Increased bone adiposity and peroxisomal proliferator-activated receptor-gamma2 expression in type I diabetic mice

Abstract

Decreased bone mass, osteoporosis, and increased fracture rates are common skeletal complications in patients with insulin-dependent diabetes mellitus (IDDM; type I diabetes). IDDM develops from little or no insulin production and is marked by elevated blood glucose levels and weight loss. In this study we use a streptozotocin-induced diabetic mouse model to examine the effect of type I diabetes on bone. Histology and microcomputed tomography demonstrate that adult diabetic mice, exhibiting increased plasma glucose and osmolality, have decreased trabecular bone mineral content compared with controls. Bone resorption could not completely account for this effect, because resorption markers (tartrate-resistant acid phosphatase 5b, urinary deoxypyridinoline excretion, and tartrate-resistant acid phosphatase 5 mRNA) are unchanged or reduced at 2 and/or 4 wk after diabetes induction. However, osteocalcin mRNA (a marker of late-stage osteoblast differentiation) and dynamic parameters of bone formation were decreased in diabetic tibias, whereas osteoblast number and runx2 and alkaline phosphatase mRNA levels did not differ. These findings suggest that the final stages of osteoblast maturation and function are suppressed. We also propose a second mechanism contributing to diabetic bone loss: increased marrow adiposity. This is supported by increased expression of adipocyte markers [peroxisome proliferator-activated receptor gamma2, resistin, and adipocyte fatty acid binding protein (alphaP2)] and the appearance of lipid-dense adipocytes in diabetic tibias. In contrast to bone marrow, adipose stores at other sites are depleted in diabetic mice, as indicated by decreased body, liver, and peripheral adipose tissue weights. These findings suggest that IDDM contributes to bone loss through changes in marrow composition resulting in decreased mature osteoblasts and increased adipose accumulation.

Figure 1. Diabetes increases serum glucose, serum osmolality and serum triglycerides and is associated with decrease body weight

Serum from 5, 14, 21, and 28 days diabetic and control Balb/c mice was analyzed for glucose concentration, osmolality, and triglyceride levels. Body weight was recorded at every time point. Values are averages +/− SE obtained from 8–11 mice per condition. *p<0.05

Figure 2. Diabetes reduces trabecular bone volume

(A) Representative μCT lateral slices (20μm) through the proximal tibia from two control and diabetic mice are shown. (B) Representative μCT transverse slices (20μm) through the proximal tibia at a distance of 0.5 mm or 1 mm from the lowest region of the growth plate.

Figure 3. Systemic and tibia specific measures of osteoclast activity in diabetic bone

(A) Diabetes decreases serum active TRAP5b levels and has no effect on urine DPD levels. Urine collected over a 24 hour period was obtained from diabetic and control mice in a metabolic cage at day 14. DPD was determined in control and diabetic urine samples and expressed relative to creatinine, to control for differences in urine volume. Serum was obtained from mice at 5, 14, 21 or 28 days after the induction of diabetes. Active TRAP5b levels were determined in control (C) and diabetic (D) mouse serum samples. (B) Histomorphometry and TRAP5 mRNA analyses were preformed on tibias obtained from mice 28 days after the induction of diabetes. All values are averages +/− SE obtained from 4–9 mice per condition. *p<0.05.

Figure 3. Systemic and tibia specific measures of osteoclast activity in diabetic bone

(A) Diabetes decreases serum active TRAP5b levels and has no effect on urine DPD levels. Urine collected over a 24 hour period was obtained from diabetic and control mice in a metabolic cage at day 14. DPD was determined in control and diabetic urine samples and expressed relative to creatinine, to control for differences in urine volume. Serum was obtained from mice at 5, 14, 21 or 28 days after the induction of diabetes. Active TRAP5b levels were determined in control (C) and diabetic (D) mouse serum samples. (B) Histomorphometry and TRAP5 mRNA analyses were preformed on tibias obtained from mice 28 days after the induction of diabetes. All values are averages +/− SE obtained from 4–9 mice per condition. *p<0.05.

Figure 4. Osteocalcin mRNA and serum levels are decreased in diabetes, in contrast to runx-2 and alkaline phosphatase (Alk Phos) mRNA levels

Total RNA, extracted from tibia isolated from control and diabetic (2 week) adult Balb/c mice, was used for real time RT-PCR analysis with SYBR green dye. Levels of amplified genes were expressed relative to cyclophilin (a housekeeping gene) and relative to control levels, which were set to 1. Serum from 2 week diabetic and control mice was analyzed for osteocalcin levels. Values are averages +/− SE obtained from 8–11 mice per condition. *p<0.05

Figure 5. Adipocyte markers, PPARγ2, aP2 and Resistin are increased in diabetic bone

Total RNA was extracted from tibia isolated from control and diabetic (2 week) adult Balb/c mice, and used for real time RT-PCR analysis with SYBR green dye. Levels of PPARγ2, aP2 and resistin are expressed relative to cyclophilin (a housekeeping gene). Values are averages +/− SE obtained from 8–11 mice per condition. *p<0.05.

Figure 6. Adiposity is increased in diabetic tibia, in contrast to peripheral adipose tissue lipolysis

Tibia (bone) obtained from control and diabetic (4 weeks) mice were fixed and processed for plastic embedding and sectioning. Representative Masson-Goldner trichrome stained sections demonstrate differences in the amount of mineralized bone (blue stain) and in the number of adipocytes (unstained large circular regions within the marrow) between control and diabetic animals. Representative fat pad and liver sections were obtained from control and diabetic mice. Sections were stained with hematoxylin & eosin and digitally photographed.

Figure 7. Peripheral adipose tissue is decreased in diabetic mice

Transverse slices of mouse legs were obtained by μCT analysis. Representative images are shown, with medium grey (marked by black arrows) indicating regions of fat. In addition, dissected femoral fat depots were photographed and weighed. The later values were pooled from 5–9 mice per condition and expressed as averages +/− SE. *p<0.01.