Bone Marrow Mesenchymal Stem Cells from Acute Myelogenous Leukemia Patients Demonstrate Adipogenic Differentiation Propensity

Background: Bone marrow mesenchymal stem cells (BMSCs) constitute one of the important cellular components of the hematopoietic microenvironmental niche. These cells are capable of differentiation into osteoprogenitors which comprise part of the endosteal niche. In vivo studies have shown that depletion of BMSCs resulted in reduction of hematopoietic stem cell content, and there is direct in vitro evidence that BMSCs are able to support both normal and leukemia progenitor cell proliferation and survival and contribute to leukemia cell resistance to cytotoxic therapies. Whether BMSCs from leukemia marrow differ from normal counterparts and how they contribute to and are influenced by the leukemia environmental niche are still incompletely understood. In this work we sought to compare normal and AML-derived BMSCs and to define the propensity of BMSCs from leukemia patients for osteogenic or adipogenic differentiation.

Methods: BMSCs were isolated from marrow aspirates of normal donor and AML patients using standard methodologies. Donors were age-matched to the leukemia patients to the extent possible in each comparative experiment. Such cells met the definition of MSCs in that they were adherent and expressed CD73, CD90, CD44, CD117 and CD105 and had osteoblastic and adipogenic differentiation capabilities. All comparisons were done at comparable passage number between samples.

Results: BMSCs from AML donors demonstrated irregular vs. spindle shaped morphology as compared with BMSCs from younger donors <50 years of age but similar morphology to those grown from older donors. AML derived BMSCs were larger and had slower growth rate as assessed by longer passage times during lower passages (41 vs. 21 days). Cell surface expression markers were similar between normal and AML. CFU-F outgrowth was less from AML as compared to normal BMSC, and no difference in osteogenic differentiation was noted by Alizarin Red S measurement. On the other hand, lipid droplet formation measured by Oil Red O during adipogenic differentiation induction was greater in the AML BMSCs as compared with age matched controls, suggesting increased adipogenic differentiation potential.

To determine if we could detect a gene signature difference between AML and normal BMSCs which would predict adipogenic propensity, RNA-Seq analysis was performed on AML and normal donor specimens(n=3 in each group), all from subjects >50 years of age using CuffDiff (v2.0.2). This identified 88 genes differentially expressed between the two groups. A heat map was generated using Euclidean distance hierarchical clustering of gene expression values from individual samples. This readily grouped AML vs. normal samples. Pathway analysis using Ingenuity Pathway Analysis (IPA) predicted dysregulation of four canonical pathways in AML-BMSCs as compared to normal BMSCs. These included 1) Role of tissue factor in cancer, 2) Airway pathology in COPD, 3) Oleate biosynthesis, and 4) Adipogenesis. The last two pathways are consistent with the biological observation of enhanced adipogenesis in AML-derived BMSCs. IPA analysis proposed a model of altered adipogenic differentiation in AML-BMSCs attributable to lower expression of two key regulatory genes, SOX9 and EGR2. Reduced expression of SOX9 and EGR2 in AML BMSCs as compared to normal BMSCs was validated by qRT-PCR and western blot analysis. SOX9 is reported to contribute to the commitment of MSCs to adipogenic phenotype through negative influence on expression of key transcription factors in adipogenesis, and ERG2 downregulation is required in some systems for adipocyte lineage commitment.

Conclusion: Understanding the role that adipogenic MSCs play during leukemia evolution and treatment could offer insight into pathogenesis and potential therapies for these disorders.