Partial agonist, telmisartan, maintains PPARγ serine 112 phosphorylation, and does not affect osteoblast differentiation and bone mass

Abstract

Peroxisome proliferator activated receptor gamma (PPARγ) controls both glucose metabolism and an allocation of marrow mesenchymal stem cells (MSCs) toward osteoblast and adipocyte lineages. Its activity is determined by interaction with a ligand which directs posttranscriptional modifications of PPARγ protein including dephosphorylation of Ser112 and Ser273, which results in acquiring of pro-adipocytic and insulin-sensitizing activities, respectively. PPARγ full agonist TZD rosiglitazone (ROSI) decreases phosphorylation of both Ser112 and Ser273 and its prolonged use causes bone loss in part due to diversion of MSCs differentiation from osteoblastic toward adipocytic lineage. Telmisartan (TEL), an anti-hypertensive drug from the class of angiotensin receptor blockers, also acts as a partial PPARγ agonist with insulin-sensitizing and a weak pro-adipocytic activity. TEL decreased S273pPPARγ and did not affect S112pPPARγ levels in a model of marrow MSC differentiation, U-33/γ2 cells. In contrast to ROSI, TEL did not affect osteoblast phenotype and actively blocked ROSI-induced anti-osteoblastic activity and dephosphorylation of S112pPPARγ. The effect of TEL on bone was tested side-by-side with ROSI. In contrast to ROSI, TEL administration did not affect bone mass and bone biomechanical properties measured by micro-indentation method and did not induce fat accumulation in bone, and it partially protected from ROSI-induced bone loss. In addition, TEL induced "browning" of epididymal white adipose tissue marked by increased expression of UCP1, FoxC2, Wnt10b and IGFBP2 and increased overall energy expenditure. These studies point to the complexity of mechanisms by which PPARγ acquires anti-osteoblastic and pro-adipocytic activities and suggest an importance of Ser112 phosphorylation status as being a part of the mechanism regulating this process. These studies showed that TEL acts as a full PPARγ agonist for insulin-sensitizing activity and as a partial agonist/partial antagonist for pro-adipocytic and anti-osteoblastic activities. They also suggest a relationship between PPARγ fat "browning" activity and a lack of anti-osteoblastic activity.

Figure 1. The effect of TEL and ROSI on adipocytic phenotype of U-33/γ2 cells.

Cells were treated for 3 days with either vehicle (DMSO) (V), or ROSI (R), or TEL (T), or ROSI and TEL (RT). A. Number of adipocytes in response to treatment with TEL at different concentrations or with 1 µM ROSI and assessed by staining of intracellular lipids with Oil Red O. B. Relative expression of WAT-specific gene markers in response to either 1 µM ROSI, or 50 µM TEL, or 1 µM ROSI and 50 µM TEL. C. Relative expression of BAT-specific gene markers in cells treated as in B. * p<0.05 vs. vehicle.

Figure 2. The effect of TEL and losartan (LOS) on proliferation of U-33/γ2 and U-33/c cells.

Cell proliferation was assessed using MTT assay after 3 days treatment with tested compounds at different concentrations. A. Dose response of U-33/γ2 cells to treatment with TEL. B. Dose response of U-33/c cells to treatment with TEL. C. Dose response of U-33/γ2 cells to treatment with LOS. V – vehicle; R – ROSI; T – TEL; L – LOS. * p<0.05 vs. vehicle.

Figure 3. The effect of TEL, LOS, and ROSI on osteoblastic phenotype of U-33/γ2 cells.

Cells were treated for 3 days with either vehicle (DMSO) (V), or ROSI (R), or TEL (T), or LOS (L), or in combination (RT or RL). A. Enzymatic activity of alkaline phosphatase (ALP) after treatment with different doses of TEL or 1 µM ROSI. B. Relative expression of osteoblast-specific transcription factors, Runx2 and Osterix, in cells treated with either 1 µM ROSI, or 50 µM TEL, or in combination. C. ALP activity in cells treated with either 1 µM ROSI, or 50 µM TEL, or in combination. D. ALP activity in cells treated with either 1 µM ROSI, or 50 µM LOS, or combination. ALP activity was normalized to the number of cells assessed in MTT proliferation assay (panels A, C, and D). * p<0.05 vs. vehicle.

Figure 4. TEL effect on expression of members of TGFβ (A) and BMP (B) signaling pathways.

U-33/γ2 cells were treated for 3 days with either vehicle (DMSO) (V), or 1 µM ROSI (R), or 50 µM TEL (T), or in combination (RT). * p<0.05 vs. vehicle.

Figure 5. Western blot analysis of ^S112pPPARγ (A) and ^S273pPPARγ (B) protein levels after treatment for 60 min with either vehicle (DMSO) (V), or 1 µM ROSI (R), or 50 µM TEL (T), or in combination (RT).

* p<0.05 vs. vehicle, # p<0.05 vs. ROSI, ∧ p<0.05 vs. TEL.

Figure 6. TEL effect on glucose disposal and bone structure.

A. Effect of 4 days administration of either regular diet (open circles), or diet supplemented with 1.5 mg/kg/d TEL (open triangles), or 3 mg/kg/d TEL (black triangles), or 20 mg/kg/d ROSI (open squares), on glucose tolerance of A^vy/a mice measured with introperitoneal glucose tolerance test (IGTT), as described in Material and Methods (n = 4 animals per group). B. Glucose disposal measured with IGTT in DIO mice at the end of 4 wks administration of either non-supplemented HFD (open circles), or HFD supplemented with 3 mg/kg/d TEL (black triangles), or with 20 mg/kg/d ROSI (open squares) (n = 8 animals per group). C. mCT analysis of L4 vertebra trabecular bone of A^vy/a mice after 4 wks administration of either control non-supplemented diet (C), or chow supplemented with 20 mg/kg/d ROSI (R), or drinking water supplemented with 3 mg/kg/d TEL (T). BV/TV – bone volume fraction in the region of interest (ROI) (%); Tb.N. – average number of trabeculae per unit length (1/mm) of ROI; Tb.Th. – trabecular thickness (mm); Tb.Sp. – trabecular separation representing mean distance between trabeculae (mm). D. Number of adipocytes in proximal tibia of experimental animals (n = 4 per group). C – control; R – ROSI; T – TEL. * p<0.05 vs. control.

Figure 7. TEL effect on energy metabolism parameters.

A. Expression of metabolic gene markers in eWAT. B. Respiratory parameters of DIO mice after 4 wks of treatment measured in CLAMS metabolic cages during a dark day cycle (12 h). C – control; R – ROSI; T – TEL. *p<0.05 vs. control.