New insights on the role of ROS in the mechanisms of sonoporation-mediated gene delivery

Highlights

• Roles of ROS are investigated using molecular dynamics simulations.

• ROS can form inside the MBs under sonoporation conditions.

• ROS could diffuse through the MB shell toward the surrounding aqueous phase.

• ROS participate in the membrane permeabilization.

• Experimental data confirm the molecular dynamics simulation predictions.

Abstract

Reactive oxygen species (ROS) are hypothesized to play a role in the sonoporation mechanisms. Nevertheless, the acoustical phenomenon behind the ROS production as well as the exact mechanisms of ROS action involved in the increased cell membrane permeability are still not fully understood. Therefore, we investigated the key processes occurring at the molecular level in and around microbubbles subjected to ultrasound using computational chemistry methods. To confirm the molecular simulation predictions, we measured the ROS production by exposing SonoVue® microbubbles (MBs) to ultrasound using biological assays. To investigate the role of ROS in cell membrane permeabilization, cells were subjected to ultrasound in presence of MBs and plasmid encoding reporter gene, and the transfection level was assessed using flow cytometry. The molecular simulations showed that under sonoporation conditions, ROS can form inside the MBs. These radicals could easily diffuse through the MB shell toward the surrounding aqueous phase and participate in the permeabilization of nearby cell membranes. Experimental data confirmed that MBs favor spontaneous formation of a host of free radicals where HO was the main ROS species after US exposure. The presence of ROS scavengers/inhibitors during the sonoporation process decreased both the production of ROS and the subsequent transfection level without significant loss of cell viability. In conclusion, the exposure of MBs to ultrasound might be the origin of chemical effects, which play a role in the cell membrane permeabilization and in the in vitro gene delivery when generated in its proximity.

Introduction

The combination of high frequency ultrasound (1–10 MHz) and ultrasound contrast agents (i.e., consisting of gas microbubbles) was introduced as a promising method for improving the therapeutic efficacy of drug by increasing their extravasation and local delivery, while minimizing side effects to healthy tissues [1]. In recent years, research in the field of microbubble-assisted ultrasound (also known as sonoporation or sonopermeabilization) aimed at delivering therapeutic molecules including nucleic acids, anti-cancer drugs, peptides and antibodies in vitro and in vivo has grown rapidly [2], [3], [4].

This method involves ultrasound-induced oscillations of gas-filled microbubbles (MBs) in the vicinity of the biological barriers (e.g., plasma membrane, endothelial barrier, blood–brain barrier) in order to transiently increase their permeability and enhance the local delivery of therapeutic molecules across these barriers in the targeted region [5]. To this effect, MBs are mixed with cells in vitro or administered in vivo intravascularly or directly into the target tissue. MB behavior in an ultrasound field has been broadly investigated, which resulted in better understanding and subsequent control of the induced bio-effects that can be exploited for drug delivery [6]. The response of MBs to ultrasound waves depends on the acoustic parameters (i.e., center frequency, acoustic pressure, duty cycle, total exposure time). Briefly, MBs stably oscillate over time upon exposure to a low mechanical index (MI < 0.2); this process is called stable cavitation. These oscillations generate fluid flows surrounding the MBs, known as acoustic microstreaming. In close contact with cells, microstreaming leads to shear stress on the cell membrane or endothelial barriers resulting in an enhancement of their native permeability to therapeutic molecules [7], [8]. At higher mechanical indices (MI > 0.2), MBs oscillate vigorously, leading to their violent collapse and destruction; this process is termed inertial cavitation. This phenomenon might be accompanied by production of shock waves, which results in greater shear stress on the cells or endothelial barriers in close proximity, leading to permeabilization of these barriers [9], [10]. The ultrasound-induced collapse of the MB may also cause the formation of high velocity jets that puncture biological barriers and thereby entail a greater permeability [11], [12]. Both cavitation regimes are exploited to transiently increase the native permeability of biological barriers (e.g., plasma membrane, endothelial barrier, blood–brain barrier) through the generation of membrane pores and/or the stimulation of endocytosis, transcellular and paracellular pathways, thus enhancing the extravasation and the intracellular uptake of therapeutic molecules [13].

Although most of the investigations suggested the major role of mechanical forces (e.g., acoustic microstreaming, shock waves, microjets) in the transient permeabilization of biological barriers, it seems however that other phenomena, including the production of reactive oxygen species (ROS), could also play a key role in the sonoporation mechanisms [14], [15], [16], [17], [18], [19], [20]. Indeed, Juffermans et al. investigated whether hydrogen peroxide (H₂O₂) is involved in the transient permeabilization of rat H9c2 cardiomyoblasts in vitro after ultrasound exposure at low diagnostic power (1.8 MHz, MI 0.1 or 0.5, during 10 s), in the presence of stable oscillating SonoVue® microbubbles, by measuring the generation of H₂O₂ and intracellular calcium influx (i.e., assessment of membrane permeabilization) [15]. They reported that MB-assisted ultrasound caused a significant increase in intracellular H₂O₂ level at both MI 0.1 and 0.5. Furthermore, they reported an increase in intracellular calcium levels at both MI in the presence of MBs, which was not detected in the absence of extracellular calcium. In addition, this calcium influx was fully inhibited at MI 0.1 and partially reduced at MI 0.5 in the presence of catalase (i.e., H₂O₂ scavenger). The authors hypothesized that microbubbles exposed to low MI ultrasound (i.e., no violent collapse of the microbubbles) induced the generation of H₂O₂ , which increased the native permeability of cell membranes to calcium ions through the formation of transient membrane nanopores [15], [21]. In agreement with previous studies [22], [23], Juffermans et al. described that acoustically-mediated calcium influx led to activation of calcium-dependent potassium channels, known as BK_Ca channels, and to a subsequent local hyperpolarization of the H9c2 cell membrane [14]. This local hyperpolarization of the cell membrane may facilitate the uptake of macromolecules through endocytosis and macropinocytosis [21]. In agreement with their in vitro studies carried out on the rat H9c2 cardiomyoblasts, Juffermans et al. demonstrated that MB-assisted ultrasound increased membrane permeability of endothelial cells to calcium ions, with an major role of H₂O₂ [16]. Further changes in ROS homeostasis involved an increase in intracellular H₂O₂ levels and protein nitrosylation as well as a decrease in total endogenous glutathione level. Furthermore, MB-assisted ultrasound significantly disturbed endothelial monolayer integrity through F-actin cytoskeletal rearrangement. These disrupted intercellular interactions were restored within 30 min, without affecting the cell viability [16]. Despite the experimental evidences, the role of ROS in the acoustically-mediated membrane permeabilization is still subject to debate. The generation of ROS and their roles in the sonoporation process has been only investigated from a biological point of view. Indeed, the consequences of the interaction between MBs and ultrasound in the production of ROS and their bioeffects on cell membranes are not known yet.

The objective of the present study is to elucidate the origins and the role of ROS in the membrane permeabilization using MB-assisted ultrasound. We harnessed the capacities of computational chemistry methods to decipher the key processes occurring at the molecular level in and around MB models subject to ultrasound. In particular, the initial aim of the work is to theoretically investigate the free-radicals generation (formation) and outcome (propagation and interaction with model cell membranes,) when microbubbles are subject to sonication conditions. Today, advances in computational techniques and HPC resources allow one to address, for the first time, these questions using molecular mechanics (MM) and high-level quantum mechanics (QM) methods as well as hybrid QM/MM strategies. We then confront these molecular simulations to our experimental data. Finally, we discuss a putative molecular model(s), which can support the role of ROS in the sonoporation mechanisms.