Introduction

In the last two decades, sonodynamic therapy (SDT) has been developed as a promising cancer treatment modality, which utilizes low-intensity ultrasound to activate certain sonosensitizers and induces cytotoxicity to cancer cells. The significant advantage of SDT is based on the fact that ultrasound is a type of mechanical wave that can penetrate soft tissues up to a depth of centimeters and then focus on a specific tumor site.16 Especially ultrasound can penetrate skull without craniotomy surgery, making SDT a potential noninvasive approach to treat brain tumors.32

Recently, several in vitro studies showed improved therapeutic effect of SDT on glioma cells. Xu et al. investigated the efficacy of SDT using Photofrin as the sonosensitizer in the glioma stem-like cells.35 It was found that the intracellular Photofrin concentration was significantly increased and the results showed that the anti-tumor effect was greatly improved in vitro by SDT with ultrasonic sonication at 1.0 MHz and 0.5 W/cm2 for 2 min. Another study by Dai et al. explored the therapeutic effect of SDT with the sonosensitizer hematoporphyrin monomethyl (HMME) on C6 glioma cells.5 The results revealed an enhanced glioma cell apoptosis with increased level of reactive oxygen species (ROS) production at the ultrasonic intensity of 1 W/cm2, suggesting that SDT might be a promising method for glioma treatment. However, the in vivo studies of SDT in the treatment of glioma-bearing animals have been reported in a small number.11,17,18,26,27 The limited delivery of sonosensitizers to the brain severely reduced the efficacy of glioma treatment in vivo and it is mainly related to the brain-blood barrier (BBB) acting as a nearly impermeable structural barrier.29 Unlike extracranial tumors, the existence of BBB presents a formidable obstacle to the crossing of sonosensitizers to the brain tumor site.19,21 Therefore, enhancing delivery of sonosensitizers across BBB is critical to the therapeutic effect of SDT on brain tumors.

Traditional methods of disrupting the BBB to improve drug delivery such as intracranial local injection and systemic administration of hypertonic solution have a lot of limitations including invasiveness and severe side effects.18 In contrast, low-intensity focused ultrasound (FUS) combined with microbubbles (MBs) has become a promising approach to open the BBB non-invasively, locally and reversibly.9,10 Our previous studies also demonstrated that FUS combined with MBs induced a significantly increased permeability of the BBB.20,23 It is worth mentioning that repeated opening of the BBB by FUS in combination with systemic injection of MBs is safe and well tolerated in clinical trial.3 Therefore, these studies provided support for the idea of using FUS with MBs to enhance the delivery of sonosensitizers in SDT for glioma.

It is generally known that the sonosensitizer is one of the most essential factors in SDT. Recent studies revealed that sinoporphyrin sodium (DVDMS) was a novel sonosensitizer which could be highly activated by ultrasound with characteristics of high chemical purity, water solubility and good singlet oxygen yield.31,34 Furthermore, it has been demonstrated that SDT using DVDMS as the sonosensitizer has strong cytotoxicity effects on several tumor cell lines, such as K562 leukemia cells, esophageal cancer ECA-109 cells, breast cancer 4T1 cells and sarcoma 180 cells.8,30,31,34 However, to the best of our knowledge, there is no report of DVDMS-mediated SDT study on the treatment of glioblastoma and the anti-tumor effect is still unknown.

Based on the fact that low-intensity ultrasound could be utilized either to open BBB or in SDT procedure, we propose a two-step strategy to treat animals bearing intracranial glioblastoma using the same experimental ultrasound apparatus in this study. As shown in Fig. 1, FUS with MBs are used to induce enhanced delivery of the sonosensitizer DVDMS through the BBB to the brain tumor site first, and then the DVDMS-mediated SDT treatment is activated by low-intensity ultrasound. The in vitro and in vivo efficacy of SDT using DVDMS as the sonosensitizer on nude mice bearing intracranial U87 MG-Red-FLuc human glioblastoma was evaluated. Bioluminescence imaging (BLI) was used to longitudinally monitor the tumor growth tendency. The proliferation and apoptosis of the intracranial glioblastoma were also assessed by immunohistochemical examination.

Figure 1
figure 1

Schematic illustration outlining the treatment of intracranial glioblastoma xenografts in nude mice using SDT. Step 1, enhanced delivery and accumulation of DVDMS using low-intensity FUS with circulating MBs. Step 2, SDT w/DVDMS was conducted to kill glioblastoma cells in irradiated regions.

Full size image

Materials and Methods

Chemicals

Sinoporphyrin sodium (DVDMS) was generously provided by Professor Qicheng Fang from the Chinese Academy of Medical Sciences (Beijing, China) and Jiangxi Qinglong Group Co., Ltd (Jiangxi, China). DVDMS was dissolved in phosphate buffer solution (PBS, pH 7.2–7.4, BioScience, Shanghai, China) and then sterilized by a 0.22 μm filter (Jet Bio-Filtration Co., Ltd., Guangzhou, China), stored with light-free at − 20 °C. Its molecular formula is C68H66N8O9Na4 (molecular weight: 1230.265) with purity > 98%. The chemical structure of DVDMS is shown in Fig. 1. All other reagents were analytical grade.

Cell Culture

A luciferase expressing human glioblastoma cell line U87 MG-Red-FLuc was used (PerkinElmer Inc., Waltham, MA, USA) in this study. The cells were cultured in Eagle’s minimum essential medium (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum (GE Healthcare Life Sciences, Victoria, Australia), 1% sodium pyruvate (Thermo Fisher Scientific, Waltham, MA, USA) as an alternative carbon source in cell culture, and 1% non-essential amino acids (Thermo Fisher Scientific, Waltham, MA, USA), in an incubator with 5% CO2 at 37 °C. Cells in the logarithmic growth phase were used.

In Vitro Study of Sonodynamic Therapy on Glioblastoma Cells

The anti-tumor effect of SDT on U87 MG-Red-FLuc cells was first evaluated in vitro. Cells (2 × 105 cells/well) were seeded into 12-well plates which were made of polystyrene and without any protein coating (Corning Inc., NY, USA) and cultured for 24 h. To determine the incubation time, cells were incubated with DVDMS in the dark for 0.5 h, 1, 2, 3, 4 and 6 h after refreshing the culture medium. Three wells were used for measurement at each time point. The concentration of 10 μg/mL was selected in this study, which was consistent with those used in prior studies.8,30,31,34 The cells were washed three times with PBS and harvested. Then the mean fluorescence intensities of these samples were determined using a microplate reader (Synergy H4; BioTek Inc., Winooski, VT, USA). The excitation and emission wavelength were 430 nm and 680 nm, respectively. In this study, the 12-well plates were covered with tinfoil to protect DVDMS from light and the experiments were performed in the dark room.

FUS was generated by a lab-made single-element spherical transducer (center frequency: 0.996 MHz, the lateral and axial full-width at half-maximum intensity of the beam: 5.0 and 65.0 mm). It was immersed within a cone filled with degassed water and sealed with a thin polyurethane membrane. The cone tip was contacted with the bottom of the plate tightly through the coupling gel to allow ultrasonic waves transmitting to the cells with minimal attenuation. A function generator (AFG3102C, Tektronix, Inc., Beaverton, OR, USA) operated in burst mode was used to produce excitation signal and connected with a 50-dB power amplifier (2100L, Electronics & Innovation, Rochester, NY, USA) to drive the FUS transducer. After incubation with DVDMS in the dark for 3 h, U87 MG-Red-FLuc cells were then subjected to sonication for 60 s with 1-Hz pulse repetition frequency (PRF) at 30% duty cycle and acoustic power of 1.7 W. The cells were insonified within the focus of the transducer (Fig. 3a), and the attenuation of ultrasound through the bottom of the 12-well plate was 17.19 ± 3.12% (n = 6, mean ± SD). The cells received four different treatments: no treatment (control), DVDMS alone (DVDMS), FUS sonication alone (SDT w/o DVDMS), FUS sonication with DVDMS (SDT w/o DVDMS).

The apoptosis of U87 MG-Red-FLuc cells undergone different treatments were then evaluated by flow cytometry method. The cells were washed three times with PBS and harvested. The collected cells were resuspended in 500 µL of binding buffer and incubated with 5 µL of Annexin V-fluorescein isothiocyanate and 5 µL of propidium iodide (BD Pharmingen, San Diego, CA, USA) for 15 min at room temperature in the dark. Then, the samples were analyzed using a flow cytometer (BD FACSCalibur, BD Biosciences, Franklin Lakes, NJ, USA).

Preparation and Characterization of MBs

MBs were coated with 1,2-distearoyl-sn-glycero- 3-phosphocholine (DSPC) and N-(carbonyl-methoxypolyethylene glycol-2000)-1,2-distearoyl-sn-glycero-3-phosphoethanolamine (DSPE-PEG2000) (Lipoid, Ludwigshafen, Germany) at a molar ratio of 9:1 with a final lipid concentration of 3 mg/mL. The phospholipids were dissolved in chloroform (Sinopharm Chemical Reagent, Shanghai, China) and the organic solvent was evaporated under vacuum (Rotavapor R210, BUCHI, Postfach, Flawil, Switzerland) overnight. Then the dried phospholipid film was hydrated with PBS mixed with 10 vol% glycerol solution and 10 vol% propylene glycol. The lipid suspension was sonicated in a water bath (SB120DT, Ningbo Scientz Biotechnology, Ningbo, China) and sealed within a 3-mL glass serum vial (SCHOTT, Suzhou, Jiangsu, China), which was then connected to a 3-way valve. Through the valve, the air was removed with a vacuum pump (AP-01P, Tianjin Autoscience Instrument Co., Ltd., Hebei, China) before inflating the headspace with perfluoropropane gas (99.999% purity, Xundong gas technology, Suzhou, China). This air exchange procedure was repeated five times. Before sonication, the microbubbles were generated by agitating the vial mechanically at the speed of 4500 oscillations/min for 20 s using an amalgamator (Zhongrun Medical Instrument, Hangzhou, China).

The concentration and particle size distribution of the prepared MBs was measured by a Coulter Counter Multisizer IV (Beckman Coulter Inc., Miami, FL, USA) after dilution with Isoton II. The structure of MBs was visualized by a microscope (BX-53, Olympus Corporation, Tokyo, Japan).

To investigate the lifetime of MBs in the brain, contrast-enhanced ultrasound imaging was performed on nude mice (n = 3) using a small animal acoustic imaging system (Vevo® LAZR; VisualSonics Inc., Toronto, Canada) with a 40-MHz transducer. To eliminate the effect of the skull, craniotomy surgery was performed on nude mice (approximately 0.5 × 0.5 cm2). Serial images were acquired before and after the injection of MBs. Through the time intensity curves of the images, the lifetime of MBs could be determined.

Delivery of DVDMS Through Opened BBB Induced by FUS with MBs

Each healthy mouse without tumor implantation was placed in the prone position and anesthetized with 1.5% isoflurane (RWD Life Science, Shenzhen, China), whose head was immobilized with a stereotaxic apparatus (RWD life science, Shenzhen, China). The body temperature of each mouse was maintained through a heating pad. The same transducer from the in vitro study with the cone was placed on the top of the right hemisphere. A thin layer of ultrasound coupling gel was used between the polyurethane membrane of the cone tip and mouse scalp skin (Fig. 1). Before sonication, a tail vein catheter was inserted using a 27-Gauge needle. Then 0.2 mL bolus of diluted MBs solution (0.2 μL/g of body weight) was co-injected with DVDMS (2 mg/kg of body weight) or PBS as the control (n = 3) and circulated for about 15 s. Then pulsed FUS was applied for 60 s with a peak-rarefactional pressure of 0.64 MPa, a burst length of 10 ms and a repetition frequency of 1 Hz.22,23

DVDMS accumulation in the brain tissue were examined at different time points, 0.5, 1, 3, 6, 14 and 28 h post BBB opening by FUS with MBs. Three mice were used for each time point. Cardiac perfusions were performed to clear DVDMS in the circulation. The brains were extracted from the skull and fixed in 10% formaldehyde for 48 h. Then the fluorescence intensity of DVDMS accumulation in the brains was acquired and quantitatively analyzed by a Xenogen IVIS spectrum system (IVIS SPECTRUM; PerkinElmer Inc., Waltham, MA, USA) with 430 nm excitation wavelength and a 680 nm emission filter.

To compare the difference between the accumulation of DVDMS that penetrate through BBB compromised by glioma and those that penetrate through BBB disrupted by FUS with MBs, six nude mice bearing ortho-topic glioblastoma xenografts were injected in the striatum with the same amount of U87 MG-Red-FLuc cells. 20 days after the tumor implantation, FUS (pulse length of 10 ms, a PRF of 1 Hz for 60 s, peak-rarefactional pressure of 0.64 MPa) was applied to the right glioblastoma of three mice after injection of DVDMS (10 mg/kg) along with MBs, while the tumors of the other three mice were unsonicated. 3 h later, cardiac perfusions were performed and the extravasation of DVDMS was detected using an IVIS imaging system (PerkinElmer Inc., Waltham, MA, USA).

Intracranial Glioblastoma Xenograft Model Preparation

Male Balb/c nude mice (6–7 weeks old, 20 ± 2 g; the Animal Center of Southern Medical University, Guangzhou, China) were used in this study. The animals were maintained under specific pathogen-free condition (relatively constant temperature: 24–26 °C, humidity: 30–50%). All surgical instruments and supplies were used after sterilization. Animal care and experiments were approved by the Animal Care and Use Committee of School of Medicine in Shenzhen University, China (Approval No.: 20160110). Animals were anesthetized with 1.5% isoflurane (RWD Life Science, Shenzhen, China), immobilized in a small animal stereotaxic apparatus (RWD Life Science). Their heads were scrubbed with betadine and alcohol. A 5-mm skin incision was made along the sagittal suture and a burr hole drilled into the skull. Then, 1 × 105 human glioblastoma cells (U87 MG-Red-FLuc) were injected into the striatum of the right hemisphere (1.3 mm anterior and 2.0 mm lateral to the bregma) at a depth of 3.5 mm from the brain surface. The drilled holes in the skull were then sealed with bone wax.

In Vivo Study of SDT on Animals Bearing Intracranial Glioblastoma

SDT efficacy of glioblastoma was then investigated in vivo on nude mice bearing intracranial tumor. On the sixth day after implantation of U87 MG-Red-FLuc cells, bioluminescence imaging (BLI) (IVIS SPECTRUM; PerkinElmer Inc., Waltham, MA, USA) was performed to verify the success of the tumor model establishment. Thirty-six tumor-bearing mice were used and randomly divided into six groups: the control group without SDT (Control, n = 6), injection of DVDMS (2 mg/kg) alone (DVDMS alone, n = 6),12,34 injection of DVDMS with BBB opening induced by FUS with MBs (DVDMS + FUSiBBBo, n = 6), focused ultrasound sonication without DVDMS injection (SDT w/o DVDMS, n = 6), focused ultrasound sonication with DVDMS injection (SDT w/DVDMS, n = 6), SDT after enhanced delivery of DVDMS through the BBB disrupted by FUS with MBs (SDT w/DVDMS + FUSiBBBo, n = 6).

Experimental timelines of the single treatment was illustrated in Fig. 2a. 3 h after enhanced delivery of DVDMS by FUS with MBs, sonication procedure was performed with a pulse length of 300 ms, a PRF of 1 Hz for 60 s. The acoustic power used in the in vivo study was 1.7 W with the consideration of the skull attenuation of ultrasound energy. Sonodynamic treatment started on the seventh day after tumor implantation and was carried out every 3 days for three times totally as shown in Fig. 2b.

Figure 2
figure 2

Experimental timelines of the in vivo study of DVDMS-mediated SDT treatment protocol. (a) The timeline of a single treatment procedure of SDT after the BBB opening by FUS with MBs. (b) The timeline of the course of SDT and BLI monitoring of glioblastoma-bearing nude mice.

Full size image

Bioluminescence Monitoring of the Intracranial Glioblastoma Growth

Longitudinal monitoring of the intracranial glioblastoma growth was conducted through BLI before and after treatments every 3 days using an in vivo imaging system (IVIS SPECTRUM; PerkinElmer Inc., Waltham, MA, USA) with the advantage of using the same cohorts of the animals. Luciferin solution (150 mg/kg of body weight, PerkinElmer Inc., Boston, MA, USA) was injected intraperitoneally 15 min before imaging. The photon emission signals were recorded and measured for quantitative analysis of the tumor cell growth.

Survival Data Analysis

The survival time of the six groups of the glioblastoma-bearing nude mice were recorded until death. Survival curves were plotted using the Kaplan–Meier method and the significances were analyzed by Log-Rank test. The median and mean survival time, the increase in median and mean survival time (ISTmedian and ISTmean), the maximal survival time were compared.

Histological Examination and Immunohistochemical Analysis

Proliferating cell nuclear antigen (PCNA) immunostaining was performed to evaluate tumor cell proliferation and cleaved-caspase-3 assay was used to examine tumor cell apoptosis. 1 week after the treatments, brain tumors were harvested and fixed using 10% formalin solution for at least 48 h. They were then embedded in paraffin and sliced into 5-µm sections. In each group, sections were also stained with hematoxylin and eosin (H&E) for histological examination.

The process of PCNA immunostaining was as follows. In brief, paraffin-embedded tumor tissue sections were dewaxed and incubated with 3% H2O2 solution and the anti-PCNA antibody (1:50 dilution, Abcam, Cambridge, MA, USA) overnight at 4 °C. After washing with PBS, antibody binding was detected with horseradish peroxidase-conjugated secondary antibody (Zhongshan Golden Bridge Biotechnology Co. Ltd., Zhongshan, China) for 1 h at 37 °C. The sections were visualized with diaminobenzidine (DAB) solution, and then lightly counterstained with hematoxylin.

To identify apoptotic cells, dewaxed sections were rehydrated and incubated with 3% H2O2 in methanol, and then incubated with monoclonal anti- cleaved-caspase-3 (Abcam, Cambridge, MA, USA) antibody and indirect immunoperoxidase technique was used. Slices were imaged using a microscope (BX53; Olympus Corporation, Tokyo, Japan). Image Pro Plus 6.0 (Media Cybernetics, Inc., Bethesda, MD, USA) was used to quantify the percentages of immunopositive cells to estimate the amount of proliferating and apoptotic cells. Five random regions per section and three sections were examined in each group.

Statistical Analysis

GraphPad Prism™ 7.0 (GraphPad Software, Inc., San Diego, CA, USA) was used for statistical analysis. All data were expressed as the mean ± standard deviation (SD). Statistical differences were evaluated by one-way analysis of variance (ANOVA) among the groups. The post-hoc analysis was used by Tukey’s multiple comparison test. A value of p  < 0.01 was considered to be significant.

Results

In Vitro Cytotoxic Effect of SDT on U87 MG-Red-FLuc Cells

The in vitro cytotoxic effect of SDT using DVDMS as the sonosensitizer was evaluated by flow cytometry. First, the cellular uptake of DVDMS experiment was conducted. The fluorescence signal of intracellular DVDMS increased with incubation time and reached its peak value at approximately 3 h after incubation with U87 MG-Red-FLuc cells (Figure S1). Thus, in the following apoptotic assay, we chose 3 h as the incubation time of DVDMS with cells. As shown in Fig. 3b, dot plot graphs showed viable cells (Q4, AV−/PI−), early apoptotic cells (Q3, AV+/PI−), late apoptotic cells (Q2, AV+/PI+) and necrotic cells (Q1, AV−/PI+). The ratios of apoptotic and necrotic cells were computed based on the total amount of cells in Q1, Q2, and Q3, divided by the total number of cells. They were 3.3 ± 1.9% in the control group, 8.9 ± 3.4% in the DVDMS alone group, 20.5 ± 5.4% in the SDT w/o DVDMS group and 63.9 ± 9.9% in the SDT w/DVDMS group respectively. There was no significant difference between the control group and DVDMS alone group. Compared with the control group, both SDT w/o DVDMS and SDT w/DVDMS with the same FUS exposure parameter induced decreased viable cells. But the apoptotic and necrotic ratio of SDT w/DVDMS was significantly higher, indicating remarkable cytotoxic effect on U87 MG-Red-FLuc cells.

Figure 3
figure 3

In vitro study of sonodynamic therapy on glioblastoma cells. (a) Schematic illustration of the in vitro study of SDT on GBM cells. (b) The in vitro study of SDT on U87 MG-Red-FLuc cells by flow cytometry and the percentages of apoptotic and necrotic cells in the four groups. *p < 0.05, **p < 0.01 and ***p < 0.001.

Full size image

Characterization and In Vivo Stability of MBs

The prepared MBs were polydispersed with sphere shapes under the brightfield microscope as shown in Fig. 4a. The mean diameter of the MBs was 1.908 μm. D90, representing the size below which 90% particles fall, was 2.976 μm after fresh preparation (Fig. 4b). The size distribution mostly ranged from 0.8 to 8 μm. The concentration of MBs in the stock vial was 2.5 × 109/mL.

Figure 4
figure 4

Characterization and in vivo stability of MBs. (a) Photomicrograph of MBs with a lipid shell and perfluoropropane core, bar: 20 μm; (b) Size distribution of MBs. (c) Contrast-enhanced B-mode ultrasound images of a mouse brain before and post MBs injection. Craniotomy surgery was performed before ultrasound imaging with a 40 MHz transducer. (d) Corresponding time intensity curve of contrast-enhanced B-mode images.

Full size image

The in vivo stability of the MBs was evaluated by ultrasound contrast-enhanced B-mode imaging of mice brains over 20 min. As shown in Fig. 4c, the brain image contrast was significantly enhanced after the injection of MBs into the tail vein. The corresponding time intensity curve of the region of interest (ROI) was plotted in Fig. 4d. The initial gray scale intensity before MBs injection was set as the baseline. The signal intensity elevated to 33 dB after MBs administration. It gradually dropped to the half level about 8 min later and to the initial amplitude about 20 min later.

Enhanced DVDMS Delivery and Accumulation in Brain Using FUS with MBs

Some studies showed that sonosensitizer DVDMS preferentially accumulated at the tumor sites likely due to the strong enhanced permeability and retention effect.14,36 Although glioblastoma vasculature has leaky property, our previous study showed that the distribution of the substance Evans blue was heterogeneous and insufficient within the tumor region.23 Thus, the enhanced effect on delivery of DVDMS by using FUS with MBs was investigated through the examination of the penetration of DVDMS through the BBB. As shown in Fig. 5, it was noticed that the accumulation of DVDMS in the right glioblastoma of sonicated mice was enhanced compared to that in the unsonicated one. The quantification of fluorescence signals showed an increased level of DVDMS deposition after treated by FUS with MBs by 3.43 folds of that in the unsonicated ones. The results confirmed that FUS with MBs effectively enhanced DVDMS delivery after BBB disruption.

Figure 5
figure 5

Comparisons of the DVDMS that penetrated through compromised BBB in glioblastoma to those that penetrated through the BBB disrupted by FUS with MBs. (a) The control group without DVDMS injection (left), DVDMS accumulation in the brain region through the compromised BBB (middle) and FUS-disrupted BBB (right). The glioblastoma xenograft was implanted in the right hemisphere. The brains were harvested 3 h after DVDMS injection or BBB opening with FUS. (b) Fluorescence intensity of the DVDMS in the brain regions in the control group, DVDMS alone group and DVDMS + FUSiBBBo group (n = 3, mean ± SD). *p < 0.05, **p < 0.01, ***p < 0.001.

Full size image

Then the accumulation of DVDMS in the brains was explored in vivo at different time points after BBB opening by FUS with MBs. Fluorescence images of accumulated DVDMS showed that the concentration of DVDMS in the brain first increased to the peak at 3 h after BBB opening and then decreased over the next hours (Fig. 6). So we chose 3 h as the best time interval for further sonodynamic treatment study.

Figure 6
figure 6

DVDMS accumulation at different time points after BBB opening induced by FUS with MBs in vivo. (a) Fluorescence images of DVDMS accumulation after BBB opening; (b) Quantitative fluorescence intensity variations of DVDMS with time (n = 3).

Full size image

In Vivo Therapeutic Efficacy of SDT on Animals Bearing Intracranial Glioblastoma

The anti-glioma effect of DVDMS-mediated SDT in vivo was investigated by monitoring intracranial glioblastoma tumor growth in nude mice through BLI before and after treatments. As shown in Fig. 7a, before the SDT treatment, there was no significant difference among the six groups in total bioluminescence photon emission on the sixth day after intracranial U87 MG-Red-FLuc cells implantation. For the untreated control group, the intracranial tumors showed rapid and exponential growth rate. In the DVDMS alone, DVDMS + FUSiBBBo and SDT w/o DVDMS group, the patterns of tumor progression of nude mice were similar to that in the control group, suggesting these treatments couldn’t inhibit tumor growth. In the SDT w/DVDMS group, the growth of glioma in nude mice rapidly progressed and had no significant difference from the control group at last (Control vs. SDT w/DVDMS: 115.05 ± 97.6 and 108.21 ± 45.97 × 107 photons/s at 23 days after tumor implantation, p > 0.05). In contrast, SDT w/DVDMS + FUSiBBBo group exhibited delayed tumor growth compared with the control group. It is noteworthy that the bioluminescence photon emission value of SDT w/DVDMS + FUSiBBBo group (1.59 ± 0.68 × 107 photons/s) was significantly lower than the control group (115.1 ± 97.6 × 107 photons/s) on the 23th day after tumor implantation (Fig. 7b). This strongly implied that SDT after enhanced delivery of DVDMS effectively retarded intracranial glioblastoma growth.

Figure 7
figure 7

In vivo study of SDT on animals bearing intracranial glioblastoma. (a) Representative BLI of intracranial U87 MG-Red-FLuc glioblastoma progression before and after different treatments. Control, the group without treatment; DVDMS alone, injection of DVDMS (2 mg/kg) alone; DVDMS + FUSiBBBo, injection of DVDMS with BBB opening induced by FUS with MBs; SDT w/o DVDMS, FUS sonication without DVDMS injection; SDT w/DVDMS, FUS sonication with DVDMS injection; SDT w/DVDMS +  FUSiBBBo, SDT after enhanced delivery of DVDMS through the BBB disrupted by FUS with MBs. (b) Bioluminescence photon emission counts of the intracranial glioblastoma treated with different treatments from day 6 to day 23 after the implantation of U87 MG-Red-FLuc cells. **p < 0.01 and ***p < 0.001. (c) The Kaplan–Meier survival curves of nude mice in each treatment group (n = 6).

Full size image

The survival times of six groups of nude mice that received different treatments were recorded. Figure 7c showed the Kaplan–Meier survival curves of the six groups and the corresponding statistical data are summarized in Table 1. It can be seen that the median survival time of DVDMS alone, DVDMS + FUSiBBBo and SDT w/o DVDMS group was 24.00, 25.00 and 23.25 days respectively, which were not statistically different from the control group (23.75 days). The median survival time of the mice treated with SDT w/DVDMS was 8.95% greater than the untreated one (p = 0.0919). Noticeably, for the SDT w/DVDMS + FUSiBBBo group, the median survival time was significantly prolonged to 30.25 days after treatment (ISTmedian = 27.37%; p < 0.0001 relative to control). Moreover, the maximal survival time in the SDT w/DVDMS + FUSiBBBo group was 39 days, compared to 26 days in the control group. These survival data of in vivo experiment further indicated that SDT with DVDMS delivered via FUS with MBs showed improved therapeutic efficacy of nude mice bearing intracranial U87 MG-Red-FLuc glioblastoma.

Table 1 Survival analysis of nude mice in the six groups.
Full size table

H&E and Immunohistochemical Analysis

The tumor sections from mice in each group 1 week after the treatments were stained with H&E and examined under the light microscope. A significantly increased degree of cell necrosis and nuclear debris could be observed in the tumors receiving SDT treatment (Fig. 8a). To evaluate tumor cell apoptosis, immunohistochemical staining with cleaved-caspase-3 antibody was performed. The results showed that cleaved-caspase-3-positive cells were significantly increased in the tumors undergone DVDMS-mediated SDT, compared with other groups (Fig. 8b). Quantitatively, there were 41.78 ± 9.92% apoptotic tumor cells found in the tumors treated by SDT w/DVDMS + FUSiBBBo (Fig. 8c). In contrast, there were 5.49 ± 1.64%, 5.39 ± 1.45%, 6.69 ± 2.33%, 9.72 ± 2.95% and 10.41 ± 2.79% apoptotic cells found in the Control group, DVDMS alone, DVDMS + FUSiBBBo, SDT w/o DVDMS and SDT w/DVDMS group respectively (p < 0.01).

Figure 8
figure 8

(a) Images of the tumor sections stained with H&E in the six groups undergone different treatments. (b) Representative immunohistochemical staining of glioma sections against cleaved-caspase-3. Cell nuclei were counterstained with hematoxylin (cleaved-caspase-3 positive cells, brown). (c) Representative immunohistochemical staining of glioma sections against PCNA. Cell nuclei were counterstained with hematoxylin (PCNA positive cells, brown). (d) Apoptosis quantification assessed by the percentage of cleaved-caspase-3-positive cells in the tumor region. The apoptotic index was determined from the percentage of positive cells out of at least 2000 tumor cells observed at × 400 magnification. (e) Quantitative analysis of PCNA-positive cells in the tumor region. The positive index was determined from the percentage of positive cells out of at least 2000 tumor cells observed at × 400 magnification. Bar: 50 µm. **p  <  0.01.

Full size image

Furthermore, to investigate the anti-proliferative effect of DVDMS-mediated SDT on glioblastoma, immunohistochemical staining with PCNA antibody was performed. The quantitative results in Fig. 8d showed markedly decreased levels of PCNA-positive cells in DVDMS w/SDT + FUSiBBBo group with the ratio of 16.07 ± 3.92%, compared with that in the Control, DVDMS alone, DVDMS + FUSiBBBo, SDT w/o DVDMS and SDT w/DVDMS group respectively (p < 0.05), which was 45.97 ± 9.64%, 42.67 ± 8.52%, 48.20 ± 11.70%, 39.97 ± 7.95%, and 34.72 ± 8.31% respectively (Fig. 8e). These findings suggested that glioma proliferation was evidently inhibited by the enhanced delivery of DVDMS combined with SDT treatment.

Discussion

Therapeutic strategies for treatment of glioblastoma including surgery, radiotherapy and chemotherapy have been used clinically, but the treatment outcomes are not satisfactory.28 In recent years, SDT has been reported to treat cancer locally and effectively as a noninvasive and repeatable approach for cancer treatment.24 Many studies showed obvious anti-tumor cytotoxicity through synergistic effects of low-intensity ultrasound in combination with the sonosensitizers.4,12,18,24,25,26,31,33 However, few studies of SDT have been performed on animal models bearing glioblastoma.11,17,18,26,27 Due to the existence of BBB, most of therapeutic agents including sonosensitizers could not be delivered sufficiently, while the extravasations and depositions of sonosensitizers from blood into tumor regions play crucial roles in SDT. Thus, this study aimed to investigate the anti-tumor effects of SDT on glioblastoma after increasing the delivery efficacy of the sonosensitizer across BBB using FUS combined with MBs.

Phospholipid-shelled MBs were prepared in our lab. The characterization result indicated that the size distribution of our MBs was similar to Definity (mean size: 1.1–3.3 μm). Through B-mode contrast imaging, we found that the duration of the signal intensity falling from the peak value (33 dB) to the half level was about 8 min (half-life of Definity: 2–10 min).13 There are several factors influencing the in vivo circulation time of MBs, including shell materials, water permeation resistances of the gas, the shell resistance to gas permeation, microbubble size etc. The shell and core gas play the crucial role in the stability of MBs. Perfluoropropane gas can increase microbubble lifetime by an order of magnitude than the air due to its higher water permeation resistances, nearly 9-fold greater than that of air, thus being used in the generation of many types of MBs, like Definity.2 For the shell, phospholipids are most commonly used for the reason that it is easy to form, biocompatible, echogenic and decreases gas permeation as well.6 The composition of the MB shell prepared in our lab (DSPE, DSPE-PEG2000) resembles that of Definity (DPPC, DPPA, MPEG5000- DPPE). The long PEGylated phospholipids could also stabilize MBs by inhibiting coalescence and passivate surface and increase the in vivo circulation time of MBs. The results indicated that our MBs were similar to Definity in several features, including lipid composition, concentration, size and stability.

Enhancing the sonosensitizer accumulation dose in glioma is critical to improve SDT efficacy. Therefore, in vivo experiments were performed to investigate the delivery of DVDMS by FUS with MBs in the present study. Low-intensity bursts of FUS in the presence of circulating MBs has been demonstrated to deliver therapeutic agents locally, noninvasively and reversibly in many studies.9,13 Our previous studies also showed that anti-cancer drugs with the size of hundred nanometers could be delivered across BBB using this approach.22 The results of this study showed that the accumulation of DVDMS in the glioblastoma region was enhanced significantly by FUS with MBs compared to the one due to the enhanced permeability and retention effect of the brain tumor itself. The time window between BBB opening and SDT, which played an important role in SDT, was also examined.7 The concentration of DVDMS in brain tissue gradually increased and then dropped to the initial value in a few hours after BBB opening which indicated that BBB was reversibly opened by pulsed FUS with MBs.1

On the basis of the above experimental results, the anti-glioma effect of DVDMS-mediated SDT was investigated in vivo subsequently on nude mice bearing intracranial glioblastoma. During the longitudinal observation through BLI, the tumor growth in SDT w/DVDMS + FUSiBBBo group was significantly retarded when compared with Control, DVDMS alone, DVDMS + FUSiBBBo, SDT w/o DVDMS and SDT w/DVDMS. It is noteworthy that the value of bioluminescence of SDT w/DVDMS + FUSiBBBo group (1.59 ± 0.68 × 107 photons/s) was significantly lower than the control group (115.1 ± 97.6 × 107 photons/s) at the ninth day after SDT treatment. Additionally, the median survival time of the SDT w/DVDMS + FUSiBBBo group was prolonged to 30.25 days after treatment by 27.37%. Moreover, the maximal survival time in the FUS + DVDMS group extended to 39 days, compared to 26 days in the control group. In Ohmura et al’s study, SDT (10 W/cm2, 1.04 MHz, 5 min) with 5-aminolevulinic acid as the sonosensitizer induced anti-tumor effect on C6 rat giloma implanted in the brains of rats. The tumor area in the SDT group decreased to nearly 60% of the control ones at 1 week after the treatment.18 However, there was no long-term survival analysis in their study. In another study by Jeong et al., 5-aminolevulinic acid -mediated SDT (2.65 W/cm2, 1 MHz, 20 min) on C6 glioma showed 67% tumor regression 3 days after the treatment.11 However, there was an obvious temperature rising in the tumor due to the thermal effect induced by the longtime continuous ultrasound wave they used. Besides, the impediment of BBB to the sonosensitizers was not taken into account in both of the above studies. In contrast, our results herein showed that human glioblastoma suppression on nude mouse models could be induced by using lower ultrasound intensity (1.7 W, equaling to 2.94 W/cm2) and less irradiation time (1 min) after the enhanced delivery of DVDMS by FUS with MBs. Besides, different from the rat tumor model they used, our study chose the U87 MG orthotopic glioblastoma model transplanted in nude mice, which is usually used to mimic human glioblastoma multiforme (GBM) closely. Recently, Suehiro et al. conducted a combination therapy of HIFU with 5-ALA-mediated SDT on a mouse glioma model.27 HIFU (2.2 MHz, 500 W/cm2, 100% duty cycle for 5 min) was performed on nude mice once a week for 3 weeks first to induce an ablation in the tumor center and then SDT was performed. The proliferative activity of the entire tumor was markedly decreased and the survival time was greatly prolonged. However, the strategy of their study was quite different from ours. Although the disparities between different sonosensitizers were not compared in the present study, our results may provide an improved approach of SDT on glioblastoma treatment.

In the present study, immunohistochemical examinations were performed to further verify the therapeutic effectiveness of SDT on glioblastoma. H&E staining showed obvious pyknotic glial cells post DVDMS-mediated SDT treatment. Furthermore, the expression level of PCNA in the SDT w/DVDMS + FUSiBBBo group was significant reduced compared with Control and other groups. Cleaved-caspase-3 staining results indicated significant tumor cell apoptosis after SDT. However, the mechanism through which DVDMS mediated SDT induced apoptosis and anti-tumor activity has not been investigated thoroughly in this study. It suggested that reactive oxygen species generated by SDT contributed to the mitochondrial pathway of apoptosis, and the key molecule cleaved-caspase-3 were involved.4 The study of Wang et al. showed that severe mitochondrial damage in ECA-109 cells was induced by SDT and the cytotoxicity was effectively remitted by reactive oxygen species scavengers.31 Therefore, oxidation stress might be involved in the mechanism of DVDMS mediated SDT on glioblastoma cells, which needs to be investigated in further study.15,36

Conclusions

This study explored the feasibility of SDT with the novel sensitizer DVDMS, which was delivered by low-intensity FUS with MBs, for the treatment of intracranial glioblastoma xenografts in nude mice. The results showed the approach of FUS with MBs induced significantly enhanced delivery outcome of DVDMS. Both in vitro and in vivo study indicated anti-tumor effects of DVDMS-mediated SDT on human glioblastoma cell line U87 MG-Red-FLuc, including apoptosis induction and cell proliferation suppression. BLI monitoring showed effective retarding of intracranial glioblastoma growth in nude mice brain. These results suggested that DVDMS-mediated SDT after enhanced delivery of the sonosensitizer by FUS with MBs may be a potential therapeutic strategy for human glioblastoma in future.