Novel compound heterozygous mutations in ENPP1 cause hypophosphataemic rickets with anterior spinal ligament ossification

Sir, Hypophosphataemic rickets (HR) is a rare disorder of renal phosphate wasting resulting from mutations in a phosphate-regulating gene with homologies to endopeptidases on the X chromosome (PHEX) in the majority of cases, fibroblast growth factor 23 (FGF23) or dentin matrix protein 1 (DMP1). In the past year, nine cases of autosomal-recessive HR have been reported with loss-of-function mutations in the ectonucleotide pyrophosphatase/phosphodiesterase 1 (ENPP1) gene [1–3]. Mutations in this gene are also associated with generalized arterial calcification of infancy (GACI), a devastating neonatal condition resulting in ectopic calcification in the vasculature [4]. We report the case of a 42-year-old Greek female who was referred with a 10-year history of progressive back stiffness and polyarthralgia. She had been diagnosed with HR at the age of 13 months when she developed bowing of her legs. She received vitamin D and occasional phosphate until the age of 20 years, and underwent femoral and tibial osteotomies as a teenager. She had an uneventful pregnancy at the age of 35 years. Her parents were non-consanguinous with no significant family history. Examination revealed short stature and severely restricted spinal movement. Radiographs demonstrated complete ossification of the anterior spinal ligament, iliosacral ligament calcification and Basstrup’s disease (Fig. 1A–C). Biochemical indices were consistent with HR, and fibroblast growth factor (FGF)-23 levels (post-treatment) were normal at 76 RU/l (<100). Calcitriol 1 μg/day rapidly improved her polyarthralgia; parathyroid hormone (PTH) corrected and other indices remained stable. A short trial of phosphate was discontinued due to intolerance and hyperparathyroidism. At 5-year follow-up spinal stiffness was unchanged, but there was radiological worsening of ligamentous ossification. Radiological appearance and predicted protein structure of ENPP1 (left) and ENPP1 T319R (right) phosphodiesterase region. (A) Dense ossification of the anterior longitudinal ligament. (B) Basstrup’s disease—contact of adjacent spinous processes (kissing spine) with resultant enlargement, flattening and reactive sclerosis of apposing interspinous surfaces. (C) The positions of residues 256 and 319 are indicated (the side chain of residue Thr256 forms a covalent intermediate with the AMP substrate, and is essential for catalytic activity). Substitution of arginine for threonine at position 319 (right) generates a patch of electrostatic charge (blue) that disrupts hydrogen bonding. Genetic screening failed to demonstrate a PHEX mutation using standard PCR conditions, exon-specific primers and automated sequencing (ABI3730; Life Technologies, Carlsbad, CA, USA; Mutation Surveyor; SoftGenetics, State College, PA, USA). PHEX dosage analysis by multiplex ligation-dependent probe amplification (MRC-Holland kit P223-B1; MRC-Holland, Amsterdam, The Netherlands) did not identify deletions or duplications. Sequence analysis of the coding and flanking intronic regions of the FGF23 and DMP1 genes did not identify any pathogenic mutations. Sequence analysis of the ENPP1 gene (25 exons on chromosome 6) identified two variants: a previously unreported missense variant, p.Thr319Arg (c.956C>G), in exon 9 and a heterozygous nonsense mutation, p.Arg782X (c.2344C>T), in exon 23. The p.Arg782X nonsense mutation results in a premature stop codon at 782 (p.Arg782X) and has been reported previously in a patient with GACI [5]. Parental saliva samples showed that the p.Thr319Arg was paternally inherited and p.Arg782X maternally inherited. The threonine 319 residue is highly conserved across species and the variant is predicted to be disease causing. In silico phenotypic prediction and 3D modelling was carried out by submitting the ENPP1 phosphodiesterase region (residues 191–591) to the Swiss-Model web server. All in silico phenotypic prediction tools reported the p.Thr319Arg substitution as damaging or not tolerated. Three-dimensional modelling based on template 3nkq chain A (mouse autotoxin) (Fig. 1D) showed that Thr319 lies on the surface, near the binding cleft for the adenosine monophosphate (AMP) substrate, and the side chain forms hydrogen bonds with the protein backbone. In the p.Thr319Arg substitution, the larger arginine side chain is predicted to disrupt the molecular surface and hydrogen bonding, generating a novel patch of positive electrostatic charge (Fig. 1D). Diseases caused by mutations in ENPP1 are perplexing. It is unclear what determines whether a patient develops GACI or HR. Interestingly, previous studies in GACI have found a variable inverse association between hypophosphataemia and aberrant calcification [4], suggesting that low phosphate may protect infants from ectopic vascular calcification. The phenotype does not appear to be related to the site of the mutation as the same mutation within a family can result in either condition [1]. Nor is it clear what determines the extent to which a patient with HR develops ligamentous ossification. Of the patients reported in the literature, only one had significant ossification, and this was in the posterior longitudinal ligament [3]. This phenotype resembles the ENPP1 knockout tiptoe walking mice that also develop intervertebral ankylosis and articular cartilage calcification [6]. In humans, ENPP1 polymorphisms have been associated with chondrocalcinosis [7] and posterior longitudinal ligament ossification [8]. Posterior longitudinal ligament ossification has been associated with X-linked HR [9], although we postulate that these patients may actually have ENPP1 mutations. Our patient is unusual in that she had marked anterior spinal ligament ossification with progressive back stiffness as her presenting symptom. Thus far, all described genetic associations of HR are thought to enhance the activity or levels of FGF-23, the key phosphaturic hormone. Although none of the described patients with ENPP1 mutations had high FGF-23 levels, these levels may be regarded as inappropriately high for the level of serum phosphate. There is no consensus for management of HR, but most clinicians stop treatment when growth has terminated. Phosphate or calcitriol may improve symptoms, but could exacerbate ligamentous ossification. It is unknown whether bisphosphonates (PPi analogues) would confer protection from or accelerate ligament ossification. Bisphosphonates have benefit in some patients with GACI [10]. Whether correction of the bone biochemistry in adults with HR improves or worsens ectopic calcification remains an important management challenge. Disclosure statement: The authors have declared no conflicts of interest.