Characterization of the PGd@tNBs nanoparticles

The technique employed to alter the crystalline phase of P(VDF-TrFE) particles was adapted from prior research [17, 22, 23]. In this studies, thermal annealing induced a transition from the ferroelectric phase to the paraelectric phase within the P(VDF-TrFE) material [24]. When subjected to a temperature of 140 °C, the X-ray diffraction (XRD) peak intensity of rP(VDF-TrFE) particles exhibited a significant increase (Fig. 2A). This enhancement was attributed to lattice expansion, leading to an augmented interchain distance. Consequently, the mobility of C-F dipoles within the crystal was improved, and friction forces were reduced [25, 26]. In the realm of Raman imaging, the peak at 858 cm^-1 corresponded to the β phase, whereas the peak at 808 cm^-1 corresponded to the α phase [27, 28]. Upon the alteration of the crystalline phase, there was a notable increase in the intensity of the β peak. Consequently, there was a substantial elevation in the ratio of β to α peak intensities (from 1.3 to 2.5) (Fig. 2B). The restructuring of copolymer chains due to annealing led to an expansion of the intermolecular spacing. This, in turn, weakened interchain interactions and minimized conformational defects [27, 29]. As a result, there was a pronounced improvement in the stability of molecular chains, as represented by the intensity of the 1433 cm^-1 peak.

Fig. 2

Characterization of PGd@tNBs nanoparticles. (A) XRD results. (B) Raman spectroscopy results. (C) Contact angle measurement of P(VDF-TrFE) nanoparticles and PGd@tNBs. (D) XPS analysis of PGd@tNBs (upper), valence state changes of Gd (below). (E) EDS analysis of PGd@tNBs nanoparticles. (F) SEM of PGd@tNBs nanoparticles. (G) Encapsulation efficiency of PGd@tNBs nanoparticles. (H) Particle size distribution of PGd@tNBs nanoparticles. (I) Zeta potential of PGd@tNBs nanoparticles

Through the surface coating of the P(VDF-TrFE) materia to enhance hydrophilicity [17, 30], a notable reduction in the contact angle of PGd@tNBs particles is achieved, measuring only 29.0° after the enhancement process, while the contact angle of P(VDF-TrFE) is 99.1° (Fig. 2C). The hydrophilicity was increased by adding DSPE-PEG-DOTA-Gd to provide MRI imaging capability and DSPE-PEG-FITC to give fluorescent properties. Consequently, XPS elemental analysis divulges the presence of elements such as Gd, S, N, and O, in addition to C and F elements (Figure S1). Following the introduction of DSPE-PEG-Mal-tHER2 and C6F14, there is no substantial alteration in the elemental composition. However, as the reaction progresses, a discernible chemical shift occurs in the Gd ions within PGd@tNBs particles. This shift is manifested by an increase in their valence electrons from + 3 in PGd NPs particles to + 4 in PGd@tNBs (Fig. 2D). Simultaneously, there is a modification in the valence energy of carbon elements as well (Figure S2). Furthermore, EDS results substantiate the homogeneous distribution of oxygen, nitrogen, and gadolinium elements on the surface of PGd@tNBs particles (Fig. 2E). The PGd@tNBs particles displayed a relatively dispersed solid liposome structure under SEM (Fig. 2F). Further, a more intuitive view of the liposomes located on the surface of the nanoparticles was characterized by TEM (Figure S3).

Upon increasing the mass ratio of PGd NPs particles to the lipid shell from 3:1 to 1:1, the encapsulation efficiency of PGd@tNBs particles reaches approximately 77% (Fig. 2G). The pure lipid shell exhibits a diameter of approximately 110 nm, while the particle size of PGd NPs measures around 150 nm. Following sonication, the size of PGd@tNBs particles is roughly 250 nm (Fig. 2H), and polydispersity index (PDI) of PGd@tNBs particles is 0.223 (Figure S4a). Intriguingly, the absolute value of ζ-potential for PGd@tNBs particles is the highest (Fig. 2I), thus accounting for their stable particle size maintained at approximately 270 nm during storage for 1 to 3 weeks at 4 °C (Figure S4b). Additionally, our separate study has also corroborated that following crystal structure remodeling, the P(VDF-TrFE) particles exhibit stabilized particle sizes of around 200 nanometers, particularly in the DMEM culture medium, owing to their enhanced hydrophilicity [31].

The imaging ability of PGd@tNBs nanoparticles was characterized in vitro

Utilizing the HER2-targeting peptide (sequence: KLRLEWNR) previously chosen by our research team [32] (Figure S5), we employed a click chemistry reaction to facilitate the conjugation of the HER2-targeting peptide with DSPE-PEG-Mal. Subsequently, this conjugate was combined with DPPC and Chol to form the lipid shell (Figure S6). With the introduction of PGd NPs and C6F14, the resultant mixture underwent ultrasonic oscillation, leading to the generation of PGd@tNBs particles (Fig. 3A).

In comparison to conventional contrast agents, nanoscale ultrasound contrast agents offer notable advantages in terms of biocompatibility, serum stability, and extended lifespan. Presently, they find utility across various domains, including musculoskeletal imaging [33], vascular plaque imaging [21, 34], and assessment of ablative therapy [35]. Utilizing Sonovue as the positive control group and double-distilled water as the negative control, our synthesized PGd@tNBs nanoparticles exhibit robust CEUS capabilities (Fig. 3B), with their imaging performance enhanced proportionately as the particle concentration increases (Fig. 3C). Gadolinium-based chelates, renowned for their capability to shorten T1 relaxation times, are widely utilized in MRI imaging [36]. The effectiveness of imaging is closely linked to the concentration of gadolinium ions [36, 37]. In our investigation, although the MRI imaging performance of PGd@tNBs particles with a gadolinium ion concentration of 0.5 mM falls short of that observed with a pure DTPA-Gd solution, it still suffices to generate contrast against the control group (Fig. 3D). The r1 relaxivity of PGd@tNBs were 2.24 mM− 1s− 1. Furthermore, the fluorescence intensity of PGd@tNBs particles demonstrates an upward trend with increasing concentrations (Fig. 3E and F). These findings highlight the potential utility of PGd@tNBs particles as a multimodal contrast agent, such as MRI, ultrasound imaging or fluorescence imaging.

Fig. 3

The imaging ability of PGd@tNBs nanoparticles was characterized in vitro. (A) Schematic representation of the synthesis process of PGd@tNBs nanoparticles. (B) Ultrasound imaging comparison between PGd@tNBs nanoparticles and Sonovue contrast agent. (C) Ultrasound imaging efficiency of PGd@tNBs nanoparticles at different concentrations. (D) MRI imaging capability comparison of PGd@tNBs nanoparticles at different concentrations with DTPA-Gd contrast agent. (E) Fluorescence imaging capability of PGd@tNBs nanoparticles at different concentrations. (F) Quantitative analysis of fluorescence intensity

Multimodal imaging of PGd@tNBs nanoparticles in vivo

Figure 4A depicts the schematic detailing tumor implantation and imaging in mice. Following the intravenous injection of a 200 µL solution containing PGd@tNBs nanoparticles at a concentration of 300 µg/mL, a swift enhancement in the tumor area was observed. Compared to the traditional contrast agent, Sonovue, the PGd@tNBs nanoparticles exhibit a more extensive and prolonged distribution within tumor tissue (Fig. 4B and Figure S7), this is attributed to the fact that nanoscale ultrasonic contrast agents typically possess superior stability, enabling them to circulate in the body for a prolonged duration without being eliminated by the immune system. Moreover, their minute size allows for easier infiltration into the extravascular space of tumor tissue, and potentially accumulation in the tumor tissue through the enhanced permeability and retention (EPR) effect [38]. Furthermore, we have employed HER2-targeted modifications on their surface, which enables these nanoparticles to actively target tumor cells. Subsequently, we executed standard ultrasound scans on the tumor, capturing Color Doppler Flow Imaging (CDFI), Tissue Doppler Imaging (TDI), Power Doppler Imaging (PDI), and Superb Microvascular Imaging (SMI) on the same cross-section of the tumor (Fig. 4C). Notably, SMI yielded a 100% success rate in its visualization (Fig. 4D). This achievement can be attributed to SMI’s specialized algorithm, adept at effectively discerning microvascular flow from motion artifacts and other noise artifacts, thus enabling users to seamlessly visualize the microvascular system [39].

Fig. 4

Multimodal imaging of PGd@tNBs nanoparticles in vivo. (A) Schematic representation of Balb/c mice tumor induction and imaging. (B) Ultrasound imaging of PGd@tNBs nanoparticles. (C) Ultrasound blood flow imaging and ultrasound elastography of mouse tumors. (D) Ultrasound blood flow imaging display rate. (E) In vivo fluorescence imaging of PGd@tNBs nanoparticles. (F) Distribution of PGd@tNBs nanoparticles in organs at different times. (G) Quantitative assessment of fluorescence intensity in mouse organs. (H) MRI of PGd@tNBs nanoparticles. (I) Quantitative analysis of T1 relaxation times in MRI imaging

Upon intravenous injection of the PGd@tNBs solution, robust fluorescence signals were observed to accumulate within the tumor area after 2 h, even after 12 h, a faint fluorescent signal can still be observed at the tumor tissue (Fig. 4E). It is worth noting that, 2 h after injection of the PGd@tNBs solution, the fluorescent signals in the liver, lungs, and kidneys further intensified. After 12 h, a slight residual fluorescence signal was still detected in the kidneys of the mice, suggesting that the PGd@tNBs particles were likely metabolized via urine excretion (Fig. 4F). Throughout the entire observation period (12 h), the fluorescent signal in the kidneys remained relatively high (Fig. 4G). After the intravenous administration of PGd@tNBs, enhancement within the vicinity of the tumor area commenced at 1 h, intensified at 3 h, began to diminish at 6 h, and by the 24-hour mark, the tumor tissue enhancement had entirely subsided, causing the tumor T1 relaxation time to revert to a level akin to that preceding injection (Fig. 4H). Conversely, the tumor tissue T1 relaxation time within the PBS group exhibited no discernible alteration throughout (Fig. 4I).

Effect of PGd@tNBs particles on membrane potential upon ultrasound stimulation

Both PGd@tNBs and non-targeted PGd@NBs nanoparticles were separately incubated with SK-BR3 cells for a duration of 3 h. A pronounced dispersion of PGd@tNBs green fluorescence was observed surrounding the cancer cells, whereas PGd@NBs particles were relatively sparse (Fig. 5A and B). Furthermore, we turned to SEM to visually characterize the interaction between nanoparticles and cells. The images clearly indicated a significantly higher count of PGd@tNBs nanoparticles binding to SK-BR3 cells compared to PGd@NBs nanoparticles (Fig. 5C). Notably, the uptake of PGd@tNBs nanoparticles to SK-BR3 cells escalated as the incubation duration was extended within a short timeframe (≤ 5 h) by flow cytometry (Fig. 5D). The phase hysteresis loop and amplitude hysteresis loop serve as the most straightforward indicators of the piezoelectric response. Upon applying an external electric field of 10 V, there is a noticeable increase in the surface potential of rP(VDF-TrFE), this increase is accompanied by a more pronounced variation in amplitude, and the particle surface morphology becomes more regular (Fig. 5E). Through the estimation of the maximum amplitude of the piezoelectric signal, the effective piezoelectric coefficient d33 of rP(VDF-TrFE) is approximately − 8.8 pC/N (Figure S8). Following this, we explored the heightened cytotoxic effect of nanoparticles with crystal restructuring under ultrasound stimulation on SK-BR3 cells. This effect was observed under conditions where individual ultrasound stimulation or nanoparticles alone had no impact on cell viability (Fig. 5F and Figures S9–S12). Continuing, we noted a decrease in the fluorescence signal intensity of the cell membrane under ultrasound exposure (Fig. 5H), indirectly indicating damage to the cell membrane potential (Fig. 5G and Figures S13). In the presence of ultrasound, PGd@tNBs particles have the capacity to elevate the concentration of intracellular free Ca2+ within cancer cells (Fig. 5I).

Fig. 5

Effect of PGd@tNBs particles on membrane potential upon ultrasound stimulation. (A) CLSM characterization of PGd@tNBs nanoparticle targeting to cell membranes (Scale bar, 20 μm). (B) Quantitative analysis of red cell membrane and green PGd@tNBs nanoparticle fluorescence. (C) SEM characterization of PGd@tNBs and PGd@NBs binding to SK-BR3 cells. (D) Uptake of PGd@tNBs nanoparticles by SK-BR3 cells within a short period (≤ 5 h). (E) Hysteresis loop, polarization-field loop, and surface potential of rP(VDF-TrFE) and P(VDF-TrFE). (F) Effects of ultrasound stimulation alone or PGd@tNBs nanoparticles alone and PGd@tNBs nanoparticles under ultrasound irradiation on the activity of SK-BR3 cells (n = 3). (G) Measurement of cell membrane potential using FluoVolt™ dye (Scale 3 μm). (H) Quantitative analysis of cell membrane fluorescence intensity. (I) Intracellular free calcium ion characterization using Fluo-3AM calcium ion probe (Scale 10 μm). (J) Impact of calcium ion channel blockers and sodium ion channel blockers on cell viability (n = 3). (K) CLSM characterization of mitochondrial membrane potential (Scale 5 μm). (L) Quantitative analysis of mitochondrial membrane potential

Increased levels of intracellular Ca2+ can precipitate alterations in cell membrane potential, disrupt cell homeostasis, exert influence over cell proliferation, gene expression, ROS generation, and even trigger mitochondrial autophagy [40,41,42]. Interestingly, even in the presence of both calcium ion channel blockers and sodium ion channel blockers, cell viability was still significantly impaired. This suggests that PGd@tNBs nanoparticles, under ultrasound exposure, induce electroporation in cells, thereby promoting the influx of calcium ions (Fig. 5J). Distinct treatments can elicit changes in mitochondrial membrane potential, a phenomenon that can be visualized through the fluorescence signal emitted by the JC-1 dye [43]. In the control group, a substantial mitochondrial membrane potential was evident, causing the JC-1 dye to aggregate and emit red fluorescence. Conversely, the PGd@tNBs + US group experienced notable damage to the mitochondrial membrane potential, leading the JC-1 dye to exist in its monomeric form and emit green fluorescence (Fig. 5K). Subsequently, flow cytometry was employed to corroborate the alterations in membrane potential across the control group, solitary PGd@tNBs treatment, and PGd@tNBs + US treatment conditions (Fig. 5L).

The piezoelectric effect of PGd@tNBs nanoparticles boost SDT

The fundamental principle underpinning SDT revolves around the generation of ROS and their consequential biological effects [44]. Notably, piezoelectric materials can amplify ROS production under ultrasound, thereby enhancing the therapeutic effectiveness of SDT [10, 44,45,46]. Hydroxyl radicals (·OH) constitute the most reactive entities among ROS components, characterized by an extraordinarily high reaction rate constant of 10^9 with neighboring molecules upon their formation, thereby inducing cellular damage [47]. Ultrasound-triggered titanium dioxide generates ·OH and augments Fenton-like reactions [48]. Under ultrasound excitation, an increase in the concentration of PGd@tNBs nanoparticles significantly enhances the generation of ROS and ·OH (Fig. 6A). Under ultrasound exposure, PGd@tNBs nanoparticles induce surface charges, oxidizing GSH into GSSH and thus diminishing the antioxidant capacity of the solution. This phenomenon mirrors ultrasound-activated barium titanate, which generates charges to enhance enzyme activity and expedite GSH consumption [49]. In alignment with these findings, the concentration of GSH decreased proportionally with increasing concentrations of PGd@tNBs particles (Figure S14).

To characterize the sonodynamic therapeutic effects of PGd@tNBs nanoparticles on SK-BR3 cells, we assessed the time point at which these nanoparticles entered the cells. At 24 h of co-incubation between the nanoparticles and cells, a noticeable fluorescence co-localization phenomenon occurred (Fig. 6B). This indicates that a significant number of nanoparticles penetrated the cells within 24 h (Figures S15 and S16). Additionally, with the extension of co-incubation time, the quantity of nanoparticles entering the cells increased (Fig. 6C). Therefore, in the subsequent characterization of SDT efficacy, a series of experiments was conducted after co-incubation of nanoparticles with cells for 24 h.

Here, we define the group receiving ultrasound stimulation after the binding of nanoparticles to cells as PGd@tNBs + US1. The group receiving ultrasound stimulation after nanoparticles enter the cells is defined as PGd@tNBs + US2. The group where nanoparticles bind to cells, enter the cells, and receive repeated ultrasound stimulation is defined as PGd@tNBs + US1 + 2.

Mild stimulation can induce the generation of intracellular ROS (Fig. 6D), while PGd@tNBs nanoparticles further increase the generation of intracellular ROS under ultrasound (Fig. 6E). Interestingly, the cell activity of the PGd@tNBs + US1 + 2 group was the lowest, and the decrease in activity was most significant with time (Figure S17). At the same time, the relative generation of ROS showed a linear increase with time (Figure S18 and Figure S19). Subsequently, the PGd@tNBs + US1 + 2 group demonstrated the most substantial decrease in GSH content, a difference that was statistically significant when compared to both the PGd@tNBs + US1 and PGd@tNBs + US2 groups (Fig. 6F and Figure S20). Importantly, cellular constituents were extracted for analysis. The superoxide dismutase (SOD) content and reduced GSH content of the different treatment groups exhibited distinct degrees of reduction (Fig. 6G). Significantly, the PGd@tNBs + US1 + 2 group exhibited notably higher efficacy in SDT compared to both the PGd@tNBs + US1 and PGd@tNBs + US2 groups. Observing through Calcein-AM/PI staining, it was noted that the PGd@tNBs + US1 + 2 group exhibited a higher number of dead cells (Fig. 6H), and the cell viability in the PGd@tNBs + US1 + 2 group was lower compared to the PGd@tNBs + US1 and PGd@tNBs + US2 groups (Fig. 6I and Figure S21).

Fig. 6

The piezoelectric effect of PGd@tNBs nanoparticles assists SDT. (A) The in vitro SDT efficacy of PGd@tNBs nanoparticles was characterized, the assessment included the generation ability of ROS and ·OH, as well as the consumption of glutathione. (B) Lysosome co-localization experiment of PGd@tNBs nanoparticles entering cells (Scale 5 μm). (C) Quantitative assessment of PGd@tNBs nanoparticle content within cells over 6–72 h. (D) Characterization of intracellular reactive oxygen species using DCFH-DA probe (Scale 5 μm). (E) Quantitative analysis of intracellular reactive oxygen species (n = 3). (F) Quantitative analysis of intracellular GSH content (n = 3). (G) UV-visible light characterization of intracellular SOD content. (H) Characterization of cell viability using Calcein-AM/PI staining (Scale 10 μm). (I) Quantitative analysis of cell viability after different treatments (n = 3). (J) Assessment of cell apoptosis after different treatments. (K) Quantitative analysis of cell apoptosis. (L) Cell count through Transwell chambers after different treatments (Scale 5 μm). (M) Quantitative analysis of cell count passing through Transwell chambers (n = 3). (N) Wound healing assay after different treatments (Scale 20 μm). (O) Quantitative assessment of wound healing after different treatments (n = 3). (P) SEM characterization of PGd@NBs nanoparticles binding to cells and subjected to ultrasound stimulation

Solely repeating ultrasound stimulation prompted apoptosis in cells, with a specific emphasis on early apoptosis (Fig. 6J). When PGd@tNBs were either adhered to the cell membrane or had penetrated the cells before ultrasound stimulation, approximately 35% of cells underwent apoptosis. After a 24-hour cycle of repeated ultrasound stimulation, roughly 80% of cells underwent apoptosis, reflecting a substantial 45% escalation in the apoptosis rate (Fig. 6K). Concurrently, PGd@tNBs particles exhibited the ability to curb the invasive potential (Fig. 6L) and migratory capacity (Fig. 6M) of SK-BR3 breast cancer cells under ultrasound stimulation. At the 48-hour mark, the count of cells that traversed the Transwell chamber in the PGd@tNBs + US1 and PGd@tNBs + US2 groups approximated 50 cells, whereas the PGd@tNBs + US1 + 2 group exhibited a mere 15 cells (Fig. 6N). Moreover, the migration rate of the PGd@tNBs + US1 + 2 group was only 12.48 ± 1.11%, notably lower than that observed in the PGd@tNBs + US1 and PGd@tNBs + US2 groups, signifying a significant statistical distinction (Fig. 6O). Even after the ultrasound treatment, the distribution of PGd@tNBs nanoparticles on the cell surface remained more discernible than that of PGd@NBs nanoparticles (Fig. 6P and Figure S22).

Anti-tumor therapeutic efficacy of PGd@NBs nanoparticles in vivo

Confirmed at the cellular level, based on the effectiveness of PGd@tNBs nanoparticle therapy, we conducted animal experiments to further verify. Figure 7A elucidates the schematic delineating the process from mouse treatment to euthanization. As the treatment regimen progressed, the PGd@tNBs + US group displayed slower tumor volume growth in comparison to the control group or the groups treated with ultrasound or PGd@tNBs nanoparticle independently. These distinctions were statistically significant (Fig. 7D). Furthermore, the tumor weight within the PGd@tNBs + US group was notably lower than that observed in the control group (Fig. 7E and F). Upon contrasting the ultrasound parameters of tumor tissues before and after treatment, it became evident that within the control group, the tumor tissue exhibited a significant increase in size and augmented blood flow on the two-dimensional images (Fig. 7G). In contrast, within the PGd@tNBs + US group, the tumor tissue’s size remained relatively unchanged and displayed a substantial reduction in blood flow signals (Fig. 7H). Collectively, the display rate of blood flow within the tumor tissue diminished (Fig. 7I). This reduction could potentially be attributed to the occlusion of tumor blood vessels due to inflammation post-treatment, consequently leading to a decreased nutrient supply to the tumor cells.

Fig. 7

Anti-tumor therapeutic efficacy of PGd@NBs nanoparticles in vivo. (A) Schematic representation of Balb/c mice treatment and imaging. (B) Changes in body weight of mice in different groups. (C) Liver function results of mice in different groups. (D) Changes in tumor growth in different groups of mice. (E) Macroscopic images of tumor tissues from different groups of mice. (F) Tumor weights of mice in different groups. (G) and (H) Ultrasound blood flow imaging and ultrasound elastography of tumor tissues before and after treatment. (I) Ultrasound blood flow imaging display rate of mouse tumors. (J) Young’s modulus values before and after treatment. (K) Ultrasound elastography scores before and after treatment. (L) MRI imaging showing changes in tumors before and after treatment. (M) The ROS levels, Ki-67 proliferation, apoptotic status, distribution of PGd@tNBs within tumors, and H&E staining of tumor tissues after different treatments (Scale 20 μm)

Post-treatment, the tumor tissue within the PGd@tNBs + US group demonstrated a remarkable elevation in hardness, with increase in elasticity score was particularly marked, displaying a significant divergence from the control group post-treatment (Fig. 7J). The Young’s modulus value manifesting a statistically significant distinction from that observed pre-treatment (Fig. 7K). The alterations in the tumor tissue post-treatment were visually captured through MRI (Fig. 7L). While individual treatments involving ultrasound stimulation or PGd@tNBs particles in isolation contributed to elevated levels of ROS within the tumor tissue, it was within the PGd@tNBs + US group that the highest ROS content was observed. Furthermore, PGd@tNBs particles exhibited proficient accumulation within the tumor region, characterized by relatively uniform distribution. Throughout the entire course of treatment, the mice’s body weight exhibited no significant fluctuations (Fig. 7B). The renal functions of the mice were maintained within the normal range (Table S1). With regard to liver function indicators, both AST and ALT levels were elevated beyond the normal range in the PGd@tNBs + US group (Fig. 7C). This phenomenon can be attributed to the primary metabolism of PGd@tNBs occurring within the liver and kidneys, leading to some degree of liver involvement. On the 1st, 3rd, and 5th days of the treatment regimen, blood routine tests were conducted through retroorbital venous blood collection. Notably, the PGd@tNBs + US group displayed an upsurge in lymphocyte and red blood cell count on the 5th day, concurrently with a decrease in mean corpuscular hemoglobin concentration (MCHC) and mean corpuscular hemoglobin (MCH)