Targeting PRMT9 Elicits Cgas-Mediated Anti-Leukemia Immunity

Immunotherapy has improved outcomes for patients with blood cancers such as B-ALL. However, for patients with immune-cold cancers such as AML, currently available immunotherapies are less effective, likely due to lack of functional T cells and/or low expression of immune checkpoint proteins like PD-L1 . For these patients, development of an effective leukemia-targeting drug with immune-stimulating activity is required. We previously reported elevated PRMT9 levels in AML patient specimens, which enhanced leukemia cell proliferation by promoting RNA translation (Blood, 2021,138 [Supplement 1]: 358). Recently, our transcriptome profiling of AML lines revealed increased expression of interferon-stimulated genes (ISGs) after PRMT9-knockdown (KD), suggesting activation of innate immunity.

To evaluate a potential PRMT9 function in anti-AML immunity, we transplanted either WT immunocompetent or Rag2^-/- immunodeficient recipient mice with inducible Prmt9-KD or control MLL-AF9 (MA9) AML donor cells and monitored leukemia development by in-vivo bioluminescence Imaging. Unlike Rag2^-/- recipients (overall survival <60 days), 5 of 7 WT recipients harboring Prmt9-KD AML survived the observation period (>120 days) and MA9 cells disappeared over time. To assess immune responses upon Prmt9 inhibition, we performed single-cell RNA-sequencing (scRNAseq) in BM and spleen cells from Prmt9-KD or control MA9-transplanted WT recipients (Fig. A). ScRNAseq analysis revealed that, relative to controls, Prmt9-KD induced T cell activation, based on increased T cell expression of Cd69, Gzmb and Ifng. After reanalyzing T cells at higher granularity based on levels of classical markers (e.g., Cd44, Sell, Fig. B, C), we observed reduced numbers of naïve T cells as well as increased effector/memory T cell subsets; increased numbers of cytotoxic T lymphocytes (CTLs) as well as decreased regulatory T cells (Tregs) subset (Fig. D). To validate T cell activation upon Prmt9 KD, we assessed an MA9/OVA transplant model (engineered with OVA to track antigen-specific T cells) and found that mice harboring Prmt9-KD MA9/OVA cells showed increased numbers of OVA-specific CD8⁺ T cells in BM tracked by flow cytometry after H-2K^b-OVA_257-264 tetramer staining (Fig. E). Notably, scRNAseq analysis of multiple immune lineages in BM engrafted with Prmt9-KD MA9 cells revealed upregulation of Isg15, Cxcl10 and other ISGs. To confirm that type-I IFN responses mediate these effects, we implanted Prmt9-KD MA9 cells into WT or type-I IFN receptor KO (Ifnar1^-/-) recipients and found that the survival advantage conferred by Prmt9-KD was lost in Ifnar1^-/- recipients (Fig. F). Interestingly, Prmt9-KD in MA9 cells also increased expression of Isg15, Cxcl10 and other ISGs by Q-PCR assay, and the effect was blocked by cGAS deletion in MA9 cells, suggesting that Prmt9-KD activates cGAS-STING signaling. Finally, to determine whether Prmt9-KD phenotypes require tumor-intrinsic cGAS activity, we knocked out cGAS in Prmt9-KD MA9/OVA cells and implanted them into WT recipients. Relative to cGAS-WT controls, cGAS-KO in the presence of Prmt9-KD abolished survival advantages and cancer-specific T cell responses seen following Prmt9 KD in MA9/OVA cells (Fig. E).

We also developed a potent new PRMT9 inhibitor (PRMT9i) based on our identified lead compound (Blood, 2021,138: 358). CyTOF analysis revealed that PRMT9i treatment ablated primary AML cells (CD34⁺CD45^dim) and relatively increased T cell frequencies in co-cultures of immune cells with leukemia cells from primary AML specimens (n=3). However, PRMT9i anti-AML effects were significantly impaired by T cell depletion. Notably, CyTOF analysis revealed significantly upregulated PD-L1 levels in 2 of 3 (CD34⁺CD45^dim) AML co-cultures after PRMT9i treatment (Fig. G), suggesting activation of the adaptive PD-1/PD-L1 axis, in agreement with the” unwanted negative effects” of cGAS-STING activation as others depicted. Accordingly, PRMT9i treatment of these 2 co-cultures combined with immune checkpoint inhibition (ICI, anti-PD1) synergistically eradicated primary AML cells (Fig. H) and promoted expansion of CD69⁺ and IFNγ⁺ CD8⁺ T cell subsets.

Overall, we demonstrate that AML cell-derived PRMT9 regulates AML immune evasion, and that combining PRMT9i with ICIs could constitute a therapeutic strategy against PRMT9-proficient cancers like AML, in which single ICIs show limited efficacy.

Salhotra:Orca Bio: Research Funding; Kadmon: Other: Advisory board meeting ; BMS: Research Funding. Marcucci:Abbvie: Other: Speaker and advisory scientific board meetings; Novartis: Other: Speaker and advisory scientific board meetings; Agios: Other: Speaker and advisory scientific board meetings.