M. Van Gils [1,2], J. Depauw , S. Verschuere [1,2], L. Nollet [1,2], O.M. Vanakker [1,2]
 Center For Medical Genetics, Ghent University Hospital;  Department Of Biomolecular Medicine, Ghent University
Introduction: Pseudoxanthoma elasticum (PXE) is an autosomal recessive soft tissue calcification disorder caused by biallelic mutations in ABCC6. In patients decreased levels of inorganic pyrophosphate (PPi) have been linked to the formation of ectopic mineralizations. However, no data has been reported on PPi levels in heterozygous carriers. Moreover, it is unknown whether a correlation exists between PPi levels and the ABCC6 genotype or the phenotype of patients.
Methods: A total of 127 citrated blood samples (91 from 62 PXE patients, 22 from 21 carriers and 14 from 14 controls) was collected and processed. Plasma PPi levels were measured via ATPase luminescence assays and samples were internally calibrated. The ABCC6 gene was screened for variants. The diagnosis of PXE was confirmed by the presence of 2 (likely) pathogenic variants (scored as C3 to C5 according to Verschuere et al. 2020) and/or histopathological changes. PXE phenotypes were assessed using the Phenodex scoring (PS) system at the time the first blood sample was obtained for a patient. Two sets of data were analyzed using SPSS26. First an analysis was performed where all available samples were considered [Bulk]. Concomitantly, a smaller subset (i.e. the samples obtained at Phenodex scoring [Single]) was analyzed. Differences in age, sex and PPi levels between the 3 cohorts were analyzed via ANOVA and T-Test. Correlations between PPi, PS, age and sex were analyzed via Spearman’s correlation analysis. After samples were subdivided based on 3 mutation archetypes: erroneous mRNA transcript (D: deletion, frameshift, splicing), nonsense (N) and missense (M) variants, genotype-PPi correlations were assessed via multiple ANOVA analyses based on more stringent variant selection (C3 - C5, C3 [Subset Likely Pathogenic] - C5, C4 - C5, C5).
Results: Compared to controls, patients and carriers had respectively ±49% and ±22% less PPi (P<0.001), though overlap in the sample ranges exists between the groups. Interestingly, PPi levels were lower in males compared to females in PXE patients (Bulk: P=0.011; Single: P=0.071), but not in the other cohorts. No significant differences in age between cohorts were observed; however, in the patient cohort PPi levels increased significantly with age (P=0.0015, r=0.279). Moreover, a significant inverse correlation between PPi levels and cardiac PS scores was identified (P=0.041, r=-0.263) as well as a weak inverse association between PPi and vascular PS (P=0.09). No correlation was found with skin or eye symptoms. A significant difference in PPi levels between D+M and N+M clusters was found during bulk, but not single, analysis (P<0.01).
Conclusion: PPi plasma levels are significantly decreased in patients and carriers and may thus also contribute to the subclinical or partial phenotype seen in the latter. Importantly, there is a region of overlap between both and with controls, suggesting PPi not to be the only relevant pro-mineralization factor in at least a subset of PXE patients. In PXE patients PPi levels appear to be influenced by sex and - to some extent unexpectedly - significantly increase with age. Whether there is a solid correlation with the ABCC6 genotype remains to be confirmed. Though unrelated to skin and eye disease, the PPi levels are inversely correlated with cardio(vascular) burden in PXE patients, making PPi promising as a cardiovascular biomarker in PXE.
Reference: Verschuere et al. 2020: PMID 32873932