Fibroblast growth factor 23 in oncogenic osteomalacia and X-linked hypophosphatemia.

BACKGROUND: Mutations in fibroblast growth factor 23 (FGF-23) cause autosomal dominant hypophosphatemic rickets. Clinical and laboratory findings in this disorder are similar to those in oncogenic osteomalacia, in which tumors abundantly express FGF-23 messenger RNA, and to those in X-linked hypophosphatemia, which is caused by inactivating mutations in a phosphate-regulating endopeptidase called PHEX. Recombinant FGF-23 induces phosphaturia and hypophosphatemia in vivo, suggesting that it has a role in phosphate regulation. To determine whether FGF-23 circulates in healthy persons and whether it is elevated in those with oncogenic osteomalacia or X-linked hypophosphatemia, an immunometric assay was developed to measure it. METHODS: Using affinity-purified, polyclonal antibodies against [Tyr223]FGF-23(206-222)amide and [Tyr224]FGF-23(225-244)amide, we developed a two-site enzyme-linked immunosorbent assay that detects equivalently recombinant human FGF-23, the mutant form in which glutamine is substituted for arginine at position 179 (R179Q), and synthetic human FGF-23(207-244)amide. Plasma or serum samples from 147 healthy adults (mean [+/-SD] age, 48.4+/-19.6 years) and 26 healthy children (mean age, 10.9+/-5.5 years) and from 17 patients with oncogenic osteomalacia (mean age, 43.0+/-13.3 years) and 21 patients with X-linked hypophosphatemia (mean age, 34.9+/-17.2 years) were studied. RESULTS: Mean FGF-23 concentrations in the healthy adults and children were 55+/-50 and 69+/-36 reference units (RU) per milliliter, respectively. Four patients with oncogenic osteomalacia had concentrations ranging from 426 to 7970 RU per milliliter, which normalized after tumor resection. FGF-23 concentrations were 481+/-528 RU per milliliter in those with suspected oncogenic osteomalacia and 353+/-510 RU per milliliter (range, 31 to 2335) in those with X-linked hypophosphatemia. CONCLUSIONS: FGF-23 is readily detectable in the plasma or serum of healthy persons and can be markedly elevated in those with oncogenic osteomalacia or X-linked hypophosphatemia, suggesting that this growth factor has a role in phosphate homeostasis. FGF-23 measurements might improve the management of phosphate-wasting disorders.

Immunohistochemical detection of FGF-23 protein in tumors that cause oncogenic osteomalacia.

OBJECTIVE: Oncogenic hypophosphatemic osteomalacia (OOM) is a rare disease characterized by hypophosphatemia, inappropriately low levels of circulating 1,25-dihydroxyvitamin D(3) and osteomalacia. The disease is most commonly caused by benign mesenchymal tumors that produce, among several other factors, fibroblast growth factor-23 (FGF-23). Current evidence thus suggests that this protein has an important role in the regulation of phosphate homeostasis. By producing polyclonal antibodies against human FGF-23 protein we wanted to determine the localization of FGF-23 protein in OOM tumors that express FGF-23 mRNA. DESIGN AND METHODS: Three polyclonal antibodies were raised in rabbits against three different peptides with sequences derived from human FGF-23: [Cys-70]FGF-23(51-69)amide, [Tyr-223]FGF-23(206-222)amide and [Tyr-224]FGF-23(225-244)amide. One of the resulting antisera was subsequently used for immunohistochemistry on sections from five different tumors causing OOM. FGF-23 mRNA expression was confirmed with in situ hybridization. RESULTS: After affinity purification, two of three antisera detected recombinant human FGF-23 by Western blot analysis. Through immunohistochemical analysis using the anti-[Tyr-224]FGF-23(225-244)amide antibody and through in situ hybridization using full-length antisense FGF-23 cRNA as a probe, we showed that abundant amounts of FGF-23 protein and mRNA are present in certain tumor cells of five different OOM tumors. CONCLUSIONS: We conclude that OOM tumors express FGF-23 protein and that the immunohistochemical detection of FGF-23 in OOM tumors is feasible and may help in establishing the diagnosis of tumor-induced hypophosphatemia through analysis of biopsies or surgical specimens.

Structure of non-(1-84) PTH fragments secreted by parathyroid glands in primary and secondary hyperparathyroidism.

BACKGROUND: Non-(1-84) parathyroid hormone (PTH) fragments are large circulating carboxyl-terminal (C) fragments with a partially preserved amino-terminal (N) structure. hPTH (7-84), a synthetic surrogate, has been demonstrated to exert biologic effects in vivo and in vitro which are opposite to those of hPTH (1-34) on the PTH/PTHrP type I receptor through a C-PTH receptor. We wanted to determine the N structure of non-(1-84) PTH fragments. METHODS: Parathyroid cells isolated from glands obtained at surgery from three patients with primary hyperparathyroidism and three patients with secondary hyperparathyroidism were incubated with 35S-methionine to internally label their secretion products. Incubations were performed for 8 hours at the patient-ionized calcium concentration and in the presence of various protease inhibitors. The supernatant was fractionated by high-performance liquid chromatography (HPLC) and fractions were analyzed with PTH assays having (1 to 4) and (12 to 23) epitopes, respectively. The serum of each patient was similarly analyzed. Peaks of immunoreactivity identified were submitted to sequence analysis to recover the 35S-methionine residues in positions 8 and 18. RESULTS: Three regions of interest were identified with PTH assays. They corresponded to non-(1-84) PTH fragments (further divided in regions 3 and 4), a peak of N-PTH migrating in front of hPTH (1-84) (region 2) and a peak of immunoreactivity corresponding to the elution position of hPTH (1-84) (region 1). The last corresponded to a single sequence starting at position 1. Region 2 gave similar results in all cases (a major signal starting at position 1) but also sometimes minor sequences starting at position 4 or 7. Regions 3 and 4 always identified a major sequence starting at positions 7 and minor sequences starting at positions 8, 10, and 15. Surprisingly, a major signal starting at position 1 was also present in region 3. The HPLC profile obtained from a given patient's parathyroid cells was qualitatively similar to the one obtained with his/her serum in each case. CONCLUSION: These results indicate that non-(1-84) PTH fragments are composed of a family of fragments which may be generated by specific or progressive cleavage at the N region. The longest fragment starts at position 4 and the shortest at position 15. A peptide starting at position 7 appears as the major component of non-(1-84) PTH fragments. The generation process is similar to the one described for smaller C-PTH fragments a number of years ago, suggesting a similar production mechanism and source for all C-PTH fragments.

Similar predictive value of bone turnover using first- and second-generation immunometric PTH assays in pediatric patients treated with peritoneal dialysis.

BACKGROUND: Accurate measurements of the concentration of parathyroid hormone (PTH) in serum or plasma are essential for the proper assessment of renal osteodystrophy. The first-generation immunometric PTH assay (1st PTH-IMA) not only detects the intact hormone, but also additional PTH fragments truncated at the amino N-terminally truncated PTH-derived fragments [ntPTH(1-84)]. A second-generation immunometric PTH assay (2nd PTH-IMA) recognizes only PTH(1-84) and possibly PTH fragments that are truncated at the carboxyl-terminus but not PTH(7-84). Whether estimates of the ratio between PTH(1-84) and ntPTH(1-84) fragments are a better predictor of bone turnover remains controversial. METHODS: Thirty-three patients aged 12.8 +/- 4.4 years treated with continuous cycling peritoneal dialysis (CCPD) for 13 +/- 9 months underwent iliac crest bone biopsy. PTH levels were measured by two newly developed first-generation and second-generation PTH-IMA. The ntPTH(1-84) fragments were calculated by subtracting PTH values determined using the 2nd PTH-IMA from values obtained using 1st PTH-IMA that detects both PTH(1-84) and relatively large ntPTH(1-84). RESULTS: Determinations of PTH levels by both assays were highly correlated (r = 0.89, P < 0.001). The relationships between first-generation and second-generation PTH-IMA and bone formation were similar (r = 0.67, P < 0.0001 and r = 0.64, P < 0.0001, respectively). When patients were grouped according to the presence or absence of secondary hyperparathyroidism, the ratio PTH(1-84) to ntPTH(1-84) did not differ between groups. CONCLUSION: PTH concentrations determined by either the first- or the second-generation PTH-IMA were found to be better predictors of bone formation than the PTH(1-84) to ntPTH(1-84) fragments ratio.