Combined Enzyme-Responsive Nano-Peptide Drug with Photothermal Function for Effective Anti-Hepatocellular Carcinoma in Mice Model

Abstract

Introduction

The efficacy of photothermal therapy (PTT) for hepatocellular carcinoma (HCC) is severely constrained by tumor heat resistance, which is driven by heat shock protein (HSP) upregulation, alongside the limited tissue penetration of near-infrared (NIR) laser.

Methods

An enzyme-responsive “peptide-drug” sequence (PD) was combined with gold nanorods (GNR) to construct a tumor-targeting microgel system (GNR/PD-M). The formulation was comprehensively evaluated in HepG2 cells and BALB/c nude mice bearing subcutaneous HCC xenografts, with key assessments including MMP-2-responsive drug release, in vitro cytotoxicity, mitochondrial and tumor targeting specificity, in vivo anti-tumor efficacy, and biosafety profiles.

Results

In both cellular and animal models, the GNR/PD-M + laser irradiation group exhibited the strongest tumor cell inhibition. Mechanistically, this treatment significantly downregulated HSP90 expression and disrupted mitochondrial function, thereby overcoming tumor heat resistance and reducing tumor cell viability. Consistently, the in vivo anti-tumor effect of GNR/PD-M + laser was markedly superior to that of the control groups.

Conclusion

The GNR/PD-M system synergistically enhances PTT efficacy against HCC by mitochondrial targeting and heat resistance reversal, representing a promising strategy for intelligent and targeted drug delivery.

Graphical Abstract

Introduction

Mitochondria, serving as the metabolic and apoptotic command centers of eukaryotic cells, dictate the proliferative, migratory and invasive phenotypes of malignancies. Consequently, active targeting of tumor cell mitochondria has emerged as an important therapeutic strategy for tumor therapy, and combing PTT with mitochondrial targeting would be more effective in tumor ablation.¹ Among the current mitochondrial targeting strategies, peptides have gained widespread attention due to their advantages of ease of synthesis, cost-effectiveness, structural tunability and functional diversity.²

Photothermal therapy (PTT) is a novel, non-invasive cellular thermal ablation method that utilizes high temperatures to destroy or eliminate tumor cells.³ However, when tumor cells are exposed to high temperatures, they have been observed to over-express protective heat shock proteins (HSPs), which increase the heat stress tolerance of tumor cells, thereby limiting the efficacy of PTT. Notably, the expression of HSPs is dependent on the energy supply of mitochondria-derived adenosine triphosphate (ATP).^4,5 Additionally, the efficacy of PTT is constrained by the limited penetration depth and range of NIR laser.⁶

Rhein (RH), an anthraquinone extracted from Rheum palmatum (family Polygonaceae), presents potential antitumor activity. Modern pharmacological studies have proved that RH can induce apoptosis by activating the caspase-dependent mitochondrial apoptotic pathway, leading to mitochondrial dysfunction and cutting off the energy supply of tumor cells, thus exerting anti-tumor effects.^7,8 Therefore, the joint action of RH and PTT on tumor cells compensates for limited penetration depth of PTT and potentiate the effect of PTT ablation.⁹ Our previous studies have demonstrated that the combination of active ingredients from traditional Chinese medicine (TCM) with photodynamic therapy (PDT) yields promising antitumor effects; however, certain limitations in targeting efficiency remain to be addressed.¹⁰

Cancer drug delivery systems based on nanomaterials have recently been shown to exert potential effect on cancer cells.^11,12 Given that unloaded drugs affect the stability of nanosystems under simple mixing,¹³ we designed a “peptide-drug” sequence (PD). PDC is an emerging targeted nanotechnology based on the concept of Paul Ehrlich’s “magic bullets”.¹⁴ By connecting peptides to cytotoxic molecules via linkers to selectively deliver drugs to tumor target sites, drug targeting is improved and adverse effects reduced. Enzyme-sensitive linkers are more frequently used in PDC drugs due to the influence of extracellular environmental pH and serum protease inhibitors, which make proteases inactive in plasma. Endogenous matrix metalloproteinases (MMPs) serve as endogenous triggers and stable PDC linkers owing to their stable overexpression.^15,16

As mentioned above, we obtained PD with cationic mitochondria-targeting peptide sequence FrFKFrFK (r = darginine) as the inner layer, which is connected to RH through MMP-2.^17,18 Coating this functional peptide segment on the photothermal material gold nanorods (GNRs), we constructed a multifunctional nano-delivery system with both photothermal and mitochondria-targeting (GNR/PD). Pre-tests revealed that the particle sizes of the nanoplatform were small. Although minor nanoparticles are shown to be more efficient, the formulation would be diluted during blood circulation and interact with plasma proteins and other components of the blood, which would compromise the stability.¹⁹ Nano-formulations with a diameter of about 100 nm maximally avoid the reticuloendothelial system, thus prolonging their exposure time in blood and facilitating the accumulation at the tumor site.²⁰ Therefore, we encapsulated GNR/PD with hybridized micelles, then modified it with folic acid on the surface to improve stability (GNR/PD-M), drug loading capacity, and tumor targeting ability.²¹

Herein, we systematically evaluated the dual tumor-and mitochondria-targeting ability, enzyme-responsive release characteristics, and synergistic antitumor efficacy of GNR/PD-M in HepG2 cells and hepatocellular carcinoma-bearing mice.

Materials and Methods

Materials

D-α-tocopheryl polyethylene glycol 1000 succinate (TPGS) and hydrogen tetrachloroaurate (III) (HAuCl₄) were purchased from Shanghai Macklin Biochemical Technology Co., Ltd (Shanghai, China). 1,2-Distearoyl-sn-glycero-3-phosphoethanolamine-N-[folate (polyethylene glycol)] (FA-DSPE-PEG) was obtained from Xi’an Ruixi Biological Technology Co., Ltd (Xi’an China). Phosphate-buffered saline (PBS, 0.01 M, powder, pH 7.2–7.4) was provided by Shanghai Acmec Biochemical Co., Ltd (Shanghai, China). Fetal bovine serum (FBS) and Dulbecco’s modified Eagle medium (DMEM) were purchased from GIBCO (Thermo Fisher Scientific, Carlsbad, California, USA).

The Cell Counting Kit-8 (CCK-8) was obtained from APExBIO Technology (Houston, TX, USA). 2’,7’-Dichlorodihydrofluorescein diacetate and JC-1 mitochondrial membrane potential fluorescent probe was obtained from Solarbio (Beijing, China). Primary antibodies against Bcl-2 or Caspase-3 were obtained from Abcam (Cambridge, UK). Penicillin-streptomycin solution (100×) was purchased from Biosharp (Hefei, China). Dichloromethane and methanol were purchased from Shanghai Macklin Biochemical Co., Ltd (Shanghai, China).

Animal Model

HepG2 cells were obtained from Shanghai Mcellbank Biotechnology Co., Ltd. Male Balb/c nude mice (4–6 weeks old, 20 ± 2 g) were purchased from Gempharmatech Co., Ltd. (production License No. SCXK [Chuan] 2020–034). All animal experiments were performed in accordance with the Ethical Guidelines for the Welfare of Experimental Animals of Southwest Medical University and approved by its Institutional Animal Care and Ethics Committee (Approval No. SWMU20240144). To establish the hepatocellular carcinoma (HCC) xenograft model, 2×10⁶ HepG2 cells were injected subcutaneously into the right axilla of each mouse. Subsequently, the mice were euthanized under deep anesthesia, followed by cervical dislocation.

Preparation and Characterization of GNR

GNR was prepared by seed growth method. HAuCl₄ (5 mL, 1 mM) was dissolved in cetyltrimethylammonium bromide (CTAB) solution (10 mL, 0.2 M), then AgNO_{3 (}250 µL, 4 mM), HCl (8 µL, 37%), and Vitamin C (70 µL) were added. The mixture was stirred rapidly until it was colorless and clarified, and added with immediate addition of NaBH₄ (15 µL, 10 mM), followed by a vigorous stir for 2 min and incubation for 6 h at 37 °C to obtain GNR.

The morphology and characteristics of GNR were observed with transmission electron microscopy (TEM). The GNR content was determined with Inductively Coupled Plasma Mass Spectrometry (ICP-MS). The particle size, polydispersity index (PDI) and zeta potential were evaluated using dynamic light scattering (Nano ZS 90, Malvern, Malvern city, UK).

Preparation and Characterization of GNR/PD-M

PD peptides were prepared via solid-phase synthesis. GNRs were mixed with a defined proportion of PD, incubated, and centrifuged to obtain GNR/PD. DSPE-PEG-FA and TPGS were dissolved in organic solvents and evaporated to form a thin film, and then GNR/PD was slowly added upon rehydration to obtain the GNR/PD-M nanomicellar system.

An aliquot of GNR/PD-M was dispersed in PBS buffer and vortexed to determine its particle size, PDI, and zeta potential. The critical micelle concentration (CMC) of the GNR/PD-M systems was also measured. The morphologies of GNRs and GNR/PD-M were observed via TEM. In addition, the concentration of PD was quantified by HPLC at a detection wavelength of 220 nm. The drug encapsulation efficiency (DEE) and drug loading efficiency (DLE) of GNR/PD-M were calculated as follows:

where M represents the total amount of PD added initially, W represents the amount of free (unencapsulated) PD in the supernatant after centrifugation, and M₀ represents the total mass of the carrier materials (including GNRs, DSPE-PEG-FA, and TPGS).

Stability Evaluation of GNR/PD-M

Aliquots of GNR/PD-M were dispersed in PBS, pure water or 10% serum to determine their particle size and zeta potential using a Malvern Zetasizer. Measurements were performed after storage for 1, 3, 5, 7, and 14 days. The stability was assessed by monitoring changes in these parameters over time in the three different media.

Photothermal Efficiency of GNR/PD-M

In vitro photothermal performance was evaluated as follows. GNR, GNR/PD, and GNR/PD-M at various concentrations (5, 10, 20, and 50 µg/mL based on GNR content) were precisely dispensed into 96-well plates and irradiated with 808 nm laser for 5 min, during which temperature changes were recorded 1 min. The photothermal efficiency of GNR/PD-M was assessed by comparing its temperature elevation with that of PBS. To examine photothermal stability, GNR/PD-M underwent three cycles of laser irradiation (5 min each), and the temperature elevation profiles were compared across cycles. For in vivo evaluation, mice bearing subcutaneous HepG2 tumors were injected with GNR/PD-M, and the tumor sites were irradiated with an 808 nm laser, while surface temperature changes were monitored with a thermal camera to assess the photothermal effect of the formulation.

Enzymatic Response of GNR/PD-M

The lyophilized MMP-2 was centrifuged at 10,000–12,000 rpm for 30s to collect the protein adhered to the tube walls at the bottom. After centrifugation, the pellet was reconstituted in re-solubilization buffer by blowing gently with a pipette gun or by gently inverting the cap of the tube and centrifuging at low speed for a few seconds. The stock solution was adjusted to a concentration of 100 µg/mL and vortexed vigorously. The reconstituted protein was allowed to stand at room temperature for a few minutes to ensure complete dissolution.

The assay buffer (TCNB buffer, pH 7.5) was composed of 50 mM Tris, 10 mM CaCl₂, 150 mM NaCl, and 0.05% (w/v) Brij 35. MMP-2 was diluted to 100 µg/mL in this buffer and activated by incubation with 1 mM p-aminophenylmercuric acetate (APMA) at 37 °C for 1 h prior to use.

Equivalent concentrations of PD, GNR/PD-M were accurately prepared and added to the TCNB buffer containing activated MMP-2. The mixtures were incubated in a constant-temperature water bath at 37 °C under shaking. The enzyme-responsive drug release from each formulation was evaluated by monitoring the increase in free drug content at predetermined time points (0, 1, 2, 4, 6, 8, 12, and 24 h).

In vitro Cellular Uptake and Internalization Analysis

To investigate the intracellular distribution, coumarin 6 was encapsulated in the nanoformulations as a fluorescent tracer instead of the drug. All procedures were performed under light-protected conditions. The cellular uptake of the nanoformulations by HepG2 cells was evaluated using confocal laser scanning microscopy (CLSM, TCS SP2, Leica) and flow cytometry (FCM, Becton Dickinson, Franklin Lakes, NJ, USA).

For CLSM observation, HepG2 cells were seeded in coverslip-lined 24-well plates at a density of 5×10⁴ cells per well and cultured overnight. Subsequently, the cells were incubated with either GNR/DiD-M or free DiD for 2 h, followed with rinsing 2–3 times with PBS. Then the cells were fixed with 4% paraformaldehyde for 30 min and 4’,6-diamidino-2-phenylindole (DAPI) for 5 min. Intracellular localization and internalization were visualized by CLSM.

For flow cytometric analysis, the cells were collected, washed with PBS, resuspended, and then analyzed by FCM.

Cell Viability and Cytotoxicity Assessment via CCK-8 Assay

The in vitro cytotoxicity of various formulations (GNR, GNR + Laser, GNR/PD, GNR/PD + Laser, GNR/PD, GNR/PD-M + Laser) against HepG2 cells were evaluated by CCK-8 assay. The principle of this assay is based on the reduction of the tetrazolium salt by mitochondrial dehydrogenases in viable cells, generating water-soluble orange formazan products that can be quantified by absorbance measurement.

Briefly, HepG2 cells were seeded into 96-well plates at 1×10⁴ cells/well and cultured overnight at 37 °C in a 5% CO₂ atmosphere. Subsequently, the cells were treated with serial dilutions of each formulation for 24 h, after which they were washed three times with PBS. A 10% CCK-8 solution in serum-free medium was freshly prepared and 100 µL was added to each well, followed by 40-min incubation. The absorbance of each well was then measured at 490 nm using a microplate reader.

Cell viability was calculated as follows:

Where OD sample is the absorbance of cells treated with the test formulations, OD _blank is the absorbance of cell-free wells containing culture medium only, and OD _control is the absorbance of untreated control cells.

Data were analyzed using GraphPad software. Cell viability curves were fitted, and the half-maximal inhibitory concentration (IC50) values were calculated.

Detection of Mitochondrial Reactive Oxygen Species (ROS)

Intracellular ROS generation was evaluated using the DCFH-DA fluorescent probe. HepG2 cells were seeded into 24-well plates at 5×10⁴ cells/well, incubated overnight. The cells were then treated with various formulations (RH, GNR + Laser, GNR/PD + Laser, GNR/PD-M, and GNR/PD-M + Laser) at equivalent GNR or PD concentrations for 8 h, with serum-free medium serving as the control. Subsequently, the cells were incubated with 2’,7’-dichlorodihydrofluorescein diacetate (DCFH-DA, 10 µM) for 30 min, washed three times with PBS, and visualized under an inverted fluorescence microscope. Fluorescence intensity was analyzed with Image J software to evaluate ROS accumulation in HepG2 cells.

Live/Dead Cell Staining Test

The Calcein AM/PI live/dead cell double staining test was performed to validate viability results obtained from the CCK-8 assay. HepG2 cells were seeded into 6-well plates at a density of 5×10⁵ per well and cultured overnight. Subsequently, the cells were harvested by trypsinzation and resuspended in assay buffer at 1×10⁵⁻⁶ cells/mL and stained with Calcein-AM (2 µL per 1 mL of cell suspension) for 20 min at 37 °C, protected from light. After that, propidium iodide (PI) was added (5 µL per 1 mL of cell suspension), and the mixture was incubated for 5 min at room temperature in the dark. After staining, the cells were visualized and photographed under an inverted fluorescence microscope.

In vitro Cell Migration Assay

The anti-migration ability of various formulations (GNR, GNR + Laser, GNR/PD, GNR/PD + Laser, GNR/PD-M, GNR/PD-M + Laser) against HepG2 cells was evaluated by the scratch wound healing assay, with serum free medium as the control. HepG2 cells were seeded into 6-well plates at a density of 4×10⁵ cells/well and cultured overnight. Once the cells were adherent to the wall and fully-grown, three parallel scratches were made vertically across the bottom of each well using a sterile 200-µL pipette tip. The detached cells were washed three times with PBS, and transferred to the microscope for visualizing and the width of the scratches were immediately recorded. Subsequently, the cells were treated with the indicated formulations at equivalent GNR or PD concentrations in serum-free medium for 12 h. After incubation, the scratches were photographed again, and cell migration was evaluated by measuring the change in scratch width using ImageJ software.

Mitochondrial Function Assessment

HepG2 cells were seeded into 6-well plates at 5×10⁴ cells per well and cultured for 24 h. The cells were then treated with various formulations (GNR + Laser, GNR/PD, GNR/PD, GNR/PD +Laser, GNR/PD-M, and GNR/PD-M + Laser) at equivalent concentrations for an additional 24 h, with serum-free medium as the control. After treatment, the cells were washed three times with PBS and incubated with JC-1 probe (diluted to 5 µg/mL in blank medium) for 20 min. The cells were then rinsed with PBS to remove excess dye, and mitochondrial membrane potential was visualized under a fluorescence microscope.

For ATP quantification, cells (5×10⁶ cells/well) were collected into centrifuge tubes, washed with cold PBS, and centrifuged (6000 rpm, 4 °C, 10 min), and the supernatant was discarded. The cell pellets were resuspended in 1 mL of deionized water and lysed by ultrasonication in an ice bath (20% amplitude, with 2 s on/1 s off cycles for 1 min). The lysates were centrifuged at 8,000 × g for 15 min at 4 °C, and the supernatants were collected and transferred to 96-well plates and mixed with the ATP assay working solution provided in the kit. The ATP content was calculated according to the manufacturer’s instructions and measured using a microplate reader.

Hemolysis Test

Blood was collected from the hearts of healthy rats via cardiac puncture. Whole blood (1 mL) was taken and diluted with saline (3 mL). The mixture was centrifuged at 2500 rpm for 10 min. These steps were repeated three times, and the supernatant was carefully removed after each centrifugation to obtain the packed erythrocytes, which was then collected with a sterile pipette and 4 drops of which was mixed with 8 mL of saline to prepare 2% (v/v) erythrocyte suspension.

Various formulations (GNR, PD, GNR/PD, GNR/PD-M) at equivalent GNR or PD concentrations were mixed with equal volume of the 2% erythrocyte suspension and incubated at 37 °C for 3 h. After incubation, the mixtures were centrifuged at 3000 rpm for 10 min, and 100 µL of the supernatant was transferred to a 96-well plate. Normal saline was used as negative control, and ultrapure water (UP water) was used as positive control. The absorbance of the supernatant was measured at 540 nm using a microplate reader. The hemolysis ratio (HR%) was calculated using the following formula:

where A _sample is the absorbance of the test formulation group, A _neg is the absorbance of the negative control (normal saline), and A _pos is the absorbance of the positive control (ultrapure water).

Western Blotting Analysis

HepG2 cells were seeded into 6-well plates (1 × 10⁵ cells per well), incubated for 24 h, followed by treatment with the indicated formulations (including untreated cells as the control) for an additional 24 h. The Cells were harvested and lysed with radio-immunoprecipitation assay (RIPA) buffer, After centrifugation at 12000 rpm for 20 min, the supernatant was collected and protein samples were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride (PVDF) membranes and blocked with Protein Free Rapid Blocking Buffer (5×) (Shanghai Epizyme Biomedical Technology Co., Ltd, Shanghai, China), diluted to 1× with PBST, for 1 h at room temperature, followed with overnight incubation at 4 °C with primary rabbit antibodies against Bcl-2, Caspase 3 and HSP90, followed by incubation for 1 h at 25 °C with goat anti-rabbit secondary antibody conjugated to horseradish peroxidase (HRP). Antibody binding was visualized using an enhanced chemiluminescence detection system (Image Quant LAS 4000 Mini; Fuji, Tokyo, Japan) and quantified with Gel Image System 4.00 (Tanon, Shanghai, China).

In vivo Biodistribution of DiD-Loaded Formulations

To track the distribution of the nanoformulations, DiD was encapsulated as a fluorescent tracer in place of the drug. HepG2 tumor-bearing mice were randomly divided into 3 groups (n=3 per group) and intravenously injected via the tail vein with free DiD, GNR/DiD, or GNR/DiD-M. DiD at an equivalent DiD dose of 5.00 mg/kg body weight. Fluorescence signals in live mice were recorded using multimodal small animal in vivo imaging system photographed at predetermined time points (1, 2, 4, 8, 12, 24, and 48 h post-injection). After the final time point, the mice were euthanized, and major organs and tumors were excised for ex vivo fluorescence imaging. In vivo distribution was assessed on the basis of the fluorescence intensity.

Pharmacokinetic Test

Sprague-Dawley (SD) rats were randomly divided into 3 groups (n=3 per group) and intravenously injected via the tail vein with free PD, GNR/PD or GNR/PD-M at an equivalent PD dose of 2.00 mg/kg. Blood samples were collected via cardiac puncture at 1, 2, 4, 8, 12 and 24 h post-injection, transferred to heparin sodium-containing microcentrifuge tubes, and centrifuged at 6000 rpm for 5 min to obtain plasma. The plasma samples were processed according to the established protocol, and the supernatant was filtered through a 0.22 µm membrane. The PD concentration in the plasma was determined by high-performance liquid chromatography (HPLC), and the drug content was calculated using a standard curve.

Pharmacokinetic parameters, including the area under the concentration-time curve AUC0-t (mg/L×h), maximum drug concentration (C_max, mg/L), half-life (T_1/2, h), and mean retention time (MRT_0-t, h) were calculated with the DAS 2.0 statistical software.

Establishment of Tumor-Bearing Mouse Model, Treatment Protocol, and Safety Evaluation

In vivo Antitumor Efficacy

BALB/c nude mice (male, specific pathogen-free) bearing subcutaneous HepG2 tumors were established as described in the Animal Model section. When the tumor volume reached approximately 100 mm³, the mice were randomly divided into 6 groups (n=5 per group) and intravenously injected via the tail vein with saline (control), free PD, GNR/PD, GNR/PD + Laser, GNR/PD-M, or GNR/PD-M + Laser at a dose of 3 mg/kg (PD equivalent). The formulations were administered every other day for a total of six injections. For the laser-treated groups, tumors were irradiated with an 808 nm laser (1.0 W/cm⁻², 5min) at 4 h post-injection on each administration day. Body weight and tumor volume were monitored every other day throughout the treatment period. Tumor volume was measured using digital calipers and calculated as (length × width²)/2. At the end of the treatment period, all mice were executed and tumors were excised and photographed. Antitumor efficacy was evaluated based on tumor volume growth curves and body weight changes.

Histopathological Analysis (H&E Staining)

At the end of the experiment, major organs (heart, liver, spleen, lung, and kidney) and tumors were excised, washed with PBS, fixed with 4% paraformaldehyde overnight, dehydrated, embedded in paraffin and sectioned. The sections were deparaffinized, rehydrated and stained with hematoxylin for 20 min, then washed with distilled water and differentiated in 1% acid alcohol for 5 s. After further washing, the sections were placed in warm water (50 °C) for bluing, then dehydrated through a graded ethanol series, cleared in xylene, and mounted with neutral balsam. Images were acquired using a digital slide scanner.

TUNEL Apoptosis Assay

Tumor tissue sections were processed for TUNEL staining according to the manufacturer’s instructions. After TUNEL staining, sections were counterstained with DAPI, mounted, and visualized under a fluorescence microscope.

Immunohistochemistry (IHC) Analysis

Tumor sections were deparaffinized, rehydrated and subjected to antigen retrieval. After blocking with 10% fetal bovine serum for 30 min at room temperature, sections were incubated overnight at 4 °C with primary antibodies against Ki67, Bax, and VEGF. After washing with PBS, sections were incubated with the corresponding HRP-conjugated secondary antibody for 1 h at room temperature. Sections were washed again, counterstained with DAPI for 5 min in the dark, and then visualized and photographed under a fluorescence microscope.

In vivo Biosafety Evaluation

At the end of treatment, blood samples were collected via retro-orbital puncture, transferred to microcentrifuge tubes, and centrifuged at 6000 rpm for 3 min to obtain serum. Serum biochemical parameter, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), urea, creatinine (CREA) and creatine kinase (CK), were measured using an automatic biochemical analyzer to evaluate hepatic, renal, and cardiac functions, respectively.

Statistical Analysis

Data were expressed as mean ± standard deviation (SD) and analyzed using GraphPad Prism 8.0.1. Differences between two groups were assessed using a two-tailed Student’s t-test, whereas comparisons among three or more groups were performed by one-way analysis of variance (ANOVA). A p-value < 0.05 was considered statistically significant.

Results

Physical and Chemical Characterization, Stability Assessment, and Enzyme Responsiveness Evaluation

The GNR solution prepared according to the described method appeared as a purplish clarified solution (Figure 1A). TEM observation revealed a rod-like morphology with relatively uniform length and diameter (Figure 1B). The content of GNR was 149.99 ± 19.8 µg/mL per portion by ICP, and the particle size of GNR was measured to be around 36.05±2.15 nm using a Malvern Zetasizer, with a PDI of 0.279 ± 0.03 (n=3), and a zeta potential of 36.37 ± 1.50 mV. The molecular weight of PD was 2808.14 Da, with a purity over 98.00% as determined by HPLC, and the mass spectrometry results are shown in Figure 1C.

Figure 1.

Preparation and characterization of GNR and PD. (A) Photograph of as-prepared GNR solution. (B) TEM image of GNR (scale bar = 100 nm). (C) The mass spectrum results of PD showing the molecular weight of the synthesized peptide.

The DEE of GNR/PD-M was measured to be 83.23 ± 0.23% (n=3) with a DLE of 8.02%. Dynamic light scattering (DLS) measurements (Figure 2A) showed that the particle size, PDI, and zeta potential of GNR/PD-M were 117.2 ± 11.01 nm, 0.259 ± 0.02, and –6.85 ± 0.66 mV, respectively. All PDI values were below 0.3, meeting the quality requirements for nanoformulations. TEM imaging (Figure 2B) revealed a spherical-like morphology of GNR/PD-M, further confirming the successful preparation of the micelles.

Figure 2.

Preparation and characterization of GNR/PD-M. (A) Particle size distribution of GNR/PD-M measured by dynamic light scattering (DLS). (B) Transmission electron microscopy (TEM) image of GNR/PD-M (scale bar = 100 nm). (C) Critical micelle concentration (CMC) of GNR/PD-M determined by fluorescence spectroscopy. (D) Particle size stability of GNR/PD-M over 14 days of storage. (E) Zeta potential stability of GNR/PD-M over 14 days of storage.

The critical micelle concentration (CMC) of GNR/PD-M was determined using pyrene as afluorescence probe. As shown in Figure 2C, the fluorescence intensity ratio (I₃₇₇/I₃₈₉) was plotted against the logarithm of micelle concentration, and the CMC value was calculated from the intersection of the two fitted lines. The CMC of GNR/PD-M was approximately 0.0022–0.008mg/mL, indicating that the micelles could remain stable after dilution.

The colloidal stability of GNR/PD-M was evaluated by monitoring particle size and zeta potential after storage at 4 °C for 1, 3, 5, 7, and 14 days in different media (PBS buffer, purified water, and 10% serum), as shown in Figure 2D and E. Compared with the initial measurements, no significant changes were observed, indicating good stability of GNR/PD-M.

The photothermal properties of GNR, GNR/PD, and GNR/PD-M at various concentrations were further investigated. The results are shown in Figure 3A–D. Under 808 nm laser irradiation at a power density of 1.0 W/cm⁻², the temperature elevation ranges were 21.3–49.3 °C for GNR, 27.1–51.1 °C for GNR/PD, and 27.1–54.2 °C for GNR/PD-M. Based on these results, GNR/PD-M at 5 µg/mL of was selected for repeated irradiation cycles (three cycles), and the heating curves showed no significant changes, confirming the photothermal stability of the formulation (Figure 3E). For in vivo photothermal evaluation, GNR/PD-M was intravenously injected into tumor-bearing mice, and the tumor site was irradiated with an 808 nm laser at 2 h post-injection (Figure 3F and G). The surface temperature of the tumors reached 43.7 °C after 5 min of irradiation, meeting the requirement for mild photothermal therapy (40–45 °C).

Figure 3.

In vitro and in vivo photothermal properties and enzyme-responsive drug release of GNR/PD-M. (A) Temperature elevation curves of GNR, GNR/PD, and GNR/PD-M at 5 μg/mL under 808 nm laser irradiation (1.0 W/cm²) over different time points (0–5 min). (B) Temperature elevation curves of GNR, GNR/PD, and GNR/PD-M at 10 μg/mL under 808 nm laser irradiation (1.0 W/cm²) over different time points (0–5 min). (C) Temperature elevation curves of GNR, GNR/PD, and GNR/PD-M at 20 μg/mL under 808 nm laser irradiation (1.0 W/cm²) over different time points (0–5 min). (D) Temperature elevation curves of GNR, GNR/PD, and GNR/PD-M at 50 μg/mL under 808 nm laser irradiation (1.0 W/cm²) over different time points (0–5 min). (E) Photothermal cycling stability of GNR/PD-M over five consecutive 5-min on/off irradiation cycles. (F) Temperature changes at the tumor site in mice after subcutaneous injection of GNR/PD-M under 808 nm laser irradiation. (G) Infrared thermal images of mice after subcutaneous injection of GNR/PD-M at different time points under 808 nm laser irradiation. (H) Enzyme-responsive release behavior: Changes in PD levels following incubation with PD, GNR/PD-M, GNR/PD-M+Laser, GNR/PD-M+MMP-2, and GNR/PD-M+Laser+MMP-2 after incubation with MMP-2.

The enzyme-responsive drug release was elevated by incubating equivalent concentrations of PD and GNR/PD-M in TCNB buffer without or without MMP-2 (Figure 3H). The free PD group exhibited the fastest depletion rate upon direct exposure to the enzyme. In the presence of laser irradiation, the PD content in both GNR/PD-M groups decreased more rapidly than that in the non-irradiated group, which might be due to accelerated drug release from the micelles and the increased enzyme activity at elevated temperatures. Notably, the PD content in the GNR/PD-M group declined more rapidly at the early time points, which may be attributed to concentration gradient driven drug penetration. At the later time points, the remaining drug amounts were similar across groups, likely due to the limited enzyme amount. Collectively, these results indicate that PD possesses enzyme-responsive release characteristics, and GNR/PD-M exhibits a sustained release profile.

Cellular Uptake Efficiency, Anti-Proliferative and Anti-Migratory Effects, and Preliminary Mechanistic Investigation

The uptake of different nanoformulations by HepG2 cells was evaluated using CLSM and FCM after co-incubated with free coumarin 6, PD/Au, or GNR/PD-M (Figure 4A–C). As shown by subsequent flow cytometric analysis, the HepG2 cells incubated with GNR/PD-M exhibited the strongest fluorescence intensity, suggesting that the micellar formulation significantly enhanced cellular uptake compared with free coumarin 6 (****P < 0.0001).

Figure 4.

Cellular uptake efficiency of GNR/PD-M in HepG2 cells. (A) Confocal laser scanning microscopy (CLSM) images of cells after incubation with GNR/PD-M (scale bar = 50 μm). (B) Flow cytometry (FCM) histograms showing the fluorescence intensity distribution of cells after incubation with GNR/PD-M. (C) Semi-quantitative analysis of mean fluorescence intensity from flow cytometry data (mean ± SD, n = 3; ****P < 0.0001).

The in vitro cytotoxicity of GNR, GNR + Laser, GNR/PD, GNR/PD + Laser, GNR/PD-M, and GNR/PD-M + Laser against HepG2 cells was assessed by the CCK-8 assay (Figure 5A). All treatment groups exhibited dose-dependent cytotoxic effect. Notably, the GNR/PD-M + Laser group demonstrated the strongest cytotoxicity, with cell viability of falling below 50% at 5 µg/mL after 24 h of administration. The IC50 value of this group was significantly different from those of the other groups, indicating potent in vitro antitumor activity.

Figure 5.

In vitro toxicity of GNR/PD-M in HepG2 cells. (A) Cell viability determined by CCK-8 assay after 24 h of treatment with different concentrations (mean ± SD, n = 3; ****P < 0.0001, ***P < 0.001, **P < 0.01). (B) Live/dead staining images of HepG2 cells (green: Calcein-AM for live cells; red: propidium iodide (PI) for dead cells; scale bar = 100 μm). (C) ROS generation in HepG2 cells detected by DCFH-DA fluorescent probe (scale bar = 100 μm).

The live/dead cell staining assay further confirmed these findings. As shown in Figure 5B, significantly more dead cells were observed the GNR/PD-M + Laser group under the same treatment conditions, corroborating the results of the CCK-8 assay.

Intracellular ROS levels were examined using the DCFH-DA fluorescent probe (Figure 5C). Compared with the control group, all treatment groups induced ROS production to varying degrees, with the GNR/PD-M + Laser group showing the highest fluorescence intensity. This indicated that the combination treatment was markedly more effective at triggering oxidative stress in HepG2 cells than the other formulations.

The anti-migratory effects of the various formulations were evaluated using a scratch wound healing assay (Figure 6A and B). At equivalent concentrations, all treatment groups inhibited tumor cell migration to varying extents. Notably, the GNR/PD-M + Laser group not only suppressed cell migration and invasion, but also induced cell death, as clearly observed in the scratch images.

Figure 6.

In vitro migration and mitochondrial function evaluation of GNR/PD-M in HepG2 cells. (A) Wound healing images of HepG2 cells after treatment across different groups at 0 and 12 h (scale bar = 100 nm). (B) Semi-quantitative analysis of the relative migration rate from the wound healing assay (mean ± SD, n = 3; ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05). (C) JC-1 staining for mitochondrial membrane potential in HepG2 cells after treatment across different groups (red: JC-1 aggregates indicating normal mitochondria; green: JC-1 monomers indicating depolarized mitochondria). (D) ATP levels in HepG2 cells after treatment across different groups (mean ± SD, n = 3).

Mitochondrial membrane potential was assessed using the JC-1 test, and the results showed a trend consistent with the ROS measurements (Figure 6C). Normal mitochondria exhibited red fluorescence, whereas damaged mitochondria showed green fluorescence. In parallel, ATP content was measured (Figure 6D), revealing a significant reduction in the GNR/PD-M + Laser group compared with the blank group (****P < 0.0001), which was inversely correlated with the ROS generation trend.

The biocompatibility of nanoformulations was evaluated by hemolysis test (Figure 7A and B). The hemolysis rates of all tested nanoformulations were below 5%, indicating excellent hemocompatibility and suggesting a low risk of hemolysis upon intravenous administration.

Figure 7.

Hemolytic activity of GNR/PD-M and other nano-formulations. (A) Photographs of red blood cells after incubation with ultrapure water (positive control), saline (negative control), GNR, PD, GNR/PD, and GNR/PD-M. (B) Hemolysis rates of different formulations (mean ± SD, n = 3; ****P < 0.0001, ***P < 0.001).

The expression level of apoptosis-and heat shock-related proteins (Bcl-2, Caspase-3, and HSP90) in HepG2 cells after various treatment were determined by Western blotting (Figure 8A–D). And analysis revealed that the GNR/PD-M + Laser group significantly down-regulated Bcl-2 and HSP90 expression while upregulating Caspase-3 expression, thereby exerting a potent antitumor effect. These results further corroborated the in vitro cytotoxicity findings.

Figure 8.

Western blot analysis of protein expression in HepG2 cells after treatment across different groups. (A) Representative Western blot bands showing the protein expression levels of Bcl-2, Caspase-3, and HSP90. (B–D) Semi-quantitative analysis of relative protein expression normalized to GAPDH (mean ± SD, n = 3; ****P < 0.0001, ***P < 0.001, **P < 0.01).

In vivo Antitumor Efficacy and Preliminary Safety Evaluation

The biodistribution of DiD-labeled formulations was monitored using a multimodal small animal in vivo imaging system. In all groups, the fluorescence signal was predominantly concentrated in the liver and kidneys, indicating that DiD was mainly metabolized and excreted via the hepatic and renal pathways after intravenous injection. Representative fluorescence images of mice at different time points post-injection revealed that the GNR/DiD-M group exhibited obvious fluorescence at the tumor site as early as 2 h post-injection, and the signal intensity gradually increased over time, suggesting progressive accumulation of DiD at the tumor site and confirming its tumor -targeting property. Among the three formulations, the GNR/DiD-M group showed the longest retention time and the strongest tumor-targeting capability, which further confirmed the pharmacokinetic results.

At 24 h post-injection, tumors and major organs of mice were excised for ex vivo fluorescence imaging (Figure 9A). The tumor fluorescence intensity in the GNR/DiD-M group was significantly higher than that in the other two groups, consistent with the results of in vivo imaging. Quantitative analysis of fluorescence intensity of tumors and organs in each group (Figure 9B) further confirmed that the GNR/DiD-M group showed statistically significant differences compared with the free DiD and GNR/DiD groups. Collectively, these results demonstrate that GNR/DiD-M possesses superior tumor targeting and prolonged circulation time in vivo.

Figure 9.

In vivo targeting and biodistribution of GNR/DiD-M in tumor-bearing nude mice. (A) In vivo near-infrared (NIR) fluorescence images of tumor-bearing nude mice at different time points after intravenous injection of free DiD, GNR/DiD, and GNR/DiD-M (DiD dose: equivalent in all groups). (B) Ex vivo NIR fluorescence images of major organs (heart, liver, spleen, lungs, and kidneys) and tumors isolated from mice after injection (mean ± SD, n = 3; ****P < 0.0001, ***P < 0.001, **P < 0.01).

The pharmacokinetic profiles of PD, GNR/PD, and GNR/PD-M are shown in Figure 10. The GNR/PD-M group exhibited the longest blood circulation time, followed by GNR/PD and free PD. The prolonged of blood circulation time may be attributed to the protective effect of the hydrophilic outer shell of the micelles and the polyethylene glycol (PEG) coating, which could reduce phagocytosis and clearance by the reticuloendothelial system. The pharmacokinetic kinetic parameters, including the half-life, were also calculated (Table 1). The GNR/PD-M group showed the longest half-life, indicating a lower in vivo clearance rate compared with the other formulations.

Figure 10.

Plasma concentration–time profiles of PD in BALB/c nude mice after intravenous administration of free PD, GNR/PD, and GNR/PD-M over 24 h (mean ± SD, n = 3).

Table 1.

Pharmacokinetic Parameters of Different Formulations After Intravenous Injection in Nude Mice

	AUC0-t (μg/L×h)	MRT0-t (h)	T1/2z (h)	Cmax (μg/L)
PD	0.82±0.0177	7.19±0.3217	4.42±1.8074	0.07±0.0038
GNR/PD	1.06±0.0046	9.69±0.1326	11.48±0.2312	0.08±0.0037
GNR/PD-M	1.29±0.0417	9.72±0.0552	19.60±1.4856	0.09±0.0016

Abbreviations: AUC_0–t, area under the plasma concentration–time curve from time 0 to the last measurable time point; MRT_0–t, mean residence time from time 0 to the last measurable time point; T_1/2z, terminal elimination half-life; C_max, maximum plasma concentration; SD, standard deviation.

The in vivo antitumor efficacy of the various formulations was evaluated by monitoring the nude carrying ectopic HepG2 tumors in different nanoformulation groups. During the treatment period, no significant change in body weight of mice were observed in any group (Figure 11A). In the saline group, tumor volume increased rapidly from 105.52 ± 22.64 mm³ to 614.40 ± 74.92 mm³ (Figure 11B). In contrast, all nanoformulation groups showed varying degrees of inhibitory effects on tumor growth, with the GNR/PD-M + Laser group exhibiting the most pronounced antitumor effect, as reflected by the smallest tumor volume and lowest tumor weight after treatment (Figure 11C and D).

Figure 11.

In vivo pharmacodynamics of GNR/PD-M in tumor-bearing nude mice. (A) Body weight changes of mice during the treatment period across different formulation groups (mean ± SD, n = 5). (B) Tumor growth curves of mice after treatment with different formulations over the experimental period (mean ± SD, n = 5). (C) Weights of excised tumors at the end of treatment (mean ± SD, n = 5; ****P < 0.0001, ***P < 0.001). (D) Representative photographs of excised tumors from each group at the end of treatment.

To further investigate the antitumor mechanisms, tumor tissues were subjected to H&E and TUNEL staining (Figure 12A and B). Compared with the saline group, which showed almost no apoptotic cells, the GNR/PD-M + Laser group exhibited more pronounced cytoplasmic degradation, nuclear damage, and pro-apoptotic effects. Meanwhile, H&E staining of major organs (heart, liver, spleen, lung and kidney) revealed no significant histopathological abnormalities between the saline and treatment groups, preliminarily indicating good targeting and low tissue toxicity of the nanoformulations.

Figure 12.

In vivo safety evaluation of GNR/PD-M in tumor-bearing nude mice. (A) H&E-stained sections of major organs (heart, liver, spleen, lungs, and kidneys) and tumors from mice after treatment across different groups (scale bar = 50 μm). (B) TUNEL-stained sections of tumor tissues across different groups (scale bar = 50 μm). (C) Immunohistochemical staining images of Ki67, Bax, and VEGF in tumor tissues across different groups (scale bar = 50 μm). (D and E) Serum biochemical parameters in mice at the end of treatment, including liver function markers (ALT and AST), cardiac function marker (CK), and renal function markers (UREA and CREA) (mean ± SD, n = 5).

Immunohistochemical (IHC) staining was performed to assess the expression of Ki67, Bax, and vascular endothelial growth factor (VEGF) in tumor tissues across experimental groups. Ki67 serves as a marker of cell proliferation, Bax as a pro-apoptotic protein, and VEGF as a key regulator of angiogenesis. As illustrated in Figure 12C, relative to the saline control group, the GNR/PD-M + Laser group displayed a significant reduction in Ki67-positive cell proportion (indicating suppressed proliferation) and a marked downregulation of VEGF expression (suggesting impaired angiogenic capacity). Additionally, extensive Bax-positive staining (predominantly in the cytoplasm) was observed in the GNR/PD-M + Laser group, indicative of robust tumor cell apoptosis. In contrast, the saline control group showed negligible Bax-positive cells. Collectively, these findings demonstrate that GNR/PD-M combined with laser treatment exerts potent anti-tumor effects by inhibiting tumor cell proliferation, downregulating pro-angiogenic factor expression, and inducing substantial apoptotic responses.

Serum biochemical parameters (ALT, AST, CK, CREA-S and UREA) were measured to assess potential organ toxicity (Figure 12D and E). Notably, the indicator levels in the treatment group were marginally elevated relative to those in the untreated tumor-bearing control group. This observation is more likely attributable to systemic metabolic stress induced by tumor burden,²² rather than inherent hepatotoxicity, nephrotoxicity, or myotoxicity associated with the GNR/PD-M. Histopathological examination of major organs (eg, liver, kidney, skeletal muscle) further corroborated this conclusion. However, the long-term safety profile of this formulation necessitates systematic assessment via recovery-phase monitoring and expanded toxicological investigations in subsequent translational research endeavors.

Discussion

Mitochondria play a critical role in the energy metabolism and apoptotic regulation of tumor cells; however, their targeted delivery efficiency remains constrained by the limited availability of carrier systems that integrate both high specificity and safety. Peptide-based delivery has emerged as a promising strategy for mitochondrial targeting, with the advantages of facile synthesis, structural adaptability, favorable biocompatibility, and efficient cellular uptake. Nevertheless, peptides inherently suffer from short circulation half-lives, potential immunogenicity, and toxicity concerns.²³ In this study, we integrated a mitochondrion-targeting peptides, an MMP-2 enzyme-responsive sequences and rhein with gold nanorods to construct a dual-targeting nano-delivery system (GNR/PD-M). Furtherly, we investigated the feasibility of this delivery platform for mitochondrion-targeted drug delivery and its potential as an efficient therapeutic system.

Our experimental results demonstrated that the GNR/PD-M + Laser group was efficiently internalized by HepG2 cells in vitro, with fluorescence intensity approximately 10⁴ -fold higher than that of the free C6 group, indicating that folic acid (FA) modification endows the formulation with robust tumor cell recognition ability. This finding was further confirmed by in vivo imaging studies in liver cancer-bearing mouse models, where GNR/PD-M exhibited effective accumulation at tumor sites following intravenous administration. JC-1 mitochondrial membrane potential and ATP level assays revealed that the local photothermal effect generated by near-infrared laser irradiation significantly induced mitochondrial membrane depolorization and ATP depletion. Moreover, both in vitro cytotoxicity and the in vivo tumor suppression experiment confirmed that this synergistic strategy exerted significantly superior anti-tumor effects against HepG2 cells and liver cancer-bearing mice compared with monotherapy, achieving a tumor suppression rate of 92.56%.

Unlike conventional PTT, which directly ablates tumors by elevating local temperature above 50 °C, mild-temperature PTT (below 45 °C) substantially reduces the risk of thermal damage to adjacent normal tissues. However, mild PTT tends to upregulate the expression of HSP, thereby enhancing tumor cell thermotolerance and compromising therapeutic efficacy. In the present study, the surface temperature of tumors in the GNR/PD-M + Laser group was maintained below 45°C during laser irradiation. Notably, HSP90 expression in the tumor tissues was significantly downregulated, suggesting that the formulation effectively attenuated the heat stress response. This observation may be attributed to mitochondrial structural damage, which reduces the energy supply required for protein synthesis of tumor cells was damaged, reducing the energy provided for protein synthesis—a notion supported by the JC-1 membrane potential and ATP level results. Although these findings support the therapeutic potential of this mitochondrial-targeting delivery platform, further ultrastructural verification (eg, TEM to assess mitochondrial cristae and outer membrane integrity) is warranted. In addition, combining confocal microscopy with JC-1/ROS probes could further clarify changes in mitochondrial membrane potential and oxidative stress levels, thereby comprehensively elucidating the subcellular mechanism underlying the observed effects.

In vivo safety of GNR/PD-M was systematically elevated through hemolysis tests, H&E staining, TUNEL staining, and serum biochemical analyses. No significant decrease in the mice’s body weight was observed during the treatment period, and H&E staining of major organs revealed no apparent pathological abnormalities, indicating that the formulation not only effectively suppresses tumor growth but also exhibits favorable in vivo biocompatibility. However, some serum biochemical parameters were slightly outside the normal reference range, which may reflect physiological alterations induced by drug administration, representing unavoidable side effects associated with tumor growth inhibition. Thus, systematic long-term safety evaluation are still required before clinical translation.

It is worth noting that the synthetic route of PD involves multiple condensation and purification steps, and the overall yield requires further optimization—an issue that may represent one of the main bottlenecks for clinical application. Additionally, the current conclusions are based solely on the HepG2 line and its corresponding xenograft model; therefore, the generalizability of this strategy to other liver cancer subtypes or other solid tumors remains to be validated. Given that MMP-2 is highly expressed in the tumor microenvironment of various malignant tumors (including breast cancer, melanoma, and pancreatic cancer), and given that the physical ablation effect of PTT is independent of specific tumor antigens, this delivery platform holds theoretical potential against solid tumors with high MMP-2 expression by loading different bioactive molecules. Further studies should verify the universality of this strategy using diverse cell lines and corresponding animal models, and explore its therapeutic efficacy in metastatic or drug-resistant tumor models.

Conclusion

In summary, we have successfully constructed a dual-targeting nano-delivery system (GNR/PD-M) that integrates mitochondrial targeting, enzyme-responsive drug release, and photothermal therapy into a single platform. Our results demonstrate that the GNR/PD-M + Laser combination significantly enhances the efficacy of PTT by co-targeting tumor cells and their mitochondria, leading to effective suppression of tumor growth in the liver cancer-bearing mouse models. These findings highlight the potential of GNR/PD-M + Laser as a promising drug delivery system for mitochondrial-targeted cancer therapy and warrant further investigation toward clinical translation.

Funding Statement

This work was supported by the State Key Laboratory of Southwestern Chinese Medicine Resources (Grant No. SKLTCM202309), the Special Project of Sichuan Traditional Chinese Medicine Administration (Grant No. 25MSZX442), the Dazhou Administration of Traditional Chinese Medicine Special Project (Grant No. 2025LHZRZD03) and The Student’s Platform for innovation and entrepreneurship training Program (Grant No. 202510632085).

Abbreviations

ALT, alanine aminotransferase; AST, aspartate aminotransferase; CCK-8, cell Counting Kit-8; Colorectal cancer; CREA, creatinine; DCFH-DA, 2’,7’-Dichlorodihydrofluorescein diacetate; DEE, drug encapsulation efficiency; DLE, drug loading efficiency; FA, folic acid; FBS, Fetal bovine serum; FCM, flow cytometer; HSPs, heat shock proteins; RH, Rhein; PTT, Photothermal therapy; ROS, reactive oxygen species.

Data Sharing Statement

The key data are contained in the manuscript. Further request can be obtained from the corresponding author Dan Zhang.

Ethical Approval and Consent to Participate

All in vivo procedures have been approved by Animal Ethics Committee of the Center for Experimental Animal Research of Southwest Medical University (Number: swum 20240307-020).

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. The author warrants that his/her contribution is original and that he/she has full power to make this consent for publication.

Disclosure

The authors declared no conflicts of interest.