Mertk Inhibition Promotes Anti-Leukemia Immunity By Reversing T Cell Suppression Via the PD-1 Axis

The efferocytic receptor MerTK is expressed on monocytes/macrophages and facilitates tolerogenic T cell suppressive clearance of apoptotic cellular debris. Given the rapid cell turn-over in leukemia, we hypothesized that ongoing efferocytosis is leukemia-permissive and inhibition of MerTK on monocytes/macrophages will reverse T cell suppression leading to enhanced anti-leukemia immunity.

To test this hypothesis, we inoculated mice with syngeneic murine BCR-ABL p185 acute lymphoblastic leukemia (ALL) cells, which do not express MerTK. The effects of MerTK inhibition were evaluated in wild-type (WT) mice treated daily with MRX-2843, a MerTK small molecule inhibitor, and in mice with a MerTK knockout mutation (Mertk^-/-). Median survival was prolonged in WT mice treated with MRX-2843 (40 days) compared to vehicle treatment (20 days, p<0.01). In littermate Mertk^-/- mice, all but one mouse survived without evidence of leukemia 150 days after inoculation (p<0.0001).

To explore the T cell suppressive effects of MerTK inhibition on CD11b+ monocytes/macrophages, we used flow cytometry to study ex vivo mixed cell cultures using Mertk^-/- or WT murine splenocytes +/- ALL cells treated +/- MRX-2843. Co-culture with ALL cells for 24 (36.0%), 48 (34.9%) and 72 (36.6%) hours resulted in a significant increase in expression of co-inhibitory ligands PD-L1 and PD-L2 on CD11b⁺ cells compared to cultures without ALL cells (2.03%, 3.2% and 7.5%, p<0.0001), or cultures with ALL cells treated with MRX-2843 (8.3%, 15.3% and 9.5%, p<0.0001). Similarly, PD-L1/PD-L2 co-expression on CD11b⁺ cells from Mertk^-/- mice was not significantly altered in the presence (16.0%) or absence (7.0%) of ALL. Furthermore, there was no significant change in PD-L1/PD-L2 expression when Mertk^-/- co-cultures were treated with MRX-2843 (12.6%). These findings were recapitulated in vivo. There was no change in PD-L1/PD-L2 expression on CD11b⁺ cells from Mertk^-/- mice inoculated with ALL (4.8% spleen; 0.8% marrow) compared to uninoculated Mertk^-/- mice (7.2% spleen; 1.2% marrow) or uninoculated WT littermates (2.0% spleen; 0.7% bone marrow), whereas CD11b⁺ cells from WT mice inoculated with ALL demonstrated a significant upregulation of PD-L1/PD-L2 (24.1% spleen, p<0.0001; and 4.2% bone marrow, p<0.001), which was downregulated upon treatment with MRX-2843 (1.9% spleen, p<0.01; 1.2% bone marrow, p<0.05)

In these experiments, expression of PD-1, the co-inhibitory receptor which binds PD-L1/PD-L2, on T cells was also assessed. In WT mice inoculated with ALL, expression of PD-1 on T cells from spleens (74.2% CD4⁺; 49.2% CD8⁺) and bone marrow (65.9% CD4⁺; 73.0% CD8⁺) was significantly increased compared to spleens (10.0% CD4⁺; 4.1% CD8⁺) and bone marrow (36.4% CD4⁺; 8.5% CD8⁺) from uninoculated control mice (p<0.0001). However, expression of PD-1 was not significantly different in spleens (28.3% CD4⁺; 7.2% CD8⁺) and bone marrow (66.2% CD4⁺; 29.2% CD8⁺) from Mertk^-/- mice inoculated with ALL compared with spleens (20.0% CD4⁺; 4.7% CD8⁺) and bone marrow (62.2% CD4⁺; 12.4% CD8⁺) from uninoculated Mertk^-/- mice.

To determine whether changes in PD-1 expression were indicative of T cell activation or exhaustion, T cells treated with α-CD3 beads were added to splenocyte and ALL cell co-cultures for 24 hours and intracellular IFN-ɣ and TNF-α levels were assessed using flow cytometry. The fraction of T cells producing IFN-ɣ/TNF-α significantly increased in co-cultures treated with MRX-2843 (44.2% CD4⁺; 51.1% CD8⁺) compared to vehicle-treated co-cultures (27.4% CD4⁺, p<0.0001; 36.8% CD8⁺, p<0.0001). Similarly, IFN-ɣ/TNF-α production increased in co-cultures with Mertk^-/- splenocytes and ALL cells (48.3% CD4⁺, p<0.0001; 51.2% CD8⁺, p<0.0001) compared to co-cultures with WT splenocytes and activation was not further augmented by treatment with MRX-2843 (48.9% CD4⁺, 51.5% CD8⁺). Similar results were obtained when T cells from Mertk^-/-mice were used, demonstrating that the effects of T cell activation were dependent on the presence of MerTK on CD11b⁺ cells, but not T cells.

In conclusion, inhibition of MerTK on CD11b⁺ monocytes/macrophages reversed T cell suppression in response to leukemia via the PD-1 axis. MerTK inhibitors are in early phase clinical trials and these data demonstrate a potential immunotherapeutic use for MerTK inhibitors in the treatment of ALL.