Epigenetics in Drug Discovery

Epigenetics in Drug Discovery

1. Definition, scope & why epigenetics is a drug-discovery frontier

• Epigenetics = heritable-yet-reversible changes in gene expression that do not alter the underlying DNA base sequence — mediated by DNA methylation, covalent histone modifications, chromatin (nucleosome) remodelling, and non-coding RNAs.
• This topic is scoped to epigenetics as a source of druggable targets and of approved drugs, and to pharmacoepigenomics — explicitly distinct from classical pharmacogenetics/pharmacogenomics (variation in the germline DNA sequence), which is the subject of G&G Ch.7 and is treated here only as the contrasting framework.
• G&G's opening framing of modern drug discovery is directly applicable: today most drugs "are not discovered, but are totally new compounds, painstakingly optimized against many criteria" — i.e. invented against a defined molecular target (G&G 14e Ch.1, p.4). Epigenetic enzymes are exactly this kind of rationally selected molecular target.
• The drug-discovery paradigm most small-molecule projects follow — "basic research that implicates a specific macromolecule, usually a protein, as a key player in a disease… [which] becomes a candidate drug target" (G&G 14e Ch.1, p.5) — is the paradigm by which chromatin-modifying enzymes (writers/erasers) and chromatin-reader proteins entered the pharmacopoeia.
• Therapeutic rationale unique to epigenetics — reversibility. Because epigenetic marks are enzymatically written and erased (unlike a fixed mutation), they are in principle pharmacologically reversible targets; a small molecule can restore a dysregulated expression state rather than having to correct a genetic lesion.
• Cancer is the beach-head indication. G&G Ch.7 establishes the precedent that "variants in the tumor (somatic) genome have emerged as a critical determinant of anticancer drug effects" and that tumour sequencing to choose targeted agents "is becoming standard of care" (G&G 14e Ch.7, p.139). Every first-generation approved epigenetic drug is an oncology drug, and the two mutation-selective agents in this topic (IDH1/2, EZH2) are prescribed exactly on the somatic-mutation-directed model G&G describes for EGFR/HER2/BRAF.

The "writers / readers / erasers" druggable-target framework

• The epigenetic machinery is conventionally partitioned into three functional classes of druggable protein:
  • Writers — enzymes that deposit a chemical mark: DNA methyltransferases (DNMTs) add methyl to cytosine; histone acetyltransferases (HATs) add acetyl to lysine; histone methyltransferases (e.g. EZH2, the catalytic subunit of PRC2) add methyl to lysine/arginine.
  • Erasers — enzymes that remove a mark: histone deacetylases (HDACs) remove acetyl; histone/lysine demethylases (KDMs, e.g. LSD1) remove methyl.
  • Readers — modules that recognise/bind a mark and translate it into a transcriptional outcome: bromodomains read acetyl-lysine (the BET proteins BRD2/3/4/T being the archetype), chromodomains and PHD fingers read methyl-lysine.
• All three classes are, in G&G's terms, enzymes or small-ligand-binding modules with defined pockets — precisely the protein architectures G&G calls most druggable: "Druggable binding pockets… usually are available in enzymes whose substrates are small molecules and in receptors that bind small-molecule hormones and transmitters" (G&G 14e Ch.1, p.6). DNMTs, HDACs and HMTs use small cofactors (SAM, acetyl-CoA, Zn²⁺); bromodomains have a defined acetyl-lysine pocket — hence their tractability to small molecules.
• Direct textual anchor: G&G Ch.1's own reference list cites structure-based work on "the first bromodomain of BRD4" (G&G 14e Ch.1, reference list, p.20) — the BET reader domain that anchors the entire BET-inhibitor pipeline discussed below.