Brain cancers are the leading cause of cancer-related mortality in children and young people. The lack of effective treatments for paediatric gliomas means these cancers are invariably fatal with a median survival of only 12 months. Sequencing of paediatric gliomas has identified two common substitution mutations (K27M and G34R) in genes encoding histone H3.3. It is not yet clear how these mutations promote oncogenesis but early studies indicate that point mutations in this histone variant are sufficient to trigger global chromatin alterations. For example, the substitution of methionine for lysine at position 27 (H3.3 K27M) acts by recruiting and inhibiting PRC2, a histone lysine methyltransferase which mediates trimethylation of H3K27 (H3K27me3). Expression of a mutant K27M transgene triggers an overall loss of H3K27me3 and aberrant distribution of this silencing modification across the genome.

We report that the H3.3 G34R mutation (glycine to arginine substitution) acts in an analogous manner by inhibiting KDM4, a histone lysine (K9/K36) demethylase. We introduced a single-copy H3.3 G34R targeted mutation in mouse ES cells to recapitulate this early oncogenic event in the absence of confounding background mutations. The expression of the H3.3 G34R mutant at endogenous levels was sufficient to increase H3K36me3 and H3K9me3 across the genome. We found that H3.3 G34R inhibits the enzymatic activity of KDM4 by blocking access to the catalytic domain, and preventing normal demethylation of H3K36me3 and H3K9me3.

This illustrates that aberrant chromatin is a key step in cancer formation and that histone point mutations can exert a whole-genome effect through inhibition of histone lysine demethylases. We propose a general principle where oncogenic histone mutations can disrupt the epigenetic profile of cancer cells through inhibition of both epigenetic writers and erasers.