First unveiled in mammals in 1984, the term ‘genomic imprinting’ (GI) describes how gene expression varies depending on the parent of origin. Genes encode proteins, but whether these proteins are produced depends on markers attached to the genes; genomic imprinting describes how the expression of specific genes – called 'imprinted genes' – depends on whether they're inherited from your mother or your father. Imprinted genes are suppressed, or 'silenced,' via robust epigenetic mechanisms, specifically on the maternal or paternal chromosome during gametogenesis (the formation of sex cells). The result is that only a single gene version, or allele, is expressed in offspring, while the other allele is silenced. Currently, about 260 genes in mice and 228 in humans are known to undergo imprinting. These genes significantly influence several physiological and behavioural aspects1.

The impact of GI is often particularly influential during in utero and postnatal development; many paternally expressed genes promote foetal exploitation of maternal resources and enhanced growth, while the maternal genome dampens growth and promotes harmony between the mother-offspring unit. However, GI's pervasive effects on the brain and periphery can extend into adulthood, occasionally giving rise to maladaptive cognitive impairment or unusually acute intelligence.

Genes from an animal’s mother and father have different incentives

The parental implications of reproduction are highly asymmetrical, pushing the ‘Sexual Conflict Theory’ into the foreground as a convincing explanation for the evolution of GI. Males, unencumbered by the physiological burdens of pregnancy and lactation, are evolutionarily driven to enhance offspring viability and competitive advantage. It is in their interests to enhance their progeny’s ability to both extract as many nutrients as possible in utero and dominate their siblings before weaning.

In contrast, even within species where females can nurture individual litters from various fathers (e.g., cats), all of their young carry their genes. This encourages the maternal genome to drive the even partitioning of resources among her young and their cooperation. The mother’s future fitness must also be protected, as reflected by maternal imprints that limit placental and foetal growth. With this arena of parental conflict present, it seems inevitable that evolution will favour the paternal silencing of genes that attenuate growth and assertiveness, aiding the propagation of the male animal's genes at no cost to him.

Even in monogamous species, where parental investment is more equal, epigenetic regulation still favours parental interests because a male always forgoes pregnancy and lactation. The earliest evidence for the conflict theory came in 1991, when the genes Igf2r and Igf2 were found to be imprinted in mice. IGF2 (insulin-like growth factor type-2 receptor) heightens cell proliferation in many tissues, while cell-surface IGF2Rs decrease its activity. Igf2 is paternally expressed, and Igf2r is maternally expressed, driving growth in diametrically opposed directions2.

Insights from chimeric mouse embryos: what happens when one parent’s genome prevails?

Healthy embryogenesis in mice, like in many mammals, requires genetic contributions from both parents. When embryos possess only maternal (parthenogenetic) or paternal (androgenetic) genomes, they typically fail to progress beyond the early post-implantation stage, leading to their demise. To investigate the roles of parental genomes in embryo development, researchers can artificially create what are called ‘chimeric embryos’ by adding normal somatic/bodily cells (N) to parthenogenetic (PG) and androgenetic (AG) embryos. This intervention allows these embryos to survive gestation, albeit with notable differences compared to embryos with both maternal and paternal genomes intact.

The resulting PG-N chimeras, which incorporate normal somatic cells into parthenogenetic embryos, typically exhibit smaller body sizes compared to controls. Conversely, AG-N embryos, where normal somatic cells are added to androgenetic embryos, tend to display larger body sizes with a significantly overdeveloped trophectoderm, a critical layer of cells in early embryo development that later becomes the placenta. But the brains of PG-N mice are significantly larger than those of controls, with particularly enhanced forebrain development (which later becomes the cortex), while AG-N chimeras present with decreased brain size.

Collectively, these results support the conflict theory, with the male genome prioritising bodily growth at any cost, and the female genome prioritising cortical development to make offspring as easy to communicate with and teach as possible. Analysing such mice in adulthood shows that maternal gene expression is absent in the hypothalamus and negligible in the brainstem, while the paternal genome contributes far less to protein levels in the frontal cortex, striatum, and hippocampus. The maternal genome therefore seems to prioritise higher-level cognition in mice, while the paternal genome is more implicated in autonomic and endocrine function, incentivising pups to demand milk from their mother and compete with their siblings3.

The mother's genome wants her young to cooperate

Certain imprinted genes promote pups’ survival by facilitating seamless mother-pup interactions. Located in proximal mouse chromosome 7, Peg3 (paternally expressed gene3 is maternally silenced and encodes a DNA-binding protein. PEG3 is abundant in the placenta and the hypothalamus, normally working simultaneously in the foetal placenta and the maternal hypothalamus to orchestrate appropriate foetal growth, and after birth, feeding and nurturing behaviours from the mother. Naturally, knockout pups show delayed post-partum growth by 2 weeks of age and delayed puberty onset, and knockout mothers are inept at normal care repertoires4.

This interaction is an example of the ‘Maternal-Offspring Coadaptation Hypothesis’, which takes into account the fact that natural selection favours congruence and synchrony between the mother and her different offspring. The Conflict Theory and this one exist parsimoniously alongside each other, since it is in the interest of both parental genomes to promote mother-infant synchrony; their gene propagation is contingent on it. However, parallel to this runs the fact that the paternal genome will always be more inclined to enhance exploitation of the mother, promoting suckling, begging, and hormonal priming of the mother to be devoted and attentive.

Genomic imprinting affects brain development

Autism is widely considered ‘the extreme male brain’, as it involves a proclivity for systemising over empathising, with high-functioning male autistics outnumbering females by 4:1. However, a more even sex ratio is seen in low-functioning autism. A theory growing in traction is that, rather than arising from the presence of high circulating levels of male hormones, autism represents the over-expression of paternally expressed genes during neurodevelopment.

Tellingly, autistic foetuses grow placentas triple the size of normal ones, pointing to the predominance of the paternal genome. Autistic children are also often difficult to verbally influence and uncooperative with their siblings, indicating a potential 'failure' of the maternal genome5. This theory may also explain how autism affects otherwise phenotypically normal females. Both males and females can be physiologically normal, yet possess overactive paternal genes – allowing autistic traits to be present in women in the absence of high testosterone levels during development. Nonetheless, hormones differentially impact the manifestation of the ‘extreme paternal brain’, producing more high-functioning males than females6.

A theoretical framework called the ‘diametric model of cognition’ tentatively considers psychosis the opposite of autism. While autism involves the absence of traits that the maternal brain should favour, psychosis entails paranoia and delusions that may reflect the ‘extreme maternal brain’. Mentalism describes our capacity to think about and reflect on ourselves and others. It allows us to recognise and attribute mental states such as beliefs, desires, intentions, and emotions to the people around us, allowing us to predict and interpret their behaviour. In stark contrast to the mechanistic cognition of autism, schizophrenia involves hypermentalism in the form of person-centred paranoia, delusions concerning oneself and others, and occasionally, magical ideation. Interestingly, the onset of schizophrenia is also much later than that of autism, potentially because hypermentalism takes years to develop to the point of psychosis7.

Prader-Willi and Angelman syndrome: when silencing one allele leads to devastating effects

As well as potentially contributing to extreme phenotypes like autism and schizophrenia, GI can lead to human neurodevelopmental disorders. This mainly happens when uniparental disomy (UPD) occurs. This is when both copies of a chromosome pair, or a portion of it, are inherited from a single parent instead of one from each parent. This can occur due to various mechanisms, such as trisomy rescue or gamete complementation. UPD can lead to genetic disorders if the affected chromosome contains imprinted genes, as the individual ends up lacking both copies of a gene that plays a crucial role.

An example of this occurring in the brain is Prader-Willi syndrome (PWS). Babies with PWS are characterised by low birth weight and impaired feeding, followed by compulsive overeating (hyperphagia) in early childhood. Mental deficits are also seen, as well as hormonal issues and stubbornness during childhood. PWS can be mapped to the lack of expression of a paternally expressed gene cluster, Snord116, on the proximal long arm of chromosome 15 (positions q11–q13); both copies of each implicated gene are inherited from the mother and, thereby, silenced.

Another condition caused by GI 'gone wrong' is Angelman syndrome. It's caused by the loss of both copies of the gene Ube3a, also found at 15q11q13. In the brain, Ube3a is only maternally expressed, meaning that its absence in Angelman syndrome can be considered a failure of the maternal genome. Individuals with AS display poor suckling abilities as infants, severe cognitive deficits, speech poverty, seizures, and random bouts of smiling/laughter8.

Paradoxically, the voracious post-weaning appetite in PWS aligns with what would be expected from an enhanced paternal epigenotype. However, considering the biphasic nature of PW, involving poor suckling during infancy abruptly followed by food obsession, casts light on how it may be advantageous to the maternal genome. Mothers often nurture numerous offspring until late adolescence; the maternal genome may, thus, promote childhood foraging, with the paternal genome suppressing the pursuit of supplemental food to encourage prolonged exploitation of the mother’s resources.

However, the poor suckling capabilities of Angelman syndrome babies flag this disorder as potentially being in discord with the conflict theory as well as the co-adaptation hypothesis. Such observations have led to alternative theories concerning the evolution of GI. Rather than attempting to explain its physiological or behavioural implications, these are centred on the effects of natural selection on the epigenetic machinery underpinning it.

Another possibility: does genomic imprinting exist to separate the mother and father's genes?

In 2000, Pardo-Manuel de Villena et al. proposed that GI may have arisen due to natural selection acting to allow for the accurate distinguishability of maternal and paternal homologous chromosomes. Mammalian cells undergo many mitotic divisions post-fertilisation; thus, until meiotic recombination occurs, the two parental genomes need to be mutually distinguishable. Given that epigenetic mechanisms alter chromatin structure, there may be a strong evolutionary drive for the introduction of epigenetic differences to the two parental genomes even without taking into consideration the implications that silencing the genes in question has on the organism9.

This selective force may explain some paradoxical cases of GI, such as Angelman syndrome infants struggling to suckle despite the paternal genome being overexpressed. But for GI to be evolutionary stable, disparities in parental commitments would still mainly keep it aligned with the conflict theory, explaining why most known placental genes that are imprinted and facilitate foetal growth, including Peg3, Peg1, Ascl2, and Igf2, are maternally silenced10.

Genomic imprinting = a hidden force shaping development

In conclusion, GI exerts a potent effect on many facets of behaviour. The conflict theory is well-poised to explain why it has evolved to affect the brain, despite permitting the emergence of disorders that are deleterious to fitness. Parallel to it is the Co-Adaptation Theory, supported by evidence that evolution favours consonance between traits that interact10. Additionally, molecular meiosis- and mitosis-related selective pressures also contribute to the pervasiveness of GI as a phenomenon, as it marks the parental genomes and keeps them separate.

In general, maternally expressed neural genes are biased towards promoting insight and receptivity (cortical development), while the paternal genome enhances the perceived salience of maternal resources (subcortical development). Usually, the maternal and paternal genomes are relatively balanced in humans. But domination of either can occur and is not without consequences. While disorders such as PWS and Angelman syndrome are highly debilitating, a clear duality underlies extreme brain phenotypes like autism and schizophrenia. Higher-functioning sufferers of autism often possess rare cognitive abilities, potentially driving group selection through innovation. Conversely, a keenness to interpret and reflect on others' states, as seen in overdrive in schizophrenia, allows us to thrive in our social milieu as humans when in proper equilibrium.

References

1 Tucci V, Isles AR, Kelsey G, Ferguson-Smith AC, Erice Imprinting Group. Genomic imprinting and physiological processes in mammals. Cell. 2019;176(5): 952–965.
2 Barlow DP, Bartolomei MS. Genomic imprinting in mammals. Cold Spring Harbor Perspectives in Biology. 2014;6(2).
3 Allen ND, Logan K, Lally G, Drage DJ, Norris ML, Keverne EB. Distribution of parthenogenetic cells in the mouse brain and their influence on brain development and behavior. Proceedings of the National Academy of Sciences of the United States of America. 1995;92(23): 10782–10786.
4 Li L, Keverne EB, Aparicio SA, Ishino F, Barton SC, Surani MA. Regulation of maternal behavior and offspring growth by paternally expressed Peg3. Science (New York, N.Y.). 1999;284(5412): 330–333.
5 Anderson GM, Jacobs-Stannard A, Chawarska K, Volkmar FR, Kliman HJ. Placental trophoblast inclusions in autism spectrum disorder. Biological Psychiatry. 2007;61(4): 487–491.
6 Li, J., Lin, X., Wang, M., Hu, Y., Xue, K., Gu, S., Lv, L., Huang, S., & Xie, W. (2020). Potential role of genomic imprinted genes and brain developmental related genes in autism. BMC Medical Genomics, 13(1), 54.
7 Badcock C. The imprinted brain: how genes set the balance between autism and psychosis. Epigenomics. 2011;3(3): 345–359.
8 Smit H. Genomic imprinting and communicative behaviour: Prader-Willi and Angelman syndrome. Netherlands Journal of Psychology. 2009;65(3): 78–88.
9 Pardo-Manuel de Villena F, de la Casa-Esperón E, Sapienza C. Natural selection and the function of genome imprinting: beyond the silenced minority. Trends in genetics: TIG. 2000;16(12): 573–579.
10 Plasschaert RN, Bartolomei MS. Genomic imprinting in development, growth, behavior and stem cells. Development. 2014;141(9): 1805–1813.