Nucleosome Binding Sites

A paper from a week or so, describes a method for detecting regions in the genome where DNA wrapping takes place. So what (you may asked) is DNA wrapped around? i-8d791d4362a4183f7a10fdf400297e06-nucleosome.gif

Nucleosomes!

Each nucleosome (red balls in the cartoon) contains 8 histone proteins (two each of Histone 2A, 2B, 3 and 4). In addition, histone 1 (yellow in the diagram) hangs out at the periphery and is involved in linking neighboring nucleosomes together (we think). Histones are not only the most conserved eukaryotic proteins but also the most abundant. In fact, if you were to run a protein gel and stain for total proteins (by coomassie stain, or ponso stain), two prominent bands will appear at the bottom of your gel at about 20kD - these are histones. Because histones are in close contact with DNA's highly negative phosphoates, histones are full of positively charged amino acids. DNA is wrapped around each nucleosome, like a string wrapped around a bead. DNA makes ~ 2 turns around each nucleosome (about 147 base pairs per nucleosome).

When a gene is "turned on", DNA is transcribed into RNA, but for this to occur, the histones must be removed and/or just simply pushed out of the way (either up or down the DNA molecule). By regulating how tight nucleosome binding is (or exactly where nucleosomes bind) the cell can control how active a gene is. Many people are trying to figure out how this nucleosome rearrangement takes place - the field is usually referred to as the chromosome remodelling field. The most studied nucleosome remodelling complex is the swi/snf complex. It was originally identified in yeast -- but more and more people are looking in mammalian cells. Many have now found actin like proteins associated with chromosome remodelling factors, so it is possible that nucleosomes can be pushed around by actin like filaments (actin is a cytoskeletal polymer not traditionally thought to function in the nucleus).

A key part to all this is that the core histones (H2A, H2B, H3, H4) themselves are actually modified and this changes how tightly the DNA is bound to them. Hisones get acetylated, methylated, phosphorylated, ubiquitinated, sumolated ... (sumo, ubiquitin = small polypeptide molecules that are covalently linked to their target proteins). Histone acetylation is associated with turnning genes on (by loosening the DNA/nucleosome interactions) and methylation is associated with turning genes off (by tightenning DNA/nucleosome interaction). Figuring out how histones are modified is a hot topics (as an example, a recent issue of Nature had two papers on how trimethylated histone 3 is demethylated). Amazingly when you copy DNA into two, the histones that reassemble onto the new strands retain the same modifications as the histones that were on the original strands. So if a piece of DNA contained five nucleosomes the first two being unmodified the next two containing one type of modification and the last nucleosome containing a second type of modification, the histones bound to the two daughter strands would carry five nucleosomes each nucleosome containing the same modification found in the parental DNA. If you think about it, this phenomenon is weird, you can inherit histone modifications. This inherited nucleosome state contributes to what is known as epigenetics - the inheritance of information that goes beyond the nucleotide sequence of the genome (in this case the histone modifications present at various positions within a genome).

i-d52418e180ca71af5712578694ecf2b4-nucleosome2.jpg
So finally where in the genome do histones bind? Well for quite some time, researchers have noted that histones tend to bind at certain spots along genes. The binding pattern is critical, it may render certain portions of the genome inaccessible to factors that bind DNA to regulate the state of gene expression. Now two groups have figured out an algorithm that predicts where nucleosomes may bind. The summary of their analysis of 199 stretches of nucleosome-bind DNA from the yeast genome is seen in the figure to the left.

[Incidentaly the results (e.i. the types of nucleotides that aid in nucleosome binding) remind me of Andrew Fire's characterization of DNA sequences present in along the C. Elegans genome and how this related to nucleosome binding and the transfectability of exogenous DNA into C. Elegans. I'm not sure that he ever published any of this though - but if he did, you can let me know.]

Ref:
Segal E, Fondufe-Mittendorf Y, Chen L, Thastrom A, Field Y, Moore IK, Wang JP, Widom J.
Genomic Code for Nucleosome Positioning.
Nature. 2006 Jul 19; [Epub ahead of print]

More like this

I got pretty excited about histones when I first ran across them, but there was precious little for the layman. I'm glad to see the number of internet-accessible resources written with a mere modicum of jargon is increasing.

Epigenetics seems like it holds a lot of promise. There are simple questions it holds the key to, like 'why does a liver cell divide into more liver cells?', but also questions like why does most animal cloning fail, and what sort of histone acetylation do humans have at conception (is histone acetylation completely reset?), which would possibly help with reproductive science, tissue cloning from adult stem cells, and even what other things make us human.

Good post.

Hi!

I'd like to use the nice nucleosome diagram in a grant application, would that be possible?

Cheers