Abstract

Acute myeloid leukemia (AML) develops by the acquisition of genomic alterations which initiate different pathways leading to full-blown malignancy. The type of genomic driver alteration governs AML biology and clinical disease course. Translocations involving the core-binding factor (CBF) complex such as t(8:21) and inv(16) represent distinct subgroups of AML genomic drivers. However, additional factors promoting CBF AML leukemogenesis remain largely unknown. Due to the hypoxic nature of the bone marrow the activation of the ubiquitous hypoxia-sensing pathway via HIF-1a might promote CBF leukemogenesis. Nevertheless, in different AML models HIF proteins were recently considered as oncogenic drivers and also as tumor suppressors. Therefore, we here aimed to clarify the role of HIF-1a in CBF AML.

Initially, we investigated the impact of hypoxia-sensing pathway activation on the growth of the CBF AML cell lines Kasumi-1 [t(8:21)] and ME-1 [inv(16)]. Kasumi-1 cells showed increased growth after 24 as well as 48 hours at 3% compared to 21% oxygen conditions. In contrast, ME-1 cellular growth was not altered by hypoxic culture conditions. Next, we genetically modified the HIF-1a pathway in the CBF AML leukemogenesis context with a previously established conditional inv(16) mouse model (Cbfb+/56M). These mice express the Cbfbeta-SMMHC oncoprotein upon Cre-mediated recombination. In preliminary studies we could show neither vav-iCre mediated Hif-1a gain- nor loss-of-function altered steady-state hematopoiesis, which was recently confirmed by others (Milica Vukovic et al., Blood 2016). We proceeded by crossing conditional inv(16) with conditional mice overexpressing a HIF-1a variant (LSL-HIF1dPA) being insensitive to oxygen-dependent degradation. The conditional inv(16) and HIF1dPA alleles were simultaneously activated using the ubiquitous hematopoietic vav-iCre transgene. Strikingly, we observed accelerated leukemia development in inv16;HIF1dPA mice compared to inv(16) mice without genetic HIF-1a pathway alterations (median survival 68 vs. 116 days; Figure 1). In a complementary approach we crossed the vavi-Cre conditional inv(16) system into a conditional Hif-1a knock-out (KO) background [inv(16);Hif-1aKO mice]. Here we observed significantly delayed leukemia onset in inv(16);Hif-1aKO mice compared to inv(16) control mice (median survival 223 vs. 116 days; Figure 1). The gross leukemic phenotype with high peripheral leukocyte count, splenomegaly and liver infiltration did not differ between the groups. Finally, we analyzed deletion of the conditional Hif-1aKO allele of leukemic inv(16);Hif-1aKO mice and observed complete recombination resulting in a Hif-1a null constellation. Thus, AML in inv(16);Hif-1aKO mice did not develop from clones which evaded vav-iCre mediated Hif-1a deletion.

In conclusion, HIF-1a represents a critical factor promoting murine inv(16) leukemogenesis and it will be worthwhile studying the impact of genetic or pharmacologic HIF-1a alteration in established inv(16) leukemia.

Figure 1

Kaplan-Meier curves of leukemia-free survival in control [Cre-negative littermates; n=55], inv(16) [n=31], inv(16);HIF1dPA [n=17] and inv(16);Hif-1aKO [n=18] mice.

Figure 1

Kaplan-Meier curves of leukemia-free survival in control [Cre-negative littermates; n=55], inv(16) [n=31], inv(16);HIF1dPA [n=17] and inv(16);Hif-1aKO [n=18] mice.

Disclosures

Dührsen:Roche: Honoraria, Research Funding; Alexion Pharmaceuticals: Honoraria, Research Funding; Amgen: Honoraria, Research Funding. Göthert:Proteros: Consultancy; Ariad Pharmaceuticals: Consultancy; Bristol-Myers Squibb: Honoraria; Novartis: Honoraria.

Author notes

*

Asterisk with author names denotes non-ASH members.