A single-amino acid substitution in the adaptor LAT accelerates TCR proofreading kinetics and alters T-cell selection, maintenance and function

Mature T cells must discriminate between brief interactions with self-peptides and prolonged binding to agonists. The kinetic proofreading model posits that certain T-cell antigen receptor signaling nodes serve as molecular timers to facilitate such discrimination. However, the physiological significance of this regulatory mechanism and the pathological consequences of disrupting it are unknown. Here we report that accelerating the normally slow phosphorylation of the linker for activation of T cells (LAT) residue Y136 by introducing an adjacent Gly135Asp alteration (LATG135D) disrupts ligand discrimination in vivo. The enhanced self-reactivity of LATG135D T cells triggers excessive thymic negative selection and promotes T-cell anergy. During Listeria infection, LATG135D T cells expand more than wild-type counterparts in response to very weak stimuli but display an imbalance between effector and memory responses. Moreover, despite their enhanced engagement of central and peripheral tolerance mechanisms, mice bearing LATG135D show features associated with autoimmunity and immunopathology. Our data reveal the importance of kinetic proofreading in balancing tolerance and immunity.

Mature T cells must discriminate between brief interactions with self-peptides and prolonged binding to agonists. The kinetic proofreading model posits that certain T-cell antigen receptor signaling nodes serve as molecular timers to facilitate such discrimination. However, the physiological significance of this regulatory mechanism and the pathological consequences of disrupting it are unknown. Here we report that accelerating the normally slow phosphorylation of the linker for activation of T cells (LAT) residue Y136 by introducing an adjacent Gly135Asp alteration (LAT G135D ) disrupts ligand discrimination in vivo. The enhanced self-reactivity of LAT G135D T cells triggers excessive thymic negative selection and promotes T-cell anergy. During Listeria infection, LAT G135D T cells expand more than wild-type counterparts in response to very weak stimuli but display an imbalance between effector and memory responses. Moreover, despite their enhanced engagement of central and peripheral tolerance mechanisms, mice bearing LAT G135D show features associated with autoimmunity and immunopathology. Our data reveal the importance of kinetic proofreading in balancing tolerance and immunity.
Adaptive T-cell immunity generates a highly diverse T-cell antigen receptor (TCR) repertoire for pathogen recognition that does not cause autoimmunity. The goal of TCR ligand discrimination is that agonist peptide bound to major histocompatibility complex (pMHC) triggers T-cell responses, whereas self-pMHC signals maintain T-cell survival 1,2 . Improper TCR ligand discrimination can cause autoimmunity and other immune-mediated diseases. Notably, TCR affinity for an agonist or self-pMHC may differ by only ten-to 15-fold 3 . In addition, TCR affinity for pMHC ligands is in the micromolar range, contrasting the binding affinities of B cell receptors, cytokine receptors and other receptor tyrosine kinases for their ligands, which are often in the nanomolar range 4 . These TCR characteristics make it difficult for T cells to reliably discriminate between self-and foreign pMHCs, to modulate the quality and quantity of resulting responses and to balance between immunity and tolerance. Several models have attempted to explain how T cells distinguish between self-peptides and foreign ligands, but in vivo evidence has been limited and the underlying mechanisms remain enigmatic. The kinetic proofreading model suggests that a series of signaling events, some of which may include nodes that function as critical Article https://doi.org/10.1038/s41590-023-01444-x which influences PLC-γ1 recruitment, phosphorylation and activation) is accelerated, endowing T cells with the ability to respond to weak or self-ligands in vitro (Extended Data Fig. 1b).

Expression of G135D LAT alters thymocyte development
Immature thymocytes require TCR signals of appropriate strength to complete thymic development. Since the thymic selection thresholds are differentially regulated in neonates and adults 24 , we analyzed polyclonal thymocyte development in LAT G135D knockin mice or wild-type mice at the neonatal (10-to 14-day-old) and adult (6-week-old) stages.
During the postnatal period, thymus size progressively increased in the wild-type mice, almost doubling between the neonatal and adult stages. In LAT G135D mice, the thymus was already enlarged at 2 weeks (Extended Data Fig. 2a). We observed less impact on the double-negative population (Extended Data Fig. 2b) but observed an effect on the double-positive (DP) and single-positive populations in LAT G135D mice ). In adult LAT G135D knockin mice, there were approximately 60% fewer single-positive CD4 (CD4SP) and single-positive CD8 (CD8SP) cells than in their wild-type littermates (Fig. 1a-c). Whereas the percentages of neonatal single-positive cells were also lower in LAT G135D mice , the absolute numbers remained comparable to those in neonate wild-type littermates (Fig. 1c). Notably, among the LAT G135D single-positive cells, the mature CD62L + H-2K b+ thymocytes ready for thymic egress were the most substantially affected population ) in mice of all ages.
To further investigate how altering the LAT Y136 kinetic proofreading step affected thymocyte development, we analyzed the DP thymocyte populations. DP cells gradually upregulate the expression of the TCR and the activation marker CD69 upon receipt of selecting signals, progressing from preselection DP1 (CD69 -TCR -) to midselection DP2 (CD69 med TCR med ) to postselection DP3 (CD69 hi TCR hi ) thymocytes. The expression of LAT G135D resulted in significantly lower frequencies and absolute numbers of DP3 cells in adult mice , which suggests that the Gly135Asp-induced defects in the CD4SP and CD8SP populations occurred at the DP2-to-DP3 thymocyte transition. Taken together, the data reveal that the expression of LAT G135D resulted in substantially smaller CD4SP and CD8SP thymocyte populations in LAT G135D mice and that immature adult thymocytes are more sensitive to LAT G135D -promoted signaling than neonatal cells.

LAT G135D expression triggers negative selection
To establish the cause of the smaller single-positive populations as defective positive selection, disrupted negative selection or death by neglect, we characterized the modifications in TCR signaling conferred by the Gly135Asp alteration. LAT G135D or wild-type preselection CD53thymocytes were isolated ex vivo and labeled with different dilutions of CellTrace Violet (Extended Data Fig. 3a). The cells were mixed and then stimulated with crosslinking anti-CD3ε antibodies. LAT G135D preselection cells exhibited a more rapid and much larger increase in cytoplasmic free calcium than that observed in wild-type preselection cells (Fig. 2a and Extended Data Fig. 3b). In contrast, wild-type preselection DP cells showed a slower, more sustained calcium increase ( Fig. 2a and Extended Data Fig. 3b). Immunoblot analysis of such ex vivo-stimulated thymocytes further demonstrated that the expression of LAT G135D led to enhanced phosphorylation of LAT Y136 and PLC-γ1 in preselection CD53thymocytes ( Fig. 2b and Extended Data Fig. 3c). Importantly, activation of the kinase Zap-70 (as evidenced by phosphorylation of Y493 in its activation loop) and phosphorylation of other LAT tyrosine residues (such as Y195) in LAT G135D thymocytes were comparable to levels in wild-type thymocytes ( Fig. 2b and Extended Data Fig. 3d). Similar results were observed in peripheral naive CD4 T cells (Extended Data Fig. 3e,f). These results suggest that the Gly135Asp alteration selectively increases the phosphorylation speed and magnitude of Y136 and PLC-γ1. molecular timers, set an activation threshold for T cells [5][6][7][8][9][10] . In essence, a ligand must bind to the TCR for long enough to initiate a series of reversible kinetic proofreading events to be considered a bona fide TCR activation signal [5][6][7][8][9][10] . Typically, when a self-pMHC engages a TCR, the binding lifetime is too short to initiate all of the necessary proofreading events to activate the T cell, although it may induce responses that contribute to cell survival 1 . Consistent with the kinetic proofreading model, adaptation to intrinsic signaling events and modification of the signaling network (for example, upregulation of programmed cell death protein 1 (PD-1) or other negative regulators) can fine-tune the activation threshold in T cells. However, it is not known how a single, specific kinetic proofreading step can influence primary T-cell function in vivo. Addressing this question requires the identification of a bona fide kinetic proofreading step.
We previously identified the tyrosine residue Y136 in mammalian linker for activation of T cells (LAT) as a molecular timer that modulates TCR ligand discrimination in T cells in vitro 11 . LAT Y136 is the only tyrosine residue that, upon phosphorylation, is able to recruit phospholipase C-γ1 (PLC-γ1) 7,12,13 . Importantly, PLC-γ1 signaling cascades activate the transcription factor nuclear factor of activated T cells (NFAT), which regulates the expression of essential development and activation genes 14,15 . Abolishing LAT Y136-mediated signals perturbs naive T-cell homeostasis and tolerance [16][17][18][19][20] . The response patterns and frequency of calcium-NFAT signals also dictate T-cell responsiveness during immune responses; for example, persistent NFAT signals may lead to T-cell exhaustion 21,22 . In addition to activating PLC-γ1 downstream signaling cascades 9,10 , Y136 has two other unique features among known Zap-70 phosphorylation sites in LAT: (1) it has markedly slower phosphorylation kinetics in vitro than other and (2) this is conferred by a glycine residue rather than an acidic residue preceding the substrate tyrosine 7,11,23 . Our recent data suggest that LAT Y136 phosphorylation constitutes an essential later kinetic proofreading step to support TCR ligand discrimination 11 . Moreover, they raise the question, 'How does LAT Y136 phosphorylation-mediated kinetic proofreading contribute to T-cell fate determination in vivo?'.
In the current study, we reveal the physiological importance and pathological consequences of tuning the phosphorylation speed of LAT Y136. We generated a mouse model featuring T cells with altered kinetic proofreading by replacing Gly residue 135 with a negatively charged Asp residue (LAT G135D ). In the thymus, LAT G135D T cells are subjected to increased negative selection. In the periphery, the expression of LAT G135D promotes specific phenotypic and functional adaptations, such as upregulation of CD5 and CD6 and induction of T-cell anergy. Strikingly, despite acquiring characteristics of enhanced tolerance in the thymus and in the periphery, LAT G135D T cells retain augmented sensitivity and proliferative fitness in response to infection with Listeria strains expressing very weak ligands. However, LAT G135D also promotes the terminal differentiation of antigen-specific CD8 T cells and impairs the formation of memory precursors in response to strong TCR stimuli. In addition, by 1 year of age, LAT G135D female mice develop higher titers of autoantibodies than wild-type mice, along with signs of colitis. Our work therefore suggests that the rate of LAT Y136 phosphorylation relative to the TCR:pMHC binding lifetime is a critical parameter of TCR ligand discrimination and contributes to T-cell fate determination upon antigen encounter.

Modifying T cells with altered TCR signal proofreading
To elucidate the consequences of disrupting a single bona fide kinetic proofreading step in an otherwise intact biological system, we utilized CRISPR-Cas9 technology to introduce a mutation into the endogenous mouse Lat locus that would be transcribed as the Gly135Asp alteration (Extended Data Fig. 1a-d). In LAT G135D mice, the kinetics of a single proofreading signaling step (that is, phosphorylation of LAT Y136, Article https://doi.org/10.1038/s41590-023-01444-x Next, we investigated TCR signaling in LAT G135D mice following physiologically relevant positive selection stimulation in the thymus [25][26][27] . Consistent with the in vitro biochemical findings (Fig. 2b), the LAT G135D T cells demonstrated stronger orphan nuclear hormone receptor Nur77 activation in vivo (Fig. 2c), probably due to encountered self-pMHCs, as indicated by flow cytometric analysis of LAT G135D Nur77-enhanced green fluorescent protein (eGFP) reporter bacterial artificial chromosome transgenic mice 28 . Notably, there were more eGFP + cells among post-DP2 LAT G135D thymocytes than among corresponding cells from wild-type littermates ( Fig. 2c and Extended Data Fig. 4a) and the geometric mean fluorescence intensity of eGFP was also higher in LAT G135D thymocytes than in wild-type cells ( Fig. 2c and Extended Data Fig. 4b). It was particularly noteworthy that twice as many LAT G135D compared with wild-type DP cells displayed a cleaved caspase-3 + (aCasp3 + ) and chemokine receptor CCR7 + phenotype, consistent with ongoing apoptosis due to clonal deletion of thymocytes migrating to the medulla 25,[29][30][31]e). Immunoblot analysis of both adult and neonatal LAT G135D DP thymocytes compared with wild-type counterparts further confirmed elevated expression of aCasp3 and proapoptotic Bcl-2 family member Bim (Fig. 2f).
Article https://doi.org/10.1038/s41590-023-01444-x (Extended Data Fig. 4d), there was a threefold increase in the aCasp3 + population in the adult LAT G135D mice compared with that in wild-type littermates (Fig. 2g,h) and an eightfold increase in LAT G135D neonates (Extended Data Fig. 4e,f). We measured the expression levels of CCR9 and CCR7 on TCR-signaled aCasp3 + thymocytes to approximate the effects of the alteration on the anatomic location and timing of clonal deletion. For both wild-type and LAT G135D thymocytes, roughly 60% of clonal deletion occurred in the cortex (Fig. 2i,j), consistent with previous reports 25,31,32 . Interestingly, in comparison with wild-type thymocytes, a larger proportion of LAT G135D thymocytes underwent clonal deletion at the semi-mature, proliferation-incompetent CCR9 + stage (Fig. 2i,k), which may explain the altered maturation pattern in LAT G135D mice (Fig. 1c,d). In addition, clonal deletion during CD4SP maturation usually correlates with failed regulatory T-cell (T reg cell) development 1 . We observed fewer T reg cells in LAT G135D mice (Extended Data Fig. 4g,h). These results indicate that the expression of LAT G135D promotes negative selection, possibly due to enhanced thymocyte reactivity to self-pMHC stimuli caused by augmented TCR-dependent LAT Y136-PLC-γ1 signaling.

LAT G135D promotes homeostatic proliferation
To investigate how T cells with altered kinetic proofreading and potentially enhanced self-reactivity respond in the periphery, we examined the phenotypic and functional characteristics of polyclonal peripheral LAT G135D CD4 and CD8 splenocytes. LAT G135D mice had fewer CD8 T cells (Extended Data Fig. 5a) than their wild-type littermates, whereas LAT G135D -expressing CD4 T cells were relatively less affected (Extended Data Fig. 5b). Nonetheless, both LAT G135D CD4 and CD8 populations included enlarged CD62L -CD44 + populations that were age dependent and obvious only in adult mice d).
Interestingly, the LAT G135D mice also harbored a substantial population of CD8 T cells that adopted a central memory-like phenotype (Fig. 3c,e and Extended Data Fig. 5e)-a population that is driven by higher self-reactivity 33,34 and exhibited enhanced responsiveness to lower-dose anti-CD3 stimulation (Extended Data Fig. 5f). Similar phenotypes were also observed in lymph nodes (Extended Data Fig. 5g,h).
To further investigate the mechanisms behind the altered cellularity in the periphery of LAT G135D mice, we adoptively transferred sorted naive LAT G135D CD4 T cells labeled with CellTrace Violet proliferation dye into sublethally irradiated congenic CD45.1 + C57BL/6 hosts to examine the homeostatic proliferation. After 4 days, we observed that LAT G135D CD4 T cells proliferated more robustly than wild-type CD4 T cells (Fig. 3f,g and Extended Data Fig. 5i); approximately 80% of LAT G135D CD4 T cells underwent proliferation compared with 49% of wild-type CD4 T cells (Fig. 3g). Naive LAT G135D CD8 T cells displayed similarly stronger proliferation than wild-type CD8 T cells (Fig. 3h,i and Extended Data Fig. 5j). Notably, transfer into sublethally irradiated MHC-II -/hosts, which prevents interaction with pMHC-II, rendered LAT G135D CD4 T cells nonproliferative (Fig. 3f,g). Restricting the repertoire of MHC-I-bound self-peptides by employing sublethally irradiated Tap1 -/-B2m -/hosts revealed similar self-pMHC-driven homeostatic proliferation of LAT G135D CD8 T cells (Fig. 3h,i). These data show that LAT G135D T cells exhibit enhanced reactivity/responsiveness to self-pMHCs, which contributes to their greater homeostatic proliferation potential.

LAT G135D T cells exhibit hyper-responsiveness to self-ligands
To more thoroughly study the effects of LAT G135D on self-pMHC reactivity, we introduced the LAT G135D mutation onto the OT-I TCR transgenic Rag1 -/background. LAT G135D .OT-I.Rag1 -/mice exhibited phenotypes consistent with those observed in polyclonal C57BL/6 mice, including a smaller CD8SP population (Extended Data Fig. 6a,b) and enhanced negative selection (Extended Data Fig. 6c,d) in the thymus, as well as augmented CD44 + populations in the periphery (Extended Data Fig. 6e-g). CD5 expression was also elevated in LAT G135D .OT-I.Rag1 -/-CD8 T cells, while the expression levels of OT-I TCR (Vα2), CD3 and CD28 were comparable to those of wild-type T cells (Extended Data Fig. 6h). Similar phenotypes resulting from LAT G135D alteration were also observed on OT-II.Rag1 -/-, SMARTA.Rag1 -/and AND.Rag1 -/-TCR transgenic backgrounds (Extended Data Fig. 7a-f).
To further test whether the expression of LAT G135D regulates T-cell ligand discrimination, we utilized four altered peptide ligands (APLs) and two self-peptides, Catnb and Cappa1, that are recognized by the OT-I TCR 35 . Using in vitro fetal thymic organ cultures (FTOCs) 36,37 , we observed that significantly fewer CD8SP cells developed in LAT G135D . OT-I.Rag1 -/-.Tap1 -/cultures than in wild-type LAT cultures treated with the agonist ovalbumin (OVA), or partial agonists Q4R7 or T4 ( Fig. 4a and Extended Data Fig. 7g). In contrast, treatment with the weak agonists V4 and G4 or the self-peptide Catnb promoted stronger positive selection of LAT G135D .OT-I.Rag1 -/-.Tap1 -/thymocytes, as indicated by a roughly twofold increase in CD8SP cells compared with those in wild-type LAT cell cultures ( Fig. 4a and Extended Data Fig. 7g). Further analysis of OVA APL two-dimensional (2D) affinity (Extended Data Fig. 7h,i), along with the frequency of CD8SP, showed that the expression of LAT G135D converts the borderline negative selectors (for example, T4 and Q4H7) into pure negative selectors and augments the selection efficiency of positive selectors (for example, V4, G4, Catnb and Cappa1) (Extended Data Fig. 7j).
Next, we isolated naive LAT G135D or wild-type LAT.OT-I.Rag1 -/peripheral CD8 T cells from 4-to 5-week-old mice (Extended Data Fig. 8a), stimulated the cells with OVA-or APL-pulsed antigen-presenting cells and examined the upregulation of CD69 (Fig. 4b). Whereas LAT G135D .OT-I.Rag1 -/-CD8 T cells responded only slightly more sensitively than wild-type OT-I.Rag1 -/-CD8 T cells to OVA or the partial agonists Q4R7, T4 and Q4H7, they responded with substantially greater sensitivity to the weak ligands V4 and G4 and self-peptides Catnb and Cappa1 (Fig. 4b). Plotting the potency by 2D (Extended Data Figs. 8b) or 3D (Extended Data Fig. 8c) affinity revealed that the expression of LAT G135D lowers the TCR discriminatory power (flattening the slope on the log-log plot) 4 , particularly in response to the weak ligands and self-peptides.
These weak ligands or self-peptides also promoted robust proliferative responses by naive LAT G135D .OT-I.Rag1 -/-CD8 T cells in contrast with wild-type cells, as revealed by the dilution of CellTrace Violet dye ( Fig. 4c and Extended Data Fig. 8d). Similarly, after culture with OVA-or APL-pulsed antigen-presenting cells, a significantly greater number of LAT G135D compared with wild-type LAT.OT-I.Rag1 -/-CD8 T cells were Ki-67 + cells the next day (Extended Data Fig. 8e,f). These Ki-67 + cells also exhibited upregulation of endogenous Nur77 (Extended Data Fig. 8f), which is evidence for TCR recognition-driven proliferation. In addition, the weak ligand V4 and self-peptide Catnb induced more LAT G135D versus wild-type LAT.OT-I.Rag1 -/-CD8 T cells to produce the cytokines interferon-γ (IFNγ) and tumor necrosis factor (TNF) (Fig. 4d). Interestingly, the expression of LAT G135D had the opposite effect on the production of interleukin-2 (IL-2) (Fig. 4e). Next, we generated cytotoxic T lymphocytes (CTLs) and found that LAT G135D .OT-I.Rag1 -/-CTLs mediated greater cytotoxicity against APL-pulsed EL4 cells at lower CTL-to-EL4 ratios than wild-type LAT.OT-I.Rag1 -/-CTLs (Extended Data Fig. 8g). Weaker ligands or self-peptides were also able to activate LAT G135D CD4 T cells expressing OT-II, SMARTA or AND TCRs (Extended Data Fig. 8h-j) to a greater degree than wild-type LAT CD4 T cells. These results suggest that the expression of LAT G135D may enable T cells to adopt a stronger effector cell program when challenged with weaker pMHCs or even self-pMHCs, suggesting that LAT G135D OT-I cells are less able to discriminate a true agonist from a weak agonist or even a self-pMHC.

LAT G135D facilitates the nuclear translocation of NFAT
To determine how the altered LAT Y136-centric kinetic proofreading step modulates the activation of specific transcription factors that are responsive to distinct signaling pathways, we examined the activation of transcription factors in isolated cell nuclei 38 from naive wild-type Article https://doi.org/10.1038/s41590-023-01444-x or LAT G135D .OT-I.Rag1 -/-CD8 T cells (Fig. 5). We first characterized nuclear NFAT1, which translocates from the cytoplasm to the nucleus following its dephosphorylation by the calcium-calmodulin-activated phosphatase calcineurin, the consequence of direct LAT-PLC-γ1-calcium downstream signaling. OVA stimulation induced rapid nuclear localization of NFAT1, and the expression of LAT G135D substantially CellTrace Violet CD5 BUV737 LAT G135D OT-I  Article https://doi.org/10.1038/s41590-023-01444-x promoted increased accumulation of NFAT1 in nuclei ( Fig. 5a and Extended Data Fig. 8k). At none of the responses of the wild-type cells did NFAT translocation equal that of the LAT G135D variant. NFAT signaling is necessary for the induction of transcripts of Nur77 (ref. 39 ), and the magnitude of the nuclear expression of Nur77 was also greatly enhanced in LAT G135D .OT-I.Rag1 -/-CD8 T cells (Fig. 5b). Interestingly, we did not observe substantial differences in nuclear translocation of nuclear factor-κB (NF-κB) (Fig. 5c) or Egr-2 ( Fig. 5d) between LAT G135D and wild-type LAT.OT-I CD8 T cells, which are regulated through costimulatory signaling in addition to TCR signals 40,41 . These data suggest that LAT G135D -promoted PLC-γ1 and calcium signals enhance the nuclear translocation of NFAT1 and transcriptional induction of Nur77, which is highly sensitive to NFAT, both of which may contribute to the hyper-responsiveness of LAT G135D T cells.

G135D LAT augments T-cell expansion to Listeria in vivo
To examine how these LAT G135D T cells balance tolerance and immune responsiveness, we used an immune challenge model. We adoptively transferred sorted CD62L + CD44naive CD45.2 + LAT G135D or wild-type LAT.OT-I.Rag1 -/spleen CD8 T cells into congenic CD45.1 + hosts (Extended Data Fig. 9a) and infected the mice with recombinant Listeria monocytogenes strains engineered to express OVA (Lm-OVA) or very weak APL V4 (Lm-V4) the next day 35 . On day 7 postinfection, LAT G135D . OT-I.Rag1 -/-CD8 T cells consistently expanded to a greater degree than wild-type OT-I.Rag1 -/-CD8 T cells in the Lm-V4 infection settings (Fig. 6a,b). Notably, OT-I TCR affinity to the V4 peptide is reported to be substantially weaker, within the range of characterized positively selecting APLs 5,35 . Lm-V4 infection resulted in the activation of only ~0.03% of wild-type T cells, but led to expansion of 0.15% of LAT G135D T cells (Fig. 6b). Interestingly, OT-I T cells expressing LAT G135D and wild-type OT-I T cells responded comparably to Lm-OVA, and activated LAT G135D . OT-I.Rag1 -/-CD8 T cells showed comparable cytokine production capacity to their wild-type counterparts (Extended Data Fig. 9b,c). These results are consistent with our in vitro data showing that modification of a kinetic proofreading step has a greater effect on weak ligand stimulation. In addition, infection with Lm-OVA emphasized the shift in effector versus memory cell fate decisions. We observed that the KLRG1 -CD127 + memory precursor population was substantially decreased by more than twofold among transferred LAT G135D .OT-I.Rag1 -/-CD8 T cells compared with transferred LAT wild-type cells in response to Lm-OVA infection (Fig. 6c,d and Extended Data Fig. 9d), whereas the short-lived KLRG1 + CD127effector cell population was consistently larger. Next, we investigated whether the enhanced proliferation of LAT G135D .OT-I.Rag1 -/-CD8 T cells was retained during recall responses. During rechallenge responses with vesicular stomatitis virus expressing OVA (VSV-OVA), LAT G135D .OT-I.Rag1 -/-T cells that had been primed with Lm-V4 maintained their expansion advantage (Fig. 6e). The skewed differentiation of KLRG1 -CD127 + versus KLRG1 + CD127cells was even more obvious upon rechallenge (Fig. 6f,g and Extended Data Fig. 9e). After rechallenge, when Lm-OVA-primed OT-I T cells were restimulated in vitro with the agonist OVA, LAT G135D .OT-I.Rag1 -/cells had inferior TNF production (Extended Data Fig. 9f)-a possible characteristic of terminally differentiated effector cells. Indeed, the T-cell factor-1-positive (TCF1 + ) population of LAT G135D .OT-I.Rag1 -/-T cells was also significantly smaller (Fig. 6h,i). These data suggest that the expression of LAT G135D augments sensitization of T cells to weak ligand stimuli in vivo and modulates cell fate decisions during immune responses.

LAT G135D female mice show signs of autoimmune pathology
To determine the effects of altered kinetic proofreading in older mice, we performed serological and histological analyses. Compared with aged wild-type littermate female mice from the same cohort, LAT G135D female mice from two cohort groups demonstrated nuclear staining for autoantibodies in indirect immunofluorescence assays (Fig. 7a). LAT G135D female mice also had higher titers of anti-double-stranded DNA (anti-dsDNA) antibodies in their sera by enzyme-linked immunosorbent assay (ELISA) (Fig. 7b) Article https://doi.org/10.1038/s41590-023-01444-x colons of wild-type female littermates (Fig. 7c). Interestingly, at 1 year of age, regulatory T cells in LAT G135D mice also exhibited enlarged TCF1 -CD62Lpopulations (Fig. 7d,e) and upregulated expression of CD44 and other phenotypic markers that are associated with effector regulatory T cells (Fig. 7f-h). Thus, these findings suggest that disruption of proper discrimination via perturbation of LAT Y136 phosphorylation results in hyper-responsiveness to self-ligands and the loss of proper maintenance of long-term tissue homeostasis, particularly at the barrier tissues.

Fig. 7 | Aged female LAT G135D mice develop higher titers of anti-dsDNA
IgG than wild-type counterparts, along with signs of colitis. a,b, Sera from aged wild-type or LAT G135D female mice (1 year old) were collected and subjected to antinuclear antibody staining (a) and anti-dsDNA IgG titers were measured by ELISA (b) (n = 15 for the wild type and n = 24 for LAT G135D ). ****P < 0.0001. The data are representative of two independent experiments. c, Histopathological analysis and hematoxylin and eosin (H&E) staining revealed signs of acute and chronic colitis, including abnormal neutrophil infiltration and crypt destruction/ distortion (arrowheads), in aged LAT G135D female mice that were absent from wild-type littermates. The data are representative of two independent experiments. Scale bars, 100 μm. d, Representative flow cytometry plots of the expression of CD62L and TCF1 on wild-type and LAT G135D CD25 + Foxp3 + regulatory T cells from 1-year-old mice. e, Bar graph summarizing the frequency of each subset as a proportion of total regulatory T cells. The regulatory T-cell subsets R1, R2 and R3 represent CD62L + TCF1 + , CD62L -TCF1 + and CD62L -TCF1cells, respectively. ***P = 0.0006 (left), ***P = 0.0002 (right) and NS = 0.1142. The data are representative of three independent experiments. f, Representative flow cytometry plots of the expression of CD62L and CD44 on wild-type and LAT G135D CD25 + Foxp3 + regulatory T cells from 1-year-old mice. g, Bar graph summarizing the frequency of CD44 hi and CD44 low regulatory T-cell populations as a proportion of total regulatory T cells. ***P = 0.0010 and ****P < 0.0001. The data are representative of three independent experiments. h, Expression of CD44, LEF1, ICOS and CD103 of wild-type and LAT G135D CD25 + Foxp3 + regulatory T cells from 1-year-old mice. In b, e and g, statistical significance was determined by two-tailed Mann-Whitney U-test.

LAT G135D T cells adapt in the periphery to maintain tolerance
LAT G135D -induced hyper-responsiveness did not result in spontaneous autoimmune disease in young adult mice. We wondered whether possible compensatory or adaptive mechanisms in the periphery prevented the autoimmune or autoinflammatory phenotypes we observed in older mice. Indeed, in the steady state, LAT G135D CD4 T cells expressed higher levels of key negative regulators of TCR-dependent T-cell responses, including Nur77, CD5, CD6, DGK-ζ and TOX (Fig. 8a). The expression levels of several well-known coinhibitory receptors were surprisingly unaffected in LAT G135D CD4 T cells, including PD-1, LAG-3, Tim-3, TIGIT and VISTA (Fig. 8b). LAT G135D CD4 T cells also developed an age-dependent anergy phenotype, as evidenced by an increase in Foxp3 -CD73 + FR4 + CD4 T cells in frequency (Fig. 8c,d) and in absolute number (Extended Data Fig. 10a) as mice aged from 2-6 weeks postnatally. These anergic CD4 cells failed to induce calcium increases in response to anti-CD3ε and anti-CD28 stimulation (Extended Data Fig. 10b) and did not upregulate CD25 or CD69 (Extended Data Fig. 10c), revealing their hyporesponsiveness. In addition, in LAT G135D .Nur77-eGFP reporter mice, the CD73 hi (Extended Data Fig. 10d) or FR4 hi cells (Extended Data Fig. 10e) were predominantly enriched in the Nur77-eGFP hi population, suggesting that continued TCR self-pMHC stimulation may have driven the emergence of the population. The CD73 + FR4 + LAT G135D CD4 T-cell population retained low expression of the activation marker PD-1 and high expression of the stemness regulator TCF1 (Extended Data Fig. 10f), consistent with clonal anergy rather than exhaustion [42][43][44] . IL-2 treatment 42,44 of the sorted CD73 + FR4 + wild-type or LAT G135D CD4 T cells at least partially reversed their unresponsive state (Extended Data Fig. 10g,h), and the formerly anergic LAT G135D T cells still mounted stronger responses than the formerly anergic wild-type T cells. Surprisingly, the frequency and size of the T reg cell population did not change significantly between wild-type and LAT G135D mice as they aged (Fig. 8e,f; absolute number in Extended Data Fig. 10i), nor did the expression of the transcription factor Helios vary between LAT G135D and wild-type T reg cells (Extended Data Fig. 10j). Interestingly, despite this, LAT G135D regulatory T cells displayed higher expression of PD-1, GITR, CD25 and Nrp-1 (Fig. 8g-i), which are markers reported to associate with superior suppressive function. Indeed, LAT G135D regulatory T cells exhibited stronger suppressive activities than wild-type regulatory T cells when cocultured with CellTrace Violet-labeled conventional CD8 T cells (Extended Data Fig. 10k,l). Taken together, our data suggest that LAT G135D augments self-pMHC sensitivity and may trigger intrinsic adaptive mechanisms to maintain peripheral tolerance ( Supplementary Fig. 1).

Discussion
A kinetic proofreading model was developed to explain the remarkable discriminatory power of TCR ligand recognition-a process central to T-cell fate decisions during development and immune responses 4,7,11 . However, until now, a lack of animal models allowing manipulation of a bona fide proofreading step has hampered our understanding of the importance of safeguarding TCR ligand discrimination in physiological and pathological settings. Here we generated a robust in vivo mouse model, harboring the LAT G135D alteration, in which T cells are hardwired to shorten the time delay for TCR-pMHC input signals to trigger activation. We showed that shortening the time of molecular engagement required for a key step in kinetic proofreading allows antigens with low signal strength that normally fail to generate effective T-cell responses to serve as activating signals. LAT G135D -expressing T cells engage robust central and peripheral tolerance and display heightened effector responses to pathogens. However, LAT G135D -mediated alterations also impair the formation of memory precursors and predispose female mice to features associated with autoimmunity. Thus, our findings suggest that the slow rate of LAT Y136 phosphorylation establishes a level of proper TCR ligand discrimination that allows T cells to scale responses accordingly to distinguish between ligands spanning a broad range of potencies and affinities. Our results emphasize the importance of slow phosphorylation of LAT Y136 to maintain T-cell unresponsiveness toward self-peptides and, therefore, tolerance. Editing to shorten the signaling delay after TCR:pMHC engagement revealed the importance of the evolutionarily conserved slow kinetics of the LAT Y136 proofreading step. The primary goal of thymic T-cell development is to generate an anticipatory T-cell repertoire of the greatest possible size to ensure efficient immune responses to foreign pathogens while precluding the development of autoimmunity. However, LAT G135D CD8 T cells exhibited restricted cell fates, with skewing toward effector cells, indicative of worse cell fate plasticity during immune responses. Such an imbalance in the ability of LAT G135D T cells to adopt various cell fates highlights the evolutionary fitness conferred by proper regulation of the TCR proofreading step. Future experiments are needed to examine the potential impact of the LAT G135D alteration upon CD8 T-cell memory responses and to explore the hypothesis that slow kinetic proofreading in mammals has created T cells that utilize TCR ligand discrimination to identify optimal agonistic signals, thereby retaining considerable plasticity to generate effector responses and form memory cells while retaining proper sensitivity to weak ligands.
T-cell ligand discrimination is particularly sensitive to the phosphorylation kinetics of LAT Y136 (among all Zap-70 substrates), plausibly because it is the sole tyrosine associated with PLC-γ1 interaction and function. Our data, together with complementary results by others on mice with a mutation conferring a Tyr136Phe alteration in LAT (LAT Y136F ) [16][17][18][19] -in which the recruitment and activation of PLC-γ1 are completely disrupted-provide an opportunity to identify the divergent signaling pathways propagated through different tyrosine residues in LAT. Selective disruption or enhancement of LAT Y136-PLC-γ1 signaling has only a modest effect on ERK signaling, allowing LAT Y136F to retain certain LAT signalosome functions 19 and LAT G135D to specifically tune PLC-γ1-specific signal transduction. Thus, we postulate that the slow phosphorylation associated with PLC-γ1 signaling and fast phosphorylation associated with Grb2/SOS and ERK/MAPK signaling enable TCR self-and nonself-discrimination to establish pathway specificity within LAT signalosomes. The LAT-PLC-γ1-calcium-NFAT pathway is, by default, the last to be activated after TCR engagement. Our single-amino acid modification of LAT uniquely facilitates NFAT signaling relative to other TCR downstream pathways, consequently altering T-cell development and homeostasis. As a result, LAT Y136 signal augmentation in LAT G135D mice results in superior induction of thymic negative selection and peripheral T-cell anergy-the tolerance mechanisms that LAT Y136F T cells fail to engage.
Interestingly, we did not observe spontaneous upregulation of coinhibitory receptors, including PD-1, in LAT G135D mice, perhaps because the coinhibitory receptors are more distal in the TCR pathway and may be most important to prevent immunopathology once an immune response has been initiated. However, signaling domains that either compete with the LAT signalosome or interact with LAT downstream signaling are upregulated. For example, upregulation of CD5 and CD6 may cause assembly of their respective signalosomes, leading to competition with LAT signalosomes for interacting proteins 45,46 . In addition, molecules that directly inhibit signaling downstream of the LAT-PLC-γ1 pathway, such as DGK-ζ 47 , are also upregulated. Together, our data reveal the elements of downstream TCR signaling that are specifically dependent on LAT-PLC-γ1-calcium-NFAT signals. These results show that the importance of the LAT Y136-PLC-γ1 pathway lies in its duality: it equips T cells with enhanced responsiveness and sensitivity while priming them for tolerance induction. The LAT-PLC-γ1 pathway requires conditions to be just right, as both signaling deficiency and hyperactivity can lead to immunodeficiency. Overall, our study demonstrates an important physiological role of a kinetic proofreading step and unmasks the importance of coordinating signal specificity within LAT signalosomes to reinforce proper TCR ligand discrimination.

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/s41590-023-01444-x.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
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 license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license 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 license, visit http://creativecommons. org/licenses/by/4.0/.

Analysis of thymic clonal deletion at the population level
Thymic clonal deletion was characterized by analysis of the expression levels of CCR7, CCR9 and cleaved caspase-3 (ref. 31 ). In brief, thymi were harvested and processed as quickly as possible to avoid nonspecific apoptosis. A total of 1 × 10 7 thymocytes were first stained with anti-CCR7 and/or anti-CCR9 antibody for 30 min at 37 °C, followed by staining of the surface markers CD5, TCRβ, CD4 and CD8. Thymocytes were then fixed with 4% freshly prepared paraformaldehyde (BeanTown Chemical) and washed with Perm/Wash buffer (BD Biosciences) twice. Next, thymocytes were stained with anti-cleaved caspase-3 (Asp175) at a 1:100 dilution for 30 min at room temperature. Cells were washed with Perm/Wash buffer and analyzed on an LSRFortessa system (BD Biosciences).

Analysis of intracellular calcium
Preselection thymocytes were obtained by depleting CD53 + cell populations. In brief, thymocytes were prepared and stained with 50 μl anti-CD53 antibody per 2 × 10 8 cells for 15 min on ice. After cells were washed twice with MCAS running buffer (Dulbecco's phosphate-buffered saline (PBS) supplemented with 0.5% bovine serum albumin (BSA) and 2 mM ethylenediaminetetraacetic acid (EDTA)), they were further stained with 20 μl anti-rat IgM biotin ( Jackson Immu-noResearch) per 2 × 10 8 cells for 15 min on ice. After they were washed twice, CD53thymocytes were enriched through a column, collected and counted. The purity of isolated preselection CD53thymocytes was confirmed by the staining of CD4, CD8, CD5, CD69 and TCRβ. The purity of cells was consistently >95%. Preselection CD53thymocytes were then loaded with 1 μM of the calcium indicator dye Indo-1 (Thermo Fisher Scientific) and 0.02% Pluronic F-127 (Thermo Fisher Scientific) at 37 °C in Roswell Park Memorial Institute (RPMI) medium supplemented with 5% FBS for 30 min, washed twice with PBS and then labeled with a 1:100 dilution of biotinylated anti-mouse CD3ε antibody on ice for 30 min. Cells were then used to perform flow cytometry-based calcium assays. Indo-1 cell-associated fluorescence was first recorded for 30 s to obtain a baseline and then monitored after additions of streptavidin (10 μg ml −1 ) at the 30th second. The calcium responses were recorded for a total of 5 min. Similarly, CD4 periphery T cells were enriched using a homemade purification kit (A.W. laboratory). CD4 T cells were then stained with the surface markers CD4, CD44, CD62L, CD73 and FR4 on ice for 30 min. Cells were washed twice and then subjected to the same experimental set up to perform experiments of calcium responses on an LSRFortessa.

In vitro T-cell stimulation assays
Naive or anergic mouse OT-I CD8 cells were sorted and cocultured at a 5:1 ratio with OVA or APL-pulsed TCR Cα -/splenocytes overnight over a titrated dose of antigens, as indicated in the figures, then plated at a concentration of 10 5 cells in round-bottomed 96-well plates in 200 μl complete RPMI media containing 10% fetal calf serum (FCS), 1× nonessential amino acids, 2 mM glutamine, 1 mM sodium pyruvate, 0.05% gentamicin and 50 μM 2-mercaptoethanol. The next day, cells were examined for their upregulation of CD69 and CD25, proliferation or cytokine production. For the proliferation assays, cells were labeled with 5 μM CellTrace Violet dye (Thermo Fisher Scientific)-a fluorescent dyes able to track proliferation-in PBS at 37 °C for 20 min in the dark. Labeled cells were incubated with complete culture medium at 37 °C for another 5 min, washed twice and cultured with peptide-pulsed splenocytes for 4 d. The proliferation ability of the cells was then analyzed by flow cytometry. As for the intracellular cytokine assays, cells were stimulated with OVA or APL-pulsed TCR Cα -/splenocytes overnight. The next day, 2 μM monensin (BioLegend) was added 4 h before the harvest of the cells. Cells were first stained with the surface markers CD4, CD8, CD25, CD69, CD73 and FR4, followed by fixation and permeabilization in Cytofix/Cytoperm (BD Biosciences) solution for 10-20 min on ice. The intracellular cytokines IFNγ, TNF and IL-2 were stained in Cytofix/Cytoperm (BD Biosciences) solution for 30 min on ice.

FTOC
The matings of wild-type or LAT G135D OT-I.Rag1 -/-.Tap1 -/mice were set up and timed to obtain embryos at an approximate gestational age of 16-17 d. The thymic lobes were harvested and cultured at 37 °C in RPMI medium supplemented with 10% FCS, 1× nonessential amino acids, 2 mM glutamine, 1 mM sodium pyruvate, 0.05% gentamicin and 50 μM 2-mercaptoethanol. After 4 d of incubation, the thymic lobes were harvested and stained for flow cytometry analysis.

Flow cytometry-based analysis of nuclear translocation of transcriptional factor
The base nuclei isolation and staining protocol was adapted from a previous LAT G135D OT-I T cells were sorted and stimulated with OVA-pulsed TCR Cα -/splenocytes for the indicated number of hours. After stimulation, cells were harvested and spun down at 300g and 4 °C and the pellets were immediately resuspended with 250 μl ice-cold PBS containing 320 mM sucrose (pH 7.4), 10 mM HEPES (Life Technologies), 8 mM MgCl 2 , 1× Roche EDTA-free cOmplete Protease Inhibitor (MilliporeSigma) and 0.1% (vol/vol) Triton X-100 (MilliporeSigma). After 15 min on ice, the plate was spun at 2,000g and 4 °C for 10 min. This was followed by two 250 μl washes with ice-cold PBS containing 320 mM sucrose (pH 7.4), 10 mM HEPES (Life Technologies), 8 mM MgCl 2 and 1× Roche EDTA-free cOmplete Protease Inhibitor (Millipore-Sigma) at 2,000g and 4 °C. After the final wash, pellets were fixed in 4% paraformaldehyde (electron microscopy grade; Electron Microscopy Sciences) and nuclei were rested on ice for 30 min for fixation, followed by two washes, resuspension in 1× PBS with 2% FBS and centrifugation at 1,000g to sufficiently pellet the nuclei. Nuclei were kept at 4 °C until flow cytometry analysis.

ELISA for serum anti-dsDNA
Serum was harvested from blood collected by lateral tail vein sampling or cardiac puncture postmortem. The serum anti-dsDNA titer was measured with a commercial ELISA kit, per the manufacturer's instructions (Alpha Diagnostic International). In brief, sera were added to plates coated with dsDNA. Anti-dsDNA titer was detected with anti-IgG-HRP. ELISA plates were developed and the absorbance was measured.

Antinuclear antibodies
Serum antinuclear antibodies (ANAs) were detected with a NOVA Lite HEp-2 ANA Substrate Slide, per the manufacturer's instructions, except for using FITC-conjugated donkey anti-mouse IgG secondary antibody. Images were captured with a Zeiss Axio Imager M2 widefield fluorescence microscope. Images were processed with ZEN pro (Zeiss). To measure the titer, the serum was serially diluted twofold from 1:40 to 1:1,280. HEp-2 ANA slides were stained with diluted serum. Images were read by a rheumatologist in a blinded manner and the titer was determined as the detectable lowest dilution of each sample. Notably, we selected only females to establish the aging cohort, to control for sex as a biological variable.

Immunoblot analysis
CD53or DP thymocytes were enriched as above. Thymocytes were washed with PBS and resuspended at 5 × 10 6 cells per ml and rested for 30 min at 37 °C. Cells were labeled with biotinylated anti-CD3ε (clone 2C11) at the indicated concentration. Cells were left unstimulated or stimulated with the addition of streptavidin over time, as described for each experiment. Cells were lysed by directly adding 10% NP-40 lysis buffer to a final concentration of 1% NP-40 (containing inhibitors of 2 mM NaVO 4 , 10 mM NaF, 5 mM EDTA, 2 mM phenylmethylsulfonyl fluoride, 10 μg ml −1 aprotinin, 1 μg ml −1 pepstatin and 1 μg ml −1 leupeptin). Lysates were placed on ice and centrifuged at 13,000g to pellet cell debris. Supernatants were run on NuPAGE 4-12% Bis-Tris Protein Gels (Thermo Fisher Scientific) and transferred to polyvinylidene difluoride membranes. Membranes were blocked using Tris-buffered saline with 0.1% Tween 20 detergent buffer containing 3% BSA and then probed with primary antibodies, as described for each experiment, overnight at 4 °C. The following day, blots were rinsed and incubated with horseradish peroxidase-conjugated secondary antibodies ( Jackson ImmunoResearch). Blots were developed using a chemiluminescent substrate and a Bio-Rad ChemiDoc imaging system (Bio-Rad).

2D K d affinity measurement
The relative 2D affinities of naive OT-I H-2K b CD8 T cells were measured using the previously characterized 2D-MP. Briefly, red blood cells were coated with biotin-LC-NHS (BioVision) and streptavidin (Thermo Fisher Scientific), together with either the biotinylated pMHC wild-type OVA (SIINFEKL WT (OVA)) or an OVA APL monomer (SIIQFERL (Q4R7), SIITFEKL (T4), SIIQFEKL (Q4), SIIQFEHL (Q4H7), SIIVFEKL (V4), SIIGFEKL (G4) or SIIRFEKL (R4)) and mouse β2-microglobulin (National Institutes of Health Tetramer Core Facility). To specifically investigate the TCR:pMHC interaction, monomers were generated with a H-2K b a3 domain with a human HLA-A2 a3 domain to mitigate CD8 coreceptor binding. A red blood cell coated with the monomer of interest and a T cell of interest were mounted onto opposing micropipettes. An electronically controlled piezoelectric actuator repeated a T-cell contact and separation cycle with the pMHC-coated red blood cell 50 times while keeping the contact area (A c ) and time (t) constant. Following retraction of the cell, binding of the TCR:pMHC was observed as a distention of the red blood cell membrane using an inverted microscope, allowing for quantification of the adhesion frequency (P a ) at equilibrium. Surface pMHC (ml) and TCR (mr) densities were determined by flow cytometry using anti-TRCβ PE antibody (clone H57-597; BD Biosciences) and PE anti-mouse β2-microglobulin antibody (clone A16041A; BioLegend), both at saturating concentrations, along with BD Quantibrite PE beads for standardization (BD Biosciences). The relative 2D affinity was calculated using the following equation: 2D affinity (A c K a ) = −ln[1 − P a (2s)]m r m l . The 'm r ' indicates TCR surface density and 'm l ' indicates ligand (pMHC) surface density. The P a (2s) means the adhesion frequency (or probability of adhesion) at 2 second (s) contact time.
Mice were infected intravenously with 1,000 colony-forming units of Listeria in log phase. Recombinant vesicular stomatitis virus (Indiana strain) expressing SIINFEKL (N4) was grown and titrated on baby hamster kidney cells. Mice were infected intravenously with 2 × 10 6 plaque-forming units.

Quantification and statistical analysis
Statistical analysis was applied to technical replicates or biologically independent mice for each experiment. All experiments described in this study have been performed at least twice and the exact numbers of independent experiments with similar results are indicated in the figure captions. All statistical analyses of experiments were performed using nonparametric, two-tailed Mann-Whitney U-tests. Image Lab (Bio-Rad) version 5.2.1 built 11 was used to acquire immunoblot data and BD FACSDiva version 8.0.1 software was used for flow cytometry. FlowJo version 9.9.3 or 10.8.1 was used for flow cytometry data analysis. SnapGene software version 4.0.8 was used to analyze DNA sequences or Sanger sequence data. GraphPad Prism 7 or 9 software (GraphPad Software) was used for data analysis and representation. All bar graphs show means with overlaid scatter dots or error bars (indicating s.d.) to show the distribution of the data, as indicated in each figure caption. P values for comparisons are provided as exact values or as P < 0.0001. 95% confidence levels were used to determine statistically significant P values. No statistical methods were used to predetermine sample sizes but our sample sizes were similar to those reported in previous publications. The data met the assumptions of the statistical tests used. Data distributions (individual data points) have been shown in all figures when applicable. Data distributions were assumed to be normal but this was not formally tested. No randomization was used in the experiments. In the animal experiments, age-matched animals were allocated based on their genotypes. In cell stimulation experiments, cells with the same genotype were pooled together and equally allocated into different groups before treatments. Data collection and analysis were not performed blind to the conditions of the experiments, except for the autoantibody ELISA and staining analysis and the hematoxylin and eosin staining-based immunopathology analysis. No data points or animals were excluded from the analysis. Article https://doi.org/10.1038/s41590-023-01444-x Extended Data Fig. 1 | Generation of LAT G135D knock-in mouse. a. The cartoon illustrates the genomic locus of mouse Lat. The numbers represent individual exons. The nucleotide and amino acid sequences of exon 7 are depicted. The coding regions are listed in capital letters and the intron immediately following exon 7 is shown in lowercase letters. The orange arrows represent the two sgRNAs used to generate the G135D knock-in mice. b. Illustration of the molecular mechanisms underlying the engineering strategy. In brief, mammalian T cells express natural (wild-type) LAT, which exhibits slow phosphorylation kinetics upon TCR recognition of ligand; this serves as a proofreading bottleneck to create the molecular time delay required for proper TCR ligand discrimination. Only bona fide activating ligands that interact with the TCR with a sufficiently long bond lifetime pass this slow signaling bottleneck to activate T cells (left). On the other hand, the bond lifetime of an interaction between the TCR and self-pMHC is too short to activate T cells (left). Importantly, the phosphorylation of LAT Y136 is regulated by the amino acid preceding Y136. Natural LAT has a small glycine at the −1 position, leading to slow phosphorylation (right); sequence-modifying LAT with a negatively charged aspartate substantially facilitates the rate and magnitude of Y136 phosphorylation (right). G135D mutant LAT therefore bestows on T cells the ability to respond to very weak ligands or self-peptides. c. The PCR genotyping results of the seven pups born to the founder generated with CRISPR/Cas9 using sgRNA #1 and the four pups born to the founder generated with sgRNA #2. Homology-directed repair (HDR) initiated by electroporation provided the repair template for CRISPR/Cas9 and was used as a positive control for PCR screening. gDNA from C57BL/6 parents was used as a negative control. The numbers along the top of the gel represent the individual pups. Pups #1 and #2 from the sgRNA #2 experiments are founders #1 and #2, respectively. The genotyping result was performed once but confirmed with Sanger sequencing analysis. d. Four-color chromatograms of the exon 7 gDNA sequence analyses of founders #1 and #2.
Article https://doi.org/10.1038/s41590-023-01444-x Extended Data Fig. 10 | Enlarged G135D anergic T cell populations are functionally hyporesponsive, but IL-2 treatment can restore their function. a. Bar graphs show the absolute number of CD73 + FR4 + CD4 T cells in the periphery at different ages. Data are representative of at least five experiments. Each dot represents one mouse. n = 20. ****P < 0.0001; ns = 0.1894. Two-tailed Mann-Whitney test. wk: weeks. b. Wild-type or G135D LAT CD4 T cells were isolated and stained with antibodies against CD62L, CD44, CD73, and FR4, and then loaded with calcium dye Indo-I and labeled with biotinylated anti-CD3. Wild-type and G135D CD4 T cells were barcoded with different titrations of CellTrace Violet and pooled together, allowing simultaneous analysis of the cells' calcium responses upon anti-CD3 crosslinking. Ionomycin treatment served as a positive control. Calcium traces recorded over 5 min are shown. Data are representative of three independent experiments. c. Naive or anergic wild-type or G135D LAT CD4 T cells were sorted, and stimulated with plate-bound anti-CD3 and soluble anti-CD28 monoclonal antibodies overnight. The upregulation of CD69 and CD25 was analyzed the next day. Representative pseudocolor contour plots are shown. Data are representative of three independent experiments. d,e. Representative flow cytometry plots show the expression of Nur77-eGFP and CD73 (d) or FR4 (e). Data are representative of at least five independent experiments. f. Representative histograms depict the expression levels of PD-1 and TCF1 in CD73 + FR4 + Foxp3anergic CD4 T cells isolated from wild-type or G135D LAT knock-in mice at 2 or 6 weeks of age. Data are representative of two independent experiments. g. CD73 + FR4 + Foxp3anergic wild-type or G135D LAT CD4 T cells were sorted and treated with 2 ng/ml or 5 ng/ml recombinant murine IL-2 (rmIL-2; concentration as indicated) overnight and then analyzed for calcium responses. Anergic T cells were labeled with biotinylated anti-CD3 and loaded with calcium dye Indo-I. Calcium responses to anti-CD3 crosslinking were analyzed by flow cytometry for 5 min. Representative calcium traces are shown. Data are representative of three independent experiments. h. CD73 + FR4 + Foxp3anergic wild-type or G135D LAT CD4 T cells were sorted and stimulated with plate-bound anti-CD3 and soluble anti-CD28 overnight along with the addition of 2 ng/ml or 5 ng/ml rmIL-2 (concentration as indicated). The upregulation of CD69 and CD25 were analyzed the next day. Representative flow pseudocolor plots are shown. Data are representative of three independent experiments. i. Bar graphs show the absolute number of CD25 + Foxp3 + CD4 T cells in the periphery at different ages. Data are representative of at least five experiments. Each dot represents one mouse. n = 15. *P = 0.0164; ns = 0.1261. Two-tailed Mann-Whitney test. j. Bar graphs show the percentages of Helios + cells among CD25 + Foxp3 + CD4 T cells in adult wild-type and G135D mice. Data are representative of at least five experiments. Each dot represents one mouse. n = 15. ns = 0.6312. Two-tailed Mann-Whitney test. k,l. Wild-type or G135D regulatory T (Treg) cells were sorted from wild-type (WT) or G135D Foxp3-RFP + mice. Polyclonal naive CD8 T cells (Tconv) from CD45.1 + C57BL/J mice were purified and labeled with CellTrace Violet dyes and used as responsive cells. CellTrace Violet-labled CD45.1 + Tconv cells were co-cultured with titrated ratios of regulatory T cells as indicated. Inhibition of Tconv cell proliferation was used as a readout for Treg cell suppressive function. Representative histograms of Tconv cell proliferation are shown in k. Bar graphs in l summarize the proliferation of Tconv cells and the suppressive activity of Treg cells. *P = 0.0286; ns = 0.8857. Two-tailed Mann-Whitney test. Data are representative of five independent experiments. Data are presented as mean values ± SD.