Negative effects of semaglutide on bone in obese mice

GLP1 receptor agonists, such as semaglutide, are used in the management of obesity. While causing weight loss, these drugs reduce both fat and lean mass. The impact of GLP1 receptor agonists on bone health has not been explored. Femur mass, stiffness, and strength are lower in obese male C57BL/6J mice treated with semaglutide, and these effects are more pronounced than in mice that are calorically restricted to the same extent. Decrements in bone are paralleled by lower circulating levels of markers of osteoblast activity. In an additional cohort, femur weight, strength, and stiffness are the same 6 weeks following the discontinuation of semaglutide compared to vehicle. Our findings provide evidence that bone mass and strength decrease with semaglutide and that this effect is dependent upon continued exposure to the drug. Clinical investigations are needed to determine if GLP1 receptor agonists have similar impacts on bone in patient populations.

Introduction

The glucagon-like peptide-1 receptor agonist (GLP1ra) semaglutide has emerged as an efficacious treatment for obesity, with recent studies demonstrating 15%–20% reductions in body weight in individuals with obesity.1,2 The effects of semaglutide on weight loss are primarily driven by increases in satiety and reductions in food intake.3,4 Interventions that lead to reductions in food intake and, subsequently, weight, such as caloric restriction, cause reductions in not only adipose tissue but also lean mass.5 Given this, concerns have been raised about the loss of muscle mass during semaglutide treatment. In this regard, we, and others, have demonstrated that semaglutide treatment reduces muscle mass and strength in mice with diet-induced obesity.6,7,8,9 GLP1ra treatment does not appear to result in a disproportionate loss of lean and skeletal muscle mass, as mice pair-fed the same amount of calories as those treated with semaglutide lost similar amounts of lean tissue.7,9 The effects of semaglutide and pair feeding on skeletal muscle reversed when treatment was discontinued with lean mass, skeletal muscle mass, and grip strength being indistinguishable from previously vehicle-treated mice 6 weeks after treatment was stopped.7

A somewhat underappreciated effect of weight loss is the impact on bone density and strength. In both obese and non-obese mice, caloric restriction reduces bone density and suppresses bone formation.10 A 10% reduction in body weight results in approximately a 1%–2% reduction in bone mineral density with a greater impact of weight loss in lean individuals compared with overweight individuals.11 While semaglutide causes marked weight loss, there is surprisingly little data exploring the effects of semaglutide on bone health. One clinical study found increased circulating C-linked telopeptide of collagen (CTX1), a marker of bone breakdown, in individuals with increased fracture risk treated with semaglutide.12 In contrast, preclinical studies using GLP1ra’s, such as liraglutide and exendin-4, have shown positive benefits on bone turnover in mouse models of osteoporosis.13,14

In the current report, we sought to further understand the impact of semaglutide on bone. We hypothesized that (1) semaglutide treatment would result in reductions in bone strength, (2) that this effect would be recapitulated by restricting caloric intake, and (3) that decrements in bone health would require continued treatment with semaglutide. To do so, we assessed bone biomechanical properties in femurs from male C57BL/6J mice with diet-induced obesity immediately following 4 weeks of either semaglutide treatment or caloric restriction (experiment 1). In an additional cohort (experiment 2), obese mice were treated with semaglutide or were calorically restricted for 4 weeks, and bone was assessed 6 weeks following the discontinuation of treatment (Figure 1).

Mice were fed a high-fat diet (45% kcal fat, 35% kcal carbohydrate, 20% kcal protein, Research diets #D12451) for 12 weeks to induce obesity. Mice were then weight-matched and assigned to vehicle, semaglutide and pair-feeding groups (n = 20 per group). Mice received daily subcutaneous semaglutide or vehicle injections for 4 weeks. A ramp protocol was used for semaglutide (MedChemExpress #HY-114118) dosing with mice receiving a 3 nmol/kg body weight dose on day 1, 6 nmol/kg dose on day 2 and 10 nmol/kg on day 3 for the remainder of the treatment period. The daily caloric intake of pair-fed mice was matched to that of the semaglutide group. For the semaglutide treatment discontinuation experiments, mice underwent the same high-fat diet and treatment periods as described above, after which semaglutide and pair-feeding were discontinued and mice given ad libitum access to high fat diet for 6 weeks (n = 10 per group).

Mice were anesthetized with sodium pentobarbital (5 mg/100 g bw) and were killed by exsanguination. Cardiac blood was collected and allowed to clot at room temperature then put on ice until further processing. Serum was collected by centrifuging blood samples at 5000 g for 10 min and stored at −80°C until further analysis. Both femurs were dissected and remaining muscle and connective tissue were carefully removed. Femurs were then left to dry at room temperature before being weighed. Data is presented as the average weight of both femurs. For bone strength testing, dissected femurs were wrapped in gauze soaked in phosphate buffered saline (PBS) and stored frozen at −20°C. For at least 12 h prior to testing, samples were thawed by hydrating in PBS at 3°C–4°C. On the day of testing, femurs were cleaned of any remaining soft tissue. Samples were tested to failure in three-point bending using a mechanical testing system (Mach-1, Biomomentum, Quebec, Canada) with a 25,000 gram-force load cell, an 8-mm span, and a displacement rate of 0.1 mm/s. The load was applied to the posterior surface of the femoral midshaft. Load and cross-displacement were recorded during testing. Maximal load (N) and stiffness (N/mm) were measured and reported according to Jepsen et al. (2015).16

All assays were analyzed using a Versamax Tunable Microplate Reader and SoftMax Pro Software (Molecular Devices). Serum Bone alkaline phosphatase (Biomatik #EKF58147), C-linked telopeptide of collagen (Novus Biologicals #NBP3-11802), Osteocalcin (Novus Biologicals #NBP2-68151), and procollagen type 1 N-propeptide (Novus Biologicals #NBP2-76466) were measured using commercially available enzyme-linked immunosorbent assay (ELISA) kits.