Histone Modification

Histone modifications are reversible, covalent post-translational modifications (PTMs) deposited predominantly along the flexible, positively charged N-terminal (and C-terminal) polypeptide tails protruding from the octameric histone core (two copies each of H2A, H2B, H3, and H4) of the eukaryotic nucleosome. Because 147 base pairs of negatively charged DNA wrap 1.65 times in a left-handed superhelix around each histone octamer, chemical modifications of specific amino acid side chains—most notably lysine ε-amino acetylation, methylation, ubiquitination, and SUMOylation; arginine methylation and citrullination; and serine / threonine phosphorylation—govern epigenetic chromatin states through two synergistic mechanisms:

1. Direct Biophysical Charge & Steric Remodeling (cis-mechanism): Neutralizing positive lysine charges (as in histone acetylation) weakens electrostatic attraction between histone tails and the negatively charged DNA phosphodiester backbone, loosening internucleosomal 30-nm fiber packing to create accessible euchromatin.
2. The "Histone Code" Effector Recruitment Platform (trans-mechanism): Formulated by Brian D. Strahl and C. David Allis (2000) (building on Vincent Allfrey's 1964 discovery of histone acetylation and Michael Grunstein's yeast genetic proofs), combinatorial patterns of covalent marks are deposited by enzymatic "writers", removed by "erasers", and recognized stereospecifically by specialized protein domains ("readers") that recruit ATP-dependent chromatin remodelers (SWI/SNF), RNA polymerase II, or transcriptional condensates.

Lysine Acetylation vs. Lysine Methylation Biophysics

The two most extensively characterized classes of histone modifications—lysine acetylation and lysine methylation—exert fundamentally different electrostatic and structural effects on the ε-amino group (–NH₃⁺, p*K*_a ≈ 10.5) of lysine residues:

• Lysine Acetylation (Kac — Charge Neutralization): Histone acetyltransferases (HATs / KATs, "writers" such as GCN5/KAT2A, PCAF, p300/CBP, and MYST-family TIP60) transfer an acetyl group (–COCH₃) from acetyl-CoA onto the ε-NH₃⁺ group of lysine, converting the cationic ammonium into a neutral amide (net charge +1 → 0). This directly disrupts salt bridges with DNA phosphates and internucleosomal H4-tail-to-H2A/H2B acidic-patch contacts, universally correlating with open euchromatin and active transcription. Histone deacetylases (HDACs, "erasers"—comprising Zn²⁺-dependent Class I, II, and IV HDAC1–11 plus NAD⁺-dependent Class III sirtuins SIRT1–SIRT7) hydrolyze the acetyl group to restore the +1 charge and condense chromatin. Acetylated lysines are read specifically by bromodomains (such as the tandem bromodomains of BRD4, targeted clinically by BET inhibitors like JQ1) and YEATS domains.
• Lysine Methylation (Kme1, Kme2, Kme3 — Charge-Preserving Steric Coding): Histone lysine methyltransferases (KMTs, "writers"—almost all containing a catalytic SET domain except the 7β-strand enzyme DOT1L for H3K79) transfer one (me1), two (me2), or three (me3) methyl groups (–CH₃) from the universal methyl donor S-adenosyl-L-methionine (SAM) onto the lysine ε-nitrogen. Crucially, methylation preserves the +1 positive charge of the lysine side chain at all three states (me1, me2, me3) while progressively increasing hydrophobicity and steric bulk. Consequently, lysine methylation does not unfold chromatin electrostatically; instead, its biological outcome (active transcription vs. deep heterochromatic silencing) depends entirely on which exact residue position is methylated and which "reader" domain (chromodomain, PHD finger, Tudor domain, PWWP, or MBT domain) binds it. Methyl groups are removed by two families of "erasers": flavin-dependent LSD1 / KDM1A (which demethylates only Kme1 and Kme2 via an imine intermediate requiring a protonated nitrogen) and Fe²⁺ / 2-oxoglutarate-dependent Jumonji C (JmjC) dioxygenases (KDM2–KDM7), which can demethylate trimethylated lysines (Kme3) via radical hydroxylation.

Embryonic "Bivalent Domains" (H3K4me3 + H3K27me3)

In 2006, bradley Bernstein, Stuart Orkin, and Eric Lander discovered that in pluripotent embryonic stem cells (ESCs), the promoters of hundreds of master developmental transcription factor genes (including the Hox gene clusters, SOX, PAX, and GATA families) simultaneously carry both the activating mark H3K4me3 (deposited by Trithorax/MLL2) and the repressive mark H3K27me3 (deposited by Polycomb PRC2) on the same nucleosome (often on opposite H3 tails of the same octamer, termed asymmetrically modified bivalent nucleosomes).

These bivalent domains keep lineage-specifying developmental genes transcriptionally silenced yet "poised" for rapid activation: when an embryonic stem cell differentiates into a neural progenitor, the UTX/JMJD3 demethylase erases repressive H3K27me3 at neural promoters (resolving them to active H3K4me3 monovalent euchromatin), while JARID1/KDM5 erases H3K4me3 at mesodermal and endodermal promoters (locking them into permanent H3K27me3 Polycomb repression).

Metabolic Control of the Epigenome (Acetyl-CoA, SAM, NAD⁺, and Oncometabolite 2-HG)

Because every chromatin-modifying enzyme requires a central cellular metabolite as its co-substrate or cofactor, histone modifications serve as a direct metabolic sensor linking nutrient availability and mitochondrial citric acid cycle flux to gene expression:

1. Acetyl-CoA and NAD⁺ (HATs vs. Sirtuins): Nuclear ATP-citrate lyase (ACLY) and pyruvate dehydrogenase convert glucose-derived citrate/pyruvate into nuclear acetyl-CoA; high glycolytic flux elevates nuclear [acetyl-CoA] within the physiological K_m range of p300/GCN5 (~1–10 μM), driving global histone hyperacetylation and growth-gene transcription. Conversely, fasting or caloric restriction elevates the NAD⁺/NADH ratio, activating SIRT1 and SIRT6 deacetylases (which deacetylate H4K16ac and H3K9ac while generating O-acetyl-ADP-ribose).
2. SAM/SAH Ratio (Methyltransferases): Folate and methionine one-carbon metabolism sets the ratio of S-adenosylmethionine (SAM) to the product inhibitor S-adenosylhomocysteine (SAH), governing H3K4me3 peak breadth.
3. 2-Oxoglutarate (α-Ketoglutarate) vs. Oncometabolites (2-HG, Succinate, Fumarate): All Jumonji C (JmjC) histone demethylases (and TET DNA 5mC hydroxylases) require the TCA cycle intermediate 2-oxoglutarate (α-KG) and Fe²⁺. In human gliomas and acute myeloid leukemias (AML) carrying somatic neomorphic mutations in isocitrate dehydrogenase (IDH1 R132H or IDH2 R140Q/R172K), the mutant enzyme reduces α-KG into millimolar concentrations of the oncometabolite D-2-hydroxyglutarate (D-2-HG) (similarly, loss of succinate dehydrogenase SDH or fumarate hydratase FH accumulates succinate or fumarate). Because D-2-HG is a close structural mimic of α-KG, it competitively inhibits JmjC histone demethylases (KDM4/KDM6) and TET2, locking progenitor cells in a hypermethylated state (↑ H3K9me3, ↑ H3K27me3) that blocks terminal differentiation (now treated clinically with the FDA-approved mutant IDH inhibitors ivosidenib, enasidenib, and vorasidenib).

Although humans encode 15 different Histone H3 genes (30 alleles per diploid genome), sequencing of pediatric brain and bone tumors by Nada Jabado and colleagues (2012) revealed that a single heterozygous point mutation in just one of the 30 Histone H3 alleles (most frequently H3F3A encoding variant H3.3, accounting for only ~5–15% of total cellular H3 protein) wipes out a specific histone modification across the entire genome via dominant-negative enzyme poisoning:

• H3K27M (Lys27 → Met) in Diffuse Midline Glioma (DIPG): Substitution of lysine 27 with methionine creates a hydrophobic side chain that binds the aromatic active-site pocket of the PRC2 methyltransferase catalytic subunit EZH2 (specifically interacting with Tyr641 in the presence of SAM) with ~16-fold higher affinity than wild-type H3, sequestering and poisoning the entire cellular pool of PRC2 and erasing global H3K27me3 across the remaining 90% wild-type H3 histones.
• H3K36M (in Chondroblastoma) and H3.3G34R/V/W (in Hemispheric Glioma & Giant Cell Tumor of Bone): H3K36M similarly poisons the SETD2 and NSD2 methyltransferases in trans (erasing global H3K36me2/me3), whereas H3G34 mutations sterically block SETD2 and KDM2A in cis on the mutated H3.3 tail itself.

How can histone modifications be inherited across mitotic cell division when nucleosomes are disrupted by the replication fork?

During S-phase DNA replication, the advancing replisome (CMG helicase and histone chaperones MCM2, Pole3/4, and ASF1) disassembles parental (H3–H4)₂ tetramers ahead of the fork and recycles them intact—In a 1:1 symmetric split between the leading and lagging daughter strands (mediated by MCM2–Ctf4–Polα on the lagging strand and Pole3/4 on the leading strand)—while newly synthesized, unmethylated H3–H4 tetramers (deposited by CAF-1) fill in the remaining 50% of nucleosome positions. To prevent repressive marks from diluting by half at every cell division, repressive histone methyltransferases use a "read–write" positive feedback mechanism:

• For H3K27me3, the EED subunit of PRC2 binds an existing parental H3K27me3 mark via its aromatic cage, which allosterically stimulates the catalytic EZH2 subunit by 5- to 10-fold to trimethylate Lys27 on the adjacent newly deposited naive H3–H4 nucleosome.
• For H3K9me3, HP1 and the chromodomain of SUV39H1 read parental H3K9me3 and write H3K9me3 onto neighboring naive nucleosomes. In contrast, narrow active marks like H3K4me3 rely heavily on sequence-specific transcription factors and unmethylated CpG islands (CFP1) to re-establish after mitosis.

How does H3K36me3 protect actively transcribed exons from mutations?

Coding exons in the human genome accumulate significantly fewer somatic mutations than intergenic DNA or untranscribed introns, even before purifying selection acts. This protection is mediated directly by H3K36me3: as RNA Polymerase II elongates through an active gene body, its phosphorylated C-terminal domain (Ser2ph) carries the methyltransferase SETD2, which coats transcribed exons with H3K36me3. The PWWP reader domain of MSH6 (a core subunit of the MutSα DNA mismatch repair complex) binds constitutively to H3K36me3, pre-loading mismatch repair machinery directly onto active exons so polymerase replication errors and oxidative lesions are repaired preferentially inside protein-coding sequences.

What is the difference between canonical replication-dependent histones (H3.1/H3.2) and histone variants (H3.3, CENP-A, H2A.Z)?

Canonical histones (H3.1, H3.2, H4, H2A, H2B) are encoded by intronless multicopy gene clusters, lack poly(A) tails (ending instead in a conserved 3' stem-loop), and are synthesized and deposited by CAF-1 strictly during S-phase DNA replication. In contrast, replacement histone variants are encoded by solitary, intron-containing polyadenylated genes expressed throughout the cell cycle (replication-independent):

• Histone H3.3 (H3F3A/B, differing from H3.1 by only 4 or 5 amino acids) is deposited into actively transcribed promoters and enhancers by the chaperone HIRA (replacing nucleosomes evicted by RNA Pol II) and into telomeres and pericentric repeats by ATRX–DAXX.
• CENP-A (Centromere Protein A) replaces H3 specifically at centromeres to epigenetically specify the kinetochore assembly site for chromosome segregation.
• H2A.X (~10% of H2A) carries a C-terminal SQEY motif phosphorylated at Ser139 (γ-H2AX) at DNA double-strand breaks.

1. Strahl, B. D., & Allis, C. D. (2000). "The language of covalent histone modifications." Nature, 403(6765), 41–45. https://doi.org/10.1038/47412
2. Bernstein, B. E., Mikkelsen, T. S., Xie, X., et al. (2006). "A Bivalent Chromatin Structure Marks Key Developmental Genes in Embryonic Stem Cells." Cell, 125(2), 315–326. https://doi.org/10.1016/j.cell.2006.02.041
3. Kouzarides, T. (2007). "Chromatin Modifications and Their Function." Cell, 128(4), 693–705. https://doi.org/10.1016/j.cell.2007.02.005
4. Millán-Zambrano, G., Burton, A., Bannister, A. J., & Schneider, R. (2022). "Histone post-translational modifications — cause and consequence of genome function." Nature Reviews Genetics, 23(9), 563–580. https://doi.org/10.1038/s41576-022-00468-7