Low Temperature and Chemical Rescue Affect Molecular Proximity of ΔF508-Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) and Epithelial Sodium Channel (ENaC)*

Background: Mutations in CFTR lead to CF, a lethal inherited disorder. Results: The rescue of mutated CFTR affects its interaction with ENaC. Conclusion: The mutated version of CFTR prevents its close association with ENaC unless ΔF508-CFTR is rescued. Significance: The nature of the CFTR-ENaC interaction is important for the management of the airway pathology, which is now the major cause of mortality for CF patients. An imbalance of chloride and sodium ion transport in several epithelia is a feature of cystic fibrosis (CF), an inherited disease that is a consequence of mutations in the cftr gene. The cftr gene codes for a Cl− channel, the cystic fibrosis transmembrane conductance regulator (CFTR). Some mutations in this gene cause the balance between Cl− secretion and Na+ absorption to be disturbed in the airways; Cl− secretion is impaired, whereas Na+ absorption is elevated. Enhanced Na+ absorption through the epithelial sodium channel (ENaC) is attributed to the failure of mutated CFTR to restrict ENaC-mediated Na+ transport. The mechanism of this regulation is controversial. Recently, we have found evidence for a close association of wild type (WT) CFTR and WT ENaC, further underscoring the role of ENaC along with CFTR in the pathophysiology of CF airway disease. In this study, we have examined the association of ENaC subunits with mutated ΔF508-CFTR, the most common mutation in CF. Deletion of phenylalanine at position 508 (ΔF508) prevents proper processing and targeting of CFTR to the plasma membrane. When ΔF508-CFTR and ENaC subunits were co-expressed in HEK293T cells, we found that individual ENaC subunits could be co-immunoprecipitated with ΔF508-CFTR, much like WT CFTR. However, when we evaluated the ΔF508-CFTR and ENaC association using fluorescence resonance energy transfer (FRET), FRET efficiencies were not significantly different from negative controls, suggesting that ΔF508-CFTR and ENaC are not in close proximity to each other under basal conditions. However, with partial correction of ΔF508-CFTR misprocessing by low temperature and chemical rescue, leading to surface expression as assessed by total internal reflection fluorescence (TIRF) microscopy, we observed a positive FRET signal. Our findings suggest that the ΔF508 mutation alters the close association of CFTR and ENaC.

Mutations in CFTR, 3 the product of the cftr gene, lead to CF, a lethal autosomal recessive disorder. Clinically, CF is characterized by multisystem involvement; however, it is the airway involvement that is the leading cause of morbidity and mortality (1)(2)(3). The mutations lead to the disruption of the CFTR function as a Cl Ϫ channel (4), which in turn interferes with proper airway hydration. Because chloride secretion and sodium absorption are responsible for the proper airway hydration, inadequate chloride secretion leads to sodium hyperabsorption. This sodium hyperabsorption, mediated by ENaC, is believed to result from CFTR mutations changing the ability of the protein to regulate sodium transport. The role of ENaC was further supported in mice that exhibited CF-like symptoms when the ␤-subunit of ENaC was overexpressed (5). We and others have shown previously that the two transport molecules are electrophysiologically coupled, where the presence of CFTR decreases the activity of channels formed by ENaCs.
In our recent work, we documented the close association between WT CFTR and ENaC subunits, showing with both co-immunoprecipitation (co-IP) and FRET that there is an interaction between these proteins (6). In this study, we investigated the association of ENaC subunits with the mutated variant of CFTR, ⌬F508-CFTR. We found that all three ENaC subunits could be co-immunoprecipitated with ⌬F508-CFTR. The results of our FRET findings, on the other hand, did not place ⌬F508-CFTR and ENaC subunits in close proximity to each other. However, both chemical and low temperature rescue of ⌬F508-CFTR led to an observable FRET signal, placing these two proteins in sufficiently close proximity to each other for a direct association to take a place. Our biochemical findings using co-IP are suggestive of an overall association between mutated CFTR and ENaC subunits. In contrast, our FRET findings suggest that the ⌬F508 mutation disrupts the close association of CFTR and ENaC, leading to excessive ENaC activity.
Transient Transfection-Transient transfections were carried out using Lipofectamine 2000 (Invitrogen) according to the manufacturer's recommended protocol. Briefly, 1 l of Lipofectamine 2000 reagent (the DNA/Lipofectamine 2000 ratio was 0.2 g/1 l) was incubated with 100 l of Opti-MEM I (Invitrogen) at room temperature for 5 min. In a separate Eppendorf tube, 0.2 g of each DNA construct was incubated with 100 l of Opti-MEM I (Invitrogen). Following the 5-min incubation time, the diluted DNA constructs and Lipofectamine 2000 reagent were combined together and incubated for another 20 min. Then the transfection solution was added to the cells in Opti-MEM I (without fetal bovine serum and antibiotics). After a 5-h incubation at 37°C in a 5% CO 2 , 95% air incubator, the transfection solution was changed to the regular growth medium without antibiotics. Forty-eight hours posttransfection, the cells were rinsed with phosphate-buffered saline (PBS) (Invitrogen) and fixed for 15 min at room temperature with 4% paraformaldehyde (prepared from 20% EM Grade solution, Electron Microscopy Services, Hatfield, PA). After fixation, the cells were rinsed three times with PBS and mounted on glass microscope slides (Fisher) using 0.2% (w/v) n-propyl gallate (Sigma) in 9:1 glycerol/PBS (v/v). The same protocol was used to transfect the cells for co-IP experiments; the DNA/Lipofectamine 2000 ratio was 1 g/2.5 l.
ENaC Antibodies-Rabbit polyclonal antibodies (Abs) were generated (6) against synthetic peptides in collaboration with Drs. Mark Knepper and Patricia A. Gonzales (National Institutes of Health, Bethesda, MD).
Western Blot Analysis of CFTR and ENaC (10, 11)-HEK293T cells were transfected with 2 g of ECFP-CFTR construct. Forty-eight hours post-transfection, cells were lysed with radioimmunoprecipitation assay buffer (Pierce) complemented with Complete protease inhibitor mixture (Roche Applied Science) at 4°C. After centrifugation (15,800 ϫ g for 10 min at 4°C), non-soluble material was discarded. 100 g of total cell lysate was subjected to 6% SDS-PAGE (Invitrogen), followed by transfer to polyvinylidene difluoride (PVDF) membrane (Bio-Rad) that was probed with 1:10,000 diluted polyclonal Ab against the second nucleotide binding domain of CFTR (kind gift of Dr. J. Hong, University of Alabama at Birmingham Cystic Fibrosis Research Center). Detection was accomplished using a secondary anti-rabbit antibody conjugated to horseradish peroxidase (HRP) (Dako) and chemiluminescence. Western blot analysis of tagged ENaC constructs was as described previously (6).
⌬F508-CFTR and ENaC Co-IP Experiments-⌬F508-CFTR and ENaC co-IP experiments were performed following a protocol described previously (11). Briefly, the HEK293T cells were transfected with CFTR and appropriate combinations of ENaC subunits and incubated for 48 h. The cells then were lysed with 0.2% Triton X-100 in PBS with Complete protease inhibitor mixture (Roche Applied Science) at 4°C. The cell lysates were centrifuged at 15,800 ϫ g for 10 min at 4°C; the non-soluble material was discarded. The supernatant was incubated for 2 h with 1 g of carboxyl-terminal CFTR monoclonal Ab clone 24 -1 (R&D Systems, Minneapolis, MN) cross-linked to A/Gagarose beads (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Then the beads were pelleted and rinsed three times in PBS with 1% Triton X-100. Immunoprecipitated CFTR and bound proteins were analyzed on SDS-PAGE and transferred to a PVDF membrane, and the samples were processed for Western blotting using Abs raised against ENaC and CFTR (R&D Systems).
Surface Co-IP Assay-A combination of surface biotinylation and co-IP (12) was used to identify changes in surface expression following low temperature rescue. HEK293T cells were co-transfected with ⌬F508-CFTR and all three ENaC subunits (one of the three subunits in ␣␤␥-ENaC was tagged with EYFP-␣-EYFP-␤␥, ␤-EYFP-␣␥, and ␥-EYFP-␣␤). The cells were incubated at 27°C for 48 h after an initial 24-h incubation at 37°C.
The cells were washed three times with ice-cold PBS containing 1 mM MgCl 2 and 0.1 mM CaCl 2 and incubated with sulfosuccinimidyl 2-(biotinamido)-ethyl-1,3-dithiopropionate (Pierce; 1.5 mg/ml in ice-cold PBS) for 30 min at 4°C. The cells were then washed twice with freshly prepared quenching solution (50 mM glycine in PBS) and lysed by radioimmunoprecipitation assay buffer (150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, and 50 mM Tris, pH 7.4) with protease inhibitor mixture tablets (Roche Applied Science). Cell lysates were homogenized by passing fifty times through a 22-gauge needle and centrifuged (13,200 rpm, 30 min at 4°C). Protein concentration of the supernatant was measured using the bicinchoninic acid (BCA) protein assay (Pierce). Immunoprecipitation of CFTR was done by adding 4 g of anti-mouse CFTR antibody to 2000 g of protein cell lysate overnight at 4°C. After washing three times with lysis buffer, Protein G beads (100 l; Pierce) were added to the antibody-lysate complex and incubated overnight at 4°C. Beads were collected after centrifuging the lysates at 5000 rpm for 5 min and washed three times with lysis buffer. After removing residual buffer, bead-IP-CFTR interactions were disrupted by incubating the beads with 150 l of 1% SDS for 2 h at 37°C. The sample was centrifuged at 5000 rpm for 5 min, and the supernatant was mixed with 450 l of lysis buffer. After washing three times with lysis buffer, streptavidin beads (100 l; Pierce) were added to the samples and incubated overnight at 4°C. The beads were collected as described above. After removing residual buffer, samples were heated at 95°C for 6 min in 1ϫ Laemmli sample buffer (25% glycerol, 2% SDS, 0.01% bromphenol blue, 10% ␤-mercaptoethanol, 62.5 mM Tris HCl, pH 6.8) and subjected to SDS-PAGE over 10% separating gels. Following transfer, membranes were blocked for 1 h with 5% bovine serum albumin (BSA) in Trisbuffered saline (100 mM Tris (pH 7.5), 150 mM NaCl), with Tween 20 (0.1%; Bio-Rad) (TBS-T) at room temperature and probed with mouse anti-GFP monoclonal antibody (Abgent, San Diego, CA) at 1:2000 in 5% BSA in TBS-T overnight at 4°C.
FRET Imaging-For FRET measurements, an acceptor photobleaching technique was utilized as described previously (7,13). Briefly, an upright DMRXE Leica Confocal Scanning System (Exton, PA) TCS SP2 with an argon laser, ϫ100 1.4 numerical aperture plan apochromatic oil immersion objective was used for imaging. ECFP and EYFP fluorescence were excited with 458-and 514-nm laser light, respectively; a dual dichroic (DD458/514) was used in the excitation path. The emissions for ECFP and EYFP were collected using 465-505-nm and 525-600-nm band pass windows, respectively. Fixed samples were imaged within 3 days. We noticed an increase in background fluorescence if fixed samples were processed after a longer period (more than 7-10 days) of storage. The regions of interest in cells expressing both fluorophores were photobleached using 514-nm laser light to 30% of original intensity. ECFP and EYFP images were taken both before and after acceptor photobleaching. FRET efficiency (E) was calculated using Leica software, E ϭ D post Ϫ D pre /D post , where D pre and D post are ECFP emission before and after regional photobleaching, respectively.
FLIM Imaging-HEK293T cells seeded on glass bottom dishes (In Vitro Scientific, Sunnyvale, CA) were transfected with appropriate constructs (see "Transient Transfection").
FLIM measurements were carried out on a Zeiss 710 confocal microscope (Carl Zeiss MicroImaging, Thornwood, NY) equipped with Becker & Hickl GmbH (Berlin, Germany) time domain FLIM system. The excitation source was a 405-nm picosecond pulsed diode laser with a 50-MHz repetition rate and a pulse width of around 70 ps integrated into the Zeiss confocal system. The fluorescence decay curves were analyzed using SPC Image software (Becker & Hickl). The apparent mean lifetime (T m ) of the donor was derived from double exponential fit of the decay curves. An apparent FRET efficiency (E) was calculated according to the equation, E ϭ 1 Ϫ (T FRET / T ECFP-CFTR ). The difference between groups was calculated using Student's two-tailed test; the significance was set at p Ͻ 0.05.
TIRF Imaging-The imaging was carried out with HEK293T cells transfected with CFTR constructs at room temperature of 25°C as described previously (14). Briefly, the cell growth medium was changed to a solution that contained 140 mM NaCl, 5 mM KCl, 2 mM CaCl 2 , 2 mM MgCl 2 , 10 mM HEPES, and 5 mM glucose (pH 7.4) before the start of the imaging. The inverted IX81 Olympus microscope (Center Valley, PA) outfitted with differential interference contrast, wide field epifluorescence, and total internal reflection fluorescence illumination was used for imaging. Images were captured with a ϫ60 plan achromatic oil immersion TIRF microscope objective with a numerical aperture of 1.45 using a CoolSNAP-HQ-cooled, charge-coupled device camera (Roper Scientific, Tucson, AZ) controlled by Metamorph imaging software version 7.0 (Molecular Devices, Sunnyvale, CA). The laser beam was brought to the correct angle for total internal reflection by focusing it on the outer edge of the objective's rear aperture. The light source was an argon laser (10 milliwatts at 488 nm, Melles Griot (Carlsbad, CA)). For epifluorescence visualization of EYFP and ECFP, we used standard fluorescein/FITC and ECFP filter sets; a xenon arc lamp (100 watts) was used as a light source. All images shown in the figures represent pseudocolored raw data with their pixel intensities without saturation and within the camera's dynamic range (0 -4095).

RESULTS
Analysis of Fluorophore-tagged Constructs-Western blot analysis demonstrated that the fusion of a 27-kDa fluorescent tag (ECFP or EYFP) to CFTR and ⌬F508-CFTR did not affect our ability to immunoprecipitate or detect the protein (Fig. 1A) (6). Also, fusion of the fluorescent tag to the WT CFTR did not interfere with its trafficking as evidenced by predominant plasma membrane delineation of the staining (supplemental Fig. S1A) (6,8) or Cl Ϫ channel function (6,8). In contrast, tagged ⌬F508-CFTR constructs show predominant intracellular staining without plasma membrane delineation (supplemental Fig. S1B), suggestive of the impaired trafficking characteristic of this CFTR mutant (15).
We have previously shown that Western blot analysis of tagged ENaC revealed the expected bands matching the molecular weight of the ENaC subunits with the fluorescent tags (6). When tagged ENaC subunits were expressed in cells, microscopy showed a more prevalent ER staining; this pattern of staining is characteristic of ENaC constructs and indicates poor traf-ficking (6). Nevertheless, functional currents with typical ENaC characteristics were recorded from cells transfected with tagged ENaC constructs (7,16,17).
ENaC Subunits Interact with ⌬F508-CFTR by IP-The recent electrophysiological findings (18), co-IP data (6), and FRET results (6) have suggested a close association between WT CFTR and ENaC. Here, we assessed a physical interaction between the mutated version of CFTR (⌬F508-CFTR) and ENaC. HEK cells were transfected with wild-type or ⌬F508 CFTR alone or in combination with ␣␤␥-ENaC, and then individual ENaC subunits were probed with appropriate subunit specific antibodies. Forty-eight hours later, the cells were lysed and immunoprecipitated with a C-terminal CFTR antibody, separated by SDS-PAGE, and transferred to PVDF membrane. Blots were incubated with antibodies directed against either the C terminus of CFTR or each individual ENaC subunit. As shown in (Fig. 1, B-D), ENaC subunits were only co-precipitated with CFTR in cells expressing both proteins, suggesting that ⌬F508-CFTR and ENaC can associate with each other. No co-IP was observed when either ⌬F508-CFTR or ENaC was omitted. These findings were comparable with WT CFTR and ENaC association studies (Fig. 1, B-D) (6).
⌬F508-CFTR is a processing mutation that interferes with folding and trafficking of CFTR, resulting in retention of mutant protein in ER. The effects of this mutation can be partially reversed with low temperature (19) incubation of the mutant CFTR. We have combined surface biotinylation with co-IP (12) to evaluate the cell surface co-IP signal of ⌬F508-CFTR and ENaC subunits following the low temperature correction. ⌬F508-CFTR and all three ENaCs (one of three subunits in ␣␤␥-ENaC was tagged with EYFP in each case) were co-transfected, and then each individual ENaC subunit was probed. Compared with the 37°C incubation (Fig. 2, left side of each panel), the low temperature incubation at 27°C (Fig. 2, right side of each panel) enhanced the co-IP signal, which was statistically significant (Fig. 2, corresponding bar graphs). These data suggest that the low temperature rescue of ⌬F508-CFTR enhances the ⌬F508-CFTR and ENaC surface co-IP signal.
Next, we explored the possibility that the random aggregation as a result of overexpression could lead to co-IP of ENaC with ⌬F508-CFTR. The lysates of the cells expressing ␤-ENaC were mixed with the lysates of the cells expressing ⌬F508-CFTR; no co-IP signal was observed, indicating that association of ␤-ENaC and ⌬F508-CFTR did not occur postlysis (Fig. 3A). As another control, we performed the co-IP experiment with ⌬F508-CFTR and another channel, ECFP-ClC-1. As Fig. 3B demonstrates, no co-IP signal was observed between ⌬F508-CFTR and ECFP-ClC-1. The anti-GFP monoclonal Ab JL-8 (BD Living Colors), which also recognizes CFP, was used to confirm the expression of ECFP-ClC-1 in the cells (Fig. 3B). Our findings indicate that ⌬F508-CFTR and ENaC can associate with FIGURE 1. Western blot analysis of tagged CFTR constructs. A, the blot was carried out using a polyclonal Ab raised against the NBD2 domain of CFTR and probed with an anti-rabbit HRP using chemiluminescence. ENaC subunits interact with both WT and ⌬F508 CFTR. ␣-ENaC (B), ␤-ENaC (C), and ␥-ENaC (D) could be co-immunoprecipitated using a C-terminal CFTR antibody when co-expressed with ⌬F508-CFTR; ␣␤␥-ENaC and ⌬F508-CFTR were co-expressed, and then individual subunits were probed. No ENaC signal was detected when ␣-, ␤, or ␥-ENaC or ⌬F508-CFTR was expressed alone. The left panels show that ␣-, ␤, or ␥-ENaC can only be detected when expressed in HEK-293 cells. Therefore, ENaC subunits are not endogenously expressed in those cells. These experiments were repeated at least three times, and similar results were obtained. each other in a specific manner when co-expressed in HEK293T cells.
FRET Measurements by Acceptor Photobleaching-Next, we utilized FRET imaging to assess the association between ⌬F508-CFTR and ENaC. FRET can only be observed when fluorescent donor and acceptor proteins are less than 10 nm apart (20 -22). In the acceptor photobleaching mode of FRET, bleaching of the acceptor will result in an increase in donor fluorescence if FRET was occurring prior to bleaching (13). The ratios of the donor (ECFP) fluorescence emission before and after selective photobleaching of the acceptor are used to calculate an apparent FRET efficiency (E).
As in our studies previously described (6), cells transfected with ECFP-linker-EYFP fusion protein served as a positive control, whereas cells transfected with separate ECFP and EYFP constructs served as a negative control for FRET (supplemental Fig. S2). A 514-nm laser was used to bleach the regions of interest. Images of ECFP and EYFP were taken before and after selective photobleaching of the acceptor. The ECFP-EYFP fusion construct (positive control) demonstrated ϳ30% FRET efficiency, whereas our negative control yielded less than ϳ2% ( Fig. 4 and supplemental Fig. S2).
The HEK293T cells were transfected with ECFP-⌬F508-CFTR and all three ENaC subunits; one of the three subunits carried EYFP as a tag. Supplemental Fig. S3 depicts a set of before and after images obtained following a selective bleaching of an acceptor (ENaC subunit tagged at the C terminus with EYFP). The green box shows the area of bleaching. We did not find an increase in CFP fluorescence after selective bleaching of the acceptor; the energy transfer efficiency values (ϳ3%) were not significantly different from those of the negative controls (ϳ2%, p Ͼ 0.44). A bar graph with a summary of E values for all combinations of tagged ENaC and ECFP-⌬F508-CFTR is given in Fig. 4. These measurements indicate that FRET did not occur between ⌬F508-CFTR and any ENaC subunits, suggesting that the misfolding of the ⌬F508-CFTR prevents its close association with ENaC. As mentioned previously, we found an appreciable increase in CFP fluorescence after selective photobleaching with all combinations of ENaC and WT CFTR; the energy transfer efficiency values averaged ϳ7% (p Ͻ 0.05), but values as high as ϳ17% were observed (6).
Previous studies have shown that low temperature (19) and some chemical correctors (23)(24)(25)(26)(27) can partially improve the folding and consequently trafficking of the ⌬F508-CFTR (28 -32). Because our model, based on lipid bilayer studies, shows a functional interaction between the two proteins at the bilayer, intracellularly retained ⌬F508-CFTR may not have a similar interaction with ENaC because the surface expressed ⌬F508-CFTR. We therefore attempted further FRET experiments in cells transfected with ⌬F508 and exposed to either low temper-FIGURE 2. Evaluation of the changes in surface expression following low temperature rescue of ⌬F508-CFTR. A combination of surface biotinylation and co-IP was utilized to monitor changes in surface expression. The cells were co-transfected with ⌬F508-CFTR and all three ENaC subunits; one of three subunits in ␣␤␥-ENaC was tagged with EYFP -␣-EYFP-␤␥ (A), ␤-EYFP-␣␥ (B), and ␥-EYFP-␣␤ (C). The low temperature rescue enhanced the co-IP signal and was statistically significant (*, p Ͻ 0.05; **, p Ͻ 0.01; ***, p Ͻ 0.001; corresponding densitometry bar graphs). The changes in densitometry value at 27°C were normalized to that at 37°C (100%). The bands at 56 kDa in each panel indicate the high molecular weight immunoglobulin band. These experiments were repeated at least three times with similar results. Error bars, S.E. ⌬F508 CFTR and ENaC Link MAY (33) reported that glafenine, an anthranilic acid derivative, was able to partially correct ⌬F508-CFTR misprocessing in several in vitro model systems. We therefore used glafenine for our ⌬F508-CFTR and ENaC FRET experiments. We found that glafenine (10 M in DMSO, 48-h incubation) resulted in a positive FRET signal (ϳ11% versus ϳ3% in negative controls, p Ͻ 0.003) between ⌬F508-CFTR and ENaC subunits (supplemental Fig. S4 and summary bar graph in Fig. 5A); this treatment, however, does not affect positive and negative FRET controls (Fig. 5A, bars 1 and 2). Next, we attempted FRET measurements between ⌬F508-CFTR and ENaC following low temperature (27°C for 48 h after an initial 24-h incubation at 37°C following transfection) rescue of ⌬F508-CFTR (supplemental Fig. S5 and summary bar graph in Fig. 5B). Again, we were able to measure a significant FRET signal (ϳ13% versus ϳ2% in negative controls, p Ͻ 0.009) between ⌬F508-CFTR and ENaC subunits (supplemental Fig. S5); this treatment does not affect positive and negative FRET controls (Fig. 5B).
Additionally, we attempted the FRET measurement between ⌬F508-CFTR and ENaC subunits following the rescue by a combination of Corr3a and Corr4a (supplemental Fig. S6 and summary bar graph in Fig. 5C). Corr3a (quinazoline compound) and Corr4a (thiazole compound) are small molecules that act as pharmacological chaperones, promoting the folding and trafficking of ⌬F508-CFTR (28 -32). However, the interpretation of these FRET findings was complicated by the fact that with these "specific" correctors, we observed increased FRET efficiency in our negative FRET controls (ϳ10% compared with Ͻ2% without the correctors). This interesting finding is probably due to the specificity and mechanism of action of these correctors, suggesting that perhaps a GFP-based FRET assay might be useful to screen/identify novel correctors.
Next, we examined the specificity of our positive FRET signal between ⌬F508-CFTR and ENaC subunits following partial correction of ⌬F508-CFTR by low temperature and chemical correctors. For this purpose, we repeated our low temperature and chemical correction experiments with ECFP-ClC-1 and ENaC subunits. Supplemental Fig. 7A demonstrates the life-time images of cells expressing positive (ECFP and EYFP fusion construct; left) and negative FLIM controls (ECFP and EYFP constructs co-expressed separately; right); the decrease in the mean ECFP lifetime to 0.8 ns (left) from 2.5 ns (right) indicates the energy transfer from ECFP to EYFP. Supplemental Fig. 7B demonstrates the lifetime images of cells expressing ClC-1 tagged with ECFP. In the absence of ENaC tagged with EYFP, the mean lifetime of ECFP tagged to ClC-1 averaged around 2.3 ns before (supplemental Fig. S7B, far left panel labeled 37°C) and following low temperature and chemical correction (supplemental Fig. S7B, panels labeled 27°C, glafenine, Corr3a, and Corr4a). We did not observe a decrease in the donor lifetime, an indication of FRET taking place, when ECFP-ClC-1 was co-expressed with any ENaC subunit tagged with EYFP (supplemental Fig. S8, A-C). The mean ECFP lifetime tagged to ClC-1 (in the presence of any EYFP-tagged ENaC) was 2.2-2.5 ns, not significantly different from that in the absence of EYFP (p Ͼ   Fig. 6 demonstrates a summary bar graph of energy transfer efficiencies under different conditions. Taken together, the presence of a FRET signal between ⌬F508-CFTR and ENaC and its absence between ClC-1 and ENaC following the low temperature and chemical correction further strengthens the specificity of energy transfer between ⌬F508-CFTR and ENaC subunits. TIRF Microscopy-We next assessed the trafficking of ⌬F508-CFTR following low temperature and chemical rescue by utilizing TIRF microscopy. This approach allows the study of tagged molecules near or at the plasma membrane. Data shown in supplemental Fig. S9 show TIRF images of cells transfected either with WT (supplemental Fig. S9A) or ⌬F508-CFTR (supplemental Fig. S9B) tagged on the N terminus with EYFP. The pattern of ⌬F508-CFTR expression shows predominant cytoplasmic distribution indicative of various membrane-bound organelles (supplemental Fig. S9B; inset depicts the epifluorescence image of the same cell) as compared with a staining pattern closer to or at the plasma membrane of WT CFTR (supplemental Fig. S9A; inset depicts the epifluorescence image of the same cell). Rescue by low temperature and chemical correctors resulted in an improved membrane distribution of ⌬F508-CFTR (supplemental Fig. S10); interestingly, the combination of the low temperature and chemical correctors had a more pronounced effect on ⌬F508-CFTR (supplemental Fig. S10D) near or at plasma membrane expression. These TIRF results verify that low temperature and chemical rescue improves the trafficking of ⌬F508-CFTR.

DISCUSSION
In our recent study (6), we demonstrated the close association between WT CFTR and ENaC subunits using biochemical and FRET approaches. In this study, we probed the possible link between the most common disease-causing variant of CFTR, ⌬F508-CFTR, and ENaC. Our co-IP findings are suggestive of a physical association between ⌬F508-CFTR and ENaC subunits. However, our FRET findings indicate that the mutated version of CFTR prevents a close association between ⌬F508-CFTR and ENaC unless ⌬F508-CFTR is rescued by low temperature and chemical correctors.
The co-IP of the two transport molecules would, at first, suggest that there is some direct interaction between the two proteins. It has often been argued that in overexpression systems, the presence of an interaction as detected by co-IP is an artifact due to the superphysiological levels of the two proteins. However, as we showed previously with WT CFTR (6), the interaction of ⌬F508-CFTR is specific for the ENaC subunits because we do not see an interaction with another transport protein, the chloride channel ClC-1. Because ⌬F508-CFTR is poorly trafficked secondary to the ⌬F508 mutation and it is well described that WT ENaC subunits are poorly trafficked (16,34), it is possible that the co-IP interaction occurs in the ER or during targeting for degradation or retention. This idea is supported by the knowledge that CFTR expression alters ENaC trafficking. Although testing this would be challenging, it is a good example of the weaknesses of co-IP to either rule in or rule out a direct physical interaction. Although a positive signal validates that in these conditions the two proteins are in some complex together, it does not necessitate a direct physical interaction, and similarly a negative signal would not rule out the possibility that there is never a direct physical interaction.
Conversely, FRET cannot occur beyond the physical limit of 10 nm, with FRET efficiency changing dependent on the relative positions of the fluorescent tags (35). Our FRET findings did not place the fluorescent tags on ⌬F508-CFTR in sufficiently close proximity to the fluorescent tags on ENaC for an association to occur, and the identity of ENaC subunits did not change the efficiency of energy transfer (Fig. 4). The simple interpretation of these data is that the deletion at residue position 508 prevents the close association of CFTR and ENaC. Indeed, this fits well with the pathogenesis of CF airway disease, which postulates that the mutation prevents the down-regulation of ENaC by CFTR (1). This CFTR-mediated inhibition of ENaC activity might involve the gating (18) of ENaC and/or the protection of ENaC from proteolysis (36) that activates near silent ENaCs (37). These possibilities are not mutually exclusive; the close proximity or direct physical association of CFTR with ENaC conceivably might represent a step to protect the ENaC subunits from proteolysis and subsequent activation.
It is possible that the physical targeting of ⌬F508-CFTR is different from that of ENaC subunits, which effectively prevents FRET. For example, ENaC might be predominantly at the surface while ⌬F508-CFTR is retained intracellularly, but perhaps if both were at the surface, the two could interact directly and be observed by FRET. However, this is not the case because a fluorescent signal was clearly visible for both proteins inside the cell and at the surface (supplemental Fig. S3, A-C). The energy transfer efficiencies measured in delineated plasma membrane versus broad regions of interest across the cells were not significantly different from each other ( Fig. 4 and supplemental Fig. S3, A-C). Interestingly, the lack of FRET signal with the ⌬F508-CFTR (without the partial rescue by the low temperature and chemical correctors) also validates that the FRET signal between WT CFTR and ENaC was not due to overexpression, tight packing, or some other artifact as could have been conjectured with our prior work (6).
Of note, however, our prior work showed WT CFTR and ENaC to FRET intracellularly as well as at the surface (6). The lack of FRET shows that there is some significant difference between the interactions of the ⌬F508-CFTR mutant and WT CFTR. The absence of these interactions could be very important for the design and implementation of therapeutics. These findings demonstrate that the ⌬F508 mutation has the potential to disrupt the CFTR and ENaC link, a disruption that may account for the observed hyperabsorption of sodium in CF airways. Thus, correcting the trafficking of this mutant, while allowing for proper chloride transport, may still not fix the sodium transport problem. Perhaps proper trafficking of the mutant will not be sufficient to correct the pathophysiology of the end stage lung disease found in CF. More acutely, perhaps measures of ENaC function as noted by the effect of amiloride on the nasal potential difference should be an important variable examined during clinical trials of CFTR correction.
If we are to believe both the FRET and the co-IP data in this study, it is possible that ⌬F508-CFTR reduces the distance between the N terminus of CFTR and the C terminus of ENaC subunits but that there is a physical interaction through another domain that allows for a positive co-IP signal. We tested this possibility with the rescue of the ⌬F508-CFTR trafficking defect by low temperature and chemical correctors. This resulted in an increase in the efficiency of energy transfer between ⌬F508-CFTR and ENaC subunits ( Fig. 5 and supplemental Figs. S4 -S6). Because the correctors help release ⌬F508-CFTR from ER retention, the positive FRET signal with rescue suggests that the interaction takes place after the channels are trafficked from the ER. At the same time, the presence of FRET signal suggests close proximity of the two proteins; it does not, however, distinguish the presence (or the absence) of the functional coupling between the rescued ⌬F508-CFTR and ENaC (38). Future experiments will need to examine other mutants of CFTR, such as those that traffic to the membrane but are electrophysiologically quiescent, such as the G551D-CFTR.
Although we were able to observe a FRET signal between ⌬F508-CFTR and ENaC subunits following low temperature and chemical rescue of ⌬F508-CFTR, they were considerably lower compared with FRET measurement between WT CFTR and ENaC subunits (6). These results suggest that even when the folding and trafficking of the ⌬F508-CFTR are corrected, the interaction with ENaC subunits is not fully restored. This may be suggestive of an impaired functional interaction between the two; however, published work regarding the consequences of the ⌬F508-CFTR correction are complex (39) and not only can lead to partial restoration of CFTR-mediated Cl Ϫ transport but may also affect the amiloride-sensitive current (40,41). However, contradictory data are presented, with certain studies suggesting that when ⌬F508-CFTR is properly trafficked by low temperature rescue, ENaC function is only corrected if ⌬F508-CFTR is potentiated with genistein (42), whereas others suggest that with recovery using chemical chaperones, such as sodium 4-phenylbutyrate, the ⌬F508-CFTR is unable to regulate ENaC function, both in vitro (38) and in vivo as measured by the nasal potential difference response to amiloride (23,26). Contrast this with in vivo results garnered by correcting a stop mutation of CFTR, which in theory folds properly and traffics without difficulty, where there is a significant correlation between the change in amiloride-sensitive nasal potential difference and recovery of CFTR function in the nasal epithelia (43). These data taken together would reiterate that ⌬F508-CFTR has a different functional relationship from WT CFTR, as captured by the differences in our FRET efficiencies.
It may also be important to determine what led to the elevation of the negative control signal with the use of Corr3a/4a. Both Corr3a and Corr4a have been shown to improve the folding and trafficking of ⌬F508-CFTR (30); it is noteworthy that Corr4a has shown more specificity for correction of ⌬F508-CFTR than Corr3a (30). One possibility is that the elevation of negative control signal may indicate some potential off-target specificities of the correctors because folding and trafficking is a complex and multistep process; further studies will be needed to address this. Another possibility is that these correctors are capable of causing artificial cross-linking or complexing of the free fluorescent proteins, bringing them into close enough proximity to cause a positive FRET signal. This is very plausible because both the positive control of the fused YFP/CFP and the free floating YFP/CFP increase by approximately equivalent amounts (5%), as shown in Fig. 5C. Unlike our transport proteins, the FRET controls are not restricted to the membrane, and thus it is possible that these small molecule correctors are able to force the control fluorophores to complex in a manner in which the tethered transport protein tags are unable to. Regardless, the data clearly establish that with correction there is statistically significant FRET between ⌬F508-CFTR and ENaC subunits. Further, it is noteworthy that our study has been carried out in a model system with recombinant ⌬F508-CFTR and ENaC subunits; it remains to be explored if the same is true in in vivo circumstances. In synopsis, WT CFTR and ENaC subunits interact at the plasma membrane and internally, as evidenced by our prior FRET studies. This study establishes that although ⌬F508-CFTR and ENaC subunits appear to interact directly via co-IP, their cytoplasmic tails do not FRET intracellularly and only weakly FRET when trafficking is corrected. This regulation of Na ϩ transport, following the ⌬F508-CFTR correction, probably occurs directly; however, our results cannot exclude the possibility of a role for an adaptor protein(s) in this association. Future work will need to examine which domains of these proteins are mediating this interaction and assess the nature of this interaction under dynamic conditions, such as when either transport protein is activated or inhibited. Understanding the structural and kinetic nature of this interaction may be critical to better managing the airway pathology that is the current life-limiting factor in patients afflicted with CF.