Hyperacidification of cellubrevin endocytic compartments and defective endosomal recycling in cystic fibrosis respiratory epithelial cells.

The cystic fibrosis transmembrane conductance regulator (CFTR), which is aberrant in patients with cystic fibrosis, normally functions both as a chloride channel and as a pleiotropic regulator of other ion transporters. Here we show, by ratiometric imaging with luminally exposed pH-sensitive green fluorescent protein, that CFTR affects the pH of cellubrevin-labeled endosomal organelles resulting in hyperacidification of these compartments in cystic fibrosis lung epithelial cells. The excessive acidification of intracellular organelles was corrected with low concentrations of weak base. Studies with proton ATPase and sodium channel inhibitors showed that the increased acidification was dependent on proton pump activity and sodium transport. These observations implicate sodium efflux in the pH homeostasis of a subset of endocytic organelles and indicate that a dysfunctional CFTR in cystic fibrosis leads to organellar hyperacidification in lung epithelial cells because of a loss of CFTR inhibitory effects on sodium transport. Furthermore, recycling of transferrin receptor was altered in CFTR mutant cells, suggesting a previously unrecognized cellular defect in cystic fibrosis, which may have functional consequences for the receptors on the plasma membrane or within endosomal compartments.

The cystic fibrosis transmembrane conductance regulator (CFTR) 1 functions as an apical membrane chloride channel (1). Different CFTR mutations causing cystic fibrosis (CF) affect the processing, intracellular localization, and function of the corresponding protein (2,3). The most common mutant form of CFTR in CF, ⌬F508 CFTR, does not enter the organelles of the secretory pathway and is not delivered to the plasma membrane as it is not properly folded and remains trapped in the endoplasmic reticulum. Mutations in CFTR result in reduced apical chloride transport but also have pleiotropic effects on the function of other ion transporters including the amiloride-sensitive epithelial sodium channel (ENaC) (4,5), outwardly rectifying chloride channels (6,7), the Na ϩ /H ϩ exchanger via EBP50 (ezrin-binding protein), Na ϩ /H ϩ exchanger regulatory factor (8), bicarbonate conductance (9,10), and aquaporin 3 (5).
It has been proposed that CFTR also plays a role in facilitating acidification of intracellular compartments, such as endosomes, by providing anions (Cl Ϫ ) and maintaining charge neutrality as protons are pumped into the lumen of these organelles (11). According to this proposal, a loss of CFTR and chloride conductance would result in increased pH (11,12). However, repeated studies have failed to detect alkalinization of intracellular compartments in CF (13)(14)(15)(16)(17).
It has been shown that CFTR is present in endosomes of stably transfected Swiss 3T3 and T84 cells, which normally express CFTR (15). The absence of CFTR on the plasma membrane and organelles of the secretory pathway, which communicate with the endocytic pathway, prompted us to re-examine potential consequences in CF on the pH of endocytic organelles by specific targeting of pH-sensitive GFP (18) to a defined endocytic compartment. Here we show that cellubrevin-labeled endosomes are hyperacidified in CF lung epithelial cells and that the pH of the recycling endosome depends on CFTR and its effects on sodium transport. In addition, we show physiological defects in the function of the endocytic pathway in CF, as recycling of receptor-mediated endocytic tracers (transferrin) is affected in CF lung epithelial cells.

EXPERIMENTAL PROCEDURES
Cells and Tissue Culture-CFT1 (19,20) is a cell line derived from the tracheal epithelium of a CF patient homozygous for the most common CFTR ⌬F508 mutation. Stably transfected derivatives of CFT1were the following: CFT1-LCFSN, expressing the wild-type CFTR gene; CFT1-⌬508, transfected with ⌬F508 mutant CFTR gene; and CFT1-LC3, the vector-transfected control cells. CFT1 and derivative cells were grown in F12 media (Invitrogen) supplemented with 10 g/ml insulin, 1 M hydrocortisone, 1 nM triiodothyronine, 10 ng/ml cholera toxin (Sigma), 3.75 g/ml endothelial cell growth supplement, 25 ng/ml epidermal growth factor, and 5 g/ml transferrin (Collaborative Research Inc., Bedford, MA) (19). IB3-1 is a human bronchial epithelial cell line derived from a CF patient with a ⌬F508/W1282X CFTR mutant genotype (21). C38 and S9 are derivatives of IB3-1 cells and are stably transfected with a functional CFTR corrected for chloride conductance (22). The physiological levels of expression of CFTR and its functionality have been established previously for C38 cells (22). The cells were * This work was supported by Grant AI31139 from the National Institutes of Health and Grant 9680 from the Cystic Fibrosis Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Transfections-Cellubrevin-pHluorin GFP and glycosylphosphatidylinositol (GPI)-pHluorin GFP DNA constructs were from J. Rothman (18). IB3-1 cells and its derivatives were seeded at 10 5 cells/ml on 25-mm coverslips in 6-well plates. Cells were transfected with 1 g/ml DNA using Lipofectin (Invitrogen) for 6 h at 37°C, 5% CO 2 . CFT1 cells and their derivatives were seeded at 10 5 cells/ml on 25-mm coverslips in 6-well plates and grown in the medium without cholera toxin. Cells were transfected with GenePorter (Gene Therapy Systems, San Diego, CA) with 2.5 g/ml DNA for 4 h at 37°C, 5% CO 2 . Transfected cells were mounted in a perfusion chamber after 48 h of expression (Harvard Instruments, Holliston, MA) set at 37°C for live microscopy or otherwise processed for colocalization studies.
Fluorescence Microscopy and pH Measurements-Fluorescence microscopy was carried out using an Olympus IX-70 microscope and Olympix KAF1400 CCD camera (LSR, Olympus, Melville, NY). The ratio of emission at 508 nm upon excitation at 410 versus 470 nm was obtained using the previously described (18) filter sets (Chroma Technology Corp., Brattleboro, VT) mounted in a Sutter filter wheel (Sutter Instruments, Novato, CA) and controlled by the Merlin program (version 1.89, LSR, Olympus, Melville, NY). For the pH standard curve, two types of calibrations were carried out. (i) Cells transfected with GPI-pHluorin GFP were mounted in a perfusion chamber and incubated in buffer A (25 mM HEPES (pH changing from 7.4 to 5.5), 119 mM NaCl, 2.5 mM KCl, 2 mM CaCl 2 , 2 mM MgCl 2 , 30 mM glucose) at 37°C. Fluorescence images were taken upon excitation at 410 and 470 nm (six consecutive exposures). Three regions of interest were selected, and the standard curve was plotted as averaged 410/470 ratio values for a given buffer pH. (ii) At the end of experiments, the pH gradient was collapsed by incubating cells in 10 M monensin and 10 M nigericin for 30 min at 37°C in buffer A at pH 7.4 or 5.5, and ratios were recorded for internal standards. Sample pH was determined the same way as for the external standard curve.
Inhibition Studies-For H ϩ -ATPase inhibition, cells were incubated with 100 nM bafilomycin A (Sigma) in buffer A at pH 7.4 for 2.5 h at 37°C. For inhibition of sodium channels, 100 M amiloride (Sigma) was added in buffer A at pH 7.4 for 60 or 120 min at 37°C. For Na ϩ /K ϩ -ATPase inhibition, cells were incubated with 10 M acetylstrophanthidin in buffer A for 60 min, and pH was measured as above.
Organellar pH in CF Cells with Rescued CFTR via Temperature or Chemically Enhanced CFTR Trafficking-In experiments where CFTR folding and trafficking were rescued (23) by low temperature, mutant IB3-1 cells were grown at 26°C, 5% CO 2 for 40 h on glass coverslips. Organellar pH was determined using ratiometric GFP-pHluorin as described in sections above. 4-Phenylbutyric acid (4-PBA; gift from Triple Crown America Inc., Perkasie, PA) was used as an agent that promotes CFTR trafficking and rescue of its function (24,25). IB3-1 cells were grown in the presence of 2.5 mM 4-PBA at 37°C, 5% CO 2 for 40 h. Cells were then subjected to ratiometric determination of organellar pH as described above.
Normalization of Organellar pH in CF Cells with Ammonia-Cells were grown for 48 h in complete LHC-8 media in the presence of 0.1-1.0 mM NH 4 Cl (from Sigma) at 37°C in 5% CO 2 , and pH measurements were carried out as describe above.
Fluorescence Microscopy Localization Studies-For localization studies with fluorescently labeled transferrin, IB3-1 cells and derivatives grown on glass slides were transfected with cellubrevin-pHluorin GFP as described above. After 48 h of expression, cells were incubated for 30 min in DMEM (BioWhittaker, Walkersville, MD), 0.2% BSA (Sigma) at 37°C followed by a change of medium and incubation at 4°C for 30 min. 20 g/ml human transferrin conjugated to Texas Red (Molecular Probes) in DMEM, 0.2% BSA was added for 30 min at 4°C followed by three washes and incubation with DMEM, 0.2% BSA at 37°C for 15 and 120 min. When indicated, cells were treated with 20 g/ml nocodazole (Sigma) in DMEM, 0.2% BSA for 60 min following a 120-min treatment with transferrin. Samples were fixed with 3.7% paraformaldehyde, 5% sucrose for 10 min at room temperature, mounted with PermaFluor (Shandon, Pittsburgh, PA), and examined by fluorescence microscopy using a 570/20 excitation filter and a dichroic mirror/emitter cube set 8300 (Chroma Technology Corp.). For localization studies with CFTR-GFP and transferrin, IB3-1 cells and derivatives were transfected with CFTR-GFP. Transfection and transferrin incubation were as described above. For localization studies with ␣2,6-sialyltransferase, cells were co-transfected with 0.5 g of cellubrevin-pHluorin GFP and Myc-tagged ␣2,6-sialyltransferase DNA using 10 l of Lipofectin. After 48 h of expression, cells were fixed with 3.7% paraformaldehyde and permeabilized with 0.2% saponin for 5 min. Mouse monoclonal antibody (9E10) against c-myc (Santa Cruz Biotechnology, Santa Cruz, CA) was followed by goat anti-mouse secondary antibody conjugated to Alexa 568 (Molecular Probes). Glass slides were mounted using PermaFluor and analyzed by fluorescence microscopy using a 570/20 excitation filter and a dichroic mirror/emitter cube set 8300. For localization studies with dextran-Texas Red, cellubrevin-pHluorin-transfected IB3-1 cells and IB3-1 derivatives were incubated with 10 g/ml dextran-Texas Red followed by three washes. Cells were either fixed or live sequences were recorded immediately after removal of dextran-Texas Red every 30 s for 30 min using a monochromator excitation light source and emission filter sets on a microscope and camera controlled by TILLvisTRAC, version 3.3 (T.I.L.L. Vision Photonics, GMBH). For localization studies of cellubrevin-pHluorin GFP and EEA-1, IB3-1 and derivative cells were transfected with cellubrevin-pHluorin GFP. EEA1 was visualized using primary human anti-EEA1 antibody (Transduction Laboratories, Lexington, KY) and secondary Alexa 568-conjugated antibody.
Transferrin Recycling-Transferrin recycling was carried out as described previously (16). IB3-1 cells and their derivatives were incubated with 125 I-labeled transferrin for 45 min in DMEM, 0.2% BSA at 37°C. Cells were then washed three times with ice-cold DMEM, 0.2% BSA. The last wash was taken as 0 time point. Cells were then incubated for 15 min at 37°C, and medium was collected and replaced with fresh DMEM, 0.2% BSA for a further 45 min. Medium was collected, and cells were lysed to establish 100% of counts. Samples were counted in a ␥-counter (Beckman, Brae, CA) and expressed as % transferrin recycled at a given time point.
Horseradish Peroxidase (HRP) Uptake and Fluid Phase Endocytosis-The assay was carried out according to Li and co-workers (26,27). Cells were seeded in 6-well plates at 5 ϫ 10 5 /well at 24 h prior to assay. After being washed in serum-free DMEM, cells were incubated for 15 min at 37°C with either DMEM or 100 ng/ml wortmannin in DMEM. After washing, cells were incubated with 5 mg/ml HRP in DMEM, 0.2% BSA for 60 min at 37°C. Uptake was stopped by washing with 4°C phosphate-buffered saline, 0.2% BSA. Cells were lysed in phosphatebuffered saline, 0.1% Triton X-100. Lysate was added to O-phenylenediamine solution (HRP substrate) in a 96-well plate and incubated at room temperature for 5 min. Reaction was stopped by adding 1 M H 2 SO 4 , and A 490 was measured using a spectrophotometer (Shimadzu UV-1601, Shimadzu, Columbia, MD). Protein concentration of the lysate was determined by BCA reaction (Pierce), and uptake was expressed as A 490 /mg protein.

Expression and Localization of Cellubrevin-and GPI-pHluorin GFP Chimeras in CF and CFTR-corrected Bronchial Epi-
thelial Cells-In this study, we employed the recently developed pH-sensitive GFP (pHluorin GFP) system for ratiometric determination of the lumenal pH in intracellular organelles (18). Two pHluorin GFP fusion constructs were used (Fig. 1, a-h), one with GPI-pHluorin GFP and another with (endosomal v-SNARE) cellubrevin (18). GPI-pHluorin GFP is expected to result in the exposure of pHluorin GFP on the plasma membrane to the extracellular fluid. The cellubrevin-pHluorin GFP fusion has GFP exposed luminally in the intracellular compartments containing cellubrevin. The endosomal (cellubrevin) and plasma membrane (GPI)-targeted pHluorin GFP probes were transfected into well characterized human bronchial epithelial cells (21,22): IB3-1 (from a compound heterozygote CFTR ⌬F508/W1282X CF patient), C38 (IB3-1 cells corrected with a functional CFTR lacking the first ecto-loop), and S9 (IB3-1 cells corrected with a full size functional CFTR cDNA). These cells have been used as standard cell lines to model the effects of CFTR (6,21,22,24,28).
The plasma membrane localization of GPI-pHluorin GFP was demonstrated by responsiveness of GFP fluorescence to pH changes of the external buffer. Fig. 1, a-d, displays the fluorescence appearance of GPI-pHluorin GFP at pH 7.4 and 5.5. The cells expressing GPI-pHluorin GFP were used to generate a standard curve (Fig. 1i). All cells showed identical dependence of the GPI-pHluorin GFP fluorescence on pH of the external buffer. In addition to the plasma membrane labeling, as evidenced in Fig. 1, a-d, all cells transfected with GPI-pHluorin GFP showed a perinuclear fluorescence corresponding to a lipid raft recycling compartment, recently described by Lippincott-Schwartz and colleagues (29). Based on our observations, this compartment responds to external buffer pH ( Fig. 1, a-d), most likely because of the previously described rapid cycling of these membranes in constant communication with plasma membrane (29). GPI-pHluorin GFP fluorescence was not dependent on changes in concentration of other ions in the medium (e.g. sodium; data not shown). There were no differences in fluorescence ratios obtained with GPI-pHluorin GFP in IB3-1, C38, and S9 cells.
Localization of cellubrevin-pHluorin GFP was examined in both CF and CFTR-corrected cells by fluorescence microscopy using EEA1 antibodies, Texas Red-conjugated endocytic tracers. First, the cells were allowed to endocytose fluorescent transferrin, which was followed by chasing this marker of receptor-mediated endocytosis into the pericentriolar/paranuclear recycling compartment. This resulted in a significant colocalization of transferrin with cellubrevin-pHluorin GFP fluorescence in the transfected cells as evidenced by a similar overall organellar distribution (Fig. 2, a-c, IB3-1 cells; d-f, CFTR-corrected S-9 cells). Both the CF and CFTR-corrected cells showed similar overall organellar distribution. The colocalization of cellubrevin-pHluorin GFP and transferrin was not absolute in either cell line, as some of the cellubrevin-and transferrin-labeled profiles did not fully overlap, consistent with previous observations of strong but incomplete colocalization between transferrin and cellubrevin labeled vesicles (30). The most complete overlap was seen in the pericentriolar recycling endosomal compartment, strongly labeled by fluorescent transferrin, which was also the site of the majority of cellubrevin-pHluorin GFP labeled intracellular organelles. In further support of the overlap between cellubrevin and the recycling endosomal compartment, the treatment of cells with nocodazole, which causes depolymerization of microtubules and dispersion of the recycling endosome, resulted in redistribution of both transferrin and cellubrevin-pHluorin GFP fluorescence with a preservation of the significant overlap between the two markers ( Fig. 2, g-i). These observations suggest that cellubrevin-pHluorin GFP is localized in human bronchial epithelial cells with similar distribution in both CF and CFTRcorrected cells in the endosomal recycling compartment equivalent to what has been observed in several model cell lines (30 -33). Importantly, CFTR partially overlapped with the recycling endosome in bronchial epithelial cells (Fig. 2, j-l). The colocalization of CFTR-GFP and transferrin was similar to the one observed with cellubrevin-pHluorin GFP and transferrin (Fig. 2, a-f).
The cellubrevin-pHluorin GFP probe did not colocalize with the early endosomal marker EEA1, although the large EEA1positive profiles and the cellubrevin recycling endosome appeared to be closely apposed (Fig. 3, a-c). Treatment of cells with nocodazole confirmed that cellubrevin-pHluorin GFP and EEA1 were in distinct compartments (Fig. 3, d-f). The organellar distribution of EEA1 and cellubrevin compartments was similar in CFTR-corrected (Fig. 3, a-c) and CF cells (Fig. 3, insets in a-c). Cellubrevin-pHluorin GFP did not colocalize with peripheral endocytic organelles labeled with the fluid phase tracer dextran-Texas Red in fixed cells (data not shown) and in live cells monitored by time lapse microscopy ( Fig. 3,  g-l). Cellubrevin-pHluorin GFP was also tightly apposed to the ␣2,6-sialyltransferase, as revealed by immunofluorescence (Fig. 4), but remained localized distinctly from the TGN marker. There were no differences in localization of cellubrevin-pHluorin GFP in the CFTR mutant cells and CFTR-corrected cells (Fig. 4, a-d, CFTR-corrected cells; insets in b-d, CF cells). Collectively, these observations indicate that cellubrevin-pHluorin GFP probe was in the identical compartments in CFTR-corrected and CF cells and that the pH probe was in the recycling endosome.
Additional experiments were also carried out to confirm these findings. Hyperacidification of cellubrevin containing endosomes in CF cells was corrected by the addition of low concentrations of the weak base NH 4 Cl (0.1 mM) bringing pH to the values matching those observed in CFTR-corrected cells (Fig.  5a). Based on these experiments, we conclude that cellubrevinlabeled compartments are hyperacidified by 0.5 pH units in CF lung epithelial cells. This phenomenon was due to defective CFTR function and trafficking, as expression of a functional CFTR, but not that of ⌬F508 CFTR, restored the normal pH in CFT1 cells (Table I). To confirm this notion, we treated mutant CFTR IB3-1 cells by growing them at a permissive temperature (26°C), which allows CFTR folding and trafficking (23), or by adding the chemical chaperone 4-PBA, which restores trafficking and CFTR function (24,25), followed by pH determination using cellubrevin-pHluorin GFP-transfected cells. The results of these experiments are shown in Fig. 6. As both treatments (low temperature and chemical chaperone) restored normal pH of the cellubrevin-endosome in CFTR-mutant cells, it is possible to conclude that defective CFTR causes aberrantly low pH in this organelle.
To examine the pH of other parts of the endocytic pathway, IB3-1, C38, and S9 cells were allowed to endocytose 5 mM HTPS, a fluid phase pH-sensitive ratiometric dye, for 5 min. The ratio of fluorescence was measured after 10 and 60 min   C38, p ϭ 0.038, IB3-1 versus S9). However, after 60 min there was no significant difference in the ratios between IB3-3 CFTR mutant cells (ratio 3.02 Ϯ S.E. 0.159) and C38 (ratio 2.88 Ϯ S.E. 0.15) or S9 (ratio 2.84 Ϯ S.E. 0.14) cells (p ϭ 0.6732 IB3-1 versus C38, p ϭ 0.5881, IB3-1 versus S9). HTPS probe responsiveness was linear over a pH range from 7.4 to 5.5. These results indicate that an early endosomal compartment accessible to the exogenously added fluid phase probe is hyperacidified in CF bronchial epithelial cells but that the late, degradative endocytic organelles are not affected.

Hyperacidification of Cellubrevin Endosomal Compartments in CF Epithelial Cells Is a Sodium-dependent Process-How
might the absence of CFTR affect endosomal pH? It has previously been suggested that chloride channel activity of CFTR may affect organellar acidification in nasal polyp epithelial cells from CF (11). Such proposals are consistent with the role that Cl Ϫ anions are believed to play in dissipating membrane potential generated by proton pumping into the lumen, which otherwise inhibits H ϩ -ATPase activity (34). However, this model would predict organellar alkalization in CF and could not explain hyperacidification observed in our experiments. Instead, we considered an alternative hypothesis, in which Na ϩ efflux from the organelles could play a role in determining lumenal pH. It is known that in CF bronchial epithelial cells, the epithelial sodium channel, ENaC, is under negative regulation by CFTR (3)(4)(5)(35)(36)(37). In the absence of CFTR, as is the case in CF, ENaC is relieved from CFTR inhibition in lung epithelial cells, leading to an increase in Na ϩ transport. To test for the possibility that altered sodium transport could play a role in affecting organellar acidification in CF, we first established whether H ϩ -ATPase played a role in hyperacidification of the cellubrevin endosomes in CF cells. Treatment with bafilomycin A 1 abrogated hyperacidification of cellubrevin-labeled compartments in CF cells (Fig. 5b). Next, a role for sodium transport in hyperacidification was tested. The addition of amiloride, a sodium channel inhibitor, led to an increase in pH of 1 unit in cellubrevin labeled endosomes of IB3-1 CFTR mutant cells after 2 h of incubation (Fig. 5c). This observation is consistent with sodium transport, (i.e. sodium efflux from the organelles) playing a role in determining the pH of the cellubrevin endosomal compartment. Additional experiments (Fig. 5d) with acetylstrophanthidin, an Na ϩ /K ϩ -ATPase inhibitor, and ion substitution studies (data not shown) confirmed the role of Na ϩ conductance in organellar hyperacidification in CF.
Recycling of Transferrin Is Affected in CF Cells-Previous studies have indicated that altering endosomal pH can affect the function of the recycling endosome (38,39). To assess the functionality of cellubrevin recycling endosome in CF, we examined recycling of transferrin in CF and CFTR-corrected cells. IB3-1 cells and their CFTR-corrected derivatives, C38 and S9, were allowed to take up 125 I-labeled transferrin. The kinetics of transferrin recycling is shown in Fig. 6. There was no difference in recycling after 15 min. However, after 60 min, recycling of transferrin in IB3-1 (CFTR mutant) cells was reduced by 21% (p ϭ 0.0244) compared with the CFTR-corrected C38 cells and by and 16% (p ϭ 0.0054) relative to S9 cells. In contrast, endocytosis of transferrin or fluid phase endocytosis (Fig. 7, inset) was not different in CF and CFTR-corrected cells and was equally sensitive to the inhibitor of bulk endocytosis,   wortmannin (27) These results indicate that endosomal recycling is impaired in CF bronchial epithelial cells. DISCUSSION The studies reported here were inspired by previous models (11,12) in which altered pH in intracellular organelles was predicted in CF. However, our observations that the recycling endosomal compartment is hyperacidified in mutant CFTR IB3-1 and CFT1 cells is at variance with the previously published values for CF nasal polyp cells reporting slight alkalinization of the endosome (pH 6.8 in CF versus pH 6.3 in CFTR corrected cells) (11). Others have observed no differences in endosomal acidification of Swiss 3T3 fibroblasts (15), CFPAC-1 (17), or Chinese hamster ovary cells (13) transfected with either functional or nonfunctional CFTR. The discrepancies between these studies and our findings can be explained by the different cell types investigated, as in our work human bronchial and tracheal epithelial cells derived from CF patients were tested. It is know that, depending upon the cell type, CFTR may have either positive (4,37) or negative (40) regulatory effects on sodium channels, and so the cell type selection for testing is critical.
It is important to note that our data cannot be easily explained by the previously proposed action of CFTR as a chloride channel in the context of organellar acidification (11). Instead, regulatory functions of CFTR must be invoked, such as the CFTR-dependent inhibition of the sodium conductance in human respiratory epithelial cells (3,4,(35)(36)(37)40). In this model, excess positive charge, caused by the accumulation of H ϩ in the lumen of the organelles, may be compensated by Na ϩ efflux into the cytosol, thus dissipating the electrogenic charge differential (41) and allowing the H ϩ -ATPase to develop a greater transmembrane pH gradient. In the context of charge gain or loss, the impact of Na ϩ exit is equivalent to the influx of Cl Ϫ with the net effect of relieving the proton pump from the inhibition associated with the build up of membrane potential. In normal cells, inactive sodium channels, and most likely active Na ϩ /K ϩ -ATPase along with potassium channels, increase the interior positive membrane potential and thus counteract acidification. In CF cells, in the absence of CFTR-dependent inhibition (3)(4)(5)(35)(36)(37), the probability for open state of the sodium channel increases and Na ϩ -efflux compensates for the H ϩ -associated positive charge build-up, thus neutralizing the membrane potential and facilitating H ϩ -ATPase action and vesicle acidification. Independent studies show that the TGN, another compartment through which CFTR and sodium channel ENaC traffic in normal cells, is also hyperacidified in CF (42).
Because, as shown here, the function of the recycling endosome is affected in CF, this defect may have repercussions on endocytic and plasma membrane trafficking processes in this disease. For example regulation of plasma membrane signaling events by endocytosis, the availability of receptors and the duration of signals may be altered in CF. This may potentially contribute to the well recognized deficiencies in pro-and antiinflammatory signaling in CF (28). In addition, the endosomal pathway may affect the interactions of respiratory epithelial cells with the bacterial pathogens responsible for recurring respiratory infections in CF (1,43). Of particular interest in CF may be the repercussions of altered pH in the recycling endosome on membrane flow to the points of bacterial entry into epithelial cells, because recycling is reduced in CF lung epithelial cells. For example, phagocytosis is inhibited when delivery of membrane from the recycling endosome to the nascent phagosome is obstructed (44,45). Thus, the dysfunction of the recycling endosome in CF could affect the uptake of microorganisms by cells and explain the reduced bacterial phagocytosis reported for CF epithelia (46). In addition, hyperacidification of the recycling endosome, and potentially that of other compartments, may have effects on other fundamental cellular functions including the transcytosis of biologically active molecules and the pH homeostasis of both intracellular and extracellular environments. The phenomena described here suggest the existence of new physiological links between the CFTR defect, via organellar hyperacidification, and pathogenesis in CF. FIG. 7. Defective endosomal recycling in CF respiratory epithelial cells. CF IB3-1 cells (E) and CFTR-corrected C38 (f) and S9 (OE) cells were allowed to endocytose 125 I-labeled transferrin for 45 min at 37°C, excess transferrin was removed by washing at 4°C, and recycling was measured after 15 and 60 min at 37°C. Shown are mean values Ϯ S.E. (n ϭ 3). After 60 min, recycling was reduced in IB3-1 (CFTR mutant) cells by 21% (p ϭ 0.0244) compared with the CFTR-corrected C38 cells and by 16% (p ϭ 0.0054) relative to S9 cells. Inset, fluid phase endocytosis (measured by HRP uptake) is not affected in CF cells (IB3-1 versus C38, p ϭ 0.4966) and is equally sensitive to wortmannin (WM) in both CF and normal cells, IB3-1 versus C38, p ϭ 0.7954; C38 versus wortmannin C38, p ϭ 0.0001; IB3-1 versus wortmannin IB3-1, p ϭ 0.0001.