AIM2 Mediates Immune Evasion of Acute Myeloid Leukemia Cells

Background

Acute myeloid leukemia (AML) is a severe hematologic malignancy with high frequency of recurrence but low sensitivity to current immunotherapy. Understanding the molecular mechanism and discovery of new targets for AML immunotherapy become urgent. This study is to elucidate a novel molecular mechanism by which leukemia cells evade anti-tumor immunosurveillance.

Methods

The AML cell-to-T cell interaction and communications were observed by live cell imaging, imaging flow cytometry, and confocal microscope. RNA sequencing analyses of wildtype or gene-knockout AML cells by CRISPR/Cas9 were used to determine the key genes and cellular signalings that regulate T cell suppression function of AML cells. The immune modulatory functions of these key genes were verified in both in vitro co-culture assays and in vivo humanized AML mouse models. Immunoprecipitation-mass spectrometry (IP-MS) analysis was used to discover the novel protein-protein interactions.

Results

In contrast with killing of AML cells by T cells, we found that ~10% AML cells that could engulf a whole T cell or acquire components of T cells in an in vitro co-culture assay. From 18 newly diagnosed AML patients, at average of 13.75% (min-max: 0.56-57.32%) primary AML cells have intracellular TCRa/b staining which suggested primary AML cells could also engulf T cell components. Next, by comparison of gene expression in between AML cells with and without T cell components, we found that absent in melanoma 2 (AIM2), a key component of inflammasome, is one of the top up-regulated genes in AML cells with T cell components. Knockout of AIM2 in AML cells significantly decreased the percentage of AML cells with T cell components in co-culture. Interestingly, loss of AIM2 in AML cells significantly increased T cell activation and cytotoxicity. Moreover, loss of AIM2 in AML cells inhibited AML progression in a T cell dependent manner, indicating an immunosuppressive function of AIM2 in AML cells. However, although knockout of other inflammasome components, such as ASC and Caspase-1, also increased T cell activation and cytotoxicity, we found that the key AIM2-inflammasome downstream molecule, GSDMD or GSDME in AML cells, didn’t suppress T cells, suggesting AIM2 may not induce immunogenic cell death pyroptosis in AML cells. Furthermore, we found that knockout of AIM2 significantly induced cGAS/STING signalings. In mechanistically, we demonstrated that AIM2 could facilitates the interaction of RalA-GTPase with its GEF protein RCC2 and promotes the activation of RalA to maintain the stability of chromosomal passenger complex (CPC) in AML cells during mitosis. Loss of AIM2 induced more micronucleus and higher chromosomal instability (CIN) that induced cGAS/STING signaling. Either knockout RalA or stimulation of STING increased T cell activation and cytotoxicity; while, knockout of STING in AML cells decreased T cell activation and cytotoxicity.

Conclusion

Together, our findings suggest that AML cells may directly engulf a whole T cell or components of T cells to escape anti-tumor T cell killing. AIM2 in AML cells may inhibit CIN-induced cGAS/STING signaling to suppress T cells. This study may help us to understand the mechanism of AML cell immune evasion.