‘Safe’ call? My thoughts on the latest mitochondrial replacement paper

This is an important paper because it may well be the last published set of experiments from the Newcastle group (which has been developing mitochondrial replacement therapy) before the HFEA makes a decision on whether or not they are given a licence to implement pronuclear transfer (PNT) clinically. The overall message they convey is one of safety. Problems with carryover are noted and success in terms of improved efficiencies in the development of the techniques are brought to the fore (they now employ an ‘early’ version called ePNT).

From my point of view the most interesting aspect of this study is the question of mitonuclear mismatching, which is something I’ve addressed with colleagues in the past. To recap, we proposed that there is a clear possibility that when moving the nuclear genome from one ova to another, the switch from one cytoplasmic/mitochondrial background to another could be detrimental to the health of the developing embryo (this concern applies to any of the variants currently proposed or developed for mitochondrial replacement therapy). We have outlined the evidence from a range of organisms showing that cytoplasmic/mitochondrial background effects are important (Reinhardt et al Science) and proposed experimental designs for how to test this in PNT (or maternal spindle transfer; Morrow et al EMBO Reports).

It is reassuring to see the authors have taken these ideas on board as they have now collected data from unmanipulated (control) oocytes, PNT from one oocyte to another (heterologous PNT), and crucially PNT within a single oocyte (remove and return). This last treatment could be used to disentangle the potentially harmful effects of switching mitochondrial background that occurs during PNT from any of the physical or technical effects of the process.

The reaction from many has been very upbeat. Prof. Mary Herbert, one of the main authors, has been reported saying “By any measure we’ve looked at, we’ve not found a harmful effect of the procedure” At a press briefing it was stated that embryos created with ePNT were “indistinguishable from those created by normal IVF”. A range of experts are quoted by the Science Media Centre London, many using that same word, “indistinguishable”. The clear conclusion of many influential reporters of science and health is that PNT (aka mitochondrial donation) is “safe” (e.g. the BBC’s Fergus Walsh and The Guardian’s Ian Sample), but not all were so convinced (see Paul Knoepfler’s blog).

My reading of the paper is very different. Again I focus on the potential for mismatching between mitochondrial and nuclear genomes (and do not deal with carryover, which is dealt with by a recent post by Paul Knoepfler).

First, early on in the paper the authors present data on the rates that blastocysts are formed following the two versions of ePNT (heterologous and autologous) and unmanipulated controls. In figure 2b they show two plots for data collected on two days (5 and 6) following the manipulations. On both days around 60% of control blastocysts (grey bars) have survived and a similar proportion of autologous ePNT blastocysts (dark purple bars). But heterologous blastocysts have a lower success rate – around 40% (light purple). For the day 6 data this difference is statistically significant, but for the smaller day 5 data set it is not, although the pattern is the same. In a third panel they show data on blastocyst quality, collected by classifying their physical appearance in terms of their morphology and how expanded they are. Here they report in the figure legend that “blastocyst quality is similar between the three groups (not significant; Fisher’s exact test)”. But a comment on PubPeer shows that this to be incorrect for day 5 data (although day 6 data is not significant).

Screen Shot 2016-06-14 at 13.32.50Screenshot of Figure 2b from Hyslop et al

The authors acknowledge that although heterologous ePNT produced lower rates of blastocyst formation, it might be explained by the necessity to use frozen oocytes that had been vitrified, instead of using fresh oocytes (as the autologous oocytes were). They provide further data in the Extended Data Figure 3 to explore this idea. Again these data show success rates of heterologous ePNT (whether fresh or vitrified) were lower than for autologous ePNT (dark purple bars higher than all others). Furthermore, vitrified oocytes had higher success rates. This suggests that the decline in blastocyst formation rates of heterologous ePNT seen in Figure 2b are not due to vitrification, and does not tally with the authors explanation for why survival of heterologous blastocysts is lower than controls.

Screen Shot 2016-06-14 at 13.37.29

Screenshot of Extended Data Figure 3 from Hyslop et al

Together, this evidence shows there is an effect of switching cytoplasmic/mitochondrial background beyond the simple technical effect of ePNT (remember autologous success rates were very similar to unmanipulated controls). These data therefore support the hypothesis that mitonuclear mismatching is a problem for embryo health, and that both the number and quality of blastocysts derived from ePNT are ‘distinguishable’ from controls. But that is not the interpretation given in the paper or at the press conference. But there’s another issue: in theory the harmful effect of mitonuclear mismatching may be less likely if the mtDNA of donor and recipient oocytes is more closely related, but it is impossible to tell from the analysis whether this was considered. Oocytes from different women were used, but were heterologous transfers carried out using occytes that differed in mtDNA haplotype? mtDNA haplotype data was collected but it does not appear to have been used for this part of the analysis.

A second important part of the paper is the analysis of gene expression of cells taken from the blastocysts using RNA sequencing. In total, 94 samples were sequenced (data available here: http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE76284) from a set of blastocysts belonging to the three treatments. The authors conduct principle component analyses, a t-SNE analysis and hierachical clustering of these data. The conclusions the authors make from all of these is that heterologous ePNT samples are “indistinguishable” (there’s that word again!) from controls (either unmanipulated or autologous). They do identify that expression profiles are grouped according to the different cell lineages within the developing blastocysts (Extended Data Figure 6). But there are several problems with this approach: (1) They do not apparently fit any statistical model to investigate the effect of the three treatments on gene expression. They therefore cannot say whether or not treatment had any effect on gene expression. The PCA plots and heatmaps provide an overview of global gene expression profiles, whether filtered to include the most variable expressed genes or not, but these are not statistical tests of the null hypothesis that treatment has no effect. There are no P-values reported, only pictures. (2) It is clear that cell lineage appears to be influencing the gene expression profiles (see Extended Data Figure 6c), and so any model to investigate treatment effects should also include cell lineage as a factor. (3) The 94 samples were also derived from a smaller set of blastocysts, and so samples should be a nested factor within any model. (4) Acknowledging this fact, it also becomes apparent that the biologically independent levels of replication for the RNAseq data is very small indeed, with data available from only 3 control blastocysts, 1 autologous, and 9 heterologous (see Extended Data Figure 5b). (5) Again this level of independence is potentially further reduced if any of the heterologous blastocysts derive from transfers between matched mtDNA haplotypes. All in all, that the authors, experts and media conclude these profiles are indistinguishable from one another is hardly surprising. They have very low power to detect differences if present, even if modelled correctly.

To summarise, the authors do find evidence for deleterious effects on embryo survival and quality (the latter not reported correctly), consistent with the hypothesis for mitonuclear mismatching. But this effect is not discussed in the paper in reference to the safety of implementing the technique clinically. In my opinion the paper does not contain a convincing analysis that gene expression profiles of the three treatments are indistinguishable from one another since they do not apparently test for those effects in a suitable statistical model, and use a dataset that is severely underpowered (even though they include a grand total of 523 oocytes in the study).

It is really concerning that despite all the warnings that mitonuclear mismatching may be an issue, and even investigated in the paper, that evidence for it is apparently glossed over by the scientists and the science communicators alike. We won’t know for sure of any effects in real people until it is implemented clinically, but any definitive conclusion of safety still seems premature.

Is gene editing mtDNA an alternative to mitochondrial replacement therapy?

The big news in genetics this week has been a report published in Protein & Cell of gene editing of human embryos, using  a technique known as CRISPR. Many news outlets and bloggers have covered the controversy but another paper out this week in Cell using a related technique has received far less attention (but see here and here), which is a shame as it is potentially a viable method for the elimination of mitochondrial diseases.

In the paper, the authors used a method called mitoTALEN to seek and destroy mitochondria containing mutant mtDNA in cells containing a mixture of diseased and healthy mitochondria. The method works by designing a probe to identify specific sequences of DNA within the mitochondrial genome and then break the strands apart. This sets in motion a process that leads the the eventual death of that particular mitochondrion. Some people have a mixture of diseased and healthy mitochondria within each of their cells (heteroplasmy) and remain healthy, but for a few unlucky individuals the proportion of diseased mitochondria is too high, which results in a number of serious health problems.

The researchers found that when the mitoTALEN technique was applied to a fertilized egg (by literally injecting it with the TALEN probe), the diseased mitochondria were eliminated and the resulting embryos were transformed from one who would experience the disease to a healthy one, as the proportion of healthy mitochondria shifts. It’s action then is really quiet similar to a drug that targets diseased cells, but in this case it’s targeting diseased organelles.

mitoTALEN therefore offers a number of advantages over mitochondrial replacement therapy (MRT, sometimes called mitochondrial donation or three-person IVF), where the nucleus from a mito-disease carrier is transferred to a donor’s egg. First, in terms of safety, it completely side-steps the issue of mitonuclear incompatibilities arsing in MRT due to genetic material being combined from 3 individuals. This has been something my colleagues and I have been vocal about, and while others discount it, we have recently shown that the rationale for rejecting these concerns are not well founded. Second, it’s a much simpler procedure which is a definite advantage.

Ethically, mito-TALEN also bypasses issues around using genetic material from three people, there is no need to produce then destroy embryos, mitonuclear interactions are not altered, and the risk of ovarian hyper-stimulation syndrome amongst donors also vanishes. For many people then (MPs and peers included?), I imagine mitoTALEN would be the preferable option. Questions remain however around its efficacy, since in carriers where proportions of diseased mtDNA are very high, the technique may result in treated embryos without enough of healthy mitochondria leading to embryo failures. So for those individuals it might not be a good option, or may show very low success rates. This would certainly be the case where the diseased mtDNA are the only type on the cell. The mitoTALEN mechanism may also cut DNA in the wrong place, perhaps most catastrophically if it was cutting nuclear DNA, since any repairs may incorporate errors, leading to other genetic (and heritable) diseases.

Overall then it looks like mitoTALEN could be widely applied but not universally so, and it clearly needs replicating and extending before applying in the clinic. I think it’s interesting though that neither mitoTALEN, nor any other gene editing technique that targets mtDNA (like CRISPR) would be classified as genetic modification by the UK government, since their definition of GM in humans is limited only to the chromosomes in the nucleus (see a previous post). With the pace of scientific progress this definition is likely to appear more and more out of date. To my mind this would also cause considerable confusion in the public if there is a moratorium amongst scientists to stop gene editing experiments on human embryos, unless mtDNA is the target.

C’est ne pas une mitochondrie

File:02 - Single Energizer Battery.jpg
(Image Wikipedia)

In September 2014 I was invited to write a blog post for BioNews based on a backbencher’s debate of mitochondrial replacement (MR) therapy. The post provoked a response by Professor Frances Flinter a Consultant in Clinical Genetics at Guy’s & St Thomas’ NHS Foundation Trust. I now take the opportunity to respond to some of the points she raises.

Prof. Flinter’s first point concerns the ethics of the debate in that while a discussing the potential risks and benefits is needed, “the debate often focuses on the potential risks”. She then goes on to ask “Is it ethical to try and prevent the development of a treatment that might enable the birth of a healthy baby for a couple for whom there are no other options?”. In asking this question she touches on a central point that the ethical arguments rest upon: that the safety and efficacy of MR has been clearly demonstrated. For example, in the popular media it is described as a “cure“, that it would “saves children’s lives“, or be a “guarantee” of having a healthy child. But the safety and efficacy of MR has not yet been established. Even the HFEA expert panels’ reviews of evidence have not made such absolute conclusions, instead stating that it is “not unsafe“. Knowing this one could ask a different question – is it ethical to discuss a new therapeutic intervention, thereby raising prospective families hopes, without actually having to hand clear evidence that it is indeed safe or effective?

A second point discussed is the use of the battery analogy for mitochondria. I do appreciate the value of an insightful model of reality, especially for a non-specialist or lay audience. After all, it is an important responsibility that scientists can communicate scientific concepts and ideas to the public effectively. But we need to be very careful that we are not misleading people with poorly constructed and inaccurate simplifications; so keep things “as simple as possible but no simpler“. For me the battery analogy is dangerously wrong. As I outlined in my BioNews post, mitochondria and the genetic variation in their genomes influence a wide variety of traits, in large part no doubt due to their crucial role in energy production (as Prof. Flinter points out, batteries store rather than produce energy). In part, I think some of the confusion comes from the reductionist view that many parties involved in the debate have taken. A view that ignores the fact that genes (whether mitochondrial or nuclear) do not work alone but within networks of other genes. Also, that these networks influence multiple biological processes at the cellular level. Similarly, cells form tissues, within organs, the totality of which are important for the whole organism to function properly. Reductionism is very useful when working on proximate level questions, but one should never forget the biological reality of the whole organism.

One important consequence of using an inaccurate analogy is that the debate is then coloured by misinformation. When mtDNA is portrayed as doing nothing other than provide the code for making more “batteries”, it is then inconceivable that it can influence anything else. Evidence showing otherwise is immediately dismissed on the basis that it doesn’t fit in logically with the initial false premise. So for example, I pointed out to the Parliamentary Science and Technology Select Committee that there is in fact evidence that genetic variation in the mtDNA influences an individual’s personality from a range of animal models, including humans (see here for my collation). But pointing this out has had absolutely no impact whatsoever, Jane Ellison MP (Parliamentary Under-Secretary of State for Health) recently repeating that “mitochondrial DNA only encodes genes responsible for energy production and will not affect the child’s appearance, personality or any other personal characteristics.”

Do I have a better analogy than the battery one? No. But that doesn’t mean should just make do and continue to use a flawed model.

Prof Flinter goes on to express surprise that I “can envisage scenarios where genetic modification in humans could be justifiable (e.g. correcting point mutations)”, and asks “is this because he thinks that genetic abnormalities of the nuclear DNA are more serious?” No, to clarify the reason is that in some cases the details of how a single point mutation may influence the phenotype are much more well resolved. As such, changing a single base will have known and predictable effects (potential examples here). Making such an adjustment could then be made in the full knowledge of its consequences. The same cannot at present be said for MR.

Finally, Prof. Flinter makes reference to the 30 or so children that were created in the 1990’s using ooplasm transfer, a technique developed by James Grifo, where mitochondria from the ova of putatively healthy women is injected into the oocyte from a woman carrying diseased mtDNA. This technique is similar, but not identical, to MR. That the health of these children has not been followed up yet is a concern, especially since reports in the media indicate that there were some problems including developmental disorders. In fact, pronuclear transfer (PNT; one of the proposed versions of MR) has been carried out before. In an abstract to a conference in 2001 the authors (including Grifo) describe how triplets were created, the pregnancy was then surgically reduced to twins, but both died prematurely. However, I see no evidence that this experiment featured in any of the HFEA expert panel reviews of PNT or MST (maternal spindle transfer), even though it was covered by the media at the time, including Nature and the Guardian (the latter which quotes Lord Winston).

So robust debate, yes. But also a debate based on a more rounded appreciation of how biological systems really work – in this respect, data from model organisms can tell us a lot about what role mtDNA plays. Also only a debate based on all available evidence, whether encouraging or not, can give clinicians and families the clearest possible insights into what the outcomes of MR might be.


The Parliamentary Science and Technology Select Committee and I

Image source: wikipedia

A recurring question for me over the past few months is how policy makers and scientists can interact in ways that ensure policies are based on the best available evidence. My current interest in this originates from my involvement in the debate over mitochondrial replacement (MR) as a potential therapy to eradicate mitochondrial disease, a therapy for which colleagues and I have reservations. More recently, I also took part in a ‘Week in Westminster’ as part of the Royal Society Parliamentary Pairing Scheme that gives researchers and policy makers the chance to see each other’s working lives in person. As well as attending seminars and seeing Parliament in action for myself, I spent some time shadowing my local MP Caroline Lucas – the UK’s first and only Green Party member of Parliament (I’ll post more detail on that experience later).

Policy makers and scientists work in very different ways and sometimes it seems they are separated by an ocean of different priorities and constraints (rather like Anna Leonowens and the King). Currently however, there is a strong drive to base political decisions on evidence that has been collected using scientific principles. For medical procedures, this is obviously the case but I have been genuinely surprised at how much sway political forces have had in the debate over the safety and efficacy of MR.

Earlier this year I was invited to attend a one-off evidence session held by the Parliamentary Science and Technology Select Committee to discuss “mitochondrial donation”. Prior to attending I also submitted some written evidence co-signed by four other academics, which as a result of teaching commitments was a rather hastily composed email with a couple of PDF attachments. At the meeting itself I appeared alongside Prof Doug Turnbull (who is developing MR in the UK) as well as Prof Lovell-Badge and Prof Peter Braude, who were both members on the HFEA expert panel convened to review scientific evidence. In later sessions there were representatives from the Nuffield Council on Bioethics and finally Jane Ellison MP, the Parliamentary Under-Secretary of State for Public Health, with Professor Dame Sally Davies, Chief Medical Officer.

In the days prior to the meeting I had been able to spend some time researching whether there was further evidence of mito-nuclear interactions influencing traits relevant to humans and was able to identify a number of studies, as well as data showing variation in the mitochondrial genome influences personality traits in a number of species, including humans. I pointed this out to the committee at the meeting and also outlined why I thought the HFEA expert panel had neglected to review the evidence adequately. See here for a full transcript of the meeting or you can even watch it here.

Needless to say no one paid much attention to my statements and I later found out when the Select Committee published the submission made to it, ours was not included (although admittedly it was not in the format requested for written submissions). The most bizarre part of the meeting was when it was pointed out by the Chair Andrew Miller MP that “I do not think I have heard anything that implies that Professor Turnbull thinks Dr Morrow is a crank. He is a legitimate scientist.” (See Q16 in transcript). Which begs the question if no-one thinks I am a crank, then why bother mentioning it? I later found out that one member of the Select Committee, David Treddinick, advocates both homeopathy and astrology! Maybe the chair was genuinely concerned in identifying any possible cranks in the room?

Overall, my impression was that the meeting was carefully managed, with the result that any concerns I was trying to raise were isolated to me as an individual rather than reflecting the views of several other scientists. Recall that our written submission did not appear in the summary of submission made to it, which also contains highly critical submissions by Professor Stuart A. Newman and Professor Justin C St John, also not referred to at the meeting.

So what’s next? Well despite Jane Ellison being urged by the committee to move forward quickly, and despite her assurance that she would press the matter with Jeremy Hunt (Secretary of State for Health), there has been no debate on the issue so far in Parliament. With only a handful of days left before the Christmas recess it seems highly unlikely the debate will occur in 2014. Which leaves a few months before the election in May 2015. But why is it taking so long? Something I learned during my Week in Westminster was that things move very fast there, the diary is filled with only a few days notice. And so given the apparent lack of other parliamentary business (MPs are reportedly even down to working a 3-day week) it seems odd to me that time has not yet been found for this important debate.

Is a puzzlement!


Influence of mitochondrial DNA on personality traits – some evidence


Recently I came across a paper in the Proceedings of the Royal Society, series B, that reports effects of mitonuclear interactions on personality traits in a seed beetle. You may ask “How can seed beetles have a personality?” – well a reasonable definition of ‘personality’ is simply within individual consistency in any measurable behaviour. The paper’s full citation and link is:

Løvlie, H., Immonen, E., Gustavsson, E., Kazancioğlu, E. & Arnqvist, G. 2014. The influence of mitonuclear genetic variation on personality in seed beetles. Proc. R. Soc. B Biol. Sci. 281: 20141039. Link

I wondered if there was any other evidence out there that mtDNA (with or without any interaction with nuclear DNA) had significant effects on other behavioural traits. Well it turns out there are *10* other studies:

Boratyński, Z., Alves, P.C., Berto, S., Koskela, E., Mappes, T. & Melo-Ferreira, J. 2011. Introgression of mitochondrial DNA among Myodes voles: consequences for energetics? BMC Evol. Biol. 11: 355. Link

Boratyński, Z., Melo-Ferreira, J., Alves, P.C., Berto, S., Koskela, E., Pentikäinen, O.T., et al. 2014. Molecular and ecological signs of mitochondrial adaptation: consequences for introgression? Heredity 113: 277–286. Link

Gimsa, U., Kanitz, E., Otten, W. & Ibrahim, S.M. 2009. Behavior and Stress Reactivity in Mouse Strains with Mitochondrial DNA Variations. Ann. N. Y. Acad. Sci. 1153: 131–138. Link

Kato, C., Umekage, T., Tochigi, M., Otowa, T., Hibino, H., Ohtani, T., et al. 2004. Mitochondrial DNA polymorphisms and extraversion. Am. J. Med. Genet. B Neuropsychiatr. Genet. 128B: 76–79. Link

Kishida, K., Tominaga, M., Matsubara, K., Taguchi, M., Noguchi, M., Tsunawake, N., et al. 2009. An Association Analysis between Mitochondrial DNA A10398G Polymorphism and Temperament in Japanese Young Adults. PLoS ONE 4: e7763. Link

Shao, L., Martin, M.V., Watson, S.J., Schatzberg, A., Akil, H., Myers, R.M., et al. 2008. Mitochondrial involvement in psychiatric disorders. Ann. Med. 40: 281–295. Link

Šíchová, K., Koskela, E., Mappes, T., Lantová, P. & Boratyński, Z. 2014. On personality, energy metabolism and mtDNA introgression in bank voles. Anim. Behav. 92: 229–237. Link

Skuder, P., Plomin, R., McClearn, G.E., Smith, D.L., Vignetti, S., Chorney, M.J., et al. 1995. A polymorphism in mitochondrial DNA associated with IQ? Intelligence 21: 1–11. Link

Tanaka, D., Nakada, K., Takao, K., Ogasawara, E., Kasahara, A., Sato, A., et al. 2008. Normal mitochondrial respiratory function is essential for spatial remote memory in mice. Mol. Brain 1: 21. Link

Yu, X., Gimsa, U., Wester-Rosenlof, L., Kanitz, E., Otten, W., Kunz, M., et al. 2009. Dissecting the effects of mtDNA variations on complex traits using mouse conplastic strains. Genome Res. 19: 159–165. Link

Mito-nuclear mismatch – some evidence

*Updated 27th October 2014*

Below are some of the papers identified so far showing evidence of mismatching between the nuclear and mtDNA resulting in a range of clinically relevant disease traits. Links to full-text versions are now added.

8 examples of a mutation in the mitochondrial DNA (mtDNA) that is modified by the nuclear background:

Ballana, E., Mercader, J.M., Fischel-Ghodsian, N. & Estivill, X. 2007. MRPS18CP2 alleles and DEFA3 absence as putative chromosome 8p23.1 modifiers of hearing loss due to mtDNA mutation A1555G in the 12S rRNA gene. BMC Med. Genet. 8: 81. Link

Bykhovskaya, Y., Mengesha, E., Wang, D., Yang, H., Estivill, X., Shohat, M., et al. 2004. Human mitochondrial transcription factor B1 as a modifier gene for hearing loss associated with the mitochondrial A1555G mutation. Mol. Genet. Metab. 82: 27–32. Link

Davidson, M.M., Walker, W.F., Hernandez-Rosa, E. & Nesti, C. 2009. Evidence for nuclear modifier gene in mitochondrial cardiomyopathy. J. Mol. Cell. Cardiol. 46: 936–942. Link

Deng, J.-H., Li, Y., Park, J.S., Wu, J., Hu, P., Lechleiter, J., et al. 2006. Nuclear Suppression of Mitochondrial Defects in Cells without the ND6 Subunit. Mol. Cell. Biol. 26: 1077–1086. Link

Hao, H., Morrison, L.E. & Moraes, C.T. 1999. Suppression of a Mitochondrial tRNA Gene Mutation Phenotype Associated with Changes in the Nuclear Background. Hum. Mol. Genet. 8: 1117–1124. Link

Hudson, G., Keers, S., Man, P.Y.W., Griffiths, P., Huoponen, K., Savontaus, M.-L., et al. 2005. Identification of an X-Chromosomal Locus and Haplotype Modulating the Phenotype of a Mitochondrial DNA Disorder. Am. J. Hum. Genet. 77: 1086–1091. Link

Johnson, K.R., Zheng, Q.Y., Bykhovskaya, Y., Spirina, O. & Fischel-Ghodsian, N. 2001. A nuclear-mitochondrial DNA interaction affecting hearing impairment in mice. Nat. Genet. 27: 191–194. Abstract only Link

Potluri, P., Davila, A., Ruiz-Pesini, E., Mishmar, D., O’Hearn, S., Hancock, S., et al. 2009. A novel NDUFA1 mutation leads to a progressive mitochondrial complex I- specific neurodegenerative disease. Mol. Genet. Metab. 96: 189–195. Link


5 examples of a mutation in the nuclear DNA that is modified by the mtDNA background or specific mutation (e.g. haplogroup/haplotype):

Bonaiti, B., Olsson, M., Hellman, U., Suhr, O., Bonaiti-Pellie, C. & Plante-Bordeneuve, V. 2010. TTR familial amyloid polyneuropathy: does a mitochondrial polymorphism entirely explain the parent-of-origin difference in penetrance? Eur. J. Hum. Genet. 18: 948–952. Link

Gershoni, M., Levin, L., Ovadia, O., Toiw, Y., Shani, N., Dadon, S., et al. 2014. Disrupting Mitochondrial–Nuclear Coevolution Affects OXPHOS Complex I Integrity and Impacts Human Health. Genome Biol. Evol. 6: 2665–2680. Link

Kim, A. Chen, C-H. Ursell, P. and Huang, T.T. 2010. Genetic modifier of mitochondrial superoxide dismutase-deficient mice delays heart failure and prolongs survival. Mammalian Genome. 21:534–542. Link

Strauss, K.A., DuBiner, L., Simon, M., Zaragoza, M., Sengupta, P.P., Li, P., et al. 2013. Severity of cardiomyopathy associated with adenine nucleotide translocator-1 deficiency correlates with mtDNA haplogroup. Proc. Natl. Acad. Sci. 110: 3453–3458. Link

Vartiainen, S., Chen, S., George, J., Tuomela, T., Luoto, K.R., O’Dell, K.M.C., et al. 2014. Phenotypic rescue of a Drosophila model of mitochondrial ANT1 disease. Dis. Model. Mech. 7: 635–648. Link


5 reviews of mutations in the mitochondrial DNA (mtDNA) that is modified by the nuclear background:

Bénit, P., El-Khoury, R., Schiff, M., Sainsard-Chanet, A. & Rustin, P. 2010. Genetic background influences mitochondrial function: modeling mitochondrial disease for therapeutic development. Trends Mol. Med. 16: 210–217. PDF

Carelli, V., Giordano, C. & d’ Amati, G. 2003. Pathogenic expression of homoplasmic mtDNA mutations needs a complex nuclear–mitochondrial interaction. Trends Genet. 19: 257–262. Link

Luo, L.-F., Hou, C.-C. & Yang, W.-X. 2013. Nuclear factors: Roles related to mitochondrial deafness. Gene 520: 79–89. Link

Taanman, J.-W. 2001. A nuclear modifier for a mitochondrial DNA disorder. Trends Genet. 17: 609–611. Link

Zhu, X., Peng, X., Guan, M.-X. & Yan, Q. 2009. Pathogenic mutations of nuclear genes associated with mitochondrial disorders. Acta Biochim. Biophys. Sin. 41: 179–187. Link


Also 1 study showing mitochondrial DNA levels are influenced by nuclear DNA mutations:

López, S., Buil, A., Souto, J.C., Casademont, J., Martinez-Perez, A., Almasy, L., et al. 2014. A genome-wide association study in the genetic analysis of idiopathic thrombophilia project suggests sex-specific regulation of mitochondrial DNA levels. Mitochondrion 18: 34–40. Link




Finally the meta-analysis showing evidence of significant cyto-nuclear (which includes mitonuclear) effects on a variety of traits:

Dobler, Rogell, Bugar and Dowling (2014) A meta-analysis of the strength and nature of cytoplasmic genetic effects. Journal of Evolutionary Biology 27: 2021–2034. Link

MPs debate the science of mitochondria

This post originally appeared in BioNews (770; 08 September 2014) under the title “Myth replacement therapy: MPs debate the science of mitochondria”. A slightly edited version is reproduced here with kind permission from Bionews.

The regulatory path to clinical trials of mitochondrial replacement therapy (MRT) was recently debated in the House of Commons. MRT is under development in the UK as way of potentially eliminating mitochondrial disease. The techniques essentially swap diseased mitochondria in the unfertilised or fertilised eggs of affected women for putatively healthy ones obtained from a donor. The outcome is unknown and evidence from animals suggest mismatching may occur between the nuclear DNA from the mother and the mitochondrial DNA (mtDNA) derived from the donor (1). The debate was called by a group of backbenchers that are unhappy about the pace with which the Government is apparently moving towards changing the regulations. The split in views in the chamber was fairly even.

From my perspective as a research scientist, what I found most interesting was how the science behind mitochondrial disease and MRT was discussed. Genetics is an extremely complex subject; while scientists are still unsure how genes and genomes cause disease and impact on our physical appearance and personality, there are clearly misconceptions about mitochondrial genetics repeated during the debate that are not supported by current scientific evidence.

For example, David Willets, the former minister for universities and science, who is widely respected within the scientific community, employed the much-used analogy that mitochondria are like batteries for the cell. He followed on to claim therefore that the DNA within ‘does not affect identity’. It’s true that mitochondria (the cellular subcomponent or organelle) are absolutely where energy is made available for all other cellular processes and so perhaps can be thought of as the cell’s ‘batteries’. But mtDNA is not a battery and the influence of genes in mtDNA on traits beyond battery-like function is well established.

Does mtDNA influence an individual’s identity then? In its review, the Nuffield Council on Bioethics thought not, but the answer to this question depends on how you define identity. Genetic variation in the mitochondrial genome has been shown to influence a range of traits including cognition, fertility, ageing and lifespan (1); arguably these are indeed part of an individual’s personal identity, they are certainly an important part of an individual’s characteristics (consider, for example, contemporary ideas about masculinity).

Chi Onwurah supplemented this idea with actual numbers of genes involved: ’13 out of 23,000′ would be replaced. The numbers are approximately correct, but the mitochondrial genes are all essential – in contrast to many of the genes in the nuclear DNA. Consider for example the Y chromosome, which contains a couple of hundred genes, and yet half of the population do quite well without any of them (females). It is therefore a contradiction to claim that mtDNA is not important for an individual’s characteristics (scientists call this the phenotype) while at the same time acknowledging that changes in the mitochondrial genetic code are important for an individual’s risk of disease (again part of an individual’s phenotype).

A second point of contention was whether MRT constituted a form of genetic modification (GM). This is a highly charged subject and there were several MP’s that were adamant that it is not. The Department of Health’s definition of GM is ‘the germ-line modification of nuclear DNA (in the chromosomes)’, thereby specifically excluding the mitochondrial genome. They do however state that MRT is a form of germ-line modification. This makes no sense logically or scientifically, since it is modification of the germ-line that is a key defining feature of genetic modification (also a point of confusion for the chief medical officer, Dame Sally Davies). It has been argued that since the entire mitochondrial genome is replaced as a whole it does not constitute GM, but there will be differences between the sequences of the mtDNA removed and donated, and these will enter the germ-line.

How will accepting these points make a difference in the debate? First, if the general public and MPs understand that mtDNA does much more than just provide the genetic code for making more battery parts, then the idea of tinkering with the genetics of mitochondria may be a much less appealing prospect. Genetic modification is obviously a highly contentious proposal for the human germ-line. I can envisage scenarios where GM in humans could perhaps be justifiable (e.g. correcting point mutations) and ethical, but an informed debate on these issues, whether in Parliament or elsewhere, needs to start with the facts, and at the moment those facts do not seem to be filtering through.


How would you define genetic modification? Comments welcome.