What is Epigenetics?
How does the cell know what parts of the information to access in order to produce what it needs?
Written by Alysia L. vandenBerg, PhD
Imagine that you are on a journey, visiting the nucleus of a cell of the human body. We can think of the nucleus as the “inner city”. In terms of activity, the inner city resembles an ant colony on caffeine—it’s incredibly busy. Access to the inner city is guarded; only certain materials are let through to the information and production headquarters housed there. About 40 miles of information, coded onto fine thread is crammed inside the inner city, but it’s only about the size of a tennis ball. The knowledge of how to produce about 23,000 products (proteins) is encompassed within this coded information (DNA), but the city only really produces around 10-20% of that for day-to-day business. How does the inner city know what parts of the information to access in order to produce what it needs? And how is this knowledge then remembered or maintained, from day to day, year to year, in order to avoid chaos? Researchers studying epigenetics are interested in exactly this type of question: how does a cell know which parts of the genetic information that’s encoded in a cell’s DNA to access in order to produce what it needs, and once these decisions are made how can they be perpetuated, or else changed, if necessary?
The roots of the term epigenetics go all the way back to Aristotle, who coined the term epigenesis. Epigenesis was used in opposition to the theory of preformation, which posited that all living organisms existed in miniature inside of the sperm (or the egg), and merely expanded in size over time [link]. In contrast, the theory of epigenesis proposed that embryos began as an undefined mass with new parts added during development (1). Conrad Waddington, described as a “renaissance biologist” (2), was a 20th century scientist-philosopher who built on Aristotle’s early notions about animal development. He is credited with using the term epigenetics in the context of development, describing it as, “the branch of biology which studies the causal interactions between genes and their products, which bring the phenotype into being” (2). This was a rather prescient observation, considering that the structure of DNA was not even known at that time.
1. J. M. Slack, Conrad Hal Waddington: the last Renaissance biologist? Nat Rev Genet3, 889 (Nov, 2002)
2. C. H. Waddington, Canalization of development and genetic assimilation of acquired characters. Nature 183, 1654 (Jun 13, 1959)
Waddington’s Classical Epigenetic LandscapeIn 1957, Conrad Waddington proposed the concept of an epigenetic landscape to represent the process of cellular decision-making during development. At various points in this dynamic visual metaphor, the cell (represented by a ball) can take specific permitted trajectories, leading to different outcomes or cell fates. Figure reprinted from Waddington, 1957.
The prefix epi- comes from Greek, meaning, “over” or “above”; therefore epigenetic information is recorded on top of the genetic information. Waddington was interested in how cells in an embryo made decisions that produced a huge number of different cell types that comprise an organism and is responsible for developing the idea of the epigenetic landscape, in which a cell is represented by a ball at the top of a hill (3). The cell makes different choices by taking only one of several possible paths downhill—this analogy provides a pleasing visual depiction of the developmental choices that a cell makes, and biologists continue to use it even today. Cells rolling down the epigenetic landscape have the same DNA throughout their journey, but they change their phenotype as they make developmental choices by deciding which genes to turn on or off (or express—as in gene expression). As with all epigenetic phenomena, the underlying DNA does not change. Cells within an organism almost all contain the same genomic information, but they have different phenotypes because they express different sets of genes. Epigenetics seeks to explain how cells “know” which genes to express.
Epigenetic processes do not change the DNA itself, but rather record developmental and environmental cues into a kind of code that signals if genes should be switched on, or expressed. Epigenetic information encoded within a cell can be transmitted to its descendants; this stable pattern of gene expression—not encoded by the DNA itself, but rather, on top of the DNA—is a cell’s epigenotype. Different epigenotypes give rise to different cell types comprising an organism; for example, a skin cell has the same DNA as a liver cell, but their epigenotypes differ.
The root of many diseases cannot be explained solely by DNA mutation and have instead been associated with epigenetic changes [link]. There are far-reaching implications of epigenetic research for human biology and disease; in recent years, researchers have discovered that epigenetic mechanisms have a hand in diseases such as cancer (4), diabetes and obesity (5), neurological disorders like schizophrenia, and function in processes like aging (6). However, the extent to which epigenetic changes are causal, rather than a consequence of genetic changes remains an open question. On the one hand, a DNA mutation or even a sequence polymorphism can impact gene expression and lead to epigenetic changes that lock in an abnormal state of gene activity. On the other hand, mutations in genes that effect epigenetic processes overall, can result in abnormal epigenetic states genome-wide.
3. A. D. Goldberg, C. D. Allis, E. Bernstein, Epigenetics: a landscape takes shape. Cell 128, 635 (Feb 23, 2007)4. M. Esteller, Epigenetics in cancer. N Engl J Med 358, 1148 (Mar 13, 2008)
5. B. E. Levin, Epigenetic influences on food intake and physical activity level: review of animal studies. Obesity(Silver Spring) 16 Suppl 3, S51 (Dec, 2008)
6. E. L. Greer et al., Transgenerational epigenetic inheritance of longevity in Caenorhabditis elegans. Nature479, 365 (Nov 17, 2011).
Scientists started uncovering a variety of mysterious inheritable phenomenon that didn’t alter the DNA sequence
Historically, the term “epigenetics” has shifted from the Waddingtonian definition of linking phenotype with genotype – to become a grab bag category to label odd examples of heritable changes in gene expression without any changes to the underlying DNA sequence (7). The laws of genetics are supposed to be able to explain how organisms inherit characteristics, dictated by the genes within an organism’s DNA. But from the 1930s to 80s, scientists started uncovering a variety of mysterious inheritable phenomenon that didn’t alter the DNA sequence. In maize, when scientists introduced a gene specifying for seeds of a deep red color into a plant containing a gene specifying for stippled red, something strange happened (8). Normally, the resulting progeny would be a mixture of genes (remember: each of us are a blending of genes from mom and dad). In other words, some offspring would have deep red seeds, while some would have stippled. But in this case, the deep red seeds never appeared again in the offspring, even though the gene was still physically there—this was a violation of the genetic laws of Mendelian inheritance and researchers were very puzzled. In another example, in 1938, Muller showed that in flies carrying chromosome rearrangements that changed the position of the white gene (which normally gives red eyes when expressed but white eyes when silent or mutated), the white gene started to behave in a rather unusual way (9). The eye color in these flies was a mosaic of red and white, with the gene being expressed in some cells and not in others. This was an example of another epigenetic phenomenon called position effect variegation. It turned out that the gene had been juxtaposed next to repressive chromatin, called heterochromatin, which would sometimes spread into the gene during development and switch it off. Once this happened, it would stay off, giving patches of white cells. This is another example of the same gene behaving in two opposite ways due to chromatin changes – similarly to the patches of coat color in calico cats, as explained later on. Recently, more examples of these phenomena have been gathered in yeast, flies, plants and vertebrates, suggesting that epigenetic inheritance is a pervasive phenomenon in nature, and it may play an important role in our lives (7, 10).
Perhaps one of the best known examples of epigenetics is mammalian X-chromosome inactivation (11). Males and females differ in their sex chromosome constitution with males having an X and a Y and females having two X chromosomes. In females, one of the two X chromosomes get shut down, or silenced, during early development to ensure that gene expression levels from the X chromosome are equal between the sexes. Once the genes on the inactive X chromosome have been silenced, this state is maintained through hundreds of cell divisions in the lifetime of the animal, but is fully reversed in the egg that will go on to produce new offspring. So this is epigenetic par excellence—two X chromosomes, with exactly the same DNA sequence, but totally different expression patterns in the same nucleus—and this differential expression can be stably maintained through cell division, but is fully reversible. Another fascinating example of epigenetics is genomic imprinting, discovered rather by accident in the 1980s when researchers were investigating if mice could be born with the genetic material of a single parent (12, 13). Artificially created embryos containing two maternal genomes or two paternal genomes, rather than one of each, were not viable. These embryos had the necessary DNA, but they didn’t grow and divide past a certain point, for some reason. This led to the idea that certain “imprints” (epigenetic marks or tags)must be laid down on the maternal and paternal genomes during oocyte (egg) and sperm generation, respectively, and that these imprints were required for normal development.
In striving to understand the nature of epigenetic information, many researchers over a period of at least two decades began to realize that in some cases, the organization and packaging of the DNA itself, into a structure known as chromatin, plays an important role in epigenetics. Chromatin was first described by scientists in the late 1800s as material that took up color (chroma) within the nucleus of a cell and was visible under a microscope [link]. Chromatin bunches up even more (or condenses) to form chromosomes, which are visible (and often depicted as cartoon-like Xs during cell division. Within every human cell around two meters of DNA is packed into the cell’s nucleus, which can be a mere 6 mm in length (that’s 0.000006 meters), meaning the DNA is compacted about 10,000 fold. Epigenetic researchers who study chromatin structure are intrigued by how cells can accomplish the feat of packing an enormous amount of DNA into a tiny nucleus.
7. E. Jablonka, M. J. Lamb, The changing concept of epigenetics. Annals of the New York Academy of Sciences981, 82 (Dec, 2002).
8. J. Brzeski, K. Brzeska, The maze of paramutation: a rough guide to the puzzling epigenetics of paramutation. Wiley Interdiscip Rev RNA2, 863 (Nov-Dec, 2011).
9. E. A. Carlson, The legacy of Hermann Joseph Muller: 1890-1967. Can J Genet Cytol9, 436 (Sep, 1967).
10. E. Jablonka, G. Raz, Transgenerational epigenetic inheritance: prevalence, mechanisms, and implications for the study of heredity and evolution. The Quarterly review of biology84, 131 (Jun, 2009).
11. M. F. Lyon, Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature190, 372 (Apr 22, 1961).
12. J. McGrath, D. Solter, Completion of mouse embryogenesis requires both the maternal and paternal genomes. Cell37, 179 (May, 1984).
13. M. A. Surani, S. C. Barton, M. L. Norris, Development of reconstituted mouse eggs suggests imprinting of the genome during gametogenesis. Nature308, 548 (Apr 5-11, 1984).
Histone proteins take up color, making chromatin visible
Photo from sciencephoto.com
Not only is it amazing how all that DNA fits in such a tiny space, it’s also fascinating to imagine how the DNA must be organized in some kind of logical way to ensure that regions that have to be accessed by various machineries can be, while those that must not, are protected. Chromatin can also be described as DNA plus the scaffolding proteins that help keep it organized [link]. Proteins called histones come together to form a kind of spool (also called a nucleosome) that the DNA winds around just under 2 times [link]. Multiple nucleosomes group together to form larger structures, which in turn make up chromosomes [link]. In fact, it’s the histone proteins that take up “chroma” color, thereby making the DNA visible. Nucleosomes can slide around, allowing other factors to access the DNA that would otherwise be unavailable—this is really important for the processes of DNA replication and gene transcription, for example. If nucleosomes were locked in place, a cell wouldn’t be able to make another copy of its DNA (or make RNA which is subsequently exported out of the nucleus to the cytoplasm to make proteins) from genes encoded within your DNA.
On top of the compaction and spatial organization provided by histones, there’s also an additional layer of information that is carried by histones and other chromatin-associated proteins [link]. Any given cell only expresses about 10-20% of all its genes, so knowing which ones to express is important. Specific modifications—we can think of them as tags—are applied to DNA and histone proteins. These tags serve as a sort of code that is in turn interpreted by other cellular players, telling them which genes to express. If your DNA were written as a book, then histone modifications and methylation marks could be imagined as a collection of post-it notes stuck throughout every chapter, providing another level of information without changing the way DNA itself was written. Methylation is one of the more well-studied modifications, since scientists have known about DNA methylation for many decades. Furthermore, DNA methylation is one of the few epigenetic marks for which we have a fairly good understanding of how it is propagated through cell division (14-16) [see also]. We also know which proteins bind to this modification, although we know less about how they influence gene expression. The tails of histone proteins (remember histones act like spools for the DNA to wind around) can be chemically modified too. Some of these modifications (like acetylation, for example) can affect the compaction or density of chromatin. [Visit this webpage for an animation of histone acetylation]. Also, these tails are thought to face outward, meaning their modified ends can be accessible so this “information” can be read by other proteins. Indeed, the list of proteins that interact with DNA, modified histone tails, histone variants and DNA methylation is growing exponentially; learning more about them will be vital for future advances in our understanding of human development and disease.
Another level of organization that is increasingly being thought of as epigenetic, is the spatial information or higher order folding of chromosomes—as well as their positions in the nucleus (17-21) . Thinking about epigenetic processes is increasingly becoming a four-dimensional adventure–the genome functions in three-dimensional space, and gene expression must be coordinated and maintained over time. The timescale can be during the lifetime of a cell, during development, between generations—even evolutionary time involves epigenetic processes!
14. L. Daxinger, E. Whitelaw, Understanding transgenerational epigenetic inheritance via the gametes in mammals. Nat Rev Genet13, 153 (Mar, 2012).
15. J. Paszkowski, U. Grossniklaus, Selected aspects of transgenerational epigenetic inheritance and resetting in plants. Current opinion in plant biology14, 195 (Apr, 2011).
16. U. C. Lange, R. Schneider, What an epigenome remembers. BioEssays : news and reviews in molecular, cellular and developmental biology32, 659 (Aug, 2010).
17. T. Sexton et al., Three-dimensional folding and functional organization principles of the Drosophila genome. Cell148, 458 (Feb 3, 2012).
18. B. D. Towbin et al., Step-wise methylation of histone H3K9 positions heterochromatin at the nuclear periphery. Cell150, 934 (Aug 31, 2012).
19. A. Delest, T. Sexton, G. Cavalli, Polycomb: a paradigm for genome organization from one to three dimensions. Current opinion in cell biology24, 405 (Jun, 2012).
20. E. Splinter et al., The inactive X chromosome adopts a unique three-dimensional conformation that is dependent on Xist RNA. Genes & development25, 1371 (Jul 1, 2011).
21. E. P. Nora et al., Spatial partitioning of the regulatory landscape of the X-inactivation centre. Nature485, 381 (May 17, 2012).
Calico and tortoiseshell cats are always females and their patches of different colored fur are due to X inactivation, or the silencing of one X chromosome
Let’s explore some examples of epigenetic processes in action. Did you know that calico and tortoiseshell cats are always females and that their patches of different colored fur are due to X inactivation, or the silencing of one X chromosome? Simple genetics cannot explain this intriguing pattern of coat color gene expression [link]. It is the fact that either of the 2 X chromosomes becomes silenced in early embryos – some cells choose one X and some cells choose the other, and if the two Xs carry different versions (or alleles) of a coat color gene, then the cell will express either one or the other version, but never both. The inactive state is propagated through cell division, leading to either orange or black patches of fur. Which X chromosome gets silenced is determined randomly, so this is the reason why no two calico cats look alike, even if they have exactly the same DNA. This is a perfect example of epigenetics at work – once the decision of which version (orange or black) is expressed, this state is maintained. This means that women are actually mosaics, with patches of cells expressing different X chromosomes. This cellular mosaicism due to X inactivation in females can also have important implications for disease. Males only have one X chromosome and so are fully affected by mutations in X-linked genes, such as muscular dystrophy, or hemophilia. Females, on the other hand, tend to be less affected, thanks to their mosaicism and the cells expressing the unmutated X chromosome – but as the mosaicism can vary substantially between individuals, so too can the disease. The precise manner in which the X chromosome gets shut down during development is still somewhat mysterious but uses various epigenetic tricks, including chromatin marks such as DNA methylation, histone modifications and their binding proteins, as well as non-coding RNAs.
Another example of epigenetics at work that is often cited concerns human twins. Identical twins have the same DNA and indeed their remarkable similarity is a powerful argument for the importance of genes in defining phenotype. However, we know that twins are not exact carbon copies of each other (22) [see also]. Studies have borne out that epigenetic marks, such as DNA methylation and histone modifications are responsible for some of these differences. Cloned animals exhibit differences from their donors too, and this arises from differences in epigenetic marks (23, 24) .
We know from such human disorders as Angelmans, Prader-Willi, and Beckwith-Wiedemann Syndrome that maternal or paternal methylation of certain genes is required for normal development (25). This methylation is called imprinting, and it leads to a rather special pattern of gene silencing. Imprinting doesn’t simply turn genes off; it turns off particular genes coming from the mother and others from the father. Why might this be useful in biology? Some paternally silenced genes favor larger offspring, while other genes that are switched off maternally instead produce smaller offspring, thereby conserving the mother’s resources, with the idea that she then has more to offer other future offspring (26). Although this is just one hypothesis among many, imprinting following this pattern would set up an epigenetic battle of the sexes even before birth!
The growing list of human diseases linked to epigenetic abnormalities is another indication of the importance of epigenetic processes (24). Faulty epigenetic marks are linked to certain forms of diabetes, lupus, asthma, various neurological disorders as well as cancer. Several FDA-approved drugs that alter epigenetic modifications of histones, or DNA methylation are currently used in the treatment of cancer (27) with many more in development (28) so it is becoming even more important for patients and doctors to have an understanding of epigenetic processes at work in cells and the nature and extent of the changes that occurs when such “epidrugs” are used.
Recent studies have also shown that longevity can be transmitted epigenetically—in worms, at least (6, 29, 30) . Researchers found that C. elegans (click here for more information on these worms) with a genetic mutation causing a defect in one particular type of histone modification lived 25-30% longer than normal. They took mutant long-lived worms and mated them with normal worms, then selected resulting “normal” offspring without the mutation to observe the effect on lifespan. If longevity were transmitted genetically, the offspring would therefore be normal. Instead they found that three generations later, the “normal” worms (without the mutation) were long-lived, showing that longevity was passed on epigenetically to offspring.
22. J. T. Bell, T. D. Spector, A twin approach to unraveling epigenetics. Trends in genetics : TIG27, 116 (Mar, 2011).
23. G. S. Archer et al., Hierarchical phenotypic and epigenetic variation in cloned swine. Biol Reprod69, 430 (Aug, 2003).
24. A. Portela, M. Esteller, Epigenetic modifications and human disease. Nat Biotechnol28, 1057 (Oct, 2010).
25. R. D. Nicholls, J. L. Knepper, Genome organization, function, and imprinting in Prader-Willi and Angelman syndromes. Annu Rev Genomics Hum Genet2, 153 (2001).
26. W. Reik, J. Walter, Genomic imprinting: parental influence on the genome. Nat Rev Genet2, 21 (Jan, 2001).
27. T. K. Kelly, D. D. De Carvalho, P. A. Jones, Epigenetic modifications as therapeutic targets. Nat Biotechnol28, 1069 (Oct 13, 2010).
28. R. A. Copeland, E. J. Olhava, M. P. Scott, Targeting epigenetic enzymes for drug discovery. Curr Opin Chem Biol14, 505 (Aug, 2010).
29. C. Luiggi, in The Scientist. (2011).
30. S. Fecht, Longevity Shown for First Time to be Inherited via a non-DNA Mechanism. Sci Am, (19 October 2011, 2011).
There are very few proven examples of heritable epigenetic changes exist in mammals to date
Emerging studies suggest that epigenetic changes might be inherited by the next generation. For the time being, epidemiological evidence in humans is only just emerging, with empirical evidence continuing to come from studies in model systems like mice. Popular media articles abound with stories of epigenetic inheritance in humans, and the possibilities for epigenetic inheritance are certainly intriguing, such as this story covered by Time Magazine. Researcher Lars Olov Bygren made a remarkable discovery about inhabitants of the remote Swedish village of Norrbotten, which historically had been subject to periods of feast and famine (31). His group studied the effect of diet on trans-generational longevity during both the “fat” and “lean” times. Remarkably, those who changed from a normal diet to a gluttonous one due to an overabundance of food had offspring who died 32 years earlier than those who experienced famine. This suggests that the environmental effects of overeating during “feast” periods were somehow passed on to the next generation [Time Magazine. Jan. 18, 2010. Why Your DNA Isn’t Your Destiny]*. However, further studies are essential to see if certain genes have indeed been marked via bona fide epigenetic mechanisms that could be responsible for these effects on longevity.
It should be noted that very few proven examples of heritable epigenetic changes exist in mammals to date (32). In mammals, epigenetics changes are by and large erased in the germ line (eggs & sperm) at each generation. Plants, on the other hand, do show trans-generational inheritance of epigenetic marks including methylation. However, there are a few examples of metastable epialleles that have been identified in mice. Metastable is used “to emphasize the probabilistic nature of expression” and epiallele, “to emphasize that the allelic forms differ with respect to epigenetic state, rather than the DNA sequence”. Notably, each of the identified metastable epialleles (such as the agouti gene involved in coat color inheritance, click here for more information)are associated with repetitive DNA elements. It will be exciting to see how future studies in model systems such as mice and Drosophila contribute to our understanding of trans-generational inheritance in humans.
We’ll finish with some thoughts from James Watson that help us to put epigenetics into a historical perspective. The sequencing of the human genome recently celebrated its 10th anniversary. “We fooled ourselves into thinking the genome was going to be a transparent blueprint, but it’s not,” says Mel Greaves, a cell biologist at the Institute of Cancer Research in Sutton, UK (33). This sentiment has been mirrored by others who thought that unraveling the genome sequence would lead to disease cures in record time. James Watson said, “The major problem, I think, is chromatin. What determines whether a given piece of DNA along the chromosome is functioning…? You can inherit something beyond the DNA sequence. That’s where the real excitement of genetics is now” (34). One thing is sure, it’s an exciting time for epigenetics research and we’re likely to learn much more in the near future about the role of chromatin and epigenetics in human health and disease.
* Time Magazine. Jan. 18, 2010. Why Your DNA Isn’t Your Destiny
* You can also check out the BBC podcasts on “The First 1000 Days: A Legacy for Life – Part I, II, III for other media coverage of epigenetic inheritance.
31. G. Kaati, L. O. Bygren, M. Pembrey, M. Sjostrom, Transgenerational response to nutrition, early life circumstances and longevity. Eur J Hum Genet15, 784 (Jul, 2007).
32. R. Feil, M. F. Fraga, Epigenetics and the environment: emerging patterns and implications. Nat Rev Genet13, 97 (Feb, 2011).
33. E. Check Hayden, Human genome at ten: Life is complicated. Nature464, 664 (Apr 1, 2010).
34. J. D. Watson, Celebrating the genetic jubilee: a conversation with James D. Watson. Interviewed by John Rennie. Sci Am288, 66 (Apr, 2003).