Heteroplasmy and Patterns of Cytonuclear Linkage Disequilibrium in Wild Carrot

Organellar genomes are considered to be strictly uniparentally-inherited. Uniparental inheritance allows for cytonuclear coevolution and the development of highly coordinated cytonuclear interactions. Yet, instances of biparental inheritance have been documented across eukaryotes. Biparental inheritance in otherwise uniparentally-inherited organelles is termed leakage (maternal or paternal) and allows for the presence of multiple variants of the same organellar genome within an individual, called heteroplasmy. It is unclear what, if any, evolutionary consequences are placed on nuclear and/or organellar genomes due to heteroplasmy. One way of accessing cytonuclear interactions and potential coevolution is through calculating cytonuclear linkage disequilibrium (cnLD), or the non-random association of alleles between nuclear and organellar genomes. Patterns of cnLD can indicate positive or negative cytonuclear selection, coevolution between the nuclear and organellar genomes, non-traditional organellar inheritance, or instances of ancestral heteroplasmy. In plants, cytonuclear interactions have been shown to play a role in cytoplasmic male sterility which occurs in gynodioecious species and is associated with leakage. We used the gynodioecious species, Daucus carota L. spp. carota, or wild carrot, to investigate cnLD. We genotyped a total of 265 individuals from two regions of the USA at 15 nuclear microsatellites, the mitochondrial genes cox1 and atp9, and an intergenic region between trnS and trnG (StoG) in the plastid genome to calculate nuclear-nuclear LD (nucLD), cnLD, and organellar LD (i.e., within the mtDNA and between mtDNA and ptDNA) within the two regions. We were further able to identify cox1 and StoG heteroplasmy and calculate some of the same LD measures within heteroplasmic and homoplasmic (non-heteroplasmic) datasets. We used a Z-transformation test to demonstrate that heteroplasmic individuals display significantly higher levels of cnLD within both regions. In spite of this, within and between organellar LD is low to moderate. Given these patterns of LD in two regions of the USA in which gene flow has been shown to occur between crop and wild carrot, we suggest that heteroplasmy is an evolutionary mechanism which permits the maintenance of cnLD while also acting to disrupt organellar LD.