Use of Ultrasound to Enhance Nonviral Lung Gene Transfer in Vivo

Gene Therapy (2007) 14, 768–774 & 2007 Nature Publishing Group All rights reserved 0969-7128/07 $30.00 www.nature.com/gt ORIGINAL ARTICLE Use of ultrasound to enhance nonviral lung gene transfer in vivo

S Xenariou1,2, U Griesenbach1,2, H-D Liang3,4, J Zhu1, R Farley1,2, L Somerton1,2, C Singh1,2, PK Jeffery1, S Ferrari1,2, RK Scheule5, SH Cheng5, DM Geddes1,2, M Blomley3 and EWFW Alton1,2 1Department of Gene Therapy, National Heart and Lung Institute, Faculty of Medicine, Imperial College, London, UK; 2UK Cystic Fibrosis Gene Therapy Consortium, London, UK; 3Ultrasound Group, Imaging Sciences Department, MRC Clinical Sciences Centre, Faculty of Medicine, Imperial College, London, UK; 4School of Engineering, University of Cardiff, Cardiff, UK and 5Genzyme Corporation, Framingham, MA, USA

We have assessed if high-frequency ultrasound (US) can being attenuated by air in the lung, sufficient energy enhance nonviral gene transfer to the mouse lung. Cationic penetrates the tissue to increase gene transfer. US-induced lipid GL67/pDNA, polyethylenimine (PEI)/pDNA and naked lung haemorrhage, assessed histologically, increased with plasmid DNA (pDNA) were delivered via intranasal instilla- prolonged US exposure. The left lung was more affected tion, mixed with Optison microbubbles, and the animals were than the right and this was mirrored by a lesser increase then exposed to 1 MHz US. Addition of Optison alone in naked pDNA gene transfer, in the left lung. The positive significantly reduced the transfection efficiency of all three effect of US was dependent on Optison, as in its absence US gene transfer agents. US exposure did not increase GL67/ did not increase naked pDNA transfection efficiency. We pDNA or PEI/pDNA gene transfer compared to Optison- have thus established proof of principle that US can increase treated animals. However, it increased naked pDNA trans- nonviral gene transfer, in the air-filled murine lung. fection efficiency by approximately 15-fold compared to Gene Therapy (2007) 14, 768–774. doi:10.1038/sj.gt.3302922; Optison-treated animals, suggesting that despite ultrasound published online 15 February 2007

Keywords: sonoporation; ultrasound; gene transfer; nonviral vectors; lung; cystic ﬁbrosis

Introduction Both viral and nonviral vectors have been used in CF clinical trials, delivered either to the nasal epithelium Our aim is to establish a clinically relevant treatment for or the lung of CF patients.5,6 Our particular focus is on cystic fibrosis (CF), a monogenic, lethal disease, in which nonviral vectors because of their capacity to be repeat- pulmonary failure is the major cause of mortality. We are, edly administered,7 a critical need in the treatment of a thus, interested in optimizing lung gene transfer, and lifelong condition. Proof of principle for the delivery and more specifically, gene transfer to the respiratory airway expression of cystic fibrosis transmembrane conductance epithelium. The lung offers the advantage of allowing for regulator (CFTR), the gene mutated in CF, with nonviral topical administration of therapeutic agents, usually vectors has been established, in preclinical models8 as via nebulization. In spite of this, it has proved to be well as in clinical trials.5,9 However, in general, nonviral a formidable target for gene delivery, with a series of gene transfer agents are not as efficient as viruses, since extracellular barriers such as mucociliary clearance and they lack the natural mechanisms to invade cells, and mucus, and in the case of CF, a layer of infected sputum. CFTR expression in clinical trials resulted in partial In addition, there are several intracellular obstacles, correction of the bioelectrical defect.10 Attempts to including clearance by uptake into endosomes, and the improve the efficiency of nonviral vectors are being nuclear membrane that must be breached.1 Efforts made, including the addition of peptide ligands to target to overcome these barriers with the use of mucolytic the vector to receptors localized on the apical membrane agents,2 viscoelastic gels,3 or agents that disrupt tight of airway epithelial cells,11 and the development of junctions, thus allowing access to receptors present on nanoparticles to improve nuclear entry.12 Here we have the basolateral membrane,4 have enhanced transfection assessed, if sonoporation, the use of high-frequency efficiency to airway epithelium, but further improve- ultrasound (US) increases nonviral gene transfer to the ments would always be welcome. lung. US is a versatile technique, and has several medical applications. High-frequency (41 MHz), low-intensity Correspondence: Dr U Griesenbach, Department of Gene Therapy, US is commonly used for imaging, for example to detect National Heart and Lung Institute, Imperial College, 1B Manresa foetal abnormalities. In contrast, US shock waves, at Road, London SW3 6LR, UK. E-mail: [email protected] lower frequencies but higher intensities, are used to Received 5 September 2006; revised 21 December 2006; accepted 21 destroy gallbladder or kidney stones (lithotripsy), December 2006; published online 15 February 2007 whereas a similar type known as high-intensity focused Use of ultrasound to enhance nonviral lung gene transfer in vivo S Xenariou et al 769 ultrasound (HIFU) is currently being evaluated in 100 preclinical models as a method to ablate tumours.13 In *** *** *** terms of gene transfer, US has been applied in several in vitro and in vivo systems. Both shock waves and HIFU 13,14 have been used. In most gene delivery studies, 10 however, the US applied has been in the range of 1–2 MHz. Sonoporation has increased gene transfer of naked plasmid DNA (pDNA) and cationic liposomes in several cell lines,15 including vascular,16 and skeletal 1 muscle cells.17 It has also been used to enhance nonviral gene transfer to tumours,18 the arterial wall,19 skeletal 20 21 muscle, and liver, in vivo. US has also been success- Luciferase (RLU/mg protein) fully used to increase adenoviral transfection efficiency to the myocardium.22 0.1 US is thought to act by transiently permeabilizing Naked pDNA PEI/pDNA GL67/pDNA the cell membrane, thus increasing vector uptake. The Figure 1 Optison reduces the transfection efficiency of naked mechanism behind this effect is thought to be cavitation, pDNA, PEI and GL67/pDNA complexes. Animals were instilled the formation and oscillation of gas bubbles, in a liquid with naked pDNA (100 mg/100 ml), PEI/pDNA (10 mg/100 ml) or medium. The size of the bubbles fluctuates in response to GL67/pDNA (40 mg/100 ml) complexes, in the absence (closed bars) or presence (open bars) of Optison (1:1 (v/v) ratio). ***Po0.001, the pressure fluctuations of the US wave, and ultimately compared to the respective control group of vector alone. n ¼ 6–8 23 the bubbles collapse, creating pores in cell membranes. mice/condition. The theory that cavitation is responsible for sonoporation effects is also supported by the fact that addition of gas- filled microbubbles, further augments US-mediated gene appeared healthy, we decided to apply US for 20 min delivery. Thus, microbubbles, and in particular, Optison, in further experiments. We, thus, went on to instill a have been used in several sonoporation studies.13,24 mixture of GL67/pDNA and Optison (40 mg pDNA/ Optison is an US contrast agent, used for medical 100 ml, 1:1 (v/v)), or PEI/pDNA and Optison (10 mg imaging, consisting of albumin microspheres, filled with pDNA/100 ml, 1:1 (v/v)) complexes with subsequent octafluoropropane.25 When compared to similar avail- exposure to US for a total of 20 min. Sonoporation, able agents, Optison was superior for US-mediated gene however, did not increase GL67 (GL67+Optison: transfer to skeletal muscle in vivo.26 A recent study also 5.4571.00 RLU (relative light units)/mg protein, GL67/ demonstrated that the effects of US on the cell membrane pDNA+Optison+US: 4.8571.69 RLU/mg protein) or are much more profound in the presence of Optison, PEI-mediated gene transfer (PEI/pDNA+Optison: providing further support for the role of cavitation.27 3.1470.90 RLU/mg protein, PEI/pDNA+Optison+US: Thus,wehavetested,forthefirsttime,theeffectof 2.9870.47). sonoporation on nonviral gene transfer to the lung. It is well established, that ultrasound is greatly attenuated when Sonoporation increases naked pDNA gene transfer in travelling through air.28 Reflection of the wave, and there- the air-filled lung when mixed with Optison fore, energy loss, will also occur as it passes through air/ tissue interfaces.28 Nonetheless, we assessed whether the We also tested the effect of sonoporation on naked pDNA m m energy that is able to penetrate the lung tissue was sufficient (100 g pDNA/100 l, 1:1 (v/v)). Importantly, a 20-min ¼ to enhance the gene delivery of cationic lipid 67 (GL67)/ US exposure led to a significant (Po0.001, n 7–8 mice/ pDNA or polyethylenimine (PEI)/pDNA complexes, and condition) increase in gene expression, of approximately naked pDNA, all previously used for lung gene transfer.29–31 1.5 logs (Figure 2). To assess if we could further enhance this effect, animals were exposed to US for 1 h. However, 2/2 animals died after 35–40 min of exposure, and thus, Results we did not continue with this. In addition, we included a group of animals exposed to US for 2 min, to determine Optison reduces nonviral gene transfer in the mouse if the positive effect previously seen with 20 min, could lung also be achieved at shorter exposures. As shown in Initial studies were carried out to investigate the effect of Figure 2, the 2-min exposure also resulted in a significant Optison on nonviral vectors. Naked pDNA, PEI/pDNA (Po0.001, n ¼ 7–8 mice/group) increase in luciferase and GL67/pDNA, mixed with Optison at 1:1 (v/v) ratio, expression, compared to the control group. However, the were delivered to the mouse lung. In all cases, the levels of expression obtained with 2 min exposure were addition of Optison significantly reduced reporter gene lower than with 20 min, although that difference was not expression (Figure 1). Naked pDNA was also mixed with statistically significant. Optison at a ratio of 1:0.25 (v/v, DNA/Optison), but this formulation also significantly reduced transfection Prolonged US exposure causes lung haemorrhage efficiency (data not shown). A common side effect of US exposure is microvessel damage and tissue haemorrhage. Therefore, we assessed Sonoporation does not increase GL67/pDNA or PEI/ if haemorrhage had occurred in the mouse lung. US pDNA transfection efficiency when mixed with Optison exposure for 2 min did not increase haemorrhage To determine suitable US conditions, mice were instilled compared to control groups (Figure 3), but this was with phosphate-buffered saline (PBS; 100 ml, n ¼ 3) and visible after 20 min of US exposure. Interestingly, the exposed to 1 MHz US for 10 or 20 min. As animals degree of haemorrhage was greater in the left lung than

Gene Therapy Use of ultrasound to enhance nonviral lung gene transfer in vivo S Xenariou et al 770 in the right (Figure 4a). The area affected also increased differences in the levels of gene expression achieved by with exposure time and, once again, the left lobe was sonoporation of naked pDNA, between the right and left more affected than the right (Figure 4b). This is likely lobes (Figure 5). Although there were no significant owing to the fact that it lies directly beneath the rib cage differences between right and left lungs in luciferase and may be more exposed than the right lobes, which expression in any group, there was a trend towards are smaller and fold onto each other. Low levels of reduced gene expression in the left lung. In keeping with haemorrhage were seen in animals treated with naked this, exposure to US for 2 min did not enhance gene pDNA alone, but the same levels were also observed in transfer in the left lung, only the 20-min exposure untreated controls (data not shown). produced a significant (Po0.05, n ¼ 7–8 animals/condition) increase of luciferase expression. US-mediated naked pDNA gene transfer produces a reduced effect Sonoporation has no effect on gene transfer in the Because of the ‘uneven’ distribution of US-mediated absence of Optison toxicity in the lung, we determined whether there were As seen above (Figure 1), Optison reduced the transfection efficiency of all three, gene transfer agents. In addition, Optison microbubbles are very ‘fragile’ and 12 would most likely not withstand nebulization, the likely 11 *** relevant method of vector delivery for CF clinical trials. 10 We, therefore, assessed whether US could improve gene 9 transfer, in the absence of Optison. 8 *** Immediately after nasal instillation, animals were 7 exposed to US for a total of 20 min. Sonoporation did 6 *** not increase reporter gene expression of any of the gene 5 transfer agents (pDNA: 4.1570.78 RLU/mg protein, 4 pDNA+US: 6.1470.79; PEI/pDNA: 3.4970.86 RLU/mg 3 protein, PEI/pDNA+US: 1.6670.23; GL67/pDNA: 2 31.0174.47 RLU/mg protein, GL67/pDNA+US: 41.017 Luciferase (RLU/mg protein) 1 8.28 RLU/mg protein; n ¼ 6–10 mice/condition). 0 pDNA No US 2 min US 20 min US pDNA + Optison Discussion Figure 2 Sonoporation of naked pDNA in the presence of Optison. Sonoporation has been successfully applied to increase Animals instilled with a mixture of naked pDNA and Optison (1:1 gene transfer in several organs, other than lung, using (v/v) ratio) were exposed to US for 2 or 20 min. The bars represent the levels of luciferase achieved in the trachea and right lobes. both viral and nonviral vectors. We assessed, for the first The open bar represents the luciferase levels with pDNA alone. time, whether the technique could also be used to ***Po0.001; n ¼ 7–8 mice/condition. enhance nonviral gene transfer to the lung. Traditionally,

Figure 3 Prolonged US exposure exacerbates lung haemorrhage. Lung sections from animals treated with (a) naked pDNA, (b) pDNA+Optison (1:1 (v/v) ratio), (c) pDNA+Optison (1:1 (v/v))+2 min US and (d) pDNA+Optison (1:1 (v/v))+20 min US, were stained with H&E to assess haemorrhage, 24 h after treatment. AW, airways, BV, blood vessel. The pictures are representative of n ¼ 6–8 animals/group.

Gene Therapy Use of ultrasound to enhance nonviral lung gene transfer in vivo S Xenariou et al 771 a 3.0 12 * 11 2.5 10 9 2.0 8 7 1.5 6 5 1.0 4 3 0.5 2 Luciferase (RLU/mg protein) 1 Score for Degree of Haemorrhage 0.0 0 pDNA pDNA + Optison 2 min U/S 20 min U/S pDNA + Optison2 min US 20 min US b 3.0 Figure 5 Comparison of naked pDNA sonoporation between the left and right lung. Animals were exposed to US for 2 or 20 min, immediately after the administration of pDNA+Optison (100 mg/ 2.5 100 ml, 1:1 (v/v)) and luciferase expression was measured 24 h later. The closed bars represent the right lung and the open bars the lower 2.0 half of the left lobe (*Po0.05, compared to the pDNA+Optison group; n ¼ 7–8 animals/condition). 1.5

1.0 barriers to gene transfer, are different. In addition, it is 0.5 not yet known how Optison interacts with gene transfer

Score for Area of Haemorrhage agents. PEI/pDNA and GL67/pDNA complexes have 0.0 been careful optimized with regard to biophysical pDNA pDNA + Optison 2 min U/S 20 min U/S properties and hence lung gene transfer efficacy. Addi- tion of Optison likely disrupts these properties, thus Figure 4 Histological score of US-induced bioeffects following naked pDNA gene transfer. The degree of haemorrhage (a) and the reducing their transfection efficiency. affected lung area (b), were assessed 24 h after sonoporation with We next assessed the effect of sonoporation on naked pDNA. The distribution of the effects between the right and nonviral gene transfer in the presence of Optison. US left lobes are shown. Each symbol represents one animal, with the exposure failed to enhance GL67/pDNA and PEI/ mean score of the group shown as a horizontal line. Closed pDNA gene transfer over and above the reduced levels diamonds represent the smallest right lobe and open diamonds the produced by Optison. However, naked pDNA transfec- upper half of the left lobe. tion efficiency was significantly enhanced. The experiments presented here were carried out by measuring luciferase reporter gene activity in lung homogenate. the lung poses challenges for US applications, including This did not allow us to determine if sonoporation imaging, because US waves are greatly attenuated when increased the number of cells transfected, or increased travelling through air. We show here, that despite these the amount of reporter gene expression per cell. limitations, sonoporation increased naked pDNA gene Although green fluorescent protein (GFP) is the most transfer, when mixed with microbubbles (Optison), but appropriate reporter gene to address these important did not enhance GL67/pDNA or PEI/pDNA transfec- questions, in our experiments, the levels of luciferase tion efficiency. expression detected would not have translated into Microbubbles, and Optison in particular, are often detectable GFP expression (Dr Lee Davies, personal used in sonoporation studies. They are thought to lower communication). the cavitation threshold and thus to enhance the effects The positive effect of sonoporation on naked pDNA of US on gene transfer. However, in contrast to suggests that in spite of the attenuation caused by air in previously published data,32 addition of Optison sig- the lungs, the remaining energy is sufficient to increase nificantly reduced the transfection efficiency of PEI gene transfer in our model. Despite this, however, the and GL67/pDNA complexes, as well as naked pDNA. overall levels of sonoporation-mediated gene transfer Optison alone has been shown to increase naked pDNA did not exceed those of naked pDNA alone, owing to the gene transfer to the skeletal muscle of young (4 weeks), inhibitory effect of Optison. Nonetheless, the positive but not older (6 months) mice.32 Interestingly, Optison effect of US on gene transfer was entirely dependent on did not affect PEI/pDNA transfection efficiency in these the presence of Optison, suggesting that the levels of studies. Although the mice in our experiments were only cavitation required to increase gene transfer can only be 6–8 weeks old, their lungs were fully developed and achieved with the use of microbubbles. Although it is differentiated. The discrepant results are most likely a unlikely that the fragile Optison microbubbles will reflection of organ-specific differences. A vector formula- withstand nebulization through standard jet nebulizers, tion that is efficient in transfecting skeletal muscle is not such as the Pari LC, the latest generation of single-pass necessarily as effective in the lung, given that the cell mesh nebulizers (e.g. Pari eFlow) might be suitable for morphology and function, as well as the extracellular nebulization of Optison.

Gene Therapy Use of ultrasound to enhance nonviral lung gene transfer in vivo S Xenariou et al 772 The vector-specific effects of US cannot be simply evidence to suggest that mice are more susceptible to attributed to restrictions in the size of the pores caused US-induced lung damage than other species, owing by sonoporation. Several studies have assessed the to differences in lung structure. For instance, mice, rats effects of US on the cell membrane, as well as the monkeys are more vulnerable than sheep, pigs and cellular uptake of different size particles and vectors humans, because they have thinner visceral pleura. In after sonoporation in vitro. These studies have confirmed addition, their airways are shorter, thinner and less that naked pDNA is directly transferred into the flexible, which also seems to contribute to their increased cytoplasm after sonoporation, probably through US- susceptibility.39 induced pores.33 This process is much faster than To address the issues of toxicity and clinical feasibility, lipofection, which involves endocytosis. In similar studies in a larger animal model may be necessary. studies, fluorescent latex nanospheres ranging from 25 Sheep, in particular, are an attractive option, as proof of to 75 nm, as well as fluorescently labelled dextran principle for nonviral gene transfer to the ovine lung has molecules with a molecular mass of 77–464 kDa, were already been established.29 Also, their lung is similar to also internalized via US-induced pores in the cell humans both in structure and size. Sonoporation studies membrane.34 The size of PEI/pDNA complexes, at a in this model could assess how US is applied, for N/P ratio of 10:1, has been shown to be around 60 nm,35 instance endobronchially or via the chest wall, and which, given the above studies, should theoretically optimize US conditions for lung gene transfer while allow them to enter cells via sonoporation-induced assessing any US-induced side effects. They would thus pores. It is not known, however, how ultrasound or the provide valuable information on the potential clinical use addition of Optison may affect the size or charge of our of sonoporation. gene transfer agents. We have shown that US-induced haemorrhage increased with exposure time, which in turn may reduce Materials and methods gene transfer efficiency. There was also some haemorrhage, albeit at low levels, in control animals that were Vectors not exposed to US. This was most likely owing to the The gene transfer agents used in these studies were: PEI, culling method, since the same observations were made 25 kDa, branched; Sigma, Poole, UK), cationic lipid 67 in untreated controls, although the actual instillation (GL67; consisting of GL67/DOPE/DMPE-PEG5000 at might also contribute. However, we had already estab- molar ratios of 1:2:0.05; Genzyme Corporation, Framing- lished that in our mouse strain culling by cervical ham, MA, USA) and naked plasmid DNA (pDNA). The dislocation causes similar lung haemorrhage as exsan- plasmid DNA carried a luciferase reporter gene under guinations, and less than lethal injection (data not the control of a cytomegalovirus immediate early shown). promoter. Equal volumes of PEI and pDNA, in sterilized US exposure affected the left lobe more severely than water for injection (Arnolds Veterinary Products, the right lung, which may explain why sonoporation had Shrewsbury, UK), were mixed at nitrogen (N):phosphate a more modest effect on naked pDNA gene transfer in (P) ratio of 10:1.40 The solution was vortexed and the former. The appearance of haemorrhage in all areas incubated for 20 min at room temperature. Similarly, of the lungs provides indirect evidence that US pene- equal volumes of GL67 (1.2 mM) and pDNA (4.8 mM, trates deep into the lung tissue of mice. However, these 1.6 mg/ml) were mixed at a molar ratio of 1:4 findings also suggest that there may be a narrow (GL67:pDNA) in sterilized water for injection and toxicity/efficacy window for US-mediated gene transfer incubated at 301C, for 15 min.41 In some cases, Optison to the mouse lung. US-induced bioeffects in the lung (Amersham Health, Oslo, Norway), an ultrasound have been previously studied but whether cavitation is contrast agent, was mixed at a 1:1 volume: volume responsible for these effects remains unclear. Although (v/v, DNA/Optison) ratio with the gene transfer agents air within the lungs could theoretically give rise to described above, in a total volume of 100 ml. ‘microbubbles’, the gas is so closely packed in the alveoli that this is unlikely.36 More likely, the mechanical stress US equipment on the tissue is caused by reflection of the US energy Pulsed US at 1 MHz frequency, with a pulse repetition at air/tissue interfaces.37 However, cavitation might frequency of 100 Hz and 20% duty cycle, was used still play a role in our system as we are delivering (Electro-Medical Supplies, Abingdon, UK). The acoustic microbubbles topically into the lung. Interestingly, the intensity was 3 W/cm2 and the peak pressure was threshold for haemorrhage in the mouse lung has been 0.74 MPa. The diameter of the probe was 1.3 cm. determined to be approximately 1 MPa.38 This is higher than the pressure applied in our studies (0.74 MPa), but In vivo transfection and sonoporation the presence of Optison might allow for bioeffects to be Male Balb/c mice (6–10-weeks old) were anaesthetized induced at lower pressures. by intraperitoneal injection of ketamine/metetomidine US-induced toxicity may be overcome by further (76 and 1 mg/kg, respectively; National Veterinary optimization of the US conditions, such as the duty Services Ltd, Stoke, UK) and the vector solution was cycle, the total exposure time and the acoustic intensity. delivered to the lungs by nasal instillation. Briefly, the However, damage may be necessary for US-mediated animals were positioned upright, and their mouth was gene delivery to take place. In retrospect, we do not held closed between the thumb and forefinger. A total of believe that mice are the most suitable model to assess 100 ml of solution were delivered drop-wise onto the tip these issues. Firstly, the size of the mouse lung and of the nose and were inhaled. The animals that were therefore, the distance that US has to travel, is much exposed to US had their back and chest shaven, before smaller than in comparison to man. Secondly, there is the nasal instillation. Immediately after the vector was

Gene Therapy Use of ultrasound to enhance nonviral lung gene transfer in vivo S Xenariou et al 773 delivered, a water-based contact gel (Aquasonic, Sonora Statistical analysis Medical Systems Inc., Longmont, CO, USA) was applied Gene expression data are expressed as mean7standard onto the shaved area and the lungs were exposed to US error of the mean (s.e.m). Independent sample t-tests or (1 MHz, 3 W/cm2, 20% duty cycle), for a total of 2 or one-way analysis of variance (ANOVA) with a Bonferroni 20 min, as indicated. The US probe covered most of the post hoc correction, were carried out for all comparisons. chest but during the exposure it was moved around in Where necessary a log10 transformation of the raw a circular motion, to ensure that both the right and left values was carried out to ensure normal distribution of lobes were exposed. During the exposure the animals the data and equal variance between groups. If normal were placed onto heated boards. At the end of the distribution and equal variance were not achieved after procedure, Antisedan (National Veterinary Services Ltd, the log transformation, a non-parametric Mann–Whitney UK), an antidote to the anaesthetic, was administered U-test or a Kruskal–Wallis plus a Dunns post hoc (1 mg/kg), by intraperitoneal injection. All procedures correction, were carried out. The null hypothesis was were approved by the Home Office under the Animals rejected at Po0.05. (Scientific Procedures) Act 1986. The pDNA doses used in the absence of Optison were: 20 mg/100 ml for PEI/pDNA,42 80 mg/100 ml for GL67/ pDNA,41 and 100 mg/100 ml for naked pDNA.31 When Acknowledgements Optison was added, the pDNA doses had to be reduced This work was supported by UK Cystic Fibrosis Trust, in some cases, in order to keep the final volume constant. through a grant to the UK Cystic Fibrosis Gene Therapy Thus, in the presence of Optison, the pDNA doses were: Consortium. 10 mg/100 ml for PEI/pDNA, 40 mg/100 ml for GL67/ pDNA and 100 mg/100 ml for naked pDNA. The animals were killed 24 h after transfection, by cervical dislocation, and the trachea and lungs removed. References The smallest right lobe and the upper half of the left lobe, were used for histological analysis. The remaining 1 Ferrari S, Geddes DM, Alton EW. Barriers to and new lung was snap-frozen in liquid nitrogen, homogenized approaches for gene therapy and gene delivery in cystic fibrosis. (Ultra-Turrax homogeniser, Science Lab, Houston, TX, Adv Drug Deliv Rev 2002; 54: 1373–1393. USA) in 300 ml (trachea and right lobes) or 150 ml (half left 2 Ferrari S, Kitson C, Farley R, Steel R, Marriott C, Parkins DA lobe) of Reporter Gene Assay Lysis Buffer (Roche et al. Mucus altering agents as adjuncts for nonviral gene transfer Diagnostics GmbH, Mannheim, Germany). The samples to airway epithelium. Gene Therapy 2001; 8: 1380–1386. were freeze–thawed three times, spun at 16 000 g for 3 Sinn PL, Shah AJ, Donovan MD, McCray Jr PB. Viscoelastic gel 5 min and the supernatant stored at À801C for luciferase formulations enhance airway epithelial gene transfer with viral measurements. vectors. Am J Respir Cell Mol Biol 2005; 32: 404–410. 4 Wang G, Zabner J, Deering C, Launspach J, Shao J, Bodner M et al. Increasing epithelial junction permeability enhances gene Reporter gene expression assays transfer to airway epithelia in vivo. Am J Respir Cell Mol Biol 2000; 22: 129–138. Luciferase activity was measured using the Luciferase 5 Caplen NJ, Alton EW, Middleton PG, Dorin JR, Stevenson BJ, Assay System (Promega, Southampton, UK) according Gao X et al. Liposome-mediated CFTR gene transfer to the to the manufacturer’s instructions, and a single-tube nasal epithelium of patients with cystic fibrosis. Nat Med 1995; 1: luminometer (TD-20e, Turner, Steptech Instruments, 39–46. Arleysey, UK) and the total protein content of the 6 Flotte TR, Zeitlin PL, Reynolds TC, Heald AE, Pedersen P, Beck S samples was quantified using the Bio-Rad protein assay et al. Phase I trial of intranasal and endobronchial administration kit (Bio-Rad, Hemel Hempstead, UK). Each sample was of a recombinant adeno-associated virus serotype 2 (rAAV2)- assayed in triplicate and expressed as RLU/mg of total CFTR vector in adult cystic fibrosis patients: a two-part clinical protein. study. Hum Gene Ther 2003; 14: 1079–1088. 7 Hyde SC, Southern KW, Gileadi U, Fitzjohn EM, Mofford KA, Waddell BE et al. Repeat administration of DNA/liposomes to Histology the nasal epithelium of patients with cystic fibrosis. Gene Therapy The smallest right lobe and the upper half of the left lobe, 2000; 7: 1156–1165. were fixed in 10% formalin, and paraffin embedded. 8 Hyde SC, Gill DR, Higgins CF, Trezise AE, MacVinish LJ, Transverse sections (5 mm) from the middle of the tissue Cuthbert AW et al. Correction of the ion transport defect in were cut and stained with haematoxylin and eosin cystic fibrosis transgenic mice by gene therapy. Nature 1993; 362: (H&E). The tissues were then scored semiquantitatively 250–255. for degree and area of haemorrhage, using an arbitrary 9 Porteous DJ, Dorin JR, McLachlan G, Vidson-Smith H, Davidson H, Stevenson BJ et al. Evidence for safety and efficacy of DOTAP scoring system. The scores assigned for degree of cationic liposome mediated CFTR gene transfer to the nasal haemorrhage were: 1 ¼ slight increase in alveolar wall epithelium of patients with cystic fibrosis. Gene Therapy 1997; 4: thickness; 2 ¼ greater increase in alveolar wall thickness, 210–218. but alveolar spaces still visible; 3 ¼ no alveolar space, 10 Alton EW, Stern M, Farley R, Jaffe A, Chadwick SL, Phillips J et lung solid. The scores for area affected by haemorrhage al. Cationic lipid-mediated CFTR gene transfer to the lungs and were: 0.5 ¼ o1/4; 1 ¼ o1/3; 2 ¼ between 1/3 and 2/3; nose of patients with cystic fibrosis: a double-blind placebo- 3 ¼ 42/3. In some cases, samples were given two scores, controlled trial. Lancet 1999; 353: 947–954. for instance 1–2 for degree of haemorrhage. In those 11 Ziady AG, Kelley TJ, Milliken E, Ferkol T, Davis PB. Functional cases the average score was used for that sample, in this evidence of CFTR gene transfer in nasal epithelium of cystic instance 1.5. fibrosis mice in vivo following luminal application of DNA

Gene Therapy Use of ultrasound to enhance nonviral lung gene transfer in vivo S Xenariou et al 774 complexes targeted to the serpin-enzyme complex receptor. Mol son, and Levovist in mice – initial results. Radiology 2003; 229: Ther 2002; 5: 413–419. 423–428. 12 Konstan MW, Davis PB, Wagener JS, Hilliard KA, Stern RC, 27 Duvshani-Eshet M, Baruch L, Kesselman E, Shimoni E, Machluf Milgram LJ et al. Compacted DNA nanoparticles administered to M. Therapeutic ultrasound-mediated DNA to cell and nucleus: the nasal mucosa of cystic fibrosis subjects are safe and bioeffects revealed by confocal and atomic force microscopy. demonstrate partial to complete cystic fibrosis transmembrane Gene Therapy 2006; 13: 163–172. regulator reconstitution. Hum Gene Ther 2004; 15: 1255–1269. 28 Martin K, Ramnarine KV. Frequency, speed and wavelength. 13 Miller DL, Song J. Tumor growth reduction and DNA transfer by In: Hoskins PR, Thrush A, Martin K, Whittingham TA (eds). cavitation-enhanced high-intensity focused ultrasound in vivo. Diagnostic Ultrasound: Physics and Equipment. London: Green- Ultrasound Med Biol 2003; 29: 887–893. wich Medical Media, 2002, pp 7–22. 14 Miller DL, Song J. Lithotripter shock waves with cavitation 29 Emerson M, Renwick L, Tate S, Rhind S, Milne E, Painter HA nucleation agents produce tumor growth reduction and gene et al. Transfection efficiency and toxicity following delivery of transfer in vivo. Ultrasound Med Biol 2002; 28: 1343–1348. naked plasmid DNA and cationic lipid-DNA complexes to ovine 15 Koch S, Pohl P, Cobet U, Rainov NG. Ultrasound enhancement lung segments. Mol Therapy 2003; 8: 646–653. of liposome-mediated cell transfection is caused by cavitation 30 Rudolph C, Ortiz A, Schillinger U, Jauernig J, Plank C, effects. Ultrasound Med Biol 2000; 26: 897–903. Rosenecker J. Methodological optimization of polyethylenimine 16 Lawrie A, Brisken AF, Francis SE, Tayler DI, Chamberlain J, (PEI)-based gene delivery to the lungs of mice via aerosol Crossman DC et al. Ultrasound enhances reporter gene expres- application. J Gene Med 2005; 7: 59–66. sion after transfection of vascular cells in vitro. Circulation 1999; 31 Gill DR, Smyth SE, Goddard CA, Pringle IA, Higgins CF, 99: 2617–2620. Colledge WH et al. Increased persistence of lung gene expression 17 Liang HD, Lu QL, Xue SA, Halliwell M, Kodama T, Cosgrove using plasmids containing the ubiquitin C or elongation factor DO et al. Optimisation of ultrasound-mediated gene transfer 1alpha promoter. Gene Therapy 2001; 8: 1539–1546. (sonoporation) in skeletal muscle cells. Ultrasound Med Biol 2004; 32 Lu QL, Liang HD, Partridge T, Blomley MJ. Microbubble 30: 1523–1529. ultrasound improves the efficiency of gene transduction in 18 Anwer K, Kao G, Proctor B, Anscombe I, Florack V, Earls R et al. skeletal muscle in vivo with reduced tissue damage. Gene Therapy Ultrasound enhancement of cationic lipid-mediated gene trans- 2003; 10: 396–405. fer to primary tumors following systemic administration. Gene 33 Mehier-Humbert S, Bettinger T, Yan F, Guy RH. Ultrasound- Therapy 2000; 7: 1833–1839. mediated gene delivery: kinetics of plasmid internalization and 19 Amabile PG, Waugh JM, Lewis TN, Elkins CJ, Janas W, Dake gene expression. J Control Release 2005; 104: 203–211. MD. High-efficiency endovascular gene delivery via therapeutic 34 Mehier-Humbert S, Bettinger T, Yan F, Guy RH. Plasma ultrasound. J Am Coll Cardiol 2001; 37: 1975–1980. membrane poration induced by ultrasound exposure: implica- 20 Christiansen JP, French BA, Klibanov AL, Kaul S, Lindner JR. tion for drug delivery. J Control Release 2005; 104: 213–222. Targeted tissue transfection with ultrasound destruction of 35 Rudolph C, Muller RH, Rosenecker J. Jet nebulization of PEI/ plasmid-bearing cationic microbubbles. Ultrasound Med Biol DNA polyplexes: physical stability and in vitro gene delivery 2003; 29: 1759–1767. efficiency. J Gene Med 2002; 4: 66–74. 21 Miao CH, Brayman AA, Loeb KR, Ye P, Zhou L, Mourad P et al. 36 Carstensen EL, Gracewski S, Dalecki D. The search for cavitation Ultrasound enhances gene delivery of human factor IX plasmid. in vivo. Ultrasound Med Biol 2000; 26: 1377–1385. Hum Gene Therapy 2005; 16: 893–905. 37 American Institute of Ultrasound in Medicine. Section 4 – Bioeffects 22 Shohet RV, Chen S, Zhou YT, Wang Z, Meidell RS, Unger RH in tissues with gas bodies. J Ultrasound Med 2000; 19: 97–108. et al. Echocardiographic destruction of albumin microbubbles 38 Dalecki D, Child SZ, Raeman CH, Cox C, Penney DP, Carstensen directs gene delivery to the myocardium. Circulation 2000; 101: EL. Age dependence of ultrasonically induced lung hemorrhage 2554–2556. in mice. Ultrasound Med Biol 1997; 23: 767–776. 23 Liu J, Lewis TN, Prausnitz MR. Non-invasive assessment and 39 American Institute of Ultrasound in Medicine. Section 3 – control of ultrasound-mediated membrane permeabilization. Selected biological properties of tissues: potential determinants Pharm Res 1998; 15: 918–924. of susceptibility to ultrasound-induced bioeffects. J Ultrasound 24 Taniyama Y, Tachibana K, Hiraoka K, Aoki M, Yamamoto S, Med 2000; 19: 85–96. Matsumoto K et al. Development of safe and efficient novel 40 Gautam A, Densmore CL, Xu B, Waldrep JC. Enhanced gene nonviral gene transfer using ultrasound: enhancement of expression in mouse lung after PEI-DNA aerosol delivery. Mol transfection efficiency of naked plasmid DNA in skeletal muscle. Therapy 2000; 2: 63–70. Gene Therapy 2002; 9: 372–380. 41 Hillery E, Cheng S, Geddes D, Alton E. Effects of altering dosing 25 Podell S, Burrascano C, Gaal M, Golec B, Maniquis J, Mehlhaff P. on cationic liposome-mediated gene transfer to the respiratory Physical and biochemical stability of Optison, an injectable epithelium. Gene Therapy 1999; 6: 1313–1316. ultrasound contrast agent. Biotechnol Appl Biochem 1999; 30: 42 Wiseman JW, Goddard CA, McLelland D, Colledge WH. 213–223. A comparison of linear and branched polyethylenimine (PEI) 26 Li T, Tachibana K, Kuroki M, Kuroki M. Gene transfer with echo- with DCChol/DOPE liposomes for gene delivery to epithelial enhanced contrast agents: comparison between Albunex, Opti- cells in vitro and in vivo. Gene Therapy 2003; 10: 1654–1662.

Gene Therapy