Electroporation is useful for gene transfer to primary cells. Primary cells are not introduced with genes with lipofection. Electroporation is the only method to introduce genes to cells, auch as normal human dermal fibroblasts[10]. Electroporation, in which high-voltage electrical currents create transient pores on the cell membranes, allows the transport of nucleic acid material into the cell[11]. Therapeutic genes are successfully transferred in vivo[12]. The electroporation method, however, has limitations such as short range of gene transfer, necessity of a surgical procedure, and tissue damage. Distance between electrodes normally requires 1 cm and is not suitable for a large area. Placement of internal electrodes requires a surgical procedure, and high voltage can damage tissue when applied. Therefore, less invasive methods are desirable.

Chemical non-viral vectors have been studied because they are generally considered safer than viral vectors. In addition to safety, liposomes can transfer larger genes with less toxicity and are relatively easy to prepare. Cationic lipids and polymers form complexes with negatively charged DNA. The complexes protect the DNA from nucleases and increase the effectiveness of transfection through the cell membrane. One major problem with lipofection is low efficiency of gene transfer. Transfection efficiency of lipofection is improved by condensing DNA with a double chain monovalent quaternary ammonium lipid[13]. Still genes are not introduced to target cells efficiently. More efficient methods have been waited.

Sonoporation refers to the formation of small pores in cell membranes by using ultrasound for the transfer of nucleic acid materials (Figure 1). The biological effects of ultrasound are categorized as thermal and non-thermal. Non-thermal effects are composed of mechanical perturbation in the vicinity of bubbles. Cavitation bubbles cause membrane poration[14]. High speed camera images reveal that the cell membrane is fractionated, and cavitation bubbles are formed[15]. The cavitation bubbles induce cell death or permeability to allow the entry of a drug or genes into the cells. Sonoporation is similar to electroporation, wherein DNA is driven by an electrical force along the electric field. Sonoporation is mediated by passive diffusion. The transfer efficiency depends on ultrasound frequency and intensity[16]. The major advantages of sonoporation are its non-invasiveness and ability to transfer genes to internal organs without a surgical procedure[17]. Targeted gene transfer can be facilitated by ultrasound irradiation of selected tissues after systemic administration[18].

Figure 1 Formation of cell membrane pores after ultrasound irradiation. Nucleic acid such as plasmids enters the cells through the membrane pores that are formed with ultrasound.

Drug delivery with ultrasound was first reported by Tachibana et al[16]. The delivery of insulin on the skin surface when exposed to ultrasound energy in the range of 3000-5000 Pa or 5000-8000 Pa at 48 kHz for 5 min decreased blood glucose levels to 22.4% of the control in 120 min. It was postulated that insulin absorption increased with ultrasound vibration after intradermal injection. This report is interesting because it demonstrates the potential of ultrasound as a method to improve absorption of therapeutic materials. With ultrasound, chemotherapeutic agents are more efficiently absorbed in the mouse xenograft model of cancer[19].

In vitro experiments are simplified models to investigate the mechanisms of sonoporation in vivo. Fechheimer et al[20] transferred plasmid DNA encoding the G418 resistance gene to cultured mouse fibroblasts by sonoporation. Colonies were observed after G418 was added to the media. This was the first report demonstrating that plasmid DNA can be transferred to cells in vitro by using ultrasound. Kim et al[21] also reported that rat joint cells can be successfully transfected with plasmid DNA. Plasmid DNA encoding green fluorescent protein (GFP) was injected as a reporter of expression into the left ventricle of mice percutaneously[22]. Mice were irradiated with transthoracic ultrasound at 1 MHz for 1 min. Histological examination showed GFP expression in the subendocardial myocardium. Intraventricular co-injection of siRNA and GFP exhibited reduced expression of GFP in the coronary artery. These data indicate that plasmid DNA and siRNA can be introduced into cells in vivo by sonoporation.

Biophysical effects of ultrasound include cavitation, radiation pressure, and microstreaming[23]. Cavitation refers to the growth and collapse of microbubbles. Radiation pressure is the force in the irradiation field. Microstreaming is the shear forces that exist near the microbubbles. Formation of cavitation increases with a rising ultrasound intensity while the frequency decreases[24]. Mechanical index (MI) is defined as PRP/√F0 (PRP: peak rarefactional pressure; √F0: transmission center frequency)[25]. MI should be lower than 1.9 for clinical use of ultrasound. Forsberg et al[26] reported that cavitation is unlikely to occur at an MI of less than 0.7. Cavitation has been well investigated with ultrasound contrast agents because microbubbles lower the threshold of cavitation[27]. Zhou et al[28] developed a device to observe the behavior of a single bubble near the cell membrane. The motion of a single bubble was monitored with a high-speed camera. Cell membrane disruption was assessed by monitoring the transmembrane current. This study showed that a single microbubble expands and contracts with ultrasound irradiation. When the bubble collapses, the cell membrane is ruptured and a pore is generated. Changes in membrane permeability are directly correlated with the formation of pores. Carugo et al[29] reported that a standing wave is involved in sonoporation in the absence of microbubbles.

The presence of microbubbles reduces the threshold of cavitation. Microbubble contrast agents are spheres filled with gas and stabilized with shells. Their size ranges between 1 and 10 μm[30]. They are small enough to circulate in blood vessels, but do not exit from the vessels. The contrast agents scatter ultrasound stronger than the surrounding blood and tissue. Thus, they are used as contrast agents for daily clinical practice. Tachibana et al[31] reported that albumin microbubbles (Albumex) accelerate thrombolysis by ultrasound. This was the first report that microbubbles improve the effects of ultrasound in tissues for purposes other than diagnostic imaging. Transfection of plasmid DNA into rat joint cells was improved in the presence of Albumex[21]. This report paved the way for the utilization of ultrasound contrast agents in sonoporation. Microbubbles expand and contract in response to compression and rarefaction of ultrasound (Figure 2). The microbubbles collapse at the high-pressure phase, emitting shock waves that perturb the cell membrane and increase permeability. Qiu et al[32] reported that pores on the cell membrane, generated by sonoporation with microbubbles, ranged from 100 nm to 1.25 μm in size. Their experiments used 1 MHz ultrasound at low acoustic pressures from 0.05 to 0.3 MPa. The pores generated with sonoporation were examined by scanning electron microscopy. The size of the pores enlarged with increased acoustic pressure or longer treatment. They concluded that the pores formed with shear stress. Liquid microjet, visualized with a high-speed camera in cultured cells exposed to single-shot short-pulsed ultrasound has been demonstrated[33]. Contrast agents are shells containing gas. Ultrasound scatters on the surface of contrast agents, and are visible as high echo on the display of diagnostic ultrasound. Physical and biological characteristics of contrast agents are basically the same as those of microbubbles. When a contrast agent (Levovist) was added to the media, the jet caused cell membrane damage. Interestingly, the cell membrane repair process suggests that the Ca²⁺-independent and Ca²⁺-trigger mechanisms are involved in rapid resealing. Positively charged microbubbles are more efficient in gene transfer than neutral ones[34]. It is hypothesized that positively charged microbubbles are more close to the cell membrane to enable more efficient gene transfer. Plasmids are introduced into cells in vitro and in vivo with sonoporation with microbubbles more efficiently than liposome that is commonly used for transfection[35,36].

Figure 2 Microbubble response to an ultrasonic pressure wave. Microbubbles expand and contract when exposed to ultrasound at rarefaction and compression, respectively. At high pressure, microbubbles collapse and a shock wave is emitted.

To date, no clinical trials have been reported using sonoporation. There are several reasons for this, including low transfer efficiency and difficulty in monitoring irradiated fields using sonoporators.

Sonoporation harbors complicated aspects. Gene transfer efficiency is low with sonoporation. 37.5-50 μg of plasmid should be applied to a single rat with sonoporation in vivo[34]. This limitation is one of the reasons that sonoporation has not been applied clinically as mentioned above. Another limitation is cell damage caused with sonoporation. Miller et al[37] irradiated cells with 2.25 MHz continuous ultrasound for 1 min. When an ultrasound contrast agent (Definity, a perflutren lipid microsphere injectable suspension; Bristol Myers Squibb Medical Imaging, N. Billerica, MA) was added to the culture media, the irradiated cells underwent apoptosis. Enzymatic activity and mitochondrial membrane are changed after sonoporation[38]. Stresses to endoplasmic reticulum and mitochondria trigger apoptosis[39,40]. Sonoporation delays DNA synthesis to arrest cell cycle[41]. It should be noted that sonoporation itself may cause apoptosis when it is applied to cancer with therapeutic genes[37].

Wang et al[27] analyzed the protection of plasmid DNA xenograft tumors in mouse hind limb after intravenous administration of a complex of plasmid DNA (luciferase) with cationic or neutral microbubbles. Luciferase activity was 3.8-fold stronger in complexes with cationic microbubbles than in those with neutral microbubbles. Factors influencing the transfection efficiency were analyzed, and the superiority of cationic microbubbles was more evident at lower doses of microbubbles and plasmid DNA[34]. This strategy appears suitable for this application because nucleic acids are negatively charged. It is expected that plasmid DNA or siRNA will interact and form better complexes with cationic microbubbles than with neutral microbubbles.

It is difficult to modify surfaces of microbubbles with functional molecules for targeting. On the other hand, liposomes are easy to modify for targeting. Suzuki et al[44] developed polyethylene glycol-modified liposomes containing perfluoropropane, a contrast agent for ultrasound imaging. These so-called “bubble liposomes”, significantly improved the transfection efficiency of plasmid DNA encoding luciferase into cultured cells. The luciferase reporter plasmid was injected into the femoral artery of mice. Two days after ultrasound irradiation, a signal was detected along the artery. These data suggest that the bubble liposome is a candidate for gene delivery both in vitro and in vivo. Bubble liposomes were transferred with the interleukin-12 gene in xenograft ovarian cancer mice, and the tumor sizes were reduced[45]. These animal experiments demonstrate that bubble liposomes may be applicable to clinical studies.

Sonoporators are typically used for sonoporation. One of the problems with sonoporators is that the irradiation field cannot be monitored. Diagnostic ultrasound is widely used in daily clinical practice. Diagnostic ultrasound is equipped with a display for image diagnosis. If diagnostic ultrasound could be utilized for sonoporation, therapeutic genes can be accurately introduced into the target fields. Miller and Quddus transfected cultured cells with plasmid DNA using diagnostic ultrasound[46]. They used a 3.5 MHz curved linear transducer of diagnostic ultrasound. A contrast agent of ultrasound (Optison) was added to the media. Although Optison may improve transfection efficiency, even without contrast agents, plasmid DNA was successfully introduced into cultured cells by using diagnostic ultrasound[47]. SiRNA of frizzled (Fz)-9, a receptor of the Wnt signaling pathway, transferred into cultured cells using diagnostic ultrasound suppresses cell proliferation[48]. SiRNA of Fz-9 transferred into cultured cells using lipofection also suppresses cell proliferation[49]. These 2 papers indicate that transfection efficiency of siRNA with diagnostic ultrasound is comparable with that of lipofection. Wang et al[27] investigated cell permeability of Evans Blue in vivo with diagnostic ultrasound. They observed the dye in xenograft hepatoma. Interestingly, they reported that efficiency of Evans Blue transfer was affected by MI sonication duration and dye dose. These data clearly demonstrate that ultrasound causes sonoporation, a biological process that is a promising new approach for the delivery of DNA/RNA for gene therapy.