Abstract

   Hydrogen therapy has shown new potential in cancer treatment,
   particularly in high‐pressure and hypoxic areas, where it demonstrates
   the ability to alter the tumor microenvironment and regulate tumor
   metabolism. Hydrogen disrupts the mitochondrial function of cancer
   cells, interferes with their energy metabolism, and ultimately leads to
   energy depletion and apoptosis. In this study, a sonocatalyst (BPM), is
   designed to generate hydrogen and oxygen in situ within tumors, further
   enhancing the therapeutic efficacy. The mesocrystalline structure of
   BPM, composed of bismuth fluoride, polyoxometalates, and molybdenum
   carbide, significantly improves charge separation and electron transfer
   efficiency under ultrasound irradiation, resulting in an efficient
   water‐splitting reaction. By simultaneously generating hydrogen and
   oxygen within the tumor microenvironment and depleting glutathione, BPM
   effectively triggers oxidative stress and alleviates hypoxia, thereby
   disrupting mitochondrial function and inhibiting energy metabolism in
   cancer cells. Additionally, BPM enhances antitumor immune responses by
   promoting dendritic cell maturation, activating T lymphocytes, and
   polarizing macrophages toward the M1 phenotype, reversing the
   immunosuppressive state of the tumor microenvironment. The results
   indicate that BPM holds potential for gas‐immunotherapy combination
   treatments, offering a multifunctional strategy to improve cancer
   therapy outcomes.

   Keywords: gas therapy, glutathione consumption, immune activation,
   Schottky junction, water splitting
     __________________________________________________________________

   A sonocatalyst, BPM (Mo[2]C‐POM‐BiF[3]), is designed for in situ water
   splitting within tumors to generate H[2] and O[2], disrupting
   mitochondrial function, depleting GSH, and alleviating hypoxia. This
   enhances tumor cell apoptosis, inhibits proliferation, and activates
   antitumor immune responses by promoting dendritic cell maturation, T
   lymphocyte activation, and M1 macrophage polarization, offering a
   promising cancer therapy strategy.

   graphic file with name ADVS-12-2413519-g003.jpg

1. Introduction

   Cancer treatment has long been a significant challenge in the global
   medical field.^[ [44]^1 ^] Although traditional therapies, such as
   radiotherapy and chemotherapy, can effectively inhibit tumor growth to
   a certain extent, they are often accompanied by severe side effects and
   can lead to tumor recurrence.^[ [45]^2 ^] In recent years, hydrogen
   (H[2]) therapy has attracted increasing attention as an adjunctive
   treatment due to its notable antioxidative, anti‐inflammatory, and
   cytoprotective effects.^[ [46]^3 ^] Hydrogen molecules can selectively
   neutralize harmful reactive oxygen species (ROS) and reactive nitrogen
   species, thereby reducing oxidative stress‐induced cellular damage and
   offering a mild and effective treatment option for cancer patients.^[
   [47]^3 , [48]^4 ^] However, despite its potential advantages, the
   widespread application of hydrogen therapy in cancer treatment faces
   certain challenges, mainly due to its poor tissue permeability and
   rapid diffusion, which limit its ability to exert sustained effects in
   deep tissues.^[ [49]^3 , [50]^4 , [51]^5 ^]

   Researchers have developed various nanomaterials for hydrogen‐based
   cancer therapy, including H[2]‐carrying materials and in situ catalytic
   nanomaterials.^[ [52]^3a,d ^] H[2]‐carrying materials release H[2]
   slowly to prolong their action time in the body and enhance therapeutic
   effects. However, their H[2]‐carrying capacity is limited, and the
   bioavailability of these materials in tumors also restricts their
   development. Producing H[2] in situ within the tumor via water
   splitting is an effective approach.^[ [53]^3 , [54]^4 ^] For instance,
   platinum‐based nanomaterials and titanium oxide have been shown to
   split water to generate H[2] under light irradiation.^[ [55]^3 , [56]^4
   ^] However, since light penetration depth is limited, ultrasound, which
   has a greater penetration depth, is an ideal choice for deep tumor
   treatment.^[ [57]^3 , [58]^5 , [59]^6 ^] Lin and Ma et al. proposed
   sonocatalytic water splitting to generate hydrogen for cancer
   treatment.^[ [60]^3c ^] A Schottky heterojunction composed of platinum
   and bismuth sulfide was used for sonocatalytic water splitting, where
   H[2] production was driven by water reduction, while the resulting
   holes oxidized and consumed glutathione (GSH), disrupting the redox
   balance in the tumor microenvironment (TME).^[ [61]^7 ^] Additionally,
   He et al. reported using ZnS nanoparticles to achieve sonocatalytic
   full water splitting into hydrogen and oxygen for immunoactivation in
   deep tumors.^[ [62]^6 ^] This suggests that water splitting to generate
   H[2] and oxygen (O[2]) within the TME for cancer treatment is feasible.

   Previous studies have shown that metal carbides, especially molybdenum
   carbide (Mo[2]C), possess semi‐metallic properties and can enhance
   semiconductor catalytic efficiency by acting as noble metals.^[ [63]^8
   ^] Feng et al. encapsulated molybdenum carbide quantum dots in
   ultrathin nitrogen‐doped graphene vesicles, demonstrating superhigh
   H[2]production from pure water.^[ [64]^9 ^] Forming a composite
   structure to regulate electron transfer and band structure was found to
   improve hydrogen evolution reaction (HER) performance.^[ [65]^8a,b ^]
   Therefore, tuning the bandgap of Mo[2]C may enable effective water
   splitting. Polyoxometalates (POMs), known for their strong redox
   activity, can act as electron donors or acceptors when combined with
   semiconductor materials, enhancing photoelectrochemical performance.^[
   [66]^10 ^] Bismuth‐based materials are semiconductors with tunable
   bandgaps that provide excellent electron–hole separation and reduce the
   recombination rate of electrons and holes.^[ [67]^2 , [68]^11 ^]
   Moreover, bismuth‐based materials have shown multiple activation
   properties for cancer catalytic treatment, serving as drug delivery
   vehicles or catalysts with high biocompatibility.^[ [69]^12 ^]

   Based on this, we designed a sonocatalyst for water splitting, composed
   of a ternary heterojunction with a mesocrystalline structure formed
   from Mo[2]C, POMs, and bismuth fluoride (Mo[2]C‐POM‐BiF[3], or BPM)
   (Scheme [70]1 ). In this catalyst, bismuth fluoride (BiF[3]), with its
   high bandgap and stability, facilitates effective charge separation;
   Mo[2]C, with its excellent conductivity and narrow bandgap, accelerates
   electron transfer and H[2] evolution; and POMs provide multiple active
   sites, forming efficient charge separation and transfer pathways. Upon
   ultrasound (US) activation, sonogenerated electrons are transferred
   from the surface of BiF[3] through POM to the Fermi level of Mo[2]C via
   the Schottky junction, significantly enhancing electron transfer
   efficiency and suppressing carrier recombination. BPM can split water
   in situ within tumors to generate H[2] and O[2], while BiF[3] degrades
   by consuming GSH. BPM inhibits cellular energy metabolism, alleviates
   tumor hypoxia, and disrupts the redox balance in the TME, inducing
   tumor cell apoptosis and inhibiting proliferation. Furthermore, this
   multifunctional strategy synergistically suppresses tumor growth in
   vivo by increasing the infiltration of dendritic cells, T cells, and M1
   macrophages in the TME, while reducing the percentage of M2
   macrophages, thus triggering a strong antitumor immune response. This
   system offers an effective catalyst design strategy for in situ
   sonocatalytic water splitting in tumors for H[2]‐based immunotherapy
   activation.

Scheme 1.

   Scheme 1
   [71]Open in a new tab

   Schematic diagram of the mechanism of BPM used for sonocatalytic gas
   therapy to activate immune synergistic therapy.

2. Results and Discussion

2.1. Synthesis and Characterization of BPM

   A ternary heterojunction BiF[3]‐POM‐Mo[2]C (BPM) was synthesized via
   oxidation and precipitation methods (Figure [72]1a; Figure [73]S1,
   Supporting Information). Transmission electron microscopy (TEM)
   revealed that the diameter of POM‐Mo[2]C (PM) particles was below
   10 nm, forming a near‐spherical structure (Figure [74]S2, Supporting
   Information). The synthesized BPM nanoparticles were ≈40 nm in diameter
   (Figure [75]1b), consisting of small spheres with a mesocrystalline
   structure, which increases the ultrasonic absorption cross‐section and
   improves catalytic performance. The Brunauer−Emmett−Teller (BET)
   results showed a specific surface area of 19.262 m^2 g^−1 and an
   average pore size of 1.4 nm (Figure [76]S3, Supporting Information). In
   Figure [77]1c, high‐resolution TEM images showed that the lattice
   spacing of BiF[3] was 0.206 nm, corresponding to the (220) plane of the
   face‐centered cubic crystal phase of BiF[3], and the lattice spacing of
   Mo[2]C was 0.303 nm, corresponding to the (111) plane of the
   orthorhombic crystal phase of Mo[2]C. Red markers indicate some
   amorphous phases, likely due to the embedding of amorphous POM within
   the material. Energy dispersive X‐ray spectroscopy element mapping
   confirmed the uniform distribution of Bi, O, C, F, and Mo elements
   throughout the nanocomposite material (Figure [78]1d). X‐ray
   diffraction Rietveld refinement analysis confirmed the cell parameters
   and structure of BiF[3] and Mo[2]C (Figure [79]1e–g). The diffraction
   peaks of the (111), (200), (220), and (311) planes matched the cubic
   phase of BiF[3] (CDS No.24522), and the diffraction peaks of the (110),
   (020), (121), and (102) planes matched the orthorhombic phase of Mo[2]C
   (CDS No.164851), consistent with the lattice spacings measured in
   high‐resolution TEM.

Figure 1.

   Figure 1
   [80]Open in a new tab

   Synthesis and characterization of BPM. a) Schematic diagram of the
   synthesis process of BPM, b) TEM image, c) high‐resolution TEM image
   (marked in red circles as the amorphous phase region), d) element
   mapping image containing elements Bi, O, C, F, and Mo. The cell
   structures and parameters of e) BiF[3] and f) Mo[2]C in BPM were
   refined by the Rietveld method. g) X‐ray diffraction pattern of BPM and
   its fitting refinement results. h) Fourier transform infrared spectra
   of BPM and PM. High‐resolution XPS spectra of i) Bi 4f, j) F 1s, k) Mo
   3d, l) C 1s, and m) O 1s elements in BPM. n) Zeta potentials of Mo[2]C,
   PM, BiF[3], and BPM.

   Fourier‐transform infrared spectroscopy (Figure [81]1h) and X‐ray
   photoelectron spectroscopy (XPS) were used to analyze the elemental
   composition of BPM and deduce the structure of POM. In the
   high‐resolution Bi XPS spectrum, the binding energy peaks at 163.5 and
   158.1 eV were attributed to Bi^3+ 4f[5/2] and Bi^3+ 4f[7/2],
   respectively (Figure [82]1i). The binding energy of F^− in the 2p
   orbital was 682.6 eV (Figure [83]1j). The Mo 3d orbital signals could
   be fitted to six characteristic peaks corresponding to three valence
   states of Mo ions: Mo^2+, Mo^5+, and Mo^6+, with ratios of 31.35%,
   62.53%, and 6.12%, respectively. Mo^2+ originated from Mo[2]C, while
   Mo^5+ and Mo^6+ were mainly derived from POM, with Mo^5+ dominating
   (Figure [84]1k). The binding energy peak at 283.3 eV in the
   high‐resolution C XPS spectrum matched the characteristic peak of C–Mo,
   confirming the presence of Mo[2]C (Figure [85]1l). The two main peaks
   at 531.5 and 530.4 eV in the high‐resolution O XPS spectrum were
   attributed to oxygen vacancies and surface‐adsorbed oxygen, which
   facilitate carrier migration and inhibit carrier recombination
   (Figure [86]1m). Dynamic light scattering measurements showed that the
   size of BPM was 120 nm (Figure [87]S4, Supporting Information). Before
   composite formation, the zeta potentials of PM and BiF[3] were −30.6
   and 23.3 mV, respectively. After composite formation, the zeta
   potential of the mixture became −27.3 mV (Figure [88]1n), further
   confirming the construction of the BPM ternary heterojunction. BPM
   could be stored in slight acid phosphate buffer saline (PBS) for
   several days, conforming to its stability (Figure [89]S5, Supporting
   Information).

2.2. Water Splitting of BPM by US and GSH Consumption

   The semiconductor mechanism of BPM sonocatalytic water splitting for
   H[2] and O[2] generation was investigated. Ultraviolet photoelectron
   spectroscopy was used to measure the work function (W) and Fermi level
   (E[f]) of BiF[3], Mo[2]C, and BPM to determine the direction of charge
   carrier transfer in the heterojunction. The difference in E[f] between
   the semiconductors is the intrinsic driving force for charge transfer
   in the heterostructure. As shown in Figure [90] 2a–c, the work
   functions of BiF[3], Mo[2]C, and BPM were calculated to be 4.05, 4.79,
   and 4.32 eV, respectively, while their Fermi levels were −0.45, 0.29,
   and −0.18 eV, respectively. Mo[2]C has a larger W and lower E[f],
   whereas BiF[3] has a smaller W and higher E[f]. When BiF[3] is in close
   contact with Mo[2]C, electrons transfer from BiF[3] to Mo[2]C. Since
   previous studies have indicated that Mo[2]C exhibits pseudo‐noble metal
   properties, a space charge region forms on the BiF[3] side when BiF[3]
   and Mo[2]C are in close contact, causing band bending and forming a
   Schottky barrier. When excited by US, the electrons in BiF[3] are
   excited from the valence band (VB) to the conduction band (CB) and
   transfer through the Schottky barrier to the Fermi level of the Mo[2]C
   nanoparticles, while the holes remain in the VB of BiF[3]
   (Figure [91]2d). Importantly, the band bending effectively prevents
   electron backflow, further accelerating carrier separation in the BPM
   catalytic system.

Figure 2.

   Figure 2
   [92]Open in a new tab

   Water splitting performance and mechanism evaluation of BPM.
   Ultraviolet photoelectron spectra of a) BiF[3], b) Mo[2]C, and c) BPM.
   d) Band diagrams and sonocatalytic mechanisms of BiF[3], Mo[2]C,
   BiF[3]‐Mo[2]C, and BPM. e) Time‐dependent photocurrent response and f)
   electrochemical impedance spectroscopy of BiF[3], BiF[3]‐Mo[2]C, and
   BPM. g) The hydrogen evolution curve and h) the oxygen production curve
   of BPM with time under different US power excitation. Linear sweep
   voltammetry i) HER and j) OER of BiF[3], Mo[2]C, and BPM. k) Tafel
   plots of BiF[3], Mo[2]C, and BPM from HER linear sweep voltammetry
   curves. l) GSH consumption in different treatment groups. Student's
   t‐test, *p < 0.05, **p < 0.01, and ***p < 0.001. Data are presented as
   mean ± SD (n = 5).

   Electrochemical impedance and transient photocurrent measurements were
   used to verify carrier separation efficiency and migration capacity. As
   shown in Figure [93]2e,f, electrochemical impedance spectroscopy
   results revealed that the Nyquist semicircles represent the electron
   transfer resistance in the materials. BPM exhibited smaller Nyquist
   semicircles and higher photocurrent density compared to BiF[3]‐Mo[2]C
   and BiF[3], indicating that the formation of the Schottky junction
   successfully reduced charge transfer resistance. More electrons and
   holes could reach the reaction interface upon irradiation, generating
   higher transient photocurrent density. Notably, BPM's impedance and
   photocurrent data were superior to those of BiF[3]‐Mo[2]C, likely due
   to the ability of POM to rapidly switch between multiple valence
   states, enhancing charge transfer efficiency when POM acts as an
   electron transfer medium. Furthermore, we utilized ultrasound to excite
   BPM, BiF[3], and BiF[3]‐Mo[2]C, BPM also presented the most obvious
   sonocurrent response, verifying the ability of sono‐excited charge
   separation (Figure [94]S6, Supporting Information).

   To determine whether the sonogenerated electrons and holes had
   sufficiently high redox potential for HER and oxygen evolution reaction
   (OER), the band structures of the nanomaterials were further measured
   using Mott–Schottky and solid‐state diffuse reflectance spectroscopy
   (Figure [95]S7, Supporting Information). The bandgaps (Eg) of BiF[3]
   and Mo[2]C were 3.52 eV and 0.46 eV, respectively. The calculated band
   structures of BiF[3] and Mo[2]C (Figure [96]S8, Supporting Information)
   exhibited typical direct bandgap semiconductor and metallic
   characteristics, respectively. The theoretical bandgap of BiF[3]
   (3.81 eV) was slightly larger than the experimental value (3.52 eV).
   The theoretical bandgap of Mo[2]C was 0 eV, with the density of states
   of Mo d‐orbitals extending across the valence band and conduction band,
   allowing d‐electrons to move freely between the bands. By drawing
   tangent lines from the Mott–Schottky plots, the flat band potentials
   (E[fb]) of BiF[3] and Mo[2]C were determined to be −0.98 and −0.5 eV
   (vs Ag/AgCl, pH 7), respectively. Therefore, the CB potentials of
   BiF[3]and Mo[2]C were −0.48 and 0 eV (vs NHE, pH = 0), respectively.
   The CB potentials of BiF[3] and Mo[2]C are more negative than the redox
   potential of H^+/H[2] (0 V vs NHE, pH 0). The VB potential of BiF[3] is
   also more positive than the redox potential of H[2]O/O[2] (1.23 V vs
   NHE, pH 0). Thus, BPM meets the conditions for water splitting to
   produce H[2] and O[2]. Besides, after contact between H[2]O and BPM,
   the contact angle was only 27.4°, indicating that BPM has
   hydrophilicity and meets the conditions for water splitting (Figure
   [97]S9, Supporting Information).

   Subsequently, gas chromatography and a dissolved oxygen meter were used
   to obtain the H[2] and O[2] production of BPM. As shown in
   Figure [98]2g,h, and Figure [99]S10 (Supporting Information), H[2] was
   detected at different US power densities (1 MHz, 1, 1.5, and 2 W
   cm^−2), with H[2] production increasing with US power densities and
   excitation time. Simultaneously, O[2] was detected. Regardless of US
   power densities, the molar ratio of H[2] to O[2] was ≈2:1, indicating
   that full water splitting occurred under US. Since the tumor‐sensitive
   concentration is in the nanomolar range, the gases produced by BPM are
   sufficient to exert therapeutic effects in tumors. Linear sweep
   voltammetry was used to measure the overpotentials for HER and OER
   (Figure [100]2i,j). BPM exhibited lower overpotentials than BiF[3] and
   Mo[2]C. At a current density of 10 mA cm^−2, the overpotentials of BPM,
   BiF[3], and Mo[2]C were 530, 582, and 583 mV, respectively, indicating
   that BPM had the highest H[2] production efficiency. Additionally, BPM
   exhibited a lower Tafel slope of 148.9 mV dec^−1, further demonstrating
   its superior catalytic efficiency (Figure [101]2k). At a current
   density of 10 mA cm^−2, BPM also had the lowest OER overpotential of
   1.74 V, lower than that of Mo[2]C (1.82 V). Therefore, BPM enables
   simultaneous HER and OER under sonocatalysis, achieving water splitting
   for gas therapy.

   Furthermore, the interaction between BPM and GSH was evaluated using
   5,5′‐dithiobis‐(2‐nitrobenzoic acid) (DTNB) probes. As shown in
   Figure [102]2l, US alone did not affect GSH, while BiF[3] and BPM
   effectively consumed GSH after 30 min of incubation, indicating that
   the GSH consumption was due to the presence of BiF[3] in BPM. Under US
   activation, BPM did not further consume GSH, suggesting that all the
   holes generated by BPM under sonocatalysis were used for OER, with no
   significant effect on GSH. Therefore, GSH consumption disrupts the
   redox balance in the TME, exacerbating oxidative damage to the tumor.
   Besides, under TME conditions (pH 5.5, H[2]O[2], and GSH), BPM could
   degreed over time, presenting a biodegradable characteristic (Figure
   [103]S11, Supporting Information).

2.3. In Vitro Therapeutic Evaluation

   The biocompatibility of BPM was assessed using the cell counting kit‐8
   (CCK‐8). Mouse breast cancer (4T1) and human embryonic kidney (293T)
   cells were incubated with different doses of BPM for 24 and 48 h. At
   concentrations below 100 µg mL^−1, cell viability was not significantly
   inhibited, indicating good safety of BPM. When the concentration is
   higher than 100 µg mL^−1, cell viability in 4T1 is lower than that in
   293T cells, which could be caused by GSH consumption and the catalytic
   performance of Mo ions (Figure [104]3a; Figure [105]S12, Supporting
   Information). As shown in Figure [106]3b, irradiation of 4T1 cells with
   US alone (0.7 W cm^−2, 5 min) had minimal impact on cell viability.
   However, when 4T1 cells were incubated with BPM and then exposed to US,
   cell viability was significantly reduced, and this reduction was
   positively correlated with the concentration of BPM. This is due to
   oxidative damage caused by H[2] and GSH consumption, which, combined
   with oxygen's ability to alleviate tumor hypoxia, synergistically
   induced apoptosis in cancer cells. Additionally, we evaluated the
   viability of BiF[3] and PM under US, and their effects were weaker than
   that of BPM.

Figure 3.

   Figure 3
   [107]Open in a new tab

   Evaluation of the in vitro tumor treatment effect of BPM. a) Cell
   viability of 4T1 cells after co‐incubation with BPM for 24 and 48 h,
   respectively. b) The cytotoxicity of 4T1 cells in different treatment
   groups. c) Images of 4T1 cells stained with MB probe after treatment in
   different groups. Fluorescence images of 4T1 cells stained with d)
   DCFH‐DA and e) JC‐1 after treatment in different groups. f) The ATP
   content inside 4T1 cells in different groups after treatment.
   Immunofluorescence staining images of (g) DNA damage, h) HMGB1, and i)
   CRT after treatment in different groups. j) MB, k) DCFH‐DA, l) JC‐1, m)
   γ‐H2AX, n) HMGB1, o) CRT corresponding fluorescence quantitative data.
   Student's t‐test, *p < 0.05, **p < 0.01, and ***p < 0.001. Data are
   presented as mean ± SD (n = 5).

   To further investigate the mechanism of BPM‐induced cell death, a
   methylene blue (MB) probe was used to analyze intracellular H[2]
   release. As shown in Figure [108]3c,j, after co‐incubation with MB
   probes, the cells displayed a blue color. Compared to the control
   group, the blue color in cells treated with US, BPM alone, or BiF[3]+US
   remained largely unchanged. However, in the BPM+US treatment group, a
   significant fading of the blue color was observed in 4T1 cells,
   indicating that BPM effectively produced H[2] within cancer cells. GSH
   consumption disrupts the oxidative balance within tumor cells, leading
   to oxidative stress in the tumor. Using
   2′,7′‐dichlorodihydrofluorescein diacetate (DCFH‐DA) as a probe to
   monitor changes in intracellular ROS levels, significant ROS production
   was observed in cells treated with BPM+US (Figure [109]3d,k). Previous
   studies have shown that H[2] can effectively target mitochondria within
   cells.^[ [110]^3 , [111]^6 ^] As such, H[2] induces apoptosis in cancer
   cells by inhibiting mitochondrial function and blocking adenosine
   triphosphate (ATP) synthesis. Mitochondrial membrane potential is a
   core indicator of mitochondrial health, and apoptosis is often
   accompanied by a reduction in mitochondrial membrane potential. JC‐1, a
   mitochondrial membrane potential assay kit, was employed to measure
   changes in mitochondrial membrane potential. When mitochondrial
   membrane potential is high, JC‐1 forms red fluorescent aggregates in
   the mitochondrial matrix. Conversely, when mitochondrial membrane
   potential is low, JC‐1 exists as green fluorescent monomers. As shown
   in Figure [112]3e,l, the BPM+US group displayed significant green
   fluorescence, indicating that the generated H[2] effectively damaged
   the mitochondria of 4T1 cells, which also led to a reduction in
   intracellular ATP levels (Figure [113]3f). Furthermore, reduced ATP
   levels and elevated ROS levels within cells may also cause DNA damage.
   By detecting the expression of γ‐H2AX protein, we investigated the
   mechanism of DNA double‐strand breaks. As shown in Figure [114]3g,m,
   the BPM+US treatment group exhibited the highest fluorescence intensity
   among all groups. These findings suggest that the produced H[2] can
   induce mitochondrial and DNA damage, inhibit tumor cell proliferation,
   and lead to cell apoptosis.

   Subsequently, we examined the expression of immunogenic cell death
   (ICD) biomarkers induced by gas therapy, including calreticulin (CRT)
   and high‐mobility group box 1 (HMGB1). These biomarkers can stimulate
   dendritic cell maturation and activate cytotoxic T lymphocytes, thereby
   enhancing antitumor immune responses. The expression of CRT and HMGB1
   in 4T1 cells was visualized using CRT/HMGB1‐specific antibodies for
   immunofluorescence staining (Figure [115]3h,i,m,n). It was found that
   the HMGB1 signal was significantly reduced in the BPM+US group compared
   to other experimental groups, while the CRT signal displayed a negative
   correlation with HMGB1 expression. These data suggest that the gas
   therapy combination of oxidative damage caused by H[2] and GSH
   consumption, along with O[2] alleviating hypoxia, can trigger a certain
   level of ICD, thereby eliciting an immune response against tumor cells.
   Thus, in tumor cells, BPM can inhibit cell viability, regulate the TME
   through HER and OER, damage tumor cells, and induce ICD under US
   activation.

2.4. In Vivo Therapeutic Evaluation

   Before evaluating the in vivo therapeutic effects, the systemic
   toxicity of BPM was assessed. Even at a high concentration of 200 µg
   mL^−1, the hemolysis rate of BPM was below 5%, indicating that the BPM
   composite nanomaterials have satisfactory blood compatibility (Figure
   [116]S13, Supporting Information). Healthy female BALB/c mice were
   intravenously injected with BPM (2 mg mL^−1, 100 µL). Fourteen days
   after administration, there were no significant differences in the
   complete blood count data of the BMP group compared to the control
   group (Figure [117]S14, Supporting Information). Meanwhile, the
   histological hematoxylin and eosin (H&E) staining of the major organs
   (heart, liver, spleen, lungs, kidneys) of mice sacrificed on day 14
   also showed no detectable signs of damage or inflammatory lesions,
   indicating that BPM has high tissue compatibility.

   The treatment process is illustrated in Figure [118] 4a. To evaluate
   the uptake of BPM by the tumor and its biodistribution in various
   organs, fluorescence imaging was used (Figure [119]4b,c). IR780 (a
   near‐infrared dye) was loaded into BPM as a fluorescent indicator to
   obtain BPM‐IR780, which was injected into tumor‐bearing 4T1 mice via
   the tail vein. Twelve hours after injection, the relative distribution
   of BPM in the tumor reached its maximum, indicating that BPM
   effectively accumulated and retained in the tumor through circulation.
   Thus, 12 hours post‐injection was determined to be the optimal time for
   US treatment. The fluorescence intensity in the liver and kidneys first
   increased and then decreased over time, indicating that BPM could be
   metabolized.

Figure 4.

   Figure 4
   [120]Open in a new tab

   In vivo treatment evaluation of BPM. a) Schematic diagram of tumor
   establishment and treatment plan. After intravenous injection of
   BPM‐IR780, 4T1 tumor‐bearing mice showed fluorescence images of b) in
   vivo and c) in various organs and tumors over time. d) The weight of
   mice in different groups during treatment. The e) growth curves and f)
   average growth curves of tumor‐bearing mice in each group during the
   treatment period. g) The average weight and photos of tumors in each
   group after treatment. h) The slice images of HIF‐1α staining, Ki‐67
   staining, and H&E staining in each group of tumors. Student's t‐test,
   *p < 0.05, **p < 0.01, and ***p < 0.001. Data are presented as mean ±
   SD (n = 5).

   Based on the therapeutic effects observed in cells, we further
   investigated the antitumor activity of BPM in a 4T1 tumor‐bearing mouse
   model. A total of 32 female BALB/c mice were randomly divided into six
   groups (n = 5): (1) Control, (2) US, (3) BPM, (4) BiF[3]+US, (5) PM+US,
   and (6) BPM+US. Treatment conditions: BPM, BiF[3], and PM (2 mg mL^−1,
   100 µL) were intravenously injected into groups (3)–(6); US (0.7 W
   cm^−2, 5 min) was applied. As shown in Figure [121]4a, treatment began
   12 h after material injection on days 0 and 3, with a total of two
   treatments over 14 days. During the 14‐day treatment, there were no
   significant differences in the body weights of mice across all groups,
   indicating that the treatment did not affect body weight and had no
   obvious side effects (Figure [122]4d). The tumor volume growth curves
   (Figure [123]4e,f) showed that the tumors in the Control and US groups
   grew rapidly without being affected. BPM alone exhibited a certain
   inhibitory effect on tumor proliferation. The BiF[3]+US and PM+US
   groups also showed some inhibitory effects. In comparison, the BPM+US
   group exhibited the most significant tumor suppression. At the end of
   the treatment, the tumors were photographed and weighed
   (Figure [124]4g). Compared to the control group, the tumor masses in
   the BiF[3]+US, PM+US, and BPM+US groups were significantly lower, with
   the BPM+US group showing the lowest tumor mass. Notably, in some mice
   in the BPM+US group, the tumors had gradually regressed, further
   demonstrating the tumor‐suppressive effect of gas combination therapy.
   After treatment, H&E and Ki67 staining of tumor tissues showed that the
   BPM+US group had the highest degree of tumor cell death, the most
   significant morphological changes, and the greatest inhibition of
   proliferation. Hypoxia inducible factor 1α (HIF‐1α) staining further
   confirmed that O[2] produced by BPM through the OER effectively
   relieved hypoxia within the tumor, reducing resistance to tumor
   immunotherapy (Figure [125]4h). Therefore, BPM exhibited superior tumor
   treatment efficacy compared to BiF[3] and PM.

2.5. Evaluation of BPM‐Induced Immune Response in Tumor Cells

   Since BPM demonstrated the ability to induce ICD by triggering the
   release of various damage‐associated molecular patterns in cell
   therapy, we evaluated BPM's ability to activate immune responses after
   in vivo treatment. The proportion of mature dendritic cells (DCs) and
   activated T lymphocytes in the treated tumor tissues was quantified by
   flow cytometry. The gating strategy for flow cytometry is shown in
   Figures [126]S15–S17 (Supporting Information). As there were no
   significant differences in therapeutic effects between the Control and
   US groups in the in vivo and in vitro experiments, we only evaluated
   (1) Control, (2) BPM, (3) BiF[3]+US, (4) PM+US, and (5) BPM+US groups.
   The co‐expression of CD80^+ and CD86^+ was used as a marker for mature
   DCs. In the BPM+US group, the proportion of mature DCs was
   significantly higher than in the control group, with an increase of
   9.9% (Figure [127] 5a,b). At the same time, compared to the control
   group, the proportion of CD8^+ T cells, which directly kill tumor
   cells, increased by 16.2% (Figure [128]5a,c). Additionally, numerous
   studies have shown that tumor hypoxia is considered one of the most
   important factors in immune suppression.^[ [129]^6 , [130]^13 ^]
   Hypoxic regions within solid tumors are infiltrated by a large number
   of immunosuppressive cells, such as M2‐type tumor‐associated
   macrophages, which suppress immune responses.^[ [131]^3 , [132]^6 ,
   [133]^14 ^] Here, we found that compared to the control group, BPM+US
   treatment reduced the proportion of anti‐inflammatory CD206^+ M2
   macrophages and increased the proportion of pro‐inflammatory CD86^+ M1
   macrophages (Figure [134]5a,d,e). This demonstrated that BPM, activated
   by US, effectively relieved hypoxia in tumor sites and reversed immune
   suppression. The same conclusions were also supported by
   immunofluorescence images of CD86^+, CD8^+, and CD206^+ in tumor
   tissues from the treatment groups (Figure [135]5f–i).

Figure 5.

   Figure 5
   [136]Open in a new tab

   a) Flow cytometry result of mature DC cells, T cells, and macrophages
   in tumors, and corresponding analysis diagrams of b) mature DC cells,
   c) T cells, d) M2 type macrophages, and e) M1 type macrophages. f)
   Results of immunofluorescence staining of tumor sections, and
   corresponding average fluorescence intensity analysis diagrams of g)
   CD8 T cells, h) CD86 DC cells, and i) CD206 macrophages. (j) Flow
   cytometry result of mature DC cells in bone, lymph, and spleen, and
   corresponding analysis diagrams of mature DC cells in k) bone, l)
   lymph, and m) spleen. n) Flow cytometry result of T cells in lymph, and
   spleen, and corresponding analysis diagrams of T cells in o) lymph, and
   p) spleen. Student t‐test, *p < 0.05, **p < 0.01, ***p < 0.001. Data
   are presented as mean ± SD (n = 5).

   When ICD occurs in tumor cells, immature DCs recognize the release of
   damage‐associated molecular patterns, leading to DC maturation,
   cytotoxic T lymphocyte activation, and the subsequent induction of
   systemic immune activation to fight tumor metastasis. To evaluate this
   immune response, we assessed the maturation of DCs in the bone marrow,
   spleen, and lymph nodes. The proportion of mature DCs in the bone
   marrow, spleen, and lymph nodes in the BPM+US group was 16.4%, 37.5%,
   and 32.8%, respectively, which were significantly higher than those in
   the other groups (Figure [137]5j–m). Additionally, we analyzed the
   activation of CD8^+ T cells. As shown in Figure [138]5n–p, the
   activation rates of CD8^+ T cells in the spleen and lymph nodes in the
   synergistic treatment group were 34.7% and 52.4%, respectively, showing
   significant enhancement compared to the control group. This indicates
   that the BPM+US group achieved systemic immune activation, leading to
   the most effective tumor suppression. These results suggest that under
   US stimulation, BPM effectively kills tumor cells by producing H[2] and
   O[2], consuming endogenous GSH, and significantly activating the immune
   system through the reversal of immune suppression, thereby enhancing
   the therapeutic efficacy of gas immunotherapy.

2.6. RNA‐Seq Analysis of BPM Synergistic Tumor Treatment Mechanism

   To further explore the mechanism behind the gas‐immunotherapy killing
   effect of BPM under US stimulation, we analyzed mRNA expression in
   tumor tissues from the Control group and the BPM+US group using
   high‐throughput sequencing. A Venn diagram showed that 26 926
   co‐expressed genes were found in both the control and BPM+US groups
   (Figure [139] 6a). Compared to the control group, ≈105 genes were
   abnormally expressed in 4T1 cells in the synergistic group, including
   87 upregulated genes and 48 downregulated genes. A genomic heatmap
   illustrated the results of the differentially expressed genes (DEGs)
   (Figure [140]6b). The DEGs were identified using screening criteria
   (Log2FoldChange ≥ 1.0 or ≤ 1.0, p‐value ≤ 0.05) (Figure [141]6c).
   BPM+US treatment induced the downregulation of the
   proliferation‐promoting gene SECTM1, the DNA repair‐promoting gene Fap,
   and the tumor growth‐promoting factor Parp12, while upregulating the
   tumor necrosis factor Tnfrsf14 and the inflammation‐related gene Hspb7.
   This may be due to BPM generating substantial H[2], which can cause
   extensive DNA damage by inhibiting mitochondrial function and blocking
   adenosine ATP synthesis, thereby promoting cell apoptosis.
   Additionally, immunosuppressive chemokines Cxcl12 and Cxcl13 were
   downregulated, while the immunostimulatory chemokine Ccl11 was
   upregulated. T‐cell activation‐inducing factors Cd8a, Icos, Gzmb, and
   Zap70 were also upregulated, suggesting that BPM by US excitation may
   activate the immune system and induce a strong immune response in the
   TME.

Figure 6.

   Figure 6
   [142]Open in a new tab

   Investigates the therapeutic mechanism of BPM through transcriptome
   high‐throughput sequencing. a) Venn plot of DEGs in RNA seq, b) DEGs
   volcano plot, and c) DEGs heatmap between the control group and BPM+US
   (Synergy) treatment group. d) Significantly enriched GO terms. e) KEGG
   pathway enrichment analysis of DEGs. f) Analyze the correlation between
   the GO enrichment pathway and DEGs. g) Analyze the correlation between
   the KEGG enrichment pathway and DEGs.

   Gene ontology (GO) enrichment analysis revealed the biological
   processes involved after synergistic treatment. As shown in
   Figure [143]6d, dysregulated genes significantly participated in
   various biological processes, including immune response, T‐cell
   migration, lymphocyte migration, receptor surface signaling pathways,
   cell proliferation, and necrosis. Furthermore, we conducted a Kyoto
   Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis to
   understand how BPM exerts its effects through H[2] and O[2] in tumor
   cells. KEGG analysis showed that the DEGs were significantly enriched
   in 10 cellular pathways (Figure [144]6e). Among them, the enrichment of
   MAPK signaling, T‐cell receptor signaling pathway, transcriptional
   misregulation in cancer, cGMP‐PKG signaling pathway, and PI3K‐Akt
   signaling pathway indicated that the synergistic treatment could
   effectively inhibit tumor proliferation and metastasis while activating
   the immune system to elicit a robust immune response. A chord diagram
   (Figure [145]6f) and a Sankey diagram (Figure [146]6g) revealed the
   correlation between DEGs and GO pathways and KEGG pathways,
   respectively.

3. Conclusion

   In summary, this study successfully designed and validated a
   sonocatalyst, BPM, for in situ H[2] and O[2] generation in the TME.
   Under ultrasonic stimulation, BPM demonstrated charge separation and
   electron transfer efficiency, improving water‐splitting efficiency and
   ensuring a continuous supply of H[2] and O[2]. The production of H[2]
   effectively disrupted the mitochondrial function of tumor cells,
   leading to energy metabolism exhaustion and apoptosis, while the
   generation of O[2] alleviated hypoxia within the tumor and enhanced the
   tumor's sensitivity to treatment. The experimental results showed that
   BPM not only inhibited tumor cell proliferation through direct
   catalytic action but also further impaired the tumor's antioxidant
   defenses by consuming GSH. Additionally, BPM significantly enhanced
   immune responses within the TME, promoting DCs maturation, T‐cell
   activation, and macrophage polarization, thereby reversing the
   immunosuppressive state and triggering a strong antitumor immune
   response. Moreover, BPM exhibited good biocompatibility and safety in
   both in vitro and in vivo experiments. In clinic, most solid tumors are
   hypoxic. In situ catalytic water splitting can relieve tumor hypoxia
   and provide auxiliary treatment. Eliminating hypoxia can enhance the
   sensitivity of tumors to other treatments and improve the efficacy.
   These findings support the design rationale of BPM as a
   gas‐immunotherapy combination strategy, offering an approach for future
   cancer treatments.

4. Experimental Section

Synthesis of POM‐Mo[2]C (PM)

   Disperse 1 mg of Mo[2]C in 15 mL of distilled water and stir rapidly at
   1000 rpm. Then, quickly add 1 mL of 30% H[2]O[2] and allow the reaction
   to proceed for 12 h. After completion, centrifuge the mixture and
   retain the supernatant. Repeat the centrifugation process 4 to 5 times.
   Finally, freeze‐dry the collected supernatant to obtain POM‐Mo[2]C
   (blue powder), which was then stored in a dry environment.

Synthesis of BiF[3]‐POM‐Mo[2]C (BPM)

   Dissolve 0.2 mmol of Bi(NO[3])[3]·5H[2]O and 40 mg of POM in 2 mL of
   PEG400 to prepare a blue mixed solution. Then, rapidly add 5 mL of EG
   solution containing 0.4 mmol of NH[4]F and stir vigorously at 25 °C for
   120 s. The reaction was terminated by adding an equal volume of
   deionized water. The resulting mixture was centrifuged and washed with
   deionized water three times. Finally, the obtained BPM was dispersed in
   deionized water and stored at 4 °C.

Cell Culture and Grouping

   4T1 murine breast cancer cells were obtained from the Shanghai
   Institute of Biological Sciences, Chinese Academy of Sciences, and
   cultured in RPMI‐1640 medium supplemented with 10% fetal bovine serum.
   The cells were maintained at 37 °C in a humidified incubator with 5%
   CO[2] (Thermo BB150, USA). The cells were divided into six experimental
   groups: 1) Control group (untreated cells); 2) US group (cells treated
   with ultrasound only); 3) BiF[3]+US group (cells treated with BiF[3]
   and ultrasound); 4) PM+US group (cells treated with PM and ultrasound);
   5) BPM group (cells treated with BPM only); and 6) BPM+US group (cells
   treated with BPM and ultrasound). The US conditions were 0.7 W cm^−2,
   1 MHz frequency, 60% duty cycle, and 5 min of exposure.

Establishment of Tumor Model

   To establish a murine breast cancer model, 5‐week‐old female BALB/c
   mice were injected subcutaneously with 4T1 cells (2 × 10^6 cells mL^−1,
   100 µL) into the forelimb. The animal facility was accredited by the
   Shanghai Municipal Commission of Science and Technology, China (SYXK
   2020‐0006). The experiments followed the guidelines of the Animal
   Ethics Committee of Shanghai East Hospital, China (Protocol No.
   2023‐036‐01).

Statistical Analysis

   Statistical analysis was performed using IBM SPSS Statistics 25
   software. A t‐test was used to assess differences between groups, and
   statistical significance was indicated as follows: *p < 0.05, **p
   < 0.01, ***p < 0.001. The data from parallel experiments were expressed
   as mean ± standard deviation (mean ± SD).

Conflict of Interest

   The authors declare no conflict of interest.

Supporting information

   Supporting Information
   [147]ADVS-12-2413519-s001.docx^ (3MB, docx)

Acknowledgements