Allis lab Science paper on histone H3.3 in pediatric brain cancer

It has been fascinating in the last year to see the unfolding story of mutations in histone variant H3.3 in pediatric glioblastoma and related tumors.

It seems amazing that in just over a year, a whole new area of cancer research has rapidly progressed.

As I discussed in a recent post, histone H3.3 is a fascinating molecule that plays critical roles in regulating cellular chromatin and epigenomic states. H3.3 is, as mentioned above, also frequently mutated in some of the worst childhood brain tumors.

Allis H3.3 GBM paper, histone H3.3
Allis lab paper on histone H3.3 in glioma.

A recent Science paper by David Allis’ lab, which has for many years done very exciting research on H3.3, greatly advances our understanding of how H3.3 mutations cause cancer.In the paper, Lewis et al., Allis’ team dug into the biochemistry of the mutant forms of H3.3, particularly a Lys27Met (K27M) mutation.This is an exceptionally good paper.The team found that a key mechanism by which mutation of H3.3 likely causes the brain tumors is through inhibiting the function of the Polycomb Repressive Complex 2 (PRC2) and reducing histone H3 K27 methylation (for example compare levels of H3K27me3 staining in tumors with WT versus K27M mutant H3.3 above in Figure 1C of the paper). This is quite interesting, even if not entirely unexpected because PRC2 normally targets K27 of histone H3 family members. Further, they reported that the K27M form of H3.3, even when greatly in the minority molecularly compared to wildtype H3.3 and canonical H3, had a dominant inhibitory function.

Exactly how K27M H3.3 acts (both in cis and trans) as a sink for PRC2 activity even when there is an apparent great molar excess of wildtype H3.3 or H3 around remains to be defined, but K27M inhibits the histone methyltransferase activity of PRC2 and specifically through the PRC2 component, EZH2.

One of the notable things about this paper from a technological perspective is that the team generated a very powerful K27M specific antibody in what must have been a relatively short period of time.

So why the mutation of K27 to M specifically in the tumors and not some other residue? In Figure 2, they tried mutation of K27 to every possible residue and found that only K27L (Leucine) gave similar results. If the mutation functioned simply to block lysine (K) trimethylation then it seems that almost any residue other than K would do. Instead the data support a model in which the K27 mutant is dominant and active. As the authors write:

our data indicate that K-to-M mutations target the active sites of diverse SET domain-containing methyltransferases, thereby effectively competing with substrate binding and turnover. 

The paper concludes:

We propose a model whereby aberrant epigenetic silencing through H3K27M-mediated inhibition of PRC2 activity promotes gliomagenesis. The broadly adaptable, yet highly specific, inhibition of SET-domain proteins through K-to-M mutation offers the intriguing possible existence of other etiological missense mutations in histones. Additionally, our work has uncovered a potentially useful mechanism to exclusively inhibit individual SET-domain methyltransferases, and conceivably other chromatin-modifying enzymes, implicated in a variety of malignancies.

Some interesting questions remain. Are there similar kinds of K-to-M histone mutations (not just in K27 of H3) in other kinds of tumors? What happens to the PRC2 pathway downstream due to the K27M mutation? Presumably there are changes in gene expression, but what are they and how do those cause cancer?

A brand new G&D paper from another lab digs more into the downstream mechanisms of these mutations as well including changes in cancer-associated gene expression.

The paper also touches on another mutation in H3.3 found in GBMs: H3.3G34R/V. While the authors make the case that this mutation at G34 functions by altering methylation of the nearby K36 by decreasing SET2 activity, one gets the feeling that there is more to the G34R/V story to be learned. For example, why isn’t the K36 residue itself simply mutated to M? One might also imagine that mutating G34 could change n-terminal structure more broadly and have other effects.

Another very recent paper on H3.3 asserts that H3.3G34R/V works through upregulation of another favorite molecule of my lab, the oncogene N-Myc.

There is sure to be much more to come from the Allis lab and others on this fascinating new area of cancer research.