Simple Summary
Myxomatous mitral valve disease (MMVD) is a common heart disease in dogs and shares similarities with mitral valve prolapse (MVP) in humans, particularly Barlow’s syndrome. However, the genetic and molecular mechanisms underlying these diseases are not yet fully understood. In this pilot study, whole-genome sequencing was performed on eight Maltese dogs with MMVD and two older Maltese dogs without MMVD, and variant frequencies in candidate genes were compared between the two groups. No variant remained significant after correction for multiple testing, which was expected given the small sample. The lowest uncorrected p-values were found in COL12A1 and COL1A2, which encode collagen proteins important for the structure of the extracellular matrix, and the canine candidate genes as a group showed slightly more differences than expected by chance. Genes linked to human MVP did not show such an excess. These candidate loci remain uncertain, and larger studies with balanced, sex-matched groups of affected and unaffected dogs are needed before they are tested.
Abstract
Myxomatous mitral valve disease (MMVD) is the most prevalent acquired cardiac disease in dogs and represents an important comparative model for human mitral valve prolapse (MVP). Although the structural pathology of MMVD is well described, its molecular and genetic underpinnings remain poorly understood. In this pilot study, we performed whole-genome sequencing (WGS) at approximately 30× coverage on eight MMVD-affected and two unaffected Maltese dogs. Variants were called following the GATK best-practices pipeline, and allele frequencies were compared between cases and controls with Fisher’s exact test across 67 canine MMVD candidate genes and 65 canine orthologs of human MVP genes. Multiple testing was assessed by false discovery rate and Bonferroni correction, and gene- and gene-set-level significance by permutation of case–control labels. No variant remained significant after correction; with 16 case and 4 control alleles, the smallest attainable p-value was 2.1 × 10−4. In the canine set, 17 genes had nominal p < 0.05, more than expected under permutation (empirical p = 0.044); the lowest p-values were in COL12A1 (p = 0.0031; empirical p = 0.022) and COL1A2 (p = 0.0035; empirical p = 0.044), both encoding collagens of the extracellular matrix (ECM). Several canine signals (CDK6, HEPACAM2 and VPS50) lay in a single haplotype block on chromosome 14, 1.2 Mb from COL1A2. The human MVP-orthologous set showed no excess of nominal genes (empirical p = 0.756). No close relatives or major population structure were detected. These pilot results point to collagen and ECM-related loci as uncertain candidates; larger studies with balanced, sex-matched case and control groups and older, echocardiographically confirmed controls are needed before these candidates are tested.