Exendin-4, a Glucagon-Like Peptide-1 Receptor Agonist, Prevents Osteopenia by Promoting Bone Formation and Suppressing Bone Resorption in Aged Ovariectomized Rats

Authors


ABSTRACT

Osteoporosis mainly affects postmenopausal women and older men. Gastrointestinal hormones released after meal ingestion, such as glucose-dependent insulinotropic peptide (GIP) and glucagon-like peptide (GLP)-2, have been shown to regulate bone turnover. However, whether GLP-1, another important gastrointestinal hormone, and its analogues also have antiosteoporotic effects, especially in aged postmenopausal situation, has not been confirmed. In the present study, we evaluated the effects of the GLP-1 receptor agonist exendin-4 on ovariectomy (OVX)-induced osteoporosis in old rats. Twelve-month-old female Sprague-Dawley rats were subjected to OVX, and exendin-4 was administrated 4 weeks after the surgery and lasted for 16 weeks. Bone characters and related serum and gene biomarkers were analyzed. Sixteen weeks of treatment with exendin-4 slowed down body weight gain by decreasing fat mass and prevented the loss of bone mass in old OVX rats. Exendin-4 also enhanced bone strength and prevented the deterioration of trabecular microarchitecture. Moreover, exendin-4 decreased the urinary deoxypyridinoline (DPD)/creatinine ratio and serum C-terminal cross-linked telopeptides of type I collagen (CTX-I) and increased serum alkaline phosphatase (ALP), osteocalcin (OC), and N-terminal propeptide of type 1 procollagen (P1NP) levels, key biochemical markers of bone turnover. Interestingly, gene expression results further showed that exendin-4 not only inhibited bone resorption by increasing the osteoprotegerin (OPG)/receptor activator of NF-κB ligand (RANKL) ratio, but also promoted bone formation by increasing the expression of OC, Col1, Runx2, and ALP, which exhibited dual regulatory effects on bone turnover as compared with previous antiosteoporotic agents. In conclusion, these findings demonstrated for the first time the antiosteoporotic effects of exendin-4 in old OVX rats and that it might be a potential candidate for treatment of aged postmenopausal osteoporosis.

Introduction

Osteoporosis, a prevalent metabolic bone disease, is characterized by low bone mass and deterioration of bone tissue. It mainly affects postmenopausal women and elderly people and results in hip and vertebral fracture. But the pathophysiological characters of osteoporosis with different etiologies (postmenopause or age) are different. Age-related bone loss represents a deficit in bone formation relative to bone resorption whereas in postmenopausal osteoporosis there is strong bone resorption relative to bone formation.[1] Therefore, the coordinated balance between osteoblastic bone formation and osteoclastic bone resorption is a key target of antiosteoporotic agents. However, among therapeutics currently used to treat osteoporosis, the vast majority are antiresorptive agents, which exert their clinical effect by decreasing the rate of bone resorption, thereby preventing further bone loss and achieving a reduction in fracture incidence.[2, 3] A significant unmet need remains for agents that stimulate bone formation or increase bone mass and bone strength (eg, anabolic agents) for osteoblastic therapy.[4] Until now the only anabolic agents approved for osteoporosis are full-length and truncated parathyroid hormone (PTH).[4] Further work is still needed for development of novel anabolic agents.

Gastrointestinal hormones, whose release is stimulated by ingestion, have recently been proved as modulators of bone growth and remodeling.[5, 6] Patients on long-term parenteral feeding displayed a reduced bone mass, which suggested that a deficit in those hormones responding to nutrient ingestion affected bone turnover.[7, 8] Among gastrointestinal hormones, glucagon-like peptide-1 (GLP-1) is an important one that is well known to play crucial roles in blood glucose control and pancreatic islet β cell proliferation.[9, 10] More recently, GLP-1 receptor knockout mice have been shown to have increased bone breakdown, suggesting that the GLP-1 signaling pathway had an antiresorptive effect.[11] Furthermore, the bone formation effect of GLP-1 was also proposed by previous work, in which activation of GLP-1 receptors has been shown to promote bone formation in streptozotocin-induced type 2 diabetic and fructose-induced insulin resistant rats, indicating an anabolic effect.[12] These findings strongly suggested that GLP-1 and its analogues may have a kind of dual regulatory effect on bone turnover as compared with those merely antiresorptive or anabolic agents. However, although the aforementioned findings suggested the dual effects of GLP-1 on bone turnover, those models did not reflect the bone properties induced by estrogen deficiency or aging. The direct effects of exogenous GLP-1 and its analogues on bone turnover in the aged postmenopausal situation remain unclear, the elucidation of which will make GLP-1 signaling a rather promising therapeutic target for osteoporosis, because a GLP-1 receptor agonist, exendin-4, has already been proved for clinical use.[13]

Exendin-4 is a peptide analogue of GLP-1 and has similar properties to GLP-1. It is resistant to degradation by dipeptidyl peptidase-IV and has a much longer plasma half-life than GLP-1,[14] whose half-life is no more than 2 minutes.[15] The prolonged duration of action, prolonged pharmacokinetics, and high apparent in vivo potency of exendin-4 make it more suitable as a potential pharmacological candidate.[14, 16] In the present study, we investigated the effects of continuous administration of exendin-4 on bone remodeling in old ovariectomized (OVX) rats and illustrated an essential role for exendin-4 in regulation of bone metabolism induced by menopause and aging.

Subjects and Methods

Animals and reagents

All experiments and animal care procedures were performed in accordance with regulations approved by the Animal Care and Use Committee of the Fourth Military Medical University (Xi'an, China). Ninety 12-month-old female Sprague-Dawley rats (body weight 300.0 ± 15.0 g) were obtained from the Animal Center of Fourth Military Medical University (Xi'an, China). The rats were housed under controlled ambient conditions (12-hour light/dark cycle, lights on at 6:00 a.m.) and were allowed free access to a standard rodent diet (containing 0.9% calcium and 0.7% phosphate) and distilled water. The acclimatized rats were randomly assigned into a sham-operated group (Sham) and five OVX subgroups: OVX with vehicle (OVX); 17β-estradiol (E2, as positive control, 25 µg/kg/d); and exendin-4 (1, 3, or 10 µg/kg/d). The doses regarding exendin-4 were selected based on previous data in other animal experiments and clinical trials.[17, 18] The treatment began 4 weeks after the surgery and lasted for 16 weeks.

Animals were weighed weekly during the experimental period. After 4 weeks of treatment, 5 rats in each group were euthanized to analyze gene expressions in femurs and vertebra. The remained rats were continuously treated for 12 weeks and then euthanized. One day prior to euthanasia, the animals were fasted in metabolic cages for 24 hours, and urine samples were collected and acidified with 2 mL 1 M HCl. After anesthesia (pentobarbital sodium, 50 mg/kg intraperitoneally [i.p.]) the animals were euthanized to weigh white fat pads and brown adipose tissue, and blood/tissues were harvested. Serum samples were prepared by centrifugation and stored at –70°C before assessment of biochemical parameters. A terminal sample was taken for determination of plasma leptin and lipids concentrations. Leptin levels were assayed by ELISA (Crystal Chem, Downers Grove, IL, USA). Femurs and lumbar vertebrae (L3–L4) were removed for bone analysis, including measurement of bone mineral content (BMC) and bone mineral density (BMD) by dual-energy X-ray absorptiometry (DXA), trabecular bone microarchitecture by micro–computed tomography (µCT), and biomechanical properties by three-point bending and axial compression tests.

Biochemical analysis

Serum calcium (S-Ca) and phosphorus (S-P) concentrations, alkaline phosphatase (S-ALP) activities, urine calcium (U-Ca), urine phosphorus (U-P) and creatinine (Cr) concentrations, and total cholesterol and triglyceride were determined by an automated biochemistry analyzer (Cobas Integra 400 Plus; Roche Diagnostics, Basel, Switzerland). Serum bone formation and resorption markers were measured by commercially available kits. Briefly, serum concentration of osteocalcin (OC) was measured using rat OC radioimmunoassay reagents (Biomedical Technologies, Stoughton, MA, USA). Urinary level of deoxypyridinoline (DPD) cross-links was analyzed using ELISA kits (Quidel Corp., San Diego, CA, USA). Serum levels of bone formation marker N-terminal propeptide of type 1 procollagen (P1NP) and bone resorption marker C-terminal cross-linked telopeptides of type I collagen (CTX-I) were measured using ELISA systems (Immunodiagnostic Systems, Boldon, UK). Leptin levels in serum were assayed by ELISA (Crystal Chem). Urinary excretion of Ca, P, and DPD were expressed as the ratio to Cr excretion. The coefficients of variation (CVs) for intraassay and interassay measurements were 5.4% and 7.6% for OC and DPD, respectively, which are similar to the manufacture's references.

Bone analysis

At the day of testing, the femurs and vertebrae were slowly thawed to room temperature and kept wrapped in the saline-soaked gauzes except during measurements. Each bone and its contralateral pair underwent the analysis on the same day in random order. All bones were analyzed by the same operator.

Dual-energy X-ray absorptiometry

Body composition and bone mineral density were assessed by DXA (Lunar Prodigy Advance DXA, GE Healthcare, Madison, WI, USA) with the small laboratory animals scan mode.[19] Animals were anesthetized with an i.p. injection of pentobarbital sodium prior to scanning. Whole-body DXA assays were conducted at the end of the experiment. Body composition was analyzed by a customized region of interest (ROI). Subsequently, the total right femur and the lumbar vertebrae (L3–L4) were used for determination of bone mineral content and density by DXA. The BMC and BMD were calculated automatically by a software package (encore 2006; GE Healthcare).

µCT

After DXA measurement, bone microarchitecture of the right femur and lumbar vertebra (L4) of rats were evaluated using a desktop µCT system (eXplore Locus SP Pre-Clinical Specimen, GE Healthcare) as described.[20] Whole lumbar vertebrae and femurs were scanned and images were reconstructed to an isotropic voxel size of 12 µm. The distal part of the femur was aligned perpendicularly to the scanning axis in a cylindrical sample holder for a total scanning length of 10 mm. Trabecular bone within the metaphysis was segmented from the cortex using a semiautomated contouring algorithm in the axial plane. The volume of interest (VOI), located 1 mm from the metaphyseal line and the 100 continuous slices above, was selected for data analysis. All 3D image manipulations and analyses were performed within the system software (MicroView, v.2.1, GE healthcare). The following densitometry and architectural parameters were determined: bone volume fraction (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), connectivity density (Conn.D), and structure model index (SMI).

Biomechanical testing

Immediately after the µCT measurements, the length of the femurs were measured using a micrometer and the mean diameter of the diaphysis was determined. The femurs and vertebrae were subjected to three-point bending and axial compression using a servohydraulic materials testing machine (MTS 858 Mini Bionix II; MTS Systems Corp., Minneapolis, MN, USA). Before testing, preload of 0.5 N was applied on the surface of the femur mid-shaft at a speed of 0.1 mm/min as contact force, and then increased to 2 mm/min when testing started. Lumbar vertebrae (L4) of rats were compressed to failure at a displacement rate of 6 mm/min. The load-deformation curve was generated and analyzed by TestStar II software. All force and displacement data were recorded until the specimen was broken. The following biomechanical parameters were determined by the load-deformation curve: maximum load (N; the load at the maximum failure point), stiffness (N/mm; the slope of the linear region), maximum stress (MPa; the maximum load per cross-sectional area) and Young's modulus (GPa; maximum slope of the stress-strain curve).

Histology and histomorphometry

For analysis of bone histology, femurs were fixed in 4% paraformaldehyde (PFA), decalcified in 10% ethylenediaminetetraacetic acid (EDTA, pH 7.0), and then embedded in paraffin. Five-µm-thick sections were stained with toluidine blue to detect osteoblasts and bone marrow adipocyte and with hematoxylin-eosin to visualize osteoclasts. For analysis of bone formation rate (BFR) and mineral apposition rate (MAR), rats were injected intraperitoneally with tetracycline (25 mg/kg) and with calcein (5 mg/kg) 10 days later. After euthanasia, tibias and femurs were isolated, fixed in 80% ethanol, and embedded in methylmethacrylate. Eight-µm-thick, double-labeled sections were evaluated using fluorescence microscopy and photographed to compute MAR and BFR.

RNA extraction and RT-PCR

Femurs and vertebra were dissected from the rats and cleaned with 0.1 M ice-cold PBS, pH 7.2. Connective tissues, muscles, and bone marrow were removed at 4°C before all bones were snap-frozen in liquid nitrogen to make them breakable. Specific primers for the rat ALP, Runx2, Collagen type 1 (Col1), Osteocalcin (OC), Osteoprotegerin (OPG), receptor activator of NF-κB ligand (RANKL), peroxisome proliferator-activated receptor γ (PPARγ), CCAAT/enhancer-binding protein α (c/EBPα), and GAPDH (a housekeeping gene) are shown in Table 1. To collect the mRNA sample, each bone was homogenized in liquid nitrogen with a ceramic grinder. Total RNA was extracted using RNeasy Mini Kit (Qiagen, Shanghai, China) according to the manufacturer's protocol, and was reverse-transcribed using PrimeScript RT reagent Kit with DNA Eraser (Takara, Kyoto, Japan). PCR was performed with the Premix Taq RT-PCR System (Takara) according to the manufacturer's instructions. Reactions were performed in a gradient thermal cycler (BioRad, Hercules, CA, USA) for 28 to 32 cycles. The amplified products were visualized by gel electrophoresis in 1% agarose and stained with 1.0 µg/mL ethidium bromide.

Table 1. The Sequence of Primers Used in This Study
GenesPrimersPrimer sequence (5′–3′)Product length (bp)
ALPForwardACGAGATGCCACCAGAGG256
 ReverseAGTTCAGTGCGGTTCCAG 
Runx2ForwardAGTCCCAACTTCCTGTGCT243
 ReverseGGTGAAACTCTTGCCTCGTC 
COL 1ForwardAACTTTGCTTCCCAGATGTCC334
 ReverseAGCCTCGGTGTCCCTTCA 
OCForwardTGCTCACTCTGCTGACCCTG109
 ReverseTTATTGCCCTCCTGCTTG 
RANKLForwardCGTACCTGCGGACTATCTTCA196
 ReverseGTTGGACACCTGGACGCTAA 
OPGForwardCATCGAAAGCACCCTGTA201
 ReverseCACTCAGCCAATTCGGTAT 
PPARγForwardTGGAGCCTAAGTTTGAGTT126
 ReverseCAATCTGCCTGAGGTCTG 
c/EBPαForwardTCAAGGGCTTGGCTGGTCC132
 ReverseTGTTGCGTTCCCGCCGTAC 
GAPDHForwardTCACTGCCACCCAGAAGA316
 ReverseAAGTCGCAGGAGACAACC 

Culture and differentiation of bone marrow mesenchymal stem cells

Bone marrow mesenchymal stem cells (BMMSCs) were isolated from the femur and tibia of 2-week-old male Sprague-Dawley rats. The cells were purified from the marrow by the Percoll density gradient centrifugation method and were cultured in α-Minimum Essential Medium (α-MEM; Hyclone, Logan, UT, USA) supplemented with 15% fetal bovine serum (FBS) (Gibco BRL, Gaithersburg, MD, USA), 100 U/mL penicillin, 100 µg/mL streptomycin, and 2 mM L-glutamine (Invitrogen, Carlsbad, CA, USA) at 37°C in a humidified atmosphere of 5% CO2. After 3 days, nonadherent hematopoietic cells were discarded and the adherent cells were washed twice with PBS. The culture medium was replenished every 3 days. When the density of cell colonies had reached approximately 90% confluence, the cells were trypsinized (0.25% trypsin) and transferred to fresh flasks at 1:2 ratio. BMMSCs at passage 3 were induced to osteogenic and adipogenic differentiation. Differentiation into osteoblasts was cultured in osteogenesis induction medium (OIM): α-MEM, 10% FBS, 2 mM glutamine, 100 U/mL penicillin, 100 mg/mL streptomycin, 50 mg/L ascorbic acid, 10 nM dexamethasone, and 10 mM β-glycerophosphate. Differentiation into adipocytes was cultured in adipogenesis induction medium (AIM): α-MEM, 2 mM glutamine, 100 U/mL penicillin, 100 mg/mL streptomycin, 10 µM insulin, 1 µM dexamethasone, 0.5 mM 3-isobutyl-1-methylxanthine (IBMX), and 200 µM indomethacin. In long-term differentiations, the medium was changed every 3 days. The experiments were terminated 10 to 21 days after the drug treatment, and then the cells were prepared for the following experiments. The osteogenic and adipogenic potentials were confirmed through gene expression and staining of their corresponding specific markers. Ten days after the adipogenic induction, cells were stained with Oil Red O (Sigma, St. Louis, MO, USA) to identify the lipid droplets. Three weeks after osteogenic induction, cells were stained with Alizarin Red (Sigma) to determine the calcium deposits.

Statistical analysis

The data were expressed as means ± SD and analyzed using SPSS for Windows version 15.0 (SPSS Inc., Chicago, IL, USA). Statistical significance was determined by one-way analysis of variance (ANOVA) followed by a post hoc multiple comparison using Student-Newman-Keuls t test. Differences were considered significant at p < 0.05.

Results

Exendin-4 reduced body weight and fat mass of rats subjected to OVX

As shown in Fig. 1A, the body weights of the OVX and sham rats were not significantly different at the beginning of the surgery. The body weights of the OVX rats became significantly increased at 2 weeks (p < 0.01) after the surgery and were sustained during the whole procedure (p < 0.01). E2 treatment significantly suppressed the OVX-induced body weight gain and decreased it to the level of sham group after 4 weeks treatment. Although exendin-4 treatment (1 µg/kg/d) did not alter body weight gain after OVX (p > 0.05), higher doses of exendin-4 treatments (3 µg/kg/d or 10 µg/kg/d) significantly decreased the body weight (p < 0.05 for 3 µg/kg/d or p < 0.01 for 10 µg/kg/d versus OVX group, respectively).

Figure 1.

Effects of 16 weeks treatment with exendin-4 or E2 on (A) body weight, (B) body composition, and (CE) plasma leptin and lipids levels in OVX rats. The body weight of the animals was recorded weekly during the experimental period. The arrow shows the time when the treatment began. Twelve-month-old female SD rats were randomly assigned into sham operated group (Sham) and five OVX subgroups: OVX treated with vehicle (OVX); OVX treated with 17β-estradiol (E2, 25 µg/kg/d); and OVX treated with exendin-4 (EX) of graded dosages (EX1, 1 µg/kg/d), (EX3, 3 µg/kg/d), and (EX10, 10 µg/kg/d). Values are expressed as means ± SD (n = 10); *p < 0.05, **p < 0.01 compared with OVX as evaluated by one-way ANOVA followed by a Student-Newman-Keuls t test.

Effects of E2 or exendin-4 on body composition were further examined. Exendin-4 significantly reduced body fat mass in a dose-dependent manner in OVX rats (p < 0.01; Fig. 1B). Furthermore, plasma leptin, total cholesterol, and triglyceride levels were also measured in rats 16 weeks after OVX. Leptin levels did not significantly change after exendin-4 treatment (p > 0.05; Fig. 1C). However, total cholesterol and triglyceride levels were found to be dose-dependently decreased in OVX rats treated with exendin-4 (p < 0.05 or p < 0.01; Fig. 1D, E).

Exendin-4 increased BMD of OVX rats

Changes in BMC and BMD of the right femur and lumbar vertebrae (L3–L4) determined by DXA analysis were shown in Fig. 2. OVX significantly lowered BMC and BMD of the right femur and lumbar vertebrae (p < 0.01 versus sham group). E2 markedly increased the BMD of right femur and lumbar vertebrae, which was significantly higher than in the OVX group (p < 0.01; Fig. 2A, C), but did not increase BMC (p > 0.05; Fig. 2B, D). However, 16-week treatment with exendin-4 increased both BMC (10 µg/kg/d; p < 0.05 or p < 0.01 versus OVX group) and BMD (dose-dependent manner; p < 0.05 or p < 0.01 versus OVX group) of the right femur and lumbar vertebrae.

Figure 2.

Effects of 16 weeks treatment with exendin-4 or E2 on (A, C) bone mineral density (BMD) and (B, D) bone mineral content (BMC) in right femur and lumbar vertebrae evaluated by DEXA in OVX rats. Twelve-month-old female SD rats were randomly assigned into sham operated group (Sham) and five OVX subgroups: OVX treated with vehicle (OVX); OVX treated with 17β-estradiol (E2, 25 µg/kg/d); and OVX treated with exendin-4(EX) of graded dosages (EX1, 1 µg/kg/d), (EX3, 3 µg/kg/d), and (EX10, 10 µg/kg/d). Values are expressed as means ± SD (n = 10); *p < 0.05, **p < 0.01 compared with OVX as evaluated by one-way ANOVA followed by a Student-Newman-Keuls t test.

Exendin-4 improved bone trabecular microarchitecture of OVX rats

Representative samples of three-dimensional images of femoral metaphysis and vertebral body are shown in Supplemental Fig. S1, which show differences in trabecular microarchitecture among the various treatment groups. After 16-week treatment with E2 or exendin-4 (3 µg/kg/d), the trabecular microarchitecture was significantly improved. According to the results, the OVX-treated specimen exhibited fewer, thinner, and more broken trabecular, when compared with the sham control, while E2 or exendin-4 treatment healed these trabecular damages induced by OVX. The parameters of the trabecular microarchitecture (BV/TV, Conn.D, Tb.N, and Tb.Th) were all decreased in OVX rats compared with their respective sham controls (Table 2). In contrast, the values of Tb.Sp and SMI in femur and vertebra were significantly increased in response to OVX (p < 0.01 for both versus sham group). Treatment with exendin-4 (3 µg/kg/d) or E2 treatment significantly improved all of the parameters of the trabecular microarchitecture (BV/TV, Tb.N, Tb.Th, Tb.Sp, Conn.D, and SMI) at similar effect (Table 2).

Table 2. µCT Assessment of Trabecular Microarchitecture in OVX Rats After 16-Week Treatment
ParameterShamOVXE2EX3
  1. Values are mean ± SD (n = 10 per group).
  2. OVX = ovariectomized; Sham = sham operated; E2 = 17β-estradiol; EX3 = exendin-4 at dosage of 3 µg/kg/d; BV/TV = bone volume fraction; Tb.N = trabecular number; Tb.Sp = trabecular spacing; Conn.D = connectivity density; SMI = structure model index.
  3. *p < 0.05 and **p < 0.01 compared with OVX as evaluated by one-way ANOVA followed by a Student-Newman-Keuls t test.
Femur    
BV/TV (%)50.30 ± 4.61**17.83 ± 2.1234.52 ± 3.40**44.20 ± 4.52**
Tb.N (1/mm)3.92 ± 0.22**1.81 ± 0.132.56 ± 0.21*3.12 ± 0.18**
Tb.Th (mm)0.15 ± 0.03**0.08 ± 0.010.10 ± 0.01*0.12 ± 0.02**
Tb.Sp (mm)0.13 ± 0.01**0.35 ± 0.040.23 ± 0.02**0.20 ± 0.03**
Conn.D (1/mm3)55.08 ± 5.86**23.76 ± 3.7148.25 ± 4.96**51.03 ± 5.36**
SMI (1)0.46 ± 0.07**1.08 ± 0.060.44 ± 0.04**0.45 ± 0.05**
Vertebra    
BV/TV (%)36.52 ± 1.82**25.22 ± 2.5233.35 ± 1.84**37.42 ± 2.05**
Tb.N (1/mm)4.36 ± 0.19**3.55 ± 0.163.96 ± 0.18*4.48 ± 0.22**
Tb.Th (mm)0.11 ± 0.01**0.08 ± 0.010.09 ± 0.020.10 ± 0.01*
Tb.Sp (mm)0.20 ± 0.01**0.25 ± 0.020.19 ± 0.02*0.18 ± 0.03**
Conn.D (1/mm3)70.11 ± 6.05**89.26 ± 8.1385.20 ± 7.0183.35 ± 9.06*
SMI (1)0.28 ± 0.08**0.61 ± 0.120.19 ± 0.05**0.18 ± 0.04**

Exendin-4 increased bone strength of OVX rats

The biomechanical competence of the femoral diaphysis and the lumbar vertebral bodies was tested after 16 weeks treatment with exendin-4 or E2 by using an ex vivo three-point bending test and axial compression test. OVX resulted in a significant decrease in Young's modulus and ultimate stress in femur and lumbar vertebra compared with sham group (p < 0.01, Fig. 3B, D), whereas 16 weeks treatments with either E2 or three doses of exendin-4 significantly increased the Young's modulus (p < 0.05 or p < 0.01, Fig. 3B, D). Furthermore, higher doses of exendin-4 (3 µg/kg/d or 10 µg/kg/d) and E2 treatment reversed the decrease of ultimate stress induced by OVX (p < 0.05 or p < 0.01, Fig. 3B, D). OVX also resulted in decrease in maximum load and stiffness, and the treatment with exendin-4 or E2 rescued the processes (p < 0.05 or p < 0.01, Fig. 3A, C).

Figure 3.

Effects of 16 weeks treatment with exendin-4 or E2 on biomechanical properties in femoral diaphysis and lumbar vertebrae evaluated by three-point bending test and axial compression test in OVX rats. (A, B) Femoral and (C, D) vertebral maximum load (N), stiffness (N/mm; the slope of the linear region), maximum stress (MPa) and Young's modulus (GPa). Twelve-month-old female SD rats were randomly assigned into sham operated group (Sham) and five OVX subgroups: OVX treated with vehicle (OVX); OVX treated with 17β-estradiol (E2, 25 µg/kg/d); OVX treated with exendin-4(EX) of graded dosages (EX1, 1 µg/kg/d), (EX3, 3 µg/kg/d), and (EX10, 10 µg/kg/d). Values are expressed as means ± SD (n = 10); *p < 0.05, **p < 0.01 compared with OVX as evaluated by one-way ANOVA followed by a Student-Newman-Keuls t test.

Exendin-4 ameliorated biochemical parameters of OVX rats

Table 3 shows the effects of exendin-4 or E2 treatment on serum and urine biochemical parameters. OVX did not alter the S-Ca and S-P levels but increased the activity of ALP (p < 0.01). Meanwhile, U-Ca and U-P levels were significantly increased in response to OVX (p < 0.01). E2 treatment reversed the increase of ALP activity induced by OVX (p < 0.01), but exendin-4 did not decrease ALP activity. Interestingly, the highest dose (10 µg/kg/d) of exendin-4 significantly increased ALP activity (p < 0.01). Furthermore, exendin-4 treatment significantly reduced the increase of U-Ca/Cr and U-P/Cr levels induced by OVX in a dose-dependent manner.

Table 3. Biochemical Parameters in OVX Rats After 16-Week Treatment
ParametersShamOVXE2EX1EX3EX10
  1. Values are mean ± SD (n = 10 per group).
  2. OVX = ovariectomized; Sham = sham operated; E2 = 17β-estradiol; EX1 = exendin-4 at dosage of 1 µg/kg/d; EX3 = exendin-4 at dosage of 3 µg/kg/d; EX10 = exendin-4 at dosage of 10 µg/kg/d; S-Ca = serum calcium; S-P, serum phosphorous, S-ALP = serum alkaline phosphatase; U-Ca = urinary calcium; U-P = urinary phosphorous; U-Cr = urinary creatinine; U-Ca/Cr = urinary calcium/creatinine ratio; U-P/Cr = urinary phosphorous/creatinine ratio.
  3. *p < 0.05, **p < 0.01 compared with OVX as evaluated by one-way ANOVA followed by a Student-Newman-Keuls t test.
S-Ca (mM)2.74 ± 0.432.58 ± 0.312.66 ± 0.442.78 ± 0.502.54 ± 0.452.56 ± 0.30
S-P (mM)1.50 ± 0.191.61 ± 0.621.70 ± 0.301.69 ± 0.191.78 ± 0.241.73 ± 0.15
S-ALP (U/L)78.06 ± 9.27**104.50 ± 10.3283.13 ± 9.26**104.20 ± 12.04109.27 ± 9.58127.79 ± 13.78**
U-Ca (mM)1.64 ± 0.79**3.18 ± 1.192.57 ± 0.77*2.52 ± 0.85*2.28 ± 0.79**1.53 ± 0.82**
U-P (mM)24.78 ± 10.49**31.28 ± 12.0626.11 ± 15.56*28.34 ± 9.69*25.78 ± 12.67*21.81 ± 9.39**
U-Cr (mM)9.66 ± 3.8210.39 ± 3.9710.62 ± 3.0210.35 ± 4.0510.05 ± 3.879.85 ± 2.64
U-Ca/Cr0.16 ± 0.06**0.35 ± 0.110.26 ± 0.07**0.26 ± 0.11**0.24 ± 0.10**0.17 ± 0.05**
U-P/Cr2.55 ± 0.78**3.06 ± 0.792.56 ± 0.45**2.85 ± 0.52*2.56 ± 0.88**2.16 ± 0.40**

Twenty weeks after OVX, analysis of bone resorption and formation markers was next carried out after various treatments with E2 and exendin-4. Urinary DPD/cyclic recombinase (CRE) and serum CTX-I, OC, and P1NP levels were higher in the OVX group compared with those in the sham group (Fig. 4). Both exendin-4 and E2 treatments significantly suppressed the urinary DPD/creatinine ratio and serum CTX-I induced by OVX (p < 0.05 or p < 0.01, Fig. 4A, B). Similarly, the increase of serum OC level induced by OVX was prevented by treatment with E2 (p < 0.05, Fig. 4C). However, different from E2 treatment, high dose of exendin-4 (10 µg/kg/d) significantly increased the OC level compared to the OVX group (p < 0.01). Although the increase in serum P1NP level induced by OVX appeared to be slightly decreased after treatment with E2, it was not statistically significant (Fig. 4D). Further significant increases in P1NP levels were observed in OVX rats receiving higher dose of exendin-4 (3 µg/kg/d, 10 µg/kg/d) treatment (p < 0.01).

Figure 4.

Effects of 16 weeks treatment with exendin-4 or E2 on bone turnover biomarkers in OVX rats. (A) urinary DPD/creatinine ratio and (B) serum CTX-I, (C) OC, and (D) P1NP. Twelve-month-old female SD rats were randomly assigned into sham operated group (Sham) and five OVX subgroups: OVX treated with vehicle (OVX); OVX treated with 17β-estradiol (E2, 25 µg/kg/d); and OVX treated with exendin-4(EX) of graded dosages (EX1, 1 µg/kg/d), (EX3, 3 µg/kg/d), and (EX10, 10 µg/kg/d). Values are expressed as means ± SD (n = 10); *p < 0.05, **p < 0.01 compared with OVX as evaluated by one-way ANOVA followed by a Student-Newman-Keuls t test.

Exendin-4 enhanced bone formation and inhibited bone resorption

The bone histomorphometry were analyzed after 16 weeks treatment with exendin-4 or E2. Osteoblast and osteoclast formation on the surface of trabecular bone were examined by bone histological analysis. OVX resulted in a significant increase in MAR, BFR, and osteoblast and osteoclast numbers compared with the sham group (p < 0.05 or p < 0.01, Fig. 5AC), whereas 16 weeks of treatment with E2 significantly decreased the number of osteoclasts (p < 0.01, Fig. 5C). Although MAR, BFR, and the number of osteoblasts appeared to be slightly decreased after treatment with E2, it was not statistically significant. However, different from E2 treatment, exendin-4 further increased MAR and BFR compared to the OVX group (p < 0.05, Fig. 5A). The bone recovery promoted by exendin-4 was associated with an increase in the number of osteoblasts and a decrease in the number of osteoclasts in OVX rats (p < 0.01, Fig. 5B, C). Together, these findings suggested that prevention of bone loss by exendin-4 is achieved through stimulation of osteoblast formation and inhibition of osteoclast formation.

Figure 5.

Effects of 16 weeks treatment with exendin-4 or E2 on osteoblast formation and osteoclast formation in OVX rats. (A) Mineral apposition rate (MAR) and bone formation rate (BFR) in the femurs were determined by double-labeling with tetracycline and calcein. Scale bar, 10 µm. (B) Femoral sections from rats were stained with toluidine blue to detect osteoblasts and the numbers of osteoblasts (N.Ob) per millimeter of trabecular bone surface (BS) were counted. Scale bar, 100 µm. (C) Femoral sections from rats were stained with hematoxylin and eosin to visualize osteoclasts and the numbers of osteoclasts (N.Oc) per millimeter of trabecular bone surface (BS) were counted. Scale bar, 100 µm. Values are expressed as means ± SD (n = 10); *p < 0.05, **p < 0.01 compared with OVX as evaluated by one-way ANOVA followed by a Student-Newman-Keuls t test.

Exendin-4 affected gene expressions related to bone formation and bone resorption

Figure 6 shows the expression of Runx2, ALP, Col1, and OC mRNA after 4 weeks of treatment with exendin-4 (3 µg/kg/d) or E2. Compared with the sham group, the expression of Runx2, ALP, Col1, and OC mRNA in the OVX control group significantly increased (p < 0.05). E2 significantly reduced Col1 and OC expressions, late markers of bone formation, to 14% and 20% of the OVX group values (p < 0.05), but did not reduce the expression of Runx2 and ALP, early markers of bone formation. Meanwhile, exendin-4 (3 µg/kg/d) markedly increased the expressions of Runx2, ALP, Col1, and OC mRNA to 19%, 26%, 22%, and 20% of the OVX group values, respectively.

Figure 6.

Runx2, ALP, Col1, and OC mRNA expressions in femur of OVX rats treated with E2 or exendin-4 (EX3, 3 µg/kg/d) for 4 weeks. Total RNA was isolated and RT-PCR was performed to determine mRNA expressions, which were normalized to that of GAPDH. Values are expressed as means ± SD (n = 5); *p < 0.05, **p < 0.01 compared with OVX as evaluated by one-way ANOVA followed by a Student-Newman-Keuls t test.

The expressions of OPG and RANKL mRNA in OVX rats after 4 weeks of treatment with exendin-4 (3 µg/kg/d) or E2 are shown in Fig. 7. In OVX rats, OPG mRNA expression was significantly decreased compared with sham group, whereas RANKL mRNA expression was significantly increased. E2 significantly decreased RANKL mRNA expression, but did not significantly affect OPG mRNA expression. However, exendin-4 not only significantly decreased RANKL mRNA expression but also increased OPG mRNA expression. Moreover, both E2 and exendin-4 significantly increased the OPG/RANKL ratio in OVX rat (p < 0.05 or p < 0.01, respectively). The results suggest that exendin-4 might affect OPG and RANKL expression and the ratio of OPG/RANKL, and thereby inhibit the process of osteoclastogenesis.

Figure 7.

OPG and RANKL mRNA expressions in femur in OVX rats treated with E2 or exendin-4 (EX3, 3 µg/kg/d) for 4 weeks. Total RNA was isolated and RT-PCR was performed to determine OPG and RANKL mRNA expressions, which were normalized to that of GAPDH. Values are expressed as means ± SD (n = 5); *p < 0.05, **p < 0.01 compared with OVX as evaluated by one-way ANOVA followed by a Student-Newman-Keuls t test.

Discussion

The entero-bone endocrine axis was proposed as a mediator of postprandial bone turnover.[21] GLP-1 and GLP-2 are important gastrointestinal hormones released from the small intestine immediately after nutrient ingestion, and GLP-2 is known to be involved in the postprandial regulation of bone turnover.[8, 22] Henriksen and colleagues[8, 23-25] found that administration of GLP-2 to human subjects inhibited bone resorption and increased bone mass in a dose-dependent manner. However, the effect of GLP-1 and its analogues on bone turnover in menopause and aging situation is still unknown. Previous investigations showed that OVX rats were an ideal animal model[26] for postmenopausal women's bone loss in old age. So we investigated the effects of continuous administration of exendin-4, an agonist of the GLP-1 receptor, on bone remodeling in this ideal animal model. We found for the first time that exenidn-4 had direct dual antiosteoporotic effects in aging and estrogen deficiency-induced osteopenia.

Osteoporosis is a condition in which loss of bone mass leads to fragility fractures. BMD is considered as an indicator of osteoporosis and fracture risk. Decreased BMD and deterioration of trabecular bone structure will result in reduced bone strength and increased osteoporotic fracture incidence.[27] In the present study, bone densitometry measurements showed that exendin-4 treatment dramatically increased BMD in a dose-dependent manner. Based on these data, we chose the dose of 3.0 µg for subsequent µCT, bone histology, and gene expression experiments. The structural analysis of trabecular bone by µCT indicated that exendin-4 protected against the deterioration of trabecular bone structure. In addition, structural changes of bone, such as adding or redistributing mass, affect its biomechanical properties. The extrinsic biomechanical properties (ultimate force and stiffness) and intrinsic biomechanical properties (Young's modulus and ultimate stress) are important parameters for description of bone state.[28] But the actual effect of treatment on biomechanical competence can only be fully evaluated if the extrinsic biomechanical parameters are corrected for changes in geometric properties, yielding intrinsic biomechanical parameters.[29] Our data showed that exendin-4 remarkably prevented loss of bone strength after OVX and produced significant dose-related improvements in all bone biomechanical properties, including ultimate force and stiffness, Young's modulus, and ultimate stress, suggesting an ideal treatment for osteoporosis according to previous opinion.[28] Moreover, the improvement of the biomechanical properties may be also due to the trabecular bone structure optimization observed in our study.

After confirming the direct effect of exendin-4 on the density, structure, and strength of bone, we next detected the effect of exendin-4 on biochemical markers and bone histology to further explore the possible underlying mechanisms of the antiosteoporotic effect. Bone is a metabolically active tissue that turns over constantly. In the aged estrogen-deficient situation, more bone is resorbed by the osteoclasts on the trabecular bone surface, which finally results in net bone loss.[30] The degree of coupling of bone formation and resorption processes can be generally reflected by calcium and phosphate balance.[31] Treatment with exendin-4 for 16 weeks prevented the osteopenia and lowered the increased rates of U-Ca and U-P excretion induced by OVX. These findings strongly suggested that exendin-4 prevented the osteopenia by modulating calcium and phosphorus balance. Furthermore, serum ALP, OC, and P1NP levels are important biomarkers for assessment of bone formation rate, whereas urinary DPD level and serum CTX-I are two major biomarkers for assessment of bone resorption rate. The present data showed that exendin-4 decreased urinary DPD/creatinine ratio and serum CTX-I and increased serum ALP, OC and P1NP levels, which evidenced the bone turnover-modulating effect of exendin-4, possibly through both promoting bone formation and suppressing bone resorption. Furthermore, we examined osteoblast and osteoclast formation on the surface of trabecular bone by bone histological analysis. Our data showed that exendin-4 further increased MAR and BFR associated with an increase in the number of osteoblasts and a decrease in the number of osteoclasts in OVX rats. These findings further demonstrated that prevention of bone loss by exendin-4 is achieved through stimulation of osteoblast formation and inhibition of osteoclast formation. This anabolic and antiresorptive effect of GLP-1 on bone has previously been described in rodents.[11, 12, 32, 33] However, data in human subjects indicated that administration of GLP-1 did not result in a reduction of the serum CTX level, suggesting that GLP-1 may not be important for reduction in bone resorption.[8] This difference between human and rodents is very probably because of the differences on the expression levels of GLP-1R.[11, 34, 35]

We then further confirmed this modulating effect of exendin-4 in gene expression level related to osteoblastic bone formation and osteoclastic bone resorption. Runx2 is a key transcription factor in osteoblast differentiation and bone formation.[36] Moreover, ALP and Col1 are early markers, and OC is a late marker in bone formation.[37] Our results displayed that exendin-4 significantly increased Runx2, ALP, Col1, and OC mRNA levels, suggesting its effects on the various stages of bone formation. Osteoblast-osteoclast coordination is critical in the maintenance of skeletal integrity. The modulation of osteoclastogenesis by immature cells of the osteoblastic lineage is mediated by RANKL, activator of NFκB receptor (RANK), and OPG. OPG is a decoy receptor that inhibits RANKL activation of osteoclastogenesis, thereby decreasing bone resorption.[38] The ratio of OPG/RANKL expression is believed to be a key determinant of osteoclastogenic activity. We found that exendin-4 increased OPG mRNA expression and reduced RANKL mRNA expression, suggesting that exendin-4 inhibits osteoclast differentiation by increasing the OPG/RANKL ratio.

Fat is a key determinant of bone loss with aging[39] and fatty deposits also contribute to weight increase induced by OVX.[40] Previous studies pointed to increased fat mass as an important factor contributing to bone deterioration in obese subjects.[41] Furthermore, a correlation between obesity and fractures, probably a consequence of an altered bone mass or bone fragility, has also been reported in aged postmenopausal women.[42] The fat mass and fat percentage data showed a loss of 19 g (13%) of total fat depots after administration of exendin-4 in the present study, which may also contribute to maintain bone mass to some extent. We also found that exendin-4 significantly decreased the number of bone marrow adipocytes in OVX rats (Supplemental Fig. S2). In addition, the significant reduction of total cholesterol and triglyceride levels in OVX rats was also consistent with the fat loss data, and the antiosteoporotic effects of exendin-4 might also be associated with reduced hyperlipidemia. Previous data showed that hyperlipidemia promoted osteoclastogenesis and bone resorption,[43] and hypercholesterolemia appears to contribute to the pathogenesis of osteoporosis in postmenopausal women, which might justify the use of cholesterol-lowering statins as putative therapies in this situation.[44, 45] Because GLP-1 and exendin-4 were reported to have beneficial effects in reducing cholesterol and triglycerides in diabetic subjects,[46] they are also likely to be promising agents to preventing osteopenia.

Although GLP-1 receptor agonist exendin-4 could protect against osteoporosis, GLP-1 receptors were reported not to be present in either osteoblasts or osteoclasts.[47] Therefore, the mechanisms underlying the GLP-1–mediated modulation of bone turnover need to be further explored in detail. It is well acknowledged that both osteoblasts and adipocytes arise from BMMSCs.[48] In humans, it has been suggested that aging lowers the levels of osteoblast differentiation and enhances adipogenesis.[49] Therefore, the differentiation of BMMSCs to osteoblasts or adipocytes may be a crucial process in bone remodeling. GLP-1 receptors are found to be expressed on BMMSCs but differentially expressed on adipocyte and osteoblast, which stimulated proliferation of BMMSCs and inhibited differentiation to adipocytes in human.[50] In accordance with these findings, we found that the osteogenic action of exendin-4 in OVX rats was also a consequence of the differentiation of BMMSCs through promoting osteogenesis while suppressing adipogenesis (Supplemental Fig. S3), strongly indicating that BMMSCs may be a key targets of GLP-1 and its analogues. It should be mentioned that GLP-1 was suggested to have different bone turnover effects in rodents compared to humans,[8, 11] the reason for this was mainly attributed to the differential expression of GLP-1 receptors between species.[11, 34, 35] However, because GLP-1 receptors were reported to be expressed on human BMMSCs,[50] GLP-1 and its analogues are very likely to exhibit similar effects in humans. Furthermore, exendin-4 did not cause hypoglycemia and hyperinsulinemia in OVX rats (Supplemental Fig. S4). Therefore, although insulin and the IGF-1 axis is important for bone, there should be certain mechanisms other than activation of the IGF-1 axis that mediate the bone formation effect of exendin-4. Further work is still needed in the future.

In conclusion, this study demonstrates for the first time the protective effects of exendin-4 directly on OVX-induced osteoporosis in old rats. The mechanism of these effects may be a result of both promoting bone formation and suppressing bone resorption. These findings raise the possibility for clinical use of GLP-1 receptor agonists in curing aged postmenopausal osteoporosis. Furthermore, given the recent observations of reduced bone density and increased fracture rates in diabetic subjects treated with thiazolidinediones,[51] the combined use of GLP-1 receptor agonists and thiazolidinediones may have clinical significance. Studies directed at understanding the antiosteoporotic mechanisms of therapies that activate GLP-1 receptor signaling seem greatly warranted.

Disclosures

All authors state that they have no conflicts of interest.

Acknowledgments

This work has been supported by Chinese National Key New Drug Creation and Development Program (2009ZX09103-663) and two grants from National Natural Science Foundation of China (81100626, 81201515).

Authors' roles: Study design: XL and XM. Study conduct: XM, JM, MJ, JH, and YZ. Data collection: XM, JM, MJ, and LB. Data analysis: XM, JM, MJ, LB, YW, and JH. Data interpretation: XL, GH, XM, JM, and MJ. Drafting manuscript: XM and GH. Revising manuscript content: XL, XM, JM, and MJ. Final manuscript approval: XM, JM, MJ, and XL. XL and GH take responsibility for the integrity of the data analysis.

Ancillary