Self-Reporting Nanoparticle Turns Tumor Hypoxia Into a Weapon Against Cancer

Solid tumors are notorious for their ability to thrive in conditions that would starve healthy tissue, and one of the most striking examples of this adaptability is hypoxia, the oxygen-depleted state that arises when rapidly dividing cancer cells outgrow their blood supply. For decades, researchers have tried to turn this weakness into a therapeutic advantage through hypoxia-activated prodrugs, inert molecules that are converted into potent cytotoxic agents only by the reductive enzymes that flourish in oxygen-starved tissue. In principle, such drugs should spare healthy cells and strike tumors with surgical precision. In practice, the strategy has repeatedly stumbled, because tumor hypoxia is patchy and unpredictable. Only a fraction of cells within a given tumor experiences oxygen levels low enough to trigger efficient drug activation, while better-oxygenated regions escape treatment and continue to drive disease progression. The clinical disappointment of evofosfamide, a hypoxia-activated prodrug that failed to deliver decisive benefits in trials, illustrates how severe this heterogeneity problem can be.

A team of researchers led by Jie Gao, Wen-Chao Geng, and Zeli Yuan, publishing in Materials Today Bio, has now unveiled a nanoplatform that tackles both of the fundamental bottlenecks of hypoxia-activated chemotherapy at once. Their system, called CL-Azo-CPT/Ce6 NPs, does not merely wait passively for endogenous hypoxia to activate the drug. Instead, it actively amplifies the oxygen starvation using light, and then reports in real time, through its own chemiluminescent glow, exactly how much drug has been released inside the tumor. The authors describe this integrated behavior as a Sense-Amplify-Report cycle, and the design represents a rare example of a therapeutic formulation that can measure its own activation while it is happening inside a living animal.

The molecular heart of the platform is a chemiluminescent prodrug named CL-Azo-CPT. It is built around Schaap’s adamantylidene-1,2-dioxetane, a luminophore famous for the persistent glow it emits when destabilized, and it is connected to the anticancer payload camptothecin through a self-immolative carbonate spacer and an azobenzene trigger. The azobenzene linkage is the gatekeeper: it is cleaved selectively by azoreductase enzymes, whose activity rises sharply under hypoxic conditions. Once the azo bond is cut, a spontaneous self-immolative cascade unfolds, simultaneously liberating camptothecin and unmasking the dioxetane luminophore. Because a single activation event produces both the drug and a photon, the light output is stoichiometrically coupled to drug release. In solution experiments, the researchers demonstrated a nearly perfect linear correlation between chemiluminescence intensity and the amount of camptothecin released, quantified independently by high-performance liquid chromatography, with a correlation coefficient of 0.997.

The optical performance of the prodrug is equally impressive. Upon exposure to sodium dithionite, a chemical mimic of azoreductase activity, CL-Azo-CPT produced a broad emission spanning 430 to 650 nanometers with a peak at 515 nanometers, roughly a ninety-fold enhancement over the background signal of the intact molecule. The signal rose to its maximum within five minutes and decayed with a half-life of eighteen minutes, kinetics well suited to dynamic monitoring. Selectivity testing against a battery of biologically relevant species, including metal ions, amino acids, glutathione, hydrogen peroxide, reducing agents, and digestive enzymes, confirmed that only azo reduction elicited a strong response. The intact azobenzene moiety efficiently quenches the dioxetane emission, so no appreciable glow appears until the hypoxia trigger has actually fired, keeping the background noise low.

Crucially, chemiluminescence requires no external excitation light. Unlike fluorescence imaging, which demands a laser or lamp that also excites endogenous molecules and generates tissue autofluorescence, chemiluminescence emerges from the chemical reaction itself. This gives it an exceptionally high signal-to-noise ratio in living tissue, where autofluorescence often buries weak fluorescent signals. For a reporter intended to track drug activation deep within a tumor microenvironment, that excitation-free property is not a luxury but a necessity, and it is one of the main reasons the authors chose a dioxetane scaffold rather than a conventional fluorophore.

To add the amplification step, the researchers co-encapsulated CL-Azo-CPT with chlorin e6, a well-known photosensitizer, inside PEGylated DSPE-mPEG2000 nanoparticles prepared by thin-film hydration. The resulting particles measured about 169 nanometers in hydrodynamic diameter with a polydispersity index of 0.22, showed uniform spherical morphology by electron microscopy, and remained colloidally stable for a week in buffer. Encapsulation efficiencies were high, at roughly 97 percent for the prodrug and 65 percent for the photosensitizer. When illuminated with white light, the encapsulated chlorin e6 generated singlet oxygen, confirmed by the rapid decay of the indicator dye ABDA, proving that co-encapsulation did not impair its photodynamic activity. That activity comes with a built-in consequence: every photon-driven reaction consumes molecular oxygen, so photodynamic therapy actively deepens the very hypoxia that the prodrug needs in order to activate.

Cellular experiments in CT26 colon carcinoma and HCT116 colorectal cancer cells mapped out each element of the cascade. Under chemically induced hypoxia, intracellular fluorescence from released camptothecin and the MCCH reporter climbed steadily over six hours, while remaining weak in well-oxygenated cells, and chemiluminescence signals increased with both time and drug concentration. Cytotoxicity assays under a four-condition matrix of oxygen status and light exposure revealed the logic of the design clearly. In the dark under normoxia, the nanoparticles were essentially harmless, confirming that the drug stayed caged and the photosensitizer stayed quiet. Under hypoxia alone, moderate killing demonstrated the Sense module, with the prodrug responding to endogenous low oxygen. Under light alone, photodynamic killing occurred but was weakened by oxygen deprivation. Only when hypoxia and light were combined did the full formulation achieve its lowest IC50 values, 3.0 micromolar in CT26 cells and 2.6 micromolar in HCT116 cells, along with near-complete cell ablation in live-dead staining.

The in vivo results were the most striking. In mice bearing subcutaneous 4T1 tumors, endogenous hypoxia alone produced a chemiluminescent response that peaked about fifty minutes after intratumoral injection. When the tumors were irradiated, the signal shifted to a later but substantially stronger maximum at 110 minutes, with peak intensity 1.69-fold higher and integrated signal 1.86-fold higher than in non-irradiated tumors, while normal tissue showed consistently weak signals. In a CT26-Luc1 peritoneal metastasis model, intraperitoneal injection of the nanoparticles produced a distinct glow that peaked at fifteen minutes in tumor-bearing mice and was undetectable in healthy animals, with ex vivo imaging confirming that the emission localized to intestinal tumor nodules. Because the chemiluminescent output is coupled to drug release, these images amount to a real-time, semi-quantitative map of where and when the chemotherapy is actually being deployed, though the authors caution that tissue absorption and scattering make the in vivo signal a relative rather than absolute measure.

Therapeutically, the platform delivered decisive results in the peritoneal metastasis model. Seventy tumor-bearing mice were divided into seven treatment groups receiving four intraperitoneal doses over fourteen days, with light-treated groups receiving abdominal illumination after each injection. Mice given saline or photosensitizer alone all died within fifteen days, and chemotherapy-containing groups without light extended survival only marginally, with all animals dying within twenty days. Photodynamic therapy alone prolonged survival to roughly twenty-five days, but the combined CL-Azo-CPT/Ce6 nanoparticles with light performed best, with one mouse surviving beyond day thirty. At the endpoint, the combined treatment produced the lowest ascites volume, the smallest tumor nodules, the highest apoptotic index at 72.81 percent by TUNEL staining, and the lowest proliferative fraction at 2.84 percent by Ki-67 staining. Serum biochemistry showed that liver stress markers, elevated in all other tumor-bearing groups, returned to healthy-control levels in the combined-treatment group, and histology of major organs revealed no off-target damage.

The authors are careful to frame the platform honestly. The current formulation is feedback-guided rather than fully closed-loop: the chemiluminescent readout can inform decisions about subsequent light dosing, but automated real-time adjustment of treatment remains a goal for future work. They also note that because chlorin e6 is excited at approximately 654 nanometers and light penetrates tissue only shallowly, the strategy is best suited to superficially accessible tumors, intraperitoneal lesions amenable to laparoscopic or intraoperative illumination, or gastrointestinal lesions reachable by endoscopy, exactly the settings modeled in their peritoneal carcinomatosis study. There is also a subtler consideration: excessive or sustained hypoxia can, in principle, promote tumor invasiveness and immunosuppression, so the self-reporting design offers a way to titrate light exposure within a therapeutic window, stopping once the readout indicates adequate activation. Even with those caveats, the work establishes a compelling framework for next-generation precision nanomedicines, one in which the treatment does not just attack the tumor microenvironment but watches itself do it, photon by photon.

Subject of Research: A self-reporting chemiluminescent nanoplatform coupling photodynamic hypoxia amplification with hypoxia-activated prodrug therapy for precision cancer treatment

Article Title: Coupling photodynamic hypoxia amplification with a self-reporting chemiluminescent prodrug for adaptive precision therapy

Article References: Gao, J., Wei, L., Wang, Y., Yang, M., Wu, Y., Geng, W.-C., & Yuan, Z. (2026). Coupling photodynamic hypoxia amplification with a self-reporting chemiluminescent prodrug for adaptive precision therapy. Materials Today Bio, 41, Article 103712. https://doi.org/10.1016/j.mtbio.2026.103712

Image Credits: AI Generated

DOI: 10.1016/j.mtbio.2026.103712

Keywords: hypoxia-activated prodrug, chemiluminescence, photodynamic therapy, nanoparticles, camptothecin, chlorin e6, tumor microenvironment, theranostics, peritoneal metastasis, azoreductase, dioxetane, precision medicine