Promoting Early Diagnosis and Precise Therapy of Hepatocellular

Jun 10, 2019 - Therefore, new targets with specific distribution in tumors and high expression even at the early stages of disease are urgently needed...
0 downloads 0 Views 10MB Size
Download PDF
Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Promoting Early Diagnosis and Precise Therapy of Hepatocellular Carcinoma by Glypican-3-Targeted Synergistic ChemoPhotothermal Theranostics Weiwei Mu, Dandan Jiang, Shengjun Mu, Shuang Liang, Yongjun Liu,* and Na Zhang* Department of Pharmaceutics, Key Laboratory of Chemical Biology (Ministry of Education), School of Pharmaceutical Sciences, Shandong University, 44 Wenhua Xi Road, Jinan, Shandong Province 250012, People’s Republic of China Downloaded via UNIV OF SOUTHERN INDIANA on July 17, 2019 at 08:31:13 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: The specific-targeting approach could promote the specificity of diagnosis and the accuracy of cancer treatment. The choice of a specific-targeting receptor is the key step in this approach. Glypican-3 (GPC3) is an oncofetal proteoglycan anchored on the cell membrane. It is overexpressed even in the early stage of hepatocellular carcinoma (HCC), whereas it shows almost no expression in the healthy adult liver. Therefore, GPC3 may be applied as a specific-targeting receptor for HCC theranostics. In this study, a GPC3 specific-targeting theranostics nanodevice, GPC3 targeting peptide (named G12)-modified liposomes co-loaded with sorafenib (SF) and IR780 iodide (IR780), was developed (GSI-Lip), which aims to realize early diagnosis and precise chemo-photothermal therapy of HCC. SF was the first-line chemotherapy drug for the treatment of HCC. IR780 was used for photothermal therapy and near-infrared fluorescence imaging. The evaluation of early diagnosis verified that early-stage tumors (3.45 ± 0.98 mm3, 2 days after 5 × 105 H22 cells’ inoculation in mice) could be clearly detected using GSILip, which was significantly more sensitive than folic acid-modified liposomes (p < 0.01, 32.90 ± 10.01 mm3, 4 days after 1 × 106 H22 cells’ inoculation in mice). The study of the endocytic pathway indicated that specific G12/GPC3 recognition may induce caveolae-mediated endocytosis of GSI-Lip. Notably, the accumulation of GSI-Lip in tumors was significantly increased compared with that observed with folic acid-modified liposomes (p < 0.01). Specific-targeting endowed the precise antitumor effect of GSI-Lip. GSI-Lip showed a higher antitumor efficacy in comparison with folic acid-modified liposomes (inhibition rate: 90.52% vs 84.22%, respectively; p < 0.01). During a period of 21 days, the synergistic chemo-photothermal therapy (GSI-Lip + laser) exhibited a better antitumor effect versus GSI-Lip without laser (inhibition rate: 94.93% vs 90.52%, respectively; p < 0.01). Overall, GPC3-targeted GSI-Lip promoted the sensitivity and specificity of HCC early diagnosis and achieved synergistic efficacy of chemo-photothermal theranostics, which has potential clinical applications. Furthermore, the present study revealed that a more specific-targeting ligand could further improve the efficacy of theranostics against HCC. KEYWORDS: specific-targeting, glypican-3, early diagnosis, chemo-photothermal therapy, hepatocellular carcinoma theranostics

■

and Drug Administration (FDA) for the detection of cancer.8,9 An ideal targeting-receptor should be expressed more selectively or uniquely in tumor cells, and must be presented at an abundant level on the cell surface. Therefore, new targets with specific distribution in tumors and high expression even at the early stages of disease are urgently needed. Glypican-3 (GPC3)an oncofetal proteoglycan located on the cytomembraneis closely correlated with tumorigenesis and poor prognosis of HCC.10−12 GPC3 is highly expressed in approximately 81% of HCC cases, whereas it is not expressed in normal adult organs.13 In addition, the expression of GPC3

INTRODUCTION

The poor prognosis associated with hepatocellular carcinoma (HCC) is a consequence of the lack of effective strategies for early diagnosis and limited treatment options.1 Researchers have devoted extensive efforts toward the development of theranostics against HCC, providing potential strategies for the improvement of the prognosis of this disease. The specifictargeting approach is a promising strategy for improving the early diagnosis of cancer and precision of therapy.2−4 The specific-targeting approach, which acts through ligand− receptor recognition, may promote the specificity of diagnosis and the accuracy of cancer theranostics.5−7 For example, BladderChek and LYMPHOSEEK, which actively target nuclear matrix protein 22 and the mannose receptor, respectively, have been approved by the United States Food © 2019 American Chemical Society

Received: March 28, 2019 Accepted: June 10, 2019 Published: June 10, 2019 23591

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces is independent of the tumor size and HCC stage, and is detectable even in the early stage of HCC.14,15 Therefore, it was hypothesized that a GPC3-targeting strategy may carry great potential for the early diagnosis and precise therapy of HCC. Apart from specific-targeting, an effective therapy strategy is also crucial to improve prognosis. Control of tumor regression using a single treatment scheme is challenging, owing to the rapid tumor proliferation, invasion, and metastasis at the advanced stage of cancer.16,17 Recently, combination therapysynergizing the efficacies of different therapeutic agentshas become the predominant theme in cancer therapy.18 For example, Dong et al. designed Au−silica nanoparticles with folic acid (FA) modification, which integrated the use of chemotherapy, radiotherapy, and computed tomography imaging for the treatment of HCC.19 Chemo-photothermal therapy (chemo-PTT) has been widely accepted as a beneficial approach to synergize efficacy in cancer therapy.20,21 PTT increases the sensitivity of cancer cells to chemotherapeutic drugs, and enhances the chemotherapeutic efficacy at the same dosage in vitro and in vivo.22 Regarding chemotherapeutic drugs, the use of sorafenib (SF)the first multikinase inhibitor approved by the FDA in 2007is preferentially recommended as the first-line treatment of “unresectable” HCC.23 Meanwhile, the exogenous chromophores IR780 iodide (IR780) can be used for PTT, and is conveniently detected through a near-infrared fluorescence (NIRF) detection system for diagnosis. IR780 absorbs the NIR light at the region of 650−900 nm, and produces localized cytotoxic heat upon irradiation with NIR light for PTT.24 Consequently, the combination of GPC3-targeting and chemoPTT is expected to achieve the early diagnosis of HCC through NIRF imaging, enhance the therapeutic efficacy of drugs against HCC, and effectively inhibit tumor growth. Nanodevices play an important role in ensuring the efficacy of tumor theranostics. At present, different nanodevices have been developed for theranostics, including lipid-based nanodevices,25 polymer-based nanodevices,26 metallic,27 and inorganic nanodevices.28 Liposomesone of the earliest types of nanodevices approved by the FDAhave received extensive attention during the previous years as carriers of pharmaceutical compounds.29 With the success of Doxil, Onivyde, and so forth, liposomes have firm industrialization foundation and favorable acceptance by clinicians and patients.30 Liposomes with excellent drug-loading capacity, favorable biocompatibility, convenient modification, and an excellent safety profile are suitable nanodevices for use in theranostics against HCC, aiming toward early diagnosis and precise therapy. Herein, a GPC3-targeting synergistic chemo-PTT theranostics strategy involving a GPC3-targeting peptide (termed G12) composed of modified liposomes co-loaded with SF and IR780 (GSI-Lip) was utilized as a versatile nanodevice for early diagnosis and precise therapy of HCC (Scheme 1). In this study, a specific-targeting peptide with high recognition capacity for GPC3-positive cells was selected and confirmed according to previous research.31,32 The specific-targeting ability of GSI-Lip was studied via evaluations of cellular uptake and NIRF imaging. In addition, the endocytic pathway of G12-modified liposomes was investigated for the first time in vitro. Initially, a feasible evaluation method under different tumor-bearing cell numbers and tumor growth time was established to evaluate the sensitivity of GSI-Lip for the early

Scheme 1. Illustration of GPC3-Specific-Targeting Synergistic Chemo-Photothermal Theranostics Design of HCCa

a

GPC3-targeting peptide-modified liposomes co-loaded with SF and IR780 (GSI-Lip) for early diagnosis and chemo-PTT of HCC.

diagnosis of HCC. Moreover, the efficacy of chemo-PTT synergistic therapy was further evaluated. To the best of our knowledge, this was the first attempt to investigate the effect of the GPC3-targeting approach on the early diagnosis and chemo-PTT therapy of HCC.

■

MATERIALS AND METHODS

Materials. SF was purchased from Shanghai Biochempartner Co., Ltd. (Shanghai, RPC). FITC-DHLASLWWGTEL [TJ12P1] (termed FITC-G12); FITC-ALNVGGTYFLTTAQ [L5] (termed FITC-G14), and FITC-YFLTTAQ [L5-2] (termed FITC-G7) were synthesized by Nanjing Leon Biological Technology Co., Ltd. (Nanjing, RPC). Egg phosphatidylcholine (EPC) was obtained from AVT Pharmaceutical Technology Co., Ltd. (Shanghai, RPC). Cholesterol (Chol) were bought from Sigma-Aldrich (St. Louis, MO, USA). Coumarin-6 (C6) was bought from Aladdin Chemical Co., Ltd. (Shanghai, RPC). Glypican 3 antibody (FITC) was provided by Wuhan Biorbyt Biotechnology Co., Ltd. (Wuhan, RPC). All other reagents were obtained from commercial sources and were of analytical-grade purity or greater. Cell Culture. Human HCC (HepG2) cells and normal hepatocellular HL-7702 cells were provided by ATCC and Institute of Biochemistry & Molecular Biology of Shandong University (Jinan, PRC), respectively. HepG2 and HL-7702 cells were cultured in Dulbecco’s modified Eagle’s medium media and RPMI-1640 media, respectively, at 37 °C under 5% CO2 humidified environment. Both the culture media were supplemented with 10% (v/v) fetal bovine serum. Animals. Female BALB/c nude mice (5−6 weeks) and female KM mice (weight: 20 ± 2 g) were provided by Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, PRC) and Medical Animals Experimental Center of Shandong University (Jinan, PRC), respectively. The experiment was authorized by the Laboratory Animal Ethical and Welfare Committee of Cheeloo College of Medicine, Shandong University. All experiments were carried out in accordance with the requirements of Animal Management Rules of PRC (document no. 55, 2001). 23592

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces Expression of GPC3. GPC3 expression in HepG2/HL-7702 cell lines from human and H22 cells/liver parenchymal cells isolated from normal mouse were analyzed by confocal laser scanning microscopy (CLSM) as well as flow cytometry (FCM) using anti-GPC3 antibody (FITC-labeled). Briefly, HepG2 and HL-7702 cells were respectively seeded with the cell numbers at 1 × 105 cells/well and incubated overnight. After the culture media was removed, the cells were handled with phosphate-buffered saline (PBS) containing GPC3 antibody (FITC) at a concentration of 200 ng/mL for 0.5 h on ice. The cells were fixed in 4% paraformaldehyde, washed with cold PBS, and then incubated with 4′,6-diamidino-2-phenylindole (DAPI) for 10 min each. Finally, the dish was examined under CLSM (LSM 780, Carl Zeiss, Germany). To quantify the expression of GPC3, HepG2 and HL-7702 cells were digested, centrifuged, and re-suspended by cold PBS and detected by a FACS Calibur flow cytometer (BD Biosciences, San Jose, USA). H22 cells were obtained from murine H22 ascites sarcoma. Liver parenchymal cells were isolated from normal mouse. Briefly, the healthy mice were sacrificed and liver was harvested. Then, the liver organ was ground by a 200 mesh and washed with cold PBS. The liver parenchymal cells were obtained after centrifugation of the separated liver cell suspensions three times (600 rpm). H22 and liver parenchymal cells were respectively seeded in 2 mL centrifuge tube with the cell number at 1 × 105, and the cells were further incubated with PBS containing fluorescein isothiocyanate (FITC)-labeled GPC3 antibody (200 ng/mL) for 0.5 h on ice. Afterwards, the cells were washed twice with cold PBS, and then stained with DAPI for 10 min each. Finally, the cells were visualized using a CLSM. To quantify the expression level of GPC3 on the cell surface, H22 and liver parenchymal cells were centrifuged, re-suspended, and collected for FCM assay. Scanning of the GPC3-Targeting Peptides. CLSM and FCM were used to analyze the relative binding ability of the three GPC3targeting peptides to different cells lines (HepG2/HL-7702 cell lines from human and H22 cells/liver parenchymal cells isolated from normal mouse). The procedure was the same as described in the “Expression of GPC3” part, except for using FITC-G12, FITC-G14, and FITC-G7 instead of the GPC3 antibody. Synthesis of Function Materials. DSPE-PEG2000-G12 was synthesized as shown in Figure S1. Briefly, G-12 (1.0 equiv) and DSPE-PEG2000-Mal (1.2 equiv) were dissolved in 2 mL of PBS (pH 7.4) and 2 mL of dimethyl sulphoxide (DMSO), respectively. Both solutions were mixed and stirred at room temperature for 8 h. The product was purified using a dialysis method (molecular weight cut-off [MWCO] 5000 Da; Millipore) for 48 h, followed by lyophilization. The purity was verified by both 1H nuclear magnetic resonance (1H NMR) (Avance DPX-300; Bruker BioSpin GmbH, Rheinstetten, Germany) and Fourier-transform infrared spectra (6700 FTIR NXR FT-Raman; Nicolet, USA). DSPE-PEG2000-FA was previously synthesized by our group (Figure S2). Briefly, Folic acid (FA, 1.2 equiv) and DSPE-PEG2000amine (1.0 equiv) were dissolved in 2 mL of distilled water (with 10% sodium hydroxide solution) and 1 mL of DMSO, respectively. After mixing the two solutions, N,N′-dicyclohexylcarbodiimide/N-hydroxysuccinimide was added to the reaction solution for 12 h. Following completion of the reaction, the solution was dialyzed using a dialysis bag (MWCO 3500 Da, Millipore) and distilled water for 48 h to remove the unconjugated reactants. All procedures were performed under limited light. The structure of DSPE-PEG2000-FA was verified through 1H NMR. Preparation of Liposomes. GSI-Lip was prepared using the film dispersion method. Briefly, 4 mL of ethanol solutions of EPC (80 mg), Chol (10 mg), SF (4 mg), and IR780 (0.4 mg) were evaporated at 40 °C to form the dried lipid film, followed by hydration with 0.001 M PBS (pH 7.4) containing DSPE-PEG2000-G12 (3% mol ratio relative to the total phospholipids). The liposomes were extruded thrice using membranes filters (0.45 and 0.22 μm), and subsequently five times using a LiposoFast-Basic extruder (0.20 and 0.10 μm) (Avestin, Ottawa, ON, Canada).

Characterization of Liposomes. Transmission electron microscopy (TEM) was used to evaluate the morphology of liposomes. The particle size, polydispersity index (PDI), and zeta potential of the liposomes were measured by a Zetasizer Nano ZS90 (Malvern, Worcestershire, UK). All samples were evaluated in triplicate. The encapsulation efficiency (EE %) and drug loading (DL %) were calculated using the following formula DL % = (Wloaded drug /Wliposomes) × 100%

EE % = (Wloaded drug /Wtotal drug) × 100% Wloaded drug represents the amounts of the loaded drug in GSI-Lip. Wtotal drug and Wliposomes were the weights of added drug and the liposome system, respectively. To further verify that SF and IR780 are effectively loaded into liposomes, ultraviolet−visible (UV−vis) spectrophotometric analyses of GSI-Lip were performed. Briefly, 2 mL (each) of free SF, free IR780, and GSI-Lip in water were subjected to scanning (wavelength: 200−900 nm) using a UV−vis spectrophotometer (U-2910, Hitachi, Japan) at a suitable concentration. The temperature profile of GSI-Lip during irradiation with NIR light was determined using a thermocouple needle. Briefly, 1 mL (each) of normal saline (NS), free IR780, I-Lip, or GSI-Lip (10 μg/ mL of IR780) was irradiated using a NIR laser (808 nm) at an intensity of 0.6 W/cm2, and the temperature of each sample was measured within 5 min. Stability of Liposomes in Different Media. Plasma (20%, v/v) was selected for the preliminary investigation of the stability of GSILip. The samples were prepared and diluted with 20% plasma, and stored at 37 °C. The particle size of GSI-Lip was measured after incubation for 0, 1, 4, 8, 12, 24, 48, and 72 h, respectively. Moreover, the long-term physical stability of GSI-Lip was evaluated at 4 °C. The particle sizes were recorded every 3 days and for a total of 30 days. In Vitro Release Profiles of SF. The release profiles of SF from the GSI-Lip and GSI-Lip + laser (0.6 W/cm2 808 nm laser for 5 min) were investigated using the dialysis bag method. Briefly, each dialysis bag containing 1.5 mL of the sample was placed into a tube, and 15 mL PBS solution (pH 7.4 containing 1% Tween-80) was added as the release medium, under 37 °C with a stirring speed at 100 rpm. The release medium was collected and replaced with fresh medium at the predetermined time points. The concentration of released SF was determined through high-performance liquid chromatography, and the amounts of cumulative release were calculated. The release profiles of SF/IR780 co-loaded liposomes (SI-Lip) and free SF were simultaneously reported for comparison. Experiments with each group were carried out in triplicate. In Vitro Cytotoxicity Assay. An analysis of the cytotoxicity of GSI-Lip was performed in HepG2 cells. Briefly, HepG2 cells (5000/ well) were seeded into 96-well plates and incubated overnight at 37 °C. Subsequently, the samples of each group with different concentrations of SF (i.e., 0.1−40 μg/mL) were added to the wells, and the cells were further incubated for 24 or 48 h. After adding MTT and DMSO, the cell viability in each group was measured using a microplate reader (model 680; Bio-Rad, CA, USA). H22 cells were incubated with free IR780, I-Lip, or GSI-Lip (with IR780 concentrations of 0.01, 0.1, 1, 2, and 4 μg/mL) to determine the cell viability under the dark condition of PTT therapy. The cells were treated using NIR laser (808 nm) irradiation at an intensity of 0.6 W/cm2 for 5 min. The cell viability of irradiated cells was evaluated through the MTT method. All experiments were performed in triplicate. The relative cell viability (%) was calculated using the following formula

Relative cell viability (%) = (A sample /Acontrol ) × 100% Assessments of Cellular Uptake in Vitro. C6 was selected as a tracer agent to analyze the cellular uptake of the formulations in vitro. HepG2 cells (1 × 105 cells/well) were incubated overnight. This was followed by addition of fresh medium containing free C6, C6-loaded liposomes (C6-Lip), or G12-modified C6-loaded liposomes (GC623593

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces

Figure 1. G12 showed the highest specific recognition ability to GPC3-positive cells. The expression of GPC3 in different cell lines and specific recognition ability of GPC3-targeting peptides to different cell lines were evaluated by CLSM and FCM. (a) CLSM images, (b) FCM histogram profiles of fluorescence intensity, and (c) FCM analysis of GPC3 expression in HepG2 cells and HL-7702 cells, respectively. Scale bar: 20 μm. (d) CLSM images, (e) FCM histogram profiles of fluorescence intensity, (f) FCM analysis of GPC3 expression in H22 cells and liver parenchymal cells, respectively. Scale bar: 20 μm. (g) CLSM images, (h) FCM histogram profiles of fluorescence intensity, (i) FCM analysis of GPC3-targeting peptides selecting in HepG2 cells and HL-7702 cells, respectively. Scale bar: 10 μm. (j) CLSM images, (k) FCM histogram profiles of fluorescence intensity, (l) FCM analysis of GPC3-targeting peptides selecting in H22 cells and liver parenchymal cells, respectively. Scale bar: 20 μm. Data were given as mean ± SD (n = 3), **p < 0.01, ##p < 0.01. Lip: C6 concentration at 200 ng/mL), and further incubation for 30 min. Subsequently, the cells were washed with cold PBS and stained with DAPI. Finally, the cells were imaged using an inverted fluorescence microscope (Axio Observer. A1; Carl Zeiss, Germany). Concerning the quantification, cells treated with free C6, C6-Lip, or GC6-Lip were collected and assessed using FCM. Furthermore, the targeting properties of GC6-Lip were evaluated after pre-incubation with excess free G12 to saturate GPC3. Following the addition of GC6-Lip or C6-Lip into the wells and incubation for 30 min, the cellassociated fluorescence was determined through CLSM and FCM, respectively. Endocytic Pathway Study. Chemical inhibitors blocking the specific endocytic pathway were used to investigate the endocytic manners of the liposomes. Briefly, 1 × 105 cells were cultured overnight in a 12-well plate, and pre-incubated with chlorpromazine (10 μg/mL), genistein (1 μg/mL), or cytochalasin D (30 mM). Subsequently, C6-Lip or GC6-Lip was added into the plate, and the cells were incubated further for 1 h. The cells were treated as described earlier in this article, and assessed through CLSM and FCM. Meanwhile, the influence on temperature block was also explored by pre-incubating the cells at 4 °C, and the following procedures were performed as described earlier in the article. In Vivo Imaging and Biodistribution. In vivo imaging and biodistribution were evaluated, performed by NIRF imaging on HepG2-bearing female BALB/c nude mice. Following tumor growth (reaching approximately 300 mm3), the mice were iv-injected with

free IR780, I-Lip, FA-modified IR780-loaded liposomes (FI-Lip), or G12-modified IR780-loaded liposomes (GI-Lip, 1 mg/kg of IR780). At each predetermined time interval, the mice were anesthetized via ip injection of 10% chloral hydrate, and observed using a real-time NIRF detector (Caliper Life Sciences, USA). Subsequently, the mice were sacrificed, and the organs as well as the tumors were harvested for further ex vivo evaluation. The nude mice were treated with free SF, S-Lip, FS-Lip, or GS-Lip (10 mg/kg of SF) through iv injection. The mice were sacrificed 24 h post administration, and the organs as well as the tumors were harvested for biodistribution analysis (n = 3). The cumulative amounts of SF in the organs and tumors were determined through high-performance liquid chromatography. In Vivo Evaluation of the Specificity of Diagnosis. H22 tumor-bearing female KM mice were used to evaluate the specificity of diagnosis of GSI-Lip in vivo. The mice were hypodermically injected with 0.1 mL of H22 cell suspensions (1 × 106). Briefly, when a tumor volume reached 100 mm3, the mice were randomly separated into four groups (n = 3), and treated with free IR780, I-Lip, FI-Lip, or GI-Lip (1 mg/kg of IR780), respectively, through intravenous injection. At 2 and 4 h post injection, the animals were anesthetized through intraperitoneal injection of 10% chloral hydrate, and evaluated using the Xenogen IVIS Lumina system as described earlier in this article. In Vivo Early Diagnosis. The ability of GSI-Lip for early diagnosis was evaluated through the different number of H22 tumor23594

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces

Figure 2. Characterization of GSI-Lip in vitro. (a) Particle size, (b) zeta potential, (c) TEM images of GSI-Lip. (d) Release profiles of SF from free SF, SI-Lip, GSI-Lip, and GSI-Lip + laser (n = 3). The stability evaluation of GSI-Lip. (e) Particle size and PDI changes of GSI-Lip at 4 °C for 30 days. (f) Particle size distribution and (g) particle size and PDI change of GSI-Lip in 20% plasma. GSI-Lip were incubated in human plasma for 0, 1, 4, 8, 12, 24, 48 and 72 h. bearing cells and time of tumor growth in a female KM mouse model. Briefly, after 2, 4, and 6 days of cell inoculation (i.e., 1 × 106, 5 × 105, and 2 × 105 H22 cells), four groups of mice (n = 3) were treated with free IR780, I-Lip, FI-Lip, and GI-Lip (1 mg/kg of IR780) by iv injection, respectively. The following procedures were performed as described in the section “In Vivo Evaluation of the Specificity of Diagnosis”. Selection of PTT Conditions for GSI-Lip in Vitro and in Vivo. For the selection of laser irradiation power in vitro, 1 mL of GSI-Lip (10 μg/mL of IR780) was irradiated using the NIR 808 nm laser at 0.2, 0.6, 1.0, or 2.0 W/cm2. To determine the concentration of IR780 in the samples in vitro, 1 mL of GSI-Lip (0.1, 1, 10, and 50 μg/mL of IR780) was irradiated using the NIR 808 nm laser at 0.6 W/cm2. The temperature of all the samples was measured over 10 min. For the NIR laser treatment in vivo, the mice received intravenous administration of NS, free IR780 (1.0 mg/kg IR780), I-Lip, or GSILip (1.0 mg/kg IR780) at 10 mg/kg SF. Twenty-four hours following injection, the animals were anesthetized via intraperitoneal injection of 10% chloral hydrate, and subsequently irradiated using the NIR 808 nm laser at 0.6, 1.0, and 2.0 W/cm2. Infrared thermal imaging was performed using an infrared camera. In Vivo Antitumor Activity. An H22 tumor-bearing model of female KM mice was used to evaluate antitumor efficacy in vivo. Following tumor growth >100 mm3, the animals were randomly separated into 14 groups (n = 6). The mice received iv administration of NS, blank Lip, free SF, S-Lip, SI-Lip, FSI-Lip, or GSI-Lip (10 mg/ kg SF) to determine the antitumor activity of chemotherapy. For chemo-PTT, the mice received iv administration of NS, free IR780 (1.0 mg/kg IR780), I-Lip, or GSI-Lip (1.0 mg/kg IR780 equivalent for liposomes, 10 mg/kg SF). Twenty-four hours post administration, the tumors in the free IR780, I-Lip, and GSI-Lip groups were irradiated with the 1 W/cm2 NIR laser for 5 min. The mice in each group received a total of seven cycles of irradiation (one cycle every 3 days; total treatment duration: 21 days). Body weights, tumor volumes, and tumor weights were recorded at predetermined time points. Immunohistochemical Analysis. Three weeks after treatment, all the mice were sacrificed, and the organs and tumors were dissected for histological analysis. All organs and tumors were fixed in 4% paraformaldehyde. Following dehydration and embedding in paraffin, the specimens were stained using hematoxylin and eosin (H&E). Concerning the evaluation of cell proliferation, the slides were in turn incubated with anti-Ki67 primary antibody as well as Rb IgG (H +

L)/horseradish peroxidase secondary antibody, and the apoptosis of tumor tissues was assessed using the TUNEL assay. In Vitro Hemolysis Assay. Red blood cell suspensions (2%) were diluted using the NS solution. The samples were mixed with the suspensions and incubated at 37 °C. Following the removal of erythrocytes from the suspensions through centrifugation (3000 rpm, 10 min), the absorbance of hemoglobin was measured using a UV−vis spectrophotometer at the wavelength of 576 nm. The hemolysis ratios of the different samples were calculated using the following formula

Hemolysis ratio (HR %) = (A x − A 0)/(A1 − A 0) × 100% Ax, A1, and A0 represent the absorbance of the samples, positive control, and negative control, respectively. Statistical Analysis. Statistically significant differences were analyzed using the paired Student’s t-test, and p < 0.05 were defined to be statistical differences. All data were presented as the mean ± standard deviation (SD).

■

RESULTS AND DISCUSSION Selection of GPC3-Targeting Peptides. HepG2 and HL-7702 cells were selected as GPC3-positive/negative human-derived cell models, respectively, according to the previous studies.33 H22 and liver parenchymal cells isolated from normal mice were used as GPC3-positive/negative mouse-derived cell models, respectively. The expression of GPC3 in different cell lines was determined using the antiGPC3 antibody (FITC-labeled) through CLSM and FCM. As shown in Figure 1a,d, fluorescence was obviously observed in HepG2 and H22 cells, whereas weak fluorescence was observed in HL-7702 and liver parenchymal cells. The quantitative results yielded by FCM showed that the expression level of GPC3 was 60.22 ± 4.86% in HepG2 cells, 54.73 ± 5.27% in H22 cells, 1.56 ± 0.36% in HL-7702 cells, and 1.66 ± 0.17% in liver parenchymal cells (Figure 1b− d,f, respectively). Thus, HepG2 and H22 cells were used as the GPC3-positive cell models, whereas HL-7702 cells and liver parenchymal cells were used as the GPC3-negative cell models. Three GPC3-targeting peptides (i.e., G12, G14, and G7) were preliminarily selected.32,34 The relative specific recognition ability of these three peptides for GPC3-positive and 23595

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces

Figure 3. G12-modified liposomes showed specific GPC3-targeting ability in vitro. Cellular uptake study of HepG2 cells and HL-7702 cells incubated for 0.5 h with free C6, GC6-Lip, and C6-Lip, respectively. (a) Inverted fluorescence microscope images. Scale bar: 20 μm. (b) FCM histogram profiles of fluorescence intensity. (c) Quantification of mean fluorescence intensity by FCM analysis. n = 3, **p < 0.01. (d) CLSM images of GC6-Lip were uptaken into HepG2 cells pre-incubated for 0.5 h with free G12. (e) FCM histogram profiles of fluorescence intensity in HepG2 cells after incubating for 0.5 h with GC6-Lip, when pre-incubated for 0.5 h with/without free G12. (f) FCM analysis of GC6-Lip was uptaken into HepG2 cells pre-incubated for 0.5 h with free G12. n = 3, **p < 0.01.

no aggregation or precipitation observed. Meanwhile, the physical stability of GSI-Lip in 20% plasma was evaluated within 72 h (Figure 2f,g). There were no changes observed in particle size or PDI (p > 0.05) in 20% plasma, indicating that GSI-Lip could remain stable in 20% plasma for 72 h. The cytotoxicity of GSI-Lip in HepG2 cells was studied in vitro for 24 and 48 h using the MTT assay (Figure S5). The IC50 value of GSI-Lip was higher compared with that measured for free SF (p < 0.05) (Table S2), indicating the lower cytotoxicity of GSI-Lip versus free SF. This result was consistent with those reported by other studies.38,39 At 48 h, the IC50 of GSI-Lip (10.51 ± 0.87 μg/mL) was significantly lower compared with that reported for SI-Lip (14.02 ± 0.79 μg/mL) (p < 0.01) (Table S2). This finding indicated that GSI-Lip provides an advantage in improving in vitro cytotoxicity versus SI-Lip. Specific GPC3-Targeting Ability of GSI-Lip in Vitro. C6-labeled liposomes were prepared to investigate the specific GPC3-targeting potential of GSI-Lip. The cellular internalization and localization of the free C6, C6-Lip, and GC6-Lip were examined in HepG2 and HL-7702 cells using an inverted fluorescence microscope and FCM (Figure 3a−c). Compared with C6-Lip, the intracellular fluorescence signals of GC6-Lip were significantly enhanced in HepG2 cells. In contrast, there was no significant difference noted in HL-7702 cells. The mean fluorescence intensity of GC6-Lip in HepG2 cells was much stronger than that reported in HL-7702 cells (Figure 3c) (p < 0.01). In addition, the mean fluorescence intensity of GC6-Lip was much stronger than that measured for C6-Lip (p < 0.01). Collectively, these results demonstrated that G12 could promote cellular uptake in GPC3-positive cells. The cellular uptake of GC6-Lip was significantly decreased in the presence of free G12 (p < 0.01), which could be interpreted as the competitively inhibited binding of free G12 to GPC3 (Figure 3d−f). These results demonstrated that G12-modified liposomes have specific-targeting ability for GPC3-positive cancer cells. Specific G12/GPC3 Recognition-Mediated Endocytosis. The cellular uptake experiments were performed at 4 °C to

-negative cells was evaluated via CLSM and FCM (Figure 1g− l). G12 showed the highest specific recognition ability for HepG2 cells (64.24 ± 6.49%) and H22 cells (55.54 ± 6.47%). Therefore, G12 was selected as the GPC3-targeting peptide for further evaluation. Characterizations of DSPE-PEG2000-G12 and GSI-Lip. DSPE-PEG2000-G12 was synthesized through the thiol group of G12-SH and maleimide of DSPE-PEG2000-Mal by additive reaction.35 DSPE-PEG2000-G12 was characterized by 1H NMR (Figure S3a) and FTIR spectroscopy (Figure S3b). G12 were successfully conjugated to DSPE-PEG2000-Mal, and the synthetic yield was 78.42 ± 0.70%. The characteristics of the S-Lip, SI-Lip, and GSI-Lip are summarized in Table S1. The EE % values of GSI-Lip were 92.44 ± 1.60% for SF and 97.58 ± 1.60% for IR780. The maximum absorption wavelength of free SF in water was 235 nm, which apparently shifted to 265 nm when SF was embedded in GSI-Lip (Figure S4b). The maximum absorption wavelength of free IR780 in water at 774 nm apparently redshifted to 800 nm when IR780 was embedded in GSI-Lip (Figure S4c). The red-shifting for maximum absorption wavelength of SF and IR780 may be due to the change of their residence environment (i.e., from polar environment to a nonpolar hybrid matrix environment).36,37 The results of the UV−vis spectrophotometric analyses and EE % values of GSILip indicated that SF and IR780 are effectively loaded into liposomes. The particle size, zeta potential, and TEM image of GSI-Lip are shown in Figure 2a−c, respectively. The in vitro release profile of SF from GSI-Lip was similar to that observed for SI-Lip (p > 0.05), indicating that the decoration of G12 did not affect the release profile of SF (Figure 2d). The cumulative release in the GSI-Lip + laser group was significantly increased versus the GSI-Lip group (p < 0.01, Figure 2d). The results indicated that PTT accelerated the release of SF, enhancing the synergistic effect of chemo-PTT. The storage stability of GSILip was evaluated over a period of 3 weeks at 4 °C. As depicted in Figure 2e, there were no appreciable changes in particle size (p > 0.05) and PDI (p > 0.05) detected. In addition, there was 23596

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces

Figure 4. Specific G12/GPC3 recognition induced the caveolae-mediated endocytosis of GC6-Lip. (a) CLSM images of GC6-Lip and C6-Lip uptake into HepG2 cells at 37 °C pre-incubated with chlorpromazine (10 μg/mL), genistein (1 μg/mL), cytochalasin D (30 mM) for 0.5 h and at 4 °C pre-incubated at 4 °C for 0.5 h. Scale bar: 50 μm. (b) FCM histogram profiles of fluorescence intensity. (c) FCM analysis of GC6-Lip and C6Lip uptake into HepG2 cells with different pre-incubations. n = 3, **p < 0.01. ##p < 0.01.

investigate whether the cellular uptake of the designed system depends on an energy-dependent process38,40,41 (Figures 4, and S6). The quantitative results of the FCM (Figure 4e) showed that the uptake of GC6-Lip and C6-Lip in HepG2 cells was reduced by 53 and 46%, respectively, versus that reported in the control group (37 °C) (p < 0.01). The results indicated that the endocytosis of both GC6-Lip and C6-Lip is mediated via energy-dependent and energy-independent processes. Following pre-incubation with chlorpromazine, cytochalasin D, and genistein, the uptake of GC6-Lip in HepG2 cells was inhibited by 23, 26, and 29%, respectively, (Figures 4c and S6b). For C6-Lip, the cell uptake into HepG2 cells was significantly inhibited by 30 and 40% after treatment with chlorpromazine and cytochalasin D, respectively. Notably, there was no obvious effect of genistein on the cellular uptake of C6-Lip. Chlorpromazine functions by blocking the clathrindependent uptake.42 In addition, cytochalasin D inhibited macropinocytosis and phagocytosis.43 Genistein has been shown to inhibit the caveolae-mediated endocytosis.44 These results indicated that clathrin-mediated endocytosis, macropinocytosis, and caveolae-mediated endocytosis were involved in the internalization process of GC6-Lip in HepG2 cells.

Moreover, micropinocytosis and clathrin-mediated endocytosis were involved in C6-Lip in HepG2 cells. Collectively, the results of the endocytic pathway experiments suggested that the internalization of both GC6-Lip and C6-Lip was mediated by more than one mechanism in HepG2 cells. For C6-Lip, micropinocytosis was the major pathway for cellular uptake. For GC6-Lip, caveolae-mediated endocytosis was the main internalization pattern. Considering the observed differences in the endocytic pathway between the GC6-Lip and C6-Lip, it could be inferred that the specific G12/GPC3 recognition may induce caveolae-mediated endocytosis.45 Specific GPC3-Targeting Effect of GSI-Lip in Vivo. The tumor-targeting ability and distribution of GI-Lip were investigated through NIRF imaging in vivo in HepG2-bearing nude mice (Figure 5). Various cancer cells (e.g., HCC, breast, uterus, ovarian, and head/neck carcinomas) overexpress folate receptors.46−48 Hence, FA is considered a targeting agent for cancer therapies and diagnostics. In this study, FI-Lip was designed as a comparison group. The characterizations of DSPE-PEG2000-FA and FI-Lip are shown in Figures S7 and S8, respectively. The in vivo NIRF images are shown in Figure 5a. Ex vivo imaging and the quantitation of fluorescence of main 23597

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces

Figure 5. GPC3-targeted GSI-Lip showed specific accumulation in the tumor site in vivo. (a) In vivo imaging of the HepG2-bearing female BALB/ c nude mice after iv administration of free IR780, I-Lip, FI-Lip and GI-Lip at 2, 4, 8, 12, and 24 h. Tumors are marked with red circles. (b) Ex vivo imaging of main organs and tumor after the mice were sacrificed at 24 h post administration. (c) Average radiant efficiency in main organs and tumor based on the ex vivo results. **p < 0.01 compared to free IR780 group, #p < 0.05, $p < 0.05. (d) Distribution of SF in main organs and tumor after iv administration of free SF, S-Lip, FS-Lip and GS-Lip to mice at 24 h. n = 3, *p < 0.05 compared with FS-Lip, **p < 0.01 compared with FS-Lip, #p < 0.05 compared with FS-Lip, ##p < 0.01 compared with FS-Lip.

Figure 6. GSI-Lip exhibited diagnosis specificity in vivo. (a) In vivo imaging of the H22-bearing KM mice after iv administration of free IR780, ILip, FI-Lip, and GI-Lip at 2 and 4 h, respectively (tumor volume ≥ 100 mm3). Tumors are marked with red circles. (b) Ex vivo imaging of tumor tissues after the mice were sacrificed. Scale bar: 5 mm. (c) Total radiant efficiency in tumor tissues. (d) Average radiant efficiency in tumor tissues. n = 3, *p < 0.05, **p < 0.01 compare to free IR780 group. #p < 0.05, ##p < 0.01.

23598

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces

Figure 7. GI-Lip showed high sensitivity and specificity for early diagnosis of HCC. The accuracy of early diagnosis for GSI-Lip was evaluated in detail by the tumor-bearing cells number (1 × 106, 5 × 105 and 2 × 105 H22 cells) and the tumor growth time (2, 4 and 6 days after cells’ incubation), respectively. (a) In vivo imaging, (b) ex vivo imaging of tumor (1 × 106 H22 cells). (c) Total radiant efficiency in tumor tissues based on the ex vivo results. (d) In vivo imaging, (e) ex vivo imaging of tumor (5 × 105 H22 cells). (f) Total radiant efficiency in tumor tissues based on the ex vivo results. (g) In vivo imaging, (h) ex vivo imaging of tumor (2 × 105 H22 cells). (i) Total radiant efficiency in tumor tissues based on the ex vivo results. Tumors are marked with red circles. Scale bar: 5 mm. n = 3, *p < 0.05, **p < 0.01 compared to the free IR780 group. #p < 0.05, ##p < 0.01 compared to the FI-Lip group.

organs and tumors are shown in Figure 5b,c, respectively. Compared with the I-Lip group, the accumulation of IR780 in tumors in the FI-Lip (p < 0.05) and GI-Lip (p < 0.05) groups was promoted, indicating the tumor-targeting ability of FI-Lip and GI-Lip. Notably, the accumulation of IR780 in tumors in

the GI-Lip group was promoted versus that observed in the FILip group (p < 0.05), indicating the excellent in vivo tumortargeting efficacy of GI-Lip. The content of SF in major internal organs was determined 24 h post injection in HepG2bearing nude mice to directly investigate the in vivo 23599

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces

Figure 8. GSI-Lip showed a higher antitumor efficacy in comparison with FSI-Lip in vivo. (a) Tumor growth curves; (b) body weight; (c) photograph of tumors; (d) tumor weight from H22-bearing mice treated with NS, blank Lip, free SF, S-Lip, SI-Lip, FSI-Lip, and GSI-Lip (10 mg/ kg of SF) via the tail vein. **p < 0.01 compared with NS; ##p < 0.01 compared with FSI-Lip. n = 6; (e) H&E, Ki67, TUNEL of tumor tissues after treatment with each treatment group. Magnification: H&E 200×, Ki67 200×, the scale bars of TUNEL represent 200 μm.

between FI-Lip, I-Lip, and free IR780 (p > 0.05). The results revealed that GI-Lip could detect an early tumor 2 days after inoculation with 1 × 106 H22 cells. In contrast, FI-Lip and ILip could not detect the tumor. On day 4, the fluorescence intensity of FI-Lip began higher than that measured for free IR780 (p < 0.01), indicating that FI-Lip plays a diagnostic role 4 days after inoculation with 1 × 106 H22 cells. Following the injection of 5 × 105 H22 cells in KM mice, in vivo NIRF imaging was performed on day 2 (3.45 ± 0.98 mm3), day 4 (13.03 ± 3.47 mm3), and day 6 (20.68 ± 6.49 mm3). The results of in vivo imaging, ex vivo imaging, and quantitation of fluorescence are shown in Figure 7d−f, respectively. On day 2, the total radiant efficiency of GI-Lip in tumor tissues was higher than those observed for FI-Lip (p < 0.05), I-Lip (p < 0.05), and free IR780 (p < 0.01). On days 4 and 6, the total radiant efficiency showed statistically significant differences between GI-Lip and FI-Lip (p < 0.01), I-Lip (p < 0.01), and free IR780 (p < 0.01). However, there were no significant differences noted between FI-Lip, I-Lip, and free IR780 (p > 0.05) on days 2, 4, and 6. These results indicated that GI-Lip could diagnose an early tumor 2 days after injection of 5 × 105 H22 cells (3.45 ± 0.98 mm3). However, FI-Lip and I-Lip could not diagnose the tumor even on day 6. Following the injection of 2 × 105 H22 cells in KM mice, in vivo NIRF imaging was performed on day 2 (2.50 ± 0.99 mm3), day 4 (5.42 ± 2.33 mm3), and day 6 (12.39 ± 6.63 mm) (Figure 7g−i, respectively). On day 2, there was no difference observed between GI-Lip, FI-Lip, I-Lip, and free IR780 (p > 0.05). On day 4, the total radiant efficiency of GILip in tumor tissues was higher than those measured for FI-Lip (p < 0.05), I-Lip (p < 0.05), and free IR780 (p < 0.05). There was no obvious difference displayed between FI-Lip, I-Lip, and free IR780 (p > 0.05) on days 2, 4, and 6. These results indicated that GI-Lip could diagnose an early tumor on day 4 (5.42 ± 2.33 mm3) after the injection of 2 × 105 H22 cells. However, it could not diagnose the tumor on day 2 (2.50 ± 0.99 mm3). Notably, FI-Lip and I-Lip could not diagnose the tumor on day 6.

distribution of GS-Lip (Figure 5d). The accumulation of SF in tumors in the GS-Lip group was significantly higher than that reported in the FS-Lip group (p < 0.01). These results demonstrated the specific tumor-targeting ability of GPC3targeted GSI-Lip in vivo. Diagnostic Specificity of GSI-Lip in Vivo. The diagnostic specificity of GSI-Lip was evaluated through NIRF imaging in H22-bearing KM mice, when the tumor volume was ≥100 mm3 (Figure 6). The in vivo and ex vivo imaging analyses are shown in Figure 6a,b, respectively. The total radiant efficiency in tumor tissues in the GI-Lip group was significantly higher than that measured in the FI-Lip group (p < 0.01, Figure 6c). In terms of the average radiant efficiency in tumor tissues, the GI-Lip group yielded a higher value versus the FI-Lip group (p < 0.05). These results demonstrated that when the tumor volume is ≥100 mm3, the tumor is detectable through I-Lip, FI-Lip, and GI-Lip. GI-Lip showed the highest diagnostic specificity in vivo versus I-Lip and FI-Lip (p < 0.05). These results indicated that GI-Lip was highly specific for HCC. Sensitivity of GSI-Lip for the Early Diagnosis of HCC in Vivo. Whether the GPC3-targeting liposomes are capable of promoting early diagnosis and the accuracy of early diagnosis is rarely involved. In this study, the accuracy of GSI-Lip for the early diagnosis of HCC was evaluated according to the number of tumor-bearing cells (i.e., 1 × 106, 5 × 105, and 2 × 105 H22 cells) and the time of tumor growth (i.e., 2, 4, and 6 days after incubation) using a female KM mouse model. Following the injection of 1 × 106 H22 cells in KM mice, in vivo NIRF imaging was performed on day 2 (14.48 ± 4.10 mm3), day 4 (32.90 ± 10.01 mm3), and day 6 (58.54 ± 10.87 mm3). The results of in vivo imaging, ex vivo imaging, and quantitation of fluorescence are shown in Figure 7a−c, respectively. On day 2, the total radiant efficiency of GI-Lip in tumor tissues was significantly higher than those observed for FI-Lip (p < 0.01), I-Lip (p < 0.01), and free IR780 (p < 0.01). However, there were no significant differences noted 23600

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces

Figure 9. Chemo-photothermal synergistic therapy (GSI-Lip + laser) exhibited superior antitumor activity. (a) Temperature changes of free IR780, I-Lip, and GSI-Lip with 808 nm laser irradiation. (b) Cell viability of H22 cells treated with free IR780, I-Lip, and GSI-Lip with or without laser. *p < 0.05, **p < 0.01. (c) Tumor growth curves of tumor-bearing mice treated with NS, free IR780, I-Lip, GSI-Lip, free IR780 + laser, I-Lip + laser, and GSI-Lip + laser. (d) Body weight changes. **p < 0.01 compared with NS; ##p < 0.01 compared with without laser. n = 6. (e) Tumor weight changes; (f) photograph of tumors from each group. (g) H&E, Ki67, TUNEL of tumor tissues after treatment with each treatment group. Magnification: H&E 200×, Ki67 200×, the scale bars of TUNEL represent 200 μm.

This finding demonstrated that GPC3-targeted liposomes exerted a better specific-targeting effect versus FA receptortargeted liposomes. Photographs of tumors (Figure 8c) and tumor weights (Figure 8d) were in accordance with tumor volumes. As shown in Figure 8b, there was no significant change in weight in the groups during the experiment (p > 0.05), indicating the low systemic toxicity of GSI-Lip. Cell proliferation and apoptosis of tumors after treatment were analyzed through H&E staining, Ki67, and the TUNEL assay (Figures 8e and S11). The tumor treated with GSI-Lip showed the highest rate of apoptosis and lowest rate of proliferation. Collectively, GSI-Lip exhibited superior antitumor efficacy versus FSI-Lip. This may be attributed to the specific-targeting ability in tumor tissues. G12 enhanced the specific internalization of the liposomes into GPC3-positive tumor cells via G12/GPC3-specific recognition. Therefore, those results suggested that specific GPC3-targeting endows the G12modified liposomes with a precise antitumor effect. Precise Chemo-PTT Synergistic Antitumor Therapy. In order to achieve the best synergistic antitumor therapy, the PTT conditions of GSI-Lip were selected in vitro and in vivo.

Collectively, GI-Lip showed high sensitivity and specificity for the early diagnosis of HCC. GI-Lip could clearly detect early tumors on day 2 after injection of 1 × 106 H22 cells (14.48 ± 4.10 mm3), on day 2 after injection of 5 × 105 H22 cells (3.45 ± 0.98 mm3), and day 4 after injection of 2 × 105 H22 cells (5.42 ± 2.33 mm3). FI-Lip showed an early diagnostic role on day 4 after injection of 1 × 106 H22 cells (32.90 ± 10.01 mm3). I-Lip and free IR780 could not detect tumor even on day 6 after injection of 1 × 106 H22 cells (58.54 ± 10.87 mm3). Collectively, these results demonstrated the higher diagnostic sensitivity and specificity of GPC3-targeted liposomes versus FA-modified liposomes, and the great potential of GPC3-targeting liposomes for the early diagnosis of HCC. Precise Antitumor Therapy with GSI-Lip in Vivo. As shown in Figure 8a, blank Lip exhibited an equivalent tumor growth tendency versus the NS group (p > 0.05). FSI-Lip and GSI-Lip showed significantly better antitumor effects compared with those observed for S-Lip and SI-Lip (p < 0.01). GSI-Lip displayed superior antitumor efficacy versus FSI-Lip (p < 0.01), with a 90.52% rate of tumor inhibition (Table S4). 23601

DOI: 10.1021/acsami.9b05526 ACS Appl. Mater. Interfaces 2019, 11, 23591−23604

Research Article

ACS Applied Materials & Interfaces As shown in Figure S9, 0.6 W/cm2 laser was chosen as laser irradiation power in vitro. GSI-Lip with 10 μg/mL IR780 was chosen for in vitro evaluation. Five minutes was chosen for the laser irradiation time. The PTT efficiency of GSI-Lip in vitro was evaluated through changes in temperature under laser irradiation (Figure 9a). The increase in temperature observed with I-Lip (48.00 ± 1.00 °C) and GSI-Lip (47.67 ± 1.15 °C) was higher than that reported for free IR780 (40.00 ± 1.00 °C, p < 0.01). This was attributed to the red shift in light absorption of IR780 loaded onto the lipid layer (Figure S4c), leading to the enhanced activation with the 808 nm NIR laser.37 The noted increase in temperature associated with the liposomes may result in irreversible damage to tumor cells.49 The PTT cytotoxicity of GSI-Lip was investigated using the MTT assay (0.6 W/cm2 808 nm laser, 5 min) (Figure 9b). Following irradiation with the laser, GSI-Lip exhibited superior inhibition of growth versus I-Lip (p < 0.01) and free IR780 (p < 0.05). The IC50 values are shown in Table S3. In summary, the results of PTT cytotoxicity in vitro further confirmed that, following irradiation with the laser, GSI-Lip promoted the antiproliferative activity. The synergistic antitumor efficacy of chemo-PTT was evaluated in H22-bearing KM mice. As Figure S10 shows, 808 nm laser irradiation at 1.0 W/cm2 could lead to a sufficient increase in temperature for GSI-Lip (the highest temperature for GSI-Lip was 43.33 ± 0.31 °C in 5 min), and avoid damage to normal cells caused by an excessive increase in temperature. Thus, the 1.0 W/cm2 808 nm laser was selected for in vivo PTT. During the course of treatment, laser irradiation was administered only once. Two days after irradiation, a clear hemorrhagic injury and severe necrotic tissue could be observed at the irradiation site in the mice treated with free IR780, I-Lip, and GSI-Lip. The tumor volume in the free IR780 group and I-Lip group did not show a significant decrease versus the NS group, suggesting that free IR780 and I-Lip have no therapeutic effect without laser radiation. All tumors treated with GSI-Lip were significantly inhibited over a period of 21 days. Notably, GSI-Lip plus irradiation with the laser exhibited superior antitumor activity versus GSI-Lip without laser (p < 0.01), and a rate of tumor inhibition of 94.93% (Table S4). There was no significant change in weight after PTT (Figure 9d), suggesting that the PTT (1 W/cm2 laser) was associated with good safety. As expected, the combination of chemo-PTT (GSI-Lip + laser) showed the lowest rate of tumor-cell proliferation (Figures 9g and S11a) and the highest level of tumor necrosis (Figures 9g and S11b). In conclusion, GPC3-targeted liposomes enhanced the targeted delivery of SF and PTT agents IR780 into GPC3positive HCC cells. Therefore, the combination of the chemotherapeutic drug SF and PTT agent IR780 enhanced the precision of the synergistic antitumor effect against HCC. The hemolytic properties were also evaluated (Figure S12). There was no hemolysis observed, even at the highest concentration of GSI-Lip (Figure S12a). Meanwhile, the hemolytic activity of GSI-Lip at concentrations of 12−60 μg/ mL was negligible (