Unique roles of co-receptor-bound LCK in helper and cytotoxic T cells

The kinase LCK and CD4/CD8 co-receptors are crucial components of the T cell antigen receptor (TCR) signaling machinery, leading to key T cell fate decisions. Despite decades of research, the roles of CD4–LCK and CD8–LCK interactions in TCR triggering in vivo remain unknown. In this study, we created animal models expressing endogenous levels of modified LCK to resolve whether and how co-receptor-bound LCK drives TCR signaling. We demonstrated that the role of LCK depends on the co-receptor to which it is bound. The CD8-bound LCK is largely dispensable for antiviral and antitumor activity of cytotoxic T cells in mice; however, it facilitates CD8+ T cell responses to suboptimal antigens in a kinase-dependent manner. By contrast, the CD4-bound LCK is required for efficient development and function of helper T cells via a kinase-independent stabilization of surface CD4. Overall, our findings reveal the role of co-receptor-bound LCK in T cell biology, show that CD4- and CD8-bound LCK drive T cell development and effector immune responses using qualitatively different mechanisms and identify the co-receptor–LCK interactions as promising targets for immunomodulation.

Article https://doi.org/10.1038/s41590-022-01366-0 Lck CA/CA mice did not show the block at the DN stage (Extended Data Fig. 3a,b). By contrast, Lck CA/CA mice had a low count of mature thymocytes, which was more pronounced in CD4 + than in CD8 + mSP thymocytes ( Fig. 1b-d). Lck CA/KR mice showed higher numbers of CD4 + mSP thymocytes than Lck CA/CA mice, but the formation of CD8 + mSP thymocytes was comparable in these two strains ( Fig. 1b-d). These results suggested a kinase-independent role of CD4-LCK, but not CD8-LCK, in thymocyte maturation. Heterozygous Lck WT/KO , Lck WT/KR and Lck WT/CA mice showed normal counts of mature thymocytes, suggesting that a single Lck WT allele is sufficient for proper T cell development (Extended Data Fig. 3c).
The numbers of mature CD4 + and CD8 + T cells in the lymph nodes (LNs) reflected their maturation in the thymus (Fig. 1e,f and Extended Data Fig. 3g). The exception was normal numbers of mature CD8 + T cells in the Lck CA/CA and Lck CA/KR mice, apparently due to lymphopenia-induced proliferation coupled with the generation of CD44 + antigen-inexperienced memory-like CD8 + T cells 18,19 in these mice (Extended Data Fig. 3h,i). Lck CA/CA mice showed a slightly higher frequency of FOXP3 + regulatory T cells among CD4 + T cells than Lck WT/WT mice, which was reverted in the Lck CA/KR mice (Extended Data Fig. 3j), indicating that regulatory T cells are less dependent on CD4-LCK than conventional T cells.
To study the intrinsic role of LCK variants in the development of CD4 + and CD8 + T cells, we generated mixed bone marrow (BM) chimeras by transplanting a 1:1 mixture of BM cells from congenic Ly5.1 mice and Lck-variant strains (Ly5.2) into irradiated Ly5.1/Ly5.2 host mice. We observed reduced numbers of peripheral Lck CA/CA CD4 + and CD8 + T cells in comparison to WT cells (Fig. 1g,h and Extended Data Fig. 3k). The co-receptor-bound kinase-dead LCK in the Lck CA/KR background partially rescued the numbers of CD4 + T cells but not CD8 + T cells (Fig. 1g,h).
Overall, Lck CA/CA mice showed an incomplete block in the maturation of CD4 + and CD8 + T cells, which was partially rescued in the Lck CA/KR mice in the CD4 + , but not CD8 + , compartment.

Role of co-receptor-LCK in double-positive (DP) thymocyte maturation
To elucidate the role of Lck variants in DP thymocytes, we assessed the expression of maturation markers by flow cytometry (Extended Data Fig.  4a,b) followed by unsupervised clustering using self-organizing maps 20,21 . This revealed a cluster of mature CD5, CD69 and TCRβ triple-high DP thymocytes ( Fig. 2a and Extended Data Fig. 4c,d). This cluster was the least abundant in Lck KR/KR and Lck KO/KO mice (Fig. 2b). The percentage of mature DP thymocytes was lower in Lck CA/CA mice than in Lck WT/WT mice, which was largely rescued in the Lck CA/KR mice (Fig. 2b). Because the overall expression of the activation markers was relatively high in DP thymocytes of Lck CA/CA and Lck CA/KR mice (Extended Data Fig. 4a,b), the partial development block of the Lck CA/CA mice probably occurs only at the final steps of the maturation of DP thymocytes. Indeed, the comparison of basal phosphorylation levels of TCRζ and ZAP70 at particular differentiation stages showed that TCRβ low thymocytes experience stronger TCR signaling in the Lck CA/CA and Lck CA/KR mice than in the Lck WT/WT mice, but this difference disappears or even reverses during their maturation into postselection TCRβ high DP thymocytes and subsequently mature TCRβ high SP4 and SP8 stages (Extended Data Fig. 4e,f and Supplementary Fig. 1c).
To assess antigenic signaling in thymocytes, we crossed our Lck-variant mice with monoclonal OT-I TCR Rag2 KO/KO (henceforth OT-I) transgenic mice specific to ovalbumin-derived H2-K b -SIINFEKL antigen (OVA). We stimulated thymocytes with T2-Kb cells presenting titrated doses of OVA or its altered peptide ligands (APL) with lower affinity to OT-I ( Supplementary Fig. 2a). Whereas Lck CA/CA and especially Lck CA/KR SP8 T cells showed weaker responses to low-affinity APLs (T4 and G4) than Lck WT/WT , we did not observe any differences in DP thymocytes (Fig. 2c). Accordingly, Lck CA/CA and Lck CA/KR thymocytes isolated from OT-I B2m KO/KO mice, which are arrested at the preselection DP stage, showed similar (if not slightly increased) response as their Lck WT/WT counterparts (Extended Data Fig. 5a). of the co-receptor-LCK interactions in the positive selection of MHC class I/MHC class II-restricted T cells, but it did not address the role of this interaction in the immune response 12 .
Overall, the contribution of co-receptor-LCK interactions to T cell signaling and eventual fate decisions is still unclear. The intuitive model is that a co-receptor-recruited LCK phosphorylates the TCR-CD3 complex 4,5,13 . An alternative model proposes that this key phosphorylation event is preferentially performed by 'free' LCK 7,9 . In the latter scenario, co-receptor-bound LCK physically stabilizes the TCR-antigen interaction from inside the cell 8 . The experimental in vivo evidence for either of these models is missing.
In this study, we characterized the role of co-receptor-bound LCK in vivo using genetically modified mouse models. The importance and mode of action of co-receptor-bound LCK differs in cytotoxic and helper T cell lineages.

Mouse models for studying the co-receptor-LCK interaction
We addressed the physiological relevance of the interaction between LCK and CD4/CD8 co-receptors using reverse genetics in mice. We generated knock-in mouse strains expressing endogenous levels of LCK bearing C20A.C23A (CA) or K273R (KR) amino acid substitutions and LCK-deficient (Lck KO/KO ) mice (Extended Data Fig. 1a-c). LCK CA does not interact with CD4 and CD8 (refs. 12,14 ; Extended Data Fig. 1d,e), and, thus, T cells in Lck CA/CA mice rely exclusively on the co-receptor-unbound 'free' LCK. LCK KR has no enzymatic activity 15 , but the putative adaptor function of LCK should be preserved. To uncouple the proposed catalytic and adaptor roles of co-receptor-bound LCK, we produced Lck CA/KR compound heterozygotes expressing one pool of strictly cytoplasmic 'free' LCK CA together with a pool of kinase-dead LCK KR interacting with co-receptors (Fig. 1a). If the TCR-CD3 complex is preferentially phosphorylated by 'free' LCK, as proposed previously 7,9 , and the co-receptor-bound LCK carries the adaptor function 8 , the Lck CA/KR mice should have normal T cell development and function.
We tested the enzymatic activity of the LCK variants in two cell lines. The cotransfection of the mouse Lck variants and their substrates CD247 (TCRζ) or ZAP70 into HEK293 cells showed that LCK CA and wild-type LCK (LCK WT ) have a comparable activity, whereas LCK KR lacks the kinase activity, as expected (Extended Data Fig. 2a,b). Accordingly, LCK-deficient Jurkat cells 16 reconstituted with human LCK WT and LCK CA showed a comparable phosphorylation of TCRζ and ZAP70 and overall tyrosine phosphorylation after stimulation with anti-TCR, whereas LCK KR was not able to restore signaling (Extended Data Fig. 2c). Moreover, the LCK KO Jurkat cells expressing OT-I TCR specific to K b -OVA antigen reconstituted with LCK WT or LCK CA showed a comparable response to the antigen (measured as CD69 upregulation), whereas LCK KR -expressing cells were unresponsive (Extended Data Fig. 2d).
Overall, we generated and validated mouse models tailored to uncover the role of co-receptor-bound LCK in vivo.

T cell maturation with uncoupled LCK and co-receptors
Lck KO/KO mice exhibited partial blocks at two key stages of T cell development in the thymus (Fig. 1b-d, Extended Data Fig. 3a-c and Supplementary Fig. 1a), as shown previously 17 . First, high frequencies of double-negative (DN) thymocytes (Fig. 1b) and, specifically, CD25 + CD44 -DN3 cells (Extended Data Fig. 3a,b) indicate inefficient pre-TCR signaling during β-selection. Second, low numbers of CD4 + or CD8 + CD24A − TCRB + (TCRβ + ) mature single-positive (mSP) thymocytes indicate defective positive selection of self-pMHC-restricted T cells ( Fig. 1b-d). Lck KR/KR mice showed an even more severe phenotype than the Lck KO Lck WT Lck KO Table 3). e-g, TCR repertoires of FACS-sorted CD4 + and CD8 + mSP cells in the LNs and thymi from indicated mice were profiled. UMIcorrected counts of TCRα (TRA) and TCRβ (TRB) CDR3 amino acid sequences were normalized after the removal of NKT TRAV11-TRAJ18 CDR3 sequences. Sample sizes are in Supplementary Table 4  Lck WT/WT thymi (Extended Data Fig. 5b). However, this was compensated by higher numbers of CD8αα SP cells, which are induced by strong signals 22 , suggesting that Q4H7 might act as a weak partial negative selector for the preselection DP thymocytes in the Lck CA/CA mice.
To study the role of the CD4-LCK interaction in the signaling of MHC class II-restricted thymocytes, we crossed our collection of the Lck-variant mice with TCR transgenic B3K508 Rag2 KO/KO (henceforth B3K508) mice specific for H2-A b -bound FEAQKAKANKAKAVD (3K) peptide 23 . The responses of DP thymocytes to Ly5.1 splenocytes presenting 3K or its APLs were comparable among the Lck WT/WT , Lck CA/CA and Lck CA/KR strains (Fig. 2d). By contrast, the responses of Lck CA/CA and Lck CA/KR SP4 thymocytes were weaker than those of Lck WT/WT thymocytes (Fig. 2d).
Overall, these data indicate that the signaling in DP T cells is relatively normal in the Lck CA/CA and Lck CA/KR mice and that their T cell developmental defects occur only during the late stages of DP development and during the maturation of SP stages.
To assess how the absence of the co-receptor-LCK interactions shapes the T cell repertoire, we analyzed TCRα and TCRβ sequences in SP thymocytes and peripheral T cells from Lck WT/WT , Lck CA/CA and Lck CA/ KR mice (Supplementary Tables 1 and 2). We did not observe major differences in TRAV and TRBV usage, with the exception of the enrichment for natural killer T (NKT) cell typical segments TRAV11, TRBV1, TRBV13-2 and TRBV29 (ref. 24 ) in the Lck CA/CA and Lck CA/KR SP4 thymocytes (Extended Data Fig. 6a,b). Accordingly, canonical NKT cell TCRα chains, TRAV11-TRAJ18 (Vα14-Jα18), were very abundant in SP4 thymocytes in Lck CA/CA and Lck CA/KR mice but not in peripheral CD4 + T cells (Extended Data Fig. 6c). After removing the NKT sequences, the repertoires of Lck CA/CA and Lck CA/KR mice were slightly less diverse than the repertoires of Lck WT/WT mice (Extended Data Fig. 6d). Principle-component analysis revealed that the TCR repertoires of SP thymocytes in some mice differ from the other samples, but the repertoires of LN cells show only subtle differences and no clear separation of the strains ( Fig. 2e and Supplementary Fig. 3). Accordingly, the most abundant peripheral TCR sequences in Lck WT/WT mice were frequent also in Lck CA/CA and Lck CA/KR mice ( Fig. 2f and Supplementary Fig. 4). Finally, there was a substantial overlap of individual sequences in peripheral T cells among the three strains ( Supplementary Fig. 5). The overlap of individual peripheral TCR sequences between the Lck CA/CA or Lck CA/KR mice and the Lck WT/WT mice was comparable to the overlap between individual Lck WT/WT mice (Fig. 2g).
Overall, the disruption of the co-receptor-LCK interaction does not reduce the development and signaling of DP thymocytes until they reach their final maturation stage. As a result, the peripheral repertoires of Lck CA/CA and Lck CA/KR mice are only minimally affected.

'Free' LCK is sufficient for largely adaptive immune responses
To study the effects of LCK variants on T cell function, we examined the Lck strains for their antiviral and antitumor immunity, which is mediated mostly by CD8 + T cells. The ability to clear lymphocytic choriomeningitis virus (LCMV) was comparable in the Lck WT/WT and Lck CA/CA mice, slightly impaired in the Lck CA/KR mice and substantially defective in the Lck KR/KR mice (Fig. 3a). The numbers of CD8 + T cells in the spleen were lower in the infected Lck CA/CA and Lck CA/KR mice than in the Lck WT/WT mice and were further reduced in the Lck KR/KR mice (Extended Data Fig. 7a). A similar reduction was observed in CD4 + T cells, with the notable difference that Lck CA/KR showed a partial rescue compared to Lck CA/CA (Extended Data Fig. 7b). We used D b -GP33 and D b -NP396 tetramers for the detection of CD8 + T cells specific to the immunodominant LCMV epitopes. The frequency of LCMV-specific CD8 + T cells was comparable in the Lck WT/WT and Lck CA/CA mice but was lower in the Lck CA/KR mice (Fig. 3b, Extended Data Fig. 7c and Supplementary  Fig. 6a,b). These LCMV-specific T cells had an antigen-experienced phenotype (CD44 + CD49d + ) and formed a comparable fraction of KLRG1 + CD127 − short-lived effectors in these strains (Extended Data Fig. 7d,e and Supplementary Fig. 6c,d). Lck KR/KR mice showed low numbers of LCMV-specific CD8 + T cells, incomplete differentiation into antigen-activated CD44 + CD49d + T cells and a bias toward the formation of short-lived effectors (Fig. 3b, Extended Data Fig. 7a-e and Supplementary Fig. 6b-d), which explained the defective viral clearance in this strain.
We observed impaired formation of CXCR5 + PD-1 + CD4 + follicular helper T cells (T FH ) in the Lck KR/KR and Lck CA/CA mice during LCMV infection, which was partially rescued in the Lck CA/KR mice (Fig. 3c,d and Supplementary Fig. 6e). Only a small percentage of these T FH cells were FOXP3 + follicular regulatory T cells (Extended Data Fig. 7f). The frequencies of CD4 + T cells specific for an immunodominant GP66 LCMV epitope were comparable between the Lck WT/WT and Lck CA/CA mice and were slightly lower in the Lck CA/KR mice (Extended Data Fig. 7g and Supplementary Fig. 6e). The counts of GP66-specific CD4 + T cells were comparable in the Lck CA/CA and Lck CA/KR mice and higher in the Lck WT/WT mice (Extended Data Fig. 7g). These GP66-specific T cells showed defective differentiation into FOXP3 − T FH cells in the Lck KR/ KR mice and to a lesser extent in Lck CA/CA , but not in Lck CA/KR , mice (Fig. 3e). Although the difference between Lck CA/CA and Lck CA/KR was not significant in this small cohort, it corresponded to the overall CD4 + T cell population (Fig. 3d).
The Lck KR/KR and Lck KO/KO mice failed to hamper the growth of MC-38 carcinomas expressing OVA (Fig. 3f,g). The Lck CA/CA and Lck CA/KR mice showed slightly or substantially faster tumor growth than Lck WT/WT mice, respectively ( Fig. 3f,g). We did not observe large differences in the number of total T cells in the tumor (Extended Data Fig. 7h and Supplementary Fig. 6f) among the strains. The numbers of antigen-specific K b -OVA tetramer + CD8 + T cells infiltrating the tumor and tumor-draining LNs were comparable among the Lck WT/WT , Lck CA/CA and Lck CA/KR mice but were lower in the Lck KR/KR strain (Extended Data Fig. 7i,j and Supplementary Fig. 6g-i). The suboptimal antitumor response in Lck CA/CA and Lck CA/KR mice is probably caused by impaired killing of tumor cells rather than by the absence of tumor-specific T cell clones.
Overall, the Lck CA/CA mice showed relatively normal antiviral and antitumor immune responses, suggesting that the interaction between CD8 and LCK is not essential for these types of immune protection. The Lck CA/KR mice showed defective tumor and viral clearance in these CD8 + T cell-based models but partially rescued the Lck CA/CA phenotype in the CD4 + T FH compartment. This indicated a differential kinase-independent function of CD8-and CD4-bound LCK.

CD8-LCK promotes responses to suboptimal antigens
To study the roles of CD4-and CD8-bound LCK separately, we used MHC class I-restricted and MHC class II-restricted monoclonal mice. First, we investigated the Lck variants in peripheral CD8 + OT-I T cells. Whereas the Lck KO/KO and Lck KR/KR OT-I mice showed a severe developmental impairment, the Lck CA/CA mice had slightly more SP8 T cells than the Lck WT/ WT mice (Fig. 4a-d), which was not observed in the polyclonal setting (Fig. 1d). This is probably connected with slightly stronger signaling of preselection DP thymocytes in the Lck CA/CA mice (Extended Data Fig. 5a,b), the absence of competing MHC class I-independent T cell clones and/or non-physiological regulation of the transgenic TCR expression. The number of SP8 T cells was reduced in the Lck CA/KR mice compared to in Lck CA/CA mice. Peripheral T cell counts were comparable in the Lck WT/WT , Lck CA/CA and Lck CA/KR mice (Fig. 4d). We observed a similar phenotype using another MHC class I-restricted TCR transgenic mouse strain F5 Rag1 KO/KO (Extended Data Fig. 8a-d).
To analyze the role of CD8-bound LCK in TCR signaling, we activated peripheral OT-I T cells with antigen-presenting cells loaded with OVA peptide or its lower-affinity APLs ex vivo using CD69 upregulation as a readout. The Lck CA/CA OT-I T cells showed a normal response to OVA but a reduced response to low-affinity OVA variants compared to the Lck WT/WT cells ( Fig. 4e and Supplementary Fig. 7a). The Lck CA/KR OT-I T cells exhibited even weaker responses than the Lck CA/CA OT-I T cells (Fig. 4e), documenting the inhibitory role of CD8-bound kinase-dead LCK. The upregulation of CD69 and proliferation induced by the Article https://doi.org/10.1038/s41590-022-01366-0 co-receptor-independent activation with anti-CD3/CD28 beads were comparable among these three strains (Extended Data Fig. 8e,f).
To separate LCK-dependent and LCK-independent roles of CD8, we analyzed the antigenic response of human OT-I Jurkat cells 25 devoid of CD8 or expressing WT CD8αβ (CD8 WT ) or LCK-binding mutant CD8α C215.217A β (CD8 CA ) 14 . Jurkat cells expressing CD8 WT and CD8 CA showed ~330-fold and ~35-fold lower responses to OVA-pulsed antigen-presenting cells than CD8cells, respectively ( Fig. 4f and Supplementary Fig. 7b). These results indicated that CD8 contributes to T cell activation in LCK-dependent and LCK-independent manners.
To elucidate the role of CD8-bound LCK in the antigenic response in vivo, we adoptively transferred the Lck-variant OT-I T cells into congenic Ly5.1 mice followed by infection with transgenic Listeria monocytogenes (Lm) expressing OVA or its lower-affinity APLs. Whereas there were no large differences in the responses to OVA, the expansion induced by low-affinity APLs followed the hierarchy Lck WT/WT > Lck CA/CA > Lck CA/KR (Fig. 5a).
We examined the ability of the Lck-variant OT-I T cells to hamper tumor progression following their adoptive transfer into T cell-deficient Cd3e KO/KO mice bearing small MC-38 OVA tumors. The antitumor activity of OT-I cells followed the hierarchy Lck WT/WT > Lck CA/CA > Lck CA/KR (Fig. 5b,c).
It has been proposed that the CD8-LCK interaction might stabilize antigen binding 8 . We assessed the role of CD8-LCK in antigen avidity using three different assays. Whereas the K b -OVA and K b -T4 tetramer staining (Fig. 5d) and the on-cell k off measurements 26 (Fig. 5e) indicated that CD8-LCK indeed stabilizes the TCR-antigen interaction, two-dimensional (2D) affinity measurements using the antigen nested in a lipid bilayer did not reveal substantial differences (Fig. 5f). Regardless of the slight discrepancies between these methods,

Fig. 3 | Role of co-receptor-LCK interaction in T cell immunity. a,b, Indicated
Lck-variant mice were infected with LCMV. a, Viral titers in the spleens were determined by RT-qPCR on day 5 or 6 after infection; n = 16   Lck WT Lck KO Lck KR  Cell count CD8α-BV421 CD4-AF700   Fig. 8g,h). This implied that the gain of function of the CD8.4 allele is mediated via LCK binding.

Fig. 4 | CD8-bound LCK is largely dispensable for positive selection and T cell activation. a-d, Thymi (a and b) and LNs (c and d) of indicated
Overall, these data suggested that the interaction of CD8 with LCK is dispensable for the responses to high-affinity antigens but enhances signaling to suboptimal antigens. The kinase-dead LCK coupled to CD8 downmodulates the T cell response to suboptimal antigens.

LCK supports surface CD4 + and helper T cell responses
We studied the role of the CD4-LCK interactions using B3K508 mice. Lck KO/KO and Lck KR/KR B3K508 thymocytes showed a developmental block ( Fig. 6a-d). In contrast to the polyclonal mice (Fig. 1b,d), the Lck CA/CA and Lck CA/KR mice had comparable (or even slightly higher) counts of SP4 and peripheral CD4 + T cells as the Lck WT/WT mice ( Fig. 6a-d). These results indicated that the CD4-LCK interaction is not required for the commitment of MHC class II-restricted T cells to the CD4 + T cell lineage.
The Lck CA/CA B3K508 T cells exhibited weaker ex vivo antigenic responses to the cognate 3K peptide and its intermediate-and low-affinity APLs (P5R and P2A, respectively) than the Lck WT/WT B3K508 T cells (Fig. 6e). The Lck CA/KR B3K508 T cells partially rescued defective responses to high-affinity 3K and intermediate-affinity P5R antigens but not to a low-affinity antigen P2A (Fig. 6e). The upregulation of CD69 and proliferation induced by the co-receptor-independent activation with anti-CD3/CD28 beads were comparable among these three strains (Extended Data Fig. 9a,b). OT-I Rag2 KO/KO , CD8 + LN cells In line with the results from the ex vivo activation, we observed weaker responses of the Lck CA/CA B3K508 T cells to Lm expressing 3K or low-affinity P2A in vivo (Fig. 6f). The Lck CA/KR T cells rescued responses to Lm-3K but not to Lm-P2A (Fig. 6f). These data indicated that CD4-coupled LCK has a kinase-independent role in T cell activation, but the response to low-affinity antigens requires the catalytic activity of CD4-bound LCK.
We observed that LCK stabilizes surface CD4 levels in a kinase-independent manner in peripheral CD4 + T cells (Fig. 7a,b). By contrast, surface CD8 levels were largely LCK independent (Extended Data Fig. 9c). The interaction with LCK stabilized surface CD4 also in thymocytes, but this effect was much weaker in DP thymocytes (Extended Data Fig. 9d) than in SP4 thymocytes (Extended Data Fig. 9e) or LN CD4 + T cells (Fig. 7a,b).
To address the role of LCK in stabilizing CD4 in human T cells, we measured surface CD4 levels in WT and LCK-deficient (LCK KO ) Jurkat cell lines 16 . LCK KO Jurkat cells expressed very low levels of surface CD4 (Extended Data Fig. 9f), which could be reverted by transducing these cells with LCK WT or LCK KR but not LCK CA (Fig. 7c). The downregulation of surface CD4 in the absence of its interaction with LCK was mediated by protein kinase C (PKC), as the PKC inhibitor elevated surface CD4 levels specifically in T cells from the Lck CA/CA , but not Lck WT/WT or Lck CA/KR , mice ( Fig. 7d and Extended Data Fig. 9g). Using electron microscopy, we observed that the absence of the CD4-LCK interaction modulates CD4 distribution in the plasma membrane. LCK is present in clusters in Lck CA/CA CD4 + T cells but is relatively uniform in Lck WT/WT CD4 + T cells (Fig. 7e,f). Overall, the interaction of CD4 with LCK is required for its proper surface localization in CD4 + T cells and SP4 thymocytes and to a lesser extent in DP thymocytes.

Discussion
We generated Lck CA/CA , Lck KR   interactions contribute to optimal T cell development and immune responses. However, the defects observed in mice with disrupted CD8-LCK and CD4-LCK interactions were much more subtle than in the LCK-deficient mice or mice expressing kinase-dead LCK, indicating that 'free' LCK can at least partially promote TCR signaling in vivo. The genetic disruption of the co-receptor-LCK interaction did not impair the development of three tested monoclonal MHC class I-or MHC class II-restricted thymocytes. It is unclear why a previous study reported a severe block in the development of thymocytes expressing MHC class II-restricted AND TCR 12 . It is possible that it was caused by their experimental system (transgenic expression of LCK) or by a unique feature of the AND TCR. However, such a feature is not a low level of self-reactivity, as the normally developing F5 T cells are weakly self-reactive 27,29,30 , whereas AND T cells are among the most self-reactive transgenic TCR clones 31 .
Polyclonal Lck CA/CA mice developed relatively normal peripheral TCR repertoires and formed virus-specific and tumor-specific MHC class I-and MHC class II-restricted T cells. However, unlike the monoclonal Lck CA/CA mice, the polyclonal Lck CA/CA mice showed a block in the development of mature SP thymocytes. This difference can be caused by the non-physiological timing of TCR expression in the TCR transgenic mice, which alters the outcome of thymic checkpoints 32,33 . Another major factor is the presence of unconventional T cells in polyclonal mice and their comparative advantage over MHC class I/MHC class II-restricted T cells in the absence of the co-receptor-LCK interaction 12 . Indeed, we observed largely increased frequencies of NKT cell clones among the Lck CA/CA thymocytes. Although it is tempting to speculate that the unconventional T cells suppress the formation MHC class I/MHC class II-restricted ones in a direct competition in the Lck CA/CA mice, our experiments with the BM chimeras show that the suboptimal formation of mature CD4 + and CD8 + T cells in the Lck CA/CA and Lck CA/KR mice is at least partially intrinsic. Overall, these results indicate that the co-receptor-LCK interaction is not essential for the formation of MHC class I/MHC class II-restricted T cells and their proper CD4/CD8 lineage commitment. However, in the absence of the co-receptor-LCK interaction, the maturation of MHC class I/MHC class II-restricted T cell clones is affected, whereas the formation of MHC class I/MHC class II-independent T cells is augmented.
We observed that the maturation and TCR signaling of early DP thymocytes is not blocked but is even slightly enhanced in the Lck CA/CA mice. The plausible explanation is that the Lck CA/CA mutation strips LCK from both CD4 and CD8 co-receptors, which leads to a large pool of 'free' LCK. For MHC class I-restricted TCRs, the loss of CD8-bound LCK probably decreases the responsiveness to self-antigens, but this is (over)compensated by the release of CD4-bound LCK into the 'free' LCK pool and vice versa for MHC class II-restricted thymocytes. It has been proposed that the co-receptor-bound LCK has a lower kinase activity than 'free' LCK 9 , which would enhance the signaling in Lck CA/CA thymocytes. However a contradictory study proposed that the co-receptor binding enhances LCK activity 34 , and we did not observe any differences between the response of Lck WT/WT and Lck CA/CA peripheral T cells to co-receptor-independent antibody-mediated TCR signaling. Lck CA/CA thymocytes have only a partial block in the formation of the most mature DP stage, which does not fully explain the relatively strong loss of SP4 and SP8 thymocytes. This suggests that a previously unappreciated signaling checkpoint might occur at the very transition  Article https://doi.org/10.1038/s41590-022-01366-0 of postselection DP thymocytes into the SP stage and/or during the maturation of SP thymocytes. Our Lck-variant mice seem to be promising tools for further elucidating the role of LCK in fate decisions of conventional and unconventional T cells at different stages of maturation.

Fig. 7 | LCK retains CD4 on the surface of T cells in a kinase-independent manner. a,b, Relative CD4 surface levels on CD4 + T cells isolated from indicated
Our data show that the importance of CD8-LCK and CD4-LCK for T cell development and immune responses differs. CD8-bound LCK plays only a relatively minor role in the development of cytotoxic T cells. Although Lck CA/CA CD8 + T cells are outcompeted by Lck WT/WT cells in the mixed BM chimeras, peripheral homeostatic proliferation 35 apparently compensates for the inefficient formation of CD8 + T cells. Lck CA/CA mice also exhibit normal or near-normal antiviral and antitumor immunity and CD8 + T cell responses to high-affinity antigens. The role of CD8-LCK in peripheral T cells seems to be largely limited to enhancing the signaling induced by low-affinity antigens. By contrast, CD8 + T cells in Lck CA/KR mice show defective antigenic responses, suggesting that CD8-bound kinase-dead LCK KR inhibits activity of the 'free' LCK CA . Plausibly, LCK phosphorylates the TCR complex only when localized in a unique site. Co-receptor-bound LCK KR might preferentially occupy this position, preventing the 'free' LCK CA from initiating TCR signaling in the compound heterozygotes. This scenario is in line with the recent observation that co-receptor-bound LCK disables TCR triggering by pMHCs with reversed docking orientations 10,36 . However, following disruption of the CD8-LCK interaction, these reversely binding antigens induce TCR signals with comparable strength to canonical ligands 10 . Further investigation is required for understanding of the relationship between the localization of LCK and its ability to phosphorylate the TCR-CD3 complex. Overall, although we observed that the CD8-bound LCK enhances pMHCI tetramer binding, the functional assays did not reveal any biological importance of a previously proposed kinase-independent adaptor role of CD8-bound LCK 8 .
CD4 + T cells are more dependent on the co-receptor-LCK interaction than CD8 + T cells. Defective maturation and activation of Lck CA/CA CD4 + T cells is partially rescued with a pool of CD4-interacting kinase-dead LCK present in the Lck CA/KR mice, implying a kinase-independent role of CD4-bound LCK. Indeed, LCK promotes surface CD4 localization and its homogenous distribution in the plasma membrane, especially in mature T cells. The regulation of CD4 stability and trafficking by LCK was observed previously in transgenic non-lymphoid cell lines 37,38 , but its relevance for T cell biology was not investigated before.
The responses to low-affinity antigens require the kinase activity of CD4-bound LCK, which is analogous to CD8 + T cells. It is plausible that the CD4-bound kinase-dead LCK might have a dominant-negative role in TCR triggering under certain conditions. The ambiguous roles of the kinase-dead LCK might explain the rescue phenotype in the Lck CA/KR mice in some aspects of T cell biology (for example, formation of mature CD4 + T cells and signaling of B3K508 CD4 + T cells) but not in some other assays (signaling of B3K508 SP4 thymocytes and formation of NKT cells).
As we did not observe impaired development of monoclonal T cells in the Lck CA/CA and Lck CA/KR mice, we concluded that the defective responses of mature CD4 + T cells and CD8 + T cells observed in monoclonal and polyclonal Lck CA/CA and Lck CA/KR mice are probably largely intrinsic. However, we cannot formally exclude that some relevant differences are imprinted already during thymic development in the knock-in strains.
The differential role of co-receptor-bound LCK in the response to high-and low-affinity antigens could be potentially used for the development of novel strategies for treating autoimmune diseases. The specific inhibition of co-receptor-bound LCK should impair autoimmune T cell clones with relatively low antigen affinity 39,40 without inhibiting protective high-affinity T cell responses to infections. Moreover, disruption of the interaction between co-receptors and LCK might modulate the balance between cytotoxic and helper T cell responses, which could be beneficial in the tumor treatment.

Online content
Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/s41590-022-01366-0.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. All TCR transgenic mice used in this study had a Rag2 KO/KO or Rag1 KO/KO background. The colonies of all transgenic strains were established de novo in our animal facility by rederivation using embryo transfer or in vitro fertilization. Mice were fed with an irradiated standard rodent breeding diet and given reverse osmosis-filtered water ad libitum. Mice were kept in a facility with a 12-h light/12-h dark cycle and temperature and relative humidity maintained at 22 ± 1 °C and 55 ± 5%, respectively.

Open Access This article is licensed under a Creative Commons
Lck C20.23A/C20.23A and Lck K273R/K273R knock-in mice and Lck KO/KO mice were generated in the Czech Centre for Phenogenomics, IMG, using one-cell-stage embryos isolated from 3-to 5-week-old females mated with 9-to 35-week-old males and stimulated with 5 IU of pregnant mare serum gonadotropin (MSD Animal Health, Folligon PMSG) and 5 IU of human chorionic gonadotropin (Sigma, CG10) 45 . Both males and females were of the C57BL/6N strain (Charles River). The one-cell-stage embryos received a pronuclear microinjection of Cas9 mRNA (100 ng ml -1 ) and guide RNA (50 ng ml -1 ) together with single-stranded DNA templates (10 ng μl -1 ) and were implanted into 7-to 15-week-old (and above 35 g of body weight) foster mothers of the CD1 strain (Charles River). The founders were back-crossed on the C57BL/6J background for at least five generations. Sequences of the oligonucleotides for the generation of mice and their genotyping are shown in the Supplementary Table 6.

Cell counting and cell lines
Cells were counted using a Z2 Coulter Counter Analyzer (Beckman Coulter) or Cytek Aurora flow cytometer (Cytek).
Primary T cells were cultured in IMDM. Jurkat T cell lines 16,25 and T2-Kb cells (provided by E. Palmer, University Hospital Basel) were cultured in RPMI. HEK293 (provided by T. Brdicka, IMG) and MC-38 cells (provided by E. Palmer, University Hospital Basel) were cultured in DMEM. Medium was supplemented with 10% fetal bovine serum (FBS; Gibco), 100 U ml -1 penicillin (BB Pharma), 100 mg ml -1 streptomycin (Sigma-Aldrich) and 40 mg ml -1 gentamicin (Sandoz). HEK293 cells are listed in the register of cell lines that are known to be misidentified through cross-contamination or other mechanisms (iclac.org/ databases/cross-contaminations/) 46 , because there was a case of their confusion with HeLa cells. We can exclude such a misidentification in our culture based on the morphology of the cells and their adhesion on the tissue culture plastic, which are clearly distinct between these two lines and which were checked in each experiment.
The parental human Jurkat leukemic line in this study was the LCK KO line 16 expressing the OT-I TCR 25 . This line was transduced with human LCK variants (LCK WT , LCK C20.23A and LCK K273R ) containing a C-terminal FLAG tag in pMSCV-IRES-LNGFR and eventually with human CD8 variants (CD8α WT CD8β, CD8α C215.217A CD8β) in pMSCV 25 . Human CD8A and CD8B genes were cloned de novo from human blood cDNA. The human LCK-encoding sequence was provided by T. Brdicka (IMG). The respective mutations in LCK and CD8A were introduced by PCR mutagenesis (Supplementary Table 6). For the unsupervised analysis of DP maturation, thymocytes were stained with TCRβ, CD4, CD8α, CD5, CD69, CD24, CD25 and LIVE/ DEAD fixable near-IR dye, and 10 5 live DP thymocytes (CD4 + CD8α + ) were downsampled from each individual mouse and concatenated together. Unbiased dimensional reduction and clustering by FlowSOM plugin was performed in FlowJo software. Graphs were mapped based on FlowSOM map using the EmbedSOM plugin in FlowJo software.

Flow cytometry analysis
The samples were analyzed using a Cytek Aurora, BD LSRII or FACSSymphony flow cytometer. The data were analyzed using Flow Jo (version 10.6.2, BD Biosciences).

Ex vivo activation assay
The human lymphoblast T2-Kb cell line expressing murine H2-K b47 was used for antigen presentation to OT-I T cells or Jurkat cell lines bearing OT-I TCR. Splenocytes from Ly5.1 mice were used for activation of B3K508 T cells. The antigen-presenting cells were pulsed with indicated Article https://doi.org/10.1038/s41590-022-01366-0 concentrations of indicated peptides and cocultured with isolated T cells at a 1:2 ratio overnight. CD69 expression was detected by flow cytometry. The results were fitted with a log (agonist) versus response (percentage of CD69 + T cells) function (least squares method) using PRISM (GraphPad Software).

Cloning of Lck variants and transfection into HEK293 cells
Lck WT, CA and KR open reading frames were amplified from cDNA obtained from the thymi of respective mice. FLAG tag was C-terminally fused to the Lck WT, CA and KR and cloned into pXJ41 vector (provided by T. Brdicka, IMG) using EcoRI/XhoI. ZAP70-and CD25-TCRζ-encoding genes (provided by T. Brdicka, IMG) were subcloned into pXJ41 (Supplementary Table 6).
HEK293 cells were grown to ~50% confluency and transfected with LCK variants and either ZAP70-or CD25-TCRζ-encoding pXJ41 plasmid. Thirty micrograms of DNA was mixed with 75 μg of polyethylenimine in 0.5 ml of DMEM/0.5% FBS for 10 min at room temperature. The mixture was then added onto cells in 3 ml of DMEM/0.5% FBS. The medium was replaced with DMEM/10% FBS/antibiotics after 3 h. Samples were collected 24 h after the transfection.
Protein concentration was equalized using the Pierce BCA protein assay kit (Thermo Scientific). Proteins were denatured in 1× Laemmli sample buffer at 93 °C.
For the analysis of Jurkat cell activation by anti-TCR, Jurkat cells expressing LCK variants were starved for 30 min at 37 °C and stimulated for 2 min with anti-Jurkat TCR c305 supernatant (kindly provided by T. Brdicka) at 37 °C. Cells were then immediately lysed and denatured in 1× Laemmli sample buffer at 93 °C. The lysates were sonicated and used for immunoblotting.
The following primary antibodies were used for the analysis of basal and induced phosphorylation in primary T cells, Jurkat cells and HEK293 cells: anti-CD3-ζ (clone 6B10. Both immunoprecipitation samples and lysates were visualized with secondary goat anti-rabbit or goat anti-mouse conjugated with horseradish peroxidase ( Jackson ImmunoResearch Labs) on Azure c200 (Azure Biosystems) or Fusion Solo S (Vilber).

BM chimeras
BM was isolated from 6-to 8-week-old Lck WT/WT , Lck CA/CA , or Lck CA/KR mice and mixed with supporting BM cells from Ly5.1 mice in a 1:1 ratio. Two million cells were transferred to lethally irradiated Ly5.1/Ly5.2 heterozygous donor mice. The mice received a dose of 6 Gy in an X-RAD 225XL Biological irradiator (Precision X-Ray). T cell development was analyzed 8 weeks after transplantation by flow cytometry.

Listeria infection
LN T cells were isolated from B3K508 and OT-I mice. Cells were adoptively transferred to Ly5.1 congenic host mice. The following day, mice were injected with 5,000 colony-forming units of transgenic Lm expressing OVA, T4, G4, 3K, P5R and P2A antigens [49][50][51] . Expansion of the responsive cells was analyzed by flow cytometry 5 d after infection.

LCMV infection
LCMV (Armstrong) was obtained from D. Pinschewer (European Virus Archive Global). For batch production, hamster BHK-21 cells were infected at a multiplicity of infection of 0.01, and the virus-containing supernatant was collected 48 h after infection. Mice were infected by intraperitoneal injection of 2 × 10 5 plaque-forming units. Detection of LCMV in the spleen was performed by quantitative PCR with reverse transcription (RT-qPCR). Total RNA was isolated by TRIzol LS (Invitrogen, 10296010), and in-column DNase digestion was performed using an RNA Clean & Concentrator kit (Zymo Research), according to manufacturers' instructions. RNA was stored at −80 °C or transcribed immediately using RevertAid reverse transcriptase (Thermo Fisher Scientific, EP0442) with oligo(dT)18 primers (Thermo Fisher Scientific, SO131) according to the manufacturer's instructions. RT-qPCR was performed using LightCycler 480 SYBR green I master mix (Roche, 04887352001) and a LightCycler 480 II machine (Roche). All samples were measured in triplicates. The LCMV titers were quantified against a standard curve from cloned S-segment of LCMV in pBlueScript vector (Supplemetary Table 6). The quality of isolated RNA was tested by RT-qPCR analysis of an endogenous reference gene Eef1a1 (Supplemetary Table 6).

Tumor growth
The mouse MC-38 cell line derived from C57BL/6 colon adenocarcinoma 52 was transduced with ovalbumin protein-coding sequence via retroviral vector pMSCV-IRES-LNGFR. Five hundred thousand cells were Article https://doi.org/10.1038/s41590-022-01366-0 injected subcutaneously to the left side of the mouse. When Cd3e -/mice were used as hosts, 2 × 10 5 OT-I cells were injected intravenously 5 d after tumor injection. Tumor size was measured by caliper, and tumor volume was estimated using the following formula: V = (L × S 2 )/2, where L and S are the longest and shortest diameters, respectively. The endpoint was the tumor volume exceeding 500 mm 3 or the end of the experiment (day 22 after MC-38 injection for polyclonal mice and day 31 for Cd3e -/mice). The approved animal protocols stated that mice with a tumor volume of 500 mm 3 or larger must be killed, which was always followed.

Cell isolation from tumors
Tumors were excised from mice, cut into small pieces and incubated with 100 μg ml -1 Liberase (Roche, 5401020001) and 50 μg ml -1 DNAse I (Roche, 101104159001) in wash buffer (1% bovine serum albumin and 1 mM EDTA in HBSS without Ca 2+ /Mg 2+ ) at 37 °C and 350 r.p.m. shaking for 45 min. The mixture was resuspended with a 1,000-μl wide-bore pipette tip every 10 min. Undigested debris was removed by filtering through a 100-μm strainer. Cells were collected by centrifugation at 350g at 4 °C for 5 min. Pellets were resuspended in 10 ml of 40% Percoll (Cytiva, 17089101) in DMEM. Ten milliliters of 80% Percoll in DMEM was carefully laid on the bottom of the tube to create a gradient. Samples were centrifuged at 320g at ~21 °C for 23 min with minimal ascending/ descending rates. Lymphocytes present at the interphase were collected, centrifuged at 400g at 4 °C for 5 min and processed for flow cytometry analysis.
Fetal thymi were isolated from fetuses of embryonic age 15.5 and placed on the prepared filters and treated with corresponding concentrations of peptides. The medium was carefully exchanged every other day, and thymi were analyzed by flow cytometry on day 7.

Flow cytometry-based TCR-ligand k off rate assay
Samples from Lck-variant OT-I mice were multiplexed by combination of staining with PE-and PerCP-Cy5.5-conjugated CD45.2 antibodies. The cells were then stained with Streptactin (IBA Lifesciences, 6-5010-001) multimerized with Alexa Fluor 488-conjugated pMHCI K b -OVA molecules 53,54 . Samples were measured at 5 °C and after 30 s of measurement, and the same volume of cold 2 mM d-biotin was added. Dissociation of the antigen was measured for an additional 10 min. For analysis, Streptactin and monomer fluorescence values of CD8 + OT-I T cells were exported from FlowJo (10.6.2) to Prism (GraphPad Software). The t 1/2 was calculated by fitting the data with a one-phase exponential decay curve.

Tetramer binding
Tetramers were produced by refolding biotinylated monomers using streptavidin-PE conjugate in a molar ratio of 1:3. Streptavidin-PE (Thermo Fisher Scientific, S866) was added in three doses with a 20-min incubation on ice after each step. The following biotinylated monomers For the detection of antigen-specific T cells, the tetramers (~100 nM) were added to the cocktail of antibodies for surface markers with the exception of the staining with I-Ab-GP66 tetramer, which was performed at ~21 °C in RPMI/2% FBS for 1 h.
For the quantitative binding analysis, peripheral T cells were isolated from OT-I Rag2 KO/KO mice and incubated with a titrated dose of H-2K b -OVA-PE and H-2K b -T4-PE tetramers for 20 min on ice. The supernatant was replaced with PBS/2% FBS, and cells were immediately analyzed using a sample cooling system.

Electron microscopy
LN cells were stained with anti-CD4 (clone RM4-5, BioLegend, diluted 50×) and washed and stained with 6-nm Colloidal Gold-AffiniPure goat anti-rat (polyclonal, Jackson ImmunoResearch 112-195-167, diluted 15×) on ice. Before processing, cell suspensions were diluted in 20% bovine serum albumin and rotated at 250g at 4 °C for 5 min. Subsequent cryofixation was done using a Leica EM ICE high-pressure freezer. Approximately 1 μl of each cell suspension variant was put into each of four type A 3-mm high-pressure freezer carrier sandwiches, which were rapidly frozen and dehydrated using a Leica EM AFS2 automatic freeze substitution unit under temperature slowly increasing from −90 °C to 0 °C over 4 d in 100% acetone enriched with 0.2% uranyl acetate, 0.2% glutaraldehyde, 0.01% osmium tetroxide and 5% water. Samples were then removed from AFS2 and infiltrated with 100% ethanol on ice and then with Quetol 651 resin diluted in 100% ethanol at 4 °C. Afterward, cells were embedded in Quetol NSA resin. After polymerization for 72 h at 60 °C, resin blocks were cut into 80-nm ultrathin sections using a Leica UC6 Ultra microtome with a diamond knife (Diatome), collected on copper slots with formvar membrane and air dried. After additional contrasting with 2% uranyl acetate in water, sections were examined with a JEOL JEM-1400Flash transmission electron microscope operated at 80 kV equipped with a Matataki Flash sCMOS camera ( JEOL).
Electron microscopy images were analyzed using the open access application Pattern (pattern.img.cas.cz) developed by the Electron Microscopy Core Facility at IMG, Prague. Images were analyzed using one-dimensional analysis, where a region of interest was manually traced along the membrane. Size calibration was defined as 1.189 nm in 1 pixel. Clustering of gold particles was determined as pair correlation function value. Values were normalized to the predicted maximum standard deviation of the simulated pair correlation function value for analyzed density of individual cell staining.
Fresh LN T cells were decorated with a site-specific AF555 C2 maleimide-conjugated (Thermo Fisher Scientific) scFV ( J1), derived from the H57-597 monoclonal antibody that is reactive against the TCRβ chain (12.5 μg ml -1 ) 58 . Stained T cells were then allowed to approach functionalized planar supported lipid bilayers to visualize and quantitate TCR-pMHCI binding using FRET donor recovery after acceptor photobleaching. For this, an image of the donor channel (AF555-H57-scFV) was recorded before and after acceptor photobleaching, and the pixel-averaged fluorescence signal f pre (before bleaching) and f post (after bleaching), respectively, was calculated for each Article https://doi.org/10.1038/s41590-022-01366-0 T cell synapse. The FRET efficiency was then given by E = (f post − f pre )/ (f post - camera background). Bleaching time was 250 ms. Illumination time was 50 ms, with a 18-ms delay. The time lag between images of donor before and after bleaching was 550 ms.

Statistical analysis
No statistical methods were used to predetermine sample sizes, but our sample sizes are similar to those reported in previous publications 6,27,59 .
The mice and transgenic cells lines were allocated to experimental groups solely based on their genotype. If more experimental conditions were used in a single experiment (such as different Listeria strains), the allocation of the mice was random (that is, based on mouse ID in the database before the experimenter had any contact with them) in the way that sex-and age-matched animals with different genotypes were compared (preferably littermates). For the cell line transfection/ transduction experiments, identical cell culture aliquots of the split parental culture were used; thus, no randomization was required.
Allocation of mice was based solely on their genotype. The experimenter processed the mice based on their ID number (that is, without the information about the genotype). ID was matched with the genotype only during data analysis at the end of the experiment. Because no subjective scoring method was used, the analysis of the mice was not explicitly blinded. Ex vivo experiments with primary cells and cell lines were not blinded. Because no subjective scoring method was used for the analysis, blinding was not necessary.
Some rare experiments/samples were excluded because of technical failures based on preestablished criteria. The only sample excluded after analysis was one sample for LCMV titer determination (Fig. 3a), which had a very late reference gene amplification in the RT-qPCR assay (Lck WT/WT day 6 after infection).
Statistical analyses were performed using Prism 5 (GraphPad Software). All the statistical tests were two tailed, if applicable. Adjustments for multiple comparisons were performed if indicated in the respective figure legends. For comparison of individual groups, a two-tailed Mann-Whitney test was used. For multirank comparison of survival curves for tumor growth, a log-rank (Mantel-Cox) test was used, and for comparison of individual survival curves, a Gehan-Breslow-Wilcoxon test was used. For multirank comparison of CD69 upregulation curves, an extra sum of squares F-test was used to test differences in the maximum and/or half-maximum effective concentration (EC 50 ) values of the non-linear regression fits. For the comparison of normalized data (to the Lck WT/WT strain), a two-tailed Wilcoxon signed-rank test or a two-tailed one-sample t-test was used. The latter was used for the exceptional cases where the low sample size did not allow us to use non-parametric tests. Data distribution was assumed to be normal in this case, but this was not formally tested because it was not possible given the sample size.

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

Data availability
Mouse strains, cell lines and vectors generated within this work are available for non-commercial research purposes following a reasonable request. Materials transfer agreements will be required. Data of the TCR repertoire analysis are available in the Sequence Read Archive (PRJNA872031). Raw flow cytometry data and microscopy images are available following a reasonable request to the corresponding author. Source data are provided with this paper. All other data generated or analyzed during this study are included in this published article (and its Supplementary Information files).

Code availability
Code for TCR repertoire analysis is available on GitHub (https://github. com/Lab-of-Adaptive-Immunity/lck-tcrseq).    Fig. 2a-b. (d) Representative mice from Experiment 3 described in Fig. 2a- Fig. 2e-g. (a) The usage of TRAV and TRBV gene segments in the indicated mice. Each bar represents the average frequency of the gene segment among all CDR3 sequences in a particular group of samples. (b) A total percentage of typical TCRβ (TRB) gene segments used by NKT cells (TRBV1,TRBV13-2, and TRBV9) in the indicated samples from the indicated mice. (c) Percentage of TCRs containing TRAV11-TRAJ18 gene segments typically used by NKT cells in the indicated cells from the indicated mice. (d) Diversity of the repertoire of the TCRα (TRA) and TCRβ (TRB) CDR3 amino acid sequences in the indicated cells from the indicated mice calculated using the Chao1 richness estimator. CD4 and CD8 T-cell CDR3 amino acid sequences were analyzed together (left) or separately (right). Invariant NKT TRAV11-TRAJ18 CDR3 amino acid sequences were removed before this analysis. The statistical significance was calculated using a Kruskal-Wallis test with multiple comparison adjustment using the Holm method. Corresponding author(s): Ondrej Stepanek Last updated by author(s): Oct 17, 2022 Reporting Summary Nature Portfolio wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency in reporting. For further information on Nature Portfolio policies, see our Editorial Policies and the Editorial Policy Checklist.

Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly The statistical test(s) used AND whether they are one-or two-sided Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals) For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Portfolio guidelines for submitting code & software for further information.

Data
Policy information about availability of data All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: -Accession codes, unique identifiers, or web links for publicly available datasets -A description of any restrictions on data availability -For clinical datasets or third party data, please ensure that the statement adheres to our policy Data of the TCR repertoire analysis are available in the Sequence Read Archive (PRJNA872031). Raw flow cytometry data and microscopy images are available upon a reasonable request to the corresponding author. All other data generated or analyzed during this study are included in this published article (and its supplementary information files).