Transfection of a reporter plasmid into cultured cells by sonoporation in vitro

Copyright

Ultrasound m Med. & Biol., Vol. 23. No. 6. pp. 953-959. 1997 0 1997 World Federation for Ultrasound in Medicme & Biology Printed in the USA. All rights reserved OXI-5629/97 $17.00 + .OO

PI1 SO301-5629(97)00025-2

ELSEVIER

@Original Contribution TRANSFECTION OF A REPORTER PLASMID INTO CULTURED CELLS BY SONOPORATION IN VITRO SHIPING BAO, * BRIAN

D. THRALL+ and DOUGLAS L. MILLER’

*USTUR Washington State University, Richland, WA; and ‘Battelle Pacific Northwest National Laboratory, Richland, WA (Received

13 December

1996:

in final

form

24 February

1997)

Abstract-Cultured Chinese hamster ovary cells were exposed to 2.25MHz ultrasound in sterile 45mL polyethylene chambers and tested for cell lysis, sonoporation and DNA transfection. Ten percent of Albunex@, a gas-body-based ultrasound contrast agent, was added to ensure cavitation nucleation, and the chambers were rotated at 60 rpm to promote cavitation activity during the 1-min exposures. Uptake of large fluorescent dextran molecules by some cells was observed for spatial peak pressure amplitudes as low as 0.1 MPa, which indicates transient permeabilization and resealing, Le., sonoporation, of these cells during exposure. Significant lysis occurred for 0.2 MPa, and increased rapidly for exposures above the apparent cavitation threshold (using the HzOz production test) of about 0.4 MPa spatial peak pressure amplitude. In the DNA transfection tests, 20 pg/mL luciferase reporter plasmid was added to the suspension during exposure, and cells were assayed for proliferation ability and luciferase gene expression 2 days after exposure. Cell proliferation was greatly reduced above the cavitation threshold. Luciferase production was significant for 0.20-MPa exposure, and reached 0.33 ng per lo6 cells at O.S-MPa exposure. The luciferase production was greater for cells exposed in medium supplemented with serum than for cells exposed in serum-free medium. Cells harvested for exposure either in the log phase or in the stationary phase of culture gave similar proliferation and transfection results. The effects essentially disappeared when the Albunex@ was omitted from the suspension and the tube was not rotated. Thus, sonoporation by ultrasonic cavitation in the rotating tube exposure system yields plasmid transfection with subsequent transient gene expression. 0 1997 World Federation for Ultrasound in Medicine & Biology Key Words: Ultrasound adverse effects, Contrast agent, Gene transfection, Gene therapy.

INTRODUCTION

utilized for opening transient pores in membranes and incorporating large molecules into surviving cells, a phenomenon termed electroporation (Neumann et al. 1989). Although the primary effect of ultrasound exposure in vitro is cell lysis induced by ultrasonic cavitation (Miller et al. 1996)) sublethal damage may also occur with passage of large molecules. Ultrasound treatment in a 20-kHz cell-disruption apparatus has been reported to induce a transient permeabilization of cell membranes in vitro, which leads to the uptake of external molecules into the cells (Fechheimer et al. 1987). Alterations in the permeability of neutrophils were induced during shock wave exposure in vitro. which produced an efflux of preloaded fluorescein from the cells and an influx of adriamycin into the cells from the extracellular medium (Holmes et al. 1992 ) . Fluorescent-labeled dextrans, which are normally not taken up by cultured cells, were loaded into the cells by shock wave exposure in the presence of the

Cell lysis results from irreversible cell membrane damage, which allows intracellular contents to leak out and vital dyes such as trypan blue to leak in and stain the interior of dead cells. Reversible membrane damage can also occur, with resealing of the holes (or pores). subsequent dye exclusion and survival of the cell. During such poration, large molecules in the medium can leak into the cells and then remain trapped after resealing. For example, red blood cells can be osmotically stressed to open membrane pores, and this phenomenon can be utilized for loading the cells with enzymes or other material of therapeutic potential (Ihler et al. 1973 ) . Several other mechanical methods, such as cell wounding, are also effective for loading molecules into the cytoplasm (McNeil 1989). Electric fields can be Address correspondence to: Dr. Douglas L. Miller, Battelle Pacific NW Laboratories, Mail Stop #P7-53. P.O. Box 999, Battelle Boulevard, Richland WA 99352, USA. 953

v53

Ukrasound

in Medicine

and Biology

molecules (Gambihler et al. 1994 ) . Similar molecularuptake effects have also been reported for cells exposed to laser-induced shock waves (Flotte et al. I995 ) Most studies have been performed with sparse suspensions of cultured cells in vitro. but erythrocytes can be loaded with fluorescent dextran of high molecular weight even in whole blood by exposure to lithotripter shock waves, which suggests that the phenomenon can occur in viva (Miller et al. 1997b). The ultrasonic induction of membrane permeabilization and resealing has been termed sonoporation. The ability to load cells with large molecules and subsequently to culture the cells opens the possibility of DNA transfection. Transfection of DNA and subsequent gene expression is well known to occur for electroporation (Neumann et al. 1989). Transfection also occurs with sonoporation. Sonication has been used for gene transfer in several types of plant cells (Joersbo and Brunstedt 1992; Zhang et al. 1991) . Many plant tissues contain gas-filled intercellular channels for respiration. which are strongly activated by ultrasound ( Miller 1987 ), and this cavitation-like phenomenon may be involved in the plant-tissue transfection ( Joersbo and Brunstedt 1992). Sonication treatment ( 20 kHz ) has been used to transfect cells with plasmid DNA and the subsequent marker-gene expression has been demonstrated both as transient and as stable transfection (Fechheimer et al. 1987). The observation of sonoporation by shock waves led to in vitro studies, which showed transfection of reporter plasmids by shock wave treatment (Lauer et al. 1994)) and to the proposed use of lithotripter treatment as a gene therapy method (Delius et al. 1995). Recently, Kim et al. ( 1996) demonstrated ultrasound-mediated transfection of mammalian cells using l-MHz ultrasound, which seemed to be cavitation related because of concomitant cell lysis and the failure to produce the effect at a higher frequency (3.5 MHz). In this research, DNA transfection by sonoporation was examined in cultured mammalian cells. The sterile cell cultures were seeded with cavitation nuclei. and a rotating tube exposure system was utilized to ensure controlled production of cavitation effects. Effects on cell membranes were examined by testing for fluorescent dextran uptake during exposure and dye exclusion after exposure. Cells were cultured after exposure to observe proliferative ability and DNA transfection. A commercial reporter plasmid, which yields luciferase enzyme production in mammalian cells, allowed sensitive luminescence assay of plasmid gene expression in surviving cells proliferating in culture. The sonoporation method of DNA transfection was tested to assess several potential advantages over other transfection methods. particularly for use in viva.

Volume

1-3, Number

6. 1997

MATERIALS

AND METHODS

Chinese hamster ovary (CHO) cells were maintained as monolayers in 150-cm’ tissue culture flasks (Coming Inc., Coming, NY, USA) at 37°C. in a humidified atmosphere of 5% CO-, in air. The growth medium was Ham’s F,? supplemented with 10% fetal bovine serum (Sigma Chemical Co., St. Louis, MO. USA) and 50 pg/mL gentamicin (Gibco BRL, Grand Island, NY, USA). To reduce genetic drift. new cultures were initiated from frozen stocks every 2 months. A luciferase reporter vector (control plasmid pGL2, Promega Corp., Madison, WI, USA) was used as the plasmid for insertion and subsequent transient expression within the cells. To generate sufficient quantities of DNA, E. coli-JM109 bacteria were transformed with the vector by the CaC& method, and plasmid DNA was isolated from the cultured bacteria (Plasmid Mega Kit, Qiagen, Inc., Chatsworth, CA, USA). The bacterial cells were harvested and lysed in NaOH/SDS (sodium dodecyl sulfate) buffer in the presence of RNase. The lysate was neutralized by the addition of acid potassium acetate. The precipitated debris were removed by high-speed centrifugation, producing a clear lysate for loading onto the Qiagen tip (anion-exchange resin). The Qiagen tip was then washed with buffer containing 1 M NaCl to remove remaining contaminants, such as traces of RNA and protein. The plasmid DNA was then eluted from the resin with buffer containing 1.25 M NaCl at pH 8.5, desalted and concentrated by isopropanol precipitation. After centrifugation, the pellet was washed with 70% ethanol to remove residual salt and to replace the isopropanol with ethanol. The purified DNA was briefly air-dried and resuspended in a small volume of TE buffer ( 10 mM Tris-HCl and 1 mM EDTA) The purification and concentration of isolated plasmid DNA were confirmed by agarose gel analysis and spectrophotometry. The ultrasound exposure system has been described previously (Miller and Thomas 1995 ) . Briefly. a 1.9-cm diameter, air-backed transducer operating at 2.25 MHz was mounted 12.5 cm from the exposure chamber in a 37°C water bath. A series of exposure levels were calibrated and arranged in steps differing by a factor of J2 of the mean spatial peak pressure amplitude ( e.g., 0.1 MPa, 0.14 MPa, 0.2 MPa, etc. ) The peak positive, peak negative and mean pressure amplitudes for several levels are listed in Table 1. The -6 dB beam width was 9.0 mm. The exposure chambers were polyethylene pipette bulbs (Sigma Chemical Co.) of approximately 4.5mL volume ( 12.5-mm inside diameter, 4.3 cm long). The ultrasound transmission through the 0.35 to 0.42-mm thick wall of the

Transfection

of a reporter

Table 1. Peak positive and negative pressure amplitudes detected at the position of the chamber for several of the exposure levels.* Mean

amplitude (MPa)

Peak

0.7 0.4 0.8 Values

positive WPa) 0.22 0.5 1 1.12

given

as mean

Peak negative W’a) 0.18 0.3 1 0.5 1

amplitudes.

chambers was checked by cutting a chamber in half lengthwise and holding one half between the transducer and the hydrophone. Transmission was 94%97% of the incident pressure amplitudes, which indicates that internal reflections (and other possible beam perturbations) were minimal in the chambers. The exposure system was evaluated for cavitation activity by observing cell lysis and sonoporation effects. Harvested cells in medium with FBS were mixed with 10% fluorescent dextran stock solution (100 mg/ mL in PBS) and 10% Albunex@ (Mallinckrodt Medical, Inc., St. Louis, MO, USA) to give a final concentration of lo6 cells mL -’ . The relatively high Albunex@ concentration was used to ensure that enough of this agent, which is sensitive to handling, remained at the time of exposure to reliably initiate cavitation activity. The FITC-dextran (FD-SOOS, fluorescein isothiocyanate-dextran, Sigma Chemical Co.) has an average molecular weight of 580,000 and is normally excluded from the cells. This mixture was loaded into the exposure chambers and exposed for 1 min with 60-rpm rotation. After exposure and removal from the chamber, the cell suspension was centrifuged, and the supernatant with the fluorescent dextran was discarded. The cells then were resuspended in fresh medium with FBS and placed on ice. Cell viability (trypan blue dye exclusion) and cells with internal fluorescent dextran were counted on a hemocytometer under a fluorescence microscope arranged to alternately observe the specimen in transmitted light or in fluorescence mode for the 490-nm excitation and 525nm emission of FITC. The exposure system was also specifically evaluated for inertial cavitation activity by assessing hydrogen peroxide production. As noted previously, gasbody-based contrast agents such as Albunex@ can be used to nucleate cavitation activity in media that lack suitable cavitation nuclei and would not normally be subject to cavitational effects (Miller and Thomas 1995). A solution of 10% Albunex@ in phosphatebuffered saline (PBS ) was exposed in the pipette-bulb chambers at 37°C. After exposure, the bulb was emptied and the contents assayed for hydrogen peroxide

plasmid

0 S. BAO et al.

955

production using the isoluminol assay method (Miller and Thomas 1995). For the transfection experiments, CHO cells were harvested by trypsinization and suspended at a final concentration of 10” cells rnI-’ with 20 pg/mL plasmid DNA and 10% Albunex@. The sterile chambers were loaded with the suspension in a laminar flow hood under aseptic conditions and sealed by heating the neck. The sealed chambers were then mounted in the water bath, which was filled with degassed water at 37°C. Ultrasound exposure was for 1 min with the chamber rotated at 60 ‘pm. After exposure. the cell suspensions were removed from the chamber under aseptic conditions, plated in 60-mm Coming tissue culture dishes with 0.5 mL of exposed suspension into 5 mL of prewarmed medium with serum per dish (i.e., a density of 5 X lo5 cells per dish based on the preexposure cell concentration) and placed in the 37°C incubator. Cell membrane integrity was checked by trypan blue exclusion. After 2 days of culture in medium with serum, the cells were harvested and assessed for cell proliferation during culture and expression of the reporter plasmid. Cell proliferation was evaluated by harvesting and counting the cells on a hemocytometer. The luciferase activity was measured using a luciferase assay system (Promega Corp.). Cells were harvested with the lysis reagent and scraped from the dish. After a brief centrifugation, the cell extract was mixed with luciferase assay reagent and light output measured using a luminometer (TD-20e, Turner Designs, Mountain View, CA. USA). To determine the actual concentration of luciferase, a standard curve was obtained using known solutions of luciferase in the lysis reagent. This sensitive measurement system provided a linear response over a range of at least 1 pg to 1 ng (per 20-PL sample ) . RESULTS Results for trypan blue exclusion and sonoporation are shown in Fig. 1, as the percent of the total intact cell number (i.e., stained or unstained by trypan blue). This experiment was initially run four times for exposures in the range of 0.2-0.8 MPa. Subsequent analysis of the results indicated a possible ultrasoundinduced effect even for the lowest exposure. The experiment was later repeated several times for exposure as low as 0.05 MPa, finally resulting in a total of eight repetitions from 0.05-0.2 MPa and seven from 0.280.8 MPa. The decline in trypan-blue-excluding cells was significantly different from shams for the 0.2-MPa exposure level. However, counts of fluorescent cells, which were zero in sham-exposed samples, reached

Ultrasoundin Medicineand Biology

4% 100

Volume 23. Number 6. 1997

T=

02000 0.0 ,ti

0.4 Pressure Amplitude

0.0

0.2

0.4

0.6

0.8

Fig. 1. Trypan blue viability (solid circles) and cells with FITC-dextran uptake (open circles) after exposure of cells suspensions with serum. The mean values are plotted with standard error bars for 7-8 repetitions (see text).

statistical significance at 0.1 MPa (p < 0.01) . Effects increased rapidly above 0.28 MPa: cell lysis reduced the intact cell population to about 75% of shams for 0.4- to 0.8-MPa exposure, and roughly half of the intact cells had intracellular fluorescent dextran. There was no significant reduction in the total numbers of counted cells (i.e., stained plus unstained). Only a few cells (data not shown) were seen with both trypan blue staining and fluorescence (probably because the dextran would leak out of cells with loss of membrane integrity). Results for the hydrogen peroxide tests are shown in Fig. 2, with significant production noted at 0.4 MPa. Hydrogen peroxide production increased gradually above this apparent threshold, reaching about 0.4 PM

,

0.0 4 0.0

0.2

0.4

0.6

0.8

06

0.8

1.0

(MPa)

1.0

Pressure Amplitude(MPa)

0.5

02

1.0

Pressure Amplitude(MPa) Fig. 2. The production of hydrogen peroxide in saline supplemented with 10% Albunex@. The mean values are plotted with standard error bars.

Fig. 3. Results of ultrasound exposure without serum and subsequent culture in fresh medium with serum for 2 days. The cell proliferation relative to shams (solid circles) declines in concert with the production of luciferase (open circles) from the transfection of marker plasmid. The mean values are plotted with standard error bars for four repetitions.

at 0.8 MPa. During these exposures it was noted that the cloudy Albunex@ suspension was vigorously stirred and cleared within a few rotations of the chamber for exposures of 0.4 MPa or higher. Cell proliferation decreased dramatically for increasing pressure amplitudes, to about 25% for 0.40.8 MPa as shown in Fig. 3 for exposures without FBS (subsequent culture was with FBS). This result indicates much greater damage, which gave lower survival or slower growth of the cultures, than was expected from the trypan blue exclusion measurements made soon after exposure (data, not shown, were similar to those shown in Fig. 1). Luciferase production was significantly above the sham-exposed sample for 0.28 MPa (the lowest exposure in this experiment), and reached 0.13 ng per million cells at 0.8 MPa. Exposure at 0.8 MPa of cell suspensions without Albunex@ and without tube rotation, which minimizes cavitation, yielded cell proliferation essentially equal to shams. Luciferase activity was very slightly increased in this test, however, from 1.7 X 10m5 ng per million shamexposed cells to 3.1 X 10 -’ ng per million exposed cells (p < 0.05 for four repetitions). Results are shown in Fig. 4 for exposures with 10% FBS in the medium. The reduction in ceil proliferation was significant for 0.28 MPa exposure (but not 0.2 MPa). Again, actual proliferation was much less than expected from the dye exclusion tests; for example, at 0.8 MPa, 60% of the cells exchlded trypan blue soon after exposure, but proliferation was only 5% of shams. Luciferase production was significantly different from sham-exposed samples even for 0.2-MPa exposure. These results with serum were larger than the results without serum. Because the two exposure-re-

Transfection

of a reporter 0.4 F 0.3 @

plasmid

l S. BAO er cd.

Table 3. Comparison of results for three 0.56-MPa exposures of cells in logarithmic growth or in stationary phase (confluent). Cell proliferation (percent of sham)

!i > 8. o.2 4 3 0.1 Jj OC0.0

=

~liferation

f 0.2

0.4

0.6

Pressure Amplitude

0.8

o\ P E

(MPa)

Fig. 4. Results of ultrasound exposure with serum after 2 days of culture. The cell proliferation relative to shams (solid circles ) and luciferase production (open circles) are plotted as mean values with standard error bars for three repetitions. sponse experiments (Figs. 3 and 4) were conducted separately, it was uncertain whether or not the apparent enhancement of effects with FBS was valid. A comparison test was therefore run with 0.56-MPa exposure of samples with or without added FBS (subsequent culture in medium with serum). Results, presented in Table 2. validate the improved results for suspensions having serum during exposure. A final comparison test was conducted to assess the influence of the culture phase on the plasmid transfection and expression. Logarithmically growing cultures were harvested 1 day after culture initiation. Stationary phase cells were harvested 5 days after culture initiation, at which time the cells had become confluent and stopped dividing. The timing of the cultures allowed exposure on the same day, and all other procedures were identical. Results are presented in Table 3 for three exposures at 0.56 MPa with serum. There were no significant differences between results for cells in log or stationary phases of culture either for proliferation or for luciferase production.

DISCUSSION Cell lysis and cytoplasmic loading with high molecular weight dextran were induced by ultrasound in

Cell proliferation (percent of sham) 10.2 (1.4) X.5 (1.6) NS

Without FBS With FBS f-test Standard

deviations

are shown

Luciferase (ng per

production 1Oh cells)

0.20 (0.05) 0.32 (0.05) p < 0.05 in parentheses.

Log growth Stationary f-test Standard

5.9 (0.32) 8.5 (1.6) NS deviations

are shown

Luciferasc trig per

production 1Oh cells)

0.3 1 10.03) 0.32 (0.06) YS in parentheses.

0.0

1.0

Table 2. Comparison of results for three repetitive exposures at 0.56 MPa with or without serum.

957

vitro. Cultured CHO cells were exposed to 2.25-MHz ultrasound in a rotating tube exposure system. with a gas-body-based contrast agent added to help initiate cavitation activity. With 10% fluorescent-labeled dextran added to the medium, about half of the cells escaping cell lysis at 0.8 MPa contained the large molecule. The transient permeabilization and resealing, i.e., sonoporation, apparently traps the large molecules. which are normally excluded by viable cells. Sonoporation also allowed the transfection and subsequent expression of luciferase reporter plasmid in the cultured cells. Cell proliferation was reduced for cells cultured after exposure, and the reduction was in excess of that expected on the basis of the trypan blue dye exclusion test. For example, exposure at 0.8 MPa with serum reduced dye exclusion to 60% of shams, but cell proliferation counts after 2 days of growth were only 5% of shams. The difference in viability indicators. which has been observed previously for in vitro studies (Miller et al. 1996). may be due to delayed cell proliferation, additional cell death after exposure, or both. Since the proliferation is so different from immediate survival, it is uncertain whether the FITC-dextran results for the rate of sonoporation accurately reflects the DNA transfection efficiency in surviving cells. DNA transfection was evaluated by adding 20 pg/ mL luciferase reporter plasmid to the suspensions and assaying the surviving cells for luciferase production 2 days after exposure. Cell proliferation decreased and transfection increased for exposures above the apparent cavitation threshold (using the HZO1 sonochemicalproduction test for inertial cavitation activity) of about 0.28- to 0.4-MPa spatial peak pressure amplitude. With reporter plasmids in the medium during exposure, subsequent luciferase production reached about 0.3 ng per 10” cells for exposures with serum in the suspension. The luciferase production was greater on a per-cell basis for cells exposed in medium supplemented with serum than for cells exposed in serum-free medium. There was no significant difference in effects on cells from cultures in logarithmic growth compared to cells in stationary cultures.

958

Ultrasound

0.00

0.10

0.05

Pressure

Amplitude

in Medicine

0.15

and Biology

0.20

(MPa)

Fig. 5. Counts of cells with membrane damage for the lower exposures (replotted from Fig. 1) . Trypan-blue stained cells (solid circles) are compared to fluorescent cells (open circles) as the mean cell counts with standard error bars for eight repetitions.

The membrane damage effects occurred at lower exposures than might have been expected on the basis of cavitation thresholds. For example, the apparent threshold for H202 production was 0.4 MPa, compared to significant FITC-dextran uptake for O.l-MPa exposure. The FITC-dextran uptake phenomenon appears to be a particularly sensitive test for membrane damage. This is illustrated by replotting the data for the lowest exposures from Fig. 1 in Fig. 5 as the actual counts of trypan blue stained cells (rather than the viable fraction) compared to the counts of fluorescent cells. Due to the zero background level of fluorescent cells in the shams, the sonoporation test appears to be a more sensitive test of membrane damage than the traditional dye exclusion test of cell viability. Only very low level effects were seen in suspensions exposed without Albunex@ and tube rotation, which further implicates the cavitation mechanism in the production of these effects. The Albunex@ was rapidly destroyed by the exposure, which implies that most of the effects were caused by ultrasonically induced cavitation rather than by any direct influence of the Albunex@. This agent is acoustically labile at amplitudes as low as 0.1 MPa (Miller et al. 1997a; Vandenberg and Melton 1994), so even the small sonoporation effects seen at low pressure amplitudes may be due to bubbles freed from the stabilized gas bodies. Cavitation activity becomes inertial above a physically defined threshold, and the occurrence of inertial cavitation (also called transient cavitation or collapse cavitation j introduces several additional bioeffects mechanisms (e.g., sonochemical production). The exposureresponse curve for hydrogen peroxide observed in this

Volume

23, Number

6. 1997

study is quite similar to results from an earlier study of inertial cavitation nucleation (Miller and Thomas 1995)) and the occurrence of sonochemical-related DNA strand breaks appears to follow this trend ( Miller et al. 199.5 ) However, the gradual increase in the exposure-response curve for the hydrogen peroxide production above the relatively high threshold of 0.4 MPa (Fig. 2) was distinctly different from the response of the reduced proliferation and DNA transfection. which indicates that mechanical processes associated with the cavitation were responsible for the effects (rather than sonochemical or other processes associated with inertial cavitation). These results are consistent with the earlier results of Fechheimer et al. ( 1987) for 20-kHz exposure and Kim et al. (1996) for l-MHz exposure. In this study, use of gas-body seeding and tube rotation further validate the association of the plasmid transfection with acoustic cavitation and sonoporation. Many biological, chemical, mechanical and electrical transfection methods with a variety of characteristics are now available (Cooper 1996; Mulligan 1993 ) . The ultrasonic transfection method may have some useful differences from the others. For example, the presence of serum in the medium can reduce the efficiency of some chemical methods but appears to enhance the ultrasonic effect. In addition, transfection was not reduced for cells in the stationary phase of culture. Kim et al. ( 1996) noted that even primary cells could be transfected by sonoporation in vitro. Furthermore, ultrasound has physical features, such as beam formation for imaging or localized therapy, which are not available to other methods. Thus, the ultrasonic transfection method has unique characteristics that may be of value in developing therapeutic procedures, particularly in viva where ultrasound can be focused on selected regions. Acknowledgement-This Institutes of Health

grant

research CA42947.

was supported

by the National

REFERENCES Cooper MJ. Noninfectious gene transfer and expression systems for cancer gene therapy. Semin Oncol 1996;23:172-187. Delius M, Hofschneider P, Lauer U, Messmer K. Extracorporeal shock waves for gene therapy? Lancet 1995;345:1377. Fechheimer M, Boylan JF, Parker S, et al. Transfection of mammalian cells with plasmid DNA by scrape loading and sonication loading. Proc Natl Acad Sci USA 1987;84:8463-8467. Flotte TJ, Lee S, Zhang H, et al. Laser-induced stress transients: Applications for molecular delivery. SPIE 1995; 2392:202-207. Gambihler S, Delius M, Ellwart JW. Permeabilization of the plasma membrane of L1210 mouse leukemia cells using lithotripter shock waves. J Membr Biol 1994; 141:267-275. Holmes RP, Yeaman LD, Taylor RG. McCullough DL. Altered neutrophil permeability following shock wave exposure in virro. J Urol 1992; 147:733-737. Ihler GM, Glew RI-I, Schnure F?V. Enzyme loading of erythrocytes. Proc Nat1 Acad Sci USA 1973;70:2633-2666.

Transfection of a reporter plasmid 0 S. BAO et al. Joersbo M, Brunstedt J. Sonication: A new method for gene transfer to plants. Physiol Plant 1992;85:230-234. Kim HJ, Greenleaf JF, Kinnick RR, Bronk JT, Bolander ME. Ultrasound-mediated transfection of mammalian cells. Human Gene Ther 1996;7:1339-1346. Lauer U, Squire 2, Burgelt E, et al. Towards a new gene transfer system: Shock wave mediated DNA transfer. J Cell Biochem Suppl 1994; 18A:226. McNeil PL. Incorporation of macromolecules into living cells. Meth Cell Biol 1989;29:153-173. Miller DL. A review of the ultrasonic bioeffects of microsonation, gas-body activation, and related cavitation-like phenomena. Ultrasound Med Biol 1987: 13:443-470. Miller DL, Thomas RM. Ultrasound contrast agents nucleate inertial cavitation in vitro. Ultrasound Med Biol 1995;21:1059- 1065. Miller DL, Gies RA, Chrisler WB. Ultrasonically induced hemolysis at high cell and gas body concentrations in a thin-disc exposure chamber. Ultrasound Med Biol 1997a; in press.

959

Miller MW, Miller DL, Brayman AA. A review of in virro bioeffects of inertial ultrasonic cavitation from a mechanistic perspective. Ultrasound Med Biol 1996;22: 1131- 1154. Miller DL, Thomas RM, Buschbom RL. Comet assay reveals DNA strand breaks induced by ultrasonic cavitation in l&-o. Ultrasound Med Biol 1995;21:841-848. Miller DL, Williams AR, Morris JE, Chrisler WB. Sonoporation of erythrocytes by lithotripter shockwaves in vitro. Ultrasonics 1997b; in press. Mulligan RC. The basic science of gene therapy. Science 1993;260:926-932. Neumann E, Sowers AE, Jordan CA. Electroporation and electrofusion in cell biology. New York: Plenum Press, 1989. Vandenberg BF, Melton HE. Acoustic lability of albumin microspheres. J Am Sot Echocardiogr 1994;7:582-589. Zhang L, Cheng L, Xu N, et al. Efficient transformation of tobacco by ultrasonication. Biotechnology 1991:9:996-997.