Reply to Annis et al.: Is quasi-Mendelian mtDNA competition enough to drive transmission of paternal mtDNA?

Mechanisms for sperm mitochondrial removal in embryos

BACKGROUND

Different animal species have different characteristics regarding the transmission of mitochondrial DNA. While some species have biparental mitochondrial inheritance, others have developed pathways to remove paternal mtDNA. These pathways guarantee the uniparental mitochondrial inheritance, so far well known in mammals, avoiding heteroplasmy, which may have the potential to cause certain mitochondrial diseases in the offspring.

SCOPE OF REVIEW

This review aims to address the main mechanisms that involve mitochondrial degradation in different animal species, as well as to describe what is present in the literature on the mechanisms involved in mitochondrial inheritance.

MAJOR CONCLUSIONS

Two theories are proposed to explain the uniparental inheritance of mtDNA: (i) active degradation, where mechanisms for paternal mitochondrial DNA elimination involve mitochondrial degradation pathway by autophagy and, in some species, may also involve the endocytic degradation pathway; and (ii) passive dilution, where the paternal mitochondria are diluted in the cells of the embryo according to cell division, until becoming undetectable.

GENERAL SIGNIFICANCE

This work brings a wide review of the already published evidence on mitochondrial inheritance in the animal kingdom and the possible mechanisms to mtDNA transmission already described in literature.

Extraordinary claims require extraordinary evidence in asserted mtDNA biparental inheritance

A breakthrough article published in PNAS by Luo et al. challenges a central dogma in biology which states that the mitochondrial DNA (mtDNA) in humans is inherited exclusively from the mother. We re-analyzed original FASTQ files and results reported by Luo et al. to investigate methodological issues (e.g. nuclear mitochondrial DNA or NUMTs, DNA rearrangements) that could lead to biological misinterpretations. A comprehensive analysis of their data reveals several methodological and analytical issues that must be carefully addressed before challenging the current paradigm. We first show that the probability of the findings described by the authors is extremely small (most likely below 10^-37). The sequencing replicates from the same donors show aberrations in the variants detected that need further investigation to exclude contributions from other sources or methodological artifacts. Applying the principle of reductio ad absurdum, we demonstrate that the nuclear factor invoked by the authors to explain the phenomenon would need to be extraordinarily complex and precise to preclude linear accumulation of mtDNA lineages across generations, which would make the appearance of mixed haplotypes a much more frequent event in the population. We discuss alternate scenarios that explain findings of the same nature as reported by Luo et al., in the context of in-vitro fertilization and therapeutic mtDNA replacement ooplasmic transplantation.