Gene Network Analysis for Osteoporosis, Sarcopenia, Diabetes, and Obesity in Human Mesenchymal Stromal Cells

Abstract

The systemic gene interactions that occur during osteoporosis and their underlying mechanisms remain to be determined. To this end, mesenchymal stromal cells (MSCs) were analyzed from bone marrow samples collected from healthy individuals (n = 5) and patients with osteoporosis (n = 5). A total of 120 osteoporosis-related genes were identified using RNA-sequencing (RNA-seq) and Ingenuity Pathway Analysis (IPA) software. In order to analyze these genes, we constructed a heatmap of one-way hierarchical clustering and grouped the gene expression patterns of the samples. The MSCs from one control participant showed a similar expression pattern to that observed in the MSCs of three patients with osteoporosis, suggesting that the differentiating genes might be important genetic determinants of osteoporosis. Then, we selected the top 38 genes based on fold change and expression, excluding osteoporosis-related genes from the control participant. We identified a network among the top 38 genes related to osteoblast and osteoclast differentiation, bone remodeling, osteoporosis, and sarcopenia using the Molecule Activity Predictor program. Among them, 25 genes were essential systemic genes involved in osteoporosis. Furthermore, we identified 24 genes also associated with diabetes and obesity, among which 10 genes were involved in a network related to bone and energy metabolism. The study findings may have implications for the treatment and prevention of osteoporosis.

Keywords: gene network analysis; mesenchymal stromal cells; osteoporosis.

Figure 1

Heatmap of DEG of mesenchymal stromal cells from control and osteoporosis groups. (A) Heatmap of one-way hierarchical clustering of all samples showing 120 differentially expressed genes complying with |FC| ≥ 2, independent t-test raw p-value < 0.05, and Z-score for normalized value (log2 based). (B) Heatmap grouped by similarity of gene expression pattern via hierarchical clustering analysis of significant genes (distance measure = Euclidean distance; linkage method = completeness).

Figure 2

Functional relationship network of genes involved in osteoporosis pathogenesis. Functional relationship network for control subject no. 3 compared with three participants in the osteoporosis group (no. 6, 8, and 9).

Figure 3

Functional relationship networks of osteoporosis-related genes involved in the canonical pathways of obesity and diabetes development. Functional relationship networks for control subject no. 3 compared with three participants in the osteoporosis group (no. 6, 8, and 9).

Figure 4

Functional relationship networks shared among osteoporosis-, sarcopenia-, obesity-, and diabetes-related genes. Control group participants no. (A) 1, (B) 2, (C) 3, (D) 4, and (E) 5.