Keeping in Shape the Dogma of Mitochondrial DNA Maternal Inheritance

It is textbook knowledge that the small multicopy mitochondrial genome (mtDNA) is maternally inherited in humans and mammals [1,2]. The uniparental mtDNA inheritance applies to most eukaryotic organisms, including animals exhibiting the doubly uniparental inheritance, such as the bivalve mollusks [3,4]. Occurrence of paternal mtDNA transmission has also been documented [5–7], and doubts on strict maternal inheritance in humans have been raised [8,9]. The best-documented case of paternal mtDNA inheritance was in a patient carrying a pathogenic mtDNA mutation [9], never replicated in following studies of patients with mitochondrial diseases due to various mtDNA defects [10–12]. The sperm mitochondria enter the oocyte during fertilization in mammals [13], but paternal mitochondria and mtDNA disappear at the initial cell divisions of the embryo in a stringently species-specific fashion [14]. In fact, the failure to efficiently eliminate paternal mtDNA from different species intercrosses [14,15] explains some of the cases of paternally inherited mtDNA [5]. Furthermore, recognition and targeted elimination of exogenous mtDNA entering the oocyte seems restricted to sperm mtDNA, not occurring with liver mtDNA, thus also displaying tissue specificity [16]. The way by which paternal mtDNA inheritance fails to occur in humans remains elusive, and it appears that several mechanisms have coevolved to avoid paternal mtDNA contribution to the embryo [17]. It has been observed that sperm mitochondria are ubiquitinated, suggestive of an “active elimination model” for paternal mtDNA [14], which may occur through different routes, such as proteosomal or lysosomal pathways [14,17]. Autophagy has been recently highlighted as the mechanism for paternal mtDNA elimination in Caenorhabditis elegans [18,19]. This was not observed in mice, for which elimination of mtDNA from prefertilization sperm and uneven persistence of paternal mtDNA in the embryo raised the possibility of a passive “dilution model” of disproportionate paternal versus maternal mtDNAs in mammals [20]. The consequent leakage of paternal mtDNA in the newborn may have remained “undetected” by the standard sequencing approaches.