Assessing the argument from personal characteristics

First, let us consider the argument from personal characteristics. While the HFEA and the Nuffield Council consider resemblances in personal characteristics to play an important role in grounding the right to know, there is room for disagreement about whether this is indeed the case. For example, Françoise Baylis argues that narrative history is important to individuals' sense of identity in a way that transcends appearances and character traits, and that consequently it is proper to think of children created using donor mitochondria as having three biological parents.11 Though Baylis does not explicitly state what implications this account of identity have for the right to know, it is not a stretch to imagine a defender of this position thinking that knowing the identity of one's biological parents is important for an accurate narrative history. On this view then, resemblances between the donor and the child are not necessary for a right to know to arise. Conversely, authors have also argued that resemblances are not sufficient for the right to know. Consider that in her defence of anonymous gamete donation and closed adoption, Sally Haslanger convincingly argues that the importance placed on the purported resemblances between children and their biological relations is greatly overstated.12

Despite the reservations we may have about the importance of resemblances in personal characteristics for the right to know, it is still useful to examine what follows if we do grant this view for the sake of argument. This is in part because, as revealed by the consultations undertaken by the HFEA and the Nuffield committee, many people in fact take this first premise of the argument to be true.

Turning then to the second premise of the argument, what is at issue is whether mtDNA donation results in similarities in personal characteristics between the donor and the resultant offspring. While both the Nuffield Council and the HFEA accept that donor mtDNA is unlikely to affect the personal characteristics of the children born following mitochondrial transfer, neither offer a definition of ‘personal characteristics’. (***The Nuffield report offers sex as an example of a personal characteristic, but does not attempt to define the category in any great detail.***) Their omission makes assessing the standing of this empirical claim difficult and forces us to speculate about what kinds of characteristics the HFEA and the Nuffield Council had in mind. I suggest that a reasonable way to proceed is to look at the kinds of characteristics that are appealed to when defending the view that there is value in having contact with one's genetic relations on the grounds of resemblances. Perhaps, the most well-known advocate of such a view is David Velleman, and on his account resemblance in physical appearance and personality traits play a prominent role.13 We can thus reasonably speculate that genetic sequences that influence personality and physical appearance would likely count as influencing personal characteristics, while genetic sequences governing deep biological processes that bring about no normally observable difference between individuals would not. For instance, a series of genes that influence one's behavioural response to stress may count as affecting personal characteristics, while genes that cause pragmatically insignificant changes to rate of hair growth likely would not.

As noted by Bredenoord et al,14 contrary to the views expressed by the HFEA, it seems plausible, if not highly likely, that mitochondrial transfer does affect the personal characteristics of the resultant offspring. Someone who does not suffer from a debilitating disease from birth is likely to have a very different personality than someone who is free of such disease. Additionally, as Bredenoord et al also state, there is evidence suggesting that mitochondrial disease affects cognition and behaviour. (***More on this below. ***) However, as stated previously, a mere effect on personal characteristics is not sufficient to ground a right to know; if it did it would lead to an implausible expansion of individuals whose personal information would have to be made available. The more plausible view is that the resemblances between the donor and the resultant child are what give rise to a right to know.

It is possible that although mtDNA affects personal characteristics as suggested by Bredenoord et al, when diseased mitochondria are replaced with those from a donor, the personal characteristics that then arise are encoded by nDNA. What then might result are resemblances between the child and the nDNA provider, but not the mitochondrial donor. Consider the following hypothetical example. Imagine that some mtDNA sequence results in children born without any hair by impeding the expression of genes involved in the production of follicles. Once this mtDNA is replaced with a donor's, the follicles develop according to gene sequences located in nDNA. The result is a child whose hair characteristics resemble the individuals who provided the nDNA and not the mtDNA, though the donated mtDNA did have an effect on the expression of the characteristic. By contrast, we could also imagine a mtDNA sequence that directly encodes a specific trait—say hair colour—and thus results in a shared characteristic with a mtDNA donor. We can therefore distinguish between mtDNA donation that only affects personal characteristics, and mtDNA donation that also results in resemblances in personal characteristics with mtDNA donors. Only the latter would give rise to a right to know.

What matters then is whether mtDNA transmits donors' characteristics to the resultant children. Recall that the Nuffield Council and the HFEA accept the claim (at least provisionally) that mtDNA has no meaningful phenotypic significance at all beyond conferring health or ill health. A plausible reading of this claim is that any phenotypic differences that might arise from mitochondrial donation are due to the restoration of traits that were inhibited by pathological mtDNA, as in the hypothetical hair growth example. However, the empirical evidence is not so clear. There is a growing body of research associating specific mtDNA variants with particular phenotypic traits including personality,15 ,16 psychological disorder17–19 and propensity for developing degenerative neurological diseases.20 Of these examples, the link between mtDNA and non-pathological behavioural traits, such as extroversion, is the weakest, and so one may argue that these examples at most suggest that certain pathologies, such as bipolar disorder and certain forms of dementia, are in fact hitherto unrecognised mitochondrial diseases. Consequently, the view that mtDNA confers nothing but either health or ill health might still remain unthreatened. Despite this possible interpretation of the evidence, there is good reason to question the conclusion that mtDNA does not impart resemblances to offspring.

Many psychiatric disorders exist on a continuum with behavioural traits that are within the non-pathological range of behavioural norms. Consequently, it is well within the realm of possibility that individuals who inherit milder forms of mutations similar to those suspected to be involved in psychiatric disorders might inherit behavioural characteristics in the non-pathological range. The possibility of non-pathological behavioural effects is made even more plausible in light of a particular feature of mtDNA inheritance previously mentioned—heteroplasmy. As noted previously, individuals can inherit multiple different mitochondrial genomes and thus can have differing degrees of penetrance of mutated mtDNA in their tissues. It thus may well be the case that some elements of personality are inherited via variants in mtDNA even in the absence of disorder.

This argument does not definitively establish that particular mtDNA variants encode specific personality traits in individuals free of mitochondrial disease, but it does show that the possibility that mtDNA has such effects is not out of keeping with current research into the role of mtDNA. Furthermore, it is worth noting that the link between nDNA and personality is itself hotly contested—a recent article argues that a meta-analysis of research examining the genetic inheritability of personality suggests a null hypothesis. The authors of this study even state that, ‘the search for genetic mechanisms of human personality, in our view, will never bear fruit’.21 There are of course others who disagree.22 The upshot here is that unless the view is that physical resemblance alone (for which there is unequivocal evidence of genetic inheritance) is sufficient to ground a right to know, it seems hard to firmly ground the ‘discordance’ between the phenotypic importance of nDNA and mtDNA on the available empirical evidence; in both cases, there is much uncertainty. Given the absence of a detailed understanding of the genetic basis of personality in general, and the evidence that mtDNA may indeed play a role, it is overly hasty to justify policy on the grounds that mtDNA has not yet conclusively been shown to play a role in the transmission of personality traits.

Taking a definitive policy stance in the face of empirical uncertainty could risk causing problems down the line, if future research establishes that mtDNA plays a role in the transmission of personal characteristics. Either children born using donated mtDNA would have to continue to be denied a right to know in spite of such findings, or a retroactive change in policy would have to impinge on the privacy interests of individuals who had donated mtDNA with the understanding that their identity would be protected. The latter is not a mere theoretical possibility, as evidenced by recent controversies in Australia23 and British Columbia.24 Neither denying children their rights nor retroactively changing the law are appealing outcomes. Given the empirical uncertainties, a less definitive policy is likely more appropriate. Instead of outright declaring that no right to know arises in mitochondria donation, policy could reflect the current empirical uncertainty by explicitly leaving open the possibility of revisiting the question. For instance, potential donors could be informed that though the current state of evidence does not establish a right to know, if this were to change then mtDNA recipients would be granted access to identifying information about their donors. This would provide individuals strongly opposed to having their identity revealed with forewarning about possible retroactive changes in policy, while at the same time building in regulatory flexibility that is appropriate, given the current state of evidence.

Assessing the argument from genetic connection

Consider now the second argument, that there is no right to know in the absence of a ‘unique genetic connection’ to a donor. Here again, the Nuffield Council declines to define their terms, so some speculation about what they mean is required. Two possibilities seem to naturally present themselves. The first is a pragmatic criterion: a donor and the child share a unique genetic connection iff it is possible to uniquely identify the donor as the child's progenitor on the basis of genetic analysis alone. The second is an ontological criterion: a donor and the child share a unique genetic connection iff it would have been impossible (or highly unlikely) for the child to have the same genetic makeup but a different donor. The argument from genetic connection can therefore be presented in the following manner:Since mitochondria do not undergo recombination, and mitochondrial inheritance is strictly maternal, any one of the donor's close female family members could have served as a substitute mitochondrial donor without affecting the mitochondrial genome inherited by the resultant child. Consequently, on both the pragmatic and ontological reading of the criterion, mitochondrial donation does not result in a unique genetic connection between donor and child. Therefore, no right to know arises.

However, on closer inspection the argument is not as plausible as it may initially seem.

First, the pragmatic criterion fails to capture our intuitions about the standing of the right to know in gamete donation cases involving monozygotic siblings. Consider a child created using the sperm of a man who was a member of a set of monozygotic triplets. Genetic testing alone would be unable to determine which member of the three siblings was in fact his/her progenitor, and so on the pragmatic reading of the criterion there would be no unique genetic connection between the child and the man who in fact donated sperm. However, it seems problematic to deny children created in such circumstances the right to know while extending this right to other donor-conceived children.

The ontological reading of the uniqueness criterion better captures our intuitions in the monozygotic sibling example because it provides justification for treating as important the link between the monozygotic sibling who actually donates gametes and the child that results. This is because the individual who in fact donates determines the exact combination of nuclear genes that are passed on to the resultant offspring. Consider that if one of the siblings would have stood in place of the actual donor, the resultant child would have, in overwhelming likelihood, inherited a different complement of paternal nDNA. Consequently, there is a kind of non-fungible link between the particular donor and the child that is created.

Someone may argue that the ontological reading demonstrates that mtDNA does not create a unique genetic link, since there is no recombination of mtDNA and so no ‘shuffling’ of the genes that get passed on to subsequent generations. We might then presume, as does the Nuffield Council, that there would be no variation in the mtDNA inherited by a child even if one of the donor's close female relatives would have donated instead. This would be unlike the sperm donation case, where the identity of the individual who in fact donates plays a role in determining the precise nDNA that is inherited by the resulting offspring.

However, this conclusion does not take into consideration the phenomena of heteroplasmy described earlier in the paper. Since different tissues possess different relative amounts of mtDNA variants, including variations in mtDNA ratios in ova, the precise makeup of mtDNA variants passed on will depend on the specific ovum used in mitochondrial donation. Consequently, the particular individual who in fact donates mitochondria does play a role in determining the specific mtDNA inheritance of the resulting child, and thus on the ontic criterion a ‘unique genetic connection’ could arise.