15-PGDH inhibition promotes muscle repair and strength recovery during GLP-1 receptor agonist–induced weight loss

Significance

Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are revolutionary treatments for obesity and diabetes, yet growing evidence indicates that the resulting weight loss includes loss of lean mass, including skeletal muscle, a major organ essential for physical activity and whole-body metabolic health. We identify 15-hydroxyprostaglandin dehydrogenase (15-PGDH), a prostaglandin-degrading “gerozyme,” as a potent, druggable target and show that its inhibition enhances muscle stem cell function and regenerative myofiber growth during semaglutide-induced weight loss. Our data suggest that PGDHi and GLP-1 RA synergize to promote effective muscle repair and functional recovery of force.

Abstract

Glucagon-like peptide-1 receptor agonists, including long-acting semaglutide, are transformative anti-obesity therapies. However, emerging evidence indicates that weight loss may come at the expense of skeletal muscle mass, a tissue essential for mobility, metabolic regulation, and overall health. Here, we show that inhibition of the gerozyme 15-hydroxyprostaglandin dehydrogenase (15-PGDH), a prostaglandin-degrading enzyme that increases with injury and aging, improves muscle repair and strength recovery in the presence of semaglutide. In a high fat diet-induced mouse model of obesity, semaglutide alone caused significant loss of muscle mass, while preserving contractile function. Following injury, obese mice exhibited pathological calcifications previously reported for the heritable myopathy, Duchenne Muscular Dystrophy. Semaglutide had both beneficial and deleterious effects, reducing calcific remodeling, but causing reduced regenerated myofiber sizes. This impaired regenerative myofiber growth in semaglutide-treated mice was surmounted by cotreatment with a 15-PGDH inhibitor (PGDHi), which stimulated muscle stem cell function and myofiber growth, leading to enhanced strength. Importantly, PGDHi synergizes with semaglutide to boost postinjury muscle quality and muscle force without compromising weight loss.

Obesity is associated with reduced life expectancy and quality of life (1–3). It is a major health challenge and contributor to the rising incidence of type 2 diabetes, cardiovascular disease, nonalcoholic fatty liver disease, and several forms of cancer. Despite sustained public-health efforts, the global prevalence of obesity has more than doubled since the 1980s, rising from approximately 6% to over 14% in adults, and it is escalating worldwide, underscoring the urgent need for effective and scalable therapeutic strategies (4). While lifestyle interventions are fundamental, pharmacological therapies that achieve sustained weight loss are critical for managing obesity and its numerous comorbidities. Among the most effective agents currently available are glucagon-like peptide-1 receptor agonists (GLP-1 RAs), such as semaglutide, which mimic endogenous incretin signaling to regulate glucose metabolism (5). These drugs act in the central nervous system to reduce appetite, and peripherally to delay gastric emptying and improve insulin secretion and sensitivity (6).

Large randomized clinical trials consistently demonstrate that GLP-1 RAs induce significant and durable reductions in body weight (7–9). For example, 68 wk of weekly semaglutide injections resulted in an average 16% body weight reduction in obese adults (7), a degree of efficacy that rivals bariatric surgery (10). Similar weight loss benefits have been reported in obese adolescents, evidence of the clinical relevance of these agents to younger populations (11). The success of GLP-1 RAs has led to their widespread adoption for multiple indications. In addition to their metabolic effects, these agents offer cardioprotective (12) and renal benefits (13), which explains their growing appeal in treating obesity as a systemic disease.

A major caveat is that the weight loss associated with semaglutide results not just from the loss of fat mass, but also from the loss of lean mass including muscle (14, 15). Large-scale randomized clinical trials have shown that GLP-1 RA–induced weight loss impacts both fat and lean tissue, with adults losing over 5 kg of lean mass after 68 wk of treatment (7). In preclinical models, semaglutide has been associated with significant muscle atrophy, further reinforcing concerns about its impact on skeletal muscle mass and integrity (16, 17). This is clinically relevant, as the loss of skeletal muscle can compromise strength, metabolic health, and quality of life. Preserving muscle during weight loss has long been a therapeutic objective.

Relatively little is known about how semaglutide affects muscle function, particularly in the context of tissue remodeling as occurs after injury, surgery, and exercise. These processes rely on cycles of minor damage and subsequent regeneration to promote gains in muscle mass and strength (18). An understanding of the impact of GLP-1 RA treatments on regenerative capacity in obese individuals is particularly important, as it may dictate their response to physical activity, rehabilitation postsurgery, and overall effectiveness of exercise-based interventions.

We previously identified the enzyme 15-hydroxyprostaglandin dehydrogenase (15-PGDH) as a gerozyme, a molecular driver of muscle wasting in aging (19–21). 15-PGDH degrades prostaglandin E2 (PGE2), a lipid metabolite that has pleiotropic beneficial effects on skeletal muscle. PGE2 promotes muscle stem cell proliferation and tissue regeneration, increases myofiber mitochondrial number and function, and restores neuromuscular connectivity (19–22). We showed that pharmacological or genetic inhibition of 15-PGDH (PGDHi) enhances PGE2 signaling and increases muscle mass and strength, especially in aged mice (19, 21).

Here we test the effects of semaglutide, alone and in combination with PGDHi, on skeletal muscle size, strength, and regeneration in young adult obese mice. Our findings reveal that semaglutide reduces muscle mass but preserves force under steady-state conditions. During muscle regeneration postinjury, semaglutide has both beneficial and deleterious effects, reducing calcifications seen in obese muscles during repair, but causing reduced regenerated myofiber sizes. In semaglutide-treated mice, PGDHi enhances MuSC proliferation, supports muscle repair, and restores myofiber size without compromising semaglutide’s metabolic benefits. Our results identify 15-PGDH as a critical node in muscle adaptation to metabolic therapies. Further, they support PGDHi as a therapeutic strategy to promote muscle repair and strength in conjunction with the use of GLP-1 RAs to induce weight loss.

Results

Semaglutide Triggers a Loss of Muscle Mass but Preserves Muscle Function.

Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) result in significant body weight loss, however, a portion of this weight loss results from a reduction in lean body mass, particularly skeletal muscle (17, 23, 24), which is a concern given the critical role of skeletal muscle in systemic metabolism and movement. Thus, there is a major unmet need for drugs that preserve muscle in conjunction with GLP-1 RA interventions. While numerous GLP-1 RA studies have assessed loss of muscle mass, few have measured muscle function in terms of strength. We reasoned that treatment with an inhibitor of 15-hydroxyprostaglandin dehydrogenase (15-PGDH) could overcome the loss of muscle mass and function observed with GLP-1 RAs. 15-PGDH degrades prostaglandin E2 (PGE2) and is a gerozyme that increases with injury and aging (19, 21, 25). To test if an inhibitor of 15-PGDH (PGDHi) could preserve muscle mass and function during semaglutide-induced weight loss, we compared the effects of semaglutide, a GLP-1 RA approved for clinical use under the brand names Ozempic or Wegovy, alone and in combination with PGDHi on body composition and skeletal muscle mass and function in a mouse model of high fat diet-induced obesity (HFD).

To induce obesity, young adult (8 wk of age) male C57BL/6 J mice were fed a high-fat diet (HFD; 60% kcal from fat) for 12 wk. Animals were then randomized by body weight and baseline plantar flexion maximal tetanic force into four treatment groups: vehicle control (Veh), PGDHi, semaglutide, and combined (PGDHi + semaglutide) (Fig. 1A). Body weight was recorded daily throughout the intervention, and maximal strength was assessed in vivo as plantar flexion torque at baseline, 3 wk, and 5 wk after initiation of treatment. In agreement with reports by others (17, 26), semaglutide treatment led to a sustained and significant reduction in body weight, reaching ~25% total body weight loss by week five (Fig. 1 B and C and SI Appendix, Fig. S1 A and B). Importantly, coadministration with PGDHi did not significantly attenuate the weight-lowering effect of semaglutide, indicating that PGDHi alone does not induce weight loss or interfere with the appetite-suppressing effect of semaglutide. Semaglutide also significantly reduced heart weight (SI Appendix, Fig. S1B) and improved glucose tolerance in agreement with previous reports (Fig. 1D and SI Appendix, Fig. S1D) (27–29). Notably, PGDHi alone had no detectable effect on either outcome. To further assess changes in adiposity, we isolated and weighed major adipose depots: inguinal (iWAT), epididymal (eWAT), and brown adipose tissue (BAT). Both semaglutide and the PGDHi + semaglutide combination significantly reduced iWAT, eWAT, and BAT weights compared to PGDHi alone or vehicle controls (Fig. 1D), in agreement with histological analyses (Fig. 1E). These data confirm that semaglutide effectively reduces fat mass and that PGDHi does not add to or interfere with this process.

Fig. 1.

Semaglutide reduces body weight and muscle mass while preserving muscle function in obese mice. (A) Experimental schematic. 8-wk-old mice were fed with a high-fat diet (HFD) for 12 wk and then treated with semaglutide and/or PGDHi for 5 wk while maintaining the HFD. Maximal tetanic force was measured at the start of the treatment, at 3 and 5 wk post treatment. (B) Body weight in grams. (C) Blood glucose levels during the glucose tolerance test. (D) Wet tissue weights for iWAT (inguinal adipose tissue), eWAT (epididymal adipose tissue), and BAT (brown adipose tissue). (E) Representative H&E staining images of iWAT (inguinal adipose tissue), eWAT (epididymal adipose tissue), and BAT (brown adipose tissue). (F) Relative wet tissue weight of gastrocnemius muscles. (G) Representative images of gastrocnemius cross-sections at 5 wk after treatment initiation stained for laminin (red). (H) Quantification of minimum Feret diameter. (I) Maximal tetanic force across the treatment. (J) Specific force at the end of treatment (5 wk). n = 8 to 10 animals per group. Data are presented as mean ± SEM. Statistical significance was determined by two-way ANOVA with multiple-comparison correction. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

To determine if the marked loss of body weight was accompanied by a loss of skeletal muscle mass and strength, we weighed the gastrocnemius (GA) and tibialis anterior (TA) muscles of obese mice and found that semaglutide-treated mice showed a significant reduction in GA and TA muscle mass (Fig. 1F and SI Appendix, Fig. S1E). However, histological analyses of GA muscle sections revealed no detectable differences in myofiber size across treatment groups (Fig. 1 G and H), suggesting that the observed reduction in muscle mass was not sufficient to produce measurable changes in fiber size, as previously reported (30). Similarly, muscle strength remained unchanged across all groups throughout the intervention (Fig. 1I). To evaluate muscle quality, we calculated specific force by normalizing plantar flexor force to GA muscle mass, as the gastrocnemius is the principal contributor to plantar flexion. We found that semaglutide, either alone or together with PGDHi, showed a trend toward increased specific force (Fig. 1J), in agreement with the observed preservation of absolute force (Fig. 1I), but a reduced muscle mass. Together, these findings suggest that semaglutide induces body weight loss predominantly through fat reduction, but also reduces skeletal muscle mass, consistent with prior mouse and human data (7, 17, 31). Importantly, PGDHi does not interfere with semaglutide-induced loss of weight, and does not significantly affect muscle mass or strength in young mice, likely because 15-PGDH levels increase with aging and are relatively low in young muscles, as previously reported (19).

15-PGDH Inhibition Significantly Increases Recovery of Muscle Strength Following Injury in Semaglutide-Treated Obese Mice.

We sought to determine if GLP-1 RA treatment would alter the regenerative capacity of muscle. Skeletal muscle possesses remarkable plasticity and can readily adapt to a range of external stimuli such as exercise, metabolic stress, or tissue injury (32–34). To test if regenerative capacity was altered, young adult obese mice were treated for 3 wk with vehicle, semaglutide, PGDHi, or the drug combination. At the end of the third week, GA muscles were injured via intramuscular injection of notexin. Semaglutide and PGDHi treatments were continued for an additional 2 wk postinjury (Fig. 2A) at which time force was assessed and GA muscles were harvested and weighed to assess muscle mass. As observed in the absence of injury, semaglutide significantly reduced gastrocnemius muscle weight compared to vehicle-treated controls, whereas PGDHi alone did not alter muscle mass (Fig. 2B). Notably, despite the reduction in muscle mass, semaglutide alone did not reduce maximal tetanic force and showed a nonsignificant trend toward increased specific force relative to vehicle-treated controls. Importantly, force measurements revealed a significant increase in muscle performance in mice cotreated with semaglutide and PGDHi, not seen with either treatment alone (Fig. 2 C and D). Additionally, specific force was markedly increased in the cotreated mice, suggesting that the increase in efficiency of regeneration resulted in enhanced muscle quality (Fig. 2E). These findings suggest that PGDHi cotreatment augments muscle functional recovery postinjury in semaglutide-treated mice.

Fig. 2.

Coadministration of semaglutide and PGDHi significantly increases muscle force recovery 2 wk after injury in obese mice. (A) Experimental schematic. Mice were fed a high-fat diet (HFD) for 12 wk and then treated with semaglutide and/or PGDHi for 5 wk while maintaining the HFD. After 3 wk of treatment, gastrocnemius muscles were injured by myotoxin injection. Maximal tetanic force was measured at the start of the treatment, at 3 and 5 wk after treatment initiation. (B) Gastrocnemius (GA) wet tissue weight. (C) Representative maximal tetanic force traces at week 5. (D) Maximal tetanic force across treatment groups at week 5. (E) Relative specific force at the end of the treatment (5 wk). n = 8 to 10 animals per group. Data are presented as mean ± SEM. Statistical significance was determined by two-way ANOVA with multiple-comparison correction. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

To elucidate potential structural correlates of this improved function, we performed histological analyses on gastrocnemius muscle cross-sections collected 2 wk postinjury. Tissue sections were immunostained for laminin to delineate fiber boundaries, as well as hematoxylin and eosin (H&E) and Sirius Red staining to assess overall morphology and fibrosis, respectively (Fig. 3 and SI Appendix, Fig. S2). Quantification of regenerating fibers, identified by centrally located nuclei (SI Appendix, Fig. S3), revealed that semaglutide-treated regenerated muscles displayed significantly smaller fiber diameters compared to all other groups (Fig. 3B), indicative of impaired myofiber growth during regeneration. Notably, this reduction in fiber size was rescued in mice cotreated with semaglutide and PGDHi. Sirius Red and H&E staining revealed regions of presumptive calcification (SI Appendix, Fig. S2 A and B), which were confirmed by Alizarin red staining as calcium deposits in the gastrocnemius of vehicle-treated obese mice following notexin injury (Fig. 3 C and D and SI Appendix, Fig. S4). These pathological features were reduced in the presence of semaglutide, with and without PGDHi. Notably, the accumulation of calcium deposits is similar to that reported for DMD muscle where it has been linked to a regenerative defect (35–37). Taken together, these data indicate that semaglutide has both beneficial and detrimental effects during muscle regeneration, limiting the accumulation of calcium deposits but reducing the size of regenerating myofibers. Cotreatment with PGDHi overcomes this deficit in growth and promotes more effective regeneration leading to increased muscle quality, evident as specific force.

Fig. 3.

Semaglutide reduces calcium deposits but impedes regenerating myofiber growth in obese mice, which is restored by cotreatment with PGDHi. (A) Representative images of gastrocnemius muscles at 2 wk post injury. Laminin (red) and DAPI (blue) staining. (B) Quantification of the minimum Feret diameter of centrally nucleated fibers in laminin-stained sections. (C) Representative images of gastrocnemius muscles at 2 wk post injury stained with Alizarin Red. (D) Quantification of calcified area in Alizarin Red-stained sections. n = 8 to 10 animals per group. Data are presented as mean ± SEM. Statistical significance was determined by two-way ANOVA with multiple-comparison correction. *P < 0.05; **P < 0.01; ***P < 0.001.

15-PGDHi Enhances Muscle Stem Cell Proliferation and Regeneration in Semaglutide-Treated Obese Mice.

We reasoned that the relatively diminished CSA of muscle fibers post injury in semaglutide-treated obese mice results from compromised regenerative capacity. Muscle regeneration relies on the activation, proliferation, and differentiation of resident muscle stem cells (MuSCs), also known as satellite cells (38, 39). We have previously demonstrated that prostaglandin E2 (PGE2) is essential to MuSC function. In mice in which the PGE2 receptor EP4 is genetically ablated, regeneration is severely blunted and strength is not restored postinjury (22). Similar results are observed if endogenous PGE2 synthesis is abrogated by treatment of mice post injury with an NSAID, indomethacin, that blocks COX-1 and COX-2 activity. This failure to regenerate and restore muscle function after injury is due to a requirement for PGE2 for MuSC proliferation and survival. In the absence of this signaling pathway they cannot meet the needs of expansion to repair the muscle damage (22).

To test if pharmacological inhibition of 15-PGDH enhances MuSC proliferation in injured muscles, gastrocnemius muscles of obese mice were injured via intramuscular notexin injection, and drug treatments were continued for 5 d. To track in vivo MuSC proliferation, mice received continuous supplementation of 5-ethynyl-2’-deoxyuridine (EdU) in the drinking water starting from the day of injury. Based on established kinetics of the MuSC proliferative response, which peaks around 5 d postinjury (38, 40), muscles were harvested at this time point for histological and molecular analyses (Fig. 4A). As expected from previous experiments (Fig. 1), semaglutide-treated animals (both semaglutide and semaglutide/PGDHi groups) exhibited significant reductions in total body weight, heart weight, and fat mass compared to vehicle controls (SI Appendix, Fig. S5 A–E). The regenerating gastrocnemius muscle mass was also significantly decreased in semaglutide-treated mice, regardless of PGDHi coadministration (Fig. 4B).

Fig. 4.

Coadministration of semaglutide and PGDHi enhances muscle stem cell function in injured obese mice. (A) Experimental schematic. Mice were fed a high-fat diet (HFD) for 12 wk and then treated with semaglutide and PGDHi, or their respective controls, while maintained on HFD. After 3 wk of treatment, gastrocnemius muscle was injured by notexin and EdU supplementation was initiated. Treatment continued for an additional 5 d, after which mice were humanely killed and tissues were collected. (B) Relative gastrocnemius wet tissue weight. (C) Representative gastrocnemius cross-sections at day 5 post injury stained for Pax7 (red), EdU (green), Laminin (white), and DAPI (blue). Yellow arrows indicate Pax7⁺EdU⁺ cells; White arrows indicate Pax7⁺EdU⁻ cells. (D) Quantification of Pax7⁺EdU⁺ cells per area. (E) Representative gastrocnemius cross-sections at day 5 post injury stained for Myogenin (red) and DAPI (blue). (F) Quantification of Myogenin⁺ cells per area. (G) Representative gastrocnemius cross-sections at day 5 post injury stained for eMyHC (green), Dystrophin (white), EdU (Red), and DAPI (blue). (H) Quantification of eMyHC⁺ fibers per area. For panels B and D, n = 6 animals per group; both gastrocnemius muscles were injured and included in the analysis. For panel F, n = 5 to 6 animals per group; only one gastrocnemius muscle per animal was used. Data are presented as mean ± SEM. Statistical significance was determined by two-way ANOVA with multiple-comparison correction. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

To determine if PGDHi promotes MuSC proliferation during regeneration, we performed immunofluorescence staining on cryosections of injured gastrocnemius muscles at 5 d postinjury. Sections were costained for EdU, a marker of DNA synthesis, and Pax7, the canonical marker of quiescent and activated MuSCs. This dual labeling enabled identification of actively proliferating MuSCs (EdU⁺Pax7⁺) (Fig. 4C). Quantification revealed a significant increase in the number of EdU⁺Pax7⁺ cells per section in the PGDHi + semaglutide group compared to semaglutide alone and vehicle controls (Fig. 4D). To assess progression through the myogenic program, we also stained muscle sections for Myogenin, a marker of myogenic differentiation (Fig. 4 E and F). Semaglutide alone reduced the number of Myogenin⁺ cells, consistent with diminished regenerative myofibers size (Fig. 3). In contrast, cotreatment with PGDHi restored Myogenin⁺ cell numbers to levels comparable to vehicle controls, indicating rescue of the semaglutide-associated differentiation deficit rather than enhancement above baseline. These data support a pro-regenerative effect of PGDHi during semaglutide treatment by enhancing MuSC proliferation and rescuing impaired myogenic differentiation.

To assess the downstream consequences of enhanced MuSC proliferation on myofiber regeneration, we analyzed the expression of embryonic myosin heavy chain (eMyHC), a marker of nascent regenerating myofibers. Immunostaining revealed a trend toward decreased numbers of eMyHC-positive fibers in semaglutide treated muscles compared to vehicle-controls (Fig. 4 G and H), and significant increase in eMyHC-positive fibers in PGDHi + semaglutide treated muscles, suggestive of more robust muscle regeneration in cotreated mice relative to those treated only with semaglutide.

To further dissect the mechanism by which PGE2 promotes MuSC proliferation, we isolated α7-integrin⁺ MuSCs from uninjured 12-wk-old C57BL/6 mice. To mimic the nutrient deficit generated by semaglutide, we cultured the MuSCs under nutrient-depleted conditions, and supplemented the media with PGE2, EP2/EP4 receptor inhibitors (EPi) to block PGE2 binding to its receptors, or a CREB phosphorylation inhibitor (pCREBi) (Fig. 5A). CREB was targeted because activation of EP2/EP4 receptors is known to increase intracellular cAMP and stimulate CREB phosphorylation, a key transcriptional mediator of PGE2-driven proliferation and survival (20, 22). Thus, EPi tests whether the proliferative effect of PGE2 is receptor-dependent, whereas pCREBi tests whether downstream cAMP–CREB signaling is required. EdU incorporation assays revealed that PGE2 treatment significantly increased MuSC proliferation compared to vehicle controls, whereas cotreatment with either a combination of EP2 or EP4 inhibitors or a pCREB inhibitor attenuated this effect (Fig. 5 B and C). These results indicate that the proliferative effects of PGE2 on MuSCs require EP receptor signaling and downstream CREB activation, providing a link between 15-PGDH inhibition, PGE2, and MuSC proliferation.

Fig. 5.

PGE2 enhances MuSC proliferation in vitro under nutrient-depleted conditions. (A) Schematic of experimental design. MuSCs (α7-integrin⁺, CD45^–, Sca1^–, CD31^–) were isolated from skeletal muscle of 12-wk-old male C57BL/6 mice by flow cytometry and cultured in regular or nutrient-depleted medium with the indicated treatments. EdU incorporation was used to assess proliferation. (B) Quantification of EdU⁺ MuSCs in nutrient-depleted medium. PGE2 increased proliferation compared with DMSO, while cotreatment with EPi (EP2/EP4 inhibitors) or pCREBi (CREB inhibitor) reduced this effect. Data are presented as mean ± SEM. Statistical significance: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. (C) Representative immunofluorescence images of EdU (green) and DAPI (blue) under the indicated conditions. (Scale bar, 200 μm.)

Together our findings show that concurrent GLP-1 receptor agonist and PGDHi treatments synergize to promote efficacious regeneration evident as increased MuSCs function, myofiber sizes, and specific force, without compromising weight loss.

Discussion

GLP-1 receptor agonists have revolutionized the treatment of obesity, leading to substantial and sustained weight loss across diverse populations (41). However, a portion of this weight loss is attributable to a reduction in lean mass, including skeletal muscle (7, 14, 15, 23, 24). There is growing concern that weight loss may come at the expense of strength, which is essential for mobility and quality of life (14, 15). Indeed, since many individuals discontinue GLP-1 RA therapy (42, 43), there is a risk that fat mass will be regained after treatment cessation while muscle mass will not be restored (44), increasing the likelihood of an obese frail population. Therefore, strategies that preserve or augment muscle during pharmacologic weight loss are of particular interest. Here, we show that 15-PGDH inhibition overcomes the muscle regenerative deficit observed with GLP-1 RA–induced weight loss by enhancing muscle stem cell function and increasing muscle strength.

Despite broad consensus that semaglutide reduces skeletal muscle mass in obese mice (17, 30, 45) and humans (7, 46, 47), its impact on muscle strength is less clear, with results varying by species and by the functional assay used. In mice, some studies report impaired muscle function using forelimb grip strength (45), or ex vivo twitch force (30), whereas other clinical studies report little to no change in physical performance or strength assessed by hand grip strength (48) or by chair rise and gait speed (31). Here, we extend prior work by measuring in vivo maximal tetanic force, a physiologically relevant endpoint that integrates neural activation with whole-muscle force production and is less confounded by motivation, learning, or limb use than voluntary tests such as grip strength. Notably, we find that semaglutide-induced reductions in muscle mass are not accompanied by a decrement in strength under steady-state conditions, as maximal tetanic force was not affected. Furthermore, cotreatment with semaglutide and a 15-PGDH inhibitor yielded similar results to semaglutide alone, indicating that 15-PGDH inhibition does not interfere with the weight loss effects of semaglutide. 15-PGDH inhibition did not affect muscle mass or force under these conditions, consistent with our previous finding that 15-PGDH activity is relatively low in young muscle (19).

With aging and with weight loss there is a concern that muscle strength will be lost and therefore building muscle through exercise is strongly advised. We reasoned that at steady state deficits may not be detected that would become apparent upon anabolic challenge. We therefore tested whether semaglutide alters MuSC-dependent regenerative capacity and whether PGDHi can enhance the muscle-building response postinjury. Satellite cells are dedicated muscle stem cells (MuSCs) that are juxtaposed to myofibers in a quiescent state, poised to spring into action when damage occurs. They are essential to skeletal muscle building in response to physiological challenges such as exercise (49–51) or trauma (52, 53). Because GLP-1 receptor agonist–induced weight loss is accompanied by a reduction in lean mass triggered by low calorie intake and high catabolism, we sought to determine whether muscle regeneration is altered in obese mice undergoing semaglutide-induced weight loss, and if so, whether 15-PGDH inhibition could surmount this deficit. To test this, we used notexin injury as an in vivo assay for myogenic repair capacity that maximally challenges MuSC activation, expansion, and fusion to rebuild myofibers. We find that although semaglutide did not reduce MuSC proliferation, it reduced myogenic differentiation and the growth and size of regenerating myofibers, an impairment that is ameliorated by 15-PGDH inhibition. Semaglutide-treated obese mice displayed significantly smaller regenerating myofibers, indicative of reduced myofiber growth. This phenotype resembles that observed in fasting models, where energy deprivation blunts regeneration and reduces fiber size compared to ad libitum-fed controls (54). Mechanistically, caloric restriction has been shown to impede regeneration by driving MuSCs into a deeper quiescent state, delaying their cell cycle entry (54), a mechanism that may also help explain the effects of semaglutide, given its potent suppression of appetite and nutrient intake. Consistent with this interpretation, we show that the reduced MuSC proliferation under nutrient-depleted culture conditions is partially rescued by PGE2, in good agreement with the known role of PGE2 in enhancing MuSC function (20, 22).

We have previously shown that MuSCs require PGE2 to function (20, 22). In the absence of this signaling pathway, MuSCs fail to proliferate and frequently die. Following injury, force is lost due to the inability of MuSCs to respond to PGE2 in the absence of the EP4 receptor or in the presence of an NSAID that blocks endogenous PGE2 synthesis (22). Moreover, the proliferative deficit of aged MuSCs is overcome upon exposure to PGE2 ex vivo as is evident by their superior engraftment following transplantation (20). Strikingly, the response of MuSCs in the muscles of aged mice in situ to a single exposure of PGE2 postinjury is so profound that muscle mass and strength are significantly increased 2 wk later (20, 22). Furthermore, inhibition of 15-PGDH, the major enzyme that degrades PGE2, elevates PGE2 levels and not only acts to enhance stem cell function, but also muscle fiber mitochondrial and contractile function and neuromuscular connectivity in aged mice (19, 21). Importantly, 15-PGDH inhibition allows physiological modulation of PGE2, to preinjury or young levels. Here we extend these findings by showing that 15-PGDH inhibition synergizes with semaglutide to promote muscle repair and strength recovery after injury in obese mice.

We found that regenerating muscles of vehicle-treated obese mice had large areas of calcification. Such calcified regions interfere with efficacious contractile function and have been previously reported in incidences of muscle regeneration where pathological remodeling occurred (55, 56), for example in Duchenne muscular dystrophy (DMD), a disease of stem cell exhaustion and regenerative failure. Calcification may arise from aberrant fibrotic, osteogenic, or inflammatory signaling, such as BMP (56) and TGF-β signaling by fibroadipogenic progenitors (FAPs) (57, 58), and appears to be indicative of failed or misdirected regeneration. Notably, semaglutide treatment reduced the accumulation of calcified lesions, likely because it reduces circulating TGF-β which is upregulated in obesity (59). This finding suggests that GLP-1 RA has a protective effect against maladaptive muscle tissue remodeling after notexin injury. However, this benefit comes at the cost of reduced myofiber growth. This balance between improved muscle quality and reduced myofiber growth may explain why maximal contractile force in semaglutide-treated mice remained similar to vehicle controls. Importantly, the administration of 15-PGDH inhibitor (PGDHi) in semaglutide-treated obese mice increased regenerating fiber size and improved muscle quality and force recovery after injury.

Several therapeutic strategies aimed at enhancing muscle mass in the context of GLP-1 RA mediated weight loss have focused on inhibiting the myostatin/activin signaling pathway (17, 60), given its well-characterized role in muscle wasting. While these approaches have demonstrated substantial hypertrophic effects in preclinical models, translation to human populations has shown mixed results. Clinical trials with myostatin inhibitors alone have reported increases in lean mass without corresponding gains in muscle strength (61, 62). More recently, Phase 2 combination trials pairing GLP-1 RA therapies with myostatin/ActRII-pathway inhibitors (e.g., Regeneron and Scholar Rock programs) have reported improved body composition and preservation of lean mass, but whether these regimens increase force remains to be determined in mice and humans.

In conclusion, our results identify 15-PGDH inhibition as a therapeutic strategy with the potential to overcome the muscle regenerative deficits associated with semaglutide-induced weight loss without blunting semaglutide’s beneficial effects on body weight reduction. Phase 1 clinical trials have shown that PGDHi is safe in humans (63). Our findings provide mechanistic evidence in support of a cotherapy of GLP-1 RA and PGDHi, that will enable robust fat loss while preserving muscle regenerative capacity and boosting strength.

Methods

Mouse Studies.

Animal experiments were performed in accordance with procedures approved by the Institutional Animal Care and Use Committee of the Stanford Animal Care and Use Committee (APLAC) protocol number #ALPAC-32982. C57BL/6 J male mice were purchased from the Jackson Laboratory and were maintained on a high-fat diet (HFD, 60% kcal from fat, # D12492, Research Diets) for 12 wk. All mice were in good health and housed in a temperature-controlled (20 to 22 °C) room on a 12-h light/dark cycle with ad libitum access to food and water. Mouse body weights were recorded daily. At the end of the experiments, body and tissue weights were recorded. Tissues were collected and frozen for further analysis.

GLP-1 RA and PGDHi Administration.

Semaglutide (Selleckchem cat #S9697) was used as GLP-1 RA at a concentration of 120 μg/kg and SW033291 (ApexBio cat #A8709; named PGDHi in this study) was used at a dose of 5 mg/kg. Mice were treated with both drugs or the corresponding vehicle controls daily via intraperitoneal injections. Semaglutide was diluted in 90% saline solution, 5% Kolliphor and 5% DMSO, and PGDHi was diluted in 10% ethanol, 5% Cremophor EL (Sigma-Aldrich cat #C5135), 85% D5W (Dextrose 5% Water) as previously described (25).

Glucose Tolerance Tests.

For glucose tolerance testing, mice were fasted for 4 h and then received an intraperitoneal injection of glucose (1.5 g/kg body weight). Blood glucose levels were recorded at 0, 15, 30, 45, 60, 90, and 120 min postinjection.

Muscle Injury.

We induced acute muscle injury by intramuscular injection of notexin into the gastrocnemius (GA) muscle (40 µL; 10 µg/mL; Latoxan, catalog #L8104). Notexin is a snake venom–derived phospholipase A² toxin that, when injected into muscle, causes rapid myofiber necrosis (64). The resulting tissue damage initiates a regenerative program characterized by activation and proliferative expansion of muscle stem cells to restore the injured muscle. Tissues were collected at 5 and 14 d post injury.

Histology and Tissue Staining.

After collection, tissues were embedded in O.C.T. compound and snap frozen in liquid nitrogen. Using a cryostat, 10 to 15 µm sections were cut. Samples were kept at −20 °C until histochemical analysis.

For histochemical analysis, sections were fixed in 4% PFA for 10 min and washed 3 times with PBS for 10 min each wash. After washing, the sections were blocked with Blocking One reagent for 45 min, and incubated overnight at 4 °C with the corresponding first antibodies: anti-laminin (ab11575), anti-embryonic myosin heavy chain (eMyHC; clone F1.652), anti-myogenin (MyoG; 556358), and anti-Pax7 (Developmental Studies Hybridoma Bank). The next day, samples were washed 3 times for 10 min with PBS-T and incubated at room temperature for 1 h with the corresponding antibodies. After incubation with secondary antibodies, samples were washed 3 times with PBS for 5 min each time, counterstained with DAPI, and mounted in Fluoromont-G (SouthernBiotech).

For hematoxylin and eosin (H&E) staining, iWAT, eWAT, and BAT were formalin-fixed, paraffin-embedded, and sectioned at 6 µm. Sections were deparaffinized and dehydrated with xylene and ethanol. Briefly, slides were stained with hematoxylin, washed with water and 95% ethanol, and stained with eosin for 30 min. Sections were then incubated with ethanol and xylene and mounted with mounting medium. Digital images were captured with an Aperio AT2 (Leica) and QuPath software was used to extract the images.

Image Analysis.

Stained sections were imaged using a Keyence microscope and analyzed with the BZ-X analyzer software (Keyence, Osaka, Japan). For quantifications, whole tissue cross sections were analyzed.

Force Measurements.

Peak isometric torque of the ankle plantar flexors was measured as previously described (19, 21, 65, 66). Briefly, anesthetized mice (3% isoflurane mixed with oxygen) were positioned with the foot secured to a footplate connected to a servomotor (Model 300C-LR, Aurora Scientific). Percutaneous Pt-Ir electrode needles (Aurora Scientific) were inserted over the tibial nerve near the posterior-medial knee. The ankle was fixed at a 90° angle, and maximal isometric torque was elicited by stimulating the tibial nerve at 150 Hz using 0.1 ms square wave pulses. Three tetanic contractions were recorded per muscle, with 1-min intervals between trials. Measurements were performed blinded to treatment groups. For relative force assessment, values were recorded both before and 1 mo after treatment, and percent change was calculated. Data acquisition and analysis were performed using Dynamic Muscle Data Acquisition and Analysis Software (Aurora Scientific). Force was measured longitudinally in mice undergoing semaglutide and/or PGDHi treatment at baseline, 3 wk and 5 wk. Specific force was calculated by normalizing plantar flexion maximal contraction to the harvested gastrocnemius muscles.

Mouse Muscle Stem Cells Isolation.

Muscle stem cells (MuSCs) were isolated from mouse hindlimb muscles using a combination of enzymatic digestion, magnetic depletion, and fluorescence-activated cell sorting (FACS). Briefly, dissected muscles were minced and digested with 0.2% collagenase type II (Worthington) for 60 min, followed by an additional 30 min digestion in collagenase/dispase (0.04 U/mL; Thermo Fisher Scientific). The resulting suspension was triturated through an 18-gauge needle to release mononuclear cells. Cells were pelleted by centrifugation at 300 g for 5 min, resuspended in FACS buffer (PBS, 2% FBS, 2 mM EDTA), and filtered through a 70 μm strainer.

For immunolabeling, cells were incubated on ice for 45 min in the dark with the following antibodies: CD11b (1:800), CD45 (1:500), Sca1 (1:200), CD31 (1:200), and α7-integrin–PE (1:200). After washing in FACS buffer, samples were incubated with streptavidin-coated magnetic beads (1:200; Miltenyi Biotec) to deplete lineage-positive cells (CD11b⁺, CD45⁺, Sca1⁺, and CD31⁺). Lineage-negative cells were then enriched using magnetic activated cell sorting (MACS). The enriched fraction was subsequently analyzed by flow cytometry, and MuSCs were defined and sorted based on α7-integrin–PE positivity, using unstained controls to establish gating. This strategy reliably yielded α7-integrin⁺CD11b^–CD45^–Sca1^–CD31^– MuSCs with high purity for downstream culture and functional assays.

Mouse Muscle Stem Cells Culture and Treatments.

Isolated MuSCs were cultured in either nutrient-depleted or regular growth medium. The nutrient-depleted medium consisted of DMEM low glucose (1 g/L) supplemented with 1 mM sodium pyruvate, 2% charcoal-stripped FBS, 5 ng/mL FGF2 (Peprotech), 1× penicillin/streptomycin, and 1.0 mM GlutaMAX (Thermo Fisher Scientific). This formulation excluded phenol red to reduce extrinsic nutrient and signaling inputs. The regular growth medium consisted of standard DMEM supplemented with 20% FBS, 5 ng/mL FGF2, and 1× penicillin/streptomycin.

Cells were plated in 96-well plates and cultured for 48 h under the indicated conditions. Pharmacological treatments were applied as follows: PGE2 (10 μM), EP2/EP4 receptor inhibitor (10 μM), or CREB inhibitor (1 μM). For proliferation assays, cells were pulsed with EdU (5 μM) during the final 12 h of culture. At the end of the treatment period, cells were fixed and processed for EdU incorporation analysis according to the manufacturer’s protocol (Thermo Fisher Scientific).

Statistical Analyses.

Data are presented as mean ± SEM and as individual data points. Statistical analyses were conducted using GraphPad Prism. Replicates and sample sizes (N), representing individual animals, are also provided in the figure legends. Prior to treatment, mice were randomized based on body weight and baseline force measurements. Group differences were analyzed using ANOVA followed by Tukey’s post hoc test for multiple comparisons. A P-value < 0.05 was considered statistically significant.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.