Enhanced sonodynamic therapy by carbon dots-shelled microbubbles with focused ultrasound

Graphical abstract

Highlights

• •This study proposed a novel sonosensitizer, using carbon dots (C-dots) to assemble microbubbles with a gas core (C-dots MBs).
• •The C-dots MBs will produce ¹O₂, •OH, and H₂O₂ and induced cellular lipid peroxidation after ultrasound irradiation.
• •In an animal solid tumor model, ultrasound with C-dots MBs increased the proportion of ROS-positive cells and apoptotic cells.

Abstract

Sonodynamic therapy involving the non-invasive and local generation of lethal reactive oxygen species (ROS) via ultrasound (US) with sonosensitizers has been proposed as an emerging tumor therapy strategy. However, such therapy is usually associated with inertial cavitation and unnecessary damage to healthy tissue because current sonosensitizers have insufficient sensitivity to US. Here, we report the use of a new proposed sonosensitizer, carbon dots (C-dots), to assemble microbubbles with a gas core (C-dots MBs). As the C-dots were directly integrated into the MB shell, they could effectively absorb the energy of inertial cavitation and transfer it to ROS. Our results revealed the appearance of ¹O₂, •OH, and H₂O₂ after US irradiation of C-dots MBs. In in vitro experiments, treatment with C-dots MBs plus US induced lipid peroxidation, elevation of intracellular ROS, and apoptosis in 32.5%, 45.3%, and 50.1% of cells respectively. In an animal solid tumor model, treatment with C-dots MBs plus US resulted in a 3-fold and 2.5-fold increase in the proportion of ROS-damaged cells and apoptotic cells, respectively, compared to C-dots MBs alone. These results will pave the way for the design of novel multifunctional sonosensitizers for SDT tumor therapy.

1. Introduction

Clinical cancer treatment modalities such as surgery, chemotherapy, radiotherapy, and immunotherapy have potential disadvantages of invasiveness, non-specificity, radiation-induced damage, or the induction of life-threatening cytokine release syndrome [1], [2], [3]. Therefore, the development of more targeted and noninvasive therapeutic approaches has attracted great interest to overcome these longstanding problems. Sonodynamic therapy (SDT) can be used to noninvasively and locally kill tumor cells because it can remotely generate cytotoxic reactive oxygen species (ROS) at a depth of several centimeters. Moreover, the cytotoxic mechanisms of SDT are mediated through oxidation of cellular substrates, which could avoid the development of tumoral drug resistance and the induction of undesired immune responses [4], [5].

As the distance of ROS diffusion is only 10–55 nm [6], SDT has been tested in various tumor animal models as a noninvasive tumor treatment approach with subcellular accuracy [7], [8]. The continuous and high acoustic pressure ultrasound (US) waves that are typically required to produce ROS make it difficult to yield sufficient ROS to kill tumor cells by US alone [4] and several materials that can generate ROS upon US excitation have been used as sonosensitizers, such as organic small molecule drugs, pyrolysable chemicals, and inorganic materials [9], [10], [11]. These sonosensitizers can interact with US to produce ROS via acoustic cavitation, sonoluminescence, or pyrolysis [8], [12]. However, due to the insufficient sonosensitivity of existing sonosensitizers, high-energy US excitation or a high dosage of sonosensitizer is often required to yield an adequate amount of ROS, raising the potential of associated tissue damage or systemic cytotoxicity [13]. Considering the mechanism of SDT, some groups have integrated sonosensitizers onto the surface of cavitation nuclei via biochemical ligands for efficient transfer of energy [14], [15], [16]; however, the immune response to those biochemical ligands raised concern. To date, there are no sonosensitizers that can produce ROS via low-energy US with high safety.

Carbon dots (C-dots) are fluorescent quasi-spherical nanoparticles with size of ∼10 nm and outer shells bearing oxygen-containing groups, including –NH₂, –OH, C-O-C, and –COOH [17], [18]. In the proposed application of C-dots as a sonosensitizer, its outer surface is first confronted with the impact of cavitating bubbles upon US activation. These cavitation-produced extremely high energies were able to break apart those oxygen-containing groups onto the C-dots, generating several transient active species and producing ROS. C-dots already have been attempted applying in biomedical field, such as bioimaging, drug delivery, biosensors, and photocatalysis, suggesting excellent biocompatibility. The sonocatalysis application of C-dots was firstly reported by Ren et al. [19]. During US sonication, the C-dots produced O- radicals and could react with hydrogen ions and water, providing 2-fold of sonocatalytic activity compared with TiO₂. Furthermore, the ROS yield of C-dots can be increased via reducing the pH values of the surrounding due to a great number of cations (e.g. H⁺, Na⁺) may: (1) interact with C-dots and cleavage their surface groups (e.g. C-O groups and C = O groups) to generate oxygen free radical [19], [20]; (2) interact with ·O– and produce ¹O₂ and •OH [19]. Previous had reported that feasibility of using C-dots with US for SDT [21], [22]. However, it has been challenging to translate C-dots–mediated SDT into in vivo applications since these C-dots required low-frequency, high-energy parameters (frequency of 20–50 kHz, energy of 200–500 W, and sonication duration of 600–1800 s) that would potentially induce thermal effect and unpredictable cavitation effect to damage biological tissues [19], [23]. The occurrence of US-induced thermal effect and cavitation effect were increased as prolonging the US irradiation duration and lowing the US frequency, respectively [24]. Besides, it’s also hard to real-time trace the distribution of the administrated C-dots in vivo.

Our group previously engineered C-dots with hydrophobic carbon chains that could assemble into liposomes for use in tumor cell targeting and the generation of ROS for photodynamic therapy [25], [26]. Here, we firstly investigated that if these C-dots liposomes could produce sufficient ROS for SDT after low-intensity US of several MHz, which has been widely applied in medical applications. Since the structure of lipid-shell gaseous microbubbles (MBs) is considered to have highly efficient US energy absorbing properties, we aimed to assemble the C-dots liposomes into MBs without additional linkage (Fig. 1). In the meantime, the structure of MBs also could provide contrast enhancement signals in US imaging. To test whether these MBs reduce the US energy for activation of C-dots in SDT, three groups were investigated in this study: (1) C-dots liposomes activated by US; (2) C-dots liposomes activated by US in combination with MBs; (3) C-dots liposomes assembled into MBs (C-dots MBs) activated by US. We then investigated the mechanism of ROS formation and identified the ROS species induced for these groups. Finally, the SDT abilities of these three frameworks were evaluated by tumor cell culture experiments and in a solid tumor–bearing animal model. We believe that this study will contribute to the development of novel SDT modalities to enhance the efficiency of oncologic treatment.

Fig. 1.

Illustration of SDT for anti-tumor application using C-dots MBs with US.

2. Materials and methods

2.1. Preparation and characterization of C-dots liposomes

The synthesis of C-dots liposomes had been reported previously [25]. Briefly, heating triolein at 220 °C for 3 days. The product was then dissolved in sodium hydroxide and sonicated for 2 h. The mixture was performed dialysis (molecular weight cut off: 15 kDa) with phosphate-buffered saline (PBS) and extracted liquid–liquid extraction in butanol/water. The C-dots liposomes were purified through dialysis (molecular weight cut off: 15 kDa) against PBS from the organic phase.

2.2. Preparation and characterization of C-dots MBs

Lipids 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[carboxy(polyethylene glycol)-5000] (DSPE-PEG5000, Avanti Polar Lipids) and C-dots were homogeneously mixed in chloroform. Six formulations (designated F1–F6, listed according to their different C-dots: DSPE-PEG5000 mass ratios in Table 1), were utilized to optimize the formulation. The chloroform was then removed by an evaporator (R‐210, Büchi Labortechnik AG, Flawil, Switzerland) to generate a dried lipid film. Glycerol, PBS, and the dried lipid film were mixed, degassed, and refilled with perfluoropropane (C₃F₈). The C-dots MBs were produced following intensive agitation. Finally, the C-dots MBs were purified through centrifugation (2000× g, 1 min).

Table 1.

The molar ratio of C-dots and DSPE-PEG 5000 for synthesis of C-dots MBs.

Formulation	C-dots: DSPE-PEG5000
(molar ratio)
F1	9:0
F2	9:0.5
F3	9:1
F4	9:2
F5	9:3
F6	9:4

For comparison, regular lipid MBs without C-dots were also prepared. The regular lipid MBs consisted of DSPC (1,2-Distearoyl-snglycero-3-phosphorylcholine; Avanti Lipids Polar, Inc) and DSPE-PEG5000 (mass ratio, DSPC: DSPE-PEG 5000 = 9: 4) and were prepared following the same fabrication processes as for C-dots MBs without centrifugation.

2.3. Characterization of C-dots liposomes and C-dots MBs

2.3.1. Size distribution and structure of C-dots liposomes

The size distribution of C-dots liposomes was determined by dynamic light scattering (Nanosizer-S, Malvern, London, UK). To verify the structure, the C-dots liposomes were observed via a transmission electron microscope (TEM) (Hitachi H-7100, Tokyo, Japan). Before imaging, the C-dots liposomes were transferred onto a 300-mesh copper grids covered with porous carbon film (HC300-Cu, PELCO, CA, USA), and dried at 60 °C for 2 h. The samples were imaged by the TEM operated at 200 kV.

2.3.2. Size distribution, concentration, and structure of C-dots MBs

The size distribution and concentration of C-dots MBs were evaluated using a coulter counter (Multisizer 3, Beckman Coulter, FL, USA). The morphology of C-dots MBs were visualized under an inverted fluorescence microscope (IX-71, Olympus, NY, USA). No additional fluorescent staining was required because the intensive intrinsic fluorescent emission of C-dots at 570 nm upon excitation at 530 nm. The C-dots MBs were also imaged by bioluminescence imaging (IVIS-200, Xenogen Corporation, CA, USA) for different concentrations (0.312×10⁶ – 10×10⁶ MBs/mL). The fluorescence intensity of the obtained images was estimated with Living Image 3.0 software (Caliper Life Sciences, MA, USA). To verify the structure, the C-dots MBs were imaged by TEM using a similar process as for C-dots liposomes.

2.3.3. Acoustic properties of C-dots MBs

The in vitro stability of the C-dots MBs was measured by estimating their echogenicity from US B-mode images. The C-dots MBs (0 – 10×10⁶ MBs/mL) were transferred into a 2% agarose phantom and imaged at intervals of 10 min at 37 °C for 1 h by a 7.5-MHz US imaging system (model t3000, Terason, MA, USA). The echogenicity of the C-dots MBs were estimated by comparing the signals intensity at each time point with 0 min by MATLAB^TM software (The MathWorks, Natick, MA, USA). The threshold of inertial cavitation at which MBs destruction occurred in the C-dots MB suspension (10×10⁶ MBs/mL) with irradiation by a 1-MHz US sonication system (ST-TM1-20, Sonitron GTS Sonoporation System, Nepa Gene Co., Ltd. Japan) was determined using the US imaging system. The 1-MHz US sonication system was operated at 50% duty cycle with different powers (0.25, 0.5, 0.75, 1 W/cm²).

2.4. Detection of ROS produced by US + C-dots liposomes or US + C-dots MBs

2.4.1. Fluorescence measurements

ROS generation was first evaluated using AUR reagent (Amplex® UltraRed Reagent, Thermo Fisher Scientific, WA, USA), which reacts with ROS and emits a red fluorescent signal. The AUR solution (concentration: 40 μg/mL) and sample (C-dots liposome: 12 mg/mL; C-dots MBs: 10×10⁶ MBs/mL) were mixed at a volume ratio of 1:1 and then sonicated by 1-MHz US (ST-TM1-20: Plane Wave Transducer Module; 50% duty cycle, 0–1.0 W/cm², 0–3 min). The 1-MHz US transducer was then driven by a system (Sonitron GTS Sonoporation System, Nepa Gene Co., Ltd. Japan). The fluorescence intensity of the sample was determined by a microplate reader (λ_ex = 530 nm, λ_em = 590 nm, SPARK 10 M, TECAN, Switzerland). The output power of the 1-MHz US transducer was presented as spatial peak-temporal average (I_spta) measured by a calibrated ultrasound power meter (OHMIC Instruments, Easton, PA). For power measurement, the power meter was filled with degas water at 25 °C and the 1-MHz US transducer was replaced above the sensor of the power meter to conduct the measurement (Fig. S1(A)).

2.4.2. EPR measurements

For electron paramagnetic resonance (EPR) analysis, hydroxyl radicals (•OH) and singlet oxygen (¹O₂) generated by US with C-dots MBs were detected by DMPO (5,5-dimethyl-1-pyrroline-N-oxide, Aladdin Biochemical Technology Co., Ltd, Shanghai, China) dye and TEMP (triacetone diamine, Aladdin Biochemical Technology Co., Ltd,) dye, respectively. PBS solution containing C-dots MBs (10×10⁶ MBs/mL) was exposed to 1-MHz US (1.0 W/cm², 50% duty cycle, 1 min) in the presence of TEMP solution (40 mM) or DMPO solution (40 mM). An aliquot of the analytical solution was then transferred to a quartz tube for EPR analysis (ELEXSYS-E580, Bruker, Germany).

2.4.3. Fluorescence measurement for H₂O₂

The production of H₂O₂ was determined using AUR reagent and horseradish peroxidase (HRP) reagent. The HRP solution, AUR solution, and PBS were mixed at a volume ratio of 2:2:996. This solution was then mixed with C-dots-MBs (10×10⁶ MBs/mL) at a volume ratio of 1:1 and sonicated by 1-MHz US (50 % duty cycle, 0–1.0 W/cm², 1 min). The HRP reagent transforms H₂O₂ into •OH and emits a red fluorescent signal. The fluorescence intensity of the sample was measured by a microplate reader (λ_ex = 530 nm, λ_em = 590 nm).

2.4.4. ROS scavenger effect analysis

Histidine (10 mM) or mannitol (100 mM) was added to a solution of C-dots MBs (10×10⁶ MBs/mL) containing AUR before 1-MHz US sonication (1.0 W/cm², 50% duty cycle, 1 min). After sonication, the fluorescence intensity of samples was detected by a microplate reader (λ _ex = 530 nm, λ_em = 590 nm).

2.5. Cell experiments

2.5.1. Cytotoxicity measurements

TRAMP (transgenic adenocarcinoma mouse prostate) cells were seeded in a 96-well plate (2×10⁴ cell/well) 24 h prior to the experiments. The cells were co-incubated with the C-dots MBs (1.25×10⁶ MBs/mL) for 4 h, then cleaned by Dulbecco’s PBS, and resuspended in fresh culture medium (Dulbecco’s Modified Eagle’s medium/Ham’s Nutrient Mixture F-12, Gibco, NY, USA) for 24 h. Cytotoxicity was assessed by a cell counting kit (Dojindo Molecular Technologies, Inc., Kumamoto, Japan) and UV absorbance measurement (450 nm).

To estimate the cytotoxicity of C-dots MBs + US treatment, the cell dish was filled C-dot MBs (1.25×10⁶ MBs/mL) containing DPBS and sealing by an US penetrable membrane (Fig. S1(B)). Then, the cell dish was flipped upside down and was sonicated by US (1.0 W/cm², 50 % duty cycle, 1 min) to facilitate the C-dot MBs approaching the cells. An US coupling gel (aquasonic 100, Parker laboratories, Fairfield, NJ, USA) was applied between the membrane and the US transducer to maximize the transmission of US between the transducer and the cells. Subsequently, the cells were cleaned by Dulbecco’s PBS and resuspended in the fresh culture medium. Six groups were included in this experiment: (1) control without treatment; (2) regular lipid MBs only; (3) FUS only; (4) C-dots MBs only; (5) regular lipid MBs + US; (6) C-dots MBs + US. Cytotoxicity was assessed by the cell counting kit and UV absorbance measurement.

2.5.2. ROS scavenger effect analysis

Histidine (10 mM) or mannitol (100 mM) was added to the C-dots MBs. Cells were co-incubated with the mixture (C-dots MBs concentration: 1.25×10⁶ MBs/mL) and sonicated by 1-MHz US (1.0 W/cm², 50 % duty cycle, 1 min). Subsequently, the cells were cleaned by Dulbecco’s PBS (Gibco) and resuspended in the fresh culture medium. Cytotoxicity was estimated using the cell counting kit and UV absorbance measurement.

2.5.3. Assessment of ROS-induced cell damage

To reveal the mechanism by which C-dots-MBs + US enhanced cell death, we designed several live-cell imaging experiments, including tests for Image-iT® lipid peroxidation kit (Thermo Fisher Scientific), intracellular ROS (DCFDA dye, 2′,7′-Dichlorofluorescin diacetate, Abcam, Cambridge, MA, USA), cell apoptosis (PI dye, propidium iodide, Abcam), and cell necrosis (Annexin V-FITC dye, Abcam).

To detect lipid peroxidation of cell membranes, at 4 h after treatment with C-dots-MBs + US, the cells were co-incubated with a lipid peroxidation kit for 30 min. Subsequently, the cells were cleaned by PBS and stained with Hoechst 33,342 (5 μg/mL) to label cell nuclei. The occurrence of lipid peroxidation was visualized by a change in the color of fluorescent dye localized on the cell membrane from red (∼590 nm) to green (∼510 nm). The number of cells with green fluorescence signals was analyzed by flow cytometry (FACScalibur, BD Biosciences, CA, USA).

To detect intracellular ROS, the cells were treated with DCFDA dye (50 μM, ab113851, Abcam) for 30 min prior to treatment. At 15 min after US treatment, the cells were cleaned by PBS and imaged by the inverted fluorescence microscope. The fluorescence intensity of DCFDA dye positively correlated with intracellular ROS levels.

To detect cell apoptosis and cell necrosis, PI solution, Annexin V-FITC solution, and PBS were mixed at a volume ratio of 1:2:37. After treatment with C-dots-MBs + US, the cells were co-incubated with the PI and Annexin V-FITC mixture for 4 h. The cells were cleaned by PBS and stained with Hoechst 33,342 (5 μg/mL) to label cell nuclei. Apoptotic cells and necrotic cells were labeled by PI and Annexin V-FITC, respectively. The populations of PI-positive cells and Annexin V-FITC-positive cells were evaluated by flow cytometry.

2.6. In vivo experiments

2.6.1. Tumor model

To establish the solid tumor animal model, the right legs of mice were subcutaneously injected TRAMP cells (5×10⁶ cells). Animals were treated 45 days after injection of cells. All animal experiments were performed following the guidelines of National Tsing Hua University Institutional Animal Care and Use Committee (approval number: NTHU10459).

2.6.2. In vivo lifetime of the C-dots MBs

The in vivo lifetime of the C-dots MBs was measured by estimating their echogenicity in the normal tissue and tumor tissue by US B-mode imaging (N = 1). A time-lapse of US B-mode images was acquired before and after intravenous injection of the C-dots MBs (1×10⁷ MBs/mouse) by the 7.5-MHz US imaging system. The echogenicity of the C-dots MBs were estimated by evaluating the average signal intensities in the targeted region of the B-mode images by MATLAB^TM software.

2.6.3. In vivo bio-distribution of the C-dots MBs after 1-MHz US sonication

The in vivo bio-distribution of C-dots MBs with and without 1-MHz US irradiation (1.0 W/cm², 50% duty cycle, 3 min) was assessed at 1 h after treatment using the bioluminescence spectrum imaging system. For ex vivo organs imaging, the organs were harvested and immediately imaged at 1 h after US irradiation (1.0 W/cm², 50% duty cycle, 3 min). The fluorescence intensity of the targeted tissue was calculated by the Living Image 3.0 software.

2.6.4. Histological assessments

At 24 h after treatment, the tumor tissues of the treated mice were harvested as well as sliced at 15 μm thickness. Hematoxylin and eosin (H&E) staining, DCFHDA staining, and TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labeling, ApopTag kit, Intergen Co., NY, USA) staining were applied to detect tumor tissue damage, intratumor ROS (green), and apoptotic cells (red), respectively. The fluorescence intensities of DCFHDA and TUNEL signal positive cells were calculated by ImageJ software to estimate the effects of ROS production and tumor cell apoptosis, individually.

2.7. Statistics

To ensure the accuracy of the experiments, at least three independent replicates were performed. All results were presented as mean value with standard deviation. Statistical significance between two groups was calculated by ANOVA analysis, and significance was defined as a p value < 0.05.

3. Results

In cell culture medium, the C-dots liposomes showed a narrow size distribution with mean size of 176.4±21.9 nm (Fig. S2(A)), suggesting good dispersity. We first tested whether the C-dots liposomes could produce ROS upon 1-MHz US sonication. In the PBS control, the fluorescence intensities increased slightly with the application of acoustic energy (0 W/cm²: 51.0±4.0 a.u.; 0.25 W/cm²: 331.4±177.7 a.u.; 0.5 W/cm²: 656.3±207.2 a.u.; 0.75 W/cm²: 981.3±303.3 a.u.; 1 W/cm²: 1433.0±450.4 a.u.) (Fig. S2(B)), indicating that US sonication alone could produce a small amount of ROS. For the C-dots liposomes, increasing the US energy (0–1 W/cm²) or US sonication duration (1–3 min) resulted in similar fluorescence intensities to those of the PBS group (Fig. S2(C)). The cytotoxicity experiments demonstrated that the ROS yield from C-dots liposomes with US did not induce cell lethal damage, confirming the low US energy conversion efficiency of C-dots liposomes (Fig. S2(D)).

The structure of lipid-shelled gaseous MBs has been considered to have high US energy absorption and emission properties due to the high acoustic impedance mismatch between the gas core and surrounding medium. We next investigated whether the C-dots liposomes could be assembled into MBs. C-dots liposomes were mixed with DSPE-PEG5000, a widely used polymer that could reduce the immunogenicity of MBs and improve the biocompatibility of C-dots MBs [27]. Fig. S3(A) shows the prepared samples of the six formulations, which corresponded to molar ratios of DSPE-PEG5000 of 0, 0.5, 1, 2, 3, and 4 (F1-F6). A foamy layer above the suspension was seen in each formulation, suggesting the successful formation of MBs. The thickness of the MBs cake layer and the flotation ability of MBs increased from F1 to F6, probably reflecting increasing size of the MBs.

We then characterized the properties of these samples, such as mean size, concentration, and percentage of MBs larger than 8 μm, to determine the optimal formulation. Fig. S3(B) shows the size distributions of these samples. As the molar ratio of DSPE-PEG5000 increased from 0 to 1 (F1, F2, F3), the average size of C-dots MBs decreased from 2.2 μm to 1.7 μm (Fig. S3(C)). However, further increasing the molar ratio of DSPE-PEG5000 over 2 resulted in the enlargement of MBs (2.1 μm to 2.2 μm for F4, F5, and F6). The concentration of MBs in F3 was statistically higher than that in other groups, indicating the high yield of this formulation (2.0±0.3×10⁹ MBs/mL, Fig. S3(D)). Although the foaming effect shown in Fig. S3(A) is a general phenomenon during MBs formation, it is possible to generate large MBs that can block capillaries. The number of MBs larger than 8 μm was significantly lower in F3 than in other groups (3.7±1.9 %, Fig. S2(E)). Based on these results, we concluded that F3 (molar ratio, C-dots: DSPE-PEG5000 = 9:1) was the optimal formulation for preparing C-dots MBs and this formulation was used in all subsequent experiments.

Next, we examined the morphology and structure of the formed C-dots-MBs by TEM and microscopy. Fig. 2A shows the TEM images of C-dots MBs and the C-dots liposomes. The C-dots MBs were well dispersed round vesicles, and the shell was a monolayer. In comparison, the shell layer of C-dots liposomes was a double layer. The bright-field image demonstrated that a single C-dots MB had a spherical structure with a light-penetrating core (Fig. 2(B)). The fluorescence images showed a dense spherical structure of fluorescent signal, which was coherent as the observations by bright field microscopy, suggesting the successful assembly of C-dots liposomes into the MB shell. Note that the wavelength of fluorescence signals emitted by C-dots MBs was similar to that of C-dots (470 nm vs. 460 nm). The bioluminescence images also showed that the fluorescence intensity of the C-dots MB suspension was enhanced as its concentration increased (R² = 0.99, Fig. 2(C)), suggesting potential applications for in vivo imaging.

Fig. 2.

The properties of C-dots MBs. (A) TEM images. (B) Microscopic images. (C) Bioluminescence images. (D) US imaging contrast enhancement ability. (E) Acoustic stability. (F) 1-MHz US destruction test to estimate the threshold of inertial cavitation.

We then investigated the existence of a gaseous core within the C-dots-MBs because most US-mediated applications rely on the gaseous core. Fig. 2D shows that the C-dots-MBs could provide B-mode imaging contrast and that contrast enhancement positively corelated with the concentration of MBs (R² = 0.94). At 37 °C, the signal intensity of C-dots-MBs gradually decreased with time (0 min: 100±0.1 % to 30 min: 90.1±4.6 %) (Fig. 2E), probably due to natural gas diffusion from MBs. Furthermore, the contrast enhancement of C-dots-MBs rapidly decreased upon 0.5 W/cm² US irradiation (pre-US sonication: 46.8±0.2 dB; post-US sonication: 10.8±3.3 dB) (Fig. 2F), suggesting the collapse of C-dots MBs. Because the inertial cavitation of MBs is accompanied by MB collapse, we inferred that 0.5 W/cm² induced inertial cavitation of the C-dots MBs. Taken together, our data showed that we had successfully transformed the C-dots liposomes into C-dots MBs. Furthermore, the C-dots MBs had several US-related applications, including contrast-enhanced imaging, vibration, and destruction.

Next, we evaluated whether the C-dots MBs could generate ROS upon US sonication. The samples were subjected to US with acoustic power of 0–1 W/cm² for 1 min and the distinct behavior of each group is exhibited in Fig. 3A. No fluorescence signals were detected without US sonication, suggesting that neither C-dots, regular-lipid MBs, nor C-dots MBs produced ROS spontaneously (59.0±10 a.u., 62.0±20 a.u., 70.0±25 a.u. for C-dots, regular-lipid MBs, and C-dots MBs respectively). The fluorescence intensities in the regular-lipid MBs group were obviously higher than in the PBS group (0.25 W/cm²: 374.3±149.3 a.u.; 0.5 W/cm²: 1218.3±343.2 a.u.; 0.75 W/cm²: 2689.3±540.3 a.u.; 1 W/cm²: 4844.0±475.2 a.u.). This scenario probably was due to the MBs inertial cavitation allow the cleavage of water molecules or the induction of chemical reactions, then resulting in the generation of ROS [28]. A mixture of C-dots liposomes with regular-lipid MBs produced higher fluorescence intensities than the regular-lipid MBs group, probably because the C-dots liposomes absorbed the inertial cavitation energy emitted from the MBs (0 W/cm²: 65.0±9 a.u.; 0.25 W/cm²: 593.0±49.5 a.u.; 0.5 W/cm²: 1266.0±400.5 a.u.; 0.75 W/cm²: 5039.6±447.3 a.u.; 1 W/cm²: 8852.3±883.5 a.u.). In the C-dots MBs group the fluorescence intensities clearly increased when the applied acoustic energy was greater than 0.5 W/cm² (0.25 W/cm²: 540.3±147.2 a.u.; 0.5 W/cm²: 4255.0±316.1 a.u; 0.75 W/cm²: 8969.7±466.3 a.u; 1 W/cm²: 14876.7±1279.1 a.u), suggesting the profound generation of ROS. These data also indicated that US energy greater than 0.5 W/cm² was required for ROS generation by C-dots MBs.

Fig. 3.

ROS production in different groups. (A) Effect of acoustic power on ROS generation by C-dots MBs with US. (B) Effect of US sonication duration on ROS generation by C-dots MBs with US. (C)-(D) Production of ¹O₂ and HO· measured by EPR. (E) Amount of H₂O₂ produced by C-dots-MBs with different power of US. (F) Identification of the ROS species produced by C-dots MBs with US using ROS scavengers.

We also assessed the effect of US sonication duration (0–3 min; 1 W/cm²). Although fluorescence signals also appeared in the PBS group with US sonication for 1 min, further prolonging the sonication duration did not significantly enhance the fluorescence intensities (1 min: 1033.0±429.5 a.u.; 2 min: 2170.7±650.1 a.u.; 3 min: 4066.0±812.8 a.u) (Fig. 3(B)). In contrast, increasing the sonication duration from 1 min to 3 min resulted in a progressive increase in fluorescence signals in the regular-lipid MBs group (1 min: 1884.0±475.2 a.u; 2 min: 6915.7±892.1 a.u.; 3 min: 8538.3±1400.5 a.u) and the C-dots liposome + regular-lipid MBs group. (1 min: 3383.0±475.2 a.u.; 2 min: 9964.0±649.5; 3 min: 14027.0±2183.5). Under these US parameters, the C-dots MBs group produced the highest fluorescence signal intensity (1 min: 15274.5±579.3 a.u.; 2 min: 21852.6±886.9 a.u.; 3 min: 25675.7±1363.3 a.u.), suggesting that our proposed strategy was effective for the production of ROS.

After verifying the ROS production ability of C-dots MBs with US, we next aimed to identify the ROS species generated using EPR spectra and fluorescence dyes. The EPR spectrum confirmed the appearance of ¹O₂ signal and •OH signal in the US + C-dots MBs group by TEMP dye and DMPO dye, respectively (Fig. 3(C)). We also added AUR dye, a fluorescence probe of H₂O₂, to the C-dots-MBs solution and found that the fluorescence intensity positively correlated with the applied US energy, suggesting the production of H₂O₂ (Fig. 3(D)). We further designed an experiment to identify the generated ROS through the addition of different types of ROS scavengers to the C-dots MBs sample before US activation. Addition of mannitol, a scavenger of •OH, reduced the fluorescence intensity by 33.4% (14853.0±4053.2 a.u. without mannitol vs. 9833.0±1774.1 a.u. with mannitol) (Fig. 3(F)). In comparison, the fluorescence intensities of the US-sonicated samples were significantly reduced to ∼12% after incubation with scavengers of ¹O₂ (1785.3±456.3 a.u. and 2276.91±670.7 a.u. for histidine and NaN₃ respectively). These results showed that the majority of ROS generated by C-dots MBs were ¹O₂, with lower levels of •OH and H₂O₂.

The cytotoxicity of the C-dots MBs was evaluated using the TRAMP cell line. Increasing the concentration of C-dots MBs from 0 to 1.25×10⁶ MBs/mL did not produce obvious cellular death (cell viability 98.8±0.6 % and 91.1±4.9 % respectively) (Fig. 4(A)). However, noticeable cytotoxicity was observed when the concentration of the C-dots MBs was higher than 5×10⁶ MBs/mL (80.7±5.8 %). We thereafter selected the MBs concentration of 1.25×10⁶ MBs/mL for the following cells experiments and animal studies to avoid C-dots MBs induced toxicity. The combination of US and C-dots MBs reduced cell viability to 49.1±5.7 %. In comparison, there was a minor decrease in viability of cells co-incubated with regular lipid MBs with US sonication (73.1±2.3 %) (Fig. 4(B)). This is most likely due to the degree of ROS induced by the regular lipid MBs was lower than that induced by the C-dots MBs. To further assess the role of ROS in the cytotoxicity mediated by C-dots MBs plus US, we incubated the cells with different ROS scavengers before US sonication. Preincubation with mannitol increased cell viability from 51.6±7.2 % to 65.4±3.6 % (Fig. 4(C)), while incubation with histidine further increased cell viability to 73.5±4.5 %, supporting the proposal that the observed cell cytotoxicity was dominated by ¹O₂.

Fig. 4.

Cytotoxicity of C-dots MBs + US in TRAMP cells. (A) Cytotoxicity of the C-dots MBs with different concentrations. (B) Cytotoxicity of different treatment groups. (C) Effect of ROS scavengers on cytotoxicity.

To reveal the mechanism by which C-dots MBs plus US enhanced cell death, we designed several live-cell imaging experiments, including tests for lipid peroxidation (fluorescence lipid peroxidation reagent), intracellular ROS (DCFDA dye), cell apoptosis (PI dye), and cell necrosis (Annexin V dye). In the lipid peroxidation test, the color of fluorescence dye localized on the cell membrane changed from red (∼590 nm) to green (∼510 nm), confirming the occurrence of lipid peroxidation on the cell membrane after treatment with C-dots MBs + US (Fig. 5(A)). Compared with the other groups, treatment with C-dots MBs + US resulted in the highest level of lipid peroxidation (32.5±8.2 %, Fig. 6(A)). Previous studies reported that lipid peroxidation would induce disruption of the cell membrane, leading to overproduction of intracellular ROS and triggering cell apoptosis. Indeed, the intracellular fluorescence intensity of the DCFDA dye significantly increased after treatment with C-dots MBs + US (10.5±2.1 a.u./μm²) compared with regular lipid MBs + US (2.1±0.4 a.u./μm²) or US only (0.6±0.2 a.u./μm²), suggesting that the extracellular activated C-dots MBs induced intracellular ROS production (Fig. 5(B), Fig. 6(B)). The proportion of PI-positive cells (apoptotic cells) was obviously higher than that of Annexin V-positive cells (necrotic cells) (50% vs. 3%) in the C-dots-MBs + US group (Fig. 5(C), Fig. 6(C)). In contrast, treatment with regular lipid MBs + US (2.1±0.4 a.u./μm²) or US only resulted in few apoptotic cells (5.3 % and 15.4 %) and necrotic cells (0.9 % and 7.8 %). Taken together, these data indicated that ROS generated from C-dots MBs + US could induce lipid peroxidation of the cell membrane, elevate intracellular ROS, and trigger cell apoptosis.

Fig. 5.

Cell experiments to determine the mechanism of cell death induced by C-dots MBs + US. (A) Occurrence of lipid peroxidation on the cell membrane. (B) The level of intracellular ROS. (C) Occurrence of cell apoptosis and necrosis indicated by staining with PI dye and Annexin V dye, respectively.

Fig. 6.

Quantification of cell experiments shown in Fig. 5 for different treatments by flow cytometry. (A) Proportion of cells with lipid peroxidation. (B) Quantification of fluorescence intensity for intracellular ROS. (C) Proportion of apoptotic cells and necrotic cells.

The lifetime within in vivo circulation of the C-dots MBs was evaluated by US imaging (Fig. 7(A)). Following bolus intravenous administration of the C-dots MBs, the signal intensity (visualized by positive signals) increased rapidly to 27.3 dB because MB wash-in. The signal intensity then gradually decreased within 400 s probably due to the MBs were eliminated during circulation or the gas inside the MBs diffused into the surrounding blood, indicating that the time window for activating SDT with C-dots-MBs is approximately 400 s. In addition, the US contrast-enhanced imaging detected positive signals within the tumor region (Fig. 7(B)), implying that the C-dots-MBs were able to enter the tumoral microcirculation for subsequent SDT application.

Fig. 7.

In vivo studies. (A) Lifetime of C-dots MBs detected by US imaging (n = 1 in each group). (B) Distribution of C-dots MBs in normal tissue and tumor tissue detected by US imaging (n = 1 in each group). (C) In vivo biodistribution of C-dots MBs with and without US sonication and (D) the corresponding fluorescence intensity (n = 1 in each group). (E) Ex vivo biodistribution of C-dots MBs with and without US sonication and (F) the corresponding fluorescence intensity (n = 3 in each group).

The biodistribution of the C-dots MBs with and without US sonication was detected by in vivo and ex vivo organ fluorescence imaging. The in vivo data showed that injection of C-dots MBs resulted in only weak signals in the tumor area (Fig. 7(C)). After US sonication, the signals of C-dots MBs within the tumor were enhanced 2.5-fold (Fig. 7(D)), suggesting that US-activated C-dots MBs underwent inertial cavitation and released C-dots within the tumor. The ex vivo organ analysis showed that the majority of the C-dots signals were still localized only in the tumor at 2 h after US + C-dots MBs treatment (Fig. 7(E and F)), indicating a low risk of off-target effects.

Finally, we assessed the capability of US + C-dots MBs treatment for increasing intratumor ROS and initiating tumor cell apoptosis by histological examination. In the US + C-dots MBs group, several tissue damages at the tumor peripheral area and core area were observed from H&E staining (Fig. 8(A and B)). In contrast, the tissue morphologies were intact in the control group and C-dots MBs only group. The intratumoral fluorescent intensity of DCFDA staining, and TUNEL staining showed 3-fold and 2.8-fold increase in the US + C-dots MBs group (11.1±0.3 a.u.) compared with the C-dots MBs only group (0.8±0.3 a.u.) and control group (0.3±0.2 a.u.), reflecting C-dots MBs cavitation-induced ROS production within tumors. TUNEL staining detected several cells positive for red fluorescence signal in the US + C-dots-MBs group, confirming that the ROS induced by US + C-dots MBs resulted in tumor cell apoptosis.

Fig. 8.

(A) Histological images of H&E staining, DCFDA staining, and TUNEL staining revealing tissue damage, ROS level, and apoptosis, respectively, after different treatment protocols (n = 1 in each group). (B) Enlargement of ROIs from (A).

4. Discussion

To the best of our knowledge, this is the first demonstration of C-dots mediated SDT using medical US at several MHz frequencies. C-dots have been engineered as promising nanomaterials for several biomedical applications due to their low toxicity, high stability, and excellent solubility [17], [29], [30]. Recent literature has reported that US-induced inertial cavitation could break up oxygen-containing groups on the surface of C-dots to generate ROS [23], [31], leading to increased interest in the development of C-dots for SDT. However, low frequency and high energy US was required to produce inertial cavitation. As an alternative approach, we report the concept of “C-dots assembled microbubbles” in which C-dots are incorporated into MBs as a novel class of sonosensitizer for SDT. First, we showed that the C-dots MBs could generate ROS in the setting of low-intensity and high-frequency US and that the ROS produced during this process included ¹O₂, •OH, and H₂O₂. Furthermore, our in vitro and in vivo experiments showed that the amount of ROS generated was sufficient to induce tumor cell apoptosis. Although we did not examine whether C-dots MBs + US would completely eradicate solid tumors, several apoptotic cells and ROS-overloaded cells were observed after treatment, suggesting the potential for tumor therapy.

We demonstrated that the C-dots MBs group produced more ROS and had higher cytotoxicity against tumor cells in vitro than a co-mixture of C-dots and MBs under the same US parameters of sonication. MBs are gas-filled, lipid or polymer-shelled bubbles of micrometer size and are approved as an US imaging contrast agent [20], [32]. Previous had reported that the MBs would perform stable cavitaion (repeated volumetric oscillations) and inertial cavitation (violent expansion, compression, and collapse) when irradiated by low energy US and high energy US, respectively [33], [34]. Inertial cavitation frequently is accompanied by a significant localized increase in temperature and/or the emission of light (sonoluminescence). Previous studies confirmed that putting sonosensitizers in close proximity to MBs undergoing inertial cavitation could improve their efficiency for ROS generation via pyrolysis-mediated processes or sonoluminescence or [14], [16]. We reasoned that C-dots directly incorporated into the MB shell would absorb the energy of inertial cavitation more effectively than a co-mixture of C-dots and MBs.

ROS production plays an important role in SDT for tumor treatment. It has been reported that three kinds of ROS are produced in SDT: •OH, ¹O₂, and H₂O₂ [12]. The effectiveness of ROS-induced cell damage mainly depends on the half-life of the ROS species. In the physiological environment, •OH radicals are easily scavenged by organic compounds (e.g., intracellular glucose and hydrophobic amino acids) present in the cells or physiological medium due to the high reactivity of •OH [35], resulting in a half-life of only 1 ns. In addition, the diffusion distance of •OH is only 6–9 μm [36], further limiting its cytotoxicity. Although ¹O₂ has a longer half-life of 4 μs and a longer diffusion distance of 200 μm [37], ¹O₂ has a strong oxidizing property and will not react with other substances [38]. Compared with •OH and ¹O₂, H₂O₂ has a longer half-life (>1 ms) and a diffusion distance greater than 1 μm, and can freely penetrate cell membranes [38], [39]. Our data showing an increase in the generation of ¹O2, •OH, and H₂O₂ in the US + C-dots MBs group suggest the potential for tumor treatment.

Although this study has provided an innovative framework to enhance C-dots–mediated SDT, the proposed method has several limitations and further modifications may improve its efficacy. First, the intensity of inertial cavitation of C-dots MBs during US sonication should be estimated to evaluate the relationship between MB inertial cavitation and ROS production. It is rational that a stronger inertial cavitation will yield more ROS; however, the inertial cavitation also leads to tissue damage [40]. Second, the gas core within MBs could potentially be replaced with oxygen. As oxygen is a key substrate for the generation of ROS in SDT, and the core of a solid tumor is frequently characterized as hypoxic, supplying oxygen during SDT can enhance ROS generation and improve the therapeutic efficiency. Third, the shell of the MBs could be conjugated with tumor-targeting ligands/nanoparticles to improve the retention and accumulation of MBs in the tumor vasculature. Increasing the quantity of MBs sonicated at the target site by US may enhance ROS generation and improve the therapeutic outcome. Fourth, the obvious cytotoxicity of C-dots MBs was observed at the concentrations of 5×10⁶ MBs/mL, suggesting that the cytotoxicity of C-dots MBs at the higher concentrations was still needed to investigate. Fifth, the bioluminescence images showed abundant deposition of C-dots within the tumor after US treatment, and the potential application of those C-dots for SDT with US should be evaluated. Performing image guided SDT is a logical extension of this work. Finally, the long-term tumor elimination and biosafety of C-dots MBs + US treatment must be investigated.

5. Conclusions

Our study demonstrated that C-dots assembled MBs can function as a novel sonosensitizer with great potential for SDT in solid tumor treatment. Under 1-MHz US activation, C-dots MBs require only 33% of the US energy used in conventional methods to produce ROS. Compared to regular lipid MBs, C-dots MBs could produce 1.9-fold more ROS, including •OH, ¹O₂, and H₂O₂. Cell experiments showed that the ROS generated by C-dots MBs + 1-MHz US caused peroxidation of cell membranes and induced apoptosis, killing 42.9% of cells. Animal experiments demonstrated that C-dots MBs + 1-MHz US increased intratumor ROS levels and promoted apoptosis of tumor cells. Future work includes the development of functionally modified C-dots MBs and the modulation of therapeutic gases encapsulated within the microbubbles to further enhance the efficacy of cancer therapy, together with their application other therapeutic concerns such as targeted therapy.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.