Mitofusin 1 is essential to oocyte fertility, whereas mitofusin 2 is required in oocytes to eliminate mutant mitochondrial DNA in mice

Mitochondria are known to play a major role in providing oocytes with developmental competence, but the mechanisms underlying this function are elusive. Compared to other cell types in an organism, the oocyte has the greatest amount of mitochondrial DNA (mtDNA) and mitochondria that distinguish by their small size and rounded shape. These unique characteristics are dependent on events of mitochondrial fusion and fission, which are determinant to organelle function and inheritance. Thus, to address the role of mitochondrial fusion in oocytes, the genes Mfn1 and Mfn2 were conditionally knocked out by Zp3-directed expression of Cre recombinase. Towards this aim, a heteroplasmic mouse lineage containing ~50% of mutant mtDNA (NZB) was used. Wild-type (WT) or knockout oocytes for Mfn1 (Mfn1-null), Mfn2 (Mfn2-null) or both (Mfn-dm) were compared regarding their developmental competence, mitochondrial function and inheritance of NZB mtDNA. Data were evaluated by oneway ANOVA followed by Duncan posthoc test. Differences with P u003c 0.05 were considered significant. Mating of females containing Mfn1-null oocytes with WT males did not result in birth. Infertility in this case was associated with a failure in ovulation, accumulation of secondary follicles (251 ± 36 vs. 54 ± 12 follicles) and arrest of meiotic progression when oocytes were in vitro cultured. Mfn1-null oocytes also presented smaller diameter (57.8 ± 0.73 vs. 85.7 ± 0.49 μm) and lower levels of mtDNA (39,263 ± 2,878 vs. 198,164 ± 18,611 copies), ATP (0.03 ± 0.016 vs. 0.46 ± 0.076 pmol) and mitochondrial membrane potential (∆Ψm; 0.71 ± 0.03 vs. 1.00 ± 0.03 a.u.). The lower abundance of Alix protein in Mfn1-null oocytes, a known marker of extracellular vesicles, suggests that secretion of oocyte-derived paracrine factors was deficient. In regards to Mfn2-null oocytes, in spite of the smaller diameter (78.6 ± 0.68 μm) and lower ovulation rate (10.3 ± 4.29 vs. 23.0 ± 3.82 oocytes), these oocytes were not affected based on the level of mtDNA, ATP and ∆Ψm. As a result, the number of offspring born to Mfn2-null and WT oocytes did not differ. Moreover, the mice born to Mfn2-null oocytes grew normally, but had higher levels of blood glucose, both baseline (135 ± 3.9 vs. 115 ± 3.5 mg/dl) and fasted (172 ± 7.2 vs. 137 ± 4.1 mg/dl). Surprisingly, Mfndm oocytes presented a mild defect when compared to WT ones, characterized by slightly smaller diameter (76.7 ± 0.79 μm), intermediate level of mtDNA (104,215 ± 12,657 copies) and ∆Ψm (0.81 ± 0.02 a.u.), and normal level of ATP, ADP and ATP/ADP. However, Mfn-dm oocytes contained smaller and fewer mitochondria, which aggregated with the endoplasmic reticulum. In spite of being ovulated, Mfn-dm showed a disrupted meiotic spindle and an incompetence to develop to term. In respect to the mitochondrial inheritance, the level of NZB mtDNA decreased in WT (-10.1 ± 1.25) and Mfn1-null (-10.3 ± 1.58) oocytes in comparison to their donor females, suggesting a selection in the germline against mutant mtDNA. However, this selection was less effective in Mfn2-null (-3.0 ± 1.50) and Mfn-dm (-4.4 ± 1.21) oocytes. In summary, this work gives evidence that Mfn1 is essential to provide oocytes with developmental competence whereas Mfn2 is involved in the elimination of mutant mtDNA. These findings support a key role of mitofusins in adapting mitochondria to meet oocyte’s needs and prevent inheritance of mutant mtDNA.