Trials in developing a nanoscale material for extravascular contrast-enhanced ultrasound targeting hepatocellular carcinoma

Ethics statement

All the experimental procedures in this study were in compliance with the National Institutes of Health guidelines and were approved by the Institutional Animal Care and Use Committee of Hainan Medical University (2018-02-27).

Preparation of liposomes and liposome microbubbles

Sulfur hexafluoride gas (SF₆) filled liposome microbubbles and conjugated with synthesized peptide (LSPMbs), SF₆ filled liposome microbubbles and conjugated with GPC3 antibody (LSGMbs), SF₆ filled liposome microbubbles and conjugated with synthesized peptide, and liposomes (LS) were prepared and assessed in the public scientific research center of Hainan medical university. To develop a GPC3 targeted liposome microbubbles, materials and formula were chosen and optimized, and the optimum formula was selected through orthogonal design test according to the enhanced ultrasound effect. 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), 1,2-Dipalmitoyl-sn-glycero-3-phospho-sn-1-glycerol (DPPG), 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[carboxy(polyethylene glycol)-2000] (DSPE-PEG2000) , GPC3, artificial synthesized peptide of DHLASLWWGTEL and SF₆ were adopted and fabricated using following protocols. 12  mg DPPC was dissolved in 3.0  mL chloroform, followed adding 1.5  mL methanol and 0.5 mL deionized water to form a mixture, next dissolved 23 mg DPPG and 11 mg DSPE-PEG2000 in the mixture. The solvent was removed by rotary evaporation and vacuum using a Vacuum Rotary Evaporator (Xian Depai17 Biotechnology Co., Ltd. Xian, China), and the mixture formed a thin film on the wall of the container. 20 mL deionized water was added to the container to harvest and suspend the lipid mixture, next it was transferred to a tube to undergo magnetic stirring at 65 °C for 20 min, followed by probe-sonication (Vibracell™, VCX130PB/VCX130 Sonics, Sonics and Materials, Inc., Newtown, USA) with a frequency of 40 kHz under 35 °C f or 5 min. A mechanical blender (Ultra-turrax T25, Janke & Kunkel, IKA-Labortechnik, Staufen, Germany) was used for liposome microbubbles preparation. The above processed lipid dispersion was put into a 50  mL Falcon tube, the air above the aqueous dispersion in the tube was replaced with SF₆ gas, and the tube was sealed with parafilm. The temperature of lipid dispersion was heated to 30 °C , the homogenizer was operated to create high shear mixing (15,000 rpm, 5 min) to form microbubbles. The mixture was centrifuged at 12,000 rpm at 4 °C for 5 min, washed with deionized water, three times. 15% (w/v) sucrose solution was added to the mixture in a 5:1 volume ratio (mixture: sucrose), small glass vial of four  mL volume was used for the loading, each with two  mL mixture. The air in the vials was replaced with SF₆ gas before lyophilized in a −85 °C lyophilizer (SP Scientific, VirTis, USA). After 24 h completely freeze-drying, vials were refilled with SF₆ gas and sealed, stored at 4 °C . Pure liposomes (LS) were prepared using the above protocols without filling SF₆ gas.

Preparation of GPC3 targeted liposome microbubbles

Firstly, after liposome microbubbles (LSMbs) were prepared using the protocols above, but did not add sucrose solution and lyophilize. Next, a carbodiimide method was used for covalent conjugation of the synthesized peptide to the free carboxyl groups on the surface of LSMbs. The prepared LSMbs were resuspended in MES buffer (0.1 mol/L, pH 6.0), and an adequate amount of EDC/NHS [1-ethyl-3(3-dimethylaminopropyl) carbodiimide (EDC): N-hydroxy succinimide (NHS) = 1:4, Sigma-Aldrich Chemical Co., Inc, USA] were added into the suspension. The mixture suspension was oscillated and incubated for 2 h at room temperature (25 °C). The remaining EDC/NHS was removed by three-time centrifugation at 12,000 rpm using MES (pH 6.0), 5 min each time. The precipitate was dispersed into MES buffer (0.1 mol/L, pH 8.0), and an adequate amount of synthesized peptide was added and incubated with stirring for 2 h at room temperature. The peptide was compounded by GL Biochem (Shanghai) Ltd. (Shanghai, China) in accordance to a 12-mer peptide with the sequence of DHLASLWWGTEL reported from previous study that it can target GPC3 of HepG2 (Zhu et al., 2016a) (DOI: 10.1021/acs.bioconjchem.6b00030). The mixture was centrifuged at 12,000 rpm at 4 °C for 5 min, washed by deionized water, three times. Next, 15% (w/v) sucrose solution was added to the mixture in a 5:1 volume ratio (mixture: sucrose), small glass vial of four  mL volume was used for the loading, each with two  mL mixture. The air in the vials was replaced with SF₆ gas before lyophilized in a −85 °C lyophilizer (SP Scientific, VirTis, USA). After 24 h completely freeze-drying, vials were refilled with SF₆ gas and sealed, stored at 4 °C. Liposome microbubbles (LSMbs), and GPC3 antibody (BM1846; Wuhan Boster Biological Technology, Ltd., Wuhan, China) conjugated liposome microbubbles (LSGMbs) were prepared in the similar protocol above.

Characterization of LSPMbs

LSPMbs Size and zeta potential measurements

LSPMbs size and Zeta potential of each sample was measured using a Zetasizer Nano S90 (Malvern Instruments Ltd., Malvern, UK) by Laser Doppler Anemometry (LDA) using electropheoretic light scattering at 25 °C . An adequate amount of LSPMbs was suspended and diluted to enable the microbubbles concentration was maintained to ensure that multiple scattering and microbubble–microbubble interactions were negligible. The microbubble size and zeta potential of each sample were measured three times, and the mean value was taken as the final microbubble size and zeta potential. LSGMbs were assessed in the same methods.

Assessment of biocompatibility of LSPMbs

Evaluation of enhanced ultrasound imaging of LSPMbs in vitro

To assess the contrast-enhanced effect of LSPMbs, LSPMbs were suspended with deionized water and diluted into various concentrations (1.6, 0.8, and 0.4  mg/mL) and placed in different plastic tubes for ultrasound evaluation. LSPMbs suspension filled tubes were fixed in a water container, and their CEU effect was assessed using GE Logiq E9 ultrasound system (GE Healthcare, Milwaukee, WI, USA), using a  mL6-15-D linear transducer with a frequency of 4–15 MHz. During the ultrasound performance, the frequency of the transducer was set to 12 MHz, the depth and focus were adjusted to optimize imaging, and the model was shifted to contrast imaging, using the default parameter (MI 0.1). Before the ultrasound scanning, the tubes were agitated slightly. Controls of commercial ultrasound contrast agent SonoVue (Shanghai Bracco Sine Pharmaceutical Corp. Ltd., Shanghai, China), degas deionized water, and air were established, and these were assessed using the same protocols above.

Enhancement ultrasound imaging effects of LSPMbs, LSGMbs, SonoVue, LS, degas deionized water, and air were determined using Photoshop software (Adobe Photoshop CS3, Adobe Systems Inc, CA, USA). To analyze ultrasound images, observers opened the image in Photoshop, activated menus of “Window, Information, and Histogram” consecutively, selected “rectangle, and statistics display ” tools, set the same region of interest to the tube, parameters yielded automatically, measured three times in each image, and adopted the mean value of scales (arbitrary units) as the result of a single image.

Assessment of affinity of LSPMbs to the liver cancer cells

Fluorescence experiment was used for the assessment of affinity of LSPMbs to the liver cancer cells. A tiny amount of 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI) (Yeasen Biotech Co., Ltd., Shanghai, China) was added into five  mL of 16  mg LSPMbs suspension to form DiI labelled LSPMbs suspension. Huh-7 cells (Cell bank of the Chinese academy of sciences (Shanghai, China)) were seeded in a six-well plate with one glass slip placed in each well at a concentration of 3 ×10⁴ cells/well. Next day, the Huh-7 cells were fixed by 90% cold ethanol for 20 min and blocked by 10% bovine serum albumin (BSA) at 37 °C for 1 h and subsequently incubated with 4 drops of DiI labelled LSPMbs suspension for 3 h in dark place. Next, the wells and slips were washed with PBS three times, and the cells on the slips were mounted with 4′, 6-diamidino-2-phenylindole (DAPI, Wuhan Boster Biological Technology, Ltd., China) for nuclei visualization and detected using a laser confocal microscope (Fluoview FV 10001000, Olympus, Japan). Images of bright, DAPI staining, DiI staining, and merged were obtained. LSGMbs were assessed in the same protocols for control. To confirm the specificity of binding of DiI labelled LSPMbs to GPC3 in the Huh-7 cells, 0.1 mL synthesized peptide (10 µg/mL) was applied before adding DiI labelled LSPMbs as a blocking control, and the other protocols were the same as the above.

Assessment of LSPMbs in vivo

All animal experimental procedures were approved by the Hainan Medical University Association for Accreditation of Laboratory Animal Care. Animals of female BALB/C mice were used for the evaluation of targeting ability and contrast-enhanced effect. Twenty Huh-7 cell xenograft tumor models of female BALB/C mice were established, and the Huh-7 cell line was acquired from the cell bank of the Chinese academy of sciences (Shanghai, China). All animals expect four mice were sacrificed by euthanasia using isoflurane after fluorescent imaging were sacrificed using carbon dioxide in a closed box at the end of the animal experiments. The criteria of animal death are that the mice were in collapse state, losing muscular tension, no breath, no heart beating, and the skin color became gray.

Enhanced ultrasound imaging assessment of LSPMbs

Fifteen of the mice were used for the targeted CEU experiments in five groups, with each group of three mice. The CEU effect was assessed using GE Logiq E9 ultrasound system (as has addressed in the previous section). During the ultrasound scanning, the frequency of the transducer was set to 12 MHz, the depth and focus were adjusted to optimize imaging, and the model was shifted to contrast imaging, using the default parameter (MI 0.1). The shape of the xenograft tumors was ovoid, and the longitudinal diameter of the tumors of 24 mice was 10.4 ± 0.53 mm in the fifth week after cells seeded. Control experiments were conducted in four groups using LSGMbs, LSPMbs (GPC3 antibody or synthesized peptide blocked previously), SonoVue (a commercial ultrasound contrast-enhanced agent), and LS, respectively. LSPMbs suspension was prepared using 14 mg LSPMbs and four mL 0.9% sodium chloride solution, agitated slightly before injection. The animals with Huh-7 xenograft tumor in five groups were intravenously injected suspension of LSPMbs, LSGMbs, LSPMbs (injected 0.1 mL synthesized peptide or 0.1 mL dilated GPC3 antibody for blocking in 3 min ahead), SonoVue (15  mg SonoVue in four  mL 0.9% sodium chloride solution), and LS (14  mg LS and four  mL 0.9% sodium chloride solution, agitated slightly before injection), respectively, each with the volume of 0.2  mL. The contrast imaging time was counted since the bolus intravenous injection of 0.2 mL LSPMbs suspension via the mouse tail vein. The images were saved in the ultrasound system and exported for later study.

Enhancement ultrasound imaging effects of LSPMbs, LSGMbs after synthesized peptide or GPC3 antibody blocking, SonoVue, LS were determined using Photoshop software, and the methods had been addressed in the previous section.

Fluorescent imaging assessment of LSPMbs

Ten mice with xenograft tumors of Huh-7 cells allotted to two groups of five each were used to conduct Cy 5.5 fluorescence experiment to test the specificity and affinity of the peptide in LSPMbs to GPC3 of the liver cancer. An IVIS Lumina image system (Xenogen) (IVIS^® Lumina XR) (Caliper life sciences) was used for the evaluation. During the fluorescence imaging, mice were under gas anesthesia with oxygen and isoflurane (Jinan Shengqi pharmaceutical Co, Ltd., Jinan, China). 0.2 mL Cy5.5 labelled LSPMbs suspension was intravenously injected into six mice with Huh-7 xenograft tumor via the tail vein, images were acquired at one minute, six hours, and 24 h. Cy5.5 labelled LSPMbs suspension was prepared using 14 mg Cy5.5 labelled LSPMbs and four  mL 0.9% sodium chloride solution, agitated slightly before injection. The control experiments were conducted in five mice with Huh-7 xenograft tumors, with the same methods after injection of 0.1 mL synthesized peptide (10 µg/mL) for blocking in three minutes. Four mice were sacrificed by euthanasia using 3 mL isoflurane in a closed small box. The tumor, liver, heart, lung, kidney, and spleen of the mice were isolated for fluorescent imaging assessment at 24 h.

Statistical analysis

Quantitative data are presented as mean ± SD (standard deviation) , and qualitative data are presented as percentile. Statistical significance of differences between groups of quantitative variables were analyzed using paired-sample t tests or univariate analysis of variance, and qualitative variables were analyzed using Chi-square test. All statistical analyses were performed using SPSS software (Version 20; IBM, Armonk, NY, USA). P < 0.05 was considered significant.

Characterization of LSPMbs

The LSPMbs appeared round shaped on a sectional view, with different size, without aggregation, identified by transmission electron microscopy (Fig. 1).

Size and Zeta potential of LSPMbs were 380.9 ± 176.5 nm and −51.4 ± 10.4 mV, respectively. The determination results showed “Good” (Figs. 2 and 3).

Assessment of biocompatibility of LSPMbs

MTT assay for cytotoxicity

The cell index had no significant differences among different LSPMbs concentrations at different time (all P >  0. 05), indicating that LSPMbs did not cause significant toxic effect on RAW 264.7 cells. As shown on Fig. 4.

Evaluation of enhanced ultrasound imaging of LSPMbs in vitro

LSPMbs and SonoVue suspension with different concentrations presented different enhancement effects, they presented almost the same enhancement effect at the same concentration (Figs. 5A and 5D were obtained from 1.6 mg/mL, Figs. 5B and 5E was obtained from 0.8 mg/mL, and Figs. 5C and 5F was obtained from 0.4 mg/mL). At higher concentration (1.6 mg/mL), they all presented homogeneous hyperechogenicity with marked attenuation (Figs. 5A and 5D), and the echogenicities became weaker when the concentrations decreased (Figs. 5B and 5E, and Figs. 5C and 5F), and they all much stronger than control of degas deionized water, which presented homogeneous anechogenicity (Fig. 5G). Control of air presented strong echogenicity at the interface of the tube, and the distal field presented marked attenuation (Fig. 5H), which was substantial different from those obtained from LSPMbs and SonoVue suspension.

Comparisons of measurements of enhanced ultrasound imaging between different concentrations and agents were as follow: (A) vs. D, P = 0.321; (B) vs. E, P = 0.472; (C) vs. F, P = 0.428; and (A) vs. G, A vs. H, B vs. H, G vs. H, A vs. B, A vs. C, B vs. C, D vs. E, and D vs. F, all P < 0.05.

Assessment of affinity of LSPMbs to the liver cancer cells

The fluorescence on the cellular membrane of Huh-7 cells was intensive (Fig. 6), indicating that there was high GPC3 expression. Cells and DiI labled LSPMbs and controls obtained from light microscope (Figs. 7A, 7E and 7I). Cell nucleus of Huh-7 cells presented blue after DAPI staining and being incubated with DiI labled LSPMbs (Figs. 7B, 7F and 7J). LSPMbs with DiI staining combined with the membrane of Huh-7 cells presented red color fluorescence on image (Figs. 7C and 7D); LSGMbs presented a very similar appearance (Figs. 7G and 7H); and LSPMbs with DiI staining after synthesized peptide blocking previously did not show red color fluorescence on the membrane of Huh-7 cells, indicating that LSPMbs had not combined with the membrane of Huh-7 cells (Figs. 7K and 7L).

Assessment of LSPMbs in vivo

Enhanced ultrasound imaging assessment of LSPMbs

All xenograft tumors in the mice of the five groups presented a complex of isoechogenicity, hypoechogenicity and anechogenicity (A), E, I, M and Q). Of the study group, after administration of LSPMbs suspension, Huh-7 xenograft tumor presented hyper-enhancement in periphery and hypo-enhancement in center (necrosis) at two seconds (Fig. 8B); the tumor enhancement lasted over 20 s with little change; at 60 s, the tumor still presented hyper-enhancement in periphery and hypo-enhancement in center (necrosis) (Fig. 8C); at 10 min, the tumor presented iso-enhancement with central hypo-enhancement (necrosis) (Fig. 8D). Of the four control groups, the mice injected with LSGMbs, LSPMbs (blocked with GPC3 antibody or synthesized peptide previously), and SonoVue, respectively, presented similar enhancement patterns and sustain time as those in Huh-7 xenograft tumors (Figs. 8F–8H, Figs. 8J–8L, and Figs. 8N–8P), and there was no significant difference; the xenograft tumors presented similar enhancement patterns and sustaining time; of the mice in the group injected with LS suspension, the xenograft tumors did not present enhancement at 2, 20, 60 s, and 10 min (Figs. 8R–8T).

Comparisons of enhanced ultrasound imaging among different agents at times of two seconds, 20 s, and 10 min, scales of images (B, C, and D of LSPMbs, J, K, and L of LSPMbs after synthesized peptide or GPC3 blocking, F, G, and H of LSGMbs, and N, O, and P of SonoVue) all had no significant difference (all P >  0.05), and there were significant difference between the above scales of images and scales of LS images (all P < 0.001). These indicate that LSPMbs has good capability in CEUS imaging, but the experiments of it did not show targeted imaging in vivo.

Fluorescent imaging assessment of LSPMbs

The fluorescent signal could be visualized all over the body soon after the administration of Cy5.5 labelled LSPMbs suspension. Images acquired at one minute (Fig. 9A), six hours (Fig. 9B) after the initial fluorescent imaging, the fluorescent signal intensity in the tumor area has no significant difference from other areas of the body expect the liver and spleen. The fluorescent signal intensity in the liver and spleen area of the mice was marked stronger than other areas, and the fluorescent signal intensity was similar in three times. The fluorescent signal could not be visualized in the mice 24 h after injection (Fig. 9C). Experiment carried out in the mice blocked previously by injection of GPC3 antibody or synthesized peptide presented the same fluorescent imaging characteristics as those in the mice with Huh-7 xenograft tumors (Figs. 9D–9F). 24 h after intravenous injection of Cy5.5 labelled LSPMbs, four mice with Huh-7 cell xenograft tumors of two in each groups were sacrificed, the tumors and visceral organs were assessed, there were fluorescence in the lungs and liver, and no fluorescence in the tumor and the heart, spleen and kidneys (Fig. 10). These indicate that the Cy5.5 labelled LSPMbs did not selective accumulated in the xenograft tumor, and the LSPMbs did not present detectable targeting ability to the tumor. The reason that the fluorescent signal intensity in the liver and spleen was higher than other areas is believed that the liver and spleen have abundant capillaries and macrophage cells, the macrophage cells can engulf the liposomes, so liposomes in these regions are richer than other regions. The more Cy5.5 labelled LSPMbs aggregated, the stronger the fluorescent signal intensity.