select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to study looking at mitochondrial DNA and the nuclear genome

Research published in Science shows that mitochondria interact with cell nucleus’ in ways previously unseen in humans. Matching mitochondrial DNA to nuclear DNA could be important when selecting potential donors for the recently-approved mitochondrial donation treatment.


Prof Peter Braude, Emeritus Professor of Obstetrics and Gynaecology, King’s College London, said:

“This complicated and detailed scientific study examines the transmission of mitochondrial DNA between generations by the study of nearly 13,000 whole genome sequences from over 1,500 mother-child pairs obtained from the NIHR BioResource (rare diseases) and the 100,000-Genome Project database.

“From this database examination, it seems that the inheritance of certain variations in mitochondrial DNA (called heteroplasmy) may be influenced by the nuclear genome with which it associated – a possibility that has been suggested previously, but not on human data of this quality.  For this reason, based both on theoretical concerns and a small amount of animal data, the 2016 report to the HFEA on using Mitochondrial Transfer to avoid transmission of mitochondrial DNA-causing diseases recommended that it would be sensible, where possible, for matching of the haplotype of the donor with that of the recipient.

“Whilst it was already understood from the work on embryonic stem cell lines that there could be a risk of an abnormal mutation increasing, rather than being eliminated or suppressed as anticipated, this risk would likely be small and would still be a substantial risk reduction compared to the certain transmission of the disease-causing mutation to the child without mitochondrial transfer.  These new results support the recommendation for prenatal diagnosis where feasible, and for mandatory long-term follow-up of any children resulting from mitochondrial donation, until the true risk is established – a detail sadly not being emphasised in those countries where Mitochondrial Transfer has already been undertaken.”


Dr Andy Greenfield, Programme Leader in Developmental Genetics, MRC’s Harwell Institute, said:

“Mitochondria are the energy-producing organelles of our cells and they contain DNA – mitochondrial DNA (mtDNA).  Like any DNA molecule, there can be mutations that disrupt the function of mtDNA and these can be responsible for devastating diseases.  Most of the proteins of the mitochondria are encoded by genes found in the nucleus, whilst a few are encoded by mtDNA itself.  The way in which energy is produced in the mitochondrial membrane requires that these two types of protein cooperate: they must work side-by-side – singing off the same hymn sheet, so to speak.  There may be problems if they fail to do so.  This is something that has been established in model organisms – like the mouse – for some time.  The mtDNA from one strain of mouse may not operate normally if required to do so alongside proteins encoded by nuclear DNA from a much more distantly related strain.  To put it another way, rat mtDNA would not work in the context of mouse nuclear genes.  The two genomes have diverged too much to allow effective cooperation – so-called mitonuclear cooperation.

“What has been missing is good evidence that this cooperation between mtDNA and nuclear DNA helps to shape the evolution of human mtDNA and the types of sequence variants (haplogroups/haplotypes) that are found in human populations, to different degrees, throughout the world i.e. data that do not derive from artificial or potentially irrelevant experimental contexts.  The paper by Wei Wei et al – which is very challenging due to its technical nature but fascinating nonetheless – claims to provide evidence that sequence variation of mtDNA in humans is shaped by its interaction with nuclear genes.  By examining the sequence of mtDNA and nuclear DNA in large numbers of individuals from distinct ancestries, they conclude that selection – either for or against certain combinations – is a significant contributory factor in the dynamic way in which mtDNA sequences have evolved in the human population over time.  The authors do not claim it is the only factor – so-called genetic drift and founder effects (and potential unknowns) are likely factors too – but they suggest that there is now evidence at the population level in humans that not all combinations of mtDNA and nuclear genome are equal.

“Mitochondrial donation (MD) is a technique (or collection of techniques) that aims to prevent the transmission of mitochondrial diseases from mother to offspring by using a mitochondrial donor (in the form of an egg or embryo from which the nuclear material has been removed but normal mitochondria remain).  MD is now lawful in the UK.  The expert panels convened by the HFEA to examine MD techniques spent considerable time on the intricacies of so-called mitonuclear matching/mismatching.  At the time of the last report, in 2016, the pre-clinical data available on this topic led the panel to recommend that any clinic offering MD should consider haplogroup matching: the matching of donor mtDNA and the mtDNA of the prospective mother, in order to minimise the likelihood of any effects arising from unusual mitonuclear interactions.  Indeed, the 2016 panel recommended the kinds of study reported by Wei Wei at al.  The authors claim that their data emphasise the potential importance of matching, perhaps in reducing the chance of so-called reversion, in which the prospective mother’s pathogenic mtDNA may come to predominate after MD.  Given these new data, and the fact that this topic was an important part of the expert panel’s deliberations and recommendations, it is reasonable that any clinic offering MD should consider anew whether haplogroup matching might be appropriate, on a case-by-case basis; this, of course, notwithstanding the potential difficulty in recruiting egg donors.  Clearly, we need ongoing research in order to understand the significance of these new data for the safety and efficacy of MD in the clinic, including the use of model organisms like the mouse to drill down and investigate specific mechanisms at play.”


Dr David Clancy, Lecturer in the Division of Biomedical and Life Sciences, Lancaster University, said:

“It is thought that mitochondria began their evolution within host cells around a billion years ago as parasites.  Mitochondrial DNA has coevolved since then with host genomes such that they now mediate or contribute to so many critical cellular processes that they have become totally indispensable.  And this is despite their relatively tiny genome size: ~20,000 base pairs compared to ~3billion base pairs of chromosomal DNA in the cell nucleus.

“Given this multiplicity of functions in addition to energy production, including nutrient and stress signalling, cell death, hormone signalling, steroid synthesis and heme synthesis for red blood cells, it would be surprising if some combinations of mitochondrial genome (mtDNA) and cell nuclear genome (chromosomal – nDNA) were not better or worse adapted to each other.

“This has become clear from studies in fruit flies in particular, across a number of traits including lifespan.

“MtDNA is maternally inherited and some individuals harbour more than one mitochondrial genotype.  Chinnery and co-workers have now shown, using human data, that the way these different types were inherited from the mother to the child was non-random.  The demands of egg cell development, and possibly later of embryo development, result in some mtDNA types being ‘preferred’ by some nDNA types.  Typically those which had existed together in populations ancestrally over many generations were preferred to those which had not.  Thus the nuclear genome seems to select its most compatible mitochondrial genome.  This selection process can be missed in the case of Mitochondrial Replacement Therapy if embryo and donor mtDNA are less optimally matched than they might be, a possibility raised by Professor Chinnery and which had been earlier sounded as a note of caution by several scientists during the debate around the procedure.  We do not know if this could lead to health implications in the adult.  As a precaution, matching of donor and recipient genomes should now be made routine for this procedure.”


Prof Robin Lovell-Badge, Group Leader, The Francis Crick Institute, said:

“This detailed piece of research by Wei et al provides evidence for co-evolution of mitochondrial and nuclear DNA in diverse human populations.

“By itself this may not have consequences relevant to health, because humans from very different origins interbreed with no obvious consequences for the children that are born with what could be a subtle mismatch between nuclear and mitochondrial genomes.

“However, it might be a problem if any mtDNA variant with a selective advantage on a particular nuclear DNA background was linked with a deleterious mutation that leads to mitochondrial disease.  This could in theory happen with mitochondrial replacement where even a small carry-over of mutant mtDNA with the patient’s nuclear DNA could lead to preferential amplification of the former at the expense of the non-mutant donor mtDNA, resulting in a child with a risk of mitochondrial disease.

“These concerns were reflected in detail in a 2016 report provided to the HFEA on the safety and efficacy of mitochondrial replacement (donation), where pre-clinical data led the panel to recommend that any clinic offering this should consider haplogroup matching: the matching of donor mtDNA and the mtDNA of the prospective mother, in order to reduce the likelihood of any effects based on unusual mito-nuclear interactions.

“However, the current paper does not provide hard evidence that this is necessary, nor does it test mechanisms by which one mtDNA haplotype could preferentially amplify over another, or identify which combinations might lead to problems.

“Nevertheless, the evidence that co-evolution of mitochondrial and nuclear genomes occurs is important for understanding human origins and diversity and the paper provides valuable data for future research into mtDNA function.”

Additional background information and analysis from Robin of the paper by Wei et al (which you can also quote from if you want to):

“Mitochondria, small organelles that provide the energy required for many biological processes within cells, have their own DNA (mtDNA), exclusively inherited maternally, which carries a few genes, the products of which collaborate with those of many hundreds of genes involved in mitochondrial function that are found in nuclear DNA, which is inherited both maternally and paternally.  It is not surprising that both mtDNA sequences (of which there are many copies present in each cell) and nuclear genes encoding mitochondrial proteins evolve; they will be subject to random genetic drift, founder effects, and, presumably, selection for aspects of energy production that will vary according to environment.  In humans this is thought to have led to a range of sequence variants that can be classified into haplogroups, that tend to vary between distinct populations, or even more detailed haplotypes within a haplogroup.  However, until now there has been no direct evidence for selection of one haplotype over another, moreover humans from very different origins interbreed with no obvious consequences for the children that are born with what could be a mismatch between nuclear and mitochondrial genomes.  The Wei et al paper makes use of data from the UK’s 100,000 genomes project to directly look at this topic.  Through focussing on families where more than one mtDNA variant is found within cells, a situation referred to as heteroplasmy, and looking at situations where the mtDNA and nuclear DNA are from widely different origins (e.g. Asian and European), they find evidence that there is indeed a tendency for new specific mtDNA variants to amplify if they are more similar to that normally associated with the nuclear DNA.

“The data, which is based on complex statistical analysis, have led the authors to make several more general conclusions, some of which are confirmatory, but nevertheless important.  For example, novel mtDNA variants were less likely to be inherited than known variants, which probably reflects that many of these novel variants compromise mtDNA replication or function and are rapidly selected against.  The paper also includes a lot of valuable data for future studies – for example it reveals that certain sequences within the D-loop, part of the mtDNA involved in its replication and transcription (i.e. mtDNA gene expression), which is overall quite variable, show a distinct absence of variants, suggesting that these have as yet unknown critical functions.

“However, the authors’ conclusion that the nuclear genetic background can influence which types of mtDNA variant are present is the most interesting.  This is not a surprise, especially given that mtDNA replication and gene expression depends on nuclear encoded proteins.  Consequently, in cases of heteroplasmy, a mtDNA variant that is less good at being replicated on a particular nuclear DNA background will tend to be lost, and one that is better will tend to predominate.  Nuclear and mitochondrial gene products also need to function together in the process of generating energy for cells, therefore any mismatch will compromise cell viability – this would also lead to retention (and perhaps selection for) the mtDNA variant that works best with a specific nuclear background.  Apart from theory, there are some previous animal and in vitro human data that supports these conjectures.

“Any selective advantage or disadvantage can be small, but it could still lead to a substantial shift in heteroplasmy where one mtDNA variant predominates over another.  However, there is no evidence that this is in itself deleterious.  Any “mismatch” might have little if any consequence.  After all, the authors find 206 individuals¥ where the nuclear and mitochondrial genomes were from different human populations (notably Asian versus European).  They found amongst these individuals (and from another 654 informative individuals) evidence for selection of new mtDNA variants that had been associated previously with their nuclear DNA type.  The fact that these new variants were still present in a heteroplasmic state suggests that any selective advantage of the new variant is small; moreover, we are not told what proportion of individuals show this.  Nevertheless, it would be a problem if any mtDNA variant with a selective advantage on a particular nuclear DNA background was linked with (i.e. present on the same mtDNA as) a deleterious mutation that leads to mitochondrial disease.  This could in theory happen with mitochondrial replacement where even a small carry-over of mutant mtDNA with the patient’s nuclear DNA could lead to preferential amplification of the former at the expense of the donor mtDNA, resulting in a child with a risk of mitochondrial disease.”

¥ The authors do not say if all 206 individuals were healthy, after all, recruitment for the 100,000 genomes project included childhood-onset neurodegenerative disorders, as well as other rare diseases, but associating any problems with a nuclear-mtDNA “mismatch” would be challenging.


‘Germline selection shapes human mitochondrial DNA diversity’ by Wei Wei et al. was published in Science at 19:00 UK time on Thursday 23 May 2019.


Declared interests

Prof Peter Braude: “Peter Braude was a member of the independent expert panel advising the HFEA on mitochondrial donation and contributed to the Singapore Bioethics Advisory committee consultation on mitochondrial donation.” 

Dr David Clancy: “No conflicting interests.”

Prof Robin Lovell-Badge: “Robin Lovell-Badge served on the HFEA’s scientific panel looking into the safety and efficacy of mitochondrial replacement (donation) techniques.  He has no conflicts of interest to declare.”

None others received.

in this section

filter RoundUps by year

search by tag