Mitochondrial replacement: Study identifies challenge for effectiveness of potential therapies

A potential hurdle to the effectiveness of mitochondrial replacement therapy, which some people refer to as ‘3-person IVF’, has been reported in Cell Stem Cell. In mitochondrial replacement, when nuclear DNA from the egg cell of one woman is transferred into an egg cell from another woman, a tiny amount of mitochondrial DNA can be carried over into the donor cell. The researchers showed that the amount of carryover DNA can occasionally increase over time, which could pose a problem for clinical therapies aiming to completely remove faulty mitochondrial DNA.


Dr. Marni Falk, Assistant Professor of Pediatrics, Division of Human Genetics, The Children’s Hospital of Philadelphia (webpage):

Expertise: Mitochondrial disease; genetic basis of mendelian disease; C. elegans models of mitochondrial disease and therapy.

“I think the paper is important. It’s addressing a question that has come up because of the whole issue of mitochondrial replacement techniques, and is an important effort to evaluate some of the basic scientific principles. I think one of the important concepts to highlight is that this paper doesn’t actually test any mitochondrial disease or DNA mutation. Instead, it’s asking whether if you have two separate mitochondrial DNA (mtDNA) haplogroups (mitochondrial genomes from different women), how likely it will  be that mixing haplogroups will influence the function of the cell, regardless of any disease causing mutations.

“In mitochondrial replacement, you basically have a donor egg which has the mtDNA that you want, and then you have the nuclear DNA that you want from the intended parent(s). It is known that you get between 0.2 and 2% carryover of mtDNA depending on the specific technique used. It is not known how important it is that the mtDNA haplogroups of the donor and parent mitochondria should match, or whether the carryover mtDNA eventually goes away or not.

“The paper is reassuring in that it shows there’s no obvious functional effect from haplogroup mixing. The paper also shows that the amount of the carryover mtDNA from the intended parent’s oocyte is pretty minimal for the most part, but in some cases it can become highly enriched. The ultimate goal is to have none of the carryover mtDNA, or so little that it would have no functional impact. We already knew that enrichment of mutated mtDNA could happen; I published a case report* where a child born with severe mitochondrial disease had a very high mtDNA mutation load in his muscle and blood, while his grandmother only had it in 1% of her blood. So we know this is inherently a very unstable and potentially unpredictable process.

“I think papers like this speak to the fact that if you really are trying to remove mutated mitochondria altogether you may need to do more technical manipulations, because just relying on low-carryover may not be enough to reliably reduce the risk of mtDNA disease developing in the child born following this procedure. People are talking about whether tools like CRISPR-Cas9 or mitoTALENs could be used to more reliably remove the 1% mtDNA carryover, but that causes concerns about increasing the number of germ cell or early zygote manipulations. Currently, none of these techniques are clinically allowable in the US or even being considered by the FDA for evaluation in clinical trials.”

*’Mitochondrial tRNA-serine (AGY) m.C12264T mutation causes severe multisystem disease with cataracts


Dr. Eric Shoubridge, Principal Investigator, Departments of Human Genetics and Neurology and Neurosurgery, McGill University (webpage):

Expertise: Molecular genetics of mitochondrial diseases, in particular those that affect the function of the respiratory chain.

“In the study by Yamada et al, nuclear genomes were transferred between human oocytes (egg cells) containing different mitochondrial DNA haplotypes (genes inherited from one parent).  The oocytes were stimulated to replicate without fertilization, and stem cell lines were derived from the resulting blastocysts. In some of the cell lines there was a shift toward the mitochondrial DNA reverting to the haplotype of the donor, due to some mitochondrial DNA being ‘carried over’ with the nuclear genome. This shows that even small amounts of carryover can have an influence.

“It is important to sort out potential pitfalls inherent in mitochondrial replacement therapy before it is applied clinically, and this study goes part way there. However, the fact that they observed genetic drift in cells undergoing a large number of divisions is not unexpected. All else being equal, the rate of genetic drift will depend on the the number of copies of mtDNA in cells, and the rate of cell division. In fact, you would predict this is simply a population genetics problem, where the proportion of cells with a rare haplotype should be the same proportion as the initial frequency of the haplotype in the initial cell. This has actually been demonstrated in an animal model in cells**.

“The implications for the clinic are simply that one should proceed with caution, but a limitation of the study is that it was carried out in artificially produced cell lines, and not in normally developing embryos. Further studies are required in developing zygotes post-fertilization to see whether mitochondrial DNA carryover is a problem in the actual embryo.”

**’Tissue-specific selection for different mtDNA genotypes in heteroplasmic mice


Declared interests (see GENeS register of interests policy):

None declared.



Genetic Drift Can Compromise MitochondrialReplacement by Nuclear Transfer in Human Oocytes‘ by Yamada et al, published in Cell Stem Cell on Thursday, May 19, 2016.


Please feel free to leave your comments below, but be aware that by doing so you agree to our Terms & Conditions.