The Genetics of Marriage - it's all in the Voles!

green

(Novembert 4th 2008) Single gene variations are now claimed to be indicative of the tendency for men to have stable martial relations. No doubt future disintegrating couples will be able to shrug off their lack of staying power because "it's a problem with my genes" but why, asks Jeremy Garwood, did these researchers look at just a single gene?

Under the title "Genetic Variation in the vasopressin receptor 1a (AVPR1A) gene associates with pair-bonding behaviour in humans", Hasse Walum and Lars Westburg, Swedish behavioural geneticists from the Karolinska Institute, Stockholm, have recently reported that "the well characterised influence of arginine vasopressin on pair-bonding in voles may be of relevance also for humans." (Proceedings of the National Academy of Sciences 2008, 105, 14153-56).

Voles? Yes, the little furry mouse-like creatures! To be more specific, montane voles (Microtus montanus), meadow voles (Microtus pennsylvanicus) and prairie voles (Microtus ochrogaster). For almost two decades, research on voles and other small mammals, has suggested that nine amino-acid long brain neuropeptides, including vasopressin and oxytocin, play hormonal roles in the social behaviour of these animals.

Prairie voles are monogamous and highly social; the first time a male prairie vole mates with a female, he forms a bond for life, breeding and raising successive litters. Conversely, montane and meadow voles are loners, the males treat sex as a succession of one-night stands, do not bond with females or help raise the offspring.

Experiments adding more vasopressin or the vasopressin receptor facilitate the social behaviour and pair-bonding observed in the male voles. The key genetic difference between these voles is a 428 base pair DNA sequence present in the upstream promoter region of the AVPR1A vasopressin receptor gene in the prairie voles.

Well, if it works for voles..."The fact that the corresponding gene has proved important for similar behaviour in voles... suggests that the thoroughly studied brain mechanisms that we know give rise to strong bonds between individual voles can also be relevant to humans", says Walum.

Walum et al. zeroed in on the human AVPR1A vasopressin receptor. Although humans lack the prairie vole's upstream sequence, they do have three repetitive sequences, RS1, RS3, and GT25, that are polymorphic in the human population, that's to say, you can be homozygous, with two copies of each repeat, heterozygous, with just one copy, or you may not have any of the repeats.

Using "mouthwash" samples, DNA was obtained from experimental subjects and the presence of one, two, or no copy of each AVPR1A repeat determined. Only the RS3 "allele 334" was found to have a statistically significant correlation: "Men with two copies of the allele had twice the risk of experiencing marital dysfunction, with a threat of divorce during the last year, compared to men carrying one or no copies," said Walum.

But how did they determine if their human subjects had pair-bonding behaviour comparable to the prairie voles? In the laboratory, voles face the 'partner preference test'. The testing apparatus consists of three chambers connected by tubes. The female 'partner' and a novel female 'stranger' are tethered in their own chambers, whereas the male subject is free to move throughout the apparatus during a 3 hour test. Pair bonding is inferred when subjects spend significantly more time in close proximity to the partner compared to the stranger (partner preference).

But, measuring differences in the affections of laboratory-raised voles is one thing, how do you reliably judge similar characteristics in men? With a subjective questionnaire and lots of statistical analysis, of course!

But who would you choose to ask and what would you ask them? Men designated as antisocial hermetic loners compared to those who have spent 50 years cuddled up to their first love?

Well, it seems unlikely that you would choose the rather special group that Walum had at hand. In effect, he simply plugged into an existing dataset from the "Twin and Offspring Study in Sweden" (TOSS). Sweden has the world's largest and oldest registry of twins, the Swedish Twin Registry. Whatever the original eugenic motivations, this registry continues to provide data for assessing the role of genetics versus environment in health, pathologies, and now, pair bonding.

However, hunting down more information about TOSS reveals that it's quite a restricted subset of twins. Furthermore, the original data was collected, as the name suggests, to search for genetic influences on parenting rather than on how tightly the men felt they were pair-bonded. TOSS intially comprised 544 pairs of adult twin women, their male partners, and adolescent children (limited to one per couple), and 317 pairs of adult twin men, their spouses, and adolescent children (one per couple). In Wallum's study, only 552 of these twin pairs were retained although we're not told how many of the twins were male or female. Most of the participants were middle class, born between 1944 and 1971, and white.

Significantly though, for a study that somehow wants to be comparable to the extremes of pair bonding in voles, that is, a life-long single mate versus zero commitment to numerous mates, TOSS's same sex twins were "required to have a relationship of at least five years with their partner" and one adolescent child!

It's at this point that one might wonder just how varied and representative of human pair bonding this TOSS dataset is. Can one really describe males who have lived for at least five years with their female partners (and probably somewhat longer if they have an adolescent child together, aged 11 to 22 years-old) as poor at pair-bonding? Walum certainly doesn't hesitate to divide the men in these families into good and bad examples of pair-bonding behaviour.

But what are the criteria that make him so certain? In the lab-raised voles, they were measuring their capacity to remain attached to the same female for a few hours, not years! His answer is "self-reporting questionnaires" and lots of statistical rearrangements until "significant" correlations appear.

To do this, Walum has selectively pruned the original TOSS data. In fact, he's severely truncated it! Of the 49 original questions, only 18 were "considered relevant measures of pair-bonding". But even this was too much and after performing a "factor analysis", just 13 questions were retained to generate his all-important 'Partner Bonding Scale' (PBS).

Rest assured though that the retained questions are "in accordance with the behavioural domains observed when studying pair-bonding among non-human primates"! Leaving us with questions from the 'Dyadic Adjustment Scale' (DAS - 7 questions), the 'Support Seeking and Giving' (SSG - 5 questions) and the 'Marital Instability Scale' (MIS - 1 whole question).

Although it's still not clear how the non-human primates may have responded, human aptitude for pair-bonding was determined from answers measured on a "6 point likert scale" to questions like: "Do you kiss your partner?" or "Have you ever regretted getting married/moving in?" Now, from personal experience, one might wonder how much such responses vary according to the mood of the experimental subjects - since when are couples uniform in their affections day-in day-out?

Similarly, just how relevant to "pair-bonding" are responses to statements like: "I find it rather easy to get close contact with other people"? Might this indicate better social interactions, tighter pair-bonding, or simply a propensity for infidelity?

"I seldom worry about being abandoned by other people" and "I don't like it when I have to be dependent on other people"? Remember, these are subjective personal responses by middle-aged men that have lived with the same woman for years - even if Walum's PBS has subsequently classified them as poor at pair-bonding.

When originally published in 2007 (Twin Res Hum Genet 10:299-313), the TOSS data addressed the question: "Can genetic factors explain the spillover of warmth and negativity across family relationships?" It found that "self-reported marital satisfaction was associated with self-reported warmth and negativity" and that "marital warmth was associated with maternal warmth", and vice versa. They even quantified their effect: "On average, genetic factors explained nearly half of the covariance between self-reported marital satisfaction and parenting" and "21% of the covariance between observed marital and maternal warmth". Leading them to the rather vague, yet somehow self-evident, conclusion that "parents' genetically influenced characteristics help shape the emotional climate of the family."

Jenae Neiderhiser, professor of psychology at Penn State, one of Walum's co-authors, said: "Our findings are particularly interesting because they show that men who are in a relatively stable relationship of five years of more who have one or two copies of allele 334 appear to be less bonded to their partners than men with other forms of this gene. We also found that the female partners of men with one or two copies of allele 334 reported less affection, consensus and cohesion in the marriage." However, in an observation that surely raises more questions about the true value of their TOSS analysis, she then noted that: "interestingly, (these women) did not report lower levels of marital satisfaction than women whose male partners had no copies of allele 334"!

Perhaps Walum et al. might do better by distinguishing between pair-bonding and monogamy. Although it isn't often mentioned, monogamy is quite rare in mammals with only 3% of mammalian species showing such bonds, and even fewer displaying obligate monogamy.

Also, some might question whether voles are actually such a good model because genetically, voles are very unusual mammals. The vole genus, Microtus, has over 60 species that have evolved in the last 500,000 to 2 million years. This means that voles are evolving 60-100 times faster than the average vertebrate in terms of creating different species. Between vole species, chromosome numbers can vary from 17 to 64! In one species, the sex-determining X (female) chromosome, contains about 20 percent of the entire genome, in another, females possess large portions of the Y (male) chromosome. In yet another species, males and females have different chromosome numbers, which is very uncommon in animals.

None of which stops Walum's team from hoping that "greater knowledge of the effect of vasopressin on human relations will one day give science a better understanding of the causes of diseases characterised by problems with social interaction, such as autism."

Strangely enough, my English dictionary defines autism as a "morbid absorption in fantasy."