A Photothermally Triggered Cascade Nanodelivery Platform for On‐Demand Nitric Oxide Release in Targeted Hepatocellular Carcinoma Therapy

ABSTRACT

Traditional nanoparticle delivery systems for hepatocellular carcinoma (HCC) rely on passive targeting and low drug release rates, limiting efficacy, while the biocompatible nitric oxide (NO) donor L‐Arginine (L‐Arg) is prone to systemic leakage, hindering precise HCC delivery. In the content, we constructed a cascade‐responsive nanoplatform based on porous hollow gold nanocages (GNCs) that integrates spatial targeting and temporal activation of multimodal treatment. By co‐loading the near‐infrared (NIR) photosensitizer indocyanine green (ICG) and the NO precursor L‐Arg into GNCs, and conjugating the HCC‐targeting peptide 9R‐P201 on the surface, we obtained the multifunctional nanoplatform GIL9R. Upon NIR irradiation, GIL9R nanoplatform simultaneously generated localized hyperthermia (photothermal therapy, PTT) and the resulting strong oxidizing agent. The resulting reactive oxygen species (ROS) catalyze L‐Arg oxidation via an enzyme‐mimetic pathway to produce NO, triggering gas therapy (GT) in situ. This PTT/PDT‐initiated NO release synchronizes three modalities, yielding potent synergistic cytotoxicity against HCC cells. In vitro, GIL9R nanoplatform showed high HCC affinity, effective photodynamic therapy (PDT), and robust NO‐mediated GT. In vivo, GIL9R nanoplatform selectively accumulated in tumors and significantly suppressed tumor growth. This rationally designed cascade platform leverages intrinsic synergy among PTT, PDT, and GT, offering a translatable strategy for precision liver cancer treatment.

We designed a gold nanocage (GNCs)‐based nanoplatform (GIL9R) integrating indocyanine green (ICG) and L‐Arginine (L‐Arg), surface‐modified with HCC‐targeting peptide 9R‐P201, enabling concurrent photothermal therapy (PTT), photodynamic therapy (PDT), and gas therapy (GT). GIL9R targeted HepG2 via 9R‐P201, generating ROS/NO under NIR to kill HCC cells while sparing normal hepatocytes. In H22 tumor‐bearing mice, it accumulated in tumors, inhibiting growth via PDT‐GT synergy and enhancing apoptosis. Excellent biosafety supported intravenous injection hepatic cancer therapy.

1. Introduction

Hepatocellular carcinoma (HCC) remains a malignant neoplasm with persistently high global incidence and mortality rates, posing substantial therapeutic challenges [1, 2, 3, 4]. These difficulties arise from the marked heterogeneity of the tumor microenvironment (TME), a pronounced propensity for metastasis and recurrence, and the frequent development of resistance to conventional chemotherapy and radiotherapy [5, 6].

With recent technological advances, gas therapy (GT) has emerged as a promising modality in oncology, offering distinctive biological effects, favorable biocompatibility, and minimally invasive characteristics [7, 8]. Nitric oxide (NO), an endogenous gaseous signaling molecule, regulates vascular tone and immune responses at low concentrations [9, 10, 11, 12], whereas at elevated concentrations (>1 µM), it effectively suppresses tumor growth via mechanisms inducing cell cycle arrest, mitochondrial dysfunction, and formation of the highly toxic peroxynitrite (ONOO⁻) [13, 14, 15, 16, 17]. The limitations of NO in clinical applications are clear: its extremely quick breakdown, its rapid diffusion throughout the body, and its inability to accumulate at target sites [18, 19, 20].

L‐Arginine (L‐Arg) is a natural and biocompatible NO precursor. It can be converted into NO through catalytic conversion, making it a viable endogenous donor [21, 22]. However, its high aqueous solubility leads to rapid systemic clearance and insufficient intratumoral retention [23, 24]. Nano‐carrier systems have therefore been extensively explored to improve L‐Arg delivery and enable spatiotemporally controllable NO release, often via co‐loading with ROS‐generating agents such as photosensitizers [25, 26, 27, 28]. For example, Chang et al. constructed a cascade nanozyme that releases β‐lapachone in a degradation‐responsive manner to generate hydrogen peroxide (H₂O₂), which subsequently drives L‐Arg oxidation into NO, while the concurrent photothermal effect amplifies ROS/NO synergy [29]. Similarly, Wang et al. showed that a polymeric vesicle system co‐loaded with PEGylated ICG and L‐Arg, when exposed to NIR light, delivers combined PDT and PTT effects to trigger NO release, resulting in complete tumor eradication in preclinical models [30]. Tang et al. utilized the strong affinity of Au–SH interactions to immobilize NO donors onto gold nanocages (GNCs), facilitating the triggered release of NO [31]. These studies underscore the feasibility of integrating NO precursors with photosensitizers within functional nanoplatforms to unify GT with PTT for multimodal antitumor efficacy [32, 33, 34].

Although nanocarrier‐based GT has witnessed remarkable advancements, existing systems remain plagued by fundamental limitations, notably poor targeting specificity, inadequate biodegradability, and restricted deep‐tissue penetration [35, 36, 37, 38, 39, 40, 41, 42, 43]. Porous hollow GNCs represent an attractive alternative, offering tunable localized surface plasmon resonance, a hollow structure conducive to high drug loading, excellent photothermal conversion efficiency, and versatile surface functionalization [44, 45, 46, 47, 48, 49, 50, 51, 52]. In addition, conjugating GNCs with ligands targeting hepatic cancer receptors, such as the 9R‐P201 peptide, can significantly enhance active tumor homing and prolong intralesional retention [2, 53, 54, 55, 56].

In this study, we designed an HCC‐targeted cascade nanoplatform (GIL9R) based on GNC co‐loaded with ICG and the NO precursor, and decorated with the liver cancer‐targeting peptide 9R‐P201 (Scheme 1). Upon NIR irradiation, GIL9R nanoplatform concurrently generated ROS and induced localized hyperthermia through the photothermal effect of GNCs and the photodynamic activity of ICG, thereby initiating both PTT and PDT. The generated ROS further catalyzed the degradation of L‐Arg to release NO, activating GT. Meanwhile, the 9R‐P201 modification conferred active targeting to HCC tissues, enhancing cellular uptake and prolonging intratumoral retention. Through the integrated action of PTT, PDT, and GT, the GIL9R nanoplatform achieved synergistic tumor ablation with enhanced specificity and reduced systemic toxicity. Extensive in‐cell and animal experiments were conducted to validate targeting efficiency, ROS/NO generation capacity, antitumor efficacy, and biocompatibility. This established a paradigm for precision and multimodal HCC therapy.

SCHEME 1.

Construction of the GIL9R nanoplatform and its cascade‐triggered burst NO release for targeted HCC treatment.

2. Research Methods

2.1. Materials and Instruments

Chloroauric acid hydrate (HAuCl4·4H₂O, 47.8%) was purchased at Shanghai Puka Chemical Technology Co., Ltd. (Shanghai, China). Ethylene glycol (EG, 99.8%), L‐Arg (98%), ICG (75%), and sodium hydrosulfide hydrate (NaHS·xH₂O, 68%) were obtained from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). Concentrated hydrochloric acid (HCl) was sourced from Luoyang Haohua Chemical Reagent Co., Ltd. (Luoyang, China). Silver trifluoroacetate (CF₃COOAg, 99%) and poly(vinylpyrrolidone) (PVP, Mw = 55,000) were obtained from Anhui Zesheng Technology Co., Ltd. (Anhui, China).The 9R‐P201 peptide was procured from QYAOBIO (chinaPeptides Co., Ltd.).  Acetone was procured from Kaifeng Fangjing Chemical Reagent Co., Ltd (Kaifeng, China). The NO assay kit and DAF‐FM diacetate (DAF‐FM DA) were acquired from Shanghai Biyuntian Bio‐Technology Co., Ltd. (Shanghai, China). All reagents were used as received without further purification.

2.2. Cell Line and Animals

All cell lines used in this study were obtained from Wuhan Servicebio Biotechnology Co., Ltd (Wuhan, China). Female BALB/c mice (6‐8 weeks of age) were maintained and handled in compliance with the Animal Ethics Committee of Henan University's approved protocols (No. HUSOM‐2024‐066).

2.3. Synthesis of AgNCs

The procedure was adapted from a reported method [57]. An oil bath was preheated to 150°C, and EG (25 mL) was added and maintained for 30 min. Then, NaSH‐EG solution (3 mM, 0.3 mL) was injected and reacted for 2 min. Subsequently, HCl‐EG (3 mM, 2.5 mL) and PVP‐EG (20 mg/mL, 6.25 mL) were added sequentially, and the reaction continued for another 2 min. Finally, CF₃COOAg‐EG solution (282 mM, 2 mL) was introduced, and the container was stirred for 90 min. Upon completion, the solution turned yellow‐green. The reaction was quenched with ice‐cold water and washed twice with acetone (volume equal to twice the reactant). The solution was then followed by three additional washes with triple‐distilled water, resulting in the collection of silver nanocubes (AgNCs).

2.4. Synthesis of GNCs

The mixture was subjected to a galvanic replacement reaction, a galvanic replacement process between presynthesized AgNC templates and HAuCl₄, following a published protocol [41]. In short, 10 mL of AgNC suspension was added to 90 mL of a PVP solution (1.5 mg/mL) in water and heated in a 100°C oil bath for 10 min. Chloroauric acid solution (1 mmol/L) was then added dropwise via a syringe until Ultraviolet‐visible spectroscopy (UV–vis) indicated that the characteristic plasmon resonance band of GNCs appeared at ∼800 nm, at which point the addition was stopped. After the process was complete, the contents were chilled in an ice water bath and saturated with NaCl. The sample was obtained by centrifuging at 9, 000 rpm for 10 min. The sample was then cleaned 3 times with deionized H₂O and labeled as GNCs.

2.5. Synthesis of GNC@ICG/L‐Arg

The GNC@ICG/L‐Arg nanocomposite (abbreviated as GIL) was synthesized by dissolving ICG (20 mg) and L‐Arg (20 mg) in 25 mL purified water, before adding 10 mg of as‐prepared GNCs. The solution was then pelleted by centrifugation in a centrifuge at 8000 rpm for 10 min. This was done after gently shaking the solution in the dark at room temperature (RT) for 12 h.

2.6. Synthesis of GNC@CG/L‐Arg@9R‐P201

After stirring GIL in the dark at room temperature for 12 h, a homogeneous GIL dispersion (0.4 mg/mL) was obtained. Subsequently, 5 mg of the 9R‐P201 targeting peptide was added to the GIL dispersion. The mixture was stirred at room temperature in the dark for 24 h to ensure sufficient conjugation. The resulting nanoplatform was collected by centrifugation at 8,000 rpm for 10 min and washed 2–3 times with deionized water. The final product was designated as GNC@ICG/L‐Arg@9R‐P201 (GIL9R).

2.7. Characterization

The images were captured using a transmission electron microscope (TEM, JEOL JEM‐F200) with an energy‐dispersive X‐ray (EDS) spectroscope. The hydrodynamic diameter and zeta potential were measured using a Nano ZS Zetasizer (Malvern Instruments, UK). UV–vis absorption peaks were measured using a UV‐3600 Plus instrument (Shimadzu) over the 200–1000 nm range. Fourier‐transform infrared (FTIR) spectra were obtained using a Bruker VERTEX 70 operating in the range of 500–4000 cm⁻¹. Cell survival was measured via a multimode platform (BioTek Synergy H1). Cellular drug uptake was observed by confocal laser scanning (LEICA TCS SP8+STED, Germany). In situ imaging was performed on a small laboratory animal imaging system (IVIS Spectrum, PerkinElmer).

2.8. Determination of Drug Loading Capacity of GIL9R

Standard stock solutions of L‐Arg (1, 5, 10, 15, or 20 µg/mL) were prepared for quantification via the Sakaguchi reaction. Briefly, 3 mL of the sample was mixed with 0.5 mL of freshly prepared 14% NaOH solution, followed by 1.5 mL of a chromogenic reagent (4% w/v, prepared by dissolving 10 mg of 2,3‐butanedione and 2 g of 1‐naphthol in 50 mL of isopropanol). After adding 1 mL of isopropanol, the mixture was vortexed vigorously, incubated in a water bath at 40°C for 10 min, and immediately cooled to room temperature in an ice‐water bath. Absorbance was measured at 515 nm using a UV–vis spectrophotometer to construct the L‐Arg standard curve. Similarly, a calibration curve for ICG was established by recording the absorbance of standard solutions (1, 5, 10, 15, or 20 µg/mL) at 780 nm.

GNC (4 mg) was dispersed in deionized water to form a 1 mg/mL stock solution. To optimize the mass ratio, predetermined amounts of L‐Arg (2, 4, or 8 mg) and ICG (2, 4, or 8 mg) were introduced to the GNC dispersion to achieve theoretical mass ratios of GNC: ICG/L‐Arg at 2:1:1, 1:1:1, and 1:2:2 (corresponding to final ICG/L‐Arg concentrations of 0.5, 1, and 2 mg/mL). The mixtures were sonicated to ensure homogeneity and subsequently stirred gently in the dark at room temperature for 24 h to facilitate drug loading. Following this, 2 mg of 9R‐P201 peptide (final concentration 0.5 mg/mL) was added, and the conjugation reaction proceeded under identical conditions for an additional 24 h.

The resulting GIL9R nanoplatform was purified by centrifugation. The supernatant was carefully collected, diluted appropriately, and analyzed via UV–vis spectroscopy. The residual concentrations of free ICG and L‐Arg were determined by measuring absorbance at 780 and 515 nm, respectively. The drug loading efficiency (DLE) and encapsulation efficiency (EE) were calculated based on the standard curves and the following equations:

$$\mathrm{Drug}\mathrm{loading}\mathrm{efficiency}(\%)=\left(\mathrm{mass}\mathrm{of}\mathrm{drug}\mathrm{input}-\mathrm{mass}\mathrm{of}\mathrm{drug}\mathrm{in}\mathrm{supernatant}\right)/\mathrm{mass}\mathrm{of}\mathrm{carrier}\times 100\%$$ (1)

$$\mathrm{Encapsulation}\mathrm{efficiency}(\%)=(\mathrm{mass}\mathrm{of}\mathrm{drug}\mathrm{input}-\mathrm{mass}\mathrm{of}\mathrm{drug}\mathrm{in}\mathrm{supernatant})/\mathrm{mass}\mathrm{of}\mathrm{drug}\mathrm{input}\times 100\%$$ (2)

2.9. In Vitro NO Release from GIL9R

The standard curve was prepared using NaNO₂ solutions (2, 20, 40, 60, and 80 µM) as the NO source, and nitrite concentration was analyzed using the Griess reaction. To investigate the kinetics of the NO release profile, GIL9R nanoplatform at different concentrations was subjected to either 808 nm light (1.5 W/cm², 10 min) or left in the absence of light. Immediately after treatment, Griess reagent was added to form a diazonium salt, and the absorbance at 540 nm was read using a microplate reader.

2.10. In Vitro Drug Release

The release profiles of ICG and L‐Arg from GIL9R nanoplatform were evaluated using a dynamic dialysis method. Briefly, GIL9R nanoplatform (equivalent to 1.0 mg of both ICG and L‐Arg) was dissolved in 1 mL of phosphate‐buffered saline (PBS) and loaded into a 1000 Da dialysis bag. The bag was submerged in a 50 mL EP tube containing 20 mL of release medium. Four centrifuge tube groups were established to simulate physiological and tumor microenvironments: pH 5.5, pH 5.5 + L, pH 7.4, and pH 7.4 + L. For groups subjected to laser irradiation, samples received NIR laser pretreatment (808 nm, 1.5 W/cm², 10 min) before the release study. The tubes were then placed in a thermostatic water bath shaker (37°C, 100 rpm). At predetermined time intervals (2, 4, 6, 8, 12, 24, 36, and 48 h), 1 mL aliquots of the external medium were withdrawn and immediately replenished with an equal volume of fresh release medium. The absorbance of the collected samples was measured using a UV–vis spectrophotometer at 780 and 515 nm.

2.11. Photothermal Conversion Characterization

Photothermal performance was assessed by irradiating GIL9R nanoplatform with NIR lasers at 808 nm and 0.25, 0.75, 1.5, 2, and 2.5 W/cm². Heating curves were recorded over 10 min at 1.5 W/cm². Subsequently, GIL9R nanoplatform of different concentrations (25, 50, 100, and 200 µg/mL) were irradiated under identical conditions (1.5 W/cm², 10 min) and temperature changes were monitored.

2.12. Photothermal Conversion Efficiency Evaluation

The photothermal conversion efficiency (h) of GIL9R nanoplatform was evaluated under 808 nm laser irradiation. Thermal images were captured every 2 min with an infrared heat camera (FLIR A35, USA) in order to record temperature changes. The GIL9R nanoplatform was irradiated at 1.5 W/cm² for 10 min, followed by natural cooling to ambient temperature. The GIL9R nanoplatform (100 µg/mL) was irradiated at 1.5 W/cm² for 10 min, followed by natural cooling to ambient conditions. The laser cycling (on/off, 30 s interval) was performed five times, with the continuous recording of the changes in temperature. The heating and cooling profiles were fitted to determine h according to the following method:

$$\eta =\frac{\mathrm{hA}\left({\mathrm{T}}_{\max }-{\mathrm{T}}_{\mathrm{surr}}\right)-{\mathrm{Q}}_{\mathrm{Dis}}}{\mathrm{I}\left(1-1^{-A_{\lambda }}\right)}$$ (3)

where h is the heat transfer coefficient, A is the laser irradiation area, T_max is the maximum steady‐state temperature of the sample, T_surr is the ambient temperature, Q_Dis is the heat dissipation by the vessel and solvent, I is the laser power, A_λ is the UV–vis absorbance at the excitation wavelength.

2.13. Cellular Uptake Study

HepG2, HeLa, SNU‐739, and HL‐7702 cells were incubated in confocal dishes at the correct concentrations and allowed to adhere overnight. Cells were treated with GIL9R nanoplatform (10 mg/mL) in the dark for 2, 4, 6, 8, and 12 h. Intracellular fluorescence was observed by CLSM (Leica TCS SP8 + STED).

2.14. Cytotoxicity Assay

HepG2, HeLa, MCF‐7, and HL‐7702 cells were placed in wells of a 96‐well dish and allowed to adhere during the night. The viability of the cells was assessed after they were treated with GIL or GIL9R at doses of 1, 2, 5, 10, and 20 µg/ml in the dark for 48 h. After incubation, the solution was poured off, and the dish was placed in a new one containing 0.1 mL of MTT solution. The cells were then left to grow in the dark at 37°C for 4 h. The supernatant was carefully poured off. Then, 0.2 mL of dimethyl sulfoxide (DMSO) was introduced to dissolve the formazan crystals. Finally, the absorbance at 450 nm was recorded using a microplate reader.

The photothermal cytotoxicity of the formulations was evaluated in HepG2 cells using an MTT assay. Briefly, HepG2 cells were seeded in 96‐well plates at a density of 5 × 10³ cells per well and allowed to adhere overnight. Following attachment, the cells were exposed to serial dilutions of free drugs or nanoformulations (n = 3 per group) and incubated for 24 h. At the 12 h time point, the cells were subjected to 808 nm laser irradiation (1.5 W/cm², 10 min). The plates were then returned to the incubator to continue the 24 h treatment cycle. Subsequently, 50 µL of MTT solution (1 mg/mL in PBS) was added to each well, and the plates were incubated at 37°C for 4 h. The supernatant was carefully aspirated, and the resulting crystals were solubilized in 100 µL of DMSO. Finally, the absorbance was measured at 570 nm using a microplate reader.

2.15. In Vitro ROS Detection

The ROS‐generating capacity of GIL9R nanoplatform was evaluated by fluorescence imaging. HepG2 cells were allocated to six groups: ICG, ICG+L, L‐Arg, L‐Arg+L, GIL9R, GIL9R+L. After adhesion, HepG2 cells were treated with the corresponding formulations and incubated for 12 h in the dark. The cultures were then gently washed, exposed to an 808 nm laser irradiation (1.5 W/cm²) for 10 min, and incubated for a further 30 min at 37°C under protection from light. Following removal from the medium and after additional washes, fresh culture medium was added, and intracellular ROS production was immediately visualized via an optical imaging system.

The DCFH‐DA probe was employed to evaluate intracellular ROS levels. HepG2 cells (1 × 10⁵ cells/well) were seeded into 6‐well plates and cultured in DMEM medium for 24 h. The growth medium was then aspirated, and the cells were treated with 1 mL of DMEM containing various formulations: PBS, PBS+L, ICG, ICG+L, L‐Arg, L‐Arg+L, GIL9R, and GIL9R+L. Following a 12 h incubation, the cells were irradiated with an 808 nm laser (1.5 W/cm², 10 min). After removing the treatment medium, the cells were incubated with the DCFH‐DA probe in the dark for 30 min. Subsequently, the cells were washed twice with PBS (pH 7.4) and centrifuged at 2000 rpm for 5 min. The resulting pellets were washed twice more with PBS and finally resuspended in 1.5 mL of PBS. Fluorescence intensity, indicative of ROS levels, was quantified via flow cytometry.

2.16. In Situ NO Release

In vitro NO production was quantified by the Griess assay. GIL9R nanoplatform dispersions at predetermined doses (100, 200, 300, 400, and 500 µg/mL) were illuminated with an 808 nm beam (1.5 W/cm², 5 min). After irradiation, an aliquot of Griess solution was mixed in, and the sample was allowed to develop in the dark at ambient temperature to allow the color to develop. The supernatant was collected, and its optical density was read at 540 nm. Furthermore, the NO standard curve was obtained via the same protocol.

2.17. Hemolysis Assay

The biocompatibility of GIL9R nanoplatform was further verified through a murine hemolysis assay. Fresh blood was obtained by retro‐orbital puncture, and erythrocytes were isolated by repeated washing with phosphate‐buffered saline (PBS) to obtain a 4% (v/v) cell suspension. A 0.5 mL volume of the erythrocyte suspension was mixed with 0.5 mL of test solutions, which included deionized water (positive control), normal saline (negative control), and GIL9R at doses of 100, 200, 300, 400, 500, and 600 µg/mL. The solutions were incubated in the dark at 37°C for 3 h, after which they were centrifuged at 2000 rpm for 5 min. Photographs were taken, and 100 µL of the supernatant was aspirated from each tube to quantify hemoglobin release at 570 nm.

2.18. In Vivo Drug Distribution

The biodistribution of GIL and GIL9R was investigated in H22 hepatoma‐bearing mice using a small animal in vivo imaging system (IVIS Spectrum, PerkinElmer). Serial fluorescence images were acquired at the time points post‐administration. After the final imaging session, mice were euthanized, and the major organs (heart, liver, spleen, lungs, kidneys, and tumor) were excised for ex vivo imaging to evaluate tissue‐specific accumulation.

2.19. In Vivo Therapeutic Evaluation

The H22 mouse tumor model was created by randomly separating the mice into seven cohorts (n = 5 per cohort): PBS control, GIL, GIL+L, GI9R, GI9R+L, GIL9R, and GIL9R+L. All treatments were delivered via tail‐vein puncture every 48 h. Changes in body mass and tumor volume were measured and recorded throughout the 14‐day study. On day 14, blood samples were collected at the endpoint by retro‐orbital puncture for serum biochemistry analysis. The mice were then killed, and their key bodily organs (heart, liver, spleen, lungs, and kidneys) were harvested for histopathological analysis of their tissue structure. Tumorous samples were fixed in formalin, embedded in paraffin wax, and sectioned for H&E, TUNEL, and Ki67 immunohistochemistry, in order to evaluate antitumorous efficacy and the proliferative index.

2.20. Data Analysis

All statistical results are presented as the mean ± standard deviation (SD). A one‐way analysis of variance (ANOVA) was performed, followed by a Student's t‐test, using Origin 2025 and GraphPad Prism (version 10.1.2). Differences were considered statistically significant if p was less than 0.05. Levels of significance were denoted as follows: ^* p < 0.05, ^** p < 0.01, and ^** p < 0.001.

3. Results and Discussion

3.1. Preparation and Characterization of GIL9R

GNCs were produced using a redox‐mediated solution galvanic reaction with AgNCs acting as sacrificial templates. The reaction was precisely monitored using UV–vis spectroscopy and was stopped when the target absorption peak at 808 nm was reached. TEM images (Figure 1a) revealed that the as‐synthesized AgNCs exhibited uniform cubic morphology with good dispersion, with edge lengths ranging from 40–50 nm. The galvanic replacement process generated hollow interiors within the GNCs, providing a structurally favorable framework for efficient encapsulation of ICG and L‐Arg. Post‐loading characterization confirmed that the morphology integrity of the GIL was preserved without significant alteration. The surface decoration with the 9R‐P201 peptide served a dual function: minimizing drug leakage during systemic circulation and conferring active targeting specificity toward HCC cells, thereby achieving both stability and targeting in a single design strategy. Elemental mapping analysis further validated the successful formation of GNCs, clearly delineating the spatial distribution of Au and residual Ag elements.

FIGURE 1.

Characterization of material structure. (a) Representative TEM images of AgNC, GNC, GIL, and GIL9R; Corresponding elemental mapping images of GNC showing the spatial distribution of Au and Ag elements. (b) Hydrodynamic diameter of AgNC, GNC, GIL, and GIL9R by DLS. (c) UV–vis absorption spectra of AgNC, GNC, GIL, ICG, L‐Arg, and GIL9R. (d) FTIR spectra of L‐Arg, GNC, GL, and GIL9R. (e) zeta potential of AgNC, GNC, GIL, and GIL9R by DLS.

UV–vis absorption spectra acquired during synthesis exhibited typical features of AgNC (Figure 1c), with a dominant LSPR peak at ∼436 nm and shoulder peaks in the 350–380 nm range, consistent with silver nanocube formation. Loading of ICG and L‐Arg onto GNCs induced a slight red shift in the absorption spectrum, while the photothermal conversion capability under NIR irradiation was retained. Successful incorporation of L‐Arg was verified by FTIR spectroscopy (Figure 1d), the characteristic vibrational bands of GIL9R at 1679 cm⁻¹ (C═N vibration), 1132 cm⁻¹ (─NH₂ bending), and 972 cm⁻¹ (C─N bonding) were unambiguously assigned to L‐Arg, verifying its presence within the nanoplatform.

Given that cellular uptake of nanoparticles (NPs) is influenced by surface charge through electrostatic interactions with the negatively charged plasma membrane, we carefully measured both the surface charge (zeta potential) and the size (hydrodynamic diameter) of the NPs. Dynamic light scattering (DLS) analysis was employed to characterize the colloidal properties of the nanoplatform. The hydrodynamic diameters were determined to be 51.7 ± 1.29 nm for AgNC, 96.2 ± 4.9 nm for GNC, 122.4 ± 7.5 nm for GIL, and 209.1 ± 10.7 nm for GIL9R. The polydispersity index (PDI) of GIL9R nanoplatform was measured as 0.225 (Figure 1b and Figure S3a), which correlates well with the morphology observed via transmission electron microscopy (TEM). Furthermore, a 7‐day longitudinal stability assay confirmed the robust colloidal stability of GIL9R in aqueous dispersion (Figure S3b). Zeta potential measurements (Figure 1e) showed values of −13.8 ± 0.86 mV (AgNC), −29.9 ± 0.61 mV (GNC), and −43.2 ± 1.1 mV (GIL). Although ICG bears a negative charge and L‐Arg is neutral at pH 7.0 (below its zero‐charge point of 10.67), the net surface charge of GIL remained negative. Conjugation of the strongly cationic 9R‐P201 peptide reversed the zeta potential of GIL9R nanoplatform to +29.8 ± 1.8 mV, facilitating electrostatic attraction to negatively charged cellular membranes and enhancing cellular internalization.

The optimal feeding ratio for drug loading was determined by establishing a standard calibration for ICG and L‐Arg (Figure S1a,b). Based on the DLE calculations for the GIL9R nanoplatform, three candidate mass ratios were evaluated. Quantitative analysis (Table S1) indicated that increasing the feeding ratio beyond a certain threshold yielded diminishing returns in loading efficiency. Consequently, a 1:2:2 mass ratio was selected as optimal to balance high loading capacity with synthetic economy. This formulation achieved DLEs of 28.59% for ICG and 7.7% for L‐Arg, with corresponding EE of 81.67% and 22.0%, respectively. This optimized formulation was employed for all subsequent experiments.

The in vitro release kinetics of both payloads were monitored (Figure S2). The GIL9R nanoplatform release profiles demonstrated a markedly faster release rate for both ICG and L‐Arg under acidic conditions (pH 5.5) compared to physiological pH (7.4), with a pronounced burst release occurring within the first 12 h. This pH‐responsive behavior is attributed to the protonation‐induced structural changes in the targeting peptide and the inherent instability of the nanocarrier in acidic environments. Furthermore, NIR laser irradiation significantly accelerated drug diffusion from the porous framework via localized hyperthermia. Notably, after 48 h at pH 5.5, cumulative release rates reached 83.41% (ICG) and 76.41% (L‐Arg) in the laser group, significantly higher than the non‐irradiated controls (73.2% and 62.2%, respectively). These findings confirm that GIL9R nanoplatform facilitates rapid, dual‐stimuli‐responsive drug release, establishing a spatiotemporally controlled therapeutic paradigm for tumor treatment.

3.2. In Vitro NO Release Profile

As depicted in Figure S1c, a standard calibration curve for NO quantification was established using the Griess assay with NaNO₂ standards. Figure S1d summarized the NO release profiles of GIL9R nanoplatform under in vitro conditions. In the absence of laser irradiation, GIL9R exhibited only basal NO release, indicating that NO release is predominantly NIR‐dependent. This low‐level background release can be attributed to the oxidation of L‐Arg by endogenous ROS present in the biological milieu. Exposure to an 808 nm laser (1.5 W/cm², 10 min) triggered substantial NO production from GIL9R nanoplatform, a substantial amount of NO in a time‐ and dose‐dependent pattern, far exceeding the basal level. These findings illustrate that laser activation effectively triggers abundant NO generation by GIL9R nanoplatform in vitro, thereby potentiating GT.

3.3. Photothermal Property Assessment

As demonstrated in Figure 2a, the GIL9R nanoplatform displayed excellent photothermal conversion performance. Increasing the NP concentration caused a corresponding increase in heating under irradiation, indicating a clear example of a concentration‐dependent effect. Figure 2b revealed that at a fixed concentration (100 µg/mL), the greater the laser power density, the greater the temperature elevation. Furthermore, the laser power densities and concentration gradients of GNC were also evaluated. The results demonstrated that its temperature increased progressively, indicating that the photothermal performance of GNC was not compromised after drug loading (Figure S4a,b). Photothermal stability was confirmed by five consecutive on/off laser cycles (30 s interval), with minimal variation in the maximum temperature attained (Figure S5). Infrared thermal imaging further visualized the concentration‐dependent heating, showing progressively higher temperatures with increasing GIL9R concentration (Figure 2d), consistent with the quantitative data. The photothermal conversion efficiency (η) of GIL9R nanoplatform was found to be 19.1%, which confirms its superior photothermal properties (Figure 2c).

FIGURE 2.

Test of photothermal efficiency. (a) Photothermal heating graphs of GIL9R at various doses under 808 nm laser treatment (1.5 W/cm², 10 min). (b) Temperature‐time profiles of GIL9R (100 µg/mL) under 808 nm irradiation at various power densities (0.25–2.5 W/cm², 10 min). (c) Time versus ‐lnθ fitting curve (blue line) and corresponding temperature change profile (black line) of GIL9R (100 µg/mL) during a single laser on/off cycle (808 nm, 1.5 W/cm²). (d) Infrared thermal photographs of GIL9R dispersions at various concentrations under 808 nm laser exposure (1.5 W/cm²).

3.4. Cellular Uptake Study

Cellular uptake and intracellular distribution of GIL9R in HepG2 cells were examined by CLSM. Uptake kinetics were assessed in four cell lines (HepG2, HeLa, SNU‐739, HL‐7702) treated with 10 µg/mL GIL9R. At predetermined time points, nuclei were stained with Hoechst 33342, and intracellular fluorescence was recorded. As shown in Figure 3a,b and Figure S6, GIL9R nanoplatform exhibited distinctive intracellular localization patterns across the tested cell lines. Time‐dependent accumulation of fluorescence intensity was prominently observed in HepG2 cells (Figure 3a), reflecting the enhanced nuclear targeting mediated by the 9R‐P201 peptide. In contrast, HeLa and SNU‐739 cells showed no significant increase in fluorescent signal accumulation over time (Figure S6a,b), confirming the targeting specificity toward HepG2. Quantitative analysis of mean fluorescence intensity over various time points further corroborates the specific targeting capability of the nanoplatform toward HepG2 cells (Figure S7a–c). Moreover, HL‐7702 normal hepatocytes exhibited negligible time‐dependent fluorescence enhancement (Figure 3b), indicating minimal non‐specific uptake, which is consistent with the results of the average fluorescence intensity analysis (Figure S7d), highlighting the biocompatibility and selectivity of GIL9R nanoplatform.

FIGURE 3.

In vitro drug distribution. CLSM images showing intracellular fluorescence of GIL9R in (a) HepG2 and (b) HL‐7702 cells at various time points post‐incubation.

3.5. Cytotoxicity Evaluation

Dark cytotoxicity was assessed in a panel of cell lines after 48 h of incubation. As shown in Figure 4a, GIL9R modified with 9R‐P201 exhibited negligible cytotoxicity toward normal hepatocytes, whereas marked cytotoxicity was observed against HCC cells, demonstrating selective recognition of HCC targets. Upon laser irradiation, GIL9R activated GT, leading to maximal cytotoxic efficacy against HCC cells. This outcome underscores the synergistic interplay between PDT and GT (Figure 4d). The low dark toxicity of GIL reflected the inherent safety of ICG and L‐Arg, ensuring good biocompatibility of the composite system (Figure S8a).

FIGURE 4.

Cytotoxicity detection and characterization. (a) Dark cytotoxicity of GIL9R toward four cell lines after 48 h incubation. (b–d) Phototoxicity of (b) L‐Arg, (c) GI9R, and (d) GIL9R toward HepG2 cells upon 808 nm laser irradiation (1.5 W/cm², 10 min). (e) Confocal light microscopy images showing live/dead cell staining under different NP treatments with or without 808 nm laser irradiation.

HepG2 cell viability following laser irradiation across all NPs is summarized in Figure 4b–d. The absence of NIR absorption by L‐Arg led to comparable cell viability with or without laser exposure (Figure 4b). Free ICG showed limited tumoricidal effect under laser irradiation owing to its non‐targeting nature (Figure S8b). Statistical analysis revealed that GNC treatment, irrespective of laser irradiation (1.5 W/cm², 10 min), failed to induce significant inhibition of HepG2 cell proliferation. Furthermore, dose‐response analysis across various laser power densities confirmed that GNCs lack the therapeutic potency to effectively suppress HCC, even under intensified irradiation. In stark contrast, the GIL9R nanoplatform demonstrated superior efficacy by leveraging a synergistic photodynamic‐GT cascade, thereby overcoming the limitations of monotherapy (Figure S8c,d). Consistent with this, GI9R+L showed markedly less cytotoxicity toward HepG2 cells (Figure 4c) compared to GIL9R+L (Figure 4d), highlighting the critical contribution of the GT‐PDT cascade.

The phototoxic effect of GIL9R upon 808 nm NIR laser irradiation was further assessed using a calcein‐AM/PI double staining. As shown in Figure 4e, the GIL9R+L treatment resulted in extensive cell death (red fluorescence), whereas the GIL+L and GIL9R without irradiation consisted predominantly of viable cells (green fluorescence). PBS and non‐irradiated controls maintained normal morphology with minimal cell death. Collectively, these data highlight the potent tumor‐inhibitory effect achieved by the synergistic PTT‐PDT cascade.

3.6. Intracellular ROS and NO Detection

Intracellular ROS levels were evaluated by inverted fluorescence microscopy using a ROS‐sensitive probe. Figure S9a showed that both free ICG and ICG+L groups exhibited only minimal fluorescence, attributable to the limited cellular internalization of unconjugated ICG. In contrast, the GIL9R group under 808 nm laser irradiation displayed intense green fluorescence, indicating robust ROS generation and efficient NIR‐triggered ROS generation. Quantitative analysis of intracellular ROS levels was performed via flow cytometry (Figure S10). Negligible ROS generation was observed in the PBS control group, irrespective of laser irradiation. While free ICG generated a measurable ROS signal upon NIR exposure, it failed to translate this into significant cytotoxicity due to inefficient cellular internalization. In stark contrast, the GIL9R+L group triggered a substantial ROS burst, resulting in lethal oxidative damage to HepG2 cells. These quantitative findings are consistent with the qualitative fluorescence microscopy data, demonstrating that GIL9R generates sufficient ROS to inflict cellular damage and simultaneously catalyze the oxidation of L‐Arg to NO, thereby establishing the desired synergistic PDT‐GT cascade.

NO production was assessed with a specific fluorescent NO probe (Figure S9b). The L‐Arg group alone showed negligible fluorescence, whereas both GIL and GIL9R treatments produced distinct green fluorescence signals, confirming that NO generation requires the synergistic action of ICG‐derived ROS and L‐Arg oxidation. These findings validate the feasibility of the PDT‐GT cascade. The 9R‐P201 peptide conferred active targeting, enabling preferential accumulation of GIL9R in liver cancer cells, thereby amplifying intracellular NO levels and suppressing cellular activity. The integrated PDT‐GT effect, mediated by this cascade mechanism, ultimately promoted liver cancer cell apoptosis.

3.7. Assessment of Biocompatibility

Hemocompatibility was evaluated by incubating GIL9R nanoplatform with freshly drawn blood obtained from H22 tumor‐bearing mice via retro‐orbital bleeding. As depicted in Figure 5a,b, GIL9R demonstrated exceptional hemocompatibility. At a dose of 600 µg/mL, the rate of lysis remained well under 5% (approximately 3.9%), markedly lower than that of the positive control (deionized water). Collectively, these findings demonstrate that GIL9R is suitable for intravenous administration in murine models without inducing significant erythrocyte damage.

FIGURE 5.

In Vivo Biosafety Detection and Drug Distribution Tracking. (a) Hemolysis images and (b) statistical graphs of the hemolysis rate of the nanoplatform. (c) In vivo biodistribution of mice intravenously treated with GIL9R and free ICG at the indicated intervals. (d) Ex vivo distribution images of the distribution of GIL9R (left panel) and free ICG (right panel) in major organs. (e) Real‐time infrared thermal imaging of H22 tumors in mice following the injection of PBS, free ICG, or GIL9R during 808 nm laser exposure (1.5 W/cm², 10 min). (f) Corresponding temperature variation at the cancer site for each treatment group.

3.8. In Vivo Biodistribution

The spatiotemporal uptake of the targeted nanoplatform was assessed in mice bearing H22 tumors using a small‐animal imaging system. Mice were intravenously administered either free ICG or GIL9R, and serial fluorescence images were acquired at predetermined time points (Figure 5c). Free ICG predominantly accumulated in the spleen, liver, and tumor, with renal clearance becoming evident by 48 h. In contrast, GIL9R nanoplatform functionalized with the HCC‐targeting peptide 9R‐P201 circulated systemically and achieved peak tumor accumulation at 24 h, indicative of active targeting. Subsequent clearance occurred predominantly via the liver. In addition, the ex vivo fluorescence imaging of primary tissue samples closely matched the in vivo biodistribution profile (Figure 5d), confirming tumor‐specific accumulation and identifying 24 h as the optimal window for initiating laser therapy.

3.9. In Vivo Antitumor Efficacy

H22 tumors were implanted into mice, which were then randomly randomized into three groups: PBS, ICG+L, and GIL9R+L. As shown in Figures 5e,f, owing to the lack of targeting specificity, free ICG exhibited poor tumor accumulation and only marginal photothermal heating upon laser irradiation. In the GIL9R+L group, 9R‐P201‐mediated targeting resulted in substantial tumor enrichment, generating robust hyperthermia and effective tumor suppression.

For comprehensive in vivo antitumor assessment, H22 hepatoma‐bearing mice were allocated into seven treatment groups (n  =  5 per group): PBS, GIL, GIL+L, GI9R, GI9R+L, GIL9R, and GIL9R+L. Mice in the irradiated groups received 808 nm laser irradiation (1.5 W/cm², for 10 min) 24 h after dosing. Bodily weight and the size of the tumor were measured before each treatment. After the 14‐day therapeutic regimen, the mice were euthanized and their cancers removed and weighed. Representative photographs of excised tumors are provided in Figure 6a. These illustrate that tumors in the GIL9R+L cohort were significantly smaller. Furthermore, tumor weights in the GIL9R+L cohort were noticeably greater than in the GI9R+L cohort, emphasizing the pivotal role of the PDT‐GT cascade in maximizing tumor suppression. The active targeting capability of 9R‐P201 enabled enhanced tumor‐specific accumulation of GIL9R, leading to potent inhibition of tumor growth and superior anticancer efficacy. A statistically notable decrease in cancerous tissue weight was observed in the GIL9R+L cohort relative to the GI9R+L and GIL+L groups, which confirms the exceptional anti‐cancer potency of laser‐activated GIL9R (Figure 6c). Variations in mouse body weight were minor and attributable to environmental factors and interindividual heterogeneity, with no significant differences among groups (Figure 6b). The tumor growth profiles in Figure 6d further demonstrated the superior outcome achieved by the PDT‐GT cascade. Integrated analysis of tumor volume and mass reveals that the synergistic combination of PDT and GT provides enhanced antitumor efficacy.

FIGURE 6.

In vivo antitumor efficacy validation. (a) Tumor growth curves in H22 cells in mice over a 14‐day course of treatment, divided into seven cohorts with or without laser irradiation. (b) Weight trends of the trial mice during the therapeutic regimen. (c) Excised tumor weights after resection at the end of treatment (day 14). (d) Tumor volume progression profiles corresponding to (a).

Histopathological analysis of the key organs (heart, liver, spleen, lungs, and kidneys) using H&E analysis revealed no discernible histopathological abnormalities or pathological changes in any of the treatment groups, which confirms the favorable biosafety profile (Figure S12). Additionally, serum biochemistry analysis (Figure S11) showed no significant abnormalities in hematological parameters across nanoplatform‐treated groups, further validating the excellent biocompatibility of the formulations.

Tumor tissue analysis by H&E and TUNEL staining (Figure S13) demonstrated that while GIL, GI9R, and GIL9R groups retained densely packed cellular architecture, the GIL9R+L group demonstrated marked tissue disintegration and sparsely arranged cells, corroborating enhanced antitumor effects. The TUNEL assay revealed intense green fluorescence in the GIL9R+L group, demonstrating efficient induction of tumor cell apoptosis. These results confirm that the optimized nanoplatform promotes cancer cell death and that the PDT‐GT cascade synergistically augments apoptotic signaling, thereby maximizing anticancer efficacy.

4. Conclusions

We developed a GNC‐based nanoplatform (GIL9R) co‐delivery of ICG and L‐Arg, with surface modification by the HCC‐targeting peptide 9R‐P201, to construct a cascade‐responsive system that orchestrates concurrent PTT, PDT, and GT. In vitro, GIL9R nanoplatform selectively accumulated in HepG2 cells via 9R‐P201‐mediated targeting, generating abundant ROS and NO upon NIR irradiation and exerting potent cytotoxicity toward HCC cells while sparing non‐malignant hepatocytes. In H22 tumor‐bearing mice, GIL9R nanoplatform showed preferential tumor accumulation and significant growth inhibition via the synergistic PDT‐GT cascade, with immunofluorescence and histopathology confirming enhanced apoptosis and tissue disruption in the GIL9R+L group. Blood biochemical and organ histopathological analysis demonstrated excellent biosafety and biocompatibility, supporting intravenous applicability in hepatic cancer therapy. This study presents a rationally designed, cascade‐enabled nanoplatform combining active targeting, multimodal treatment, and biocompatibility. This offers a new, successful approach to treating HCC with precision.

Funding

The National Natural Science Foundation of China (21501044), Natural Science Foundation of Henan Province (222300420411), Youth Cross Fund Key Project of Henan University (HDP20230337), and Special Professor Project of Henan University. (Please include all funding sources including grant numbers).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: smll73927‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.