A Single Amino Acid Swap in Histone H2B Makes Breast Cancer Spread to the Lungs

On October 12, 2026, a research team led by professor Chan Kui-ming at the City University of Hong Kong reported a result that reframes how we think about what makes cancer lethal. In a paper published in Science Advances, the team showed that H2BE113K, a missense mutation in the histone protein H2B, promotes lung metastasis in breast cancer. The surprising part: the mutation does nothing to the growth of the primary tumor. It simply makes the cancer more likely to spread, and spread is what kills.

The mutation swaps a single amino acid, glutamate for lysine, at position 113 of histone H2B, one of the protein spools that DNA wraps around. H2BE113K has been spotted in cancer genome data primarily in breast and lung tumors, but until now nobody had shown experimentally what it actually does. This study moves it from a statistical curiosity to a confirmed driver of metastasis, a bona fide oncohistone.

What the team actually did

The researchers started by engineering breast cancer cells to carry the H2BE113K mutation, using CRISPR-Cas9 to create a knock-in rather than an artificial overexpression system. That detail matters. A knock-in means the mutation sits at its natural location in the genome and is produced at natural levels, which makes the results far harder to dismiss as a lab artifact.

The first results were, on the surface, a letdown. Cells carrying H2BE113K divided at the same rate as unmodified cells. They did not migrate faster in migration assays either. If the team had stopped there, the mutation would have looked like a passenger, one of the thousands of harmless changes that accumulate in cancer genomes.

But colony formation assays told a different story. H2BE113K cells were significantly better at forming colonies from sparse, isolated cells, a classic laboratory readout for the ability of a single cancer cell to establish a new outpost. Metastasis, after all, is exactly that: a lone cell or small cluster surviving in foreign tissue and rebuilding a tumor from scratch.

Two hundred and twenty-one genes and one culprit

To find the mechanism, the team ran RNA sequencing and CUT&RUN sequencing, a technique that maps where proteins bind DNA. They identified 221 candidate genes whose expression changed in the presence of the mutation, including RAD50 and G3BP2, both previously implicated in various cancers.

One gene, G3BP2, emerged as the key downstream player. The study found that the mutant histone recruits SMARCA5, a chromatin remodeling protein, to the G3BP2 promoter region. SMARCA5's presence switches G3BP2 expression up, and higher G3BP2 in turn enhances the colony-forming ability of the cells. The full chain reads like a short causal sentence: mutant histone recruits remodeler, remodeler turns up a gene, the gene makes cells better colonizers.

The mutation does not make the tumor grow faster. It makes the cancer better at leaving.

Growth and spread turn out to be separable

The team did not rely on cell culture alone. They built gene-edited mice and confirmed the same pattern in living animals: H2BE113K did not affect the development or growth of primary breast tumors, but it promoted lung metastasis. Lung tumors in the mutant mice showed higher levels of Ki67, a standard marker of cell proliferation, than those in control animals, suggesting that once the cells arrive in the lung, the mutation also helps them multiply in their new environment.

This separation between growth and spread is the conceptual center of the paper. For decades, the default assumption in cancer biology was that aggressive tumors grow fast and spread because they grow fast, two consequences of the same runaway engine. H2BE113K shows the programs can be uncoupled. A tumor can look unremarkable at its primary site while carrying a molecular upgrade that improves its odds of colonizing distant organs. Chan's group has been working this seam for years; the lab previously identified roles for the H2BG53D and H2BE76K mutations in pancreatic and breast cancer, part of a growing catalog of H2B oncohistones.

A drug target hiding in plain sight

The mechanism also hands researchers something they can aim a drug at. If SMARCA5 is the remodeler that the mutant histone exploits, then blocking SMARCA5 should blunt the effect. The team tested ED2-AD101, described as a first-in-class SMARCA5 inhibitor, and found that it effectively diminished the excessive colony formation seen in H2BE113K mutant cells.

That is promising but should be read with the usual caution. ED2-AD101 was tested in cell experiments, not in patients, and blocking a chromatin remodeler is a blunt instrument: SMARCA5 does essential work in healthy cells too. The finding is best understood as a proof of principle, that the pathway from mutant histone to metastatic behavior has a druggable node, and a starting point for precision approaches aimed at tumors carrying this specific mutation.

Where this fits in the oncohistone story

Oncohistones entered the scientific mainstream in the early 2010s, when mutations like H3K27M and H3G34R were tied to diffuse intrinsic pontine gliomas, aggressive childhood brain tumors. Since then, genome sequencing projects have cataloged thousands of missense mutations in histone genes, and functional studies keep finding that many converge on the same theme: perturbed chromatin regulation and aberrant gene expression. H2B mutations have been relatively understudied compared with their H3 cousins, which makes systematic work like this one, which the authors published under the title 'H2BE113K mutation promotes breast cancer metastasis through modulating chromatin dynamics,' particularly valuable.

For patients and clinicians, the near-term implications are diagnostic rather than therapeutic. If H2BE113K reliably marks tumors with higher metastatic potential, testing for it could one day help stratify risk and guide surveillance. That is still a long way off, and the mutation's frequency in patient populations, its interaction with existing therapies, and whether the SMARCA5 inhibitor path pans out in animals and eventually humans are all open questions. As Chan put it, the work confirms that a single amino acid substitution in histone H2B is sufficient to reshape chromatin regulation, and that confirmation opens the door to precision medicine. The door is open. Walking through it is the next decade's work.