Blocking UXS1 Turns Metformin's Metabolic Side Effect Into a Lung Cancer Weakness
Metformin pushes lung cancer cells to make toxic UDPGA. Blocking UXS1 with plantainoside traps them, boosting tumor kill and immunotherapy in mice.
Summary
Metformin's anticancer effects have been inconsistent in clinical trials. This preclinical study found that metformin activates the enzyme UGDH through AMPK, PDK4 and a newly identified phosphorylation site (S476). This converts UDP-glucose into UDP-glucuronic acid (UDPGA), which is toxic at high levels. Tumor cells survive by using UXS1 to convert UDPGA into UDP-xylose, so metformin-treated cells depend more on UXS1. Virtual screening of 17,676 compounds identified plantainoside as a UXS1-targeting agent. Combined with metformin, it killed lung adenocarcinoma cells synergistically and shrank tumors in organoids, xenografts and a spontaneous mouse model. The combination also increased CD8+ T cells and improved immunotherapy response in mice. The findings are preclinical, and plantainoside's direct target engagement still needs fuller validation.
Detailed Summary
Metformin is among the most widely used drugs in the world and has long been studied as a possible anticancer and longevity-related agent. Its results in cancer patients are inconsistent, though. Some reports suggest benefit, and others suggest no effect or worse outcomes alongside immune checkpoint inhibitors. Understanding what makes tumors respond or resist could help define when metformin is useful.
The researchers ran metabolomics on lung adenocarcinoma (LUAD) patient tumors, mouse xenografts and A549 cells, comparing metformin exposure with none. UDP-glucose (UDPG) fell consistently in all three. Follow-up experiments showed metformin drives UDPG into UDP-glucuronic acid (UDPGA) via UGDH. Phosphoproteomics identified increased phosphorylation of UGDH at serine 476. Metformin triggers this through AMPK and an intermediate kinase, PDK4, which physically interacts with UGDH. Docking analysis suggests the phosphorylation weakens the binding of UDP-xylose, UGDH's natural feedback inhibitor, so the enzyme stays active. Phospho-dead and phospho-mimetic mutants blunted metformin's metabolic effect, supporting S476's role.
High UDPGA is toxic, and cells normally clear it with UXS1, which converts it to UDP-xylose (the 'kitchen-sink' or drain model). UXS1 knockout lowered metformin's IC50 in LUAD cells, raised UDPGA and cut UDP-xylose. In 60 metformin-treated LUAD patients, low UXS1 expression tracked with better survival. Removing UGDH or UGP2 in UXS1-knockout cells erased metformin's added effect, consistent with the pathway model.
To exploit this dependency, the team virtually screened 17,676 compounds, narrowed to 126 candidates, and used high-content screening to select plantainoside. It showed synergy with metformin in cell assays, lung cancer organoids, xenografts and a spontaneous lung cancer mouse model, where the combination gave the slowest tumor growth and longest survival. UGDH-knockout cells were largely resistant to plantainoside, consistent with action through UXS1 flux. The combination also increased CD8+ T cells, reduced macrophage differentiation and improved immunotherapy efficacy in mice.
The work is preclinical, and no human trial of this combination exists. The S476 phosphorylation had only borderline prognostic value in patient cohorts (OS p = 0.11 in CPTAC; log-rank p = 0.0917 in the in-house cohort). Plantainoside's direct binding and selectivity, plus safety in normal tissues, remain to be established. Even so, the study offers a mechanistic explanation for metformin's variable effects and a rational, biomarker-linked combination strategy.
Key Findings
- Metformin lowered UDP-glucose in patient tumors, xenografts and cells by driving its conversion to UDP-glucuronic acid via UGDH.
- AMPK-PDK4 signaling phosphorylates UGDH at S476, a newly identified site that appears to relieve UDP-xylose feedback inhibition.
- UXS1 knockout lowered metformin's IC50 and raised UDPGA; low UXS1 tracked with better survival in 60 metformin-treated LUAD patients.
- Virtual screening of 17,676 compounds identified plantainoside, which synergized with metformin in cells, organoids, xenografts and spontaneous lung tumors.
- The combination increased CD8+ T cells, suppressed macrophage differentiation and improved immunotherapy efficacy in mouse models.
Methodology
Multi-level preclinical study combining metabolomics and phosphoproteomics in LUAD patient tumors, xenografts and cell lines (A549, PC9), CRISPR knockouts and phospho-mutants, docking, and virtual screening of 17,676 compounds. Efficacy was tested in organoids, nude mouse xenografts and a spontaneous lung cancer model, with a 60-patient in-house survival cohort. Patient sample sizes for metabolomics were small (n = 6 per group).
Study Limitations
All efficacy data come from cell lines, organoids and mouse models; no human trial of the combination exists. Plantainoside's direct binding to UXS1, selectivity and toxicity need further validation, and p-UGDH(S476) prognostic associations were not statistically significant. Metformin concentrations used in vitro (1 mM) exceed typical plasma levels in patients.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: