Dual-responsive polypyridyl Ru(II)-scaffolded hydrogel for trinitarian tumor therapy activated by mono-wavelength photo-irradiation Authors: Zhendong Chen, Jiao Zhou, Sizhen Wang, Tianheng Chen, Francisco Aznarez, Xintong Yan, Xiaoxiao Hu, Yiling Xia, Linhong Sun, Chen Wang, Feng Yang, Beibei Guo Published: 2026 ABSTRACT Metal-coordinated hydrogels offer a strategy to integrate local drug retention, responsive release, and multimodal therapy, but assigning structural and therapeutic roles to a single metal complex remains challenging. Here, we report an injectable polypyridyl Ru(II)-scaffolded hydrogel, Ru(fmb)3HADH-PDA@IR780, formed by dynamic Schiff-base crosslinking and loaded with polydopamine nanoparticles carrying IR780. In this platform, PDA@IR780 enables 808 nm near-infrared-activated photothermal and photodynamic outputs whereas the Ru(II) complex functions as both a hydrogel crosslinking unit and a redox-active therapeutic component. This formulation demonstrates dual pH/photo-responsiveness: acidic tumor microenvironment (TME) promoted formulation release, and NIR irradiation triggered rapid local heating and reactive oxygen species (ROS) generation. In vitro, the hydrogel produced enhanced tumor-cell killing through combined photothermal, photodynamic, and Ru-associated oxidative and chemodynamic therapy (CDT), accompanied by glutathione depletion and lipid-peroxidation-related stress. In vivo, local administration followed by short NIR irradiation markedly suppressed tumor growth, increased tumor-cell apoptosis, prolonged tumor-site retention, and showed no obvious short-term systemic toxicity under the tested regimen. Transcriptomic and histological analyses further suggested treatment-associated redox disruption and favorable modulation of the local immune microenvironment. This Ru(II)-scaffolded hydrogel provides a component-defined, single-wavelength-activated platform for locally retained multimodal tumor therapy. Notably, this work represents the first report of a polypyridyl Ru(II)-scaffolded hydrogel that unifies Ru(II)’s structural coordination with intrinsic therapeutic action, offering a translational strategy for synergistic, site-selective tumor therapy.Graphical Abstract Schematic illustration of the overall fabrication and application of Ru(fmb)3HADH-PDA@IR780 hydrogel in combination therapy, including the components of the hydrogel and three therapeutic modalities: PDT, PTT, and CDT. Supplementary InformationThe online version contains supplementary material available at 10.1186/s12951-026-04658-x. Introduction Cancer remains one of the leading causes of death globally, with conventional therapies such as surgery, chemotherapy, and radiotherapy facing inherent bottlenecks. Surgery is curative for early resectable tumors but inapplicable to unresectable or metastatic cases; chemotherapy improves survival across cancer types yet induces severe off-target toxicity [1, 2]; radiotherapy suffers from tissue damage and radioresistance, collectively limiting clinical outcomes. Ru(II) complexes, particularly Ru(II) polypyridyl complexes, have emerged as promising alternatives to platinum-based chemotherapeutics [3–6]. Their appeal lies in superior biocompatibility, tumor-selective uptake, prolonged intratumoral retention, and unique photophysical properties [7, 8], which collectively underpin multi-faceted mechanisms of action including enzyme-mimetic catalysis, DNA damage induction, and immunomodulation [9–12]. Such multi-faceted mechanisms enable their diverse application [13, 14] as chemotherapeutics, photosensitizers (PSs) and immunoregulants for photodynamic/photothermal therapy (PDT/PTT) and immunotherapy, effectively addressing tumors refractory to conventional treatments. Notably, the Ru(II)-based photosensitizer TLD1433 has advanced to Phases I/II clinical trials for bladder cancer PDT [15]. However, free Ru(II) agents are associated with significant challenges, including drug resistance, dose-dependent systemic toxicity and short circulatory half-lives [16, 17], which hinder multimodal therapy implementation. Although polymer-based multimodal therapeutic platforms have been developed [18–20], they remain limited in maintaining therapeutically effective concentrations at the tumor site, underscoring that controllable drug delivery and release systems are critical to unlocking Ru(II)’s full therapeutic potential [21]. Injectable and responsive hydrogels stand out as ideal carriers for localized, smart drug delivery [22–25]. By integrating responsive components into the hydrogel matrix, such as pH-sensitive moieties [26, 27], the release of therapeutic agents can be precisely modulated in response to specific environmental stimuli. This characteristic facilitates targeted drug delivery, enabling selective release at the intended site of action, thereby mitigating systemic side effects and optimising therapeutic efficacy. Recent advances have led to the design of hydrogels responsive to the mildly acidic conditions of the tumor microenvironment (TME) [28, 29]. Specifically, we innovatively integrated a polypyridyl Ru(II) complex Ru(fmb)3Cl2 bearing aldehyde groups, into a hydrogel network via Schiff-base reactions with acyl hydrazide-functionalized hyaluronic acid (HADH) and hydrogen bonding, yielding an injectable and pH-responsive Ru(II)-scaffolded hydrogel (Ru(fmb)3HADH). This gel features an advanced three-dimensional scaffold, serving as an efficient carrier platform for diverse agents. Incorporating encapsulated functional agents is crucial to constructing multimodal therapeutic systems for enhanced efficacy. Recently, a series of photothermal agents (PTAs) and PSs responding to near-infrared (NIR) light have emerged as formidable candidates in various tumor therapeutic systems, helping in the destruction of tumor cells through PDT and PTT [30–33]. To enable multimodal therapy, newly designed NIR-responsive PDA@IR780 nanoparticles were encapsulated into Ru(fmb)3HADH, forming the Ru(fmb)3HADH-PDA@IR780 system. Previous studies have shown that PDA possesses light absorption, photothermal conversion, adhesive, and drug-loading capabilities, making it suitable for biomedical delivery and PTT applications [34]. PDA can enhance photosensitizer delivery and support combined PDT/PTT [35]. In contrast, IR780 can generate heat and ROS under NIR irradiation. Notably, IR780 exhibits selective accumulation in tumor cell mitochondria [36], improving its precision in PDT cancer treatment. However, its inherent poor water solubility and photoinstability under light exposure have long plagued its translational application and therapeutic efficacy [37, 38], which can be mitigated by encapsulation in PDA@IR780 nanoparticles. Furthermore, the nanoparticles can passively accumulate in tumor tissues via the enhanced permeation and retention (EPR) effect, further boosting the combined PDT/PTT efficacy. Moreover, one additional innovation lies in the hydrogel’s ability to amplify therapeutic synergy via glutathione (GSH) depletion. Tripeptide GSH in tumors exerts a significant influence on the clearance of ROS to sustain redox homeostasis within the organism, ultimately undermining the therapeutic efficacy of PDT [39, 40]. To overcome this challenge and simultaneously enhance the functionality of hydrogel systems, the Ru(fmb)3HADH-PDA@IR780 system incorporates pre-designed aldehyde groups. Upon direct in situ injection into the tumor site, this hydrogel system exhibited efficient, sustained pH-responsive release of Ru(II) complex and PDA@IR780 within the weakly acidic TME owing to the reversible properties of imine bonds within the hydrogel network. The aldehyde groups in the released Ru(II) complexes of the system depletes cysteine (Cys) effectively, a critical rate-limiting precursor in GSH biosynthesis [41]. This system harnesses the reactivity of aldehyde groups toward amino and thiol groups, thereby consuming cysteine and reducing its availability for GSH biosynthesis [42]. The resulting decrease in cysteine levels limits GSH production, weakens ROS scavenging capacity, and impairs the activity of glutathione peroxidase 4 (GPX4) [43, 44]. Hence, the Ru(II) complexes featuring aldehyde groups within this hydrogel network exert their well-established chemotherapeutic efficacy on their own and induce ferroptosis-mediated cytotoxicity against cancer cells through promoting the accumulation of lipid peroxides (LPO) [45]. Consequently, the newly designed injectable pH/NIR-dual-responsive polypyridyl Ru(II)-scaffolded hydrogel system (Ru(fmb)3HADH-PDA@IR780) demonstrated a significant enhancement in the efficacy of synergistic tumor PDT/PTT/CDT via mono-wavelength 808 nm irradiation. It simultaneously minimizes systemic toxicity and enhances localized drug release, fully tapping into the potential of hydrogels as efficient drug delivery systems in cancer therapy. This work pioneers the use of a coordination-synthesized Ru(II) complex as both a hydrogel crosslinker and therapeutic agent, establishing a novel paradigm for precision synergistic tumor therapy (Fig. 1). Fig. 1Schematic illustration of the overall fabrication and application of Ru(fmb)3HADH-PDA@IR780 hydrogel in combination therapy, including the components of the hydrogel and three therapeutic modalities: PDT, PTT, and CDT Schematic illustration of the overall fabrication and application of Ru(fmb)3HADH-PDA@IR780 hydrogel in combination therapy, including the components of the hydrogel and three therapeutic modalities: PDT, PTT, and CDT Results Preparation and characterization of HADH, Ru(fmb)3Cl2, and PDA@IR780 The synthetic route, preparation process and stimuli responsiveness are illustrated in. Fig. 2. Upon the addition of the ligand to the precursor solution, the mixture rapidly formed a gel. Aldehyde-functionalized Ru(II) polypyridyl complexes have been reported as versatile precursors for further conjugation, with the aldehyde functionality retaining useful chemical reactivity within the Ru coordination framework, related aldehyde-containing Ru polypyridine complexes have also been used for the recognition of Cys/Hcy [46, 47]. Owing to the injectability of the hydrogel, the mixture could be extruded from a syringe for precise pattern formation (Fig. 3a). The water absorption characteristics of the hydrogel were evaluated based on the ratio of aldehyde groups in the ligand and amino groups in the precursor. The hydrazine (from HADH precursor) and aldehyde group contents (from Ru(fmb)3Cl2 complex) were controlled at 4:1 (3.4%), 3:1 (4.6%), and 2:1 (6.8%) (v/v). As the hydrazine group content increased, the crosslinking density and water absorption capacity decreased. The hydrogels exhibited swelling ratios from 7 to 14 times (Fig. 3b[i]), which facilitate the absorption of local bleeding and tissue fluid after tumor resection. The water absorption capacity of hydrogels designed for tumor microenvironments must strike a balance—neither too low nor too high to ensure optimal performance [48]. This is critically influenced by the elevated interstitial fluid pressure (IFP) characteristic of tumor tissues. Fig. 2Schematic synthesis workflow of Ru(fmb)3HADH hydrogel. (a) Synthetic strategy of Ru(fmb)3Cl2, HADH, and Ru(fmb)3HADH hydrogel. (b) Construction of PDANPs and IR780-loaded PDA@IR780. (c) Schematic illustration of the gelation process of the Ru(fmb)3HADH hydrogel and micro three-dimensional structure of porous hydrogel. (d) Schematic illustration of dual-responsiveness, including pH-responsiveness conferred by dynamic Schiff base linkages and the photo-responsiveness achieved via the incorporation of the photosensitizer IR780 and PDANPs under conditions involving irradiation at 808 nm Schematic synthesis workflow of Ru(fmb)3HADH hydrogel. (a) Synthetic strategy of Ru(fmb)3Cl2, HADH, and Ru(fmb)3HADH hydrogel. (b) Construction of PDANPs and IR780-loaded PDA@IR780. (c) Schematic illustration of the gelation process of the Ru(fmb)3HADH hydrogel and micro three-dimensional structure of porous hydrogel. (d) Schematic illustration of dual-responsiveness, including pH-responsiveness conferred by dynamic Schiff base linkages and the photo-responsiveness achieved via the incorporation of the photosensitizer IR780 and PDANPs under conditions involving irradiation at 808 nm The 6.8% complex ratio hydrogel showed a degradation rate of 69.75% in phosphate buffer solution (PBS) after 72 h (Fig. 3b[ii]), indicating good biodegradability. Therefore, all subsequent experiments utilized the Ru(fmb)3HADH hydrogel with a 6.8% complex. The microscopic structural details of the hydrogel were investigated using SEM. The analysis revealed that the Ru(fmb)3HADH hydrogel exhibited a well-defined network structure, with high-resolution images displaying a smooth surface texture and an approximate pore size of 100 μm, whereas the encapsulation of PDA@IR780 nanoparticles preserved the inherent structural integrity of the material (Fig. 3c). These characteristics indicate that the hydrogel possesses a suitable structural foundation for its application as a drug delivery carrier. Furthermore, energy-dispersive X-ray spectroscopy analysis revealed the elemental composition of the gel (Fig. 3e), revealing a ruthenium (Ru) content of 0.60% with uniform distribution. Transmission electron microscopy provided detailed insights into the microscopic structure of the PDA nanoparticles (Fig. 3f). The rheological properties of the Ru(fmb)3HADH hydrogel were systematically investigated. Time sweep and strain sweep analyses were conducted to evaluate variations in the elastic (G′) and viscous moduli (G″) under different strain conditions, thereby defining the linear viscoelastic region (Fig. 3d[i] and [ii]). Thixotropic testing demonstrated the ability of the hydrogel to recover its viscosity following high shear stress (Fig. 3d[iii]). Frequency sweep measurements highlighted the solid-like behavior of the hydrogel, characterised by a dominant storage modulus (G′ > G″) over short time scales (Fig. 3d[iv]). Flow curves provided insights into the fluidity of the hydrogel and its flow behavior (Fig. 3d[v]). Finally, step-strain experiments revealed the excellent self-healing capabilities of the hydrogel, as evidenced by the reversible changes in G′ and G″ (Fig. 3d(vi)). Dynamic light scattering analysis revealed the hydrodynamic size of and zeta potential of PDA (Fig. 3g). DLS values are expected to be larger than TEM-derived sizes because DLS reports hydrated hydrodynamic diameter in dispersion. Recorded [1] H NMR spectra confirmed the synthesis of the Ru(fmb)3Cl₂ complex and HADH precursor solution (Fig. S1). FT-IR spectroscopy revealed a strong, narrow absorption band in the 1700–1650 cm⁻¹ range, assigned to the stretching vibration of the -N-N- bond of the diazanyl group typically appears in the range of 1080–1020 cm⁻¹, corresponding to the characteristic -C = O stretching vibration of the diazanyl group. Additionally, a characteristic bending vibration absorption peak for the -N-N-H group in the hydrazide appeared in the 1600–1500 cm⁻¹ region. Comparison with the IR spectrum of HA corroborated the synthesis of the HADH precursor (Fig. S2). Mass spectrometry (Fig. S3) and 1H-NMR (Fig. S4) confirmed the synthesis of ruthenium complex Ru(fmb)3Cl2. IR780 release from Ru(fmb)3HADH-PDA@IR780 hydrogel was measured at pH 7.4 and 5.5 (Fig. S6) [49]. After 10 h, the cumulative release at pH 5.5 was 75.43%, whereas it was 49.12% at pH 7.4. These results indicate that the hydrogel responds to the acidic TME, enhancing IR780 release for targeted delivery. Fluorescence staining revealed that in B16F10 cells co-incubated with Ru(fmb)3HADH hydrogel under pH 5.5 and pH 7.4 conditions for 15 min and 60 min, there was clear colocalisation of the Ru ligand with the nuclear dye Hoechst 33,342 and the mitochondrial dye MitoTracker Green (MTG). Notably, after 60 min of incubation the uptake of the Ru ligand by B16F10 cells at pH 5.5 was markedly higher than in cells incubated at pH 7.4. In contrast, in L929 cells virtually no red fluorescence signal of the Ru ligand was detected after either 15–60 min, and both the plasma-membrane and mitochondrial morphology remained intact. These results indicate that under physiological extracellular pH conditions, normal tissue-derived cells take up exceedingly low amounts of the Ru(fmb)3Cl2 ligand (Fig. S9). Fig. 3Characterization of physiochemical and rheological properties of Ru(fmb)3HADH hydrogel and PDANPs. (a) Injectability of Ru(fmb)3HADH hydrogel and images of precursor and complex. (b) Swelling ratio and degradation rate of Ru(fmb)3HADH hydrogels with 3.4%, 4.6%, and 6.8% Ru(fmb)3Cl2 (v/v; n = 3). (c) SEM images of the hydrogels. (d) Rheological properties of Ru(fmb)3HADH hydrogels: (i) time sweep, (ii) strain sweep, (iii) thixotropy, (iv) frequency sweep, (v) flow curves, and (vi) step-strain tests. (e) EDS analysis showing element distribution in Ru(fmb)3HADH hydrogels, including ruthenium. (f) TEM images of PDANPs. (g) Hydrodynamic size and zeta potential of PDANPs and PDA@IR780 (n = 3) Characterization of physiochemical and rheological properties of Ru(fmb)3HADH hydrogel and PDANPs. (a) Injectability of Ru(fmb)3HADH hydrogel and images of precursor and complex. (b) Swelling ratio and degradation rate of Ru(fmb)3HADH hydrogels with 3.4%, 4.6%, and 6.8% Ru(fmb)3Cl2 (v/v; n = 3). (c) SEM images of the hydrogels. (d) Rheological properties of Ru(fmb)3HADH hydrogels: (i) time sweep, (ii) strain sweep, (iii) thixotropy, (iv) frequency sweep, (v) flow curves, and (vi) step-strain tests. (e) EDS analysis showing element distribution in Ru(fmb)3HADH hydrogels, including ruthenium. (f) TEM images of PDANPs. (g) Hydrodynamic size and zeta potential of PDANPs and PDA@IR780 (n = 3) In vitro and in vivo photothermal property The photothermal performance of the hydrogel system was evaluated both in vitro and in vivo. PDANPs contributed substantially to the PTT/photo-heating performance of the formulation. Various concentrations of PDANP suspensions were tested, and optimal conditions were determined using an 808 nm laser at different power densities. As shown in Fig. 4a, the laser power significantly influences the photothermal heating of the gel, with heating efficiency increasing as the power density rises. Nevertheless, excessively high laser powers should be avoided in practical applications to prevent thermal damage to skin and other tissues. Additionally, the photothermal performance of PDANPs reaches saturation at higher concentrations, suggesting that light absorption efficiency may decrease at elevated concentrations. Under irradiation with an 808-nm laser at a power density of 1.5 W/cm² for 240 s, a 1000 µg/mL PDANP suspension was allowed to cool naturally at 25 °C. The photothermal heating-cooling cycle was repeated, and temperature changes were monitored to generate the heating curve. After multiple cycles of photothermal heating and cooling, the hydrogel demonstrated stable and superior photothermal performance. This experimental observation indicates that the gel system undergoes repeated laser irradiation without the need for reinjection during the drug delivery intervals, enhancing its biological safety (Fig. 4a[iii]). Notably, this result also reflects the excellent photostability of PDA@IR780 within the hydrogel, as the polydopamine encapsulation effectively protects the IR780 dye from photobleaching during cyclic laser exposure. This protective role of PDA is further supported by previous studies [50, 51]. Furthermore, in vitro photothermal imaging results (Fig. 4b) revealed that at the same PDA@IR780 concentration, the Ru(fmb)3HADH-PDA@IR780 system exhibited the most pronounced photothermal effect, with the temperature increasing by 24.76 °C within 2 min of laser irradiation. These findings indicate that the Ru(fmb)3HADH-PDA@IR780 ruthenium-based hydrogel drug delivery system exhibited excellent and reliable photothermal properties. In the in vivo photothermal performance evaluation, the Ru(fmb)3HADH-PDA@IR780 group exhibited a temperature increase of 13.6 °C at the tumor site after 2 min of irradiation with the 808-nm laser at 1.0 W/cm², reaching a local temperature of 48.90 °C (Fig. 4c). The Ru(fmb)3HADH-PDA@IR780 group demonstrated superior photothermal performance compared to that of the PDA@IR780 group both in vitro and in vivo. Fig. 4Photothermal conversion evaluation. (a) Photothermal conversion capacity of PDANPs: (i) Temperature rise curves under 808 nm NIR irradiation at different power intensities (1.5, 2.0, 2.5, 3.0 W/cm2), (ii) Temperature rise curves under 808-nm NIR irradiation at various PDANPs concentration (200, 500, 750, 1000, 1500, and 2000 μg/mL), (iii) Photothermal elasticity measurement. (b) In vitro photothermal image (n = 3). (c) In vivo photothermal image (n = 3). Data are presented as mean ± standard deviation. Statistical significance was determined by one-way analysis of variance with Dunnett's post hoc test; *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001 Photothermal conversion evaluation. (a) Photothermal conversion capacity of PDANPs: (i) Temperature rise curves under 808 nm NIR irradiation at different power intensities (1.5, 2.0, 2.5, 3.0 W/cm2), (ii) Temperature rise curves under 808-nm NIR irradiation at various PDANPs concentration (200, 500, 750, 1000, 1500, and 2000 μg/mL), (iii) Photothermal elasticity measurement. (b) In vitro photothermal image (n = 3). (c) In vivo photothermal image (n = 3). Data are presented as mean ± standard deviation. Statistical significance was determined by one-way analysis of variance with Dunnett's post hoc test; *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001 Aldehyde-functionalised polypyridyl ruthenium complex reduces Cys, GSH, and GPX levels in vitro Cys and GSH play critical roles in regulating ferroptosis through a GPX4-dependent mechanism. While Cys primarily acts as a precursor for GSH synthesis, thereby mitigating ferroptosis, it also demonstrates an independent inhibitory effect that extends beyond its role in GSH biosynthesis [42]. Notably, in comparison to the control group, the Cys levels in the Free IR780 (L+) group showed a slight increase, whereas a decrease in Cys levels was observed across all other experimental groups. The Ru(fmb)3HADH (L-) group exhibited a reduction of 17.59% in Cys levels. The most significant decrease occurred in the Ru(fmb)3HADH-PDA@IR780 (L+) group, where Cys levels dropped by 78.34% (**P < 0.01). The Ru(fmb)3HADH-PDA@IR780 (L-) group also showed a remarkable reduction in Cys content (Fig. 5a[i]). A comparison of these results suggests that the light-responsive behavior of the hydrogel facilitated the dissociation of the complex, leading to the release of more Ru(fmb)3 complexes. This promoted the reaction between the aldehyde group on the complex and thiol group of Cys. The trend in total GSH levels reflected that of Cys levels. In the Free IR780 (L+) group, total GSH levels increased by 8.66%. However, the most substantial reduction was observed in the Ru(fmb)3HADH-PDA@IR780 (L+) group, where total GSH levels decreased by 94.85% (****P < 0.0001). The Ru(fmb)3HADH (L-), Ru(fmb)3HADH-PDA@IR780 (L-), and PDA@IR780 (L+) groups also induced significant reductions in total GSH levels, with decreases of 72.29%, 72.35%, and 70.76%, respectively (Fig. 5a[ii]). These results suggest that the uptake of the ruthenium complex and photodynamic effect of the photosensitizer, which generates ROS, contribute to the depletion of total GSH levels. Glutathione disulfide (GSSG), the oxidised form of GSH, serves as a key indicator of intracellular oxidative stress. An elevation in the GSSG level typically reflects an increase in oxidative stress within the cell. The dynamic changes in the GSSG/GSH ratio, particularly in the TME, are instrumental in regulating tumor initiation, progression, and therapeutic response. Elevated GSSG levels contribute to heightened oxidative stress, which can inhibit tumor growth by triggering programmed cell death mechanisms, such as apoptosis or autophagy. In contrast to the control group, where GSSG constituted only 4.77% of the total GSH, the proportion of GSSG in the Ru(fmb)3HADH-PDA@IR780 (L+) group significantly increased to 26.17% (Fig. 5a[iii]) (**P < 0.01), suggesting a pronounced oxidative stress effect in this treatment group. In comparison to the control group, the activity of GPX was significantly reduced in the other four experimental groups, with the Ru(fmb)3HADH-PDA@IR780 (L+) group showing a notable decrease of 73.88% (Fig. 5a[iv]; ****P < 0.0001). GPX enzymes are instrumental in maintaining cellular redox homeostasis by reducing hydrogen peroxide and other ROS. Tumor cells typically exhibit enhanced oxidative stress, and GPX activity mitigates this stress. However, excessively elevated GPX activity may hinder the generation of sufficient oxidative stress within tumor cells, thereby reducing their susceptibility to therapeutic interventions such as chemotherapy or radiotherapy. Remarkably, in the Ru(fmb)3HADH (L-) group, GPX activity decreased by 35.76% (**P < 0.01), which was more pronounced compared to the 33.92% reduction observed in the PDA@IR780 (L+) group and 17.75% decrease in the Free IR780 (L+) group. This experimental observation suggests that the Ru(fmb3 complex may directly inhibit GPX activity through mechanisms that are not solely dependent on its influence on GSH biosynthesis pathways. Recent Ru-based studies have also reported ferroptosis-associated mechanisms involving GSH depletion and GPX4 inactivation [52]. In vitro antitumor efficiency and biocompatibility The cytocompatibility and cytotoxicity of the hydrogel were evaluated using the Cell Counting Kit-8 (CCK-8) assay via co-incubation with L929 and B16F10 cells. Ru(fmb)3HADH demonstrated favourable biocompatibility, maintaining cell viability above 85.86% (24 h, 250 µg/mL) after 12 and 24 h of co-culture (Fig. 5b[i]). Upon light irradiation, the Ru(fmb)3HADH-PDA@IR780 hydrogel exhibited pronounced cytotoxicity against B16F10 cells, with a 12-h IC50 value of 3.35 µM for IR780 (Fig. 5b[ii]). In contrast, the IC50 values for the PDA-IR780(L-), PDA-IR780(L+), and HA-PDA-IR780(L-) groups were 15.82, 9.54, and 12.64 µM, respectively (Fig. 5c). These results suggest that the hydrogel effectively enhances the therapeutic efficacy of IR780 upon light irradiation, with superior cytotoxicity observed in the Ru(fmb)3HADH-PDA@IR780 group. In vitro ROS generation after NIR irradiation is influenced by pH-responsiveness Fluorescence microscopy analysis revealed that the Ru(fmb)3HADH-PDA@IR780 (L+) group exhibited the highest levels of ROS generation and B16F10 cell cytotoxicity (Fig. 5d). Significantly, under mild acidic conditions (pH = 5.5), both ROS production and the cytotoxic effect tested by live/dead staining of B16F10 cells were significantly more pronounced than under neutral conditions (pH = 7.4). The flow cytometry assay results were consistent with these findings (Fig. S11). This observation indicates that the Schiff base structure within the gel responds to the acidic environment, facilitating dissociation and the subsequent release of increased amounts of IR780 and Ru ions, thereby enhancing the therapeutic efficacy. Live/Dead staining further verified antitumor properties of Ru(fmb)3HADH-PDA@IR780 hydrogel The live/dead staining results further demonstrate that the Ru(fmb)3HADH-PDA@IR780 hydrogel exhibited remarkable tumor cell cytotoxicity under 808-nm NIR irradiation (Fig. 5d[ii]). Moreover, upon encapsulation within the hydrogel, IR780 is more efficiently internalised by tumor cells and, upon light activation, generates ROS. The results of the PDA@IR780 L+ group highlighted that the encapsulation of IR780 within PDANPs enabled the combination of PDT and PTT, which demonstrates enhanced cytotoxic efficacy in tumor cells. Fig. 5Molecular mechanisms at the cellular level, tumoricidal efficacy evaluated by CCK-8 assay, biocompatibility, ROS generation, and live/dead staining results. (a) Post-treatment analysis of each experimental group examining (i) cysteine levels, (ii) total glutathione content, (iii) GSH/GSSG ratio, and (iv) GPX activity. (b) Viability assessed by CCK-8 assay after co-culture of cells with hydrogels: (i) Cell viability following co-incubation of Ru(fmb)3HADH hydrogel, serially diluted from 500 to 7.5 μg/mL, with the mouse fibroblast cell line L929 for 12 and 24 h; (ii) Cell viability following co-incubation of Ru(fmb)3HADH-PDA@IR780 and Ru(fmb)3HADH gel/PDANPs/IR780 at a ratio of 500/200/100 μg/mL, with serial dilution of IR780 from 100 to 0.78 μg/mL (IR780 concentration used as reference), with the mouse melanoma cell line B16F10 for 12 and 24 h, followed by 3 min of 1.5 W/cm² 808 nm NIR irradiation. (c) Cell viability measured by the CCK-8 assay in other control experimental groups. The treatment groups were categorised into PDA@IR780 and Ru(fmb)3HADH-PDA@IR780, with IR780 concentration as the reference, serially diluted from 60 to 0.47 μg/mL. Subsequently, the groups were subjected to either dark incubation or 3 min of 1.5 W/cm² 808 nm NIR irradiation. (d) (i) Fluorescence microscopy images showing fluorescence staining (scale bar = 0.5 μm); (ii) live/dead staining. Data are presented as mean ± standard deviation. Statistical significance was determined by one-way analysis of variance with Dunnett's post hoc test; *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001 Molecular mechanisms at the cellular level, tumoricidal efficacy evaluated by CCK-8 assay, biocompatibility, ROS generation, and live/dead staining results. (a) Post-treatment analysis of each experimental group examining (i) cysteine levels, (ii) total glutathione content, (iii) GSH/GSSG ratio, and (iv) GPX activity. (b) Viability assessed by CCK-8 assay after co-culture of cells with hydrogels: (i) Cell viability following co-incubation of Ru(fmb)3HADH hydrogel, serially diluted from 500 to 7.5 μg/mL, with the mouse fibroblast cell line L929 for 12 and 24 h; (ii) Cell viability following co-incubation of Ru(fmb)3HADH-PDA@IR780 and Ru(fmb)3HADH gel/PDANPs/IR780 at a ratio of 500/200/100 μg/mL, with serial dilution of IR780 from 100 to 0.78 μg/mL (IR780 concentration used as reference), with the mouse melanoma cell line B16F10 for 12 and 24 h, followed by 3 min of 1.5 W/cm² 808 nm NIR irradiation. (c) Cell viability measured by the CCK-8 assay in other control experimental groups. The treatment groups were categorised into PDA@IR780 and Ru(fmb)3HADH-PDA@IR780, with IR780 concentration as the reference, serially diluted from 60 to 0.47 μg/mL. Subsequently, the groups were subjected to either dark incubation or 3 min of 1.5 W/cm² 808 nm NIR irradiation. (d) (i) Fluorescence microscopy images showing fluorescence staining (scale bar = 0.5 μm); (ii) live/dead staining. Data are presented as mean ± standard deviation. Statistical significance was determined by one-way analysis of variance with Dunnett's post hoc test; *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001 Affinity of Ru(fmb)3 complex to GPX4 computed by molecular docking The GSH/GPX4 axis is a central antioxidant defense pathway that suppresses lipid peroxide accumulation and ferroptosis [53], we investigated the binding interactions of Cys (Fig. 6a), GSSG (Fig. 6b), GSH (Fig. 6c), and our synthesised ruthenium complex Ru(fmb)3 (Fig. 6d) with GPX4 using molecular docking analysis, since chloride ions do not participate in the interaction between the ruthenium complex and GPX4, and are not expected to significantly affect the binding mode, the structure of Ru(fmb)3 was used for the molecular docking calculations. The docking sites for these compounds on GPX4 are illustrated. While the complexes formed by Cys and Ru(fmb)3 with GPX4 exhibited lower binding affinities compared to that of GSH, their interactions with GPX4 remain significant. Remarkably, no evident binding affinity was observed between GSSG and GPX4. Additionally, we computed the binding energies between GPX4 and three relevant upstream molecules, which inversely correlates with binding affinity. Among these molecules, Ru(fmb)3 complex showed the highest affinity for GPX4 (−8.31 kcal/mol), even higher than natural ligand GSH (−4.76). The key amino acid residues in the GPX4 active site, including Lys125, Lys127, Ile129, and Gly154, mediated the hydrophobic interactions between GSH and GPX4 (Fig. 6c[iii]). In contrast, the affinity of Cys (−3.99 kcal/mol) for GPX4 is lower than that of GSH. These interactions are primarily mediated by the key amino acids Asp6, Asp7, Phe17, Trp8, Glu16, Ala11, Ser18, Ala19, Met26, His114, Pro115 for Ru(fmb)3 complex (Fig. 6a[iii] and 6b[iii]). These findings suggest that Ru(fmb)3 complex may interact with GPX4 through a binding region distinct from that of GSH, suggesting a possible non-classical modulatory interaction, thereby influencing GPX4 activity. Cys can both serve as a substrate for GPX4 and exert antioxidant effects in the absence of GSH, as well as act as a rate-limiting compound in GSH synthesis, thereby modulating GSH levels. The aldehyde functionality of Ru(fmb)3 complex may participate in thiol-related reactions and thereby potentially affect cysteine/GSH homeostasis, thereby limiting GSH synthesis and restricting GPX4-mediated antioxidant activity. This limitation led to the excessive accumulation of lipid peroxidation (LPO), thereby promoting ferroptosis. Furthermore, Ru(fmb)3 complex showed a favorable predicted interaction with GPX4, which may contribute to the experimentally observed reduction in GPX-related antioxidant activity. Molecular docking results further suggest that the binding site of Ru(fmb)3 complex on GPX4 was distinct from the binding site of GSH, supporting the potential of Ru(fmb)3 complex as a modulator of GPX4 function. Fig. 6Comprehensive molecular docking analysis of GPX4 with Cys, GSH, GSSG, and Ru(fmb)3 complex. (a) Docking analysis of Cys-GPX4: (i) three-dimensional ribbon representation of the binding conformation, (ii) surface representation of the docking site, (iii) two-dimensional interaction diagram showing hydrogen bonding and hydrophobic interactions between Cys and key amino acid residues of GPX4. (b) Docking analysis of GSSG-GPX4: (i) three-dimensional ribbon representation of the binding conformation, (ii) surface representation of the docking site, (iii) two-dimensional interaction diagram showing the molecular interactions between GSSG and GPX4 residues. (c) Docking analysis of GSH-GPX4: (i) three-dimensional ribbon representation of the binding conformation, (ii) surface representation of the docking site, (iii) two-dimensional interaction diagram showing the interactions between GSH and GPX4 residues. (d) Docking analysis of Ru(fmb)3-GPX4: (i) three-dimensional ribbon representation of the binding conformation, (ii) surface representation of the docking site, (iii) two-dimensional interaction diagram illustrating the interactions between Ru(fmb)3 complex and GPX4 residues. (e) Comparative docking energy analysis, including (i) internal energy, (ii) torsional energy, and (iii) binding energy for Cys, GSH, GSSG, and Ru(fmb)3 complex with GPX4. The binding conformations and interaction patterns indicate the contributions of hydrogen bonding, hydrophobic interactions, and other non-covalent forces to the overall binding stability and specificity Comprehensive molecular docking analysis of GPX4 with Cys, GSH, GSSG, and Ru(fmb)3 complex. (a) Docking analysis of Cys-GPX4: (i) three-dimensional ribbon representation of the binding conformation, (ii) surface representation of the docking site, (iii) two-dimensional interaction diagram showing hydrogen bonding and hydrophobic interactions between Cys and key amino acid residues of GPX4. (b) Docking analysis of GSSG-GPX4: (i) three-dimensional ribbon representation of the binding conformation, (ii) surface representation of the docking site, (iii) two-dimensional interaction diagram showing the molecular interactions between GSSG and GPX4 residues. (c) Docking analysis of GSH-GPX4: (i) three-dimensional ribbon representation of the binding conformation, (ii) surface representation of the docking site, (iii) two-dimensional interaction diagram showing the interactions between GSH and GPX4 residues. (d) Docking analysis of Ru(fmb)3-GPX4: (i) three-dimensional ribbon representation of the binding conformation, (ii) surface representation of the docking site, (iii) two-dimensional interaction diagram illustrating the interactions between Ru(fmb)3 complex and GPX4 residues. (e) Comparative docking energy analysis, including (i) internal energy, (ii) torsional energy, and (iii) binding energy for Cys, GSH, GSSG, and Ru(fmb)3 complex with GPX4. The binding conformations and interaction patterns indicate the contributions of hydrogen bonding, hydrophobic interactions, and other non-covalent forces to the overall binding stability and specificity In vivo biosafety Verified by haematoxylin and eosin (H&E) staining of tissue section and biodistribution analysis The systemic toxicity of the designed hydrogel reactor system was evaluated via blood biomarker analysis and histological staining following subcutaneous injection in mice (Fig. 7). Liver and kidney toxicities were assessed by measuring serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and blood urea nitrogen (BUN), total bilirubin (TBIL), creatinine (CREA), and uric acid (Fig. 7a). No significant differences were noted regarding biomarker analysis between groups, suggesting that the Ru(fmb)3HADH-PDA@IR780 hydrogel is unlikely to induce severe liver or kidney toxicity when compared to that of the control group (tumor-bearing mice with no treatment). Histopathological changes in the heart, kidney, liver, lung, and spleen were examined via H&E staining (Fig. 7b). No significant pathological alterations were observed in the Ru(fmb)3HADH-PDA@IR780 group compared to those in the control group (no treatment), indicating the absence of acute toxicity, such as inflammation. Considering the relevant toxicity data from animal studies and the concentrations of IR780 and Ru(II) complex in the Ru(fmb)3HADH-PDA@IR780 hydrogel, these results support the acceptable short-term tolerability of the Ru(fmb)3HADH-PDA@IR780 hydrogel under the present intratumoral administration regimen and tested therapeutic dose, while longer-term and dose-escalation studies remain necessary to fully evaluate potential IR780- and Ru-associated toxicity. To investigate the short-term biodistribution and retention behavior of the formulation in vivo, ex vivo NIR fluorescence imaging and ICP-MS-based Ru quantification were conducted at predefined time points after administration (Fig. S10). The fluorescence signals were mainly observed at the tumor site over the monitored period. ICP-MS analysis further revealed that the Ru content in tumors was consistently higher than that in major organs from 0 to 96 h. The gradual decrease in intratumoral Ru content indicated progressive release and redistribution rather than abrupt loss of the formulation from the injection site. Meanwhile, the relatively low Ru levels in major organs, together with detectable signals in the liver and kidney, suggested limited systemic exposure and possible elimination through hepatobiliary and renal pathways. These findings provide evidence for prolonged local retention of the hydrogel system and support its role as a tumor-localized therapeutic depot. Fig. 7In vivo systemic biosafety data of Ru(fmb)3HADH-PDA@IR780. (a) Blood biochemistry profiles of the control, IR780 with laser, PDA@IR780 with laser, Ru(fmb)3HADH-PDA@IR780 with laser, and Ru(fmb)3HADH-PDA@IR780 without laser treatment groups. ALT, AST, TBIL, BUN, CREA, and uric acid levels are presented with mean ± standard deviation values (n = 3). (b) H&E staining of the heart, liver, spleen, lung, and kidney in the control, IR780 with laser, PDA@IR780 with laser, Ru(fmb)3HADH-PDA@IR780 with laser, and Ru(fmb)3HADH-PDA@IR780 without laser treatment groups. Scale bar = 200 μm. Statistical significance was determined by one-way analysis of variance with Dunnett's post hoc test; *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001 In vivo systemic biosafety data of Ru(fmb)3HADH-PDA@IR780. (a) Blood biochemistry profiles of the control, IR780 with laser, PDA@IR780 with laser, Ru(fmb)3HADH-PDA@IR780 with laser, and Ru(fmb)3HADH-PDA@IR780 without laser treatment groups. ALT, AST, TBIL, BUN, CREA, and uric acid levels are presented with mean ± standard deviation values (n = 3). (b) H&E staining of the heart, liver, spleen, lung, and kidney in the control, IR780 with laser, PDA@IR780 with laser, Ru(fmb)3HADH-PDA@IR780 with laser, and Ru(fmb)3HADH-PDA@IR780 without laser treatment groups. Scale bar = 200 μm. Statistical significance was determined by one-way analysis of variance with Dunnett's post hoc test; *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001 Combination therapy of PDT, PTT, and CDT in melanoma-bearing mice The therapeutic efficacy of different treatment strategies combined with NIR laser irradiation was evaluated in a syngeneic melanoma mouse model (Fig. 8a). Briefly, the loaded hydrogel was administered by intratumoral injection, and the tumor site was irradiated with an 808 nm NIR laser for 2 min at 1 h after injection. The same treatment procedure was repeated at 4-day intervals, and the overall therapeutic efficacy was evaluated after three treatment cycles. During the treatment period, no significant differences in body weight were observed among the groups, except on the final day of the study, when the control group exhibited a slight increase in body weight, likely due to tumor burden (**P < 0.01; Fig. 8b[i]). On Day 12, at the end of the treatment, the groups treated with Ru(fmb)3HADH-PDA@IR780 + NIR laser, Ru(fmb)3HADH-PDA@IR780 (dark), free IR780 + NIR laser, and PDA@IR780 + NIR laser all demonstrated prolonged survival compared to that of the control group, with the Ru(fmb)3HADH-PDA@IR780 + NIR laser group showing the longest survival (Fig. 8b[ii]). As shown in Fig. 8b(iii), the Ru(fmb)3HADH-PDA@IR780 + NIR laser group exhibited the most pronounced inhibition of tumor growth (****P < 0.0001), the tumor volume in this group decreased from 163.96 mm³ after the first treatment to 96.49 mm³ after three treatment cycles, corresponding to only 5.9% of the tumor volume observed in the control group on Day 12. The PDA@IR780 + NIR laser group exhibited a relative tumor volume of 37.9% compared to that of the control group, with significant inhibition (**P < 0.01). Although the Ru(fmb)3HADH-PDA@IR780 (dark) treatment group and free IR780 + NIR laser group also showed smaller tumor volumes than the control group, no significant differences were observed. Figure 8c further highlights the superior antitumor efficacy of the Ru(fmb)3HADH-PDA@IR780 + NIR laser group. This hydrogel system employs a multi-modal therapeutic approach, combining PDT, CDT, and PTT, all contributing to its overall antitumor effect. Under experimental conditions, the tri-modal combination of PDT, CDT, and PTT demonstrated superior tumor growth inhibition compared to that of other treatments. Histological analysis, including H&E and terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining, was performed to investigate the mechanisms underlying cancer cell destruction induced by the hydrogel-based treatment (Fig. 8d). The Ru(fmb)3HADH-PDA@IR780 + NIR laser group showed a notably larger necrotic area compared to that of the other groups. H&E staining revealed more disorganised and loose tumor tissue in this group, whereas TUNEL staining showed minimal apoptosis in the control group. In contrast, the Ru(fmb)3HADH-PDA@IR780 + NIR laser group showed a marked increase in apoptotic regions. In contrast, apoptosis in the free IR780 + NIR and PDA@IR780 + NIR laser groups was primarily limited to the tumor surface, with limited penetration into deeper tissues. The Ru(fmb)3HADH-PDA@IR780 (dark) treatment group also induced significant apoptosis. These findings collectively demonstrate that the Ru(fmb)3HADH-PDA@IR780 + NIR laser treatment, via the effective integration of multiple therapeutic modalities, exhibited superior antitumor efficacy. To further assess whether treatment was associated with local immune changes in the tumor tissue, tumor sections were additionally stained for CD8, CD86, and CD206. As shown in Fig. 8e, the Ru(fmb)3HADH-PDA@IR780 + laser group showed stronger CD8 and CD86 staining and weaker CD206 staining than the control and comparison groups. These observations provide histological evidence suggestive of enhanced CD8⁺ T-cel infiltration and a shift in macrophage polarization toward a pro-inflammatory phenotype, as evidenced by elevated CD86 and reduced CD206 signals. Together with the marked TUNEL-positive signals in this group, the data support that the superior antitumor activity of the complete formulation is associated with both direct tumor cell damage and a more favorable local tumor microenvironment. RNA sequencing To further elucidate the underlying mechanisms of multimodal therapy, we conducted transcriptomic sequencing of tumor tissues harvested from three groups of tumor-bearing mice: the Ru(fmb)3HADH-PDA@IR780 (L⁺) group, the Ru(fmb)3HADH (L⁻) group, and an untreated control group (Fig. 9). Compared with the control group, the Ru(fmb)3HADH-PDA@IR780 (L⁺) treatment induced marked transcriptional alterations, with 1,924 significantly upregulated and 168 significantly downregulated genes (q-value < 0.05, |log₂FC| > 1). Similarly, the Ru(fmb)3HADH (L⁻) group showed 1,062 upregulated and 150 downregulated genes relative to control group, indicating that both treatments primarily activated intracellular gene expression programs (Fig. 9a). The differences in gene regulation between the Ru(fmb)3HADH-PDA@IR780 (L⁺) group and the Ru(fmb)3HADH (L⁻) group were compared simultaneously (Fig. S12). Consistent with the findings in Fig. 5a, the transcriptomic data strongly support that the multimodal combination therapy triggers irreversible oxidative stress in tumor cells. Gene Set Enrichment Analysis (GSEA) revealed a profound collapse of the cellular antioxidant defense axis: the lipid peroxidase scavenging enzyme GPX4 was markedly downregulated, while intracellular Cys and reduced GSH levels sharply decreased, accompanied by an elevated GSSG/GSH ratio. In parallel, the lipid oxidation–driven ferroptotic pathway was robustly activated. Upregulation of ACSL4 enhanced the oxidation sensitivity of polyunsaturated fatty acids (PUFAs), and increased expression of lipoxygenases such as ALOX5 and ALOX15 accelerated PUFA peroxidation into cytotoxic lipid hydroperoxides. This “inhibited clearance + amplified substrate” synergy drove explosive accumulation of lipid ROS, effectively committing tumor cells to ferroptosis rather than apoptosis or necrosis. GO, KEGG, and GSEA enrichment analyses collectively demonstrated strong activation of innate immune response pathways (Fig. 9c). Canonical inflammatory signaling cascades, including NF-κB, TNF, and MAPK, were significantly enriched. Upregulated genes such as CSF3, IL1A, CXCL2/3, S100A9, and LCN2 indicated a potent inflammatory response and possible recruitment of immune cells—particularly neutrophils—to the tumor microenvironment. Furthermore, enrichment of the type II interferon signaling pathway, along with upregulation of the IFN-β–responsive gene ISG15, suggested activation of interferon-mediated cascades that may amplify antitumor immunity. Enrichment of the phagosome pathway further confirmed that extensive oxidative injury and cell death released damage-associated molecular patterns (DAMPs), which likely intensified inflammation and promoted immune-mediated clearance of damaged tumor cells. Fig. 8In vivo assessment of antitumor therapeutic efficacy. (a) Schematic of the melanoma C57 syngeneic mouse melanoma model and treatment protocol, outlining the specific treatment time points and methods. (b) Evaluation of tumor progression and individual mouse parameters: (i) Changes in the body weight in mice at each time point during treatment (n = 7). (ii) Survival curves for mice in each group (n = 5). (iii) Tumor volume measurements at each time point during the treatment period (n = 7). (iv) Tumor weight assessment at the end of the treatment. (c) Representative images of tumors at the conclusion of treatment for each group. (d) H&E staining and TUNEL staining of tumor tissues after treatment. TUNEL staining was performed with fluorescence labeling using DAPI and FITC. (e) Representative immunohistochemical staining of CD8, CD86, and CD206 in tumor tissues from different treatment groups. The Ru(fmb)3HADH-PDA@IR780 + laser group showed increased CD8 and CD86 signals and reduced CD206 staining relative to the control and comparison groups. Scale bar = 100 μm. Data are presented as mean ± standard deviation. Statistical significance was determined by one-way analysis of variance with Dunnett's post hoc test; *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001 In vivo assessment of antitumor therapeutic efficacy. (a) Schematic of the melanoma C57 syngeneic mouse melanoma model and treatment protocol, outlining the specific treatment time points and methods. (b) Evaluation of tumor progression and individual mouse parameters: (i) Changes in the body weight in mice at each time point during treatment (n = 7). (ii) Survival curves for mice in each group (n = 5). (iii) Tumor volume measurements at each time point during the treatment period (n = 7). (iv) Tumor weight assessment at the end of the treatment. (c) Representative images of tumors at the conclusion of treatment for each group. (d) H&E staining and TUNEL staining of tumor tissues after treatment. TUNEL staining was performed with fluorescence labeling using DAPI and FITC. (e) Representative immunohistochemical staining of CD8, CD86, and CD206 in tumor tissues from different treatment groups. The Ru(fmb)3HADH-PDA@IR780 + laser group showed increased CD8 and CD86 signals and reduced CD206 staining relative to the control and comparison groups. Scale bar = 100 μm. Data are presented as mean ± standard deviation. Statistical significance was determined by one-way analysis of variance with Dunnett's post hoc test; *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001 At the cellular structural and homeostatic levels, multimodal therapy induced a state of defensive fortification with attenuated invasiveness. Although compensatory upregulation of cytoskeletal proteins such as XIRP1/2 was observed, key mediators of tumor migration and immune evasion were suppressed. Spock1, a known promigratory factor, was significantly downregulated, while molecules involved in neuronal signaling and intercellular communication, such as NLGN1, were inhibited. Regarding DNA integrity, although stress checkpoint molecules like PLK3 were upregulated, the downtrend of DNA repair pathways and KEGG enrichment of transcriptional misregulation in cancer suggest that sustained ruthenium complex, photothermal, and oxidative stress overwhelmed genomic repair capacity, leading to irreversible damage and cell death. Collectively, the RNA-seq analysis was used to obtain an unbiased overview of treatment-associated biological responses and to complement the experimentally validated redox and cell-death assays. Rather than identifying fully validated new therapeutic targets, these data suggest that the superior antitumor efficacy of Ru(fmb)3HADH-PDA@IR780 under NIR irradiation may be associated with coordinated disruption of redox homeostasis, enhancement of lipid peroxidation-related stress, suppression of tumor aggressiveness-related programs, and activation of inflammatory and immune-response pathways. These findings provide mechanistic context for the multimodal antitumor effects of the hydrogel system and generate hypotheses for future functional validation. Discussion In summary, this study developed an injectable dual-responsive Ru(fmb)3HADH-PDA@IR780 hydrogel in which each component contributes a distinct but complementary function to the light-activated therapeutic system. The Ru(II) polypyridyl complex served not only as a structural component of the dynamic hydrogel network but also as a therapeutic module associated with intracellular redox disturbance, GSH depletion, and CDT-related oxidative stress. PDA nanoparticles primarily functioned as the photothermal transducer under 808 nm irradiation and as a carrier matrix for IR780 loading, thereby contributing to localized PTT and formulation stability. IR780 acted mainly as a NIR-responsive photosensitizing component, contributing to PDT-related ROS generation and providing an imaging-related fluorescence signal, while possibly offering an auxiliary photothermal contribution. Through these coordinated but nonredundant roles, the hydrogel integrated local retention, pH-responsive release, and single-wavelength NIR-triggered PTT/PDT/CDT-like multimodal therapy. Nevertheless, the present findings should be interpreted as preclinical evidence of short-term therapeutic efficacy and tolerability, and further studies on long-term clearance, dose optimization, and component-specific pharmacokinetics will be required before clinical translation. This study represents the first utilisation of a polypyridyl ruthenium complex with metal-based anticancer activity, synthesised through coordination chemistry, as a crosslinker for hydrogels. Unlike inert traditional crosslinkers, the coordination-synthesized polypyridyl-Ru(II) complex not only stabilizes the hydrogel network via Schiff-base reactions and hydrogen bonding, but also contributes to Ru-associated redox disturbance, GSH depletion, and ferroptosis-related lipid peroxidation. Furthermore, the loading of PDA@IR780 nanoparticles endow the hydrogel with high photothermal heating performance, enhanced the drug-loading capacity, and simultaneously promote interactions with phenolic hydroxyl groups, thereby facilitating the generation of ROS for fenton-type reactions. RNA sequencing provided pathway-level support for treatment-associated redox disruption, inflammatory signaling, and immune-related changes. Beyond direct induction of tumor cell death and suppression of invasiveness, the multimodal therapy fundamentally reshapes the tumor microenvironment, through triggering irreversible oxidative stress, driving ferroptosis via lipid peroxidation accumulation, and activating innate immune responses, and abrogates metastatic potential. This innovative system not only lays the foundation for advanced cancer treatments, but also enables the development of other ruthenium ligand-based hydrogel systems, which may be tailored to address various pathological conditions and deliver diverse therapeutic agents, offering valuable insights for future formulations in drug delivery and cancer therapy. Furthermore, this work provides a new strategy for hydrogel construction by chemically modifying therapeutically active metal-coordination complexes into hydrogel crosslinkers. In this design, the metal complex functions not only as a structural crosslinking unit within the hydrogel network but also as an effective therapeutic component. Fig. 9RNA-seq analysis of the tumor tissue. (a) Volcano plot of differentially expressed genes (DEGs): (i) Between Ru(fmb)3HADH-PDA@IR780 (L⁺) group and control group (n = 3). (ii) Between Ru(fmb)3HADH gel group and control group (n = 3). (b) Heatmap of expression levels of inflammatory factors, immune and cell skeleton regulation related genes. (c) GO enrichment and (d) KEGG enrichment analysis of key pathway: (i) Between Ru(fmb)3HADH-PDA@IR780 (L⁺) group and control group. (ii) Between Ru(fmb)3HADH gel group and control group. (e) GO enrichment chord diagram, showing gene entries with the (i) minimum and (ii) maximum logFC values. (f) GSEA enrichment analysis of cell cycle, DNA replication, immune response and inflammatory response (n = 3) RNA-seq analysis of the tumor tissue. (a) Volcano plot of differentially expressed genes (DEGs): (i) Between Ru(fmb)3HADH-PDA@IR780 (L⁺) group and control group (n = 3). (ii) Between Ru(fmb)3HADH gel group and control group (n = 3). (b) Heatmap of expression levels of inflammatory factors, immune and cell skeleton regulation related genes. (c) GO enrichment and (d) KEGG enrichment analysis of key pathway: (i) Between Ru(fmb)3HADH-PDA@IR780 (L⁺) group and control group. (ii) Between Ru(fmb)3HADH gel group and control group. (e) GO enrichment chord diagram, showing gene entries with the (i) minimum and (ii) maximum logFC values. (f) GSEA enrichment analysis of cell cycle, DNA replication, immune response and inflammatory response (n = 3) Despite the encouraging antitumor efficacy observed in this study, several limitations should be acknowledged. The current work mainly demonstrates short-term therapeutic efficacy and biosafety in a subcutaneous tumor model, whereas the long-term fate of the released Ru(II) complex, PDA nanoparticles, and IR780 after hydrogel degradation remains to be fully defined. In addition, because the antitumor activity of the system relies on 808 nm NIR irradiation, its clinical application may be more suitable for superficial, endoscopically accessible, or intraoperatively exposed tumors, while treatment of deeply located lesions would require further optimization of light-delivery strategies. From a translational perspective, the multicomponent nature of Ru(fmb)3HADH-PDA@IR780 also raises practical challenges related to scalable synthesis, batch-to-batch reproducibility, sterilization, storage stability, and standardized quality control. Future studies involving extended pharmacokinetic tracking, long-term toxicity evaluation, component-specific clearance analysis, and validation in clinically relevant tumor models will be necessary to further advance this formulation toward clinical translation. Methods Materials Dopamine hydrochloride, eugenol, hyaluronic acid sodium (Mw = 2000 kDa), adipic acid dihydrazide 1-ethyl-3-(3-(dimethylamino)propyl) carbodiimide, N-hydroxysuccinimide, DMSO, HOBt, and IR780 were purchased from Aladdin Reagent Inc. (Shanghai, China). CCK-8 was purchased from Dojindo laboratory (Kumamoto, Japan), and 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA), PBS, and 4’,6-diamidino-2-phenylindole (DAPI) were purchased from Tansole (Shanghai, China). Dulbecco’s modified Eagle medium (DMEM), foetal bovine serum (FBS), penicillin-streptomycin (100X), cell freezing solution, and polyoxymethylene (POM). Cell culture The mouse epithelioid fibroblast cell line L929 (RRID: CVCL_0462) and mouse melanoma cell line B16F10 (RRID: CVCL_0159) (ScienCell Research Laboratories, Inc., California, USA) were cultured in RPMI-1640 medium (Gibco, USA) supplemented with 10% FBS (Atlanta Biologicals, Lawrenceville, GA, USA) and 1% penicillin–streptomycin (Gibco). Synthesis of Ru(fmb)3HADH-PDA@IR780 hydrogel The detailed synthesis procedure for the Ru(fmb)3HADH-PDA@IR780 hydrogel is provided in the supplementary information. Hydrogel and nanoparticles characterization The basic properties of the hydrogel and nanoparticles—including hydrogel morphology, element distribution, rheological behavior, swelling and degradation properties, as well as nanoparticle morphology, particle size, and zeta potential—are described in detail in the Supplementary Information. In vitro cellular biocompatibility The biocompatibility of the hydrogel composites was evaluated using the CCK-8 assay in mouse fibroblast (L929) cells. The culture medium comprised RPMI-1640 supplemented with 1% (v/v) penicillin–streptomycin and 10% (v/v) FBS. L929 cells were seeded at a density of 8.0 × 10³ cells per well in 96-well plates. Cell viability was assessed after co-incubation with Ru(fmb)3HADH-PDA@IR780, composed of Ru(fmb)3HADH gel/PDANPs/IR780 at a ratio of 500/200/100 µg/mL, with IR780 concentrations serially diluted from 100 to 0.78 µg/mL. Using IR780 concentration as the reference, the mixture was incubated with the B16F10 mouse melanoma cell line under standard conditions (37 °C, 5% CO₂) for 12 and 24 h, followed by 3 min of 808 nm NIR irradiation at 1.5 W/cm². Six replicates were prepared for each concentration. Following incubation, the cells were washed three times with PBS and treated with 10% (v/v) CCK-8 reagent diluted in serum-free DMEM for 30 min. Absorbance at 450 nm was measured using a microplate reader to determine cell viability and calculate the IC50 values for IR780. In vitro cell cytotoxicity and death The antiproliferative efficacy of the synthesised hydrogel composites was assessed using the CCK-8 assay in B16F10 (murine melanoma) cells. B16F10 cells were cultured in DMEM supplemented with 1% (v/v) penicillin–streptomycin and 10% (v/v) FBS. Cells were seeded into 96-well plates at a density of 8.0 × 10³ cells per well, and cell viability was measured using the CCK-8 assay across various control and experimental groups. The treatment groups included PDA@IR780 and Ru(fmb)3HADH-PDA@IR780, with IR780 concentration as the reference, serially diluted from 60 to 0.47 µg/mL. The groups were divided into two conditions: one underwent dark incubation, whereas the other was exposed to 3 min of 808 nm NIR irradiation at 1.5 W/cm² after 24 h of co-incubation with cells. After laser exposure, cells were further incubated for 24 h under the same conditions. Subsequently, the cells were washed three times with PBS and treated with a 10% (v/v) dilution of the CCK-8 reagent diluted in serum-free DMEM for 30 min. Absorbance at 450 nm was measured using a microplate reader to assess cell viability and determine the IC50 values for IR780. Cell viability was calculated via the following formula.1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}&\rm Cell\;\;V{\rm{ia}}bility\;(\% )\; \\&= (OD{450_{Experiment}}\hbox{-}OD{450_{Blank}}) \div (OD{450_{C{\rm{ontrol}}}}\hbox{-}OD{450_{Blank}}) \\& \times 100\% \quad \quad \quad \quad\end{aligned}$$\end{document} Furthermore, cell viability was assessed after co-incubation of Ru(fmb)3HADH-PDA@IR780, formulated as a gel composed of Ru(fmb)3HADH, PDANPs, and IR780 at a ratio of 500/200/100 µg/mL. The IR780 was serially diluted, with concentrations ranging from 100 to 0.78 µg/mL, and used as the reference. The experiment was conducted using the B16F10 mouse melanoma cell line, with incubation periods of 12 and 24 h under standard conditions (37 °C, 5% CO₂), followed by a 3-min NIR irradiation at 808 nm and 1.5 W/cm². The cells were then washed thrice with PBS and treated with a 10% (v/v) dilution of the CCK-8 reagent in serum-free DMEM for 30 min. Absorbance at 450 nm was measured using a microplate reader to assess cell viability. Photothermal heating performance evaluation The in vitro photothermal heating performance of the hydrogel was evaluated through laser irradiation time-dependent temperature measurements. Aliquots of Ru(fmb)3HADH-PDA@IR780 at various concentrations (based on PDA concentration) were transferred into 15 mL transparent centrifuge tubes, with each aliquot being 2 mL. Each formulation was subjected to irradiation with an 808 nm NIR laser at a power density of 1.5 W/cm², while temperature fluctuations were meticulously monitored using an electronic temperature detector. Following the assessment of in vitro biocompatibility and cytotoxicity results, the optimal concentration was identified. Subsequently, the hydrogel at this optimal concentration (2 mL) was transferred to 15 mL transparent centrifuge tubes and irradiated with an 808-nm NIR laser at power densities of 1.5, 2.0, 2.5, and 3.0 W/cm², with temperature variations continuously recorded by the electronic temperature detector. Upon determining the optimal power and concentration, the light-on-off cycle was repeated to investigate the photothermal elasticity of the material. The in vivo photothermal heating performance of the hydrogel was evaluated through intratumoral injection assays in mice, with Ru(fmb)3HADH-PDA@IR780, and Ru(fmb)3HADH, after injection. Following the injection, the targeted region was irradiated with NIR (808 nm, 1.0 W/cm²) for 2 min. Thermal imaging of the mouse’s body surface was conducted throughout the irradiation process using a thermal camera, and the temperature fluctuations at the injection site were recorded. ROS production induced by NIR irradiation In a six-well plate, 1.0 × 105 B16F10 cells were seeded in each well and cultured for 24 h. The wells were categorised into five experimental groups: Ru(fmb)3HADH (L-), Free IR780 (L+), PDA@IR780 (L+), Ru(fmb)3HADH-PDA@IR780 at pH 7.4 (L+), Ru(fmb)3HADH-PDA@IR780 at pH 5.5 (L+) and control group CTRL(L-). Each L+ group was irradiated with a laser at 1.5 W/cm² and 808 nm for 2 min, whereas L- groups were kept in the dark. After aspirating the spent culture medium, the wells were washed with PBS. DCFH-DA was diluted to 1000X in serum-free DMEM, and 2.0 mL of this diluted solution was added to each well. The plate was subsequently incubated in a CO2 incubator for 30 min, after which the cells were washed three times with serum-free culture medium. Finally, 1.0 mL of DAPI staining solution was added to each well, and after a 10-min incubation, the wells were rinsed with PBS before being examined under a fluorescence microscope. Biosafety and systemic toxicity The biosafety and systemic toxicity of the hydrogel mixtures were assessed by performing a histological staining and blood biochemical assays. On days 0, 4, and 8, a dose of 2 mg/kg of Ru-HADH, Free IR780, PDA@IR780, Ru-HADH@IR780, and saline control was administered in situ into the tumor at a volume of 0.1 mL each. On day 12, the heart, kidneys, liver, lungs, and spleen were excised. The harvested tissues were fixed in a 4% (v/v) formaldehyde solution and subsequently processed for H&E staining. Following established protocols, the fixed tissues were stained with H&E reagent for histological evaluation. On day 12, blood samples were acquired through ocular puncture, with three samples obtained from each group. The serum concentrations of albumin, ALT, AST, and BUN were measured to assess the hepatic and renal toxicity of the hydrogel. When the tumor volume reached approximately 100 mm³, 0.1 mL of Ru(fmb)3HADH-PDA@IR780 (L+) hydrogel pre-mixed with the near-infrared fluorescent dye DiR was immediately administered by intratumoral in situ injection. At 0, 6, 12, 24, 48, and 96 h after injection, fluorescence images of the mice were acquired using a small-animal in vivo fluorescence imaging system, the images at different time points were obtained from independent representative mice. At the same time points, the heart, liver, spleen, lung, kidney, and tumor tissues were collected, and the Ru content in each tissue was quantified by ICP-MS. Cysteine level detection Cysteine content was determined using the Cysteine Content Assay Kit (D799571, Sangon Biotech, Shanghai) according to the manufacturer’s instructions. After collecting the cells, Cys was extracted. The sample and standard solutions with different concentrations were mixed with the reaction buffer containing phosphotungstic acid. The Cys in the sample and standard solutions reduce phosphotungstic acid to produce tungsten blue. Then, the cysteine content can be calculated by measuring the absorbance of tungsten blue at 600 nm. GSH and GPX4 level detection Intracellular glutathione content was measured using the Glutathione Assay Kit (S0053, Beyotime), following the manufacturer’s protocol. Cell lysates treated with a protein removal reagent were mixed with an assay cocktail containing glutathione reductase, NADPH, and DTNB. The measured GSH included both intrinsic GSH and GSH regenerated from oxidised glutathione (GSSG) via glutathione reductase. In this process, the thiol group (-SH) of GSH reacted with DTNB to produce yellow-colored TNB (2-nitro-5-thiobenzoic acid), while the mixed disulfide GSTNB (formed between GSH and TNB) was reduced by glutathione reductase, regenerating GSH and producing additional TNB. The absorbance of TNB at 405 nm was proportional to the total levels of GSH and GSSG, and the results were normalised to cell counts to ensure comparability across samples. GPX activity in cell samples was measured using the Total Glutathione Peroxidase Assay Kit (S0058, Beyotime). GPX catalyses the conversion of GSH to GSSG, and glutathione reductase reduces GSSG to GSH using NADPH. GPX activity was quantified by monitoring NADPH consumption. The content of GSH + GSSG, GSH/GSSG ratio, and GPX activity were calculated from the recorded experimental results. Protein docking analysis The protein and docking ligand structures were initially assessed using Chimera 1.13.1. Following this, the initial structures were processed with AutoDock Tools 1.5.7, where hydrogens and charges were added, and pdbqt files were generated. Molecular docking simulations were conducted using AutoDock 4.2.6 and Autogrid to calculate the binding energies [54, 55]. Subsequently, the docking results were analysed using Ligplus, where the interactions between amino acid residues and ligands were evaluated in terms of hydrophobic interactions and hydrogen bond strengths. Tumor xenograft and in vivo antitumor efficacy The potential of the hydrogel system to inhibit tumor growth was investigated by using the B16F10 tumor mouse model. A suspension of B16F10 cells (1.0 × 105 cells in 0.1 mL) was injected into the right hind limb of 6-week-old male C57 mice (Hangzhou Ziyuan Experimental Animal Technology Co., ltd, Zhejiang, China). When the tumor volume reached approximately 100 mm3, all mice were divided into five groups: Ru(fmb)3HADH-PDA@IR780 (L+), Ru(fmb)3HADH-PDA@IR780 (L-), PDA@IR780 (L+), Free IR780 (L+), and a control group (n = 5). Each mouse received an intratumoral injection of 0.1 mL at a dosage equivalent to 2 mg/kg body weight into the tumor on days 0, 4, and 8, followed by exposure to near-infrared laser (808 nm, 1.5 W/cm², 3 min). Starting from day 0, tumor volume and mouse body weight were measured every 2 d. On day 12, all tumor tissues were excised and fixed in a 4% (v/v) formaldehyde solution. H&E reagent was added to the tissue samples according to standard protocols, while additional samples were treated with terminal deoxynucleotidyl TUNEL assay reagents. Fluorescein isothiocyanate and DAPI were used for color development in TUNEL staining for assessing apoptosis in tumor tissue by DNA fragmentation staining. The tumor volume was calculated using the following formula. CD8, CD86, and CD206 immunohistochemical staining were performed on tumor tissues to evaluate the extent of tumor-infiltrating immune cells.2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\rm V\; = \;{L_l} \times {L_w} \times {L_w} \div 2\quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad \quad$$\end{document} Ll: tumor length, Lw: tumor width. Statistical analysis Statistical analyses were conducted using GraphPad Prism, version 8.4.2. Data are presented as the mean ± standard deviation. Comparisons between two groups were performed using a t-test, whereas one-way analysis of variance was applied for comparisons involving more than two groups. A P-value of < 0.05 was considered statistically significant. P values are indicated on the graphs as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, and ns means nonsignificant. Supplementary Information Supplementary Material 1. Supplementary Material 1.