In the tumour microenvironment (TME), infiltrating effector T (Teff) cells function to control tumour growth whilst infiltrating regulatory T (Treg) cells act to suppress the immune responses elicited by Teff cells, therein promoting tumour progression. As a consequence the maximal antitumour activity of Teff cells will likely be limited even under immune checkpoint blockade. Hence, one approach to improve patient outcomes to immune checkpoint therapies could be to target tumour-reactive Treg cells. However, complete Treg depletion would result in systemic autoimmunity. To avoid this obstacle, Di Pilato et al. selectively modulated the function of Treg cells in tumours by targeting the CARMA1–BCL10–MALT1 (CBM) signalosome complex in only a fraction of Treg cells and found that this was sufficient to impede tumour growth and prime the TME for immune checkpoint blockade.

Credit: Lara Crow/Springer Nature Limited

In searching for T cell receptor signalling components that could be disrupted to weaken intratumoural Treg cells, the authors focused their attention on the CBM complex, given its key role in controlling lymphocyte activation, proliferation and function. They observed that conditionally deleting one or both copies of the gene encoding the scaffold protein CARMA1 in mature FOXP3+ Treg cells in mice proportionately decreased levels of CARMA1 in CD4+FOXP3+ Treg cells in the lymph nodes (LNs). As expected, complete loss of CARMA1 in Treg cells was fatal, in line with the essential role played by Treg cells in sustaining immune tolerance. However, up to a 50% loss in the expression of CARMA1 could be tolerated.

Next, to investigate how CARMA1-deficient Treg cells might influence tumour development, D4M.3A mouse melanoma or MC38 mouse colon adenocarcinoma cells were subcutaneously implanted into mice with partially or fully CARMA1-depleted Treg cells. This resulted in a pronounced reduction in tumour growth compared with mice carrying CARMA1-sufficient Treg cells. Interestingly the CARMA1-deficient Treg cells present in the tumour tissue secreted the inflammatory cytokines interferon-γ (IFNγ) and tumour necrosis factor (TNF), indicative of a gain of effector function in situ. Strikingly, this reprogramming of Treg cells was specific to the TME as CARMA1-deficient Treg cells in tumour draining LNs (tdLNs) or non-lymphoid tissues did not secrete effector cytokines.

Furthermore, intravenous injection of CARMA1-depleted Treg cells but not CARMA1-sufficient Treg cells into D4M.3A tumour-bearing wild-type or Ifng–/– mice could attenuate tumour growth in both hosts. Yet, the inhibition of tumour growth in Ifng-/- mice could be reversed upon neutralization of IFNγ with an IFNγ antibody, confirming the requirement for this cytokine in the antitumour activity of the converted intratumoural Treg cells. Even if tumours were already established and further Treg cell egress from tdLNs was blocked prior to CARMA1 deletion in Treg cells, subsequent tumour growth was impeded. Additionally, the antitumour phenotype was associated with increased major histocompatibility complex class II (MHC II) expression on macrophages as well as increased MHC I and programmed cell death 1 ligand 1 (PDL1) expression on tumour cells.

To assess whether this strategy had potential therapeutic applicability, the authors chose to pharmacologically inhibit the CBM complex. As CARMA1 inhibitors do not currently exist, the activity of two MALT1 protease inhibitors, mepazine and MI-2, were instead tested on D4M.3A allografts. Both inhibitors attenuated melanoma growth to the same degree as that following genetic targeting of CARMA1 in Treg cells. As MALT1 inhibitors would be expected to block lymphocyte effector functions, this outcome is likely to be specific to the weakened Treg cells — this is reinforced by the finding that decreases in tumour growth were not observed in RAG1-deficient mice, which lack lymphocytes. Mepazine treatment also induced expression of MHC I and PDL1 on tumour cells, expression of Ifng and IFNγ-associated genes in tumour tissue, and, importantly, increased infiltration of cytotoxic T lymphocytes and natural killer cells into tumours.

As T helper 1 (TH1)-type inflammation and adaptive immune resistance simultaneously ensued following targeting of the CBM complex, this suggested tumours could be sensitized to subsequent programmed cell death protein 1 (PD1) blockade. Indeed, combining a PD1 antibody with CARMA1 deletion or MALT1 inhibition in Treg cells produced much more effective, synergistic control of the otherwise poorly immunogenic D4M.3A tumours than with mepazine alone. This outcome was also reproduced by treating mice bearing MC38 tumours with the combination; in this case, most mice were free from relapse for more than 12 months after treatment had stopped.

intratumoural Treg cells lose their immunosuppressive function

An interesting question that arises from this study is why only intratumoural Treg cells lose their immunosuppressive function when the CBM complex is disrupted. Once answered, there is promise that the reprogramming of Treg cells into IFNγ-secreting inflammatory effector cells could be successfully translated to the clinic.