A Photosensitizer‐Intercalated Programmable DNA System for Enhanced Photoimmunotherapy via Tumor Microenvironment Remodeling

ABSTRACT

Photoimmunotherapy (PIT) has emerged as a minimally invasive therapeutic modality for cancer, offering high spatial precision and favorable therapeutic outcomes. However, the aggregation‐caused quenching (ACQ) of photosensitizers and the hypoxic tumor microenvironment (TME) limit its overall efficacy. Herein, a DNA tetrahedron‐based nanoplatform (HPTM) functionalized with hypoxia‐inducible factor‐1α antisense oligonucleotide (HIF‐1α ASO) and PD‐L1 aptamers was engineered to load TMPyP for enhanced tumor PIT. The intercalation of TMPyP within the DNA framework effectively prevented photosensitizer aggregation, thus improving reactive oxygen species (ROS) production and photodynamic therapy (PDT) efficacy. Furthermore, the aptamer‐functionalized structure facilitated tumor targeting and blocked the PD‐1/PD‐L1 immune checkpoint on the cell surface. Concurrently, the intracellular delivery of HIF‐1α ASO downregulated endogenous PD‐L1 expression by disrupting the HIF‐1α/PD‐L1 signaling axis. This collaborative endogenous and exogenous strategy effectively remodeled the immunosuppressive TME. The downregulation of HIF‐1α also alleviated the hypoxic microenvironment, further supporting PDT performance. Upon irradiation, HPTM triggered immune activation via PDT‐induced immunogenic cell death (ICD) and promoted long‐term immunological memory. Overall, this multifunctional DNA‐programmed nanoplatform provides efficient photosensitizer loading and addresses the immune imbalance associated with PDT, offering a promising strategy for cancer PIT.

This DNA tetrahedron‐based nanoplatform mitigates the aggregation‐caused quenching of photosensitizers via the framework intercalation effect to support photodynamic therapy. Additionally, it employs an endogenous and exogenous collaborative strategy, integrating HIF‐1α antisense oligonucleotides with PD‐L1 aptamers, to modulate the hypoxia‐induced immunosuppressive tumor microenvironment, thus facilitating an enhanced photoimmunotherapeutic efficiency for cancer therapy.

1. Introduction

As a combination of phototherapy and immunotherapy, photoimmunotherapy (PIT) has emerged as a highly promising anticancer strategy owing to its high precision, excellent biosafety, favorable prognosis, and minimal invasiveness [1, 2, 3]. Photodynamic therapy (PDT), a widely used phototherapy modality, exerts selective cytotoxic effects and induces antitumor immune responses via immunogenic cell death (ICD). Its therapeutic efficacy largely depends on the properties of the photosensitizers employed [4, 5, 6, 7]. Type I photosensitizers generate cytotoxic species through electron transfer interactions with surrounding substrates, thus inducing tumor cell death. However, their complex synthetic procedures and suboptimal biocompatibility limit their widespread application [8, 9, 10]. Thus, current PIT strategies tend to favor Type II photosensitizers, such as porphyrins, phthalocyanines, azopyrroles, and their derivatives [11]. These photosensitizers operate through energy transfer to molecular oxygen and produce singlet oxygen (¹O₂) to induce cellular necrosis and apoptosis [12]. Despite the ideal biosafety, there are two major predicaments for the further application of Type II photosensitizers [13]. First, their intrinsic tendency to aggregate in biological environments leads to aggregation‐caused quenching (ACQ), which reduces the production of reactive oxygen species (ROS) [14]. Second, oxygen depletion during PIT treatment exacerbates tumor hypoxia and immunosuppression within the tumor microenvironment (TME), ultimately impairing therapeutic outcomes [15]. Therefore, the development of PIT strategies that prevent photosensitizer aggregation and regulate hypoxic and immunosuppressive TME is vital for therapeutic efficacy enhancement [16, 17, 18].

Chemical modification is an effective strategy to improve the performance and dispersibility of photosensitizers, thus preventing ACQ. However, it often requires sophisticated molecular design and complex synthetic procedures [19]. Drug delivery systems offer an alternative strategy to address the ACQ effect of photosensitizers. Various nanoplatforms have been developed for photosensitizer delivery, including liposomes, polymeric nanoparticles, inorganic carriers, and metal‐organic frameworks [20, 21, 22, 23, 24]. Nevertheless, photosensitizer molecules still undergo a certain degree of aggregation within drug delivery systems, and the premature leakage of photosensitizers may also lead to side effects [25]. The development of drug delivery platforms with simple molecular design, facile synthesis, and the ability to effectively alleviate ACQ of photosensitizers is of critical importance.

Additionally, the synergistic use of immune checkpoint blockade (ICB) and PDT is a favorable form of PIT, which improves the cancer treatment efficiency with anti‐metastasis capability [26, 27, 28]. Our previous study has revealed that a PD‐L1 aptamer can bind to PD‐L1 proteins on tumor cell membranes for PD‐1/PD‐L1 pathway blockade from an external direction, and the PD‐L1 aptamer demonstrates the value of immune checkpoint inhibitors with a certain immunotherapy effect [29, 30]. However, the hypoxic and immunosuppressive TME induced by PDT increases the abundance of endogenous PD‐L1 via the HIF‐1α/PD‐L1 axis and reduces the therapeutic efficacy of PIT [31, 32, 33]. Negative feedback is presented between PDT and immunotherapy [34, 35].

In recent years, DNA‐based nanomaterials offer new opportunities to address these challenges because of their highly programmable base sequences, facile chemical functionalization, along with superior biocompatibility [36, 37, 38]. DNA nanostructures are able to be precisely engineered as multifunctional delivery vehicles, and functional oligonucleotides endow these structures with a variety of therapeutic functional modules via reasonable design [39, 40]. Herein, an HIF‐1α ASO and PD‐L1 aptamer‐modified multifunctional DNA tetrahedron system (termed HPTM) was established via integrating oligonucleotide chains with the photosensitizer to improve the therapeutic potency of PIT. As illustrated in Scheme 1, the DNA tetrahedral scaffold spatially separated positively charged TMPyP molecules via an intercalation structure, effectively preventing their aggregation for PDT. Following systemic administration, HPTM preferentially accumulated at tumor sites through aptamer‐mediated active targeting. Within the acidic TME, the PD‐L1 aptamer dissociated from DNA frameworks and selectively bound to PD‐L1 proteins localized on tumor cell membranes, thus interrupting the PD‐1/PD‐L1 immune checkpoint pathway. After cellular internalization, HIF‐1α ASO suppressed the expression level of HIF‐1α, which led to a down‐regulation of PD‐L1 expression. This dual‐action mechanism, comprising endogenous and exogenous modulation of PD‐L1 expression, alleviated the immunosuppressive TME to facilitate immunotherapy. Furthermore, the down‐regulation of HIF‐1α via HIF‐1α ASO alleviated tumor hypoxia by promoting vascular normalization and reducing cellular oxygen consumption. This improvement in the hypoxic TME consequently enhanced the efficacy of PDT [41, 42, 43, 44]. Moreover, HPTM could also initiate an immune response via ICD and promote an immune memory effect with anti‐metastatic capacity. Taken together, HPTM presented a synergistic approach to augment the therapeutic efficiency of PIT by simultaneously preventing photosensitizer aggregation and remodeling the TME, offering a promising strategy for cancer photoimmunotherapy.

SCHEME 1.

The scheme for the construction of HPTM, and its application for tumor PIT with 660 nm irradiation.

2. Results and Discussion

2.1. Synthesis and Structural Characterization of HPTM

Figure 1a illustrates the assembly of the multifunctional DNA tetrahedron nanoplatform, which consisted of HIF‐1α ASO, PD‐L1‐targeting aptamer, DNA tetrahedron scaffold, and photosensitizer TMPyP (abbreviated as HPTM; a complete list of all abbreviations in this study was provided in Table S1). HIF‐1α ASO was designed to suppress endogenous PD‐L1 generation via disrupting the HIF‐1α/PD‐L1 signaling axis. Upon stimulation by the acidic TME, the PD‐L1 aptamer was released and blocked cell‐surface PD‐L1 via an exogenous pathway. This endogenous and exogenous coordinated regulation mitigated the immunosuppression induced by PIT. TMPyP was intercalated into the DNA tetrahedron for PDT, and its aggregation was minimized to enhance ROS generation upon light irradiation. For the nanoplatform assembly, HIF‐1α ASO and pH‐responsive hairpin sequence‐modified DNA tetrahedron (abbreviated as HT) were first synthesized with four rationally designed strands. The sequences of all oligonucleotides were provided in Table S2. Under neutral pH conditions, the PD‐L1 aptamer hybridized with the complementary hairpin sequence on HT, yielding the complex composed of HIF‐1α ASO, PD‐L1 aptamer, and DNA tetrahedron (abbreviated as HPT). Finally, HPT was incubated with TMPyP to produce HPTM.

FIGURE 1.

Construction and characterization of the HPTM nanoplatform. (a) Schematic illustration of the stepwise preparation of HPTM. (b) PAGE analysis of the stepwise assembly of HPT. (c) AFM image of the assembled HPT nanostructure. (d) Schematic illustration of the pH‐responsive PD‐L1 aptamer release process triggered by the acidic TME. (e) UV–vis absorption spectra of free TMPyP and HPTM. (f) CD spectra of free TMPyP and HPTM. (g) Molecular docking model showing the binding mode and key interacting nucleotide residues between TMPyP and the HPT framework. (h) Fluorescence emission spectra of free TMPyP at gradient concentrations (4, 6, 8, 10, 12, and 14 µM). (i) Fluorescence emission spectra of HPTM at gradient equivalent TMPyP concentrations (4, 6, 8, 10, 12, and 14 µM).

The stepwise construction of HPT was confirmed via polyacrylamide gel electrophoresis (PAGE) (Figure 1b). As additional strands were integrated, a gradual reduction in electrophoretic mobility was observed. This was attributed to the increased molecular weights and enlarged structural dimensions, which confirmed the sequential assembly and successful synthesis of HPT. A stability test showed that HPT maintained its structural integrity for 72 h in 10% FBS and remained stable in DNase I (0.01 U/L) for 48 h (Figure S1), indicating its robustness for potential applications in vivo. Furthermore, atomic force microscopy (AFM) imaging revealed that HPT exhibited a rigid and well‐defined tetrahedral structure (Figure 1c). Dynamic light scattering (DLS) analysis demonstrated an average hydrodynamic diameter of approximately 30.1 nm for HPT (Figure S2). After binding with TMPyP, the hydrodynamic diameter slightly increased to 31.0 nm, while the zeta potential shifted from −10.7 to + 0.3 mV, suggesting that TMPyP altered its surface charge without affecting the particle size of HPT. The appropriate particle sizes of HPT endowed its feasibility for cellular uptake via endocytosis. To investigate the pH‐responsive release of PD‐L1 aptamer from HPT structure, a fluorophore‐quencher system comprising FAM and BHQ2 was employed (Figure 1d). The release behavior was evaluated across four pH conditions (7.4, 7.0, 6.5, and 6.0). Specifically, pH 7.4 was used to mimic normal blood circulation. pH 7.0 and 6.5 represented the mildly acidic perivascular region and the more acidic and hypoxic regions within tumors, respectively. pH 6.0 was used to mimic lysosomal compartments. As illustrated in Figure S3a, the progressive enhancement of FAM fluorescence from pH 7.4 to 6.5 revealed a pH‐responsive release mechanism. These findings suggested the pH‐triggered dissociation of the PD‐L1 aptamer under tumor‐relevant acidic conditions, which facilitated exogenous blockade of the PD‐1/PD‐L1 pathway. Furthermore, a 5‐cycle pH‐switching assay was also conducted (Figure S3b). The system exhibited reversible fluorescence variations upon alternating the pH between 7.4 and 6.0, without noticeable signal attenuation over five cycles. These results indicated that the response originated from the reversible structural transformation of the i‐motif, rather than the degradation of the DNA tetrahedron.

The successful intercalation of TMPyP into HPT was critical to prevent the quenching of the photosensitizer caused by aggregation. Accordingly, UV–vis spectroscopy absorption spectra and circular dichroism (CD) spectra were measured to verify the intercalation of TMPyP. As shown in Figure 1e, the characteristic absorption peak of free TMPyP was redshifted from 420 to 430 nm upon complexation with HPT. This redshift was attributed to the interaction of TMPyP with DNA, reducing the energy required for the π–π ^* electronic transition of TMPyP molecules. The distinct negative CD signal of HPTM also indicated the successful intercalation (Figure 1f). Molecular docking calculations were employed to visualize the binding pattern and molecular interactions of TMPyP with HPT. As shown in Figure 1g, the spatial localization of the binding interface was clearly illustrated. Key nucleotide residues involved in the interaction were identified, including A26, C25, T12, G15, T22, and G21. These residues delineated the specific region where TMPyP interacted with HPT. Figure S4 further highlighted the types of intermolecular forces responsible for the binding. Multiple interactions, including π‐anion interactions, attractive electrostatic interactions, π‐cation interactions, and π–π stacking interactions, were clarified. Collectively, these forces constituted the molecular basis for the stable association between TMPyP and HPT. Among them, π‐related interactions, namely π‐anion, π‐cation, and π–π stacking, were considered to contribute significantly to preserving the stability of the binding conformation, while the electrostatic attraction likely enhanced the binding affinity. Figure 1h showed that the fluorescence intensity of free TMPyP was increased gradually as the dosage was increased from 4 to 10 µM, and fluorescence was quenched at the concentration of 12 µM due to aggregation. However, the fluorescence intensity of HPTM presented a continuous enhancement as the concentration of TMPyP increased, indicating that the aggregation challenge of the photosensitizer was effectively resolved via the intercalative incorporation of TMPyP into HPT (Figure 1i). To determine a suitable molar ratio between HPT and TMPyP that presented the most effective PDT performance of HPTM, a gradient of HPT‐to‐TMPyP molar ratios (1:1, 1:5, 1:10, 1:15, and 1:20) was evaluated. A TMPyP standard curve was first established to quantify the encapsulation efficiency (Figure S5a). As shown in Figure S5b, the TMPyP encapsulation capacity increased with the proportion of TMPyP and plateaued at a 1:10 ratio. This plateau reflected the maximum binding capacity of HPT frameworks, with the encapsulation efficiency of 63.1% (Figure S5c). At higher ratios (1:15 and 1:20), no further increases in encapsulation capacity or corresponding ROS generation capacity were observed with DPBF as a ROS probe (Figure S5d). Additionally, all tested HPTM with different HPT‐to‐TMPyP molar ratios maintained a consistent average hydrodynamic diameter of 30 nm (Figure S5e) and demonstrated structural stability over 7 days of storage at 4°C, with no prominent changes in particle size (Figure S5f). Consequently, the HPT‐to‐TMPyP molar ratio of 1:10 was selected for subsequent experiments, which offered a favorable performance among encapsulation efficiency, ROS generation, structural uniformity, and stability. Compared with TMPyP, HPTM exhibited a more significant PDT capacity under equivalent photosensitizer concentrations, which manifested the better ROS generation efficiency of HPTM due to the suppression of TMPyP aggregation (Figure S6). All these results confirmed the successful construction of HPTM and highlighted its potential for enhanced PIT.

2.2. In Vitro Cellular Uptake, Cytotoxicity and ROS Generation Studies

To validate the targeting capability of HPTM in vitro, LLC cells with high PD‐L1 expression levels were selected as the PD‐L1 positive cell model, and MLE‐12 cells served as a negative control. The confocal laser scanning microscopy (CLSM) images and flow cytometry analysis indicated that HPTM presented a significantly enhanced accumulation in LLC cells compared to MLE‐12 cells, which was attributed to the specific recognition of the PD‐L1 aptamer (Figure 2a and Figure S7). Kinetic analysis showed that the dissociation constant of the PD‐L1 aptamer with PD‐L1 protein was 2.5 × 10⁻⁸ mol/L, indicating a strong binding affinity between them for targeted identification and stable combination (Figure S8). Subsequently, the intracellular uptake of HPTM by LLC cells was further explored, and the groups of PBS, free TMPyP, and HPTM without HIF‐1α ASO modification (PTM) were conducted as comparisons. As shown in CLSM images and flow cytometry results, minimal green TMPyP fluorescence was observed in PBS and free TMPyP groups (Figure 2b and Figure S9). This weak signal was due to the non‐specific endocytosis of LLC cells during co‐incubation. In contrast, LLC cells treated with HPTM and PTM exhibited significant fluorescence of TMPyP, confirming the robust tumor‐targeting ability of the PD‐L1 aptamer and the efficient cellular uptake mediated by the DNA nanostructure. Figures S10 and S11 revealed that the cellular uptake of HPTM displayed both concentration‐dependent and time‐dependent behaviors, which supported the continuous internalization process and demonstrated its potential in tumor therapy. Given that real tumor tissues showed more complex pathophysiological features, we further constructed a tumor microsphere model to validate the penetration capability of HPTM in 3D LLC tumor microspheres (Figure 2d and Figure S12). The results showed that, at the same penetration depth within the tumor spheroids, the HPTM group exhibited higher green fluorescence intensity than the free TMPyP group, verifying the improved penetration of HPTM to infiltrate deeper regions of solid tumors. To confirm whether HPTM could escape from lysosomes to avoid enzymatic degradation via ROS generation, the colocalization of HPTM with lysosomes was investigated (Figure S13). After co‐incubation with LLC cells for 4 h in the absence of light, significant overlap was observed between the fluorescence of HPTM and LysoTracker signals, which indicated the retention of HPTM within lysosomes under non‐irradiated conditions. Nevertheless, upon 660 nm laser irradiation for 5 min, a notable separation of fluorescence signals between HPTM and LysoTracker was observed, illustrating the escape of HPTM from lysosomes under light irradiation. This phenomenon was because of the generation of ROS via PIT, which disrupted lysosomal membranes and facilitated the release of HPTM into the cytoplasm. LLC cells receiving various treatments were stained with an ROS probe, and cellular ROS levels were determined by CLSM. CLSM images revealed a stronger green fluorescence in HPTM‐treated cells compared to those treated with free TMPyP or PBS after laser irradiation (Figure 2c), reflecting an elevated level of intracellular ROS. Moreover, since HPTM relieved the hypoxic microenvironment via HIF‐1α regulation, it exhibited higher ROS generation efficiency than PTM. These results demonstrated that HPTM significantly enhanced ROS generation under irradiation for satisfactory PDT.

FIGURE 2.

In vitro cellular targeting, uptake, and gene regulation performance of HPTM. (a) Representative CLSM images of the PD‐L1 targeting capability of HPTM in LLC cells and MLE‐12 cells. (b) CLSM images of cellular uptake of free TMPyP, PTM, and HPTM by LLC cells after 4 h incubation. (c) Representative CLSM images of intracellular ROS generation in LLC cells after different treatments. (d) CLSM z‐stack images and 3D reconstruction showing the penetration capability of free TMPyP and HPTM in 3D LLC tumor spheroids at a scanning depth of 100 µm. (e) Representative CLSM images of PD‐L1 expression on LLC cell surfaces after different treatments with 660 nm laser irradiation. (f) Relative mRNA expression levels of HIF‐1α and PD‐L1 in LLC cells measured by RT‐qPCR after different treatments (n = 5). (g) Western blot analysis of HIF‐1α and PD‐L1 protein expression levels in LLC cells after different treatments. Data in quantitative plots were presented as mean ± SD. Statistical analysis was performed by one‐way ANOVA with Tukey's post‐hoc test. ^*** p < 0.001.

2.3. In Vitro Investigations on the Regulation of HIF‐1α and PD‐L1 Expressions

HPTM‐mediated down‐regulation of intracellular PD‐L1 expression was subsequently evaluated. As shown in the immunofluorescence staining results in Figure 2e and Figure S14, the intense green fluorescence signals of TMPyP + L and PTM + L groups after PDT indicated the overexpression of PD‐L1 induced by a hypoxic microenvironment, and specific binding of PD‐L1 aptamers to PD‐L1 proteins could not modulate the overall expression of PD‐L1 with an exogenous method. Given that HIF‐1α modulates PD‐L1 expression through binding to the hypoxia response element (HRE), the intermediation of HIF‐1α was considered a potential strategy for modulating intracellular PD‐L1 levels. Therefore, HIF‐1α ASO was introduced into the DNA tetrahedron nanoplatform to inhibit HIF‐1α. HPTM + L group displayed the weakest green fluorescence, suggesting that HPTM + L could effectively down‐regulate overall PD‐L1 expression via the HIF‐1α/PD‐L1 pathway for better PIT effect. The mRNA levels of HIF‐1α and PD‐L1 in different treatment groups after irradiation were also assessed by RT‐qPCR to further clarify the mediation of PD‐L1 via the HIF‐1α/PD‐L1 signaling axis (Figure 2f). The experimental results indicated that HPTM + L exhibited a significant decrease in both HIF‐1α and PD‐L1 mRNA expressions compared with other groups. This suppression was attributed to the degradation of mRNA and promoted by the hybridization between HIF‐1α ASO and target HIF‐1α mRNA. Consistent with the RT‐qPCR results, the western blot experiment also confirmed that HPTM + L effectively decreased intracellular HIF‐1α and PD‐L1 expression (Figure 2g and Figure S15). Thus, HPTM exhibited the capacity of immunotherapy by synergistically down‐regulating exogenous and endogenous PD‐L1. Moreover, a scrambled ASO‐loaded DNA tetrahedron (scHPTM) was further prepared as a negative control group to investigate the non‐specific down‐regulation effect of oligonucleotides. As shown in Figure S16, HPTM effectively decreased the mRNA expression levels of HIF‐1α and PD‐L1, while scHPTM treatment did not induce any notable down‐regulation. These results indicated that the modulation of the HIF‐1α/PD‐L1 axis was mediated by the sequence‐specific targeting of the designed HIF‐1α ASO, rather than non‐specific effects of nucleic acids.

2.4. Anti‐Tumor Effects of HPTM In Vitro

HPTM exhibited the capacity to produce ROS and concurrently modulated the immunosuppressive TME, offering a promising approach for targeted PIT. Therefore, the in vitro anti‐tumor activity of HPTM was evaluated. First, cell viability was measured with the CCK‑8 kit (Figure S17). Under dark conditions, HPTM at different concentrations exerted minimal cytotoxicity toward LLC cells. . In addition, when MLE‐12 cells were employed as the control cell line, low cytotoxicity was observed after PBS + L, TMPyP + L, PTM + L, and HPTM + L treatments. Collectively, these results supported the good biocompatibility of HPTM. As shown in Figure 3a, no significant apoptosis was observed under dark conditions for any of these groups, manifesting the biocompatibility of HPTM without photoactivation. However, upon laser irradiation, the HPTM + L group induced an increase in both early and late apoptosis rates of LLC cells compared with the TMPyP + L group and the PTM + L group. This enhanced apoptotic effect of the HPTM + L group was attributed to the efficient utilization of TMPyP through DNA tetrahedron intercalation and the downregulation of HIF‐1α for sensitized PDT. Cell viability results measured with CCK‐8 were also consistent with the apoptosis data and demonstrated the potential PDT capacity of HPTM (Figure 3b). The live/dead cell staining showed similar results and further validated the cytotoxic potential of HPTM under laser treatment (Figure 3d and Figure S18). HPTM‐treated cells exhibited intense red PI fluorescence, contrasting with the weaker signals in the free TMPyP and PTM groups. These results unequivocally demonstrated that HPTM exhibited the strongest cytotoxic effect with laser irradiation.

FIGURE 3.

In vitro antitumor activity of HPTM upon 660 nm laser irradiation. (a) Flow cytometric analysis of apoptosis in LLC cells detected by Annexin V‐FITC/PI double staining after different treatments. (b) Viability of LLC cells after different treatments determined by CCK‐8 assay (n = 5). (c) Western blot analysis of apoptosis‐related proteins (pro‐apoptotic Bax and anti‐apoptotic Bcl‐2) in LLC cells after different treatments. (d) Calcein‐AM/PI live/dead cell staining of LLC cells after different treatments. (e) CLSM images of mitochondrial membrane potential in LLC cells detected by JC‐1 probe after different treatments. Data in quantitative plots were presented as mean ± SD. Statistical analysis was performed by one‐way ANOVA with Tukey's post‐hoc test. ^*** p < 0.001.

Given the critical role of mitochondria in the apoptosis process, mitochondrial functions were examined to verify the cell‐killing capacity of HPTM with 660 nm irradiation. Mitochondrial membrane potential (ΔΨm) was detected using the JC‐1 fluorescent probe. Figure 3e and Figure S19 showed that the red fluorescence of PBS, TMPyP, PTM, and HPTM groups indicated intact mitochondrial function without damage, while the most notable green fluorescence in the HPTM + L group indicated the most severe loss of ΔΨm and mitochondrial depolarization. This observation confirmed that HPTM effectively induced mitochondrial dysfunction and activated the mitochondrial apoptosis pathway under 660 nm irradiation. The levels of BCL‐2 family proteins, which were linked to mitochondrial membrane permeability, indicated that HPTM + L led to a remarkable upregulation of pro‐apoptotic BAX and a downregulation of anti‐apoptotic BCL‐2 (Figure 3c). These findings suggested enhanced mitochondrial membrane permeability and the activation of apoptotic signaling cascades. Overall, HPTM effectively induced apoptosis in LLC cells through ROS‐mediated mitochondrial dysfunction with 660 nm irradiation, underscoring its potential as a highly promising candidate for targeted cancer therapy.

2.5. The Induction of ICD In Vitro

Beyond direct tumor cell death, laser‐irradiated HPTM also triggered ICD and elicited potent antitumor immune responses for immunotherapy. The ICD‐inducing capability of HPTM was explored by detecting key damage‐associated molecular patterns (DAMPs), CRT, HMGB1, and extracellular ATP. The extracellular secretion of HMGB1 was examined by immunofluorescence staining. Relative to the PBS group, treatments with TMPyP + L and PTM + L resulted in partial translocation of HMGB1 from the nucleus of LLC cells, while the intracellular fluorescence signal of the HPTM + L treated group was nearly undetectable (Figure 4a,b), indicating that HMGB1 was translocated from the nucleus to the cytoplasm and extracellular space. Moreover, as depicted in Figure 4c,d, the most intense green fluorescence signal on the cell surface of LLC cells in the HPTM + L group indicated enhanced CRT externalization. The level of extracellular ATP released by LLC cells in each treatment group was quantified using a commercially available ATP detection assay kit. As illustrated in Figure S20, the HPTM + L group showed the most pronounced ATP efflux following laser irradiation compared with the TMPyP + L and PTM + L groups. This finding was attributed to HPTM's dual effects of alleviating the ACQ effect of TMPyP and ameliorating the hypoxic TME via HIF‐1α modulation, consistent with our previous in vitro antitumor activity results. Consequently, laser‐irradiated HPTM significantly enhanced tumor PIT efficacy by inducing robust ICD in tumor cells.

FIGURE 4.

Induction of ICD and adaptive immune activation in vitro. (a) Representative CLSM images of HMGB1 nucleoplasmic translocation and extracellular release in LLC cells after different treatments. (b) Quantitative analysis of intracellular HMGB1 MFI in each group (n = 5). (c) Representative CLSM images of CRT exposure on the membrane surface of LLC cells after different treatments. (d) Quantitative analysis of CRT MFI in each group (n = 5). (e) Representative flow cytometric plots of CD80⁺CD86⁺ double‐positive mature DCs in each group. (f) Concentrations of pro‐inflammatory cytokines TNF‐α, IL‐6, and IL‐12 in the co‐incubation supernatant detected by ELISA (n = 5). (g) Schematic workflow of the T cell‐mediated tumor cell killing assay. Data in all quantitative plots were presented as mean ± SD. Statistical analysis was performed by one‐way ANOVA with Tukey's post‐hoc test. ^*** p < 0.001.

2.6. In Vitro Promotion of Dendritic Cell Maturation

DCs represented the most potent professional antigen‐presenting cells (APCs) in the immune system and were recognized as initiators of immune responses triggered by ICD. A transwell co‐incubation system of DCs and LLC cells was constructed to evaluate the capacity of HPTM to promote DCs maturation. As shown in Figure 4e, after 660 nm laser irradiation, LLC cells in the upper chamber of the transwell system yielded increased proportions of CD80⁺CD86⁺ double‐positive mature DCs in the lower chamber. The percentage reached 28.7% in the TMPyP + L group and 32.8% in the PTM + L group, indicating moderate DCs maturation. In the HPTM + L treatment group, laser‐irradiated LLC cells induced a marked elevation in the proportion of CD80⁺CD86⁺ double‐positive DCs, with a peak value of 50.1%, reflecting more robust DCs maturation. Moreover, the expression levels of major histocompatibility complex class I (MHC‐I) and class II (MHC‐II) on the surface of DCs were also assessed, whose upregulation was another key indicator of DCs maturation. Figure S21 showed that HPTM + L‐treated LLC cells remarkably upregulated both MHC‐I and MHC‐II expression on DCs. The potential direct activation of DCs by HPTM was investigated using a tumor cell‐free co‐incubation system. After incubation of DCs with 1 µM HPTM for 24 h, no obvious DCs activation was observed (Figure S22a,b). These results suggested that HPTM did not directly induce DCs maturation.

Furthermore, the secretion levels of pro‐inflammatory cytokines including tumor necrosis factor‐α (TNF‐α), interleukin‐6 (IL‐6) and interleukin‐12 (IL‐12) in co‐incubation supernatant were higher in the HPTM + L group than those in the control group, further confirming the potent ICD‐inducing ability of laser‐activated HPTM (Figure 4f). Collectively, these findings demonstrated that laser‐activated HPTM effectively triggered ICD and promoted DCs maturation, thus facilitating antigen cross‐presentation and eventually amplifying the systemic antitumor immune response.

2.7. Improved Tumor Cell Killing and Cytokine Release In Vitro

Mature DCs presented tumor‐associated antigens (TAAs) and secreted cytokines to recruit and activate T cells for adaptive immune responses during the ICD process. Thus, the effects of various treatments on T cell‐mediated tumor cell killing were investigated.

T cells were first activated by DCs harvested from various treatment groups, then isolated and co‐cultured with LLC cells. T cell cytotoxicity was subsequently quantified using a lactate dehydrogenase (LDH) release assay. The experimental procedure is presented in Figure 4g. To ensure the consistency of T cell activation status prior to co‐incubation, the expression of the early activation marker CD25 was analyzed by flow cytometry. As shown in Figure S23a,b, no significant differences in the proportion of CD25⁺ cells were observed among the tested groups. This indicated a consistent baseline activation status of T cells across all groups before the establishment of the co‐incubation system. Figure S23c showed that T cells from the HPTM + L group demonstrated the highest tumor cell killing efficiency with an approximately 58.6% cell killing rate at an effector‐to‐target ratio of 1:1. It was confirmed that HPTM could significantly enhance the PDT effect under 660 nm irradiation, which further induced ICD for adaptive immune responses.

Moreover, the designed DNA tetrahedron of HPT could regulate the expression of PD‐L1 on LLC cell surfaces through both endogenous and exogenous pathways for immune escape inhibition, therefore improving the immune efficacy of T cells for immunotherapy. The influence of HPT on T cell‐mediated tumor cell killing was evaluated with a co‐incubation cell model. LLC cells were treated with PBS, PT, HPT, and FDA‐approved PD‐L1 antibody drug atezolizumab, respectively. Then, the LLC cells of various groups were co‐incubated with activated T cells at an effector‐to‐target ratio of 1:1 overnight. As shown in Figure S24, the PT‐treated group enhanced the killing ability of T cells against LLC cells (32.1%) in an exogenous regulatory manner by blocking the PD‐1/PD‐L1 recognition pathway due to the binding of PD‐L1 aptamers to PD‐L1 proteins on the LLC cell membrane. The HPT group was loaded with HIF‐1α ASO, which reduced intracellular PD‐L1 expression via the HIF‐1α/PD‐L1 axis. Dual exogenous and endogenous regulation further enhanced the cytotoxicity of T cells against LLC cells with a 42.1% cell killing rate. This effect was just slightly inferior to that of PD‐L1 antibody drug used in clinical practice (44.8%), suggesting that HPT could serve as a promising candidate for cancer immunotherapy.

2.8. Antitumor Efficacy of HPTM In Vivo

In vivo PIT studies were subsequently performed using LLC xenografts in C57BL/6 mice. To begin with, the biocompatibility of HPTM was assessed with a hemolysis assay. As presented in Figure S25, the hemolysis ratio of HPTM at concentrations from 0 to 50 µg/mL was less than 2%, verifying its excellent blood compatibility. The tumor accumulation of HPTM following intravenous injection was then evaluated by an in vivo imaging system (IVIS) to determine the optimal time window for laser irradiation (Figure 5a). Tumor fluorescence reached its peak at 6 h post‐injection, at which point laser treatment was applied. Notably, strong fluorescent signals were still detectable in the tumor area 12 h after administration, suggesting a high retention rate of HPTM within the tumor microenvironment. At 24 h, biodistribution analysis further revealed preferential accumulation of HPTM in tumors compared with major organs including the heart, liver, spleen, lungs, and kidneys (Figure 5b), which illustrated the satisfactory targeting ability and biological safety.

FIGURE 5.

In vivo systemic antitumor efficacy of HPTM in LLC tumor‐bearing mice. (a) In vivo fluorescence imaging of tumor‐bearing mice at 2, 4, 6, 12, and 24 h after intravenous injection of HPTM. (b) In vivo fluorescence imaging of major organs (heart, liver, spleen, lung, kidney) and tumor tissue at 24 h after HPTM. (c) Schematic illustration of the LLC tumor allograft model establishment and treatment regimen. (d) Relative tumor volume growth curves of LLC tumor‐bearing mice in each group during the 14‐day treatment period (n = 5). Tumor volume was calculated as TV = (L × W²)/2 (L: tumor major axis, W: tumor minor axis). (e) Survival curves of LLC tumor‐bearing mice in each group (n = 5). (f) Body weight curves of mice in each group during the treatment period (n = 5). (g) Representative photographs of excised LLC tumor tissues from each group at the experimental endpoint (day 14). (h) Representative images of H&E staining, Ki‐67 immunohistochemical staining, and TUNEL staining of tumor sections from each group. Data in quantitative plots were presented as mean ± SD. Statistical analysis was performed by one‐way ANOVA with Tukey's post‐hoc test. ^*** p < 0.001.

When tumors reached approximately 80 mm³, tumor‐bearing mice received different treatment regimens to investigate the PIT potential of HPTM, as outlined in Figure 5c. Their physiological status was monitored every two days thereafter. The findings showed that the mice treated with HPTM + L exhibited pronounced inhibition of tumor growth, which was in agreement with the in vitro data (Figure 5d,g, and Figure S26). Moreover, the survival analysis revealed that the HPTM + L group achieved the highest survival rate among all treatment groups (Figure 5e). Taken together, these findings confirmed that HPTM combined with laser irradiation displayed superior antitumor efficacy compared with TMPyP + L and PTM + L, which demonstrated the synergistic effect of the alleviation of TMPyP quenching via DNA tetrahedral intercalation and the tumor microenvironment regulation via HIF‐1α ASO for PIT enhancement. Furthermore, the impacts of HPTM on tumor cell proliferation and apoptosis were also determined using three staining methods: Hematoxylin and Eosin (H&E) staining, Ki‐67 immunostaining, and TdT‐mediated dUTP Nick‐End Labeling (TUNEL) staining. Results from Figure 5h indicated that in the PBS + L group, H&E staining of tumor tissues, Ki‐67 index, and TUNEL fluorescence exhibited only minor differences, indicating a relatively low level of apoptosis. In contrast, in the H&E‐stained sections of tumors treated with HPTM + L, significant cellular disorganization, vacuolization, and nuclear fragmentation were observed. Consistently, a markedly reduced Ki‐67 index and enhanced TUNEL fluorescence were detected, confirming the suppressed proliferation and enhanced apoptosis process. These findings further substantiated the potent tumor‐inhibitory activity of HPTM + L. This effect was attributed to the tumor‐targeting capability derived from the PD‐L1 aptamer, which enabled HPTM to accumulate more efficiently at the tumor site. The intercalative binding between TMPyP and HPT alleviated the ACQ of TMPyP, thus enhancing the therapeutic efficacy of PDT. In addition, the improvement of immune effect through the modulation of the HIF‐1α/PD‐L1 axis also promoted the PIT efficacy of HPTM.

The regulatory effects of HPTM on HIF‐1α and PD‐L1 expressions in vivo were further examined by immunofluorescence staining. Figure S27 illustrated that the green fluorescence signals associated with HIF‐1α in tumor tissues were decreased after HPTM + L treatment, indicating the downregulation of HIF‐1α expression. By contrast, tumors in TMPyP + L and PTM + L groups exhibited elevated HIF‐1α fluorescence intensities compared with the control group, suggesting that Type II photosensitizer‐based PDT induced tumor hypoxia and compensatorily improved expression of HIF‐1α, which would impair the therapeutic outcomes. Similarly, PD‐L1 expression displayed a comparable pattern with downregulation in the HPTM + L group via the HIF‐1α/PD‐L1 signaling axis for immune escape prevention. Furthermore, although the PTM + L group reduced the expression of PD‐L1 to a certain extent via the exogenous interaction between aptamers to PD‐L1 protein, endogenous PD‐L1 translocated to the tumor cell surface under hypoxic conditions for a weakened immune response. Relative to the control group, PD‐L1 expression in the PTM + L group was elevated by 48.9% (Figure S28). Thus, the endogenous and exogenous coordinated regulation strategy of HPTM highlighted the ability to relieve the immunosuppressive TME for better PIT efficiency.

To further evaluate the biosafety of HPTM, blood specimens were gathered from mice for hematological and biochemical tests after 14 days of treatment. No abnormalities were observed in any of the experimental groups, confirming that HPTM + L was well tolerated and suitable for intravenous administration (Figures S29 and S30). Throughout the treatment period, body weights remained stable across all groups (Figure 5f), suggesting minimal systemic toxicity. Additionally, H&E staining of key organs exhibited no obvious pathological lesions after various interventions (Figure S31). All these results demonstrated that HPTM not only exhibited satisfactory PIT efficacy with 660 nm irradiation, but also exhibited excellent biocompatibility and safety in vivo.

2.9. The Stimulation of Robust Anti‐Tumor Immunity In Vivo

Immunofluorescence staining of CRT and HMGB1 was employed to evaluate the capacity of HPTM + L for ICD induction in vivo. Figure 6a showed that in the HPTM + L group, CRT exposure and HMGB1 release were significantly increased, implying that the ICD process was successfully activated with HPTM + L treatment. The changes of the relevant immune cells within TME were further evaluated with flow cytometry. DCs mainly captured tumor‐derived antigens and presented them on their cell surface to trigger the activation of functional T cells. Concurrently, the activated immature DCs differentiated into mature DCs. Thus, the frequency of mature DCs in the lymph nodes of mice in each group was examined (Figure 6b and Figure S32). The frequency of mature DCs was 27.5% in the TMPyP + L group, while the frequency of mature DCs increased to 29.1% and 51.8% in the PTM + L and HPTM + L groups, respectively. HPTM + L group exhibited higher DCs maturation than the TMPyP + L group and the PTM + L group, indicating that HPTM + L induced a stronger ICD process. This improvement was attributed to the efficient uptake and utilization of TMPyP with the delivery platform of DNA tetrahedron, and the modulation of the hypoxic TME via the HIF‐1α ASO signal axis for enhanced PDT.

FIGURE 6.

In vivo antitumor immune activation and anti‐metastasis efficacy of HPTM. (a) Representative immunofluorescence images of CRT exposure and HMGB1 release in tumor tissues from each group. (b) Flow cytometric quantification of CD80⁺CD86⁺ mature DCs in tumor‐draining lymph nodes of mice in each group. (c) Flow cytometric quantification of M1‐type tumor‐associated macrophages (TAMs, F4/80⁺CD86⁺) in tumor tissues. (d) Flow cytometric quantification of M2‐type TAMs (F4/80⁺CD206⁺) in tumor tissues. (e) Flow cytometric quantification of CD4⁺ and CD8⁺ T cells infiltrated in tumor tissues. (f) Representative images of Bouin's fixative‐stained whole lungs and H&E staining of lung sections from the lung metastasis model. (g) Weight of the lungs from mice in each group at the endpoint of the anti‐metastasis experiment (n = 5). (h) Number of pulmonary metastatic nodules in the lungs of mice in each group (n = 5). Data in quantitative plots were presented as mean ± SD. Statistical analysis was performed by one‐way ANOVA with Tukey's post‐hoc test. ^*** p < 0.001.

Within the TME, tumor‐associated macrophages (TAMs) typically exhibited an immunosuppressive M2 phenotype, contributing to tumor immune evasion and progression. Thus, the elevated percentage and count of antitumor M1 macrophages within the TME acted as a biomarker to evaluate the degree of tumor immune activation. Figure 6c,d demonstrated that the control group exhibited minimal M1 macrophage infiltration, accounting for only about 10.9%, whereas M2 macrophages constituted approximately 40.2%. After HPTM + L treatment, the proportion of M1 macrophages increased to about 33.7%, while that of M2 macrophages decreased to roughly 17.9%. We found that HPTM + L was able to effectively improve the immune function of macrophages through boosting their tumor cell killing ability (M1 phenotype) while lessening their facilitation of tumor cell growth (M2 phenotype). Maturation of DCs and an increased proportion of M1 macrophages facilitated the activation of T cells. Flow cytometry was performed to examine T cell infiltration within tumor tissues and evaluate the impact of HPTM + L on the activation of antitumor immune responses. The infiltrating frequencies of CD4 and CD8 T cells in the HPTM + L group reached 45.8% and 32.7%, respectively, which were higher than the corresponding values of 12.7% and 16.3% observed in the PTM + L group (Figure 6e). These results indicated that the introduction of HIF‐1α ASO could promote the stimulation and infiltration of T cells through regulation of the hypoxic TME and the endogenous PD‐L1 expression, leading to improved therapeutic efficacy. The immunofluorescence analysis of CD4 and CD8 T cells in Figure S33 was consistent with the conclusion obtained from flow cytometry. Further assessment of IFN‐γ expression in CD8 T cells was performed by staining with anti‐CD8 and anti‐IFN‐γ antibodies. As shown in Figure S34, the HPTM + L group exhibited increased frequencies of 52.6% CD8 IFN‐γ double‐positive T cells compared to 15.4% in the control group. Moreover, the proportion of Tregs, which suppress effective anti‐tumor immunity, significantly decreased from 61.5% in the control group to 16.6% in the HPTM + L group (Figure S35). Peripheral blood was collected to measure the secretion of pro‐inflammatory cytokines, and the HPTM + L group presented the obvious increments of IL‐6 and TNF‐α (Figure S36). Therefore, HPTM + L reshaped the tumor‐infiltrating immune cell population by promoting DCs maturation, increasing the proportion of M1 macrophages, and augmenting CD4 and CD8 T cells, while simultaneously inhibiting Tregs. All these immune responses contributed to a strengthened anti‐tumor capacity of HPTM with 660 nm irradiation.

2.10. In Vivo Anti‐Metastatic Efficacy and Long‐Term Immune Memory Evaluation

To assess the anti‐metastatic effects and the long‐term immune memory effects of HPTM + L, a lung metastatic model was successfully constructed. On day 7 following the completion of different treatments, 1 × 10⁵ LLC cells were injected intravenously via the tail vein to induce lung metastasis in mice, and the mice were then humanely euthanized and dissected on day 21 post‐injection. The lungs were stained with Bouin's solution and revealed that numerous characteristic gray‐white pulmonary lesions were observed in all groups. Histopathological evaluation also confirmed the presence of pulmonary metastatic foci after various treatments (Figure 6f). Then, lung weights and the quantity of lung nodules were measured. The lung weights (Figure 6g) in the HPTM + L group were notably less than those in the remaining groups, and the number of lung nodules (Figure 6h) was also reduced. These results indicated that HPTM + L treatment not only exerted curative effects on primary tumors but also possessed the ability to suppress the development of metastatic foci via triggering the immune memory response. Subsequently, DCs and memory T cells in the spleen were examined following various treatments to verify the long‐term immune memory effect elicited by HPTM + L. Compared with the PBS group, the maturation rate of splenic DCs in the HPTM + L group increased from 2.0% to 18.6% after treatment (Figure S37). As illustrated in Figure S38, the population of effector memory T cells (TEM) in the spleen was increased and reached 26.7% after HPTM + L treatment. This finding confirmed the successful establishment of durable immune memory for antitumor therapy and indicated a favorable prognosis of PIT by HPTM + L treatment.

2.11. Transcriptomic Analysis

RNA sequencing (RNA‐seq) was conducted on tumor specimens of the control group and the HPTM + L group to elucidate the potential mechanisms underlying its antitumor effects. As shown in Figure 7a, correlation analysis demonstrated high consistency among the 3 biological replicates within each group, verifying the robust biological reproducibility. A clear divergence in gene expression profiles was observed in the treated group relative to the control group, which was attributed to the intervention of HPTM + L. Principal component analysis (PCA) additionally showed the clear clustering of the two groups (Figure 7b). Differential expression analysis identified 1,431 upregulated and 409 downregulated genes in response to HPTM + L treatment (Figure 7c), indicating that this therapeutic strategy profoundly reshaped the tumor transcriptome.

FIGURE 7.

RNA‐seq analysis of the potential therapeutic mechanism of HPTM. (a) Heatmap of Pearson correlation analysis between biological replicates in the control group and the HPTM + L treatment group (n = 3). (b) PCA illustrated the distribution of DEGs in the control and HPTM + L groups. (c) Volcano plot of DEGs in the control and HPTM + L groups. (d) GO enrichment analysis circle plot displayed the 15 categories with the smallest q‐values. (e) Bubble plot of DEGs associated with multiple signaling pathways. (f) Bubble plot of DEGs enriched in KEGG. (g) KEGG chord diagram depicting the DEGs associated with specific pathways. Statistical analysis was performed by DESeq2 with adjusted q‐value < 0.05.

To delineate the biological roles of the differentially expressed genes (DEGs) identified in the treated group relative to the control group, Gene Ontology (GO) functional enrichment analysis was performed (Figure 7d). The data revealed that DEGs were significantly clustered into pathways associated with immune responses, cell death, and apoptosis (Figure 7e). It suggested that HPTM + L exerted antitumor effects through the inhibition of cell proliferation and the activation of apoptotic programs. Concomitantly, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment profiling highlighted the enrichment of core signaling pathways, such as MAPK, NF‐κB, PI3K‐Akt, p53, and TNF signaling pathways, which have been validated to modulate apoptosis and immune responses (Figure 7f). Notably, 41 DEGs were mapped to apoptosis‐, HIF‐1‐, TNF‐, MAPK‐, and p53‐related signaling pathways. Downregulation of the HIF‐1 and NF‐κB signaling pathways was associated with reduced PD‐L1 expression and inhibition of immune escape. HIF‐1α ASO alleviated the hypoxic microenvironment and regulated endogenous PD‐L1 via the HIF‐1α/PD‐L1 signaling axis for enhanced PIT. Moreover, several key DEGs involved in the PI3K‐Akt and p53 pathways, such as CACNG6, CSF3, and PMAIP1, exhibited significant dysregulation following HPTM + L treatment (Figure 7g). These transcriptomic analyses demonstrated that laser‐activated HPTM profoundly altered gene expression in tumor tissues, leading to the activation of apoptotic and immune signaling pathways, thus contributing to its potent antitumor efficacy.

3. Conclusion

Herein, we developed a DNA tetrahedron‐based nanoplatform (HPTM) to strengthen the therapeutic efficacy of PIT by overcoming two critical challenges, photosensitizer ACQ and the immune imbalance of TME. Incorporation of the photosensitizer TMPyP into the DNA tetrahedron framework via intercalation effectively prevented ACQ, thus improving ROS generation upon 660 nm light activation. By blocking the PD‐1/PD‐L1 pathway with PD‐L1 aptamers and regulating PD‐L1 expression through HIF‐1α suppression, this endogenous and exogenous synergistic strategy effectively alleviated the immunosuppressive TME for immunotherapy. Furthermore, the regulation of HIF‐1α in tumor tissues additionally relieved the hypoxic microenvironment induced by PDT and improved the effect of PDT. As a result, HPTM markedly enhanced PIT efficacy, achieving both direct tumor cell destruction and immune activation. These findings highlight HPTM as a promising platform for synergistically improving PIT outcomes and advancing the development of effective cancer therapies.

4. Materials and Methods

4.1. Materials

All oligonucleotides utilized in the present work were synthesized and purified by high‐performance liquid chromatography (HPLC) and commercially obtained from Sangon Biotech Co., Ltd. (Shanghai, China). The sequences of all oligonucleotides were listed in Table S2. Photosensitizer 5,10,15,20‐Tetrakis (1‐methyl‐4‐pyridinio) porphyrin tetra (p‐toluenesulfonate) (TMPyP) and singlet oxygen probe 1,3‐diphenylisobenzofuran (DPBF) were purchased from Macklin Biochemical Technology Co., Ltd. (Shanghai, China). The mitochondrial membrane potential assay kit, cell counting kit‐8 (CCK‐8), adenosine triphosphate (ATP) assay kit, lactate dehydrogenase (LDH) assay kit, reactive oxygen species (ROS) assay kit, Calcein‐AM/PI live/dead cell staining kit, Annexin V‐FITC apoptosis detection kit, and primary antibodies against HIF‐1α, PD‐L1, B‐cell lymphoma 2 (Bcl‐2), Bcl‐2‐associated X protein (Bax), and β‐actin were purchased from Beyotime Biotechnology Co., Ltd. (Shanghai, China). Fluorochrome‐conjugated antibodies for flow cytometry, including APC anti‐Mouse CD3, PE anti‐Mouse CD4, FITC anti‐Mouse CD8, PE anti‐Mouse CD86, FITC anti‐Mouse CD80, APC anti‐Mouse F4/80, and FITC anti‐Mouse CD206, were obtained from Elabscience Biotechnology Co., Ltd. (Wuhan, China). Enzyme‐linked immunosorbent assay (ELISA) kits for mouse TNF‐α, IL‐6, and IL‐12 were purchased from Ruixin Biotech Co., Ltd. (Quanzhou, China). Horseradish peroxidase (HRP)‐conjugated goat anti‐rabbit secondary antibody and goat anti‐mouse secondary antibody were purchased from Signalway Antibody LLC (College Park, MD, USA). Fetal bovine serum (FBS), Dulbecco's Modified Eagle Medium (DMEM), RPMI 1640 medium, penicillin‐streptomycin solution, and trypsin‐EDTA digestion solution were purchased from Gibco (Thermo Fisher Scientific, Waltham, MA, USA). 4% paraformaldehyde, DAPI nuclear staining solution, TRITC phalloidin for cytoskeleton staining, and all other analytical grade chemical reagents were purchased from Sigma‐Aldrich (Merck KGaA, Darmstadt, Germany).

4.2. Cell Lines

Mouse lung epithelial cells (MLE‐12), RAW264.7 macrophage cell line, DCs dendritic cell line, and Lewis lung carcinoma (LLC) cell line were purchased from the Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences (Beijing, China). Cytotoxic T lymphocyte line 2 (CTLL‐2) was purchased from Procell Life Science & Technology Co., Ltd. (Wuhan, China).

MLE‐12, RAW264.7, DCs, and LLC cells were cultured in DMEM supplemented with 10% heat‐inactivated FBS and 1% penicillin‐streptomycin solution. CTLL‐2 cells were cultured in RPMI 1640 medium supplemented with 10% FBS and 1% penicillin‐streptomycin solution. All cells were incubated at 37°C in a humidified incubator with 5% CO₂, and all cell lines were verified to be free of mycoplasma contamination before formal experiments.

4.3. The Preparation of HPTM

HIF‐1α ASO and PD‐L1‐targeting aptamer‐modified DNA tetrahedral structure (HPT; all abbreviations used herein are summarized in Table S1) were constructed via Watson‐Crick base pairing. Briefly, four single‐stranded oligonucleotides (2 µM) in Tris‐EDTA (TE) buffer were incubated at 95°C for 10 min, then cooled at 4°C for 30 min to form HIF‐1α ASO modified DNA tetrahedron (HT). The nucleotide sequences were listed in Table S2. Subsequently, 100 µL of 1 µM PD‐L1 aptamer was introduced into 100 µL of the aforementioned HT solution (500 nM), with the solution incubated for 2 h at 37°C to yield HPT. For loading the photosensitizer, 200 µL of TMPyP (250 nM) was introduced into HPT solution, and the solution was stirred at 25°C for 10 h. Ultimately, free TMPyP was eliminated via centrifugal separation (12 000 rpm, 15 min) to yield HPTM. HPTM was characterized by 8% polyacrylamide gel electrophoresis, atomic force microscopy (AFM), UV–vis spectroscopy, dynamic light scattering (DLS), and circular dichroism (CD).

4.4. Determination of the Optimal Proportion for HPTM Assembly

1 µM HPT was employed to prepare a series of HPTM complexes with molar ratios of 1:1, 1:5, 1:10, 1:15, and 1:20 between HPT and TMPyP. Centrifugation of the mixtures was carried out at 12 000 rpm for 15 min to remove unbound TMPyP, and the resulting HPTM complexes with different HPT‐to‐TMPyP molar ratios were resuspended. The encapsulation efficiency of TMPyP was quantified via UV–vis spectrophotometry based on a pre‐established standard curve of TMPyP absorbance at 422 nm. The hydrodynamic diameter of the assembled products at each ratio was measured via DLS at 25°C. The ROS generation capacity of each group was evaluated using the DPBF probe by monitoring the decrease in absorbance of DPBF at 410 nm under 660 nm laser irradiation. The storage stability of each group was assessed by monitoring the changes in hydrodynamic diameter via DLS every day during 7 days of storage at 4°C. The encapsulation efficiency (EE) was determined via the equation presented as follows:

$$\mathrm{EE}=\frac{c_{loaded}}{c_{total}}\times 100\%$$

4.5. The Generation of ROS In Vitro

DPBF was solubilized in dimethyl sulfoxide (DMSO) and served as a specific probe for monitoring the production of ROS. TMPyP and HPTM were introduced into the above DMSO solution, respectively. TMPyP and HPTM were kept at concentrations of 2 µM, while DPBF was kept at 50 µM. After exposure to 660 nm laser irradiation at 20 mW/cm², the absorbance variations at 415 nm were recorded via UV–vis spectroscopy.

4.6. pH‐Responsive PD‐L1 Aptamer Release Assay

The pH‐responsive release profile of PD‐L1 aptamer from HPT was evaluated across a series of physiologically relevant pH values (pH 6.0, 6.5, 7.0, and 7.4). HPT was pre‐labeled with FAM on S3 and BHQ2 on the PD‐L1 aptamer. The pre‐labeled HPT solution was added to buffers at various pH values, and the resulting mixture was incubated at 37°C with constant shaking. The fluorescence intensity was subsequently measured using a fluorescence spectrophotometer. The reversible pH‐responsive property of the system was assessed via cyclic pH switching experiments. Briefly, HPT solution was alternately incubated in 1 × TE buffer at pH 7.4 and pH 6.5 for 1 h per pH condition, with 5 consecutive cycles. The changes in fluorescence intensity were monitored at the end of each incubation cycle using the same instrument.

4.7. In Vitro Cellular Uptake

The PD‑L1 targeting ability of HPTM was assessed by CLSM. LLC cells were plated into confocal dishes with a density of 1 × 10⁵ cells per dish and cultured for 12 h. The original culture medium was thereafter substituted with fresh culture medium supplemented with 6 µM TMPyP, 1 µM PTM and 1 µM HPTM. Following 4 h of co‐incubation, the culture medium was aspirated, and cells were rinsed three times with PBS. The cells were subsequently fixed using 4% paraformaldehyde, stained with TRITC (for F‐actin) and DAPI (for nuclei), and analyzed for cellular uptake with CLSM.

4.8. Tumor Microspheres Model Analysis

LLC cell suspensions (3000 cells per well) were plated into 96‐well ultra‐low‐attachment plates and incubated for 48 h. After mature spheroids had formed, the samples were incubated with 200 nM TMPyP or HPTM for 4 h. CLSM was used to acquire z‐stack images with 20 µm z‐axis intervals.

4.9. Intracellular ROS Generation

DCFH‑DA served as the probe for ROS detection in vitro. LLC cells were plated onto glass‐bottom confocal culture dishes at a seeding density of 1 × 10⁵ cells in each dish, followed by incubation for 12 h. After that, cells were divided into eight separate experimental groups: (1) PBS, (2) TMPyP, (3) PTM, (4) HPTM, (5) PBS + Laser, (6) TMPyP + Laser, (7) PTM + Laser, and (8) HPTM + Laser. Following 4 h of incubation, the laser‐treated groups received 660 nm laser irradiation at 200 mW/cm² for 5 min. The concentrations of TMPyP, PTM, and HPTM were kept at 6, 1, and 1 µM in all the in vitro experiments, respectively. After a 2 h pre‐incubation step, the cells were treated with 10 µM DCFH‐DA, with incubation performed at 37°C for a 20 min duration. Following medium removal, the cells were washed 3 times using PBS.DCF green fluorescence was analyzed using CLSM.

4.10. Western Blotting and Real‐Time‐qPCR

At a density of 1 × 10⁶ cells per well, LLC cell suspensions were plated into 6‐well plates and incubated overnight. The treatment protocol was consistent with that described in the “Intracellular ROS generation” section. After the designated treatment, LLC cells underwent lysis on ice for 15 min in cell lysis buffer containing protease inhibitor phenylmethanesulfonyl fluoride (PMSF), prior to centrifugation at 12 000 rpm for 10 min under 4°C. After adding 6 × loading buffer at a 5:1 volume ratio (v/v), the extracted proteins were denatured through 10 min of boiling at 95°C. Protein separation was performed via SDS‐PAGE, with running time adjusted according to the target protein's molecular weight and gel concentration. The resolved proteins were subsequently blotted onto polyvinylidene difluoride (PVDF) membranes under a constant current of 400 mA for 30 min. The PVDF membranes were blocked with 1× Tris‐buffered saline containing Tween 20 (TBST) plus 5% bovine serum albumin (BSA) at 25°C for 2 h. Incubation with primary antibodies (1:1000 dilution in TBST) was performed overnight at 4°C, and then washed three 5‐min washes with TBST. Subsequently, the membranes underwent incubation for 2 h at 25°C with shaking, using horseradish peroxidase (HRP)‐conjugated goat anti‐rabbit or anti‐mouse secondary antibodies diluted 1:2000 in TBST. β‐actin served as a loading control. Finally, protein expression levels were quantified using an enhanced chemiluminescence detection system (ECL).

For RT‐qPCR, cells were pretreated as described above. Using the HP Total RNA Kit, total RNA was isolated according to the manufacturer's instructions. 1 µg of extracted total RNA was reverse‐transcribed into complementary DNA (cDNA) with the PrimeScript First‐Strand cDNA Synthesis Kit. Quantitative PCR was then performed using the TB Green Premix Ex Taq. The ΔCt method was applied to calculate the relative expression levels of target genes. scHPTM was prepared identically to HPTM and subjected to the same RT‐qPCR assays described above. The nucleotide sequences of the forward and reverse primers used in the assays were detailed in Table S2.

4.11. Immunofluorescence Assay

LLC cells were seeded into confocal microscopy‐compatible culture dishes at a density of 1 × 10⁵ cells per dish and incubated for 12 h to allow adherence. The treatment protocol was consistent with that described in the “Intracellular ROS generation” section. Following a further 2 h incubation, cells were blocked using 10% goat serum for 20 min at 25°C, washed with PBS, and then subjected to overnight incubation at 4°C with antibodies against CRT, HMGB1, and PD‐L1. Primary antibodies were diluted in TBST solution for a 200‐fold dilution. The next day, following equilibration of the cells at 25°C for 2 h, the cells were washed and probed with species‐appropriate secondary antibodies (1:500 dilution) for 2 h at 37°C. Cytoskeletal staining was performed using TRITC for 20 min, after which cell nuclei were stained with DAPI for a 10 min period. Following rinsing with PBS, fluorescent images were captured via CLSM.

4.12. Apoptosis Assay

Apoptosis of LLC tumor cells triggered by HPTM was determined using the Annexin V‐FITC apoptosis detection kit. For this assay, LLC cells were inoculated into 6‐well plates and cultivated for 12 h. Then, the cells were subjected to the following eight treatments: (1) PBS, (2) TMPyP, (3) PTM, (4) HPTM, (5) PBS + Laser, (6) TMPyP + Laser, (7) PTM + Laser, and (8) HPTM + Laser. The radiation intensity of the 660 nm laser was 200 mW/cm² for 5 min, and concentrations of TMPyP, PTM, and HPTM were kept at 6, 1, and 1 µM in all the in vitro experiments, respectively. After an extra 12 h incubation period, trypsin digestion was performed on cells, which were thereafter washed 3 times using PBS to clear away leftover trypsin. Following the preceding experimental steps, the pre‐formulated Annexin V‐FITC/PI staining solution was applied for cell staining over 15 min, in full compliance with the operating instructions detailed in the kit manual, and the cells were assessed via flow cytometry to quantify apoptotic cell populations.

4.13. Cytotoxicity Assay

The CCK‐8 assay was performed to determine cellular cytotoxicity in 96‐well plates inoculated with LLC cell suspensions at a density of 1 × 10⁴ cells per well, followed by overnight incubation to achieve complete cell adhesion. The cells were treated with various treatments of (1) PBS, (2) TMPyP, (3) PTM, (4) HPTM, (5) PBS + Laser, (6) TMPyP + Laser, (7) PTM + Laser, and (8) HPTM + Laser. The radiation intensity of 660 nm laser was 200 mW/cm² for 5 min, and concentrations of TMPyP, PTM, HPTM were kept at 6, 1, and 1 µM, respectively. An additional 12 h of incubation was carried out after the preceding experimental steps. Prior to cytotoxicity assessment, following the addition of 10 µL CCK‐8 reagent to each well, the culture plates were kept for a 1 h period at 37°C under light‐protected conditions. Absorbance at 450 nm was measured via a microplate reader to quantitatively determine the viability of the treated cells.

4.14. Live/Dead Cell Viability Staining Assay

Single‐cell suspensions of LLC cells were inoculated into 6‐well plates and cultivated for a 12 h period. The treatment protocol was consistent with previous treatments. The incubation period was extended for another 12 h. The cells of eight groups were rinsed 3 times with PBS and stained using a pre‐prepared Calcein‐AM/PI dual‐staining solution. After co‐incubating the cells with the staining solution over a 30 min period, the residual fluorescent probe was thoroughly removed by washing. A fluorescence microscope was then utilized to capture fluorescence images for both viable and dead cells in the test samples.

4.15. Mitochondrial Membrane Potential

JC‐1 assay kit was utilized to determine the mitochondrial membrane potential of harvested cells. The treatment protocol was consistent with the above description, and the cells were treated with (1) PBS, (2) TMPyP, (3) PTM, (4) HPTM, (5) PBS + Laser, (6) TMPyP + Laser, (7) PTM + Laser, and (8) HPTM + Laser. After incubation for an additional 2 h, JC‐1 staining was carried out following the manufacturer's instructions included in the JC‐1 detection kit. Subsequently, alterations in mitochondrial membrane potential were visualized and documented via CLSM.

4.16. Extracellular ATP Assays

A commercial ATP detection kit was utilized to determine the level of extracellular ATP secreted by the treated cells. For this assay, LLC cell suspensions were inoculated into 96‐well plates and cultivated overnight to achieve full cell adherence. Following exposure to the eight treatments described below (1) PBS, (2) TMPyP, (3) PTM, (4) HPTM, (5) PBS + Laser, (6) TMPyP + Laser, (7) PTM + Laser, and (8) HPTM + Laser, the incubation was continued for an additional 2 h, and the culture medium was carefully collected. ATP content in the supernatant was finally detected per the manufacturer's protocol included in the commercial detection kit.

4.17. The Detection of DCs Maturation In Vitro

A transwell chamber was inserted into each well of a 12‐well plate. DCs were plated into the lower compartment of the transwell system. LLC cells were plated into the upper compartment and subsequently treated with various groups as mentioned above. Following irradiation, the co‐culture system was maintained for an additional 24 h, allowing interaction between the treated LLC cells and the underlying DCs. Following the completion of the incubation period, the DCs in the lower compartment were harvested, washed 3 times with PBS, and immunostaining was performed using fluorochrome‐conjugated antibodies targeting CD86 and CD80 on the harvested cell samples. Thereafter, the expression profiles of these cell surface markers were quantified by flow cytometry to evaluate the activation status of DCs. To exclude the direct stimulatory effect of HPTM on DCs, a tumor cell‐free negative control group was also conducted. DCs were co‐incubated with 1 µM HPTM for 24 h directly.

4.18. The Assessment of T Cell Recognition and Cytotoxicity In Vitro

Prior to co‐culture, the activation status of T cells was assessed by flow cytometry. CTLL‐2 cells were incubated with CD25 antibody at 4°C in the dark for 20 min, then washed twice with ice‐cold PBS to remove unbound antibody. The cells were subsequently analyzed by flow cytometry. To evaluate T cell activation via ICD induced by HPTM under 660 nm irradiation, LLC cells were placed into 6‐well plates and cultured for 12 h. Then, cells were subjected to the various treatments described above. Incubation was continued for an additional 12 h. After various treatments, the supernatants from each group were collected and used to co‐incubate with DCs to promote their maturation. Following 24 h of co‐culture, CTLL‐2 cells were added to the system and incubated for 48 h for activation. The activation status of CTLL‐2 cells was subsequently analyzed by flow cytometry. Afterward, the activated CTLL‐2 cells (defined as effector cells) were collected and co‐incubated with LLC cells (defined as target cells) in 96‐well plates overnight at an effector‐to‐target ratio of 1:1. The cytotoxic effects of the activated T cells on LLC cells were quantified using an LDH cytotoxicity assay kit.

To evaluate the influence of HPT on T cell‐mediated tumor cell killing, LLC cells were placed into a 6‐well plate overnight and treated with PBS, 1 µM PT, 1 µM HPT, and 2 µM FDA‐approved PD‐L1 antibody drug atezolizumab. After incubation for another 12 h, the supernatants were aspirated and discarded, and activated T cells were added to the system and co‐incubated overnight at an effector‐to‐target ratio of 1:1. The cytotoxic effects of the activated T cells on LLC cells were quantified using an LDH cytotoxicity assay kit.

4.19. Hemolysis Analysis

To evaluate the hemocompatibility of HPTM and its suitability for in vivo tumor therapy, an in vitro hemolysis assay was performed. Fresh mouse blood anticoagulated with ethylenediaminetetraacetic acid (EDTA) was collected, and centrifugation at 2000 rpm for a 5 min duration was performed to isolate red blood cells (RBCs) from the processed samples. The RBCs were then washed repeatedly with PBS until the supernatant became clear. The RBC suspension was diluted to an appropriate concentration, and 200 µL of the diluted RBC solution was individually mixed with PBS, deionized water, or varying concentrations of HPTM (5, 10, 20, 30, 40, and 50 µM), yielding a final volume of 1 mL in each tube. After incubation at a constant 25°C for 4 h, the mixtures were centrifuged to sediment undamaged RBCs. A microplate reader was then utilized to detect the absorbance of the harvested supernatant at the 541 nm detection wavelength.

4.20. In Vivo Anti‐Tumor Effect Evaluation

To establish a tumor‐bearing mouse model, 100 µL of LLC cell suspension (containing 1 × 10⁶ cells per animal) was subcutaneously administered into the right hind thigh of six‐week‐old male C57BL/6 mice. All experiments were performed on 6‐week‐old male C57BL/6 mice with an initial body weight of 18–22 g at tumor inoculation, purchased from GemPharmatech Co., Ltd., with SPF‐grade certification. Mice were housed in an SPF‐grade barrier environment at Chongqing University School of Medicine, with constant temperature (22 ± 2°C), relative humidity (50% ± 5%), a strict 12 h light/12 h dark cycle, and ad libitum access to sterilized chow and water. All animal procedures strictly complied with the ethical guidelines approved by the Chongqing University Laboratory Animal Welfare and Ethics Committee (Approval No. CQU‐IACUC‐RE‐202211‐002). Strict randomization was implemented to avoid selection bias. When the subcutaneous tumor volume reached 80 mm³, tumor‐bearing mice were randomly assigned to 4 experimental groups (n = 5). Each mouse received an intravenous injection via the tail vein of 80 µL drug solutions with certain concentrations (PBS, TMPyP: 60 µM, PTM: 10 µM, HPTM: 10 µM). Therapeutic interventions were performed according to the predefined treatment schedule. For groups subjected to laser irradiation, tumor sites were irradiated with a 660 nm laser at a power density of 0.2 W/cm² for 10 min. For a total study period of two weeks, tumor volumes and body weights were recorded every other day during the full treatment cycle. The tumor volume was calculated using the following formula: tumor volume = (L × W²)/2 (L = tumor major axis, W = tumor minor axis). At the experimental endpoint, blood samples were collected via retro‐orbital bleeding. The levels of AST, ALT, ALP, ALB, CK, LDH, CREA, UREA, and functional blood cells and proteins were determined by a seralyzer for complete blood count and serum biochemical analyses. Following humane euthanasia of all animals, tumors were resected, weighed, and photographed for experimental documentation. Major organs, including the heart, liver, spleen, lungs, and kidneys, were collected and processed for hematoxylin and eosin (H&E) staining for histopathological evaluation. Tumor tissues were further analyzed using H&E staining, Ki67 immunohistochemistry for proliferative activity, terminal deoxynucleotidyl transferase‐mediated dUTP nick‐end labeling (TUNEL) assay for apoptosis, and immunofluorescence staining for PD‐L1, HIF‐1α, HMGB1, and CRT expression.

4.21. Immune Response Evaluation In Vivo

Consistent with the experimental protocol used for antitumor efficacy assessment, C57BL/6 mice were randomly assigned to four groups a week after subcutaneous inoculation with LLC cells for immune response evaluation in vivo. Each group received the assigned treatments, and laser irradiation was applied three times according to the experimental protocol. On day 5 of the treatment administration period, blood was harvested from the retro‐orbital sinus of five mice per group for ELISA to quantify the levels of IL‐6 and TNF‐α. Following sample collection, the mice were humanely euthanized, and the right inguinal lymph nodes were harvested for flow cytometry analysis to assess the maturation status of DCs in every experimental group. In addition, tumor tissues were collected to evaluate the immune landscape, including macrophage polarization and infiltration of T lymphocytes, using flow cytometric profiling.

4.22. In Vivo Anti‐Lung Metastasis Efficacy Evaluation

The lung metastasis model was developed to assess anti‐metastasis efficacy. LLC cells (100 µL, 1 × 10⁵ cells) were administered via the tail vein into the treated mice (HPTM + L) (n = 5). Fourteen days later, the mice were euthanized and excised to remove lungs for tumor metastasis observation. The removed lungs in various treatment groups were soaked in Bouin's solution, and the pulmonary nodules were counted. Tumor metastasis lesions of the lungs were also evaluated via H&E staining. Flow cytometry was also employed to analyze DCs and effector memory T cells isolated from the spleens.

4.23. Transcriptomic Analysis

Tumor tissue samples were collected from LLC tumor‐bearing C57BL/6 mice in the PBS + L group and the HPTM + L group at the experimental endpoint (n = 3), immediately snap‐frozen in liquid nitrogen and stored at −80°C until use. Total RNA was extracted from frozen tumor tissues using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's standard protocol, the purity and concentration of extracted RNA were determined with a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and RNA integrity was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Only RNA samples with an RNA Integrity number (RIN) ≥ 8.0 were used for subsequent library construction. Strand‐specific mRNA libraries were constructed using the VAHTS Stranded mRNA‐seq Library Prep Kit for Illumina V2 (Vazyme Biotech Co., Ltd, Nanjing, China) according to the manufacturer's instructions, including mRNA purification, double‐stranded cDNA synthesis, adapter ligation, fragment size selection, and PCR amplification. Qualified libraries were sequenced on the Illumina NovaSeq 6000 platform with a paired‐end 150 bp (PE150) strategy, and library construction and sequencing were completed by OE Biotech Co., Ltd. (Shanghai, China). For bioinformatic analysis, raw sequencing reads were processed for quality control using fastp (v0.39). Clean reads were aligned to the mouse reference genome (GRCm39/mm39) using HISAT2 (v2.2.1). Gene expression levels were quantified with featureCounts (v2.0.1). Differentially expressed genes (DEGs) were identified using DESeq2 (v1.38.3) with the threshold of |log2 fold change| ≥ 1 and adjusted q‐value < 0.05. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional enrichment analysis of DEGs were performed using clusterProfiler (v4.6.2), and all statistical analysis and data visualization were completed using R (v4.2.3).

4.24. Data Analysis

All schematic illustrations, including the Table of Contents graphic, Scheme 1, and all mechanistic schematic diagrams in the figures, were created, edited, and optimized using Adobe Photoshop 2018 (Adobe Inc., San Jose, CA, USA). The layout, typesetting, and integrated arrangement of all visual elements in figures and schemes were completed using Adobe Illustrator 2022 (Adobe Inc., San Jose, CA, USA). For quantitative data processing and statistical plotting, all experimental data were analyzed, and statistical graphs including bar charts, line graphs, and quantitative analysis plots were generated using GraphPad Prism 10.1.2 (GraphPad Software, San Diego, CA, USA). For other visualization and data analysis work, the molecular docking simulation and visualization of the binding mode between TMPyP and DNA tetrahedral frameworks were performed via AutoDock Vina 1.2.3 for docking calculation, and PyMOL 2.5 (Schrödinger, LLC, New York, NY, USA) for structural rendering and interaction visualization. Image processing and quantitative analysis of confocal laser scanning microscopy images and western blot bands were conducted with ImageJ 1.54f (National Institutes of Health, Bethesda, MD, USA). The raw data acquisition and analysis of flow cytometry assays were performed using FlowJo v10.8.1 (BD Biosciences, Franklin Lakes, NJ, USA).

Independent replicate experiments were performed for all in vitro assays, with a minimum of 3 replicates per assay (n ≥3). All in vivo animal studies included 5 mice per experimental group (n = 5). All quantitative data throughout the manuscript were presented as mean ± standard deviation (SD), and error bars in all figures denote mean ± SD unless otherwise specified. All statistical tests were performed according to pre‐specified criteria. The unpaired Student's t‐test was used for comparisons between two independent experimental groups, while one‐way analysis of variance (ANOVA) was applied for comparisons among three or more groups. Statistical significance was defined as ^* p < 0.05, ^** p < 0.01, and ^*** p < 0.001. Non‐significant differences were denoted as ns (not significant). All statistical analyses were performed using GraphPad Prism 10.1.2.

Author Contributions

Jichun Yang: Writing – review & editing, Methodology, Formal analysis, Supervision, Resources, Funding acquisition, Conceptualization. Xin Cui: Writing – original draft, Methodology, Formal analysis, Conceptualization. Danfeng Zhang: Writing – original draft, Methodology, Formal analysis, Conceptualization. Yang Wang: Writing – original draft, Methodology, Formal analysis, Conceptualization. Yao Luo: Investigation, Data curation, Conceptualization. Sitong Wang: Validation, Methodology, Conceptualization. Linke Shi: Validation, Methodology.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: adhm71350‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.