Ultrasound-Mediated Gene Delivery: Influence of Contrast Agent on

Apr 10, 2007 - MAT B III cells were insonified at ultrasound frequencies of 1.15 and 2.25 .... Journal of Drug Delivery Science and Technology 2013 23...
0 downloads 3 Views 455KB Size
Bioconjugate Chem. 2007, 18, 652−662

652

Ultrasound-Mediated Gene Delivery: Influence of Contrast Agent on Transfection Sophie Mehier-Humbert,†,‡ Feng Yan,‡ Peter Frinking,‡ Michel Schneider,‡ Richard H. Guy,†,§ and Thierry Bettinger*,‡ University of Geneva, Department of Pharmacy and Biopharmacy, 30 quai Ansermet, 1211 Geneva 4, Switzerland, Bracco Research SA, 31 route de la Galaise, 1228 Plan-les Ouates, Switzerland, and Department of Pharmacy and Pharmacology, University of Bath, Claverton Down, Bath, BA2 7AY, U.K. Received August 7, 2006; Revised Manuscript Received October 20, 2006

MAT B III cells were insonified at ultrasound frequencies of 1.15 and 2.25 MHz in the presence of different ultrasound contrast agents (UCA) and a plasmid encoding for the green fluorescent protein. The transfection efficiency was assessed by flow cytometry, while contrast agent destruction by ultrasound was evaluated using optical and scanning electronic microscopies. It was found that the gas and shell properties of the UCA have an important influence on cell transfection. A good correlation was observed between bubble destruction and transfection rate, and it was demonstrated, for the first time, that hard-shelled contrast agents (gas microcapsules) are promising candidates for ultrasound-mediated gene delivery.

INTRODUCTION Ultrasound contrast agents (UCA) are widely used in diagnostic applications in medical research and clinical practice. Recently, they have been exploited in ultrasound-mediated gene delivery (1-3). This new approach uses acoustic cavitation effects of gas microbubbles to induce transient membrane permeabilization (sonoporation) at the cellular level, thereby facilitating the transfer of genetic materials. In comparison to other transfection techniques (viral or nonviral), sonoporation may allow in vivo gene transfer more specifically with high spatial and temporal control, since the site and levels of transfection can be precisely localized and controlled by ultrasound (4, 5). This approach has been applied in vitro (69) and in vivo (3, 10-12). It is thought that sonoporation is induced by acoustic cavitation (13), which generates microstreaming or microjets in the surrounding medium (14-16), resulting in permeabilization of the cell membrane. Most UCA are made of gas microbubbles and are referred to as “soft-shelled” agents, e.g., Albunex, Optison, and Definity (17). These UCA have been used in sonoporation as cavitation nuclei (9, 18-20). Agents specially designed for gene delivery have also been reported (21, 22), and are generally soft-shelled, comprising phospholipids or proteins with a plasmid-bearing surface and a perfluorocarbon gas core. Hard-shelled UCA (i.e., containing a rigid lipid or polymeric shell) may also be used for ultrasound-mediated gene delivery. These agents, although not yet commercially available, have been extensively studied for diagnostic imaging (23-25), drug delivery (26), acoustic properties (27, 28), and other biophysical applications (29). While it is generally believed that UCA destruction at high acoustic pressure is the main cause of sonoporation, there have been few studies in which cell transfection and UCA destruction have been assessed at the same time and then related to the formulation properties of the contrast agent. In the present work, it was hypothesized that the physicochemical parameters of * Corresponding author. Dr. T. Bettinger, Bracco Research SA, 31 route de la Galaise, 1228 Plan-les Ouates, Switzerland; thierry.bettinger@ brg.bracco.com; fax# 0041 22 884 88 85. † University of Geneva, Department of Pharmacy and Biopharmacy. ‡ Bracco Research SA. § Department of Pharmacy and Pharmacology, University of Bath.

UCA, including bubble size, shell properties, and encapsulated gas, may be modulated to improve the efficiency of gene transfection in sonoporation. Two types of experimental UCA were investigated for in vitro transfection: gas microbubbles, stabilized with phospholipids, and gas-filled microcapsules with a rigid shell made of either triglyceride or polystyrene. Contrast agent destruction was also evaluated and correlated with transfection.

EXPERIMENTAL PROCEDURES Cell Culture. Rat mammary carcinoma cells (MAT B III) were incubated at 37 °C under 5% CO2 in 225 cm2 tissue culture flasks, in a MacCoy’s 5A medium containing Glutamax-I (Life Technologie, Switzerland), supplemented with 10% v/v heatinactivated fetal calf serum (FCS) and 1% v/v antibiotics (initial concentration: 10 000 IU/mL penicillin, 10 000 µg/mL streptomycin, 25 µg/mL fungizone). Ultrasound Contrast Agents (UCA). UCA were formulated with various shell materials and gases (Table 1). The particle concentration and average number diameter of each agent were determined using a Coulter Multisizer II (Coulter Electronics Limited, Luton, Bedfordshire, England). Table 1 includes the resistance of each UCA to hydrostatic pressure (Pc50, expressed in mmHg), a critical value at which 50% of UCA are destroyed/ collapsed. This pressure was determined as described elsewhere (30). BR14 and the commercial agent SonoVue are both phospholipid-stabilized gas microbubbles containing, respectively, C4F10 and SF6. These microbubbles were prepared accordingly to published procedure (31). Studies performed using Coulter counter and HPLC analyses confirm that BR14 microbubbles are coated with a monolayer of phospholipid molecules (unpublished results). It is well-known that the thickness of a monolayer of mixed phospholipids spread at a gas-liquid interface is about 2-3 nm (e.g., determined by Langmuir through or Langmuir-Blodgett) (32). Figure 1A depicts BR14 microbubbles. Gas-filled microcapsules were prepared by an emulsification procedure according to published protocol (33). The characteristics of similar gas-filled microcapsules (also called microballoons) were already reported by Schneider et al. (23, 34, 35). A comparative study on the behavior of lipid- and polymer-shelled agents (the two agents

10.1021/bc0602432 CCC: $37.00 © 2007 American Chemical Society Published on Web 04/10/2007

Bioconjugate Chem., Vol. 18, No. 3, 2007 653

Contrast Agents for Ultrasound-Mediated Gene Delivery

Figure 1. (A) Freeze-fracture electron micrograph of a microbubble (BR14) stabilized by a thin phospholipid film. (B) Scanning electronic micrograph of microcapsules (BG1768)sthe shell is made of polystyrene (100 kDa) and has a thickness of about 50-200 nm. The openshelled structure was obtained after ultrasound exposure of the microcapsules.

used in the present study) in response to acoustic exposure was performed by Bloch et al. (36). For hard-shelled agents, formulated with either triglyceride (Figure 1B) or polystyrene, the estimated shell thickness was 50-200 nm. The soft-shelled agents are hereafter referred to as microbubbles, and the rigid-shelled UCA as microcapsules. Ultrasound Exposure. A 3 mL polystyrene round-bottom tube was used as an exposure chamber. 500 µL of a cell suspension, at a final concentration of 1 × 106 cells/mL, was mixed with culture medium containing the plasmid gWiz-GFP (5757 base pairs; Aldevron, Fargo, ND), encoding for the green fluorescent protein (GFP). Most sonoporation experiments (except those studying the effects of bubble concentration and size) were performed with a final plasmid concentration of 10 µg/mL and a bubble-to-cell ratio of 30. The tube was mounted in a rotating exposure system and immersed in a water bath at 37 °C. The distance between the transducer and the tube was 7.6 cm. MAT B III cells were insonified for 10 s, using two different transducers (Figure 2). Peak-negative pressures were measured with a membrane hydrophone (Precision Acoustics, Ltd.) for each transducer at a distance of 76 mm (near-field

distance for both transducers). The polystyrene tube was cut in half along its axis and was positioned in front of the membrane hydrophone to account for attenuation of the ultrasound beam by the tube. A duty cycle of 20% and a pulse repetition frequency of 100 Hz were applied for all experiments. After insonification, the mixture was placed in 12-well plates and supplemented with 2 mL of medium containing 10% FCS. Finally, the cells were incubated at 37 °C under a 5% CO2 atmosphere for 24 h. Assay of Reporter Gene. GFP-positive cells were analyzed 24 h after ultrasound exposure by flow cytometry (FACS Calibur, Becton Dickinson AG, Switzerland). The cell suspension was placed in 5 mL polystyrene round-bottom tubes and washed twice with phosphate-buffered saline (PBS). The pellet was resuspended with 250-300 µL of PBS, and 20 µL of a propidium iodide (PI) solution (40 µg/mL, Sigma-Aldrich Chemie GmbH, Schnelldorf, Germany) was added to assess cell viability. Results are expressed as a percentage of GFP-positive cells and relative fluorescence intensity (arbitrary units, au), using the software CellQuest Pro (Becton Dickinson Bioscience, Binninggerstrasse 94, CH-4123 Allschwill, Switzerland). The percentage of positive cells is calculated on the basis of the total number of cells, including dead cells. Microbubble Destruction. Microbubbles (12.5 × 106 BR14 bubbles, corresponding to a bubble-to-cell ratio of 25) were suspended in 500 µL of a MAT B III cell suspension (final concentration 1 × 106 cells/mL in medium without FCS) in the sample tube. The suspensions were then exposed to ultrasound using identical parameters to those employed in the transfection experiments. Microbubbles were counted before and after insonification using computer-assisted microscopic analysis (Leica - Program Routine BR20.Q5R, Leica Microsystem AG, Glattbrugg, Switzerland). The percentage of intact bubbles and the bubble size distribution were determined. Scanning Electron Microscopy (SEM). Microcapsule destruction was assessed using scanning electron microscopy (LEO S470; experiments were performed at Surcotec SA, Plan-lesOuates, Switzerland). Briefly, 15 × 106 particles were suspended in 500 µL of PBS in a sample tube and exposed to ultrasound as before. After insonification, the suspensions of ten tubes were pooled in a 15 mL polystyrene tube. The microcapsules were then washed twice in distilled water by centrifugation (10 min, 800 g), and the pellet comprising the ruptured microcapsules was collected to be freeze-dried. The washed capsules were then stuck onto an adhesive tape, fitted onto the metal stand of the microscope, and placed under vacuum at room temperature overnight. After gold sputtering, the samples were analyzed at 20 kV acceleration voltages. Microcapsules nonexposed to ultrasound were used as control. Statistical Analysis. All results are reported and displayed as mean ( standard deviation (n g 6). Tests of significance were performed using a one-way analysis of variance (Minitab software v 13.20, Minitab, Inc., Paris, France), with p < 0.05 considered to be statistically significant.

RESULTS Transfection Efficiency and Cell Viability. Transfection efficiency was assessed as (i) the percentage of transfected cells calculated from the ratio of GFP-positive cells to the total number (including dead cells); (ii) the fluorescence intensity of the cells determined by flow cytometery, relative to average amount of GFP produced per transfected cell; and (iii) the cell viability, expressed as viable cells/total cells and determined by propidium iodide exclusion. Effect of UCA Concentration and Size. MAT B III cells (1 × 106/mL) were sonoporated using the 1.15 or 2.25 MHz transducer at acoustic pressures of 570 and 207 kPa (peak

654 Bioconjugate Chem., Vol. 18, No. 3, 2007

Mehier-Humbert et al.

Figure 2. Ultrasound beam simulation for single-element transducers. Upper panel: 1.15 MHz unfocused transducer (Valpey Fisher; Hopkinton, MA); aperture ) 0.75 in, narrow band excitation. Lower panel: 2.25 MHz focused transducer (Panametrics; Waltham, MA) (3 in); aperture ) 1 in, narrow band excitation. Both plots are normalized to their maximum peak negative pressure in the field. The contours indicate the -6 dB values relative to the maximum. Table 1. Physicochemical Characteristics of UCA Studied formulation 1A009 1A009-like 1A009-like BG1126 (1A009-like) BG1127 (1A009-like) BG1382; BG1766 BG1431 BG1767; BG1768; BG1769

shell composition

gas

mean diameter (µm)

Microbubbles; Soft-Shelled (2-3 nm) phospholipid monolayer C4F10 phospholipid monolayer air phospholipid monolayer C3F8 phospholipid monolayer C4F10 phospholipid monolayer C4F10 Microcapsules; Hard-Shelled (∼100 nm) triacylglyceride air triacylglyceride C4F10 polystyrene(30; 100; 300 kDa) air

negative pressure, P_), respectively. The exposure duration was 10 s. Perfluorobutane gas microbubbles (BR14) were used as UCA, and the bubble-to-cell ratio was varied from 0 to 150. As shown in Figure 3A, the transfection rate first increased with the bubble-to-cell ratio, peaked at 30 bubbles/cell, and then decreased progressively at higher bubble concentrations. A transfection rate of 8-17% (expressed in % of GFP-positive cells) was obtained depending on the transducer frequency and bubble concentration, with a cell viability varying between 60% and 80% (Figure 3B). Cell viability decreased with bubble-tocell ratio and reached a plateau at 30 bubbles/cells. Interestingly, both transducers gave similar trends for transfection and cell viability. Suspensions of BR14-like phospholipid-stabilized microbubbles were fractioned into two narrow-sized populations by decantation. One had a median number diameter, Dn50, of 0.99 µm (BG1126), and for the other, Dn50 was 1.9 µm (BG1127). Transfection was performed with these bubble fractions and with the original BR14 bubbles, using different bubble-to-cell ratios. The results are summarized in Table 3 and show that, under

concentration (part/mL)

Pc50 (mmHg)

2.0 2.0 2.0 1.0 2.1

3.0 × 108 2.0 × 108 3.0 × 108 1.5 × 109 5.7 × 108

630 40-50 580 450 700

1.6 1.6 1.7

2.1-2.6 × 109 1.2 × 109 1.3-2.6 × 109

>1200 >1200 >1200

the experimental conditions considered, bubble size appeared to have little effect on transfection efficiency (no statistically significant difference between the two decanted samples at p < 0.05). The study focused therefore on UCA shell and gas properties. Effect of UCA Shell Properties. Microbubbles Vs Microcapsules. The effect of UCA shell properties on MAT B III transfection was first studied using the 2.25 MHz transducer, with an ultrasound exposure of 10 s and an applied peak negative pressure varying from 0 to 800 kPa. MAT B III cells were sonoporated in the presence of plasmid and either soft-shelled microbubbles (BR14) or hard-shelled microcapsules (BG1382) made of triglycerides. Air was the encapsulated gas for both agents. As shown in Figure 4A, the percentage of GFP-positive cells increased with acoustic pressure and reached a plateau of 16% and 25% at 570 kPa, respectively, for the two agents. An important difference in the cell transfection rate was observed below 570 kPa between the soft-shelled and hard-shelled agents (p < 0.05). For instance, at a pressure of 247 kPa, 10% of GFP-

Bioconjugate Chem., Vol. 18, No. 3, 2007 655

Contrast Agents for Ultrasound-Mediated Gene Delivery

Figure 3. Effect of bubble concentration (bubble-to-cell ratio) on transfection efficiency. MAT B III cells were insonified using C4F10 microbubbles (BR14) at either 1.15 MHz at 571 kPa (-b-) or 2.25 MHz at 207 kPa (-O-). (A) Percentage of transfected cells. (B) Cell viability (%). Table 2. Effect of Bubble Size on Transfection (viability and % GFP-positive cells) near field aperture distance [mm] [mm]

frequency [MHz]

transducer type

1.15 2.25

unfocused/air-backed focused

19.1 25.4

76 76

manufacturer Valpey Fisher Panametrics V304

Table 3 viability UCA

median diameter (µm)

1A009

Dn50 ) 1.4; Dv50 ) 8.9

BG1126

Dn50 ) 0.99; Dv50 ) 1.6

BG1127

Dn50 ) 1.9; Dv50 ) 3.9

bubble-to-cell ratio

% GFP positive cells

mean

sd

mean

sd

91.9

0.0

0.02

0.0

25 0.12

77.3 83.2

2.8 4.0

23.1 15.5

1.5 1.5

12 25 0.12

76.2 79.4 85.8

2.2 0.4 1.0

17.7 21.0 13.1

3.9 4.5 1.3

12 25

78.9 80.3

0.9 0.3

20.7 26.5

0.3 2.2

0

positive cells were obtained for microbubbles vs only 0.4% for microcapsules. These results suggest that soft-shelled microbubbles made of phospholipids require a lower pressure to induce acoustic cavitation than microcapsules to facilitate DNA delivery into cells. Microcapsules, being more pressureresistant, appeared to have a higher cavitation threshold (400-

600 kPa for 2.25 MHz), below which only few cells were sonoporated. The GFP fluorescence from transfected cells is plotted in Figure 4B as a function of ultrasound pressure. Surprisingly, the microcapsules, although showing a lower transfection rate over the entire range of acoustic pressures, gave significantly higher fluorescence (by a factor about 1.5). This means that more plasmid copies per cell were delivered into the cell nucleus in the presence of the hard-shelled agent. Cell viability decreased with increasing ultrasound pressure for both agents with a viability varying from 93% to 72%. The microcapsules resulted in generally lower cell viability as compared to the microbubbles (about 10-15%; data not shown). A similar experiment was performed using the air-backed transducer with a center frequency of 1.15 MHz and at acoustic pressures varying from 0 to 480 kPa (Figure 4C,D). The same UCAs were compared, except that the encapsulated gas, in this case, was a perfluorocarbon (C4F10). Transfection rates increased steadily with ultrasound pressure for the two agents. The ultrasound pressure required for maximum transfection was much lower for this frequency; i.e., only about 50% of that required with the 2.25 MHz transducer. This difference in pressure threshold could be due to the beam profile of the two transducers: the 1.15 MHz transducer is unfocused, providing a more homogeneous and a larger insonification field than the 2.25 MHz focused transducer (see Figure 2). In comparison to sonoporation at 2.25 MHz, even higher levels of fluorescence were obtained with the microcapsules (Figure 4D). However, cell viability was 15-25% lower with the capsules (60-79%) than with the microbubbles (85-94%) for the same applied pressure (data not shown). Microcapsules Made of Different MW Polystyrenes. The shell properties of microcapsules are generally determined by the nature and amount of the materials employed. Three formulations of hard-shelled microcapsules, made with polystyrene of different molecular weights (30, 100, and 300 kDa), were compared. A constant amount of polymer was used in the three preparations, and air was the encapsulated gas core. It was expected that using a polymer, such as polystyrene, would provide a much more rigid shell than that obtained with triglycerides (BG1382, BG1431; see Table 1), particularly when using a high molecular weight polymer of 300 kDa. MAT B III cells were sonoporated in the presence of polystyrene microcapsules using the focused 2.25 MHz transducer with ultrasound pressure varying from 0 to 809 kPa. The results are summarized in Figure 5. For all three formulations, transfection increased with increasing acoustic pressure from 360 to 570 kPa and reached a plateau of 14-18% GFP-positive cells (Figure 5A). Interestingly, at a pressure of 360 kPa, the transfection rate obtained was significantly higher with the low MW polystyrene: 8% of GFP-positive cells for 30 kDa vs only 2% for the 100 and 300 kDa polymers. The low MW microcapsules also showed a lower acoustic cavitation threshold, similar to that of microbubbles (see Figure 4A). In contrast, the results in Figure 5B showed that the 100 and 300 kDa microcapsules gave significantly higher fluorescence intensities. For example at 364 kPa, these microcapsules, despite their 4-fold lower transfection rates, resulted in 60% more GFP produced per transfected cell as compared to the 30 kDa microcapsules. Effect of UCA Encapsulated Gas. MAT B III cells were insonified using the 2.25 MHz-focused transducer. Microbubbles used in this experiment were filled with air, perfluoropropane (C3F8), or perfluorobutane (C4F10). As shown in Figure 6A, similar transfection rates were obtained for UCA filled with C3F8 and C4F10 and were significantly higher than those for the air-filled microbubbles at all applied acoustic pressures. A small

656 Bioconjugate Chem., Vol. 18, No. 3, 2007

Mehier-Humbert et al.

Figure 4. Effect of UCA shell properties on transfection efficiency. MAT B III cells were insonified at various peak negative pressures at 2.25 MHz (A and B) and 1.15 MHz (C and D), in the presence of either a soft-shelled agent, BR14 (-b-), or a hard-shelled agent, BG1766 or BG1382 (-O-). (A,C) Percentage of transfected cells. (B,D) Fluorescence intensity (expressed in arbitrary units (au)). Both UCA contained air as the encapsulated gas.

increase of fluorescence intensity was also observed for the perfluorocarbon gases (Figure 6B). A similar study was performed with hard-shelled agents (BG1382, 1431) using the 1.15 MHz transducer. The results are in Figure 6. In this case, the encapsulated gas had no impact on transfection (Figure 6A) but significantly improved GFP production (Figure 6B): at pressures above 300 kPa, the fluorescence intensity was almost 2-fold higher with C4F10 compared to air. These results clearly show that, in addition to shell properties, the gas encapsulated in UCA may also play an important role in sonoporation. Contrast Agent Destruction. It is generally believed that acoustic cavitation and subsequent sonoporation can be enhanced by UCA destruction. It is still unclear, however, whether microbubble or microcapsule destruction is a prerequisite for successful sonoporation and whether there is a relationship between bubble destruction and transfection efficiency. To address these points, microbubbles and microcapsules were insonified using the same ultrasound exposure conditions as described above with the 1.15 and 2.25 MHz transducers. Bubble size distribution was analyzed by optical microscopy before and after ultrasound exposure. As microcapsule destruction could not be directly assessed by optical microscopy or Coulter analysis due to their rigid shells, scanning electron microscopy (SEM) was used to evaluate these agents. Microbubbles. After a 10 s ultrasound exposure at 1.15 MHz and 402 kPa, more than 90% of gas microbubbles were destroyed (determined using computer-assisted microscopic analysis; data not shown). Figure 7A shows that the bubble size distribution was significantly modified by ultrasound exposure. The bubbles of 4-7 µm seemed to be most sensitive. An

increase from 12% to 38% in the large bubble fraction was noted upon ultrasound exposure. This could be explained by the fact that large bubbles (>8 µm) are generally more persistent and less easily destroyed than small ones. The 6 µm bubbles showed the lowest count after insonification; interestingly enough, this diameter roughly corresponds to the calculated resonance diameter for ultrasound at 1.15 MHz. Bubble destruction and transfection experiments were further performed in parallel using BR14 microbubbles and the focused 2.25 MHz transducer with the acoustic pressure varying from 0 to 809 kPa. As shown in Figure 7B, the 3 µm bubbles appeared most sensitive, and this again correlated well with the calculated resonance diameter of microbubbles for this frequency. Size distribution analysis indicated that 63% to 74% of the bubbles were destroyed during exposure, depending on applied pressure with the 2.25 MHz transducer (Figure 7B). There was a good correlation between transfection rate and bubble destruction as a function of the applied pressure: the higher the pressure, the larger the bubble destruction and the better the transfection (Figure 7C). Microcapsules. Microcapsules of polystyrene 100 kDa were exposed to 570 kPa ultrasound insonification at 2.25 MHz. Scanning electron microscopy (SEM) revealed the morphology of the microcapsules before (Figure 8A) and after (Figure 8B) ultrasound exposure. Although sample preparation for SEM could alter the appearance of hollow microcapsules (the control microcapsules seemed to be flat rather than spherical in Figure 8A), differences were evident between insonified and control samples. Most ultrasound-treated microcapsules showed an open or deflated structure with a folded shell. In contrast, the control microcapsules presented a regular shape with smooth surfaces

Contrast Agents for Ultrasound-Mediated Gene Delivery

Figure 5. Effect of UCA shell properties on transfection efficiency. Microcapsule shell(s) were polystyrene of different molecular weights: 30 kDa (-×-), 100 kDa (-0-), and 300 kDa (-2-). MAT B III cells were insonified at various peak negative pressures at 2.25 MHz, in the presence of 30 capsules/cell. (A) Percentage of transfected cells. (B) Fluorescence intensity (expressed in arbitrary units (au)). Both agents contained air as the encapsulated gas.

(Figure 8A). The open or deflated structures were observed for both small and large microcapsules, suggesting that shell properties such as stiffness, thickness, and surface homogeneity may play a more important role in microcapsule destruction than size. The SEM in Figure 8B shows that some microcapsules remained intact (a few had a small hole on the shell surface), but many were broken as a result of ultrasound exposure. Similar experiments were also performed with the microcapsules of polystyrene 30 and 300 kDa under the same exposure conditions. Generally, the 300 kDa microcapsules were less altered, in terms of their number and shape, than those of 30 kDa (data not shown). It seemed that the shells made of high MW polystyrene were more stable to ultrasound exposure; however, upon destruction, these capsules may collapse more violently, projecting DNA molecules directly toward the cell nucleus. This may explain why fluorescent intensity (GFP production) was higher for these high MW agents (Figure 4D).

DISCUSSION In vitro sonoporation has been widely investigated. It has been shown that ultrasound exposure parameters, such as acoustic pressure or intensity (18, 37), frequency (9, 38), duration of exposure, pulse-repetition frequency and duty cycle (37, 39), and temperature (40, 41) are important for transfection and delivery of small molecules into cells. Effects of contrast agents and bubble concentration have been also studied (39, 42, 43). In vitro gene transfection has equally been demonstrated with various cell types, such as rat fibroblasts, chondrocytes (6) and cultured Chinese hamster ovary cells (9), human cancer cells

Bioconjugate Chem., Vol. 18, No. 3, 2007 657

HeLa, NIH/3T3 C127I (44), and prostate cancer cells DU145 (41). Other aspects involved in sonoporation have also been evaluated, such as bubble-to-cell ratio or distance (19), cell-tocell heterogeneity of molecular uptake (37), ultrasound modalities (43), DNA integrity (41), and plasmid-attached microbubbles (21, 22, 45, 46). On the other hand, UCA destruction has been evaluated initially for diagnostic imaging purposes (47-51), before being extended to UCA-induced bioeffects (52-55) and therapeutic applications (5, 56-58). There have been very few studies in which the influence of UCA properties on gene/drug delivery has been systematically assessed, and UCA formulations dedicated to this application have not been optimized in any way. The present study aimed at addressing this issue by assessing, in particular, the effect of UCA size, shell, and gas composition on transfection efficiency. These parameters have been shown to be important in both contrast agent destruction and enhanced acoustic cavitation. Bubble Concentration and Size. UCA bioeffects on cell lysis appear to decrease with the cubic power of the distance between microbubbles and cells (19). A high bubble concentration generally favors cell transfection, but it may also induce strong acoustic attenuation in vitro and in vivo, thereby reducing transfection efficiency. High bubble concentration may also lead to more cell damage, reducing in vitro cell viability and inducing in vivo side effects. For a given sonoporation sample volume, bubble-to-cell distance decreases with both bubble-to-cell ratio and cell density. It was thought that the bubble/cell ratio would be more relevant to assess the effect of UCA concentration on transfection, rather than the number of microbubbles (or microcapsules) alone. In the literature, the use of a UCA dilution factor has also been reported; however, this parameter is meaningful only for a given contrast agent with a defined bubble concentration. The in vitro results showed a maximum transfection rate with a ratio of 30 bubbles per cell for the two different transducers used in the present study. At this ratio, the bubble/cell distance was ca. 30 µm (19). A further increase of the bubble-to-cell ratio tended to reduce transfection rate (Figure 3A). Since the high bubble-to-cell ratios (60, 75, and 150) did not induce more cell damage, this would suggest that the lower percentage of GFP-positive cells obtained was only due to a strong acoustic attenuation caused by microbubbles. These observations are consistent with other studies (13, 37). The high concentration used in this study might yield to secondary effect, such as Bjerknes forces, by which aggregates of bubbles are formed (59). Aggregates of bubbles might interact differently with cells compared to individual bubbles. This possible interaction was not taken into account in the study. It is well-known that bubble size is an important parameter for UCA stability and acoustic response. For instance, contrast agent destruction and its mechanism depend on acoustic pressure and bubble size (60). The results obtained in the present study showed little influence of microbubble size on transfection efficiency (Table 3). This lack of effect may be due to heterogeneity of bubble size; indeed, the three bubble samples all had relatively large size distributions: while the median diameters of three preparations were quite different (1.6, 3.9, and 8.9 µm), their Dn values were rather similar: 1.0, 1.4, and 1.9 µm, respectively. Bubbles smaller than 3 µm were predominant: 85% for BR14, 84% for BG1126 (large bubbles), and 99% for BG1127 (small bubbles). The bubble destruction experiment performed at the same frequency and pressure showed that more than 70% of microbubbles were destroyed, in particular, the smaller ones between 1 and 4 µm (see Figure 7A). This is in agreement with a previous report (50) which showed that, at a frequency of 2.25 MHz and a peak negative

658 Bioconjugate Chem., Vol. 18, No. 3, 2007

Mehier-Humbert et al.

Figure 6. Effect of the encapsulated gas on transfection efficiency. Microbubbles (BR14) were filled with air (-×-), C3F8 (-0-), or C4F10 (-2-) (A,B); and microcapsules were filled with air (-O-) or C4F10 (-b-) (C,D). MAT B III cells were insonified at various peak negative pressures at 2.25 MHz (A,B) or 1.15 MHz (C,D), in the presence of 30 bubbles/cell. (A,C) Percentage of transfected cells. (B,D) Fluorescence intensity (expressed in arbitrary units (au)).

pressure of 800 kPa, small microbubbles (300 kPa, 2.25 MHz), bubble destruction became significant, resulting in a maximum transfection rate (15-25% of GFPpositive cells/total cells) and a lower cell viability (∼70%). Microcapsules, on the other hand, showed only a negligible transfection at this pressure due to the stiffness of their shell. Hard-shelled agents do not significantly respond to ultrasound pressure waves until the shell is destabilized. Consequently, microstreaming effects will be negligible before UCA destruction. This is especially true for microcapsules made of high molecular weight polystyrene (100 and 300 kDa), for which destruction and gene transfection only occurred at pressures approaching 600 kPa (with the 2.25 MHz focused transducer). Transfection rates with microcapsules improved by using the unfocused air-backed transducer at lower frequency (1.15 MHz). The requirement of UCA agent destruction to induce transfection implies that only inertial (transient) cavitation is effective for these agents. The results in this study also showed that rigid microcapsules generally delivered more DNA molecules into the cell, and subsequently into the cell nucleus, and produced more GFP per cell than microbubbles (by a factor of 1.5 to 2). The exact mechanism involved remains unknown; it is possible that microstreaming/microjets, and subsequent shear stress, caused by destruction/fragmentation of microcapsules, are much stronger due to the large expansion and rapid collapse of the released free gas bubbles. It is interesting to note that the level of GFP/ cell can be further improved by combining a hard-shelled agent and a gas with low water solubility such as perfluorobutane (C4F10). By contrast, for microbubbles, the nature of the entrapped gas had little effect. While significant progress has been made in both the mechanistic understanding and experimental optimization of sonoporation, more research is required to improve efficacy for in vivo applications. Gene or drug delivery in vivo will be limited by the extremely unfavorable bubble-to-cell ratio, low plasmid concentration, low or nondividing properties of targeted cells, and the accessibility of the cells to the UCA. For instance, at diagnostic imaging doses of UCA, the bubble-to-cell ratio may be lower than 1 ppm. Different approaches have been proposed to tackle these problems (including optimization of

ultrasound exposure parameters and settings, exploring new ultrasound modalities, designing novel and dedicated transducers, use of plasmid-loaded microbubbles, targeted UCA and radiation force-assisted targeting, combination of lipofection and sonoporation, etc.). The optimization of UCA formulation parameters will no doubt contribute to the development of this exciting technology.

ACKNOWLEDGMENT We would like to thank Philippe Bussat for the preparation of the hard-shelled agents.

LITERATURE CITED (1) Bekeredjian, R., Chen, S., Frenkel, P. A., Grayburn, P. A., and Shohet, R. V. (2003) Ultrasound-targeted microbubble destruction can repeatedly direct highly specific plasmid expression to the heart. Circulation 108 (8), 1022-1026. (2) Bekeredjian, R., Chen, S., Pan, W., Grayburn, P. A., and Shohet, R. V. (2004) Effects of ultrasound-targeted microbubble destruction on cardiac gene expression. Ultrasound Med. Biol. 30 (4), 539543. (3) Lu, Q. L., Liang, H. D., Partridge, T., and Blomley, M. J. (2003) Microbubble ultrasound improves the efficiency of gene transduction in skeletal muscle in vivo with reduced tissue damage. Gene Ther. 10 (5), 396-405. (4) Unger, E. C., Hersh, E., Vannan, M., and McCreery, T. (2001) Gene delivery using ultrasound contrast agents. Echocardiography 18 (4), 355-361. (5) May, D., Allen, J. S., and Ferrara, K. (2002) Dynamics and fragmentation of thick-shelled microbubbles. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 49 (10), 14001410. (6) Kim, H. J., Greenleaf, J. F., Kinnick, R. R., Bronk, J. T., and Bolander, M. E. (1996) Ultrasound-mediated transfection of mammalian cells. Hum. Gene Ther. 7 (11), 1339-1346. (7) Lawrie, A., Brisken, A. F., Francis, S. E., Tayler, D. I., Chamberlain, J., Crossman, D. C., Cumberland, D. C., and Newman, C. M. (1999) Ultrasound enhances reporter gene expression after transfection of vascular cells in vitro. Circulation 99 (20), 2617-2620. (8) Miller, D. L., and Gies, R. A. (2000) The influence of ultrasound frequency and gas-body composition on the contrast agent-mediated enhancement of vascular bioeffects in mouse intestine. Ultrasound Med. Biol. 26 (2), 307-313. (9) Miller, D. L., Bao, S., and Morris, J. E. (1999) Sonoporation of cultured cells in the rotating tube exposure system. Ultrasound Med. Biol. 25 (1), 143-149. (10) Lawrie, A., Brisken, A. F., Francis, S. E., Cumberland, D. C., Crossman, D. C., and Newman, C. M. (2000) Microbubble-enhanced ultrasound for vascular gene delivery. Gene Ther. 7 (23), 20232027. (11) Anwer, K., Kao, G., Proctor, B., Anscombe, I., Florack, V., Earls, R., Wilson, E., McCreery, T., Unger, E., Rolland, A., and Sullivan, S. M. (2000) Ultrasound enhancement of cationic lipid-mediated gene transfer to primary tumors following systemic administration. Gene Ther. 7 (21), 1833-1839.

Contrast Agents for Ultrasound-Mediated Gene Delivery (12) Miller, D. L., and Song, J. (2003) Tumor growth reduction and DNA transfer by cavitation-enhanced high-intensity focused ultrasound in vivo. Ultrasound Med. Biol. 29 (6), 887-893. (13) Greenleaf, W. J., Bolander, M. E., Sarkar, G., Goldring, M. B., and Greenleaf, J. F. (1998) Artificial cavitation nuclei significantly enhance acoustically induced cell transfection. Ultrasound Med. Biol. 24 (4), 587-595. (14) Koch, S., Pohl, P., Cobet, U., and Rainov, N. G. (2000) Ultrasound enhancement of liposome-mediated cell transfection is caused by cavitation effects. Ultrasound Med. Biol. 26 (5), 897-903. (15) Wu, J. (2002) Theoretical study on shear stress generated by microstreaming surrounding contrast agents attached to living cells. Ultrasound Med. Biol. 28 (1), 125-129. (16) Wu, J., Ross, J. P., and Chiu, J. F. (2002) Reparable sonoporation generated by microstreaming. J. Acoust. Soc. Am. 111 (3), 14601464. (17) Miller, A. P., and Nanda, N. C. (2004) Contrast echocardiography: new agents. Ultrasound Med. Biol. 30 (4), 425-434. (18) Bao, S., Thrall, B. D., and Miller, D. L. (1997) Transfection of a reporter plasmid into cultured cells by sonoporation in vitro. Ultrasound Med. Biol. 23 (6), 953-959. (19) Ward, M., Wu, J., and Chiu, J. F. (2000) Experimental study of the effects of Optison concentration on sonoporation in vitro. Ultrasound Med. Biol. 26 (7), 1169-1175. (20) Danialou, G., Comtois, A. S., Dudley, R. W., Nalbantoglu, J., Gilbert, R., Karpati, G., Jones, D. H., and Petrof, B. J. (2002) Ultrasound increases plasmid-mediated gene transfer to dystrophic muscles without collateral damage. Mol. Ther. 6 (5), 687-693. (21) Unger, E. C., Hersh, E., Vannan, M., Matsunaga, T. O., and McCreery, T. (2001) Local drug and gene delivery through microbubbles. Prog. CardioVasc. Dis. 44 (1), 45-54. (22) Christiansen, J. P., French, B. A., Klibanov, A. L., Kaul, S., and Lindner, J. R. (2003) Targeted tissue transfection with ultrasound destruction of plasmid-bearing cationic microbubbles. Ultrasound Med. Biol. 29 (12), 1759-1767. (23) Schneider, M., Bussat, P., Barrau, M. B., Bodino, F., Gotti, C., Hybl, E., Pelaprat, M. L., and Yan, F. (1991) A new ultrasound contrast agent based on biodegradable polymeric microballoons. InVest. Radiol. 26 Suppl. 1, S190-S191. (24) Bauer, A., Blomley, M., Leen, E., Cosgrove, D., and Schlief, R. (1999) Liver-specific imaging with SHU 563A: diagnostic potential of a new class of ultrasound contrast media. Eur. Radiol. 9 Suppl. 3, S349-S352. (25) Schlosser, T., Pohl, C., Kuntz-Hehner, S., Omran, H., Becher, H., and Tiemann, K. (2003) Echoscintigraphy: a new imaging modality for the reduction of color blooming and acoustic shadowing in contrast sonography. Ultrasound Med. Biol. 29 (7), 985-991. (26) Frinking, P. J., Bouakaz, A., de Jong, N., Ten Cate, F. J., and Keating, S. (1998) Effect of ultrasound on the release of microencapsulated drugs. Ultrasonics 36 (1-5), 709-712. (27) Hoff, L., Sontum, P. C., and Hovem, J. M. (2000) Oscillations of polymeric microbubbles: effect of the encapsulating shell. J. Acoust. Soc. Am. 107 (4), 2272-2280. (28) Frinking, P. J., and de Jong, N. (1998) Acoustic modeling of shellencapsulated gas bubbles. Ultrasound Med. Biol. 24 (4), 523-533. (29) Bouakaz, A., Frinking, P. J., de Jong, N., and Bom, N. (1999) Noninvasive measurement of the hydrostatic pressure in a fluidfilled cavity based on the disappearance time of micrometer-sized free gas bubbles. Ultrasound Med. Biol. 25 (9), 1407-1415. (30) Schneider, M., Arditi, M., Barrau, M. B., Brochot, J., Broillet, A., Ventrone, R., and Yan, F. (1995) BR1: a new ultrasonographic contrast agent based on sulfur hexafluoride-filled microbubbles. InVest. Radiol. 30 (8), 451-457. (31) Schneider, M., Bichon, D., Bussat, P., Puginier, J., and Hybl, E. (1991) Patent WO 91/15244. (32) Dufrene, Y., Barger, W., Green, J., and Lee, G. (1997) Nanometerscale surface properties of mixed phospholipid monolayers and bilayers. Langmiur 13, 4779-4784. (33) Bichon, D., Bussat, P., and Schneider, M. (1991) Eur. Patent EU 0,458,745 A1. (34) Schneider, M., Broillet, A., Bussat, P., and Ventrone, R. (1994) The use of polymeric microballoons as ultrasound contrast agents for liver imaging. InVest. Radiol. 29 Suppl. 2, S149-S151. (35) Schneider, M., Bussat, P., Barrau, M. B., Arditi, M., Yan, F., and Hybl, E. (1992) Polymeric microballoons as ultrasound contrast

Bioconjugate Chem., Vol. 18, No. 3, 2007 661 agents. Physical and ultrasonic properties compared with sonicated albumin. InVest. Radiol. 27 (2), 134-139. (36) Bloch, S. H., Wan, M., Dayton, P. A., and Ferrara, K. W. (2004) Optical observation of lipid- and polymer-shelled ultrasound microbubble contrast agents. Appl. Phys. Lett. 84 (4), 631-633. (37) Guzman, H. R., Nguyen, D. X., Khan, S., and Prausnitz, M. R. (2001) Ultrasound-mediated disruption of cell membranes. II. Heterogeneous effects on cells. J. Acoust. Soc. Am. 110 (1), 597606. (38) Huber, P. E., Jenne, J., Debus, J., Wannenmacher, M. F., and Pfisterer, P. (1999) A comparison of shock wave and sinusoidalfocused ultrasound-induced localized transfection of HeLa cells. Ultrasound Med. Biol. 25 (9), 1451-1457. (39) Miller, D. L., and Quddus, J. (2000) Sonoporation of monolayer cells by diagnostic ultrasound activation of contrast-agent gas bodies. Ultrasound Med. Biol. 26 (4), 661-667. (40) Nozaki, T., Ogawa, R., Loreto, F. B., Jr., Kagiya, G., Fuse, H., and Kondo, T. (2003) Enhancement of ultrasound-mediated gene transfection by membrane modification. J. Gene. Med. 5 (12), 10461055. (41) Zarnitsyn, V. G., and Prausnitz, M. R. (2004) Physical parameters influencing optimization of ultrasound-mediated DNA transfection. Ultrasound Med. Biol. 30 (4), 527-538. (42) Guzman, H. R., McNamara, A. J., Nguyen, D. X., and Prausnitz, M. R. (2003) Bioeffects caused by changes in acoustic cavitation bubble density and cell concentration: a unified explanation based on cell-to-bubble ratio and blast radius. Ultrasound Med. Biol. 29 (8), 1211-1222. (43) Pislaru, S. V., Pislaru, C., Kinnick, R. R., Singh, R., Gulati, R., Greenleaf, J. F., and Simari, R. D. (2003) Optimization of ultrasoundmediated gene transfer: comparison of contrast agents and ultrasound modalities(1). Eur. Heart J. 24 (18), 1690-1698. (44) Unger, E. C., McCreery, T., and Sweitzer, R. H. (1997) Ultrasound enhances gene expression of liposomal transfection. InVest. Radiol. 32 (12), 723-727. (45) Frenkel, P. A., Chen, S., Thai, T., Shohet, R. V., and Grayburn, P. A. (2002) DNA-loaded albumin microbubbles enhance ultrasoundmediated transfection in vitro. Ultrasound Med. Biol. 28 (6), 817822. (46) Guo, X., and Szoka, F. C., Jr. (2003) Chemical approaches to triggerable lipid vesicles for drug and gene delivery. Acc. Chem. Res. 36 (5), 335-341. (47) Wei, K., Skyba, D. M., Firschke, C., Jayaweera, A. R., Lindner, J. R., and Kaul, S. (1997) Interactions between microbubbles and ultrasound: in vitro and in vivo observations. J. Am. Coll. Cardiol. 29 (5), 1081-1088. (48) Moran, C. M., Anderson, T., Pye, S. D., Sboros, V., and McDicken, W. N. (2000) Quantification of microbubble destruction of three fluorocarbon-filled ultrasonic contrast agents. Ultrasound Med. Biol. 26 (4), 629-639. (49) Shi, W. T., Forsberg, F., Tornes, A., Ostensen, J., and Goldberg, B. B. (2000) Destruction of contrast microbubbles and the association with inertial cavitation. Ultrasound Med. Biol. 26 (6), 1009-1019. (50) Chomas, J. E., Dayton, P., Allen, J., Morgan, K., and Ferrara, K. W. (2001) Mechanisms of contrast agent destruction. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 48 (1), 232-248. (51) Chen, Q., Zagzebski, J., Wilson, T., and Stiles, T. (2002) Pressuredependent attenuation in ultrasound contrast agents. Ultrasound Med. Biol. 28 (8), 1041-1051. (52) Ward, M., Wu, J., and Chiu, J. F. (1999) Ultrasound-induced cell lysis and sonoporation enhanced by contrast agents. J. Acoust. Soc. Am. 105 (5), 2951-2957. (53) Miller, M. W., Miller, D. L., and Brayman, A. A. (1996) A review of in vitro bioeffects of inertial ultrasonic cavitation from a mechanistic perspective. Ultrasound Med. Biol. 22 (9), 1131-1154. (54) Dalecki, D., Child, S. Z., Raeman, C. H., Xing, C., Gracewski, S., and Carstensen, E. L. (2000) Bioeffects of positive and negative acoustic pressures in mice infused with microbubbles. Ultrasound Med. Biol. 26 (8), 1327-1332. (55) Dalecki, D. (2004) Mechanical bioeffects of ultrasound. Annu. ReV. Biomed. Eng 6, 229-248. (56) Lindner, J. R., and Kaul, S. (2001) Delivery of drugs with ultrasound. Echocardiography 18 (4), 329-337.

662 Bioconjugate Chem., Vol. 18, No. 3, 2007 (57) Stride, E., and Saffari, N. (2003) On the destruction of microbubble ultrasound contrast agents. Ultrasound Med. Biol. 29 (4), 563-573. (58) Dijkmans, P. A., Juffermans, L. J., Musters, R. J., van Wamel, A., Ten, Cate, F. J., van Gilst, W., Visser, C. A., de Jong, N., and Kamp, O. (2004) Microbubbles and ultrasound: from diagnosis to therapy. Eur. J. Echocardiogr. 5 (4), 245-256. (59) Dayton, P. A., Morgan, K., Klibanov, A. L., Brandenburger, G. H., Nightingale, K., and Ferrara, K. (1997) A preliminary evaluation of the effects of primary and secondary radiation forces on acoustic contrast agents. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 44 (6), 1264-1277. (60) Chomas, J. E., Dayton, P., May, D., and Ferrara, K. (2001) Threshold of fragmentation for ultrasonic contrast agents. J. Biomed. Opt. 6 (2), 141-150. (61) Marmottant, P., and Hilgenfeldt, S. (2003) Controlled vesicle deformation and lysis by single oscillating bubbles. Nature (London) 423 (6936), 153-156. (62) Lechardeur, D., Sohn, K. J., Haardt, M., Joshi, P. B., Monck, M., Graham, R. W., Beatty, B., Squire, J., O’Brodovich, H., and Lukacs,

Mehier-Humbert et al. G. L. (1999) Metabolic instability of plasmid DNA in the cytosol: a potential barrier to gene transfer. Gene Ther. 6 (4), 482-497. (63) Li, T., Tachibana, K., Kuroki, M., and Kuroki, M. (2003) Gene transfer with echo-enhanced contrast agents: comparison between Albunex, Optison, and Levovist in mice-initial results. Radiology 229 (2), 423-428. (64) Dolan, M. S., El Shafei, A., Puri, S., Tamirisa, K., St. Vrain, J., Flanagan, J., Havens, E., and Labovitz, A. J. (2001) For left ventricular opacification and endocardial border definition: is it really important which contrast agent we use, or is it the imaging modality we choose? Eur. J. Echocardiogr. 2 (3), 154-162. (65) Dayton, P. A., Chomas, J. E., Lum, A. F., Allen, J. S., Lindner, J. R., Simon, S. I., and Ferrara, K. W. (2001) Optical and acoustical dynamics of microbubble contrast agents inside neutrophils. Biophys. J. 80 (3), 1547-1556. BC0602432