By Kathy Wang ’14 and Logan Chestnut ’14
With the rise of the field of genetics, our genes and the environment we grow up in became polarized forces, competing determinants of who we are. Ongoing scientific research, however, continues to illuminate the fact that they are not two opposing forces; instead, they continuously interact as they drive the development of living organisms. While Mendel’s elegant experiments with peas paved the way for the contemporary model of inheritance, Punnett squares do not tell the whole story of how the expression of our genes leads to our appearance and behavior. For example, while each of our cells contains the same genetic code, each cell expresses different genes to function properly. To unravel some of these complexities, research in epigenetics investigates genetics beyond just inheritance.
To understand how this phenomenon both influences and is influenced by our development, researchers in the lab of Catherine Dulac, Higgins Professor of Molecular and Cellular Biology at Harvard, are uncovering the mechanisms by which differential gene expression influences nervous system structure and function. In two papers published late last year in Nature, Dulac and her colleagues discuss the intricacies behind this process.
Mammals inherit half of their chromosomes from the father and half from the mother. In this way, two sets of genes are inherited and two alleles exist for each trait. However, scientists have discovered the surprising result that mammals do not always express maternal and paternal alleles equally. In fact, alleles from one parent may not be expressed at all. At times, the expression of either the maternal or paternal allele is silenced by chemical modification of DNA, so there is only one copy left that can be expressed. This pattern of gene expression is passed down through generations and can also be altered depending on environmental factors. Dulac and her colleagues are studying this phenomenon, called genomic imprinting. Through DNA methylation and interaction with modified histone proteins, this epigenetic mechanism causes organism-wide silencing of a maternal or paternal allele. Knowing that both copies of genes are usually expressed, the scientists were prompted by this discovery to investigate why this phenomenon exists, and how imprinted genes affect us.
Dr. David Haig, George Putnam Professor of Biology, has worked closely with the Dulac lab on this research, and has pioneered one leading theory behind genomic imprinting. More than twenty years ago, Haig and Mark Westoby, a professor in the Department of Biological Sciences at Macquarie University in Sydney, Australia, proposed a theory based on two behaviors typical of many mammals. First, females are intimately attached to their young for long periods before giving birth. Second, males mate with many females during the course of their life. From these behaviors arise competing goals that the mother and father have for their offspring. The mother would like to produce as many offspring as possible while the father would like to make sure that his offspring survive in the womb and outcompete those of other males. In terms of maximizing evolutionary fitness, the father wants to maximize the health of his offspring by enabling it to consume as much of the mother’s fat resources as possible, while the mother will take into account the toll that further development of an offspring will take on her system and her ability to reproduce later.
This divergence is the justification for epigenetic effects that arise. It is known as the kinship theory of genomic imprinting—conflict between maternal and paternal genes lead to the offspring evolving changes in gene expression. This biological arms race rooted in the difference between the mother’s and the father’s reproductive goals causes some genes in the offspring to change expression patterns depending on which parent contributed them. For example, some genes are completely on if inherited from the mother and completely off if inherited from the father. These are the genes previously known to be subject to parent-of-origin allelic affects. Not long after Haig and Westoby first proposed their theory, the first parental-specific expression of a gene was discovered, lending their idea much support.
Recent work in the Dulac lab has found that this genetic phenomenon is more prevalent than previously expected. Past work with mice had resulted in the discovery of only about 100 such genes. Using new technology, post-doctoral fellow Christopher Gregg and Professor Dulac were able to find more than 1300 new genes that are subject to parent-of-origin allelic expression, and further analysis of these genes has led to some surprising discoveries.
According to Gregg, there are many ways to find imprinting effects, and Dulac’s lab has been taking advantage of next-generation sequencing technology, which allows scientists to sequence the entire human genome in just one or two weeks. With this powerful technology, scientists can “ask a lot of different questions beyond just the sequence of a human genome.” By specifically marking genes and cross-breeding different strains of mice, this high-throughput technology gives researchers a “very sensitive and high resolution way of looking at the balance between maternal and paternal gene expression” that is much cheaper and faster than alternative methods such as genetic linkage studies.
To find these differences in gene expression, Dulac and Gregg performed reciprocal crosses on two strains of mice to obtain an F1 generation with two divisions. The team then used DNA tags unique to each strain to distinguish the regions of DNA that were being transcribed in the F1 generation. They analyzed tissue samples collected from mouse embryonic day 15 brains, and the preoptic area (POA) of the hypothalamus and medial prefrontal cortex (mPFC) of adult mice.
They chose these two brain areas because of previous pioneering work in epigenetics. Using techniques in artificial fertilization, researchers created embryos from two eggs, called gynogenotes, and from two sperm, called androgenotes. These embryos did not survive, but when they added these doomed embryos to normal embryos, the organisms did develop, though with strange physical appearance. The embryos with the androgenic cells added had disproportionately large bodies and small heads, while the gynogenote chimeras had small bodies and large heads. This difference occurs because the androgenote and gynogenote cells go in opposite directions in the brain during development. While the cells with two paternal complements stay in the hypothalamic regions, they disappear from most of the cortical and limbic regions. On the other hand, the cells with two female complements do the opposite. Since researchers had not investigated this phenomenon for a decade, Dulac and Gregg took a chance with next-generation sequencing to see if there was a maternal cortex or a paternal hypothalamus. Gregg notes that “it turned out to be different, but much more interesting.”
The researchers found that in the brains of 15-day-old mouse embryos, 61% of the genes identified were maternally expressed. In the adult brain, however, more than two-thirds of the identified genes were paternally expressed. Instead of finding a static, one-time phenomenon, Dulac and Gregg showed that parent-of-origin allelic expression is a dynamic process. Moreover, most of these maternally or paternally expressed genes are only expressed differently in specific types and areas of the brain. This suggests an extraordinary level of spatial and temporal regulation of differential parental gene expression, but the lab still had more to test.
Following the discovery of the large number of genes subject to parent-of-origin allelic effects, the Dulac group was then interested in how the gender of the offspring influenced the imprinting patterns inherited from their parents. The second paper identifies three processes that can lead to sexually dimorphic genomic imprinting: non-random X inactivation, changes in the expression of the active paternal X (Xp) caused by imprinting specific loci of the inactive maternal X (Xm) that differs from the expression of the active Xm, and autosomal genes being imprinted in only one sex.
To investigate the sex-based bias in X inactivation, the lab compared global levels of gene expression of Xm and Xp. While paternal inactivation was favored in both areas, Xm activation was a significantly more favored in the mPFC. To test for the second mechanism, they looked for the imprinting of individual genes, finding nine in the POA and three in the mPFC. For the final investigation, they tested to find tagged sites that were imprinted in only one sex and identified 347 candidate genes. They found that there were three times the number of sex-specific imprinted genes in the female POA than the male POA, but they found no significant difference between the total number of sex-based imprinted genes in the mPFCs of males and females. This data fits snugly with the previously widely held belief that the POA is a highly sexually dimorphic area thought to be involved in maternal and mating behaviors.
Dulac and Gregg have made many novel inroads with their work, from increasing the number of known parent-of-origin specific genes in mice tenfold to discovering new paths of sexually dimorphic brain development. In light of these discoveries, their work is raising even greater questions. As Dulac said, “The problem to some extent is that we are the victims of our own success, which is that we never expected to find that many genes to be imprinted.” Now that the Dulac lab has discovered this amazing array of new genes, the goal is to illuminate their functions.
For example, in selecting a set of interesting genes to start further lines of inquiry. Dulac has chosen one gene in particular involved in cell death, proceeding with the idea that if it is imprinted in some areas of the brain and not others, the rates of cell death between those areas will vary. It is an idea that, if supported, has great potential to advance ongoing dementia research. These new mechanisms for genomic imprinting are very versatile. But beyond what we currently know, as Gregg notes, “[t]here are some big ‘ifs’ in this field. If epigenetic effects are major regulators of behavior—and it’s not clear that they are yet—these epigenetic mechanisms really could have a big impact on how we understand the propagation of health problems in society.” Thus, research in genomic imprinting and other epigenetic mechanisms may have even broader roles to play in defining how we view society.