On the website of the Kennell lab, this page on mitochondrial plasmids. The senDNAs may be relevant to the origin of the mitochondrial DNA. I have to look further into it, for instance the possibility that the mechanisms surrounding senescence are the reverse that happened in evolution.
In addition to mobile genetic elements associated with mtDNA, such as mobile introns and certain repetitive elements, extragenomic DNA molecules are frequently detected in fungal mitochondria. These elements can be divided into two groups; those that are derived from the mitochondrial genome via intra- or inter-molecular recombination and true plasmids that are autonomously-replicating genetic elements having little or no homology with the host genome. The former group has been intensely studied in the filamentous fungus Podospora anserina where it has been demonstrated that they are associated with a cell death phenomenon, called senescence. Called senDNAs (among other names), the closed circular, multimeric DNA molecules are derived from one of several regions of the mtDNA (reviewed by Griffiths 1992). Other well-studied examples of the formation of circular DNAs derived from mtDNA occur in Neurospora crassa (Bertrand et al. 1980; de Vries et al. 1986) and Aspergillus nidulans (Lazarus et al. 1980). Called “stoppers” or “ragged,” respectively, these DNAs are found in certain growth-defective mutants and show similarities to some of the senDNAs.