To construct HMoO@SA@COS nanomachines, MoO nanospheres ( ~ 200 nm in diameter) were synthesized via a citrate-assisted hydrothermal method (Fig. 2a). Active hydrogen atoms were subsequently loaded into MoO nanospheres following established protocol. The hydrogen-loaded MoO (HMoO) nanoparticles retained spherical structure of the original MoO nanospheres (Fig. 2b). High-resolution TEM (HRTEM) images revealed a significant increase in lattice spacing of MoO from 0.2 nm to 0.27 nm upon hydrogen loading (Fig. 2a inset, Fig. 2b inset, Supplementary Fig. 1). Additionally, the aqueous dispersion of MoO changed from white to black upon hydrogenation, indicating successfully hydrogen doping (Fig. 2a inset, Fig. 2b inset). To enhance biocompatibility, the HMoO nanospheres were coated with sodium alginate (SA) and chitosan oligosaccharide (COS) to construct the HMoO@SA@COS nanomaterials. TEM images confirmed the successful coating of the biopolymer on the surface of HMoO nanospheres (Fig. 2c, Supplementary Fig. 2a). The significant increase in hydrodynamic size from HMoO to HMoO@SA@COS can be primarily attributed to the thickness of the polymer coating, the extension of polymer chains, electrostatic hydration-induced swelling, and the moderate assembly of the HMoO@SA@COS nanomachines. Zeta potential analysis showed that hydrogenation had minimal effect on the MoO component's potential, suggesting that protons and electrons were introduced in pairs. Following modification with negatively charged sodium alginate, the uptake of the nanomaterial by negatively charged intestinal epithelial cells was impeded. To overcome this limitation, the surface was subsequently functionalized with positively charged chitosan oligosaccharide, facilitating enhanced cellular uptake (Supplementary Fig. 2b, c). Meanwhile, to preserve the intrinsic advantages of biocompatibility and minimized non-specific interactions, which in turn reduce toxicity risks compared with strongly positively charged particles, a slightly negatively charged HMoO@SA@COS (-6 mV) was selected as the active hydrogen delivery nanomachine for this study.

X-ray powder diffraction (XRD) confirmed the successful synthesis of highly crystalline MoO. Upon hydrogenation, the diffraction peaks matched those of hydrogen-doped molybdenum oxides (HMoO, PDF#35-0604; HMoO, PDF#36-0605). Notably, the MoO (020) peak exhibited a slight blue shift and peak splitting, attributed to lattice expansion induced by hydrogen insertion (Fig. 2d). X-ray photoelectron spectroscopy (XPS) further revealed partial reduction of MoO, with the appearance of Mo and Mo states (Fig. 2e, Supplementary Fig. 3a). Hydrogenation also shifted the MoO valence band toward the Fermi level, indicating a modification in the electronic energy levels (Supplementary Fig. 3b). The O 1 s spectrum confirmed Mo-O-H bond formation, validating HMoO fabrication (Fig. 2f). Solid-state ¹H nuclear magnetic resonance (NMR) spectroscopy further verified hydrogen incorporation and release. Hydrogenated HMoO exhibited a sharp hydrogen peak, which disappeared after complete reaction with methylene blue (MB), returning to a signal level similar to pristine MoO (Supplementary Fig. 7a).

Hydrogenation significantly enhanced the optical absorption of HMoO nanospheres beyond 500 nm, while the SA and COS coating did not affect their absorbance (Fig. 2g). Fourier transform infrared (FTIR) analysis confirmed the successful surface modification of HMoO nanospheres with SA, as evidenced by the characteristic peaks at 1629 cm, 1416 cm, and 998 cm, corresponding to the asymmetric stretching, symmetric stretching, and stretching vibrations of -COOH group, respectively, in the FTIR spectrum of HMoO@SA@COSs (Fig. 2h). Stepwise changes in zeta potential demonstrated the successful layer-by-layer functionalization of HMoO nanospheres with SA and COS (Supplementary Fig. 2b, c). Additionally, the content of SA and COS in HMoO@SA@COSs was approximately 22%, as determined by thermogravimetric analysis (TGA) (Fig. 2i). For oral administration in the treatment of intestinal radiation injury, stability studies in varying gastrointestinal pH environments are critical. HMoO@SA@COSs demonstrated excellent colloidal stability without aggregation in PBS and cell culture medium (pH 7.4) (Supplementary Fig. 4). Crucially, it also demonstrated strong acid resistance in PBS at pH 1.2. In contrast, gradual degradation of HMoO@SA@COSs occurred over 24 h in pH 7.4 or under intestinal mucosa conditions (pH 6.8) due to the formation of MoO in alkaline conditions (Supplementary Fig. 5, 6). This pH-responsive degradation property prevents long-term accumulation of the nanomachine in the body, ensuring superior biosafety in biomedical applications.

MoO nanospheres, with their typical layered structures, are capable of loading active hydrogens. Due to the low reduction potential (H/H· (active hydrogen), -0.34 V vs NHE), these active hydrogens efficiently scavenge various cytotoxic RONS, including ·OH, HO, ·O, and ONOO (Fig. 3a, b, Supplementary Fig. 7b, Supplementary Table 1). To confirm the presence and reductive capacity of active hydrogens, methylene blue (MB) was employed as a probe, as it can be reduced to colorless leuco-methylene blue (LMB) upon receiving electrons and re-oxidized to blue when exposed to HO (Fig. 3c). Using MB and CoQ as reagents, the active hydrogen content in HMoO@SA@COSs was quantified as 1.59 µmol/g (Supplementary Fig. 7c-j). Notably, H alone does not induce this color change in MB, but the introduction of Pt nanoparticles (Pt NPs) catalyzes the reduction of MB to LMB due to the active hydrogens generated by Pt NPs. In this study, we observed that HMoO@SA@COSs rapidly reduced MB to LMB, causing a color change from dark blue to black (Fig. 3d, Supplementary video 1). Upon adding HO, the solution gradually reverted to blue, indicating that HMoO@SA@COSs can store and release active hydrogens, displaying reductive activity similar to the H and Pt NPs catalytic system. The reversible changes of MB-LMB-MB transitions observed in UV-vis-NIR absorption spectra further confirmed the reductive capabilities of HMoO@SA@COSs (Fig. 3e). To verify that MB reduction is driven by active hydrogen rather than Mo valence state transitions, a defective MoO (containing Mo(IV) and Mo(V)) was synthesized (Supplementary Fig. 7k) and evaluated in MB decolorization experiments. MoO exhibited no significant impact on MB absorbance (Supplementary Fig. 7l), indicating that Mo oxidation state transitions do not contribute to MB reduction, further substantiating the critical role of active hydrogen in this process. Additionally, the MB probe was used to assess the reductive activity of HMoO@SA@COSs under simulated gastric fluid (SGF) for 2 h and simulated intestinal fluid (SIF) for 8 h. As anticipated, HMoO@SAs maintained strong reductive activity after 2 h in SGF and 8 h in SIF (Supplementary Fig. 8).

HMoO@SA@COSs demonstrate superior and sustained reducibility, effectively scavenging RONS such as ·OH, HO, ·O¯, and ONOO (Fig. 3f-i, Supplementary Fig. 9-12). The ESR results further confirmed the scavenging of ·OH and ·O¯ (Supplementary Fig. 6m, n). Furthermore, ESR spectroscopy revealed disappearance of Mo(V) during this process. This is due to the formation of Mo-OH bonds upon hydrogen insertion, where the electron-withdrawing effect of hydroxyl groups reduces the oxidation state of Mo. After full reaction with RONS, hydrogen removal led to the restoration of the Mo oxidation state. The change in the Mo oxidation state follows a similar trend to that of hydrogen, suggesting that the Mo valence state change is induced by the insertion and release of hydrogen (Supplementary Fig. 6o). Additionally, HMoO@SA@COSs exhibit a strong total antioxidant capacity, as evidenced by the ABTS assay (Fig. 3j, Supplementary Fig. 13). Radar charts comparing the antioxidant capabilities of HMoO@SA@COS and MoO highlight the broad-spectrum RONS scavenging ability of HMoO@SA@COSs (Fig. 3k). Moreover, the absence of detectable hydrogen gas in the HMoO@SA system in PBS indicates that its effect is attributed to active hydrogen (Supplementary Fig. 14). The decolorization ability of HMoO@SA@COSs on MB was monitored under different temperatures (25 °C and -20 °C) over a continuous period of 9 days (Supplementary Fig. 15). These results indicate that HMoO@SA@COSs retain approximately 80% of its reductive activity after 9 days at both temperatures. Overall, the excellent, robust and long-acting reducibility of HMoO@SA@COSs endows it with superior RONS scavenging capability, showing great promise for mitigating oxidative stress.

The bandgap of HMoO@SA@COSs, as calculated by the Tauc method, was significantly reduced from 2.55 eV to 0.52 eV after hydrogen doping, indicating enhanced light absorption in the NIR region (Fig. 4a). These findings were highly consistent with the UV-vis-NIR spectra of samples before and after hydrogen doping (Fig. 2g). Upon exposure to 808 nm light irradiation (0.2 mg/mL, 1 W·cm), the temperature of the HMoO@SA@COS aqueous dispersion increased by 32 °C within 10 minutes, showing excellent photothermal properties (Fig. 4b). The photothermal properties of HMoO@SA@COSs exhibited notable dependence on concentration and light power, akin to other photothermal agents (Supplementary Fig. 16). The photothermal conversion efficiency of HMoO@SA@COSs, calculated from the cooling phase of the heating-cooling curve, was approximately 46.5% (Fig. 4c). Moreover, the photothermal performance of HMoO@SA@COSs remained stable after five heating-cooling cycles, highlighting its outstanding photothermal stability (Fig. 4d). The tissue penetration depth of NIR light was assessed by infrared thermal imaging with varying thicknesses of chicken breast tissue as barriers (Fig. 4e). Previous studies have proved that NIR light, within the 650 - 950 nm wavelength range, can penetrate up to approximately 10 mm into tissue. Consistent with this, exposure to 808 nm light results in a temperature increase of approximately 10 °C through a 5 mm thick chicken breast barrier (Fig. 4f, g), confirming that 808 nm NIR light possesses excellent tissue penetration, suitable for in vivo biomedical applications.

Considering the excellent photothermal performance of HMoO@SA@COSs, the in vitro motility driven by 808 nm light was analyzed. Finite element simulations revealed that exposure to 808 nm light induced the formation of an asymmetric thermal field across the 200 nm HMoO nanosphere (Fig. 4h). This manifests that, in addition to asymmetric nanostructures, spherical HMoO nanoparticles can also exhibit strong propulsion via self-thermophoresis (Fig. 4i). The directional motion of HMoO@SA@COSs under 808 nm light was further evaluated using MB as an indicator, with HMoO@SA@COSs injected at the bottom of a cuvette. Intriguingly, upon exposure to NIR light, HMoO@SA@COSs displayed a ballistic motion trajectory with a fast velocity of approximately 2058 μm/s (Fig. 4j, Supplementary Video 2, Supplementary Fig. 17). After 6 min of NIR irradiation, the blue color of the system was completely decolorized, primarily due to the enhanced reduction rate of MB molecules by the active hydrogens released through the nanomachines' movement. This motion not only enables HMoO@SA@COSs to overcome intestinal barriers but also significantly accelerates RONS scavenging. The motility of HMoO@SAs under 808 nm light was further investigated in solution with different viscosities, with agarose gels used to simulate intestinal mucus. The velocity (mm/s) of HMoO@SA@COS nanomachines in agarose gels is inversely proportional to the gel viscosity (Supplementary Fig. 18). The viscosity-concentration relationship is illustrated in Fig. 4k, with reported small intestine mucosal viscosity ranging from 10 to 19 mPa·s. Notably, in agarose gel with a viscosity of 26.5 mPa·s, HMoO@SA@COSs exhibited excellent directional motion, covering a distance of approximately 23 mm in 3 min under NIR irradiation (Fig. 4l, Supplementary Fig. 19, Supplementary Video 3). This robust mechanical mobility allows HMoO@SA@COSs to effectively traverse intestinal mucus barriers and deliver active hydrogens to hard-to-reach injury sites. In summary, the simple spherical HMoO@SA@COS nanomachines demonstrate outstanding NIR-driven motility, comparable to that of reported Janus self-thermophoresis-powered micro/nanomotors, which significantly enhances their potential for practical applications. As a conceptual validation, an 808 nm laser, widely used in the literature, was employed for the experiments. Given that HMoO@SA@COSs exhibit broad absorption in the near-infrared region, to further explore the application potential of HMoO@SA@COSs in deeper tissues, 1064 nm and 1550 nm lasers were used for the driving experiments of HMoO@SA@COSs. The results showed that HMoO@SA@COSs could be driven to move rapidly by 1064 nm and 1550 nm lasers, further demonstrating their potential for applications in deep tissue (Supplementary Fig. 20, Supplementary Video 4).

The HMoO@SA@COS nanomachines exhibit remarkable RONS scavenging capacity and exceptional motility, positioning them as promising candidates for intestinal radioprotection. To further investigate their radioprotective effects and underlying mechanisms at the cellular level, we assessed their RONS scavenging ability, inhibition of cell apoptosis, protection against DNA damage, and modulation of macrophage polarization (Fig. 5a). Cytotoxicity was evaluated using the CCK-8 assay, where HMoO@SA@COSs presented excellent biocompatibility across three cell lines-HIEC-6, MODE-K, and RAW264.7 with cell viability exceeding 90% even at high concentrations (Fig. 5b). Notably, HMoO@SA@COSs provided significant radioprotection against acute radiation-induced damage in HIEC-6 cells (Fig. 5c), as evidenced by CCK-8 assay results. After radiation exposure to 4 Gy or 10 Gy, cell viability was significantly decreased, while cells treated with HMoO@SA@COSs exhibited higher viability, indicating effective protection against radiation-induced apoptosis. Additionally, the NIR-light-propelled motion of HMoO@SA@COSs enhanced their ability to penetrate cell membranes, leading to enhanced cellular internalization (Fig. 5d, g, Supplementary Fig. 21). Quantitative internalization analysis via ICP-MS revealed a 2.3-fold increase in the cellular uptake due to the nanomachine's motility (Supplementary Fig. 22). This enhanced internalization would significantly improve RONS scavenging and provide greater protection against cell apoptosis.

The intracellular RONS scavenging capability of HMoO@SA@COS was assessed using the DCFH-DA fluorescence probe (Fig. 5e, h). After 4 Gy radiation exposure, intracellular RONS levels, indicated by green fluorescence signals, were significantly increased. In contrast, pretreatment with HMoO@SA@COS significantly reduced the fluorescence signals, demonstrating its effectiveness in scavenging overexpressed RONS. Notably, the fluorescence intensity in the IR+HMoO@SA@COS+NIR group was significantly lower than that in the IR+HMoO@SA@COS group, approaching that of control group. This indicates that NIR light-powered movement of HMoO@SA@COS enhances cellular internalization, thereby improving RONS scavenging efficiency. Additionally, as radiation induces oxidative stress in cells, it can lead to various forms of cell death, such as apoptosis and ferroptosis. Annexin Ⅴ-FITC/PI flow cytometry results showed that HMoO@SA@COS significantly reduced radiation-induced apoptosis, with the IR+ HMoO@SA@COS + NIR group exhibiting superior inhibition of cell apoptosis compared to the IR+HMoO@SA@COS group (Fig. 5f, i, Supplementary Fig. 23). By detecting lipid peroxidation markers associated with ferroptosis, such as MDA and 4-HNE, it was found that radiation caused a significant increase in these markers, while the levels in the material-treated group were similar to the control group (Supplementary Fig. 24a, b). This suggests that HMoO@SA@COS may have a potential effect in alleviating radiation-induced ferroptosis. Analysis of cell sections from the control, IR, and IR+HMoO@SA@COS groups revealed significant vacuolization and damage in the mitochondria of IR-treated cells, while mitochondria in the IR+HMoO@SA@COS group remained relatively intact, further supporting the notion that radiation induced ferroptosis, and HMoO@SA@COS can mitigate this effect (Supplementary Fig. 24c). The γ-H2AX immunofluorescence staining results at 4 Gy (Fig. 5j, l) doses confirmed that HMoO@SA@COS significantly mitigates DNA damage, with the IR+HMoO@SA@COS+NIR group offering the most effective protection against DNA damage.

Surprisingly, we discovered that active hydrogens could promote macrophages polarization from the pro-inflammatory M1 phenotype to the anti-inflammatory M2 phenotype, a crucial mechanism for reducing inflammation in radiation enteritis therapy. Immunofluorescence staining of CD86 (an M1 marker) and CD206 (an M2 marker) was performed to assess macrophage phenotype (Fig. 5k, m, n). After radiation exposure, strong CD86 signals were detected, indicating that the majority of RAW264.7 cells were M1 phenotype, which correlates with elevated inflammation levels. In contrast, treatment with HMoO@SA@COSs led to a marked reduction in CD86 signals and a significant increase in CD206 signals, demonstrating that the active hydrogen released from HMoO@SA@COS s played a pivotal role in immunoregulation. Previous studies have shown that heat can induce the transition from M1 to M2 macrophages. Accordingly, NIR irradiation further reduced CD86 expression and enhanced CD206 expression in the IR+HMoO@SA@COS + NIR group. Immunofluorescence staining results of other M2 markers (CD163, Arg1) further support this conclusion (Supplementary Fig. 25). To investigate the long-term protective effects of HMoO@SA@COS, we conducted a colony formation assay. The ability to form colonies relies on the normal proliferation capacity of cells, which can reflect DNA damage and long-term protection to some extent. The experimental results showed that the colony formation ability of irradiated cells was significantly reduced, while the colony formation ability of cells protected by HMoO@SA@COS was significantly improved (Supplementary Fig. 26). This indicates that HMoO@SA@COS provides overall radiation protection, effectively alleviating radiation-induced inhibition of cell proliferation. These results demonstrate that HMoO@SA@COS nanomachines offer significant protection against radiation enteritis by combining their excellent RONS scavenging ability, potent immunoregulatory effects, and movement-enhanced cellular internalization to reduce oxidative stress, apoptosis, and radiation-induced DNA damage.

Figure 4 illustrates the excellent photothermal properties and in vitro directional motility of HMoO@SA@COS nanomachines upon exposure to NIR light. To further explore their in vivo motion enhanced penetration, we conducted studies in C57BL/6 J mice. One hour after the oral administration of Cy5-labeled HMoO@SA@COSs, the mice underwent three rounds of 808 nm NIR irradiation (10 min per irradiation, with 10 min intervals, at a power density of 1.0 W·cm). At various time points, intestines were extracted and imaged using an in vivo imaging system. The results indicated that NIR-treated mice exhibited significantly higher intestinal retention of HMoO@SA@COSs compared to those without NIR exposure (Fig. 6a). To quantify the NIR-enhanced retention of HMoO@SA@COS, more detailed experiments were conducted. After the mice orally received HMoO@SA@COSs for 2 hours, NIR was applied, and fluorescence images of the mouse intestine were captured at 2, 4, and 8 hours (n = 4). Total fluorescence intensity in the small intestine, quantified using imaging software, revealed significantly higher fluorescence signals in the 4 hour and 8 hour +NIR groups compared to the corresponding -NIR groups, further confirming the role of NIR in enhancing the retention of HMoO@SA@COS (Supplementary Fig. 27). HMoO@SA@COSs retained in the small intestine for up to 8 hours under NIR irradiation, with minimal accumulation in other organs (Supplementary Fig. 28). After thorough rinsing with HEPES buffer, partial intestinal segments were imaged, demonstrating a marked increase in fluorescence intensity in the NIR-treated group compared to the untreated group (Fig. 6b, Supplementary Fig. 29). These observations confirm that NIR irradiation significantly enhanced retention within the intestines. Moreover, frozen sections of the small intestine, obtained 8 hours post-administration, further corroborated these findings, showing notably stronger fluorescence in the NIR group compared to the untreated group. (Fig. 6c). These findings indicate that the directional motility of HMoO@SA@COSs under NIR stimulation effectively prolongs intestinal retention, thereby enhancing the therapeutic efficacy of active hydrogens in radiation enteritis treatment.

To verify whether 808 nm light could propel HMoO@SA@COS nanomachines in the intestine, we monitored temperature changes in the abdominal area of mice using an infrared thermal imaging camera at 1 hour after oral administration of HMoO@SA@COSs. As expected, the temperature in the HMoO@SA@COS group was obviously higher than that of the PBS group, demonstrating that 808 nm light was able to penetrate body tissue and was absorbed by HMoO@SA@COSs in the gut (Fig. 6d, e). The maximum temperature in the HMoO@SA@COS group reached approximately 42.5 °C, while the PBS group showed about 35.0 °C, showing a notable temperature difference of 7.5 °C (Fig. 6f). This result demonstrates that the externally applied 808 nm light effectively propelled the HMoO@SA@COS nanomachines through the intestinal mucus barrier, facilitating the targeted delivery of active hydrogens to otherwise inaccessible injury sites.

The penetration of HMoO@SA@COSs into ex vivo intestinal tissue was investigated. HMoO@SA@COS nanomachines were transferred to the intestinal surface, followed by NIR irradiation. FITC-labeled HMoO@SA@COSs and rhodamine-labeled wheat germ agglutinin were used to stain the intestinal mucus. Confocal laser scanning microscope (CLSM) images clearly revealed that the NIR-powered directional motion of HMoO@SA@COS nanomachines allowed them to successfully traverse the mucus barrier (Fig. 6g). Transmission electron microscopy (TEM) results further confirmed that these nanomachines, driven by NIR light, penetrated the mucus layer and entered intestinal epithelial cells, whereas most HMoO@SA@COSs without NIR irradiation were removed during the washing process (Fig. 6h). This unique motility presents a highly promising approach for treating enteric diseases, as it enhances sample retention in the intestine tract and facilitates overcoming the gastrointestinal mucus barrier.

During radiation therapy for gastrointestinal tumor patients, both tumor cells and normal cells may suffer radiation-induced damage. To verify the DNA damage caused by radiation, comet assays were performed on colorectal cancer cells (CT26) and intestinal epithelial cells (HIEC-6). The results showed that both CT26 cells and HIEC-6 cells formed clear comet images, indicating that radiation induces DNA damage in both cancerous and normal cells (Supplementary Fig. 30). Further analysis of immunogenic cell death (ICD) markers in HIEC-6 cells revealed that after radiation treatment, compared to the control group, HMGB1 expression was reduced in the nucleus and increased in the cytoplasm of the irradiated ICD cells. Additionally, CRT expression in the cytoplasm significantly increased (Supplementary Fig. 31a, b). The HMGB1 and CRT IHC detection in irradiated mouse small intestine tissue further strengthened this observation (Supplementary Fig. 31c, d). These results suggest that during radiation therapy, radiation not only induces DNA damage in normal cells but also triggers immunogenic cell death (ICD), which could contribute to the severe side effects of radiation on normal tissues. Given the superior radioprotection effects, prolonged intestinal retention, and enhanced mucus barrier penetration capabilities of HMoO@SA@COS nanomachines, their protective efficacy against acute radiation enteritis in C57BL/6 J mice was investigated. To assess the oral safety of HMoO@SA@COSs, mice were administered 40 mg/kg daily for three consecutive days. During a 14 day observation period, the mice body weight increased similarly to the control group (Supplementary Fig. 32). Histological evaluation of major organs and blood tests on day 14 showed no significant changes, indicating that oral administration of HMoO@SA@COSs did not adversely affect hematology, liver, or kidney function (Supplementary Fig. 33, 34). A lethal dose of 16 Gy γ-ray total abdominal irradiation (TAI) was employed to evaluate the mice survival rate, while 12 Gy γ-ray TAI was used to model acute intestine injury. C57BL/6 J mice were divided into four groups: control, IR, IR+HMoO@SA@COS, and IR+HMoO@SA@COS+NIR, with 10 mice per group. Mice received 40 mg/kg HMoO@SA@COS orally before γ-ray irradiation, followed by three days of treatment. NIR irradiation was administered 1 h post-treatment for 10 minutes per cycle, with three cycles total (Fig. 7a). As shown in Fig. 7b, the IR group exhibited a 30-day survival rate of 10%, while the IR+HMoO@SA@COS group achieved a significantly higher survival rate of 46.5%. The IR+HMoO@SA@COS+NIR group demonstrated the best radioprotective effects, with an 80% survival rate, consistent with in vitro results. Body weight records revealed a sharp decline in the IR group, while the IR+HMoO@SA@COS and IR+HMoO@SA@COS+NIR groups exhibited sustained weight gain from day 6 onwards, indicating the onset of recovery in treated mice (Fig. 7c).

We systematically investigated the radioprotective effects of HMoO@SA@COSs on acute radiation-induced injury in mice exposed to 12 Gy γ-ray TAI. Colon length analysis served as a marker of intestinal inflammation, where the IR group exhibited significant colon shortening compared to the control group. In contrast, colon length was restored in the HMoO@SA@COS-treated group, with the best results observed under NIR irradiation, suggesting that NIR-mediated motion plays a critical role (Fig. 7d, e). H&E staining analysis of small intestinal tissues further revealed that the IR group experienced villous atrophy, structural disorganization, and significant inflammatory cell infiltration (Fig. 7f-h). However, the IR+HMoO@SA@COS and IR+HMoO@SA@COS + NIR groups' intestines displayed preserved villous architecture, minimal mucosal damage, and nearly intact villi with organized crypt structures. Additionally, the number and depth of crypts were well restored, demonstrating strong intestinal protection. Antioxidative and anti-inflammatory properties of HMoO@SA@COSs in vivo were also investigated. Superoxide dismutase (SOD) activity and malondialdehyde (MDA) levels were used as indicators of oxidative stress. In the IR group, a marked decrease in antioxidative capacity was observed, while treatment with HMoO@SA@COSs notably restored SOD activity and reduced MDA levels due to its robust RONS scavenging and immunomodulatory effects (Fig. 7i, j). To further investigate the antioxidant environment in the intestine, ELISA kits were used to detect the oxidative stress biomarkers 3-NT and 4-HNE in the small intestine and large intestine. The results showed that IR increased the levels of 3-NT and 4-HNE in both the small intestine and large intestine, whereas treatment with HMoO@SA@COS resulted in a decrease in the expression of 3-NT and 4-HNE, further supporting the improvement of the antioxidant environment in the intestine (Supplementary Fig. 35). Enzyme-linked immunosorbent assays (ELISA) results indicated that the treatment with HMoO@SA@COSs remarkably reduced the expression of inflammatory cytokines TNF-α and IL-6 in radiation-damaged intestinal tissues (Fig. 7k, l). The IR+HMoO@SA@COS + NIR group showed superior antioxidative and anti-inflammatory effects compared to the IR+HMoO@SA@COS group, attributed to the NIR-driven motion enhancing the precise delivery of active hydrogens to injury sites. In summary, HMoO@SA@COS nanomachines possessed excellent therapeutic effects on acute radiation-induced intestine injury in C57BL/6 J mice. The NIR-enhanced motility further amplified their efficacy, providing a promising drug-free strategy for treating radiation enteritis by leveraging active hydrogens and precision delivery of therapeutics.

Immunohistochemistry (IHC) and immunofluorescence (IF) analyses were performed on mouse small intestine tissues to evaluate the radioprotective effects of HMoO@SA@COSs, focusing on antioxidant stress response, DNA damage resistance, and macrophage polarization. Nrf2, a key regulator of the antioxidant pathway, showed low expression in the control group. After radiation exposure, Nrf2 levels showed a slight increase, indicating an oxidative stress response. In the IR+HMoO@SA@COS and IR+HMoO@SA@COS+NIR groups, Nrf2 expression was significantly elevated, suggesting that HMoO@SA@COSs further activated the Nrf2 pathway to combat radiation-induced oxidative stress. Severe DNA damage was observed in the IR group, as indicated by high levels of γ-H2AX, whereas in the IR+HMoO@SA@COS and IR+HMoO@SA@COS + NIR groups, γ-H2AX levels were significantly reduced, likely due to the broad-spectrum RONS scavenging capabilities of HMoO@SA@COSs, which mitigated RONS-induced DNA damage. Immunofluorescence staining revealed an increase in iNOS signals, a marker for the pro-inflammatory M1 macrophage phenotype, in the IR group. In contrast, iNOS expression was significantly reduced in the IR+HMoO@SA@COS and IR+HMoO@SA@COS+NIR groups. Moreover, the introduction of HMoO@SA@COSs significantly upregulated CD206, a marker of the anti-inflammatory M2 macrophage phenotype. The IR+HMoO@SA@COS+NIR group exhibited the highest CD206 expression and the lowest iNOS signal, indicating that HMoO@SA@COSs promoted macrophage polarization toward the anti-inflammatory M2 phenotype, with NIR serving as a synergistic enhancer. These results are consistent with in vitro findings (Fig. 5m-q) and highlight the innovative role of active hydrogens-induced immune modulation in inflammation. In addition, the immunohistochemical results of TGF-β showed that HMoO@SA@COSs significantly downregulated the radiation-induced increase in TGF-β expression (Supplementary Fig. 36), demonstrating that HMoO@SA@COSs helps prevent intestinal tissue fibrosis. This underscores the substantial potential of HMoO@SA@COS nanomachines for radioprotection.

Immunohistochemical (IHC) staining of small intestinal tissues from acute radiation enteritis mouse models was conducted to assess intestinal recovery under different treatments. TUNEL staining results revealed that a 12 Gy high dose of γ-ray total abdominal irradiation (TAI) induced severe apoptosis of intestinal epithelial cells, whereas the application of HMoO@SA@COS nanomachines significantly inhibited cell apoptosis (Fig. 8a-b). Notably, the best radioprotective effects occurred in the IR+HMoO@SA@COS+NIR group, highly identical to the above results (Fig. 7). Periodic acid-Schiff (PAS) staining demonstrated that radiation drastically reduced the number of goblet cells, which are responsible for mucin secretion in the intestinal villi. However, treatments with HMoO@SA@COS or HMoO@SA@COS+NIR significantly restored goblet cell counts, promoting mucin secretion and protecting the epithelial layer (Fig. 8c-d). Goblet cells are crucial for rebuilding the intestinal mucus barrier and preventing bacterial invasion. The modest increase in goblet cells in the IR group may be related to neurotransmitter release from pain receptors, which may promote goblet cell differentiation. The significant increase in goblet cells in the IR+HMoO@SA@COS and IR+HMoO@SA@COS + NIR groups likely resulted from the modulatory effects of active hydrogens or thermal stimulation on pain receptors. Lgr5 and Ki67 markers indicated that radiation severely impaired intestinal crypts, hindering epithelial cell proliferation and differentiation (Fig. 8e-h). Treatment with HMoO@SA@COS or HMoO@SA@COS+NIR effectively restored the proliferative capacity of epithelial cells, facilitating intestinal regeneration post-injury. Remarkably, the extent of recovery in the IR+HMoO@SA@COS+NIR group was nearly comparable to that in the control group. Additionally, the increase in Paneth cell numbers further highlighted the ongoing intestinal repair process after the treatment with HMoO@SA@COSs (Fig. 8i, j). Collectively, these findings confirm the exceptional radioprotective efficacy of HMoO@SA@COS nanomachines in mitigating radiation-induced intestinal damage, underscoring their promising potential for clinical applications.

To evaluate the long-term safety and protective effects of HMoO@SA@COS on the intestine, mice that survived for more than 6 months were used for the following studies. Hematoxylin and eosin (H&E) staining of major organs showed no significant pathological damage in the HMoO@SA@COS treatment group (Supplementary Fig. 37-41), and hematological parameters were normal (Supplementary Fig. 42). This indicates that active hydrogen exposure does not pose a health risk to the mice. In addition, we collected small intestine crypt cells from long-term surviving mice in the IR group and IR+HMoO@SA@COS group for comet assay. The results showed that, even after long-term survival following irradiation, most of the crypt cells in the intestine exhibited comet tails, indicating significant DNA damage. In contrast, the proportion of crypt cells with comet tails in the HMoO@SA@COS treatment group was lower, suggesting that HMoO@SA@COS treatment alleviated DNA damage and provided long-lasting protection (Supplementary Fig. 43).

The gut microbiota plays a crucial role in maintaining intestinal health. In this study, 16S rRNA gene sequencing was utilized to examine the effects of HMoO@SA@COSs on regulating intestinal microbiota. Mice in the IR group (n = 4) exhibited reduced Chao1 and Shannon indices, indicating decreased microbial diversity. In contrast, treatment with HMoO@SA@COSs restored microbial richness and diversity (Fig. 9a, b). Principal coordinates analysis (PCoA) revealed a significant shift in microbial community structure between the IR group and healthy control group (Fig. 9c). At the phylum level (Fig. 9d, e), the IR group exhibited an increase in Proteobacteria and a reduction in Firmicutes, indicative of radiation-induced dysbiosis. Notably, the treatments with HMoO@SA@COS and HMoO@SA@COS+NIR reversed these imbalances, suggesting a regulatory effect on gut microbiota. Further analysis via Lefse identified microbial groups driving these differences. LDA scores and cladogram analysis showed a notable increase in Verrucomicrobia abundance in the IR+HMoO@SA@COS+NIR group, with Verrucomicrobia also serving as a key differential bacterium in the IR+HMoO@SA@COS group compared to the IR group (Fig. 9f, g, Supplementary Fig. 43). Additionally, a significant rise in Akkermansia, a mucin-degrading bacterium within the Verrucomicrobia phylum, was observed (Fig. 9h). Akkermansia, considered a next-generation probiotic, is known to improve metabolic disorders, reduce inflammation, and strengthen intestinal barrier function. It is hypothesized that treatments with HMoO@SA@COS and HMoO@SA@COS+NIR may promote intestinal recovery by increasing the abundance of Akkermansia, a bacterium that thrives on mucin from the intestinal mucus layer. Notably, an increase in goblet cells key producers of mucus was observed following both treatments, suggesting a potential correlation between the elevated levels of Akkermansia and the rise in goblet cell numbers. Yet, this relationship requires further experimental validation to substantiate these findings.

The HMoO@SA@COS and HMoO@SA@COS+NIR treatments significantly increased the abundance of beneficial probiotics while suppressing pathogenic bacteria (Fig. 9i, j), suggesting a positive regulatory effect on the gut microbiota. Notably, the abundance of Muribaculaceae, a family known to secrete propionate a short chain fatty acid with radioprotective properties, was enhanced, facilitating intestinal recovery post-radiation damage. Propionate has also been shown to induce goblet cell differentiation, consistent with our previous findings that demonstrated an increase in goblet cell numbers. In summary, HMoO@SA@COS nanomachines play an important role in restoring intestinal microbiota by enhancing microbial diversity, reducing pathogenic bacteria, and promoting gut health. The observed increase in beneficial bacteria such as Akkermansia significantly contributes to the radioprotection and recovery from radiation enteritis.