Abstracts

Rationale: Protocadherin-19 (PCDH19) encodes a transmembrane protein with six cadherin domains, which creates an extensive adhesive interface. Over 100 pathogenic variants have been identified in patients with a female-restricted form of infantile-onset epilepsy. In contrast to typical X-linked diseases, hemizygous males are spared. Our aim was to characterize structural determinants associated with PCDH19-related epilepsy variants. Methods: Variants were ascertained from the literature with the search parameters “PCDH19" AND “Epilepsy” (PubMed, accessed May 2018). Subsequently, we assembled a 3-dimensional structural model of the PCDH19 protein, mapped all reported patient-associated missense PCDH19 variants on the model using PyMOL, and compared the spatial distribution of variants in patients to those reported in the population database (gnomAD). Furthermore, we determined whether the patient variants affect buried amino acids in the core of the protein or amino acids at the surface. Finally, we performed an enrichment analysis comparing amino acid positions with patient vs. population missense variants with respect to the predicted functional domains of the protein model, such as the calcium-binding sites. Results: We mapped 106 missense variants reported in literature to 77 locations on the 3-dimensional structural protein model. We show that the missense patient variants are present in distinct regions from population variants (p<0.0001, Fisher’s exact test). Moreover, 47/77 (61%) of the missense patient variants are enriched in regions where control individuals from gnomAD show a paucity of variants (p < 0.0001, Fisher’s exact test). Among those variants, 62% (29/47) are enriched in calcium binding sites in between the cadherin domains (p = 0.038, Fisher’s exact test). Conclusions: Mapping all known disease-associated variants on a computationally derived structural model of PCDH19 reveals that missense variants in PCDH19-related epilepsy cluster at regions distinct from those where general population variants are located, between the cadherin domains at the adhesive interface of the protein. The present results provide further support to the idea that it is the abnormal binding of these cadherin domains between cells that mediate at least some of the neurological dysfunction seen in patients. We thus demonstrate that mapping genetic variants onto a 3-dimensional protein structure can facilitate variant interpretation and also provide molecular insight into the pathogenesis of disease. Funding: Dr. Anne Rochtus is supported by a Fellowship of the Belgian American Educational Foundation and by a Fulbright Program grant sponsored by the Bureau of Educational and Cultural Affairs of the United States Department of State and administered by the Institute of International Education. Dr. Annapurna Poduri is supported by the Boston Children’s Hospital Translational Research Program.