INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

TRANSIENT TRANSFECTION of epithelial cells with original lipid reagents such as Lipofectin or LipofectAMINE is limited in efficacy in epithelial cells grown in nonpolarized culture. However, in this study our laboratories have shown that more recently developed lipid compounds are quite efficacious in transiently transfecting non-cystic fibrosis (non-CF) and CF airway epithelial cell lines. In particular, LipofectAMINE PLUS and LipofectAMINE 2000 are the most efficacious and showed little or no toxicity. In early stages of this work, two studies by our laboratories used this emerging technology on nonpolarized epithelial and heterologous cells (1, 5). Because of this breakthrough in epithelial "lipofection" efficiency, we hypothesized that these more newly developed lipid transfecting reagents might also prove efficacious in transducing polarized epithelial monolayers.

Few studies have shown successful transfection of polarized epithelia. Snyder (6) recently reported a study in which -, -, and -isoforms of the epithelial sodium channel (ENaC) were expressed together and transiently in Fischer rat thyroid epithelial cells grown as polarized monolayers. Transient transfection was undertaken on forming monolayers that did not disrupt transepithelial resistance and resulted in expression of the ENaC constructs for 3-7 days following transfection (6). Both electrophysiology and biochemistry could be performed on these polarized monolayers (6). The lipid reagent mixture used by Snyder was proprietary, but it contained DOPE- and DMRIE-based cationic lipids. Moreover, Li and Mrsny (3) transiently transfected a rat salivary gland epithelial cell line grown as a monolayer with LipofectAMINE PLUS and cDNAs encoding the proteins Raf-1, occludin, and claudin-1 to affect tight junction and monolayer integrity. Uduehi et al. (9) transiently transfected Caco-2 intestinal epithelial monolayers with DNA:DOTAP complexes early during monolayer formation, although the Caco-2 monolayers became resistant to lipid-mediated transfection as they aged.

Based on our previous results and the handful of studies outlined above, our working hypothesis was that newly developed lipid reagents, especially those that used an enhancer or priming lipid to complex with mammalian expression vectors before mixing with transfecting lipid (LipofectAMINE PLUS, for example), may have efficacy in transiently transfecting epithelial cell monolayers. The data described herein show that mammalian expression vectors bearing the reporter genes luciferase (LUC) or green fluorescent protein (GFP) or the epithelium-specific gene, the cystic fibrosis transmembrane conductance regulator (CFTR), are transfected transiently into forming epithelial monolayers without damage or toxicity to the monolayer and with positive gene transduction. This work derives from efficient gene transfer in nonpolarized human airway epithelial cells. These results as well as the establishment of these transfection systems may have wide utility to investigators who wish to study the activity of their epithelial protein of interest in a polarized epithelium. These systems may also obviate the need for establishing stably transfected clones, where levels of gene and protein expression vary widely from clone to clone or in which the loss of expression over passage or time in culture may occur.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell and Monolayer Culture

Culture in flasks. All substrates for all epithelial cell lines were coated with Vitrogen 100 solution (Cohesion) that was diluted 1:15 in Ca²⁺,Mg²⁺-free Dulbecco's PBS (Life Technologies, Grand Island, NY). The diluted Vitrogen was added to the substrate and incubated for 1-2 min before removal. For filter supports, the incubation was 5-10 min. Coating was performed a minimum of 8 h before plating of epithelial cells. Four epithelial cell lines were used for this study: a 16HBE14o non-CF human bronchial epithelial cell line and a CFBE41o CF human bronchial epithelial cell line (2, 10); an MDCK C7 canine kidney epithelial cell line (4); an IB3-1 CF human bronchial epithelial cell line (12); and a CFT-1 CF human tracheal epithelial cell line (11). MDCK cells, CFBE41o cells, and 16HBE14o cells were grown in tissue culture flasks in a minimum essential medium (MEM) with Earle's salts and L-glutamine (Life Technologies) supplemented with 5% fetal bovine serum (certified FBS, heat-inactivated at 55°C for 30 min; Life Technologies), 6 ml of penicillin-streptomycin 100× solution (penicillin 100 U/ml final; streptomycin 100 µg/ml final; Life Technologies), 6 ml of 200 mM L-glutamine 100× solution (2 mM final; Life Technologies), 2 ml of fungizone solution (amphotericin B, 250 µg/ml stock, 1 µg/ml final; Life Technologies), and 1 ml of gentamycin stock solution (100 µg/ml final; Life Technologies). IB3-1 cells were grown in LHC-8 medium (Biofluids, Rockville, MD) supplemented with 5% FBS, 10 ml of penicillin-streptomycin 100× solution, 5 ml of 200 mM L-glutamine 100× solution (2 mM final; Life Technologies), and 1 ml of fungizone solution. CFT-1 cells were grown in the same LHC-8 medium used for IB3-1 cells, the only exception being no FBS supplementation. CFT-1 and CFBE41o cells are homozygous for the F508-CFTR mutation, whereas IB3-1 bears the F508-CFTR and W1282X-CFTR mutations.

Culture on permeable filter supports. For 16HBE14o, CFBE41o, IB3-1, and CFT-1 airway epithelial cells, cells were seeded onto 6.5-mm-diameter Transwell filters (Corning-Costar, Corning, NY). Despite standardized seeding of cells (1 × 10⁶ cells per 6.5-mm-diameter filter; 2× cells for 12-mm-diameter filters; 4× cells for 24-mm-diameter filters), transepithelial resistance (R_TE) was always greatest for the airway epithelial cell lines on the 6.5-mm-diameter filters. R_TE was routinely twofold greater on clear polyester Transwell filter supports than on more standard polycarbonate Transwell filters. For MDCK cells, the same findings also held true; however, sufficient R_TE values were observed on the different size filters (3,000-6,000  · cm² for 6.5-mm filter monolayers; 1,000-2,000  · cm² for 12-mm filter monolayers; 500-1,000  · cm² for 24-mm filter monolayers). Data from different size filters are shown for MDCK cells and is indicated in the figure legends. Lots of certified FBS are routinely screened by our laboratory for optimal support of 16HBE14o cell monolayer formation and are routinely available from Life Technologies (now a division of Invitrogen). These lots of FBS are key ingredients in helping the epithelial cell lines establish polarized monolayers. Other lots of FBS from Life Technologies as well as lots of FBS from other sources failed to support monolayer formation. This is not a different strategy compared with other laboratories in the CF research community who rely on medium supplements such as Ultraser G and PC-1 supplement for growth and monolayer formation in cell lines and primary cultures of CF and non-CF airway epithelial cells.

Transient Transfections of Nonpolarized Epithelial Cells

Per well of a six-well plate, the following protocol was performed for each lipid. Only the lipids that were effective in our study are shown here. Amounts of DNA, lipid, and medium were increased in proportion to surface area of culture.

LipofectAMINE PLUS. A 15-ml tube was used to mix 1 µg of cDNA and 6 µl of PLUS reagent in 150 µl of OptiMEM-I medium. In a separate 15-ml tube, 4-6 µl of LipofectAMINE reagent were mixed with 150 µl of OptiMEM-I medium. Tthe two tubes were incubated separately at room temperature for 15 min. The contents of the LipofectAMINE reagent tube were then transferred and mixed with the cDNA and PLUS reagent. The single tube containing both lipids and the cDNA was incubated at room temperature for another 15 min. During the two 15-min incubations, the cells were washed three times with OptiMEM-I medium. The volume of the mixture was diluted to 1.5 ml with OptiMEM-I, and the contents of the tube were added to the well of the cells.

LipofectAMINE 2000. A 15-ml tube was used to mix 1 µg of cDNA in 150 µl of OptiMEM-I medium. In a separate 15-ml tube, 3-4 µl of LipofectAMINE 2000 reagent were mixed with 150 µl of OptiMEM-I medium. The two tubes were incubated separately at room temperature for 15 min. The contents of the LipofectAMINE 2000 reagent tube were then transferred and mixed with the cDNA. The single tube containing LipofectAMINE 2000 and cDNA was incubated at room temperature for another 15 min. During the two 15-min incubations, the cells were washed three times with OptiMEM-I medium. The volume of the mixture was diluted to 1.5 ml with OptiMEM-I, and the contents of the tube were added to the well of the cells.

Effectene. A 15-ml tube was used to mix 1 µg of cDNA and 6 µl of Enhancer lipid in 150 µl of OptiMEM-I medium. The tube was incubated at room temperature for 10 min. To the same tube, 12 µl of Effectene lipid were added to the mixture containing cDNA and Enhancer lipid. The single tube containing both lipids and the cDNA was incubated at room temperature for another 10 min. During the two 15-min incubations, the cells were washed three times with OptiMEM-I medium. The volume of the mixture was diluted to 1.5 ml with OptiMEM-I, and the contents of the tube were added to the well of the cells.

FuGENE 6.0. A 15-ml tube was used to mix 1 µg of cDNA and 6 µl of FuGENE 6.0 reagent in 150 µl of OptiMEM-I medium. The tube was incubated at room temperature for 15 min. During the 15-min incubation, the cells were washed three times with OptiMEM-I medium. The volume of the mixture was diluted to 1.5 ml with OptiMEM-I, and the contents of the tube were added to the well of the cells.

Transient Transfections of Epithelial Monolayers

The following lipid reagents were screened: Lipofectin, LipofectAMINE, LipofectAMINE PLUS, LipofectAMINE 2000, and DMRIE-C (Life Technologies); FuGENE 6.0 (Roche Diagnostics); Tfx-10, Tfx-20, and Tfx-50 (Promega); and Effectene (Qiagen). Per 6-mm-diameter monolayer, 6 µl of each lipid reagent were used (with the exception of Effectene, where 12 µl were added per monolayer). This 6-µl amount also applies to the Enhancer lipid reagent for Effectene and the PLUS reagent for LipofectAMINE PLUS. Lipid:DNA complexes mixed as a cocktail were added to both sides of the monolayer (1-ml total volume; 0.5 ml to apical and basolateral sides of the monolayer), except where sidedness of lipid-mediated gene transfer into the monolayer was tested. Transfection cocktails were mixed according to the manufacturer's instructions; however, OptiMEM-I was used as the medium for making the mixtures in all cases. Monolayers were incubated for 6 h at 37°C in a humidified, 5% CO₂ incubator. R_TE was measured before and after transfection. Routinely, R_TE dropped 100-200  · cm² during the 6-h transfection; however, after feeding with FBS-containing medium and an overnight recovery, the monolayer R_TE returned to or exceeded pretransfection values (see RESULTS). The chemistry of the lipid mixtures is known for Lipofectin, LipofectAMINE, and DMRIE-C. Unfortunately, these lipids failed to transduce monolayers well. The information is proprietary for the other reagents, which, interestingly enough, were the most efficacious.

LUC and GFP Reporter Assays

GFP-transfected monolayers (grown always on clear polyester filters to visualize GFP fluorescence in the monolayer) were visualized every 24 h following transfection using a Nikon Eclipse TE-200 inverted microscope equipped with GFP epifluorescence. The relative transfection efficiency was approximated visually and was confirmed by a second investigator. Digital images were taken using a Nikon Eclipse TS100 microscope equipped with a digital camera port. A Nikon CoolPix E990 digital camera was attached to the camera port via an MDC CoolPix lens. Live cells were imaged in standard culture medium 48 h posttransfection; no fixation procedures were needed. 16HBE14o cell monolayer images were taken at ×10 magnification to maximize the field of cells imaged, whereas IB3-1 CF cells were imaged at ×20 magnification.

Voltohmmeter Open-Circuit and Ussing Chamber Short-Circuit Current Measurements of Monolayer Electrical Properties

RANTES Chemokine ELISA

Immunoprecipitation of CFTR

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Efficacy of Transient Transfection of Epithelial and Heterologous Cells with LipofectAMINE PLUS, LipofectAMINE 2000, FuGENE 6.0, and Effectene

View this table: [in this window] [in a new window]   Table 1.   Efficacy of lipid-mediated transient transfection in non-polarized 16HBE14o cells transfected at subconfluence

IB3-1 cells were transfected transiently under nonpolarized conditions to visualize and image GFP to show that the majority of transfected cells were GFP positive under the most optimal conditions (Fig. 1). Again, normalizing for the surface area of a well of a six-well plate, 3 µl of LipofectAMINE 2000 or 4 or 6 µl of both LipofectAMINE and PLUS reagent (in 1.5 ml of OptiMEM-I medium volume) were optimal in transfecting this CF human airway epithelial cell line without toxicity. A conservative estimate showed that at least 50% of the cells were GFP positive when transfected with 6 µl of both LipofectAMINE and PLUS reagent (Fig. 1A). Together, these data show that at least 50% of the cells in the culture are transfected positively with GFP, that the LUC protein signal is robust in the lysates, and that LipofectAMINE PLUS, when used in a limited amount, is not toxic to epithelial cells grown under nonpolarized conditions.

View larger version (55K): [in this window] [in a new window]   Fig. 1.   Images of green fluorescent protein (GFP) reporter protein fluorescence in live, nonpolarized IB3-1 cystic fibrosis (CF) human bronchial epithelial cells to assess the percentage of positively transfected cells in the culture and immunoprecipitation of cystic fibrosis transmembrance conductance regulator (CFTR) protein from a transiently transfected IB3-1 CF cell culture. A: IB3-1 cells imaged under GFP fluorescence (a) and phase microscopy (b). One microgram of EGFP-C1 vector and 6 µl of both LipofectAMINE and PLUS reagent were used for this culture. B: duplicate transfections were performed on separate days, and cells were prepared for immunoprecipitation. "Band C" refers to the fully processed and fully glycosylated form of CFTR, whereas "Band B" refers to the immature, core glycosylated form of CFTR that is retained in the endoplasmic reticulum (ER) or an intermediate ER-associated compartment. No butyrate induction, chemical chaperones, or proteasomal inhibitors were used in these experiments. WT, wild-type CFTR; F, F508 CFTR mutation; EV, empty vector control (pcDNA 3.1 plasmid without CFTR gene). Note the lack of any endogenous CFTR protein in this IB3-1 system.

Transfer of reporter genes is promising and shows efficacy; however, these gene cassettes are small. It is possible that smaller size vectors may overestimate transfection efficiency, because they gain entry to cells more easily. Thus we also assessed whether vectors bearing the CFTR gene gain as efficient entry into nonpolarized IB3-1 CF cells that are null for biochemically or functionally detectable CFTR. Figure 1B shows two representative immunoprecipitation experiments where vectors bearing the wild-type form of CFTR and the most common disease-associated CFTR mutation, F508, were transfected transiently into IB3-1 CF cell cultures with LipofectAMINE PLUS. Whereas empty vector-transfected cultures had no forms of CFTR protein (see Fig. 1B), wild-type CFTR showed the fully glycosylated "band C" form of CFTR protein, with a hint of the partially glycosylated "band B" form of CFTR. In contrast, F508-CFTR, an ER retention mutant because of a folding defect in NBD-1, showed only the immature "band B" form of CFTR. Together, these data show that 1) IB3-1 cells are an ideal CFTR-null human airway epithelial expression system for CFTR expression studies, and 2) a strong biochemical signal for CFTR is achieved upon transient transfection with LipofectAMINE PLUS, indicative of efficient and robust gene transfer.

Optimization of Monolayer Formation in Non-CF and CF Airway Epithelial Cell Lines

Transient Transfection of Epithelial Cell Lines Grown as Monolayers with LUC Reporter Genes

Figure 2 shows representative data sets for three different efficacious lipid reagents on 16HBE14o monolayers. Surprisingly, no lipid transfecting reagent had a sustained effect on R_TE; only a transient and short-lived drop in R_TE was observed that was reversed upon feeding and overnight incubation in standard culture medium (Fig. 2). R_TE was measured for each set of monolayers studied throughout the project, as shown in Table 2 and before and after transient transfection. This data set is large; thus only typical examples are shown. Because this transient reduction was seen with every lipid reagent tested, we hypothesized that the OptiMEM-1 medium is the cause of the transient dip in R_TE, which resolves when the monolayers are washed and fed with standard culture medium. This hypothesis was supported by an experiment performed in Fig. 7.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Lack of effect on lipid transfecting reagents on transepithelial resistance (RTE) development in 16HBE14o non-CF airway epithelial monolayers. In these experiments, monolayers were transfected transiently with LipofectAMINE PLUS (A), FuGENE 6.0 (B), or Effectene (C) and pRL-CMV-LUC. RTE was measured 2 days after seeding, before and after transfection on day 3, and each day (D) thereafter. Monolayers were harvested 48, 72, and 96 h later to perform LUC assays on the lysates (see Fig. 3). These are 3 examples of effects on transient transfection with different lipids on 12 monolayers. This procedure was performed for every monolayer transiently transfected in this study.

Because it was known that lipid reagents were not toxic to epithelial monolayers (even some lipids that were toxic to nonpolarized epithelial cells), LUC was transfected transiently into 16HBE14o monolayers, and LUC activity was measured in monolayer lysates 48, 72, and 96 h after transfection. Lipofectin and LipofectAMINE had little efficacy in 16HBE14o monolayers (Fig. 3A). In sharp contrast, LipofectAMINE PLUS transduced 16HBE14o monolayers with LUC significantly at all three time points (Fig. 3A). DMRIE-C reagent had little effect, as did Tfx-20 and Tfx-50. Because DMRIE reagent was mixed with DOPE reagent in the study by Snyder (6), we also mixed DMRIE-C reagent with all of the lipid reagents listed in Fig. 3. DMRIE had no effect on the ability of any other lipid reagent to transfect 16HBE14o monolayers (data not shown). In addition to LipofectAMINE PLUS, Effectene also transduced 16HBE14o monolayers significantly at the 48- and 72-h time points (Fig. 3A). In a second series of experiments, similar results were observed for LipofectAMINE PLUS and Effectene (Fig. 3B). However, in addition to these reagents, LipofectAMINE 2000 and FuGENE 6.0, two additional proprietary lipid transfecting reagents that became available during this project, were also efficient at transducing 16HBE14o monolayers with LUC (Fig. 3B). Of interest, combinations of two different lipid reagents had no additive effect on LUC gene transfer. Together, more newly developed lipid transfection reagents, LipofectAMINE PLUS, LipofectAMINE 2000, Effectene, and FuGENE 6.0, transfected 16HBE14o monolayers transiently.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Transient transfection of epithelial cell lines grown as monolayers with lipid transfecting reagents and Luciferase (LUC) reporter gene. In these experiments performed in triplicate, 16HBE14o monolayers were transfected transiently with a subset of lipid reagents (A). Monolayers were harvested 48, 72, and 96 h later to perform LUC assays on the lysates. B: single lipid reagents or combinations of reagents were used for transducing 16HBE14o monolayers with LUC, and lysates were harvested and examined for LUC activity only at the 48-h time point (again, 3 monolayers per condition). C: experiments were performed in triplicate on Madin-Darby canine kidney (MDCK) cell monolayers at the 48-h time point on 6.5-mm-diameter (left) and 12-mm-diameter filters (right). D: experiments were performed in triplicate on CFT-1 CF human tracheal epithelial cell monolayers at the 48-, 72-, and 96-h time points on 6.5-mm-diameter filters. E: experiments were performed on 16HBE14o monolayers in triplicate to assess the sidedness of effective lipid transfecting reagents in LUC reporter gene transfer. Ap, apical; Bl, basolateral. Six 6.5-mm-diameter monolayers were tested for each condition. With LipofectAMINE 2000 and Effectene, the signal was significantly enhanced when complexes were added to both sides vs. apical only; however, the signal was not observed with the other lipids. We hypothesize that the OptiMEM-1 medium actually may loosen tight junctions (as reflected in the transient decrease in monolayer resistance) to allow some enhanced transduction.

Figure 3C shows a similar analysis of transient transfection of MDCK monolayers. In contrast to the studies of 16HBE14o monolayers, only Effectene reagent transfected MDCK monolayers transiently and effectively in an initial series of experiments on 6-mm-diameter filter supports (Fig. 3C, left). LipofectAMINE PLUS produced a lower LUC signal. A second series performed on MDCK monolayers grown on 12-mm-diameter filters showed that Effectene, LipofectAMINE 2000, FuGENE 6.0, and, to a lesser extent, LipofectAMINE PLUS transfected transiently and effectively MDCK monolayers (Fig. 3C, right). Together with data for the 16HBE14o monolayers, these data show that the above-mentioned lipid transfecting reagents transfect epithelial monolayers transiently. The heterogeneity of the LipofectAMINE PLUS data suggests that surface area of the monolayer may be an important factor in designing transient transfection experiments. This only came to light in MDCK monolayer experiments, because non-CF and CF human airway epithelial cell lines failed to form monolayers on filter supports larger than 6.5 mm in diameter.

To study CFTR in a polarized system that is truly null for CFTR functionally, we needed to 1) grow CF human airway epithelial cells as polarized monolayers (see Table 2) and 2) transiently transfect the monolayers with CFTR constructs. To assess the feasibility of transiently transfecting CF airway epithelial monolayers with lipid:DNA complexes, LUC transfer into CFT-1 monolayers (Fig. 3D) and IB3-1 monolayers (not shown) was performed in a manner similar to that for MDCK and 16HBE14o monolayers. As with the above-mentioned systems, Effectene, LipofectAMINE PLUS, and FuGENE 6.0 were all effective in transduction.

Transient Transfection of Airway Epithelial Monolayers Occurs from the Apical Side of the Monolayer

Transient Transfection of Non-CF and CF Epithelial Cell Primary Cultures Grown as Monolayers with LUC Reporter Genes

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Transient transfection of epithelial cell primary cultures grown as monolayers with lipid transfecting reagents and LUC reporter gene. A: experiments were performed on 16HBE14o monolayers in parallel to primary cell monolayers to assess the effect of air:fluid (Air:F) interface culture vs. fluid:fluid (F:F) culture as well as lipid combination cocktails (n = 4 per condition). Of course, during the course of the transient transfection, fluid containing lipid:DNA complexes was added to both sides of the epithelium for the 6-h transfection period. After several washes, the monolayer was subjected again to air:fluid interface culture, whereas fluid:fluid cultured monolayers never deviated from this condition. B: non-CF primary human small airway epithelial (NHSAE; n = 4 per condition) and sinus epithelial (NHSE; n = 1 per condition) cell monolayers were grown in air:fluid interface culture and transfected transiently. C: CF primary human nasal polyp (CFNP) epithelial cell monolayers were also tested (n = 4 per condition).

Figure 4, B and C, shows results from transient transfection of non-CF and CF human primary airway epithelial monolayers done in parallel with the 16HBE14o monolayers grown in air:fluid interface culture in Fig. 4A. Primary cultures of non-CF human small airway epithelial cells (NHSAE) and non-CF human sinus epithelial cells (NHSE) were grown as monolayers in air:fluid interface culture and were transiently transfected with LUC (Fig. 4B), while primary cultures of CF human nasal polyp epithelial cells grown as monolayers were also examined (Fig. 4C). As with 16HBE14o monolayers, non-CF primary monolayers were transduced best with LipofectAMINE 2000, and using a combination of LipofectAMINE PLUS and LipofectAMINE 2000 gave no added benefit or, actually, inhibited transduction. Interestingly, and in contrast, monolayers of CF human nasal polyp epithelial cells were transduced by both LipofectAMINE PLUS and LipofectAMINE 2000 individually; however, these primary CF monolayers were transduced best by a combination of LipofectAMINE PLUS and LipofectAMINE 2000.

Transient Transfection of Epithelial Monolayers with GFP As a Reporter

However, we could assess GFP gene transfer in 16HBE14o cells grown packed together as a very confluent, pseudopolarized monolayer (Fig. 5, in contrast to subconfluent cells studied in Table 1) on a larger filter support. It should be noted that the R_TE was significantly less on the 24-mm filter supports (200-300  · cm²) than on the 6.5-mm filter supports (600-1200  · cm²). LipofectAMINE 2000 was the best lipid transfecting reagent and showed little toxicity. A collage of GFP fluorescent images with increasing amounts of LipofectAMINE 2000 shows that 6 µl of this lipid gave maximal GFP reporter gene transfer (Fig. 5). Again, a conservative estimate shows that ~20-30% of the surface area of the monolayer showed GFP fluorescence in a given field. For these monolayers, LipofectAMINE PLUS was not as efficient (only 10% of cells were transfected with 6 and 8 µl of both lipids); enhanced GFP (EGFP)-only and lipid-only controls showed no GFP fluorescence (data not shown). In contrast to nonpolarized CF and non-CF airway epithelial cells where LipofectAMINE PLUS was the most useful lipid transfecting reagent, these results show that LipofectAMINE 2000 was the most efficient lipid transfecting reagent as the epithelial cells pack together and become polarized. We did not attempt to add greater amounts of LipofectAMINE 2000 to achieve greater GFP-positive surface area; however, this may be possible without encountering toxicity.

View larger version (95K): [in this window] [in a new window]   Fig. 5.   GFP gene transfer with LipofectAMINE 2000 in 16HBE14o cell monolayers. A collage of images of GFP fluorescence of 16HBE14o cell monolayers, taken 48 h after transfection, shows efficacy of 1 through 6 µl of LipofectAMINE 2000 (C-H; phase images are shown to left of each condition). One microgram of EGFP-C1 vector was used in each condition. DNA-only (A) and lipid-only (B) controls are shown. Of note, in Table 1, data for LUC, %GFP-positive cells, and toxicity were generated on 16HBE14o cell cultures that were transfected at subconfluence and processed 48 h later at 80-90% confluence (e.g., never reached confluence). In the studies summarized in Table 1, LipofectAMINE PLUS was as efficient as LipofectAMINE 2000; however, LipofectAMINE PLUS was not as efficacious in 16HBE14o cell monolayers grown on 24-mm supports.

Chloride Channel Function of CFTR in Transiently Transfected Monolayers

View larger version (16K): [in this window] [in a new window]   Fig. 6.   CFTR chloride channel function in the apical membrane of transiently transfected epithelial monolayers assessed under open-circuit conditions. After RTE measurements under basal conditions and after amiloride treatment of the apical side of the monolayers for 10 min, forskolin was added to both sides of the epithelium for a 10-min incubation. The decrease in RTE is a measure of cAMP-activated CFTR chloride channels or chloride channels positively regulated by CFTR. n = 6 for WT-CFTR-transfected monolayers (pRSV-WT) and n = 2 for untreated, empty vector-transfected, and F508-CFTR-transfected monolayers (pRSV-F508). Absolute RTE values are provided in text. Similar data were generated in CFT-1 monolayers; however, the overall RTE and changes in RTE were smaller in magnitude.

CFTR-Dependent Rescue of RANTES Expression in Transiently Transfected Monolayers

View larger version (16K): [in this window] [in a new window]   Fig. 7.   CFTR-dependent rescue of RANTES expression in transiently transfected CFT-1 CF human airway epithelial monolayers. Experiments were performed in triplicate by sampling apical and basolateral medium 48 h after transfection of WT-CFTR or empty vector (mock) and after overnight treatment with the cytokines TNF- and IFN- (20 ng/ml each) or with carrier alone to both sides of the monolayer. All samples were transfected with pRL-CMV-LUC to standardize for transfection efficiency. Effectene was used as the transfecting lipid reagent for these experiments. RANTES ELISA was performed, and data were normalized to LUC activity measured after monolayers were lysed, after medium aliquots were harvested.

Immunoprecipitation of CFTR protein from transiently transfected monolayers. Ultimately, an important utility of this work is to transfect airway or heterologous epithelial monolayers transiently to study the cell biology and biochemistry of the CFTR protein in its normal physiological setting, the apical membrane of a polarized epithelium. Figure 8 includes representative SDS-PAGE gels of CFTR immunoprecipitation from IB3-1 and CFT-1 CF human airway epithelial monolayers transiently transfected with several different lipid transfecting reagents or combinations of lipid transfecting reagents. In IB3-1 monolayers, LipofectAMINE PLUS and LipofectAMINE 2000 alone or together with Effectene transfected monolayers effectively so that CFTR protein could be detected by immunoprecipitation (Fig. 8). Effectene and FuGENE 6.0 alone or in combination failed to transduce IB3-1 monolayers efficiently enough to observe a CFTR protein signal. In CFT-1 cells, Effectene alone failed to yield a CFTR protein signal; however, LipofectAMINE PLUS and LipofectAMINE 2000 in combination with Effectene produced a CFTR immunoprecipitation signal (Fig. 8). Together, these data show that transient transfection of resistive CF human airway epithelial monolayers is feasible and is an ideal system to study CFTR protein cell biology and biochemistry in a polarized epithelium.

View larger version (36K): [in this window] [in a new window]   Fig. 8.   Detection of transiently transfected CFTR protein via immunoprecipitation of CFTR from CF human airway epithelial monolayer lysates. A: transiently transfected IB3-1 monolayers were harvested 48 h afterollowing transfection with the different lipid cocktails listed (empty vector was pEGFP-C1 without the CFTR cDNA; the CFTR vector used was pGFP-WT-CFTR). Lane A, T84 monolayers included as a positive control; lane B, empty vector with Effectene reagent; lane C, CFTR with Effectene reagent; lane D, CFTR with Effectene and FuGene 6.0; lane E, FuGene 6.0; lane F, CFTR with Effectene and LipofectAMINE PLUS; lane G, CFTR with LipofectAMINE PLUS; lane H, CFTR with Effectene and LipofectAMINE 2000; and lane I, CFTR with LipofectAMINE 2000. B: transiently transfected CFT-1 monolayers were harvested 48 h after transfection with the different lipid cocktails listed (empty vector was pEGFP-C1 without the CFTR cDNA; the CFTR vector used was pGFP-WT-CFTR). Lane A, T84 monolayers as a positive control; lane B, empty vector with Effectene reagent; lane C, CFTR with Effectene; lane D, CFTR with Effectene and LipofectAMINE PLUS, and lane E, CFTR with Effectene and LipofectAMINE 2000. A hint of less core glycosylated B band is seen in the T84-positive control and in the positively transfected samples (but, interestingly, not in mock controls or unsuccessful transfections). These data are representative of 3 such experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

These studies are the first to show effective transient transfer of reporter genes and CFTR into human primary or immortalized non-CF and CF airway epithelial cells or MDCK epithelial cells grown as polarized epithelial monolayers. Very few studies have documented such findings (3, 6, 9) in polarized epithelial cell systems. This work benefited from the optimization of transient transfection in nonpolarized epithelial cell systems that supported our previous work with CFTR biochemistry, cell biology, and physiology. With this polarized monolayer system, we can ask many questions about CFTR cell biology and protein biochemistry, CFTR-dependent signaling and effects on gene expression, and CFTR physiology in a native and more in vivo-like polarized system. Early work suggests that CFTR behaves very differently in its cell biology and function in human airway epithelial cells vs. heterologous cells (Bebok Z and Collawn JF, unpublished observations).

With the exception of LipofectAMINE PLUS and LipofectAMINE 2000, however, many lipids were toxic to epithelial and heterologous cells grown in nonpolarized conditions. In particular, Effectene was very toxic to nonpolarized epithelial cells. Why then is a lipid like Effectene not toxic to polarized epithelial monolayers? The simplest explanation is that the epithelial cells, when packed tightly together as a resistive monolayer, are more resistant to toxicity. At no time in any transient transfection of a polarized epithelial monolayer did we have a large, sustained decrement in R_TE that did not recover. In fact, the transient decrease in R_TE caused by the lipid-mediated transfection was due to 6-h exposure to the serum-free medium rather than the DNA:lipid complexes.

Although it was not the intent of these studies to discover a new vehicle of gene transfer for CF, we may have, in fact, stumbled onto some lead candidates for lipids that transduce human airway epithelial cells from the apical side. Experiments examining the sidedness of LipofectAMINE reagent-based gene transfer showed that the side of entry was apical. Lack of basolaterally mediated transfer may have been caused by the collagen-coated filter support itself; however, the success of apically mediated delivery has led us to begin experiments with instillation of these more newly developed lipid transfecting reagents in mice.

It is important to emphasize that the same efficacy with a given lipid reagent in these epithelial cell systems may not hold for an epithelial cell from a different tissue. A screen of a panel of lipid reagents is necessary to select the reagent that is most effective for a given epithelial model. The four reagents shown to be effective in this study would be a blueprint from which to begin such an analysis. For example, early work with T84 and HT-29 monolayers suggests that LipofectAMINE 2000 is the best lipid for transient transfection of these human intestinal epithelial systems, whereas the others are less efficacious (Schwiebert EM, unpublished observations). In contrast, CFPAC-1 monolayers preferred LipofectAMINE PLUS over the other lipids (Zsembery A and Schwiebert EM, unpublished observations). Nevertheless, there are even newer lipid reagents that have come to market since we began this study, and these should be tested. However, because we were successful with mammalian expression vectors bearing LUC and GFP as well as multiple vectors bearing the CFTR cDNA, we feel that this system could be applied to most mammalian expression vector constructs as large as 10 kb.

In conclusion, transient transfection of polarized epithelial monolayers is feasible and effective. It is hoped that investigators will use these foundations to apply this system to epitheliim-specific cDNAs and proteins of interest to study the cell biology and physiology of epithelium-derived proteins in nonpolarized epithelial cells and, ultimately, in a polarized epithelium, their native and natural environment. Future studies are needed to also use this novel technology to assess whether these lipid transfecting reagents are suitable for gene transfer for CF in vivo.