Intracellular and horizontal transfer of mitochondrial genes in grass evolution: Pseudogenes, retroprocessing and chimeric genes

As an energy powerhouse in all animal and plant cells, the mitochondrion was at one time a free-living bacterium until recruited by a nucleated cell to co-exist symbiotically. Following this process, the mitochondrion has since surrendered its cellular autonomy by giving up its genes to be encoded in the nucleus of the host cell. These transfers of genetic materials have essentially ceased in animal cells but are ongoing in flowering plants. While the complex pathways of mitochondrial gene transfer have been well studied in an array of plants, there are remaining questions regarding the actual timing, frequency and mechanism of this process. In this dissertation, I investigated mitochondrial gene transfer in the grass family. Two tightly linked mitochondrial genes, rps14 and rpl5, were found to have different gene transfer histories. While rps14 was functionally transferred to the nucleus very early in grass evolution, a transcribed and edited rps14 pseudogene continues to be in the mitochondrion for over 80 million years. In contrast, rpl5 underwent a retroprocessing early in grass evolution that allowed the gene to bypass critical steps required for functional mitochondrial gene transfer and led to several separate transfers to the nucleus in the grass family alone. Rpl5 and rps14 was also found to have been horizontally transferred in the grasses. My results indicate that even short fragments of mitochondrial DNA can be horizontally transferred from one grass to another, resulting in chimeric genes that contain segments of foreign genes embedded within native genes. My findings suggest that mitochondrial genes can survive as pseudogenes for millions of years, undergo repeated transfers to the nucleus, and are horizontally transferred even more frequently in plant mitochondrial genomes than was previously suspected.