Mutations at existing cis-regulatory elements (primarily at enhancers) are thought to be critical in generating most evolutionarily novel structures. However, there are other mechanisms, and these processes, albeit rare, may be responsible for reorganizing the types of genes expressed in a particular cell (Lynch et al. 2011).
Imagine a system where new enhancers can be added to a genome by a virus. These viruses would integrate into the host DNA and place the enhancers next to the host’s genes. If the host cell were making a transcription factor that bound to the new enhancer, the gene could now be regulated by that transcription factor.
This actually seems to have happened, only the mobile genetic element was not a virus—it was a transposon, a transposable DNA element. Transposons are DNA sequences that can travel from one species to another and integrate into the host’s DNA; once there, they become part of the host genome. They are simpler than viruses, being a DNA sequence with ends that can be ligated into the genome. It is thought that transposons are transmitted by parasites or symbionts that can move between hosts (Gilbert et al. 2010; Schaack et al. 2010).
DNA from transposons may play significant roles, since nearly 50% of the human genome consists of such transposable DNA sequences (Lander et al. 2001; Ostertag and Kazazian 2001; Cordaux and Batzer 2009). This is not all “junk DNA”; about 8% of functional human genes contain sequences derived from transposons (Wagner and Lynch 2010). In some instances, the transposons carry enhancer sequences that can be bound by transcription factors, and it appears that such transposons have been extremely important in the evolution of the placental mammals (Figure 1; Fleschotte 2008).
All mammals have hair and mammary glands. Female placental mammals also have a uterus that can retain the growing fetus and provide it with nutrition, oxygen, and protection from the mother’s own immune system. For this to happen, the mammary glands and uterus respond to progesterone, the hormone that maintains pregnancy and initiates milk production. Progesterone in the uterus activates the Foxo1a gene (Kyo et al. 2011) and increases the expression of Hoxa11 (Taylor et al. 1999). The Hoxa11 and Foxo1a transcription factors were mentioned earlier as binding together on the enhancer of the Prolactin gene to activate that gene. But that enhancer is mammalian-specific. More than that, it is located on MER20, a mammal-specific transposon.
MER20 is a transposable element found only in placental mammals. It contains binding sites for Hoxa11 and Foxo1a, as well as for the transcription insulator CTCF. Chromatin immunoprecipitation finds Hoxa11, Foxo1a, and CTCF, as well as transcriptional activator p300 and RNA polymerase II, together at the MER20-derived enhancers. MER20 is found in numerous places in the human genome, including the progesterone-responsive enhancer of the Prolactin gene (Lynch et al. 2011). Indeed, progesterone induces about 1500 genes in human uterine stroma cells, and about 13% of them are very close to MER20 DNA elements.
MER20 may not be the only transposable element involved in mammalian pregnancy. A second transposable element, MER39, is a primate-specific transposon found (among other places) in the promoter of the Prolactin gene (Emera et al. 2011). There is also evidence that some of the proteins used in the formation of the mammalian placenta were originally retroviral envelope proteins that became recruited for a new function (Dupressoir et al. 2011). Thus, transposable DNAs may have “rewired” gene regulatory networks and contributed to the evolution of the mammalian uterus.