©
Tuan-Anh Nguyen
This world-first discovery, published today in the journal Science, rewrites our understanding of how cells control the production of DNA’s building blocks - and how this process affects the response to widely used cancer and autoimmune drugs.
The study, led by Professor Kilian Huber and researchers from the Centre for Medicines Discovery, Nuffield Department of Medicine, University of Oxford in collaboration with scientists from the CeMM (Research Centre for Molecular Medicine), Austria focused on the enzyme NUDT5, known to have a role in cellular energy metabolism and signalling. This new work uncovered a previously unknown function where NUDT5 acts as a molecular ‘handbrake’ on another enzyme, PPAT, which controls the rate of purine synthesis - the pathway that generates a key building block for DNA and RNA base pairs. By restraining PPAT, NUDT5 fine-tunes the cell’s nucleotide supply, limiting DNA replication.
When this ‘handbrake’ is lost - through genetic variation, disease, or chemical removal - cells overproduce purines. These purines can out-compete certain cancer drug treatments that mimic these molecules (such as thiopurines), thus rendering the treatment ineffective. This discovery helps explain why some patients respond better than others to long-used drugs such as 6-thioguanine, used in leukaemia and autoimmune disease.
This study builds on previous work from the Oxford team and others showing that NUDT5 can be targeted by drug-like small molecules, opening the door to probe its biological roles in cells.
To investigate its function, the Oxford team developed dNUDT5, a first-in-class small-molecule degrader that removes NUDT5 entirely from cultured human cells. Unlike traditional enzyme blockers, this approach revealed that the protein also acts as a scaffold, tightly binding and thus inhibiting PPAT, the key enzyme required for purine production. Removing NUDT5 abolished the brake on purine synthesis and provided direct evidence of the mechanism.
Using proteomics, the researchers then discovered that NUDT5 physically interacts with PPAT, explaining how this enzyme can control metabolism through structure rather than chemistry. Additionally, researchers found that dNUDT5 could rescue adenosine-induced toxicity in patient-derived fibroblasts from individuals with MTHFD1 deficiency, showing the pathway’s broader biomedical relevance.
'This discovery was a big surprise, in the best possible way,' said Professor Kilian Huber, 'An enzyme long thought to have a single, well-defined role turned out to moonlight as a molecular scaffold - something never seen before in this enzyme family. It challenges the textbook view of how cells regulate the production of DNA building blocks and offers a fundamental insight into cellular control. It’s also a reminder that in science, you should never take anything for granted - even familiar proteins can still surprise us. This work shows what can be achieved when curiosity-driven research meets international collaboration.
'Our team is now evaluating NUDT5-PPAT as a potential biomarker in clinical samples, advancing NUDT5 degraders through preclinical testing, and exploring whether other enzymes might have similar hidden scaffold roles in cell metabolism.'
Dr Anne-Sophie Marques, co-first author on the study, said: 'In this study we have shown an unprecedented role of NUDT5 in repressing the purine de novo biosynthetic pathway. In addition to proving that the scaffolding role of NUDT5 is essential in this phenotype, we have demonstrated that the interaction between NUDT5 and PPAT, the rate-limiting enzyme of purine de novo synthesis, acts like a switch to repress the pathway.
'It is undeniable that our results pave the way for future discoveries around the NUDT5 protein and its role in cancer.'
The preclinical findings were independently validated by multiple research teams, including parallel work led by Professor Ralph DeBerardinis and colleagues at the University of Texas Southwestern Medical Center. The convergence of evidence across independent studies underscores the robustness and integrity of these data, pointing to the status of NUDT5-PPAT as a potential biomarker for personalised therapy and as a target for new treatments in cancer and metabolic disorders.
The work was carried out with collaborators from ETH and the University of Zurich, Switzerland, McGill University, Canada, and supported by the EU Innovative Medicines Initiative programme.
The study 'A non-enzymatic role of Nudix hydrolase 5 in repressing purine de novo synthesis’ is published in Science.