CDK6 is essential for mesenchymal stem cell proliferation and adipocyte differentiation

Abstract

Background: Overweight or obesity poses a significant risk of many obesity-related metabolic diseases. Among all the potential new therapies, stem cell-based treatments hold great promise for treating many obesity-related metabolic diseases. However, the mechanisms regulating adipocyte stem cells/progenitors (precursors) are unknown. The aim of this study is to investigate if CDK6 is required for mesenchymal stem cell proliferation and adipocyte differentiation. Methods: Cyclin-dependent kinase 6 (Cdk6) mouse models together with stem cells derived from stromal vascular fraction (SVF) or mouse embryonic fibroblasts (MEFs) of Cdk6 mutant mice were used to determine if CDK6 is required for mesenchymal stem cell proliferation and adipocyte differentiation. Results: We found that mice with a kinase inactive CDK6 mutants (K43M) had fewer precursor residents in the SVF of adult white adipose tissue (WAT). Stem cells from the SVF or MEFs of K43M mice had defects in proliferation and differentiation into the functional fat cells. In contrast, mice with a constitutively active kinase CDK6 mutant (R31C) had the opposite traits. Ablation of RUNX1 in both mature and precursor K43M cells, reversed the phenotypes. Conclusion: These results represent a novel role of CDK6 in regulating precursor numbers, proliferation, and differentiation, suggesting a potential pharmacological intervention for using CDK6 inhibitors in the treatment of obesity-related metabolic diseases.

Keywords: Cdk6; Runx1; obesity; progenitors; stem cells.

FIGURE 1

Reduced fat mass in K43M mice and increased fat mass in R31C mice. (A) Ventral view of female WT, K43M, and R31C mice fed with NCD. Appearance of a close view of the eWAT (B) and gWAT (C) from the mice of the three genotypes. (D) Mass of various fat pads was normalized to body weight of male mice on NCD at 18–20 weeks of age. (E) Mass of various fat pads was normalized to body weight of female mice on NCD at 18–20 weeks of age. For (D, E), data shown are mean ± SE (n = 6 for each group), *p < 0.05, t-test, vs. WT of the corresponding adipocyte tissue.

FIGURE 2

Reduced adipocyte precursors in K43M mice and increased adipocyte precursors in R31C mice. (A, C, E) Representative flow cytometric profiles of Sca-1⁺ (A), Sca-1⁺CD36⁺ (C), and Lin⁻Sca-1⁺CD24⁺CD29⁺CD34⁺ (E) cells isolated from eWAT of WT, K43M, and R31C mice at 18–20 weeks of age. (B, D, F) Histograms summarizing the Sca-1⁺ (B), Sca-1⁺CD36⁺ (D), and Lin⁻Sca-1⁺CD24⁺CD29⁺CD34⁺ (F) cells in panel (A, C, E), respectively. For (B, D, F), data shown are mean ± SE (n = 6–10). *p < 0.05, t-test. For B, p-value < 0.0001, one way ANOVA, for D, p-value < 0.0001, one way ANOVA, and for F, p-value = 0.0008, one way ANOVA.

FIGURE 3

Role of CDK6 kinase activity in BrdU incorporation and cell cycle profiles. (A-C) Representative immunofluorescent detection of BrdU-labeled cells (green) in eWAT derived from 4-month-old male WT, K43M, and R31C (n = 4) mice which were administered BrdU for three consecutive days. Cell nuclei counterstained with DAPI (blue). Magnification ×40. (D, E) Representative flow cytometric profiles of negative control with SVF cells stained with normal mouse IgG-FITC and Sca-1-APC (D) and Sca-1⁺BrdU⁺ cells isolated from eWAT of male WT, K43M, and R31C mice at 18–20 weeks of age (E). (F) Histograms summarizing the Sca-1⁺BrdU⁺ cells in panel (E). (G) Representative flow cytometric profile profiles of Sca-1⁺7AAD⁺ cells isolated from eWAT of male WT, K43M, and R31C mice at 18–20 weeks of age. (H) Histograms summarizing the Sca-1⁺7AAD⁺ cells in panel G. For F, data shown are mean ± SE (n = 4–5), *p < 0.05, t-test. For H, data shown are fold change of cells in different cell cycle phases normalized to the relative WT controls, which was arbitrarily defined as 1 unit. Data shown are mean ± SE (n = 5–9). *p < 0.05, t-test. For 3F, p-value = 0.0002, one way ANOVA, and for 3H, p-value of S-G2-M = 0.0027, p-value of G1 = 0.9521, and p-value of dead cells = 0.0013.

FIGURE 4

CDK6 kinase activity is required for differentiation of precursors derived from MEFs and ADSCs into adipocytes. (A–C, E–G) Representative images of Oil Red O staining of differentiated cells from MEFs (A–C) or SVFs (E–G) isolated from WT, K43M, and R31C mice in the presence of WAT inducers. (D, H, L) qRT-PCR analyses for expression levels of WAT-associated transcriptional factors and WAT-related genes with 36B4 mRNA as an internal control. Data shown are fold changes of each mRNA normalized to the relative controls, either WT (D, H) or (K43M + V) (L), which was arbitrarily defined as 1 unit. (I, J) K43M MEFs transduced with MigR1-Vctor [(I), K43M + V] and with MigR1-Cdk6 [(J), K43M + Cdk6] were grown to confluent and then induced to differentiate. At day 8 post-induction, cells were stained for lipid droplets with Oil Red O. (K), Immunoblots of the protein levels of CDK6 from the differentiated cells (I, J). Tubulin was used as loading control. For (D, H, L), data shown are mean ± SE (n = 3). *, p < 0.05, T-test, significantly different from the respective control. For 4D and 4H, p value was labeled on the top of each group, one way ANOVA.

FIGURE 5

Knock-down of CDK6 in 3T3-L1 cells copied the defect of K43M in differentiation. (A, D) Immunoblots of the protein levels of CDK6 and CDK4 in pre-adipocytes (A) and differentiated cells (D) of control (CTR) and KO cells. Vinculin was used as loading control. (B) Representative flow cytometric cell cycle profiles of PI+ pre-adipocytes. (C) Histograms summarizing the cell cycle distribution of the cells in (B). Data shown are mean ± s.e. (n = 3); *p <0.05 vs CTR, T-test. (E, F) Representative images of Oil Red O staining of differentiated cells from CTR (E) or KO (F) cells in the presence of WAT inducers for 7 days. (G) qRT-PCR analyses for expression levels of WAT-associated transcriptional factor and WAT-related genes. Data shown are mean ± SE (n = 3). *, p < 0.05, T-test, significantly different from the CTR.

FIGURE 6

Ablation of RUNX1 in mature adipocytes rescued the defect of precursor numbers in K43M mice. (A,C, E) Representative flow cytometric profiles of Sca-1⁺ (A), Sca-1⁺CD36⁺ (C), and Lin⁻Sca-1⁺CD24⁺CD29⁺CD34⁺ (E) cells isolated from eWAT of WT, K43M, and KR mice at 4–5 months of age. (B, D, F) Histograms summarizing Sca-1⁺ cells in panel (A, B), Sca-1⁺CD36⁺ cells in panel (C, D), and Lin⁻Sca-1⁺CD24⁺CD29⁺CD34⁺ cells in panel (E, F). For (B, D, F), data shown are mean ± SE (n = 6–9). *p < 0.05, t-test. For B, p-value = 0.0009, one way ANOVA, for D, p-value = 0.0406, one way ANOVA, and for F, p-value = 0.0164, one way ANOVA.

FIGURE 7

Ablation of RUNX1 in K43M mature adipocytes rescued the defects in reduced BrdU incorporation, reduced proliferation, and increased dead cells in K43M mice. (A-C) Representative immunofluorescent detection of BrdU-labeled cells (green) in eWAT derived from 4-month-old male WT, K43M, and KR (n = 4) mice which were administered BrdU for three consecutive days. Cell nuclei counterstained with DAPI (blue). Magnification ×40. (D, E) Representative flow cytometric profiles of negative control with SVF cells stained with normal mouse IgG-FITC and Sca-1-APC (D) and Sca-1⁺BrdU⁺ cells isolated from eWAT of male WT, K43M, and KR mice at 4–5 months of age (E). (F) Histograms summarizing the Sca-1⁺BrdU⁺ cells in panel (E). (G) Representative flow cytometric cell cycle profiles of Sca-1⁺7AAD⁺ cells isolated from eWAT of male WT, K43M, and KR mice at 4–5 months of age. (H) Histograms summarizing the Sca-1⁺7AAD⁺ cells in panel (G). For (F, H), data shown are fold change of Sca-1⁺BrdU⁺cells (E) or Sca-1⁺7AAD⁺ (G) normalized to the relative WT controls, which was arbitrarily defined as 1 unit. Data shown are mean ± SE (n = 4–5). *p < 0.05, t-test. For 7F, p-value = 0.0065, one way ANOVA, and for 7H, p-value of S-G2-M = 0.0010, p-value of G1 = 0.8963, and p-value of dead cells = 0.0170.

FIGURE 8

Ablation of RUNX1 in K43M precursors rescued the defects of K43M in differentiation into white adipocytes. (A, B) Fluorescent photomicrographs of differentiated cells from K43M;Runx1 ^fl/fl + GFP-CRE (A) or K43M;Runx1 ^+/+ + GFP-CRE (B) in the presence of WAT inducers. Red fluorescence indicates positive Oil Red O staining. Green fluorescence indicates the expression of GFP-CRE. The yellow fluorescence indicates merged images from red and green fluorescence. Scale bar: 50 μm. (C) Relative mRNA levels of WAT related transcriptional factors and WAT markers in differentiated cells in the presence of WAT inducers. Data shown are fold change of different mutants normalized to their relative K43M controls, which was arbitrarily defined as 1 unit, *p < 0.05, vs. K43M, t-test (n = 5). (D) Immunoblots of the protein levels of RUNX1 and CDK6 in differentiated cells from 100 μg of cell lysates after 7 days in the presence of WAT inducers. α-tubulin was used as loading control. For 8c, p-value was labeled on the top of each group, one way ANOVA.