Endocrine regulation of energy metabolism by the skeleton

Abstract

The regulation of bone remodeling by an adipocyte-derived hormone implies that bone may exert a feedback control of energy homeostasis. To test this hypothesis we looked for genes expressed in osteoblasts, encoding signaling molecules and affecting energy metabolism. We show here that mice lacking the protein tyrosine phosphatase OST-PTP are hypoglycemic and are protected from obesity and glucose intolerance because of an increase in beta-cell proliferation, insulin secretion, and insulin sensitivity. In contrast, mice lacking the osteoblast-secreted molecule osteocalcin display decreased beta-cell proliferation, glucose intolerance, and insulin resistance. Removing one Osteocalcin allele from OST-PTP-deficient mice corrects their metabolic phenotype. Ex vivo, osteocalcin can stimulate CyclinD1 and Insulin expression in beta-cells and Adiponectin, an insulin-sensitizing adipokine, in adipocytes; in vivo osteocalcin can improve glucose tolerance. By revealing that the skeleton exerts an endocrine regulation of sugar homeostasis this study expands the biological importance of this organ and our understanding of energy metabolism.

Figure 1. Increased insulin secretion and β-cell proliferation in Esp −/− mice

(A) LacZ stained tissues from newborn Esp −/ − mice demonstrating Esp locus activity in bone and testis but not in pancreas or fat pads. (B) Expression of Esp in osteoblasts, adipocytes, and pancreatic islets by real time PCR in 1 month-old mice. (C) Southern blot analysis showing efficient recombination at the Esp locus in osteoblasts of Esp_osb −/ − mice. (D) Using real time PCR Esp expression is 90% decreased in osteoblasts but not altered in testis of Esp_osb −/ − mice. (E) Decreased percentage at weaning of Esp −/ − pups born from crosses between Esp+/ − mice. (F) Lower survival at birth and at weaning of Esp −/ − pups born from Esp+/ − and Esp −/ − mothers. (G and H) Blood glucose levels (G) and serum insulin levels (H) in WT and Esp −/ − newborn before feeding (P0) or after random feeding at indicated ages. (I-J) GSIS (I) and GTT (J) test in 1 month-old WT and Esp −/ − mice. (K) H&E staining, insulin immunostaining and insulin/Ki67 double immunostaining showing larger islets and increased β-cell proliferation in pancreas of WT and 1 month-old Esp −/ − mice. Arrowheads indicate islets and arrows point at Ki67 positive cells. Scale bars, 100μm except upper panels, 800μm. Histomorphometric comparisons of islet number, size and β-cell mass between 1 month-old WT and Esp −/ − mice (lowest panel). (L) Pancreas insulin content in 1 month-old WT and Esp −/ − mice. (M) Quantification of the number of Ki67 immunoreactive cells in pancreatic islets of P5 and 1 month-old WT and Esp −/ − mice. All panels except I and J, °p<0.05 and *p<0.01 vs WT (Student’s t test). Panels I and J, °p<0.05 vs WT and *p 0.001 vs WT (ANOVA followed by post hoc analysis).

Figure 2. Increased insulin sensitivity and Adiponectin expression in Esp −/ − mice

All experiments compare 1 month-old mice WT and Esp −/ − unless otherwise indicated. (A) ITT. (B) Glucose infusion rate during hyperinsulinemic-euglycemic clamp. (C) Expression of markers of insulin sensitivity in skeletal muscle measured by real-time PCR. (D) Electron microscopy images (upper panel, 20,000X) and corresponding quantification (lower panel) of mitochondrial area in gastrocnemius muscle. Scale bars, 1μm .(E) Decreased number of lipid droplets on oil red O stained liver sections (upper panel) and modified expression of insulin target genes by real time PCR (lower panel) in Esp_osb −/ − mice. Scale bars, 50 μm. (F) Fat pad mass (fat pad weight over body weight). (G) Energy expenditure. (H) Serum triglyceride levels after an overnight fast. (I) H&E staining of adipose tissues of WT and Esp −/ − mice (upper panel) and respective distribution of diameters for 100 measured adipocytes per slide (lower panel). Scale bars, 50μm. (J) Expression of markers of adipogenesis, lipogenesis, fat uptake, and lipolysis in fat. (K) Serum free fatty acid (FFA) in fed and overnight-fasted mice. (L) Expression of Leptin, Resistin and Adiponectin in fat. (M) Serum levels of adiponectin in newborn mice before feeding (P0) and after random feeding at other indicated ages. (N) Expression of adiponectin target genes in tissues of WT and Esp −/ − mice. Panel A, °p<0.05 vs WT and *p 0.001 vs WT (ANOVA followed by post hoc analysis); panels B-N, *p<0.01 vs WT (Student’s t test).

Figure 3. Esp −/− mice are protected from obesity and glucose intolerance

(A-F) Food intake per day (A), body weight curve (B), fat pad mass (C), serum triglyceride levels (D), GTT (E) and ITT (F) in 4 month-old WT and Esp −/ − mice 3 months after GTG or vehicle injection. (G-I) Body weight curve (G), GTT (H), and ITT (I) in 3 month-old WT and Esp −/ − mice fed a high fat diet for 6 weeks. (J and K) Serum insulin levels (J) and pancreatic insulin content (K) in 1 month-old WT and Esp −/ − mice 8 days after STZ or vehicle injection. (L and M) Survival of mice (L) and change of blood glucose levels (M) in 1 month-old WT and Esp −/ − mice during the 8 days following STZ injection. (N) Urinary glucose assays in 1 month-old WT and Esp −/ − mice 8 days after STZ injection. Panels A-F, J and K: a, WT vs Esp −/ −; b, WT+GTG(or STZ) vs WT+vehicle; c, WT+GTG(or STZ) vs Esp −/ − +GTG(or STZ); d, Esp −/ − +GTG(or STZ) vs Esp −/ − +vehicle. Panels G-I and M, *p<0.05 WT vs Esp −/ −. Panels A, C, D, J and K: Student’s t test, p<0.05 for a-d; panels B, E-I, L and M: ANOVA followed by post hoc analysis when number of groups>2, p 0.001 for a-d.

Figure 4. Osteoblasts secrete a factor regulating Insulin and Adiponectin expression

(A-E) All experiments compare 1 month-old WT and α1(I)–Esp mice. (A) Insulin immunostaining (upper panel) and histomorphometric comparisons of islet number, size, β-cell mass and Ki67 immunoreactive cells in pancreas (lower panel). Scale bars, 100μm. (B) Blood glucose, and serum insulin and adiponectin levels. (C) GSIS test. (D) GTT. (E) ITT. (G) Expression of Insulin and Glucagon in WT islets co-cultured with fibroblasts or osteoblasts. (H) Expression of Adiponectin and Leptin in WT adipocytes co-cultured with fibroblasts or osteoblasts. (I) Expression of Insulin and Adiponectin in Esp −/ − indicated cells co-cultured with fibroblasts or osteoblasts. (J and K) Expression of Insulin (J) and Adiponectin (K) in WT indicated cells co-cultured with or without osteoblasts in presence of a filter preventing cell-cell contact or in presence of conditioned medium (CM) collected from osteoblast cultures. Panels A, B and F-J: *p<0.05 vs WT (Student’s t test); panels C-E: °p<0.05 vs WT and *p 0.001 vs WT (ANOVA).

Figure 5. Osteocalcin regulatesβ-cell proliferation, insulin secretion and insulin sensitivity

All experiments compare 3 month-old mice WT and Ocn −/ − mice unless otherwise indicated. (A) Blood glucose levels after random feeding. (B) Insulin levels. (C) GSIS test. (D) GTT. (E) ITT. (F) Glucose infusion rate during hyperinsulinemic-euglycemic clamp. (G) Energy expenditure. (H) Expression of insulin target genes by real time PCR. (I) Histomorphometric comparisons of islet numbers, islet size, β-cell mass, insulin content in pancreas and Ki67 immunoreactive cells in pancreatic islets. P5, 5 day-old pups; 3M, 3 month-old mice. (J) Fat pad mass (fat pad weight over body weight). (K) Serum triglyceride levels after an overnight fast. (L and M) serum levels (L) and gene expression (M) of adiponectin. (N) Expression of adiponectin target genes by real time PCR. (O) Expression of Insulin and Glucagon in WT pancreatic islets co-cultured with osteoblasts of indicated genotypes. (P) Expression of Adiponectin and Leptin in WT adipocytes co-cultured with osteoblasts of indicated genotypes. (Q) Expression of Insulin and Adiponectin in WT indicated cells cultured in presence of conditioned media from COS cells transfected with an Osteocalcin expression vector or its empty counterpart. (R) Expression of Insulin and Adiponectin in WT islets and adipocytes co-cultured with fibroblasts in presence of recombinant osteocalcin (3ng/ml) or vehicle, or with osteoblasts expressing (5d) or not (1d) Osteocalcin. (S and T) Dynamic of glucose (S) and insulin levels (T) in Ocn −/ − mice injected simultaneously with glucose and 20ng of recombinant osteocalcin or vehicle. Panels A, B, F-R: *p<0.05 vs WT (Student’s t test); panels C-E, S and T, °p 0.01 vs WT and *p 0.001 vs WT (ANOVA).

Figure 6. Osteocalcin regulates insulin sensitivity via adiponectin

(A-E) Comparison between 6 week-old WT, Adiponectin+/ − (Adipo+/ −), Osteocalcin+/ − (Ocn+/ −) , and Ocn+/ −; Adipo+/ − mice. (A) ITT. (B) Insulin serum levels. (C) Blood glucose levels. (D) GSIS test. (E) Adiponectin serum levels. Panels A and D, *p 0.001 vs WT (ANOVA followed by post hoc analysis); panels B, C and E: *p<0.05 vs WT (Student’s t test).

Figure 7. Esp −/− mice are a model of increased osteocalcin bioactivity

(A-G) Comparison between 6 week-old WT, Esp −/ −, Ocn+/ −, and Esp −/ −;Ocn+/ − mice. (A) Blood glucose levels. (B) Serum insulin levels. (C) Serum adiponectin levels. (D) GTT. (E) ITT. (F) GSIS test. (G) Quantification of the number of Ki67 immunoreactive cells in pancreatic islets. (H and I) Quantification of the percentage of osteocalcin bound to hydroxyapatite (HA) resin after a 15 min. incubation of serum of 1 month-old mice of indicated genotypes (H) or of conditioned medium from osteoblast cultures treated with warfarin or vehicle (I). (J) Expression of Adiponectin in WT adipocytes co-cultured with osteoblasts treated with warfarin or vehicle. (K) Expression of Adiponectin in WT adipocytes cultured in presence of vehicle or of 1ng/ml of commercially available carboxylated osteocalcin (Immunotopics) or bacterially produced uncarboxylated osteocalcin. (L) Expression of Insulin and CyclinD1 in WT islets cultured in presence of 3ng/ml of bacterially produced uncarboxylated osteocalcin or vehicle. Panels A-C and G-L: *p<0.05 vs WT (Student’s t test); panels D-F, °p<0.05 vs WT and *p 0.001 vs WT (ANOVA followed by post hoc analysis).