The Interplay between Macrophages and Pediatric AML Blasts

Patients suffering from AML still lack significant benefits from immunotherapeutic approaches (Pfister et al., 2022, Subklewe et al., 2023). One reason for this is an incomplete understanding of the AML microenvironment (ME), particularly the role of the innate immune compartment. Ample evidence in solid tumors shows that innate immune cells, such as tumor-associated macrophages, neutrophils and myeloid-derived suppressor cells, contribute to a suppressive ME, while relatively little is known about the innate immunity in AML (Perzolli et al., 2024). In this project, we focus on the characterization of macrophages and their respective cellular interaction with leukemic and non-leukemic cells in the pediatric AML ME.

In two independent pediatric AML cohorts (Maxima, n=140 and TARGET, n=159, bulk bone marrow RNA-sequencing data), we observed a higher abundance of anti-inflammatory M2-like macrophages compared to pro-inflammatory M1-like macrophages. Additionally, we noted that AML patients with fusion genes, in particular KMT2A-rearrangements (KMT2Ar), CBFB::MYH11 and NUP98::NSD1, had a higher abundance of M2-like macrophages compared to those with normal karyotypes with or without mutations in known leukemic regulators such as NPM1 and CEBPA. Furthermore, pediatric AML patients showed a higher amount of M2-like macrophages than healthy donors (n=7). By flow cytometry, we confirmed M2/M1 ratios >1 in five primary AML patients (KMT2A::MLLT10, KMT2A::MLLT9, NUP98::TOP1, RUNX1::RUNX1T1 and CBFB::MYH11) and M2/M1 ratios <1 in three healthy donors.

RUNX1::RUNX1T1 and KMT2Ar AML cell lines, as well as patient-derived xenograft (PDX) cells, can alter the phenotype of M1-like macrophages generated from the monocytic cell line THP-1, skewing them towards an M2-like phenotype. We, therefore, investigated the interaction between macrophages and AML blasts using primary and PDX cells. We differentiated monocytes from healthy donors into five different macrophage subtypes (M0, M1, M2a, M2c, M2d) and co-cultured them with PDX or primary AML cells. These co-culture experiments revealed that primary M2-like macrophages (M2a and/or M2c) support RUNX1::RUNX1T1 and KMT2Ar AML cell proliferation, while M1-like macrophages transiently inhibited leukemic proliferation. Notably, this inhibitory effect disappeared after six days of co-culture. Flow cytometry showed reduced surface expression of the M1 markers CD80 and HLA-DR and an increase of the M2 markers CD206 and CD163. Gene expression analysis demonstrated reduced expression of pro-inflammatory genes (IL1α, ILβ, IL6, CXCL9) and increased expression of anti-inflammatory and angiogenic genes (ARG2, TGFB2, VEGFA, VEGFC). We next investigated whether these transcriptional changes lead to functional alterations. Using a bioluminescence T cell activation assay, we found that M1-like macrophages lose their ability to activate T cells when co-cultured with AML blasts. Additionally, flow cytometry showed that M1-like macrophages inhibit T cell proliferation when co-cultured with AML blasts, behaving similarly to M2-like macrophages. These data suggest that AML blasts reeducate human-monocyte derived M1-like macrophages to a pro-tumoral and immunosuppressive phenotype.

Finally, we examined whether the interaction between primary human monocyte-derived macrophages and AML blasts depends on cell-cell contact. Transwell experiments revealed that AML blasts require direct contact with macrophages to influence their phenotypes and functions. In the transwell setting, M1-like macrophages retained their ability to suppress AML cell proliferation and did not show increased expression of M2 surface markers (CD206 and CD163), anti-inflammatory (ARG2 and TGFB2) or angiogenic genes (VEGFA and VEGFC). Further studies are underway to elucidate the specific nature of the interactions between macrophages and AML blasts, including the characterization of ligand-receptor interactions between the two populations and the identification of perturbed molecular mechanisms.

Overall, our data suggest that AML blasts reorganize the microenvironment by re-educating pro-inflammatory macrophages into an anti-inflammatory state through direct contact. We predict that an understanding of the interaction between AML blasts and BM macrophages will identify novel targets for reversing the immunosuppressive ME in pediatric AML.

Zwaan:Incyte Corporation: Consultancy; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees, Other: Institutional grant ; Pfizer: Other: Institutional grant; Beigene: Consultancy; Astra Zeneca: Consultancy; Daiichi Sankyo, Inc.: Other: Institutional grant; Takeda Pharmaceuticals: Other: Institutional grant; AbbVie: Other: Institutional grant ; Jazz Pharmaceuticals: Other: Institutional grant; Kura Oncology: Consultancy, Other: Institutional grant; Innovative Therapies for Children with Cancer: Other: Leadership role; Sanofi: Membership on an entity's Board of Directors or advisory committees, Other: Institutional fees; Bristol Myers Squibb: Consultancy; Gilead Sciences, Inc.: Consultancy; Chair of the Medical Research committee MERC Utrech: Other: Leadership role. Heidenreich:Syndax: Other: institutional funding; Roche: Other: institutional funding.