Abstract
Introduction: Recent studies have elucidated the importance of using 3-dimensional rather than 2-dimensional models in order to create an experimental system recapitulating the specialized properties of the bone marrow microenvironment. Since the neoplastic bone marrow (BM) milieu plays important roles in multiple myeloma (MM) pathogenesis, novel models to study the MM cell in its neoplastic microenvironment are needed.
Methods: To mimic the neoplastic BM microenvironment of MM patients, we have established a special hydrogel-based 3-dimensional (3-D) model by ex-vivo culturing MM patient-derived mesenchymal stem cells (MM-MSCs), the predominant cellular component of the marrow niche, which promotes greater mineralization and differentiation than a 2-dimensional (2-D) system.
Results: To characterize MM-MSCs in different stages of MM, we utilized an 11 multi-color flow cytometry panel. The percentage of MSCs (CD73+CD90+CD105+lin-CD45-CD34-HLA-DR-) population in BM aspirate samples of 50 MM patients (MGUS, smoldering MM, newly diagnosed MM, and relapsed or relapsed/refractory MM) was evaluated, and correlated with the distribution of (CD38+ CD138+) plasma cells. MSCs were less frequent (10x) than plasma cells, and increased with disease progression to relapsed/refractory MM. We seeded MM-MSCs (N=34) which had been expanded by adhesion methods in 2-D versus 3-D models in order to create an ex-vivo MM niche-like structure. In the hydrogel-based 3-D model, MM-MSCs formed compact clusters with active fibrous connections and meshwork-like structures at day 3 to 7. Moreover, calcium mineralization of clusters was observed, associated with the capacity for differentiation towards the osteoblastogenic or adipogenic lineage when cultured with differentiation media. Furthermore, the production of osteopontin (OPN) and angiopoietin-2 (Ang-2) was significantly higher in 3-D vs. 2-D MM-MSCs, assessed by multiplex luminex technology. Phenotypic profiling of 3-D MM-MSCs clusters revealed high expression of CD73+CD90+CD105+ and lack of expression of CD45, CD34 and HLA-DR, as in to 2-D MM-MSCs. MSC-specific markers including CD166 and HLA-ABC did not reveal any significant changes in 3-D vs. 2-D MM-MSCs; however, 3-D MM-MSCs had significantly decreased expression of CD271 and CD146 compared to 2-D cultures. We also observed significantly higher expression of extracellular matrix (ECM) molecules including fibronectin, laminin, collagen I, and collagen IV (p<0.001) in 3-D vs. 2-D MM-MSCs. Similarly, activation of integrins including VLA-2, VLA-4 and VLA-5 on the MSCs surface was also increased in 3-D MM-MSCs, as determined by confocal microscopy and flow cytometry analysis. Importantly, MM-MSCs cultured in 3-D vs. 2-D model have higher expression of N-cadherin and CXCL12 and decreased expression of nestin, reflecting the MM BM niche. Gene expression analyses of 3-D MM-MSCs revealed upregulation of BMP-2, MGP, PTGIS, COL14A1 and other genes and down-regulation of DKK1, ADAM9, OPCML genes and others compared to 2-D MM-MSCs. We also measured significantly higher production of IL-6 (p=0.002), IL-8, MCP-1(MCAF), RANTES, VEGF and HGF (p<0.001) in 3-D vs. 2-D MM-MSCs, by multiplex luminex analysis. Next, we co-cultured tumor cells from MM patients (12 MM patients) with either autologous or allogeneic MM-MSCs in 3-D vs. 2-D model. Plasma (CD38/CD138+) cells in 3-D co-culture were increased in 8/12 MM patients and equivalent to 2-D in 4/12 patients. By co-culturing MM cell lines (OPM1, RPMI-S, OCIMY5 and KMS11) labeled with CFSE fluorescent dye with various MSCs, we evaluated expression of side population (SP) cells, identified by Hoechst staining, and gating on CFSE positive MM cells, as low Hoechst stained cells. Our results showed that the SP fraction was significantly lower in 3-D compared to 2-D in co-cultures of various MM-MSCs with all 4 MM cell lines. Finally, we validated drug resistance to melphalan, bortezomib, lenalidomide, and carfilzomib in 3-D co-cultures of CFSE labeled primary tumor cells with various MM-MSCs.
Conclusions: This 3-D co-culture system closely mimics the myeloma BM niche, and therefore may be useful to identify and validate novel targeted therapies.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.