X-linked hypophosphatemic rickets/osteomalacia (XLH), autosomal dominant and recessive hypophosphatemic rickets/osteomalacia (ADHR and ARHR) share common clinical features including high fibroblast growth factor 23 (FGF23) levels. These diseases are caused by mutations in phosphate regulating endopeptidase homolog, X-linked (PHEX), FGF23, and dentin matrix acidic phosphoprotein 1 (DMP1) gene respectively. It remains unclear whether these diseases can be clinically discriminated.
To clarify the underlying mechanism of patients with hypophosphatemic rickets whose parents showed no physical findings suggesting rickets.
Design and patients
The proband is a 39-year-old woman. She and her 37-year-old brother show the same clinical features such as bowing of legs together with hypophosphatemia (sister: P 1.8 mg/dl, brother: P 1.6 mg/dl) and high FGF23 levels (sister: 542 pg/ml, brother: 96 pg/ml). Physical findings of their parents are normal and ARHR was suspected.
Sequencing of all coding exons and exon–intron junctions of DMP1 and FGF23 genes showed no mutation. Subsequent analysis revealed that there is a deletion of 52 143 bp including exons 1–3 in PHEX gene in the brother. His sister was found to be a heterozygote for the same deletion indicating that they are suffering from XLH. The same deletion was detected in the mother. However, the amount of the wild-type allele was more and that of the mutant one was less in genomic DNA from the mother compared with those from the sister. Single nucleotide polymorphism (SNP) analysis indicated that the mother has three kinds of PHEX alleles suggesting a somatic mosaicism.
Careful genetic analysis is mandatory for correct differential diagnosis of hypophosphatemic rickets with high FGF23 levels.
Hypophosphatemia is a well-known cause of rickets/osteomalacia. There are several kinds of hypophosphatemic rickets/osteomalacia, which share common clinical features. X-linked hypophosphatemic rickets/osteomalacia (XLH), autosomal dominant and recessive hypophosphatemic rickets/osteomalacia (ADHR and ARHR) are characterized by hypophosphatemia associated with impaired proximal tubular phosphate reabsorption. In addition, hypophosphatemia usually enhances 1,25-dihydroxyvitamin D (1,25(OH)2D) production and increases serum 1,25(OH)2D level. However, 1,25(OH)2D remains to be low to low normal in patients with these diseases (1). Recent investigations indicated that excess actions of fibroblast growth factor 23 (FGF23) underlie the pathogenesis of these hypophosphatemic diseases (2, 3, 4, 5, 6). FGF23 suppresses proximal tubular phosphate reabsorption by reducing the expression of type 2a and 2c sodium–phosphate cotransporters, which mediate physiological phosphate reabsorption (7). FGF23 also decreases serum 1,25(OH)2D at least in part by suppressing 1,25(OH)2D production (7). Responsible genes for XLH, ADHR, and ARHR have been shown to be phosphate regulating endopeptidase homolog, X-linked (PHEX), FGF23, and dentin matrix acidic phosphoprotein 1 (DMP1) respectively (4, 5, 6, 8).
While patients with XLH, ADHR, and ARHR have similar clinical and biochemical features, these diseases show different modes of inheritance. Therefore, it is anticipated that these diseases can be clinically discriminated, especially in familial cases. Here, we report familial cases of hypophosphatemic rickets/osteomalacia. Our case indicates the difficulty of proper differential diagnosis of hypophosphatemic rickets/osteomalacia by clinical analysis alone.
Subjects and methods
The proband is a 39-year-old female. She was born from nonconsanguineous parents who showed no signs of metabolic bone diseases. She visited a hospital because she showed bowing of legs when she was 21 months old. She started to take oral vitamin D3 and phosphate. She switched to active vitamin D3 and phosphate when she was about 10 years old. She received 2 g of phosphate per day at most. She received wedge osteotomy of tibia at the age of 14. Hypercalcemia appeared when she was 16 years old and she received parathyroidectomy and transplantation of a part of her parathyroid glands for tertiary hyperparathyroidism when she was 20. Since then, intact PTH levels have been within the reference range with active vitamin D3 treatment.
Her 37-year-old brother also showed bowing of legs when he was 8 months old and started to take oral vitamin D3 and phosphate. The maximal dose of phosphate was 2.5 g/day. He later developed hypercalcemia with increased PTH levels when he was 13 years old and he stopped taking vitamin D3. However, hypercalcemia from tertiary hyperparathyroidism persisted and he started 0.1 μg/kg per day of 1,25(OH)2D3 when he was 17 years old with a partial reduction in PTH levels. He finally had parathyroidectomy and transplantation of a part of his parathyroid glands. Since then, he has taken oral active vitamin D3 and his symptoms have been stable except for mild renal dysfunction (estimated glomerular filtration rate of 55.5 ml/min per 1.73 m2 from the formula established by the Japanese Society of Nephrology).
Their mother is 68 years old and physical examination revealed no signs of rickets/osteomalacia. Their father showed no physical findings of rickets/osteomalacia, either, and did not approve to give blood samples for analysis. Biochemical parameters and clinical features of this family are summarized in Table 1. Because parents of these two patients showed no physical signs suggesting rickets/osteomalacia, both patients were presumed to have ARHR caused by mutations in DMP1 gene.
Clinical and biochemical features of the family.
|Lower limb (cm)||82||73||83|
|Arm span (cm)||157||155||175|
|Serum Ca (mg/dl)||9.1||9.1||9.9||8.5–10.2|
|Serum phosphate (mg/dl)||2.4||1.8||1.6||2.5–4.5|
|Intact PTH (pg/ml)||59||44||54||10–65|
|Estimated GFR (ml/min per 1.73 m2)||91.2||63.0||55.5|
GFR, glomerular filtration rate.
Sample collection and biochemical measurement
After written informed consent was obtained, blood samples were collected and frozen at −20 °C until analysis. Serum phosphate and creatinine were measured by autoanalyzer. Serum FGF23 was evaluated using FGF23 ELISA kit that detects only full-length, biologically active FGF23 (Kainos, Tokyo, Japan). This study was approved by the institutional review board of the University of Tokyo.
Sequencing DMP1, FGF23, and PHEX genes
Genomic DNA was extracted from peripheral blood leukocytes using a DNA extraction kit (QIAamp DNA Blood Mini Kit, Qiagen). We amplified all the coding exons and exon–intron junctions of DMP1, FGF23, and PHEX genes using PCR; primer sequences can be obtained upon request. PCR products were separated by electrophoresis with 1% agarose gel, purified using Wizard SV Gel and PCR Clean-Up system (Promega), and directly sequenced with the same primers by dye deoxy termination cycle sequencing.
Analysis of 5′ region of PHEX gene
Initial study suggested the presence of a deletion in 5′ region of PHEX gene in the male patient (see Results). It is known that spermine synthase (SMS) gene is the adjacent one present in the 5′ region of PHEX (9). Therefore, we prepared several primers corresponding to the intervening sequence between SMS and PHEX, and analyzed 5′ region of PHEX gene by PCR.
Semi-quantitative PCR of genomic DNA was conducted to compare the amount of wild-type and mutant PHEX alleles. Primer pairs amplifying either the deleted or undeleted regions were used to amplify the mutant or wild-type allele respectively. PCR was done using the same amount of genomic DNA (1 μg/tube) and 25 cycles were chosen not to saturate the amplification.
We have selected eight SNPs that exist between exons 1 and 3 of PHEX gene (rs178711, rs178712, rs2027822, rs2027824, rs2027825, rs178715, rs2301304, and rs178716) and prepared five sets of primers to cover these SNPs. PCR products were directly sequenced to analyze haplotypes of PHEX gene.
Direct sequencing of PCR products using genomic DNA from the proband and her brother identified no disease-causing mutation in all coding exons and the exon–intron junctions of FGF23 or DMP1 gene. In addition, no missense mutation in PHEX gene was found in PCR products from genomic DNA of the proband. However, no PCR products corresponding to exons 1–3 were obtained from genomic DNA of the male patient (Fig. 1a). Because PCR products corresponding to exons 1–3 were obtained using DNA from the proband (Fig. 1a), these results suggested that there is a deletion including exons 1–3 of PHEX gene in DNA of the male patient.
Analysis of deleted region of PHEX gene
In order to confirm that there is a deletion in PHEX gene, we tried to identify the actual deleted region using PCR from DNA of the male patient. The intervening sequence between SMS and PHEX is about 38 kb, and intron 3 of PHEX gene is about 29 kb. We prepared several pairs of primers covering these regions that were expected to amplify PCR products of about 200–400 bps (Fig. 2). PCR using these primer pairs indicated that the starting point of the deletion exists between DNA regions corresponding to the reverse primer for PCR product of 5c and the forward primer of 5d. Similarly, the deletion ends between DNA regions corresponding to the reverse primer for PCR product of 3c and the forward primer for PCR product of 3d. Genomic PCR from the brother using the forward primer for PCR product of 5c and the reverse primer for PCR product of 3d produced a PCR product of about 2600 bp (Fig. 1b). Direct sequencing of the PCR product showed that there is a 52 143 bp deletion including exon 1–3 in this patient. Because there are identical nucleotide sequence of seven bases in the 5′ and 3′ deleted regions (Fig. 2), it was impossible to accurately specify the starting and the ending nucleotide of the deletion.
The same deletion was observed in DNA from the proband (Fig. 1b). These results indicated that the male patient is hemizygous, while the female patient is heterozygous for the same deletion of PHEX gene and these patients are suffering from XLH.
However, it is considered to be quite unlikely that the same mutations develop de novo in two patients of this family. Therefore, we examined whether the mother of these patients has the same deletion of PHEX gene. While she showed no physical findings suggesting XLH, the same deletion was also observed in DNA obtained from the mother (Fig. 1b). Furthermore, semi-quantitative PCR suggested that the amount of the mutant allele in the mother is less than those of the patients (Fig. 3a). In addition, comparison of PHEX gene in the mother and her daughter indicated that the amount of wild-type allele corresponding to the deleted region (exons 2 and 3) in the daughter is less than that of the mother. In contrast, there is no difference in the amount of wild-type allele not affected by the mutation (exon 7) in these subjects (Fig. 3b). These results indicated that the ratio of the wild-type and the mutant allele is different between the mother and her children, even between the mother and the daughter. Because the daughter is considered to be a heterozygote, these results suggested that the mother is not a heterozygote for the mutation.
The above results suggested that the mother has more wild-type allele than the proband who is a heterozygote for the mutation. In order to evaluate this possibility, we examined SNPs existing in the deleted region of PHEX gene. It was found that there are two alleles in DNA from the mother in one of the eight SNPs (rs178712 T and C). These results indicate that there are two wild-type alleles and one deleted mutant allele in genomic DNA from the mother.
XLH is the most common cause of vitamin D-resistant hypophosphatemic rickets/osteomalacia. Since the identification of PHEX, more than 100 kinds of mutations in PEHX gene have been reported (10). Various mutations including missense mutations, insertions, and deletions have been reported in PHEX gene (11). Recent investigations have indicated that mutant PHEX in bone underlie the pathogenesis of XLH probably through enhancing the production of FGF23 (12, 13). We have identified a 52 143 bp deletion including exons 1–3 of PHEX gene. While a deletion containing exons 1–3 has been already reported in one family (14), the precise length and location of this deletion were not described. It is unclear whether the present deletion is the same as the previously reported one or not.
While the proband and her brother showed bowing of legs and had been treated with vitamin D metabolites under the diagnosis of hypophosphatemic rickets, the mother did not show any physical features of rickets. Her height is normal and her arm span is almost equal to her height. Her legs are straight without any treatment. Therefore, ARHR was suspected in this family. However, genetic analysis indicated that the mother also has the same deletion in the PHEX gene as that of her children. This indicates that the deletion is present in germ cells of the mother. Further analysis indicated that the mother is heterozygous for one SNP located in the deleted region of PHEX gene showing that she has one mutant allele and two kinds of wild-type alleles of PHEX gene. A possible explanation of this would be mosaicism developed in the early stage of the zygote. The lack of family history of rickets in the ancestors of the mother supports this possibility. Already, mosaicism of mutations in PHEX gene has been reported in several papers (15, 16). In one report, the mosaicism was described to cause a hypophosphatemic disease mimicking an autosomal dominant inheritance (15). Otherwise, she may be suffering from triple X syndrome. Because she declined further chromosomal analysis, we could not discriminate these two possibilities.
It is of interest that the mother with a deletion in PHEX gene has no physical signs of XLH. XLH is an X-linked dominant disease with almost complete penetrance (17). This means that loss of one allele of wild-type PHEX gene causes XLH phenotype. As shown in the table, the mother was mildly hypophosphatemic with slightly elevated FGF23. These results could be explained by a small number of mutant cells in bone, which resulted in marginal increase in FGF23 if she is mosaic. Because semi-quantitative PCR indicated that the amount of the mutant allele in the mother is less than that of the heterozygous proband, the percentage of the mutant cells would be <50%. If she is suffering from triple X syndrome, one-third of the cells are expected to express the mutant allele because of lyonization. In any case, these results suggest that there is a threshold in the amount of mutant PHEX allele to produce XLH phenotype.
While XLH is by far more frequent than ADHR or ARHR and patients with hypophosphatemic rickets with high FGF23 are most likely to be suffering from XLH (18), mutations in PHEX gene have not been detected in all clinically diagnosed patients with XLH (14, 19, 20, 21, 22). Usual direct sequencing of PCR products for all exons in PHEX gene failed to detect any mutation in the mother and the proband in this family; therefore, it would have been impossible to identify the deletion in this family if there were no male patients.
In conclusion, we have reported a family with hypophosphatemic rickets. While ARHR was clinically suspected, genetic analysis indicated there is a deletion in PHEX gene. Careful genetic analysis is mandatory for correct differential diagnosis of hypophosphatemic rickets with high FGF23 levels.
Declaration of interest
There is no conflict of interest that could be perceived as prejudicing the impartiality of the scientific work reported.
This study was supported in part by grants from the Ministry of Education, Culture, Sports, Science and Technology, and from the Ministry of Health, Labour and Welfare, Japan.
JonssonKBZahradnikRLarssonTWhiteKESugimotoTImanishiYYamamotoTHampsonGKoshiyamaHLjunggrenOObaKYangIMMiyauchiAEconsMJLavigneJJuppnerH. Fibroblast growth factor 23 in oncogenic osteomalacia and X-linked hypophosphatemia. New England Journal of Medicine20033481656–1663.
YamazakiYOkazakiRShibataMHasegawaYSatohKTajimaTTakeuchiYFujitaTNakaharaKYamashitaTFukumotoS. Increased circulatory level of biologically active full-length FGF-23 in patients with hypophosphatemic rickets/osteomalacia. Journal of Clinical Endocrinology and Metabolism2002874957–4960.
Lorenz-DepiereuxBBastepeMBenet-PagesAAmyereMWagenstallerJMuller-BarthUBadenhoopKKaiserSMRittmasterRSShlossbergAHOlivaresJLLorisCRamosFJGlorieuxFVikkulaMJuppnerHStromTM. DMP1 mutations in autosomal recessive hypophosphatemia implicate a bone matrix protein in the regulation of phosphate homeostasis. Nature Genetics2006381248–1250.
GojiKOzakiKSadewaAHNishioHMatsuoM. Somatic and germline mosaicism for a mutation of the PHEX gene can lead to genetic transmission of X-linked hypophosphatemic rickets that mimics an autosomal dominant trait. Journal of Clinical Endocrinology and Metabolism200691365–370.
RowePSOudetCLFrancisFSindingCPannetierSEconsMJStromTMMeitingerTGarabedianMDavidAMacherMAQuestiauxEPopowskaEPronickaEReadAPMokrzyckiAGlorieuxFHDreznerMKHanauerALehrachHGouldingJNO'RiordanJL. Distribution of mutations in the PEX gene in families with X-linked hypophosphataemic rickets (HYP). Human Molecular Genetics19976539–549.
FrancisFStromTMHennigSBoddrichALorenzBBrandauOMohnikeKLCagnoliMSteffensCKlagesSBorzymKPohlTOudetCEconsMJRowePSReinhardtRMeitingerTLehrachH. Genomic organization of the human PEX gene mutated in X-linked dominant hypophosphatemic rickets. Genome Research19977573–585.