Our mutational analysis identified 15 different PHEX gene mutations in ten sporadic XLH cases and six XLH pedigrees (Fig. 6). There were 20 female and five male patients in this study. The sex ratio of four females to one male conforms to the mode of XLH transmission [7]. Although the 20 female patients were heterozygous and had one normal X chromosome, there were no significant differences between the two genders in terms of clinical and biochemical measurements. The phenotypes among family members of the same pedigree did not offer necessary connections. In families 2 and 3, “X”-type bending of the lower extremities was found in the sons, while “O”-type bending on lower extremities was found in their mothers. These conditions were consistent with a previous report [8]. Recently, a report observed that hyperparathyroidism might be a frequent complication in adult XLH patients [9]. However, we did not find a similar symptom in our preliminary testing; however, we will follow patients for this condition by long-term observation.
As of last year, 623 different mutations have been identified and recorded in the HGMD (2020.4 version). Although several reports have presented the statistics of the different mutations, there were no explicit hotspots confirmed. Variant c.1601C > T(p.Pro534Leu) was mentioned as a potential hotspot, as it was found in five unrelated patients in southern China, two patients in other areas of China, and two patients in other counties [10–15]. Three other mutations (Glu579Arg, Arg747*, c.1645 + 1G > A) might also represent hotspot mutations [13]. None of these mutations were detected in our study. However, another variant, c.2078G > A(p.Cys693Tyr), appeared twice in the current study. This specific mutation has also been reported in three previous studies, including in one Chinese patient [13, 16, 17]. Thus, it might represent a hotspot.
Missense and nonsense mutations, splicing mutations, regulatory mutations, small deletions, insertions, or indels, gross deletions, insertions or duplications, and complex rearrangements have been detected in the PHEX gene. Splicing mutations are the third most frequent type of mutation found in the PHEX gene. Splicing mutations have been found in every intron (Fig. 7). However, mutations have not been reported in the acceptor sites of introns 12 and 18. Splicing mutations in introns 3, 4, 10, 15, and 17 have been reported more frequently. Most splicing mutations occur within ten bases upstream or downstream of a splice site. Interestingly, a mutation located 73 bases upstream of the donor site of intron 5 and a mutation found 1268 bases downstream of the donor site of intron 7 could impact the recognition of the splice site [18, 19]. These mutations were detected in the RNA from patient blood samples; however, patient blood RNA samples are not always available. Therefore, we used a splicing reporter minigene assay in the current study. According to other studies, this approach is a relatively accurate method [20, 21]. In this study, we analyzed six splicing mutations using the minigene approach, including one small deletion in the donor site of intron 14 (c.1586_1586 + 1delAG). Coincidentally, the two bases before the deletion were also AG and recognized as an alternative donor site. Thus, we prefer to describe this mutation as c.1585_1586delGA (p.Glu529Valfs*52). In the other five splicing mutations, two variants (IVS14 + 4A > T and IVS20-1G > T) caused exon-skipping, resulting in a frameshift (p.Ile495Valfs*52 and p.Val691Glyfs*6). Two other variants (IVS15 + 1G > A and IVS5 + 1G > A) activated a preexisting pseudo splicing site within the exon, causing a frameshift (p.Arg549Hisfs*21 and p.Leu206Phefs*11). The variant IVS20 + 1G > A resulted in two outcomes. One resulted in the recognition of a new splice site in intron 20, and the other outcome was exon-skipping, which only accounted for a small portion comparing with the exon-skipping.
In previous studies, more than half of the identified mutations resulted in truncated proteins. In this study, all the splicing mutations caused frameshifts and truncated protein. However, no phenotypic differences have been observed between patients with truncating or non-truncating mutations except the condition that mutations in the first 649 amino acids led to earlier disease onset and higher iFGF23 levels [11, 13]. Our 3D homology modeling of the PHEX mutant proteins demonstrated a difference in the structures of wild-type and mutant PHEX proteins. Even the mutant proteins of IVS20 + 1G > A, IVS20-1G > T, c.1586_1586 + 1delAG, and c.1709_1710insTGAG(p.Tyr570Glufs*12), which contained most of the wild-type protein (i.e., they were not truncated by the mutation), were structurally different from the wild-type PHEX, including changes in the position of the helix and the length of the loop. These alterations might affect the exposure of the functional domain and cause loss of function.
Patients in our study have been treated at the time of diagnosis. Oral phosphate supplements and active vitamin D analogs (calcitriol) were gradually given for medical management in the first one or two months. Oral phosphate was administered four times per day, with the dosage varying from 15 to 100 mg/kg/day depending on individual patient requirements and the occurrence of phosphate-induced diarrhea. ALP was measured one hour after taking the phosphate to ensure its levels had returned to the normal range. It was also important to monitor PTH levels in the patients because they might increase when phosphate is used alone, or the calcitriol dosage used in combination is insufficient, leading to secondary hyperparathyroidism [22, 23]. We did not detect high PTH levels in our study. Because increased FGF23 levels are responsible for phosphate wasting, the development of the anti-FGF23 monoclonal antibody (burosumab) has been met with great expectations [24, 25]. Clinical trials for burosumab in children have received market approval by the Food and Drug Administration (FDA) and European Medicines Agency (EMA) and are underway [26–30]. A phase 3 trial evaluating the efficacy of burosumab in adult XLH patients is also being performed [31]. Thus far, the reported burosumab side effects are generally tolerable. The long-term effects require further investigation.