SEARCH

SEARCH BY CITATION

Keywords:

  • hypophosphatemia;
  • X-linked hypophosphatemic rickets;
  • tumor-induced osteomalacia;
  • end stage renal disease;
  • calcium × phosphorus product

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We investigated if the circulating levels of the phosphaturic factor FGF23 are elevated in subjects with XLH. Although we failed to find a statistically significant increase, FGF23 levels were significantly correlated with the degree of hypophosphatemia in XLH. In contrast, FGF23 levels were markedly increased in subjects with ESRD and correlated inversely with the degree of hyperphosphatemia.

Introduction: Inactivating mutations of PHEX cause renal phosphate wasting in X-linked hypophosphatemic rickets (XLH) because of the accumulation of a phosphaturic hormone called phosphatonin. The recent discovery that FGF23 is the circulating phosphaturic factor in autosomal dominant hypophosphatemia raises the possibility that FGF23 is phosphatonin.

Methods: Fasting serum FGF23 levels and serum biochemical parameters were measured using a human FGF23 (C-terminal) ELISA assay in 11 subjects with XLH and 42 age-matched controls, 5 subjects with hypophosphatemia of unknown cause, and 14 hyperphosphatemic subjects with end stage renal disease (ESRD). Associations between variables were examined using the Spearman's correlation coefficient and linear regression analysis.

Results and Conclusions: FGF23 (RU/ml) concentrations were not different (p = 0.11) between control and hypophosphatemic XLH subjects, but were significantly increased in hyperphosphatemic subjects with ESRD (p < 0.001). Western blot analysis found the presence of both full-length and C-terminal FGF23 fragments in serum from ESRD subjects. There was a strong inverse correlation between FGF23 and serum phosphorus (r = −0.60) and calcium and phosphorus (Ca × P) product (r = −0.65) in XLH, and a strong positive relationship between FGF23 and Pi (r = 0.50) and Ca × P product (r = 0.62) in ESRD. FGF23 levels were variably elevated in subjects with hypophosphatemia of unknown cause, one of which had tumor-induced osteomalacia (TIO). Removal of the tumor resulted in rapid reduction in serum FGF23 levels. These findings suggest that FGF23 has a possible role in mediating hypophosphatemia in XLH and TIO, but the overlapping levels of FGF23 in hypophosphatemic disorders and normal subjects indicate that serum phosphorus and FGF23 can also be independently regulated.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

The parathyroid hormone-vitamin D axis plays an important role in regulating renal and gastrointestinal absorption of phosphate, but it is not sufficient to account for the complexity of phosphate regulation.(1) There is emerging evidence for primary hormonal pathways for maintaining systemic phosphate homeostasis.(2–6) Identification of the disease gene in autosomal dominant hypophosphatemia (ADH) has identified FGF23 as a novel phosphaturic hormone.(4,5) Missense mutations of the 176-RXXR-179 processing site of FGF23 leads to its impaired degradation of FGF23 in ADH.(4) FGF23, as well as other potential phosphaturic factors, has also been identified in tumors from subjects with the acquired hypophosphatemic disorder tumor-induced osteomalacia (TIO).(4,6,7)

X-linked hypophosphatemia (XLH) shares a common clinical phenotype with ADH and TIO, suggesting that renal phosphate wasting in these disorders may have a common pathogenesis.(8) Although there is evidence for the accumulation of a circulating phosphaturic factor, referred to as phosphatonin,(9) in XLH, this disorder is caused by inactivating mutations of PHEX, a member of the M13 family of zinc-dependent type II cell-surface membrane metalloproteases.(2,8,10,), 11 Some(12) but not all studies(11) have implicated FGF23 as a PHEX substrate and the phosphaturic factor in XLH. It is not clear that FGF23 is increased in XLH as a consequence of inactivating PHEX mutations or whether FGF23 or other phosphaturic factors are responsible for hypophosphatemia in XLH.

In the current study, we used a recently developed commercial assay for FGF23 to measure circulating levels of this factor in subjects with normal and disordered phosphate homeostasis. Specifically, we measured FGF23 levels in normal controls, subjects with hypophosphatemia caused by XLH and unknown causes, and hyperphosphatemic subjects with end stage renal disease (ESRD). We also sought to define what factors may be determinants of FGF23 levels in an attempt to shed further insight into the physiologic regulation of FGF23 in health and disease.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Subject population

A total of 72 subjects were prospectively recruited for participation in the study, which was approved by the Institutional Review Board of Duke University Medical Center. Eleven subjects with XLH from eight separate families were followed at Duke University Medical Center. Seven of these subject were part of the original cases identifying PHEX as the gene causing XLH,(2) whereas four nongeneotyped subjects were identified through the clinical practice of one clinician (TJW). All patients with XLH had a history of documented hypophosphatemia (serum phosphorus < 2.5 mg/dl), and characteristic family of X-linked inheritance. Five patients with XLH were from separate families, and the remaining six patients with XLH were from three separate families as follows: a mother and her son, a sister and brother, and two first cousins. Five subjects with hypophosphatemia of unknown cause were also identified. Subjects with ESRD were recruited from the hemodialysis clinic at Duke University Medical Center where they were undergoing thrice weekly hemodialysis. Control subjects were collected through advertisements, did not have any identifiable disorders of calcium or phosphorus homeostasis, and were age and sex matched with XLH subjects at an approximately 4:1 ratio. One subject with XLH and three subjects with unknown hypophosphatemia were on calcitriol and phosphorus therapy. All subjects with ESRD were treated with either calcitriol or paracalcitol as well as phosphate binders.

Diagnosis of TIO

During the course of these investigations, one subject in the hypophosphatemia of unknown cause category was suspected to have TIO based on an increased FGF23 level and the absence of a family history of hypophosphatemia. To identify the tumor, we performed scintigraphic studies using penetreotide (Sandostatin; Novartis, East Hanover, NJ, USA) labeled with Indium (111) according to standard techniques. An enhancing lesion in the left lower extremity was identified, and the subject underwent successful resection of a mesenchymal phosphaturic tumor. Sectioning and hematoxylin and eosin staining of paraffin-embedded tissue were performed according to standard techniques.

Biochemical analysis

A single blood sample was collected after a 12-h overnight fast from all subjects, with the exception of the subject with TIO, who underwent serial blood sampling immediately before and at times 0.25, 0.5, 1, 2, 4, 8, 24, and 48 h after resection of the tumor. Individual serum samples were isolated by centrifugation and frozen at −20°C before biochemical analysis. Serum FGF23 levels were measured using the human C-terminal ELISA Kit (Immutopics Inc., San Clemente, CA, USA). The interassay CV for this assay is 12% and the intra-assay CV is 14%. Intact parathyroid hormone (iPTH) and whole (wPTH) were measured using the Duo PTH immunoradiometric assay (Scantibodies, Santee, CA, USA). Serum chemistries were measured by routine methods using an Ortho Diagnostics Vitros 950 autoanalyzer.

SDS-PAGE and Western blot analysis

The human serum samples were concentrated by filtration through a Centricon-50 (Millipore Corp., Bedford, MA, USA) to separate proteins less than 50 kDa. Proteins in 25 μl of concentrated serum were separated on 4–10%NuPAGE Tris-Bis gel (Invitrogen Corp., Carlsbad, CA, USA) under reducing conditions and transferred onto nitrocellulose membrane (Bio-Rad, Chicago, IL, USA) using the Xcell II Blot Module (Invitrogen Corp.). The membrane was incubated with the 1:1000 anti-FGF23 polyclonal goat antibody that recognizes the C-terminal of FGF23 (gift from Immutopics, Inc.). Blots were incubated with I:2000 diluted donkey anti-goat horseradish perioxidase (HRP) conjugated secondary antibody (Santa Cruz Biotech, Inc., Santa Cruz, CA, USA) and visualized by enhanced chemiluminescence (ECL; Amersham, Piscataway, NJ, USA).

Statistical analysis

For baseline variables, we used the Kruskal-Wallis and Wilcoxon-Mann-Whitney tests to compare groups with regard to continuous variables and χ2 and Fisher's exact test with regard to categorical variables. For further analysis, we excluded the subjects with unknown disease, and because of the inherent differences between the three remaining groups, we performed all analyses separately for each one of them. Association between variables was examined using the Spearman's correlation coefficient and linear regression analysis. Before the regression analysis, FGF23 was transformed using natural logarithm to obtain normality. Data are presented as mean ± SE for normally distributed variables and as median (range) for variables that are not normally distributed. The statistical analysis was performed with SSPS.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Circulating FGF23 levels in normal subjects

Table 1 shows the demographics and serum biochemistries for the four groups. Control subjects had normal calcium, phosphorus, and PTH levels, as well as normal renal function, as assessed by serum creatinine levels. In normal controls, the median FGF23 level was 13 RU/ml and ranged from 5 to 210 RU/ml (Table 1), in a pattern that was not normally distributed (Fig. 1). Thirty-eight of the 42 controls had levels less than 30 RU/ml, while there were 3 notable outliers in the control group (FGF23 levels: 67, 120, and 210 RU/ml), all with normal serum phosphorus and creatinine levels (Fig. 1).

Table Table 1. Characteristics of Normal Controls, Hypophosphatemic Subjects with XLH, and Unknown Cause and Hyperphosphatemic Subjects With End Stage Renal Failure
Thumbnail image of
thumbnail image

Figure FIG. 1.. Individual values of serum FGF23 levels in control, XLH, unknown hypophosphatemic, and ESRD subjects. FGF23 levels are displayed on a logarithmic scale. The horizontal line indicates the median value.

Download figure to PowerPoint

Circulating FGF23 levels subjects with hypophosphatemic disorders

We evaluated two groups with hypophosphatemia (Table 1). Subjects with XLH had significantly lower serum phosphorus levels compared with controls (Table 1). Subjects with XLH also had significant increments in alkaline phosphatase, a marker of osteoblast activity, and serum PTH levels, indicative of mild hyperparathyroidism (Table 1). All subjects with XLH had serum calcium and creatinine levels in the normal range. Only one of the subjects with XLH was receiving treatment with calcitriol and elemental phosphorus. The median serum FGF23 level was 35 RU/ml (range, 5–195 RU/ml) in XLH subjects (Fig. 1). Although the difference was not statistically significant (Wilcoxon-Mann-Whitney test, p = 0.11) from controls, seven of the eleven subjects with XLH had values greater than 30 RU/ml. Differences in FGF23 levels did not track with the family of origin; affected members of the same XLH family showed variability in FGF23 levels. Indeed, affected individuals in the same kindred had both normal and elevated FGF23 levels.

The other group with hypophosphatemia of unknown cause also had lower serum phosphorus levels compared with controls (Table 1). Serum calcium was normal, but serum creatinine was increased in one subject and normal in the other four. In addition, PTH levels were increased in this group compared with controls. Similar to XLH subjects, serum FGF23 levels were variably elevated (Fig. 1). In two subjects, the FGF23 levels were less than 30 RU/liter, while in the other three subjects, values ranged from 57 to greater than 4000 RU/liter. Only one subject with a markedly elevated FGF23 levels had impaired renal function.

Further evaluation of one subject with unexplained hypophosphatemia and an FGF23 level above the median value for controls identified TIO as the cause of renal phosphate wasting. An In (111) octreotide scan identified a tumor in the left lower leg (Fig. 2A). Histological analysis of the resected tumor revealed mixed connective tissue features, hypervascularity, and multinucleated giant cells consistent with a mesenchymal phosphaturic tumor (Fig. 2B), similar to that reported by Weidner and Santa Cruz.(13) After tumor resection, serum FGF23 levels fell rapidly (Fig. 2C), whereas 48 h was required for the serum phosphorus to normalize (Fig. 2D).

thumbnail image

Figure FIG. 2.. Imaging, histology, and biochemical studies in a subject with TIO. (A) Scinitigraphic images obtained 24 h after injection of 7.5 mCi of octreotide labeled with In-111. Pooling of radiolabeled octreotide is demonstrated in the proximal medial soft tissue of the left lower leg. (B) Hematoxylin and eosin staining of resected tissue reveals a mesenchymal-type tumor with abundant extracellular matrix/chondroid type material (arrow) and giant multinucleated cells (*). (C) Time course of serum FGF23 levels after tumor resection. (D) Time course of serum phosphorus levels after tumor resection.

Download figure to PowerPoint

Circulating FGF23 levels in ESRD

To evaluate subjects with elevated serum phosphorus, we assessed 14 subjects with ESRD treated with thrice weekly hemodialysis. Subjects with ESRD were older than the other groups and had significantly higher levels of phosphorus and calcium × phosphorus product than all other groups (Table 1). Consistent with mild to moderate secondary hyperparathyroidism, they also demonstrated elevations of both whole PTH, which measures PTH(1–84), and intact PTH, which measures PTH(1–84) and circulating N-terminal PTH fragments. Serum levels of alkaline phosphatase levels were also increased. Compared with the other groups, FGF23 levels were markedly elevated, with a median FGF23 level of 12,630 RU/ml (range, 1157–94,046; Fig. 1).

To assess the proportion of elevated FGF23 caused by the presence of circulating C-terminal fragments versus bioactive full-length FGF23, we performed Western blot analysis of serum using an FGF-23 specific antibody (Fig. 3). Sera derived from representative control, ESRD, XLH, and TIO subjects were compared with recombinant FGF23 in conditioned media from Cos-7 cells expressing a human FGF23 cDNA. In cell-conditioned media (Fig. 3, lane 1) the 32-kDa full-length FGF23 protein is barely detectable because of the proteolytically processing at the RXXR site to generate the predominant 18-kDa fragment detected by the C-terminal antibody (the N-terminal fragment is not detected with this antibody). In ESRD subjects, we observed an ∼32-kDa band corresponding to the full-length FGF23 protein and a more abundant ∼18-kDa C-terminal fragment (Fig. 3, lane 4). The greater ratio of the C-terminal fragment to the full-length FGF23 suggests that less than one-quarter of the circulating FGF23 is bioactive in ESRD subjects. In contrast, FGF23 protein could not be detected in other subjects, reflecting the lower circulating concentrations of FGF23 (Fig. 3, lanes 2, 3, and 5).

thumbnail image

Figure FIG. 3.. Western blot analysis of FGF23 proteins in human serum samples. Recombinant human FGF23 derived from conditioned medium from FGF23-producing Cos-7 cells was used as a positive control (lane 1). Size selected serum samples from a control (lane 2), XLH (lane 3), ESRD (lane 4), and TIO (lane 5) subject were loaded onto each lane. Western blot analysis was carried with anti C-terminal human FGF23 antibody as described in the Materials and Methods section. Both full length (∼32 kDa) and C-terminal fragments (∼18 kDa) were detected from positive control conditioned medium and the serum of ESRD subjects. No bands were detected from serum samples of derived from control, XLH, and TIO subjects.

Download figure to PowerPoint

Determinants of serum FGF23 levels

Next, we investigated which factors might determine the serum FGF23 levels in normal control, XLH, and ESRD subjects. Initially we examined associations between specific variables using Spearman's correlation coefficient (Table 2). In the control subjects, only age was significantly correlated with FGF23 levels (Table 2). There was no significant correlation between FGF23 and any of the other variables measured, including gender, and more importantly, serum phosphorus.

Table Table 2. Correlation Between FGF23 and Other Variables in Control, XLH, and ESRD Subjects
Thumbnail image of

In subjects with XLH, FGF23 levels were significantly and inversely correlated with phosphorus (Table 2; Fig. 4A). Equally strong negative and positive correlations were also seen with the Ca × P product and intact PTH, respectively; although neither calcium alone or whole PTH had significant correlations with FGF23 levels. Because PTH was also elevated in XLH subjects, we performed additional analyses to examine the relative role of FGF23 and other potential determinants of phosphorus levels in XLH subjects (Table 3). We found that only FGF23 levels (with and without log transformation) were inversely related to serum phosphorus, whereas no other factors studied showed a significant association. Finally, we performed stepwise linear multiregression analyses to determine which factors contributed most significantly to serum phosphorus levels in XLH subjects. In a model including phosphorus as the dependent variable and lnFGF23, whole PTH, calcium, and alkaline phosphatase as independent variables, only lnFGF23 was independently associated with serum phosphorus (R2 = −0.4, β = −0.199, p = 0.038).

Table Table 3. Correlation Between Serum Phosphorus and Other Variables in XLH Subjects
Thumbnail image of
thumbnail image

Figure FIG. 4.. Different relationship between serum FGF23 and phosphorus in XLH and ESRD. (A) Regression analysis between serum phosphorus and FGF23 in XLH patients. FGF23 is inversely related to serum phosphorus (LnFGF23 = 8.53–1.98 × phosphorus; p = 0.038). (B) Regression analysis between serum phosphorus and FGF23 levels in ESRD subjects. FGF23 is positively correlated with serum phosphorus (LnFGF23 = 5.86 + 0.46 × phosphorus; p = 0.037).

Download figure to PowerPoint

In ESRD subjects, we found the opposite relationship between serum FGF23 and phosphorus (Table 2, Fig. 4B), observing a significant positive correlation of serum phosphorus with FGF23 levels. We also observed a significant positive correlation with the Ca × P product, which is a parameter used to assess the risk of extraskeletal calcification in ESRD. We did not observe a significant relationship between either intact PTH or whole PTH and FGF23 in subjects with ESRD. Multivariable linear regression analysis revealed that FGF23 levels were best explained by the Ca × P product in ESRD (r2 = 0.51, β = 0.062, p = 0.004), suggesting an interaction between calcium and phosphorus in regulating FGF23 levels in this setting.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

FGF23 is a new member of the FGF family of proteins that has phosphaturic and hypophosphatemic actions.(3,4) In the current study, we found evidence that immunoreactive FGF23 circulates in normal subjects, albeit at low levels in most subjects. In a few subjects, however, FGF23 was elevated without apparent abnormalities of phosphate homeostasis or renal function (Fig. 1). Using an ELISA assay that detects both biologically active full-length and inactive C-terminal fragments, the median, mean ± SE, and range were 13, 25 ± 6, and 5–210 RU/liter, respectively, in fasting normal subjects (Table 1; Fig. 1). Because our assay is directed toward the C-terminal region and FGF23 is processed into N- and C-terminal fragments, inactive C-terminal fragments and biologically intact FGF23 levels were measured. Consequently, the absolute values do not represent biologically active FGF23. Although expressed in different units, our results, however, are similar to the mean ± SE value of 29 ± 1 ng/liter and range of 8.2–54.3 ng/liter reported in control subjects by Yamazaki et al., using another two-site ELISA assay that detects full-length FGF23.(14) FGFs typically act as autocrine/paracrine factors and do not normally circulate in nonpregnant humans, but can be detected in serum under pathological conditions.(15–19) The structural features of FGF23, which consists of an N-terminal FGF homology domain and a unique C-terminal region,(4) may account for the ability of this new member of the FGF family to circulate under normal conditions. We did not identify gender- or serum phosphorus-dependent differences in FGF23, but we did observe that FGF23 increases with age. Our results contrast with those of Yamazaki et al., who did not observe any relationship with age.(14) These differences likely reflect altered metabolism and/or clearance of the C-terminal fragments that are detected by the assay used in our study.

While autosomal dominant hypophophatemic rickets (ADHR) results from mutant FGF23 that is resistant to proteolysis,(5) recent studies have implicated overproduction of FGF23 by TIO tumors that overwhelms the capacity of degradative enzymes(6) and impaired metabolism of FGF23 in XLH as the cause of hypophosphatemia in these related disorders.(3,12) Similar to other studies that have assessed FGF23 levels in XLH,(14) we failed to show increased FGF23 in all subjects with XLH or established a threshold level of FGF23 required to cause hypophosphatemia in an individual subject (Table 1; Fig. 1). Because of the variability in FGF23 levels, we also could not establish significant differences in the median value or range between normal and XLH subjects (Table 1; Fig. 1). While we had greater number of subjects with XLH than prior reports(20) and compared our results to age-matched controls, an even larger sample size might reveal significant differences between FGF23 levels in XLH and normal subjects.

One could interpret our data as failing to support a role of FGF23 in mediating the hypophosphatemia in XLH. This view is inconsistent with the in vitro observation that PHEX cleaves FGF23(12) but is consistent with other observations that the RXXR site controlling FGF23 metabolism is the processing site of enzymes other than PHEX.(4,6,11,), 21 Indeed, FGF23 is a very labile protein that is degraded into inactive N- and C-terminal fragments in cell culture systems in which it is produced, presumably by a ubiquitous subtilisin-like protease that is not PHEX (Fig. 3). On the other hand, our study revealed a correlation between serum phosphorus and FGF23 levels that is stronger than between phosphorus and PTH (Fig. 4), suggesting that FGF23 is the most important determinant of serum phosphorus in XLH. Therefore, it is conceivable that “normal” levels of FGF23 that we observed in some XLH subjects are inappropriate for the degree of hypophosphatemia, analogous to “normal” PTH levels found in a subset of hypercalcemic subjects with primary hyperparathyroidism. Nevertheless, given the overlap between FGF23 levels in normal and hypophosphatemic XLH subjects (Fig. 1), it is necessary to either postulate that other factors maintain serum phosphorus in normal subjects with high FGF23 levels or that FGF23 is not entirely responsible for hypophosphatemia in XLH. Further studies are needed to establish the role of FGF23 in the pathogenesis of XLH and confirm that FGF23 is a physiological substrate for PHEX.

FGF23 also has been implicated in the pathogenesis of hypophosphatemia in TIO.(20) In this study, production by the tumor causing TIO was confirmed by the rapid reductions in FGF23 levels after tumor removal (Fig. 2), similar to that reported by Yamazaki et al.(14) Although we did not measure fractional tubular reabsorption of phosphate, which may have permitted the detection of earlier effects of FGF23 reductions on phosphate transport after tumor resection, normalization of the serum phosphorus levels was delayed. Serial analysis of gene expression (SAGE) analysis of TIO tumors have identified other phosphaturic factors, such as secreted frizzled related protein-4 (sFRP-4) and matrix extracellular phosphoglycoprotein (MEPE), in greater abundance than FGF23, suggesting that these additional phosphaturic factors may also have been removed with the tumor.(7,22) Consequently, the reduction in FGF23 after removal of the tumor, while confirming FGF23 production by the tumor, does not prove that FGF23 is responsible for the hypophosphatemia.

Surprisingly, we found elevated PTH levels in some subjects with XLH as well as in several subjects with hypophosphatemia of unknown cause (Table 1). Elevated PTH levels, however, do not seem to account for the hypophosphatemia in XLH. In this regard, multiregression analysis demonstrated that FGF23 was a more important predictor of serum phosphorus than FGF23 in XLH (Table 3). In addition, previous studies have shown that parathyroidectomy does not cure hypophosphatemia in the Hyp mouse homolog of this disorder.(23) Nevertheless, the increased PTH levels suggest abnormal parathyroid gland function in these hypophosphatemic disorders as reported previously.(24,25) One possibility is that prior treatment of hypophosphatemia resulted in secondary hyperparathyroidism. However, only one subject was being treated. Because PHEX is expressed in parathyroid glands, it is possible that inactivation of mutations in this endopeptidase might affect PTH metabolism.(24) In support of this possibility, only intact PTH, which measures full-length and (7–84) fragments, but not whole PTH, which measures only full-length PTH, correlated with FGF23 levels in XLH. This suggests that XLH may share a common abnormality in metabolism of these bioactive hormones.

We also found that FGF23 levels are elevated in ESRD, consistent with impaired renal clearance of inactive FGF23 fragments. Analysis within normal controls and other groups, however, failed to observe any significant correlation between FGF23 accumulation and serum creatinine, most likely because these groups represented either normal or absent renal function and lacked subjects with differing degrees of renal insufficiency. The detection of C-terminal fragments of FGF23 confounds the interpretation, analogous to detection of inactive PTH fragments with C-terminal radio-immunoreactive assays. The presence of full-length FGF23 protein in the serum of ESRD subjects by Western blot analysis, however, suggests that bioactive FGF23 circulates in these subjects. Moreover, the finding that serum phosphorus is inversely related to FGF23 levels in this setting also suggests that FGF23 production and/or clearance are regulated by hyperphosphatemia in ESRD. That phosphorus may stimulate the increase in FGF23 is supported by preliminary clinical observations showing an effect of phosphate loading to increase circulating FGF23 levels(26) and the preliminary findings showing that intact FGF23 increases commensurately in rats with progressive azotemia and hyperphosphatemia.(27) In addition, the observation that the Ca × P product was the best predictor of FGF23 levels in ESRD (Table 2) is consistent with an interaction with calcium.

Although overlapping clinical features of ADHR, XLH, and TIO suggest a common pathogenesis of these disorders mediated by FGF23, the apparent high circulating FGF23 levels in some normal individuals, the variable increase in FGF23 in XLH subjects, and the report of other phosphaturic factors in TIO suggest that these disorders may have a more complex pathogenesis. Therefore, despite the observed associations between serum phosphorus and FGF23, a causal relationship between FGF23 and hypophosphatemia in XLH and TIO cannot be established by our studies. While FGF23 and PHEX are likely two pieces of a novel hormonal cascade controlling phosphate homeostasis, further advances in our knowledge of FGF23 metabolism, PHEX substrates, and other phosphaturic factors are still needed to piece together the common pathogenesis of TIO, ADH, and XLH.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

This work was supported in part by National Institute of Arthritis and Musculoskeletal and Skin Diseases Grants RO1-AR37308 and RO1-AR45955. We thank Amy Wright for ascertainment of subject samples and Cristy McGranahan for secretarial assistance in the preparation of this manuscript.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Yanagawa N, Lee DBN 1992 Renal handling of calcium and phosphorus. In: CoeFL, FavusMJ (ed.) Disorders of Bone and Mineral Metabolism. Raven Press, New York, NY, USA, pp. 340.
  • 2
    Anonymous 1995 A gene (PEX) with homologies to endopeptidases is mutated in subjects with X-linked hypophosphatemic rickets. The Hyp Consortium. Nat Genet 11:130136.
  • 3
    Strewler GJ 2001 FGF23, hypophosphatemia, and rickets: Has phosphatonin been found? Proc Natl Acad Sci USA 98:59455946.
  • 4
    Shimada T, Mizutani S, Muto T, Yoneya T, Hino R, Takeda S, Takeuchi Y, Fujita T, Fukumoto S, Yamashita T 2001 Cloning and characterization of FGF23 as a causative factor of tumor-induced osteomalacia. Proc Natl Acad Sci USA 98:65006505.
  • 5
    White KE, Evans WE, O'Riordan JL, Speer MC, Econs MJ, Lorenz-Depiereux B, Grabowski M, Meitinger T, Strom TM 2000 Autosomal dominant hypophosphatemic rickets is associated with mutations in FGF23. Nat Genet 26:345348.
  • 6
    White KE, Jonsson DG, Carn G, Hampson G, Spector TD, Mannstadt M, Lorenz-Depiereux B, Miyauchi A, Yang IM, Ljunggren O, Meitinger T, Strom TM, Juppner H, Econs MJ 2001 The autosomal dominant hypophosphatemic rickets (ADHR) gene is a secreted polypeptide overexpressed by tumors that cause phosphate wasting. J Clin Endocrinol Metab 86:497500.
  • 7
    DeBeur DM, Fennegan RB, Vassiliadis J, Cook B, Barberio D, Estes S, Manavalan P, Petroziello J, Madden SL, Cho JY, Kumar R, Levine MA, Schiavi SC 2002 Tumors associated with oncogenic osteomalacia express genes important in bone and mineral metabolism. J Bone Miner Res 17:11021110.
  • 8
    Quarles LD, Drezner MK 2001 Pathophysiology of X-linked hypophosphatemia, tumor-induced osteomalacia, and autosomal dominant hypophosphatemia: A perPHEXing problem. J Clin Endocrinol Metab 86:494496.
  • 9
    Kumar R 1997 Phosphatonin-a new phosphaturetic hormone? (Lessons from tumor-induced osteomalacia and X-linked hypophosphatasemia.) Nephrol Dial Transplant 12:1113.
  • 10
    Grieff M, Mumm S, Waeltz P, Mazzarella R, Whyte MP, Thakkar RV, Schlessinger D 1997 Expression and cloning of the human X-linked hypophosphatemia gene cDNA. Biochem Biophys Res Commun 231:635639.
  • 11
    Guo R, Liu S, Spurney RF, Quarles LD 2001 Analysis of recombinant Phex: An endopeptidase in search of a substrate. Am J Physiol Endocrinol Metab 281:E837E847.
  • 12
    Bowe AE, Finnegan R, Jan de Beur SM, Cho J, Levine MA, Kumar R, Schaivi SC 2001 FGF-23 inhibits renal tubular phosphate transport and is a PHEX substrate. Biochem Biophys Res Commun 284:977981.
  • 13
    Weidner N, Santa Cruz D 1987 Phosphaturic mesenchymal tumors. A polymorphous group causing osteomalacia or rickets. Cancer 59:14421454.
  • 14
    Yamazaki Y, Okasaki R, Shibata M, Hasegawa Y, Satoh K, Tajima T, Takeuchi Y, Fujita T, Nakahara K, Yamashita T, Fukumoto S 2002 Increased circulatory level of biologically active full-length FGF-23 in subjects with hypophosphatemic rickets/osteomalacia. J Clin Endocrinol Metab 87:49574960.
  • 15
    Hill DJ, Flyvbjerg A, Arany E, Lauszus FF, Klebe JG 1997 Increased levels of serum fibroblast growth factor-2 in diabetic pregnant women with retinopathy. J Clin Endocrinol Metab 82:14521457.
  • 16
    Kurz C, Hefler L, Zeisler H, Schatten C, Husslein P, Tempfer C 2001 Maternal basic fibroblast growth factor serum levels are associated with pregnancy-induced hypertension. J Soc Gynecol Investig 8:2426.
  • 17
    Katinioti AA, Tousoulis D, Economou E, Stefanadis C, Trikas A, Tentolouris C, Pitsavos C, Androulakis A, Toutouzas P 2002 Basic fibroblast growth factor changes in response to coronary angioplasty in subjects with stable angina. Int J Cardiol 84:195199.
  • 18
    Sasaki H, Yukiue H, Kobayashi Y, Nakashima Y, Moriyama S, Kaji M, Kiriyama M, Fukai I, Yamakawa Y, Fujii Y 2001 Elevated serum vascular endothelial growth factor and basic fibroblast growth factor levels in subjects with thymic epithelial neoplasms. Surg Today 31:10381040.
  • 19
    Brattstrom D, Bergqvist M, Larsson A, Holmertz J, Hesselius P, Rosenberg L, Brodin O, Wagenius G 1998 Basic fibroblast growth factor and vascular endothelial growth factor in sera from non-small cell lung cancer subjects. Anticancer Res 18:11231127.
  • 20
    Yamazaki Y, Okazaki R, Shibata M, Hasegawa Y, Satoh K, Tajimi T, Takeuchi Y, Futia T, Nakahara K, Yamashita T, Fukumoto S 2002 Increased circulatory level of biologically active full-length FGF-23 in subjects with hypophosphatemic rickets/osteomalacia. J Clin Endocrinol Metab 87:49574960.
  • 21
    Nakayama K 1997 Furin: A mammalian subtilisin/Kex2p-like endoprotease involved in processing a wide variety to precursor proteins. Biochem J 327:625635.
  • 22
    Kumar R 2002 New insights into phosphate homeostasis: Fibroblast growth factor 23 and frizzled-related protein-4 are phosphaturic factors derived from tumors associated with osteomalacia. Curr Opin Nephrol Hypertens 11:547553.
  • 23
    Meyer RA Jr, Tenenhouse HS, Meyer MH, Klugerman AH 1989 The renal phosphate transport defect in normal mice parabiosed to X-linked hypophosphatemic mice persists after parathyroidectomy. J Bone Miner Res 4:523532.
  • 24
    Blydt-Hansen TD, Tenenhouse HS, Goodyer P 1999 PHEX expression in parathyroid gland and parathyroid hormone dysregulation in X-linked hypophosphatemia. Pediatr Nephrol 13:607611.
  • 25
    Posillico JT, Lobaugh B, Muhlbaier LH, Drezner MK 1985 Abnormal parathyroid function in the X-linked hypophosphatemic mouse. Calcif Tissue Int 37:418422.
  • 26
    Allen HC, Whybro A, Barker ME, Estell R, Bulumsohn A 2002 Endocrine response to escalating-dose phosphate supplementation in MEN: Is FGF-23 phosphatonin. J Bone Miner Res 17:S1;S159.
  • 27
    Yamashita T, Hasegawa H, Yamazake Y, Kawata T, Urakawa I, Shimada T, Takeuchi Y, Fujita T, Fukumoto S, Nagano N 2002 Involvement of FGF-23 in abnormal vitamin D and mineral metabolism associates with renal insufficiency. J Am Soc Nephrol 13:577A.