SMYD3 is a histone methyltransferase, an enzyme that adds chemical tags to histone proteins around which DNA is wound, thereby influencing which genes are switched on. It was first characterized as a methyltransferase for histone H3 lysine 4 and H4 lysine 5, associated with RNA polymerase complexes and the activation of genes governing proliferation, cell cycle progression and epithelial-mesenchymal transition. But SMYD3 also methylates non-histone proteins, modulating signaling pathways involved in cell survival, stemness and self-renewal. Crucially, the enzyme is overexpressed in many tumor types, including colorectal, breast, pancreatic and lung cancers, hepatocellular carcinoma and gastric cancer, where its levels correlate with aggressive biological behavior and poor prognosis. The new study builds on the team’s earlier work showing that SMYD3 physically interacts with the DNA repair proteins ATM, CHK2 and BRCA2 and is required for the repair of double-strand breaks, the lesions in which both strands of the DNA helix are severed simultaneously.

To define the clinical landscape, the researchers first mined data from 407 stomach adenocarcinoma samples in The Cancer Genome Atlas Pan-Cancer Atlas cohort. Tumors were split into SMYD3-high and SMYD3-low groups based on mRNA levels, and their genomes were annotated for clinically actionable alterations using the OncoKB precision oncology database. The result was striking: SMYD3-high tumors showed significant mutual exclusivity with actionable alterations in PTEN, ATM and BRCA2, as well as with ERBB2 amplification, high microsatellite instability or tumor mutation burden, and MTAP deletion. In other words, tumors that overproduce SMYD3 rarely carry the genomic hallmarks that currently guide targeted therapy, and more than half of them lacked any actionable alteration at all. This positions SMYD3 overexpression as a defining feature of a patient subgroup that today has few precision options, and it suggests that SMYD3 itself could serve as both a biomarker and a drug target for this population.

The mechanistic core of the study relied on an ingenious experimental system. The team engineered AGS gastric cancer cells, which rank among the highest SMYD3-expressing gastric lines in the DepMap database, to carry a version of the AsiSI endonuclease fused to an estrogen receptor domain. Adding the hormone analog 4-hydroxytamoxifen shuttles the enzyme into the nucleus, where it cuts the genome at roughly 150 predetermined sites, generating a synchronized wave of double-strand breaks whose chromatin environment can be interrogated with base-pair precision. Using chromatin immunoprecipitation followed by quantitative PCR, the researchers showed that SMYD3 itself is recruited to homologous recombination-prone break sites, and that its recruitment is abolished by EM127, a potent covalent SMYD3 inhibitor developed through medicinal chemistry work published in 2022.

What happens at the break site when SMYD3 is active turned out to be a coordinated choreography of histone marks. Within hours of damage induction, the enzyme’s activity was required for the enrichment of H4K20me2, a dimethylation mark critical for the earliest recognition of DNA damage, and for H3K9me3, a trimethylation mark that in turn recruits the acetyltransferase TIP60 to deposit H4K16ac, another modification essential for repair. Blocking SMYD3 significantly reduced all three marks at damaged chromatin. Interestingly, H3K79me2, a mark that can compensate for H4K20me2 loss, rose when SMYD3 was inhibited, hinting that cells attempt a backup remodeling program when the primary pathway fails. Downstream of the chromatin changes, the recruitment of the entire homologous recombination machinery, including RAD50, ATM, CHK2, BRCA1, RPA32, BRCA2 and RAD51, was markedly impaired by SMYD3 inhibition, while the repair sensor 53BP1 was also reduced at a non-homologous end joining-prone break site. When the team monitored repair kinetics after treatment with the DNA-breaking agent neocarzinostatin, both pharmacological inhibition and genetic silencing of SMYD3 left cells littered with unresolved gamma-H2AX foci, the microscopic signatures of unrepaired breaks.

The study then identified the molecular switch that activates SMYD3 during repair. Bioinformatic prediction across four phosphorylation-site algorithms pointed to threonine 22, a surface-exposed residue, as the most likely ATM target. Mass spectrometry of recombinant SMYD3 phosphorylated by ATM in vitro confirmed T22 as the residue modified, and a custom antibody raised against the phospho-T22 epitope, validated by enzymatic dephosphorylation, revealed that the modification appears in AGS cells after DNA damage, disappears when ATM is blocked with KU60019 or SMYD3 is silenced, and, critically, is detectable in tumor samples from gastric cancer patients who had received neoadjuvant chemotherapy, which induces double-strand breaks and activates ATM in vivo. Co-immunoprecipitation experiments showed that phospho-T22-SMYD3 associates with phospho-ATM, CHK2, BRCA2 and RAD51, whereas a phospho-deficient T22A mutant failed to bind the repair complex altogether. Cells forced to rely on the T22A mutant proliferated less after DNA damage and showed elevated cleaved PARP, a hallmark of apoptosis, demonstrating that this single phosphorylation event is essential for both repair complex assembly and survival.

Functionally, SMYD3 inhibition proved to be a selective strike against homologous recombination. In the DR-GFP reporter assay, both siRNA knockdown and EM127 treatment reduced HR efficiency in AGS cells, while a luciferase-based assay showed only a partial impairment of non-homologous end joining. That asymmetry is the key to the therapeutic strategy: PARP inhibitors exploit exactly this kind of repair imbalance, and they work best in tumors whose homologous recombination is already defective, a state known as BRCAness. By pharmacologically imposing BRCAness on HR-proficient tumors, SMYD3 inhibition should widen the applicability of PARP inhibitors beyond the BRCA-mutated setting. The data bore this out. In high-SMYD3 AGS and NCI-N87 cells, combining EM127 with olaparib or with rucaparib, another clinically relevant PARP inhibitor currently in gastric cancer trials, dramatically reduced proliferation and increased cell death, confirmed by PARP cleavage, whereas low-SMYD3 KATOIII cells were largely unresponsive, underscoring SMYD3 expression as a predictive biomarker.

Perhaps the most clinically resonant result came from a model of acquired resistance. The researchers generated an olaparib-resistant derivative of the HGC-27 gastric cancer line by chronic exposure to escalating drug concentrations over six months. Resistant cells displayed significantly increased nuclear SMYD3 compared with the parental line, suggesting the enzyme helps mediate PARP inhibitor resistance. When SMYD3 was inhibited with EM127, or silenced genetically, olaparib sensitivity was restored, with PARP cleavage rising sharply. The combination also worked in three-dimensional settings that better mimic real tumors: gastric tumorspheres, which showed elevated SMYD3 alongside the stemness markers KLF4 and OCT4, were killed far more effectively by the dual treatment than by either drug alone, and the same held true for olaparib-resistant tumorspheres.

The final layer of evidence came directly from patients. Working with surgical specimens from gastric cancer patients at the IRCCS Saverio de Bellis Research Hospital in Castellana Grotte, the team used droplet digital PCR to stratify tumors by SMYD3 expression, finding elevated levels in roughly half of the samples analyzed, and grew matched normal and tumor organoids from the SMYD3-high cases. These patient-derived tumor organoids retained their tumor-specific molecular profile, with enrichment of the epithelial marker EpCAM and reduction of CDH1, reflecting loss of epithelial integrity and invasiveness. Treating them with EM127 plus olaparib or rucaparib produced marked cell death and PARP cleavage, mirroring the two-dimensional and tumorsphere results. Taken together, the findings delineate a complete arc from molecular mechanism to translational opportunity: ATM phosphorylates SMYD3 at threonine 22 in response to DNA damage, phosphorylated SMYD3 remodels chromatin at break sites to assemble the homologous recombination machinery, and blocking this cascade sensitizes SMYD3-high gastric cancers, including PARP inhibitor-resistant ones, to synthetic lethality. If the strategy advances toward clinical testing, SMYD3 expression could guide patient selection, extending the benefits of PARP inhibition to a subgroup of gastric cancer patients who currently have no targeted option.

Subject of Research: SMYD3-dependent chromatin remodeling in DNA double-strand break repair and its targeting for epigenetics-based therapy of gastric cancer

Article Title: Targeting the DNA-repair chromatin-modifier protein SMYD3 as a novel epigenetics-based therapy for gastric cancer

Article References: De Marco, K., Latrofa, M., Forte, G., Lepore Signorile, M., Di Nicola, E., Sanese, P., Fasano, C., Disciglio, V., Candela, E., Coletta, S., Grossi, V., & Simone, C. (2026). Targeting the DNA-repair chromatin-modifier protein SMYD3 as a novel epigenetics-based therapy for gastric cancer. Journal of Experimental & Clinical Cancer Research, 45(1), Article 202. https://doi.org/10.1186/s13046-026-03823-2

Image Credits: AI Generated

DOI: 10.1186/s13046-026-03823-2

Keywords: gastric cancer, SMYD3, histone methyltransferase, DNA damage response, homologous recombination, double-strand breaks, ATM, PARP inhibitors, synthetic lethality, epigenetics, patient-derived organoids, drug resistance

The enzyme in question is lysine methyltransferase 9, or KMT9, a histone-modifying protein that previous work from the same group had already implicated in prostate tumor growth and androgen receptor signaling. In the new study, published in the journal Molecular Cancer, Jon Peñarando, Eric Metzger, Roland Schüle, and colleagues demonstrate that KMT9 does far more than drive cancer cell proliferation. It actively constructs the molecular barriers that prevent the immune system from recognizing and destroying the tumor, making it a uniquely attractive target for combination approaches designed to sensitize prostate cancer to immunotherapy.

At the center of the discovery is a chemokine signaling axis that functions as a cellular summons for immunosuppressive cells. The researchers found that KMT9 regulates the expression of C-X-C motif chemokine ligands, including CXCL5, which are secreted by prostate tumor cells and bind to the C-X-C motif chemokine receptor 2, CXCR2, on the surface of circulating myeloid cells. This ligand-receptor interaction acts as a homing beacon, drawing polymorphonuclear myeloid-derived suppressor cells, or PMN-MDSCs, into the tumor. Once recruited, these cells populate the tumor immune microenvironment in large numbers, creating a dense myeloid shield around the malignant glands.

PMN-MDSCs are among the most potent enemies of antitumor immunity. They suppress cytotoxic T cell responses through multiple mechanisms, including depletion of the amino acid arginine, production of reactive oxygen species, and interference with T cell trafficking. Using genetically engineered mouse models of prostate cancer in which Pten and Trp53, two frequently altered tumor suppressor genes, are deleted in prostate epithelium, the team showed that tumors with intact KMT9 were heavily infiltrated by these suppressive myeloid cells. When the researchers ablated the Kmt9a gene specifically in the prostate, the picture changed dramatically: PMN-MDSC recruitment collapsed, and cytotoxic T cells flooded into the tumor glands, showing clear signs of activation such as granzyme B expression.

The spatial dimension of this effect proved just as important as the cellular one. Advanced single-cell spatial phenotyping revealed that in tumors with functional KMT9, cytotoxic T cells were largely excluded from the tumor epithelium, lingering instead in the surrounding stroma where they could not make contact with their targets. This immune-excluded pattern is one of the recognized hallmarks of tumors that resist checkpoint blockade, since drugs like anti-PD-1 antibodies can only reinvigorate T cells that are physically close to the cancer cells they are meant to kill. Loss of KMT9 converted this excluded architecture into an inflamed, T cell-infiltrated landscape, precisely the configuration associated with immunotherapy responsiveness.

But KMT9’s contribution to immune evasion did not end with recruitment of suppressor cells. In a second, mechanistically distinct arm of the study, the researchers found that KMT9 also arms the tumor cells themselves against T cell attack. Chromatin immunoprecipitation sequencing showed that KMT9 binds directly to the promoter of the gene encoding arginase 1, ARG1, an enzyme that catabolizes L-arginine, an amino acid essential for T cell function and proliferation. By driving ARG1 expression in prostate tumor cells, KMT9 renders them resistant to T cell-mediated cytotoxicity, effectively allowing them to survive even when killer lymphocytes do manage to engage them.

The therapeutic implications of this dual mechanism were tested in preclinical experiments. When mice carrying Kmt9a-deficient prostate tumors were treated with SB225002, a CXCR2 inhibitor, tumor growth was inhibited to a greater degree than with either intervention alone, confirming that the chemokine axis is a clinically relevant vulnerability downstream of KMT9. Similarly, combining Kmt9a loss with numidargistat, a pharmacological inhibitor of ARG1, produced enhanced suppression of tumor growth. These combination experiments suggest that even partial disruption of the KMT9 pathway could be amplified by drugs that target the individual immune-evasion mechanisms it controls.

Importantly, the team connected their mouse findings to human disease by analyzing data from the TCGA prostate adenocarcinoma cohort, which comprises 500 prostate tumor samples and 52 healthy prostate controls. The analysis showed that KMT9 expression is elevated in prostate tumors relative to normal tissue and that high KMT9 levels correlate with shorter progression-free survival, underscoring the clinical relevance of the pathway. The consistency between the murine genetic models, the human genomic data, and the pharmacological studies lends considerable weight to the conclusion that KMT9 is not an artifact of a single experimental system but a genuine driver of immune evasion in prostate cancer.

For a field searching for ways to unlock immunotherapy in prostate cancer, the study offers a compelling conceptual framework. Rather than targeting the tumor’s defenses one at a time, inhibiting KMT9 would strike at the epigenetic master switch that controls both the recruitment of immunosuppressive myeloid cells through CXCR2 ligand expression and the intrinsic resistance of tumor cells through ARG1. Ablation of KMT9α in the mouse models produced inhibition of prostate tumor growth accompanied by a massive reduction in PMN-MDSC recruitment and a significant increase in cytotoxic T cell activation and infiltration of the tumor glands, a triple effect that few single agents have achieved in this disease.

Considerable work remains before these findings reach the clinic. KMT9 inhibitors are still in early stages of development, and the safety of systemically targeting a methyltransferase with functions in normal tissue is not yet established. The study’s authors note that their findings establish KMT9 as a therapeutic target to reprogram the immunosuppressive landscape and potentially improve the clinical efficacy of current immunotherapies, a formulation that anticipates future trials combining KMT9-directed agents with checkpoint blockade. If those efforts succeed, the stubborn resistance of prostate cancer to immunotherapy, one of the most frustrating puzzles in modern oncology, may finally begin to yield, not to a smarter antibody or a stronger checkpoint inhibitor, but to an epigenetic enzyme that had been quietly building the tumor’s defenses all along.

Subject of Research: The role of the lysine methyltransferase KMT9 in shaping the immunosuppressive tumor immune microenvironment of prostate cancer

Article Title: KMT9 drives T cell exclusion and dysfunction by promoting PMN-MDSCs infiltration and ARG1 expression in prostate cancer

Article References: Peñarando, J., Willmann, D., Sum, M., Jia, Y., Berlin, C., Braun, L. M., Chen, Z., Urban, S., Jung, M., Duteil, D., Metzger, D., Gratzke, C., Zeiser, R., Greschik, H., Schüle, R., & Metzger, E. (2026). KMT9 drives T cell exclusion and dysfunction by promoting PMN-MDSCs infiltration and ARG1 expression in prostate cancer. Molecular Cancer. https://doi.org/10.1186/s12943-026-02801-8

Image Credits: AI Generated

DOI: 10.1186/s12943-026-02801-8

Keywords: KMT9, prostate cancer, tumor immune microenvironment, PMN-MDSCs, CXCL5, CXCR2, ARG1, cytotoxic T cells, cancer immunotherapy, epigenetics, histone methyltransferase, immune evasion