The Direct Interactions with Bone Marrow Microenvironment Confer Resistance to the Inhibition of Oxidative Phosphorylation in AML

Acute myeloid leukemia (AML) cells are highly dependent on oxidative phosphorylation (OxPhos) for survival and continually adapt to the bone marrow (BM) microenvironment. We investigated how the BM microenvironment impacts the response to energy-depriving OxPhos inhibition in AML using a novel complex I OxPhos inhibitor (OxPhosi), IACS-010759. We have reported that OxPhosi-resistant primary AML samples demonstrated higher baseline transcription of genes related to cell adhesion, integrin and/or Rho GTPase family genes that modulate intracellular actin dynamics. (Yang et al. ASH 2019)

In this study, we performed Cap Analysis of Gene Expression (CAGE) transcriptome analyses using IACS-010759-sensitive and -resistant AML PDXs. CAGE identifies and quantifies the 5' ends of capped mRNA transcripts (= transcription start sites) and allows investigating promoter structures necessary for gene expression. Primary AML cells from 9 AML PDXs were injected into irradiated NSG mice, which were randomized upon documented engraftment to receive IACS-010759 or vehicle (n = 3/group). The antileukemia efficacy of the treatment was monitored by serial measurements of circulating AML cells. Of the 9 models tested, we defined 4 PDXs as sensitive and 5 as resistant to OxPhos inhibitor therapy. In the resistant models, CAGE analysis of OxPhosi-induced changes (comparing pretreatment with posttreatment) identified upregulation of 77 promoters and downregulation of 207 promoters (log 2-fold change > 3.0, FDR < 0.05, EdgeR), including increased promoter expression (>3.0 fold) of genes associated with adhesion (CCR8, ADGRB2, LAG3, BMF, ATN1, PLXDC1), migration (CCR8, NKX3-2, TMEM123, IGLV7-43, FAM171A1, LBX2, TRAV21, PPP2R5C, BMF, PLXDC1), and actin cytoskeleton dynamics (FAM171A1, BMF, BEST1, PLXDC1). Of note, the 6 adhesion-associated promoters that were upregulated by OxPhosi in 5 of the OxPhosi-resistant mouse models were unchanged or downregulated in the 4 OxPhosi-sensitive models. We then used DEGseq, an R package for identifying differentially expressed genes, to identify promoters whose expression was different between OxPhosi-treated and vehicle-treated groups in the OxPhosi-resistant mouse models. DEGseq detected consistent changes of 214 upregulated and 626 downregulated promoters with OxPhosi treatment in all 5 mouse models. KEGG pathway enrichment analysis was performed with these consistently changed genes and revealed that OxPhos inhibitor treatment significantly upregulated the transcripts of cell adhesion pathway.

We then confirmed that BM derived mesenchymal stem cells (MSC) protected OxPhosi-sensitive OCI-AML3 cells; the IC50 of IACS-010759 under MSC coculture was 80-fold higher than in monoculture conditions (IC50; 0.04 nM in monoculture vs. 3.25 nM in coculture), and IACS-010759 (10nM) induced 55% reduction of viable cells in coculture condition as compared to 70% reduction in monoculture. We further observed that OCI-AML3 cells adhered to MSCs were more profoundly protected from OxPhosi induced apoptosis than nonadherent cells. These results indicate that BM stromal cells, in particular those in direct contact with leukemia cells, play a key role in the microenvironment-mediated protection of AML cells from metabolic stress caused by OxPhos blockade.

We further observed promoter upregulation of ASS1, coding Argininosuccinate Synthase 1 and of LRP1, coding LDL Receptor Related Protein 1. Argininosuccinate Synthase 1 is an epigenetically regulated key enzyme in the biosynthesis of arginine and energy starvation that induces adaptive transcriptional upregulation of ASS1. LDL Receptor Related Protein 1 plays a major role in lipid metabolism and has been reported to be responsible for hemin-induced autophagy in leukemia cells. These might contribute to intrinsic AML resistance to OxPhosi via activation of compensatory metabolic pathways, amino acid metabolism and lipid metabolism.

Taken together, our data highlight the importance of direct interaction with BM stromal cells as well as complementally modification of amino acid- and lipid metabolism for the resistance of AML cells to OxPhos inhibition. While the mechanisms of stroma-leukemia interactions are likely complex, reducing the adhesion of AML cells to nurturing stromal cells ameliorates the resistance to the metabolic and energetic consequences of OxPhos inhibition.