NIR-II imaging-guided nanoplatform for synergistic mitochondria-targeted pyroptosis and macrophage reprogramming immunotherapy

Abstract

Cancer immunotherapy has revolutionized modern oncology by mobilizing the body’s immune system, yet its efficacy remains severely limited in immunologically “cold” tumors, which are defined by poor immune infiltration and low tumor immunogenicity. Here, we report a multi-functional nanoplatform that integrates a new second near-infrared (NIR-II) aggregation-induced emission luminogen (AIEgen), a mitochondria-targeted lonidamine prodrug, and cryo-shocked M1 macrophage membranes (CSMs) to achieve synergistic tumor microenvironment (TME) reprogramming and precision image-guided immunotherapy. The bright NIR-II AIEgen enables high-resolution fluorescence and photoacoustic imaging for real-time tumor visualization and photothermal therapy. The prodrug LND-1-PEG-24, cleavable by TME-overexpressed cathepsin B, preferentially accumulates in mitochondria to trigger caspase-3/GSDME-mediated pyroptosis, leading to the release of danger-associated molecular patterns that markedly enhance tumor immunogenicity. Simultaneously, CSMs promote durable polarization of tumor-associated macrophages (TAMs) toward the tumoricidal M1 phenotype via the TLR2/MAPK pathway, thereby alleviating TME immunosuppression. In tumor-bearing mice, this nanoplatform synergistically enhances cytotoxic T cell infiltration, reverses immune suppression, and effectively inhibits both primary tumor growth and metastatic progression through the activation of systemic antitumor immunity. This work establishes a versatile strategy that unifies NIR-II phototheranostics, mitochondria-targeting pyroptosis, and TAM reprogramming, providing a robust and targeted approach for cancer immunotherapy.

Graphical abstract

We report a multifunctional nanoplatform that integrates an NIR-II aggregation-induced emission luminogen, a mitochondria-targeted prodrug, and M1 macrophage membranes, enabling high-resolution NIR-II fluorescence and photoacoustic imaging, photothermal therapy, pyroptosis-induced immunogenicity, and tumor-associated macrophage reprogramming. This synergistic strategy enhances T cell infiltration, reverses immunosuppression, and suppresses tumor growth and metastasis, offering precise, durable, and systemic cancer immunotherapy.

Highlights

• •A novel AIEgen enables FLI/PAI/PTI imaging and PTT therapy.
• •A TME-responsive mitochondria-targeting prodrug triggers caspase-3/GSDME pyroptosis.
• •This nanoplatform enables TAM reprogramming for immunotherapy.
• •This nanoplatform enhances cytotoxic T-cell infiltration and suppresses metastasis.

1. Introduction

Malignant tumors remain a formidable and escalating threat to global health, underscoring the urgent need for precision diagnostic tools and innovative therapeutic strategies that transcend the limitations of conventional interventions [[1], [2], [3]]. Over the past decades, cancer immunotherapy has profoundly reshaped oncology by mobilizing the host immune system to selectively recognize and eradicate malignant cells [[4], [5], [6]]. Despite these advances, its clinical efficacy is markedly constrained by the poor response rates observed in a large subset of patients [7]. This resistance is primarily driven by the prevalence of immunologically “cold” tumors, which are characterized by scarce tumor-infiltrating lymphocytes (TILs) and an abundance of immunosuppressive cell populations within the tumor microenvironment (TME) [8,9]. Addressing this challenge requires restoring tumor immunogenicity and reprogramming the TME toward an immune-activated state, thereby enabling more effective and personalized immunotherapy [10,11].

Induction of immunogenic cell death (ICD) has emerged as a promising strategy to enhance tumor immunogenicity and amplify antitumor immunity [12,13]. During ICD, dying tumor cells release tumor-associated antigens (TAAs) and damage-associated molecular patterns (DAMPs), which function as “danger signals” to recruit and activate immune cells [[14], [15], [16]]. This process promotes cytotoxic T lymphocyte (CTL) infiltration, ultimately converting immune-desert “cold” tumors into immune-active “hot” tumors [[17], [18], [19]]. Among ICD modalities, pyroptosis, a highly inflammatory form of programmed cell death mediated by gasdermin proteins (e.g., GSDMD and GSDME), has gained increasing attention [20,21]. Upon cleavage and activation, gasdermins oligomerize to form membrane pores, causing rapid lytic cell death and extensive DAMP release, thereby magnifying tumor immunogenicity [[22], [23], [24]]. Recent evidence further indicates that mitochondrial permeability transition (MPT) serves as a potent trigger of cell death pathways with superior efficacy and speed [25,26]. Consequently, the development of mitochondria-targeted pyroptosis inducers represents a critical unmet demand in advancing cancer immunotherapy [[27], [28], [29]].

In parallel, tumor-associated macrophages (TAMs), a predominant immune population within the TME, play a decisive role in shaping the immune milieu. Under the influence of tumor-derived metabolites (e.g., lactate) and signaling cues, TAMs are frequently polarized toward an M2-like phenotype, which fosters tumor progression and suppresses TIL function, thereby exacerbating immune evasion [30,31]. While reprogramming TAMs toward the tumoricidal M1 phenotype has shown encouraging potential in stimulating T cell infiltration, enhancing antitumor immunity, and establishing durable immune memory [32], strategies that safely and effectively amplify this effect remain limited. Intriguingly, recent studies have revealed that macrophages undergoing regulated cell death can robustly promote TAM M1 polarization while preserving biocompatibility [33,34]. This observation resonates with the ICD paradigm, in which the release of TAAs and DAMPs further augments antitumor immune responses [35,36]. Based on this rationale, we hypothesize that engineered apoptotic M1 macrophages could act as safe and potent triggers to reprogram TAMs, thereby synergizing with tumor immunogenic restoration to overcome TME immunosuppression [[37], [38], [39], [40]].

To maximize therapeutic efficacy, precise spatiotemporal monitoring of tumor localization and immune response is indispensable. Precision image-guided immunotherapy has thus emerged as a transformative strategy for improving treatment outcomes. In this context, fluorescence imaging in the second near-infrared window (NIR-II, 1000–1700 nm) has attracted growing interest owing to its deep tissue penetration, high sensitivity, and real-time imaging capability. Nevertheless, conventional NIR-II fluorophores often suffer from aggregation-caused quenching and systemic toxicity, which limit their translational potential [41,42]. Aggregation-induced emission (AIE) has provided an elegant solution by transforming aggregation into an advantage, enabling bright emission and improved photostability [43,44]. Furthermore, NIR-II AIE luminogens (AIEgens) often possess strong NIR absorption and efficient nonradiative decay pathways, making them suitable for complementary photoacoustic (PA) imaging and photothermal therapy (PTT) [[45], [46], [47]]. PA imaging offers high spatial resolution and favorable penetration depth, thereby providing structural context to fluorescence signals, while PTT enables precise, noninvasive ablation of tumors and can concurrently trigger ICD to augment immune activation [[48], [49], [50]].

In this contribution, we present a mitochondria-targeting, pyroptosis-inducing nanoplatform that integrates a NIR-II AIEgen, a TME-responsive prodrug, and cryo-shocked M1 macrophage (CSM1) membranes. As outlined in Fig. 1, we first designed a novel AIEgen with bright NIR-II emission and excellent photothermal conversion efficiency. We then synthesized a TME-activated prodrug (LND-1-PEG-24) by conjugating lonidamine (LND) to polyethylene glycol (PEG) through a cathepsin B (CTSB)-cleavable peptide linker. Encapsulation within CSM1 membranes not only enhanced tumor-targeting capability but also promoted TAM repolarization toward the immunostimulatory M1 phenotype via the TLR2/MAPK pathway. This design facilitated mitochondrial delivery of LND-1-PEG-24, where LND triggered tumor-specific pyroptosis through the caspase-3/GSDME axis, accompanied by massive DAMP release. Collectively, this integrated nanoplatform, combining NIR-II fluorescence/PA imaging, PTT, and mitochondria-targeting pyroptosis, enabled precise tumor visualization, robust immune activation, and reversal of TME immunosuppression. By promoting durable immune memory against tumor recurrence and metastasis, this strategy introduces a paradigm shift in cancer immunotherapy through the synergistic convergence of NIR-II phototheranostics and mitochondrial pyroptosis induction.

Fig. 1.

Schematic diagram of a peptide self-assembly nanoplatform for synergistic tumor microenvironment reprogramming and precision NIR-II imaging-guided immunotherapy.

2. Results and discussion

2.1. Synthesis and properties of molecular probe

To construct a high-performance NIR-II AIEgen, we adopted a donor-acceptor (D-A) molecular design strategy, a well-established approach for tuning the photophysical properties. As illustrated in Fig. 2a, the alkoxylated triphenylmine was chosen as the D unit for the strong electron-donating ability and excellent solubility, and benzothiadiazoloselenadiazole (BTSe) served as the A moiety due to its highly conjugated framework and robust electron-withdrawing capability. This combination facilitates intramolecular charge transfer (ICT), thereby extending absorption and emission into the long-wavelength spectral region. Moreover, 2,3-dihydrothieno[3,4-b] [1,4]dioxine (EDOT) was introduced both as an auxiliary donor and a π-bridge. Importantly, the steric hindrance from the cyclohexyl ether group in EDOT perturbs molecular planarity, suppresses aggregation-caused quenching, and promotes AIE. The detailed syntheses and characterzations of the intermediates and TPA-Se were presented in Figs. S1–S10. Density function theory calculations were performed to elucidate the molecular geometry and electronic energy levels. As shown in Fig. 2b, TPA-Se adopted a highly twisted conformation, with dihedral angles exceeding 40° between EDOT and BTSe, which effectively hindered π–π stacking. The optical properties of TPA-Se were systematically evaluated to validate its optical features. As presented in Fig. 2c and d, TPA-Se in tetrahydrofuran (THF) displayed absorption maximum at ∼850 nm and emission maximum at ∼1100 nm, placing it within NIR-II window and favorable for deep-tissue imaging. Furthermore, the photoluminescence (PL) spectra in THF/water mixtures with varying water fractions (f_w) were recorded. As depicted in Fig. 2e and f, the PL intensity of TPA-Se increased significantly with f_w increasing from 20 % to 90 %, demonstrating a typical AIE behavior. Collectively, these results identify TPA-Se as a bright NIR-II AIEgen with superior optical properties, holding great promise for image-guided theranostics.

Fig. 2.

Photophysical properties of molecular probe and NPs. a) Synthetic route for TPA-Se. b) Molecular geometric configuration, HOMO, LUMO, and electronic levels of TPA-Se. c) Absorption and d) PL spectra of TPA-Se and ANPs. e) A plot of PL peak intensity versus the water fraction in THF/water mixtures, where I₀ and I represent the PL intensity of TPA-Se in pure THF and THF/water mixtures with various water fractions. f) PL intensity changes of TPA-Se in THF/water mixtures with different water fractions. g) Photothermal behavior of ANPs, which were irradiated with 808 nm light for 5 min, then the laser was removed, and the samples naturally cooled down. h) The calculation of τ_s from the linear regression curve in the cooling curve and represents the characteristic thermal time constant. i) Photothermal heating curves of ANPs across various concentrations (0–100 μg/mL based on TPA-Se) under the irradiation of 808 nm light (1.0 W/cm²). j) The photothermal heating curves of ANPs (50 μg/mL based on TPA-Se) under irradiation of 808 nm light wiht different power densities. k) Photothermal stability of ANPs and ICG aqueous solution during five cycles of 808 nm laser (1.0 W/cm²) on-off processes.

To improve solubility and biocompatibility, amphiphilic DSPE-PEG₂₀₀₀ was used to encapsulate TPA-Se via nanoprecipitation, yielding water-dispersible nanoparticles (ANPs). The absorption and emission spectra of ANPs closely resembled those of free TPA-Se, with only minor bathochromic shifts. Then we investigated the photothermal property of ANPs under 808 nm laser irradiation. As depicted in Fig. 2g, ANPs exhibited a rapid temperature rise upon laser exposure, reaching a stable plateau within 5 min, indicating efficient light-to-heat conversion kinetics (further details provided in the Supporting Information). The photothermal conversion efficiency of ANPs was calculated to be 53.9 % (Fig. 2h), a value substantially higher than that of most reported NIR photothermal agents and indicative of excellent PTT efficacy [51]. The amplitude of photo-induced temperature elevation increased with the concentrations of ANPs and the laser power intensity (Fig. 2i and j). Notably, the temperature reached approximately 80 °C when 100 μg/mL of ANPs (based on TPA-Se) were irradiated with 808 nm laser (1.0 W/cm²) for 5 min. ANPs had good stability as the photothertmal performance remained constant after five cycles of consecutive NIR laser irradiation. In stark contrast, indocyanine green (ICG), an FDA-approved agent widely used in clinic, exhibited a severe decline in photothermal activity after successive irradiation cycles, likely due to the photobleaching effect (Fig. 2k). PA signal is closely linked to the absorption and photothermal conversion properties, thus we further investigated the PA property. As shown in Fig. S11, ANPs showed maximal PA signal at about 880 nm, consistent with their absorption profile. We subsequently evaluated the photostability and photoacoustic (PA) stability of the molecules, using indocyanine green (ICG) as a reference. As shown in Figs. S12 and S13, the absorption of TPA-Se remained stable after 30 min of continuous light exposure. In contrast, under identical experimental conditions, the absorption of ICG significantly decreased, accompanied by a visible color change from green to yellow. Furthermore, TPA-Se exhibited a substantially stronger photoacoustic signal compared to ICG, with a more favorable emission wavelength for deep-tissue imaging applications. The fluorescence quantum yield of TPA-Se in the liquid state was determined to be 3.7 % (Fig. S14). Collectively, these results confirm that ANPs exhibit excellent NIR-II optical properties, high photothermal conversion efficiency, and robust photothermal stability, making them promising candidates for theranostic applications.

2.2. Characterization of nanoformulations and macrophage polarization mechanism

To enable TME-responsive drug release and synergistic therapy, we synthesized an activatable prodrug (LND-1-PEG-24) by conjugating LND modified with a mitochondria-targeted peptide (Phe-Arg-Phe-Lys) with PEG through a CTSB-cleavable peptide linker (Gly-Phe-Leu-Gly) [52]. The high expression of CTSB catalyzes the cleavage of Gly-Phe-Leu-Gly, while the sequence Phe-Arg-Phe-Lys can precisely target mitochondria, thereby enabling precise release of LND. This design leverages the overexpression of CTSB in TME, ensurig site-specific prodrug activation and minimizing off-target toxicity. Although CTSB is typically lysosome-confined, endosomal escape may allow its cytosolic release to activate the prodrug, enabling subsequent mitochondrial targeting and pharmacological action. The synthetic route and characterization of LND-1-PEG-24 were depicted in Figs. S15 and S16.

Self-assembly of TPA-Se with LND-1-PEG-24 yielded ALNPs, while LNPs (containing only prodrug) served as the control. To enhance the stability and targeting capability, ALNPs were cloaked with M1 macrophage membranes, a bioinspired strategy that leveraged the intrinsic homing ability of macrophages to tumor sites. Cryo-shocked M1 macrophages offer dual advantages: enhanced tumor targeting and evasion of phagocytic clearance, supporting more effective nanoparticle binding at tumor sites. To obtain M1 macrophage membranes, RAW264.7 cells were co-stimulated with lipopolysaccharide (LPS) and interferon-γ (IFN-γ) for 24 h, a well-established protocol to induce polarization toward the pro-inflammatory M1 phenotype. After stimulation, the cell supernatant was discarded, and M1 macrophage membranes were extracted via hypotonic lysis and ultracentrifugation. The extracted membranes were then co-extruded with ALNPs to fabricate M1 membrane-coated ALNPs (M1-ALNPs). Dynamic light scattering (DLS) analyses revealed that the hydrodynamic diameters of ALNPs and M1-ALNPs were about 100 nm and 120 nm, respectively. The size increment was consistent with the thickness of macrophage membrane (∼10 nm), suggesting successful coating. Transmission electron microscopy (TEM) images confirmed the presence of a continuous membrane layer on the surface of M1-ALNPs (Fig. 3a and b). Next, as shown in Fig. 3c and d, the M1-ALNPs exhibited absorption and emission spectra similar to those of the THF solution, with a slight bathochromic shift observed in the absorption spectrum and a corresponding redshift in the photoluminescence (PL) spectrum.

Fig. 3.

Characterization of nanoformulations and macrophage polarization mechanism. DLS results and (inset) TEM images of (a) ALNPs and (b) M1-ALNPs. Scale bars: 100 nm. c) Absorption and d) PL spectra of TPA-Se and M1-ALNPs. e) CLSM images of 4T1 cells treated with LND-1(NBD)-PEG-24 after staining with Mito-tracker™. Scale bar: 10 μm. f) Schematic diagram of mitochondrial co-localization. g) Immunoblotting of specific proteins related to the MAPK signaling pathway in different formulations. h) Biological TEM images of 4T1 cells treated with M1-ALNPs for different time. i) KEGG enrichment analysis of pathways involved in M1-ALNPs-induced macrophage polarization. Quantification of (j) IL-10, (k) TGF-β1, (l) IL-12 p40, and (m) TNF-α by ELISA. Data are presented as the mean ± SD (n = 3). Statistical significance was determined using a two-tailed Student’s t-test. **p < 0.01, ***p < 0.001, ****p < 0.0001. n) Stability of ALNPs and M1-ALNPs in PBS or PBS supplemented with 10 % of FBS. o) The release of LND-1 from LND-1-PEG-24 at pH 7.4 or pH 5.5, with or without the treatment of CTSB. Data are presented as the mean ± SD (n = 3).

To elucidate the spatiotemporal intracellular transport behavior of LND-1-PEG-24, we conjugated an environment-sensitive fluorescent dye, 7-nitrobenz-2-oxa-1,3-diazole (NBD), to the self-assembled peptide domain, generating LND-1(NBD)-PEG-24. This modification enabled real-time tracking of the prodrug, as NBD exhibits enhanced green fluorescence upon self-assembly into nanostructures, providing a direct readout of its intracellular localization [[53], [54], [55]]. After incubating 4T1 cells with LND-1(NBD)-PEG-24 for 6 h, confocal laser scanning microscopy (CLSM) observations revealed that the NBD fluorescence signal gradually dissociated from the nanocarriers and translocated to mitochondria, identifiable by their characteristic filamentous morphology. This spatial redistribution indicated that the majority of LND-1(NBD)-PEG-24 accumulated within mitochondria post-internalization. Quantitative analysis demonstrated a Pearson’s correlation coefficient (PCC) of 0.796 between NBD fluorescence and MitoTracker Red signals (Fig. 3e and f), confirming the efficient mitochondrial accumulation of LND-1-PEG-24 after cellular internalization. Furthermore, the confocal images clearly demonstrate that the control group (without mitochondria-targeted drugs) is only detected on the cell surface, while the mitochondria-targeted prodrugs are extensively taken up by the cells (Fig. S17).

We further utilized Western blotting to investigate the role of TLR-2/MAPK signaling pathway in this process. The activation of MAPK pathway was confirmed by assessing key members of the MAPK superfamily, including extracellular signal-regulated kinase (ERK), c-Jun N-terminal kinase (JNK), and p38 MAP kinase (Fig. 3g), indicating that TLR-2 played a crucial role in initiating CSM1-mediated macrophage polarization. Following the validation of macrophage polarization mechanism, we employed TEM to investigate the morphology changes in 4T1 cells after M1-ALNPs treatment. In the untreated cells, mitochondria exhibited intact outer membranes and well-organized cristae (indicated by white arrows). In contrast, at 4 h post-treatment, TEM images revealed dense nanofibers at the mitochondrial periphery (red arrows) alongside clear signs of mitochondrial damage: outer membrane disruption, cristae disorganization, mitochondrial swelling, and abnormal vacuolization (yellow arrows). By 6 h post-treatment, mitochondrial damage was further exacerbated, with widespread outer membrane rupture, cristae degradation, and extensive vacuolization (Fig. 3h). These nanofibers can be attributed to the structural transformation of M1-ALNPs upon targeting mitochondria, and their formation is likely dependent on CTSB-mediated cleavage of the prodrug linker—consistent with our earlier observation of TME-responsive prodrug activation. To elucidate the molecular pathway underlying CSM-mediated M1 polarization, we conducted KEGG enrichment analysis on differentially expressed genes (Fig. S18). The results indicated that TLR-2 was significantly upregulated after M1-ALNPs treatment, while the expression of other TLRs (TLR-4, TLR-7, TLR-8, and TLR-9) remained unchanged (Fig. 3i). Next, we evaluated the effect of cryoshock-treated M1 macrophage cell membranes on the expression levels of membrane surface proteins before and after freeze-thaw cycles. The experimental results demonstrated that the cryoshock-treated cell membranes exhibited superior stability, maintaining consistent protein expression levels even after ten freeze-thaw cycles (Figs. S19 and S20). Finally, we evaluated the impact of M1 polarization on the inflammatory immune response, focusing on cytokine secretion. Compared with untreated macrophages, CSM1-treated macrophages exhibited markedly reduced levels of immunosuppressive cytokines TGF-β1 and IL-10 (Fig. 3j and k), along with elevated secretion of pro-inflammatory cytokines IL-12p40 and TNF-α (Fig. 3l and m). These findings underscore the regulatory role of M1 polarization in immune responses. Next, we assessed the colloidal stability of M1-ALNPs. As shown in Fig. 3n, the hydrodynamic diameter of M1-ALNPs remained nearly unchanged over a 5-day period, indicating excellent colloidal stability. Finally, the CTSB-responsive drug release behavior of M1-ALNPs was evaluated under simulated physiological (pH 7.4, without CTSB) and TME (1 U/mL of CTSB) conditions. As depicted in Fig. 3o, in the presence of CTSB in an acidic condition (pH 5.5), M1-ALNPs exhibited significantly accelerated release of LND-1, with ∼70 % of LDN released within 24 h. In contrast, minimal drug release occurred under physiological conditions (pH 7.4, without CTSB). As shown in Fig. S21, Western blot analysis confirmed that nanoparticles coated with membranes derived from non-polarized macrophages (M0) did not express toll-like receptor 2 (TLR-2), whereas those coated with membranes from cold shock–polarized M1 macrophages exhibited significant TLR-2 expression. These results confirmed the successful fabrication of M1-ALNPs with favorable physicochemical properties, excellent stability, and TME-responsive drug release behavior, laying the foundation for subsequent in vitro and in vivo therapeutic evaluations.

2.3. In vitro cellular investigation

To investigate the tumor-targeting capability of M1-ALNPs, we evaluated their cellular uptake in 4T1 breast cancer cells and MCF-10a normal mammary epithelial cells, using uncoated ALNPs as the control. As depicted in Fig. 4a and b, CLSM images revealed that the fluorescence signal of M1-ALNPs in 4T1 cells was much stronger than that of pristine ALNPs. The enhanced uptake can be attributed to the M1 macrophage membrane coating, which presents α4β1 integrin that specifically binds to vascular cell adhesion molecule-1 (VCAM-1), a receptor highlt expressed on 4T1 tumor cells but minimally present on normal epithelial cells. Flow cytometry analysis further confirmed that cellular internalization of M1-ALNPs after 6 h was approximately 2.3-fold higher than that of ALNPs-treated cells (Fig. 4c). In contrast, MCF-10a cells exhibited significantly weaker fluorescence following M1-ALNPs treatment compared to ALNPs (Fig. 4d and e), demonstrating selective tumor cell targeting while sparing healthy cells.

Fig. 4.

In vitro cellular investigation. Representative CLSM images of 4T1 cells incubated with (a) ALNPs or (b) M1-ALNPs for different time. Scale bars: 20 μm. c) Quantitative data of fluorescence intensity in different cells. Data are presented as mean ± SD (n = 3). Statistical significance was determined using a two-tailed Student’s t-test. d) Flow cytometry analysis of 4T1 cells uptake after treatment with ALNPs or M1-ALNPs. e) Representative CLSM images of MCF-10a cells incubated with M1-ALNPs for different time. Scale bar: 20 μm. f) Hemolysis rate of red blood cells treated with water and different concentrations of M1-ALNPs. Data are presented as mean ± SD (n = 3). g) Cell viability of 4T1 cells treated with various NPs. Data are presented as mean ± SD (n = 3). h) 4T1 cells co-stained with Calcein-AM and propidium iodide (PI) after different treatments. Scale bar: 100 μm. i) Colony formation assay and j) bright-field images showing the morphology of 4T1 cells after various treatments. Scale bar: 20 μm. k) Representative bright-field images showing the morphology of 4T1 cells treated with M1-ALNPs and 808 nm light irradiation (1.0 W/cm², 2 min) after different time. Scale bar: 20 μm ***p < 0.001, ****p < 0.0001.

We conducted a hemolysis test to assess the compatibility of M1-ALNPs across a concentration range of 0–200 μg/mL (based on TPA-Se). As depicted in Fig. 4f, the hemolysis rate of M1-ALNPs remained below 0.5 % even at the highest concentration, indicating excellent blood compatibility. Next, we examined the cytotoxicity of different formulations against 4T1 cells using the cell counting kit-8 (CCK-8) assay, with or without 808 nm laser irradiation (1.0 W/cm², 2 min). As shown in Fig. 4g, LNPs alone exhibited minimal cytotoxicity, reflecting the limited efficacy of LND monotherapy. In comparison, ALNPs under laser irradiation displayed higher cytotoxicity than LNPs, primarily driven by the photothermal effect of TPA-Se. Remarkably, M1-ALNPs, with M1 macrophage membrane, exhibited the strongest tumor cell inhibition under laser irradiation, which could be attributed to three synergistic mechanisms: enhanced tumor targeting and uptake by M1 membranes, photothermal ablation by TPA-Se, and mitochondria-targeted pyroptosis induced by LND. To visually confirm the cytotoxic effect of M1-ALNPs, we performed live-dead staining on 4T1 cells with different treatments. As shown in Fig. 4h, ANPs under laser irradiation induced moderate cell death, while ALNPs + NIR caused a slight increase in dead cells. In the M1-ALNPs + NIR group, nearly all 4T1 cells exhibited signs of cell death, with Image-J quantification confirming an 83 % cell death rate, consistent with the CCK-8 results. We further assessed the long-term anti-proliferative activity using a colony formation assay. As shown in Fig. 4i, M1-ALNPs + NIR significantly suppressed colony formation, whereas LNPs had minimal impact on cell proliferation.

Mechanistically, LND is known to upregulate gasdermin E (GSDME), a pyroptosis executor, while PTT activates caspase-3, which cleaves GSDME to release the pore-forming GSDME-N domain, thereby inducing plasma membrane perforation and pyroptosis. In agreement, distinct pyroptotic features, including cell swelling and membrane bubble formation (white arrows), were observed exclusively in the M1-ALNPs + NIR group (Fig. 4j). By contrast, the LNPs group showed limited pyroptosis, while the ANPs + NIR group primarily displayed ferroptotic morphology, including cell shrinkage and formation of apoptotic-like bodies, indicating that neither LND monotherapy nor PTT alone was sufficient to trigger robust pyroptosis. To delineate the temporal progression of pyroptosis, we monitored 4T1 cells treated with M1-ALNPs + NIR at different time points. As shown in Fig. 4k, most cells exhibited early features of pyroptosis at 4 h post-irradiation, while extensive cell rupture, characteristic of late-stage pyroptosis, was observed at 8 h. Collectively, these results confirm that M1-ALNPs effectively integrate mitochondria-targeted prodrug activation with PTT to induce robust pyroptosis, thereby amplifying tumor immunogenicity and laying the foundation for systemic anti-tumor immune activation.

2.4. In vitro cellular immune response

To evaluate the potential of M1-ALNPs in enhancing anti-tumor immune responses via pyroptosis induction, we investigated the release of immunogenic cell death (ICD)-associated danger signals. During ICD, dying tumor cells release damage-associated molecular patterns (DAMPs), such as high mobility group box 1 (HMGB1) and surface-exposed calreticulin (ecto-CRT), which are critical for promoting phagocytosis by antigen-presenting cells (APCs) and initiating tumor-specific immune responses [56]. We first evaluated HMGB1 translocation using immunofluorescence staining. HMGB1 was predominantly confined to the nucleus in untreated 4T1 cells. In contrast, cells treated with M1-ALNPs under NIR irradiation exhibited marked translocation of HMGB1 to the cytoplasm and subsequent release into the extracellular space, evidenced by diffused cytoplasmic fluorescence and decreased nuclear signal (Fig. 5a). Concurrently, surface exposure of ecto-CRT was assessed to further characterize ICD. As shown in Fig. 5b, M1-ALNPs + NIR treatment significantly elevated the proportion of CRT-positive 4T1 compared with other groups. Quantitative analyses revealed that the HMGB1 release and CRT exposure in the M1-ALNPs + NIR group were 5.1 times and 5.8 times higher than those observed in the PBS control group, respectively (Fig. 5c and d), highlighting the robust immunogenic potential of the treatment. Given that the released LND preferentially targets mitochondria, we next examined mitochondrial dysfunction, a critical upstream trigger of pyroptosis, using JC-1, a fluorescent probe sensitive to mitochondrial membrane potential. As shown in Fig. 5e, untreated 4T1 cells exhibited strong orange fluorescence, indicative of intact mitochondrial membrane potential. By contrast, cells treated with ALNPs + NIR or M1-ALNPs + NIR displayed a substantial increase in green fluorescence (JC-1 monomers) alongside decreased orange fluorescence (JC-1 aggregates), indicating mitochondrial depolarization. Quantitative analysis of the monomer-to-aggregate fluorescence ratio confirmed significant mitochondrial depolarization in both treatment groups, with the M1-ALNPs + NIR group showing the most pronounced effect (Fig. 5f). Collectively, these results demonstrated that M1-ALNPs under NIR irradiation effectively induced ICD through mitochondrial dysfunction and pyroptosis, thereby releasing potent immunostimulatory signals that were likely to enhance subsequent anti-tumor immune responses.

Fig. 5.

In vitro cellular immune response. a) Representative CLSM images showing the expression of HMGB1 (red fluorescence) in 4T1 cells after different treatments. The cell nuclei were stained with DAPI (blue fluorescence). b) Representative CLSM images showing the expression of ecto-CRT (green fluorescence) in 4T1 cells after different treatments. The cell nuclei were stained with DAPI (blue fluorescence). Scale bar: 20 μm. Quantitative data for mean fluorescence intensity (MFI) of c) HMGB1 and d) ecto-CRT, presented as mean ± SD (n = 3). e) The detection of mitochondrial potential changes in 4T1 cells after different treatments using JC-1 staining. The red fluorescence indicates JC-1 aggregate, while the green fluorescence indicates JC-1 monomer. Scale bar: 20 μm. f) Quantitative data representing the ratio of JC-1 monomer to aggregate fluorescence for 4T1 cells treated with various formulations (n = 3). g) LDH release assay of 4T1 cells receiving different treatments (n = 3). h) ATP concentrations released from 4T1 cells after various treatments (n = 3). i) Immunoblotting for the expression of pyroptosis-related proteins (Caspase-3, Cleaved Caspase-3, GSDME, and GSDME-N) in 4T1 cells after different treatments. j) Schematic illustration of in vitro assessment of BMDCs maturation. k) Representative flow cytometry results and l) quantitative analyses of mature DCs (CD 80⁺/86⁺) after different treatments (n = 3). Statistical significance was determined using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

LND functions as a glycolysis inhibitor by suppressing lactate efflux via monocarboxylate transporter 4 (MCT4), thereby potentially preventing M2 polarization of tumor-associated macrophages (TAMs) (Figs. S22 and S23). Building on this mechanism, we investigated whether M1-ALNPs could synergistically promote M1-mediated TAM polarization. Lactate dehydrogenase (LDH) release, a critical indicator of membrane integrity during pyroptosis, was markedly elevated in the M1-ALNPs + NIR group, exhibiting approximately 3.0-fold and 1.8-fold increases relative to the PBS and ANPs groups, respectively (Fig. 5g). Concurrently, intracellular ATP levels were significantly decreased in the M1-ALNPs + NIR group (Fig. 5h), reflecting extracellular ATP release as a DAMP signal during pyroptotic cell death. Western blot analysis further demonstrated that M1-ALNPs + NIR treatment significantly upregulated GSDME-N, the membrane-permeabilizing active domain of gasdermin E (GSDME) (Fig. 5i). Compared with LNPs alone, GSDME-N expression was substantially higher in the M1-ALNPs + NIR group, underscoring the pivotal role of PTT in amplifying LND-induced pyroptosis. Precisely controlled mild hyperthermia induces caspase-3 activation via the mitochondrial apoptotic pathway. This temperature-dependent process is essential for GSDME cleavage and pyroptosis. Mechanistically, PTT-induced cellular stress activated caspase-3, which subsequently cleaved GSDME into its active GSDME-N form, establishing a synergistic mechanism wherein LND-1-PEG-24-mediated GSDME upregulation combined with PTT-induced cleavage to efficiently trigger robust pyroptosis under NIR irradiation. Activated GSDME forms pores on mitochondrial membranes, disrupting function and causing loss of membrane potential, reduced oxidative phosphorylation, and release of mitochondrial contents. This damage generates additional mtROS. As Fig. S24 shown, M1-ALNPs induced substantial mtROS production which promoted GSDME activation. The resulting GSDME pores further damaged mitochondria, leading to sustained mtROS bursts and exacerbated pyroptosis.

To assess downstream immune activation, we evaluated the maturation of bone marrow-derived dendritic cells (BMDCs) using a Transwell co-culture system. 4T1 tumor cells pretreated with various nanoformulations were seeded in the upper chamber, while BMDCs were cultured in the lower chamber (Fig. 5j). Flow cytometry analysis revealed that M1-ALNPs + NIR-treated tumor cells markedly enhanced BMDC maturation relative to all other groups (Fig. 5k). Quantitative analysis showed that the proportion of mature BMDCs (CD11c⁺CD80⁺CD86⁺) in the M1-ALNPs + NIR group was 4.1-, 3.4-, 2.5-, 1.3-, and 1.1-fold higher than those in the PBS, LNPs, M1-ALNPs, ANPs + NIR, and ALNPs + NIR groups, respectively (Fig. 5l). These results provide direct evidence that pyroptosis induced by M1-ALNPs under NIR irradiation significantly potentiates dendritic cell maturation, which is essential for efficient antigen presentation and subsequent T cell-mediated anti-tumor immunity.

2.5. In vivo NIR-II fluorescence and PA imaging of tumor

To evaluate the in vivo tumor-targeting efficiency and imaging performance of M1-ALNPs, we conducted in vivo NIR-II fluorescence and PA imaging. Tumor-bearing mice were randomly divided into two groups (n = 3 mice per group) and intravenously administered M1-ALNPs or uncoated ALNPs. Temporal monitoring of NIR-II fluorescence at the tumor site was performed to determine the optimal imaging and treatment window. As illustrated in Fig. 6a, the NIR-II fluorescence signal in the M1-ALNPs group gradually increased over time, reaching a peak at approximately 24 h post-injection. This time point was therefore selected as the optimal window for both imaging and therapeutic interventions. Notably, the NIR-II fluorescence intensity of tumor site in M1-ALNP-treated mice was more than two-fold higher than that in the ALNPs group (Fig. 6b), indicating that the M1 macrophage membrane coating markedly enhanced tumor accumulation. Furthermore, strong NIR-II signals persisted at the tumor site up to 48 h post-injection, demonstrating prolonged in vivo retention of M1-ALNPs. At 24 h post-administration, mice were euthanized, and major organs (heart, liver, spleen, lungs, kidneys) along with tumor tissues were harvested for ex vivo NIR-II imaging. As shown in Fig. 6c and d, tumors from M1-ALNP-treated mice exhibited significantly higher fluorescence signals compared to those from the ALNPs group, providing direct evidence of enhanced tumor-targeting capability. Interestingly, the accumulation of M1-ALNPs in the liver and spleen was reduced relative to ALNPs, likely attributable to the inherent properties of M1 macrophage membranes, which facilitate nanoparticle evasion from normal tissues. This phenomenon was likely due to the proteins inherited from the macrophage membrane—for instance, CD47. This protein can block unwanted macrophage-driven phagocytosis by interacting with SIRPα, a molecule expressed on the surface of macrophages. This observation further supporting the selective tumor-homing ability of M1-ALNPs.

Fig. 6.

In vivo NIR-II fluorescence and PA imaging of tumor-bearing mice. a) Representative fluorescence images and b) corresponding quantative analyses of 4T1 tumor-bearing mice at different time points after intravenous injection of ALNPs or M1-ALNPs. c) Representative ex vivo fluorescence imaging and d) corresponding fluorescence intensity of major organs and tumors isolated from mice at 24 h after intravenous injection of ALNPs or M1-ALNPs. e) Representative PA images and f) corresponding PA intensity of tumor site in 4T1 tumor-bearing mice after i.v. injection of ALNPs or M1-ALNPs for various time. Data are presented as mean ± SD (n = 3 mice). Statistical significance was determined using Student’s t-test. **p < 0.01, ***p < 0.001, ****p < 0.0001.

To complement the high-sensitivity temporal information provided by NIR-II fluorescence imaging with high-resolution anatomical details, we performed in vivo photoacoustic (PA) imaging in tumor-bearing mice. PA imaging offers superior spatial resolution and deep-tissue penetration compared to fluorescence imaging, enabling precise visualization of tumor morphology. As shown in Fig. 6e, the PA signal at the tumor site exhibited a time-dependent profile closely matching the temporal evolution of NIR-II fluorescence, reaching a maximum approximately 24 h post-intravenous administration. Under 830 nm excitation, PA imaging clearly delineated the tumor location and morphological features with high contrast (Fig. 6f). Collectively, the combination of in vivo NIR-II fluorescence and PA imaging provided complementary temporal and spatial information, enabling precise tumor localization and offering robust guidance for subsequent therapeutic interventions.

2.6. In vivo anti-tumor studies

Encouraged by the robust in vitro tumor-killing efficacy of M1-ALNPs, we next evaluated their therapeutic performance in vivo using a subcutaneous 4T1 breast cancer mouse model, a well-established preclinical system characterized by low immunogenicity, aggressive proliferation, and high metastatic potential. Female BALB/c mice were subcutaneously inoculated with 4T1 cells. Seven days post-inoculation, when tumors reached a volume of ∼100 mm³, mice were randomly assigned to six groups (n = 5 mice per group) and received different treatments: PBS, LNPs, M1-ALNPs, ANPs + NIR, ALNPs + NIR, and M1-ALNPs + NIR. At 24 h post intravenous injection of various nanoformulations (200 μL, 500 μg/mL), tumors in the groups with “NIR” were subjected to 808 nm (1.0 W/cm²) laser irradiation for 7 min. The treatment regimen was repeated on days 0, 3 and 6, while tumor growth, volume, and body weight were monitored to evaluate therapeutic efficacy (Fig. 7a). First, we validated the in vivo photothermal performance of M1-ALNPs using real-time IR thermal imaging. As shown in Fig. 7b and c, tumors in the M1-ALNPs + NIR group displayed a rapid and sustained temperature increase, peaking at ∼54 °C, whereas tumors in the ALNPs + NIR group exhibited only moderate heating (∼42 °C). These results underscored the superior photothermal capability of M1-ALNPs, highlighting their potential for spatiotemporally controlled tumor ablation.

Fig. 7.

In vivo anti-tumor studies. a) Schematic illustration showing the treatment procedure and assessment of therapeutic efficacy of M1-ALNPs in 4T1 tumor-bearing mice. The “i.v.” represents “intravenous injection”. b) IR thermal images and c) temperature changes at tumor sites of 4T1 tumors-bearing mice after intravenous injection with M1-ALNPs under 808 nm laser irradiaiton (1.0 W/cm²) for different time. Data are presented as mean ± SD (n = 3 mice). d) Individual tumor growth curves and e) average tumor volumes after different treatments. Data are presented as mean ± SD (n = 5 mice). Statistical significance was determined using one-way ANOVA. f) Body weight changes and g) tumor weights of 4T1 tumor-bearing mice subjected to various treatments. Data are presented as mean ± SD (n = 5 mice). h) Images of tumors excised on day 14 from mice receiving different treatments. i) Representative H&E staining and TUNEL staining of tumor sections harvested from mice receiving different treatments on day 14. Scale bars: 50 μm. Statistical significance was determined using one-way ANOVA. **p < 0.01, ***p < 0.001, ****p < 0.0001.

Tumor growth kinetics revealed distinct therapeutic outcomes across groups (Fig. 7d and e). LNPs and M1-ALNPs without NIR produced only modest inhibition, reflecting the limited chemotherapeutic effect of low-dose LND monotherapy. In stark contrast, the M1-ALNPs + NIR group achieved the most pronounced tumor suppression. By day 14, mean tumor volumes in this group were 16.26-, 9.32-, 7.22-, and 4.05-fold smaller than those in the PBS, LNPs, ALNPs, and M1-ALNPs groups, respectively. This enhanced efficacy is attributed to the synergistic integration of M1 membrane-mediated tumor targeting, TPA-Se-induced photothermal ablation, and LND-driven mitochondria-targeted pyroptosis. Notably, all groups exhibited minimal body weight fluctuations throughout the study, indicating favorable systemic biocompatibility of M1-ALNPs (Fig. 7f). Final tumor weights further corroborated the superior therapeutic performance of M1-ALNPs + NIR, with significantly reduced tumor burden relative to all other groups (Fig. 7g). Ex vivo tumor photographs visually confirmed these results, revealing substantially smaller tumors in the M1-ALNPs + NIR group (Fig. 7h). Histological analyses provided mechanistic insights into tumor cell destruction. H&E staining demonstrated dense, intact tissue architecture in the PBS and LNPs groups, whereas tumors treated with M1-ALNPs + NIR exhibited extensive necrosis, disrupted architecture, and markedly reduced cell density (Fig. 7i). Complementary TUNEL staining revealed a pronounced increase in apoptotic cells in the M1-ALNPs + NIR group, confirming robust induction of tumor cell death in vivo, likely through the combined actions of photothermal ablation and pyroptosis (Fig. 7i).

2.7. In vivo anti-metastasis effect

To elucidate the immunological mechanisms underlying the potent antitumor efficacy of M1-ALNPs + NIR, we conducted comprehensive immunological analyses. A key early step in initiating adaptive immunity is the maturation of DCs within tumor-draining lymph nodes (TDLNs), as mature DCs serve as the primary APCs for priming tumor-specific T cell responses. Flow cytometry analysis revealed that M1-ALNPs + NIR treatment markedly enhanced DC maturation, with the proportion of mature DCs (CD11c⁺CD80⁺CD86⁺) in TDLNs reaching 33.8 %, significantly higher than in all other groups (Fig. 8a and b). This observation suggests that M1-ALNPs + NIR effectively promoted DC activation, likely facilitated by the release of DAMPs such as HMGB1 and ATP from pyroptotic tumor cells. These signals drive upregulation of co-stimulatory molecules (CD80/CD86), enabling DCs to efficiently present tumor antigens to naïve T cells, thereby initiating robust adaptive immune responses. We next assessed T cell infiltration within the TME, as T cells represent the principal effectors of adaptive anti-tumor immunity. Flow cytometry demonstrated that M1-ALNPs + NIR treatment significantly increased the frequency of tumor-infiltrating CD8⁺ cytotoxic T lymphocytes (CTLs, CD3⁺CD4^‒CD8⁺), approximately 2.48-fold higher than that in the PBS group (Fig. 8c and d). Simultaneously, the population of CD4⁺ helper T cells (CD3⁺CD4⁺CD8^‒), which enhance CTL effector function by secreting pro-inflammatory cytokines such as IFN-γ, was also markedly elevated in the M1-ALNPs + NIR group. Importantly, this was accompanied by a substantial reduction in immunosuppressive regulatory T cells (Tregs, CD3⁺CD4⁺Foxp3⁺) compared with other treatment groups (Fig. 8e and f). Given that Tregs suppress CTL activity and reinforce immune tolerance within the TME, their depletion effectively relieved immunosuppression and further amplified tumor-specific T cell responses [57,58].

Fig. 8.

In vivo anti-metastasis effect. a) Representative flow cytometric analysis and b) quantitative analyses of mature DCs (CD11c⁺CD80⁺CD86⁺) in TDLNs of 4T1 tumor-bearing mice after different treatments (n = 3 mice). c) Representative flow cytometric analysis and d) quantitative analyses of tumor infiltrating CD8⁺ in CD3⁺ T cells in mice after different treatments (n = 3 mice). e) Representative flow cytometric analysis and f) quantitative analyses of tumor-infiltrating CD4⁺Foxp3⁺ Tregs in different groups (n = 3 mice). g) Representative flow cytometric analysis and h) quantitative analyses of tumor-infiltrating M1-likemacrophages (CD11b⁺ F4/80⁺ CD86⁺) in tumor site after different treatments (n = 3 mice). i) Representative flow cytometric analysis and j) quantitative analyses of tumor-infiltrating M2-like macrophages (CD206⁺CD11b⁺F4/80⁺) in 4T1 tumor-bearing mice after different treatments (n = 3 mice). k) The schematic diagram of anti-metastasis assessment after surgical resection. l) Representative photograph and H&E staining image of lungs collected at the end of experiment. m) Number of metastatic nodules on the lungs. Data are presented as mean ± SD (n = 5 mice). Statistical significance was determined using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001. n) Representative flow cytometric analyses of CD3⁺CD8⁺CD62L⁻CD44⁺ TEM in spleen collected from 4T1 tumor-bearing mice in different groups (n = 3 mice). Source data are provided as a Source Data file.

Given the pivotal role of TAMs in remodeling the TME, with M1-like TAMs exerting anti-tumor functions and M2-like TAMs promoting tumor progression and therapeutic resistance, we next evaluated TAM polarization in vivo via flow cytometry [59,60]. As shown in Fig. 8g–j, treatment with M1-ALNPs + NIR markedly decreased the proportion of M2-like TAMs (CD11b⁺F4/80⁺CD206⁺, indicative of immunosuppressive TAMs), while concomitantly increasing the fraction of M1-like TAMs (CD11b⁺F4/80⁺CD86⁺, indicative of pro-inflammatory TAMs). Notably, this treatment achieved the highest M1/M2 TAM ratio among all experimental groups, confirming the effective reprogramming of pro-tumoral M2-like TAMs toward the anti-tumoral M1 phenotype. Mechanistically, this polarization shift is likely driven by the synergistic effects of M1 macrophage membrane-derived signals (e.g., TLR2 ligands) activating the TLR2/MAPK pathway and LND-mediated glycolysis inhibition reducing lactate-driven M2 polarization, collectively transforming the TME from immunosuppressive to immunostimulatory [61,62].

Activation of systemic anti-tumor immunity offers the additional advantage of inhibiting metastasis, a major contributor to cancer-related mortality. To assess the anti-metastatic potential of M1-ALNPs + NIR, we employed a 4T1 lung metastasis model, established by intravenous injection of 4T1 cells into tumor-bearing mice on day 14 post-primary tumor inoculation (Fig. 8k). Two weeks after 4T1 cell injection, H&E staining revealed extensive metastatic nodules and focal necrosis in the PBS-treated group (Fig. 8l). The LNPs and M1-ALNPs (without NIR) groups exhibited similarly abundant metastatic lesions, while ANPs + NIR and ALNPs + NIR groups showed moderate improvement but retained sizable metastatic foci. In contrast, M1-ALNPs + NIR treatment nearly abolished observable lung metastases (Fig. 8m), likely resulting from activation of systemic, tumor-specific CTLs capable of eliminating circulating tumor cells and micrometastases. Beyond the immediate anti-tumor immune response, the establishment of immune memory plays a crucial role in preventing tumor recurrence. The results showed that the numbers of the memory T cell subsets in the spleen and blood of the treatment group peaked after treatment, which were significantly higher than those in the control group, and remained at a relatively high level. (Fig. 8n). These results underscore that M1-ALNPs + NIR not only mediates local tumor ablation but also elicits durable systemic anti-tumor immunity. However, this study is primarily based on subcutaneous xenograft models, which may not fully replicate the complex tumor microenvironment and metastatic processes observed in clinical settings. To address this limitation, orthotopic tumor models will be employed in future research to more accurately evaluate therapeutic efficacy and organ-specific targeting. This approach is expected to enhance the clinical relevance of the findings.

Finally, we evaluated the in vivo safety profile of M1-ALNPs. Healthy mice were randomly assigned to receive intravenous injections of PBS or M1-ALNPs (n = 3 per group). Ten days after the last administration, major organs (heart, liver, spleen, lungs, and kidneys) were harvested for H&E staining. As shown in Fig. S25, no evident organ damage or inflammatory lesions were observed in either group. Additionally, routine blood analyses and liver/kidney function tests performed on day 10 (Figs. S26 and S27) showed no significant differences between PBS- and M1-ALNPs-treated mice. Collectively, these findings confirmed that M1-ALNPs possess excellent in vivo biocompatibility and safety.

3. Conclusion

In summary, we engineered a multifunctional nanoplatform that integrates NIR-II molecular imaging, mitochondria-targeted pyroptosis, and TAM reprogramming for precise cancer theranostics. We first developed a new high-performance AIEgen that enabled dual NIR-II fluorescence and PA imaging for real-time tumor delineation, while also exhibiting superior photothermal conversion for effective PTT application. The AIEgen was co-assembled with a CTSB-responsive mitochondria-targeted prodrug (LND-1-PEG-24) and coated with CSMs to form M1-ALNPs with potent synergistic functionality. Within TME, CSMs promoted durable TAMs polarization toward the immunostimulatory M1 phenotype via the TLR2/MAPK pathway, mitigating immunosuppression and enhancing tumor-targeted drug delivery. Subsequent mitochondrial accumulation of LND triggered caspase-3/GSDME-mediated pyroptosis and DAMP release, which—together with photothermal ablation—augmented tumor immunogenicity, facilitated dendritic cell maturation, and promoted robust CD8⁺ T cell infiltration while suppressing Tregs. In vivo, M1-ALNPs not only eradicated primary tumors but also inhibited metastatic dissemination and induced durable immune memory, effectively suppressing 4T1 breast cancer lung metastasis.

The development of immunotherapies for cancer treatment has marked a turning point in quest to fight cancer with immune checkpoint inhibitors being approved for more than 25 different types of cancer. Combining immunotherapy with precision medicine and advanced biomaterials or nanoparticles to combine multiple therapeutic strategies into one treatment, may overcome the limitations of traditional immunotherapies using a single approach. Our approach of developing this multifunctional nanoplatform harmonizes NIR-II phototheranostics, mitochondria-targeted pyroptosis, and TAM reprogramming into a single system. It provides a paradigm for reversing tumor immune suppression, activating systemic anti-tumor immunity, and preventing metastasis—offering a promising strategy for precision oncology and advancing the clinical translation of combination immunotherapies. This study highlights a paradigm in precision oncology by harmonizing NIR-II phototheranostics, mitochondria-targeted pyroptosis, and TAM reprogramming within a single nanoplatform. By reversing TME immunosuppression and activating systemic antitumor immunity, this approach provides a robust strategy for long-term tumor control and metastasis prevention, thereby advancing the translational potential of combination immunotherapies.

CRediT authorship contribution statement

Di Zhang: Formal analysis, Funding acquisition, Investigation, Methodology, Writing – original draft. Xu He: Data curation, Methodology, Software, Visualization. Kannappan Vinodh: Data curation, Software. Zhehan Yao: Data curation, Software, Visualization. Wanyu Wei: Data curation, Validation. Jingxiang Liang: Methodology, Validation. Ningbo Li: Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing. Zhifang Wu: Conceptualization, Supervision, Writing – review & editing. Sijin Li: Conceptualization, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the supplementary data to this article:

Data availability

No data was used for the research described in the article.