Gene Transfer of the Na+,K+-ATPase [beta]1 Subunit Using Electroporation Increases Lung Liquid Clearance

Over the past decade, numerous viral and nonviral approaches have been proposed and developed for transferring genes to the lung, but most have significant limitations. The inefficiency of gene transfer, immunologie responses, and nonspecificity of cell targeting are just a few of the problems. For example, although adenovirus appears to be the vector of choice for pulmonary gene therapy in the laboratory, its use can cause inflammation and cell damage (1). Furthermore, immune responses developed against the viral vector limit the success of repeated administration. These drawbacks make its clinical use doubtful. In contrast, much less inflammation and immune response are generated against DNA, either as naked plasmid or when complexed with liposomes or other polymers, such as polyethylenimine (2). Plasmid production is simple and yields high levels of contaminantfree, pure vector. Unfortunately, the efficiency of nonviral gene transfer remains low (2). Thus, for nonviral vectors to be of use in the lung, their ability to transfect cells in vivo must be increased.

Recent research from our laboratory and others has demonstrated that electroporation can be used to efficiently deliver genes to various tissues in vivo without damage and yields highlevel gene expression (3). The application of electric fields to tissues transiently opens pores in the plasma membrane for the lifetime of the pulse, allowing exogenous DNA or other extracellular molecules to enter the cell (4). Electroporation causes up to a 1,000-fold increase in gene expression compared with DNA injection alone (5-9). At the appropriate field strengths, electroporation has proved to be a safe and effective method.

In the current study, we have developed a relatively noninvasive procedure to transfer genes to the lungs of rats using electric fields. The procedure is safe and nontraumatic and results in levels of reporter gene expression that approach those obtained with the latest generation adenoviruses. Moreover, this technique is appropriate for the delivery of therapeutic genes without associated inflammatory responses that are common to most other methods for pulmonary gene delivery. Some of the results of these studies have been previously reported in the form of abstracts (10, 11).

METHODS

Plasmids

The plasmids pEGFP-N1 (Clontech, Palo Alto, CA), pCMV-lux-DTS, pCMV-lacZ-DTS, and pCMV-ß1 express green fluorescent protein (GFP), firefly luciferase, ß-galactosidase, and the rat Na+,K+-ATPase ß1 subunit, respectively, from the cytomegalovirus (CMV) immediate early promoter/enhancer (9, 12). pGFP-ß1 expresses a GFP-rat Na+,K+-ATPase ß1 subunit fusion protein from the CMV promoter. Plasmids were purified using Qiagen Giga-prep kits (Qiagen, Chatsworth, CA) and suspended in 10 mM Tris (pH 8.0), 1 mM ethylenediaminetetraacetic acid, and 140 mM NaCl.

In Vivo Gene Transfer to the Lung

Male Sprague-Dawley rats (200-350 g) were anesthetized and hailunder the forelimbs was removed. Pediatric cutaneous pacemaker electrodes (Quik-Combo RTS; Medtronic Physio-Control Corporation, Redmond, WA) were cut to 3 × 4 cm and placed on either side of the chest under the forelimbs. A small amount of K-Y jelly (Johnson and Johnson) was placed on the skin, and the electrodes were held in place using surgical tape. A solution of 500 µl of plasmid in saline was administered between breaths over a 2-second period via a 16-gauge Angiocath catheter (BD, Franklin Lakes, NJ) used as an endotracheal tube, and the animal was allowed to recover breathing for 20 seconds. Immediately after this step, a series of eight square-wave electric pulses were administered using an ECM830 electroporator (BTX, San Diego, CA). Unless otherwise stated, a field strength of 200 V/cm was applied (calculated as through the chest, electrode center to electrode center). Later, the lungs were removed and analyzed. All experiments were conducted in accordance with institutional guidelines in compliance with the recommendations of the Guide for Care and Use of Laboratory Animals (35).

Adenoviral gene transfer was performed as previously described, using 5 × 108 plaque-forming units (pfu) of either a null virus (InVivogen, San Diego, CA) or a virus expressing the rat Na+,K+-ATPase ß1 subunit (Ad-ß1), delivered with surfactant (Survanta; Abbott Laboratories, Columbus, OH) (13).

Measurement of Lung Liquid Clearance

The isolated, fluid-filled, perfused lung preparation was performed immediately after brief ventilation of rats as previously described (14-16). Changes in concentration of Evan's blue-tagged albumin instilled into the airspace were used to estimate the volume of fluid cleared from the alveolar airspaces. The unidirectional flux of Na+ from the alveolar space (i.e., active transport and passive movement) was calculated from the rate of loss of ^sup 22^Na+ from the airspaces. Passive Na+ flux was calculated by subtracting the active Na+ flux (calculated from the rate of net fluid clearance) from total Na+ flux (16). Similarly, the flux of mannitol was calculated from the rate of loss of 3H-mannitol from the airspaces (16). Albumin flux from the pulmonary circulation into the alveolar space was determined from the fraction of fluorescein isothiocyanate (FITC-)-labeled albumin, placed in the perfusate, that appeared in the alveolar instillate during the experimental protocol.

Histologie Analysis

Paraffin-embedded thin sections were cut from lungs inflated to total lung capacity with 3% paraformaldehyde immediately following euthanasia. Lungs were X-gal stained as described (9). Immunohistochemistry was performed using the Vcclastain ABC-AP system (Vector Laboratories, Foster City, CA). Hematoxylin and eosin-stained sections (three sections each from three animals per condition) were blinded and reviewed by a pathologist for lung injury.

Measurement of Luciferase Expression

The lungs were frozen in liquid nitrogen immediately after removal, and extracts were prepared as previously described (9). Luciferase activity was measured in duplicate using the Luciferase Assay System (Promega, Madison, WI) in a Turner luminometer (Turner Biosciences, Sunnyvale, CA). Purified recombinant luciferase (Promega) was used to produce a standard curve for each experiment.

Measurement of Interleukin 6

Interleukin 6 (IL-6) was measured in lung extracts by ELISA (R&D Systems, Minneapolis, MN).

Isolation of Alveolar Type II Cells

Three days after lung electroporation, alveolar epithelial type II cells (ATII) were isolated as previously described (17-20). Productively transfected cells were scored for GFP-ß1 expression by counting GFP ' cells in multiple fields by fluorescence microscopy.

Measurement of Rubidium Uptake in Isolated Cells

Ouabain-sensitive ^sup 86^Rb+ uptake was used to estimate Na+,K+-ATPase activity in ATII cells isolated from treated animals as previously described (21).

Quantitation of Na+,K+-ATPase ß1 Subunit

Western blots were performed on basolateral membranes isolated from the lungs of treated animals as previously described (21).

RESULTS

In Vivo Gene Transfer to the Rat Lung Using Electroporation

We have developed a noninvasive electroporation procedure for gene transfer to rat lungs for in vivo studies (Figure 1A). Briefly, DNA was delivered to the lungs via an endotracheal tube, the electric field was applied transthoracically using electrodes placed directly on the chest, and gene expression was assessed 3 days later. In our initial studies, we attempted to apply ??-millisecond square-wave pulses at a field strength of 200 V/cm, as we did ex vivo and in previous in vivo studies in other tissues. However, because the size of the rats necessitated a large gap (4 cm between electrodes requiring 800 V), we were unable to administer the field for more than 10 microseconds using the ECM830 electroporator, because of instrument constraints. The ECM830 will allow the administration of up to 500 V at long pulse lengths (millisecond range), but above this voltage, only shorter pulse lengths can be used (microsecond range). Thus, we investigated the ability of this reduced pulse length to mediate gene transfer (Figure 1B). Even at a pulse length of f O microseconds, statistically significant gene transfer and expression were observed over DNA delivered in the absence of an electric field. Electroporation-mediated gene transfer was dose-dependent, with 170 ± 9 pg of luciferase per gram of wel weight (mean ± SEM; n = 6) being expressed when 600 µg of DNA was electroporated into the lungs. Expression was detected as early as 6 hours post electroporation (but not at 1 hour) and persisted for 3 days; by 5 and 7 days, gene expression could no longer be detected (not shown).

To determine whether longer pulse durations or lower field strengths could be used to increase gene delivery and expression, a set of experiments was performed using small rats (200-225 g) whose chests were small enough to electroporate with higher pulse times (Figure 1C). Pulse times of 10 microseconds, 0.5 milliseconds, and 10 milliseconds were used. The mortality rate using all conditions was very low (<5%). In all cases, death appeared to be caused by fluid delivery, and, on necropsy, there was no evidence of hemorrhaging or bleeding. Essentially, no gene expression was detected in the absence of an electric field when 100 µg of DNA was delivered to the lungs, but the application of an electric field with all pulse durations promoted gene transfer and expression. Maximal gene transfer and expression was detected using a field strength of 100 V/cm with 8 pulses at 10 milliseconds each. This amount of expression was similar to that obtained using higher doses of DNA, but with the 10-microsecond pulse duration (compare Figures 1B and 1C). Consequently, either approach (i.e., high-dose DNA with low pulse length or low-dose DNA with long pulse length) promotes high-level gene transfer to the rat lung in vivo with very little mortality. However, for all subsequent studies, the low pulse length method (10 microseconds) was used to be able to use rats of all sizes.

Localization of Gene Transfer and Expression

To determine the distribution of gene transfer and expression in the lungs, several different reporter plasmids were used. When a plasmid expressing a fusion protein between GFP and the Na+,K+-ATPase ß1 subunit was electroporated into the rat lung, GFP-ß1 expression could be detected throughout multiple lobes of the lung but was absent from surrounding tissues (Figures 2A and 2B). As can be seen, GFP-ß1 expression is relatively uniform throughout the lungs. When ATII epithelial cells were isolated from the lungs of electroporated animals 3 days after electroporation and observed by fluorescence microscopy, 54.7 ± 7.4% of cells were GFP-ß1^sup +^. Similarly, 54% of macrophages that were panned during isolation were GFP-ß1^sup +^ as well. To better distinguish what cell types received DNA in the intact lung, plasmids expressing ß-galactosidase were electroporated into the lungs of rats, and 3 days later, the lungs were removed and reacted with X-gal to visualize gene expression (Figures 2C and 2D). ß-Galactosidase expression was dose-dependent, with increasing amounts of X-gal-reactive tissue in all lobes of the lungs as the dose of DNA was increased. Lungs treated with 600 µg of pCMV-lacZ DNA in the absence of an electric field and naive lungs showed no X-gal reactivity (not shown in whole lungs). To determine which cells in the lungs expressed the gene product, lungs were paraffin-embedded, sectioned, and stained for ß-galactosidase expression by immunohistochemistry (Figure 3). In naive lungs (Figure 3A), lungs electroporated without added DNA (200 V/cm, 10 microseconds; Figure 3B) and lungs receiving DNA without electroporation (Figure 3C) did not express any detectable ß-galactosidase in any cell type. By contrast, electroporation of the reporter plasmid gave significant gene transfer and expression (Figures 3D-3I). At high magnification, expression is detectable in airway epithelial and smooth muscle cells (Figure 3G), ATI and ATH cells (Figure 3H), and vascular endothelial and smooth muscle cells (Figure 3I). However, less gene expression was consistently detected in the smooth muscle layers of large blood vessels (Figure 3D). This expression pattern was present throughout the lung in all lobes, although some areas showed greater expression than others. Taken together, these results demonstrate that electroporation can be used effectively to target genes to multiple cell types throughout the lung.

Inflammatory Response and Histologie Analysis of Electroporated Lungs

To assess whether electroporation elicited any damage in the rat lung, sections of treated and naive lungs were examined 3 days after treatment for histologie changes and cytokine responses from the procedure (Figure 4 and Table 1). Additional animals were examined at 1,24, and 72 hours following treatment (Table 1). Blinded pathologic examination could not distinguish between naive lungs (Figures 4A and 4B), lungs from animals that received Tris-ethylenediaminetetraacetic acid-saline with electroporation (Figures 4C and 4D), and those that received DNA (600 µg of pcDNA3) with electroporation using the following four criteria on a 5-point scale: vascular congestion, hyaline membranes, polymorphonuclear cell infiltrates, and interstitial infiltrates. Furthermore, there was no difference between lungs from animals that were electroporated using 10-microsecond (not shown) or 10-millisecond pulses (Figures 4E and 4F). By contrast, animals receiving 5 × 10^sup 8^ pfu of a control null firstgeneration adenovirus showed profound vascular congestion and polymorphonuclear cell and interstitial infiltrates, as has been previously described (14). As another indicator of inflammation, levels of IL-6 were also measured in treated and naive lungs. Again, no differences in levels of this cytokine were noted between animals that were treated under any of the previously described conditions, with the exception of the adenovirustreated animals (Table 1).

Because transgene expression was detected throughout the lung in cells within and below the epithelial layer (Figure 3), DNA must be gaining access to cells across the epithelial layer. Thus, we asked whether the process of electroporation caused a transient increase in epithelial barrier permeability. When FITClabeled albumin was instilled into the lungs together with DNA and electroporated, there was no significant accumulation of the alveolar tracer in the blood of the rats within 15 minutes of the procedure, compared with animals that were instilled with the tracer and DNA but not electroporated (0.011 ± 0.01% vs. 0.09 ± 0.01 % of the instilled tracer in the blood, mean ± SEM; n = 5). These results suggest that although the DNA may be gaining access to the subcpilhelial cells through the epithelial tight junctions, there is no significant increase in epithelial barrier permeability.

Transfer of the Na+,K+-ATPase ß1 Subunit to Rat Lungs

On the basis of the results that reporter genes could be transferred efficiently to the lungs with little or no inflammation or trauma to the tissue, we evaluated electroporation as a means to transfer the rat Na+,K+-ATPase ß1 subunit to the lung and measured alveolar fluid clearance (AFC) after gene transfer. Plasmids were transferred to the lungs of animals with or without electroporation; 3 days later, the animals were euthanized and AFC was measured in the isolated lungs (Figure 5A). Electroporation in the absence of DNA, transfer of the ß subunitexpressingplasmid (pCMV-ß1 subunit) without electroporation, or electroporation of an "empty" plasmid that does not code for a gene product (pcDNA3) did not increase AFC compared with that seen in naive animals. By contrast, electroporation of the pCMV-ß1 subunit into lungs increased AFC by 74%. Similarly, a recombinanl adenovirus expressing the Na+,K+-ATPase ß1 subunit (Ad-ß1) also increased AFC by almost a factor of 2. However, fluid clearance could not be assessed in the adenovirustreated lungs until 10 days after transfer because of severe inflammation between Days 2 and 7 (14). Furthermore, electroporation did not alter the permeability of FITC-albumin, mannitol, or sodium across the alveolar-capillary barrier under any of the conditions tested (Table 2).

To study whether the increases in AFC were from the functional expression of the ß1 subunit, the activity of the Na+, K+-ATPase was measured in cells isolated from treated lungs. Three days after gene transfer by electroporation, ATII epithelial cells were isolated from the lungs and the uptake of ^sup 86^Rb+ was measured as an assay for Na+,K+-ATPase activity (Figure 5B). Cells isolated from naive lungs, lungs electroporated without DNA, lungs receiving pCMV-ß1 subunit DNA alone (no electroporation), or lungs electroporated with the empty vector or a luciferase-expressing plasmid showed essentially the same levels of ouabain-sensitive ^sup 86^Rb+ uptake. However, cells isolated from lungs electroporated with the pCMV-ß1 subunit plasmid showed a twofold increase in ouabain-sensitive ^sup 86^Rb+ transport, confirming that the increased fluid clearance paralleled the increased Na+ pump activity. Finally, as shown in Figure 5C, the levels of the ß1 subunit, quantified by Western blot, from the lungs of rats that were electroporated with the pCMV-ß1 subunit plasmid showed a twofold increase in the Na+,K+-ATPase ß1 subunit protein abundance at the basolateral membranes isolated from the lungs at 3 days after treatment relative to naive lungs. Taken together, these results demonstrate that electroporation can effectively deliver the Na+,K+-ATPase ß1 subunit, resulting in increased lung liquid clearance.

DISCUSSION

One obstacle to the use of in vivo gene delivery is the level of technical difficulty associated with many delivery methods. The production of high-quality, helper-free recombinant viral vectors can be difficult, costly, and time-consuming. Although polyplex and lipoplex formulations are easier to create, as are the plasmid constructs that are delivered with them, effective formulation of these agents can be cumbersome. By contrast, electroporation offers an alternative that only requires purified plasmids and an electric field generator. Large-scale plasmid production and purification is easy and inexpensive. Furthermore, the application of this approach to the lung is fast and relatively noninvasive. Once the animal is anesthetized, the entire procedure takes less than 10 minutes, and recovery is rapid. The use of an endotracheal tube circumvents any incision being made to deliver the DNA to the lungs, as we have reported in mice (9).

Perhaps the major limitation of viral vectors is the potent proinflammatory response. The application of an electric field (200 V/cm) to the lungs, either to animals receiving vehicle or DNA, caused no increases in IL-6 levels and showed no histopathologic changes compared with naive animals, either immediately after electroporation (1 hour) or 1 and 3 days later. By contrast, lungs from animals treated with recombinant adenoviruses show increases in all of these criteria between Days 2 and 7 compared with naive lungs.

We report here that electroporation achieves high levels of gene expression. Furthermore, the physiologic effects after electroporation-mediated delivery of the Na+,K+-ATPase ß1 subunit were indistinguishable from those using recombinant adenoviruses expressing the ß1 subunit, which have been very effective for pulmonary gene transfer (14).

One additional drawback to most viral and nonviral methods of gene delivery is that they are limited in delivery to only those cells with which they come into contact. Thus, delivery from the airways results overwhelmingly in gene transfer to the airway and alveolar epithelial cells only, whereas systemic delivery targets only endothelial cells in the lung (22-26). However, electroporation mediates gene delivery to multiple cell layers within the lung following DNA delivery via the airways. In addition, we detected relatively uniform gene expression throughout the periphery and parenchyma in airway and alveolar epithelial cells, airway smooth muscle cells, vascular endothelial cells, and some vascular smooth muscle cells. Similar distributions of gene transfer and expression were observed using GFP and lacZ genes with detection by direct fluorescence and immunohistochemistry, respectively. When lungs that had been electroporatcd with a lacZ construct were reacted with X-gal, gene expression did not appear as evenly distributed as it did in sections from the same lungs using immunohistochemistry. A similar observation was made in the mouse lung where most X-gal reactivity congregated in ATI and ATII epithelial cells, although the gene product appeared uniformly distributed throughout all cell types when immunohistochemistry was performed using antibodies against ß-galactosidase (9). The likely reason for this apparent discrepancy is that the use X-gal reactivity to measure ß-galactosidase expression resulting from gene transfer routinely underestimates transfection efficiency during transient transfections (27). By contrast, GFP fluorescence or immunohistochemistry using an amplification system such as the Vector ABC system (Vector Laboratories, Foster City, CA) are much more sensitive and more reliable indicators of real transfection efficiency (28). Thus, it is possible that more DNA is delivered to the alveolar epithelium or that these cells may express more protein than other pulmonary cells, but all cells in the lung do receive and express the transgene.

How the DNA is gaining access to the subepithelial cells in the electroporated lungs is currently unclear. Electroporation has both electrophoretic and electropermeabilization components that aid DNA delivery to cells within tissues (4). Thus, the DNA is electrophoresed to and into cells via membrane pores that exist for the lifetime of the electric pulse (29). However, it is unlikely that the DNA moves transcellularly by electrophoresis to access cells on the other side of the epithelium. Rather, it is more likely that the DNA moves across and/or through the epithelial tight junctions with the applied electric field. This would imply that a transient increase in epithelial barrier permeability would be coincident with the electric field. Although we did not detect changes in alveolar epithelial permeability, or see histologic evidence of pulmonary edema, we cannot exclude the possibility of a very transient, but physiologically insignificant, barrier permeability. Furthermore, if electrophoresis and membrane permeabilization occur only during the duration of the electric field that is between 10 microseconds and 10 milliseconds in our experiments, it is unlikely that any significant permeability changes would be detected or detrimental.

It has been demonstrated previously that transfer of the ß1 or a2 subunit of the Na+,K+-ATPase using recombinant adenoviruses can increase expression of the Na+,K+-ATPase in lungs of rats and result in increased AFC (13, 14, 21, 30). A drawback to these studies is that the effects of gene transfer could only be experimentally measured 7 days after infection because of the induction of inflammatory responses by the viral vector. As such, the timing of this therapeutic response corresponds more to the subacute or "proliferative" phase of acute lung injury, rather than the acute phase when therapy may be more meaningful (31). The fact that gene transfer using electroporation causes no inflammatory response greatly widens the therapeutic window of gene expression from between 1 and 5 days after electroporation to cover the entire acute phase of lung injury (9).

Several groups have reported that gene transfer of either the Na+,K+-ATPase ß1 or a2 subunit can protect from lung injury initiated after gene delivery; in the clinical setting, however, this is not the way that therapy would be administered (13, 14, 21, 30, 32). Rather, the ideal therapy would be administered after lung injury and edema have developed. However, the injured lung presents many additional barriers to gene transfer that are not encountered in the normal lung (33). It has been shown recently that recombinant adenoviruses can be used effectively in edematous lungs after hyperoxia-induced acute lung injury (34). It remains to be seen whether electroporation will work as efficiently in the injured lung as it does in the healthy lung, but because electroporated DNA moves through mucous and multiple cell and extracellular matrix layers to target cells beyond the pulmonary epithelium, it could be effective and warrants further research. Thus, on the basis of the ease, efficiency, and nontraumatic nature of this electroporation method for pulmonary gene transfer, its use may have clinical potential.