Cancer Immunotherapy with Immunomodulatory Anti-CD137 and Anti-PD-1 Monoclonal Antibodies Requires BATF3-Dependent Dendritic Cells

Abstract

Weak and ineffective antitumor cytotoxic T lymphocyte (CTL) responses can be rescued by immunomodulatory mAbs targeting PD-1 or CD137. Using Batf3(-/-) mice, which are defective for cross-presentation of cell-associated antigens, we show that BATF3-dependent dendritic cells (DC) are essential for the response to therapy with anti-CD137 or anti-PD-1 mAbs. Batf3(-/-) mice failed to prime an endogenous CTL-mediated immune response toward tumor-associated antigens, including neoantigens. As a result, the immunomodulatory mAbs could not amplify any therapeutically functional immune response in these mice. Moreover, administration of systemic sFLT3L and local poly-ICLC enhanced DC-mediated cross-priming and synergized with anti-CD137- and anti-PD-1-mediated immunostimulation in tumor therapy against B16-ovalbumin-derived melanomas, whereas this function was lost in Batf3(-/-) mice. These experiments show that cross-priming of tumor antigens by FLT3L- and BATF3-dependent DCs is crucial to the efficacy of immunostimulatory mAbs and represents a very attractive point of intervention to enhance their clinical antitumor effects.

Significance: Immunotherapy with immunostimulatory mAbs is currently achieving durable clinical responses in different types of cancer. We show that cross-priming of tumor antigens by BATF3-dependent DCs is a key limiting factor that can be exploited to enhance the antitumor efficacy of anti-PD-1 and anti-CD137 immunostimulatory mAbs.

Fig. 1. Antitumor therapy with immunomodulatory mAbs is abrogated in Batf3^-/- mice and is not rescued by IL-12 administration.

WT or Batf3^-/- mice were s.c. inoculated with 5 x 10⁵ MC38 cells. (A, B) Mice were injected i.p. with 100 µg anti-PD-1 and anti-CD137 mAbs, alone or in combination (100 µg each), or with vehicle (untreated) on days 4, 7 and 10 after tumor cell inoculation. (A) Growth plots of individual tumors. (B) Overall survival charts show pooled results from 3 independent experiments with similar results. (C) Tumor-inoculated mice were injected i.p. with 100 µg anti-CD137 mAb on days 7, 10 and 13. The indicated groups of mice additionally received i.t. injections of recombinant mouse IL-12 or saline on days 7, 9 and 11. IL-12 was injected at 25 ng/dose into the tumor nodules. On the left, tumor area (mean ± SEM); on the right, overall survival. Fractions indicate the number of animals surviving at the end of the protocol. * p < 0.05; ** p < 0.01; *** p < 0.001.

Fig. 2. Reduced ability of Batf3^-/- DC to cross-prime CTLs against tumor antigens both in steady state and after treatment with anti-CD137 and anti-PD-1 mAbs.

(A-C) CD11c⁺ DCs from WT and Batf3^-/- mice bearing MC38-OVA tumors were magnetically sorted from tumor-draining LNs and cocultured (see Methods) with purified naïve CD8⁺ OT-I TCR transgenic T cells over a range of DC:T cell ratios. (A) Left: representative flow cytometry dot plots of intracellular IFN-γ staining in OT-I T cells cultured at a 1:4 DC:T cell ratio. Right: percentages of IFN-γ-positive OT-I T cells at all ratios tested. (B) IFN-γ concentrations in the culture supernatants. (C) Percentages of proliferating OT-I cells by dilution of Cell Violet dye (D-F) WT and Batf3^-/- mice grafted with MC38-OVA cells were treated with anti-CD137 (days 5 and 7) and tumor-draining LN analyzed on day 9 (see Methods). (D) Frequency of H-2K^b-OVA-tetramer⁺ cells among CD8⁺ T cells. (E) Intracellular IFN-γ production induced by restimulation with OVA_257-264 peptide in CD8⁺ T cells from tumor-draining LN. (F) PD-1 surface staining on tumor-draining LN CD8⁺ T cells. (G) Frequency of PD-1⁺ lymphocytes among CD8⁺ TILs in mice treated as in D. (H) WT and Batf3^-/- mice grafted with MC38 cells were treated with anti-CD137 and anti-PD-1 mAbs on days 12 and 14 and tumor-infiltrating lymphocytes were analyzed on day 16 to detect CD8 T lymphocytes specific for gp70 antigen (A-C) Two-way and (D-H) one-way ANOVA with Bonferroni post-hoc test. * p < 0.05; ** p< 0.01; *** p < 0.001

Figure 3. sFlt3L and poly-ICLC induce a Batf3-dependent increase in the numbers of tumor-antigen-specific TILs expressing CD137 and PD-1.

WT or Batf3^-/- mice were inoculated with B16-OVA melanoma cells on day 0, concomitant with hydrodynamic gene transfer of sFlt3L or control empty plasmid. On day 7, tumors were injected with poly-ICLC or control. Tumors were retrieved and TILs analyzed on day 10. (A) H2Kb-OVA_257-264 tetramer staining in CD8⁺ TILs. Left: Representative plots. Right: Graphs corresponding to a representative experiment (n = 3) (B) Surface CD137 and PD-1 immunostaining in CD8 TILs. (C) PD-1 and CD137 surface immunostaining in SIINFEKL tetramer⁺ gated T cells. One-way ANOVA with Bonferroni post-hoc test, * p < 0.05; ** p < 0.01; *** p < 0.001.

Figure 4. sFlt3L and poly-ICLC do not control the progression of B16-OVA derived tumors in Batf3^-/- mice.

WT B16-OVA-bearing mice administered with hydrodynamic gene transfer with sFlt3L or control empty plasmid received i.p. injections of anti-CD137 mAb (A) or anti-PD-1 mAb (B), controlled by vehicle buffer, on days 4, 7 and 10. Poly-ICLC or control was administered i.t. on day 7. On the left, tumor areas (mean ± SEM). On the right, overall survival. (C-D) Comparison of the combined efficacy of sFlt3L + poly-ICLC with anti-CD137 mAb (C) or anti-PD-1 (D) in WT and Batf3^-/- mice. Graphs represent pooled data from 4 (A,C) or 2 (B,D) independent experiments with similar results, for a total of 10-15 mice per group. * p < 0.05; ** p < 0.01; *** p < 0.001