Submission Details

The CPA6 gene was first reported in association with a human phenotype in an individual with Duane anomaly and a de novo balanced reciprocal translocation interrupting the coding sequence of this gene (Pizzuti et al., 2002, PMID: 12454025). Subsequent reports have suggested a relationship between CPA6 and both autosomal dominant and autosomal recessive epilepsy (MONDO:0005027). Autosomal recessive epilepsy is the focus of this curation.
CPA6 variants were initially reported in association with seizures by Salzmann et al. in 2012 (PMID: 23105115). This case report describes a consanguineous Moroccan family where p.A270V homozygotes presented with variable seizure phenotypes, including simple and complex febrile seizures and temporal lobe epilepsy. In this family, linkage analysis and autozygosity mapping identified a large region of linkage at 8q12.1-8q13.2 (maximum 2-point LOD score of 2.18). Sequencing of an additional 38 genes within the linkage region did not reveal any other likely causative variant. Although segregation was complete in unaffected and affected individuals, affected individuals presented with variable seizure and developmental phenotypes. Additionally, two individuals homozygous for this variant are present in the gnomAD population database (gnomAD v3.1.1); therefore, this variant was not scored. In an additional cohort of individuals with temporal lobe epilepsy screened for variants in CPA6, an individual carrying both p.G267R and p.Q207E was identified, but phase was not determined and these variants have been reported as occurring in cis in another individual in the literature (PMID: 26648591). An individual with intellectual disability and seizures has been reported with a homozygous T291A variant (PMID: 31130284). This variant is present at a low frequency in the gnomAD database and was scored 0.25. A number of the reported missense variants demonstrated significantly decreased CPA6 activity via expression studies in HEK293T cells; however similar reductions in activity were identified for missense variants too common in general population databases to be considered pathogenic, making the significance of these findings difficult to interpret (PubMed: 21922598, PubMed: 23105115, PubMed: 25875328). A zebrafish knockdown model demonstrated reduced response to proconvulsant stimulants despite the presumed loss of function disease mechanism in humans (PMID: 27050163). Additionally, the seizure, clinical and radiologic phenotypes associated with CPA6 variants are variable and CPA6 expression in brain tissue is weak. Case level data provides a score of 0.25 points and there is no scored experimental evidence, resulting in a total score of 0.25 points. In summary, there is convincing evidence disputing the relationship between CPA6 and autosomal recessive epilepsy, which significantly outweighs the evidence supporting the relationship. This classification was approved by the ClinGen Epilepsy GCEP on 5/4/21 (SOP Version 8).