0.1% sodium azide (FACSwash). Viable cells were stained with trypan blue and counted in a hemocytometer. The total number of T or B cells was calculated from the frequency estimated by FACS analysis and the total number of living cells recovered per organ.
Flow cytometry
The following mAbs were used: anti-B220 (RA3-6B2), anti-CD4 (RM4-5 or GK1.5), anti-CD5 (53-7.3), anti-CD8 (53-6.7), anti-CD19 (1D3), anti-CD25 (7D4 or PC61), anti-CD44 (IM7), anti-CD45RB (16A), anti-CD69 (H1.2F3), anti-CD62L (Mel-14), CD103 (2E7), anti-CD152 (9H10), anti-Ki67(B56), anti-Ly6C (AL-21), anti-Ly5.1 (A20), and anti-Ly5.2 (104) (BD Pharmingen). Cell surface staining was performed as previously described (7, 15), and analyses were performed on a FACSCalibur interfaced to a Macintosh computer with BD Biosciences/CellQuest software or WinMDI2.8 software (Dr. J. Trotter, The Scripps Research Institute, San Diego, CA). Dead cells were excluded from analysis by a combination of light scatter and/or absence of propidium iodide (PI) staining. Intracellular staining for cytoplasmic TCR expression was conducted by standard procedures. Intracellular staining for FoxP3 was conducted using an anti-FoxP3 (mAb FJK-16S) staining kit according to the manufacturer’s instructions (eBioscience). BrdU labeling, staining, and analysis was conducted as previously described (7). Cell sorting was performed either on a Moflo (DakoCytomation) or FACSAria (BD Biosciences) and the purity of sorted cells was always >98%.
Proliferation and inhibition assays
A standard TReg assay was used (26). This involved culturing 2.5 x 104 CD4+CD25– naive B6 T cells either alone, or with titrated numbers of CD4+CD25+ T cells from B6 or pT–/– mice. B6 or pT–/– CD4+CD25+ T cells were also cultured alone. Cultures were stimulated with 2 µg/ml anti-CD3
(2C-11) and 5 x 104 irradiated syngeneic spleen cells and after 48 h pulsed for 16–18 h with 1 µCi/well [3H]thymidine (Amersham Biosciences; specific activity 185 GBq/mM).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
–/– mice
pT–/– mice have been extensively studied regarding thymus development (6), but little information is available regarding their peripheral T cells. As shown in Fig. 1a, we observed a considerable reduction in CD4 and CD8 T cell numbers in pooled LN of pT–/– mice at all ages tested. At 6 mo, T cells in both LN and spleen comprised 
10% that in age-matched WT controls. In contrast, staining with CD19 showed that pT–/– B cell numbers (Fig. 1a) and phenotype (data not shown) were normal and that peripheral lymphopenia was restricted to T cells.
|
–/– T cells showed that the CD4 subpopulation was enriched in CD44highCD45RBlowCD62Llow cells (Fig. 1b) and the CD8 subpopulation in CD122+CD44highLy6Chigh cells. Both these phenotypes are characteristic of so-called memory-like cells (27). BrdU-labeling experiments showed that after 48 h, incorporation was 3-fold higher in total CD4+, CD4+CD25+, and CD8+ T cells in pT–/– mice (Fig. 1c). Because WT CD44high cells are relatively enriched in cells incorporating BrdU (28), the difference between WT and pT–/– mice was considerably less for this subpopulation. Combined staining for annexin V and PI (Fig. 1d) showed that whereas the proportion of annexin+PI+CD19+ B cells was about the same, the proportion of annexin+PI+CD3+ T cells was 1.5-fold increased in pT–/– mice. Thus, T cells in pT–/– mice were enriched in memory-like cells, showed an increased incorporation of BrdU, an increased proportion of apoptotic cells, and their number remained low. Indeed, the pT–/– in vivo environment had a dominant, inhibitory effect on transferred WT T cells. Thus, whereas after transfer into CD3KO recipients, 106 WT CD8+ T cells after 6 days had undergone a mean of 3.5 cell divisions with only 5% cells remaining undivided, in pT–/– recipients they had only divided a mean of 1.5-fold with 80% cells remaining undivided (data not shown).
Previous studies had indicated that the TCR repertoire of pT–/– T cells is as diverse as WT mice (4, 29). As recently published (15) and using the small available panel of TCR V Abs, we compared the expressed V repertoire of CD25+ and CD25– subpopulations of WT and pT–/– CD4+ T cells. In WT (or pT–/–) mice, the percentages of Va11, Va8, Va3, and Va2-positive cells were fairly similar, namely 6.1 (8.1), 3.9 (3.6), 3.3 (4.7), 15.2 (12.6) for CD25–CD4 cells and 6.3 (8.1), 4.0 (4.7), 2.8 (4.5), 13.1 (10.3) for CD25+CD4 cells. To determine the functionality of peripheral T cells, numerous tests were undertaken, including allogeneic MLRs, where on a per-cell basis, the responses of WT and pT–/– mice were similar and T cell-dependent Ab responses to trinitrophenyl-coupled OVA, where primary and secondary Ag-specific Ab titers of WT and pT–/– mice were indistinguishable (data not shown). Transfer of CFSE-labeled pT–/– T cells into nonirradiated RAG-2–/– or CD3–/– recipients showed that both CD8 and CD4 T cells could undergo so-called lymphopenia-induced proliferation (data not shown). Three-color confocal microscopy showed that the few T cells in the spleens of pT–/– mice were correctly localized to the centre of B cell follicles (data not shown). In addition, after immunization with trinitrophenyl-OVA, germinal centers, as identified by combined staining for B220, IgM, and peanut lectin agglutinin, could be seen in the spleens of WT and pT–/– mice. Activation of T cells in vitro resulted in the rapid induction of intracytoplasmic IFN-, TNF-, and IL-2 expression, and release of cytokines into the supernatant (data not shown). Cytoplasmic staining for Bcl-2 failed to reveal any difference between WT and pT–/– T cells (data not shown)
Increased proportion of TReg in pT–/– mice
Recently, an association has been made between T cell lymphopenia and autoimmune disease (30, 31). However, throughout these studies, all pT–/– mice showed peripheral T cell lymphopenia, but none had manifestations of autoimmunity. Thus, there was no diarrhea or signs of inflammatory bowel disease, no lymphocytic infiltrations in organs such as intestine, liver, lung, or salivary gland and by indirect immunofluorescence on tissue sections from RAG-2–/– mice, no elevated titers of autoantibodies in the serum (data not shown).
Lymphopenia-induced autoimmunity may be controlled by so-called naturally arising CD4+CD25+ FoxP3+ TReg (32). Indeed, spleen and LN from pT–/– mice were enriched in CD25+ cells and these cells had the characteristic CD4low phenotype (15). When stained for cytoplasmic FoxP3 protein, CD4+ cells from WT mice contained the expected distribution of CD25 and FoxP3 (33) whereas the proportion of FoxP3+ cells was elevated in pT–/– mice (Fig. 1e, left cytograms). As reported by others (33), by FACS, there was no absolute correlation between CD25 and FoxP3 staining. Additional staining for CD5, CD44, CD45RB, CD62L, and CTLA4 did not reveal any phenotypic difference between CD25+ cells in pT–/– vs WT controls. However, 90% CD4+FoxP3+ cells were CD103+ in pT–/– mice, compared with only 55% in WT controls (Fig. 1e, right histograms).
Table I summarizes the results from a series of experiments with spleen or LN cells from adult mice stained for CD4, CD25, and FoxP3. Thus, the proportion of FoxP3+CD25– spleen cells was increased on average 3.1-fold and FoxP3+CD25+ cells 1.6-fold in pT–/– mice. In the pT–/– LN, the proportion of FoxP3+CD25– was increased 2.3-fold and FoxP3+CD25+ cells 3.3-fold. Note that despite an increased proportion, the absolute number of FoxP3+ cells was still lower in pT–/– mice than in WT controls (Table I). These differences were seen irrespective of mouse age (data not shown).
|
–/– mice were regulatory, a standard in vitro TReg assay was performed (26) using as responders, a fixed number of WT naive CD4+CD25– cells. Results obtained showed that CD4+CD25+ cells from WT or pT–/– mice blocked, in a dose-dependent fashion, the proliferation of naive responders (see Fig. 1f). CD4+CD25– cells from pT–/– mice seemed more potent at blocking proliferation than the equivalent number of WT cells. In addition, as would be expected for TReg, cultures containing WT or pT–/– CD4+CD25+ cells alone showed no proliferation under these culture conditions (right columns of Fig. 1f) (11, 20).
Detailed analysis of pT–/– thymocytes
To investigate why peripheral lymphopenia was so severe in pT–/– mice, we decided to analyze their thymuses in more detail. As reported previously, DP cells in pT–/– mice contained a reduced proportion (mean 37%) (4) of intracytoplasmic (i.c.) TCR+ cells and many DP were CD25+ (Fig. 2a) (4, 34).
|
and either surface CD25 or CD5, DP cells from WT mice were essentially all i.c.TCR+ CD25– and CD5+ (Fig. 2a). However, DP cells from pT–/– mice were quite different. They contained the reported 70 ± 1.5% (mean ± SD, n = 4) i.c.TCR– cells, of which 94% were CD5–. The 30% i.c.TCR+ cells could be subdivided into 14.6 ± 1.1% CD5– and 14.8 ± 1.6% CD5+ cells. As expected, 40% of pT–/– DP cells were CD25+ (34), but surface CD25 expression did not distinguish i.c.TCR– from i.c.TCR+ pT–/– DP subpopulations. In WT mice, CD5 was initially up-regulated at the CD8 immature SP (ISP) stage but in pT–/– mice, CD8 ISP cells were still CD5– (Fig. 2b). Whereas CD5 expression was heterogeneous on DP cells (see above), it was uniformly high on TCR+ CD8 and CD4 SP of pT–/– mice (data not shown and Fig. 2b). CD5 was also brightly expressed by WT and pT–/– TCR thymocytes (data not shown). Because 20% of pT–/– CD4 SP thymocytes were i.c.TCR– and 7% surface TCR+ (data not shown), CD4 SP profiles in Fig. 2b are those of cells additionally gated for TCR surface expression.
To demonstrate that the phenotype of DP thymocytes in pT–/– mice was cell autonomous, radiation BM chimeras were generated. Thus, reconstitution of lethally irradiated Ly5.1 WT recipients with pT–/– BM showed that the CD5 and CD25 distribution on Ly5.2+ pT–/– DP thymocytes was not due to some coincidental thymic stromal abnormality (Fig. 2a). Cell numbers of adult WT and pT–/– thymocytes are summarized in Table II.
|
–/– mice could be positively selected into the TCR cell lineage, we decided to analyze their TCR gene rearrangement status using a semiquantitative multiplex genomic DNA PCR assay (15). Results obtained indicated that at the genomic level, TCR rearrangements in i.c.TCR– CD5–, i.c.TCR+CD5–, and i.c.TCR+CD5+ DP subpopulations were indistinguishable (Fig. 2c). Taking into account thymic cellularity and the combined distribution of i.c.TCR and surface CD5 expression, we estimate that the number of CD5high i.c.TCR+ DP thymocytes in pT–/– mice capable of being positively selected to be <1% that in WT controls. Is the peripheral T cell phenotype cell autonomous?
To determine to what extent the peripheral T cell phenotype of pT–/– mice was cell autonomous, lethally irradiated Ly5.2 RAG-2–/– mice were reconstituted with a 1:1 mixture of Ly5.2 pT–/– and Ly5.1 WT BM cells. Whereas the peripheral B cell compartments were reconstituted by pT–/– and WT cells in a ratio similar to that of the BM inoculum (data not shown), at 6 wk after reconstitution, 95% of spleen CD4 T cells were derived from Ly5.1+ WT progenitors (Fig. 3a). Thus, WT cells outcompeted pT–/– T cells for generation and survival in the periphery; this is particularly true for the CD4+CD25+ subpopulation where their proportion decreased from 17.2 ± 1.2% in the spleen of pT–/– mice (Table I) to 5 ± 3.3% in 50% chimeras (Fig. 3a). This shows that the composition of the pT–/–CD25+ cell compartment can be influenced in trans by WT cells.
|
The lymphopenia in pT–/– mice could be due either to a pT–/–-specific defect in mature T cell behavior or to a direct consequence of their severe and chronic reduced thymic output. To distinguish between these possibilities and to try to recapitulate the reduced thymic output of the pT–/– mouse experimentally, we generated mixed BM chimeras in which thymic output of mature T cells could be varied (35). To this end, lethally irradiated Ly5.2 RAG-2–/– mice were reconstituted with different ratios of BM from Ly5.2 TCR–/– and Ly5.1 WT mice. As previously shown (35), in such mixed chimeras, both progenitor types contribute equally to the development of DP cells, but only WT DP cells are capable of being selected to the SP compartment. Analysis of peripheral T cells, which were all Ly5.1+, showed there was an inverse correlation between the total number of peripheral T cells and the proportion of CD25+ cells (Fig. 3b).
Thus, when WT progenitors comprised 1% of the BM inoculum, between 25 and 30% of peripheral CD4+ T cells were CD25+ and enriched in CD44high cells. Staining for intracellular Ki67, as a marker of proliferation, indicated that there was an inverse correlation between cell number and percent Ki67+ in both CD4+ and CD8+ cells (Fig. 3b, lower panels). Based on CD5, CD44, CD62L, and i.c. CTLA-4 expression, the TReg cells in mixed chimeras were identical with controls (data not shown). Thus, the elevated proportion of TReg in situations of peripheral T cell lymphopenia was not unique to pT–/– mice but could be reproduced in 1% WT:99% TCR–/– mixed BM chimeras. Indeed, in two other mutant mice where there is peripheral T cell lymphopenia, namely IL-7–/– and IL-7R–/– mice (22, 23), an increased proportion of FoxP3+ TReg was also found (Fig. 3c) a result recently confirmed in IL-7–/– mice by Peffault de Latour et al. (36).
Does the pT–/– thymus overproduce FoxP3+ cells?
To determine whether there was an increased proportion of FoxP3+ cells among pT–/– thymocytes, cells from WT or pT–/– mice were stained with CD4, CD8, CD25, and i.c.FoxP3, taking care to exclude doublets from analysis (Fig. 4a). In WT mice, 0.17 ± 0.05% DP and 4.3 ± 0.5% CD4 SP were FoxP3+ whereas there were 0.27 ± 0.09% FoxP3+ DP and 12.3 ± 4.8% FoxP3+ CD4 SP cells (mean ± SD, n = 3 for WT and n = 4 for pT–/–) in pT–/– mice. Taking into account cell numbers, there were 20-fold fewer FoxP3+ DP cells and 6-fold fewer FoxP3+ CD4 SP (Fig. 4a, right) in pT–/– mice. The ratio of FoxP3+ DP:CD4 SP was 4.3 in WT and 1.6 in pT–/– mice. Numerically, pT–/– DP thymocytes were enriched 2-fold in FoxP3+ cells.
|
+CD8+ and CD4+ pT–/– thymocytes were enriched in CD44high, so-called memory phenotype cells (27, 28) (Fig. 4b). BrdU-labeling experiments (Fig. 4c) showed that whereas the proportion of labeled DP cells was about the same in WT (50%) and pT–/– (65%) mice (upper panel), values for both CD25–CD4 SP (1.6 vs 38.5%) and CD25+CD4 SP (0.03 vs 22.3%) were dramatically different and in pT–/– mice, resembled that of peripheral T cells (Fig. 1c). Cell numbers from these experiments are summarized in the lower panel. These BrdU-labeling experiments revealed first that despite differences in CD25 expression, the fraction of DP cells labeled in WT and pT–/– mice was indistinguishable and second, that there was about a 1000-fold increase in the number of BrdU-labeled CD4+CD25+ pT–/– thymocytes. Are pTa–/– SP thymocytes recirculating T cells?
To distinguish between increased thymic production and increased recirculation as an explanation for the abundance of FoxP3+ SP in the pT–/– thymus, parabiosis experiments were established where GFP+ WT mice were parabiosed with pT–/– partners (24, 37). After 7 days, whereas the spleens of both mice contained a significant proportion of cells derived from the partner mouse (Fig. 5), only the thymus of the pT–/– mouse contained a significant fraction of GFP+ cells which were either CD4 or CD8 SP and not DN or DP cells (Fig. 5, upper panels). To positively identify pT–/–-derived T cells in WT parabionts, Ly5.2+ pT–/– mice were parabiosed with Ly5.1+ WT mice. Results were essentially similar (Fig. 5, lower panel) showing extensive chimerism in the pT–/– thymic SP compartment but very little migration of pT–/– cells to the WT thymus. Analysis of control parabionts between Ly5-congenic B6 mice showed that whereas there was extensive peripheral T cell chimerism (38), very few SP thymocytes were derived from the partner mouse. Staining for CD44 showed that the WT-derived CD4 SP in the pT–/– thymus were representative of WT peripheral T cells, being mostly CD44low, whereas the few pT–/–-derived T cells in the WT thymus were CD44high (data not shown). Thus, the rate of migration of WT TReg to the thymus of pT–/– mice was comparable to what had been previously observed for Treg migration between secondary lymphoid organs (38), i.e., slightly less than for naive T cells (data not shown). In addition, because pT–/– mice are lymphopenic, spleen T cell chimerism was considerably less in WT vs the pT–/– partner (38). Taken together, it would seem that the majority of pT–/– SP thymocytes represent recirculating peripheral T cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
–/– mice showing that they are enriched in CD4+CD25+FoxP3+ so-called naturally arising regulatory T cells. Deletion of the pT–/– gene results in the absence of pre-TCR signaling at the DN3 stage of thymocyte development and consequent decrease in the number of DP cells potentially capable of being positively selected. This results in a drastic and chronic reduction in mature T cell production by the thymus and the peripheral T cell compartment remains lymphopenic.
As recently reviewed (4), most DP thymocytes in pT–/– mice do not express intracytoplasmic TCR chains and in addition, i.c.TCR– DP disappear in pT–/– x TCR–/– double-deficient mice, suggesting that they are generated via a TCR-independent (39), TCR-dependent pathway (4, 21). At the genomic level, they nevertheless contained TCR rearrangements (Fig. 2c). Whatever the precise mechanism for their development, this did not result in their CD5 up-regulation. Initial CD5 up-regulation is a consequence of pre-TCR signaling initiated at the DN3 stage (8) and is visible at the CD8 ISP stage in WT but clearly absent in pT–/– mice (Fig. 2). Further CD5 up-regulation at the DP stage before positive selection may then reflect the functional expression of a TCR pair capable of supporting positive selection. Importantly, the 30% i.c.TCR+ pT–/– DP cells could be subdivided into CD5– and CD5+ subpopulations of approximately equal size. Some i.c.TCR+ DP cells may remain CD5– because they either lack correct TCR pairing or functional TCR rearrangements. Secondary TCR rearrangements might allow some CD5– i.c.TCR+ DP cells to up-regulate CD5 expression (40). However, our results suggest that only 15% of pT–/– DP, namely, i.c.TCR+CD5+ cells, are potentially capable of being positively selected. Numerically, their number is 1% that in the WT thymus. Thus, pT–/– mice have a chronic deficiency in thymic output with consequent persistent T cell lymphopenia.
Given the considerable interest in characterizing the consequences of reduced thymic output on peripheral T cell behavior (41), we analyzed the peripheral T cell compartment of pT–/– mice. Despite severe T cell lymphopenia, there were no obvious signs or symptoms of autoimmunity, presumably because of the relative enrichment in functional CD4+CD25+FoxP3+ TRegs which were enriched in CD103+ cells, a subpopulation of TRegs with increased regulatory T cell function (42, 43). The number and relative proportion of TReg among peripheral T cells are important parameters in the context of controlling autoimmunity (32). Adoptive transfer of small numbers of TReg can prevent the development of autoimmunity in either T lymphopenic normal (44) or mutant mice where the TReg lineage is affected (33). Intriguingly, the diverse TCR repertoire of TRegs appears to have an autoreactive bias (13, 33) and therefore, their bright expression of CD5 may be a mechanism by which they escaped negative selection in the thymus (9).
The increased proportion of CD4+CD25+FoxP3+ cells in pT–/– mice was probably not only due to increased thymic production. Thus, as in WT mice (45), the proportion of FoxP3+ DP cells was only increased 2-fold in adult or newborn pT–/– mice (Fig. 4 and data not shown). As shown by parabiosis experiments, the increased proportion of FoxP3+CD4+pT–/– SP thymocytes presumably reflected the increased recirculation of peripheral T cells to the thymus. Recirculation of peripheral mature T cells back to the neonatal thymus (46) and of activated cells to the adult thymus (47) was shown some time ago. Very recently, recirculation of naive peripheral T cells back to the aging thymus has been demonstrated (48). In the pT–/– thymus, CD4 SP were enriched in memory-like cells. Therefore, in pT–/– mice, most CD4SP "thymocytes" are not recently derived from DP precursors but rather represent recirculating peripheral T cells. Therefore, in situations of reduced thymic production or aging (48), the generation of SP cells from DP precursors could be overestimated; also, the distinction between "recent thymic migrants" and recirculating peripheral T cells becomes problematic.
The origin of peripheral TReg is still not fully resolved (20, 45, 49). Evidence, including those from fetal thymus organ cultures (R. Ceredig, unpublished data), indicates that some are generated intrathymically, but they can also be generated in the periphery (50, 51, 52), a situation that is perhaps favored by lymphopenia. Postthymic expansion of cells initially generated in the thymus was shown some time ago (53) and this may be particularly true for TRegs. Given that both CD4 and CD8 T cells in pT–/– mice have a memory-like phenotype (27), our results would best fit with the idea that there is considerable postthymic expansion of T cells, particularly TReg, and that they also recirculate back to the thymus (38).
One possibly trivial explanation for the pT–/– phenotype was that the absence of pre-TCR signaling at the DN3 stage of development somehow compromised mature T cell behavior, thereby resulting in peripheral T cell lymphopenia. However, by all functional criteria, including MLR, T-dependent Ab responses, T cell transfer experiments and the in vitro TReg assay, pT–/– T cells appeared normal. In addition, pT–/– T cells localized normally to T cell areas of spleen lymphoid follicles (data not shown). To show that the pT–/– phenotype was not due to a pT-specific defect but a consequence of a chronic reduction in thymic output, we generated TCR–/–:WT mixed BM chimeras where the proportion of selectable DP thymocytes approached the 1% of pT–/– mice. In the present study, there was an inverse correlation between the total peripheral CD4 T cell number and proportion of TRegs (Fig. 3). In two other lymphopenic mouse strains, namely IL-7–/– or IL-7R–/– (22, 23) mice, peripheral T lymphopenia was associated with an increased proportion of FoxP3+ cells. However, one could argue that the absence of IL-7 signaling was somehow responsible for their phenotype. The presence of FoxP3+ cells in both IL-7–/– and IL-7R–/– mice nevertheless demonstrates that the generation of TReg is not crucially dependent upon IL-7.
As recently reviewed (20), TReg cells reveal the paradox that the control of autoimmunity is mediated by a subpopulation of cells that is itself potentially autoreactive. Clearly, the potential beneficial effects of TReg are both to limit unwanted clonal expansion and to control autoimmunity. A potential harmful effect of such cells is that they may prevent antitumor immune responses (54). In the mammalian adaptive immune system, apparently a delicate balance exists between the presence or absence of autoimmunity. It is presently unclear whether the increased proportion of TRegs seen in association with peripheral T cell lymphopenia in pT–/– mice and BM chimeras is a phenomenon that can be generalized to other situations of T lymphopenia. Depletion of CD25+ cells in vivo failed to precipitate autoimmunity in pT–/– mice, that is, no detectable loss of weight or diarrhea (data not shown) was possible because they contain a sufficient pool of FoxP3+CD25– cells (Fig. 1e).
It seems surprising that despite some continuous thymic output, the peripheral lymphoid compartment of pT–/– mice remains lymphopenic. Our analysis shows that adult pT–/– T cells are enriched in memory-like cells, show an increased incorporation of BrdU, an increased proportion of apoptotic cells, and an enrichment in TReg. This altered state of equilibrium is maintained without any obvious signs of autoimmunity. To account for this, we propose a model outlined in Fig. 6. Thus, in the face of lymphopenia (upper right panels) and to control potential autoimmunity, the proportion and sphere of influence (as represented by larger circles) of TReg must increase. Indeed, based on the in vitro Treg assay (Fig. 1f) and increased CD103 expression (Fig. 1e) (43), it would seem that on a per cell basis, pT–/– TReg were more potent. The pT–/– lymphopenic environment seemed to have a dominant effect on transferred WT T cells (data not shown). It is generally accepted that in situations of T lymphopenia, there is increased cell turnover (55) (Fig. 1c), and consequent enrichment in "memory-like" cells (27) (Fig. 1b). As we also show, there is also a concomitant increase in cell loss (Fig. 1d). Situations of lymphopenia associated with reduced B or T lymphocyte production are associated with both increased cell turnover and increased cell death (35, 56). Thus, despite increased cellular turnover, the periphery remains lymphopenic. One possibility to consider is that reduced lymphocyte viability is associated with telomere erosion (57). Whether the pools of conventional and TReg cells are completely independent is unclear, but it is known that TReg can arise from activated peripheral T cells (50, 51, 52). By inducing proliferation, the generation of TReg may be favored, thereby increasing their proportion. Soluble factors, for example IL-2, produced by proliferating conventional T cells may also promote TReg survival (51). Finally, upon adoptive transfer, TReg are known to prevent the expansion of naive/conventional T cells (58). This "desirable" function of TReg in a T cell-replete environment may nevertheless have a detrimental effect in situations of lymphopenia, thereby preventing the expansion of potentially beneficial T cells and resulting in persistent lymphopenia.
|
–/– mice are quite different. In the former (10, 11, 30), the reduced peripheral T cell compartment is derived from a cohort of mature thymocytes relatively deficient in TReg that undergo extensive postthymic expansion. In the latter, the thymus is continually producing small numbers of T cells with a normal proportion of TReg, a situation analogous to that seen following HIV infection or adult BM transplantation. Thus, the pT–/– mouse may be a unique model with which to study peripheral T cell lymphopenia as a consequence of reduced thymic output. |
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was carried out while N.B. was supported by a PhD scholarship from the Commissariat à l’Energie Atomique, France. A.G.R. is holder of the chair in Immunology endowed by R. Hoffman-La Roche (Basel, Switzerland). 
2 N.B. and F.A. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Rhodri Ceredig, Department of Clinical and Biological Science, Developmental and Molecular Immunology, Center for Biomedicine, University of Basel, Mattenstrasse 28, 4058 Basel, Switzerland. E-mail address: Rod.Ceredig{at}unibas.ch
4 Abbreviations used in this paper: pT, pre-T; DN, double negative; DP, double positive; SP, single positive; TReg, regulatory T cell; WT, wild type; BM, bone marrow; LN, lymph node; PI, propidium iodide; i.c., intracytoplasmic; ISP, immature SP.
Received for publication May 18, 2006. Accepted for publication July 24, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
chain and a 33 kd glycoprotein. Cell 75: 283-294. [Medline]-chain: not just a surrogate. Nat. Rev. Immunol. 5: 571-577. [Medline] gene in development of but not T cells. Nature 375: 795-798. [Medline]-chains (CD25): breakdown of a single mechanism of self-tolerance causes various autoimmune diseases. J. Immunol. 155: 1151-1164. [Abstract] chain in the development and phenotype of naturally arising CD4+CD25+ T cells. Mol. Immunol. 43: 246-254. [Medline]-chain repertoire in pT-deficient mice is diverse and developmentally regulated: implications for pre-TCR functions and TCRA gene rearrangement. J. Immunol. 163: 6053-6059. expression generates TCR complexes that signal the DN-to-DP transition and impair development. Nat. Immunol. 3: 564-569. [Medline]
) are controlled and the background is gray. In T cell lymphopenic mice (right) where the naive:TReg ratio is 10:1 and the sphere of TReg influence is unchanged (upper panel), areas of the background appear white and FoxP3– cells located here are not controlled by TReg (hatched