Chromatin's structure plays an important regulatory role in DNA template-dependent processes including transcription, replication, recombination, repair, segregation, chromosomal stability, cell cycle progression, and epigenetic silencing.1-3 Many factors can induce remodeling (changes in chromatin structure) including histone modification and the binding of numerous non-histone proteins that are loosely termed the transcriptional apparatus.
Histones contain two distinct domains. The proteins' amino-terminal tails, which protrude from the nucleosome core, are unstructured and highly positively charged owing to the presence of several lysine and arginine residues. Conversely, the histone core domain is globular and responsible for the histone:histone interactions involved in nucleosome formation.2 A variety of well-conserved post-translational modifications occur in the histone tail domain, including acetylation, phosphorylation, methylation, ribosylation, ubiquitinylation, and glycosylation; the most well-studied of these is the acetylation of histone tail domains on specific lysine residues.1-5 The conventional wisdom holds that these modifications alter the strength of the interaction between DNA and the histone core, thereby changing the accessibility of this DNA to transcription factors, leading to changes in gene expression. However, a new hypothesis has recently emerged that challenges the simplicity this model.
Reinvigorating Chromatin Research
C. David Allis and colleagues reinvigorated chromatin research in 1996 when they demonstrated that the yeast Gcn5p protein, already known to positively regulate gene transcription, was a histone acetyltransferase.6 Acetylation of histone tail domains on specific lysine residues neutralizes the amino acid's positive charge.1 This event may induce a conformational change that weakens nucleosomal DNA:histone interactions, making the DNA more accessible to transcription factors.2 Chromatin-based research has surged in the wake of Allis' publication. Other coactivators, such as TAFII250, SRC-1, ACTR, and CBP/p300 were subsequently also found to contain histone acetyltransferase (HAT) activities critical to their functioning.5
Researchers have also identified the enzymes that can undo histone acetylation, called nuclear histone deacetylases (HDACs). This family includes the mammalian HDACs 1-8, as well as large multi-protein complexes involving mSin3, HDACs 1 and 2, RbAp-46 and -48 (retinoblastoma and histone binding proteins), Mi2 (autoantigen), and SAP-30 and -18 (mSin3-associated proteins).5
Whereas histone acetylation is generally associated with transcriptional activation, deacetylation is associated with silencing.5 The interplay of HAT and HDAC activity therefore yields a steady-state level of histone acetylation that defines the transcriptional activity of a given region; certain stimuli can disrupt this equilibrium to change gene expression levels.
It has been proposed that HATs and HDACs recognize specific consensus motifs, analogous to the functioning of protein kinases and phosphatases. In fact, the bromodomain of a human HAT, PCAF (p300/CBP-associated factor), binds to acetyl-lysine in the context of H3 and H4 tails, a phenomenon reminiscent of the binding of phosphorylated tyrosine residues by SH2-containing molecules in certain contexts.1
Other post-translational histone modifications of specific tail-domain residues exist in addition to acetylation. These include serine (Ser) phosphorylation and the more recently identified methylation of arginine (Arg) and/or lysine (Lys) residues, which are implicated as components of a transcriptional co-activation pathway for steroid-sensitive nuclear receptors.2
Histone methylation occurs predominantly on H3 and H4, and lysines can be mono-, di-, or trimethylated. Disregulation of these modification processes may play a role in disease states. For example, Rsk-2 has recently been identified as a kinase that phosphorylates H3 in vitro, and mutations in Rsk-2 are linked to Coffin-Lowry syndrome in humans, which involves a defect in EGF-stimulated H3 phosphorylation.1
Enter the "Histone Code"
Some modifications are highly conserved among eukaryotes, such as the phosphorylation of histone H3 at invariant Ser10 during rapid and transient transcriptional activation.1,2 Studies show that this phosphorylation event is sometimes followed by Lys14 acetylation, giving rise to a multiply-modified H3.2 In fact, many HATs have a substrate affinity for Ser10-phosphorylated H3. For example, Ser10 phosphorylation also combines with specific acetylation on Lys9 and Lys14 on the same H3 tail during immediate-early gene expression activation upon EGF stimulation. Likewise, the combination of Ser10 and Ser28 phosphorylation on the same H3 tail is a hallmark of mitosis in some mammalian cells.1,2
HATs appear to act synergistically with another group of coactivators, the ATP-dependent family of chromatin remodeling enzymes (ATPases). The energy produced by ATP hydrolysis evidently disrupts DNA:histone interactions and thereby tends to enhance local gene expression. As noted above, sequence-specific DNA-binding transcriptional activators act by recruiting several coactivators--including HATs and ATPases--to particular promoter regions in addition to the components of the general transcriptional machinery.
In some instances, for example, in yeast, ATP-dependent remodeling must precede HAT action, but the reverse has also been shown for some mammalian genes.4 Thus, these two families of remodeling enzymes act synergistically to bring about a change in chromatin structure to a state more permissive for subsequent transcription.
The recruitment of histone remodeling enzymes is not random. Instead, a strict and perhaps sequential order must be followed that may be determined by either the promoter involved or position within the cell cycle.4 While there are a few instances in which interdependent chromatin-remodeling enzyme recruitment occurs, most studies show independent recruitment with an obligatory sequence of events.
Histone methylation is another integral facet of the histone code. The gene-silencing family of Su(var) proteins, which contain histone methyltransferase activity (HMT)--as well as HDACs, protein phosphatases, and heterochromatin-associated structural proteins--selectively modifies H3 by adding methyl groups to Lys9.2 This event appears to antagonize Ser10 phosphorylation and is associated with chromosomal dysfunction during mitosis.2 Conversely, the Ser10 phosphorylation of H3 antagonizes the methylation of Lys9.2 Thus, one modification can effect subsequent modifications either synergistically or adversely.
One interpretation of this data is that histone deacetylation may be required to free lysine amino groups for subsequent methylation by HMTs.2 Thus, expression patterns at any given time may be controlled in part by characteristic combinations of histone modifications that in turn determine which trans-acting transcriptional components are permitted access to particular DNA sequences.2 The histone code is far from completely unraveled. It may involve additional elements, such as the spacing between modification sites within histone tails.1
Chip-ing Away at the Code
One approach to studying transcriptional regulation at a specific promoter at the chromatin level is to examine the protein composition of such complexes, including identification of specifically modified histone residues and other trans-acting, non-histone proteins. Early chromatin-analysis methods involved identification of DNA sequences and proteins after biochemical fractionation of transcriptionally active from inactive chromatin.3 However, chromatin is a dynamic entity and the composition of DNA-binding proteins and histone modifications at any one locus is constantly changing. Such transient interactions and events are difficult to detect by fractionation and many essential factors can be lost in the process.
A relatively new, powerful, and versatile tool for studying such DNA:protein interactions within a native chromatin environment is the chromatin immunoprecipitation (ChIP) assay. ChIP overcomes the limitations of other protocols by freezing DNA:protein interactions in situ with covalent crosslinking using formaldehyde.3 This process is efficient, reversible, and faithfully preserves chromatin structure.3 Chromatin analysis using ChIP is possible thanks to the wide variety of high-titer, residue- and modification-specific histone antibodies currently available. The advent of such antibodies has made ChIP the most popular method for studying effects of chromatin topology on transcriptional activity and other DNA template-dependent processes. Scientists have successfully used the technique with many different model systems including human cells, Drosophila, the sea urchin embryo, Archaea, protozoa, and budding yeast.3
The basic ChIP protocol involves several steps.2,3 Following treatment under conditions in which the gene of interest is transcriptionally affected nuclear protein:DNA complexes are cross-linked with one percent formaldehyde. Nuclei are then isolated and the DNA is sheared, often by sonication, to produce chromatin fragments approximately 500-1,000 basepairs in length. An alternative method uses micrococcal nuclease, which cuts between nucleosomes to create a discrete, 146-basepair chromatin fragment ladder.8 Micrococcal nuclease digestion is more targeted than sonication and, by yielding nucleosomal unit-sized fragments, allows the subsequent use of primers designed across nucleosomes, a somewhat more specific approach. After DNA fragmentation, chromatin immunoprecipitation is achieved using the appropriate modification-specific histone antibody. The crosslinking is then reversed and the DNA is purified and quantified. Finally, the enriched DNA sequences are examined for regions of interest, generally by PCR.
Given the protease-sensitivity of histone tails, researchers must use protease inhibitors throughout the protocol. They must also consider the stability of the modifications being studied. The success of the technique depends in part upon the abundance of the DNA:protein interaction of interest, the quality of the antibody being used, including its specificity and affinity, and the complexity of the genome being assayed.3
An Antibody for Everybody
|Courtesy of W.M. Bonner, National Cancer Institute|
Jelinek explains that a recent company breakthrough has come with the development of two new antibodies against dimethylated histone H3: anti-dimethyl-H3 (Lys4) and anti-dimethyl-H3 (Lys9). The excitement regarding production of these two new antibodies involves the functional outcome of H3 dimethylation at Lys4 versus Lys9. Studies have shown that dimethylation of Lys4 on histone H3 strictly occurs during a transcriptionally active state, whereas dimethylation of Lys9 on histone H3 strictly occurs during transcriptionally inactive states. Upstate Biotechnology is currently preparing to release monoclonal forms of these two antibodies. Also, given the role of these two modifications as a master switch for the histone code, the company is looking to incorporate these two specific antibodies into a microarray kit.
Santa Cruz, Calif.-based Santa Cruz Biotechnology also offers a wide variety of antibodies directed against core histones, specifically modified histones, HATs, HDACs, and other DNA-binding proteins known to be involved in chromatin-remodeling complexes. All of the company's antibodies have been used successfully in Western blot (WB) and immunohistochemistry (IHC) studies, but only some have been tested in immunoprecipitation (IP) studies. Since formaldehyde cross-linking can destroy or mask certain epitopes, not all antibodies will work in a ChIP assay. The only way to be sure is to test the antibody directly.
Warren Shore, president of Swampscott, Mass.-based US Biological, explains that his firm has recently completed a database of antibodies that includes those directed against core histones and different varieties of acetylated histones H3 and H4. While none of US Biological's histone antibodies have been company-tested in ChIP studies, they have been used successfully in WB, IHC, and ELISA assays.
Some companies carry just one or a few high-quality antibodies directed against a specific modification that is associated with the research area they are focused on. For example, Gaithersburg, Md.-based Trevigen's main focus is DNA damage and repair. As such, the company offers an antibody against phosphorylated histone H2AX, a ubiquitous member of the histone H2A family. Serine 139 of histone H2AX becomes rapidly phosphorylated in response to double-strand DNA damage. Therefore, this antibody is unique in that it only detects phosphorylated histones at sites of double-strand DNA breaks. Researchers have used this antibody successfully in WB and IHC studies, but it has not yet been tested in a ChIP assay.
Golden, Colo.-based Affinity Bioreagents and Alexis Corp. of San Diego offer antibodies directed against HDAC 1-4 and HDAC 1-7, respectively. Both companies recommend their antibodies' use in Western blotting, but like, like Trevigen, have not tested them in ChIP studies.
According to Meddi Awalom, San Diego-based Calbiochem production manager, the company has begun marketing several new antibodies for its 2002 catalogue. Adding to the company's previously available monoclonal anti-H1 and polyclonal anti-HDAC1 antibodies, Calbiochem now offers antibodies directed to HDACs 2-5 as well as anti-H3, anti-acetyl-H3, anti-phospho-H3(Ser10) and anti-acetyl-H4. While all of Calbiochem's antibodies have been used successfully in WB and IHC studies, the anti-phospho-H3(Ser10) and anti-acetyl-H4 antibodies have also been tested and proven compatible with IP techniques. Calbiochem offers its antibodies in a 5 µg trial size, for customers wanting to test their success in particular experimental designs before ordering the full 50 µg aliquot.
1.B.D. Strahl, C.D. Allis, "The language of covalent histone modifications," Nature, 403:41-5, 2000.
2. Upstate Biotechnology, "Histones: From structural proteins to chromatin regulators," www.upstatebiotech.com/support/PDFs/histones.pdf.
3. M.-H. Kuo, C.D. Allis, "In vivo cross-linking and immunoprecipitation for studying dynamic protein:DNA associations in a chromatin environment," Methods: A Companion to Methods in Enzymology, 19:425-33, 1999.
4. C.J. Fry, C.L. Peterson, "Chromatin remodeling enzymes: Who's on first?," Current Biology, 11:R185-97, 2001.
5. Santa Cruz Biotechnology Inc., "Transcription regulators: Histone acetylases and deacetylases," www.scbt.com.
6. J.E. Brownell et al., "Tetrahymena histone acetyltransferase A: A homolog to yeast Gcn5p linking histone acetylation to gene activation," Cell, 84:843-51, 1996.
7. B.A. Maher, "Researchers focus on histone code," The Scientist, 15:15, Sept. 17, 2001.
8. G. Fragoso, G.L. Hager, "Analysis of in vivo nucleosome positions by determination of nucleosome-linker boundaries in crosslinked chromatin," Methods: A Companion to Methods in Enzymology, 11:246-52, 1997.