By comparing genes in the opossum and the platypus, Duke University Medical Center researchers have uncovered evidence that questions the origin of "genomic imprinting" -- a process by which a gene's expression is governed solely by which parent donated the gene copy, rather than by the classic laws of Mendelian genetics, in which genes are either dominant or recessive. The findings also resolve questions surrounding the structure of the mammalian evolutionary tree and may open the door to better evaluation of carcinogenic agents.
The theory of the evolution of genomic imprinting holds that imprinted genes, which usually are related to growth and development, represent a genetic "battle of the sexes" -- a competition between paternally imprinted genes that lead to enhanced fetal growth and maternally imprinted genes that restrict growth, saving nutrients for the mother herself. According to this theory, animals without a lengthy fetal development stage -- such as marsupials, whose infants leave the womb while still embryonic and develop in external pouches, and egg-laying monotremes, the most primitive order of mammals -- were not expected to be imprinted.
However, in the April 21 issue of the journal Molecular Cell, a research team led by graduate student Keith Killian and principal investigator Randy Jirtle reports that a particular gene, called M6P/IGF2R, is imprinted in the marsupial opossum but not in the platypus, a monotreme. Until now, imprinted genes had not been examined in such primitive mammals. "This finding suggests that M6P/IGF2R imprinting evolved a lot farther back than we thought, in a precursor animal between what are currently the monotremes and marsupials," Jirtle said. He is a professor of radiation oncology and associate professor of pathology at Duke University Medical Center and a member of the Duke Comprehensive Cancer Center.
The M6P/IGF2R gene codes for the receptor for another molecule, called "insulin-like growth factor 2" (IGF2), whose gene is also imprinted in many mammals. Both genes are involved in growth, but imprinted M6P/IGF2R is maternally expressed, while IGF2 is paternally expressed.
The researchers' finding that M6P/IGF2R is imprinted in the opossum, but not in the platypus, supports the idea that imprinting is due to a "battle of the sexes," but the primary requirement seems to be competition for survival among the young, rather than a lengthy fetal development period. "Young opossum must crawl to reach the pouch and latch on, and this is consistent with a battle of the sexes for nutrient allocation as a reason for imprinted genes," said Jirtle. "For the opossum, it's literally a race for life. But since opossum have only a 13-day gestation period, a lengthy fetal development stage is not necessary for imprinting to arise." Unlike puppies or piglets, once newborn opossum latch on to the mother's nipple, they can't let go until they grow large enough to open their mouths. Since there are only a dozen nipples for up to 50 babies in a single litter, competition is important. The faster a baby opossum can climb, the more likely it is to survive, and the more likely its genes will be passed on. On the other hand, there's no active competition on the part of young platypus.
While competition to ensure genes' survival can explain why imprinting first evolved for marsupials, it doesn't answer why imprinted genes still exist in higher mammals, such as humans. In lower mammals a single group of offspring can be fathered by more than one male, but in higher mammals that is very unlikely, making competition between offspring to keep their fathers' genes in the pool unnecessary. "About 30 imprinted genes are known, but there may be as many as 500. We don't know if humans need these imprinted genes, or if we would be OK without them," Jirtle said. "Nevertheless, we have them and as a result of their limited expression, they become targets for developmental diseases and cancer." If something happens to the functional copy of an imprinted gene, there's no back-up as with other genes. "These are growth promoting and growth inhibiting genes, for the most part," Jirtle explained. "If the balance is altered, you could lose a tumor suppressing gene or turn on an oncogenic one."
In fact, mutations of IGF2 and M6P/IGF2R appear to be early steps in a wide variety of cancers, scientists say. In many tumors, both copies of growth-inducing IGF2 are turned on, despite the gene being imprinted in humans. For M6P/IGF2R, which is not imprinted in humans, both copies normally function. However, in more than 60 percent of human liver cancers, 30 percent of breast cancers and 50 percent of lung cancers at least one copy of this growth-suppressing gene doesn't work.
Since humans have two functioning M6P/IGF2R genes, but mice have only one due to imprinting, it's reasonable to think that humans should be more resistant to cancer-causing agents. Right now, however, possible carcinogens are tested on mice. Jirtle and Killian want to develop a "better" mouse -- one with two working copies of M6P/IGF2R that might better reflect human susceptibility to cancer-causing agents.
Another of the study's findings should help them figure out how to turn on the mouse's silent M6P/IGF2R gene. While other researchers have reported finding the region of DNA that controls the imprinting of M6P/IGF2R in mice, this entire region was missing in the opossum. Since Killian has no doubt that the gene they cloned in the opossum is in fact the homologue of M6P/IGF2R, the current finding likely means that the actual controlling region has yet to be found. "The proposed controlling element is not there in the opossum," Jirtle explained. "Either the opossum has a completely different, unique, and as-yet-unidentified mechanism to control imprinting of this gene, or the proposed region in the mouse is not the actual controlling region." If they can find the region in the opossum that turns the imprinted genes on and off, and if they can then locate that region in the mouse, the researchers could potentially develop a strain of mice that can't imprint the M6P/IGF2R gene. If so, they'll have found their better mouse.
The scientists' studies of imprinting also have important implications for understanding the mammalian evolutionary tree, Jirtle said. The presence of imprinted genes in the opossum but not platypus suggests that marsupials are more closely related to eutherians -- the name for mammals whose offspring develop in the womb -- than they are to monotremes. This relationship supports the proposed version of the mammalian evolutionary tree in which monotremes branched off the main evolutionary trunk before marsupials; rather than another version that places marsupials and monotremes as two twigs on a single branch. In order for the second proposed evolutionary tree to be correct, given the current finding, imprinting would have had to evolve twice -- once on the marsupial twig and again on the eutherian branch -- in a process called convergent evolution. That's very unlikely, say the scientists.
Jirtle and Killian also say the findings demonstrate the validity of using primitive mammals to learn about genes' evolution and function. "Marsupials can be an important tool in understanding genetics and imprinting," said Killian, an MD/PhD candidate. "Mice and humans are too closely related to distinguish some evolutionary developments from random events."
Co-authors on the paper are James Jirtle of Duke University Medical Center, James Byrd and Richard MacDonald of the University of Nebraska Medical Center, Barry Munday of the University of Tasmania (Australia), and Michael Stoskopf of North Carolina State University.
Additional information on genomic imprinting can be found on the Internet at http://www.geneimprint.com.
Journal
Molecular Cell