Multiple origin of organellar genomes of cultivated pea (Pisum sativum L. subsp. sativum)

The initial area and gene pool involved in plant domestication are a matter of ongoing debates. Earlier we found that pea cultivar Cameor had plastid and mitochondrial genomes related to wild peas of different provenance. We sequenced complete plastid and mitochondrial genomes from 27 accessions to compile a sample of 91 peas including 26 landraces of traditional cultivation. The vast majority of plastid genomes of cultivated peas tightly clustered and was closely related to wild peas primarily from Ponto-Caspian area and Zagros. However, two accessions from Central Asia showed affinity to a different wild pea lineage. Mitochondrial genomes of most cultivated peas were found in three clusters. Accessions most related to wild peas from the domestication ‘Core Area’ originated from periphery of traditional pea cultivation: Africa, Central Asia and Himalaya. Another cluster, related to wild peas from the Balkan Peninsula and Sicily, was present in Central Asia and Greece. Accessions most related to the cultivar Cameor were found throughout the pea cultivation range. We hypothesise that the pea cultivation area, initially occupied by peas domesticated in the ‘Core Area’, underwent two subsequent waves of invasion of cultivated peas with mitochondria introgressed from wild peas from elsewhere. One of the waves spread from South-East Europe and/or West Asia, and the second was associated with relatively recent expansion of crops of European origin. Mitochondrial genomes were supposed to introgress readily from wild to cultivated peas. Knowledge on diversity of organellar genomes of wild and cultivated peas may facilitate appropriate choice of co-adapted nuclear-organellar combinations for breeding programs.