The Mammalian Epigenome

As according to UCSC's web site for bioinformatics course, they discriminate bioinformatics and computational biology. Of course, discrimination of these things would be meaningless because boundary between them is vague and we cannot weigh up the computational skill and the biological knowledge. Maybe this is only my own problem, my effort as a bioinformatician is leaning toward to raise computational skill. Even in the recent situation where contents is emphasizing, lack of biological insight can be serious problem for being leader of the field which I belong. Thus I planned to make group with lab members, who are related with my ex-firm, for studying epigenomics. I ask them to be active on this meeting, and I hope that this meeting will be more than studying groups. This is the first posting for epigenomics.

Bradley E. Bernstein, Alexander Meissner, and Eric S.Lander

-Epigenetic modifications in mammalian genomes

Epigenetic modifications fall into two categories

DNA methylation : In vertebrate, DNA methylation occurs almost exclusively in CpG and most CpG are methylated.

Histone modification : The core histones are subject to more than 100 posttranslational modification (acetylation, methylation, phosphorylation and ubiquitination). This occurs primarily in N-terminal. Lysine acetylation almost correlate to chromatin accessiblity and transcriptional activity, but lysine methylation have various effects (H3K4, H3K36 -> transcribed chromatin, H3K9, H3K27, H4K20 -> repression).

--Epigenetic Inheritance

Epigenetic modifications are heritable. By methyl-transferase Dnmt1 which can methylate hemi-methylated CpG, newly synthesized DNA strand are methylated and the pattern are preserved.

Histone modifications are a little bit hard to explain and need more reference. so I will defer this to next posting.

-A Dynamic Landscape of Cytosine Methylation DNA Methylation

--DNA Methylation in Development and Disease

DNA methylation work as

1.accelerator of recruitment of regulatory protein : methyl-CpG binding proteins recognize methylated cytosine and mediate transcriptional repression through interactions with histone deacetylases.

2.repressor for protein binding to DNA : CTCF binding at the H19 locus.

In development

In cancer, genome-wide loss or aberrant local gain of methylation. Tumor suppressor gene promoter is the target of the latter cases.

--CpG Islands

CpG island cover 0.7 % of genome and contain 7 % of total CpG. In the germ line, CpG in island are unmethylated, so can be recognized and corrected by repair system(deamination of cytosine(uracil), deamination of methylated cytosine(thymine), uracil can be corrected). 60% of promoter are related to CpG islands. The subset of these are tissue specific methylated during development. Some CpG reside within repetitive elements, and these are heavily methylated in somatic tissues. Oct-4 and Nanog promoter's methylation are associated with transcriptional activation while IL-2 and Sry's demethylation are associated.

--Studying the DNA Methylome

bisulfite sequencing : treat DNA with NaHSO3 (this convert unmethylated cytosine to uracil) and then sequence the DNA(uracil read as thymine, methylated cytosine read as cytosine).With this method survey on 6,20,22 chromosome found that small fraction(9.2%) of 511 CpG island are methylated while almost 50 % of non-CpG islands containing 5' UTRs were hypermethylated, and about 70 % conserved methylation profile between mouse and human.

Using fractionation of methylated and unmethylated portions by restriction or antibody : some found that the average methylation level CpG(selected from 5' regions of 371 genes) wa about 35 %. some group detect novel methyaltion at 135 promoters(of which 127 contain CpG islands) when comparing colon cell line with normal colon.

-Global Insights into Histone Biology

--Methods for Large-Scale Analysis of Histone Modifications

--Chromatin Domains and Cellular Memory

--Epigenetic Mechanisms of Genome Plasticity

-The Relationship between Genome and Epigenome

--The Newly Reprogrammed Epigenetic State

ES cells from inner cell mass where remethylation begins in early development, are representative of the reprogrammed state. The role of DNA in defining ES cell epigenome is significant.

H3K4 methylation coincide with CpG islands : trxG complexes that catalyze H3K4 methylation contain domains that bind unmethylated CpG dinucleotide.

H3K27 methylation correlates with transposon-free regions : 1.transposition is incompatible with the chromatin structure or negative slection. OR 2. transposons could interfere with the spreading of H3K27 methylation.

--Repetitive Elements

Transposon-derived DNA sequence is typically highly methylated in somatic tissues. Within given repeat class, modification status appeared to depend on the cellular differentiation state.

--Predicting Cytosine Methylation from DNA Sequences

-Emerging Technologies in Epigenome Research

--Cytosine Methylation

1.MSREs (methylation-sensitive restriction enzymes) , but has limitation on outside the restriction site.

2.Immunoprecipitation with a methylcytosine antibody, also has limitation on resolution

3.bisulfite-treat and sequence analysis, has little bias

--Histone Modifications

Chromatin IP (XChIP) : cross-link DNA and nearby protein by formaldehyde, sonicating the chromatin and then immunoprecipitating with antibody against a particular histone modification. but this has limitations, which are efficacy of antibody, fragmentation bias, and "masked" by other proteins specific to histone.

Native ChIP (N-ChIP) : shearing chromatin by micrococcal nuclease digestion which cut DNA at the length of the linker instead of sonication and no cross-link.

Biotin-tag affinity purification,

DamID (DNA adenine methltransferase identification) : using fusion protein consisting of E.coli Dam(Dam methylates adenines in the sequence GATC) and a certain chromatin protein or TF.

-Future Perspectives and Challenges

There are important issues

1.reduction of the numer of cells required for histone modification studies.

2.better antibody reagents for increasing sensitivity.

3.development of highly parallel DNA analyses.

4.computational tools in need.

I just want to summarize and rearrange the article in my way, but the most of sentences are compressed because of characteristic of review paper, so I couldn't do that.

further study

epigenome project, genomic imprinting & DNA methylation, inheritance of epigenome.