Study identifies a major DHODH liability for QC6352 and its clinical analogue Zavondemstat, highlighting the importance of orthogonal chemical probes and ongoing probe curation.

New research published in Nature Chemical Biology provides important evidence that the widely used KDM4 chemical probe QC6352 has a major additional pharmacological activity that can explain its anticancer effects in the glioblastoma models investigated. The study, led by researchers at the University of Sydney in collaboration with Goethe University, the University of Oxford and the Institute of Cancer Research, shows that QC6352 and its clinical analogue Zavondemstat (TACH-101) are potent dual KDM4–DHODH inhibitors. In the glioblastoma models studied, the researchers found that inhibition of dihydroorotate dehydrogenase (DHODH), rather than KDM4 inhibition, is primarily responsible for the compounds’ antiproliferative activity.

The findings have implications not only for the interpretation of studies using QC6352 as a KDM4 probe, but also for the broader use and assessment of KDM4-directed chemical tools, including compounds being developed clinically.

A highly regarded probe reveals an important liability.

QC6352 has been widely used as a chemical probe for the KDM4 family and was previously considered a best-in-class KDM4 inhibitor among the probes listed on the Chemical Probes Portal. It has been used in studies across several cancer models and has also served as an important tool for investigating KDM4 biology.

The new study began with an investigation of QC6352 in glioblastoma stem cells. The compound showed potent activity across six patient-derived glioblastoma stem-cell models, with growth-inhibitory concentrations in the low-nanomolar range.

However, the researchers found that structurally distinct KDM4 inhibition did not reproduce the effects of QC6352. In particular, the orthogonal KDM4 inhibitor ML324 did not phenocopy QC6352’s antiproliferative activity in the glioblastoma models examined. This prompted a more detailed investigation of QC6352’s mechanism of action.

The researchers identified DHODH as a major additional target.

QC6352 inhibited human DHODH with sub-nanomolar potency, and structural studies demonstrated a direct interaction between QC6352 and DHODH. Importantly, the evidence extended beyond biochemical binding. Addition of uridine, which can bypass inhibition of the de novo pyrimidine biosynthesis pathway, fully rescued the antiproliferative effects of QC6352 in the glioblastoma models tested. Expression of an inhibitor-resistant DHODH A58T mutant also strongly rescued the effects of QC6352. Together, these findings support DHODH inhibition as the primary driver of QC6352 activity in the glioblastoma models investigated.

The finding extends to the clinical analogue Zavondemstat.

The study also examined Zavondemstat, a clinical analogue of QC6352 that has progressed into clinical trials. The researchers found that Zavondemstat potently inhibits DHODH. The difference in cellular potency between its R- and S-enantiomers correlated much more closely with their difference in DHODH potency than with their relatively small difference in KDM4A potency. Both enantiomers also showed cellular effects that could be rescued by uridine.

These findings raise important questions for the interpretation of the mechanism of action of Zavondemstat and therefore have implications for its ongoing clinical development.

Importantly, the findings do not by themselves establish that DHODH inhibition accounts for every clinical or biological effect of Zavondemstat. Rather, they identify a potent and previously underappreciated pharmacological activity that needs to be considered when interpreting the compound’s biology and clinical development.

What about other KDM4-directed compounds?

The implications of the work extend beyond QC6352 and Zavondemstat. The new findings therefore reinforce the importance of considering DHODH activity when evaluating KDM4-directed compounds, particularly where cellular phenotypes are being attributed specifically to KDM4 inhibition.

However, it is important not to overextend the conclusions of the new study.

The experiments reported in the paper provide strong evidence for DHODH-driven activity in the glioblastoma models investigated. They do not demonstrate that DHODH inhibition necessarily explains every phenotype previously reported for QC6352 or Zavondemstat in every biological setting.

The authors themselves note that the relative contributions of KDM4 and DHODH may vary between cellular and disease contexts, depending on the expression and functional dependency of the two targets.

This distinction will be important as the wider literature is reassessed.

Why orthogonal probes matter

One of the central lessons of the study is the importance of using orthogonal chemical tools when assigning a cellular phenotype to a particular target.

QC6352 was a particularly compelling KDM4 probe based on the evidence available when it was characterised and assessed. Its potency, selectivity within the Jumonji demethylase family and drug-like properties contributed to its previous status on the Chemical Probes Portal. The new work demonstrates how even a highly regarded probe can have an important pharmacological liability that becomes apparent only through additional mechanistic investigation.

The authors specifically highlight the risk of relying on a single chemical entity and the need for multiple, high-quality chemical probes when investigating protein function.

This is particularly relevant when a compound produces a strong cellular phenotype. Potency alone does not establish mechanism, and a phenotype produced by one chemical probe should ideally be tested using structurally distinct compounds, genetic approaches and appropriate pathway-level controls.

The Chemical Probes Portal response

The Chemical Probes Portal provides researchers with curated information about the strengths, limitations and appropriate use of chemical probes. A key part of this role is ensuring that probe assessments evolve as new evidence becomes available.

In response to the new findings, the Portal will:

Update the QC6352 entry. The Portal record will be revised to incorporate the new evidence concerning DHODH activity and its contribution to the cellular phenotypes observed in the study.
Add a clear curator warning. Researchers will be alerted to the DHODH liability and the implications for interpreting experiments that use QC6352 to investigate KDM4 biology.
Review the QC6352 probe assessment. The existing probe rating will be reconsidered in light of the new evidence and following appropriate review and discussion. Consider the implications for related compounds.

The DHODH liability will also be taken into account when assessing related KDM4-directed compounds, while maintaining a distinction between demonstrated findings and mechanisms that require further investigation. Highlight the need for orthogonal approaches.

The Portal will emphasise the importance of using multiple chemical probes and appropriate controls when assigning cellular phenotypes to KDM4 inhibition.

The aim is not simply to label a compound as “good” or “bad”. Chemical-probe assessment is an ongoing process, and the usefulness of a probe depends on understanding both its intended activity and its limitations.

A new opportunity for better KDM4 probes

The study also provides a constructive way forward.

Using structural information about the interaction of QC6352 with DHODH and KDM4, the researchers developed new compounds designed to retain KDM4 inhibition while avoiding DHODH inhibition.

Two compounds demonstrated potent KDM4A inhibition while showing poor DHODH inhibition, providing starting points for the development of more selective KDM4 chemical probes.

This is an important outcome because identifying a liability does not have to be the end of a chemical probe programme. It can provide an opportunity to design better tools that separate the intended target activity from unintended pharmacology.

Keeping the evidence up to date

The story of QC6352 illustrates an important principle in chemical biology: probe quality is not a permanent designation.

A compound can be extensively characterised and provide valuable information, while later experiments reveal limitations that change how its biological effects should be interpreted.

The appropriate response is not to disregard the previous literature, nor to assume that every historical observation was caused by the newly identified target. Instead, researchers need access to the new evidence so they can reassess individual findings in context.

The Chemical Probes Portal will therefore continue to update its assessments as new evidence emerges, making important limitations visible to researchers and supporting the use of appropriate orthogonal tools.

The new study is a strong reminder that rigorous target validation requires more than a potent compound and a compelling phenotype.

Good chemical biology depends on good chemical tools — and good chemical tools depend on continuous scrutiny.

Read the new paper here DOI: 10.1038/s41589-026-02306-x

New evidence prompts reassessment of widely used KDM4 chemical probes