This post describes the three main theories of genomic imprinting, interpreting the explanations within the review article “The evolution of genomic imprinting: theories, predictions and empirical tests” by MM Patten et al (2014) in Heredity.
What is genomic imprinting? In sexually reproducing species, generations inherit one full set genetic blueprints from each parent. Half of their DNA comes from their mother and half from their father. Not all of these genetic blueprints are used, however. Some genes express (or use such that it shows up in the physical makeup or phenotype of the organism) one copy preferentially over another based on DNA sequence or the product of the gene, some express one copy randomly, and some express both. For a very special set of genes, which allele is expressed depends on which parent it came from. An allele is a copy of a gene that has a sequence variation that distinguishes it from other types of copies that might be inherited. Two alleles in an organism represent the same gene, but each was inherited from a different parent and might display a different phenotype. Genomic imprinting occurs when an imprinted gene is expressed only when inherited from the parent of a particular sex: it is expressed only if inherited from the mother or only if inherited from the father. This is different than other modes of expression because it is not based upon the sequence of the allele but rather upon the allele’s origin. But the parental origin of the allele is not encoded in the DNA! This allele could have been passed to the parent from either sex in the grandparent’s generation, so there is no consistency in which gendered parent passes it down, as there is with Y chromosome genes, where only the father passes them on. There then need to be epigenetic markers on the DNA outside of the code that label it as maternally inherited and expressed or paternally inherited and expressed. An epigenetic marker is a protein or chemical addition to the external structure of the DNA that encodes information about how that section of DNA should be used. Because the epigenetic markers can be altered, the pattern of genomic imprinting can be different in different tissues in the organism. For example, the maternal copy might be expressed alone in some tissues while the paternal copy is expressed alone in others. This means that genomic imprinting can lead to very different phenotypes deriving from the same gene, depending on which allele is silenced, or not expressed.
Why are we interested in the evolutionary origins of genomic imprinting? Like most phenomenon in epigenetics, social context plays a large role in explaining the presence and evolution of imprinting. Knowing what selection forces may have driven the evolution of imprinting will tell us something about the relative significance of various elements of the social contexts in which mammals, including humans, develop. Understanding the purpose and nature of these epigenetic mechanisms also affords some hope of manipulating them to remedy damage and improve overall health.
The three main theories of the evolution of genomic imprinting are the kinship theory, the parental conflict theory, and the co-adaptation theory.
The Kinship Theory, pioneered by Haig (2002), focuses on genes that govern some physiological or behavioral interaction between individuals. This is especially important when individuals are differently related, more or less closely, to other individuals in their environment. For example, if you live in a household with only siblings with whom you share both your mother and your father, you all are equally related. However, if you live in a household in which you live with some full siblings and some siblings who all share your mother but have different fathers, you are more closely related to some siblings than others. This situation changes the effect your behavior and physiology have on the fitness of the whole group of siblings. (Fitness is an individual organism’s likelihood of survival and reproduction.) If, for example in a mouse litter, one pup fetus grows larger than its half-siblings in the same gestation, it may demand more than an equal share of maternal resources and decrease the fitness of the half-sibling fetuses that lack that paternally inherited gene variant or allele allowing that pup fetus to demand more resources. While all pups in this scenario share equal relatedness through the genes the inherited from the mother, the possible genes inherited from the father are more variable and may provide a competitive edge.
Imprinting in this case is an evolutionary attempt to maximize the benefit a pup will gain from the competitive edge it might get over siblings with a different parent. While the mother maximizes her fitness by distributing resources to her pups equally, the father maximizes his fitness by making sure each of his pups takes as many resources from the mother as possible. He can use the resources of many mothers to produce pups whereas each mother can only use her own resources to produce her pups. Therefore, if a pup’s paternally-inherited allele (gene variant) improves its ability to acquire resources relative to its half-siblings, there is selective pressure to express that particular copy. The copy inherited from the mother is more likely to be related to its half-siblings (50/50 chance sharing the same maternal copy because the mother passes down one of two copies she possesses and 0 chance of being related in the paternal copy because they are different fathers) and therefore will have almost no beneficial edge. Natural selection will therefore, in this case, try to silence the maternal copy of the gene in favor of expressing only the paternal, competitive, copy. Keep in mind, however, that in the case of a gene which improves an organism’s fitness because it more closely matches its maternal half-siblings, the paternally-inherited gene will instead undergo selective pressure to be silenced.
The Parental Conflict Theory or Sexual Antagonism, by Day and Bonduriansky (2004), relies on sex-specific selection pressure. If, for highest fitness, males and females of a species require two very different sets of phenotypic qualities or genotypes, they are experiencing sex-specific selection pressures. This means that for some genes, the alleles most likely get passed down to the next generation by mothers is a different set of alleles than those most likely to get passed by fathers. This is because more mothers with that set of alleles survived and reproduced and more fathers with the other set of alleles survived and reproduced. However, the next generation, which now possesses one allele from each parent, also needs to survive under the same sex-specific selection pressures. Possessing one maternally inherited allele more likely to be helpful to females and one paternally inherited allele more likely to be helpful to males means that the offspring is not adapted to the sex-specific selection pressures for either sex unless it chooses to express one or the other allele. The offspring needs to express only the allele most suitable for its sex. Imprinting improves, over random 50/50 selection, the offspring’s chances of expressing an allele more adaptive for its sex because it selects for expression the allele passed down by the same sex parent. On a population-level average, the same-sex parent is more likely to have passed down an allele advantageous for the individual organism, simply because more parents of that sex that survived possessed and passed down that allele.
The Mother-Offspring Coadaptation Theory, generated by Hager (1999), highlights the impact of interactions between mother and offspring on the fitness of a phenotype. In this case, an increase in fitness due either to matching or diverging expression of a phenotype that influences mother-offspring interaction could drive the evolution of specific imprinting as a mechanism to achieve this match or mismatch.
What this means is that if the mother and the offspring are more “fit” (more likely to survive and create more offspring) if they both express the same allele, it is better if the offspring expresses only the copy of that gene it received from its mother because it is more likely to match the particular copy the mother expresses (50/50). If it expressed the copy it received from its father, it is less likely to match because the father is assumed to likely have two entirely different variants of that particular gene than those the mother possesses. The chance that the offspring’s expressed allele will match the mother’s expressed allele will be close to 0 if it expresses the copy it received from its father. Evolutionary pressures therefore would favor imprinting because the offspring assisted by imprinted expression of the maternally inherited gene copy would be more likely to survive and reproduce than those who chose randomly or chose only the copy received from the father.
This mechanism can also work for cases in which the mother and offspring should mismatch, expressing very different variants of the gene. Here, the selective pressure works the opposite way, making it more favorable to choose to express only the copy inherited from the father because it would be less likely to also be shared with the mother.
The key to understanding the selection pressures in this mechanism is to understand that the model assumes not a system in which there are only two possible variants (alleles) of the gene but rather a system in which there are many, many possible variants of the gene and that the two parents are not likely to share variants of the gene in common.
The differences in the theories of the evolution of genomic imprinting do not mean that one mechanism excludes the others. It is very likely that the theories apply to very different sets of genes. These sets of genes may be unified by a common theme: sexual antagonism might explain the imprinting of genes involved in placental and fetal growth whereas mother-offspring coadaptation might explain the imprinting of genes governing nursing and pup-rearing behaviors. The sets of imprinted genes each of these theories explain may also differ between species. What works in the mouse model may be useful for exploring the human model, but it is an imperfect facsimile. Mouse evolution differs from human evolution, so while it may provide clues into how we work, we must be careful in extrapolating results in one model to explain something more complex in the other.