Mitochondrial replacement therapy

Mitochondrial replacement therapy is a group of in vitro fertilisation techniques that combine the nuclear DNA of a prospective mother with the cytoplasm, and therefore the mitochondria, of a donor egg. The purpose is to prevent a child from inheriting a pathogenic mutation in mitochondrial DNA. It is the only form of heritable human genetic modification that any country has legalised, and the number of children born from it worldwide remains small.

§The disease it addresses

Mitochondria carry their own small circular genome of about 16,500 base pairs encoding thirteen proteins of the respiratory chain plus the RNAs needed to make them. It is inherited almost exclusively from the egg. Mutations in it cause a heterogeneous group of disorders, among them Leigh syndrome, MELAS, MERRF and Leber's hereditary optic neuropathy. These typically strike tissues with high energy demand: brain, heart, skeletal muscle, retina.

Two features complicate inheritance. Cells carry many mitochondrial genomes, so a person can be heteroplasmic, carrying a mixture of normal and mutant copies. And most mitochondrial diseases show a threshold effect: symptoms appear only above a high mutant fraction, often well over half. During oogenesis a genetic bottleneck randomly samples a small number of mitochondrial genomes to seed each egg, so a heteroplasmic mother's eggs vary widely and unpredictably in mutant load. She may have several children with very different outcomes.

The bottleneck also explains why replacement is not always the right answer. Because eggs differ so much, Embryo selection for low heteroplasmy is often sufficient and is technically far simpler. Replacement is aimed principally at women who are homoplasmic for a pathogenic variant, where every egg carries it, and at those whose embryos all exceed the threshold.

§How it works

In maternal spindle transfer, the chromosome-bearing spindle is removed from the patient's unfertilised egg with a micropipette and inserted into a donor egg whose own spindle has been removed. The reconstructed egg is then fertilised. In pronuclear transfer, both eggs are fertilised first and the two pronuclei are moved from the patient's zygote into an enucleated donor zygote. A third approach, polar body transfer, uses the chromosomes discarded during meiosis and carries less cytoplasm, but has been used less.

All of these are variants of the micromanipulation developed for somatic cell nuclear transfer, the technique behind Human cloning, and they inherit its practical difficulties: the spindle is invisible without polarised optics, and the manipulation must not activate the egg prematurely. Some cytoplasm inevitably travels with the nuclear material, which is the source of the technique's central problem.

§Development history

1. 1997–2001
   Ooplasmic transfer
   A New Jersey clinic injects donor cytoplasm into eggs of women with repeated IVF failure. Around seventeen children are born before the FDA requires an investigational new drug application, halting the practice.
2. 2009
   Spindle transfer in monkeys
   Shoukhrat Mitalipov's group reports healthy macaques born after maternal spindle transfer, the first primate demonstration.
3. 2010
   Pronuclear transfer in human embryos
   A Newcastle team shows the technique works in abnormally fertilised human zygotes with low mitochondrial carryover.
4. 2015
   The UK legalises it
   Parliament approves regulations permitting mitochondrial donation under licence, the first statutory authorisation of heritable modification anywhere.
5. 2016
   First reported birth
   A US clinician performs spindle transfer in Mexico for a couple with Leigh syndrome; the FDA later issues a warning letter.
6. 2022
   Australia follows
   Maeve's Law permits mitochondrial donation in a staged research and clinical pathway.
7. 2025
   The UK cohort reports
   The Newcastle programme publishes outcomes for a small number of children born after pronuclear transfer, with low but not always undetectable mitochondrial carryover.

The scientific groundwork was laid in two places. Shoukhrat Mitalipov's laboratory in Oregon showed that spindle transfer produced healthy macaques,¹ and a Newcastle group led by Doug Turnbull and Mary Herbert established that pronuclear transfer worked in human zygotes with low carryover of the original mitochondria.²

The United Kingdom then took the deliberate route. After several public consultations and reviews of safety by the fertility regulator, Parliament approved regulations in 2015 under the existing Human Fertilisation and Embryology Act; the Newcastle Fertility Centre received the first licence in 2017 and the first patient approvals followed. In 2025 the Newcastle group published outcomes for the children born under the programme: eight births, with the mother's pathogenic mitochondrial DNA undetectable in most of the children and present at low levels in the rest, in every case below the level associated with disease.³ The same report described a larger number of births achieved by preimplantation testing rather than by donation, consistent with the pathway's design, in which selection among embryos is attempted first and donation reserved for the women it cannot help. None of the children had developed a mitochondrial disease phenotype at the time of reporting, though the follow-up was short and, as with every heritable intervention, the informative period runs into the next generation.

The unregulated route ran in parallel. The 2016 Mexico birth was performed by a US-based clinician outside American jurisdiction and reported afterwards.⁴ Clinics in Ukraine and Greece subsequently used nuclear transfer techniques not for mitochondrial disease but for infertility, on the hypothesis that donor cytoplasm improves poor-quality eggs. That indication is not supported by controlled evidence, and it accounts for a substantial share of the children born from the technique worldwide. The pattern, of a procedure debated for a decade in one country and sold commercially in another, is the same jurisdictional arbitrage that later framed the reaction to the He Jiankui affair, with the important difference that mitochondrial donation had a legitimate clinical rationale and a legal home.

§The reversion problem

Carryover of a few percent of the mother's mitochondria is unavoidable. If those genomes simply persisted at that level, the technique would be safe. They do not always.

Embryonic stem cell lines derived from replacement embryos showed that in a fraction of lines the carried-over genotype expanded until it dominated, sometimes completely.⁵ The mechanism is not settled; proposed explanations include replicative advantage conferred by particular sequence variants in the mitochondrial control region, and interaction between the donor mitochondrial haplotype and the patient's nuclear genome. One response has been to match donor and recipient mitochondrial haplogroups, which is intuitively appealing and not clearly supported.

The nuclear-mitochondrial interaction question is broader than reversion. Mouse experiments that placed one nuclear genome against a mismatched mitochondrial background found effects on metabolism, ageing markers and lifespan, suggesting the two genomes are co-adapted in ways that a transfer disrupts.⁶ Whether this matters at the magnitude of variation seen between human haplogroups is unresolved, and it is one of the main scientific reasons for continued caution about the technique.

Why "three-parent baby" misleads in both directions

The donor contributes 37 genes out of roughly 20,000, none of them influencing the traits people associate with parenthood, so the phrase overstates the biological relationship. But it understates the legal point: the modification is heritable, since a girl born after the procedure passes the donor mitochondria to her own children. That is precisely why it required primary legislation.

§Regulation

The United Kingdom and Australia have explicit statutory frameworks with case-by-case licensing. Most other countries have neither authorised nor specifically prohibited the technique, leaving it to general rules on embryo research.

The United States has a distinctive blockage. Since late 2015, an annual appropriations rider has barred the Food and Drug Administration from acknowledging any application for a clinical investigation in which a human embryo is intentionally modified to include a heritable genetic modification. Because the FDA asserts jurisdiction over the technique, the rider functions as a prohibition without Congress ever having debated mitochondrial donation on its merits. A National Academies committee had earlier recommended a cautious path forward, restricted initially to male embryos so that any error would not be transmitted; the rider superseded the recommendation.

The wider pattern is the one described in Governance of human genome editing: a technology permitted under supervision in a few jurisdictions, prohibited by side-effect in another, and available commercially in several with little oversight. Mitochondrial donation is the clearest existing test of whether the line drawn around heritable modification in Human germline editing debates is a line about mechanism or about consequence. It crosses the germline while touching no nuclear gene, which is why some commentators treat it as an exception and others as a precedent.

Ethically it has been comparatively uncontroversial, which is itself informative. It prevents a severe, often fatal childhood disease; it is not an enhancement in any sense recognised by Bioethics of enhancement; and it introduces no allele that a healthy person does not already have. The objections that were raised concerned donor identity and the child's interest in knowing about the procedure, questions familiar from gamete donation, together with the disability-rights argument set out in Disability rights and enhancement that preventing the birth of children with a condition carries an implicit judgement about people living with it. Because eligible families are rare and the procedure requires a specialist licensed centre, the Access and inequality questions that dominate other reproductive technologies have played a smaller part.

§Alternatives and outlook

Two technical alternatives could make the transfer approach obsolete. Editing mitochondrial DNA directly has become possible with cytosine base editors derived from a bacterial toxin that acts on double-stranded DNA, since conventional CRISPR guide RNAs cannot be imported into mitochondria; the chemistry is described under Base editing and has been demonstrated in cells and in mice, not in a clinical setting. Such editors could in principle shift heteroplasmy below the disease threshold without any donor egg. Separately, In vitro gametogenesis from reprogrammed somatic cells would permit gametes to be generated and screened in quantity, though it would not by itself remove a homoplasmic mutation.

For the disease itself, the fundamental question is unchanged by any of this. Mitochondrial dysfunction is also a feature of ordinary ageing, one of the Hallmarks of aging and the subject of Mitochondrial dysfunction in aging, and no intervention yet restores function to a tissue that has already lost it. Replacement therapy prevents transmission. It does nothing for the patients who prompted its development.

§See also

• Embryo selection
• Human germline editing
• In vitro gametogenesis
• Mitochondrial dysfunction in aging
• Governance of human genome editing
• Human cloning
• Designer babies
• Reproductive longevity

§References