Chronic kidney disease (CKD) is associated with various pathologic changes, including elevations in serum phosphate levels (hyperphosphatemia), vascular calcification, and skeletal muscle atrophy. Elevated phosphate can damage vascular smooth muscle cells and cause vascular calcification. Here, we determined whether high phosphate can also affect skeletal muscle cells and whether hyperphosphatemia, in the context of CKD or by itself, is associated with skeletal muscle atrophy. As models of hyperphosphatemia with CKD, we studied mice receiving an adenine-rich diet for 14 weeks and mice with deletion of Collagen 4a3 (Col4a3^−/−). As models of hyperphosphatemia without CKD, we analyzed mice receiving a high-phosphate diet for three and six months as well as a genetic model for klotho deficiency (kl/kl). We found that adenine, Col4a3^−/−, and kl/kl mice have reduced skeletal muscle mass and function and develop atrophy. Mice on a high-phosphate diet for six months also had lower skeletal muscle mass and function but no significant signs of atrophy, indicating less severe damage compared with the other three models. To determine the potential direct actions of phosphate on skeletal muscle, we cultured primary mouse myotubes in high phosphate concentrations, and we detected the induction of atrophy. We conclude that in experimental mouse models, hyperphosphatemia is sufficient to induce skeletal muscle atrophy and that, among various other factors, elevated phosphate levels might contribute to skeletal muscle injury in CKD.

Key Points: Our study identified skeletal muscle as a novel source of fibroblast growth factor 23 (FGF23) in scenarios of hyperphosphatemia, including high dietary phosphate intake and CKD. Skeletal muscle–derived FGF23 had endocrine actions lowering serum phosphate levels, which may protect from hyperphosphatemia-induced tissue damages. Skeletal muscle–derived FGF23 might also have paracrine effects counteracting atrophy. Background: Fibroblast growth factor 23 (FGF23) is a hormone that reduces the renal reabsorption of phosphate in response to systemic phosphate elevations. In CKD, serum levels of phosphate and FGF23 reach levels that can harm various tissues. While the bone acts as the main production site for FGF23, bone-specific gene deletion studies in mice suggest the existence of other FGF23 sources. Here, we determine if skeletal muscle produces FGF23 in response to phosphate elevations. Methods: We studied four mouse models with phosphate elevations, two with CKD (global Col4a3 deletion and adenine-rich diet) and two without CKD (genetic klotho deficiency and high-phosphate diet). Furthermore, we generated a new mouse line with skeletal muscle–specific Fgf23 deletion (KO), which received an adenine-rich or a high-phosphate diet. We measured skeletal muscle FGF23 by quantitative real-time PCR, ELISA, and immunohistochemistry, as well as serum levels of phosphate, FGF23, and parathyroid hormone (PTH). We determined FGF23 mRNA levels in muscle biopsies from patients with CKD, and we studied the effects of phosphate and PTH elevations on FGF23 expression in cultured myotubes isolated from mice and patients with CKD. Finally, we studied the effects of acute phosphate loading on urine phosphate levels in Fgf23 KO mice. Results: All four mouse models with phosphate and PTH elevations showed FGF23 expression in skeletal muscle tissue on mRNA and protein level. Phosphate, but not PTH, induced FGF23 expression in cultured myotubes. Furthermore, patients with CKD with higher serum phosphate levels expressed more FGF23 in skeletal muscle. Fgf23 KO mice had elevated serum phosphate levels when administered a high-phosphate diet and decreased urine phosphate levels after acute phosphate loading. Conclusions: Phosphate elevations induced FGF23 expression in skeletal muscle, independent of the absence or presence of CKD. Skeletal muscle–derived FGF23 reduced renal phosphate reabsorption.

Fibroblast growth factor (FGF) 23 is a bone-derived hormone that promotes renal phosphate excretion. Serum FGF23 is increased in chronic kidney disease (CKD) and contributes to pathologic cardiac hypertrophy by activating FGF receptor (FGFR) 4 on cardiac myocytes, which might lead to the high cardiovascular mortality in CKD patients. Increases in serum FGF23 levels have also been observed following endurance exercise and in pregnancy, which are scenarios of physiologic cardiac hypertrophy as an adaptive response of the heart to increased demand. To determine whether FGF23/FGFR4 contributes to physiologic cardiac hypertrophy, we studied FGFR4 knockout mice (FGFR4^−/−) during late pregnancy. In comparison to virgin littermates, pregnant wild-type and FGFR4^−/− mice showed increases in serum FGF23 levels and heart weight; however, the elevation in myocyte area observed in pregnant wild-type mice was abrogated in pregnant FGFR4^−/− mice. This outcome was supported by treatments of cultured cardiac myocytes with serum from fed Burmese pythons, another model of physiologic hypertrophy, where the co-treatment with an FGFR4-specific inhibitor abrogated the serum-induced increase in cell area. Interestingly, we found that in pregnant mice, the heart, and not the bone, shows elevated FGF23 expression, and that increases in serum FGF23 are not accompanied by changes in phosphate metabolism. Our study suggests that in physiologic cardiac hypertrophy, the heart produces FGF23 that contributes to hypertrophic growth of cardiac myocytes in a paracrine and FGFR4-dependent manner, and that the kidney does not respond to heart-derived FGF23.