Page 220 - Haematologica Vol. 110 - January 2025

LETTER TO THE EDITOR
Cd39 and P2rx7-Wnt signaling enhance blast pathogenicity in an experimental model of acute myeloid leukemia
Acute myeloid leukemia (AML) is the most common type of acute leukemia in adults and is caused by uncontrolled clonal expansion of immature myeloid cells. Elements of the leukemic microenvironment, as well as genetic and metabolic alterations of blast cells, are all linked to ad- verse outcomes.1 High CD39 expression on leukemic blast cells is associated with poor prognosis in AML patients.2 CD39/ENTPD1 (ectonucleoside triphosphate diphospho- hydrolase-1), in tandem with CD73, converts extracellular adenosine triphosphate (eATP) and adenosine diphosphate to ultimately generate adenosine.3 Extracellular nucleotides drive type-2 purinergic receptor (e.g., P2RX7) responses that underpin tumor immunity.4 Genetic, pharmacological and immunological studies have documented the therapeutic potential of targeting CD39 and purinergic responses in solid tumors.5-8 In this current work, we explored the pathogenic roles of Cd39 and P2rx7 and demonstrate novel downstream elements of purinergic signaling in an experimental model of AML. Animal experimentation protocols were reviewed and approved by the Institutional Animal Care and Use Committee of Beth Israel Deaconess Medical Center.
We established a model of aggressive AML by transplanting TIB-49 cells into immunocompetent syngeneic C57BL/6 wild-type (WT) mice. Heterogeneous expression patterns of Cd39 on immune cells throughout blood, spleen and bone marrow from control mice and TIB-49-bearing mice were revealed (Online Supplementary Figure S1A). TIB-49 cells from spleen and bone marrow displayed high levels of Cd39 (Figure 1A), even though this was not detected on cultured TIB-49 cells in vitro (Online Supplementary Figure S1B). Furthermore, there were no Cd39 transcripts present in TIB-49 cells sorted from spleen and bone marrow of WT AML mice (Online Supplementary Figure S1C).
We considered that TIB-49 cells acquired Cd39 via tro- gocytosis in specialized niches (e.g., spleen and bone marrow) in vivo. To test this, TIB-49 cells were inoculated into Cd39-/- mice. We noted that these TIB-49 cells from Entpd1 null blood, spleen and bone marrow did not express Cd39 (Figure 1A). Our data indicated that Cd39 on TIB-49 cells was acquired from Cd39+ host cells, likely through trogocytosis (Figure 1B).9,10 Additionally, circulating TIB-49 cells from blood of WT mice did not express Cd39 despite high levels of this ectoenzyme on the vasculature (Figure 1A), suggesting an essential role of cell-cell contact be- tween acceptor cells and donor cells within the leukemic microenvironment.
TIB-49 cells were then inoculated into WT and Cd39-/- mice. Higher levels of TIB-49 cells were detected by fluorescence activated cell sorting (FACS) in blood and spleen from
WT mice than in Cd39-/- mice (Figure 1C), and there were delays in engraftment in Cd39-/- mice (Figure 1D). TIB-49 cells overexpressing Cd39 with TdTomato as an indicator (viz. TIB-Cd39high cells) were then generated. Compared with parental TIB-49 cells, TIB-Cd39high cells inoculated into WT mice demonstrated faster engraftment (Figure 1E). Furthermore, higher levels of Cd39 on TIB-Cd39high than on TIB-49 cells was confirmed in vivo (Online Supplementary Figure S1D). Importantly, TIB-Cd39high inoculation resulted in more rapid disease progression with shorter times to euthanasia (Figure 1F) despite the proliferation rates of these two cell lines being similar in vitro (Online Supple- mentary Figure S1E).
To the TIB-49-inoculated mice we then administered aCd39 monoclonal antibody, which had been shown to deplete Cd39high cells and diminish cell surface Cd39 expression, through FcgRIV-dependent trogocytosis in the MC38 mod- el.9 Although the aCd39 monoclonal antibody selectively depleted myeloid-derived suppressor cells and downreg- ulated surface Cd39 expression in multiple immune cells and TIB-49 cells (data not shown), this form of aCd39 antibody monotherapy did not alter the experimental dis- ease course in mice inoculated with either TIB-49 cells or TIB-Cd39high cells (Online Supplementary Figure S1F). When aCd39 monoclonal antibody treatment was combined with cytarabine, we found no additional effects in the TIB-49- bearing WT mice (data not shown). We also tested the ef- fects of aCd39 monoclonal antibody in a TIB-49 chloroma model. In this instance, aCd39 monotherapy effectively boosted the chemotherapeutic effects of low-dose cytar- abine with respect to growth of TIB-49 chloroma in vivo (data not shown). These studies suggest that the benefits and strategy of targeting Cd39 in solid cancer cannot be simply extrapolated to therapeutic applications in liquid cancers, such as AML.
Bulk RNA sequencing of TIB-49 cells and TIB-Cd39high cells was conducted to explore purinergic and other mechanisms dictating TIB-Cd39high pathogenicity (Online Supplementary Figure S2A). We found that P2rx1, P2rx7, P2ry10 and P2ry14 were substantively upregulated in TIB-Cd39high cells. We then explored RNA-sequencing data of human CD39+ AML cells in The Cancer Genome Atlas database. We subdivided 151 AML samples into two groups based on median CD39 levels and found that P2RX1 and P2RX7 were upregulated in the group with higher CD39 expression. These increases of P2RX7 were greater than those seen for P2RX1 (Figure 2A). We therefore explored the roles of P2rx7 (P2rx7-v1 (NM_011027), which is considered the canonical full-length P2rx7 in mediating TIB-49 pathogenicity. FACS did not detect
Haematologica | 110 January 2025
212