1. Immunology and Inflammation
Download icon

CD8+ T cell self-tolerance permits responsiveness but limits tissue damage

  1. Emily N Truckenbrod
  2. Kristina S Burrack
  3. Todd P Knutson
  4. Henrique Borges da Silva
  5. Katharine E Block
  6. Stephen D O'Flanagan
  7. Katie R Stagliano
  8. Arthur A Hurwitz
  9. Ross B Fulton
  10. Kristin R Renkema  Is a corresponding author
  11. Stephen C Jameson  Is a corresponding author
  1. Center for Immunology, University of Minnesota, United States
  2. Minnesota Supercomputing Institute, University of Minnesota, United States
Research Article
  • Cited 0
  • Views 761
  • Annotations
Cite this article as: eLife 2021;10:e65615 doi: 10.7554/eLife.65615

Abstract

Self-specific CD8+T cells can escape clonal deletion, but the properties and capabilities of such cells in a physiological setting are unclear. We characterized polyclonal CD8+ T cells specific for the melanocyte antigen tyrosinase-related protein 2 (Trp2) in mice expressing or lacking this enzyme (due to deficiency in Dct, which encodes Trp2). Phenotypic and gene expression profiles of pre-immune Trp2/Kb-specific cells were similar; the size of this population was only slightly reduced in wild-type (WT) compared to Dct-deficient (Dct-/-) mice. Despite comparable initial responses to Trp2 immunization, WT Trp2/Kb-specific cells showed blunted expansion and less readily differentiated into a CD25+proliferative population. Functional self-tolerance clearly emerged when assessing immunopathology: adoptively transferred WT Trp2/Kb-specific cells mediated vitiligo much less efficiently. Hence, CD8+ T cell self-specificity is poorly predicted by precursor frequency, phenotype, or even initial responsiveness, while deficient activation-induced CD25 expression and other gene expression characteristics may help to identify functionally tolerant cells.

Introduction

Accurate discrimination between harmful (pathogens, toxins, cancerous cells) and non-harmful entities (self, innocuous environmental components, non-pathogenic microbes) underlies effective functioning of the immune system. Understanding the mechanisms that normally enforce immunological tolerance to self is a prerequisite for safely and effectively manipulating the immune system to therapeutically induce or break self-tolerance.

Tolerance can be mediated by the clonal deletion of developing self-reactive T cells (Hogquist et al., 2005; Kappler et al., 1987). Largely based on studies in transgenic mouse models, this process has long been regarded as common and highly efficient (Palmer, 2003). However, recent studies have revealed that thymic clonal deletion is less effective than previously thought (Richards et al., 2016). Self-reactive CD8+ T cells have been shown to escape negative selection in mice (Bouneaud et al., 2000; Zehn and Bevan, 2006), with one group proposing that up to 4% of peripheral CD8+ T cells are self-specific (Richards et al., 2015). Furthermore, studies in humans indicated that precursor frequencies of blood CD8+ T cells specific for certain self-peptides were comparable to those demonstrated for foreign peptides (Yu et al., 2015) and suggested that such cells might be capable of overt autoreactivity if suitably stimulated (Maeda et al., 2014).

Aside from clonal deletion, tolerance mechanisms include ignorance of antigen, suppression by regulatory T cells (Tregs), and induction of a functionally unresponsive or hyporesponsive anergic state (Mescher et al., 2007; Mueller, 2010; Redmond and Sherman, 2005; Schietinger and Greenberg, 2014). However, different models have produced conflicting evidence regarding the contribution of each of these mechanisms and whether non-deletional CD8+ T cell tolerance is an intrinsic property of tolerant cells (Yu et al., 2015) or dependent on restraint by Tregs (Maeda et al., 2014; Richards et al., 2015). It is also unclear how the presence and reactivity of self-specific CD8+ T cells relates to their ability to drive immunopathology. The majority of commonly used mouse models of tolerance have the drawbacks of relying on T cell receptor (TCR) transgenic animals that may not recapitulate normal physiology or utilizing in vitro analyses for characterization of functionality.

Our studies are intended to provide a better understanding of non-deletional CD8+ T cell tolerance by utilizing a more physiological and translationally relevant mouse model in which an epitope from the melanocyte differentiation enzyme tyrosinase-related protein 2 (Trp2) is recognized by CD8+ T cells as either self or foreign. Trp2, an enzyme involved in melanin biosynthesis encoded by the dopachrome tautomerase (Dct) gene, is normally expressed by melanocytes in the skin in both humans and C57BL/6 mice and is overexpressed by many melanomas (Avogadri et al., 2016; Wang et al., 1996). Using wild-type (WT) mice and a novel Dct-deficient (Dct-/-) strain, we compared responses to Trp2180–188/Kb (Trp2/Kb) as a self- versus foreign antigen. This model is relevant to human health, as Trp2 is a common target in cancer immunotherapy directed against melanoma (Cho and Celis, 2009; Liu et al., 2014; Parkhurst et al., 1998), and Trp2/Kb-specific responses can be induced in WT mice with vigorous priming approaches (Bowne et al., 1999; Byrne et al., 2011; Cho and Celis, 2009). Instead of utilizing TCR transgenic models, we focus on the polyclonal Trp2/Kb-specific CD8+ T cell repertoire to maximize applicability to normal physiology.

Here, we show that tolerance among Trp2/Kb-specific CD8+ T cells is manifest primarily at the level of minimizing overt autoimmunity; differences in the size and initial Trp2 responsiveness of the precursor pool in WT and Dct-/- mice were relatively modest, although there was evidence for ‘pruning’ of cells with the highest avidity Trp2/Kb-specific TCRs from the WT precursor population. The underlying tolerance mechanism does not depend on cell-extrinsic regulation but rather correlates with a cell-intrinsic failure of WT Trp2/Kb-specific CD8+ T cells to sustain optimal proliferation. However, while differences in the responsiveness of WT and Dct-/- cells to Trp2 immunization were mostly subtle, a notable difference emerged when the cells were assessed for their ability to provoke autoimmune vitiligo: cells primed in Dct-/- mice were much more effective than those primed in WT animals. Accordingly, we conclude that tolerance in this setting does not lead to marked changes in the presence, phenotype, or initial reactivity of Trp2/Kb-specific CD8+ T cells but limits these cells’ capacity for overt autoreactivity. Moreover, our polyclonal model reveals that certain characteristics of Trp2/Kb-responsive effector cells—reduced CD25 expression and impaired differentiation into a highly proliferative subpopulation—correlate with functional tolerance of a T cell population, providing a framework for future characterization of self-specific CD8+ T cells.

Results

The pre-immune population of Trp2/Kb-specific cells is similarly sized and appears naïve in WT and Dct-/- mice

Clonal deletion is a well-studied tolerance mechanism that may result in the near-total culling of self-specific cells or a reduction in the number and apparent TCR affinity of surviving self-specific cells (Bouneaud et al., 2000; Cheng and Anderson, 2018; Enouz et al., 2012; Hogquist et al., 2005; Zehn and Bevan, 2006). We assessed the number and phenotype of Trp2/Kb-specific cells in pre-immune (naïve) mice to examine deletional central tolerance in our model. For mice in which Trp2 would not be a self-antigen, we used a novel Dct-/- strain that carries a large deletion encompassing the exon (exon 2) encoding Trp2180–188 (Figure 1—figure supplement 1, Figure 1—source data 2 and 3), unlike a previously described Dct-targeted strain that retains the coding sequence for that epitope (Guyonneau et al., 2004). We performed tetramer enrichment from the spleen and lymph nodes of pre-immune WT and Dct-/- mice to quantify the number of Trp2/Kb-specific CD8+ T cell precursors (Figure 1A, Figure 1—figure supplement 2A). While on average we identified more antigen-specific cells in Dct-/- mice (~1.4-fold increased), we nevertheless found relatively large numbers of Trp2/Kb-specific CD8+ T cells (>1500) in both strains, evidence that most Trp2/Kb-specific cells escape thymic or peripheral clonal deletion in WT mice.

Figure 1 with 2 supplements see all
Trp2/Kb-specific CD8+ T cells in pre-immune wild-type (WT) and Dct-/- mice share a naïve phenotype while showing modest differences in frequency and tetramer staining.

(A) Tetramer enrichment was performed to enumerate Trp2/Kb-specific CD8+ T cells per mouse. Median tetramer fluorescence intensity (MFI) was used to estimate the avidity of enriched Trp2/Kb-specific (B) or B8R/Kb-specific cells (C). (D) CD44/CD122 expression of Trp2/Kb-specific cells. (E) CD69 expression of Trp2/Kb-specific cells. (F) PD-1, CTLA-4, LAG3, and TIM3 expression of Trp2/Kb-specific cells. (G) CD5 expression of Trp2/Kb-specific cells. Data are compiled from three independent experiments in A and D. Individual experiments are shown in B and C; results are representative of other experiments. The graphs in (E, F, and G) represent individual experiments with four mice per group. Squares indicate male animals. **p<0.01 by unpaired t test.

Figure 1—source data 1

Data file related to Figure 1.

https://cdn.elifesciences.org/articles/65615/elife-65615-fig1-data1-v2.pzfx
Figure 1—source data 2

Primers used to screen prospective clones for homologous recombination of the Dct knockout construct.

https://cdn.elifesciences.org/articles/65615/elife-65615-fig1-data2-v2.docx
Figure 1—source data 3

Primers used to screen mice to confirm deletion of the Dct gene.

https://cdn.elifesciences.org/articles/65615/elife-65615-fig1-data3-v2.docx

In some systems, T cells bearing TCRs with low-affinity for self-antigens avoid deletion; low-affinity TCRs can often be identified by reduced peptide/MHC tetramer binding to these cells (Bouneaud et al., 2000; Daniels and Jameson, 2000; Daniels et al., 2006; Enouz et al., 2012; Hogquist et al., 2005; Zehn and Bevan, 2006; Zehn et al., 2009). We compared the Trp2/Kb tetramer median fluorescence intensity (MFI) in pre-immune WT and Dct-/- mice. The average Trp2/Kb tetramer staining was higher on Dct-/- cells, but the MFI largely overlapped between the two populations (Figure 1B), suggesting that the range of TCR avidities did not markedly differ between the Trp2/Kb-specific pools. Indeed, the tetramer staining difference we observed (a WT:Dct-/- tetramer MFI ratio of ~0.8) is more subtle than that noted in a previous study using a transgenic mouse model, which reported a tetramer ratio of ~0.35 between mice with versus without self-antigen expression (Bouneaud et al., 2000). As a control to make sure there were no staining differences between the strains, we also assessed the avidity of cells specific for an irrelevant foreign epitope—B8R/Kb from vaccinia virus—in WT and Dct-/- mice; the tetramer MFI of B8R/Kb-specific cells was comparable between the strains as expected (Figure 1C).

We also examined the phenotype of Trp2/Kb-specific cells in pre-immune WT and Dct-/- mice. No consistent differences in the expression of activation/memory markers (CD69, CD44, CD122) or anergy/exhaustion markers (PD-1, LAG3, CTLA-4, TIM3) were identified between Dct-/- and WT Trp2/Kb-specific cells (Figure 1D–F, Figure 1—figure supplement 2B–D). The majority of cells exhibited low expression of the memory markers CD44 and CD122, and anergy/exhaustion marker expression was low in both populations. CD5 can indicate self-antigen recognition (Azzam et al., 1998; Fulton et al., 2015), but we did not detect significant differences in expression between the groups (Figure 1G). RNAseq analysis of Trp2/Kb tetramer-binding cells isolated from pre-immune mice by fluorescence-activated cell sorting (FACS) showed no consistent differences in gene expression related to their derivation from WT versus Dct-/- mice (Figure 1—figure supplement 2E), although this does not rule out the possibility of epigenetic differences between the populations.

To ensure our findings were not unique to Trp2180/Kb-specific cells, we used tetramer enrichment to isolate CD8+ T cells specific for other skin antigens—a distinct Trp2 epitope (Trp2363/Db) and a tyrosinase-related protein 1 epitope (Trp1455/Db)—in mice expressing or lacking these antigens. We were able to identify cells with these specificities present at numbers similar to slightly less in mice expressing antigen relative to mutant mice (Figure 1—figure supplement 2F). This suggests that CD8+ T cells specific for other melanocyte self-epitopes also largely escape clonal deletion.

Hence, although we identified some minor differences between Trp2/Kb-specific cells from WT versus Dct-/- mice, these pre-immune populations generally resembled each other in number, phenotype, and gene expression, arguing against extensive clonal deletion or overt steady-state anergy induction as the mechanisms dictating tolerance to this antigen. These findings resonate with studies in humans, which have shown that the precursor frequency and average peptide/MHC tetramer staining intensity are only modestly reduced (or ‘pruned’) among self-antigen-specific CD8+ T cells (Yu et al., 2015) and that self-specific cells can be phenotypically naïve (Maeda et al., 2014; Yu et al., 2015). Accordingly, these data suggested that analysis of Trp2/Kb-specific responses in mice could serve as a useful model for investigating the characteristics and responsiveness of self-specific CD8+ T cells that escape deletional tolerance.

Differences in the magnitude of the response to Trp2 immunization in WT and Dct-/- mice

It was possible that the lack of substantial clonal deletion or signs of prior activation in WT Trp2/Kb-specific cells indicated a form of ‘ignorance’ toward Trp2—that is, T cells capable of strong responses may simply not have encountered or recognized the self-antigen during normal homeostasis, as has been reported in some transgenic models (Heath et al., 1992; Ohashi et al., 1991). To investigate this, we challenged WT and Dct-/- mice with Trp2 in an immunogenic context using TriVax, a subunit immunization strategy comprising peptide, agonist anti-CD40 antibody, and poly(I:C) (Cho and Celis, 2009). It should be noted that the TriVax approach uses the minimal peptide for priming, which likely excludes antigen-specific Treg involvement. We included B8R peptide in addition to Trp2 peptide in these experiments as an internal control. While WT and Dct-/- mice responded similarly to B8R, WT mice showed a more limited response to Trp2 at an effector time point (day 7) relative to Dct-/- mice (Figure 2A,B), ruling out this type of ignorance as the dominant tolerance mechanism. We observed a significantly larger number and frequency of Trp2/Kb-specific cells in Dct-/- mice, and the Dct-/- cells exhibited higher apparent Trp2/Kb avidity (as measured by tetramer MFI; Figure 2C–E); TCRß MFI was similar between WT and Dct-/- Trp2/Kb-specific cells, indicating that the difference in tetramer MFI between the strains did not reflect differential modulation of cell-surface TCR (Figure 2—figure supplement 1A). Still, WT Trp2/Kb-specific cells expanded >1000-fold (Dct-/- cells expanded ~4000-fold). The WT:Dct-/- tetramer ratio was little changed relative to the pre-immune populations, suggesting the difference in avidity between WT and Dct-/- cells had not been amplified by activation.

Figure 2 with 3 supplements see all
Differences in the magnitude of the response to Trp2 immunization in wild-type (WT) and Dct-/- mice.

Mice were primed with TriVax (50 μg each of Trp2 and B8R peptides; A–E). The number (A) or percent (B) of splenic Trp2/Kb or B8R/Kb-specific cells was assessed at day 7. (C, D) The tetramer fluorescence intensity of splenic Trp2/Kb-specific cells was compared. (E) Gating for dual Trp2/Kb tetramer-positive CD8+ (samples were not enriched for Trp2/Kb-specific cells). (F) The frequency of the indicated splenic population expressing PD-1 is shown. Data in A and B are compiled from more than three experiments. Data in C–F are representative of three or more similar experiments. Squares indicate male animals. *p<0.05, ****p<0.0001 by unpaired t test (C) or one-way ANOVA with Sidak’s multiple comparisons test (A, B, F).

Aside from the difference in avidity, we were unable to detect meaningful differences in the expression of activation markers, chemokine receptors, or anergy/exhaustion markers at this time point (Figure 2—figure supplement 1B–D and data not shown). The frequency of PD-1+ cells was comparable between WT and Dct-/- Trp2/Kb-specific populations (Figure 2F), suggesting similar exposure to antigen. We also assessed cytokine production and degranulation following ex vivo Trp2 stimulation at day 7 after TriVax and found that a greater frequency of Dct-/- CD8+ T cells produced IFN-γ, TNF-α, and CD107a (an indicator of degranulation), in approximate proportion to the frequency of Trp2/Kb-specific cells (Figure 2—figure supplement 1E). Among IFN-γ-producing cells, a slightly larger proportion of Dct-/- cells co-produced TNF-α, indicating that Trp2/Kb-specific Dct-/- cells may have modestly improved polyfunctionality (Figure 2—figure supplement 1E).

As the skin is the site of Trp2 expression (in WT mice), we looked at Trp2/Kb-specific cells in the skin at day 7 after TriVax to determine whether we would see a more divergent phenotype between WT and Dct-/- cells in this location and whether larger numbers of Trp2/Kb-specific cells might be attracted to the skin in WT mice, potentially explaining the difference in splenic representation. Interestingly, this was not the case: the number of Trp2/Kb-specific cells in the skin was similar between the strains (Figure 2—figure supplement 2A). Tetramer MFI was again greater among the antigen-specific Dct-/- population; whether the difference in local antigen expression impacted cells’ responses was unclear—the proportion expressing CD69 ± CD103 was slightly greater in WT Trp2/Kb-specific cells, but we did not detect a clear difference in PD-1 expression (Figure 2—figure supplement 2B–D).

Since the Trp2180–188 epitope shows suboptimal binding to Kb (McWilliams et al., 2006), it was possible that our findings were influenced by the high doses of Trp2 peptide used in the TriVax immunization approach. To explore Trp2 responses in a more physiological context, we infected mice with a recombinant Listeria monocytogenes strain expressing Trp2 (LmTrp2) (Bruhn et al., 2005) and sacrificed the mice at effector (day 7) and memory (day 45) time points, assessing the percentage and number of Trp2/Kb-specific CD8+ T cells and cytokine production in response to ex vivo Trp2 stimulation. Again, the Trp2/Kb-specific response was greater in Dct-/- mice at both effector and memory time points (Figure 2—figure supplement 3A–C). As with TriVax, the frequency of all CD8+ T cells responding to ex vivo Trp2 stimulation with cytokine production (IFN-γ, TNF-α) was larger in Dct-/- mice; the percent producing IFN-γ approximated the tetramer-positive population, suggesting that the majority of Trp2/Kb-specific cells were able to produce this cytokine in both strains of mice (Figure 2—figure supplement 3D). Among IFN-γ-producing cells, those from Dct-/- mice tended to produce increased amounts of cytokine on a per-cell basis (as assessed by IFN-γ MFI).

With both the TriVax and LmTrp2 approaches, polyclonal Trp2/Kb-specific cells from WT mice showed evidence of tolerance, that is, submaximal responsiveness to Trp2, while those from Dct-/- mice mounted a stronger response consistent with a typical response to a foreign antigen. Because cells from Dct-/- mice were able to respond robustly in these experiments, the poor responsiveness of WT Trp2/Kb-specific cells could not be attributed solely to ignorance or ineffective immunization.

Tolerance to Trp2/Kb is CD8T cell-intrinsic

Both cell-intrinsic and cell-extrinsic mechanisms of CD8+ T cell tolerance have been previously described. Sakaguchi’s group (Maeda et al., 2014) identified anergic CD8+ T cells specific for melanocyte antigens in healthy human donors and concluded that these cells were restrained by Tregs. In contrast, other groups have shown cell-intrinsic deficits among self-reactive CD8+ T cells. For example, Davis’ group Yu et al., 2015 found human self-antigen-specific T cells to be poorly responsive to antigenic stimulation even in the absence of Tregs, and Greenberg and colleagues (Schietinger et al., 2012) showed that tolerant self-reactive murine CD8+ T cells remained tolerant when transferred into new hosts that lacked antigen expression.

Accordingly, we investigated whether cell-intrinsic or cell-extrinsic mechanisms were active in restraining Trp2/Kb-specific CD8+ T cells in WT mice. To assess this, we transferred bulk WT polyclonal CD8+ T cells to both WT and Dct-/- recipients, then primed the mice with TriVax and examined the effector response at day 7 post-immunization (Figure 3A, B and D). Transferred WT Trp2/Kb-specific cells did proliferate (~100-fold expansion), albeit to a much lesser degree than endogenous Dct-/- Trp2/Kb-specific cells. Importantly, their expansion was comparable in both WT and Dct-/- recipients (Figure 3B,D), suggesting that the WT CD8+ T cells remained hyporesponsive even in an environment where endogenous cells were not tolerant to Trp2, supporting a cell-intrinsic basis for the impaired reactivity of WT Trp2/Kb-specific CD8+ T cells. These findings also argue against a model in which the reduced number of splenic Trp2/Kb-specific cells arising after priming in WT mice reflects recruitment into sites of self-antigen exposure (e.g., the skin), since similar numbers of transferred WT cells are found in the spleen in WT hosts and Dct-/- hosts lacking antigen expression.

Figure 3 with 1 supplement see all
Wild-type (WT) Trp2/Kb-specific cells exhibit cell-intrinsic tolerance.

(A) We performed negative enrichment for CD8+ T cells from WT or Dct-/- donors and transferred bulk CD8+ T cells into congenically distinct WT or Dct-/- recipients. One day later, mice were immunized with TriVax (100 μg of Trp2 and B8R peptide). Donor and endogenous cells were collected from the spleens of recipient mice on day 7 following immunization and assessed for the percent (B, C) and number (D, E) of Trp2/Kb-binding cells. Data in B–E were compiled from three or more experiments. Squares indicate male animals. **p<0.05, ***p<0.001, ****p<0.0001 by one-way ANOVA with Sidak’s multiple comparisons test. Endog, endogenous; recip, recipients.

We also assessed the performance of Dct-/- CD8+ T cells when transferred into Dct-/- and WT hosts prior to priming to determine whether they would acquire tolerance in the WT environment (Figure 3A). These Trp2/Kb-specific donor cells were able to expand robustly in both Dct-/- and WT recipients (Figure 3C,E), further demonstrating a lack of extrinsic regulation in the WT environment. Dct-/- cells actually performed better in WT recipients than in Dct-/- recipients; the basis for this outcome is not clear but could be due to reduced competition by endogenous Trp2/Kb-specific cells in WT hosts. Preliminary studies indicated that Dct-/- cells still showed strong expansion when the interval between cell transfer and TriVax was extended from 1 day to 1 week, suggesting that these cells did not acquire tolerance characteristics within this timeframe (data not shown).

We conducted similar transfers utilizing LmTrp2 instead of TriVax, again finding evidence of cell-intrinsic tolerance. The transferred cells behaved in accordance with the donors’ Trp2 expression rather than that of the recipients: WT cells remained tolerant when primed in Dct-/- recipients, while Dct-/- cells retained the ability to expand when primed in WT recipients (Figure 3—figure supplement 1A,B). Collectively, these data indicate that cell-intrinsic mechanism(s) enforce tolerance among WT Trp2/Kb-specific cells.

WT Trp2/Kb-specific cells are capable of an acute response to Trp2

Although the response to Trp2 immunization was weaker in WT versus Dct-/- mice, the WT response was still substantial (Figure 2A). Studies on T cells with low affinity TCRs have shown a normal initial proliferative response that stalls prematurely relative to the response by high-affinity T cells (Enouz et al., 2012; Ozga et al., 2016; Zehn et al., 2009). Alternatively, it was possible that fewer clones would be recruited into the Trp2 response in WT mice, leading to decreased expansion relative to Dct-/- animals from the initiation of an immune response. To distinguish between these possibilities, we studied the expansion kinetics of the Trp2/Kb-specific response in WT and Dct-/- mice. In order to track early polyclonal responses, TriVax with a higher dose of Trp2 peptide was used in these studies, and tetramer enrichment was used to isolate Trp2/Kb-specific cells. Interestingly, WT Trp2/Kb-specific cells were capable of an initial response that largely paralleled that shown by their Dct-/- counterparts (Figure 4A). One day after TriVax immunization, few cells were isolated, likely due to either trapping within the tissues (Weninger et al., 2001) or TCR downregulation (Cai et al., 1997). Slightly more Trp2/Kb-specific cells were identified in WT mice on day 2, while increased numbers of Trp2/Kb-specific cells were seen in Dct-/- mice on days 3 through 5. By days 6 and 7 after high-dose TriVax immunization, Trp2/Kb-specific cells in Dct-/- mice outnumbered those in WT mice by an average ratio of 4:1. Although significant, these differences in expansion were modest in comparison with the >1000-fold expansion of Trp2/Kb-specific cells in both strains (Figure 4A). Preliminary assessment of apoptosis induction (annexin V staining) showed no differences between the strains at 1 or 3 days after TriVax (data not shown).

Figure 4 with 1 supplement see all
Wild-type (WT) Trp2/Kb-specific cells are capable of an initial response to Trp2 similar to that of Dct-/- cells.

WT and Dct-/- mice received intravenous injections of TriVax with 200 μg Trp2 peptide. Tetramer enrichment was used to enumerate Trp2/Kb-specific cells and assess their phenotype at the indicated time points following immunization (A–C, E). The ratio between the mean experimental PE median fluorescence intensity (MFI) of Trp2/Kb-specific cells in WT mice relative to Dct-/- mice is plotted in D, with each symbol representing one experiment comprising two to five individual mice. Data are compiled from three or more experiments in A, C, and D. Representative flow plots from 1 day four experiment are shown in B, and the same representative day four experiment is shown in E. Squares indicate male animals; the dotted line indicates the average naïve precursor frequency from the spleen and lymph nodes. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 by one-way ANOVA with Sidak’s multiple comparisons test (performed on log-transformed data in [a]).

We also assessed the phenotype of responding Trp2/Kb-specific cells acutely after TriVax. With this approach, CD69 did not serve as a reliable indicator of activation due to the type I interferon response induced by poly(I:C) leading to CD69 upregulation (Shiow et al., 2006), and widespread CD44 expression was seen in both tetramer-positive and tetramer-negative cells because of the potent inflammatory response unleashed by this method of immunization. Accordingly, we tracked CD25 expression as an indicator of activation. CD25, the high-affinity alpha component of the IL-2 receptor, is upregulated with activation in certain situations (Valenzuela et al., 2002) and may enable a stronger effector response by cells expressing it (Obar et al., 2010; Williams et al., 2006). The proportion of Trp2/Kb-specific cells expressing CD25 was significantly greater in Dct-/- mice on day 4, and trended higher on days 2 and 3 (Figure 4B and C). The CD25 MFI of CD25+ cells was also higher on Dct-/- Trp2/Kb-specific cells on day 4 (Figure 4B and C), suggesting that Dct-/- cells expressed more CD25 on a per-cell basis. Once again, Dct-/- Trp2/Kb-specific cells displayed significantly higher tetramer MFI than WT cells on days 2 through 7, but the avidity differences detected by tetramer staining did not demonstrate a progressive increase with time; the ratio between the WT and Dct-/- tetramer MFI transiently dropped at days 4–6, but the ratio at day 7 was similar to that of pre-immune cells (Figure 4D, Figure 4—figure supplement 1A). As with pre-immune cells, the tetramer MFI of B8R/Kb-specific cells was similar between the strains at day 7 (Figure 4—figure supplement 1B). Tetramer MFI was highest among the CD25+ subset for both WT and Dct-/- cells; the tetramer MFI of CD25+ WT cells was similar to the MFI of the overall tetramer-binding Dct-/- population on day 4 (Figure 4E).

We also assessed the early response following peptide stimulation alone, since this would be analogous to encountering Trp2 in a non-inflammatory context. We again found the early response to be similar between WT and Dct-/- Trp2/Kb-specific cells. The number (Figure 4—figure supplement 1C) and phenotype of WT Trp2/Kb-specific cells was comparable to that of Dct-/- cells on day 1 post-peptide. The activation markers CD44 and CD69 were similarly upregulated in both (Figure 4—figure supplement 1D,E), however, responses began to diverge by day 2 after peptide stimulation, and by day 3 significant differences in number and CD44 expression had emerged, with Dct-/- cells clearly outperforming WT cells (Figure 4—figure supplement 1C,D).

These findings indicate that the response to Trp2/Kb in WT and Dct-/- mice follows similar kinetics and magnitude but that expansion in WT animals terminates prematurely. Interestingly, we did not observe a progressive increase in apparent TCR avidity over time among the Dct-/- responder pool relative to WT cells, as might be expected if the subset of Dct-/- T cells with higher avidity TCRs were the only cells capable of responding strongly to Trp2/Kb.

Single-cell sequencing reveals an impaired ability to differentiate into a highly proliferative population early after priming among WT Trp2/Kb-specific cells

To better understand the defects in expansion and functionality observed among WT Trp2/Kb-specific cells and assess the heterogeneity within this population, we performed single-cell RNA sequencing on Trp2/Kb-specific cells from WT and Dct-/- mice at day 7 after TriVax priming. After initial data processing, the WT and Dct-/- datasets were merged (Stuart et al., 2019); clusters based on the cells’ transcriptomes were generated in an unbiased manner and visualized using uniform manifold approximation and projection (UMAP). Surprisingly, the distribution of WT and Dct-/- cells was quite similar (Figure 5—figure supplement 1A); the majority of clusters showed similar frequencies of cells from the two strains, although a slightly larger proportion of Dct-/- cells (14% versus 6% of WT cells) were located in cluster 1 and a slightly smaller proportion in cluster 2 (5% versus 12% of WT cells). We assessed the expression of genes associated with activation, anergy, or exhaustion among these populations and found generally similar patterns of expression between sample groups (Figure 5—figure supplement 1B). Aligning with our analysis of anergy/exhaustion marker expression by flow cytometry, we found low expression of these genes in WT (and Dct-/-) cells at this effector time point (Figure 2—figure supplement 1D; Figure 5—figure supplement 1B).

Based on our finding that WT cells began to fall behind Dct-/- cells in number around day 4 after TriVax, it was possible that these populations had diverged earlier after activation. Accordingly, we also performed single-cell sequencing at day 3 after TriVax. At this time point, cells clustered into two major groups separated along the x-axis, each comprised of smaller clusters (Figure 5A). Interestingly, over half of the cells (58%) from Dct-/- mice were localized in cluster 0, but this cluster was nearly devoid of WT cells, representing only a small subset (13%) of WT cells (Figure 5B). Suspecting that this cluster might contain a more functional subset poorly represented in the WT population, we assessed its characteristics in more detail. Because this cluster made up the majority of Group A (left group), we performed differential gene expression (DE) analysis between the two major groups in the merged dataset: A and B (right group).

Figure 5 with 2 supplements see all
Wild-type (WT) Trp2/Kb-specific cells show proliferative defects in the early effector phase.

Trp2/Kb-specific CD8+ T cells were isolated from WT and Dct-/- mice on day 3 after TriVax and submitted for scRNA-seq. After initial processing, the WT and Dct-/- datasets were merged and further analyzed. (A) Uniform manifold approximation and projection (UMAP) representation of gene expression from merged datasets determined using Seurat; each dot represents one cell. Clusters are indicated by color. (B) Cells from the Dct-/- sample are shown on the left and cells from the WT sample on the right using the same UMAP (generated from merged data) shown in Figure 6A. (C) The most differentially expressed genes between Groups A and B (see Figure 6A); histone genes and other genes associated with proliferation are indicated. A positive average log-fold change value indicates higher expression in Group A. (D) Cell cycle analysis indicates the cell cycle phase for each cell on the UMAP plot (merged dataset). (E) Pie charts show the frequencies of cells within cluster 0 in each stage of the cell cycle (left: Dct-/- sample, right: WT sample). (F) Expression of CD25 (Il2ra) by cell is indicated on the clusters by color.

Histone genes (e.g., Hist1h1b, Hist1h1e, Hist1h1d, Hist2h2ac) were among the most upregulated in Group A compared to Group B; these genes are commonly induced in association with cellular replication (Mei et al., 2017). Other genes associated with proliferation, such as Myc, Nolc1, Npm1, and Ccne2, were also upregulated in Group A (Figure 5C), and cell cycle analysis revealed that the majority of cells in Group A were in stages G2/M or S of the cell cycle (Figure 5D). Among cells in cluster 0 (Group A), 76% of Dct-/- cells were in G2/M or S versus 42% of WT cells (Figure 5E). Gene set enrichment analysis revealed a strong enrichment of gene sets comprising Myc targets, E2F targets, and genes related to mTORC1 signaling and the G2/M checkpoint (Figure 5—figure supplement 2A).

Many of the cells in Group A expressed CD25, with the majority of the remainder located in cluster 2 of Group B (Figure 5F); cells in these clusters also showed enrichment for a gene signature associated with IL-2 receptor signaling (Figure 5—figure supplement 2B). This aligns with our finding that the frequency of cells expressing CD25 was greater in Dct-/- than WT Trp2/Kb-specific effectors (Figure 4C). Indeed, Group A cells showed significantly higher expression of certain genes relevant to the IL-2 signaling pathway with known impacts on T cell function, such as Irf4 and Myc; previous work has demonstrated that signaling through the IL-2 receptor is important for sustained Myc expression (Preston et al., 2015). Lag3 and Pdcd1 (PD-1) expression were seen among cells in some clusters, but expression of these markers was higher among Dct-/- cells (Figure 5—figure supplement 2C); other anergy and exhaustion markers were not widely expressed in either population, similar to the data from day 7.

Based on the association of IL-2 signaling with Dct-/- Trp2/Kb-specific effector cells, we administered IL-2 complex (IL-2 + anti-IL-2 S4B6 antibody) to WT and Dct-/- mice previously primed with TriVax to determine whether this would correct the defective proliferation of the WT Trp2/Kb-specific cells. IL-2 complex acts through the β and γ components of the IL-2 receptor, negating the impact of differential CD25 expression. IL-2 complex treatment on day 5 after TriVax or LmTrp2 improved the expansion of both WT and Dct-/- Trp2/Kb-specific cells to a similar extent (Figure 5—figure supplement 2D,E). Although this treatment did not correct the expansion defect of WT Trp2/Kb-specific cells in a selective manner, it did improve their numbers to the level seen among untreated Dct-/- cells, supporting the use of IL-2R-directed therapies in cancer immunotherapy designed to engage tolerant cells (Moynihan et al., 2016; Rosenberg, 2014; Waithman et al., 2008).

Taken as a whole, the RNA-sequencing data suggest that WT Trp2/Kb-specific cells are deficient in their ability to form the more proliferative subpopulation that comprises a majority of the Dct-/- Trp2/Kb-specific population on day 3 after priming. Nevertheless, proliferation of the WT population was not entirely constrained, since this pool continued to expand over successive days (Figure 4).

WT Trp2/Kb-specific cells are inefficient at mediating vitiligo

T cells that escape self-tolerance mechanisms can sometimes elicit autoimmunity. Even CD8+ T cells with very-low-affinity TCRs that avoid deletional tolerance have been found to drive tissue destruction following activation (Enouz et al., 2012; Sabatino et al., 2011; Zehn and Bevan, 2006). Furthermore, vigorous immunization against Trp2 can break tolerance and lead to vitiligo (Bowne et al., 1999; Cho and Celis, 2009; Moynihan et al., 2016). Our data indicated that the proliferative response of Trp2/Kb-specific cells was only slightly impaired in WT relative to Dct-/- mice, but the ability of these expanded cells to mediate overt tissue damage, as indicated by autoimmune vitiligo, was unclear.

To investigate this, we primed WT and Dct-/- donors with TriVax, then transferred day 7 effectors to congenically distinct WT recipient mice in parallel and immunized these recipients with TriVax. Equal numbers of Dct-/- and WT cells were transferred to compensate for the reduced response in WT mice. The recipients were treated with dinitrofluorobenzene (DNFB) on the left flank 6 days after transfer as a local inflammatory stimulus (Haas et al., 1992; Mackay et al., 2012; Zhang et al., 2009); vehicle (acetone/olive oil) was applied to the right flank (Figure 6A). Analysis of the blood 6 days after transfer and boosting revealed expansion of both types of donor cells; although there was a trend for transferred Dct-/- cells to expand to a greater degree than donor WT cells, the difference was not statistically significant (Figure 6B). Recipient mice were subsequently monitored for vitiligo development on a weekly basis and scored using a numeric metric (Figure 6—figure supplement 1A).

Figure 6 with 2 supplements see all
Wild-type (WT) Trp2/Kb-specific cells are unable to mediate efficient anti-melanocyte activity.

(A) WT mice were monitored for vitiligo after receiving 50,000 Trp2/Kb-specific cells from WT or Dct-/- donors primed with TriVax 7 days prior; recipient mice received TriVax (100 μg Trp2) on the day of transfer and were treated with dinitrofluorobenzene (DNFB) (left flank) 6 days later. No cell transfer controls (not shown in schematic) received TriVax and DNFB but no transferred cells. (B) Recipient mice were bled on day 6 after transfer and TriVax; the number of transferred Trp2/Kb-positive cells per μL blood is shown. (C) Kaplan-Meier curve of vitiligo development; mice were considered to have vitiligo when they first had a vitiligo score of two that was sustained. Mean group vitiligo scores over time are shown in (D), with a dotted line indicating definite vitiligo. Data in C and D are compiled from three experiments with 4–10 mice per group. Data in B are compiled from two experiments with 4–10 mice per group. ****p<0.0001 by unpaired t test (B), log-rank survival analysis (C), or two-way ANOVA followed by Tukey’s multiple comparisons test (D).

Recipients of Dct-/- cells developed vitiligo more rapidly and more extensively than mice receiving WT cells, beginning around day 20 after cell transfer (Figure 6C). Vitiligo was most frequently initiated at the DNFB-treated site and would often progress over the following weeks to involve the right flank, hair around the eyes, and—in some cases—hair distributed over the body. Vitiligo progressed more rapidly and to a greater extent (higher numeric score) in recipients of Dct-/- cells (Figure 6D; Figure 6—figure supplement 1B,C), although low-grade vitiligo was observed in some mice receiving WT Trp2/Kb-specific effector cells or TriVax and DNFB without cell transfer. It is possible that initial melanocyte destruction mediated by the transferred cells facilitated antigen release and a broadening of the anti-melanocyte response to include endogenous T cells; nevertheless, limited studies using irrelevant TCR transgenic mice (P14) as recipients showed that Trp2/Kb-specific donor cells from Dct-/- mice were still able to induce vitiligo in this setting (Figure 6—figure supplement 2A,B).

Vitiligo severity (average vitiligo score) was positively correlated with the number of donor Trp2/Kb-specific cells in the blood on day 6 after transfer and TriVax boost (Figure 6—figure supplement 1D). This suggests that the enhanced proliferative capacity of Dct-/- cells was a factor in their superior ability to induce vitiligo, although it does not rule out additional qualitative differences between the WT and Dct-/- populations.

We also assessed the number and phenotype of transferred Trp2/Kb-specific cells in the skin of recipient P14 mice, which lacked a large population of endogenous antigen-specific cells. At day 11 or 12 after cell transfer, we identified similar numbers of cells from both donors in the DNFB-treated (left) flank; interestingly, this number was also similar to the number of cells identified in the vehicle-treated (right) flank (Figure 6—figure supplement 2C). The phenotype (PD-1, CD49a expression) of the transferred Trp2/Kb-specific cells was also similar across donor group and both flanks (Figure 6—figure supplement 2D). As in previous experiments, antigen-specific cells from Dct-/- donors exhibited higher tetramer MFI; the tetramer MFI on donor cells in the skin was similar to or slightly higher than that observed in the spleen (Figure 6—figure supplement 2E and data not shown). At a memory time point, fewer transferred Trp2/Kb-specific cells were identified in the skin, but this reduction was similar in recipients of WT and Dct-/- cells, as was the number of transferred Trp2/Kb-specific cells in the spleen at day 27 after transfer (data not shown). These findings contrasted with the clear difference in vitiligo elicited by Dct-/- versus WT cells and indicate that differences in skin recruitment and retention are unlikely to account for the distinct disease progression induced by these populations.

In summary, in contrast to the relatively modest differences in the expansion of Trp2/Kb responders in WT and Dct-/- mice, the ability of these populations to mediate autoimmune damage—melanocyte destruction—was strikingly different.

Discussion

A number of groups have demonstrated the existence of self-reactive CD8+ and CD4+ T cells in the periphery of mice and healthy human adults (Anderson et al., 2000; Bloom et al., 1997; Delluc et al., 2010; Maeda et al., 2014; Su et al., 2013; Yu et al., 2015). In some cases, self-reactive cells display indicators of reduced functionality, revealing them as tolerant and unlikely to cause spontaneous pathology. For example, self-reactive cells are often reported to express inhibitory receptors such as CTLA-4, PD-1, and LAG-3 (Fife and Bluestone, 2008; Maeda et al., 2014; Nelson et al., 2019; Schietinger et al., 2012). However, studies in human adults have identified self-reactive cells with a phenotype similar to that of naïve CD8+ T cells specific for foreign antigens (Yu et al., 2015); these cells did not display an overtly anergic phenotype but still responded poorly to stimulation. It is important to understand the mechanisms restraining these cells under normal conditions as well as their potential to cause pathology; such knowledge is critical to designing effective therapies to restrain these cells (e.g., to control autoimmune disease) or induce their responses (e.g., for cancer immunotherapy). As described in this report, we developed a polyclonal mouse model for CD8+ T cell self-tolerance, enabling us to define the characteristics of these cells and their reactivity in a physiological setting.

We found that the pre-immune populations of Trp2/Kb-specific cells in WT and Dct-/- strains were qualitatively similar, sharing a naïve phenotype and indistinguishable gene expression profile; there were no clear signs of prior antigen exposure among the WT cells. The size of the Trp2/Kb-specific precursor pool was only slightly (although significantly) smaller in WT mice, by less than twofold, and the response to Trp2 immunization was substantial in both strains, leading to a >1000-fold expansion of Trp2/Kb-specific cells in both WT and Dct-/- mice. Despite these commonalities, the population primed in Dct-/- mice showed greater expansion and elicited more rapid and widespread tissue destruction, read out as vitiligo, after adoptive transfer. Since the enhanced induction of vitiligo by transferred Dct-/- cells did not align with their improved recruitment to the skin, these findings suggest that Dct-/- cells are more efficient at mediating melanocyte destruction within the tissue.

Our adoptive transfer studies showed that the observed restraint in the WT Trp2/Kb-specific response did not depend on extrinsic factors but was a cell-intrinsic feature of pre-immune CD8+ T cells. This implies that other cell populations, including CD4+ Tregs or regulatory CD8+ T cells (Saligrama et al., 2019) are neither required for nor capable of affecting the responses of tolerant and non-tolerant Trp2/Kb-specific cells during priming. These findings also effectively eliminate the possibility that self-antigen presentation during Trp2 priming alters the nature of the immune response. However, while our studies argue that cell-extrinsic regulation is not required for enforcement or maintenance of tolerance by Trp2/Kb-specific cells, this does not exclude a potential role for Treg populations in establishing the initial tolerant state in WT mice. It is currently unclear whether tolerance to Trp2 is enforced during thymic development or in the periphery of WT mice: one report suggested that Dct expression is undetectable in thymic mTECs (Träger et al., 2012), but Trp2 could be brought into the thymus by dendritic cell populations to induce tolerance in WT animals. The site of tolerance induction was not a focus of the current study, but it will be interesting to determine whether instances of CD8+ T cell self-tolerance correlate with self-antigen expression patterns in the thymus (e.g., AIRE-regulated tissue-specific antigen expression).

Many of the characteristics we report for Trp2/Kb-specific CD8+ T cells in WT mice are strongly reminiscent of T cells with low affinity/avidity for antigen (Bouneaud et al., 2000; Enouz et al., 2012; Zehn and Bevan, 2006; Zehn et al., 2009), and we did observe modestly higher Trp2/Kb tetramer staining intensity on a subset of Dct-/- cells compared to WT responder cells. The ‘pruning’ of WT CD8+ T cells with high-avidity Trp2/Kb-specific TCRs remains the most likely explanation for this finding. If so, however, the effect of such pruning appears to be remarkably subtle. While analysis of unimmunized mice revealed that the average number and tetramer staining intensity of Trp2/Kb-specific CD8+ T cells were significantly lower in WT than Dct-/- animals, the degree of overlap suggests such measures would be unreliable for detecting tolerant cells beyond this optimized experimental model. Furthermore, the relatively consistent WT to Dct-/- tetramer MFI ratio in our time course experiments argues against an imbalanced outgrowth of higher affinity/avidity TCR clones in Dct-/- relative to WT animals (as might have been expected if TCR affinity was the primary factor driving improved proliferation in the Dct-/- population). At a practical level, the relatively modest differences in cell number and tetramer staining between tolerant and non-tolerant cells that we observed in this carefully controlled model system would not be sufficient to accurately predict self-reactivity versus tolerance in a clinical setting. Broadly similar conclusions were drawn from earlier studies of self-tolerance in humans (Maeda et al., 2014; Yu et al., 2015). This finding highlights the limitations of currently available assays for accurately predicting responsiveness to self-antigens.

Other studies utilizing mouse models have reported that even CD8+ T cells with very-low-affinity/avidity TCRs (including those undetectable by normal peptide/MHC tetramer staining) can provoke overt tissue damage that may reach or exceed the response observed for non-tolerant, high-affinity/avidity cells (Enouz et al., 2012; Sabatino et al., 2011; Zehn and Bevan, 2006). Our data indicate that the opposite can also occur: despite largely overlapping tetramer staining profiles, WT and Dct-/- Trp2/Kb-specific cells exhibit markedly different abilities to mediate widespread vitiligo. Hence, the impact of CD8+ T cell tolerance toward some self-antigens only partially limits expansion but can prevent the generation of cells readily capable of potent tissue destruction: tolerance is not a binary state.

Although enumeration, phenotyping, and pre-immune gene expression profiling failed to provide a robust metric for identifying functionally tolerant CD8+ T cells, we were able to delineate an inflection point following priming at which the responses of tolerant and non-tolerant cells diverged. Flow cytometry and single-cell RNA sequencing of Trp2/Kb-specific CD8+ T cells soon after priming demonstrated that WT responders failed to differentiate into a CD25+, IRF4+ population (a characteristic of most Dct-/- responder cells) and indicated that WT cells showed poor commitment to sustained proliferation. These combined features may be useful for further defining the responses by self-antigen-specific cells that are or are not capable of overt tissue destruction.

The early effector population of Trp2/Kb-specific cells demonstrates heterogeneity on a transcriptomic level in both strains. Whereas the majority of Dct-/- cells show a highly proliferative phenotype characterized by active cell cycling and responsiveness to mTOR and Myc, few WT Trp2/Kb-specific cells fall into this group. The reason(s) underlying the inability of WT cells to optimally engage these important pathways and proliferate efficiently require(s) further investigation but may relate to impaired sensitivity to endogenous IL-2 or other cytokines, the composition of the TCR repertoire, and/or altered TCR signaling. How these or other factors relate to the relative inability of primed WT Trp2/Kb-specific cells to mediate overt tissue damage is currently unclear, but it will be critical to identify the cellular and molecular mechanisms involved in future studies. Recent studies on dysfunctional tumor-specific or exhausted CD8+ T cells have shown that epigenetic changes in chromatin accessibility or methylation can maintain such states (Ghoneim et al., 2017; Pauken et al., 2016; Philip et al., 2017), which is one potential explanation for the cell-intrinsic nature of the tolerance seen in our model.

We were able to identify CD8+ T cells specific for other melanocyte epitopes/antigens in pre-immune mice; these cells had a similar phenotype to WT Trp2/Kb-specific cells. Accordingly, we predict that our results will apply to other populations of CD8+ T cells specific for melanocyte and potentially other tissue-restricted antigens. Similar populations of self-specific CD8+ T cells may exist in humans, and the ability of such cells to respond to self-antigen immunization while not causing autoimmune damage is relevant for understanding the limits of ‘breaking’ tolerance, for example, for cancer immunotherapy. Indeed, our results align with work examining polyclonal self-antigen-specific cells in human adults (Yu et al., 2015) with regard to the phenotype (modestly lower tetramer MFI, lower CD25 expression) and response to cognate peptide (diminished) observed among tolerant cells. Another study examining self-specific CD8+ T cells (Maeda et al., 2014) attributed their restrained responsiveness to Treg-mediated suppression; while we did not detect a cell-extrinsic regulatory mechanism in our studies, it is certainly possible that this mechanism limits the response to some self-antigens.

Our finding that polyclonal melanocyte-specific cells exhibit covert cell-intrinsic tolerance characterized by a partial defect in proliferation and a profound defect in tissue damage has implications for utilizing such cells therapeutically. This model has clear relevance to human physiology and will be useful in exploring methods of correcting the proliferative defects of tolerant cells to more effectively mobilize them in cancer immunotherapy approaches targeting tumor antigens shared with self.

Materials and methods

Key resources table
Reagent type
(species) or
resource
DesignationSource or
reference
IdentifiersAdditional
information
Strain, strain background (Mus musculus)Dct-/- (Exon 2–6-/-)Dr. A. Andy Hurwitz, NCI
Strain, strain background (Mus musculus)C57BL/6NCrl (C57BL/6)NCI Charles RiverStrain code: 556
Strain, strain background (Mus musculus)
B6.SJL-PtprcaPepcb/BoyCrCrl (CD45.1)NCI Charles RiverStrain code: 564
Strain, strain background (Mus musculus)P14Dr. R. Ahmed, Emory University
Chemical compound, drugPoly(I:C)InvivogenCat. #: vac-pic 
AntibodyInVivoMAb anti-mouse CD40BioXCellCat. #: BE0016-2;
RRID:AB_1107647
Clone FGK4.5
Peptide, recombinant proteinTrp2 peptideNew England PeptideTrp2180-188 H2N-SVYDFFVWL-OH
Chemical compound, drug1-Fluoro-2,4-dinitrobenzene (DNFB)Sigma AldrichCat. #: D-1529
AntibodyAnti-mouse CD8a FITCTonbo BiosciencesCat. #: 35-0081;
RRID:AB_2621671
Clone 53-6.7
AntibodyAnti-mouse CD8a vf450Tonbo BiosciencesCat. #: 75-0081; RRID:AB_2621931Clone 53-6.7
AntibodyAnti-mouse CD8a PerCP-Cy5.5Tonbo BiosciencesCat. #: 65-0081; RRID:AB_2621882Clone 53-6.7
AntibodyAnti-mouse CD8a APCef780eBioscienceCat. #: 47-0081-80; RRID:AB_1272221Clone 53-6.7
AntibodyAnti-mouse CD4 BV605BioLegendCat. #: 100548; RRID:AB_2563054Clone RM4-5
AntibodyAnti-mouse CD4 PE-Cy7Tonbo BiosciencesCat. #: 60-0042; RRID:AB_2621829Clone RM4-5
AntibodyAnti-mouse CD44 BV786BD BiosciencesCat. #: 563736; RRID:AB_2738395Clone IM7
AntibodyAnti-mouse CD44 FITCeBioscienceCat. #: 11-0441-85; RRID:AB_465046Clone IM7
AntibodyAnti-mouse CD44 rf710Tonbo BiosciencesCat. #:80-0441; RRID:AB_2621985Clone IM7
AntibodyAnti-mouse CD45.2 FITCTonbo BiosciencesCat. #: 35-0454; RRID:AB_2621692Clone 104
AntibodyAnti-mouse CD45.1 PE-Cy7Tonbo BiosciencesCat. #: 60-0453; RRID:AB_2621850Clone A20
AntibodyAnti-mouse CD90.1 ef450eBioscienceCat. #: 48-0900-82; RRID:AB_1272254Clone HIS51
AntibodyAnti-mouse CD90.2 PE-Cy7Tonbo BiosciencesCat. #: 60-0903; RRID:AB_2621857Clone 30-H12
AntibodyAnti-mouse MHC Class II (I-A/I-E) APC ef780Thermo Fisher ScientificCat. #: 47-5321-82; RRID:AB_1548783Clones M5/114.15.2
AntibodyAnti-mouse MHC Class II (I-A/I-E) BV510BioLegendCat. #: 107635; RRID:AB_2561397Clones M5/114.15.2
AntibodyAnti-mouse F4/80 BV510BioLegendCat. #: 123135; RRID:AB_2562622Clone BM8
AntibodyAnti-mouse F4/80 APC ef780Thermo Fisher ScientificCat. #: 47-4801-82; RRID:AB_2735036Clone BM8
AntibodyAnti-mouse CD122 BV421BD BiosciencesCat. #: 562960; RRID:AB_2737918Clone TM-β1
AntibodyAnti-human Granzyme B PEInvitrogen/ThermoCat. #: GRB04; RRID:AB_2536538Clone GB11
AntibodyAnti-mouse IFN-g PE-Cy7Tonbo BiosciencesCat. #: 60-7311-U100; RRID:AB_2621871Clone XMG1.2
Antibody
Anti-mouse TNFa APCeBioscienceCat. #: 17-7321-81; RRID:AB_469507Clone MP6-XT22
AntibodyAnti-mouse CD107a APCBioLegendCat. #: 121614; RRID:AB_2234505Clone ID4B
AntibodyAnti-mouse IL-2 APCBioLegendCat. #: 503810; RRID:AB_315304Clone JES6-5H4
AntibodyAnti-mouse CD62L BV510BD BiosciencesCat. #: 563117; RRID:AB_2738013Clone MEL-14
AntibodyAnti-mouse CD62L PerCP-Cy5.5Tonbo BiosciencesCat. #: 65-0621Clone MEL-14
AntibodyAnti-mouse KLRG1 PE-Cy7eBioscience/ThermoCat. #: 25-5893-82; RRID:AB_1518768Clone 2F1
AntibodyAnti-mouse CD127 BV786BD BiosciencesCat. #: 563748; RRID:AB_2738403Clone SB/199
AntibodyAnti-mouse CD69 FITCTonbo BiosciencesCat. #: 35-0691; RRID:AB_2621698Clone H1.2F3
AntibodyAnti-mouse CD25 PE-Cy7Thermo FisherCat. #: 25-0251-82; RRID:AB_469608Clone PC61.5
AntibodyAnti-mouse CD103 BV510BD BiosciencesCat. #: 563087; RRID:AB_2721775Clone M290
AntibodyAnti-mouse PD-1 PerCP-Cy5.5BioLegendCat. #: 135208; RRID:AB_2159184Clone 29F.1A12
AntibodyAnti-mouse PD-1 FITCBioLegendCat. #: 135214; RRID:AB_10680238Clone 29F.1A12
AntibodyAnti-mouse Tim-3 BV421BioLegendCat. #: 119723; RRID:AB_2616908Clone RMT3-23
AntibodyAnti-mouse CTLA-4 PE-Cy7BioLegendCat. #: 106314; RRID:AB_2564238Clone UC10-4B9
AntibodyAnti-mouse CD5 APCef780Thermo Fisher scientificCat. #: 47-0051-82; RRID:AB_2573940Clone 53-7.3
AntibodyAnti-mouse TCRβ APCef780Thermo Fisher ScientificCat. #: 47-5961-82; RRID:AB_1272173Clone H57-597
AntibodyAnti-mouse LAG3 PE-Cy7BioLegendCat. #: 125226; RRID:AB_2715764Clone C9B7W
AntibodyAnti-mouse CD49a BUV395BD BiosciencesCat. #: 740262; RRID:AB_2740005Clone Ha31/8
AntibodyAnti-mouse CXCR3 PerCP-Cy5.5BioLegendCat. #: 126514; RRID:AB_1186015Clone CXCR3-173
AntibodyInVivoMab anti-mouse IL-2 antibodyBioXCellCat. #: BE0043-1; RRID:AB_1107705Clone S4B6
Chemical compound, drugRecombinant Mouse IL-2 Protein, CFR&D SystemsCat. #: 402-ML-500/CF
Chemical compound, drugStreptavidin, R-Phycoerythrin Conjugate, premium gradeInvivogenCat. #: S21388
Chemical compound, drugStreptavidin, Allophycocyanin Conjugate, premium gradeInvivogenCat. #: S32362
Chemical compound, drugH-2K(b) SVYDFFVWL MonomerNIH Tetramer CoreBiotinylated Monomer
Chemical compound, drugGhost Dye Red 780 Viability DyeTonbo BiosciencesCat. #: 13-0865-T100
Chemical compound, drugLIVE/DEAD Fixable Aqua Dead Cell Stain KitThermo FisherCat. #: L34966
Chemical compound, drugAnnexin V FITC Apoptosis Detection KitBD BiosciencesCat no. 556570; RRID:AB_2869085

Mice

C57BL/6 (WT) mice were obtained from Charles River laboratories and housed in specific pathogen-free conditions at the University of Minnesota. Dct-/- mice on a C57BL/6 background were developed by Katie Stagliano and A. Andy Hurwitz at the NCI (see below); the mice were subsequently bred in-house on different congenic backgrounds and housed in specific pathogen-free conditions. Animals were used at 6–14 weeks of age. All animal experiments were approved by the Institutional Animal Care and Use Committee at the University of Minnesota. In accordance with NIH guidelines, both male and female animals were used in experiments; males are indicated by square symbols in the figures.

Generation of DctΔ2-6 (Dct-/-) mice

Request a detailed protocol

The construct used to generate an exon 2–6 deletion mutant of Dct (DctΔ2-6—referred to as Dct-/- in the manuscript) was synthesized by IDT (Coralville, IA). Electroporation of the gene construct into mouse embryonic stem cells (mESCs) was performed by the Transgenic Mouse Model Laboratory (TMML) of the Laboratory Animal Services Program at the National Cancer Institute, Frederick, MD. Polymerase chain reaction (PCR) primers were designed to screen for the presence of the recombined construct in the proper genomic location of prospective clones. pDRAW32 deoxyribonucleic acid (DNA) analysis software (http://www.acalone.com) and Primer3 (http://www.simgene.com) were used to design primers, and the NCBI Primer-BLAST tool was used to confirm the specificity of primers in the mouse genome (https://www.ncbi.nlm.nih.gov/tools/primer-blast/).

Genomic DNA from candidate mESC samples was purified using standard phenol chloroform extraction. Clones were screened by PCR using the primer pairs listed in Figure 1—source data 2. Hot-Start Taq Blue Master Mix from Denville Scientific (Holliston, MA) was used with the following conditions: 5 min at 95°C; followed by 40 cycles of 1 min at 95°C, 1 min at 55°C, and 7 min at 72°C; followed by cooling and storage at 4°C. PCR products were visualized by agarose gel electrophoresis. After screening ~350 mESC samples, four candidates that showed recombination were found. These cells were then transferred into blastocysts and inserted into pseudo-pregnant females by the TMML. Ultimately, two chimera lines were generated and transferred into the Hurwitz laboratory’s mouse colony where we oversaw the breeding of chimeras and the intercrossing of pups. Pups were screened for slate coat color, which is the phenotype expected for mice with a homozygous deletion of Dct. Tail clips were used to test for the presence or absence of exons by PCR. Primer pair sequences are listed in Figure 1—source data 3. Hot-Start Taq Blue Master Mix from Denville Scientific (Holliston, MA) was used with the following conditions: 5 min at 95°C; followed by 30 cycles of 1 min at 95°C, 1 min at 60°C, and 1 min at 72°C; followed by cooling and storage at 4°C. All primers were purchased from Integrated DNA Technologies, Inc (Coralville, IA).

Tetramer enrichment

Request a detailed protocol

Tetramer enrichment was used to isolate antigen-specific cells from pre-immune or acutely challenged mice. A modification of the method used by Obar et al., 2008 was employed. Following digestion with collagenase D, single-cell suspensions were prepared from the spleens (acutely challenged mice) or spleen and cervical, axillary, brachial, inguinal, and mesenteric lymph nodes (pre-immune mice). When possible, the same tetramer (Trp2180–188/Kb) was used in both APC and PE to ensure specificity. Anti-PE and anti-APC beads and magnetized columns (both from Miltenyi Biotec) were used to enrich for tetramer-bound cells. Samples were stained and analyzed by flow cytometry; CountBright counting beads (Invitrogen) were used for enumeration.

In vivo priming with Trp2

Request a detailed protocol

TriVax immunization was used as previously described (Cho and Celis, 2009); mice were immunized intravenously (via tail vein injections) with Trp2180–188 peptide or Trp2 and B8R20–27 peptides, agonist-anti CD40 antibody (BioXCell), and vaccine-grade poly(I:C), a toll-like receptor 3 agonist (InvivoGen). Peptide doses of 50, 100, and 200 μg per mouse were used for effector time points, transfer experiments, and acute time points, respectively, unless otherwise noted. Animals that received TriVax immunization via intraperitoneal instead of intravenous injection were removed from the analysis unless otherwise noted.

Infections with LmTrp2

Request a detailed protocol

Frozen stocks of LmTrp2 (Bruhn et al., 2005) were thawed and grown to log-phase in tryptic soy broth supplemented with streptomycin (50 μg/mL). Mice were typically injected with ~105–106 colony-forming units intravenously or intraperitoneally. Infectious doses were verified by colony counts on tryptic soy broth-streptomycin plates.

Ex vivo stimulation

Request a detailed protocol

In some experiments, splenocytes were stimulated ex vivo after isolation from infected mice. Splenocytes were incubated with Trp2 peptide (10−6 M) and Golgiplug (BD Biosciences) for 4–6 hr at 37°C; parallel wells with no peptide were used as a control. Cells were washed and stained with surface antibodies, followed by fixation and permeabilization with a FoxP3 Fix/Perm kit (eBioscience) or FoxP3/transcription factor staining buffer kit (Tonbo Biosciences) and staining with intracellular antibodies.

Skin harvests

Request a detailed protocol

Skin was shaved and dissected from the underlying tissue, then cut into small pieces and digested in a Collagenase Type III solution. Subsequently, samples were processed using a gentleMACS dissociator (Miltenyi Biotec), filtered, and washed before staining for flow cytometry.

Adoptive transfer experiments

Request a detailed protocol

Bulk polyclonal CD8+ T cells were isolated from the spleen and lymph nodes of WT or Dct-/- mice using negative magnetic enrichment (CD8a+ T cell Isolation Kit; Miltenyi Biotec). Enriched CD8+ T cells (typically ~75–90% pure) were resuspended in sterile PBS and 2–2.5 × 106 CD8+ T cells were injected intravenously per recipient mouse; recipient mice were congenically distinct (by CD45 and/or Thy-1 alleles). One day later, the recipient mice were immunized with TriVax or LmTrp2 intravenously or intraperitoneally. Mice were sacrificed for analysis 7 days later.

Bulk RNA sequencing of pre-immune mice

Request a detailed protocol

Trp2/Kb-specific cells were isolated from pre-immune WT and Dct-/- mice using tetramer enrichment followed by FACS on double tetramer-positive cells. Cells were isolated from three separate cohorts, with each cohort comprising eight WT and eight Dct-/- mice. The Clontech StrandedRNA Pico Mammalian workflow was used for library preparation, and samples were sequenced using an Illumina NextSeq instrument (2 × 75 bp paired end reads).

Bulk RNAseq analysis

Request a detailed protocol

Raw sequencing data were demultiplexed by sample into FASTQs (mean 24.6 million reads/sample) and mapped against the mouse genome (Ensembl GRCm38 release 95) using Hisat2 software (v 2.1.0). Gene level quantification was completed using Subread featureCounts software (v 1.6.2) and the read counts table was processed in R (v 3.5.2). Differentially expressed genes were identified with DESeq2 software (v 1.22.2) using a negative binomial model with effect size estimation completed by apeglm algorithm via the lfcShrink function. Group comparison p-values were adjusted by the Benjamini and Hochberg method to account for multiple hypothesis testing where genes with a false discovery rate (FDR) q < 0.05 were investigated in downstream analyses.

Single-cell RNA sequencing

Request a detailed protocol

Day 7: WT and Dct-/- mice were primed with TriVax (50 μg Trp2), and Trp2/Kb-specific cells were isolated using negative enrichment for CD8+ T cells followed by FACS for tetramer-positive cells. Day 3: WT and Dct-/- mice were primed with TriVax (200 μg Trp2), and Trp2/Kb-specific cells were isolated using tetramer enrichment followed by FACS. At both time points, cells were submitted for barcoding and library preparation using the 10× Genomics platform (Chromium Single Cell 5’ Library and Gel Bead Kit; Zheng et al., 2017), and samples were sequenced using an Illumina NovaSeq instrument with 2× 150 bp paired end protocol.

Single-cell RNAseq analysis

Request a detailed protocol

Raw sequencing data were processed using Cell Ranger (v 3.0.2; 10× Genomics) software programs ‘mkfastq’ for demultiplexing the WT and Dct-/- Illumina libraries and ‘count’ for read alignment against the mouse genome (mm10, provided by 10× Genomics, v 3.0.0) and generation of the mRNA transcript count table. Raw count data were loaded into R (v 3.6.1) and analyzed with the Seurat R package (v 3.0.3.9039) (Butler et al., 2018; Stuart et al., 2019). All scRNA datasets (WT or Dct-/-) at each time point (3 or 7 days after TriVax immunization) were independently filtered to include only cells (i.e., uniquely barcoded transcripts) expressing more than 300 genes and genes expressed in more than three cells (e.g., counts > 0). The proportion of mitochondrial RNA in each cell was calculated and cells with extreme levels (top or bottom 2% of all cells) were removed from the analysis. Genes with extreme expression levels (top or bottom 1% of all genes) were removed. Contaminating cells in the day 3 dataset expressing high levels of B cell or myeloid lineage marker genes and low levels of T cell markers were removed using empirically derived thresholds (675 B cells and 26 myeloid cells removed from WT and 117 B cells removed from Dct-/-). Downstream analysis of the day 3 dataset included a total of 4539 WT cells (19,326 genes) and 11,680 Dct-/- cells (19,416 genes); the day 7 dataset included a total of 6254 WT cells (12,902 genes) and 4784 Dct-/- cells (12,437 genes). The datasets from each time point were analyzed similarly in parallel, unless otherwise noted. Raw RNA counts were normalized with the LogNormalize function and each cell was classified according to its expression of canonical cell cycle genes using the CellCycleScoring function from Seurat (S-phase and G2/M-phase gene sets provided by Seurat were originally developed by Tirosh et al., 2016). For each cell, a cell cycle score was computed by subtracting the average expression of a random control gene set from the average expression of either S or G2/M cell cycle gene sets. If the difference score was negative for both comparisons, the cell was labeled G1. Otherwise, the cell was labeled according to which difference score was larger (S or G2/M). Raw RNA counts were normalized and transformed using the Seurat SCTransform function (Hafemeister and Satija, 2019) including the percent of mitochondria expression as a regression factor. Principal components analysis (PCA) was performed using the normalized, mean-centered, and scaled SCT dataset (RunPCA function). The top 3000 variable genes from each dataset were identified using the FindVariableFeatures function (vst method) and were used for WT and Dct-/- sample integration (Stuart et al., 2019). Two-dimensional projections were generated using the top 30 PCA vectors as input to the RunTSNE and RunUMAP functions. Cells were clustered using the FindNeighbors (top 30 PCA vectors) and FindClusters functions (testing a range of possible resolutions: 0.2, 0.4, 0.8, 1.2, 1.6). Pairwise DE testing (Wilcox rank-sum) with the FindMarkers function was performed between all initial clusters; any two clusters were merged if there were fewer than five significant DE genes (i.e., absolute value of log2-fold-change≥0.25 and Bonferroni adjusted p-value≤0.01). Pairwise DE testing continued on subsequently merged clusters. A final resolution of 0.2 was chosen (merging of initial clusters by DE testing was not required) to best represent the biological processes within both datasets. Cluster-specific pathway expression testing was completed using the VISION R package (DeTomaso et al., 2019) and figures were generated using the ggplot2 R package (Wickham, 2016). Gene Set Enrichment Analysis (Mootha et al., 2003; Subramanian et al., 2005) was performed using pre-ranked gene lists (sorted from largest to smallest log2-fold change between clusters compared). Gene set enrichment statistics were calculated for two gene set collections in the Molecular Signatures Database (hallmarks and c2 curated) derived for mouse symbols using the R package msigdbr, v 6.2.1 using the R package clusterProfiler (v 3.12.0). Interesting gene sets with an FDR q < 0.05 were evaluated. Raw and processed data have been deposited at Gene Expression Omnibus and are available via GEO.

IL2 complex treatment

Request a detailed protocol

Ten micrograms antibody (S4B6-1; Bio XCell) plus 1 μg murine recombinant carrier-free murine IL-2 (R and D) was administered per mouse via intraperitoneal injection on day 5 after priming with TriVax or LmTrp2; control mice received an equal volume of PBS.

Vitiligo induction

Request a detailed protocol

Donor mice (WT and Dct-/-) were primed with TriVax (100 μg Trp2); in one experiment, donor mice received ~50% less of the other TriVax components to minimize adverse reactions. Negative enrichment for CD8+ T cells was performed on day 7. Live cells were counted using a hemocytometer, and the percentages of CD8+ and Trp2/Kb tetramer-binding cells were applied to enumerate Trp2/Kb-specific cells; equal numbers (50,000) of WT or Dct-/- Trp2/Kb-specific cells were transferred to WT or P14 recipients. Recipients were treated with TriVax (100 μg Trp2) later the same day. On day 6 after cell transfer, the recipients were bled to assess donor populations. The mice were then treated with DNFB (0.15% in 4:1 acetone:olive oil) on the left flank; 30 μL was applied to a shaved patch of skin ~1.5 cm×1.5 cm in size; 30 μL of vehicle was applied to the right flank in the same manner. Control mice did not receive cell transfers, but did receive TriVax immunization and DNFB treatment at the same time as mice receiving cell transfers. Mice were monitored for vitiligo development on a weekly basis by an observer blinded to the experimental groups or by two independent observers.

Tetramers and flow cytometry

Request a detailed protocol

H-2Kb tetramers loaded with Trp2180–188 or B8R20–27 were obtained from the NIH tetramer core facility and labeled with streptavidin-fluorophore conjugates in house. Single-cell suspensions were stained with tetramers (when applicable) and fluorescent dye-conjugated antibodies purchased from BD Biosciences, Tonbo Biosciences, eBioscience, or BioLegend. In many experiments, Live/Dead Fixable Aqua Dead Cell Stain Kit (Thermo Fisher Scientific) was used for dead cell exclusion. When applicable, cells were fixed with a FoxP3 Fix/Perm kit (eBioscience) or FoxP3/transcription factor staining buffer kit (Tonbo Biosciences). These kits were also used for permeabilization prior to staining with intracellular antibodies. Samples were run on a BD LSR II or BD Fortessa instrument using BD FACSDiva (BD Bioscience), and data were analyzed with FlowJo (BD).

Statistical analysis

Request a detailed protocol

Initial sample size estimates were based on use of G*Power software (University of Dusseldorf). Prism software (GraphPad) was used to plot data and conduct statistical analyses. An unpaired t test was used for two-way comparisons between two groups. A one-way ANOVA with Sidak’s or Tukey’s multiple comparisons test was used when multiple comparisons were performed. Log-rank (Mantel-Cox) tests were used to evaluate Kaplan-Meier curves. A two-way ANOVA with Tukey’s multiple comparisons test was used to evaluated vitiligo scores over time. p-Values are represented as follows: *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Data availability

NextGen sequencing data has being deposited at GEO: Code GSE171221.

The following data sets were generated
    1. Truckenbrod EN
    2. Burrack KS
    3. Knutson TP
    4. Block KE
    5. Stagliano KR
    6. Hurwitz AA
    7. Fulton RB
    8. Renkema KR
    9. Jameson SC
    (2021) NCBI Gene Expression Omnibus
    ID GSE171221. Non-deletional CD8+ T cell self-tolerance permits responsiveness but limits tissue damage.

References

    1. Parkhurst MR
    2. Fitzgerald EB
    3. Southwood S
    4. Sette A
    5. Rosenberg SA
    6. Kawakami Y
    (1998)
    Identification of a shared HLA-A*0201 restricted T-cell epitope from the melanoma antigen tyrosinase-related protein 2 (TRP2)
    Cancer Research 58:4895–4901.
  1. Book
    1. Wickham H
    (2016)
    Ggplot2: Elegant Graphics for Data Analysis
    New York: Springer-Verlag.

Decision letter

  1. Gabrielle T Belz
    Reviewing Editor; The University of Queensland, Australia
  2. Satyajit Rath
    Senior Editor; Indian Institute of Science Education and Research (IISER), India
  3. Emma Hamilton-Williams
    Reviewer; University of Queensland
  4. Ian A Parish
    Reviewer; Australian National University, Australia

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

This is a very interesting manuscript which aims to identify distinguishing features of cells that undergo self tolerance. The authors have developed a novel system involving a new T cell receptor transgenic mouse to do this. The work is elegant and indicates that there are few differences between tolerized cells and non-tolerized cells. Interestingly, the authors demonstrate that regardless of tolerization, cells proliferate in response to antigen but tolerized cells prematurely fail to continue to proliferate to the same extent as non-tolerized cells. Most tolerance models are highly contrived and this study represents a significant effort to generate a more physiological system. The authors have provided a detailed and very thorough response to the comments of the reviewers. This work will be a valuable addition to the field.

Decision letter after peer review:

Thank you for submitting your article "Non-deletional CD8+ T cell self-tolerance permits responsiveness but limits tissue damage" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Satyajit Rath as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Emma Hamilton-Williams (Reviewer #2); Ian A Parish (Reviewer #3).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential Revisions:

1. A number of sections require more extensive phenotyping in the context of the activation setting and across different tissues.

2. The Dct-/- mouse requires much more extensive description and characterisation as outlined by the reviewers.

3. The proposed mechanism is not fully supported, thus the authors are asked to provide stronger molecular/cellular data to support their claim, and balance other deletional/tolerance possibilities in discussing the outcomes and interpretations of their work.

Reviewer #1:

This is a very interesting manuscript which aims to identify distinguishing features of cells that undergo self tolerance. The authors have developed a new and novel system involving a new T cell receptor transgenic mouse to do this. The work is elegant and indicates that there are few differences between tolerized cells and non-tolerized cells. Interesting the authors demonstrate that regardless of tolerization, cells proliferate in response to antigen but tolerized cells prematurely fail to continue to proliferate to the same extent as non-tolerized cells.

1. Both males and females were used in experiments – were differences evaluated between these two sexes? or expected?

2. Were cells stained for Lag3 and CTLA-4 expression?

3. RNAseq was performed but few genes were proffered as regulating the early termination of proliferation. This was performed on day 3, have cells a few days later been analysed or key genes interrogated? Are cytotoxic genes amplified at the same rate in tolerized vs non-tolerized cells?

4. CD25 expression is postulated as a mechanism of regulation. However, CD25 expression is highly variable, even between different infectious settings and thus it is unclear that it might be the defining feature.

5. This work is elegant but seems to lack a strong mechanism explaining the outcomes for tolerized cells and how these might be different from non-tolerized cells.

Reviewer #2:

Truckenbrod et al. evaluate tolerance induction in endogenous melanocyte antigen-specific CD8+ T cells from mice with and without expression of the specific self-antigen (Trp2). They show that while knock-out mice still have ample antigen-specific cells present (albeit in reduced numbers and with lower tetramer binding) they appeared phenotypically naïve. Upon immunisation, the antigen-specific cell population still expands dramatically (1000x) but with slower proliferation kinetics and the proliferation could be recovered by providing IL-2. Thy use single cell sequencing to identify that the most highly proliferative cell clusters are greatly reduced in the WT mice. They use a disease model of vitiligo induction and show that the cells from WT mice with endogenous antigen exposure are less able to induce disease. The strengths of the paper are a very thorough phenotypic analysis including using sophisticated single cell technology to elucidate the responding populations. They also validated their findings using both a vaccination model and an infection model. The model is highly relevant to human disease, particularly in the setting of cancer immunotherapy when self-antigen specific T cell populations must be recruited to the anti-tumour response. The paper is well written and clearly presented. The weakness was the inability to identify phenotypic differences in the pre-immune T cell response that programmed the difference in tolerance outcomes, which did not appear to be derived from recruitment of different affinity cells into the response (though this was not directly tested and closer analysis of TCR usage or clonotype sequencing may reveal this as a potential mechanism). A lack of gene expression differences in the pre-immune population may have been due to analysis of combined cell populations from multiple immune sites (spleen, peripheral and mesenteric LN). While presumably, antigen experience was occurring in the skin-draining LN where Trp2 antigen would be present. This may have diluted their ability to observe phenotypic differences/signs of antigen experience in the pre-immune population. Antigen exposure in the periphery may have resulted in reprogramming of the signalling machinery rather than gene expression changes. However, the authors do not claim to have determined the mechanism behind the tolerogenic phenotype and will investigate this in future studies. overall, the findings are interesting and novel.

My key issue was combining analysis of spleen and various LN in the analysis of the pre-immune population. Combining these different cell sources may well have masked the ability to detect a population with signs of antigen experience. Phenotyping of activation markers in skin-draining LN compared to other LN/spleen where Trp2 is not expected to drain should be included.

I am not sure that the conclusion from Figure 4D/E that there is no preferential response of higher affinity cells in the KO compared to the WT can be made from this data. It is hard to compare MFI between the two genotypes with low numbers of mice when small differences in staining efficiency can amplify differences between the samples. Ideally, the staining here needs to be normalised to an internal control within the same sample (total TCR staining?) or the two genotypes should be mixed and stained together.

Figure 1B and D are the histograms gated on tetramer+ cells? It is hard to compare Figure 1B and Suppl. Figure 1A as tick marks are not clear/no numbers on axis. In Figure 1B and 1C are male KO vs female WT mice compared or does the legend comment that squares represent male mice only apply to panel A and D?

Figure 2E: is this tetramer staining from enriched or non-enriched populations?

Methods: please explicitly state which lymph nodes are included in 'peripheral' lymph nodes.

Methods: please add a description of how the cell-cycle stage is assessed from the single-cell data.

Methods: Please provide details of the congenic markers used in the mice.

Suppl Figure 1A: please include example of WT vs KO pre-immune tetramer staining.

Reviewer #3:

Truckenbrod et al. examine how immune tolerance processes impact upon an endogenous population of self-reactive CD8+ T cells specific for the Trp2 self-peptide. They demonstrate persistence of a population of self-reactive CD8+ T cells in wild-type mice, albeit at lower numbers and avidity compared to those cells found within a mouse lacking the target self-antigen. This self-reactive T cell population could be expanded upon antigen encounter during either vaccination or infection, although interestingly the expanded T cell populations that developed in wild-type mice bearing the self-antigen exhibited some defects early during activation, and were less capable of causing autoimmune disease, compared to cells from mice lacking the target self-antigen. These defects were cell intrinsic and apparently not due to suppression by antigen-specific regulatory T cells. The authors conclude that a novel tolerance mechanism independent of deletion preserves responsiveness of cells to antigenic challenge but restrains their capacity to cause tissue damage.

Strengths

The key strength of this manuscript is the experimental model used. They have examined tolerance within endogenous T cell populations specific for a natural self-antigen in wild-type mice, and in parallel they examine control mice that lack the self-antigen to enable quantification of tolerance. This study is important as most tolerance phenomena have been described in far more artificial settings (eg. using high affinity TCR transgenic cells and/or studying tolerance to model self-antigens). By examining tolerance in more physiological models, the real-world relevance of phenomena previously described in more artificial tolerance models can be assessed. Their data suggest that tolerance within endogenous repertoires is more complex than that observed in more artificial models, consistent with recent findings in humans, and are thus a valuable addition to the field.

The authors have also conducted a comprehensive and impressive set of experiments to thoroughly probe the effect of tolerance on immune responsiveness in their models.

Weaknesses

The main weakness of this manuscript is that the argument that the observed phenomena cannot be explained by previously described tolerance mechanisms such as deletion and ignorance are not supported by the data. They argue that antigen exposure potentially leads to epigenetic conditioning of the remaining cells such that they can respond to antigenic challenge but cannot cause disease. However, their data do provide strong evidence that high avidity self-reactive cells are deleted by tolerance mechanisms and that the remaining cells are simply lower avidity cells that fall below the threshold for selection. These residual cells appear naïve, likely because they receive insufficient signals to be activated (ie. they are ignorant). This shift towards a lower avidity population is just as likely an explanation for the observed phenotypes, particularly given that high avidity cells are likely to be more destructive in autoimmune models. Thus, the findings can be explained by both deletion and ignorance, which are previously reported phenomena.

The authors also study the autoimmune potential of tolerized or control cells in a vitiligo model by first expanding the cells by vaccination prior to transfer into the disease model. However, little information is provided on the phenotype of these expanded cells prior to transfer. This is important as the phenotypes are likely to be different given that there is an expansion defect in tolerized cells. Phenotypic differences at this stage could explain differences in autoimmune potential. Additionally, the authors have not examined the skin infiltrating potential of the transferred cells, which may also be different due to differences in initial priming.

Finally, the authors have not examined whether there is evidence of migration of high avidity self-reactive cells into the skin in mice that bear the self-antigen vs those that do not. Differences in skin migration could explain the lower cell counts of tolerized cells in the spleen either in the steady-state or after vaccination.

The following changes are required to the manuscript

i. The manuscript should be revised to substantially temper the conclusions. There is clearly some deletion occurring (~2 fold) so it is misleading to suggest that no deletional tolerance has occurred. Given that this reduction in numbers is associated with lower avidity, this argues that the high avidity cells are deleted in WT mice. This is reflected in other data in the paper (eg. appearance of an IRF4hi scRNAseq cluster in KO mice suggestive of stronger TCR signal strength after vaccination, and enhanced expansion of KO cells in WT mice where there is presumably less competition from high avidity cells). The remaining lower avidity cells appear naïve by all measures, and the simplest explanation here is ignorance. While it is possible that other factors may be at play (eg. epigenetic conditioning), without identifying a shared endogenous TCR present in WT and KO mice and showing that it behaves differentially in both contexts (eg. via retrogenic mice), this is very hard to prove. Overall, there should be a diminished focus on an unknown tolerance mechanism contributing to the phenotypes – this can be replaced by a section in the Discussion considering the relative likelihood of deletion/ignorance vs epigenetic conditioning as an explanation for the findings. "Non-deletional" should be removed from the title, and the potential for deletion of high avidity clones to explain the phenotype should not be so readily dismissed within the manuscript.

ii. Do cells end up in the skin in WT mice (ie. the site of Ag)? This hasn't been ruled out as an explanation for reduced cells in the spleen (either after vaccination or in the steady-state). There could be subclinical inflammation/autoimmunity drawing high avidity cells into this site.

iii. More detail on the Dct KO strain is needed (schematic showing region deleted, more explanation on generation, whether or not a truncated protein is produced etc)

iv. Basic phenotyping of d7/8 cells post vaccination is needed, particularly as this has important implications for the vitiligo experiment. Eg. %KLRG1+ vs CD127+KLRG1-, GzmB expression, chemokine receptor expression (CXCR3, CX3CR1), cytokine polyfunctionality (IFNγ/TNFa/IL-2 co-production), degranulation etc. It's likely that the cells transferred in the vitiligo experiments are in different states of effector differentiation given the expansion and avidity differences, which is a major caveat in this experiment that is not considered/discussed. Also, Dct-/- CD8s have less competition from endogenous high avidity Trp2-responsive cells in WT mice, so will likely fare better in this transfer system based on Figure 3. Although there was no significant difference in persistence/expansion at d6, they may persist better long-term in this model, and this point should at least be discussed as an explanation for increased disease.

v. Was there a difference in early skin infiltration of Dct-/- derived CD8s in the vitiligo experiments? And did only higher avidity cells make it into the skin? Differences in chemokine receptor/integrin expression downstream of altered differentiation could affect homing capacity.

vi. Re avidity – although there is no change in relative avidity between the groups, only cells above a certain avidity threshold (~1000 MFI) appear to expand in both groups (compare Figure 2D to 1B). This would suggest that higher avidity cells are preferentially selected for expansion, meaning the statement that there is no selection for high avidity cells is technically incorrect and should be qualified.

vii. Figure 1 – FR4 and CD73 should be included for completeness given that they are bona fide anergy markers. Compiled MFIs of all markers in this figure should also be plotted. Additionally, CD25 MFIs in Figure 4B should be plotted.

https://doi.org/10.7554/eLife.65615.sa1

Author response

Essential Revisions:

1. A number of sections require more extensive phenotyping in the context of the activation setting and across different tissues.

2. The Dct-/- mouse requires much more extensive description and characterisation as outlined by the reviewers.

3. The proposed mechanism is not fully supported, thus the authors are asked to provide stronger molecular/cellular data to support their claim, and balance other deletional/tolerance possibilities in discussing the outcomes and interpretations of their work.

We have responded to all of these points in the specific comments below. In summary: (1) we present additional phenotypic (and gene expression) data to extend our analysis; (2) an extensive description of the generation and characterization of the Dct-/- mice is now included; (3) we expand the discussion of potential mechanisms that underpin tolerance in this system throughout the manuscript, particularly in the discussion.

Reviewer #1:

This is a very interesting manuscript which aims to identify distinguishing features of cells that undergo self tolerance. The authors have developed a new and novel system involving a new T cell receptor transgenic mouse to do this. The work is elegant and indicates that there are few differences between tolerized cells and non-tolerized cells. Interesting the authors demonstrate that regardless of tolerization, cells proliferate in response to antigen but tolerized cells prematurely fail to continue to proliferate to the same extent as non-tolerized cells.

We thank the reviewer for their enthusiastic comments. Just to clarify, we did not create a new TCR transgenic model in this work (which is focused strictly on responses of the natural polyclonal T cell pool), but we did generate and analyze a novel Dct knockout strain in these studies.

1. Both males and females were used in experiments – were differences evaluated between these two sexes? or expected?

We did not anticipate differences between males and females in these studies – we used both in order to comply with NIH guidelines and because susceptibility to some autoimmune diseases does indeed vary between the sexes. We did not conduct stringent statistical analysis comparing responses in males and females but did not observe striking differences. We can withdraw the indication of animal sex from the data if this is considered to be confusing to the readers.

2. Were cells stained for Lag3 and CTLA-4 expression?

CTLA-4 and LAG3 staining was shown for pre-immune cells (Figure 1F) and new data show gene expression for these and other exhaustion/anergy markers in primed Trp2/Kb-specific cells from Dct-/- and WT mice (Figures S7B, S8C) – in none of these situations did we see substantially increased expression in WT (tolerant) cells, which would have been anticipated if these cells were exhausted.

3. RNAseq was performed but few genes were proffered as regulating the early termination of proliferation. This was performed on day 3, have cells a few days later been analysed or key genes interrogated? Are cytotoxic genes amplified at the same rate in tolerized vs non-tolerized cells?

We apologize for the paucity of information on the phenotype and gene expression profiles of Trp2/Kbspecific cells at day 7 post-priming of WT and Dct-/-, and this was also requested by Rev. 3. As now shown in Figures S3 and S7, we observe only moderate differences between WT and Dct-/- populations at day 7 (fewer than we observed at day 3); by this stage, which represents the approximate peak of clonal expansion, it is expected that many cells will be out of cycle. As discussed in the manuscript, we expect that reduced early expression of CD25, limiting responsiveness to IL-2, contributes to the reduced expansion of Trp2/Kb-specific cells in WT mice. Regarding cytotoxicity, we did not see differential expression of the gene encoding Perforin but there was modestly higher expression of transcripts for Granzyme B in KO cells (at day 7 post-priming; Figure S7B and data not shown). We have expanded our description of phenotypic and gene expression characteristics in the text.

4. CD25 expression is postulated as a mechanism of regulation. However, CD25 expression is highly variable, even between different infectious settings and thus it is unclear that it might be the defining feature.

This is a fair point – indeed, while CD25 expression contributes to the expansion of CD8+ T cells in response to pathogens, this role is fairly limited. However, reduced expression of CD25 has also been noted in other studies on T cell tolerance, including the responses of low affinity/avidity CD8+ T cells, and we did observe that IL-2 therapy enhanced the expansion of Trp2/Kb-specific CD8+ T cells in WT mice (and also in Dct-/- mice)—both in the context of TriVax and LmTrp2 infection.

5. This work is elegant but seems to lack a strong mechanism explaining the outcomes for tolerized cells and how these might be different from non-tolerized cells.

We appreciate the reviewer’s comment (shared by Rev. 3). As made clearer in the revised manuscript, our chief goal in this report was to employ a mouse model to investigate whether non-deletional mechanisms dominate in CD8+ T cell tolerance toward self-antigens, as was suggested by some studies in humans (Maeda et al. 2014; Yu et al. 2015). By careful analysis in a well-controlled system, we conclude that there is quite modest “pruning” of CD8+ T cells with the highest affinity/avidity TCRs for Trp2/Kb in WT mice, and that this curtails but does not prevent considerable expansion in response to priming yet has a more substantial effect on restraining induction of immunopathology (the latter being something that could not be tested directly in the human studies). This general conclusion aligns with prior studies on the responses of low affinity/avidity TCR CD8+ T cells, but our data suggest the differences in TCR repertoire are remarkably subtle.

Reviewer #2:

[…] The weakness was the inability to identify phenotypic differences in the pre-immune T cell response that programmed the difference in tolerance outcomes, which did not appear to be derived from recruitment of different affinity cells into the response (though this was not directly tested and closer analysis of TCR usage or clonotype sequencing may reveal this as a potential mechanism). A lack of gene expression differences in the pre-immune population may have been due to analysis of combined cell populations from multiple immune sites (spleen, peripheral and mesenteric LN). While presumably, antigen experience was occurring in the skin-draining LN where Trp2 antigen would be present. This may have diluted their ability to observe phenotypic differences/signs of antigen experience in the pre-immune population. Antigen exposure in the periphery may have resulted in reprogramming of the signalling machinery rather than gene expression changes. However, the authors do not claim to have determined the mechanism behind the tolerogenic phenotype and will investigate this in future studies. overall, the findings are interesting and novel.

As the reviewer mentions, our studies on gene expression or phenotype in the pre-immune pool did not detect substantial differences between Trp2/Kb-specific cells from WT and Dct-/- mice. As the reviewer raised the idea that this could be affected by analysis of cells that were combined from multiple lymphoid tissues, we now provide data on this point for the reviewers. This is discussed in detail in the response to specific concerns, but preliminary analysis of skin-draining (vs non-skin-draining) lymph nodes showed similar numbers of Trp2/Kb-specific cells with similar phenotypic characteristics in WT and Dct-/- mice.

As the reviewer mentions, we do not claim to have defined the mechanism for tolerance in this system – although apparently subtle elimination of T cells with high-affinity/avidity TCRs remains a strong possibility – but we agree that these studies, focusing on the natural polyclonal T cell repertoire and testing the limits of tolerance through to immunopathology, are novel and interesting.

My key issue was combining analysis of spleen and various LN in the analysis of the pre-immune population. Combining these different cell sources may well have masked the ability to detect a population with signs of antigen experience. Phenotyping of activation markers in skin-draining LN compared to other LN/spleen where Trp2 is not expected to drain should be included.

This is an interesting idea, and we thank the reviewer for raising it. Since we are tracking the natural polyclonal T cell pool in these studies there is, however, a substantial technical barrier in analysis of skin-draining lymph nodes since very few cells would be expected to be Trp2/Kb specific. Still, as the reviewer indicated, that assumption might be incorrect – if there were, for example, accumulation of and/or prior activation of Trp2/Kb-specific CD8+ T cells in skin-draining lymph nodes from WT (but not Dct-/-) mice due to constitutive Trp2 antigen presentation. We have explored this in pilot experiments presented to the reviewers: by pooling skin-draining lymph nodes (axillary, brachial, cervical and inguinal) from 6 mice and performing Trp2/Kbtetramer enrichment, we could identify sufficient numbers of Trp2/Kb-specific cells to assess whether there were substantial differences between the populations in WT and Dct-/- mice. From the same mice, we pooled non-skin draining lymph nodes (mesenteric) separately and also examined splenic cells. We observed that the numbers and cell surface phenotype of Trp2/Kbtetramer-binding CD8+ T cells from skin-draining vs non-skin draining lymph nodes, and from WT vs Dct-/- mice, were all quite similar: in particular, there was scant evidence for enrichment or prior activation of Trp2/Kb-specific CD8+ T cells isolated in WT skin-draining nodes (preliminary data, not shown). Even with pooling samples from 6 mice, however, we detected only several hundred Trp2/Kb-specific events per group of lymph nodes (as would be predicted based on these cells being ~0.01% of the preimmune CD8+ T cell repertoire), making it hazardous to be dogmatic about these findings. As a result, we would like to limit our description of these preliminary experiments to “data not shown”.

In addition to these direct studies, it is also worth noting that if it were indeed the case that Trp2/Kb-specific cells in skin-draining nodes (but not other lymphoid sites) of WT mice were tolerized, this would not account for the poor response of WT CD8+ T cells after adoptive transfer into Dct-/- host mice and Trp2 immunization (Figure 3) – presumably, this population of donor T cells would be overwhelmingly non-tolerant Trp2/Kb-specific cells from nonskin draining sites. The same set of experiments preclude the idea that a few tolerant Trp2/Kb-specific CD8+ T cells dominantly impair responses (note the robust response of Dct-/- donor CD8+ T cells in WT host mice).

I am not sure that the conclusion from Figure 4D/E that there is no preferential response of higher affinity cells in the KO compared to the WT can be made from this data. It is hard to compare MFI between the two genotypes with low numbers of mice when small differences in staining efficiency can amplify differences between the samples. Ideally, the staining here needs to be normalised to an internal control within the same sample (total TCR staining?) or the two genotypes should be mixed and stained together.

First, it is important to note that the tetramer MFI ratios do not reflect low numbers of mice – each data point in Figure D is a summary of the average tetramer MFI ratio for groups of mice. So, for example, the d7 timepoint is reflective of a total of 44 mice, and the total for figure 4D is > 150. Second, while the reviewer’s comment is well taken, we showed significant differences in Trp2/Kbtetramer MFI between WT and Dct-/- mice in pre-immune mice and all stages of priming in dozens of experiments yet saw no significant difference in the MFIs for the B8R/Kb-specific cells (a group which happens to be at a similar precursor frequency as Trp2/Kb cells but is not likely to be affected by ablation of Dct; see Figures 1 and S6). Lastly, we have performed limited assessment of TCRß surface expression in parallel with Trp2/Kbtetramer binding – as shown in Figure S3A (and preliminary data, not shown), we saw similar TCR expression levels but distinct Trp2/Kbtetramer staining on the WT vs Dct-/- cells.

Figure 1B and D are the histograms gated on tetramer+ cells? It is hard to compare Figure 1B and Suppl. Figure 1A as tick marks are not clear/no numbers on axis. In Figure 1B and 1C are male KO vs female WT mice compared or does the legend comment that squares represent male mice only apply to panel A and D?

Yes, the histograms in 1B and 1D were gated on tetramer-positive cells. We have revised the data shown in Figures 1Band S2A (formerly S1A) to improve comparison.

Figure 2E: is this tetramer staining from enriched or non-enriched populations?

Non-enriched. This is now stated in the legend.

Methods: please explicitly state which lymph nodes are included in 'peripheral' lymph nodes.

Methods: please add a description of how the cell-cycle stage is assessed from the single-cell data.

Methods: Please provide details of the congenic markers used in the mice.

As requested we have stated which lymph nodes were harvested (p. 24), how cell-cycle stage was assessed in the scRNAseq studies (p. 26), and the details of the congenic markers used (CD45 and Thy-1 alleles) (p. 25) in the manuscript.

Suppl Figure 1A: please include example of WT vs KO pre-immune tetramer staining.

This has been added to Figure S2A (formerly S1A)

Reviewer #3:

[…] The main weakness of this manuscript is that the argument that the observed phenomena cannot be explained by previously described tolerance mechanisms such as deletion and ignorance are not supported by the data. They argue that antigen exposure potentially leads to epigenetic conditioning of the remaining cells such that they can respond to antigenic challenge but cannot cause disease. However, their data do provide strong evidence that high avidity self-reactive cells are deleted by tolerance mechanisms and that the remaining cells are simply lower avidity cells that fall below the threshold for selection. These residual cells appear naïve, likely because they receive insufficient signals to be activated (ie. they are ignorant). This shift towards a lower avidity population is just as likely an explanation for the observed phenotypes, particularly given that high avidity cells are likely to be more destructive in autoimmune models. Thus, the findings can be explained by both deletion and ignorance, which are previously reported phenomena.

The reviewer raises several important issues. To clarify, we did not intend to convey that we believe epigenetic regulation is the root cause of tolerance in this system, but simply raised this a possibility among several. We would also note that our definition of “ignorance” appears to differ from that of the reviewer – our usage of the term is based on the concept of T cells that are fully competent to respond to an antigen yet do not encounter it during normal homeostasis. This is similar, for example, to the model described by Ohashi and colleagues in 1991 (Ohashi et al., 1991), in which TCR transgenic CD8+ T cells maturing in an animal that also expresses the cognate antigen are neither deleted nor impaired in their response to priming with the same antigen. This is distinct from a case in which the T cells are unable to respond to the antigen (due, for example, to their TCRs being too low affinity to be activated). We have included a clarification about how we define the term on p. 7 (starting on line 187).

Finally, we agree with the reviewer that our data indicate some degree of deletion occurs in WT mice, culling T cells with the highest avidity TCRs for Trp2/Kb from the repertoire. While this has certainly been observed before, we believe the novelty and impact of our work is to show that this pruning of the TCR repertoire is quite subtle, leading to a largely overlapping peptide/MHC tetramer staining on tolerant and non-tolerant populations which does not become exaggerated during the immune response. Accordingly, our data highlight that what appears to be relatively modest editing of the polyclonal TCR repertoire can have substantial effects on the capacity of these cells to mediate immunopathology, which we believe to be novel.

The authors also study the autoimmune potential of tolerized or control cells in a vitiligo model by first expanding the cells by vaccination prior to transfer into the disease model. However, little information is provided on the phenotype of these expanded cells prior to transfer. This is important as the phenotypes are likely to be different given that there is an expansion defect in tolerized cells. Phenotypic differences at this stage could explain differences in autoimmune potential. Additionally, the authors have not examined the skin infiltrating potential of the transferred cells, which may also be different due to differences in initial priming.

We apologize for not including more data on the Trp2/Kb-primed populations used for adoptive transfer in the vitiligo studies. We now include phenotypic and gene expression analysis of this population (Figures S3 and S7), which suggest the differences observed between WT and Dct-/- populations in the early phase of the response to priming are largely resolved by the peak of the immune response. We also have examined infiltration of donor cells into the skin (see next point).

Finally, the authors have not examined whether there is evidence of migration of high avidity self-reactive cells into the skin in mice that bear the self-antigen vs those that do not. Differences in skin migration could explain the lower cell counts of tolerized cells in the spleen either in the steady-state or after vaccination.

These are also important points, and we thank the reviewer for raising them. We have now examined the recruitment of donor cells into the skin after TriVax immunization, and these data indicate similar infiltration of WT and Dct-/- cells in the skin (Figure S4). Additionally, we have added data regarding the number and phenotype of transferred WT and Dct-/- cells infiltrating the skin in our vitiligo transfer model, which we also find to be comparable regardless of donor origin (Figure S10).

Regarding infiltration at steady state, we and others observe extremely low numbers of CD8+ T cells in the skin in unmanipulated B6 mice, making it difficult to confidently assess whether any of these are specific for Trp2/Kb. Likewise, as we outlined in the response to Rev. 2, we do not see convincing evidence for activation of the (rare) Trp2/Kb specific CD8+ T cells found in the skin-draining lymph nodes of B6 mice at steady state.

The following changes are required to the manuscript

i. The manuscript should be revised to substantially temper the conclusions. There is clearly some deletion occurring (~2 fold) so it is misleading to suggest that no deletional tolerance has occurred. Given that this reduction in numbers is associated with lower avidity, this argues that the high avidity cells are deleted in WT mice. This is reflected in other data in the paper (eg. appearance of an IRF4hi scRNAseq cluster in KO mice suggestive of stronger TCR signal strength after vaccination, and enhanced expansion of KO cells in WT mice where there is presumably less competition from high avidity cells). The remaining lower avidity cells appear naïve by all measures, and the simplest explanation here is ignorance. While it is possible that other factors may be at play (eg. epigenetic conditioning), without identifying a shared endogenous TCR present in WT and KO mice and showing that it behaves differentially in both contexts (eg. via retrogenic mice), this is very hard to prove. Overall, there should be a diminished focus on an unknown tolerance mechanism contributing to the phenotypes – this can be replaced by a section in the Discussion considering the relative likelihood of deletion/ignorance vs epigenetic conditioning as an explanation for the findings. "Non-deletional" should be removed from the title, and the potential for deletion of high avidity clones to explain the phenotype should not be so readily dismissed within the manuscript.

As mentioned in the previous section (public review), we agree with the reviewer that some measure of clonal deletion (which leads to a reduction of the average number of cells by ~1.4-fold in the pre-immune pool) likely occurs in WT mice, and that the reduced Trp2/Kbtetramer MFI on these cells as well as other characteristics of their responses (impaired expression of CD25 and IRF4) align with this population being of lower avidity. Indeed, we had discussed our findings in the context of WT cells behaving like low affinity/avidity cells in the original discussion. We do not consider this to be a form of “ignorance” – at least with our intended meaning for the term, which we have clarified in the text. We have also removed the term “non-deletional” from the title and edited the rest of the manuscript to be consistent with our finding that some deletion does occur in WT mice.

However, we think it equally important to highlight that the degree of deletion or “pruning” of the TCR repertoire, if this is indeed the basis for tolerance, is remarkably subtle. Although we observe significant differences in the average Trp2/Kb-specific precursor number and tetramer MFIs, there is considerable overlap between the Trp2/Kb-specific populations from WT and Dct-/- mice that would severely limit the ability to predict a given animal’s tolerance status. Hence, our data indicate that assessment of clonal deletion would be an unreliable way to identify self-tolerance in, for example, a clinical setting. These points are made in the discussion.

Finally, a central goal of our studies was to model findings that had been published from studies in humans, which suggested minimal clonal deletion among self-specific CD8+ T cells and proposed varied mechanistic explanations for tolerance (ranging from a form of anergy to a key role for Tregs) (Maeda et al. 2014; Yu et al. 2015). Our studies indicate that modest editing of the TCR repertoire may minimally impact the magnitude of the response against self-antigen immunization – yet can lead to a more substantial effect on ensuing immunopathology.

ii. Do cells end up in the skin in WT mice (ie. the site of Ag)? This hasn't been ruled out as an explanation for reduced cells in the spleen (either after vaccination or in the steady-state). There could be subclinical inflammation/autoimmunity drawing high avidity cells into this site.

As mentioned under public comments, we have examined the recruitment of cells into the skin after immunization (Figures S4 and S10); the number of CD8+ T cells we can recover from the skin in unimmunized mice is too low for accurate analysis. We believe the adoptive transfer studies (Figure 3) argue against recruitment to the site of antigen as a basis for the low numbers of Trp2/Kb-specific splenocytes observed in WT mice – in those studies, WT donor cells expanded to a similar extent in the spleens of mice that did (WT hosts) and did not (Dct-/- hosts) express the antigen in the skin. The expansion of Dct-/- donor cells in the spleens of WT host mice was in fact increased relative to Dct-/- hosts, which is inconsistent with recruitment to the skin depleting the splenic pool of responder cells.

iii. More detail on the Dct KO strain is needed (schematic showing region deleted, more explanation on generation, whether or not a truncated protein is produced etc)

We apologize that these details were not provided – a schematic is now shown (Figure S1A), outlining the deleted regions, and details about the generation of the knockout are provided in the methods (p. 23). This knockout allele removes 5 of the 8 coding exons, so we suspect no truncated protein is produced. In any case, the coding sequence for the Kb-restricted Trp2 epitope (Trp2 180-188) is in the deleted exon 2.

iv. Basic phenotyping of d7/8 cells post vaccination is needed, particularly as this has important implications for the vitiligo experiment. Eg. %KLRG1+ vs CD127+KLRG1-, GzmB expression, chemokine receptor expression (CXCR3, CX3CR1), cytokine polyfunctionality (IFNγ/TNFa/IL-2 co-production), degranulation etc. It's likely that the cells transferred in the vitiligo experiments are in different states of effector differentiation given the expansion and avidity differences, which is a major caveat in this experiment that is not considered/discussed. Also, Dct-/- CD8s have less competition from endogenous high avidity Trp2-responsive cells in WT mice, so will likely fare better in this transfer system based on Figure 3. Although there was no significant difference in persistence/expansion at d6, they may persist better long-term in this model, and this point should at least be discussed as an explanation for increased disease.

As mentioned in the public comments, we apologize for the absence of these data. We examined d7 primed WT and Dct-/- cells by both phenotypic assays and by scRNAseq (Figures S3 and S7). Since the latter provides a more comprehensive assessment of the status of the cells (and also alignment of subpopulations via clustering), we focus on that in the revised manuscript. However, we also show and refer to data regarding most of the phenotypic and functional characteristics mentioned. Overall, and perhaps surprisingly, these data show that by day 7 of TriVax priming the Trp2/Kb-specific populations in WT and Dct-/- mice are more closely related than they were at day 3 of priming – the scRNAseq data indicates less distinct clusters (although there certainly are still some minor differences in clustering). This conclusion is amplified by phenotyping – we did not see clear differences in the expression of markers associated with activation or anergy/exhaustion between the populations at this time point. In terms of function, we did observe slightly higher Granzyme B gene expression in Dct-/- cells, and we saw a slight but significant increase in polyfunctionality in Dct-/- vs WT cells in ex vivo stimulation assays (in this case, polyfunctional cells made IFN-g and TNFα – we did not see clear production of IL-2, consistent with effector stage cells). The use of KLRG1 as a marker to distinguish short-lived effector- versus memory precursors is difficult in priming with TriVax; as has been published previously, this type of adjuvanted subunit vaccine does not generate a KLRG1hi population (Klarquist et al., 2018). We hope these additional data provide a more comprehensive view of the WT and Dct-/- cell populations in the late effector phase (these same cells being used for adoptive transfer in the vitiligo studies).

v. Was there a difference in early skin infiltration of Dct-/- derived CD8s in the vitiligo experiments? And did only higher avidity cells make it into the skin? Differences in chemokine receptor/integrin expression downstream of altered differentiation could affect homing capacity.

These are interesting but technically challenging experiments. We have had success in isolating donor cells from the skin (both the original DNFB-treated and contralateral flanks) from various timepoints in the vitiligo experiments, and we observe overall similar representation of donor WT and Dct-/- Trp2/Kb-specific cells (in terms of both percentage of all infiltrating CD8+ T cells and in terms of numbers). These data suggest that it is not infiltration into the skin per se that limits vitiligo induction by WT donor cells. Studies on the tetramer binding of these cells suggested that recruited cells were modestly increased for tetramer staining relative to splenic cells, but this was similar in WT and Dct-/- donor populations (and directly comparing tetramer staining MFIs on cells extracted from the skin versus those isolated from lymphoid tissues may be difficult). We included these data in Figure S10.

vi. Re avidity – although there is no change in relative avidity between the groups, only cells above a certain avidity threshold (~1000 MFI) appear to expand in both groups (compare Figure 2D to 1B). This would suggest that higher avidity cells are preferentially selected for expansion, meaning the statement that there is no selection for high avidity cells is technically incorrect and should be qualified.

We take the reviewer’s point – there is indeed a trend toward increased tetramer staining (often used as a surrogate for TCR avidity) in the Trp2/Kb-specific cells during the primary response relative to naïve cells, although it is difficult to know if this is truly a selection for a high-avidity population or relates to the activation state of the cells influencing staining. Regardless, the point of our statement was that there is not a widening gap in terms of the relative tetramer MFIs on Dct-/- and WT cells during the response – i.e., we do not see an imbalanced outgrowth of presumed high avidity cells in Dct-/- relative to WT animals. We have clarified this in the discussion.

vii. Figure 1 – FR4 and CD73 should be included for completeness given that they are bona fide anergy markers. Compiled MFIs of all markers in this figure should also be plotted. Additionally, CD25 MFIs in Figure 4B should be plotted.

In our RNA-seq analysis of preimmune cells, FR4 and CD73 were not differentially expressed between Trp2/Kb-specific cells from WT and Dct-/- mice. Among Trp2/Kb-specific cells from immunized mice, very few (< 1.5%) expressed these markers in limited examination by flow cytometry at day 7 after TriVax (data not shown). Small populations of WT Trp2/Kb specific cells (< 25% of WT cells in clusters 2 and 3) do express Izumo1r (FR4) at day 3 after TriVax by scRNA-seq, but neither marker is differentially expressed in a statistically significant manner at either day 3 or day 7 by scRNA-seq (Figures S7, S8, and data not shown).

We have now plotted compiled MFIs for the markers indicated in Figure 1 (Figure S2D) and CD25 MFIs (Figure 4C).

References:

Klarquist, J., Chitrakar, A., Pennock, N. D., Kilgore, A. M., Blain, T., Zheng, C.,... Kedl, R. M. (2018). Clonal expansion of vaccine-elicited T cells is independent of aerobic glycolysis. Sci Immunol, 3(27). doi:10.1126/sciimmunol.aas9822

Ohashi, P. S., Oehen, S., Buerki, K., Pircher, H., Ohashi, C. T., Odermatt, B.,... Hengartner, H. (1991). Ablation of "tolerance" and induction of diabetes by virus infection in viral antigen transgenic mice. Cell, 65(2), 305-317. doi:10.1016/0092-8674(91)90164-t

Williams, M. A., Tyznik, A. J., and Bevan, M. J. (2006). Interleukin-2 signals during priming are required for secondary expansion of CD8+ memory T cells. Nature, 441(7095), 890-893.

doi:10.1038/nature04790

https://doi.org/10.7554/eLife.65615.sa2

Article and author information

Author details

  1. Emily N Truckenbrod

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Contribution
    Conceptualization, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-3819-6307
  2. Kristina S Burrack

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Present address
    Hennepin Healthcare Research Institute, Saint Paul, United States
    Contribution
    Data curation, Validation, Investigation, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared
  3. Todd P Knutson

    Minnesota Supercomputing Institute, University of Minnesota, Saint Paul, United States
    Contribution
    Data curation, Formal analysis, Validation, Visualization, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8431-9964
  4. Henrique Borges da Silva

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Present address
    Mayo Clinic Arizona, Phoenix, United States
    Contribution
    Formal analysis, Investigation, Writing - review and editing
    Competing interests
    No competing interests declared
  5. Katharine E Block

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Contribution
    Formal analysis, Validation, Investigation, Writing - review and editing
    Competing interests
    No competing interests declared
  6. Stephen D O'Flanagan

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Contribution
    Investigation
    Competing interests
    No competing interests declared
  7. Katie R Stagliano

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Present address
    NIAID, NIH, Rockville, United States
    Contribution
    Conceptualization, Resources, Investigation, Writing - review and editing
    Competing interests
    No competing interests declared
  8. Arthur A Hurwitz

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Present address
    AgenTus Therapeutics, Inc, Boston, United States
    Contribution
    Conceptualization, Resources, Methodology, Writing - review and editing
    Competing interests
    is affiliated with AgenTus Therapeutics, Inc. The author has no financial interests to declare.
  9. Ross B Fulton

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Present address
    iFiBiO Therapeutics, Rockville, United States
    Contribution
    Resources, Formal analysis, Investigation, Methodology, Writing - review and editing
    Competing interests
    is affiliated with HiFiBio, Inc. The author has no financial interests to declare.
  10. Kristin R Renkema

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Present address
    Grand Valley State University, Allendale, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Supervision, Validation, Investigation, Visualization, Methodology, Project administration, Writing - review and editing
    For correspondence
    renkemak@gvsu.edu
    Competing interests
    No competing interests declared
  11. Stephen C Jameson

    Center for Immunology, University of Minnesota, Saint Paul, United States
    Contribution
    Conceptualization, Resources, Supervision, Funding acquisition, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    james024@umn.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-9137-1146

Funding

National Institute of Allergy and Infectious Diseases (R01AI140631)

  • Stephen C Jameson

National Institute of Allergy and Infectious Diseases (P01AI035296)

  • Stephen C Jameson

NIH Office of the Director (T32OD010993)

  • Emily N Truckenbrod

Cancer Research Institute

  • Ross B Fulton

National Cancer Institute (T32CA009138)

  • Kristin R Renkema

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank members of the Jamequist lab for helpful discussions and technical assistance. We thank the UMN Flow Cytometry Resource for cell sorting, the UMN Genomics Center for assistance with sequencing experiments, and the NIH tetramer core for peptide/MHC monomers. We thank the Celis and Restifo labs for providing reagents and samples used in enumerating cells specific for alternative melanocyte epitopes. We thank Dietmar Zehn for reviewing an early draft of this manuscript.

This work was supported by NIH grants P01 AI035296 (SCJ), R01 AI140631 (SCJ), and T32 OD010993 (ENT).

Ethics

Animal experimentation: This study was performed in strict accordance with the NIH Guide for the Care and Use of Laboratory Animals and handled according to protocols approved but the University of Minnesota IACUC (#1709-35136A and #2007-38243A).

Senior Editor

  1. Satyajit Rath, Indian Institute of Science Education and Research (IISER), India

Reviewing Editor

  1. Gabrielle T Belz, The University of Queensland, Australia

Reviewers

  1. Emma Hamilton-Williams, University of Queensland
  2. Ian A Parish, Australian National University, Australia

Publication history

  1. Received: December 9, 2020
  2. Accepted: April 29, 2021
  3. Accepted Manuscript published: April 30, 2021 (version 1)
  4. Version of Record published: May 21, 2021 (version 2)

Copyright

This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

Metrics

  • 761
    Page views
  • 103
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Developmental Biology
    2. Immunology and Inflammation
    Xia Liu et al.
    Research Article Updated

    Signals from the pre-T cell receptor and Notch coordinately instruct β-selection of CD4CD8double negative (DN) thymocytes to generate αβ T cells in the thymus. However, how these signals ensure a high-fidelity proteome and safeguard the clonal diversification of the pre-selection TCR repertoire given the considerable translational activity imposed by β-selection is largely unknown. Here, we identify the endoplasmic reticulum (ER)-associated degradation (ERAD) machinery as a critical proteostasis checkpoint during β-selection. Expression of the SEL1L-HRD1 complex, the most conserved branch of ERAD, is directly regulated by the transcriptional activity of the Notch intracellular domain. Deletion of Sel1l impaired DN3 to DN4 thymocyte transition and severely impaired mouse αβ T cell development. Mechanistically, Sel1l deficiency induced unresolved ER stress that triggered thymocyte apoptosis through the PERK pathway. Accordingly, genetically inactivating PERK rescued T cell development from Sel1l-deficient thymocytes. In contrast, IRE1α/XBP1 pathway was induced as a compensatory adaptation to alleviate Sel1l-deficiency-induced ER stress. Dual loss of Sel1l and Xbp1 markedly exacerbated the thymic defect. Our study reveals a critical developmental signal controlled proteostasis mechanism that enforces T cell development to ensure a healthy adaptive immunity.

    1. Cell Biology
    2. Immunology and Inflammation
    Ziwei Yang et al.
    Research Article

    Upon viral RNA recognition, the RIG-I signalosome continuously generates IFNs and cytokines, leading to neutrophil recruitment and inflammation. Thus, attenuation of excessive immune and inflammatory responses is crucial to restore immune homeostasis and prevent unwarranted damage, yet few resolving mediators have been identified. In the present study, we demonstrated that RTN3 is strongly upregulated during RNA viral infection and acts as an inflammation-resolving regulator. Increased RTN3 aggregates on the endoplasmic reticulum and interacts with both TRIM25 and RIG-I, subsequently impairing K63-linked polyubiquitination and resulting in both IRF3 and NF-κB inhibition. Rtn3 overexpression in mice causes an obvious inflammation resolving phenomenon when challenged with VSV, Rtn3-overexpressing mice display significantly decreased neutrophil numbers and inflammatory cell infiltration, which is accompanied by reduced tissue edema in the liver and thinner alveolar interstitium. Taken together, our findings identify RTN3 as a conserved negative regulator of immune and inflammatory responses and provide insights into the negative feedback that maintains immune and inflammatory homeostasis.