Sonodynamic therapy (SDT) involves the use of sonosensitizers that only become cytotoxic upon exposure to ultrasound (US) (Yumita et al., 1989; Dolmans et al., 2003; Rengeng et al., 2017). SDT has displayed impressive anticancer effects (Rosenthal et al., 2004; Trendowski, 2014; Wood and Sehgal, 2015; Wan et al., 2016) and inhibition of atherosclerotic plaque progression (Sun et al., 2019) in various studies. Capitalizing on the deep penetration into biological tissues noninvasively of US, the superiority of SDT in treatment emerges from the ability to locally activate preloaded sonosensitizers, which have little toxicity to normal tissues, on lesions buried deeply in tissues.

Pulsed low-intensity ultrasound (PLIU), with an acoustic intensity under 3.0 W/cm², has minimum thermal and mechanical effects on human tissues. Around the frequency of 1 megahertz, low-intensity US is capable of generating acoustic cavitation and therefore apparent sonochemical activities (Brotchie et al., 2009), which makes it recommended to be employed in SDT (Jiang et al., 2019). With appropriate sonication apparatus, a potent sonosensitizer is indispensable for the optimization of SDT efficacy. As the most concerned type of sensitizer, porphyrin and its derivatives valid their ability for selective accumulation inside disease cells and little toxicity when applied to both SDT and photodynamic therapy (PDT) (Chen et al., 2014). Numerous novel porphyrin-related sonosensitizers (Kuroki et al., 2007; Tsuru et al., 2012), including metal-based porphyrin (Giuntini et al., 2018) and porphyrin-based nanoparticles (Qian et al., 2016; Canavese et al., 2018) have been synthesized and proven with enhanced cell toxicity under US.

Diverse forms of cytotoxic mechanisms that predominately contribute to the cytotoxicity of porphyrin-mediated SDT have been reported (McHale et al., 2016; Choi et al., 2020). Generating excessive amounts of reactive oxygen species (ROS) that could exert lethal oxidative damage to cellular organs has been considered the most effective way to achieve proper cytotoxicity (Wan et al., 2016). Although ROS always plays a harmful role, moderate levels of ROS are pivotal for normal cellular life (Holmström and Finkel, 2014). Even if certain levels of ROS could promote tumor progression, the tumor cells are more fragile when facing oxidative disorder. When the oxidative balance tilts to excessive levels of ROS, it can still be detrimental to cancer cells (Perillo et al., 2020). Cell membranes and membranous organelles such as mitochondria are particularly vulnerable to ROS due to lipid peroxidation, which could end up with cell apoptosis (Kiang et al., 2012).

Massive efforts have been made to elucidate the relationship between the structure of sonosensitizer and its sonosensitivity and to discover more effective sonosensitizers. It has been reported that the sono-cytotoxicity of metal porphyrin derivatives varied according to the metal moiety in the macrocycle. It seems that the effectiveness of sonosensitizers was closely related to the photosensitivity of the porphyrin macrocycle’s electronic structure to sonoluminescence. However, the light intensity of sonoluminescence induced by the low-intensity US or the therapeutic US is serval orders lower than lasers in PDT applications (Beguin et al., 2019). As a result, this strategy could hardly be valid for PLIU based SDT applications. On the other hand, there were speculations that the outcome might depend on the side-chain structure rather than the macrocycle (Tsuru et al., 2012). The synergistic effect of US and drugs on cells was greatly influenced by the amphiphilicity of the side chain molecules (Dougherty, 1987; Miyoshi et al., 2003; Kale et al., 2017). Several results showed that the cytotoxicity of agents that are devoid of photosensitivity could still be enhanced by US (Hachimine et al., 2007).

With this in mind, we suppose that the tumor cell-killing efficacy of SDT could be enhanced by improving the sonochemical reactivity of the side chains of porphyrin sonosensitizers. In this study, we have introduced and tested a potent novel sonosensitizer (a porphyrin derivative)-BBTPP [butyl 3-(3-(3-(2-(2-hydroxyethoxy)ethoxy)-3-oxopropyl)-12,17-bis(1-(2-(2-hydroxyethoxy)ethoxy)ethyl)-2,8,13,18-tetramethylporphyrin-7-yl)propanoate]. In the side-chain structure of BBTPP, we have introduced multiple ether groups. The ether groups could be easily oxidized due to elevated temperature conditions that shall be formed during ultrasonic cavitation and generate peroxyl radicals. The cell-killing efficacy of BBTPP under the irradiation of pulsed low-intensity focused ultrasound (PLIFU) was evaluated and compared with hematoporphyrin monomethyl ether (HMME), which had a similar porphyrin ring structure but different side chains. Possible mechanisms of BBTPP-mediated SDT were also evaluated.

Our results showed that BBTPP could be effectively accumulated in PC-9 cells. In the presence of 4 μg/ml BBTPP, PLIFU provoked apparent apoptosis to PC-9 cells, while the cytotoxicity was not obvious when PLIFU was acting alone. HMME has been explored as a potent sonosensitizer in serval reports. The apoptosis rate of U937 cells under sonication (1.1 MHz, 1 W/cm², continuous focused ultrasound) raised from 15.8 to 35.6% with the aid of HMME (Su et al., 2013). Less than 15% apoptotic rate in QBC939 cells induced by (1.2 MHz, 1 W/cm², PLIU) in the presence of a wide dose range of HMME was also reported (Liang et al., 2013). Under our US apparatus, the sono-toxicity of BBTPP is significantly more effective than HMME. In our study, the cell death induced by SDT was mainly early apoptosis rather than late apoptosis and necrosis, which could be mainly attributed to that low-intensity ultrasound were more likely to cause early apoptosis (Guo et al., 2019; Zhang et al., 2021). The ratio of necrosis in cells increased with raising ultrasound intensity (Ferril et al., 2005). In addition, ultrasound in continuous mode or in tone burst mode with higher intensity or duty cycle was more prone to induce late apoptosis and necrosis (Su et al., 2013; Shen et al., 2020). Another study showed that PLIU related SDT mainly induced apoptosis, while high intensity induced necrosis (Lv et al., 2017). Besides ultrasonic apparatus, the time of incubation after treatment might also be responsible. Our apoptotic detection was immediately prepared after the completion of all treatments. Studies with 6 h incubation period after SDT showed a higher percentage of late apoptosis and necrosis (Sun et al., 2015a; Sun et al., 2015b).

We need to make clear that the overall apoptotic rate in our study had been greatly limited by the volume of the focal spot of the PLIFU system. The volume of the focal region is about 6.9 mm³ in our PLIFU apparatus. Thus, only a small percentage of cell suspensions could be irradiated. In fact, a substantial percentage of necrosis cells were observed in flow cytometry after we had employed a plane wave US to seek better coverage of US exposure (data not shown). In the meanwhile, the consistency of the results was drastically reduced as well. This situation might be attributed to the formation of standing waves which could cause more severe cavitation damage to tumor cells.

The only structural difference between BBTPP and HMME was that the BBTPP molecule contains a bigger number of ether groups. It is reasonable to deduce that the superior sono-toxicity of BBTPP is contributed by the active sites in the side chains. Ultrasonic cavitation is known to induce localized high-temperature environments-near cavitation cavities which are called “hotspots” (Suslick et al., 2002). At this moment, the cavity interface can be viewed as suitable chemical reactors for the oxidation of ether groups where they could be easily oxidized and then converted to peroxyl radicals. It is worth noting that the ether groups process much lower oxidative stability which led to several orders higher reaction rate than hydrocarbons at high-temperature conditions (Burke et al., 2015; Rodriguez et al., 2015). Porphyrin-based peroxyl radicals have long lifetimes compared to most other radical species and hence are able to induce oxidative damage to both surface and intracellular organelles (Misík and Riesz, 2000).

Possible participation of peroxyl radicals was also suggested as evidenced by a significant increase of both intracellular ROS and lipid peroxidation levels of BBTPP-SDT treated cells. ROS acts as a double-edged sword for cancer life: at moderated levels, ROS could promote cancer progression; at high levels, ROS could trigger programmed cell death of cancer cells (Perillo et al., 2020). Multiple pathways have been involved in ROS-induced cell death. Overproduction of ROS could cause direct damage to lipids, proteins, and nucleic acids, leading to cell death (Uttara et al., 2009; Liu et al., 2018). Various studies demonstrated that the excessive production of ROS in SDT plays an important part in killing cancer cells (Chen et al., 2014; Giuntini et al., 2018).

Lipid peroxidation of biological membranes can disturb the physical properties of lipid bilayers, which inevitably will alter lipid-lipid, and lipid-protein interactions, ion gradients, membrane fluidity and permeability, membrane-initiated signaling pathways (Gaschler and Stockwell, 2017). To gain further insight into the cytotoxic mechanism of BBTPP-SDT, specific signaling pathways involved in apoptosis were evaluated by western blot. There are two main apoptotic pathways: the extrinsic and intrinsic pathways (Elmore, 2007). Regulation of apoptosis is controlled by two classes of caspases: initiator caspases including caspase-8 and caspase-9, and effector caspases including caspase-3. Caspase-8 activation is related to the extrinsic apoptotic pathway, while caspase-9 is activated in the intrinsic pathway (Olsson and Zhivotovsky, 2011). Peroxidation of membrane lipids could cause cell apoptosis through both the extrinsic and intrinsic apoptotic pathways (Gaschler and Stockwell, 2017; Elkin et al., 2018). A classic feature of apoptotic cells is the dissipation of membrane lipid asymmetry and exposure of phosphatidylserine on the outer surface of the plasma membrane. The oxidized lipid could influence the membrane function (Volinsky and Kinnunen, 2013) and could depolarize mitochondria and disturb permeability by changing the expression of pro-apoptotic and anti-apoptotic proteins. Both Bcl-2 and Bax proteins belong to the Bcl-2 family, the former protein is anti-apoptotic and the latter one is pro-apoptotic pore former (Kale et al., 2017). In our study, the intrinsic pathway was activated through downing regulating Bcl-2 and releasing Bax, which increased mitochondrial membrane permeability ultimately causing the activation of caspase-9. The extrinsic pathway is activated through caspase-8. Both caspase-9 and caspase-8 then activate caspase-3, which executed apoptosis.

PDT, as a minimally invasive therapeutic modality, has been used for the management of a variety of diseases. However, this treatment could only apply to superficial lesions for the lacking penetration of light, which could be compensated by US. To improve the efficacy of SDT, optimizing US apparatus and sonosensitizers is equally important. For acquiring more effective sonosensitizers, in recent years, serval strategies associated with the photosensitivity of porphyrin macrocycles had been adopted (Qian et al., 2016; F et al., 2018; Pan et al., 2018). These strategies rely on the participation of sonoluminescence which only occurs in the most violent cavitation bubbles. It is well known that the structure and elastic properties of cells are widely different. Efforts even have been made to selectively damage tumor cells by means of the properties difference between tumor cells and normal cells (Mittelstein et al., 2020). Theories have been developed to describe the mechanism for bioeffects of therapeutic US (Zinin et al., 2005; Krasovitski et al., 2011). For instance, an intramembrane cavitation theory suggested that the cavitation cavities may be surrounded by elastic lipid membranes (Krasovitski et al., 2011). Meanwhile, there is little evidence to report the observation of intracellular sonoluminescence in the absence of artificial bubbles. We believe that this might be attributed to the viscoelasticity of the cellular components which greatly impairs the resonance strength of the intracellular cavities. Our in vitro data presented here indicate the importance of the side-chain structure of porphyrins on the US-induced ROS generation as well as the sono-toxicity of PC-9 cells. This method of improving sonosensitivity does not rely on sonoluminescence, so we believe that it is a promising approach to enhance the efficacy of low-intensity ultrasound-mediated SDT. There are limitations in our study. We only validate our approach in vitro. A systematic and comprehensive study containing in vivo evaluations can provide more detailed information. Furthermore, chemical analysis on the structure change of sonosensitizer in this study should be carried out in the future to further elucidate the reaction dynamics of the side chain.

Our results indicate that BBTPP could accumulate in PC-9 cells after incubation. The sono-toxicity on PC-9 cells of BBTPP is more effective than HMME under the irradiation of PLIFU suggesting a pivotal role of the side-chain structure. Excessive intracellular ROS production and subsequent lipid peroxidation were induced by BBTPP-mediated SDT. The apoptosis of PC-9 was followed, and both extrinsic and intrinsic apoptotic pathways were involved. Our study indicates that, under PLIFU, BBTPP is valuable as a sonosensitizer in PC-9 cells.