Skip To Content
Cambridge University Science Magazine

It is a scientific cliché that humans share 99.9% of our DNA with each other, making us remarkably similar. Yet, despite this apparent homogeneity, even 0.1% difference confers a huge degree of variation, largely centred around the Major Histocompatibility Complex (MHC). This region of the genome encodes around 200 genes implicated in many functions which primarily involve immunity, but also influence the brain, gut, and other organs. In fact, the MHC may even bias whom we are attracted to and who is attracted to us.

The MHC is a section, or locus, of our DNA encoding cell-surface proteins called MHC molecules. These are expressed on nearly all cells and present protein fragments, or peptides, to the immune system to help identify pathogens. The MHC locus is both highly polygenic, meaning it includes many genes, and highly polymorphic, meaning each gene has many alleles. Individuals can be grouped by their particular combination of MHC alleles, known as their MHC haplotype. Importantly, MHC genes are expressed codominantly, so if chromosomes with different haplotypes are inherited from each parent, children will express both haplotypes and have a wider range of MHC molecules. This extends the repertoire of peptides that can be presented to immune cells, leaving us better equipped, both individually and as a population, to survive disease.

In this context, the importance of maintaining this variation is clear. Firstly, progeny of parents with dissimilar MHC loci will have an increased range of peptides to prime immune cells so the body is better prepared to attack pathogens and resist attacking itself. Secondly, similar MHC haplotypes may indicate familial relation, so choosing a partner with a dissimilar haplotype, a process known as MHC disassortative mating, helps to prevent inbreeding. For these reasons, disassortative mating is expected to improve species survival.

Given this predicted evolutionary benefit, scientists wondered whether MHC haplotypes contribute to mate-selection in practice. To investigate this, Kunio Yamazaki and colleagues observed the mating choices of caged inbred mice. They isolated the effect of MHC variation by using mice strains with different MHC haplotypes against an otherwise indistinguishable genetic background. When a male mouse was presented with two females, MHC genes influenced mate choice. Similar results were seen for stickleback fish, showing that at least for non-human animals, the MHC has a definite role to play in partner selection.

How can these MHC differences be detected? One much-researched theory is by smell. Olfactory stimulation is a well characterised method of sexual attraction, the archetypal example being sex pheromones. MHC or bound peptides could act in a similar way. Several theories have been proposed to explain the effect of MHC on body odour, which tend to centre around volatile MHC-derived or MHC-bound molecules, or, more indirectly, around MHC genes influencing the ‘friendly bacteria’ in our gut. Perhaps a likely explanation is that MHC molecules restrict the available pool of MHC-binding peptides, which are then degraded and made volatile by our gut bacteria to produce a scent. This olfactory mediation is affected by both partners, meaning that odours may not be universally attractive - instead, it may be that individuals are attracted to particular smells given off by individuals with dissimilar MHC genes.

Several years after their initial experiment, Yamakazi and colleagues conducted follow-up experiments demonstrating that mice can differentiate the odours of mice differing only at MHC loci. This ability is conserved between species: mice and rats are able to distinguish between urine from humans with different MHC genes, possibly through volatile molecules.

Human pheromones?

However, whereas many mammals rely heavily on smell, humans prioritise sight and sound. Additionally, human mate selection includes more factors (or so we would like to think). The next step, therefore, was finding evidence for olfactory-mediated MHC-disassortative mating in humans.

One way of investigating this is analysis of MHC haplotypes in isolated populations. Christopher Ober and his team studied mate choice in Hutterite populations — colonies near the USA-Canada border isolated by religious beliefs. To find partners, adolescents travel to other Hutterite colonies, resulting in repeated patterns of marriages and higher levels of homozygosity in Hutterite genomes. Despite this, MHC haplotypes of Hutterite spouses are more different than would be expected by chance, and their progeny have more heterogeneous MHC loci than the rest of their genomes. This indicates a preference for MHC-dissimilar partners, confirming the results from mice. Furthermore, Ober found that foetal loss increased significantly when parents had similar MHC haplotypes. Whilst he could not conclude that MHC similarity caused this (rather than linked genes, for example), MHC may be implicated in the role of immune cells in distinguishing foetal cells as ‘self’ rather than invasive. This is supported by further research showing this bias is solely dependent on maternal, rather than paternal, MHC genes. Put simply, humans are attracted to partners with MHC genes different from their mothers. (Take that, Freud!)

Through this analysis, Ober showed a tendency towards MHC-disassortative mating in humans, but it was Claus Wedekind who conducted a now-fabled experiment to assess smell-dependence. Wedekind instructed 44 males and 49 females to prime their sense of smell using nasal spray for 14 days and, amusingly, to read Perfume by Patrick Süskind. Following this, the men wore the same T-shirts for two nights and abstained from activities that might change their scent, including sex, smoking, or drinking. The women then ranked the T-shirts’ attractiveness. Wedekind found that women found the smell of partners with dissimilar MHCs more attractive. This demonstrated that MHC preference aims to maximise progeny heterozygosity rather than favouring specific alleles or MHC combinations.

Since these experiments, access to genomic data has increased massively, allowing large-scale analysis of MHC variation. A 2019 study published by the Royal Society analysed the relative contribution of MHC haplotypes in partner selection in 833 European and Middle Eastern couples. Across Northern Europe, spouses were more MHC-dissimilar than random pairs of individuals, and this pattern of dissimilarity was exceptional compared to the rest of the genome. These results were significant in terms of reliability, but the effect was small, suggesting that MHC has a minor but definite contribution to mate-selection. Interestingly, this trend was not as strong in the Middle East, and in Israel there was even evidence for MHC-assortative mating. This could be explained by common marriages between cousins and social homogeneity. Additionally, the similarities between MHC genes were not significantly stronger than would be seen with totally random mating, supporting this hypothesis. These observations indicate that, whilst the MHC may play a role in sexual decisions, contextual factors can supersede this.

So how do all of these findings affect how we will find our match? Already, websites such as invite users to take a DNA test and have their MHC analysed. Users are then matched with someone who has ‘compatible’, i.e., dissimilar, MHC genes. It is possible that in the future, catfishes on Tinder may soon be manipulating their MHC sequencing results instead of profile pictures, but, for now, it is hard to imagine preferring a date who reeks of a repeatedly worn T-shirt over the cool, refreshing scent of deodorant.

Clodagh Bottomley is a second-year undergraduate at Trinity College studying Natural Sciences.