Validation of the reproduction-related SNPs and expression differences between the testis and oviduct
Differential expressions identified in the RNAseq data were validated by RT-PCR of selected genes in the testis and oviduct tissue samples. For the validation, we selected DE genes associated with reproduction that contained nonsynonymous SNPs, because these genes represent potential candidate genes for an effect in reproduction traits. We analyzed the expression differences between the testis and oviduct samples for four genes (FRAS1, TCF4, ADAT1, and SPAG6) by qPCR and for two additional genes (PIWIL2 and DNAH8) by RT-PCR and an agarose gel. All genes showed a similar expression pattern in the RT-PCR analysis, as detected by RNAseq (Figure 5). Testis-specific genes ADAT1, SPAG6, PIWIL2, and DNAH8 exhibited none or extremely low expression in the oviduct. The genes with higher expression in the oviduct compared to the testis in the RNAseq data appeared to be present in the testis samples, but at a much lower level (Figure 5B). Furthermore, the polymorphisms within genes FRAS1, ADAT1, SPAG6, DNAH8, and PGR were confirmed by Sanger sequencing. Thus, these polymorphisms represent potential candidates for gene-assisted selection.
Figure 5  Differential gene expression of identified reproduction-related genes with polymorphisms in the testis and oviduct. (A) PIWIL2 and DNAH8 are predominantly expressed in the testis. (B) qPCR results of genes SPAG6, ADAT1, FRAS1, and TCF4 confirm identified expression differences between the testis and oviduct in the RNAseq data. A possible effect of the validated SNPs was analyzed using the SIFT and CFSSP prediction tools. All nonsynonymous SNPs were tolerated, but the SIFT score indicated a possible effect on protein function due to the level of conservation of the amino acid sequence at the SNP location for FRAS1, ADAT1, SPAG6, and PIWIL2 (SIFT score <0.35) (Table S8). The effect on protein secondary structure was also explored by CFSSP, which indicated a shift in the helix structure at the SNP position for SPAG6 and FRAS1 and the removal of a helix in PIWIL2. Although the effect of the identified SNPs on protein function and phenotypic differences is yet to be investigated, our data suggest possible causative mutations for differences in reproductive performance in the Large White pig population.