Research Highlights:

5-hydroxymethylcytosine, a modified mammalian DNA base with a potential regulatory role

Evaluation of: Kriaucionis S, Heintz N: The nuclear DNA base 5-hydroxymethylcytosine is present in Purkinje neurons and the brain. Science 324(5929), 929–930 (2009).

Tahiliani M, Koh KP, Shen Y et al.: Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science 324(5929), 930–935 (2009).

Since the discovery of 5-methylcytosine in DNA approximately six decades ago, researchers have assumed that only five DNA bases constitute mammalian DNA. Of course, there are many other minor DNA bases that can be detected using ultra-sensitive methodology, but they all have been thought to be the products of DNA damage produced by endogenous or exogenous sources of reactive chemicals or radiation. 5-hydroxymethylcytosine is no exception. This base was first identified as a natural DNA base in bacteriophages in the 1950s [1]. Its later measurement in rodent and frog brain, and liver DNA has gone largely unnoticed and unconfirmed [2]. The similarity of 5-hydroxymethylcytosine and 5-methylcytosine has prompted studies that have shown, for example, that the hydroxylated base can inhibit the binding of methyl-CpG binding domain proteins to DNA in vitro[3].

The perception of 5-hydroxymethylcytosine was changed drastically earlier this year when two research groups reported the presence of substantial amounts of this modified base in Purkinje cells of the brain and in embryonic stem cells [4,5]. In one study, the presence of an unidentified spot on thin-layer chromatograms led to the subsequent identification and confirmation of 5-hydroxymethylcytosine by mass spectrometry approaches [4]. In the other landmark study, the enzymatic activity that produces 5-hydroxymethylcytosine was identified. Starting with sequence homology searches with domains of trypanosomal thymidine hydroxylase proteins, Tahiliani et al. identified three related proteins of the iron-dependent 2-oxoglutarate oxygenase family, TET1–3 [5]. The authors went on to show that transfection of human cells with TET1 unambiguously produced the modified base 5-hydroxymethylcytosine from 5-methylcytosine in DNA.

The unexpected presence of 5-hydroxymethylcytosine in DNA and the discovery of an enzyme that produces it raise important questions about the functional role of this base modification. For example, is this base further modified by glycosylation as in bacteriophages? What is the distribution of the base in different cell types and in different compartments of the genome? Is the modified base recognized by specific proteins? Moreover, it is quite possible that 5-hydroxymethylcytosine is an intermediate that occurs during active enzymatic demethylation of 5-methylcytosine. The existence of a mammalian DNA demethylase has remained enigmatic and controversial [6], but 5-hydroxymethylcytosine provides a reasonable potential intermediate in a direct demethylation reaction that does not require base excision repair and DNA strand cleavage. In this context, the observation that DNA methyltransferases can add aldehydes to DNA, providing a hydroxymethyl intermediate, provides a hypothetical mechanism of how formaldehyde may be released from 5-hydroxymethylcytosine in a reversible enzymatic reaction [7]. Finally, the discovery of 5-hydroxymethylcytosine raises important issues of how this modified base will be distinguishable from 5-methylcytosine, either by sodium bisulfite sequencing, restriction enzyme cleavage or by any other techniques, before epigenome mapping studies of 5-hydroxymethylcytosine will become feasible.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.

Enhancers in the spotlight

Evaluation of: Heintzman ND, Hon GC, Hawkins RD et al.: Histone modifications at human enhancers reflect global cell-typespecific gene expression. Nature 459, 108–112 (2009).

The ultimate goal of epigenomics research is to understand in detail the mechanisms that so precisely determine the patterns of gene expression that differ between the approximately 200 cell types of the human body. The prime candidate genomic elements that may play important regulatory roles in this process are promoters, enhancers, silencers and insulators. In human cells, transcription start sites of active genes are generally characterized by high levels of histone H3 lysine 4 trimethylation (H3K4me3), high levels of various lysine-acetylated histones and by the presence of the histone variant H2A.Z. In contrast, inactive genes have low levels of H3K4 trimethylation and histone acetylation at promoter regions, and often contain high levels of H3K27me3 [1–4]. In previous work, enhancers were shown to have a characteristic peak of H3K4 monomethylation (H3K4me1) along with histone H3K27 acetylation and presence of the histone acetyltransferase p300 [3]. In order to understand how these histone modification patterns vary across different cell types, Heintzman et al. have analyzed chromatin states in five different cell lines using chromatin immunoprecipitation microarray (ChIP-chip) technology [5]. To their surprise, the chromatin signatures at promoters (H3K4me1, H3K4me3 and H3K27 acetylation) were remarkably similar across all cell types analyzed. On the other hand, the same chromatin marks, when present at enhancers, were highly cell type-dependent. In addition, the localization of the insulator protein CTCF, which in some cases is known to regulate promoter–enhancer interactions, was shown to be largely cell type-independent as well. At a global level within CTCF-delineated domains, enhancers predicted by their specific chromatin marks were enriched near HeLa cell-specifically expressed genes. Whereas most genes in HeLa and K562 cells were not expressed cell-type-dependently, the chromatin modification patterns at most of the over 55,000 predicted enhancers were in fact cell-type-specific [5]. These data suggest that enhancers, rather than promoters, are largely responsible for setting up cell-type-specific patterns of gene expression, and that for many genes different cell-specific enhancers may activate expression of the same gene. By analyzing only two cell lines in detail, the authors identified over 50,000 enhancers, and only approximately 10% of those were common to both cell lines. This implies that the human genome probably contains millions of cell-type-specific enhancers that regulate and fine-tune gene-expression programs in the many different cells and tissues of the body.

Sperm histones are organized to carry epigenetic clues into the zygote

Evaluation of: Hammoud SS, Nix DA, Zhang H, Purwar J, Carrell DT, Cairns BR: Distinctive chromatin in human sperm packages genes for embryo development. Nature 460, 473–478 (2009).

Spermatozoa are highly specialized cells for transmitting the haploid male genome and paternal DNA methylation marks into the oocyte. Sperm DNA is tightly bound by protamines. A small fraction of sperm DNA, however, is associated with histones [1], and the distribution of histones in sperm is not random [2–5]. There is evidence that sperm-specific chromatin can be transmitted into the zygote [6]. If epigenetic clues are present in the sperm chromatin, these are likely to be transmitted into the zygote to facilitate zygotic gene expression and the recognition of the paternal alleles of imprinted genes [7]. Using high-resolution genomic approaches Hammoud and colleagues now provide a comprehensive genome-wide map of human sperm histones [8]. They provide clear evidence that sperm chromatin is very well organized to aid sperm-specific functions and to provide epigenetic clues to early embryonic development and genomic imprinting. This chromatin organization consists of the placement of nucleosomes, placement of histone variants and also the specific post-translational modifications of histones at regions of functional importance. Nucleosomes in sperm are localized at specific loci – for example, at promoters of genes that are essential for early development, and also at imprinted regions. Testis-specific histone variant TH2B is localized to genes with functions in sperm biology, capacitation and fertilization. The covalent histone tail modification, H3K4me3, is present at spermatogenesis-specific gene promoters, at developmentally critical genes, such as HOX clusters, at regulatory microRNAs and also at critical transcription factor gene promoters. Approximately half of the ‘poised‘ promoters, which exhibit bivalent chromatin in embryonic stem cells [9], are marked by H3K4me3/H3K27me3 bivalency in sperm chromatin. H3K27me3, but not H3K4me3, is found at development-regulating genes that are silent in the early embryo. Opposite DNA methylation marks in sperm and oocyte at the differentially methylated regions (DMRs) of imprinted regions constitute the gametic imprints. Maternally methylated DMRs carry H3K4me2 and H3K9ac signals in spermatocytes and spermatids [10]. Hammoud and colleagues now show that paternally expressed imprinted genes and imprinted noncoding RNAs are marked with large blocks of H3K4m3, whereas paternally silenced genes carry moderate enrichment of H3K9me3 in spermatozoa [8]. Repressive chromatin marks at paternally methylated DMRs, when transmitted from sperm into the zygote, could be critical in maintaining CpG methylation [7] during the global zygotic demethylation events [11]. Similarly, active chromatin marks could be important in protecting the paternal allele of maternally methylated DMRs from DNA methylation during the wave of global de novo methylation around implantation. It will be very interesting to see when these paternal chromatin marks are established at DMRs during male germ cell development, how they escape the histone–protamine transition, and how long they can survive in the zygote and in the early embryo. In addition, it will be important to see if the oocyte carries specific histone modification marks at DMRs and other important loci.