SIRT1 regulates differentiation of mesenchymal stem cells by deacetylating β-catenin

Abstract

Mesenchymal stem cells (MSCs) are multi-potent cells that can differentiate into osteoblasts, adipocytes, chondrocytes and myocytes. This potential declines with aging. We investigated whether the sirtuin SIRT1 had a function in MSCs by creating MSC specific SIRT1 knock-out (MSCKO) mice. Aged MSCKO mice (2.2 years old) showed defects in tissues derived from MSCs; i.e. a reduction in subcutaneous fat, cortical bone thickness and trabecular volume. Young mice showed related but less pronounced effects. MSCs isolated from MSCKO mice showed reduced differentiation towards osteoblasts and chondrocytes in vitro, but no difference in proliferation or apoptosis. Expression of β-catenin targets important for differentiation was reduced in MSCKO cells. Moreover, while β-catenin itself (T41A mutant resistant to cytosolic turnover) accumulated in the nuclei of wild-type MSCs, it was unable to do so in MSCKO cells. However, mutating K49R or K345R in β-catenin to mimic deacetylation restored nuclear localization and differentiation potential in MSCKO cells. We conclude that SIRT1 deacetylates β-catenin to promote its accumulation in the nucleus leading to transcription of genes for MSC differentiation.

Figure 1. The role of SIRT1 in mesenchymal progenitor cells in aged (2.2 years old) mice in vivo

1. A. Western blot analysis of SIRT-1 in mesenchymal progenitor cells from the bone marrow.

2. B. Subcutaneous fat weight of fl/fl and MSCKO mice. Bars, SD; *p = 0.01, t-test, n = 8 per group.

3. C. Haemalaun–eosin staining of subcutaneous adipose tissue, 20× magnification.

4. D. Body weight of fl/fl and MSCKO mice. Bars, SD; *p = 0.006 versus fl/fl, t-test, n = 8 per group.

5. E. Blood tryglyceride level in fl/fl and MSCKO mice. Bars, SD; *p = 0.04, t-test, n = 8 per group.

6. F. Bone marrow histology of fl/fl and MSCKO mice (haemalaun–eosin staining, 5× magnification).

7. G–I. Longitudinal microCT image (G) through femur depicting areas of midshaft and distal femur analysed in experiments. MicroCT images of the (H) midshaft femur (left) and (I) distal femur, Ct.Th, cortical thickness; TV/BV trabecular volume/bone volume; results are mean ± SD; *p = 0.01, t-test, n = 6 per group.

8. J, K. Histomorphometry of femurs. (J) BFR/BS, bone formation rate/bone surface, Bars, SD; *p = 0.02, t-test, n = 6 per group; (K) ObS/BS, osteoblast surface/bone surface, Bars, SD; *p = 0.04, t-test, n = 6 per group.

Figure 2. The role of SIRT1 in mesenchymal progenitor cells in young (2-month-old) mice in vivo

1. A, B. Subcutaneous fat (A) and overall body weight (B) of fl/fl and MSCKO mice. Bars, SD; p = 0.05, t-test, n = 6 per group.

2. C. Haemalaun–eosin staining of subcutaneous adipose tissue, 20× magnification.

3. D. Blood tryglyceride level in fl/fl and MSCKO mice. Bars, SD; p = 0.04, t-test, n = 6 per group.

4. E. Bone marrow histology of fl/fl and MSCKO mice (haemalaun–eosin staining, 5× magnification).

5. F. MicroCT images of the distal femur, transversal section on the left, 3D longitudinal reconstruction on the right.

6. G. MicroCT analysis of distal femur. TV/BV, trabecular volume/bone volume, Tb.N, trabecular number, Tb.Th. trabecular thickness, Bars, SD; *p = 0.05, t-test, n = 6 per group.

Figure 3. The effect of SIRT1 on differentiation of MSCs in vitro

1. A–D. Single cell MSC proliferation and differentiation. Differentiation to osteoblasts on the left and to adipocytes on the right. (A) Cell number per well (n = 12 wells per group), (B) FACS analysis of apoptotic annexin V positive cells per well (n = 6 wells per group), (C) FACS analysis of ALP⁺ osteoblasts per well (*p = 0.04, t-test, n = 12 wells per group) on the left and FACS analysis of FABP4⁺ adipocytes per well (n = 12 wells per group) on the right, (D) FACS analysis of STRO-1⁺ mesenchymal progenitor cells per well (*p = 0.03 on the left and p = 0.05 on the right, t-test, n = 6 wells per group); bars, SD.

2. E. FACS analysis of percentage of MSCs in the bone marrow of old (left, *p = 0.01, t-test, n = 6 per group) and young (right) mice.

3. F. Alkaline phosphatase (ALP) staining of MSCs (isolated by the plating method) differentiated towards osteoblasts. BFU bone forming units, numbers are mean ± SD; *p = 0.001, t-test, n = 12 wells per group.

4. G. Alcian blue staining of MSCs differentiated towards adipocytes. CFU/alcian blue⁺, colony forming units/Alcian blue positive, numbers are mean ± SD; *p = 0.03, t-test, n = 6 wells per group.

5. H. LipidTOX staining of MSCs differentiated towards adipocytes, *p = 0.05, t-test, n = 6 wells per group.

6. I. ALP staining of MSCs (isolated by the plating method) after fl/fl and MSCKO bone marrow transplantation (BMT) and differentiated towards osteoblasts. Numbers are mean ± SD; *p = 0.01, t-test, n = 12 wells per group.

Figure 4. SIRT1 promotes nuclear localization of β-catenin and its activity in mesenchymal progenitor cells

1. Alkaline phosphatase staining of fl/fl MSCs treated with SIRT1 inhibitor EX527. BFU bone forming units, mean ± SD, *p = 0.004, t-test, n = 6 wells per group.

2. Western blot analysis of p53 acetylation as a marker of SIRT1 activity and bar graph showing acetyl53/p53 densitometry fold change.

3. Alkaline phosphatase staining of MSCKO mesenchymal progenitor cells differentiated towards bone after transfection with control vector, SIRT1 overexpressing virus, and SIRT1 deacetylase inactive HY mutant. BFU bone forming units, mean ± SD, *p = 0.001, ANOVA Dunnett test, n = 6 wells per group.

4. qPCR analysis of β-catenin targets, n = 4 mesenchymal progenitor samples per group, *p = 0.01–0.05, t-test.

5. Luciferase reporter assays of fl/fl and MSCKO MSCs. Transcriptional activity of β-catenin in MSCs was detected using pTOP-flash (containing triple Tcf/Lef1 binding sites, the basic thymidine kinase promoter and firefly luciferase reporter gene) or pFOP-flash (containing mutated Tcf/Lef1 binding sites) plasmids and was normalized to Renilla luciferase. *p = 0.003, t-test, n = 12 wells per group, three independent experiments.

6. Immunofluorescent staining of β-catenin and DAPI in fl/fl and MSCKO MSCs transfected with pCMV- β-cateninT41A (β-catenin*). Thirty-four plus or minus eight percent of cells showed nuclear localization in fl/fl MSCs, while 0% showed nuclear staining in MSCKO MSCs.

Figure 5. SIRT1 deacetylates β-catenin leading to nuclear localization and differentiation of MSCs

1. Western blot analysis of old MSCs treated with SIRT1 deacetylase inhibitor EX527 before and after differentiation to osteoblasts. SIRT1 was blotted before immunoprecipitation of β-catenin and acetyl-lysine following immunoprecipitation of β-catenin.

2. Acetylation of β-catenin in young MSCs following the treatment with SIRT1 inhibitor, EX527. MSC samples were immunoprecipitated with β-catenin antibody and membranes were probed with acetyl-lysine and β-catenin antibody.

3. Immunofluorescent staining of β-catenin and DAPI in fl/fl and MSCKO. MSCs were transfected with the stable form of β-catenin (T41A) and a β-catenin mutant K49R K345R. Nuclear staining was 0% of cells for T41A and 29 ± 9% of cells for T41A K49R K345R.

4. Alkaline phosphatase staining of MSCKO mesenchymal progenitor cells differentiated towards bone after transfection with the control vector, stable form of β-catenin (T41A) and β-catenin mutant where K49R, K345R. %+ area, percentage of alkaline phosphatase positive area of the well; numbers are mean ± SD, *p = 0.005 versus lacZ, ANOVA Dunnett test.