Spotlight: Epigenetics - heredity not only in genes / Dorit Ferns

Professor Haim Sider, one of the fathers of the field of epigenetics, changed the way we understand how genes work

"Once upon a time when doctors didn't know the origin of a disease, they said it was a virus, then they moved on to blame the immune system and today they blame it on changes in DNA methylation," laughs Professor Haim Sider, from the Department of Developmental Biology and Cancer Research at the Hebrew University of Jerusalem. who discovered about 30 years ago that the DNA sequence is not the only factor influencing gene expression. What is that methylation and how was it discovered?

We met with Sider to hear from him about the field of research that he pioneered and which earned him many prestigious awards, including the Israel Prize, the AMT Prize, the Wolf Prize, and this year - the Rothschild Prize.

The basic question - to express or not to express?

All cells in the body contain the same genetic information. And yet, there are hundreds of different types of cells in our body, all created from a single cell - the fertilized egg. The differences between the cells are created during the development of the embryo, in a process known as differentiation. From a molecular point of view, the differences that differentiation creates lie in the "expression" of the genes: a cell in a certain tissue uses only some of the genes found in the genome and silences the rest. How is it determined during differentiation which genes to activate and which to silence? This is the basic and weighty question that a researcher has sorted out.

Usually, an "active" gene is a gene whose encoded protein is actually produced in the cell. The gene is transcribed, meaning that RNA molecules are formed, they carry the information to the ribosomes and these, in turn, translate the information and build the protein (see "Polar Bears and Drunk Ribosomes", Scientific American Israel, December 2008). When the gene control mechanisms in the cell silence the gene, the protein is not made.

When students learn about the subject, they start with the classic studies of François Jacob and Jacques Monaud that won them the Nobel Prize in 1965. They discovered for the first time control proteins that bind to certain DNA sequences in bacteria and are able to silence genes or activate them. "To this day," smiles Cedar, "many think that this is the main mechanism responsible for activating genes and silencing them here as well." But Chaim Sidor and his long-time research partner, Professor Aharon Razin, also from the Department of Developmental Biology and Cancer Research at the Hebrew University, were clear that this could not possibly be the main mechanism responsible for cell differentiation in humans.

There are two main reasons for this, Cedar explains. First, humans have many more genes than bacteria. Most bacteria only have about 5,000 genes, most of which are active all the time. The bacteria silence only about 100 genes and it is possible to assign 100 control proteins for this purpose. But with us, it is necessary to silence more than 12,000 genes in each cell, and it does not make sense for each of them to have its own unique control protein. Second, the level of gene silencing in bacteria is weak compared to mammals. For example, when the gene studied by Jacob and Mono is silenced, its expression level drops 1,000-fold. In mammalian cells, on the other hand, silencing in liver cells of the gene for the globin protein, for example, lowers its activity 10 million times compared to blood cells where the gene is active.

A different kind of control

Cedar and Resin therefore set out to search for the mechanism responsible for silencing such a large amount of genes in mammals. When they started, in the 70s, it was already known that DNA undergoes methylation: a process in which a methyl group (-CH3) is attached to cytosine, C, one of the four DNA bases marked with the letters A, G, C and T. So the scientists believed that the methylation protects the DNA from self-degradation, as studies in bacteria have shown.

But Cedar began to understand that the role of methylation in mammals is different: it prevents the expression of genes. Before him, they had already noticed a correlation between the level of methylation of genes and their level of activity, but it was not clear if there was a causal relationship between the things. Cedar discovered that methylation in mammals, unlike bacteria, is done exclusively on C bases that are next to G bases, that is, only in the CG pair. The frequency of the pair in the human genome is not random, and in fact, in most of the genome it appears much less often than usual. However, certain regions, called CG islands, are richer than expected in this pair, and they are often used as control regions for gene expression.

Another characteristic that distinguishes methylation in mammals is its symmetry: according to the rules of base pairing in DNA, opposite each CG pair in the complementary strand is a GC pair, and there the C will undergo methylation as well. Cedar showed that there is a special enzymatic mechanism that maintains this methylation pattern even when the cell divides. The mechanism detects methylated CG sequences only in the old strand, which served as a template for DNA replication during cell division, but not in the new strand, and corrects this. Cedar discovered that a methylated sequence would remain marked with this chemical label even after 50 or even 100 cell divisions.

In the end, Cedar and his partners directly proved that methylation inhibits gene activity, by causing the DNA in these regions to be compressed in a way that prevents the transcription enzymes from accessing the genes - and a gene that does not undergo transcription is a silenced gene. (See details in the article: "The hidden switches of the soul", page 54 of this issue.)

Cedar and his colleagues therefore discovered a new mechanism for controlling gene expression, which is not "written" directly in the DNA sequence, but rather results from chemical changes of the DNA bases. Most of the time, this control received the name epigenetics (Greek: above genetics), and it also includes other molecular mechanisms and not just methylation.

Commemoration of decisions

After characterizing the phenomenon, Seder turned to investigate methylation during embryonic development - the stage where gene control directs tissue differentiation. Cedar discovered that after the egg is fertilized, all the methylations are erased, to allow reprogramming of the cell. At the time of rooting, about two weeks after fertilization, all the CG sites in the genome are methylated, with the exception of the CG islands. Many of these CG islands are the activity control areas of genes necessary for the general maintenance of all cells, regardless of differentiation, and are therefore called "housekeeping genes". The methylation pattern changes during embryonic development to enable the silencing of genes whose function has ended and to enable the activation of genes necessary for continued differentiation. In adult human cells, for example, a whole series of genes is silenced by methylation, which allows the cells to maintain the unlimited ability to differentiate, which characterizes embryonic stem cells. Reactivation of one of these key genes, Oct4, underlies the generation of induced stem cells, iPS (see “Turning back the cell clock”, Scientific American Israel, February-March 2009).

The findings shed light on the methylation strategy. It is not a matter of normal inheritance, which is passed down from the parents in the DNA sequence, but of an expression plan of the genome. And more importantly, methylation makes it possible to sign decisions made during development without the need to maintain these decisions through the constant activity of control proteins as in bacteria. For example, once a stem cell has "decided" that it loses its unlimited ability to differentiate, there is no need to repeat the decision at every division. And vice versa: to keep a gene in an active state, you simply have to leave the DNA in its area loose and accessible to transcription proteins. In fact, the transcription proteins encounter entire compressed areas in the DNA that are blocked to them and only here and there do they reach an open area where they can act. In this way, from the moment of differentiation there is no need for specific control proteins, neither for activation nor for silencing.

Methylation diseases

Like any biological mechanism, methylation is also prone to defects that cause hereditary diseases that do not arise from DNA mutations. One of the well-known examples is the fragile X syndrome characterized by unique facial features and mental retardation. The syndrome results from the fact that a certain sequence, CGG, found in the FMR1 gene on the X chromosome, repeats itself more times than usual. The repetition causes unnecessary silencing of the gene even though its protein is normal.

Recently, there has also been increasing interest in methylation errors in cancer cells. Cedar et al showed that cancer cells have targeted methylation of CG islands that control transcription of cancer suppressor genes. The methylation silences these genes and allows the cancer cell to multiply. In fact, most of the properties of cancer cells are related to methylation, such as their ability to divide rapidly and the ability to invade other tissues or secrete toxins.

The connection between cancer cells and methylation is not new, says Cedar, and describes an amazing experiment, according to him, that was conducted already 16 years ago. The experiment was conducted on mice with a mutation in the APC gene (the murine version of which is known as MIN), which appears in most colon cancer cases in humans. The mutant mice develop dozens of tumors (mostly benign) already at the age of 4-5 months, but if they are treated with a methylation inhibitor, they have no tumors at all. The treatment is not effective after there is already a tumor, but the experiment indicates that it is worth examining treatments that affect methylation at least in people with congenital mutations that significantly increase the risk of developing cancer at a young age.

The genome and the methylome

Although methylation is epigenetic, all methylation decisions still depend, of course, on the DNA sequence. And now Sidder believes that the time has come to understand how the DNA sequence dictates the epigenetic model, and he is engaged in developing computer models for this purpose. In the first phase, his laboratory focuses on understanding the methylation model of CG islands, and they have already developed an algorithm that succeeds in identifying sequences protected from methylation.

The final goal is to predict differences in the methylation pattern between cells in different tissues, between an embryonic cell and an adult cell or between a cancerous cell and a normal cell. Methylation differences are also found between different humans. Even in identical twins, on average, 5,000 differences in the methylation pattern were discovered (out of about 30 million methylations in the human genome). There may also be other influences on the methylation pattern, such as diseases, age, and perhaps even environmental influences.

When Sider started his research, not much importance was attached to DNA methylation in mammals. Today, following his groundbreaking research, thousands of articles on the subject are published every year. This is usually the stage in various biological fields where it is necessary to start "putting order in the mess", and this is exactly what Sider is trying to do in his Metilum project.

Post to Twitter