Diploid aposporous sunflower forms triploid BIII progeny displaying increased apospory levels and non-random genetic mutations

Apomixis, i.e. asexual reproduction via seeds, could represent a revolution for sunflower breeding. In previous work, we identified a diploid sunflower inbred line (Rf975) naturally displaying aposporous-like embryo sacs (AES-like). Here, we explored the nature (reduced vs. unreduced) and the viability of these AES-like gametophytes by studying the formation of triploid (3x) BIII hybrids (2n + n) in the progeny of Rf975. Flow cytometry of immature seeds revealed that a considerable proportion (42.8 %) of the self-pollination Rf975 progeny were triploids. However, only 36.6% of these triploids reached maturity. Reproductive characterization revealed that 100% of the analyzed triploids displayed AES, and that the average proportion of ovules with AES was 61.9%, a value significantly higher than that detected in line Rf975 (4.6%). Abnormal pollen grains and scarce viable seeds were also detected. A segregant F2 progeny (including diploid and triploid individuals) was obtained by crossing Rf975 and HA89 (a genetically divergent sexual diploid inbred line). Methylation content sensitive enzyme ddRAD (MCSeEd) analysis of the diploid F2 did not reveal individuals with the F1 genetic constitution, proving that aposporous Rf975 is not capable of parthenogenesis. Moreover, both genetic and epigenetic non-random changes were detected in the F2 triploids, in specific genomic loci. Our results indicate that the AES-like gametophytes detected in Rf975 are supernumerary non-reduced embryo sacs unable to carry on parthenogenesis. Moreover, we discovered non-random (epi)genetic changes occurring in sunflower during triploidization events. This research will contribute to the development of essential knowledge and tools to enable the future implementation of apomixis-based breeding programs for this crop.