Antibody-mediated activation of FGFR1 induces FGF23 production and hypophosphatemia

Abstract

The phosphaturic hormone Fibroblast Growth Factor 23 (FGF23) controls phosphate homeostasis by regulating renal expression of sodium-dependent phosphate co-transporters and cytochrome P450 enzymes involved in vitamin D catabolism. Multiple FGF Receptors (FGFRs) can act as receptors for FGF23 when bound by the co-receptor Klotho expressed in the renal tubular epithelium. FGFRs also regulate skeletal FGF23 secretion; ectopic FGFR activation is implicated in genetic conditions associated with FGF23 overproduction and hypophosphatemia. The identity of FGFRs that mediate the activity of FGF23 or that regulate skeletal FGF23 secretion remains ill defined. Here we report that pharmacological activation of FGFR1 with monoclonal anti-FGFR1 antibodies (R1MAb) in adult mice is sufficient to cause an elevation in serum FGF23 and mild hypophosphatemia. In cultured rat calvariae osteoblasts, R1MAb induces FGF23 mRNA expression and FGF23 protein secretion into the culture medium. In a cultured kidney epithelial cell line, R1MAb acts as a functional FGF23 mimetic and activates the FGF23 program. siRNA-mediated Fgfr1 knockdown induced the opposite effects. Taken together, our work reveals the central role of FGFR1 in the regulation of FGF23 production and signal transduction, and has implications in the pathogenesis of FGF23-related hypophosphatemic disorders.

Figure 1. R1MAbs induce hypophosphatemia.

(A) Serum phosphate and calcium levels in male C57BL/6 mice intraperioneally injected with R1MAb1 or isotype control (Control IgG) at 3 mg/kg. HFD-fed mice were on the diet for 15 weeks at the end of the study. Serum phosphate and calcium levels were determined at 7 days post injection. N = 8 mice/group. (B) Serum phosphate and calcium levels in female db/db mice intraperioneally injected with R1MAb1, OA-R1MAb1 or isotype control (Control IgG) at 3 mg/kg. Serum phosphate and calcium levels were determined at 7 days post injection. N = 5–7 mice/group. (C) Body weight, and serum phosphate and calcium levels in male C57BL/6 mice intraperioneally injected with R1MAb2 or isotype control (Control IgG) at 1 mg/kg. Control mice were subjected to pair feeding to adjust body weight. Body weight, serum phosphate and calcium levels were determined at 48 hour post injection. N = 8 mice/group. (A–C) * p<0.01, ** p<0.005 (versus control IgG). #<0.005 versus chow-fed group with the same antibody treatment.

Figure 2. R1MAb2 activates the FGF23 pathway in kidney cortex.

(A) mRNA expression analysis by qPCR in kidney cortex at 48 hour post injection. The same animals in Figure 1C were analyzed. N = 8 mice/group. * p<0.01, ** p<0.005. (B) Protein expression by western blot on the membrane fraction of kidney cortex homogenate at 48 hour post injection. Each lane represents an independent sample from different animals. N = 3 mice/group.

Figure 3. R1MAb2 induces FGF23 production.

(A and B) Serum FGF23 (A) and PTH (B) levels in male C57BL/6 mice intraperioneally injected with R1MAb2 or isotype control (Control IgG) at 1 mg/kg. The same animals described in Figure 1C and Figure 2 were analyzed at 48 hour post injection. N = 8 mice/group. (C) Serum FGF23 levels in female db/db mice intraperioneally injected with R1MAb1 or isotype control (Control IgG) at 2 mg/kg. The samples were collected at 7 days post injection. N = 6 mice/group. (D and E) Serum FGF23 levels (D) and phosphate levels (E) in male C57BL/6 mice intraperioneally injected with an indicated antibody at 1 mg/kg. The samples were collected at 3 days post injection. N = 8 mice/group. (F) FGF23 levels in culture medium after treatment of differentiate rat osteoblast with vitamin D (100 nM), R1MAb1, or isotype control IgG (26.7 nM). The cells were incubated for 48 hours in the presence of the indicated ligand. N = 6 samples/treatment. (A–F) * p<0.05, **<p<0.005, ***<p<0.0005. (G) Differentiated rat osteoblasts were treated with R1MAb2, or isotype control IgG (26.7 nM), for 1 hour and subjected to Western blot analysis to examine phosphorylation of MAPK pathway proteins, CREB and STAT3.

Figure 4. Gene expression in differentiated osteoblasts.

(A) mRNA was isolated from differentiated osteoblasts treated with the indicated ligands (vitamin D (100 nM), R1MAb2, or isotype control IgG (26.7 nM)) for 48 hours, and subjected to qPCR analysis. Data represent means ± SEM (N = 3). *P<0.05, **P<0.01, versus mock (for vitamin D) or versus control IgG (for R1MAb2). (B) Schematic summary of the data presented in (A). Vitamin D and R1MAb2 induce overlapping, but distinct sets of target genes.

Figure 5. Bone effects of R1MAb2 treatment.

(A) Blood glucose levels of female db/db mice (N = 8 mice/group) during the study. The mice received intraperitoneal injection of R1MAb1 or control IgG at 3 mg/kg doses on day 0 and day 42 (Arrow). Statistical significance in glucose reduction (p<0.05) was observed between day 3–31 and day 43–49. (B) Serum FGF23 levels of mice in (A) on day 49. p<0.01, N = 7–8 mice/group. (*# p<0.05, versus Control IgG (*) or versus Control IgG, PF (#)) (C) Bone phenotype of mice described in (A–B). The bones were dissected on day 49, and subjected to μCT analysis. Statistical significance (p<0.05) was observed only for total volume, but not other parameters shown. # p<0.05 (versus Control IgG, PF).

Figure 6. R1MAb2 activates FGFR1 in the presence of Klotho.

The activation of FGFR was assessed using GAL-Elk1 luciferase assay in rat L6 myoblast cells. Cells were co-transfected with an expression vector for Klotho (KL) and the indicated FGFR isoform together with GAL-Elk1, SV40-renilla Luciferase, and GAL-responsive firefly luciferase reporter. Transfected cells were incubated with increasing concentrations of R1MAb2, human FGF23, or a positive control, FGF1, for 8 h before luciferase assays. Firefly luciferase activity was normalized to renilla luciferase activity and expressed as fold induction over untreated control. The data represents means ± SEM (N = 3).

Figure 7. R1MAb2 activates the FGF23 pathway in kidney epithelial cells.

(A) OK cells were treated with vehicle (Mock), R1MAb2 (0.5, 5, or 50 nM), or isotype control IgG (50 nM) for 24 hours, and the mRNA expression of indicated genes were determined by qPCR. The expression of each gene was normalized by the expression of actin in the same sample and shown as relative expression. N = 3. (B) Similar gene expression analysis in OK cells after treatment with an indicated antibody at 50 nM. N = 6. (C) mRNA expression in OK cells treated with scrambled or FGFR1 siRNA oligos, determined by qPCR. N = 6. (D) Cyp24a1 gene expression in OK cells after treatment with siRNA oligos and an indicated ligand. N = 6. The data represents means ± SEM. * p<0.05 or *** p<0.001 compared with the control group.

Figure 8. A model for the mechanism of perturbation of phosphate homeostasis by anti-FGFR1 agonists.

See text for explanations.