In reply

In their letter, Attardi et al. claim that the observation of inter-mitochondrial complementation by Ono et al.1 is a rare phenomenon and cannot be generalized, particularly to an in vivo system.

However, using mito-mice carrying exogenously-introduced mutant mtDNA with a deletion of 4,696 bp (ΔmtDNA4696)2, Nakada et al.3 recently reported unambiguous evidence for the presence of extensive in vivo inter-mitochondrial complementation in all tissues examined: all mitochondria in tissues with ΔmtDNA4696 showed normal COX activity until it accumulated, preventing those mice from expressing disease phenotypes. Moreover, coexistence of COX-positive and -negative mitochondria within single cells was not observed. Therefore, these results suggest the occurrence of in vivo inter-mitochondrial complementation by the exchange of mitochondrial contents between exogenously introduced mitochondria with ΔmtDNA4696 and host mitochondria with normal mtDNA.

They also claim that the observations of Ono et al.1 are in striking contrast with their previous observations4,5, which indicated absence of inter-mitochondrial complementation, and suggest that nuclear background would be responsible for these discrepancies.

Their previous observations, however, do not necessarily prove the absence of inter-mitochondrial complementation. In the paper by Enriquez et al.4, for example, the authors fused respiration-deficient cells carrying a mutation in MTTK (also known as MERRF) and respiration-deficient cells carrying a mutation in NDUFA4 (also known as ND4), and showed that very small numbers of colonies with restored respiratory function grew in a medium that selects for respiration competence. To prove that complementation is a rare event, however, one must show that there is no increase in the frequency of transcomplementing clones even after 10–14 days of growth in nonselective medium. This evidence must be obtained before suggesting the involvement of nuclear factors.

On the other hand, the paper by Yoneda et al.5 showed that completely respiration-deficient cells without mitochondrial translation activity were obtained by the fusion of parent cells carrying a mutation in MELAS (with 60% mitochondrial translation activity) and parent cells carrying an MTTK mutation (with about 5% mitochondrial translation activity). However, if there was no interaction between these mitochondria, fused cells with both parental mutant mtDNAs should have 5–60% of normal mitochondrial translation activity, but they did not. These observations should therefore be interpreted as showing that parental mitochondria with 60% activity and those with 5% activity fused to produce no activity, suggesting interaction between mitochondria that resulted in complete inhibition of mitochondrial translation activity.

Attardi et al. also claim that the small fraction of transcomplementing clones would not increase in number, even by extending the period of growth in non-selective medium to 10–14 days after fusion, as they could not observe an increase after 6 days of growth in this medium. This is not correct. As we outlined in Table 2 of our paper1, no transcomplementation was observed for 7 days after fusion, with an additional 4–7 days being critical for the restoration of respiratory function.

Based on the in vitro1 and in vivo3 transcomplementation of mitochondria, we recently proposed a new hypothesis, the interaction theory of mammalian mitochondria6. In this hypothesis, we suggest that the presence of intermitochondrial cooperation would rescue aged tissues, with various kinds of somatic mutant mtDNAs, from age-associated mitochondrial dysfunction.

See “Inter-mitochondrial complementation of mtDNA mutations and nuclear context” by Attardi et al.