Gated Nanoreactor Turns Tumor Glucose Into Carbon Monoxide to Boost Mild Heat Therapy

Photothermal therapy has long promised a way to cook tumors with near-infrared light, but the approach carries an awkward trade-off: the hotter the tumor gets, the more likely healthy tissue nearby is to burn. A team of researchers in China now reports a nanoscale solution that lets cancer cells be killed with surprisingly gentle heat by sabotaging their own metabolism first. Writing in Materials Today Bio, the group describes a cascade nanoreactor, designated ZIF/HMPB-CO@GOx, that converts the glucose feeding a tumor into hydrogen peroxide, carbon monoxide and a storm of lipid-destroying radicals, all while keeping the tumor temperature near 45 degrees Celsius rather than the 50-plus degrees typically needed for thermal ablation.

The central problem the researchers set out to solve is thermotolerance. When cancer cells experience sublethal heating, they switch on heat-shock responses, most notably the molecular chaperone HSP70, which refolds damaged proteins and shields the cell from thermal stress. That chaperone cycle is expensive: HSP70 depends on the binding and hydrolysis of ATP to do its work. The team reasoned that if a therapy could drain the tumor cell’s ATP supply from two directions at once, blocking glycolysis in the cytoplasm and choking mitochondrial respiration, the heat-shock defense would collapse and mild heat would become lethal. That logic turned a limitation of gentle photothermal therapy into an opportunity for metabolic intervention.

The nanoreactor is built in layers, each with a distinct job. At its core sits hollow mesoporous Prussian blue, an iron-rich nanoparticle chosen for its strong near-infrared absorption, its biocompatibility and its multienzyme-like redox activity. The hollow interior was loaded with dicobalt octacarbonyl, a compact metal-carbonyl complex carrying eight carbon monoxide ligands, and with glucose oxidase, the enzyme that oxidizes glucose into gluconic acid and hydrogen peroxide. A shell of ZIF-8, a zinc-imidazolate metal-organic framework, was then grown over the whole assembly. This shell is the gatekeeper: it stays intact at physiological pH, protecting the payloads during circulation, but dissolves rapidly in the acidic environment of endosomes and lysosomes once the particle is taken up by a cell.

The gating behavior proved central to the design. In buffer at pH 7.4, the particles released little zinc, indicating a stable shell, whereas at pH 5.5 the framework disassembled and zinc release accelerated sharply. Glucose oxidase release followed the same pH dependence and was further promoted by near-infrared irradiation. Encapsulation efficiency for the enzyme reached 75.4 percent, with a loading capacity of 21.6 percent, while spectroscopic analysis of the final particles confirmed the coexistence of the iron-based core, the cobalt carbonyl donor and the zinc shell. Electron microscopy showed the expected porous nanocubic morphology, with an additional outer layer appearing after coating, and blood-cell testing showed red cells retained intact morphology at the highest concentration tested.

The most striking chemistry is the self-amplifying link between glucose and carbon monoxide. In sealed cuvettes, carbon monoxide liberation from the particles was measured by tracking the conversion of deoxygenated hemoglobin to carboxyhemoglobin. Release was modest without a trigger, increased in the presence of hydrogen peroxide, and rose further when glucose and near-infrared light were applied together, reaching roughly 0.062 micromoles of CO per milligram of nanoparticles after 40 minutes. Adding catalase, an enzyme that destroys hydrogen peroxide, sharply blunted the glucose-triggered response, confirming that the hydrogen peroxide produced upstream by glucose oxidase is the chemical trigger that unlocks the cobalt carbonyl donor. In other words, the tumor’s own fuel supply is converted into the signal that releases the gas.

Under an 808-nanometer laser, the particles heated efficiently, reaching a photothermal conversion efficiency of 20.5 percent and maintaining stable performance across repeated on-off cycles. Crucially, the fully coated formulation topped out at 45.9 degrees Celsius in solution, within the 42-to-46-degree nonablative window the authors define as mild photothermal therapy, while bare Prussian blue cores climbed to 53.7 degrees. The coating attenuated but did not abolish heating, and the team argues this moderation is a feature rather than a flaw, because the killing power is meant to come from the biochemical cascade rather than from brute-force thermal ablation.

Experiments in 4T1 murine breast cancer cells showed the cascade firing as intended. Confocal imaging confirmed time-dependent uptake of the coated particles, and a fluorescent probe verified intracellular carbon monoxide generation. Metabolic profiling with a Seahorse analyzer revealed that the full treatment suppressed both the extracellular acidification rate, a proxy for glycolysis, and the oxygen consumption rate, a proxy for mitochondrial respiration. Glycolytic and mitochondrial ATP production both fell, lactate output dropped, and reactive oxygen species fluorescence climbed to its highest level in the laser-treated group. Mitochondrial membrane potential collapsed, electron microscopy revealed damaged organelles, and the lipid peroxidation marker malondialdehyde accumulated to its peak, a signature of ferroptosis, the iron-dependent cell death pathway.

Two control experiments pinned down the death mechanism. Pretreating cells with ferrostatin-1, a ferroptosis inhibitor, partially rescued viability, indicating that ferroptosis contributes substantially to the killing but is not the whole story, with photothermal stress and mitochondrial injury also at play. Western blotting showed that GPX4, the enzyme that detoxifies lipid peroxides, was reduced, and that HSP70 protein abundance fell alongside ATP depletion, consistent with a weakened heat-shock defense. Transcriptomic sequencing added a nuance: the gene encoding HSP70 was strongly induced at the RNA level even as the protein declined, which the authors interpret as a compensatory stress response that could not keep pace with the combined metabolic, oxidative and thermal damage.

In mice bearing 4T1 tumors, the coated particles circulated longer and accumulated more strongly in tumors than uncoated versions, with the ZIF-8 shell extending the plasma half-life and nearly tripling drug exposure over 24 hours. When treated animals received intravenous nanoparticles followed by laser irradiation, tumor temperatures settled around 44.5 to 45.1 degrees Celsius, and tumor growth was suppressed more strongly than in any control group, with a statistically robust difference at day 17. Body weights remained stable, and histological examination of major organs showed no apparent treatment-related abnormalities, though the authors are careful to frame these as preliminary tolerability evidence rather than proof of long-term safety.

The study’s authors acknowledge real limitations before clinical translation: the formulation is complex, the exposure of tumors and healthy tissue to carbon monoxide and zinc ions must be controlled, several causal links in the pathway remain incompletely resolved, and more clinically representative tumor models will be needed. Even so, the conceptual achievement is notable. Rather than fighting a tumor with heat alone, the nanoreactor enlists the cancer cell’s own glucose as fuel for a chain reaction, one that starves glycolysis, poisons respiration with a gas, strips away antioxidant defenses and melts the molecular shield that normally protects cells from warmth. If such designs can be made safe and reproducible, the future of photothermal cancer therapy may belong not to hotter lasers, but to smarter chemistry.

Subject of Research: A pH-responsive cascade nanoreactor combining glucose oxidation, carbon monoxide release and ferroptosis to sensitize tumors to mild photothermal therapy

Article Title: ZIF-8-gated cascade nanoreactor amplifies glucose-triggered CO release and ferroptosis for mild photothermal cancer therapy

Article References: Chen, H., Jiang, T., Zhang, Y., Mei, L., & Mao, Y. (2026). ZIF-8-gated cascade nanoreactor amplifies glucose-triggered CO release and ferroptosis for mild photothermal cancer therapy. Materials Today Bio, 41, Article 103724. https://doi.org/10.1016/j.mtbio.2026.103724

Image Credits: AI Generated

DOI: 10.1016/j.mtbio.2026.103724

Keywords: photothermal therapy, nanoreactor, ZIF-8, glucose oxidase, carbon monoxide, ferroptosis, Prussian blue, HSP70, ATP depletion, tumor metabolism, hydrogen peroxide, cancer nanomedicine