Thanks to your parents, you have two copies of each chromosome, which means that you have maternal and paternal copies of every gene. In most cases, having two copies of a gene is no problem, but in some cases, two is too much, and your cells have to shut one copy down. How does a cell do it?
Shutting down one copy of a gene (or an entire section of a chromosome) is called genomic imprinting. (This is not the same thing as the newly hatched duckling that latches on to the first thing it sees, obviously). Genomic imprinting is a critical process used by placental and marsupial mammals to control the dosage of many genes, but how did this process evolve?
The answer, in part, has been discovered by an analysis of the platypus genome. Genomic imprinting appears to have evolved from a defense mechanism used by cells to knock down parasitic DNA.
How does genomic imprinting work? The basic idea of imprinting is simple: surrounding some genes are stretches of DNA sequence that recruit molecular regulatory machinery which shuts the genes down. Genes that are near these DNA sequences get shut down, or at least one copy does, (either the maternal copy or the paternal copy). The result is that, for whatever reason (and the functional purpose of imprinting is still debated), cells shut down one of two copies of a gene.
Many imprinted genes are expressed in the placenta, and one proposal tossed around is that genomic imprinting evolved to control the regulation of nutrient exchange between the mother and the gestating fetus. If that idea is right, then egg-laying mammals (which don't exchange nutrients with the fetus once the egg is laid) don't need genomic imprinting, and as far as researchers can tell, the egg-laying platypus in fact does not engage in genomic imprinting.
This suggests that genomic imprinting evolved after egg-laying mammals split off from the evolutionary lineage leading to marsupial and placental mammals. That, in turn, indicates that the recently sequenced platypus genome might be a good place to look for clues about the evolution of genomic imprinting. The same genes subject to imprinting in placental mammals exist in the platypus, but in the platypus these genes are not imprinted. What causes this difference? A group of researchers from Australia and the US examined the platypus genome to find out.
It turns out that imprinted genes are surrounded by parasitic DNA elements (also called transposable elements) - essentially dead or crippled viruses which are not infectious, but occasionally still have the ability to hop around within the genome. All of our genes are surrounded by transposable DNA elements, since transposable elements make up a huge fraction of our genomes (we have more than 10 times as much DNA for transposable elements as we do for protein-coding genes and their associated regulatory sequences). While all of our genes are surrounded by this stuff, imprinted genes are more mired in transposable elements than other genes, which suggests that these transposable elements may have something to do with genomic imprinting.
How do things look in the platypus genome? Interestingly, genes that are are surrounded by more transposable elements and imprinted in other mammals are not surrounded by extra transposable elements in the platypus. This again suggests that it's the transposable elements which direct genomic imprinting.
So how can transposable elements control the shut-off of a gene? This is a phenomenon that has been well-studied, so the results of this platypus genome analysis fit in nicely with what's already known. Transposable elements, which are like viruses that can hop around within the genome and cause potentially harmful mutations, are suppressed by a host defense mechanism, which zeros in on these DNA parasites and shuts them down. This shut-down process can also silence nearby host genes.
Thus, imprinting occurs because the DNA parasites sitting next to imprinted genes trigger host defenses that shut down the entire region of DNA. In some cases this effect can be beneficial, and thus favored by natural selection. In the case of imprinting, the evolutionary story might have gone like this: as placental mammals evolved, there was a need for better regulation of nutrient exchange between the mother and the fetus. Fetuses that happened to have transposable elements shutting down certain genes were more likely to survive to birth. Over generations, this process of selecting for certain arrangements of transposable elements would lead to the imprinting system that exists today.
This study of the platypus genome appears to be another example of evolution's opportunism: regulating the dosage of certain genes was beneficial for the developing fetus in placental mammals, and parasitic transposable elements could be recruited to do the job.