Cathepsin L-dependent positive selection shapes clonal composition and functional fitness of CD4+ T cells

Petrozziello, Elisabetta; Sayed, Amina; Freitas, João A.; Federle, Christine; Nedjic, Jelena; Ravens, Sarina; Akçabozan, Batuhan; Schulz, Anna M.; Zehn, Dietmar; Schmidt-Supprian, Marc; Obst, Reinhard; Prinz, Immo; Verdoes, Martijn; Kisielow, Jan; Reinheckel, Thomas; Straub, Tobias; Daley, Stephen R.; Klein, Ludger

doi:10.1038/s41590-025-02182-y

Download PDF

Article
Open access
Published: 13 June 2025

Cathepsin L-dependent positive selection shapes clonal composition and functional fitness of CD4⁺ T cells

Nature Immunology volume 26, pages 1127–1138 (2025) Cite this article

9557 Accesses
7 Citations
9 Altmetric
Metrics details

Subjects

Abstract

The physiological significance of thymic positive selection and its reliance on a single stromal cell type, cortical thymic epithelial cells, remain incompletely understood. The lysosomal cysteine protease cathepsin L (CTSL) has been implicated in generating major histocompatibility complex class II-bound peptides in cortical thymic epithelial cells for efficient CD4⁺ T cell differentiation. Here, we addressed the extent and nature of the CD4⁺ T cell repertoire changes associated with CTSL deficiency. In the absence of CTSL, a highly selective loss of T cell receptors resulted in a markedly reduced repertoire diversity. However, a similarly large proportion of nominally ‘CTSL-independent’ T cell receptors were retained. Clones representative of the second category experienced weaker positive selection signals in the absence of CTSL, which were sufficient for further maturation yet imprinted aberrant responsiveness to agonist stimulation and impaired homeostatic behavior. Together, these findings demonstrate that CTSL is crucial for both shaping full repertoire diversity and optimizing CD4⁺ T cell functionality.

CD4 expression in effector T cells depends on DNA demethylation over a developmentally established stimulus-responsive element

Article Open access 18 March 2022

Modulation of TCR signalling components occurs prior to positive selection and lineage commitment in iNKT cells

Article Open access 08 December 2021

Integrin CD103 reveals a distinct developmental pathway of autoreactive thymocytes in TCR transgenic mice

Article Open access 18 July 2025

Main

During thymic selection, thymocytes test their T cell receptor (TCR) on self-peptide major histocompatibility complex (pMHC) ligands presented by thymic antigen-presenting cells (APCs). ‘Weak’ TCR–pMHC interactions promote developmental progression and CD4⁺/CD8⁺ T cell lineage commitment (termed positive selection), whereas ‘strong’ signals trigger negative selection¹. Although negative selection eliminates autoreactive T cells², how positive selection shapes a ‘useful’ repertoire remains unclear³.

Traditionally, positive selection was thought to enforce self-MHC restriction. However, T cells recognizing specific pMHC ligands arise at similar frequencies regardless of the selecting MHC allele⁴, and the repertoire’s apparent self-MHC restriction may indirectly result from negative selection of overly MHC-reactive TCRs⁵. Increasing evidence also indicates that positive selection imprints T cell functionality, such as tuning of the inhibitory TCR rheostat CD5 (ref. ⁶). CD5 levels are thought to reflect the strength of the selecting TCR–pMHC interactions and have been linked to responsiveness to foreign antigens^3,7,8, contribution to primary versus memory responses⁹ and differentiation potential into helper T (T_H) cell subsets or regulatory T (T_reg) cells^{10,11,12,13,14}.

Positive selection relies on a single stromal cell type, cortical thymic epithelial cells (cTECs). This specialized role appears to stem from unique pathways of self-antigen handling and processing, likely generating a partially ‘private’ pMHC ligandome^1,15. For MHC class I (MHCI), cTECs express ‘thymoproteasomes’ containing the β5t subunit, which is absent from other APCs. β5t deficiency profoundly affects thymic development of CD8⁺ T cells¹⁵. For MHC class II (MHCII), cTECs use autophagy-associated mechanisms for unconventional endogenous MHCII loading. Disruption of these pathways perturbs CD4⁺ T cell selection^16,17,18. Mice lacking the thymus-specific serine protease PRSS16 exhibit impaired polyclonal CD4⁺ T cell responses and diminished positive selection of some transgenic TCRs^19,20. Of the pathways implicated in shaping the cTEC pMHC class II (pMHCII) ligandome, the most profound reduction in the thymic CD4⁺ T cell population is caused by ablation of cathepsin L (CTSL)²¹. Cathepsins are a family of lysosomal proteases. CTSL is strongly expressed in cTECs but is barely present in other MHCII⁺ APCs, whereas cathepsin S exhibits the opposite expression pattern. Both enzymes serve dual functions in the MHCII pathway by degrading the MHCII-associated invariant chain (Ii) and processing antigens to generate peptides for MHCII loading²². Importantly, the Ii-degrading function of CTSL alone cannot explain its requirement for efficient CD4⁺ T cell selection²³. Limited resolution and cell numbers have so far precluded a comprehensive characterization of the cTEC pMHCII ligandome beyond relatively abundant peptides²⁴. However, analysis of pMHCII ligands in fibroblasts engineered to express CTSL or cathepsin S indicated qualitative and quantitative differences²⁵, suggesting that CTSL-dependent ‘private’ pMHCII ligands on cTECs may be crucial for CD4⁺ T cell positive selection. However, the impact of CTSL on the diversity and functionality of the CD4⁺ T cell repertoire remains unclear.

Here, we analyzed the CD4⁺ T cell repertoire selected in the absence of CTSL and characterized the clonal composition and reactivity of residual CD4⁺ T cells specific for a prototypical foreign antigen. Using high-resolution repertoire sequencing and re-expression of selected clones in TCR transgenic mice, we found that CTSL shapes CD4⁺ T cell selection by promoting full repertoire diversity and fine-tuning CD4⁺ T cell functionality.

Results

CTSL deficiency impairs positive selection of CD4⁺ T cells

Full genomic deletion of Ctsl in Ctsl^−/− mice has pleiotropic effects, most prominently alopecia and epithelial hyperplasia in the skin²¹, reflecting ‘nonimmune’ functions. To selectively delete Ctsl in TECs, we generated Ctsl^ΔTEC mice, which carry a conditional Ctsl allele and a Foxn1-cre transgene (Supplementary Fig. 1a,b). Compared to Ctsl^+/+ mice, Ctsl^ΔTEC mice showed a reduction in CD4 single-positive (CD4SP) thymocytes, as described in Ctsl^−/− mice²¹ (Fig. 1a,b and Extended Data Fig. 1a). The phenotypic segregation of CD4SP thymocytes into three consecutive maturation stages (CD69⁺MHCI^– semimature, CD69⁺MHCI⁺ mature 1 and CD69⁻MHCI⁺ mature 2; SM, M1 and M2, respectively)²⁶ was largely preserved, although there was a reduction in the most mature M2 stage (Extended Data Fig. 1b). The proportion of Foxp3⁺ T_reg cells among CD4SP thymocytes was unchanged; however, there was a trend toward an increased proportion of CD73⁺CCR7⁻ reimmigrants from the periphery (Extended Data Fig. 1c,d). Peripheral CD4⁺ T cell populations were diminished and contained more ‘memory-like’ Foxp3⁻CD44⁺CD62L⁻ cells and Foxp3⁺ T_reg cells (Fig. 1c and Extended Data Fig. 1e–g). CD8⁺ T cells developed in normal proportions, and the thymic architecture was indistinguishable from Ctsl^+/+ mice (Fig. 1a,b and Supplementary Fig. 1e). Ctsl^ΔTEC mice did not show skin defects (Supplementary Fig. 1c,d).

Fig. 1: CTSL deficiency alters the cTEC pMHCII ligandome and impairs CD4+ T cell positive selection. — **Fig. 1: CTSL deficiency alters the cTEC pMHCII ligandome and impairs CD4⁺ T cell positive selection.**

Total MHCII on cTECs remained unchanged (Fig. 1d). However, the fraction of pMHCII ligands on Ctsl^ΔTEC cTECs that consisted of complexes with the invariant chain-derived peptide CLIP, as detected using the monoclonal antibody (mAb) 15G4, was increased, and, conversely, non-CLIP pMHCII ligands, as indicated by the mAb BP107, were reduced (Fig. 1e,f). Nevertheless, compared to H2-Ab1^−/− (hereafter MHCII^−/−) cTECs, non-CLIP pMHCII ligands still constituted a major fraction of the pMHCII ligandome in Ctsl^ΔTEC cTECs (Fig. 1f). For instance, cTECs from Ctsl^ΔTEC C57BL/6 × BALB/c F1 mice presented substantial amounts of the ‘frequent’ non-CLIP ligand I-A^b:Eα_52–68, recognized by the mAb Y-Ae, albeit moderately less than Ctsl^+/+ control mice (Fig. 1g). These effects were seen in cTECs but not medullary TECs (mTECs; Fig. 1e–g and Extended Data Fig. 1h,i), consistent with the differential expression of CTSL between cTECs and mTECs.

The diminished CD4SP thymocyte population in Ctsl^−/− mice has been suggested to result, at least in part, from positive selection of an altered TCR repertoire that is hypersusceptible to negative selection²³. To test whether interference with negative selection by hematopoietic APCs ‘rescued’ the CD4SP compartment, we reconstituted Ctsl^+/+, Ctsl^ΔTEC or Ctsl^−/− mice with H2-Ab1^+/+ (hereafter MHCII^+/+) or MHCII^−/− bone marrow (BM). MHCII^−/− → Ctsl^+/+ chimeras harbored a significantly increased percentage of CD4SP thymocytes compared to MHCII^+/+ → Ctsl^+/+ controls (Fig. 1h), reflecting diminished negative selection²⁷. By contrast, the CD4SP compartment of MHCII^−/− → Ctsl^ΔTEC or MHCII^−/− → Ctsl^−/− chimeras was not, or only marginally, increased compared to the respective MHCII^+/+ BM controls (Fig. 1h and Extended Data Fig. 1j). To interfere with negative selection by mTECs, we generated Ctsl^+/+, Ctsl^ΔTEC and Ctsl^−/− mice carrying a transgene (Ciita^kd) that mediates tissue-specific knockdown of C2TA, a transcription factor that controls multiple MHCII pathway components, leading to reduced MHCII expression on mTECs²⁸. The Ciita^kd transgene resulted in a significantly increased CD4SP compartment in Ctsl^+/+ mice²⁸ but did not ‘rescue’ the CD4SP compartment in Ctsl^ΔTEC or Ctsl^−/− mice (Fig. 1i and Extended Data Fig. 1k).

The proportion of TCRβ⁺CD69⁺ cells among CD4⁺CD8⁺ (double-positive; DP) thymocytes (representing signal-selection intermediates) was reported to be normal in Ctsl^−/− mice²³; however, TCRβ⁺CD69⁺ DP cells also include CD8⁺ T cell lineage selection intermediates that engage pMHCI ligands. To exclude these, we generated MHCI-deficient B2m^−/−Ctsl^ΔTEC mice, which recapitulated the reduced CD4SP compartment associated with CTSL deficiency (Extended Data Fig. 1l). In these mice, where positively selecting interactions could be exclusively attributed to pMHCII ligands, the proportion of ‘signaled’ TCRβ⁺CD69⁺ DP cells was reduced to about half that of B2m^−/−Ctsl^+/+ controls (Fig. 1j). CD5 expression on these DP cells (Extended Data Fig. 1m) and bulk Ctsl^ΔTEC CD4SP cells was lower than on their counterparts in Ctsl^+/+ mice (Fig. 1k), suggesting that positive selection occurred through relatively weak interactions or nonselection of ‘natural’ CD5^hi clones. Together, these findings indicate that the contraction of the CD4⁺ T cell compartment in CTSL-deficient mice was not secondary to selection of an altered repertoire that was overly susceptible to negative selection but most likely reflected a bona fide numerical constraint in positive selection as a consequence of an altered cTEC pMHCII ligandome.

CTSL deficiency causes ‘clonal holes’ and ‘newcomers’

We next addressed whether the bottleneck in positive selection in Ctsl^ΔTEC mice was TCR selective, leading to the disappearance of some clones while allowing others to persist within the repertoire. Across seven transgenic MHCII-restricted TCRs with diverse antigen specificities (OT-II (chicken ovalbumin), Dep (human C-reactive protein), AND and AD10 (pigeon cytochrome c), LLO56 and LLO118 (Listeria monocytogenes listeriolysin O; LLO) and PLP1 (myelin proteolipid protein)), all exhibited a profound blockade in the emergence of CD4SP thymocytes in Ctsl^ΔTEC mice, whereas the MHCI-restricted OT-I TCR was efficiently selected (Fig. 2a,b and Extended Data Fig. 2a–e). For each MHCII-restricted TCR transgene, CD5 expression on DP thymocytes, which is upregulated concomitant with positive selection, was substantially reduced in Ctsl^ΔTEC mice (Fig. 2c and Extended Data Fig. 2b), suggesting that these TCRs did not, or did not sufficiently, interact with pMHCII ligands to elicit positive selection. When normally selected in Ctsl^+/+ mice, OT-II, AND, Dep or AD10 CD4SP thymocytes each displayed distinct CD5 levels, which varied widely between these clones and spanned the entire range of CD5 expression observed in polyclonal CD4SP cells (Fig. 2d). Thus, nonselection in Ctsl^ΔTEC mice was not tied to specific CD5 characteristics and, by inference, was not confined to a particular window in the affinity range of positively selecting TCR–pMHC interactions.

**Fig. 2: CTSL is essential for positive selection of multiple MHCII-restricted transgenic TCRs.**

To globally characterize TCR repertoire perturbations caused by CTSL deficiency, we crossed Ctsl^ΔTEC mice with mice expressing a transgenic TCRβ chain (hereafter Tcrb^Fixed), enabling high-throughput sequencing of variable TCRα chains paired with the ‘fixed’ β-chain. Tcrb^FixedCtsl^ΔTEC mice had fewer CD4SP thymocytes and peripheral CD4⁺ T cells (Fig. 3a and Extended Data Fig. 3a–c), with reduced CD5 expression on CD4SP cells (Fig. 3b) compared to Tcrb^FixedCtsl^+/+ mice. Tcra sequencing was performed on CD4SP cells in the most mature M2 stage²⁶, ensuring that the repertoire was fully shaped by thymic selection. Sampling depth approached saturation for both genotypes (Fig. 3c). Repertoire diversity, as assessed using the Shannon index, was significantly lower in Tcrb^FixedCtsl^ΔTEC mice than in Tcrb^FixedCtsl^+/+ mice (Fig. 3d). Based on the Morisita–Horn index, repertoires were highly stereotypical between genotype-matched replicates but varied significantly between genotypes (Fig. 3e). Of 9,626 ‘recurrent’ TCRs (defined as TCRs found in three or more of all six samples (three Tcrb^FixedCtsl^ΔTEC and three Tcrb^FixedCtsl^+/+)), 4,765 were shared between the two genotypes (Fig. 3f). Almost half (3,848 of 8,613) of all recurrent TCRs in the Tcrb^FixedCtsl^+/+ repertoire were entirely absent from the Tcrb^FixedCtsl^ΔTEC repertoire (Fig. 3f), and these ‘CTSL-dependent’ clones disproportionately contributed to the diversity of the ‘normal’ repertoire (Fig. 3g). Conversely, about 20% (1,013 of 5,778) of clones in the Tcrb^FixedCtsl^ΔTEC repertoire were not found in the ‘normal’ Tcrb^FixedCtsl^+/+ repertoire (Fig. 3f). These ‘newcomer’ TCRs displayed a bias toward more distal TCRα variable (V) and joining (J) elements and increased nucleotide additions or deletions at the V–J joint (Extended Data Fig. 3d–f), suggesting a selection bias for unusual TCR features. Thus, the loss of TCRs in the absence of CTSL was highly selective, affecting roughly half of the ‘normal’ TCR repertoire, whereas a similarly large array of seemingly CTSL-independent TCRs was retained.

**Fig. 3: CTSL deficiency results in nonselection of ~50% of TCR clonotypes and enables emergence of ‘newcomer’ TCRs.**

CTSL deficiency blunts CD4⁺ T cell responses

Having identified ‘clonal holes’ with various TCR transgenes and in the Tcrb^Fixed repertoire, we asked whether CTSL deficiency caused corresponding ‘antigenic holes’ in a fully polyclonal setting. MHCII tetramer (Tet) staining revealed a marked reduction (or near absence) of cells recognizing the epitopes 2W, human invariant chain residues 277–285 (huCLIP) and PLP_11–19 in Ctsl^ΔTEC mice (Fig. 4a,b). However, numbers of LLO_190–201:I-A^b-specific CD4⁺ T cells were comparable between Ctsl^ΔTEC and Ctsl^+/+ mice, both among peripheral CD4⁺ T cells and thymic CD4SP cells (Fig. 4c and Extended Data Fig. 4). Given the approximately fourfold reduction in total CD4⁺ T cell counts in Ctsl^ΔTEC mice (Extended Data Fig. 1e), the frequency of LLO-specific cells was thus even increased by a corresponding factor. To assess the functionality of these cells, we immunized mice with LLO peptide in adjuvant. At day 7 after challenge, ~1 × 10⁴ LLO-Tet⁺ T cells were detectable in the draining lymph nodes of Ctsl^+/+ mice (Fig. 4d), reflecting a >20-fold expansion following antigen exposure. By contrast, despite higher precursor frequencies, LLO-Tet⁺ cells were approximately tenfold less abundant in Ctsl^ΔTEC mice (Fig. 4d).

Fig. 4: CTSL deficiency imparts ‘antigenic holes’ in the repertoire and blunts CD4+ T cell responsiveness to immunization. — **Fig. 4: CTSL deficiency imparts ‘antigenic holes’ in the repertoire and blunts CD4⁺ T cell responsiveness to immunization.**

We performed adoptive co-transfer experiments to determine whether the diminished expansion of LLO-Tet⁺ cells in Ctsl^ΔTEC mice reflected a cell-intrinsic defect or an indirect consequence of lymphopenia. Bulk M2 thymocytes from CD45.2 Ctsl^ΔTEC and CD45.1 Ctsl^+/+ donors were co-transferred at a 1:1 ratio (thereby establishing a ~4:1 ratio of Ctsl^ΔTEC to Ctsl^+/+ LLO-Tet⁺ cells in the input population) into CD45.1/CD45.2 Ctsl^+/+ recipients that had been intravenously (i.v.) immunized with LLO peptide plus polyinosinic–polycytidylic acid (poly(I:C)) 6 h before. At day 7 after challenge, Ctsl^+/+CD45.1⁺ donor-derived cells accounted for 2.6 ± 0.3% of splenic LLO-Tet⁺ cells, whereas Ctsl^ΔTECCD45.2⁺ donor-derived cells contributed only marginally (0.006 ± 0.004%; Fig. 4e). This corresponded to a ratio of Ctsl^ΔTEC to Ctsl^+/+ LLO-Tet⁺ cells of ~1:400 (Fig. 4f), indicating a marked competitive disadvantage in antigen-driven expansion and/or defective homeostatic maintenance of cells selected in Ctsl^ΔTEC donors. Thus, positive selection in Ctsl^ΔTEC mice not only created ‘antigenic gaps’ but also resulted in impaired expansion and/or persistence of retained cells following antigen encounter.

Nonselection affects clones across the CD5 range

Polyclonal ‘natural’ CD5^hi clones have been reported to respond more robustly to immunization than CD5^lo clones⁸, suggesting a direct link between the modalities of positive selection and responsiveness to foreign antigens^3,8. To assess whether the nonselection of CTSL-dependent clones in the Ctsl^ΔTEC thymus correlated with their CD5 level, we generated TCR inventories from sorted Tcrb^FixedCtsl^+/+ CD4SP cells at both extremes of the CD5 spectrum (Fig. 5a). The TCR compositions were remarkably stereotypic within the four replicates of CD5^lo or CD5^hi CD4SP cells, respectively, yet highly distinct between the two groups, as evidenced by Morisita–Horn comparisons (Fig. 5b). This supports the notion that partitioning of a given clone into the CD5^lo or CD5^hi subset of the CD4⁺ T cell compartment is not stochastic but specified by TCR identity. We classified TCRs found in three or more of four CD5^lo samples and absent from all four CD5^hi datasets as ‘natural CD5^lo TCRs’ and those exhibiting a reciprocal pattern as ‘natural CD5^hi TCRs’. Cross-comparison with the recurrent TCRs in our previously established Tcrb^FixedCtsl^+/+ versus Tcrb^FixedCtsl^ΔTEC datasets revealed that 70% of the natural CD5^lo TCR clones and 43.5% of the natural CD5^hi TCR clones were not selected in the absence of CTSL compared to a loss of 44.7% across all TCRs (Fig. 5c). Thus, at the global repertoire level, nonselection in the absence of CTSL was more pronounced among the natural CD5^lo subrepertoire, yet affected clones across the entire spectrum of natural CD5 expression.

**Fig. 5: CTSL deficiency impedes selection of clones at both extremes of the CD5 spectrum but retains a substantial proportion of nominal ‘good-responder’ TCRs.**

We next assessed whether the diminished LLO-specific response in Ctsl^ΔTEC mice reflected a CTSL dependency of ‘good-responder’ TCR clones within the ‘normal’ repertoire. To this end, we i.v. immunized Tcrb^FixedCtsl^+/+ mice with LLO and performed Tcra sequencing on the expanded LLO-Tet⁺ population at day 7 after challenge, identifying the top ten expanded clonotypes (Fig. 5d). These clonotypes were cross-referenced with our previously established inventories of CTSL-dependent and CTSL-independent TCRs in the naive repertoire of Tcrb^FixedCtsl^+/+ mice, revealing that five TCRs could be assigned to the CTSL-dependent category and five to the CTSL-independent category (Fig. 5d). Thus, a significant proportion of ‘good-responder’ TCRs were retained in the Ctsl^ΔTEC repertoire, suggesting that the blunted response of LLO-Tet⁺ cells in Ctsl^ΔTEC mice could not be explained solely by the physical absence of all such clones.

CTSL specifies selection signals in CTSL-independent clones

To assess whether and which LLO-Tet⁺ clonotypes were retained in the absence of CTSL in a fully TCRαβ polyclonal setting, we performed single-cell TCR sequencing of LLO-Tet⁺ cells sorted from Ctsl^+/+ and Ctsl^ΔTEC mice, yielding 316 and 156 paired TCRαβ clonotypes, respectively (Fig. 6a). Most of these were detected only once within each genotype; however, five TCRs were shared between Ctsl^+/+ and Ctsl^ΔTEC mice (Fig. 6a). To explore the characteristics of such CTSL-independent CD4⁺ T cell clones, we generated two TCR transgenic mouse lines, hereafter Lm54^Tg and Lm6^Tg. Lm54^Tg or Lm6^Tg mice gave rise to CD4SP cells on both the Rag1^−/−Ctsl^+/+ and Rag1^−/−Ctsl^ΔTEC backgrounds (Fig. 6b,c). However, CD4SP cells were reduced for both TCRs on the Ctsl^ΔTEC background (markedly for Lm6 cells and subtly for Lm54 cells; Fig. 6b,c and Extended Data Fig. 5a,b), with a corresponding reduction in peripheral CD4⁺ T cell numbers (Fig. 6d,e), indicating a graded impairment in positive selection in the absence of CTSL, even for these apparently CTSL-independent TCRs.

**Fig. 6: LLO-specific TCRs retained in the absence of CTSL indicate a role for CTSL in calibrating positive selection signal strength.**

Although CD4SP cells were reduced in Rag1^−/−Ctsl^ΔTECLm54^Tg mice compared to Rag1^−/−Ctsl^+/+Lm54^Tg mice (hereafter Ctsl^ΔTECLm54^Tg and Ctsl^+/+Lm54^Tg, respectively), their segregation into the SM, M1 and M2 subsets was virtually identical (Fig. 6f). To determine whether selection of Lm54 thymocytes in the presence or absence of CTSL affected the underlying TCR signals (despite seemingly identical developmental progression downstream of the positive selection checkpoint), we assessed the expression of the nuclear receptor Nur77, whose levels reflect ongoing or recent TCR signaling²⁹. Although Lm54 thymocytes in Ctsl^+/+ mice showed typical dynamic Nur77 modulation, with upregulation in the signaled DP cells and return to baseline in M2 CD4SP cells, Nur77 expression was markedly attenuated in cells selected in Ctsl^ΔTEC mice (Fig. 6g), consistent with weaker selecting TCR–pMHC interactions in the absence of CTSL. In line with this, both Lm54 and Lm6 thymocytes selected in Ctsl^ΔTEC mice showed markedly lower CD5 upregulation than their counterparts selected in Ctsl^+/+ mice (Fig. 6g,h), and this pattern was also observed for CD6, whose expression likewise correlates with TCR signal strength³⁰ (Extended Data Fig. 5c). Thus, even for TCR clones that appeared CTSL independent in their repertoire seeding, CTSL was still crucial for calibrating the intensity of the positive selection signal.

Functional tuning by CTSL regulates homeostatic fitness

We next compared gene expression profiles in M2 CD4SP cells from Ctsl^ΔTECLm54^Tg and Ctsl^+/+Lm54^Tg mice. Genes more highly expressed in Ctsl^ΔTEC cells included the genes encoding the two subunits of co-receptor CD8, as well as ion channels or solute carriers such as TMIE and SLC16A5 (Fig. 7a), which typically peak in DP cells, whereas Ccr8, which is upregulated in CD4SP cells³¹, was reduced, consistent with ‘less mature’ traits in thymocytes selected in the absence of CTSL. Gene set enrichment analysis (GSEA) further revealed underrepresented transcripts in Ctsl^ΔTEC cells linked to translation, mTORC1 signaling and Myc targets (Fig. 7b), and these correlated with reduced cell size and diminished upregulation of CD44, a marker of mTOR signaling^32,33 (Extended Data Fig. 6a,b), suggesting reduced metabolic activity. Consistent with this, ex vivo assessment of basal protein biosynthesis showed diminished translation in the absence of CTSL, first emerging in CD69⁺ signaled DP cells and persisting through the M2 CD4SP cell stage (Fig. 7c).

Fig. 7: CTSL deficiency causes aberrant functional tuning and impaired homeostatic fitness of CD4+ T cells. — **Fig. 7: CTSL deficiency causes aberrant functional tuning and impaired homeostatic fitness of CD4⁺ T cells.**

After 18 h of stimulation with plate-bound I-A^b:LLO_190–201 in vitro, M2 CD4SP cells from Ctsl^ΔTECLm54^Tg mice upregulated the activation marker CD69 with an approximately fivefold lower half-maximum inhibitory concentration than their counterparts from Ctsl^+/+Lm54^Tg mice (Fig. 7d), indicating enhanced TCR sensitivity. Elevated responsiveness to TCR stimulation was likewise observed in polyclonal Ctsl^ΔTEC M2 CD4SP cells (Extended Data Fig. 6c). By contrast, short-term stimulation with phorbol 12-myristate 13-acetate (PMA), which bypasses upstream TCR signaling, elicited comparable levels of ERK phosphorylation in M2 CD4SP cells selected in Ctsl^ΔTECLm54^Tg and Ctsl^+/+Lm54^Tg mice (Extended Data Fig. 6d), suggesting that the hyperresponsiveness of CD4SP cells selected in the absence of CTSL was confined to the proximal TCR signaling cascade. Flow cytometric analysis showed that, in addition to CD5 and CD6, two further negative regulators of TCR signaling (PD-1 and BTLA^34,35) also exhibited reduced expression downstream of positive selection in Ctsl^ΔTECLm54^Tg mice compared to Ctsl^+/+Lm54^Tg mice (Fig. 7e). These findings suggest that the observed TCR-proximal hyperresponsiveness of cells selected in the absence of CTSL may be attributable to reduced expression of multiple TCR signaling attenuators.

Gene expression profiling of M2 CD4SP cells from Ctsl^ΔTECLm54^Tg or Ctsl^+/+Lm54^Tg mice after stimulation with plate-bound I-A^b:LLO_190–201 revealed a relative enrichment of transcriptional modules related to mRNA export, processing and splicing (processes potentially more directly linked to TCR activation) in cells selected in the absence of CTSL (Fig. 7f). By contrast, these cells exhibited comparably less efficient implementation of transcriptional programs related to key metabolic pathways crucial for sustaining T cell activation, including proliferation, glycolysis, mTORC1 signaling, cholesterol homeostasis and fatty acid metabolism (Fig. 7f). Many of these differences were not exclusively triggered following TCR activation but were already evident before stimulation and became more pronounced following TCR engagement (Fig. 7f), suggesting that these traits had been differentially ‘imprinted’ during positive selection in the presence or absence of CTSL.

Given the aberrant ‘tuning’ of multiple basal metabolic programs when Lm54 CD4⁺ T cells were selected in the absence of CTSL, we next examined additional hallmarks associated with CD4⁺ T cell survivability. Flow cytometric analysis showed that, across all maturation stages downstream of positive selection, expression of the interleukin-7 receptor (IL-7R) and BCL-2 (both key orchestrators of CD4⁺ T cell survival³⁶) was reduced in Ctsl^ΔTECLm54^Tg mice compared to Ctsl^+/+Lm54^Tg mice (Fig. 7g).

We therefore next assessed whether M2 CD4SP cells from these two genotypes differed in their homeostatic properties. Following 24-h in vitro culture without added growth factors or TCR stimulation, cells selected in the absence of CTSL exhibited substantially higher apoptosis (Fig. 7h). Administration of high-dose IL-7 rescued this in vitro survival defect (Fig. 7i). To address whether selection in the absence of CTSL impaired homeostasis in vivo, we conducted competitive co-transfer experiments using M2 CD4SP cells from Ctsl^ΔTECLm54^Tg and Ctsl^+/+Lm54^Tg mice. Two weeks after co-transfer at a 1:1 ratio into sublethally irradiated wild-type recipients, the donor cell ratio had markedly shifted in disfavor of cells selected in the absence of CTSL (Fig. 7j). Analogous co-transfers of polyclonal M2 CD4SP cells from Ctsl^ΔTEC and Ctsl^+/+ donors to wild-type recipients recapitulated these observations and revealed a tendency toward diminished homeostatic proliferation of Ctsl^ΔTEC cells (Extended Data Fig. 6e,f).

To disentangle the complex interplay between tonic TCR–pMHCII interactions and competition for soluble cues such as IL-7, which together sustain naive CD4⁺ T cell maintenance and may also drive homeostatic proliferation³⁶, we repeated the co-transfer of M2 CD4SP cells from Ctsl^ΔTECLm54^Tg and Ctsl^+/+Lm54^Tg mice using MHCII^−/− or Rag1^−/− recipients. In MHCII^−/− hosts (where homeostatic MHCII contacts are abolished but survival factors such as IL-7 are readily available due to the absence of endogenous CD4⁺ T cells), cells selected in the absence of CTSL again exhibited a competitive disadvantage, although neither donor population showed evidence of proliferation (Fig. 7k). This indicated a diminished survivability of cells selected in the absence of CTSL that was not explained by altered responsiveness to tonic TCR signaling and/or reduced homeostatic proliferation. In Rag1^−/− recipients, which similarly provide ample access to soluble survival factors owing to their lack of an endogenous CD4⁺ T cell compartment but, in contrast to MHCII^−/− recipients, permit homeostatic TCR–pMHCII contacts, the donor cell ratio once more shifted in disfavor of cells selected in the absence of CTSL (Fig. 7l). However, unlike in MHCII^−/− recipients (Fig. 7k), a significantly smaller proportion of these cells underwent at least one cell division (Fig. 7m), revealing defective homeostatic proliferation as an additional layer of impaired functional fitness. Thus, CD4⁺ T cells selected in Ctsl^ΔTEC mice in a seemingly CTSL-independent manner (based on their progression to the mature CD4SP stage) retained a lasting functional imprint of their selection history, manifesting as a metabolically less poised state and diminished homeostatic responsiveness to both ‘tonic’ TCR signals and TCR-independent survival cues.

Discussion

Our findings highlight two key roles for CTSL in shaping the CD4⁺ T cell compartment by establishing full repertoire diversity and optimizing the fitness of clones that enter the repertoire in a seemingly CTSL-independent manner. We refer to the complete loss of certain clones as ‘essential CTSL dependency’ and to impaired functionality of retained clones as ‘functional CTSL dependency’.

Essentially CTSL-dependent clones contributed disproportionately to repertoire diversity, conceivably reflecting reliance of low-abundance clones on few (or even a single) selecting pMHCII ligand(s). High-abundance clones may be more flexible in their range of selecting pMHCII ligands. If CTSL is required for only a subset of these ligands, the frequency of TCR–pMHC contacts may be reduced, allowing such clones to be retained in the repertoire, albeit with defects characteristic of functionally CTSL-dependent clones. CD5 expression in bulk CD4SP cells was diminished in the absence of CTSL. This was not due to preferential loss of ‘natural’ CD5^hi clones, suggesting that reduced signal strength affected the full range of selecting interactions. As a consequence, natural CD5^lo clones may more frequently fail to reach signaling thresholds for positive selection. Indeed, the proportion of essentially CTSL-dependent clones was higher among natural CD5^lo clones. By contrast, higher-affinity natural CD5^hi clones may persist through compensatory sensitization of proximal TCR signaling. This may involve reduced expression not only of CD5 but also of other negative regulators such as CD6, PD-1 and BTLA, all evident in clones Lm54 and Lm6.

Although reduced peptide diversity may explain some consequences of CTSL deficiency, this does not preclude the possibility that the key role of CTSL lies in generating ‘qualitatively special’ peptides optimal for selection. Despite the loss of ~50% of TCRs in Tcrb^FixedCtsl^ΔTEC mice, numerous ‘newcomer TCRs’ emerged, suggesting that the pMHCII ligandome remained diverse yet was enriched in otherwise absent or outcompeted ‘newcomer ligands’. The unusual V–J features of newcomer TCRs imply that they arose through an atypical selection process and may not be functionally equivalent to normally selected CD4⁺ T cells. As cTECs express at least two other cathepsins, cathepsin B and D²⁴, these may generate diverse, yet ‘suboptimal’, selecting peptides in the absence of CTSL. Circumstantial support for this notion comes from experiments showing that occupancy of <5% of MHCII molecules on cTECs by pMHC ligands generated by the normal proteolytic machinery, including CTSL, was sufficient to sustain CD4⁺ T cell numbers near wild-type levels³⁷.

Essential and functional CTSL dependency likely both contribute to the blunted anti-LLO CD4⁺ T cell response in Ctsl^ΔTEC mice. Half of the ten ‘best-responder’ clones in Tcrb^FixedCtsl^+/+ mice were physically absent from the repertoire of Tcrb^FixedCtsl^ΔTEC mice. We deem it likely that both in Tcrb^Fixed mice and in the fully αβ polyclonal repertoire, a sizeable proportion of clones retained in the absence of CTSL are functionally compromised. This is exemplified by the clones Lm54 and Lm6, whose altered features when selected in the absence of CTSL may be the rule rather than the exception. However, their marked homeostatic defects complicate efforts to resolve how functional CTSL dependency affects responsiveness and effector functions following immunogenic challenge with cognate antigen.

The exact mechanism underlying the impaired homeostatic fitness of functionally CTSL-dependent clones remains unclear. These clones appeared to progress ‘on autopilot’ through differentiation stages downstream of positive selection in the absence of CTSL yet exhibited alterations in multiple cellular programs. A downshift in CD5 expression may directly contribute to their inferior homeostatic behavior. In normally selected thymocytes, elevated CD5 expression calibrates the NF-κB pool, promoting viability and responsiveness despite attenuating TCR signals³⁸. NF-κB also establishes proliferation competence²⁶ and augments IL-7 responsiveness, triggering prosurvival transcriptional programs including BCL-2 (ref. ³⁹). IL-7R expression is fine-tuned by TCR signals during thymic development⁴⁰, and reduced IL-7R and BCL-2 expression in functionally CTSL-dependent clones suggests that selection in the absence of CTSL led to diminished IL-7 responsiveness and compromised survivability.

The ‘altered peptide’ model, proposed over three decades ago, suggested that developing thymocytes engage unique pMHC combinations for positive selection⁴¹. This model was later abandoned when it was found that several abundant MHCII-associated peptides were shared between cTECs and other APCs⁴², and the ‘affinity model’ has become the prevailing concept to explain the dual role of self-pMHC ligands in both positive and negative selection. However, the discovery of distinct proteolytic pathways in cTECs has revived interest in the possibility of ‘private’ pMHC ligands in cTECs having crucial physiological relevance. The ‘peptide-switch’ model proposes that minimizing the overlap between the pMHC ligandomes of cTECs and those of negatively selecting APCs prevents positive and negative selection from canceling each other out^15,43. Although we cannot formally exclude contributions from negative selection, our findings demonstrate that the diminished CD4SP compartment in Ctsl^ΔTEC mice primarily reflects a genuine bottleneck in positive selection.

How ‘private’ peptides generated by CTSL contribute to optimal CD4⁺ T cell selection remains speculative. Their unique role may be specified through conserved TCR contact and/or MHC anchor residues or via an allosteric influence on the ‘nonpeptide’ MHC–TCR interface. Methodological advances will be required to comprehensively characterize the cTEC pMHC ligandome. Another intriguing question is how the requirement for ‘private’ peptides in positive selection can be reconciled with the hypothesis that the very same self-peptides mediating positive selection also support naive T cell homeostasis in the periphery and act as coagonists when T cells respond to foreign antigens^44,45.

Methods

Mice

Ctsl^fl/fl⁴⁶, Ctsl^−/−⁴⁷, Tcr-Dep⁴⁸, Ciita^[kd28, Tcr-Plp1 (ref. ⁴⁹), Tcr-LLO56 and Tcr-LLO118 (ref. ⁵⁰), Tcr-AND and Tcr-AD10 (ref. ⁵¹), Tcr-OT-II⁵², Foxn1-cre⁵³, MHCI^−/− (B2m^−/−)⁵⁴, MHCII^−/− (H2-Ab1^−/−)⁵⁵, Rag1^−/−⁵⁶, Plp1^−/−⁵⁷, Tcra^−/−⁵⁸ and Foxp3^GFP reporter mice (DEREG)⁵⁹ have been described previously. For the generation of Lm54^Tg and Lm6^Tg mice, pTα and pTβ cassette vectors⁶⁰ were modified to contain the V(D)J regions of the respective TCRs identified by single-cell TCR sequencing. Transgenic mice were generated by injection of linearized DNA into the pronuclei of C57BL/6 zygotes. Tcrb^Fixed transgenic mice only express the TCRβ chain of the Lm54 clone. All mice used were on a C57BL/6J background. Mice were maintained under specific pathogen-free conditions in individually ventilated cages at an ambient temperature of 22 °C and 55% humidity with standard light cycle conditions. All phenotypic analyses were performed in mice of 8–12 weeks of age, unless otherwise indicated, and animals of both sexes were included, as we did not find any evidence for sex differences in the parameters addressed. Animal studies and procedures were approved by local authorities (Regierung von Oberbayern; Az Vet_02-22-66).

CTSL active site labeling and western blotting

Sorted TECs were incubated for 1 h at 37 °C in conditioned culture medium containing 1 μM BMV109 and processed as previously described⁶¹. Briefly, cell pellets obtained from 3 × 10⁴ TECs were lysed in hypotonic buffer in the presence of 4 mM DTT for 15 min on ice and centrifuged for 30 min at 10,000g. Supernatants were denatured by the addition of 3× SDS sample buffer (containing 10% β-mercaptoethanol) and resolved by SDS–PAGE (15%) at 120 V for 60 min, together with the protein marker ROTI Mark Tricolor (Roth). The gel was scanned with a Typhoon FLA9500 imager in the Cy5 channel (GE Healthcare). Proteins were either stained with Coomassie or transferred in an exact replica of the gel onto a nitrocellulose membrane via semidry western blotting. Transfer was performed with transfer buffer containing 8% methanol and at a maximum of 50 V and 160 mA for 90 min. Following protein transfer, the membrane was incubated with goat anti-mouse CTSL polyclonal IgG (AF1515, R&D; 1 μg ml^–1), followed by secondary mouse anti-goat horseradish peroxidase-conjugated polyclonal IgG (205-035-108, Jackson ImmunoResearch; 40 ng ml^–1) and SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific). Membranes were imaged with an iBright1500 scanner (Invitrogen). For the housekeeping control, mouse mAb to β-actin (clone AC-15, Sigma; 200 ng ml^–1) was used as primary antibody, followed by secondary rabbit anti-mouse horseradish peroxidase-conjugated polyclonal IgG (P0260, Agilent Dako).

Histology

Five-micron sections of paraffin-embedded material were stained with hematoxylin (Mayer) and eosin (Morphisto). Microscopy was performed with a Leica DM2500 brightfield microscope, equipped with a DMC2900 CMOS camera and HC PL FL ×10/0.30-NA PH1 or HC PL FLUOTAR ×5/0.15-NA objective. The resulting image pixel sizes were 581 nm (×10) and 1.162 nm (×5). Microscopy images were acquired with LAS X Office v1.4.6 (Leica Microsystems).

Determination of cell size

Cell size measurements were conducted on a Countess 3 Automated Cell Counter (Thermo Fisher).

Preparation of TECs

Thymi from 3- to 5-week-old animals were cut into pieces, and thymocytes were mechanically released by pipetting up and down. The supernatant containing thymocytes was discarded. The thymus fragments were digested with liberase (0.5 U ml^–1; Roche) and DNase I (10 mg ml^–1; Roche) at 37 °C in three consecutive rounds of 15 min. Cells were washed and resuspended in 1 ml of high-density Percoll (ρ = 1.115; GE Healthcare) and overlaid with 1 ml of low-density Percoll (ρ = 1.055), followed by a layer of 1 ml of RPMI (Gibco). The gradient was centrifuged at 1,350g for 30 min at 4 °C (without brake). The top interphase, containing the low-density cell fraction, was collected, washed and subjected to CD45 magnetic-activated cell sorting depletion, using CD45 MicroBeads (Miltenyi Biotech). The CD45^– fraction was stained with DAPI and surface antibodies. TECs were analyzed or sorted according to the expression of CD45, Ly51, EpCAM, CD80 and MHCII as follows: cTECs (CD45⁻EpCAM⁺Ly51⁺), total mTECs (CD45⁻EpCAM⁺Ly51⁻) and mTEC^hi (CD45⁻EpCAM⁺Ly51⁻MHCII^hi, CD80^hi).

Flow cytometry

Antibodies were purchased from Biolegend, unless otherwise specified, and used as conjugates with various fluorochromes: anti-CD4 (clone RM4-5), anti-CD8α (53-7.3), anti-CD326/EpCAM (G8.8), anti-Ly51 (6C3), anti-CD80 (16-10A1), anti-CD5 (53-7.3), anti-TCRβ (H57-597), anti-CD69 (H1.2F3), anti-H-2K^b (AF6-88.5), anti-CD45.1 (A20), anti-CD45.2 (104), anti-CD44 (IM7), anti-CD62L (MEL-14), anti-TCRvα2 (B20.1), anti-CD127/IL-7Rα (A7R34), anti-PD-1 (RMP1-30), anti-BTLA (6A6) and anti-I-A/I-E (M5/114.15.2). The following antibodies were used to distinguish MHCII structural epitopes: anti-CLIP:I-A^b (15G4)⁶², anti-non-CLIP:I-A^b (BP107.2.2, a kind gift from A. Rudensky, Memorial Sloan Kettering Cancer Center)⁶³ and anti-Eα_52–68:I-A^b (Y-Ae)⁶⁴. To determine translation activity, thymocytes were pulsed for 5 min with 5 µg ml^–1 puromycin (Merck) immediately before intracellular staining with phycoerythrin (PE)-conjugated mAb to puromycin (2A4)⁶⁵. Control cells were pretreated for 15 min with 5 µg ml^–1 harringtonine (MedChem Express) and, during the puromycin pulse, treated with 100 µg ml^–1 cycloheximide (Roth) and 150 µg ml^–1 chloramphenicol (Merck). For intracellular staining, cells were fixed and permeabilized using reagents from a Foxp3 staining kit (eBioscience) and stained with PE-conjugated anti-Nur77 (12.14, eBioscience) or PE-conjugated anti-BCL-2 (BCL/10C4). Cells were analyzed using a BD FACSCANTOII or LSRFortessa flow cytometer or sorted using a BD FACSAriaFusion sorter. Flow cytometry data were acquired using FACSDiva v6.2 software (BD Bioscience). Flow cytometry data were analyzed with FlowJo v10.9.0 software.

Tet staining

Tets conjugated to APC or PE were kindly provided by M. Jenkins (University of Minnesota)⁶⁶. Single-cell suspensions from thymi or pooled lymph nodes and spleens were incubated for 1 h at 25 °C with 10 nM of both APC- and PE-conjugated Tets, as previously described⁶⁷. Cells were next washed and enriched with anti-APC and anti-PE MicroBeads (Miltenyi Biotech). AccuCheck Counting beads (Thermo Fisher) were used to determine the number of Tet⁺ cells in the column-bound fraction. Tet staining on samples from LLO-immunized mice was followed by enrichment with anti-CD4 MicroBeads (Miltenyi Biotech), instead of anti-APC/anti-PE MicroBeads. A dump cocktail of antibodies, containing anti-CD11b (M1/70), anti-CD11c (N418), anti-B220 (RA3-6B2) and anti-F4/80 (BM8), was used to exclude non-T cells and autofluorescent cells in flow cytometric analyses.

BM chimeras

Recipient mice were irradiated with two split doses of 4.5 Gy, at least 2 h apart, the day before reconstitution. BM was depleted of differentiated cells using biotinylated mAbs to CD4 (clone RM4-5), CD8α (53-7.3), B220 (RA3-6B2), CD11b (M1/70), CD11c (N418), Gr1 (RB6-8C5) and F4/80 (BM8; Biolegend) together with streptavidin magnetic-activated cell sorting MicroBeads (Miltenyi Biotech). Recipient mice were injected i.v. with 5 × 10⁶–10 × 10⁶ BM cells. Neomycin (Belapharm) was supplemented in the drinking water for the first 4 weeks, and mice were killed 6 weeks after reconstitution.

Large-scale Tcra sequencing

For experiments depicted in Fig. 3 (global M2 repertoire comparison in Ctsl^ΔTEC versus Ctsl^+/+ Tcrb^Fixed mice), 3 × 10⁵–4 × 10⁵ mature M2 CD4SP (CD4⁺CD8α⁻CD69⁻MHCI⁺CD25⁻FoxP3⁻) cells from 5- to 8-week-old Ctsl^ΔTECTcra^+/−Tcrb^FixedFoxp3^GFP mice or corresponding controls were bulk sorted. Material from two to three mice was pooled to obtain comparable total cell counts in each sample. For experiments depicted in Fig. 5a–c (global ‘natural’ CD5^lo versus ‘natural’ CD5^hi M2 repertoire comparison in Ctsl^+/+Tcrb^Fixed mice), 1 × 10⁵–2 × 10⁵ mature M2 CD4SP cells (CD4⁺CD8α⁻CD69⁻MHCI⁺CD25⁻FoxP3⁻), corresponding to the 15% lowest or 15% highest levels of CD5 expression on total CD4SP cells, were bulk sorted from individual 5- to 8-week-old Ctsl^+/+Tcra^+/−Tcrb^FixedFoxp3^GFP mice. For experiments depicted in Fig. 5d,e (TCR analysis of expanded LLO-Tet⁺ cells), 1 × 10⁵–2 × 10⁵ LLO-Tet⁺ cells were bulk sorted from pooled spleen and lymph node cells of individual 6- to 8-week-old Ctsl^+/+Tcra^+/−Tcrb^FixedFoxp3^GFP mice 7 days after systemic immunization with LLO_190–201. Samples were stored at −80 °C in RNAprotect (Qiagen). All further sample preparation steps were performed by Qiagen. RNA was isolated using an RNeasy Mini/Micro kit (Qiagen), and library preparation was performed using a QIAseq Immune Repertoire RNA Library kit (Qiagen), modified to include only Tcra-specific primers for both the reverse transcription and target enrichment steps. Library preparation quality control was performed using an Agilent DNA 7500 Chip and Qubit dsDNA HS. Libraries that passed quality control were finally sequenced on a NovaSeq 6000 (Illumina) sequencing instrument according to the manufacturer’s instructions, with a read length of 2 × 250 bp and an SP flow cell. Raw data were demultiplexed, and FASTQ files for each sample were generated using bcl2fastq software (Illumina).

Single-cell TCR sequencing

To increase the likelihood of finding shared TCRs and to ensure subsequent detection of transgenically re-expressed TCRs by fluorescence-activated cell sorting, we restricted this analysis to cells stainable with the mAbs MR9-4 (anti-TCR-Vβ5; genes Trbv5.n) and B20.1 (anti-TCR-vα2; genes Trav14.n). Dump cocktail-negative MR9-4⁺B20.1⁺Tet⁺ cells were single-cell sorted into 96-well plates and immediately frozen at −80 °C. All subsequent steps were adapted from Dössinger et al.⁶⁸. Reverse transcription was performed using an iScript Select cDNA Synthesis kit (Bio-Rad) with a mix of primers specific for the TCRα (Trac) and TCRβ (Trbc) constant regions. Reverse transcription was followed by a digestion step with exonuclease I (Thermo Scientific) to remove single-stranded primers. The exonuclease digestion product was further subjected to a dGTP tailing step and split into two 96-well plates to continue with separate PCRs for Trac and Trbc sequencing. The first PCR was performed with an anchor primer complementary to the introduced 3′-guanosine overhang together with primers specific for Trac or Trbc, respectively. The following two rounds of nested PCR were performed with primers binding Trav14 and Trac or Trbv5 and Trbc, respectively. PCR products were Sanger sequenced by Eurofins Genomics by using either a Trav14-specific primer (for TCRα) or a Trbv5-specific primer (for TCRβ). Both Trac and Trbc sequences were annotated using IMGT/V-Quest⁶⁹ with the C57BL/6-specific library. TCR clonotypes were defined at the level of amino acid sequence (V-region, CDR3 and J-region), and only those consisting of both an in-frame Trac and an in-frame Trbc were retained. As each TCR clonotype found in a given mouse was counted only once, counts in the Source Data related to Fig. 6a refer to the number of individual mice where the respective clonotype was detected.

TCRα repertoire analysis

The CLC Genomics Workbench software (v23.0.3) provided by Qiagen was used to generate clonotype reads. Briefly, sequences with the same unique molecular index were merged. Further quality control steps included merging and trimming of overlapping paired-end sequences. Results were annotated using the C57BL/6J-specific functional Tcra genes extracted from the IMGT mouse TCR database. Analyses were performed using RStudio (version 2024.04.2+764). A clonotype was defined as a unique combination of Trav gene and Traj gene and an in-frame CDR3 amino acid sequence. Clonotype abundance was defined as the number of clonotype-encoding unique cDNA molecules (distinguished by unique molecular index) detected in a sample. TCR diversity was assessed using the approach of Chao et al.⁷⁰, wherein one unique cDNA molecule was defined as one ‘individual’ and one clonotype as one ‘species’. Rarefaction with Hill numbers was performed using ‘iNEXT’ (versions 2.0.20 and 3.0.1). Shannon diversity is the exponential of Shannon entropy and is calculated in ‘iNEXT’ using q = 1. Shannon diversity (or effective number of TCR clonotypes) is the number of equally abundant clonotypes that would be needed to give the same Shannon diversity as the sample. As sampling completeness approached the maximum (coverage = 1) for each sample and subrepertoire examined, the observed Shannon diversities were used to calculate ‘relative diversity’. As adding 95% confidence intervals constructed from five bootstrap replications made a negligible difference to the appearance of the rarefaction curves and were smaller than the symbols on the ‘relative diversity’ summaries, these details were omitted. To assess TCR overlap between pairs of samples, we used ‘abdiv’ (version 0.2.0) to calculate the Morisita–Horn index. Quantification of nucleotide deletion or addition at Trav–Traj junctions was performed using results from the IMGT Junction Analysis tool. The ‘no. nucleotides deleted or added at the Trav–Traj junction’ was defined as the sum of the absolute values of the number of nucleotides deleted from the germline Trav and/or Traj gene(s) plus the number of P and/or N nucleotides added between the remaining germline Trav and/or Traj nucleotides. Due to duplication events at the Tcra locus in C57BL/6 mice, 23–30% of unique cDNA molecules per sample aligned with greater than one Trav paralog; these sequences were excluded from the analyses of chromosomal location of Trav gene usage and nucleotide deletion or addition at Trav–Traj junctions.

Immunization

For hock immunization, isoflurane-anesthetized mice were subcutaneously injected with an emulsion of LLO_190–201 peptide (NEKYAQAYPNVS, GenScript) in PBS and Freund’s adjuvant (Sigma). Systemic immunization was performed by i.v. injection of 50 μg of LLO_190–201 peptide and 50 μg of high-molecular-weight poly(I:C) (InvivoGen) in PBS. When combined with adoptive cell transfers, mice were immunized 6 h before the injection of donor cells.

Adoptive T cell transfer

For experiments depicted in Fig. 4d, mature CD4SP conventional thymocytes from Ctsl^ΔTEC Foxp3^GFP (CD45.2) and Foxp3^GFP littermate control (CD45.1) mice were sorted (CD4⁺CD8α⁻CD69⁻MHCI⁺CD25⁻FoxP3⁻) and mixed at a 1:1 ratio. Cells were injected i.v. into CD45.1/CD45.2 recipients (10⁶ total cells per mouse), previously immunized with LLO plus poly(I:C). For transfer experiments depicted in Fig. 7, M2 CD4SP thymocytes from Ctsl^ΔTECLm54^TgRag1^−/− and Ctsl^+/+Lm54^TgRag1^−/− mice were sorted, mixed at a 1:1 ratio and i.v. injected into CD45.1/CD45.2 recipients (1 × 10⁶ total cells per mouse). Where indicated, the cells were labeled with CellTrace Violet (Invitrogen; 5 μM for 5 min at 37 °C) before adoptive transfer. All results were confirmed with inverted combinations of congenic markers.

In vitro stimulation of Lm54 CD4SP thymocytes

For antigen-specific in vitro stimulation, M2 CD4SP thymocytes were stimulated for 18 h at 37 °C in the presence of soluble anti-CD28 (clone 37.51, Bio X Cell; 250 ng ml^–1) and titrated amounts of plate-bound LLO_190–201:I-A^b monomer. All stimulations were performed with 1.5 × 10⁵ cells per well in 96-well, U-bottom plates and HL-1 serum-free medium (Lonza). The viability dyes Zombie NIR and Apotracker Green (Biolegend) were used to exclude dead and apoptotic cells. Short-term PMA stimulation was performed with 50 ng ml^–1 PMA for 5 min at 37 °C.

RNA sequencing and analysis

Cells were collected and stored at −80 °C in RNAprotect (Qiagen). All further sample preparation steps (RNA isolation, rRNA depletion and library preparation) were performed by Eurofins Genomics, using the INVIEW transcriptome discovery package. Libraries that passed quality control were sequenced on a NovaSeq 6000 (Illumina) sequencing instrument, with 2 × 150 bp read length and S4 flow cells. All data processing methods were applied using default parameters unless specified. Expression quantification was performed using kallisto (version 0.48) with Ensembl release version 106 for M. musculus. In R/Bioconductor, expression data were collapsed from the isoform level to the gene level for downstream processing. Differential expression was assessed using DESeq2 (version 1.36). GSEAs were conducted using fgsea (version 1.22).

Statistical analysis

Statistical analyses were performed using GraphPad Prism (v9), except for calculations of Shannon diversity and Morisita–Horn indexes. The specific statistical tests used are indicated in the corresponding figure legends. Data distribution was assumed to be normal, but this was not formally tested. No a priori sample size calculations were performed; sample sizes were based on prior experience with similar experiments and were deemed sufficient to detect biologically meaningful differences. Measurements were not conducted blind to the conditions of the experiment, as flow cytometric outputs were analyzed using standardized, objective gating strategies. No data points were excluded from analysis.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Sequencing data from this study have been deposited at the Gene Expression Omnibus and will be publicly available from the date of publication. The accession numbers are GSE269202 for the bulk TCR-sequencing data (relating to Figs. 3c–g and 5b–e) and GSE269197 for the RNA-sequencing data (relating to Fig. 7a,b,f). Source data are provided with this paper.

Code availability

An executable capsule containing code and bulk TCR-sequencing data is available at https://doi.org/10.24433/CO.5359169.v1.

References

Klein, L., Kyewski, B., Allen, P. M. & Hogquist, K. A. Positive and negative selection of the T cell repertoire: what thymocytes see (and don’t see). Nat. Rev. Immunol. 14, 377–391 (2014).
CAS PubMed PubMed Central Google Scholar
Klein, L. & Petrozziello, E. Antigen presentation for central tolerance induction. Nat. Rev. Immunol. 25, 57–72 (2025).
CAS PubMed Google Scholar
Vrisekoop, N., Monteiro, J. P., Mandl, J. N. & Germain, R. N. Revisiting thymic positive selection and the mature T cell repertoire for antigen. Immunity 41, 181–190 (2014).
CAS PubMed PubMed Central Google Scholar
Chu, H. H. et al. Positive selection optimizes the number and function of MHCII-restricted CD4⁺ T cell clones in the naive polyclonal repertoire. Proc. Natl Acad. Sci. USA 106, 11241–11245 (2009).
CAS PubMed PubMed Central Google Scholar
Huseby, E. S. et al. How the T cell repertoire becomes peptide and MHC specific. Cell 122, 247–260 (2005).
CAS PubMed Google Scholar
Hogquist, K. A. & Jameson, S. C. The self-obsession of T cells: how TCR signaling thresholds affect fate ‘decisions’ and effector function. Nat. Immunol. 15, 815–823 (2014).
CAS PubMed PubMed Central Google Scholar
Fulton, R. B. et al. The TCR’s sensitivity to self peptide–MHC dictates the ability of naive CD8⁺ T cells to respond to foreign antigens. Nat. Immunol. 16, 107–117 (2015).
CAS PubMed Google Scholar
Mandl, J. N., Monteiro, J. P., Vrisekoop, N. & Germain, R. N. T cell-positive selection uses self-ligand binding strength to optimize repertoire recognition of foreign antigens. Immunity 38, 263–274 (2013).
CAS PubMed PubMed Central Google Scholar
Persaud, S. P., Parker, C. R., Lo, W. L., Weber, K. S. & Allen, P. M. Intrinsic CD4⁺ T cell sensitivity and response to a pathogen are set and sustained by avidity for thymic and peripheral complexes of self peptide and MHC. Nat. Immunol. 15, 266–274 (2014).
CAS PubMed PubMed Central Google Scholar
Bartleson, J. M. et al. Strength of tonic T cell receptor signaling instructs T follicular helper cell-fate decisions. Nat. Immunol. 21, 1384–1396 (2020).
CAS PubMed PubMed Central Google Scholar
Rogers, D. et al. Pre-existing chromatin accessibility and gene expression differences among naive CD4⁺ T cells influence effector potential. Cell. Rep. 37, 110064 (2021).
CAS PubMed Google Scholar
Sood, A. et al. Differential interferon-γ production potential among naive CD4⁺ T cells exists prior to antigen encounter. Immunol. Cell. Biol. 97, 931–940 (2019).
CAS PubMed Google Scholar
Martin, B. et al. Highly self-reactive naive CD4 T cells are prone to differentiate into regulatory T cells. Nat. Commun. 4, 2209 (2013).
PubMed Google Scholar
Zinzow-Kramer, W. M., Weiss, A. & Au-Yeung, B. B. Adaptation by naive CD4⁺ T cells to self-antigen-dependent TCR signaling induces functional heterogeneity and tolerance. Proc. Natl Acad. Sci. USA 116, 15160–15169 (2019).
CAS PubMed PubMed Central Google Scholar
Takahama, Y. The thymoproteasome in shaping the CD8⁺ T-cell repertoire. Curr. Opin. Immunol. 83, 102336 (2023).
CAS PubMed PubMed Central Google Scholar
Nedjic, J., Aichinger, M., Emmerich, J., Mizushima, N. & Klein, L. Autophagy in thymic epithelium shapes the T-cell repertoire and is essential for tolerance. Nature 455, 396–400 (2008).
CAS PubMed Google Scholar
Postoak, J. L. et al. Thymic epithelial cells require lipid kinase Vps34 for CD4 but not CD8 T cell selection. J. Exp. Med. 219, e20212554 (2022).
CAS PubMed PubMed Central Google Scholar
Rodrigues, P. M. et al. LAMP2 regulates autophagy in the thymic epithelium and thymic stroma-dependent CD4 T cell development. Autophagy 19, 426–439 (2023).
CAS PubMed Google Scholar
Gommeaux, J. et al. Thymus-specific serine protease regulates positive selection of a subset of CD4⁺ thymocytes. Eur. J. Immunol. 39, 956–964 (2009).
CAS PubMed Google Scholar
Viret, C. et al. Thymus-specific serine protease contributes to the diversification of the functional endogenous CD4 T cell receptor repertoire. J. Exp. Med. 208, 3–11 (2011).
CAS PubMed PubMed Central Google Scholar
Nakagawa, T. et al. Cathepsin L: critical role in Ii degradation and CD4 T cell selection in the thymus. Science 280, 450–453 (1998).
CAS PubMed Google Scholar
Honey, K. & Rudensky, A. Y. Lysosomal cysteine proteases regulate antigen presentation. Nat. Rev. Immunol. 3, 472–482 (2003).
CAS PubMed Google Scholar
Honey, K., Nakagawa, T., Peters, C. & Rudensky, A. Cathepsin L regulates CD4⁺ T cell selection independently of its effect on invariant chain: a role in the generation of positively selecting peptide ligands. J. Exp. Med. 195, 1349–1358 (2002).
CAS PubMed PubMed Central Google Scholar
Ohigashi, I., Matsuda-Lennikov, M. & Takahama, Y. Peptides for T cell selection in the thymus. Peptides 146, 170671 (2021).
CAS PubMed PubMed Central Google Scholar
Hsieh, C. S., deRoos, P., Honey, K., Beers, C. & Rudensky, A. Y. A role for cathepsin L and cathepsin S in peptide generation for MHC class II presentation. J. Immunol. 168, 2618–2625 (2002).
CAS PubMed Google Scholar
Xing, Y., Wang, X., Jameson, S. C. & Hogquist, K. A. Late stages of T cell maturation in the thymus involve NF-κB and tonic type I interferon signaling. Nat. Immunol. 17, 565–573 (2016).
CAS PubMed PubMed Central Google Scholar
van Meerwijk, J. P. et al. Quantitative impact of thymic clonal deletion on the T cell repertoire. J. Exp. Med. 185, 377–383 (1997).
PubMed PubMed Central Google Scholar
Hinterberger, M. et al. Autonomous role of medullary thymic epithelial cells in central CD4⁺ T cell tolerance. Nat. Immunol. 11, 512–519 (2010).
CAS PubMed Google Scholar
Moran, A. E. et al. T cell receptor signal strength in Treg and iNKT cell development demonstrated by a novel fluorescent reporter mouse. J. Exp. Med. 208, 1279–1289 (2011).
CAS PubMed PubMed Central Google Scholar
Goncalves, C. M., Henriques, S. N., Santos, R. F. & Carmo, A. M. CD6, a rheostat-type signalosome that tunes T cell activation. Front. Immunol. 9, 2994 (2018).
CAS PubMed PubMed Central Google Scholar
Thyagarajan, H. M., Lancaster, J. N., Lira, S. A. & Ehrlich, L. I. R. CCR8 is expressed by post-positive selection CD4-lineage thymocytes but is dispensable for central tolerance induction. PLoS ONE 13, e0200765 (2018).
PubMed PubMed Central Google Scholar
Daley, S. R. et al. Rasgrp1 mutation increases naive T-cell CD44 expression and drives mTOR-dependent accumulation of Helios⁺ T cells and autoantibodies. eLife 2, e01020 (2013).
PubMed PubMed Central Google Scholar
Rosenlehner, T. et al. Reciprocal regulation of mTORC1 signaling and ribosomal biosynthesis determines cell cycle progression in activated T cells. Sci. Signal. 17, eadi8753 (2024).
CAS PubMed Google Scholar
Nishimura, H., Nose, M., Hiai, H., Minato, N. & Honjo, T. Development of lupus-like autoimmune diseases by disruption of the PD-1 gene encoding an ITIM motif-carrying immunoreceptor. Immunity 11, 141–151 (1999).
CAS PubMed Google Scholar
Watanabe, N. et al. BTLA is a lymphocyte inhibitory receptor with similarities to CTLA-4 and PD-1. Nat. Immunol. 4, 670–679 (2003).
CAS PubMed Google Scholar
Surh, C. D. & Sprent, J. Homeostasis of naive and memory T cells. Immunity 29, 848–862 (2008).
CAS PubMed Google Scholar
Barton, G. M. & Rudensky, A. Y. Requirement for diverse, low-abundance peptides in positive selection of T cells. Science 283, 67–70 (1999).
CAS PubMed Google Scholar
Matson, C. A. et al. CD5 dynamically calibrates basal NF-κB signaling in T cells during thymic development and peripheral activation. Proc. Natl Acad. Sci. USA 117, 14342–14353 (2020).
CAS PubMed PubMed Central Google Scholar
Miller, M. L. et al. Basal NF-κB controls IL-7 responsiveness of quiescent naive T cells. Proc. Natl Acad. Sci. USA 111, 7397–7402 (2014).
CAS PubMed PubMed Central Google Scholar
Sinclair, C., Saini, M., Schim van der Loeff, I., Sakaguchi, S. & Seddon, B. The long-term survival potential of mature T lymphocytes is programmed during development in the thymus. Sci. Signal. 4, ra77 (2011).
PubMed Google Scholar
Marrack, P. & Kappler, J. The T cell receptor. Science 238, 1073–1079 (1987).
CAS PubMed Google Scholar
Marrack, P., Ignatowicz, L., Kappler, J. W., Boymel, J. & Freed, J. H. Comparison of peptides bound to spleen and thymus class II. J. Exp. Med. 178, 2173–2183 (1993).
CAS PubMed Google Scholar
Kincaid, E. Z., Murata, S., Tanaka, K. & Rock, K. L. Specialized proteasome subunits have an essential role in the thymic selection of CD8⁺ T cells. Nat. Immunol. 17, 938–945 (2016).
CAS PubMed PubMed Central Google Scholar
Ebert, P. J., Jiang, S., Xie, J., Li, Q. J. & Davis, M. M. An endogenous positively selecting peptide enhances mature T cell responses and becomes an autoantigen in the absence of microRNA miR-181a. Nat. Immunol. 10, 1162–1169 (2009).
CAS PubMed PubMed Central Google Scholar
Lo, W. L. et al. An endogenous peptide positively selects and augments the activation and survival of peripheral CD4⁺ T cells. Nat. Immunol. 10, 1155–1161 (2009).
CAS PubMed PubMed Central Google Scholar
Parigiani, M. A. et al. Conditional gene targeting reveals cell type-specific roles of the lysosomal protease cathepsin L in mammary tumor progression. Cancers 12, 2004 (2020).
CAS PubMed PubMed Central Google Scholar
Roth, W. et al. Cathepsin L deficiency as molecular defect of furless: hyperproliferation of keratinocytes and pertubation of hair follicle cycling. FASEB J. 14, 2075–2086 (2000).
CAS PubMed Google Scholar
Klein, L., Klein, T., Ruther, U. & Kyewski, B. CD4 T cell tolerance to human C-reactive protein, an inducible serum protein, is mediated by medullary thymic epithelium. J. Exp. Med. 188, 5–16 (1998).
CAS PubMed PubMed Central Google Scholar
Wang, L. et al. Epitope-specific tolerance modes differentially specify susceptibility to proteolipid protein-induced experimental autoimmune encephalomyelitis. Front. Immunol. 8, 1511 (2017).
PubMed PubMed Central Google Scholar
Weber, K. S. et al. Distinct CD4⁺ helper T cells involved in primary and secondary responses to infection. Proc. Natl Acad. Sci. USA 109, 9511–9516 (2012).
CAS PubMed PubMed Central Google Scholar
Kaye, J., Vasquez, N. J. & Hedrick, S. M. Involvement of the same region of the T cell antigen receptor in thymic selection and foreign peptide recognition. J. Immunol. 148, 3342–3353 (1992).
CAS PubMed Google Scholar
Barnden, M. J., Allison, J., Heath, W. R. & Carbone, F. R. Defective TCR expression in transgenic mice constructed using cDNA-based α- and β-chain genes under the control of heterologous regulatory elements. Immunol. Cell. Biol. 76, 34–40 (1998).
CAS PubMed Google Scholar
Zuklys, S. et al. Stabilized β-catenin in thymic epithelial cells blocks thymus development and function. J. Immunol. 182, 2997–3007 (2009).
CAS PubMed Google Scholar
Zijlstra, M. et al. β₂-Microglobulin deficient mice lack CD4–8⁺ cytolytic T cells. Nature 344, 742–746 (1990).
CAS PubMed Google Scholar
Cosgrove, D. et al. Mice lacking MHC class II molecules. Cell 66, 1051–1066 (1991).
CAS PubMed Google Scholar
Mombaerts, P. et al. RAG-1-deficient mice have no mature B and T lymphocytes. Cell 68, 869–877 (1992).
CAS PubMed Google Scholar
Klugmann, M. et al. Assembly of CNS myelin in the absence of proteolipid protein. Neuron 18, 59–70 (1997).
CAS PubMed Google Scholar
Mombaerts, P. et al. Mutations in T-cell antigen receptor genes α and β block thymocyte development at different stages. Nature 360, 225–231 (1992).
CAS PubMed Google Scholar
Lahl, K. et al. Selective depletion of Foxp3⁺ regulatory T cells induces a scurfy-like disease. J. Exp. Med. 204, 57–63 (2007).
CAS PubMed PubMed Central Google Scholar
Kouskoff, V., Signorelli, K., Benoist, C. & Mathis, D. Cassette vectors directing expression of T cell receptor genes in transgenic mice. J. Immunol. Methods 180, 273–280 (1995).
CAS PubMed Google Scholar
Verdoes, M. et al. Improved quenched fluorescent probe for imaging of cysteine cathepsin activity. J. Am. Chem. Soc. 135, 14726–14730 (2013).
CAS PubMed Google Scholar
Nakagawa, T. Y. et al. Impaired invariant chain degradation and antigen presentation and diminished collagen-induced arthritis in cathepsin S null mice. Immunity 10, 207–217 (1999).
CAS PubMed Google Scholar
Miyazaki, T. et al. Mice lacking H2-M complexes, enigmatic elements of the MHC class II peptide-loading pathway. Cell 84, 531–541 (1996).
CAS PubMed Google Scholar
Rudensky, A., Rath, S., Preston-Hurlburt, P., Murphy, D. B. & Janeway, C. A. Jr. On the complexity of self. Nature 353, 660–662 (1991).
CAS Google Scholar
Seedhom, M. O., Hickman, H. D., Wei, J., David, A. & Yewdell, J. W. Protein translation activity: a new measure of host immune cell activation. J. Immunol. 197, 1498–1506 (2016).
CAS PubMed Google Scholar
Nelson, R. W. et al. T cell receptor cross-reactivity between similar foreign and self peptides influences naive cell population size and autoimmunity. Immunity 42, 95–107 (2015).
CAS PubMed PubMed Central Google Scholar
Moon, J. J. et al. Naive CD4⁺ T cell frequency varies for different epitopes and predicts repertoire diversity and response magnitude. Immunity 27, 203–213 (2007).
CAS PubMed PubMed Central Google Scholar
Dossinger, G. et al. MHC multimer-guided and cell culture-independent isolation of functional T cell receptors from single cells facilitates TCR identification for immunotherapy. PLoS ONE 8, e61384 (2013).
PubMed PubMed Central Google Scholar
Alamyar, E., Duroux, P., Lefranc, M. P. & Giudicelli, V. IMGT tools for the nucleotide analysis of immunoglobulin (IG) and T cell receptor (TR) V-(D)-J repertoires, polymorphisms, and IG mutations: IMGT/V-QUEST and IMGT/HighV-QUEST for NGS. Methods Mol. Biol. 882, 569–604 (2012).
CAS PubMed Google Scholar
Chao, A. et al. Rarefaction and extrapolation with Hill numbers: a framework for sampling and estimation in species diversity studies. Ecol. Monogr. 84, 45–67 (2014).
Google Scholar

Download references

Acknowledgements

E.P. was supported by the Boehringer Ingelheim Fonds. E.P., A.S. and L.K. received funding from the Deutsche Forschungsgemeinschaft (DFG; German Research Goundation) under project number 456882036. A.S. and L.K. received funding from the DFG under SFB-TRR 355/1 Project B01 (490846870). J.A.F. and L.K. were supported by the European Union Horizon 2020 Research and Innovation Program under the Marie Skłodowska-Curie grant, agreement number 675395 (ENLIGHT-TEN+). T.R. received funding from the DFG GRK 2606 (423813989). S.R.D. received funding from the Australian National Health and Medical Research Council (grant 1188589) and Queensland University of Technology. A.M.S. was supported by European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement number 754462 and by the DFG (419162346 and SFB1371). We thank P. Allen (Washington University) for LLO118 and LLO56 mice, M. Jenkins (University of Minnesota) for MHCII Tets and A. Rudensky (Memorial Sloan Kettering Cancer Center) for the BP107.2.2 hybridoma. We thank B. Tast and the staff of the Flow Cytometry Core Facility at the Biomedical Center for flow cytometry support.

Funding

Open access funding provided by Ludwig-Maximilians-Universität München.

Author information

Anna M. Schulz
Present address: Onkologisches Zentrum Freising MVZ, Freising, Germany
Jan Kisielow
Present address: Repertoire Immune Medicines, Schlieren, Switzerland
These authors contributed equally: Stephen R. Daley, Ludger Klein.

Authors and Affiliations

Institute for Immunology, Biomedical Center, Faculty of Medicine, LMU Munich, Planegg-Martinsried, Germany
Elisabetta Petrozziello, Amina Sayed, João A. Freitas, Christine Federle, Jelena Nedjic, Batuhan Akçabozan, Reinhard Obst & Ludger Klein
Institute of Immunology, Hannover Medical School, Hannover, Germany
Sarina Ravens & Immo Prinz
Division of Animal Physiology and Immunology, School of Life Sciences Weihenstephan and TUM Center for Infection Prevention, Technical University of Munich, Freising, Germany
Anna M. Schulz & Dietmar Zehn
Institute for Experimental Hematology, Center for Translational Cancer Research (TranslaTUM), School of Medicine and Health, Technical University of Munich, Munich, Germany
Marc Schmidt-Supprian
Institute of Systems Immunology, Hamburg Center for Translational Immunology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
Immo Prinz
Department of Tumor Immunology, Radboud Institute for Molecular Life Sciences, Radboud University Medical Center, Nijmegen, the Netherlands
Martijn Verdoes
Institute for Molecular Health Sciences, ETH Zürich, Zürich, Switzerland
Jan Kisielow
Institute of Molecular Medicine and Cell Research, Faculty of Medicine, University of Freiburg, Freiburg, Germany
Thomas Reinheckel
Bioinformatics Unit, Biomedical Center, Faculty of Medicine, LMU Munich, Planegg-Martinsried, Germany
Tobias Straub
Centre for Immunology and Infection Control, School of Biomedical Sciences, Faculty of Health, Queensland University of Technology, Brisbane, Queensland, Australia
Stephen R. Daley

Authors

Elisabetta Petrozziello
View author publications
Search author on:PubMed Google Scholar
Amina Sayed
View author publications
Search author on:PubMed Google Scholar
João A. Freitas
View author publications
Search author on:PubMed Google Scholar
Christine Federle
View author publications
Search author on:PubMed Google Scholar
Jelena Nedjic
View author publications
Search author on:PubMed Google Scholar
Sarina Ravens
View author publications
Search author on:PubMed Google Scholar
Batuhan Akçabozan
View author publications
Search author on:PubMed Google Scholar
Anna M. Schulz
View author publications
Search author on:PubMed Google Scholar
Dietmar Zehn
View author publications
Search author on:PubMed Google Scholar
Marc Schmidt-Supprian
View author publications
Search author on:PubMed Google Scholar
Reinhard Obst
View author publications
Search author on:PubMed Google Scholar
Immo Prinz
View author publications
Search author on:PubMed Google Scholar
Martijn Verdoes
View author publications
Search author on:PubMed Google Scholar
Jan Kisielow
View author publications
Search author on:PubMed Google Scholar
Thomas Reinheckel
View author publications
Search author on:PubMed Google Scholar
Tobias Straub
View author publications
Search author on:PubMed Google Scholar
Stephen R. Daley
View author publications
Search author on:PubMed Google Scholar
Ludger Klein
View author publications
Search author on:PubMed Google Scholar

Contributions

E.P., J.N., S.R.D. and L.K. designed the experimental strategies. E.P., A.S., C.F., S.R., A.M.S. and B.A. performed the experiments. J.A.F., I.P., S.R. and S.R.D. analyzed TCR repertoires. M.V. supported active site labeling experiments. J.K. contributed to the characterization of TCRs. M.S.-S. designed transgenic constructs. T.S. performed gene expression analyses. D.Z. designed immunization strategies. T.R. generated the conditional knockout model. M.V., R.O. and T.R. contributed essential materials. L.K., S.R.D. and E.P. wrote the manuscript.

Corresponding author

Correspondence to Ludger Klein.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Immunology thanks Nuno Alves, Judith N. Mandl and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available. Primary Handling Editor: Ioana Staicu, in collaboration with the Nature Immunology team.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Ctsl-deficiency alters the cTEC pMHCII ligandome and impairs CD4⁺ T cell positive selection.

a, Thymocyte subset cell numbers (Mean ± SEM) in Ctsl^+/+ (n = 8) and Ctsl^ΔTEC mice (n = 10). b, Representative flow cytometry plots of MHCI and CD69 on CD4SPs and percentage ± SEM of M2 cells in Ctsl^+/+ (n = 7) and Ctsl^ΔTEC mice (n = 7). c, Representative flow cytometry plots of CD25 versus Foxp3 and percentage ± SEM of Foxp3⁺CD25⁺ T_reg cells in CD4SPs in Ctsl^+/+ (n = 7) and Ctsl^ΔTEC mice (n = 7); difference not significant. d, Representative flow cytometry plots of CD73 versus CCR7 on Foxp3⁺CD25⁺ CD4SPs in Ctsl^+/+ (n = 3) and Ctsl^ΔTEC mice (n = 4). Percentage ± SEM of CD73⁺CCR7⁻ re-immigrants is indicated; P = 0.1. e, Mean ± SEM of lymph node CD4⁺ T cells in Ctsl^+/+ (n = 3) and Ctsl^ΔTEC mice (n = 3). f, Representative flow cytometry plots of CD44 versus CD62L on CD4⁺ T cells from Ctsl^+/+ (n = 3) and Ctsl^ΔTEC mice (n = 3). g, Representative flow cytometry plots of CD25 versus Foxp3 and percentage ± SEM of Foxp3⁺CD25⁺ T_reg cells in CD4⁺ T cells from Ctsl^+/+ (n = 3) and Ctsl^ΔTEC mice (n = 3); P = 0.016. h, Representative flow cytometry plots of MHCII:CLIP on CD80⁻ (immature) mTECs from Ctsl^+/+ (n = 13) and Ctsl^ΔTEC mice (n = 13); MHCII^−/− defines background. i, Representative flow cytometry plots of MHCII:non-CLIP on immature mTECs from Ctsl^+/+ (n = 13) and Ctsl^ΔTEC mice (n = 13); MHCII^−/− defines background. j, CD4SP percentages ± SEM in Ctsl^+/+ mice (n = 10 or 6) or Ctsl^−/− mice (n = 9 or 7) reconstituted with bone marrow from MHCII^+/+ or MHCII^−/− mice. k, CD4SP percentages ± SEM in Ctsl^+/+ mice (n = 3 or 4) or Ctsl^−/− mice (n = 3 each) on wild-type or C2ta^kd transgenic background. l, Representative flow cytometry plots of thymocyte subsets and frequency ± SEM of CD4SPs in Ctsl^+/+ B2m^−/− (n = 9) and Ctsl^ΔTEC B2m^−/− mice (n = 13); P < 0.0001. m, Representative flow cytometry analysis and MFI ± SEM relative to Ctsl^+/+ cells of CD5 expression on TCRβ^intCD69⁺ DPs from Ctsl^+/+ B2m^−/− (n = 9) and Ctsl^ΔTEC B2m^−/− mice (n = 13). P-values: (a) by 2-way ANOVA and Sidak’s test for multiple comparisons; (b-g) and (j-m) by Student’s two-tailed t-test.

Source data

Extended Data Fig. 2 Ctsl is essential for positive selection of multiple MHCII-restricted transgenic TCRs.

a, Representative flow cytometry plots of thymocyte subsets in Ctsl^+/+ and Ctsl^ΔTEC mice reconstituted with bone marrow (BM) from the indicated Rag1^−/− TCR transgenic donors and corresponding percentages ± SEM of CD4SPs. (n = 3 Ctsl^+/+ and 3 Ctsl^ΔTEC for LLO56; n = 3 and 3 for LLO118). b, Representative flow cytometry plots and MFI ± SEM of CD5 expression in DP thymocytes from BM chimeras as in (a), relative to cells selected in Ctsl^+/+ chimeras. c, Representative flow cytometry plots of thymocyte subsets in Ctsl^+/+ Plp1^−/− (n = 3) and Ctsl^ΔTEC Plp1^−/− mice (n = 3) reconstituted with BM from the PLP1 TCR transgenic donors and corresponding percentages ± SEM of CD4SPs. d, Representative flow cytometry plots and MFI ± SEM of CD5 expression in DP thymocytes from BM chimeras as in (c), relative to cells selected in Ctsl^+/+ Plp1^−/− chimeras. e, CD4SP numbers ± SEM in Ctsl^+/+ and Ctsl^ΔTEC mice (or Ctsl^+/+ Plp1^−/− and Ctsl^ΔTEC Plp1^−/− mice in case of PLP1) reconstituted with BM from the indicated Rag1^−/− TCR transgenic donors as in main Fig. 2a or a or c, respectively). All P-values by Student’s two-tailed t-test.

Source data

Extended Data Fig. 3 Ctsl-deficiency results in an altered repertoire in TCRβ^Fixed mice.

a, CD4SP cell number ± SEM in TCRβ^Fixed Ctsl^+/+ (n = 10) and TCRβ^Fixed Ctsl^ΔTEC mice (n = 9). b, Representative flow cytometry plots of CD4 versus CD8 and percentage ± SEM of CD4⁺ T cells in lymph nodes from TCRβ^Fixed Ctsl^+/+ (n = 5) and TCRβ^Fixed Ctsl^ΔTEC mice (n = 4). c, Cell number ± SEM of CD4⁺ T cells in lymph nodes from TCRβ^Fixed Ctsl^+/+ (n = 7) and TCRβ^Fixed Ctsl^ΔTEC mice (n = 5). d, (left) V-element (Trav gene) usage in recurrent TCRα clones detected exclusively in Fixed-β Ctsl^+/+ TCRα inventories (n = 3) (blue; ‘Ctsl^+/+-only’ Ctsl-dependent TCRs), or exclusively in Fixed-β Ctsl^ΔTEC TCRα inventories (n = 3) (red; ‘Ctsl^ΔTEC-only’ newcomer TCRs). Curves show the cumulative proportion of TCR cDNAs using the respective Trav segment ordered from proximal to distal. (right) Area under the curve (Mean ± SEM), determined by dividing the sum of the cumulative proportions by the number of distinct Trav segments detected. e, (left) J-element (Traj gene) usage in ‘Ctsl^+/+-only’ TCRα clones or ‘Ctsl^ΔTEC-only’ newcomer TCRα clones as in (d). (right) Area under the curve (Mean ± SEM), determined by dividing the sum of the cumulative proportions by the number of distinct Traj segments detected. f, (left) Cumulative proportion of TCRα cDNAs whose V-J junction was modified by deletion or addition of the number of nucleotides specified on the x-axis. (right) Area under the curve (Mean ± SEM), determined by dividing the sum of the cumulative proportions by the number of nucleotide deletion/addition categories (17 possibilities in total). Data in d-f are from n = 3 biological replicates per genotype; each replicate represents pooled cells from 2 or 3 mice. All P-values by Student’s two-tailed t-test.

Source data

Extended Data Fig. 4 Ctsl-deficiency does not affect the number of LLO-Tet⁺ CD4SPs.

(left) Representative flow cytometry plots and total number ± SEM of LLO_190-201-specific CD4SPs in the thymus of Ctsl^+/+ (n = 16) and Ctsl^ΔTEC mice (n = 16) Dot plots are gated on dump-negative (CD11b^–CD11c^–B220^–F4/80^–) CD4⁺CD8^– cells after magnetic enrichment of Tet⁺ cells. P-value by Student’s two-tailed t-test.

Source data

Extended Data Fig. 5 Ctsl calibrates the positively selecting signal strength.

a, Thymocyte subset cell numbers (Mean ± SEM) in Ctsl^+/+Rag1^−/−Lm54^Tg (n = 3) and Ctsl^ΔTECRag1^−/−Lm54^Tg mice (n = 4). b, Thymocyte subset cell numbers (Mean ± SEM) in Ctsl^+/+Rag1^−/−Lm6^Tg (n = 6) and Ctsl^ΔTECRag1^−/−Lm6^Tg mice (n = 7). c, CD6 surface expression (MFI ± SEM) at consecutive DP and CD4SP differentiation stages in Ctsl^+/+Rag1^−/−Lm54^Tg (n = 4) and Ctsl^ΔTECRag1^−/−Lm54^Tg mice (n = 5), assessed by flow cytometry. P-value by Student’s two-tailed t-test.

Source data

Extended Data Fig. 6 Selection in the absence of Ctsl causes aberrant functional tuning and impaired homeostatic fitness of CD4⁺ T cells.

a, Cell size (Mean ± SEM) of M2 thymocytes in Ctsl^+/+Rag1^−/−Lm54^Tg (n = 3) and Ctsl^ΔTECRag1^−/−Lm54^Tg mice (n = 4). b, CD44 surface expression (MFI ± SEM) at consecutive DP and CD4SP differentiation stages in Ctsl^+/+Rag1^−/−Lm54^Tg (n = 4) and Ctsl^ΔTECRag1^−/−Lm54^Tg mice (n = 5); assessed by flow cytometry. c, CD69 and CD25 expression (MFI ± SEM) in polyclonal M2 thymocytes from Ctsl^ΔTEC (n = 3) or Ctsl^+/+ mice (n = 3) after 40 h culture with or without (unstim.) plate-bound antibodies to CD3 and CD28; assessed by flow cytometry. d, Representative flow cytometry plots of intracellular phospho-ERK staining of M2 thymocytes from Ctsl^+/+Rag1^−/−Lm54^Tg (n = 3) and Ctsl^ΔTECRag1^−/−Lm54^Tg mice (n = 3) after 5 min stimulation with PMA. Controls without stimulation define background. e, Donor cell ratio (Mean) in the spleen of 4.5 Gy-irradiated recipients (n = 2) after transfer of a 1:1 mixture of CTV-labelled Ctsl^+/+ and Ctsl^ΔTEC M2 CD4SPs; assessed by flow cytometry. f, Replication index of donor cells on day 21 after transfer as in (e) assessed with FlowJo_v10.9.0 software, assessed by flow cytometry on the basis of CTV-dilution profiles. P-values by Student’s two-tailed t-test.

Supplementary information

Supplementary Information (download PDF )

Supplementary Figs. 1 and 2.

Reporting Summary (download PDF )

Peer Review File (download PDF )

Source data

Source Data Fig. 1 (download XLSX )

Statistical source data.

Source Data Fig. 2 (download XLSX )

Statistical source data.

Source Data Fig. 3 (download XLSX )

Bulk TCR-sequencing data table.

Source Data Fig. 4 (download XLSX )

Statistical source data.

Source Data Fig. 5 (download XLSX )

Bulk TCR-sequencing data table.

Source Data Fig. 6 (download XLSX )

Single-cell TCR-sequencing data table.

Source Data Fig. 7 (download XLSX )

Bulk RNA-sequencing data table.

Source Data Extended Data Fig. 1 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 2 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 3 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 4 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 5 (download XLSX )

Statistical source data.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Petrozziello, E., Sayed, A., Freitas, J.A. et al. Cathepsin L-dependent positive selection shapes clonal composition and functional fitness of CD4⁺ T cells. Nat Immunol 26, 1127–1138 (2025). https://doi.org/10.1038/s41590-025-02182-y

Download citation

Received: 22 May 2024
Accepted: 08 May 2025
Published: 13 June 2025
Version of record: 13 June 2025
Issue date: July 2025
DOI: https://doi.org/10.1038/s41590-025-02182-y

This article is cited by

Let the CAThepsin L out of the thymus
- Nuno L. Alves
Nature Immunology (2025)

Subjects

Abstract

Similar content being viewed by others

Main

Results

CTSL deficiency impairs positive selection of CD4+ T cells

CTSL deficiency causes ‘clonal holes’ and ‘newcomers’

CTSL deficiency blunts CD4+ T cell responses

Nonselection affects clones across the CD5 range

CTSL specifies selection signals in CTSL-independent clones

Functional tuning by CTSL regulates homeostatic fitness

Discussion

Methods

Mice

CTSL active site labeling and western blotting

Histology

Determination of cell size

Preparation of TECs

Flow cytometry

Tet staining

BM chimeras

Large-scale Tcra sequencing

Single-cell TCR sequencing

TCRα repertoire analysis

Immunization

Adoptive T cell transfer

In vitro stimulation of Lm54 CD4SP thymocytes

RNA sequencing and analysis

Statistical analysis

Reporting summary

Data availability

Code availability

References

Acknowledgements

Funding

Author information

Authors and Affiliations

Contributions

Corresponding author

Ethics declarations

Competing interests

Peer review

Peer review information

Additional information

Extended data

Supplementary information

Source data

Rights and permissions

About this article

Cite this article

Share this article

This article is cited by

Search

Quick links

CTSL deficiency impairs positive selection of CD4⁺ T cells

CTSL deficiency blunts CD4⁺ T cell responses