Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Eur J Immunol. Author manuscript; available in PMC 2010 December 1.
Published in final edited form as:
PMCID: PMC2829603

Thymocyte deletion can bias regulatory T cell formation toward low abundance self-peptide


Autoreactive CD4+ T cells can undergo deletion and/or become CD25+Foxp3+ Treg cells as they develop intrathymically, but how these alternative developmental fates are specified based on interactions with self-peptide(s) is not understood. We show here that thymocytes expressing an autoreactive TCR can be subjected to varying degrees of deletion that correlate with the amount of self-peptide. Strikingly, among thymocytes that evade deletion, similar proportions acquire Foxp3 expression. These findings provide evidence that Foxp3+ Treg cells can develop among members of a cohort of autoreactive thymocytes that have evaded deletion by a self-peptide, and that deletion and Treg cell formation can act together to bias the Treg cell repertoire toward low abundance self-peptide(s).

Keywords: Tolerance, Thymic selection, Transgenic models


During their development, thymocytes encounter an array of self-peptides that are expressed in different amounts and by distinct cell types. Self-peptides with which the TCR is strongly reactive can induce thymocyte deletion, which eliminates potentially autoagressive clones from the CD4+ and CD8+ T cell repertoires [1]. Thymocytes can also undergo a program of differentiation to become CD4+CD25+Foxp3+ Treg cells, which additionally participate in preventing autoimmunity [2, 3]. The processes that instruct thymocytes to undergo deletion or to develop along a pathway to become Treg cells remain poorly understood.

Evidence that interactions with self-peptides can promote Treg cell formation came from studies showing increased CD4+CD25+ Treg cell development in TCR Tg mice that co-express their agonist peptide as a self-peptide [4-6]. The idea that signaling from self-peptides can promote Treg cell formation is supported by studies showing that TCR signaling of progenitor CD4+CD8- (CD4SP) thymocytes can promote Foxp3+ expression [7]. However, as noted above, signaling from self-peptides can also promote thymocyte deletion, and the relationship between thymocyte deletion and Treg cell formation remains to be defined. One possibility is that differentiation along the Treg pathway is induced by cues other than recognition of self-peptides [8, 9]. For example, non-peptide-mediated cues might induce differentiation along a Treg pathway and cause a subset of thymocytes to be more resistant to deletion than are the remaining thymocytes, which are fated to develop into conventional CD4+Foxp3- T cells unless they are signaled by self-peptides to undergo deletion [8]. An alternative possibility is that the fate of an autoreactive thymocyte is dictated by the mode of self-peptide presentation e.g. expression at high or low densities, or by particular thymic cell types, although recent studies argued against a role for specialized thymic cell types in promoting Treg cell formation [10, 11].

We have examined these questions by determining the extent of thymocyte deletion and/or Foxp3+ Treg cell formation in Tg mice expressing a MHC class II-restricted TCR, and co-expressing the cognate peptide for this TCR under the control of a variety of different promoter/enhancer sequences. We found that Treg cell formation can occur among a subset of thymocytes that evade deletion by a self-peptide, and that the extent of this deletion is sensitive to the amount of self-peptide that is encountered. We propose that these processes are likely to produce a Treg cell repertoire that is focused toward low abundance self-peptides.

Results and discussion

Foxp3+ Treg cell formation accompanies thymocyte deletion in TS1× HA Tg mice

TS1 mice express a Tg TCR (detected with the mAb 6.5) that recognizes the major I-Ed-restricted CD4+ T cell determinant of PR8 HA (termed S1) as an agonist peptide [4, 12]. HA28, PevHA, βmyoHA, HA12 and HA104 mice are Tg mice that express the influenza virus PR8 hemagglutinin (HA) as a neo-self Ag under the control of the following promoters/enhancers: HA28, SV40 early region; PevHA, β-globin locus control region; βmyoHA, cardiac myosin heavy chain; HA12 and HA104, SV40 early region [13-15]. We mated TS1 mice with the different lineages of HA Tg mice and analyzed 6.5hiCD4SP thymocytes and 6.5hiCD4+ T cells for their frequency, and for the expression of Foxp3, which is closely associated with Treg cell function [16].

The frequencies of 6.5hiCD4SP thymocytes were decreased in all of the TS1× HA lineages relative to TS1 mice, indicating deletion of 6.5hiCD4SP thymocytes. However, the degree of 6.5hiCDSP thymocyte deletion varied in the different lineages; the most modest reductions were in TS1× HA28 and TS1× PevHA mice, which contained approximately one half and one third as many 6.5hiCD4SP thymocytes, respectively, as TS1 mice (Fig. 1A). By contrast, TS1× βmyoHA mice contained approximately one tenth as many 6.5hiCD4SP thymocytes, and TS1× HA12 and TS1× HA104 mice each contained less than one percent as many 6.5hiCD4SP thymocytes as were present in TS1 mice. Strikingly, quite similar proportions of the 6.5hiCD4SP thymocytes that evaded deletion in the different TS1× HA lineages acquired Foxp3 expression (Fig. 1A). Thus, despite these wide differences in the extent of deletion of 6.5hiCD4SP thymocytes, the percentages of 6.5hiCD4SP thymocytes that were Foxp3+ ranged from 20% to 35% in the different lineages, and in all cases greatly exceeded the 0.2% of 6.5hiCD4SP thymocytes that were Foxp3+ in TS1 mice (Fig. 1A). Moreover, all of the TS1× HA lineages contained significantly higher numbers of 6.5hiFoxp3+CD4SP thymocytes than were present in TS1 mice, with average increases ranging from 300% in TS1× HA12 mice to 7,000% in TS1× HA28 mice. The lymph nodes of TS1× HA28 and TS1× PevHA mice also contained significantly more 6.5hiFoxp3+CD25+CD4+ T cells than were present in TS1 mice, and the percentages of 6.5hiCD4+ lymph node cells that were also CD25+Foxp3+ were higher in all of the TS1× HA lineages than in TS1 mice (Fig. 1B). However, the absolute numbers of these cells did not differ significantly in TS1× βmyoHA, TS1× HA12 and TS1× HA104 mice relative to TS1 mice, in part because allelic inclusion of TCR alpha chains allowed for the expansion of Foxp3+CD25+CD4+ T cells co-expressing the 6.5 TCR along with non-6.5 TCR chains in TS1 mice [17].

Figure 1
Differing degrees of Treg cell formation in TS1× HA Tg mice. (A) Histograms show 6.5 expression by CD4+CD8- thymocytes (left panels) and Foxp3 expression by 6.5hiCD4+CD8- thymocytes (right panels) from TS1× HA (black lines) overlaying ...

Thus, thymocytes expressing the 6.5 TCR are subjected to varying degrees of deletion by the S1 peptide in the different TS1× HA lineages. However, among the 6.5hiCD4SP thymocytes that evade deletion in the each lineage, the proportion that acquires Foxp3 expression is quite similar (Fig. 1C). As a result of these two processes, the numbers of 6.5hiFoxp3+CD4SP thymocytes (and of 6.5hiCD25+Foxp3+CD4+ T cells) varied substantially in the different TS1× HA lineages. In all cases, however, significantly higher numbers of 6.5hiFoxp3+CD4SP thymocytes developed in TS1× HA mice than in TS1 mice, which do not co-express the S1 peptide.

Degree of thymocyte deletion correlates with amount of self-peptide

To examine how variations of the expression of the S1 peptide can affect 6.5 CD4SP thymocyte development in the different TS1× HA lineages we first examined the amount of HA transgene mRNA present in total thymic mRNA preparations from all of the different HA Tg lineages. HA28 and PevHA mice contained the lowest amounts of HA mRNA (Fig. 2), and the most modest deletion of 6.5hiCD4SP thymocytes (accompanied by abundant 6.5hiFoxp3+CD4SP thymocyte formation) occurred in TS1× HA28 and TS1× PevHA mice (Fig. 1A). Higher levels were found in βmyoHA mice, and the highest levels of HA mRNA were found in HA12 and HA104 mice (Fig. 2), corresponding with the increasingly severe deletion of 6.5hiCD4SP thymocytes observed in TS1× βmyoHA, TS1× HA12 and TS1× HA104 mice (Fig. 1A). Notably, then, the relative level of HA mRNA expression correlated well with the degree of deletion of 6.5hiCD4SP thymocytes, suggesting that the extent of thymocyte deletion is sensitive to the amount of self-peptide that is expressed in the thymus.

Figure 2
HA Tg mice express differing amounts of HA transgene mRNA. Levels of HA transgene mRNA in total thymic mRNA from indicated mouse lineages determined by real time PCR. Bars indicate means±SEM. Data are representative of three independent experiments ...

To examine whether expression of the HA in different cell types might contribute to 6.5hiCD4SP thymocyte deletion and/or 6.5hiFoxp3+CD4SP thymocyte formation we generated radiation bone marrow chimeras using HA28 and HA12 mice. These lineages contain the same transgene construct, but were derived from different founders [4, 18], and the distinct transgene integration sites in these lineages affect the expression of the S1 peptide such that it induces quite different degrees of 6.5hiCD4SP thymocyte deletion/6.5hiFoxp3+CD4SP thymocyte formation in TS1× HA28 versus TS1× HA12 mice (see Fig. 1A). The 6.5 TCR underwent minimal deletion and efficient development into 6.5hiFoxp3+CD4SP thymocytes in TS1→HA28 chimeras (as was observed in intact TS1× HA28 mice), while 6.5hiCD4SP thymocyte development appeared little affected in TS1× HA28→BALB/c mice (Fig. 3A and B). Interestingly, in TS1→HA12 mice, the 6.5 TCR was subjected to efficient deletion (accompanied by modest 6.5hiFoxp3+CD4SP thymocyte formation), indicating that the same self-peptide can induce either efficient thymocyte deletion or abundant Treg cell formation when it is synthesized by radioresistant thymic epithelial cells in HA12 versus HA28 mice (Fig. 3C). Moreover, increased formation of 6.5hiFoxp3+CD4SP thymocytes was observed in TS1× HA12→BALB/c mice, indicating that self-peptides synthesized by bone marrow-derived cells can also contribute to Foxp3+CD4SP thymocyte formation (Fig. 3D). Collectively, these data argue against a role for distinct thymic cell types in inducing 6.5hiCD4SP thymocyte deletion versus Foxp3+ Treg cell formation in TS1× HA Tg mice, and resemble recent findings indicating that multiple accessory cell types can contribute to the thymic generation of Foxp3+ Treg cells [10, 11]. Rather, higher levels of expression of the S1 self-peptide correlate with more extensive deletion of thymocytes expressing the 6.5 TCR, which is accompanied by Foxp3+ expression by a subset of the 6.5hiCD4SP thymocytes that evade deletion.

Figure 3
Diverse cell types can induce thymocyte deletion and/or Foxp3 expression. (A) Histograms show 6.5 expression by CD4+CD8- thymocytes (left panel) and Foxp3 expression by 6.5hiCD4+CD8- thymocytes (right panel) from TS1→HA28 bone marrow chimeras ...

Concluding remarks

The findings here are most readily accommodated into a model of Foxp3+ Treg cell formation in which interactions with self-peptides can induce autoreactive CD4SP thymocytes to undergo deletion with a varying efficiency that is sensitive to the amount of the self-peptide that is expressed. Among the autoreactive thymocytes that evade deletion, a subset acquires Foxp3 expression. A significant feature of this model is that thymocyte deletion and Treg cell formation do not reflect distinct outcomes of thymocyte development; we did not, for example, find that the S1-self peptide could induce deletion in the absence of Treg cell formation in some lineages, while in others there was minimal deletion but abundant Treg cell formation. Instead, we found that deletion and Treg cell formation behaved as intertwined processes that are sensitive to the relative abundance of a self-peptide, and have the effect of enhancing 6.5+Foxp3+ Treg cell formation in those mice in which HA is a relatively low abundance self-peptide. Thus, even though the proportions of 6.5hiCD4SP thymocytes that became Foxp3+ in the different TS1× HA lineages were relatively similar, fewer such cells evaded deletion in backgrounds (such as TS1× HA12 and TS1× HA104 mice) containing relatively high levels of HA mRNA. Substantially higher numbers of 6.5hiFoxp3+CD4SP thymocytes were formed in mice containing less transgene mRNA (such as TS1× HA28 and TS1× PevHA mice), where the deletion of 6.5hiCD4SP thymocytes was less complete.

This model of Treg cell formation is supported by observations in a variety of other systems. Thus, the notion that thymocytes expressing an MHC class II-restricted TCR can undergo varying degrees of deletion in response to differing levels of a self-peptide is consistent with other in vitro and in vivo studies [8, 18, 19]. Likewise, the conclusion that only a subset of the thymocytes that evade deletion acquire Foxp3 expression resembles studies in other systems showing Foxp3 expression in only a subset of susceptible thymocytes [7, 11], and may reflect heterogeneity in the expression of additional molecules and/or epigenetic modifications of target genes that are required for Foxp3 expression. The acquisition of Foxp3 predominantly among maturing CD4SP thymocytes in TS1× HA Tg mice is consistent with the timing of Foxp3 expression in non-Tg mice [20], but in the studies here it is possible to further conclude that cells that are acquiring Foxp3 expression belong to a cohort that has evaded deletion by a self-peptide. Finally, the conclusion that Treg differentiation can accompany deletion of autoreactive thymocytes resembles findings obtained using BDC2.5 Tg mice, where increased numbers of Foxp3+CD4SP thymocytes were found to develop in fetal thymic organ cultures in response to ranges of agonist stimulation where thymocyte deletion also occurred [21].

The findings here are, by contrast, inconsistent with an alternative model of Foxp3+ thymocyte development that posits that non-TCR-mediated cues induce a subset of developing thymocytes to become Foxp3+, which are then selectively protected from deletion [8]. It is notable that this conclusion was based on analyses using AND TCR transgenic mice, which contain a much higher frequency of CD4SP thymocytes expressing the transgene-encoded clonotypic TCR than is the case for the TS1 mice used here. A recent study showed that high frequencies of thymocytes expressing a clonotypic TCR can impair Treg cell formation through intraclonal competition [22], and the failure to observe self-peptide-induced Treg cell formation in the AND system could therefore be a result of the quasi-monoclonal TCR repertoire in AND mice. In this respect, then, the more efficient formation of Treg cells in TS1× HA mice could be in part due to the relatively poor allelic exclusion that is imposed by the 6.5 TCR transgene, which could have the effect of decreasing the degree of intraclonal competition for self-peptides. There is however an important distinction between the studies here and those demonstrating the effects of clonal competition of Treg cell formation; in TS1× HA Tg mice, 6.5hiFoxp3+CD4SP thymocyte formation occurred among thymocytes that had evaded deletion by the S1 self-peptide, while the studies by Bautista et al found no evidence for clonal deletion under conditions in which Foxp3+CD4SP thymocyte formation could occur [22]. Additional studies in the TS1× HA Tg system may allow us to more fully understand the effects of clonal competition on Treg cell formation and/or thymocyte deletion. At this stage, however, the increased numbers of 6.5hiFoxp3+CD4SP thymocytes that are formed in TS1× HA Tg mice clearly show that Foxp3+CD4SP thymocyte formation can be a consequence of peptide-induced formation, rather than selective survival of thymocytes that had acquired Foxp3 expression in response to other cues.

The findings in TS1× HA Tg mice also provide evidence that the Treg cell repertoire that is generated in response to bona fide self-peptides may be biased toward low abundance self-peptides. Thus, some low abundance self-peptides may behave like the S1 peptide in HA28 and PevHA mice and generate relatively large numbers of Treg cells, while other more abundant peptides may induce more severe deletion and in turn generate fewer Treg cells. A focusing of the Treg cell repertoire toward low abundance self-peptides may explain a failure to detect Treg cell activity against self-peptides [23], since the target self-peptides recognized by Treg cells might be too rare to induce detectable activation of Treg hybridomas in in vitro culture systems. It also suggests that the Treg cell repertoires that are expressed by outbred individuals may be quite idiosyncratic, since their composition would be imposed by low abundance peptides that may show greater variation between individuals than is the case for highly abundant self-peptides.

Materials and Methods


TS1, HA28, PevHA, βmyoHA, HA12 and HA104 mice have been described previously [4, 12, 15, 17, 18]. Radiation bone marrow chimeras were generated by injecting 5×106 T cell-depleted BM cells i.v. into 6 to 8 week-old irradiated (900 rad) mice and analyzed 8 weeks later. All mice were maintained in sterile microisolaters in the Association for the Assessment and Accreditation of Laboratory Animal Care International-accredited animal facility at The Wistar Institue.

Flow cytometry

Anti-clonotypic antibody 6.5 [12] was biotinylated and detected with streptavidin-allophycocyanin, and remaining antibodies were obtained from BD Biosciences or eBioscience. Intracellular staining for Foxp3 was performed according to the eBioscience protocol. Flow cytometry was performed using a FACSCalibur (BD Biosciences) and analyzed using FlowJo software (Tree Star).

Real-time PCR analysis

Total thymic RNA was isolated using Trizol LS (Invitrogen) and RNeasy columns (Qiagen) and reverse transcribed to generate cDNA. HA transgene sequences were quantitated relative to HPRT using Taqman and the Assays-by-Design kit FLUPR8HA-TC4 (forward primer: AGGCAAATGGAAATCTAATAGCACCAA; reverse primer: GATGCCGGACCCAAAGC; probe: CTCAGTGCGAAAGCAT) (Applied Biosystems) to detect HA, and the Assays-on-Demand kit Mm00446968_m1 (Applied Biosystems) to detect HPRT. Samples were analyzed using a PRISM 7700 sequence detection system (Applied Biosystems).


We thank Dr. A. Bhandoola for his thoughtful reading of this manuscript. Supported by grant AI59166 from the National Institutes of Health, from the Lupus Foundation of Southeastern Pennsylvania, and by the Commonwealth Universal Research Enhancement Program, Pennsylvania Department of Health.




1. Starr TK, Jameson SC, Hogquist KA. Positive and negative selection of T cells. Annu Rev Immunol. 2003;21:139–176. [PubMed]
2. Sakaguchi S. Naturally arising CD4+ regulatory T cells for immunologic self-tolerance and negative control of immune responses. Annu Rev Immunol. 2004;22:531–562. [PubMed]
3. Fontenot JD, Rudensky AY. A well adapted regulatory contrivance: regulatory T cell development and the forkhead family transcription factor Foxp3. Nat Immunol. 2005;6:331–337. [PubMed]
4. Jordan MS, Boesteanu A, Reed AJ, Petrone AL, Holenbeck AE, Lerman MA, Naji A, Caton AJ. Thymic selection of CD4+CD25+ regulatory T cells induced by an agonist self-peptide. Nat Immunol. 2001;2:301–306. [PubMed]
5. Apostolou I, Sarukhan A, Klein L, von Boehmer H. Origin of regulatory T cells with known specificity for antigen. Nat Immunol. 2002;3:756–763. [PubMed]
6. Walker LS, Chodos A, Eggena M, Dooms H, Abbas AK. Antigen-dependent proliferation of CD4+ CD25+ regulatory T cells in vivo. J Exp Med. 2003;198:249–258. [PMC free article] [PubMed]
7. Lio CWJ, Hsieh CS. A two-step process for thymic regulatory T cell development. Immunity. 2008;28:100–111. [PMC free article] [PubMed]
8. van Santen HM, Benoist C, Mathis D. Number of Treg cells that differentiate does not increase upon encounter of agonist ligand on thymic epithelial cells. J Exp Med. 2004;200:1221–1230. [PMC free article] [PubMed]
9. Pennington DJ, Silva-Santos B, Silberzahn T, Escorcio-Correia M, Woodward MJ, Roberts SJ, Smith AL, Dyson PJ, Hayday AC. Early events in the thymus affect the balance of effector and regulatory T cells. Nature. 2006;444:1073–1077. [PubMed]
10. Liston A, Nutsch KM, Farr AG, Lund JM, Rasmussen JP, Koni PA, Rudensky AY. Differentiation of regulatory Foxp3+ T cells in the thymic cortex. Proc Natl Acad Sci USA. 2008;105:11903–11908. [PubMed]
11. Wirnsberger G, Mair F, Klein L. Regulatory T cell differentiation of thymocytes does not require a dedicated antigen-presenting cell but is under T cell-intrinsic developmental control. Proc Natl Acad Sci USA. 2009 doi: 10.1073/pnas.0901877106. published online before print June 10 2009. [PubMed] [Cross Ref]
12. Kirberg J, Baron A, Jakob S, Rolink A, Karjalainen K, von Boehmer H. Thymic selection of CD8+ single positive cells with a class II major histocompatibility complex-restricted receptor. J Exp Med. 1994;180:25–34. [PMC free article] [PubMed]
13. Cerasoli DM, McGrath J, Carding SR, Shih FF, Knowles BB, Caton AJ. Low avidity recognition of a class II-restricted neo-self peptide by virus-specific T cells. Int Immunol. 1995;7:935–945. [PubMed]
14. Shih FF, Cerasoli DM, Caton AJ. A major T cell determinant from the influenza virus hemagglutinin (HA) can be a cryptic self peptide in HA transgenic mice. Int Immunol. 1997;9:249–261. [PubMed]
15. Lerman MA, Larkin J, 3rd, Cozzo C, Jordan MS, Caton AJ. CD4+ CD25+ regulatory T cell repertoire formation in response to varying expression of a neo-self-antigen. J Immunol. 2004;173:236–244. [PubMed]
16. Zheng Y, Rudensky AY. Foxp3 in control of the regulatory T cell lineage. Nat Immunol. 2007;8:457–462. [PubMed]
17. Larkin J, III, Rankin AL, Picca CC, Riley MP, Jenks SA, Sant AJ, Caton AJ. CD4+CD25+ regulatory T cell repertoire formation shaped by differential presentation of peptides from a self-antigen. J Immunol. 2008;180:2149–2157. [PubMed]
18. Riley MP, Cerasoli DM, Jordan MS, Petrone AL, Shih FF, Caton AJ. Graded deletion and virus-induced activation of autoreactive CD4(+) T cells. J Immunol. 2000;165:4870–4876. [PubMed]
19. Spain LM, Berg LJ. Developmental regulation of thymocyte susceptibility to deletion by “self”-peptide. J Exp Med. 1992;176:213–223. [PMC free article] [PubMed]
20. Fontenot JD, Rasmussen JP, Williams LM, Dooley JL, Farr AG, Rudensky AY. Regulatory T cell lineage specification by the forkhead transcription factor foxp3. Immunity. 2005;22:329–341. [PubMed]
21. Feuerer M, Jiang W, Holler PD, Satpathy A, Campbell C, Bogue M, Mathis D, Benoist C. Enhanced thymic selection of FoxP3+ regulatory T cells in the NOD mouse model of autoimmune diabetes. Proc Natl Acad Sci USA. 2007;104:18181–18186. [PubMed]
22. Bautista JL, Lio CWJ, Lathrop SK, Forbush K, Liang Y, Luo J, Rudensky AY, Hsieh CS. Intraclonal competition limits the fate determination of regulatory T cells in the thymus. Nat Immunol. 2009;10:610–617. [PMC free article] [PubMed]
23. Pacholczyk R, Kern J, Singh N, Iwashima M, Kraj P, Ignatowicz L. Nonself-antigens are the cognate specificities of Foxp3+ regulatory T cells. Immunity. 2007;27:493–504. [PMC free article] [PubMed]