Repression Makes Us What We Are

Notes on polycomb and histone interaction in developmental gene control.

While we have shockingly few genes in the genomes which engender our complexity, most of them are also turned off most of the time. It is repression, just as much as activation, that generates the patterns of gene expression that allow developmental ramification and specialization. This repression comes in several forms, from individual gene switches to the sequestering of whole chromosomes, such as in female X-inactivation. The stronger forms of repression can be what is called "epigenetic", which simply means they can last a long time, such as through several cell division cycles, or even across a generation. That means that they begin to mimick true genetic effects. However unlike true mutations, they can be programmatically reversed at some future time- otherwise they would not have any function. Thus "epigenetic" features are part of an organism's phenotype, not its genotype.

One general mechanism of long-term and long-range (that is, may extend over many neighboring genes, and large chromosomal regions) gene repression is the polycomb system, named after phenotypes conferred by some of its mutations in fruit flies, whose males have sex combs. This is a system that may descend on a chromosomal region if it is never needed again in the development of an organism, and keeps those genes off, locked up tight. This effect is also called "gene silencing", for obvious reasons.

There are several proteins that make up polycomb complexes, each with its own super-powers. Most of these powers revolve around working with histones- the small proteins that are the smallest packaging units of DNA in eukaryotic cells. Histones are not passive spools around which DNA wraps, however, but through a vast number of possible chemical modifications (methylation, acetylation, ubiquitination, at numerous different positions) are pivotal levers of control over on the availability of DNA to all the other denizens of the nucleus.

One typical core complex of histones, acting as a spool for our DNA. The DNA is actually much larger, covering most of the histone surface. Yet the histone tails stick out, ready to be modified and recognized by various other regulatory proteins.

One of the special characteristics of polycomb action is that it can spread along the DNA from a initiating site to nearby locations, in a progressive fashion. This is unlike a normal gene repressor, which just acts locally within a mixture of activators and repressors for one enhancer site upstream of one gene. How does polycomb do this, and how does it know where to stop?

There are two parts to the system, polycomb complex 1 and complex 2 (PRC1, PRC2). PRC2 acts first, binding to other proteins at particular DNA sites called polycomb response elements, and methylates the local histones H3 on lysines (K) 9 and 27, which are on the tails sticking away from the DNA and thus accessible. How does this get turned on? The targetting proteins are the ones that are themselves regulated to initiate the whole repression process. A second activating change is methylation of the local DNA, at CG dinucleotides, which tends to concentrate at silenced genomic locations, and helps initiate that silencing.

Methylation has a specific charge effect, eliminating the positive charge on the affected lysines, (or negative charge at the CpG DNA sites), thus helping the histones repel each other less, and pack together. But the marking of histones by various methyl, acetyl, and other groups on their tails has more subtle effects, since each modification is specifically recognized by other proteins, creating a complex code that regulates gene activity at very fine scales.

The second step, carried out by PRC1, which finds and binds to histone H3 methylated on lysine 27, is ubiquitination of histone H2A on lysine 119. As ubiquitin is a small protein, its attachement is a much more dramatic change than modifications like methylation or ethylation. And though ubiquitin is generally associated with marking proteins for destruction, here is doesn't have that effect, but rather has a regulatory role in stablizing and compacting chromatin structure. However, as a large complex with several activities, PRC1 may do other things to promote repression which are not yet known.

A recent paper delved into this a bit to ask how PRC1 is composed, and what activates it. It is not a stable or uniform complex, but a consortium of several proteins that converge when needed and whose components come in several flavors. It is apparent that repression even in this general polycomb class comes in different forms. Representative components are:

• RING1 or RING1B- this is the ligase that enzymatically attaches ubiquitin to Histone H2A
• PCGF1- This is a helper for the RING ubiquitin ligase activity.
• CBX2 or RYBP- a protein that binds to the H3 methylation site, and binds other proteins, especially PCGF1, and YY1. YY1 is one of the targeting transcription factors that can bind the polycomb response elements and help initiate repression.
• KDM2B- an enzyme that can de-methylate histone lysines, and binds to the CpG dinucleotides that, in part, target polycomb repression. It also has a protein domain with a role in targeting ubiquitination of other proteins (the F-box).
• BCOR- This protein interacts with other histone de-acetylases.

The point of this particular paper was to demonstrate the composition of one particular version of the PRC1 complex, and to show that the core subunits of RING1B and PCGF1 are sufficient for histone ubuquitination, but that they are stimulated by the addition of the subunit RYBP. The other subunits don't help ubiquitination in vitro, but have other roles (whether known or unknown) in regulating and directing the complex's activity in cells.

One example of a PRC1 complex, which ubiquitinates H2A histones. B shows an electrophoretic gel that separates the proteins by size. kD is kilo-Daltons. C shows a mass-spectrographic study of cross-linked complexes showing which parts of which components interact directly with which other ones.

Another finding is that the PRC2 complex recognizes not only the initiating factors at the polycomb response elements, but also the ubiquitinated histones left by PRC1. This is likely to be part of the positive feedback "spreading" mechanism by which polycomb extends its area of repression from those initiating sites on the DNA / chromatin. Unfortunately, the details of initiation, the exact mechanism of spreading, the implications of ubiquitination, and the reasons for limits on the dimensions of polycomb-repressed regions are still largely unknown, or only hinted at, so far.

That gives you a taste of the state of the field, from this recent paper. The polycomb system has been known for a long time, having been established genetically in fruit flies over 70 years ago, with the discovery of the original polycomb mutation. It is unfortunate that this field is not farther along, in the understanding of the individual components, and how this form of repression is initiated and limited.

• Yes, he has his own brown shirts.
• Meet your new friends.
• Cultures of stupidity.
• To hell with the whole thing...
• We can make all the jobs we want.
• Apparently, a feminized culture is a bad thing.
• Scientific epistemology, the value of negative results, and the canonization of facts.
• Institutional development in China, vs freedom.
• Can we call it treason, already?
• Religion is part of the reason for our new love of Putin.
• Mergers are good for someone, but not you.

"By utilizing new techniques to isolate the effects of mergers, they find no evidence that mergers increase efficiency, but do find evidence that they increase market power, meaning they allow companies to generate higher profits by raising prices."