Allelic haplotype combinations at the region, including P-class pentatricopeptide repeat family genes, influence wide phenotypic variation in pollen grain number through a cytoplasmic male sterility model in citrus.

In citrus breeding programs, male sterility is an important trait for developing seedless varieties. Sterility associated with the male sterile cytoplasm of Kishu mandarin (Kishu-cytoplasm) has been proposed to fit the cytoplasmic male sterility (CMS) model. However, it remains undetermined whether CMS in citrus is controlled by interactions between sterile cytoplasm and nuclear restorer-of-fertility () genes. Accordingly, mechanisms underlying the control of the wide phenotypic variation in pollen number for breeding germplasm should be elucidated. This study aimed to identify complete linkage DNA markers responsible for male sterility at the region based on fine mapping. Two P-class pentatricopeptide repeat (PPR) family genes were identified as candidates for based on predicted mitochondrial localization and higher expression in a male fertile variety/selected strain than in a male sterile variety. Eleven haplotypes (HT1-HT11) at the region were defined based on genotyping of DNA markers. Association analysis of diplotypes at the region and the number of pollen grains per anther (NPG) in breeding germplasms harboring Kishu-cytoplasm revealed that the diplotypes in this region influenced NPG. Among these haplotypes, HT1 is a non-functional restorer-of-fertility () haplotype; HT2, a less-functional ; HT3-HT5 are semi-functional ; and HT6 and HT7 are functional . However, the rare haplotypes HT8-HT11 could not be characterized. Therefore, P-class PPR family genes in the region may constitute the nuclear genes within the CMS model, and a combination of the seven haplotypes could contribute to phenotypic variation in the NPG of breeding germplasms. These findings reveal the genomic mechanisms of CMS in citrus and will contribute to seedless citrus breeding programs by selecting candidate seedless seedlings using the DNA markers at the region.