Leukemic Stress Targets the mTOR Pathway to Suppress Residual HSC in the BM Microenvironment

The remodeling of the bone marrow (BM) microenvironment that occurs along with the progressive spread of acute myeloid leukemia (AML) cells can be considered a constitutive aspect of leukemogenesis. To date most studies have focused on the functional and in part inflammatory adaptation of stroma, and its potential role in extrinsic chemotherapy resistance. Much less is known about the impact of leukemic stress on residual hematopoietic cells. We previously identified the trafficking of select microRNAs (miRs-) in extracellular vesicles (EVs) between AML cells and hematopoietic progenitor cells (HPC). These studies revealed the mechanism underlying the suppression of HPC function in the AML niche (Hornick, Doron et. al., Science Signaling 2016). Several groups, including ours also noted the relative resistance of residual hematopoietic stem cells (HSC) to elimination from the BM of xenografted animals. In the current study we set out to understand how leukemic stress in the AML xenograft niche shapes HSC fate and function. Using AML cell lines (Molm14, U937, HL60) we established NSG xenografts, systematically tracking peripheral blood AML chimerism to recover murine hematopoietic cells at low or negative tumor burden, and replicating key assays using purified EV for intrafemoral injections. In immunofluorescent studies we initially confirmed the uptake of GFP labeled xenograft-derived EVs across the spectrum of HPC and HSC (KSL/CD150+/CD48-), as well as the successive loss and peripheral displacement of HPCs, and gains in HSC frequency in the leukemic niche. These HSC were found to be enriched for G₀ cell cycle status with an increase in phospho- p53, but showed no evidence of apoptosis or senescence. To understand the mechanism underlying their apparent quiescence, we performed in vitro proteomics studies of AML EV exposed HPSC identified downregulation of ribosomal biogenesis pathways. We then confirmed in vivo that residual HSC from AML xenografts experienced a loss of protein synthesis (OPP assay). We next reasoned that deficits in ribosome dysfunction and protein synthesis may reflect deregulation by specific miRNAs highly abundant in AML EV. Here, we had an opportunity to profile EV miRNA from the plasma of 12 unselected AML patients at diagnosis versus 12 control samples, and we confirmed a significant enrichment for specific miRNAs, including miR-1246. Raptor is a component of the mTOR pathway and an annotated target of miR-1246. We demonstrated in a series of experiments that miR-1246 translationally suppresses Raptor and downregulates protein synthesis in residual HSC from AML xenografts. The transfection of synthetic anti-miR1246 sequences on the other hand reversed the effects of AML EV in murine HSC. In aggregate we show that direct crosstalk between AML and hematopoietic cells adds to the adaptive changes that occur in the AML niche. Our experiments suggest a functional significance for EV miRNA that can be detected in AML patient plasma in the regulation of residual BM HSC. More broadly, the mechanisms by which leukemic stress alters hematopoietic function remain underexplored, but our observations suggest that leukemia derived EV contribute to changes in competitive fitness of residual HSC.