Histone proteins are subject to a variety of posttranslational modifications (PTMs), many of which (i.e., methylation, acetylation, ubiquitylation and SUMOylation) occur at lysines [4], or citrullination at arginine [5], or serotonylation at glutamine [6]. While only a handful of modification sites have been identified in the buried histone-fold domains, PTMs on the flexible N-terminal tail regions are quite common. Many histone modifications mediate functionally significant changes in chromatin structure and do so either by altering chromatin structure/dynamics directly or through the recruitment of histone modifiers and/or nucleosome remodeling complexes [4, 7]. Histones are the components of nucleosomes, which turn over several times within each cell cycle, suggesting that histone modifications are rapidly erased by nucleosome turnover. This turnover of modified nucleosomes may propagate epigenetic changes and determine the valency of methylation [8].

Histone PTMs have been found to affect a host of chromatin-based reactions, including transcription, heterochromatic gene silencing and genome stability [9-12]. For example, C Weber et al showed that Kdm6b, a histone demethylase, mediated the temperature-dependent sex determination in reptiles [13]. Modifications implicated in gene expression are especially significant as they have the potential to influence whole transcription programs. H3K4me3 initiates transcription, H3K27ac promotes gene transcription, while H3K27me3 represses the transcription of developmental genes [14, 15], although such a characterization remains debatable [16]. Defects in histone PTM metabolism have been linked to misregulated gene expression in various in vitro models and, in some cases, also correlate with human disease, as has been demonstrated for immunodeficiency disorders and a variety of human cancers [17-22]. Thus, how histone markers are regulated and influence the association of PTM-specific binding proteins will continue to be an area of significant investigation [7, 23-25].

Determining the function of histone PTMs often involves investigating the modification's abundance and interacting partners. The methods described here address these areas and include summaries and references for protocols detailing histone purification methods, the production of recombinant site-specifically modified histones, peptide-based systems to characterize PTM-binding proteins and histone modification-specific antibodies, in vivo histone modifications imaging, and different methods for analyzing chromatin immunoprecipitation samples. Histones and their modifications such as those with circulating cell-free nucleosomes have been proposed as diagnostic tools for cancers or kidney transplant rejection [26].

For detection of histone modifications by Western blotting, it often is possible to use whole-cell lysates made by extraction with SDS Laemmli sample buffer. In the case of animal cell lines, cells collected by centrifugation can be directly resuspended in sample buffer and boiled for loading [27, 28] ; note, however, that some protocols additionally recommend sonicating the samples after the extraction step. Protein extracts of fungi can be prepared the same way with the exception of an optional alkali pre-treatment step [29]. This step can be omitted, however, if it is experimentally necessary to minimize sample handling time, so long as the omission is noted and samples for follow-up experiments are treated the same way. Following extraction, the sample's insoluble fraction is removed by centrifugation, leaving behind the soluble whole cell extract in the supernatant.

For some applications, it is necessary to examine either histone-enriched fractions or purified histone proteins; examples of an enriched sample would be isolated nuclei or a crude chromatin preparation. The isolation of nuclei from cultured metazoan cells or yeast is very straightforward and requires essentially three steps: hypotonic swelling (after prior cell wall digestion in the case of yeast), cell membrane lysis by mechanical shearing (i.e., breakage with a Dounce homogenizer or mild agitation on a rotator), and isolation of nuclei by centrifugation [30, 31] (Figure 1A). The isolation of crude chromatin is also quite simple and includes only a detergent lysis step followed by the sedimentation of chromatin by centrifugation [32-34] (Figure 1B).

Several of the available histone purification protocols are excellent and easy to follow [31]. In the method described here, histones are extracted from nuclei using a dilute sulfuric acid solution and then purified by column chromatography (Figure 1C). The beauty of this protocol is that nucleic acids and many of the non-histone proteins can be readily removed by centrifugation because of their insolubility at acidic pH. The soluble histone fraction is then precipitated with trichloroacetic acid (TCA) and, if desired, purified over a reversed-phase HPLC column. The histones at this point can be used in a variety of applications, including Western blotting and mass spectrometry. Please note that another publication [31] also describes an alternate, high-salt histone extraction method.

[enlarge]

Figure 1. Protocols for isolating intact nuclei (A), preparing crude chromatin fractions (B), and purifying histones (C).

In addition, the technique of Chromatin Affinity Purification with Mass Spectrometry (ChAP-MS) can be used for affinity purification of ~1 kb regions of chromatin for analysis of histone PTMs [35].

Histone PTMs are generally detected through the antibodies. Phospho-histone H3 (ser10) is commonly used as a marker for mitosis [14, 36]. M Coolen et al identified mitotic killifish cells through phospho-histone H3 staining [36]. Rhodes JDP et al stained mouse ESCs with CST phospho-histone H3 (ser10) mouse mAb ( 9706) to detect mitotic cells [37] and Y Shwartz et al used CST rabbit monoclonal antibody 3377 [38]. Butler AA et al detected histone methylation through MilliporeSigma antibodies against H3K9me2 ( 07-441), H3K27me3 ( 07-449) and H3K4me3 ( 04-745) in Western blots [39] ; Batie M et al used Cell Signaling Technology antibodies against H3K4me3 (9751), H3K9me3 (9754 / 13969), H3K27me3 (9733), H3K36me2 (2901) and H3K36me3 (4909) in immunoblotting Hela and HFF cell lysates [40]. The quality and specificity of anti-PTM antibodies should be carefully evaluated prior to experimental application. Concerns to be addressed include cross-reactivity with alternate histone modification sites, recognition of the unmodified (recombinant) protein and cross-reactivity with other nuclear species. The procedures for this type of evaluation are very straightforward and involve either Western blotting of nuclear extract preparations against recombinant histones or immunoblotting an array of modified and unmodified peptides spotted on nitrocellulose membrane. Table 1 lists antibodies against histones and histone-related products cited among the formal articles Labome has surveyed for Validated Antibody Database.

Recent initiatives to address anti-PTM histone antibody quality issue have resulted in the characterization of more than 200 antibodies against 57 different histone modifications [41]. About 20% of antibodies were deemed not effective. The supplementary data to this report catalogues the performance of the different antibodies tested in dot blots, Western blots (across different species), and chromatin immunoprecipitation (ChIP) assays. Xia W et al used CST H3K27me3 antibody ( 9733) for CUT&RUN and immunocytochemistry and Active motif H3K27ac antibody ( 39133) on human oocytes or embryos [15].

A new peptide array application, the MODified™ Histone Peptide Array, has recently become available for the analysis of the binding specificity of histone modification-specific antibodies [42]. Additionally, a Histone Antibody Specificity Database has been created [43]. The database catalogs commercially available histone antibodies and their behavior in peptide microarray-based experiments.

One method for producing site-specifically modified histones involves the ligation of protein fragment in vitro [44-48] (Figure 2A). The premise of this method is to chemically ligate a synthetic, modified peptide (corresponding to either the very N- or C-terminal end of the protein) to a recombinant fragment containing the complementary part of the protein, followed by purification of the full-length ligation product [49]. A related method, published more recently, describes the use of native chemical ligation to produce a fully synthetic modified histone by the sequential addition of synthetic peptides corresponding to consecutive parts of the protein [50].

[enlarge]

Figure 2. Production of site-specifically modified histones by expressed protein ligation (A) and direct incorporation in vivo (B).

Besides by chemical ligation, lysine-acetylated histones can also be produced by direct incorporation in vivo using E.coli that have been genetically engineered to incorporate acetyl-lysine at UAG codons [51] (Figure 2B). The premise of this system is to introduce into E. coli an M. barkeri pyrrolysyl tRNA synthetase variant that has been sequence-optimized to charge its cognate tRNACUA with acetyl-lysine. Thus, to produce a site-specifically acetylated histone, one would need only to add to the strain a construct containing a UAG codon at the desired modification site.

In cases where it is of interest to study the interaction of a protein with endogenous histones, it is necessary to first convert the chromatin to mononucleosome-sized pieces by digestion with micrococcal nuclease (MNase). The protein of interest can then be immunoprecipitated from the soluble chromatin fraction and evaluated for association with the core histones and different histone modifications by Western blotting [52, 53]. This experiment can also be done as a pulldown assay using a recombinant bait protein that has been incubated in the solubilized nucleosome fraction and then purified [54]. If it is obvious that a particular modification is enriched in complexes containing the target protein, it may then be of interest to characterize the interaction in binding assays using modified peptides (see below). In cases where no such preference is determined, yet it is still of interest to know whether an interaction is modification dependent, relative affinities for modified versus unmodified histones can be determined in binding assays using native and recombinant histones [55].

One common method of purifying native histones from tissue culture cells is to produce oligonucleosome-sized pieces of chromatin by either limited digestion with MNase or mechanical shearing, followed by chromatography of the fragments on a hydroxyapatite column and elution at high salt [56, 57] (Figure 3A). Recombinant histones are insoluble when overproduced as monomers; however, it is possible to recover the overexpressed proteins from inclusion bodies [58, 59] (Figure 3B). To start, preparations of inclusion bodies are made from bacterial cultures individually overexpressing each histone. The histones are then extracted from the inclusion bodies, purified over sequential ion-exchange resins, and eluted using a linear salt gradient. To generate histone octamers, equimolar amounts of H3, H4, H2A and H2B are first unfolded, combined, refolded by dialysis, and then purified over a sizing column.

[enlarge]

Figure 3. Purification schemes for isolating native (A) and recombinant histones (B).

The modification preferences of a protein can be determined by comparing the relative binding affinities of the protein for differentially modified peptides. This can be accomplished in a few different ways. In one format, bead-immobilized peptides are used to pull down a test protein - using either recombinant protein or a nuclear extract – and the relative recoveries of the protein are determined by Western blotting [60-62] (Figure 4A). The principle of short peptides as bait molecules has also been applied in unbiased experiments to discover previously unknown histone PTM-binding proteins [53, 54, 63, 64]. In this assay, binding proteins from a nuclear extract are identified on the basis of enrichment on the modified peptide relative to an unmodified peptide or a peptide carrying the same modification at a different amino acid.

Also described in the literature are array-based methods capable of screening the interaction of probe proteins with multiple peptides simultaneously (Figure 4B). For such experiments, the protein of interest is incubated over the surface of a peptide microarray, which is essentially a streptavidin-coated slide spotted with different biotinylated peptides, and the protein peptide complexes are visualized by detection with a fluorophore-conjugated antibody and an array scanner [65-67]. Studies using an inverse setup, i.e., protein arrays incubated with fluorescently-labeled peptide, have been described as well [68].

[enlarge]

Figure 4. Characterization of histone modifications using peptide-based pulldown assays (A) and peptide microarrays (B).

Genomic sites that are enriched for a particular PTM can be characterized by immunoprecipitating chromatin fragments containing the mark of interest and then quantifying the relative proportion of different loci that the PTM is associated with (Figure 5). Detailed ChIP protocols are available [69-71]. The first step of most ChIP protocols is to treat cells with formaldehyde to "freeze" the position of chromatin-associated proteins by crosslinking. Whereas cells of metazoan origin are then lysed directly, yeast cells must first undergo cell wall breakage, which can be accomplished by either mechanical shearing or enzymatic digestion. Depending on the nucleotide resolution desired at detection, chromatin is cut into short pieces in one of two ways: shearing by sonication to yield fragments of between 200 and 500 bp or digestion with MNase down to individual nucleosomes. Immunoprecipitation of the extracts is performed the same way as in a conventional IP experiment except that after the elution step, the eluate is incubated overnight at 65°C to reverse cross-links. On the following day, the samples are treated with Proteinase K and then phenol/chloroform extracted to recover the co-precipitating DNA fragments.

[enlarge]

Figure 5. Chromatin immunoprecipitation (A) results in the isolation of co-purifying DNA fragments, whose sequence and enrichment can be analyzed by PCR (B), ChIP-chip (C), and/or Chip-seq (D).

A variation of the ChIP procedure, ZipChip (Figure 6), was recently designed by Harmeyer et al [1]. The method facilitates the detection of histone modifications that are more difficult to identify, while reduces the procedure time. Moreover, it allows for the screening of multiple loci. Another variation of CHiP is CUT&RUN [15, 72].

If it is of interest to analyze PTM levels at only a handful of loci, this can be accomplished by either real-time PCR or quantification of PCR products on an ethidium-stained gel (Figure 5B). Briefly, the loci to be analyzed are amplified from dilutions of the IP and input fractions and compared, ideally, against these fractions from a parallel IP of the histone the PTM is attached to. To determine whether the PTM is enriched at a particular location, the normalized PTM/histone ratios should be compared with those at a region nearby that the PTM is not predicted to be associated with. For example, Butler AA et al immunoprecipitated formaldehyde-treated N2a cell chromatins with an H3K9me2 antibody and assayed the c-Fos gene promoter through RT-PCR [39].

[enlarge]

Figure 6. ZipChip procedure diagram as in [1].

Microarray-based methods enable the analysis of histone modification enrichment at large numbers of loci simultaneously. The microarray itself is a coated glass slide affixed with different oligonucleotides – ranging in number from the tens of thousands to the tens of millions. Sites where a protein is enriched are determined by co-hybridizing to the array fluorescently labeled DNAs derived from the IP and input fractions and comparing the normalized IP/input intensities across the array (Figure 5C).

The first steps of preparing DNAs for microarray analysis are to amplify the DNA from each test (IP) and reference (input) sample and differentially label the amplification products with either of two fluorophores. DNA amplification can be accomplished in several ways, and the methods that are PCR based begin with the addition of primer binding sites to the DNA ends. This part can be done by either ligating the DNAs to short linkers of a known sequence [73] or performing an initial two rounds of annealing and extension using primers that have degenerate 3' sequence and known 5' sequence [69, 74]. The end-tagged products are then PCR amplified for several rounds using primers that recognize the added linker sequence.

Another widely used method for DNA amplification involves the conversion of DNAs into transcription templates, followed by linear amplification of the products by transcription with T7 RNA polymerase [75]. To start, short polyT tails are added to the DNA 3' ends to create a priming site for the first-strand synthesis by Klenow. The oligo used for priming contains a stretch of A's at the 3' end of the minimal T7 RNA polymerase promoter sequence, which enables subsequent amplification of the first-strand products by in vitro transcription.

Samples that have been amplified by PCR can be fluorescently labeled in one of two ways: (1) by incorporating a fluorophore-modified dNTP during the final set of PCR cycles or (2) indirectly, using an amine-modified dNTP that can be dye coupled later [76]. The same labeling principles apply to products amplified by transcription, except that the modified nucleotides are incorporated during a final reverse transcription step [76]. When preparing samples for labeling, it is important also to label an appropriate reference sample that the test sample can be co-hybridized with. Thus, the labelings are set up such that one of the two samples is labeled with Cy3 and the other with Cy5. After the final amplification or post-labeling step, the samples are purified, and the labeled test and reference samples are mixed and prepared for hybridization.

Large-scale enrichment analysis can also be performed using a variety of massively parallel DNA sequencing methods. Such methods make possible the parallel sequencing of millions of DNA molecules in real time. The majority of ChIP-seq studies published so far were done using the "sequencing by synthesis" platform by Illumina. The premise of this platform is to perform parallel sequencing of millions of clonal DNA clusters on the surface of a flow-cell (Figure 5D).

ChIP-seq samples are prepared by ligating the IP'd DNAs to oligonucleotide adapter molecules, after which the ligation products (in some protocols) are amplified for several rounds by PCR and then purified [77, 78]. Samples are then injected into a flow cell whose surface is coated with oligonucleotides complementary to the ligation product adapter sequences. The density of the tethered oligonucleotides is such that, during the amplification steps, the DNAs remain spatially close to the parent template, as the newly synthesized molecules originate from primers that are attached to the flow-cell surface adjacently. The sequencing phase of the protocol is accomplished by single-base extension using nucleotides that are fluorescently labeled and can terminate elongation reversibly. The sequencing reaction therefore proceeds as follows: a labeled nucleotide is added to the free 3' end, after which elongation pauses to enable detection of the incorporated nucleotide. Subsequently, the terminator group is cleaved to permit the addition of the next nucleotide. Nucleotide extension is iterated for several more cycles, which typically results in read-lengths in the 40-bp range. For an in-depth review of next-generation sequencing methods, and comparison with microarray-based approaches, please see [79-82].

Bias in ChIP-seq can be introduced due to the chromatin state [83]. Open chromatin regions often lead to false positive, and algorithmic adjustment should be made.

ChIP-seq proved very useful for the detection of the presence or absence binding sites. Determining any changes in the binding intensity caused by different biological condition would be more informative. Chen et al have recently developed a statistical method, ChIPComp, to perform a quantitative comparison of multiple ChIP-seq datasets in order to detect differential protein binding or histone modification [84]. In this method, read counts from all datasets are compared, and peaks are selected to generate a set of candidate regions, which are further analyzed via an R software package freely available at: http://web1.sph.emory.edu/users/hwu30/software/ChIPComp.html. Computational workflows for differential binding analysis are also emerging [85].

Quantitative information on the level of histone modification at any particular chromosomal loci can now be obtained experimentally. Van Galen et al suggest a quantitative multiplexed ChIP-seq technology, Mint-Chip-seq, to profile chromatin quantitatively [86]. They map the relative levels of multiple histone modifications across multiple samples, while monitoring the dynamic changes of the chromatin following different cellular treatments, both at global and locus-specific levels. A detailed protocol as used in [86], as well as a further optimized version are available online from the Bernstein laboratory: http://bernstein.mgh.harvard.edu/resources/.

A method has been developed to identify specific histone PTM for a specific gene in a particular cell type in histological sections [87]. The cell type is marked through cell-specific marker with an antibody. The gene of interest is detected through in situ hybridization with a biotin-conjugated probe, and histone with specific PTM is labeled through a histone modification-specific antibody. Proximity ligation assay [88] is used to identify the co-location of the biotin label and histone antibody. The method enables detection at the single cell level.

Mass spectrometry is a technique that proved effective in identifying and quantifying histone PTMs and their binding proteins, beyond the limitations of specific antibodies. The method has proved useful in the study of host-pathogen interactions [89], epigenetics-associated diseases [90], as well as in the analysis of clinical samples [91]. Histone PTM analysis can be done either by using intact histones (the top-down approach), intact histone tails (the middle-down approach) or shorter fragments (the bottom-up approach) with equal reliability [92, 93]. These approaches have been previously described [90, 94], while a new workflow for analysis by the bottom-up approach became available as of 2016 [95].

Apart from the identification of novel PTMs, the technique was shown to be useful in the analysis of the histone modification dynamics during the cell cycle [96]. Meharena et al used Global Chromatin Profiling (GCP) assay, an approach of quantitative targeted mass spectrometry with spiked-in internal control peptides, to quantify changes of histone H3 modifications [97]. Comprehensive reviews of the most influential PTMs, PTM purification methods, the fundamentals of PTMs-related proteomics experiments have recently been published [98, 99], and include suggestions for accurate interpretation of data in the context of the biological processes that are being studied.

[enlarge]

Figure 7. A. Schematic diagram of the split-reporter complementation system; protein-protein interactions between domains A and B lead to the formation of an integral, functional luciferase B. RLuc system for H3-methylation imaging; Chr – histone chromodomain; K9 – Lys9; Met-K9 – methylated Lys9.

Post-translational modifications of histone proteins regulate gene expression during development, as well as the pathogenesis of cellular diseases. Several techniques for in vitro and in vivo imaging of PT-modified histones are known. Most of them are based on the use of a split-reporter complementation system (Figure 7A). Most recently, the potential of the Renilla luciferase (RLuc) split reporter complementation system was extended to image histone 3 methylation [100] (Figure 7B) . In addition, Kanno et al [101], Lin et al [102] and others [103] used the fluorescence resonance energy transfer (FRET) method to image histone modifications in live cells. Histone modifications have also been tracked with the help of fluorescent modification-specific antibodies [104].

[enlarge]

Figure 8. Overview of the EpiTOF platform. Adapted from [2].

EpiTOF (epigenetic landscape profiling using cytometry by time-of-flight) is a method that measures the overall cellular levels of chromatin marks at a single-cell level and in high-throughput format (Figure 8). The authors used it to measure the levels of 8 classes of histone PTMs and 4 histone variants in 22 subsets of immune cells [2].

The key steps of the procedure are the following: PBMCs (isolated from healthy volunteers and cryopreserved) were labelled with lanthanide-labelled antibodies against chromatin marks and in the presence of Fc receptor blocker. This was followed by extracellular marker staining, fixation in 1.6% paraformaldehyde (PFA), permeabilisation and mass-tag sample barcoding was performed. After an overnight incubation with intracellular antibodies in CyTOF buffer containing the Fc receptor blocker, they were stained with 250 nM 191/193Ir DNA intercalator. Cells were resuspended in ddH2O containing four element calibration beads and the final cell suspension was analysed on CyTOF2 (Fluidigm) in Stanford Shared FACS Facility while raw data were analysed with FlowJo software following a dedicated gating hierarchy. Single-cell signal from chromatin markers was normalised to basal levels of histone H3 and H4 proteins using a multivariate linear regression model [2] :

M_i,j = β₀ + β₁H3_i + β₂H4_i

where H3 and H4 represent the raw values for the respective histone proteins in cell I; and M_i,j is the raw value for a given chromatin mark j in cell i across all cell types and subjects. The normalized chromatin score S_i,j was defined as the residual of the regression, corresponding to:

S_i,j = M_i,j–(M_i,j)^

where (M_i,j)^ indicates the best fit of the multivariate regression model [2].

The key findings of the study using 40 lanthanide-labelled antibodies against chromatin marks. showed that the chromatin profiles of immune cells is distinct in each cell type indicative of the hematopoietic lineages from which individual populations are derived. Interestingly, a signature of chromatin marks could predict the identity of immune cells.

In conclusion, EpiTOF is a versatile method with the ability to detect different immunophenotypic markers in immune single cells or any other sample with single cells in suspension derived from solid tissues and detect chromatin dysregulation in a variety of human pathological conditions.

Protein acetylation is the addition of an acetyl group (CH₃CO) on the ε-amino group of lysine residues by substituting an active hydrogen atom. Histones are heavily acetylated and the reaction is catalysed by histone acetyltransferases (HATs) using the metabolite acetyl-coenzyme A (acetyl CoA) as a substrate and it is reversed by deacetylases (HDAC) [105]. ENL protein, a chromatin reader, interacts with acetylated histoness and recruits/stabilizes the transcriptional machinery to increate gene expression [106].

[enlarge]

Figure 9. Lysine acetylation.

HATs comprise two groups: type A, HATs localised in the nucleus acetylating chromatin, and type B, HATs localised in the cytoplasm acetylating nascent histone proteins [107] ; and 5 families of which the p300/CBP is the most prevalent [108]. HDACs are divided into 4 classes based on their sequence similarity, Class I, II, III, IV [109].

While lysine is a positively charged amino acid, acetylation removes the charge thereby reducing the affinity to the negatively charged DNA molecule. This usually results in a more relaxed chromatin structure and is associated with chromatin loosening and increased gene transcription [110]. Besides, acetylated lysines can promote or inhibit the binding of transcriptional regulators [111]. These regulated acetylation and deacetylation cycles control chromatin reorganisation in response to various signals and regulate transcription, DNA repair and DNA replication [4].

Detection of acetylated lysines is achieved by mass spectrometry and acetylation-specific antibodies: The substitution of the hydrogen group by the acetyl group causes a +42.01056 Da mass shift which is easily detectable by mass spectrometers. As histones are lysine-rich, a couple of additional steps can improve the identification of acetylated lysines. This involves the chemical propionylation of unmodified (and monomethylated) lysines before protein digestion inhibiting cleavage by trypsin thereby allowing the generation of larger peptides [112]. Besides, trypsin does not cleave after acetylated lysines. Middle-down proteomics strategies can also be valuable [113].

Furthermore, anti-acetyl antibodies have been produced in animals immunizing with synthetic histone peptides bearing an acetylated lysine. Noteworthy, extra care needs to be taken to validate such antibodies [43]. Histones richness in lysines, adjacent amino acid modifications, and the repetition of certain motifs e.g., K9S10T11, K27S28T29, K56S57T58 on H3, render antibodies less specific. A terrific example is an anti-H3K56ac antibody which has caused significant irritation to the scientific community with its cross-reactivity towards the H3K9ac [114, 115].

Detection of increased histone acetylation is an established way for detection of euchromatic regions, promoters of active genes [116] and active enhancers (H3K27ac) [117]. Saito T et al used anti-H3K27ac antibody MABI0309 from MAB Institute for ChIP-qPCR assay to study the promoter regions of PPAR-alpha target genes [118]. Recently, hypoacetylated coding gene regions have been found to inhibit chromatin phase separation [119], promote transcriptional elongation [120]. Acetylation of specific amino acids on histones can have crucial roles in DNA replication and repair, either by serving as docking motifs for replication and repair proteins or as activators of the DNA damage and repair pathways [121-124].

For their detrimental effect on DNA metabolism and cell proliferation, the regulation of histone modification is associated with pathological conditions, including cancer [125], Rubinstein–Taybi syndrome (RTS) [126], inflammation [127], allergy [128], neurodegenerative diseases [129] and others [130]. The regulators of histone acetylation, namely HATs and HDACs are found mutated in multiple cancer types and are targets of several therapeutic strategies [131, 132] thus specific histone PTMs can serve as predictors for prognosis and classification of cancer subtypes [125].

Protein methylation is the transfer of one to three methyl groups to lysine or arginine residues. Histones are heavily methylated and its regulation is more complex than the other modifications because lysines can be mono-, di-, or tri-methylated while arginines can be mono- or di-methylated. Histone methyltransferases (HMTs) are the enzymes that use an S-adenosyl-L-methionine (SAM) to catalyse the methylation reaction on lysines while protein arginine methyltransferases (PRMTs) to arginines. HMTs are divided into the SET domain-containing and non-SET domain-containing. The methylation reactions are reversible through the action of histone demethylases, comprised of two families: lysine-specific demethylase 1 (LSD1) and Jumonji domain-containing histone demethylases (JmjC domain)

[enlarge]

Figure 10. Lysine and arginine methylation.

Histone arginines are the most frequent hydrogen bond donors in the association of histones with the DNA. While the methylation does not alter the charge of arginine or lysines significantly [133], methyl groups are hydrophobic and bulkier, therefore the main change they incur to histones may be to inhibit or promote the binding of proteins [134].

Methylation is detected by mass spectrometry and anti-methyl specific antibodies. Mono-methylation results in 14.0156 Da, di-methylation in 28.0312 Da and tri-methylation in 42.0468 Da increase on the mass of the modified peptide [135]. Pan anti-methyl antibodies can enrich methylated species prior to mass spectrometry and improve detection of methylation events of low stoichiometry [136]. As with most histone PTMs, histone propionylation improves their detection by mass spectrometry significantly. Likewise, the specificity of anti-methyl specific antibodies should be validated not only for identifying the amino acid of interest, but also to distinguish one methylation status from another [137].

H3K9me has been established as a marker of heterochromatic regions [92]. H3K9me serves as a docking motif to the inner nuclear membrane [92]. H3K4 methylation patterns can signal different chromatin events e.g., H3K4me3 is associated with transcriptional start sites and actively transcribed genes, especially during development [77, 138], H3K4me with enhancers but can also have roles in gene repression in a cell-type specific manner [139]. A handful of histone lysine methylations also play a role in RNA splicing [140, 141].

As with acetylation, histone methylation is associated with a variety of diseases, including poor survival in cancer [142], mental retardation [143], neurodevelopmental disorders [144], ageing [145] and others [146].

The addition of a phosphoryl group (PO4) on serine, threonine, tyrosine and histidine residues is a reaction termed phosphorylation. Phosphorylation is the most abundant post-translational modification in most organisms mediated by kinases which transfer a phosphate from an ATP molecule to the hydroxyl group of the amino acid’s side chain. The reaction is reversed through hydrolysis by phosphatases. Phosphorylation regulates cellular signalling by switching on or off receptor and enzyme functions. Histone phosphorylation, albeit not as prevalent, it has a great impact on chromatin structure.

[enlarge]

Figure 11. Phosphorylated serine, threonine and tyrosine residues. From [3].

Serine, threonine and tyrosines have uncharged polar R groups thus the addition of the phosphate group gives a negative charge thereby reducing the affinity to the negatively charged DNA molecule. There is no rule that predicts the chromatin state from its phosphorylation status, however there are important histone phosphorylation events that are cell cycle-regulated and mark certain chromatin functions or cell cycle phases. Notable examples are the H3S10ph, H3S28ph and H3T3ph in mitosis or the H2AS139ph (γH2A.X) as one of the very first events in the formation of double strands breaks (DSBs) and in DNA damage signalling [147, 148]. γH2A.X Ser139 labeling is commonly used to indicate DNA double-stranded break [149, 150].

The main mechanisms through which histone phosphorylations affect chromatin are: i) the interplay with adjacent modifications, especially acetylation and methylation events [151] ; ii) the inhibition [152, 153] or promotion of binding of “reader” proteins which exert their function [154] ; iii) the potential change in nucleosome dynamics [155].

Detection of phosphorylation events on histones is achieved, as always, by mass spectrometry or phosphor-specific antibodies. In tandem mass spectrometry during the Collision Induced Dissociation (CID), the fragmentation of the ionised peptide that bears the phosphor group results in the loss of a H₃PO₄ molecule and the neutral loss is detected by mass spectrometers. As the phosphorylation is quite labile, enrichment is a usually prerequisite. This is achieved by metal oxide based methods, such as Titanium dioxide, Fe or Zn [156]. In addition, chromatographic fractionation methods, such as hydrophilic interaction chromatography (HILIC) or electrostatic repulsion HILIC (ERLIC) can further enrich the peptide sample for phosphopeptides before mass spectrometry [157]. With respect to phosphotyrosines, pan-phosphotyrosine antibodies work well in enriching tyrosine phosphopeptides [158]. Furthermore, anti-phospho-specific antibodies have been produced in and purified from animals injected with synthetic histone peptides bearing the phosphorylated amino acid. Noteworthy, as with any histone PTM antibody, extra care needs to be taken to validate such antibodies.

Most chemotherapeutic strategies target the stage of DNA replication resulting in double-strand breaks. Thus, γH2A.X can be a marker for patient responses and the efficiency of a therapy [159]. Microscopy examinations of biopsies [160] or peripheral blood mononuclear cells (PBMCs; [161] ) can be excellent tools for the detection and quantification of gH2A.X foci. H4S47ph has been associated with resistance to temozolomide (TMZ) treatment against glioblastoma [162].

Protein citrullination, the conversion of the amino acid arginine (Arg) to the non-conventional amino acid citrulline, can also take place on histones. This epigenetic mechanism includes the deimination of arginine’s side chain and is mediated by a specific enzyme called peptidylarginine deiminases (PADIs), a family that includes PAD1, PAD2, PAD3, PAD4, and PAD6. Besides, no enzymes have been found yet that converts citrulline back to arginine.

[enlarge]

Figure 12. The chemical conversion (deimination) of arginine to citrulline. From Wikipedia.

PAD4 is the main PADI (PAD2 to a lesser extent [5] ) responsible for histone H3 and H4 citrullination (H3cit/H4cit) and have the capacity to citrullinate unmodified as well as monomethylated arginines. Citrullination of H3R2, H3R8, H3R17, H3R26, H2AR3 and H4R3 has been shown to inhibit further methylation of these arginines thus altering the transcriptional activity at target genes [163, 164]. While arginine is positively charged, citrulline has no net charge thus increasing the hydrophobicity. This change from basic to neutral isoelectric point may alter the tertiary structure of or inhibit binding of histone “readers” at the protein domain [165]. Knockdown of PAD2 results in the decrease of H3R26cit and reverses the chromatin state from closed to open locally at ER-regulated genes [166].

Detection of citrullinated arginines can be achieved by anti-citrullinated antibodies and by mass spectrometry. In the former case, commercial antibodies facilitate detection of citrullinated proteins [167]. Likewise, fluorescent chemical probes have been developed for use in western blots and immunoassays [168, 169]. However, the currently available antibodies are characterised by lack of sensitivity and specificity for broad citrullination events as they may recognise epitopes that are dependent on the surrounding amino acids of the citrulline [170].

In proteomics strategies, the conversion of arginine to citrulline causes a mass shift of +0.98 Da which, albeit small, can be detected by state-of-the-art mass spectrometers. Besides, trypsin (the classical enzyme used in bottom-up proteomics, which normally digests the peptide bonds of polypeptides right after the positively charged arginines and lysines) can offer a significant advantage as it doesn’t cleave after uncharged citrulline. This, together with the mass shift will change the time of elution in HPLC columns and facilitate detection of deimination events [171]. Other methods that biotinylate the citrullinated residue have been successful in enriching the detection of citrullinated peptides prior to mass spectrometry [172, 173]. However, as with most histone PTM identifications, the richness of histones in lysines and arginines is a bottleneck for most proteomics approaches [170]. A recent study has reported new histone citrullination events, H2AR88, H3R42, H3R116, H4R23, H4R55, H4R92, [174] which await validation and their functions to be uncovered.

Detection of increased protein citrullination by antibodies is an established way for detection of rheumatoid arthritis (RA). More recent studies have implicated citrullination of histones with other pathologies and is associated with the formation of neutrophil and macrophage extracellular traps (NETs and METs) [175, 176]. H3cit, for example, is increased upon stimulation with various agents, including LPS, TNF, etc and is found elevated in plasma of patients with advanced cancer [177] and acute kidney injury [178] while pan-PAD inhibitors suppressed miRNA expression affecting oncogene expression involved in prolactinoma and somatoprolactinoma tumors [179]. H4cit is increased in obese mice models of acute pancreatitis [180].

Ultrasonic cell disrupter (200W, 4 min for twice) or mechanical microfluidics. Add some glycerin in the buffer, absolutely no SDS in the buffer.

If it is the whole cell lysates, then the regular cytoplasmic controls such as beta-actin can be used. If it is the nuclear fraction, then the nuclear controls such as lamin B1 or PCNA can be used. If it is the purified histone fraction, then other members of the histone family, such as H2A can be used.

One of the important considerations is that histone proteins have low molecular weights and can easily be transferred through the membrane during Western blotting. Therefore, it's important to reduce the transfer time, and also make sure the markers with similar molecular weights show up on the membrane.

Dr. Nikolaos Parisis contributed the sections about acetylation, methylation, phosphorylation, citrullination of histones in 2018 and contributed the section EpiTOF in October 2020.