The research, led by Gui-Ming Zhang, Le Yu and colleagues at Southern Medical University and collaborating institutions in China, took a deliberately indirect route. Rather than trying to block the oncogenic signal at its source, the team asked a different question: which genes does a cancer cell become fatally dependent on once GNAQ has been mutated? This concept, known as synthetic lethality, exploits the idea that a mutation that is harmless on its own can become lethal when a second, compensating function is removed. In practice, it means that a drug targeting the partner vulnerability can kill cancer cells while sparing normal tissue, which lacks the initiating mutation.

To find such vulnerabilities systematically, the researchers used an isogenic CRISPR screening approach. They engineered pairs of cell lines that were genetically identical except for the presence or absence of oncogenic GNAQ signaling, and then used CRISPR-based gene knockout across the genome to see which losses were selectively catastrophic only in the mutant cells. This design filters out genes that are generally essential for all cells and isolates those whose importance is specifically created by the cancer-driving mutation. Among the top-ranked hits, one gene stood out: EP300, which encodes p300, a protein lysine acetyltransferase that decorates histones and other proteins with acetyl groups to regulate gene expression.

Notably, the screen also flagged CREBBP, the gene encoding CBP, a closely related paralogue of p300 that shares much of its catalytic machinery and often partners with it at regulatory DNA. The convergence of both paralogues near the top of the candidate list suggested that the dependency was not a screening artifact but a genuine, redundant axis that GNAQ-mutant cells rely upon. Because p300 and CBP sit at the interface between signaling pathways and the epigenetic machinery that executes gene expression programs, their identification pointed the investigators toward a specific mechanism worth dissecting in detail.

That mechanism turned out to involve YAP, the transcriptional co-activator that serves as the main effector of the Hippo signaling pathway. When the Hippo pathway is switched off, YAP accumulates in the nucleus and teams up with DNA-binding transcription factors to activate genes promoting cell growth, survival and migration. Oncogenic Gαq is known to feed into this pathway, driving unrestrained YAP activity in uveal melanoma. The new work shows that p300 and CBP are not passive bystanders in this process: they physically interact with YAP at chromatin, the packaged form of DNA inside the nucleus, and this interaction is what allows the YAP program to be fully executed.

At the chromatin level, the researchers found that p300 and CBP enhance the acetylation of histone H3 at lysine 27, a chemical mark abbreviated H3K27ac that flags active promoters and putative enhancers, the regulatory regions that boost gene transcription. In GNAQ-mutant cells, this acetylation is concentrated at the regulatory elements of YAP target genes. The histone mark, in turn, promotes the recruitment of RNA polymerase II, the enzyme that reads genes into messenger RNA, together with CDK9, a kinase that helps release the paused polymerase into productive elongation. In other words, p300 and CBP act as the chromatin-level engine that converts YAP’s presence at DNA into actual gene output, and without them the oncogenic transcriptional program stalls.

The functional consequences of this dependency were tested rigorously in animal models. The team employed multiple in vivo models representing clinically relevant stages of Gαq-mutant uveal melanoma, from primary tumor growth to metastatic spread, the stage that matters most for patient survival. Both genetic approaches, which remove p300 function directly, and pharmacological inhibitors of its acetyltransferase activity profoundly impaired tumor growth and metastasis in these models. The consistency between genetic and drug-based suppression strengthens the case that the effect is genuinely mediated by p300’s enzymatic function rather than some scaffolding role or off-target toxicity, and it provides a direct preclinical rationale for moving toward therapeutic inhibition.

The significance of the finding lies in its strategic positioning. Because the dependency is created by the oncogenic Gαq pathway itself, p300 inhibition represents a form of indirect targeting: the drug does not need to shut down the mutated G protein, which has proved chemically difficult, but instead blocks an essential downstream enabler of the cancer’s transcriptional addiction. Epigenetic enzymes such as p300 and CBP are increasingly considered druggable, and small-molecule acetyltransferase inhibitors are an active area of medicinal chemistry. The study thereby converts a hard-to-drug oncogene problem into a potentially tractable epigenetic one, a reframing that could reshape how the field approaches GNAQ- and GNA11-driven cancers more broadly.

There are, of course, caveats that temper immediate clinical translation. p300 and CBP perform essential functions in many normal cell types, including roles in hematopoiesis and general transcriptional regulation, so therapeutic windows will need to be carefully defined, and the authors note that their work establishes p300 as a potential therapeutic target rather than a validated drug. The pharmacological results come from preclinical models, and the behavior of human metastatic disease in patients may introduce additional dependencies and resistances not captured in the laboratory. Nonetheless, the synthetic lethal logic demonstrated here, in which the very mutation that drives the cancer also creates its Achilles’ heel, is exactly the kind of therapeutic principle that has succeeded in other hard-to-treat tumors, such as PARP inhibition in BRCA-mutant breast and ovarian cancers.

For patients with metastatic uveal melanoma, whose median survival has remained dismal for decades, the study offers something the field has lacked: a mechanistically grounded, experimentally validated vulnerability that is specific to the molecular engine of their disease. By showing that GNAQ-mutant tumors depend on the p300/CBP chromatin machinery to sustain YAP-driven transcription, the researchers have charted a path from genome-wide screening to a defined mechanism to in vivo efficacy. The next steps, developing or deploying clinically suitable acetyltransferase inhibitors and testing them in appropriately designed trials, will determine whether this laboratory insight can be converted into a real therapeutic option for one of oncology’s most neglected cancers.

Subject of Research: Synthetic lethal identification of p300/CBP as a therapeutic vulnerability in GNAQ-mutant uveal melanoma

Article Title: Synthetic lethal screens reveal p300/CBP as a vulnerability for GNAQ-driven uveal melanomas

Article References: Zhang, G.-M., Yuan, J., Chen, L., Jin, X.-Y., Ren, Z.-L., Zhou, Y.-L., Zhou, R., Xiao, Y.-F., Wang, J.-W., Zhang, X.-A., Bai, C., Li, Z., Li, Y.-L., & Yu, L. (2026). Synthetic lethal screens reveal p300/CBP as a vulnerability for GNAQ-driven uveal melanomas. Nature Communications. https://doi.org/10.1038/s41467-026-77999-8

Image Credits: AI Generated

DOI: 10.1038/s41467-026-77999-8

Keywords: uveal melanoma, GNAQ, p300, CBP, YAP, synthetic lethality, CRISPR screen, epigenetics, H3K27 acetylation, Hippo pathway, targeted therapy, eye cancer

The story begins with lactate, a molecule long dismissed as little more than a metabolic waste product. Cancer cells famously consume enormous amounts of glucose and ferment it into lactate even in the presence of oxygen, a phenomenon known as aerobic glycolysis or the Warburg effect. In 2019, researchers discovered that lactate is not merely discarded; it can be chemically attached to lysine amino acids on proteins, including the histones around which DNA is wound. This modification, called lactylation, can loosen chromatin and activate gene expression, effectively allowing a metabolic byproduct to rewrite the instructions a cell follows. Since that discovery, scientists have raced to determine which lactylation marks matter most in which cancers, and whether blocking them could stall tumor progression.

In the new study, a research team led by Chuang Fu and Zhi-Zhou Shi of Kunming University of Science and Technology focused on a specific mark: lactylation at lysine 18 of histone H3, abbreviated H3K18la, in head and neck squamous cell carcinoma, or HNSCC. The team first showed that hypoxia, the oxygen-starved conditions typical of solid tumors, increased both H3K18la levels and the abundance of EGFR, the epidermal growth factor receptor that is a major therapeutic target in this disease. Single-cell RNA sequencing analysis of patient-derived data revealed a high-lactylation state within HNSCC cells, confirming that the modification is a prominent feature of the tumor landscape rather than a laboratory artifact.

To test whether this mark actually drives malignant behavior, the researchers manipulated lactylation pharmacologically. When they treated HNSCC cells with 2-deoxyglucose, a compound that blocks glycolysis, or with oxamate, an inhibitor of lactate dehydrogenase that prevents lactate production, the cells proliferated more slowly and formed fewer colonies. Critically, adding lactate back into the culture medium reversed these effects, restoring the cells’ growth advantage. This gain-and-loss evidence established that lactate availability, funneled into histone lactylation, directly supports the proliferative engine of the cancer cells, rather than being an incidental byproduct of their altered metabolism.

The next challenge was identifying which genes the mark actually controls. The team combined RNA sequencing, which measures gene expression, with CUT&Tag, a technique that maps where specific histone modifications sit along the genome. The integrated analysis revealed that H3K18 lactylation was strongly enriched at the locus encoding GLI2, a transcription factor best known as a downstream effector of the Hedgehog signaling pathway. The lactylation mark at the GLI2 locus was associated with transcriptional activation, meaning the epigenetic tag was literally switching on a potent pro-growth gene. Functional experiments confirmed the point: knocking down or inhibiting GLI2 suppressed HNSCC cell proliferation and colony formation, while overexpressing GLI2 promoted those same malignant phenotypes.

Having identified the mark and one of its key targets, the researchers asked which enzyme writes it. Attention turned to CBP, a well-known histone acetyltransferase that had previously been implicated as a lactyltransferase in other contexts. Molecular docking simulations predicted a binding interaction between CBP and the lactylated H3K18 mark, and co-immunoprecipitation experiments confirmed that the two proteins physically associate. When the team knocked down CBP, or treated cells with the CBP inhibitor SGC-CBP30 or the compound C646, H3K18la levels fell, GLI2 expression dropped, and HNSCC cells lost much of their capacity to proliferate and form colonies. In short, CBP emerged as the writer of the lactylation mark that sustains the GLI2 program.

The most clinically consequential part of the study concerns EGFR-targeted therapy. Drugs such as gefitinib and cetuximab, which block EGFR signaling, are standard options for subsets of HNSCC patients, but resistance frequently develops and limits their long-term benefit. The researchers found that EGFR itself sits upstream of the lactylation machinery: knocking down EGFR, or treating cells with gefitinib or cetuximab, reduced global protein lactylation, lowered H3K18la, and suppressed GLI2 expression. EGFR inhibition also decreased intracellular lactate concentrations and reduced glucose uptake, suggesting that EGFR signaling feeds the glycolytic flux that generates the lactate substrate for histone modification. Mechanistically, the team showed that EGFR binds to and stabilizes GLUT3, a glucose transporter, thereby helping the cancer cells maintain their sugar supply and, by extension, their lactylation program.

This wiring suggested a strategy: if the lactylation-GLI2 axis helps tumors survive EGFR blockade, then disabling that axis should resensitize them to the drugs. That is precisely what the experiments showed. Inhibitors acting on the H3K18la-GLI2 axis made HNSCC cells more vulnerable to both gefitinib and cetuximab in culture. More strikingly, in mouse xenograft models derived from CAL27 HNSCC cells, combining the CBP inhibitor SGC-CBP30 with gefitinib greatly improved the anti-tumor effects of the EGFR drug compared with either agent alone. The team also established gefitinib-resistant HNSCC cell lines and found that the H3K18la-GLI2 axis was activated in the resistant cells, and that blocking the pathway overcame that resistance. Together, these results position histone lactylation not just as a marker of aggressive disease but as an actionable driver of therapeutic failure.

The implications extend beyond head and neck cancer. Histone lactylation has now been implicated in multiple tumor types and in inflammatory and fibrotic diseases, and the identification of a defined writer-enzyme, mark, and target-gene cascade gives researchers a concrete axis to interrogate in other settings where EGFR inhibitors are used, including lung cancer. The study also highlights the value of pairing metabolic interventions with epigenetic drugs: because the lactylation mark depends on glycolysis-derived lactate, combinations that attack both the metabolic supply line and the epigenetic reader-writer machinery may prove more effective than either approach alone. Compounds such as 2-deoxyglucose and SGC-CBP30 are experimental tools rather than approved therapies, so translating these findings into the clinic will require careful dose-finding, safety evaluation, and biomarker development.

For patients with HNSCC, whose treatment options narrow sharply once EGFR-targeted agents stop working, the research offers a mechanistic explanation for resistance and a rational combination strategy to test in future trials. The work, funded by programs in Yunnan Province, China, and conducted under approved animal ethics protocols, exemplifies a broader shift in cancer biology: the recognition that the chemical state of chromatin is not fixed but is continuously shaped by the metabolic environment of the tumor. As lactylation research matures, the dream of turning a cancer cell’s own waste product against it, by blocking the enzymes that convert that waste into a growth-promoting epigenetic signal, moves one step closer to reality.

Subject of Research: Histone H3K18 lactylation in head and neck squamous cell carcinoma progression and EGFR-targeted therapy resistance

Article Title: CBP-mediated H3K18 lactylation drives GLI2 transcription to promote progression and EGFR-targeted therapy resistance in head and neck squamous cell carcinoma

Article References: Fu, C., Zhang, X.-G., Li, W.-T., Jiang, W., Fan, S.-C., & Shi, Z.-Z. (2026). CBP-mediated H3K18 lactylation drives GLI2 transcription to promote progression and EGFR-targeted therapy resistance in head and neck squamous cell carcinoma. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-09049-2

Image Credits: AI Generated

DOI: 10.1186/s12967-026-09049-2

Keywords: histone lactylation, H3K18la, head and neck squamous cell carcinoma, EGFR, GLI2, CBP, gefitinib resistance, lactate, GLUT3, epigenetics, targeted therapy, cancer metabolism