Bone marrow mesenchymal stem cells are abnormal in multiple myeloma

Abstract

Recent literature suggested that cells of the microenvironment of tumors could be abnormal as well. To address this hypothesis in multiple myeloma (MM), we studied bone marrow mesenchymal stem cells (BMMSCs), the only long-lived cells of the bone marrow microenvironment, by gene expression profiling and phenotypic and functional studies in three groups of individuals: patients with MM, patients with monoclonal gamopathy of undefined significance (MGUS) and healthy age-matched subjects. Gene expression profile independently classified the BMMSCs of these individuals in a normal and in an MM group. MGUS BMMSCs were interspersed between these two groups. Among the 145 distinct genes differentially expressed in MM and normal BMMSCs, 46% may account for a tumor-microenvironment cross-talk. Known soluble factors implicated in MM pathophysiologic features (i.e. IL (interleukin)-6, DKK1) were revealed and new ones were found which are involved in angiogenesis, osteogenic differentiation or tumor growth. In particular, GDF15 was found to induce dose-dependent growth of MOLP-6, a stromal cell-dependent myeloma cell line. Functionally, MM BMMSCs induced an overgrowth of MOLP-6, and their capacity to differentiate into an osteoblastic lineage was impaired. Thus, MM BMMSCs are abnormal and could create a very efficient niche to support the survival and proliferation of the myeloma cells.

Figure 1. Co-culture of CD34⁺ cells with BMMSCs

CD34⁺ cells (4.10⁴) were cultured on normal or MM BMMSCs (5.10³/cm²) in 12-well plates. The co-culture was performed in Myelocult medium with 10 μM of hydrocortisone as described in Materials and Methods. At days 7, 14, 21 and 28, nonadherent cells were counted and a clonogenic test was performed. At day 35, nonadherent and adherent cells were harvested, counted and assayed in the clonogenic test.

Figure 2. Unsupervised data analysis of BMMSCs of the 3 groups of subjects

A: Principal component analysis (PCA) of BMMSCs from 6 myeloma patients (●) and from 7 normal individuals ( ). Data were obtained from U133 2plus microarrays (Affymetrix) as described in Materials and Methods. Analysis was performed on 28746 probe sets declared as present by the call detection algorithm. The variation coefficient (VC), defined as the ratio of the SD to the mean of signal values for the considered probe set, was used to select genes incorporated in the PCA. By varying VC from 0% to 100%, the normal and MM BMMSC clusters can be more or less separated. At a VC of 60% the separation was the best and includes ~ 2,000 genes. The boxed scheme corresponds to the cluster analysis performed on the same set of data with Cluster and TreeView softwares (http://rana.lbl.gov/EisenSoftware.htm) B: MGUS BMMSC ( ) were added to the analysis.

Figure 2. Unsupervised data analysis of BMMSCs of the 3 groups of subjects

A: Principal component analysis (PCA) of BMMSCs from 6 myeloma patients (●) and from 7 normal individuals ( ). Data were obtained from U133 2plus microarrays (Affymetrix) as described in Materials and Methods. Analysis was performed on 28746 probe sets declared as present by the call detection algorithm. The variation coefficient (VC), defined as the ratio of the SD to the mean of signal values for the considered probe set, was used to select genes incorporated in the PCA. By varying VC from 0% to 100%, the normal and MM BMMSC clusters can be more or less separated. At a VC of 60% the separation was the best and includes ~ 2,000 genes. The boxed scheme corresponds to the cluster analysis performed on the same set of data with Cluster and TreeView softwares (http://rana.lbl.gov/EisenSoftware.htm) B: MGUS BMMSC ( ) were added to the analysis.

Figure 3. Differentially expressed genes classification

The genes found to be differentially expressed between normal (n = 7) and MM BMMSCs (n = 6) were classified by use of gene ontology through the Fatigo platform (http://fatigo.bioinfo.cnio.es/) into 7 classes. The tumor microenvironment class comprised genes that belong to the intercellular communications (27%), to receptor signalization (9%) and to extracellular matrix and other secreted molecules (9%).

Figure 4. Myeloma cell line and BMMSC co-culture

MOLP-6 cells were cultured in RPMI-1640 medium with 10% FCS at 2 × 10⁴ cells/well in 24-well plates with or without an increasing concentration of GDF-15 (10, 50 and 100 ng/ml). After 7 days, the cells were counted by trypan blue exclusion, and the expansion coefficient was calculated (d7 cell number/d0 cell number). MOLP-6 expansion with 100 ng/ml of GDF-15 was 2 times greater than that without GDF-15 (n = 7, R = 0.38 and p = 0.048)

Figure 5. Osteoblastic differentiation of BMMPCs

The cells were cultured in osteoblastic medium as described in Materials and Methods during 21 days. (A) Matrix mineralization was then revealed by von Kossa staining, and cells were counter colored by use of giemsa. Pictures were taken on an Olympus 1X71 inverted microscope (magnification 100X) and underwent analysis with Analysis B software. (B) Plates were stained with alizarin red and extensively washed in distilled water. The intensity of the staining was quantified according to Gregory.