Association between polymorphisms in the SOX9 region and canine disorder of sex development (78,XX; SRY-negative) revisited in a multibreed case-control study

Testicular or ovotesticular disorders of sex development (DSD) in individuals with female karyotype (XX) lacking the SRY gene has been observed in several mammalian species, including dogs. A genetic background for this abnormality has been extensively sought, and the region harboring the SOX9 gene has often been considered key in canine DSD. Three types of polymorphism have been studied in this region to date: a) copy number variation (CNV) in a region about 400 kb upstream of SOX9, named CNVR1; b) duplication of SOX9; and c) insertion of a single G-nucleotide (rs852549625) approximately 2.2 Mb upstream of SOX9. The aim of this study was thus to comprehensively analyze these polymorphisms in a large multibreed case-control cohort containing 45 XX DSD dogs, representing 23 breeds. The control set contained 57 fertile females. Droplet digital PCR (ddPCR) was used to study CNVR1 and the duplication of SOX9. Fluorescent in situ hybridization (FISH) was used to visualize copy numbers on a cellular level. The Sanger sequencing approach was performed to analyze the region harboring the G-insertion. We confirmed that CNVR1 is highly polymorphic and that copy numbers varied between 0 and 7 in the case and control cohorts. Interestingly, the number of copies was significantly higher (P = 0.038) in XX DSD dogs (mean = 2.7) than in the control females (mean = 2.0) but not in all studied breeds. Duplication of the SOX9 gene was noted only in a single XX DSD dog (an American Bully), which had three copies of SOX9. Distribution of the G-nucleotide insertion was similar in the XX DSD (frequency 0.20) and control (frequency 0.14) cohorts. Concluding, our study showed that CNVR1, located upstream of SOX9, is associated with the XX DSD phenotype, though in a breed-specific manner. Duplication of the SOX9 gene is a rare cause of this disorder in dogs. Moreover, we did not observe any association of G-insertion with the DSD phenotype. We assume that the genetic background of XX DSD can be different in certain breeds