PGC-1α Controls Skeletal Stem Cell Fate and Bone-Fat Balance in Osteoporosis and Skeletal Aging by Inducing TAZ

Abstract

Aberrant lineage specification of skeletal stem cells (SSCs) contributes to reduced bone mass and increased marrow adipose tissue (MAT) in osteoporosis and skeletal aging. Although master regulators of osteoblastic and adipogenic lineages have been identified, little is known about factors that are associated with MAT accumulation and osteoporotic bone loss. Here, we identify peroxisome-proliferator-activated receptor γ coactivator 1-α (PGC-1α) as a critical switch of cell fate decisions whose expression decreases with aging in human and mouse SSCs. Loss of PGC-1α promoted adipogenic differentiation of murine SSCs at the expense of osteoblastic differentiation. Deletion of PGC-1α in SSCs impaired bone formation and indirectly promoted bone resorption while enhancing MAT accumulation. Conversely, induction of PGC-1α attenuated osteoporotic bone loss and MAT accumulation. Mechanistically, PGC-1α maintains bone and fat balance by inducing TAZ. Our results suggest that PGC-1α is a potentially important therapeutic target in the treatment of osteoporosis and skeletal aging.

Keywords: PGC-1α; TAZ; aging; bone; fat; lineage decision; mesenchymal stem cells; osteoporosis; skeletal stem cells.

Figure 1.. Loss of PGC-1α Potentiated Bone Loss and MAT Accumulation in Skeletal Aging

(A) mRNA expression of Ppargc1a from bone marrow SSCs of 3- and 18-month-old mice, as determined by qRT-PCR. *p < 0.05; **p < 0.01 by Student’s t test. (B) mRNA expression of Ppargc1a from marrow monocytes and/or macrophages of 3- and 18-month-old mice, as determined by qRT-PCR. *p < 0.05; **p < 0.01 by Student’s t test. (C) IHC of PGC-1α in femurs of 4- and 16-month-old mice and IOD quantification. Scale bars, 50 μm. (D) IHC of PGC-1α in Pgc1a^+/+ and Pgc1a^−/− mouse bone marrow and IOD quantification. Scale bar, 50 μm. (E) Expression of Ppargc1a in bone marrow homogenates of Pgc1a^+/+ and Pgc1a^−/− mice was determined by RT-PCR. (F) Quantitative measurements of BMD and BV/TV of femurs from 3- and 18-month-old Pgc1a^+/+ and Pgc1a^−/− mice by μCT. Scale bar, 0.3 mm. (G) Osteoblast number (Ob.N/BS) and osteoblast surface (Ob.S/BS) in young and aged Pgc1a^+/+ and Pgc1a^−/− mice. (H) ELISA of serum OCN in 3-and 18-month-old Pgc1a^+/+ and Pgc1a^−/− mice. (I) Bone formation rate (BFR) of 3-and 18-month-old Pgc1a^+/+ and Pgc1a^−/− mice as determined by dual labeling. (J) Representative H&E staining of 3-and 18-month-old mouse femurs. (K) Adipocyte number per tissue area and area of adipocytes per tissue area measured based on H&E images. (L) Representative IHC of FABP4 and IOD quantification. Scale bar, 80 μm. (M) PPARGC1A mRNA expression of SSCs isolated from human bone marrow (n > 13 subjects per group) of two age groups. All data are presented as means ± SD. **p < 0.01 by two-way ANOVA with Holm-Sidak post hoc test. For (G)–(L), n = 8 mice per group. See also Figure S1.

Figure 2.. Depletion of PGC-1α in Mesenchymal Cells using Prx1-cre Exacerbated Bone-Fat Imbalance in OVX-Induced Osteoporosis

(A) Ppargc1a mRNA in Sca1⁺CD29⁺CD45⁻CD11b⁻ SSCs (left) and bone marrow monocytes and/or macrophages (BMMΦs, right) from 2-month-old Prx1;Pgc^f/f and Prx1;Pgc^w/w mice was assessed by qRT-PCR. (B) Quantitative measurements of BMD and BV/TV of femurs from Prx1;Pgc1a^f/f and Prx1;Pgc1a^w/w mice following sham and OVX by μCT. Scale bar, 0.3 mm. (C) Osteoblast number (Ob.N/BS) and osteoblast surface (Ob.S/BS) in Prx1;Pgc1a^f/f and Prx1;Pgc1a^w/w mice. (D) Serum OCN levels in Prx1;Pgc1a^f/f and Prx1;Pgc1a^w/w mice. (E) BFR of Prx1;Pgc^f/f and Prx1;Pgc^w/w mice after OVX. (F) Representative H&E staining images of Prx1;Pgc1a^f/f and Prx1;Pgc1a^w/w mouse femurs. (G) Representative immunostaining of FABP4. Scale bar, 80 μm. (H) Adipocyte number per tissue area and area of adipocytes per tissue area measured based on H&E images. (I) Quantitative measurements of adipocyte volume and representative μCT images of rMAT in the proximal tibiae and cMAT in the distal tibiae (n = 6 mice per group). Scale bar, 0.3 mm. *p < 0.05; **p < 0.01 by two-way ANOVA with Holm-Sidak post hoc test. For (B)–(I), n = 8 mice per group. Also see Figure S2.

Figure 3.. Depletion of PGC-1α in Mesenchymal Cells using LepR-cre Exacerbated Bone-Fat Imbalance in OVX-Induced Osteoporosis

(A) Ppargc1a mRNA in Sca1⁺CD29⁺CD45⁻CD11b⁻ SSCs (left) and BMMFΦs (right) from 2-month-old LepR;Pgc1a^f/f and LepR;Pgc1a^w/w mice was assessed via qRT-PCR (left). (B) Representative μCT images of femurs from LepR;Pgc1a^f/f and LepR;Pgc1a^w/w mice following sham and OVX. Scale bar, 0.3 mm. Quantitative measurements of BMD and BV/TV by μCT. (C) Osteoblast number (Ob.N/BS) and osteoblast surface (Ob.S/BS) in LepR;Pgc1a^f/f and LepR;Pgc1a^w/w mice. (D) BFR of LepR;Pgc1a^f/f and LepR;Pgc1a^w/w mice. (E) Serum OCN level in LepR;Pgc1a^f/f and LepR;Pgc1a^w/w mice. (F) Representative H&E staining images of LepR;Pgc1a^f/f and LepR;Pgc1a^w/w mouse femurs. (G) Representative IHC of FABP4. Scale bar, 80 μm. (H) Adipocyte number per tissue area and area of adipocytes per tissue area measured based on H&E images. (I) Quantitative measurements of adipocyte volume and representative μCT images of rMAT and cMAT in tibiae of LepR;Pgc1a^f/f and LepR;Pgc1a^w/w mice following sham and OVX (n = 6 mice per group). Scale bar, 0.3 mm. All data are presented as means ± SD. *p < 0.05; **p < 0.01 by two-way ANOVA with Holm-Sidak post hoc test. For (B)–(I), n = 8 mice per group. See also Figure S3.

Figure 4.. PGC-1α Controls Osteoblastic and Adipogenic Differentiation of SSCs In Vitro

(A) Osteoblastic differentiation of SSCs isolated from Pgc1a^+/+ and Pgc1a^−/− mouse bone marrow was determined by ALP staining. (B) Mineralized nodule formation of SSCs from Pgc1a^+/+ and Pgc1a^−/− mice was assessed by alizarin red staining. (C) Adenovirus expressing FLAG-tagged PGC-1α (adPGC-1α) or control CMV (adCMV) was transfected to overexpress PGC-1α in Pgc1a^+/+ (WT) and Pgc1a^−/− (KO) SSCs. Top panel shows western blot confirming overexpression. Bottom panel shows mRNA expression of Ppargc1a in SSCs, as assessed using qRT-PCR. (D) Restoration of PGC-1α in Pgc1𝛼^−/− SSCs restored osteogenic differentiation, as determined by ALP and ARS staining. (E) qRT-PCR showing that restoration of PGC-α1 in Pgc1α^−/− SSCs restored the expression of the key osteogenic marker genes, including Runx2, Ibsp, Col1a1, and Bglap. (F) Oil red O staining of Pgc1a^+/+ and Pgc1a^−/− SSCs after 14 days of adipogenic induction. (G) The restoration of PGC-1α in Pgc1a^−/− SSCs inhibited adipogenic differentiation as determined by oil red O staining. (H) qRT-PCR showing that restoration of PGC-1α in Pgc1α^−/− SSCs inhibited the expression of the key adipogenic marker genes Pparg and Cd36. All experiments were repeated three times, and a representative dataset is presented. All data are presented as means ± SD. *p < 0.05; **p < 0.01 by Student’s t test for (A and B) and one-way ANOVA for (C and E–H). See also Figure S4.

Figure 5.. Conditional Deletion of PGC-1α in Osteoblasts Only Impaired Bone Formation but Did Not Affect MAT Accumulation after OVX

(A) RT-PCR showing that Ppargc1a was deleted in bone tissues, but not in other tissues, from 2-month-old 2.3crePgc1a^f/f compared to 2.3crePgc1a^w/w mice. (B) Quantitative measurements of BMD and BV/TV of femurs from 2.3crePgc1a^f/f and 2.3crePgc1a^w/w mice after OVX by μCT. Scale bar, 0.3 mm. (C) Histomorphometric measurements of osteoblasts in 2.3crePgc1a^f/f and 2.3crePgc1a^w/w mice. (D) BFR of 2.3cre;Pgc1a^f/f and 2.3cre;Pgc1a^w/w mice. (E) ELISA of serum OCN in 2.3cre;Pgc1a^f/f and 2.3cre;Pgc1a^w/w mice. (F) Representative H&E staining in femurs from 2.3cre;Pgc1a^f/f and 2.3cre;Pgc1a^w/w mice. (G) Adipocyte number per tissue area and area of adipocytes per tissue area in the distal end of femurs. Scale bar, 150 μm. (H) Quantitative measurements of adipocyte volume andr μCT images of rMAT and cMAT in tibiae of 2.3cre;Pgc1a^f/f and 2.3cre;Pgc1a^w/w mice (n = 6 mice per group). Scale bar, 0.3 mm. All data are presented as means ± SD. *p < 0.05; **p < 0.01 by two-way ANOVA with Holm-Sidak post hoc test. For (B)–(G), n = 8 mice per group. Also see Figure S5.

Figure 6.. Induction of PGC-1α Mitigates Bone Loss and MAT Accumulation in Osteoporosis

(A) Immunostaining showing PGC-1α expression in femurs of wt;tetO-Pgc1a and rtTA;tetO-Pgc1a mice following doxycycline induction. Scale bar, 60 μm. (B) Representative μCT images of femurs from wt;tetO-Pgc1a and rtTA;tetO-Pgc1a mice following doxycycline induction and then OVX or sham operation. Scale bar, 0.3 mm. (C) BMD and BV/TV of wt;tetO-Pgc1a and rtTA;tetO-Pgc1a mice were measured by μCT following doxycycline induction and OVX. (D) Morphometric measurements of osteoblasts in wt;tetO-Pgc1a and rtTA;tetO-Pgc1a mice following doxycycline induction and OVX. (E) BFR of wt;tetO-Pgc1a and rtTA;tetO-Pgc1a mice. (F) ELISA of serum OCN in wt;tetO-Pgc1a and rtTA;tetO-Pgc1a mice. (G) Representative tolidine blue staining delineating trabecular bone and adipocytes in the distal end of femurs. Scale bar, 150 μm. (H) Adipocyte number per tissue area and area of adipocytes per tissue area in wt;tetO-Pgc1a and rtTA;tetO-Pgc1a mice following doxycycline induction and OVX. (I) Representative IHC of FABP4. Scale bar, 50 μm. (J) Quantitative measurements of adipocyte volume and representative mCT images of rMAT and cMAT in the tibiae of wt;tetO-Pgc1a and rtTA;tetO-Pgc1a mice following sham and OVX (n = 6 mice per group). Scale bar, 0.3 mm. All data are presented as means ± SD. *p < 0.05; **p < 0.01. For (C)-(H), n = 8 mice per group. See also Figure S6.

Figure 7.. PGC-1aα Controls SSC Cell Fate Decisions through TAZ

(A) The expression of TAZ and YAP in Pgc1a^+/+ and Pgc1a^−/− SSCs was examined by western blot following osteogenic induction. Relative band intensity was normalized against α-tubulin and determined using ImageJ. The YAP image was cropped from two parts of the same blot, with two irrelevant lanes removed. (B) qRT-PCR showing Taz mRNA expression in Pgc1a^+/+ and Pgc1a^−/− SSCs following osteogenic induction. (C) qRT-PCR showing Taz mRNA expression in Pgc1a^+/+ and Pgc1a^−/− SSCs following adipogenic induction. (D) Overexpression of TAZ in Pgc1a^+/+ and Pgc1a^−/− SSCs potentiated osteogenic differentiation, as determined by ALP staining. (E) Overexpression of TAZ in Pgc1a^+/+ and Pgc1a^−/− SSCs potentiated osteogenic differentiation, as determined by alizarin red staining. (F) Overexpression of TAZ in Pgc1a^+/+ and Pgc1a^−/− SSCs inhibited adipogenic differentiation, as determined by oil red O staining. (G) ChIP assays showed that PGC-1α and NRF2 were recruited to the Taz promoter following osteogenic induction. Top panel indicates the primer position relative to the three ETS binding sites at the Taz promoter. (H) ChIP assays showed that TAZ binding to the Bglap promoter was induced following osteogenic induction. (I) ChIP assays of PGC-1α and NRF2 on the Taz promoter following adipogenic induction. (J) mRNA expression of Taz and Yap in tissue RNA isolated from bone cells in 3-month-old Pgc1a^+/+ and Pgc1a^−/− mice. (K) IHC of TAZ in 18-month-old Pgc1a^+/+ and Pgc1a^−/− mouse femurs. Scale bar, 150 μm. (L) IHC of TAZ in LepR;Pgc1a^f/f and LepR;Pgc1a^w/w mice after OVX and sham operations. Scale bar, 150 μm. (M) IHC of TAZ in wt;tetO-Pgc1a and rtTA;tetO-Pgc1a mice after sham operation. Scale bar, 100 μm. All experiments were repeated three times, and a representative dataset is presented. All data are presented as means ± SD. *p < 0.05; **p < 0.01 by Student’s t test. See also Figure S7.