Stem Cell Rev and Rep DOI 10.1007/s12015-015-9639-z
Adipogenic Mesenchymal Stromal Cells from Bone Marrow and Their Hematopoietic Supportive Role: Towards Understanding the Permissive Marrow Microenvironment in Acute Myeloid Leukemia Yevgeniya Le 1 & Sylvain Fraineau 2 & Priya Chandran 2 & Mitchell Sabloff 3 & Marjorie Brand 2,3 & Jessie R. Lavoie 4 & Rémi Gagne 4,5 & Michael Rosu-Myles 4 & Carole L. Yauk 4,5 & Richard B. Richardson 1,6 & David S. Allan 2,3
# Springer Science+Business Media New York 2015
Abstract Purpose The role of bone marrow-derived mesenchymal stem/stromal cells (MSCs) in creating a permissive microenvironment that supports the emergence and progression of acute myeloid leukemia (AML) is not well established. We investigated the extent to which adipogenic differentiation in normal MSCs alters hematopoietic supportive capacity and we undertook an in-depth comparative study of human bone marrow MSCs derived from newly diagnosed AML patients and healthy donors, including an assessment of adipogenic differentiation capacity.
Electronic supplementary material The online version of this article (doi:10.1007/s12015-015-9639-z) contains supplementary material, which is available to authorized users. * Richard B. Richardson
[email protected] * David S. Allan
[email protected] 1
Canadian Nuclear Laboratories, Chalk River, ON K0J 1 J0, Canada
2
Regenerative Medicine Program, Ottawa Hospital Research Institute, 501 Smyth Rd., Box 704, Ottawa, ON K1H 8L6, Canada
3
Department of Medicine, University of Ottawa, Ottawa, ON, Canada
4
Centre for Biologics Evaluation, Biologics and Genetic Therapies Directorate, Health Products and Food Branch, Health Canada, Ottawa, ON, Canada
5
6
Environmental Health Science and Research Bureau, Healthy Environments and Consumer Safety Branch, Health Canada, Ottawa, ON, Canada McGill Medical Physics Unit, Montreal General Hospital, Montreal, QC, Canada
Findings MSCs from healthy controls with partial induction of adipogenic differentiation, in comparison to MSCs undergoing partial osteogenic differentiation, expressed increased levels of hematopoietic factors and induced greater proliferation, decreased quiescence and reduced in vitro hematopoietic colony forming capacity of CD34+ hematopoietic stem and progenitor cells (HSPCs). Moreover, we observed that AML-derived MSCs had markedly increased adipogenic potential and delayed osteogenic differentiation, while maintaining normal morphology and viability. AML-derived MSCs, however, possessed reduced proliferative capacity and decreased frequency of subendothelial quiescent MSCs compared to controls. Conclusion Our results support the notion of a bone marrow microenvironment characterized by increased propensity toward adipogenesis in AML, which may negatively impact normal hematopoiesis. Larger confirmatory studies are needed to understand the impact of various clinical factors. Novel leukemia treatments aimed at normalizing bone marrow niches may enhance the competitive advantage of normal hematopoietic progenitors over leukemia cells. Keywords Mesenchymal stromal cells . AML . Hematopoietic niche . Bone marrow microenvironment . Adipogenesis . Osteogenesis
Introduction Understanding the mechanisms involved in the maintenance of malignant hematopoiesis is crucial in designing effective therapy for eradicating acute myeloid leukemia (AML). AML represents a group of genetically heterogeneous neoplasms
Stem Cell Rev and Rep
and is characterized by a low rate of survival due to chemorefractoriness, disease relapse and high therapy-associated morbidity [1]. Both clinical [2, 3] and experimental [4–6] observations suggest an important role of the bone marrow hematopoietic niche [7] in disease etiology. Regulation of the bone marrow microenvironment is largely attributed to stromal elements, and more specifically, mesenchymal stromal/stem cells (MSCs) [8]. Bone marrow MSCs represent a heterogeneous population of perivascular non-hematopoietic stem and progenitor cells and include CD146+ cells in humans [9], CXC chemokine ligand 12 (CXCL12)-abundant reticular cells (CARs) [10], Nestin+ cells [11], NG2+ pericytes [12] and LepR+ perivascular cells [13]. These have been widely implicated in support of normal hematopoiesis and shown to secrete stem cell factor (SCF) [14] and CXCL12 [10], molecules required for HSC maintenance and retention within the bone marrow, respectively. MSCs are also able to differentiate and contribute to specific stromal elements that regulate bone marrow niches, such as endosteum (osteoblasts) and adipose tissue (adipocytes) [15, 16]. While the importance of mesenchymal stromal cells in supporting normal hematopoiesis continues to be refined, the role of MSCs in the emergence or progression of leukemia has only been explored more recently [17–20]. Experimental evidence to date suggests perturbed function, denervation and decreased number of Nestin+ and NG2+ MSCs in murine studies of myeloproliferative disorders and AML [18, 19]. Studies by Medyouf [17] and Geyh [20] have demonstrated that MSCs derived from patients with myelodysplastic syndrome (MDS) possess abnormal differentiation profiles, propensity to expand dysplastic cells and decreased hematopoietic support functions. Few studies have looked at changes in MSCs derived from AML patients [21, 22] but these studies have not addressed MSC differentiation capacity and its impact on hematopoietic supportive function. To address how altered differentiation capacity may impact hematopoietic supportive function of MSCs, we compared adipo-biased MSCs with osteo-biased MSCs and observed that adipo-biased MSCs expressed increased levels of hematopoietic factors, supported greater expansion of differentiated hematopoietic cells, depleted the number of quiescent hematopoietic progenitor cells and attenuated in vitro hematopoietic colony-forming capacity of CD34+ cells. We also performed a comprehensive phenotypic and functional analysis of MSC samples derived from AML patients at diagnosis in comparison to healthy donors. Our findings confirm that marrow-derived MSCs in patients with AML exhibit increased adipogenesis and delayed osteogenesis. Our findings are consistent with a model that suggests MSCs are reprogrammed towards increased adipogenic differentiation, which exhausts normal hematopoietic progenitors and introduces a competitive advantage for leukemia cells to occupy
quiescent bone marrow niches where they may be protected from cytoreductive treatment.
Materials and Methods Human Bone Marrow Samples and Mesenchymal Stromal Cells (MSCs) Cells were obtained from filters following normal bone marrow harvests and bone marrow aspirates from newly diagnosed AML patients, in accordance with the Ottawa Health Sciences Network Research Ethics Board. Bone marrow mononuclear cells (MNCs) were prepared using Ficoll-Paque Plus density gradient as per manufacturer’s instructions (GE Healthcare, Pittsburg, PA, USA). MNCs were seeded in complete Dulbecco’s Modified Eagle Medium (DMEM, Gibco/Life Technologies, Carlsbad, CA, USA) supplemented with 10 % fetal bovine serum (FBS, Hyclone/ Thermo Scientific, Waltham, MA, USA) and 1 % penicillin/ streptomycin (Multicell/Wisent Inc., St. Bruno, QC, CA), at a density of 1.0–1.5×107 cells per T-75 flask. The emergence of plastic-adherent MSCs was followed for 3–6 weeks in a humidified tissue culture incubator (5 % CO2, 21 % O2, 37 ° C) with fresh medium changes performed every 7–10 days. Flow Cytometry Analysis Immunophenotyping: MSC immunophenotype, as established by the International Society for Cellular Therapy (ISCT) [23], was analyzed by flow cytometry using the following panel of antibodies: FITC-CD14, APC-CD19, PE-CD73, PE-CD90, APC-CD105, PE-Cy7CD146 (BD Biosciences, Mississauga, ON, CA); PE-Cy7CD34, FITC-CD45 (eBioscience, San Diego, CA, USA). Viability: The percent of pre-apoptotic, apoptotic and necrotic cells was measured using Alexa Fluor 488 Annexin V flow cytometry staining kit (Invitrogen/Life Technologies) according to manufacturer’s protocol. Cell cycle status: The proportion of cells in G0/G1, G2/M and S phases was determined using BrdU incorporation assay over 72 h using FITC-BrdU kit according to manufacturer’s instructions (BD Biosciences). To distinguish between cells in G0 vs. G1 phases, 1× 106 MSCs were stained with 2 μg/ml Hoechst 33342 (BD Biosciences) for 1 h followed by 30-min incubation with 0.5 μg/ml Pyronin Y (Sigma, Oakville, ON, CA). All incubations were performed in 1 ml of complete DMEM medium, at 37 °C. Immunophenotype and viability analyses utilized CyAn ADP 9 flow cytometer and Summit software (v.4.3.02, Beckman Coulter, Inc., Mississauga, ON, CA). Pyronin assay samples were analyzed on BD LSR Fortessa flow cytometer using FACSDiva software (v. 8.0, BD Biosciences). Minimum of 10 000 events were collected per each sample in duplicate. Morphology Assay Cells were stained with mouse α-tubulin primary antibody (#T9026, clone DM1A, Sigma, 1:2 000
Stem Cell Rev and Rep
dilution), Alexa Fluor 488 goat anti-mouse secondary antibody (1:200, Invitrogen/Life Technologies) and DAPI (1:2 000, Pierce Biotechnology, Inc., Rockford, IL, USA). Fluorescent images of, at least, 10 independent fields/well were acquired in duplicate using Axio Observer Z1 inverted microscope and AxioCam MRm camera (Carl Zeiss Microscopy, Pleasanton, CA, USA). MSC Differentiation and Western Blot Analysis MSCs were differentiated using Human MSC Functional Identification Kit (R&D Systems, Minneapolis, MN, USA) according to manufacturer’s instructions. Adipogenic differentiation was performed for 14 days, osteogenic differentiation for 17 days and chondrogenic differentiation for 21 days. In the case where no osteogenic differentiation occurred within 17 days, cells were maintained in differentiation medium until differentiation was observed. Chondrocytes and adipocytes were fluorescently stained with goat anti-human aggrecan and fatty acid binding protein 4 (FABP4, R&D Systems, manufacturer’s protocol), respectively. Images were acquired for, at least, 5 chondrocyte sections and 20 fields/well in duplicate for adipocyte monolayers using Axio Observer Z1 inverted microscope and AxioCam MRm camera (Carl Zeiss Microscopy). Adipocytes and osteocytes were chemically stained with Oil Red O for fat lipids and Alizarin Red S for calcium deposits, respectively, according to manufacturer’s protocol (Electron Microscopy Sciences, Hatfield, PA, USA). At least 15 transmitted light phase contrast images were obtained for each well in duplicate using: 1) Axio Observer A1 inverted microscope and AxioCam MRc camera (Carl Zeiss Microscopy) for osteocytes; and 2) Nikon Eclipse TS-100 inverted microscope and camera (Nikon Instruments Inc., Melville, NY, USA) for adipocytes. The adipogenic and osteogenic differentiation capacity of MSCs was quantitatively assessed using Western Blot analysis for: 1) adipo-specific FABP4 (monoclonal rabbit, #92501, 1:1 000, Abcam, Toronto, ON, CA); 2) osteo-specific osteopontin (monoclonal mouse, #21742, 1:100, Santa Cruz Biotechnology, Dallas, TX, USA); and 3) tubulin loading control (monoclonal mouse, 1:5 000, Developmental Studies Hybridoma Bank, The University of Iowa, IA, USA). Intensity of bands was quantified by densitometry using Scion Image software (v. Alpha 4.0.3.2, Scion Corporation, Frederick, MD, USA) and expressed as a ratio to tubulin loading control. Adipo/osteo-MSC Co-culture with Umbilical Cord Blood (UCB) CD34+ Cells and Colony Forming Unit (CFU) Assay MSCs from 5 normal donors were differentiated for 7 days resulting in normal bone marrow-derived adipo- and osteoMSCs. CD34+ cells were isolated from umbilical cord blood of consented donors using EasySep human CD34+ positive selection kit, as per manufacturer’s instructions (StemCell Technologies, Vancouver, BC, CA). UCB-CD34+ cells (3×
105) were co-cultured on the monolayer of either adipo- or osteo-driven MSCs or alone for 8 days in α-MEM complete medium without cytokines (Gibco/Life Technologies, 10 % FBS and 1 % penicillin/streptomycin). Following the coculture non-adherent fraction of cells (CD34+ hematopoietic progenitor cells, HPCs) was collected and degree of cell expansion and percent of quiescent cells were determined by live cell counting and Pyronin Y/Hoechst staining, respectively. CD34 + cells derived from co-culture were seeded in Methocult for CFU assay (#H4434, StemCell Technologies, manufacturer’s protocol). Plating efficiency – the number of colonies formed in a CFU assay per 500 CD34+ cells was determined. All results were expressed relative to CD34+ cells cultured alone. qRT-PCR The expression of the following hematopoietic maintenance genes by adipo- vs. osteo-driven MSCs was measured by qRT-PCR: scf, cxcl12/sdf1, il7, vcam1 and angpt1. RNA from adipo- and osteo-MSCs was extracted using a Qiagen RNeasy mini kit (Qiagen, Germantown, MD, USA). All values were expressed relative to β2 microglobulin (β2m) housekeeping gene. For a detailed list of primers used, refer to the supplementary information. All primers were from Integrated DNA Technologies (IDT, Coralville, IA, USA). RNA-Sequencing Analysis Total RNA was extracted from 6 control and 5 AML samples (miRvana miRNA isolation kit, Life Technologies, Carlsbad, CA, USA) and subjected to next-generation sequencing at Health Canada (Ottawa, Canada) using Ion Proton™ sequencer (Life Technologies, Carlsbad, CA). For a detailed description, see Supplementary Information. Briefly, Poly-A RNA enrichment (DynaBeads® mRNA DIRECT Micro Kit, Life Technologies) was performed followed by fragmentation and library construction (Ion Total RNA-seq kit v.2, Life Technologies) with ligated barcode adapters (Ion Xpress™ RNA-seq Barcode 1–16 kit, Life Technologies). Bioinformatics analysis included read alignment to reference human genome (GRCh38 rel. 77) and Ingenuity Pathway Analysis (v. 23814503) for differentially expressed genes associated with adipogenesis, osteogenesis and chondrogenesis. Ingenuity Knowledge Base was queried using the Disease/Function View tool to identify functional roles of identified genes in adipo-/osteo-/chondrogenesis pathways and Molecule Activity Predictor tool was used for predicted effect of differentially expressed genes on those pathways (increased, decreased, unchanged). The cut-off used for differential gene expression was: false discovery rate (FDR)
