Targeting the CBM complex causes T_reg cells to prime tumours for immune checkpoint therapy

Abstract

Solid tumours are infiltrated by effector T cells with the potential to control or reject them, as well as by regulatory T (T_reg) cells that restrict the function of effector T cells and thereby promote tumour growth¹. The anti-tumour activity of effector T cells can be therapeutically unleashed, and is now being exploited for the treatment of some forms of human cancer. However, weak tumour-associated inflammatory responses and the immune-suppressive function of T_reg cells remain major hurdles to broader effectiveness of tumour immunotherapy². Here we show that, after disruption of the CARMA1–BCL10–MALT1 (CBM) signalosome complex, most tumour-infiltrating T_reg cells produce IFNγ, resulting in stunted tumour growth. Notably, genetic deletion of both or even just one allele of CARMA1 (also known as Card11) in only a fraction of T_reg cells—which avoided systemic autoimmunity—was sufficient to produce this anti-tumour effect, showing that it is not the mere loss of suppressive function but the gain of effector activity by T_reg cells that initiates tumour control. The production of IFNγ by T_reg cells was accompanied by activation of macrophages and upregulation of class I molecules of the major histocompatibility complex on tumour cells. However, tumour cells also upregulated the expression of PD-L1, which indicates activation of adaptive immune resistance³. Consequently, blockade of PD-1 together with CARMA1 deletion caused rejection of tumours that otherwise do not respond to anti-PD-1 monotherapy. This effect was reproduced by pharmacological inhibition of the CBM protein MALT1. Our results demonstrate that partial disruption of the CBM complex and induction of IFNγ secretion in the preferentially self-reactive T_reg cell pool does not cause systemic autoimmunity but is sufficient to prime the tumour environment for successful immune checkpoint therapy.

Main

Local exposure of tumour-infiltrating T_reg cells to their cognate antigens sustains their tumour-promoting immunosuppressive functions⁴. We therefore explored which T-cell-receptor (TCR)-dependent signalling pathways could be targeted to disable tumour-reactive T_reg cells. The scaffold protein CARMA1 nucleates assembly of the CBM complex in T cells in response to TCR-dependent PKCθ activity and promotes several functions, including activation of the AP-1, mTOR and classical NF-κB pathways, as well as stabilization of mRNA⁵. Constitutive genetic deletion of CARMA1, BCL10 or MALT1 abrogates thymic development of T_reg cells^6,7,8,9,10, but their role in mature T_reg cells is unknown.

When we conditionally deleted either one or both alleles of CARMA1 in mature T_reg cells by crossing Foxp3^YFP-cre to CARMA1^flox/flox mice (hereafter referred to as F^cre × C1^f/+ or F^cre × C1^f/f mice), levels of CARMA1 protein were proportionally reduced in CD4⁺FOXP3⁺ T_reg cells from lymph nodes (LNs) (Extended Data Fig. 1a). F^cre × C1^f/f, but not F^cre × C1^f/+ or C1^+/+ control mice, stopped thriving at 17 days; most died before 4 weeks of age after a T-helper-1 (T_H1)-dominated multiorgan inflammatory disease characterized by splenomegaly, lymphadenopathy, effector differentiation and inflammatory cytokine secretion by conventional T (T_conv) cells, production of autoreactive IgG, and activation of the myeloid compartment (Fig. 1a, Extended Data Figs. 1b–f, 2a–f). Hence, CARMA1 is essential for T_reg cells to maintain immune homeostasis, but a reduction in its expression at least up to 50% is tolerated.

Fig. 1: Loss of CARMA1 in T_reg cells is fatal, but reduced expression is sufficient to maintain immune tolerance.

a, Survival of F^cre × C1^+/+, C1^f/+ and C1^f/f mice (n = 8, 10 and 20 per group, respectively). b, Frequency of T_reg cells among CD4⁺ T cells and of eT_reg cells among total T_reg cells in LNs. c, d, Expression of cytokines (c) and transcription factors (TF) (d) in LN T_reg cells after ex vivo stimulation. e, Survival of F^cre × C1^f/f mice treated with anti-IFNγ antibodies from day 14 of life, compared with F^cre × C1^+/+ and scurfy mice. f, Cytokine expression of YFP⁺ T_reg cells from LNs of nine-week-old female heterozygous F^cre/+ × C1^+/+, C1^f/+ and C1^f/f mice after ex vivo stimulation. g, Frequency of YFP⁺ T_reg cells among CD4⁺ T cells and of YFP⁺ eT_reg cells among total YFP⁺ T_reg cells in LNs. h, Expression of indicated proteins in YFP⁺ eT_reg cells from 9-week-old mice. Data in b–h represent two independent experiments with similar results. Data are mean and individual replicates. In a and e, *P < 0.05 versus wild type (+/+), &P < 0.05 versus scurfy, and #P < 0.05 versus anti-IFNγ (log-rank (Mantel–Cox) test). In all other panels, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (one-way analysis of variance (ANOVA) with Tukey post hoc test). i, j, Bulk RNA sequencing analysis of YFP⁺ cT_reg and eT_reg cells from LNs of F^cre/+ × C1^+/+, C1^f/+ and C1^f/f mice. i, Principal component (PC) analysis of transcriptomes. j, Scaled expression of eT_reg signature genes by eT_reg cells (defined by fold change > 2 and P_adj < 0.01 between C1^+/+ cT_reg and C1^+/+ eT_reg cells). Selected eT_reg cell genes are annotated.

Absolute numbers of T_reg cells were increased in LNs of F^cre × C1^f/f mice, concomitant with an increase in overall LN cellularity (Extended Data Fig. 2g, h). However, the overall frequency of T_reg cells among CD4⁺ T cells did not vary with CARMA1 expression, whereas the proportion of CD44^highCD62L^– effector T_reg (eT_reg) was strongly reduced in its absence (Fig. 1b). Notably, CARMA1-deficient T_reg cells, while retaining expression of FOXP3, almost uniformly secreted IFNγ and—at lower frequencies—secreted IL-4, IL-17 and TNF (Fig. 1c). Although nearly all CARMA1-deficient T_reg cells secreted the T_H1-cytokine IFNγ, far fewer—and only eT_reg cells—expressed the T_H1 lineage-defining transcription factor T-bet along with RORγt and, to a lesser degree, GATA-3 (Fig. 1d and Extended Data Fig. 2i, j). Hence, complete—but not partial—deletion of CARMA1 in T_reg cells markedly dysregulates their cytokine expression that (in the case of IFNγ) is dissociated from expression of T-bet, and may contribute to the pathogenesis of inflammatory disease. Indeed, F^cre × C1^f/f mice died more rapidly than T_reg cell-deficient scurfy mice, but their lifespans were similar when IFNγ was neutralized (Fig. 1e). Thus, under inflammatory conditions, CARMA1-deficient T_reg cells convert from an immunoregulatory cell type into an IFNγ-secreting pathogenic cell type.

In heterozygous female F^cre/+ × C1^f/f mice, random inactivation of the X chromosome causes the yellow fluorescent protein–Cre recombinase (YFP–Cre) fusion protein to be expressed and CARMA1 to be deleted in only half of the T_reg cells, whereas the other half maintains immune homeostasis (Extended Data Fig. 3a–c). Under such non-inflammatory conditions, CARMA1-deficient T_reg cells did not secrete effector cytokines (Fig. 1f). However, in competition with CARMA1-sufficient T_reg cells for niche space, we observed a proportional decline specifically in the frequency of eT_reg cells—but not of central T_reg (cT_reg) cells—that lacked one or both alleles of CARMA1 (Fig. 1g and Extended Data Fig. 3e). The remaining YFP⁺ eT_reg cells expressed less FOXP3 and other proteins that are characteristically increased in expression during the differentiation of cT_reg cells into eT_reg cells (Fig. 1h). They expressed more of the pro-apoptotic protein BIM, but also more of the anti-apoptotic protein BCL2, which probably reflects impaired differentiation of eT_reg cells, as control eT_reg cells strongly downregulated both BCL2 and BIM relative to cT_reg cells (Extended Data Fig. 3d–h). The in vitro suppressive function of CARMA1-deficient T_reg cells was reduced, but not abrogated (Extended Data Fig. 4a, b), and they failed to persist and did not suppress lymphopenia-induced expansion of effector T cells after transfer into RAG1-deficient hosts (Extended Data Fig. 4c). Failure to persist in vivo did not seem to result from accelerated apoptosis, because the previously described high apoptotic rate of eT_reg cells¹¹ was not further enhanced in the absence of CARMA1 (Extended Data Fig. 4d). Lack of CARMA1 did not lead to an increase in the formation of FOXP3-negative ‘exT_reg’ cells (Extended Data Fig. 4e).

On the basis of global gene expression analyses, CARMA1-deficient eT_reg cells were equally dissimilar to control eT_reg cells as they were to control cT_reg cells, whereas CARMA1-deficient cT_reg cells were only moderately dissimilar to control cT_reg cells (Fig. 1i). In the latter, only 96 genes were differentially expressed, compared with 344 genes in eT_reg cells (Extended Data Fig. 5a and Supplementary Table 1). Based on differences between control cT_reg and eT_reg cells, we defined an ‘eT_reg signature’, which largely overlapped with a previously reported gene set^12,13. When examining these 689 genes, hemizygous CARMA1-deletion had only a moderate effect, whereas homozygous deletion induced major changes specifically in the eT_reg cell gene expression program (Fig. 1j and Extended Data Fig. 5b). Minor changes in the expression of the Bcl2 family of apoptotic regulator genes occurred during eT_reg cell differentiation, but CARMA1-deficient eT_reg cells did not deviate from this pattern apart from having less pronounced downregulation of BCL2 and BIM, confirming our observations on protein expression (Extended Data Fig. 6). Thus, a complete—and even a partial—loss of CARMA1 expression impairs eT_reg cell differentiation and persistence, but does not induce the cells to become pathogenic or convert to exT_reg cells under non-inflammatory conditions. However, in the context of incipient inflammation trigged by a global loss of the suppressive function of T_reg cells, CARMA1-deficient T_reg cells secrete IFNγ, which further accelerates inflammatory disease.

Failed thymic development of T_reg cells in the absence of CARMA1 results from disabled NF-κB signalling and is restored through the expression of a constitutively active form of IκB kinase 2 (referred to as IKK2ca)¹². Furthermore, the NF-κB proteins REL (also known as c-Rel) and RELA (p65) have important roles in T_reg cell function^13,14,15, which suggests that failed activation of NF-κB may primarily account for the effects of CARMA1-deletion in T_reg cells. However, the expression of IKK2ca in T_reg cells neither prolonged the lifespan of F^cre × C1^f/f mice, nor reduced the differentiation of effector T_conv cells (Extended Data Fig. 7a, b). Restoring NF-κB activation did not restore differentiation of eT_reg cells, and did not limit the secretion of IFNγ, but only of TNF (Extended Data Fig. 7c, d). Therefore, although NF-κB activation is evidently essential^13,16, additional CBM-complex effector functions are also crucial for maintaining T_reg cell function. The relevance of these additional functions will need to be investigated, but an initial examination of CARMA1-deficient T_reg cells from healthy F^cre/+ × C1^f/f mice already revealed decreased expression and TCR-induced phosphorylation of the AP-1 family protein JUN, as well as changes to phosphorylation of FOXO1, possibly reflecting CARMA1- and TBK1-driven regulation of AKT activity¹⁷ (Extended Data Fig. 7e, f).

Considering the pro-inflammatory potential of CARMA1-deficient T_reg cells, we examined their tumour response. Subcutaneous implantation of the poorly immunogenic Braf^V600E × Pten^−/− melanoma D4M.3A¹⁸ into female F^cre/+ × C1^f/f hosts amplified the effects of CARMA1-deficiency on T_reg cells, because the frequency not only of eT_reg cells but also of total T_reg cells was reduced in tumours and tumour-draining LNs (tdLNs) as a function of decreasing CARMA1 expression, accompanied also by a more pronounced reduction in FOXP3 expression (Extended Data Fig. 8a–d). Notably, we observed growth deceleration of D4M.3A melanoma, and of MC38 colon carcinoma, when half of the T_reg cells lacked one or both alleles of CARMA1 (Fig. 2a and Extended Data Fig. 8e). Because a mere loss of function in only half of the T_reg cells is not predicted to cause loss of tumour tolerance¹⁹, this suggested active T_reg-cell-mediated anti-tumour activity. Indeed, a large fraction of completely or even partially CARMA1-deficient T_reg cells secreted both TNF and IFNγ in situ, whereas these effector cytokines were undetectable in tumour-infiltrating CD4⁺ and CD8⁺ T_conv cells (Fig. 2b, c). Importantly, no increase in cytokine secretion by T_reg cells was observed in tdLNs or in non-lymphoid tissues, such as skin or lung (Fig. 2d and Extended Data Fig. 8f, g). IFNγ expression in tumour tissue correlated with downregulation, but not loss of FOXP3 in both partially and fully CARMA1-deficient T_reg cells (Extended Data Fig. 8h). Notably, destabilization of control T_reg cells by IFNγ-producing T_reg cells, as described in other settings²⁰, did not occur because no increase in cytokine expression was detectable in YFP⁻ CARMA1-sufficient T_reg cells in the same tumours (not shown). Neutralization of IFNγ fully restored tumour growth in F^cre/+ × C1^f/f mice (Extended Data Fig. 8i), which suggests a crucial role for this cytokine in T_reg-cell-mediated anti-tumour immunity. However, IFNγ may also derive from other cellular sources after T_reg cell destabilization. To test the role of T_reg-cell-produced IFNγ specifically, we transferred T_reg cells with reduced expression of CARMA1 into tumour-bearing C57BL/6 or Ifng^–/– mice. In both hosts, T_reg cells stunted tumour growth similarly, but not when IFNγ was neutralized, which indicates that T_reg-cell-derived IFNγ is both necessary and sufficient for anti-tumour effects (Fig. 2e, f and Extended Data Fig. 8j–m). Therefore, although neither partially nor fully CARMA1-deficient T_reg cells cause inflammation in healthy mice, both are selectively destabilized in tumour tissue and secrete IFNγ to decelerate tumour growth.

Fig. 2: Reduced CARMA1 expression converts tumour-infiltrating T_reg cells into IFNγ-secreting effector cells that dominantly control tumour growth.

a, Female F^cre/+ × C1^+/+, C1^f/+ and C1^f/f mice were implanted with D4M.3A melanoma, and tumour growth was recorded. b–d, In situ expression in tumour tissue (b, c) or tdLNs (d) of effector cytokines in YFP⁺ T_reg cells lacking one or both alleles of CARMA1 as well as in CD4⁺ and CD8⁺ T_conv cells 18 days after tumour implantation. e, f, One million YFP⁺ T_reg cells from F^cre/+ × C1^f/+ or C1^+/+ mice were injected intravenously (i.v.) into either C57BL/6 (e) or IFNγ-deficient (f) hosts, which were implanted with D4M.3A melanoma the next day, and tumour growth was recorded. Some IFNγ-deficient hosts were treated with neutralizing anti-IFNγ antibody. Data represent two independent replicates with similar results. Data are mean and either individual replicates (c, d) or s.e.m. (a, e, f). *P < 0.05, **P < 0.01, ****P < 0.0001 (two-way ANOVA with Bonferroni post hoc test in a, f; one-way ANOVA with Tukey post hoc test in c, d; two-tailed Student’s t-test in e).

Expression of IKK2ca restored the frequencies of total T_reg and eT_reg cells in tdLNs, but not in tumour tissue (Extended Data Fig. 9a). It did not restore FOXP3 expression, only partially reduced TNF and IFNγ co-expression by tumour-infiltrating CARMA1-deficient T_reg cells, and did not prevent their anti-tumour activity (Extended Data Fig. 9b–d), which emphasizes the importance of CBM-complex effector functions other than NF-κB activation¹⁶ in stabilizing tumour-reactive T_reg cells.

To examine whether CARMA1 deletion acutely destabilizes intratumoral T_reg cells, we generated Foxp3^GFP-creERT2 × CARMA1^f/f (hereafter termed F^creERT2 × C1^f/f) mice and treated these with tamoxifen to trigger Cre-mediated CARMA1 deletion when tumours were already established (Fig. 3a and Extended Data Fig. 9e). To prevent subsequent recruitment of additional T_reg cells from tdLNs, we concurrently blocked lymphocyte tissue egress using the functional S1P receptor antagonist FTY720, as previously described⁴. Within two days of treatment, tumour growth decelerated (Fig. 3b). A similarly rapid, albeit slightly less pronounced, growth effect as well as increased secretion of effector cytokines by T_reg cells resulted from deletion of CARMA1 in only half of the T_reg cells (Extended Data Fig. 9f, g). Intratumoral destabilization of T_reg cells was accompanied by pronounced induction of macrophage expression of cell-surface major histocompatibility complex class II (MHC-II) molecules, both after constitutive or acute deletion of one or both alleles of CARMA1 in T_reg cells (Fig. 3c and Extended Data Fig. 9h). Furthermore, the expression of MHC-I molecules on tumour cells increased, predictably sensitizing them to cytotoxic-T-lymphocyte (CTL)-mediated lysis (Fig. 3d). Although T_reg cell-derived IFNγ thus caused widespread tumour inflammation, it also triggered tumour cell-expression of PD-L1, a ligand for the T cell inhibitory receptor PD-1, which suggests that concurrent induction of adaptive immune resistance³ limited improved tumour control resulting from enhanced anti-tumour immune effector functions (Fig. 3d).

Fig. 3: CARMA1-deleted T_reg cells rapidly induce tumour inflammation but also adaptive immune resistance.

a, b, D4M.3A melanoma growth in F^creERT2 × C1^+/+ and C1^f/f mice treated with tamoxifen from days 8–12 as well as with FTY720 starting the same day until the end of the experiment. Arrow in b indicates the start of treatment. c, d, MHC-II surface expression on F4/80⁺ tumour macrophages (c), and MHC-I and PD-L1 expression on D4M.3A tumour cells (expressing blue-fluorescent H2B-Cerulean) (d) three days after the initiation of tamoxifen treatment. MFI, mean fluorescence intensity. Data represent two independent replicates with similar results. Data are mean and individual replicates (c, d) or s.e.m. (b). *P < 0.05 (two-tailed Student’s t-test).

Considering the induction of PD-L1 on tumour cells, we hypothesized that antibody-mediated blockade of PD-1 may synergize with the anti-tumour effects of IFNγ-secreting T_reg cells. Indeed, treatment with anti-PD-1 antibody simultaneously with CARMA1 deletion in T_reg cells enabled much more rapid and consistent control of D4M.3A melanoma than either treatment alone (Fig. 4a). Targeting the CBM complex in T_reg cells may thus be highly effective at enhancing the potency of immune checkpoint therapy (ICT) in patients with cancer.

Fig. 4: CARMA1 deletion in T_reg cells and pharmacological inhibition of MALT1 protease synergize with anti-PD-1 ICT.

a, Female F^creERT2 × C1^+/+ and C1^f/f mice were implanted with D4M.3A melanoma, and treated with tamoxifen starting on day 9 until the end of the experiment, as well as with three doses of the anti-PD-1-antibody 29F.1A12 or isotype control, and tumour growth was recorded. b, CBM complex effector pathways and predicted effects (red arrows) of the MALT1 protease inhibitors mepazine and MI-2. c, d, D4M.3A tumour growth in C57BL/6 (c) or RAG1-deficient (d) hosts treated with MALT1 inhibitors (MI-2 or mepazine). Dimethylsulfoxide (DMSO) was used as a vehicle control. e, f, Effects of mepazine treatment for three days on intratumoral T_reg cell frequency and their in situ expression of effector cytokines (e), and on the expression of MHC-I and PD-L1 on tumour cells (f). g, h, Synergistic control of tumours by combined anti-PD-1 and mepazine treatment of poorly immunogenic D4M.3A (g) and immunogenic D4M.3A-SIINFEKL (h) tumours in male C57BL/6 hosts. Numbers in parentheses indicate fraction of mice without relapse for more than 12 months after discontinuation of treatment. Data in a, c–f represent two independent replicates with similar results. Data are mean and individual replicates (e, f) or s.e.m. (a, c, d, g, h). Arrows in tumour growth charts indicate the start of treatment. In a, *P < 0.05 versus C1^+/+, #P < 0.05 versus C1^+/+ + anti-PD-1, and &P < 0.05 versus C1^f/f. In g and h, *P < 0.05 versus vehicle (DMSO), #P < 0.05 versus anti-PD-1, and &P < 0.05 versus and mepazine, respectively. In e, *P < 0.05 (two-way ANOVA with Bonferroni post hoc test in a, c, g, h; two-tailed Student’s t-test in d–f).

Although pharmacological inhibitors of the scaffold protein CARMA1 are, to our knowledge, currently not available, inhibitors of MALT1 paracaspase are predicted to attenuate the majority of CBM-complex-dependent effector pathways (Fig. 4b). Indeed, similar to CARMA1-deficient mice, T_reg cells are virtually absent in mice that express mutant MALT1 proteins that lack paracaspase activity (replicating continual and complete pharmacological inhibition)^21,22,23. We therefore tested the allosteric MALT1 inhibitor mepazine^24,25 and the catalytic site binder MI-2²⁶ for activity against solid tumours. Both inhibitors produced a similar deceleration in melanoma growth to that observed after deletion of CARMA1 in T_reg cells (Fig. 4c), even when CD8⁺ T cells were depleted (Extended Data Fig. 9i). Systemic inhibition of MALT1 will also target cells other than T_reg cells, including melanoma cells²⁷. However, no effect on tumour growth occurred in RAG1-deficient mice that lack lymphocytes (Fig. 4d). Inhibition of MALT1 did not synergize with deletion of CARMA1 in T_reg cells, which suggests that the anti-tumour activity of MALT1 does not result from effects other than attenuated CBM complex function in T_reg cells (Extended Data Fig. 9j). Because MALT1 inhibition is predicted to attenuate, and not enhance lymphocyte effector functions²⁸, we conclude that its effect on tumour growth is probably mediated through destabilization of T_reg cells. Similar to T_reg-cell-specific deletion of CARMA1, treatment with mepazine caused a rapid, albeit less pronounced, induction of TNF and IFNγ expression by tumour-infiltrating T_reg cells (Fig. 4e). Short-term in vitro treatment of T_reg cells triggered only a minor reduction of FOXP3, GITR and CTLA-4 expression (Extended Data Fig. 10a) and did not induce IFNγ secretion (data not shown), which indicates that the latter occurs only under the conditions of the tumour microenvironment. Accordingly, mepazine caused upregulation of the expression of PD-L1 and MHC-I molecules on tumour cells in vivo (Fig. 4f) and induction of Ifng and a wide range of IFNγ-regulated genes indicative of both T_H1 inflammation and adaptive immune resistance in tumour tissue (Extended Data Fig. 10b). In contrast to constitutive deletion of CARMA1, short-term inhibition of MALT1 did not reduce the frequency of T_reg cells, and the expression of T_reg-cell-associated genes in tumour tissue was not reduced (Fig. 4e and Extended Data Fig. 10c). Nevertheless, in addition to overall enhanced infiltration of immune cells, treatment with mepazine specifically increased the frequencies of CTL and natural killer cells in tumour tissue (Extended Data Fig. 10d–h).

A high tumour mutational load favours response to ICT in patients with cancer^29,30, and a low mutational burden remains a major challenge that limits the success of this form of immunotherapy to some cancer types and to a minority of patients. Accordingly, D4M.3A melanoma, which carries a negligible mutational load relative to the C57BL/6J reference exome (D. E. Fisher, personal communication), is completely resistant to anti-PD-1 monotherapy in male hosts (Fig. 4g), in contrast to female hosts, in which Y-antigen-expressing male D4M.3A tumours showed a partial response (Fig. 4a). Concurrent MALT1 inhibition, however, synergized with anti-PD-1 treatment, and arrested tumour growth even in male hosts (Fig. 4g). Anti-PD-1 treatment did not further increase T_reg cell expression of IFNγ, indicating that PD-1 did not restrict the pro-inflammatory function of destabilized T_reg cells (Extended Data Fig. 10i). Furthermore, when we raised the immunogenicity of D4M.3A tumours by expressing the chicken ovalbumin-derived SIINFEKL epitope as a surrogate mutational neoantigen, we observed an initial response to anti-PD-1 monotherapy, but 40% of tumours relapsed. A combination of anti-PD-1 antibodies with mepazine, however, produced accelerated rejection and prevented relapse (Fig. 4h). Finally, to explore the effects of MALT1 inhibition on other cancer types, we treated mice implanted with MC38 colon carcinoma. Although anti-PD-1 monotherapy had only a moderate effect on late-stage tumours, combination with mepazine enabled profound tumour control and relapse-free rejection in most mice (Extended Data Fig. 10j). Hence, systemic inhibition of MALT1 inflames the tumour environment and renders poorly immunogenic tumours responsive to anti-PD-1 therapy while enhancing responses of immunogenic tumours and minimizing the frequency of relapse, a common problem in clinical ICT³¹.

We propose that inhibition of MALT1 protease or of other CBM complex functions could be a useful therapeutic strategy to elicit an intratumoral T_H1 autoimmune reaction mediated by locally destabilized, preferentially self-reactive T_reg cells. Pro-inflammatory effects of destabilized T_reg cells seem to outweigh any potential attenuation of immune effector cell activities through MALT1 inhibition. Owing to its selectivity for intratumoral T_reg cells, this treatment may increase the fraction of patients with cancer who respond to PD-1/PD-L1-targeted ICT or other forms of immunotherapy, without inducing systemic autoimmune toxicity.

Methods

Mice

Foxp3^YFP-cre (ref. ³²), Foxp3^GFP-creERT2 (ref. ³³), ROSA26-stop^f/f-YFP (ref. ³⁴), ROSA26-stop^f/f-IKK2ca (ref. ³⁵), Ifng^−/−36, B6-scurfy (ref. ³⁷) and C57BL/6/J mice were purchased from Jackson Laboratories. R. J. Xavier and J. J. Moon provided CARMA1^f/f (ref. ³⁸) and Rag1^−/− mice, respectively. Animals were housed in specific-pathogen-free facilities at the Massachusetts General Hospital (MGH) and all experimental studies were approved and performed in accordance with guidelines and regulations implemented by the MGH Institutional Animal Care and Use Committee (IACUC). For survival studies, the age of mice at euthanasia mandated by a moribund state of health was recorded in Kaplan–Meyer plots.

Tumour cell lines

The Braf^V600E × Pten^−/− melanoma cell line D4M.3A (ref. ¹⁸) was provided by D. E. Fisher. For some experiments, D4M.3A cells were lentivirally transduced to express a blue fluorescent histone H2B-Cerulean fusion protein (D4M.3A-H2B-Cerulean), as previously described³⁹, to facilitate detection by flow cytometry. To generate D4M.3A-SIINFEKL tumours expressing the chicken ovalbumin-derived H-2K^b-restricted SIINFEKL peptide, we transduced D4M.3A cells with a VSV-G pseudotyped pHAGE-EF1α lentiviral vector engineered to express a fusion of histone H2B and Cerulean separated by two copies of the SIINFEKL minigene and its native flanking sequences in the ovalbumin protein to facilitate processing for antigen presentation. The colon adenocarcinoma cell line MC38⁴⁰ was obtained from A. D. Luster. All tumour lines were grown in DMEM with 10% fetal calf serum (FCS) and used for experiments when in exponential growth phase.

Tumour growth studies and treatments

One million D4M.3A, D4M.3A-H2B-Cerulean, D4M.3A-SIINFEKL or MC38 tumour cells were injected subcutaneously in 100 μl HBSS without Ca²⁺ into the flanks of mice. Wherever possible, animals were randomized into treatment groups. Tumour volumes were measured every second to third day after the start of treatments and calculated as V = (length × width²)/2.

Tamoxifen (1 mg per mouse in 100 μl of a 9:1 mixture of olive oil and ethanol) was injected intraperitoneally daily as indicated. FTY720 (1 mg kg⁻¹ bodyweight) in 150 μl H₂O was injected intraperitoneally every other day until the end of the experiment. Anti-IFNγ antibody (500 μg per mouse; clone XMG1.2) was injected intraperitoneally on day 14 after birth or on the day of tumour implantation and then every other day thereafter until the end of the experiment. Anti-PD-1 (200 μg; clone 29F.1A12) or rat IgG2a isotype control (200 μg; clone 2A3) was injected intraperitoneally three times in 100 μl PBS every other day at the indicated time points. Anti-CD8α (150 μg; clone YTS169.4) was injected in 100 μl PBS every other day from the indicated time point until the end of the experiment. Mepazine (16 mg kg⁻¹ bodyweight in 5% dimethylsulfoxide (DMSO)) or MI-2 (20 mg kg⁻¹ in 5% DMSO in purfied H₂O) was injected intraperitoneally daily starting at the indicated time points until the end of the experiment, unless indicated otherwise. For adoptive T_reg cell transfer studies, CD4⁺ YFP⁺ T_reg cells were purified to more than 95% purity through magnetic-activated cell sorting (Miltenyi) from LNs and spleen of F^cre × C1^f/+ or C1^+/+ mice and 10⁶ cells per mouse were injected intravenously into the tail vein the day before tumour implantation.

Preparation of single-cell suspensions, antibody staining and flow cytometry

Heparinized peripheral blood collected through sub-mandibular vein puncture was treated with ACK red blood cell lysis buffer. LNs and spleens were passed through 40-μm cell strainers, followed by red blood cell lysis (spleens only). Tumours and lung tissue were minced into small fragments and treated with 1.5 mg ml⁻¹ collagenase IV and 50 U ml⁻¹ DNase I for 30 min at 37 °C under agitation. Skin tissue was digested in medium containing 2% FCS, 10 mM HEPES, 0.5 mg ml⁻¹ hyaluronidase, 1.5 mg ml⁻¹ collagenase IV, and 50 U ml⁻¹ DNase I for 45 min at 37 °C under agitation. Residual tissue fragments were mechanically dissociated.

Cell surface proteins were stained for 20 min at 4 °C with the following antibodies against: CD11b (M1/70), CD120b/TNFR2 (polyclonal Armenian hamster IgG), CD274/PD-L1 (10F.9G2), CD357/GITR (DTA-1), CD4 (GK1.5), CD45 (30-F11), CD62L (MEL-14), CD73 (TY/11.8), CD8α (53-6.7), CD90.2 (30-H12), F4/80 (BM8), H-2K^b (AF6-88.5), -I-A/I-E (M5/114.15.2), Ly-6C (HK1.4), Ly-6G (1A8), CD45R/B220 (RA3-6B2), CD64 (FcγRI) (X54-5/7.1), CD11c (N418), CD103 (2E7), NK-1.1 (PK136), CD335 (NKp46) (29A1.4), CD3 (17A2), CD19 (1D3/CD19), CD45RB (16A), and CD44 (IM7) (BioLegend), CD11c (HL3) and CD25 (PC61.5) (eBIoscience).

Intracellular and nuclear proteins were stained for 60 min at room temperature after permeabilization and fixation (Mouse regulatory T cell staining Kit; eBioscience) using antibodies against: CD152/CTLA-4-(UC10-4B9), TNF (MP6-XT22), IL-4 (11B11), IL-17A (TC11-18H10.1), IFNγ (XMG1.2), T-bet (4B10), and Ki67 (16A8) (BioLegend), BIM (C34C5), CARD11/CARMA1 (1D12) (Cell Signaling), FOXP3 (FJK-16 s, eBioscience), GATA-3 (L50-823) and Ki67 (B56) (BD Biosciences), RORγt (AFKJS-9) and GFP (rabbit polyclonal) (Invitrogen). Polyclonal goat anti-rabbit Ig (H+L) secondary antibody (Life Technologies) was used to reveal primary anti-CARMA1 staining.

Preceding antibody staining, dead cells were stained using the fixable viability violet dye Zombie Red (Biolegend) for 15 min at room temperature, followed by blocking of Fc receptors with TruStain FcX (Biolegend) for 20 min at 4 °C. Cells were analysed on LSR II, LSRFortessa or LSRFortessa X-20 flow cytometers (BD Biosciences), and data were analysed with FlowJo software v.9.9.5.

Phospho-protein analysis

LN single-cell suspensions were stained using the fixable viability dye ZombieRed (Biolegend) for 15 min at room temperature, and added for 30 min at 37 °C to tissue culture plates pre-coated overnight with anti-CD3ε (clone 145-2C11) and anti-CD28 (clone 37.51) antibodies (at 10 μg ml⁻¹ of each antibody), or to uncoated control plates. Samples were then fixed in 4% paraformaldehyde (PFA) for 10 min at room temperature, and permeabilized for 20 min through dropwise addition of 1 ml ice-cold methanol. Cells were then stained for CD90.2 (30-H12), CD4 (GK1.5), CD8α (53-6.7), CD44 (IM7) (BioLegend), Foxp3 (FJK-16 s, eBioscience), pFoxo1 (Thr24)/Foxo3a (Thr32), p-c-Jun (Ser73) (D47G9) (Cell Signaling) and GFP (rabbit polyclonal Ab) (Invitrogen).

Analysis of in situ and ex vivo stimulated cytokine secretion

To detect in situ cytokine secretion, mice were slowly injected intravenously with 500 μg of brefeldin A in 250 μl PBS 6 h before euthanization and intracellular cytokine staining.

To detect cytokine secretion in T cells upon ex vivo re-stimulation, single-cell suspensions from tumours and LNs were resuspended in RPMI 1640 with 10% FCS and added to anti-CD3 (clone 145-2C11)/anti-CD28 (clone 37.51) antibody-coated (overnight at 10 μg ml⁻¹ antibody) tissue culture plates for 8 h at 37 °C in the presence of 1 μg ml⁻¹ Golgiplug and Monensin (both from Biolegend) and cells processed for intracellular cytokine staining.

Analysis of exT_reg cells

CD4⁺ YFP^bright cells were first purified by FACS from LNs and spleens of Foxp3^YFP-cre/+ × CARMA1^f/f (or ^{f/+ or +/+}) × ROSA26^YFP mice and stained for FOXP3 expression for flow cytometry analysis, as described above.

In vivo and in vitro suppression

For in vivo suppression studies, 3 × 10⁵ Miltenyi (negative selection) enriched CD4⁺ and FACS sorted (>98% purity) CD45RB^high YFP⁻ cells from LNs and spleens of Foxp3^YFP-cre/cre mice were intravenously injected into the tail vein of Rag1^−/− mice with or without 1 × 10⁵ Miltenyi (negative selection) enriched CD4⁺ and FACS sorted (>98% purity) YFP^bright T_reg cells from LNs and spleens of Foxp3^YFP-cre/+ × CARMA1^f/f (or CARMA1^f/+ or CARMA1^+/+) × ROSA26^YFP mice.

For in vitro suppression studies, 1 × 10⁴ FACS-sorted (>98% purity) CD4⁺ YFP⁻ conventional T cells from LNs and spleens of Foxp3^YFP-cre/cre mice were labelled with 5 μM CellTrace Violet and stimulated with 250 ng ml⁻¹ of anti-CD3 monoclonal antibody (145-2c11, Biolegend) in presence of 2.5 × 10⁴ T-cell depleted splenocytes and different concentrations (from 1:1 to 1:16) of Miltenyi (negative selection) enriched CD4⁺ and FACS sorted (>98% purity) YFP^bright Treg cells from LNs and spleens of Foxp3^YFP-cre/+ × CARMA1^f/f (or CARMA1^f/+ or CARMA1^+/+) × ROSA26-stop^f/f-YFP mice. CD4⁺ YFP^– conventional T cell proliferation was read out after 72 h, as previously described⁴¹. In brief, percentage of suppression was scaled from 0 (proliferation of conventional T cell in absence of T_reg cells) to 100 (complete absence of proliferation).

In vitro apoptosis

Enriched CD4⁺ and FACS-purified (>99% purity) YFP^bright T_reg cells from LNs of Foxp3^YFP-cre/+ × CARMA1^f/f (or CARMA1^+/+) × ROSA26-stop^f/f-YFP mice were added for 6 and 18 h at 37 °C to tissue culture plates pre-coated overnight with anti-CD3ε (clone 145-2C11) and anti-CD28 (clone 37.51) antibodies (at 10 μg ml⁻¹ of each antibody). Viability of CD44^lowCD62L⁺ cT_reg and CD44^highCD62L^– eT_reg cells was then read out by annexin V and Zombie Red staining (Biolegend).

RNA sequencing studies: sample collection

CD4⁺ T cells from LNs and spleens of F^cre/+ × C1^+/+, C1 ^f/+ or C1 ^f/f mice were pre-enriched by immunomagnetic cell sorting (Miltenyi negative selection) and then 5 × 10³ YFP⁺CD4⁺CD44^lowCD62L⁺ cT_reg cells per animal and the same number of YFP⁺CD4⁺CD44^highCD62L^– eT_reg cells were sorted to >99% purity directly into 10 μl lysis buffer consisting of TCL buffer (Qiagen) and 1% of β-mercaptoethanol. Samples were then flash-frozen and kept at −80 °C before further processing following a modified version of the Smart-Seq2 protocol^11,12,14, as described below. A total of 18 samples was collected, but 2 samples were discarded for technical reasons.

Reverse transcription

Samples were thawed on ice for 2 min, and then centrifuged at 2,500 rpm at 4 °C for 1 min and the RNA concentration normalized. RNA (1.9 μl per sample) was moved to a full-skirt 96-well plate (Eppendorf). Each sample was then mixed with 1 μl 10 μM RT primer 5′-AAGCAGTGGTATCAACGCAGAGTACTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTVN-3′ (IDT), 1 μl 10 mM dNTP (Life Technologies/Thermo Fisher Scientific), and 0.1 μl SUPERase·In RNase-Inhibitor (20 U μl⁻¹, Life Technologies/Thermo Fisher Scientific). Samples were denatured at 72 °C for 3 min using an Eppendorf Mastercycler and placed immediately on ice afterwards. The Reverse Transcription Mix (7 μl) was subsequently added to every well, consisting of: 2 μl 5× RT buffer (Thermo Fisher Scientific), 2 μl 5 M betaine (Sigma-Aldrich), 0.9 μl 100 mM MgCl₂ (Sigma-Aldrich), 1 μl 10 μM TSO (5′-AAGCAGTGGTATCAACGCAGAGTACATrGrG+G-3′, Exiqon), 0.25 μl SUPERase·In RNase-Inhibitor (20 U μl⁻¹, Life Technologies/Thermo Fisher Scientific), 0.1 μl Maxima H Minus Reverse Transcriptase (200 U μl⁻¹, Thermo Fisher Scientific) and 0.75 μl nuclease-free water. Reverse transcription was carried out by incubating the plate at 50 °C for 90 min, followed by heat inactivation at 85 °C for 5 min.

PCR pre-amplification and cDNA purification

PCR Mix (14 μl, consisting of 0.5 μl 10 μM PCR primer 5′-AAGCAGTGGTATCAACGCAGAGT-3′ (IDT), 12.5 μl 2× KAPA HiFi HotStart ReadyMix (KAPA Biosystems) and 1 μl nuclease-free water) was added to each well for a final PCR reaction volume of 25 μl. The reaction was carried out with an initial incubation at 98 °C for 3 min, followed by 16 cycles at 98 °C for 15 s, 67 °C for 20 s, and 72 °C for 6 min, and a final extension at 72 °C for 5 min. PCR products were purified by mixing them with 20 μl (0.8×) of Agencourt AMPureXP SPRI beads (Beckman-Coulter), followed by a 6-min incubation period at room temperature. The plate was then placed onto a magnet for 6 min before removing the supernatant. SPRI beads were washed twice with 100 μl of freshly prepared 70% ethanol, with care being taken to avoid loss of beads during the washes. After removing all residual ethanol traces, SPRI beads were left to dry at room temperature for 10 min. The beads were then resuspended in 20 μl of TE buffer (Teknova) and incubated at room temperature for 5 min. The plate was placed on the magnet for 5 min before transferring the supernatant containing the amplified cDNA to a new 96-well plate. This cDNA SPRI clean-up procedure was repeated a second time to remove all residual primer dimers. The concentration of amplified cDNA was measured on the Synergy H1 Hybrid Microplate Reader (BioTek) using the Qubit dsDNA High Sensitivity Assay Kit (Life Technologies/Thermo Fisher Scientific). The cDNA size distribution of few selected wells was assessed on a High-Sensitivity Bioanalyzer Chip (Agilent), and the expected size distribution sharply peaked around 2 kb.

Sequencing library preparation

Library preparation was carried out using the Nextera XT DNA Sample Kit (Illumina) with custom indexing adapters, allowing the 18 libraries to be simultaneously generated in a 384-well PCR plate (Eppendorf). For each library, the amplified cDNA was normalized to a 0.15–0.20 ng μl⁻¹ concentration range. The tagmentation reaction consisted of mixing 0.625 μl of normalized cDNA with 1.25 μl of Tagmentation DNA (TD) buffer and 0.625 μl of Amplicon Tagment enzyme Mix (ATM). The 2.5-μl reaction was incubated at 55 °C for 10 min and then immediately placed on ice upon completing this incubation step. The reaction was quenched with 0.625 μl of Neutralize Tagment (NT) buffer and incubated at room temperature for 10 min. The libraries were amplified by adding 1.875 μl of Nexstera PCR Master (NPM) Mix, 0.625 μl of 10 μM i5 adaptor 5′-AATGATACGGCGACCACCGAGATCTACAC[i5]TCGTCGGCAGCGTC-3′ (IDT), in which [i5] signifies the 8-bp i5 barcode sequence (see below for sequences), and 0.625 μl of 10 μM i7 adaptor 5′CAAGCAGAAGACGGCATACGAGAT[i7]GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGGG-3′ (IDT), in which [i7] represents the reverse-complement of the 8-bp i7 barcode sequence (see below for sequences used). The PCR was carried out at an initial incubation at 72 °C for 3 min, 95 °C for 30 s, followed by 12 cycles of (95 °C for 10 s, 55 °C for 30 s, 72 °C for 1 min), and a final extension at 72 °C for 5 min. Following PCR amplification, 2.5 μl of each library were pooled together in a 1.5-ml Eppendorf tube. The pool was mixed with 67.5 μl (0.9× ratio for 2.5 μl of 30 samples pooled together) of Agencourt AMPureXP SPRI beads (Beckman-Coulter) and incubated at room temperature for 5 min. The pool was then placed on a magnet (DynaMag-2, Life Technologies) and incubated for 5 min. The supernatant was removed and the SPRI beads were washed twice with 1 ml of freshly prepared 70% ethanol. After removing all residual ethanol traces, the SPRI beads were left to dry at room temperature for 10 min. The beads were resuspended in 100 μl of nuclease-free water and incubated at room temperature for 5 min. The tube was then placed back on the magnet for 3 min before transferring the supernatant to a new 1.5-ml Eppendorf tube. This SPRI clean-up procedure of the library was repeated a second time to remove all residual primer dimers, using the same approach. The concentration of the pooled libraries was measured using the Qubit dsDNA High Sensitivity Assay Kit (Life Technologies/Thermo Fisher Scientific), and the library size distribution measured on a High-Sensitivity Bioanalyzer Chip (Agilent), showing the expected size distribution of 300–500 bp. The 18 pooled samples were sequenced as paired-end on an Illumina NextSeq 500 instrument using the NextSeq 500/550 High Output v2 kit (75 cycles).

i5 barcodes: AAGTAGAG, ACACGATC, TGTTCCGA. i7 barcodes: GAATTGCT, GTCAAGTT, ATCCGACA, CAAGGCGA, AGTGTCTT, GACCGAGA.

RNA sequencing analysis

Raw sequencing reads were demultiplexed and converted to FASTQ files using Illumina bcl2fastq2 Illumina software (version 2.17.1.14). FASTQ sequencing reads were then aligned to mm10 reference genome using the STAR aligner with default parameters⁴². RSEM (v.1.2.8) was used to quantify gene expression level from aligned reads and generate count expression matrices for each experimental condition⁴³. We filtered out lowly expressed genes with a count per million (CPM) < 0.5 in more than two conditions, leaving a total of 14,168 genes for further analysis. The distribution of log₂ normalized CPM data was visualized to assess for coverage, and all conditions had similar distributions.

Gene expression analysis

Gene expression matrices were analysed using the limma package in R⁴⁴. The global topology of quantile normalized data was visualized using the multidimensional scaling (plotMDS) function in limma after removing batch effects using the removebatchEffect function in limma with default parameters taking into account design and batch matrices. Differential gene expression was performed using empirical Bayesian statistics (eBayes) function in limma simultaneously correcting for batch using blocking terms for batch covariates. Differentially expressed genes with log fold change greater than 1 and a P value below cut-off were visualized using the heatmap.2 function in gplots. All P values were corrected for multiple hypothesis testing using Benjamini–Hochberg correction. For R scripts used to perform the gene expression analyses see Supplementary Methods. The same differential expression steps were used to re-analyse the gene expression data from GEO accession GSE82008 to obtain the list of differentially expressed genes between c-Rel and p65 knockout versus wild-type resting and activated T_reg cells. A list of 831 ‘eTreg signature’ genes from ref. ¹⁶ was obtained by direct correspondence with the authors. Overlap between differentially expressed genes, including the list of eT_reg cell signatures from the current study and ref. ¹⁶, was visualized using the vennDiagram function in limma.

Quantitative RT–PCR

For analysis of gene expression, RNA was isolated (AllPrep, DNA/RNA Mini kit; Qiagen) from CD4⁺ GFP⁺ T_reg cells sorted to >99% purity from tdLNs and tumours, or from homogenized tumour tissue, and reverse transcribed using iScript cDNA Synthesis Kit (Bio-RAD). Quantitative reverse transcription (RT–PCR) was performed using iQ SYBR green supermix (Bio-RAD) and primers: CARMA1 Fwd 5′-ACATGCTGAGCCGTTACATCA-3′, CARMA1 Rev 5′-CCACATAGCCCCTTTGTCCC-3′, Ifng Fwd 5′-CGGCACAGT CATTGAAAGCCTA-3′, Ifng Rev 5′-GTTG CTGATGGCC TGATTGTC-3′, Ctla4 Fwd 5′-GCTTCCTAGATTACCCCTTCTGC-3′, Ctla4 Rev 5′-CGGGCATGG TTCTGGATCA-3′, CD25-Fwd 5′-CCACATTCAAAGCC CTCTCCTA-3′, CD25-Rev 5′-GTTTTCCCACACTTCATCTTGC-3′, Foxp3 Fwd 5′-TTGG CCAGCGCCA TCTT-3′, Foxp3 Rev 5′-TGCCTCCTCCAGAGAGAAGTG-3′, GITR (also known as Tnfrsf18) Fwd 5′-AAGGTTCAGAACGGAAGTG-3′, GITR Rev 5′-GGGTCTCCACAGTGGTACT-3′, CD73 (also known as Nt5e) Fwd 5′-CAA ATCCCACACAACCACTG-3′, CD73 Rev 5′-TGCTCACTTGGTCACA GGAC-3′, Gzmb Fwd 5′-CATGTAGGGTCGAGAGTGGG-3′, Gzmb Rev 5′-CCTCCTGC TACTGCTGAC CT-3′, Pdl1 (also known as Cd274) Fwd 5′-TGCTGCATAATCAGCTACGG-3′, Pdl1 Rev 5′-GCTGGTCACATT GAGAAGCA-3′, Socs1-Fwd 5′-ACAAGCTGCTACAACCAGG G-3′, Socs1 Rev 5′-ACT TCTGGCTGGAGACCTCA-3′, Tap1 Fwd 5′-GTGGCCGCAGTG GGA CAAGAG-3′, Tap1 Rev 5′-AGGGCACTGGTGGCATCATC-3′, Stat1 Fwd 5′-TGGTGAAATTGCAAG AGCTG-3′, Stat1 Rev 5′-CAGACTTCCG TTGGTGGATT-3′, Irf1 Fwd 5′-CAG AGGAAAG AGAGAAAGTCC-3′, Irf1 Rev 5′-CACACGGTGACAGTGCTGG, Cxcl10 Fwd 5′-CATC CTGCTGGGT CTGAGTG-3′, Cxcl10-Rev 5′-ATTCTCACTGGCCCGTCATC, Nos2 Fwd 5′-CAAGAGAGTGCTGTTCCAGGT-3′ and Nos2 Rev 5′-GAGCACGCTGAGT ACC TCATT-3′, Gapdh Fwd 5′-TGGTGAAGGTCGGTGAAC-3′ and Gapdh Rev 5′-CC ATGTAGTTGAGGTCAATGAAGG-3′. Results were expressed as 2^−ΔCt relative to the house keeping gene Gapdh.

Histology

Tissue samples obtained from all organs were fixed in 10% buffered formalin for 48 h, trimmed and placed into microcassettes, and embedded in paraffin wax. Sections of 5 μm were stained with haematoxylin and eosin according to standard procedures.

Immunofluorescence

Kidney, liver, and stomach from a Rag1^−/− mouse were embedded in OCT and flash-frozen in cold methylbutane equilibrated on dry ice. Sections of 10 μm were permeabilized with pre-cooled 90% methanol for 10 min at −20 °C, blocked in TruStain FcX (93, Biolegend) with 1% goat serum and 0.25% BSA in PBS for 60 min, incubated with sera (1:100 dilution) from Foxp3^YFP-cre × CARMA1^f/f (or CARMA1 ^f/+ or CARMA1^+/+) mice for 120 min and stained with anti-mouse IgG (H+L)-Alexa Fluor 647 (1:500) (A-21235, Thermo Fisher) and DAPI (Sigma) for 120 min. Sections were mounted on coverslips in Prolong (Thermo Fisher) and imaged with LSM 780 AxioObserver confocal microscope (Carl Zeiss) using a 20× lens (Apochromat, 0.8 W).

Statistical analysis

A two-tailed Student’s t-test was used for comparisons between two groups, and a two-way ANOVA with Bonferroni post hoc test (multiple time points) or one-way ANOVA with Tukey post hoc test (single time points) was used for comparisons across multiple groups, unless otherwise indicated. A log-rank (Mantel–Cox) test was used to compare survival curves. All statistical tests were performed with GraphPad Prism software, and P < 0.05 was considered statistically significant. No statistical methods were used to predetermine sample size. Investigators were not blinded to allocation during experiments and outcome assessment.

Reporting summary

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

Extended data figures and tables

Extended Data Fig. 1 Lymphoproliferative disease after T_reg-cell-specific deletion of CARMA1.

a, CARMA1 protein in T_reg cells and CD4⁺ T_conv cells from LNs of F^cre × C1^+/+, C1^f/+ and C1^f/f mice. b, Weight curves (n = 5 per group). c, d, Appearance of 21-day-old mice (c), and their spleens and LNs (d). e, Histological appearance of liver, skin and lung at 21 days of age of indicated mice. Scale bars, 150 μm and 50 μm (insets). f, Kidney, liver and stomach tissue sections of healthy C57BL/6 Rag1^−/− mice were reacted with serum from 21-day-old mice of the indicated genotypes, and self-tissue-reactive IgG revealed by anti-mouse IgG staining (green). Nuclei were stained with DAPI (red). In b, *P < 0.05 versus C1^+/+ and C1^f/+ (two-way ANOVA with Bonferroni post hoc test).

Extended Data Fig. 2 Myeloid cell expansion and effector cytokine secretion by T_conv and T_reg cells after T_reg-cell-specific deletion of CARMA1.

a–c, Size of the CD11b⁺ splenic myeloid compartment and proportions of Ly6G⁺ neutrophils, CD11c⁺MHC II^high dendritic cells (DCs), Ly6C^high monocytes, Lyc6G^lowSSC^high eosinophils, and Ly6C^lowSSC^low macrophages in F^cre × C1^+/+, C1^f/+ or C1^f/f mice. d, Expression of MHC-I, MHC-II and PD-L1 on splenic myeloid subsets. e, Frequency of CD4⁺ and CD8⁺ T_conv cells with a CD44^highCD62L⁻ effector memory phenotype in LNs of indicated mice at age 12 and 21 days. f, Effector cytokine expression of T_conv cells from 21-day-old mice after 8-h ex vivo stimulation on anti-CD3/CD28-coated plates. g, LN cellularity. h, Absolute numbers of T_reg cells in LNs. i, Co-expression of indicated transcription factors by T_reg cells from LNs of indicated mice. j, Expression of CD44 and CD62L by F^cre × C1^f/f T_reg cells expressing T-bet (green dots), GATA-3 (blue dots), or RORγt (red dots), compared to total C1^f/f T_reg cells (contour plots). *P < 0.05, **P < 0.01, ***P < 0.001, ***P < 0.0001 (one-way ANOVA with Tukey post hoc test).

Extended Data Fig. 3 Role of CARMA1 in eT_reg cell differentiation.

a, Female heterozygous F^cre/+ × C1^f/f mice express YFP–Cre and delete C1^f/f in half of the T_reg cells owing to X-chomosomal location of the Foxp3^YFP-cre allele and random X chromosome inactivation, whereas the other half of the T_reg cells remains functional. b, Frequency of CD4⁺ and CD8⁺ T_conv cells with a CD44^highCD62L^– effector memory phenotype in peripheral blood of ageing F^cre/+ × C1^+/+, C1^f/+ or C1^f/f mice (n = 4 per group). c, Appearance of spleens and LNs of indicated mice at one year of age. d–f, Frequency and absolute numbers (e) of eT_reg cells and expression of FOXP3, indicated markers of eT_reg cell differentiation, as well as proliferation marker Ki67, pro-apoptotic protein BIM and anti-apoptotic protein BCL2 by YFP⁺ cT_reg and eT_reg cells (d, f) from nine-week-old F^cre/+ × C1^+/+, C1^f/+ or C1^f/f mice. Note that some data on eT_reg cells in e and f are also shown in Fig. 1g, h and shown here to facilitate the comparison to cT_reg and YFP^– T_reg cells in g and h. g, h, Frequency and absolute numbers of eT_reg cells (g) and eT_reg cell markers on YFP^– cT_reg and eT_reg cells (h) from the same mice as shown in d and f. *P < 0.05, **P < 0.01, ***P < 0.001, ***P < 0.0001 (one-way ANOVA with Tukey post hoc test).

Extended Data Fig. 4 In vitro and in vivo suppression, apoptotic rate and exT_reg cell formation of CARMA1-deficient T_reg cells.

a, CD4⁺CD45RB^highYFP^– T_conv cells and CD4⁺CD45RB^lowYFP^bright T_reg cells were double-sorted to more than 98% purity from LNs and spleens of F^cre × C1^+/+ × ROSA26-stop^f/f-YFP mice, which allow for clear differentiation of Cre-expressing T_reg cells based on high expression of soluble enhanced yellow fluorescent protein (eYFP) in addition to the YFP–Cre fusion protein. b, YFP^bright T_reg from F^cre × C1^+/+ or F^cre/+ × C1^f/+ or C1 ^f/f mice and CellTrace Violet-labelled T_conv cells from F^cre × C1^+/+ mice were co-cultured at indicated ratios for three days in the presence of anti-CD3 antibodies and T-cell-depleted splenocytes, and suppression measured as reduction of T_conv cell proliferation. c, T_reg cells of various genotypes and T_conv cells were co-adoptively transferred into Rag-deficient hosts and their respective frequency in peripheral blood was determined eight weeks later. d, CD4⁺YFP⁺ T_reg cells of indicated genotypes were cultured without exogenous IL-2 on anti-CD3/CD28-coated or uncoated plates for 6 or 18 h and examined for reactivity with annexin V and the viability dye ZombieRed. e, CD4⁺YFP^bright cells were sorted from LNs of one-year-old and F^cre/+ × C1^+/+ (or C1 ^f/+, C1 ^f/f) × ROSA26-stop^f/f-YFP mice and subsequently stained for expression of FOXP3 protein to determine the frequency of FOXP3^– exT_reg cells. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (two-way ANOVA with Sidak post hoc test in b, c; two-tailed Student’s t-test in d).

Extended Data Fig. 5 Bulk RNA sequencing analysis of YFP⁺ cT_reg and eT_reg cells from LNs of F^cre/+ × C1^+/+, C1^f/+ and C1^f/f mice.

a, Scaled expression in cT_reg cells (top) and eT_reg cells (bottom) of genes differentially expressed (fold change > 2 and P_adj < 0.05) between C1^+/+ and C1^f/f mice. b, High-resolution, fully annotated heat map of eT_reg signature genes as shown in Fig. 1j.

Extended Data Fig. 6 Expression of apoptotic regulator genes.

a, Comparison of normalized, log-transformed mRNA expression levels of anti- and pro-apoptotic Bcl2 family genes in YFP⁺ cT_reg and eT_reg from F^cre/+ × C1^+/+ mice, based on RNA sequencing analyses. b, Comparison of expression of the same genes in cT_reg (top) and eT_reg (bottom) cells of the indicated genotypes. *P_adj < 0.05.

Extended Data Fig. 7 Restoring NF-κB activation in CARMA1-deficient T_reg cells.

a, Survival of F^cre × C1^+/+ (or C1^f/f) × ROSA26-stop^f/f-IKK2ca mice that express one allele of IKK2ca after expression of Foxp3^cre. b–d, Frequency of CD4⁺ and CD8⁺ T_conv cells with a CD44^highCD62L^– effector memory phenotype in LNs (b), frequency of T_reg among CD4⁺ T cells and of eT_reg cells among total T_reg cells in LNs (c), and effector cytokine expression of T_conv cells after 8-h ex vivo stimulation on anti-CD3/CD28-coated plates (d) in indicated mice. e, f, Expression of indicated phospho-proteins by YFP⁺ T_reg cells from F^cre/+ × C1^+/+ or × C1^f/f mice at baseline and after 30 min. Anti-CD3/CD28 in vitro stimulation. Grey solid histograms show unstained control cells. T_conv cells were used as internal controls, and showed no differences (not shown). *P < 0.05, ***P < 0.001, ****P < 0.0001 (log-rank (Mantel–Cox) test in a; one-way ANOVA with Tukey post hoc test in b–d, f).

Extended Data Fig. 8 Role of IFNγ secretion by FOXP3^int unstable T_reg cells selectively in tumour tissue.

a–c, Frequencies of total YFP⁺ T_reg cells and of YFP⁺ eT_reg cells among CD4⁺ T cells (a), absolute YFP⁺ T_reg and eT_reg cell numbers (b), and normalized FOXP3 expression in YFP⁺ T_reg and eT_reg (c) in tdLN and tumour tissue of 18-day-old D4M.3A tumours in female heterozygous F^cre/+ × C1^+/+, C1 ^f/+ or C1 ^f/f mice. d, tdLN and tumour cellularities. e, Growth of MC38 tumours in female heterozygous F^cre/+ × C1^+/+ or C1 ^f/+ mice. f, g, Tumour-bearing mice were treated with brefeldin A for 5 h, injected intravenously with 3 μg of anti-THY1.2 monoclonal antibodies, and 3 min later collected for direct ex vivo analysis of IFNγ expression in extravascular YFP⁺ T_reg cells in the lung (f) and skin (g). Gates for IFNγ⁺ cells drawn based on fluorescence-minus-one (FMO) controls. h, Normalized FOXP3 expression in IFNγ⁺ and IFNγ^– T_reg cells from tumour tissue. n.d., not detectable. i, Tumour growth in indicated mice implanted with D4M.3A melanoma and19 treated with or without neutralizing anti-IFNγ antibody. j, Frequency of adoptively transferred, YFP⁺ T_reg cells of indicated genotypes in tdLNs of Ifng^−/− hosts at day 18 of tumour growth. k–m, Frequency (k, l) and effector cytokine expression (m) of adoptively transferred, YFP⁺ T_reg cells in tumours in Ifng^−/− hosts. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (one-way ANOVA with Tukey post hoc test in a–d, f, g; two-tailed Student’s t-test in e, h, j, l, m; two-way ANOVA with Bonferroni post hoc test in i).

Extended Data Fig. 9 Tumour response of CARMA1-deficient T_reg cells after restoration of NF-κB activation.

a–d, D4M.3A melanoma cells were implanted into F^cre/+ × C1^+/+ (or C1 ^f/f) × ROSA26-stop^f/f-IKK2ca mice to record frequencies of YFP⁺ T_reg and eT_reg cells among CD4⁺ T cells (a) and their normalized FOXP3 expression (b) in tdLNs and tumour tissue, effector cytokine expression by tumour-infiltrating T_reg cells (c), and tumour growth (d). e, YFP⁺ T_reg cells were sorted from D4M.3A melanoma tissue and tdLNs after five days of treatment of F^creERT2 × C1^+/+ and C1^f/f with tamoxifen and analysed for CARMA1 expression by RT–qPCR. f, Tumour growth in female F^creERT2 or F^creERT2/+ × C1^+/+ and C1 ^f/f mice, in which CARMA1 was deleted in all (F^creERT2) or half (F^creERT2/+) of T_reg cells. Arrow indicates tamoxifen treatment start. g, In situ expression in tumour tissue of effector cytokines by YFP⁺ T_reg five days after CARMA1 deletion in half or all T_reg cells. h, MHC-II expression on tumour-associated macrophages in D4M.3A-implanted F^cre/+ × C1^+/+, C1^f/+ or C1^f/f mice. i, D4M.3A tumour growth in mice treated with depleting anti-CD8 antibody from day 8 and treated with mepazine or vehicle from day 9. j, D4M.3A tumour growth in F^cre/+ × C1^+/+ or C1^f/f mice treated with mepazine or vehicle starting on day 9. Data are mean and individual replicates or s.e.m. In a–c, g–i, P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. In d, *P < 0.05 versus C1^+/+, &P < 0.05 versus C1^+/+ + IKK2ca. In f, *P < 0.05 versus F^creERT2 × C1^+/+, #P < 0.05 versus F^creERT2/+ × C1^f/f. In j, *P < 0.05 versus C1^+/+, #P < 0.05 versus C1^+/+ + mepazine (one-way ANOVA with Tukey post hoc test in a–c, g, h; two-way ANOVA with Bonferroni post hoc test in d, f, j: two-tailed Student’s t-test in i).

Extended Data Fig. 10 Mepazine effects on the tumour microenvironment.

a, YFP⁺ T_reg cells were sorted from F^cre × C1^+/+ mice and treated with 10 μM mepazine or vehicle for 8 or 24 h with or without concurrent anti-CD3/28 monoclonal antibody TCR stimulation (8-h time point only). Expression of FOXP3, markers of eT_reg cell differentiation, cell viability, and frequency of eT_reg cells were recorded. b, c, RT–qPCR analysis of expression of Ifng and genes of adaptive immune resistance (the PD-L1 genes Cd274 and Socs1), antigen presentation (Tap1), IFNγ signalling (Stat1 and Irf1), T-cell recruitment (Cxcl10), M1 macrophage-activation (Nos2) (b) and of Foxp3 and various T_reg-cell-associated genes (c) in whole tumour tissue lysate after three days of treatment with mepazine or vehicle control. d–h, Composition of the tumour tissue immune infiltrate and frequencies of CD45⁺ cells (e) and of various immune cell subsets (f) as well Ki67 expression by T_conv cells (g) and MHC-II expression by macrophages (h) after three days of treatment with mepazine or vehicle control. i, Effector cytokine co-expression by tumour-infiltrating T_reg cells after 12 days of treatment with mepazine and anti-PD-1 antibody. j, Synergistic tumour control of MC38 colon carcinoma through anti-PD-1 and mepazine combination treatment in female C57BL/6 hosts. Numbers in parentheses indicate fraction of mice without relapse for more than 12 months after discontinuation of treatment. In j, *P < 0.05 versus vehicle, #P < 0.05 versus anti-PD-1, and &P < 0.05 versus mepazine. In all other panels, *P < 0.05, **P < 0.01, ***P < 0.001 (two-tailed Student’s t-test in a–c, e–h; one-way ANOVA with Tukey post hoc test in i; two-way ANOVA with Bonferroni post hoc test in j).