Modulating the body's networks could become mainstream therapy for many health issues.
Image credit: The Feinstein Institutes for Medicine Research.
New Research In
Physical Sciences
Featured Portals
Articles by Topic
Biological Sciences
Featured Portals
Articles by Topic
Contributed by Michael J. Welsh, March 25, 2016 (sent for review January 29, 2016; reviewed by Ronald G. Crystal and James M. Wilson)
Significance
Cystic fibrosis transmembrane conductance regulator (CFTR) gene mutations cause cystic fibrosis (CF), and airway infections cause most morbidity. In airways, the CFTR forms a channel that secretes chloride and bicarbonate. A persistent question for CF gene- and cell-based therapies is how much CFTR is needed to correct host defense defects that predispose to infection. In addressing this question, we were informed by discoveries that, without CFTR-mediated bicarbonate secretion, liquid covering airways becomes abnormally acidic, which impairs airway host defenses. By studying airway epithelia from CF, non-CF, and CFTR heterozygote piglets, we found a relatively linear relationship between amounts of CFTR, bicarbonate secretion, and host defense properties. The results have direct implications for developing therapeutics. They may also explain the risk of airway disease in CF carriers.
Abstract
Cystic fibrosis (CF) is caused by mutations in the gene encoding the cystic fibrosis transmembrane conductance regulator (CFTR) anion channel. Airway disease is the major source of morbidity and mortality. Successful implementation of gene- and cell-based therapies for CF airway disease requires knowledge of relationships among percentages of targeted cells, levels of CFTR expression, correction of electrolyte transport, and rescue of host defense defects. Previous studies suggested that, when ∼10–50% of airway epithelial cells expressed CFTR, they generated nearly wild-type levels of Cl− secretion; overexpressing CFTR offered no advantage compared with endogenous expression levels. However, recent discoveries focused attention on CFTR-mediated HCO3− secretion and airway surface liquid (ASL) pH as critical for host defense and CF pathogenesis. Therefore, we generated porcine airway epithelia with varying ratios of CF and wild-type cells. Epithelia with a 50:50 mix secreted HCO3− at half the rate of wild-type epithelia. Likewise, heterozygous epithelia (CFTR+/− or CFTR+/∆F508) expressed CFTR and secreted HCO3− at ∼50% of wild-type values. ASL pH, antimicrobial activity, and viscosity showed similar relationships to the amount of CFTR. Overexpressing CFTR increased HCO3− secretion to rates greater than wild type, but ASL pH did not exceed wild-type values. Thus, in contrast to Cl− secretion, the amount of CFTR is rate-limiting for HCO3− secretion and for correcting host defense abnormalities. In addition, overexpressing CFTR might produce a greater benefit than expressing CFTR at wild-type levels when targeting small fractions of cells. These findings may also explain the risk of airway disease in CF carriers.
Cystic fibrosis (CF) is an autosomal recessive disease caused by mutations in the gene encoding the cystic fibrosis transmembrane conductance regulator (CFTR) (1⇓–3). CFTR is an anion channel expressed in multiple cell types, including the apical membrane of airway epithelia. Loss of CFTR function in airway epithelia impairs host defense, which causes bacterial infection, inflammation, mucus accumulation, obstruction, and remodeling, the major sources of CF morbidity and mortality. Although there has been significant progress in treating people with CF, the disease continues to shorten lives.
Potential genetic-based treatments for CF lung disease include gene transfer of a wild-type (WT) CFTR cDNA (gene therapy) (4⇓–6), correction of the CFTR gene sequence (genome editing) (7), repair of CFTR mRNA (8), and stem-cell transplantation to airways (9). Gene- and cell-based approaches have potential for long-lasting benefits and are applicable for all CFTR genotypes, but they are limited by CFTR restoration to a limited number of cells. Consequently, a major question is what percentage of cells must express CFTR to reverse CF airway host defense defects.
Previous studies have begun to address this question. Several studies used viral vectors to overexpress CFTR in CF airway epithelial cells and then examined the relationship between the percentage of transduced cells and transepithelial Cl− secretion. Johnson et al. (10) used a retrovirus to overexpress CFTR in an immortalized CF epithelial cell line and measured forskolin-stimulated Cl− secretion as the cAMP-induced change in short-circuit current [∆Isc(cAMP)]. They found that, when ∼10% of the cells overexpressed CFTR, ∆Isc(cAMP) was ∼60% of WT levels and that ∼50% of cells overexpressing CFTR generated ∼80% of WT ∆Isc(cAMP). Goldman et al. (11) used an adenovirus vector to overexpress CFTR in airway epithelial xenografts. They found that CFTR expression in ∼7% of cells generated changes in transepithelial voltage, in the presence of a Cl− concentration gradient, that were ∼75% of voltage changes in non-CF xenografts. Zhang et al. (12) used a parainfluenza virus to overexpress CFTR in cultured human CF airway epithelia. ∆Isc(cAMP) progressively increased as the percentage of transduced epithelial cells increased. When 60% of the cells were transduced (the highest level tested), the ∆Isc(cAMP) was plateauing, although there was no comparison with WT epithelia. Farmen et al. (13) approached this question differently by generating airway epithelia with varying ratios of CF and non-CF airway epithelial cells (which express CFTR at endogenous levels) and measuring transepithelial Cl− secretion. They found that epithelia containing 50–60% of non-CF cells had Cl− secretion rates equal to non-CF epithelia. Dannhoffer et al. (14) mixed 10% non-CF cells with 90% CF cells and found that ∆Isc(cAMP) was ∼90% of non-CF values. Taken together, the presence of CFTR in ∼10–50% of airway epithelial cells, even at endogenous levels, generated a transepithelial Cl− secretory current that was approximately the same as that in non-CF airway epithelia. Finding that overexpressing CFTR in a fraction of cells and that expressing CFTR at endogenous levels in a fraction of cells generated similar Cl− secretion rates suggested that, for purposes of gene therapy, overexpressing CFTR confers no advantage compared with endogenous levels of expression. The plateau of Cl− secretion when 10–50% of the cells express CFTR is explained by a limitation to Cl− entry into cells at the basolateral membrane and anion movement between cells through gap junctions (13).
A related question is how much CFTR should be expressed in individual cells. That is, on a per cell basis, are greater than WT levels of CFTR more efficacious at correcting host defense defects? One approach to answering this question is to study airway epithelia from people who are heterozygous for a CF-causing mutation (CF carriers). A few studies in humans and mice measured transepithelial voltage (Vt) across nasal epithelia before and during perfusion of a solution that is Cl−-free and contains an agent to increase cellular levels of cAMP to phosphorylate and activate CFTR (15⇓⇓–18). Two reports also studied cultured mouse and human cells (17, 18). Most, but not all, studies found no differences between control and CFTR heterozygotes. However, conclusions from those studies are limited because Vt does not provide a quantitative measure of ion transport, mice do not develop CF airway disease, and infection and inflammation may produce secondary alterations in transepithelial electrolyte transport. Despite evidence of similar Cl− transport, CF carriers are predisposed to airway sinus disease (19⇓–21), bronchiectasis (22, 23), and asthma (24⇓⇓–27). Those findings suggest that airway host defense might be impaired in CF carriers and that measures of transepithelial Cl− secretion may not be sufficiently sensitive to detect a mild abnormality.
We recently developed CFTR−/− and CFTR∆F508/∆F508 pigs (28, 29). At birth, their airways lack infection and inflammation, but, over the ensuing weeks and months, they develop the hallmark features of CF airway disease (30). By studying newborn CF piglets, we identified at least two airway host defense defects (2). Mucociliary transport is impaired by mucus with abnormal biophysical properties (31, 32), and the activity of airway surface liquid (ASL) antimicrobials and synergism between antimicrobials are impaired (33, 34). CF ASL has an abnormally low pH, which reduces the activity of ASL antimicrobials and increases the viscosity of ASL. The abnormally acidic ASL pH results from loss of CFTR-mediated HCO3− secretion in the presence of continued H+ secretion (35). CFTR is also key for HCO3− secretion and alkalization of ASL pH in small airways (36).
The importance of HCO3− secretion for airway host defense suggested that knowing the relationship between CFTR expression and HCO3− secretion might inform development of gene- and cell-based therapies. Therefore, we asked what percentage of airway epithelial cells expressing CFTR and what level of CFTR expression would rescue defective HCO3− secretion and restore abnormalities related to airway host defense. We studied three epithelial models. (i) We cultured porcine airway epithelia composed of mixtures of WT and CF (CFTR−/− or CFTR∆F508/∆F508) cells in varying ratios. In the WT cells, CFTR was expressed at endogenous levels; endogenous expression levels are low, with an estimate of 1–2 CFTR transcripts per cell (37). (ii) We cultured epithelia composed entirely of heterozygous (CFTR+/− or CFTR+/∆F508) cells. (iii) We produced CF, heterozygous, and WT epithelia that overexpressed CFTR in a small fraction of cells after adenovirus-mediated gene transfer.
Results
HCO3− Secretion and Cl− Secretion Show Different Relationships to CFTR Expression.
We generated primary cultures of airway epithelia from newborn pigs. They were grown at the air–liquid interface and differentiated and were studied at least 2 wk after seeding as previously reported (38, 39). We used newborn pigs to avoid secondary changes that might ensue from infection, inflammation, and airway remodeling. To determine the relationship between CFTR expression and anion secretion, we studied epithelia with varying ratios of CFTR−/− (CF) and CFTR+/+ (WT) cells. Because the most common CFTR mutation in humans is deletion of phenylalanine at position 508 (ΔF508) (3), we also examined airway epithelia produced by mixing cells from CFTRΔF508/ΔF508 pigs with cells from CFTR+/+ littermates (29). After epithelia were studied, quantitative PCR of genomic DNA indicated that the percentages of CF and WT cells in epithelia were as predicted by the seeding ratio (Fig. 1A). In addition, CFTR mRNA and CFTR protein paralleled the percentage of WT cells (Fig. 1 B–D). We measured short-circuit current (Isc) in Ussing chambers after addition of forskolin and 3-isobutyl-1-methylxanthine (IBMX). These agents elevate cellular levels of cAMP leading to phosphorylation and activation of CFTR and stimulation of anion secretion (40).
Fig. 1.
The amount of CFTR varies with the percentage of WT cells and with the heterozygous state. (A) Airway epithelia were generated with varying ratios of CFTR∆F508/∆F508 and CFTR+/+ cells. Quantitative polymerase chain reaction (qPCR) of genomic DNA was used to determine the fraction of WT CFTR DNA. Data are normalized to 0% for CFTR∆F508/∆F508 epithelia and 100% for WT epithelia. Data are mean ± SEM; n = 4. Line is linear regression, R2 = 0.88, P < 0.0001. (B) Quantitative reverse transcription polymerase chain reaction was used to assay CFTR mRNA in CFTR+/+, CFTR+/−, and CFTR−/− airway epithelia. Data are mean ± SEM normalized to CFTR mRNA in CFTR+/+ epithelia; n = 4. The * indicates P < 0.05 relative to CFTR+/+ by unpaired Student's t test. (C) Example of Western blot of CFTR in airway epithelia generated with varying ratios of CFTR∆F508/∆F508 and CFTR+/+ cells and in CFTR+/∆F508 epithelia. (D) Quantification of Western blots of CFTR. CFTR+/∆F508 is the red square. Data are mean ± SEM normalized to 0% for CFTR∆F508/∆F508 and to 100% for CFTR+/+ epithelia; n = 5. The line is linear regression, R2 = 0.82, P < 0.0001. The amount of CFTR in CFTR+/∆F508 epithelia and epithelia composed of a 50:50 mix of CFTR∆F508/∆F508 and CFTR+/+ cells did not differ by unpaired Student's t test.
As the percentage of WT cells increased, Cl− secretion increased and plateaued at about a 50:50 mixture of CFTR−/−:CFTR+/+ cells (Fig. 2A) and CFTRΔF508/ΔF508:CFTR+/+ cells (Fig. 2B). These data are consistent with results from human airway epithelia expressing either endogenous levels of CFTR or overexpressing CFTR (10⇓⇓⇓–14). However, as the percentage of WT cells increased, cAMP-stimulated HCO3− secretion continued to increase until 100% of the cells were WT (Fig. 2 C and D). Fifty % of WT cells generated ∼50% of WT HCO3− secretion. This pattern was similar when the CF cells were either CFTR−/− or CFTRΔF508/ΔF508.
Fig. 2.
HCO3− secretion varies proportionally with the percentage of WT epithelial cells. Epithelia were composed of mixtures of CFTR−/− and CFTR+/+ cells (A and C) or mixtures of CFTR∆F508/∆F508 and CFTR+/+ cells (B and D) in varying proportions. Red squares indicate epithelia composed entirely of heterozygous epithelia (CFTR+/− or CFTR+/∆F508). (A and B) Solutions contained 130 mM Cl− and were HCO3−/CO2-free. (C and D) Solutions contained 25 mM HCO3−/5% CO2 and were Cl−-free. Data are mean ± SEM of changes in short-circuit current (∆Isc) induced by adding 10 μM forskolin and 100 μM IBMX apically to increase intracellular levels of cAMP. To allow comparison between multiple groups of mixed epithelia, data were normalized for each group of epithelia so that ∆Isc for CF (CFTR−/− and CFTR∆F508/∆F508) epithelia was set at 0% and ∆Isc for epithelia composed entirely of WT cells was set at 100%. For epithelia containing 100% CF cells and 100% WT cells, respectively, the mean ± SEM ∆Isc were as follows: (A) 0.7 ± 0.1 μA⋅cm−2 and 14.2 ± 1.0 μA⋅cm−2; (B) −0.5 ± 0.2 μA⋅cm−2 and 4.0 ± 0.2 μA⋅cm−2; (C) 0.3 ± 0.1 μA⋅cm−2 and 2.3 ± 0.2 μA⋅cm−2; (D) −0.2 ± 0.2 μA⋅cm−2 and 2.0 ± 0.2 μA⋅cm−2. In A and C, n = 13 for CFTR−/−, n = 13 for CFTR+/+, n = 6–9 for mixtures of CFTR−/− and CFTR+/+, and n = 8–9 for CFTR+/−. In B and D, n = 6 for CFTR∆F508/∆F508, n = 6 for CFTR+/+, n = 6 for mixtures of CFTR∆F508/∆F508 and CFTR+/+, and n = 1–4 for CFTR+/∆F508. (E and F) The amiloride-induced change in Isc (∆Isc(amiloride)) and Gt (∆Gt(amiloride)) in a Cl− (HCO3−-free) solution or an HCO3− (Cl−-free) solution. Heterozygote epithelia are identified in red. n = 7–19. In Cl− containing solution, ∆Isc(amiloride) increased as the percentage of WT epithelia increased (P < 0.02 by linear regression). Other data in E and F showed no relationship to the percentage of WT cells. The * indicates the difference from value for epithelia composed of 100% WT cells; P < 0.05 by ANOVA. The # indicates that the value for ∆Isc(HCO3−) differed from the corresponding value for ∆Isc(Cl−) by unpaired Student’s t test, P < 0.05.
We also studied epithelia composed entirely of heterozygous (CFTR+/− and CFTR+/ΔF508) cells. Consistent with the finding that heterozygous epithelia contain half as much WT CFTR mRNA as WT epithelia (Fig. 1B) (41), heterozygous epithelia had half as much CFTR protein as CFTR+/+ epithelia (Fig. 1 C and D). cAMP stimulated Cl− secretion to levels similar to CI− secretion in WT epithelia (Fig. 2 A and B). In contrast, cAMP stimulated HCO3− secretion to levels half those generated in WT epithelia (Fig. 2 C and D).
Thus, 50% of CFTR generated ∼50% of the HCO3− secretion observed in WT epithelia. This result occurred both when 100% of the cells expressed CFTR at 50% of WT levels (i.e., heterozygous epithelia) and when 50% of the cells expressed CFTR at 100% of WT levels (i.e., mixtures of CF and WT cells). These results contrast with measures of Cl− secretion in which 50% of CFTR generated ∼100% of the Cl− secretion observed in WT epithelia.
Airway epithelia also absorb Na+ through amiloride-inhibitable apical epithelial Na+ channels (ENaCs) (2). In porcine and human epithelia, apical amiloride produces a greater change in Isc in CF than non-CF because of the lack of a Cl− conductance (42, 43). Previous studies used gene transfer vectors to overexpress CFTR in human CF epithelia and measured basal transepithelial electrical properties or the amiloride-induced reduction in Isc [∆Isc(amiloride)]. The results have varied widely. One study found that expressing CFTR in ∼10% of cells produced a near maximal reduction in ∆Isc(amiloride) (10). In contrast, a different study reported that CFTR expression in all of the cells reduced ∆Isc(amiloride), but there was no relationship to vector dose (44). Between those extremes, one study reported that, as the percentage of transduced cells increased (∼60% was the highest percentage tested), ∆Isc(amiloride) progressively decreased (12). Another study reported variable effects of CFTR expression on basal transepithelial voltage in CF xenografts (11).
In a Cl− (HCO3−-free) solution, ∆Isc(amiloride) decreased as the percentage of WT cells increased (Fig. 2E). However, the amiloride-induced reduction in transepithelial electrical conductance ∆Gt(amiloride) was not affected (Fig. 2F). These results are consistent with an effect of the Cl− conductance on ∆Isc(amiloride), rather than changes in ENaC activity (42, 43). Those data and that conclusion predicted that removing Cl− would eliminate the relationship between percentage of WT cells and ∆Isc(amiloride). Indeed, compared with a 140-mM Cl− solution, in a solution containing 25 mM HCO3−, which has a lower permeability through CFTR than Cl−, there was no relationship between the percentage of WT cells and ∆Isc(amiloride) or ∆Gt(amiloride) (Fig. 2 E and F).
Increasing CFTR Expression Progressively Enhances ASL Host Defense Properties.
To establish the relationship between the percentage of WT cells in an epithelium and assays related to host defense, we did several studies.
As previously reported (32, 33), CFTR−/− epithelia had a lower ASL pH than CFTR+/+ epithelia (Fig. 3A). Mixtures of CFTR−/− and CFTR+/+ cells generated intermediate ASL pH values, and a 50:50 mixture produced an ASL pH that was approximately half way between CF and WT.
Fig. 3.
Increasing the percentage of non-CF cells in CF epithelia increases ASL pH, enhances antimicrobial activity, and reduces ASL viscosity. Airway epithelia were composed of varying ratios of CFTR−/− and CFTR+/+ cells (A–D) or CFTR∆F508/∆F508 and CFTR+/+ cells (E). Epithelia composed entirely of heterozygous cells (CFTR+/− or CFTR+/∆F508) are indicated by red squares. Measurements were made 2 h after basolateral addition of 10 μM forskolin and 100 μM IBMX. Data are mean ± SEM; in some cases, error bars are hidden by symbols. (A) ASL pH was measured using a ratiometric pH indicator, SNARF-conjugated dextran distributed in the ASL. n = 5–6. (B) ASL antimicrobial activity was assayed by touching bacteria-coated grids to the ASL for 1 min and then determining the percentage of bacteria killed. n = 4–6. (C) ASL viscosity (τASL/τsaline) was assessed by fluorescence recovery after photobleaching (FRAP). n = 4. (D) ASL depth was measured using confocal microscopy to detect a fluorescent tracer in ASL. n = 4. (E) ASL pH was measured as in A, but with mixtures of CFTR∆F508/∆F508 and CFTR+/+ cells. n = 7–8. The * indicates the difference from epithelia with 100% WT cells by repeated measures ANOVA, P < 0.05.
A reduced ASL pH inhibits the activity of ASL antimicrobials, which impairs bacterial killing (33, 34). We previously developed an assay of ASL antimicrobial activity in which a small gold grid with attached bacteria is briefly touched to the ASL in vivo or in vitro and removed, and the percentage of nonviable bacteria is counted (33, 34). We routinely use Staphylococcus aureus because it more commonly infects the airways of young children with CF and neonatal CF piglets than Pseudomonas aeruginosa (30, 45). However, we have observed similar pH-dependent killing of P. aeruginosa (33, 34). Compared with WT epithelia, bacterial killing was reduced in CF epithelia (Fig. 3B). Epithelia with mixed cell types showed intermediate levels of bacterial killing, consistent with the rates of HCO3− secretion and ASL pH.
CF mucus has altered properties that impair mucociliary transport (31). Recent studies showed that CF ASL has an increased viscosity due to a reduced ASL pH (32, 35). In epithelia composed of mixtures of CFTR−/− and CFTR+/+ cells, the relationship between the percentage of WT cells and ASL viscosity (τASL/τsaline) approximately paralleled that of ASL pH (Fig. 3C). Variations in viscosity were not due to differences in ASL depth (Fig. 3D) (42).
We also studied epithelia containing mixtures of CFTRΔF508/ΔF508 and CFTR+/+ cells (Fig. 3E). Like mixtures of CFTR−/− and CFTR+/+ cells, those mixtures produced intermediate values of ASL pH.
Consistent with the finding that CFTR+/− epithelia had reduced rates of HCO3− secretion, heterozygous epithelia exhibited intermediate values of ASL pH, bacterial killing, and viscosity (Fig. 3 A, B, C, and E). These results with heterozygous epithelia are similar to results obtained with 50:50 mixtures of CF and WT epithelia.
CFTR Overexpression in a Small Percentage of Cells Increases HCO3− Secretion and ASL pH.
If CFTR is the rate-limiting step for HCO3− secretion, then increasing CFTR expression to supernormal levels, even in a limited number of cells, should increase HCO3− secretion. To test this prediction, we used an adenovirus vector to deliver CFTR linked to GFP (46). A cytomegalovirus promoter was used to drive high-level expression in transduced cells. We determined percentages of transduced cells based on GFP expression. To understand the consequences of overexpressing CFTR, we transduced CFTR−/−, CFTR+/−, and CFTR+/+ epithelia. Previous studies indicate that the adenovirus vector itself does not affect CFTR function or the function of other epithelial channels (11, 44, 47).
The vector transduced varying percentages of cells (Fig. S1). In epithelia of all three genotypes, HCO3− secretion increased to levels greater than levels obtained when we generated epithelia with cells expressing endogenous levels of CFTR. For example, when an epithelium comprised ∼10% CFTR+/+ cells and ∼90% CFTR−/− cells, cAMP-stimulated HCO3− secretion was ∼10% of WT levels (Fig. 2 C and D). In contrast, when a CFTR−/− epithelium comprised ∼10% of transduced cells overexpressing CFTR, cAMP-stimulated HCO3− secretion was ∼60–70% of WT levels (Fig. 4A). CFTR overexpression also increased HCO3− secretion in WT epithelia. For example, in some epithelia, HCO3− secretion approached values almost 400% of CFTR+/+ epithelia.
Fig. 4.
Overexpressing CFTR increases HCO3− secretion, but ASL pH does not exceed WT values. CFTR−/− (red), CFTR+/− (gray), and CFTR+/+ (blue) epithelia were treated with adenovirus encoding CFTR-GFP and studied 3 d later. GFP fluorescence and DAPI fluorescence (nuclear stain) were assayed to determine the percentage of epithelial cells that were transduced (circles). Mean ± SEM of control epithelia not treated with adenovirus are shown as squares. (A) HCO3− secretion induced by adding forskolin (10 μM) and IBMX (100 μM) basolaterally. Data are the change in Isc in Cl−-free, HCO3−/CO2-containing solutions [∆Isc(HCO3−)]. Data are normalized to ∆Isc(HCO3−) in nontransduced CFTR+/+ epithelia as 100% (dotted line, from Fig. 2) and in nontransduced CFTR−/− epithelia as 0%. ∆Isc(HCO3−) in transduced CFTR−/− epithelia and in transduced CFTR+/+ epithelia was greater than in control, nontransduced CFTR−/− and CFTR+/+ epithelia, respectively, by unpaired Student's t test, P < 0.001. (B) ASL pH was measured 2 h after basolateral addition of 10 μM forskolin and 100 μM IBMX. ASL pH of transduced CFTR−/− epithelia was greater than that of control, nontransduced CFTR−/− epithelia (from Fig. 3) by unpaired Student's t test, P < 0.001. (C) Relationship between cAMP-stimulated HCO3− secretion and ASL pH. Data are from A and B.
Fig. S1.
The relationship between CFTR mRNA expression and the percentage of GFP-positive cells. CFTR−/−, CFTR+/−, and CFTR+/+ epithelia were treated with adenovirus encoding CFTR-GFP and studied 3 d later. CFTR mRNA was measured by qPCR and normalized to WT epithelia as 100%. GFP fluorescence and DAPI fluorescence (nuclear stain) were assayed to determine the percentage of GFP-positive epithelial cells. The line is linear regression, R2 = 0.75, and P < 0.0001.
Overexpressing CFTR elevated ASL pH in CF and heterozygous epithelia (Fig. 4B). However, despite the increase in HCO3− secretion, CFTR overexpression did not further increase ASL pH in WT epithelia. These consequences of CFTR overexpression are apparent from an examination of the relationship between cAMP-stimulated HCO3− secretion and ASL pH (Fig. 4C). As HCO3− secretion increased above WT levels, there was no additional increase in ASL pH.
Discussion
The Rate of HCO3− Secretion Depends on CFTR Expression.
Previous studies have shown that, when 10–50% of the cells in an airway epithelium are WT or overexpress CFTR, Cl− secretion rises to the level of WT epithelia (10⇓⇓⇓–14). However, knowledge that CFTR conducts HCO3− (48), that HCO3− is the major pH buffer of ASL (49), and that HCO3− secretion plays a key role in airway host defense (31⇓⇓⇓–35, 50) led us to test the relationship between CFTR expression and HCO3− secretion. We found that, unlike Cl− secretion, HCO3− secretion was directly proportional to CFTR expression. Direct proportionality was the case for mixing studies, for CFTR+/− epithelia, and for overexpression of CFTR.
Different relationships for Cl− secretion and HCO3− secretion can be attributed to limitations at the basolateral membrane. Cl− entry into the cell occurs primarily through the basolateral Na+/K+/2Cl− cotransporter (NKCC) (51). After CFTR reaches >50% of WT levels, Cl− secretion plateaus as Cl− entry into the cell becomes rate-limiting. The fact that the plateau occurs in cell-mixing studies indicates that this phenomenon is not cell autonomous and depends on Cl− movement between cells through gap junctions (10, 13). In contrast, the relatively proportional relationship between CFTR expression and the rate of HCO3− secretion and the ability of CFTR overexpression to further increase the rate of HCO3− secretion in WT epithelia indicate that the basolateral membrane is not rate-limiting, even when CFTR expression exceeds WT levels. Perhaps the basolateral membrane is not rate limiting because there are multiple mechanisms for generating intracellular HCO3− and basolateral HCO3− entry and/or expression of those processes at relatively high levels (52, 53). Although we cannot exclude the possibility that HCO3− does not readily move through gap junctions, this concept seems unlikely given their permeability to charged molecules and many molecules smaller than 500 Da (54).
As HCO3− Secretion Increases, ASL pH Reaches a Maximum.
Loss of HCO3− secretion in the face of continued proton secretion contributes to the abnormally reduced pH of CF ASL (35). Although increasing HCO3− secretion increased ASL pH, pH reached a maximum at ∼7.50. Note that HCO3− secretion is a rate, and it was measured with transepithelial voltage clamped at zero. In addition, apical and basolateral solutions had constant and identical pH, HCO3− concentration, and other ion concentrations. In contrast, ASL pH is assayed at equilibrium. Consequently, ASL pH is influenced by the transepithelial voltage, transepithelial chemical gradients (including H+ and HCO3− gradients), and paracellular permselectivities. In addition, changes in the composition of the ASL might have a regulatory influence on transport processes. Those factors, together with the presence and localization of the cellular transport processes, impose a physiologic limitation on the maximum ASL pH (53).
Reduced HCO3− Secretion in CFTR+/− Epithelia May Explain the Risk of Airway Diseases in CF Carriers.
The observation that people heterozygous for a CF-causing CFTR mutation had normal rates of airway epithelial Cl− secretion made it difficult to understand why such individuals have an increased risk of airway disease (19⇓⇓⇓⇓⇓⇓⇓–27). Our data provide a potential explanation. Compared with WT controls, heterozygous epithelia had a reduced HCO3− secretion rate, a reduced ASL pH, and impaired properties associated with airway host defense. Thus, although most CF carriers have no observed lung disease, we speculate that mild impairment of host defense might predispose to airway disease late in life. If that is the case, perhaps increasing HCO3− secretion or elevating ASL pH might benefit airway host defense in CF carriers.
This Study Has Advantages and Limitations.
An advantage is that we used a porcine model of CF that develops the characteristic features of CF lung disease (29, 30). To avoid the effects of long-term infection and inflammation, including potential epigenetic changes in epithelial cells, we generated airway epithelia from newborn animals, which we previously showed exhibit airway host defense defects (30, 31, 33). We found comparable results in CFTR−/− and CFTRΔF508/ΔF508 epithelia. We examined the effect of multiple interventions to alter HCO3− secretion—mixing cells, studying heterozygote epithelia, and overexpressing CFTR. We measured electrolyte secretion and related it to properties associated with host defense.
A limitation of this study is that loss of CFTR may influence the function of other cell types and structures, including submucosal glands (31, 55), airway smooth muscle (56), cartilage (57, 58), and myeloid-derived cells (59). Impairment of their function might also contribute to the pathogenesis of CF airway disease. It is also possible that, in a mixed epithelium, substances in a WT cell type might modify functions in a CF cell, or the converse. Another limitation is that we do not know which cells (ciliated, goblet, or nonciliated columnar) are most important for HCO3− secretion and ASL pH control. Pertinent to our gene transfer studies, adenovirus vectors are reported to transduce all of the cell types in airway epithelia (44, 60), and we therefore suspect the same occurred in this study. Finally, these studies were done in vitro rather than in vivo. With the development of future animal models, it may be possible to address some of these questions and the consequences for airway disease in vivo.
These Findings Have Implications for Therapy Development.
A major challenge for potential gene- and cell-based therapies is targeting a sufficient number of cells. Our data suggest that expressing CFTR at supernormal levels in a small percentage of cells might restore epithelial HCO3− secretion, ASL pH, and abnormalities associated with host defense. This rescue might be the case for gene transfer approaches, but it is not likely to occur with genome editing, CFTR mRNA repair, or stem cell transplantation. Thus, for the same numbers of airway cells targeted, our data suggest that there might be greater benefit from gene transfer approaches that overexpress CFTR. However, there are additional considerations. For example, high-level expression can target some CFTR to the basolateral membrane (13). Basolateral CFTR might be less of a limitation for HCO3− than for Cl− secretion; the transepithelial electrochemical gradients for HCO3− suggest that basolateral CFTR expression might not have adverse effects on HCO3− secretion or ASL pH under equilibrium conditions. Another consideration is that overexpressing CFTR has been reported to inhibit cell proliferation, arrest cell growth, increase cell volume, depolarize the cell membrane, and alter CFTR function (61⇓–63).
There are also implications for other potential CF treatments. Our data suggest that the therapeutic window for augmenting CFTR function might be broader than previously thought, extending all of the way from no CFTR function to WT levels of CFTR function. However, increasing CFTR activity above WT levels may not further elevate ASL pH. Nevertheless, previous studies showed that increasing ASL pH above levels observed in WT epithelia further enhances antimicrobial activity and reduces ASL viscosity (32, 34). Therefore, elevating ASL pH by adding a pH buffer, inhibiting H+ secretion, or some other intervention might have additional therapeutic benefit and might be of value in other diseases with a reduced ASL pH.
Materials and Methods
Also see SI Materials and Methods.
CFTR−/− (CF), CFTR+/− (heterozygote), and CFTR+/+ (WT) pigs were the product of CFTR+/− matings (28). CFTRΔF508/ΔF508 (CF), CFTR+/ΔF508 (heterozygote), and CFTR+/+ pigs were the product of CFTR+/ΔF508 matings (29). These studies were approved by the University of Iowa Animal Care and Use Committee. Airway epithelial cells from the inferior turbinate were grown at an air–liquid interface as previously described (38). Epithelia were generated with varying proportions of CF and WT cells and studied at least 14 d after seeding. Nasal epithelia have electrolyte transport characteristics comparable with tracheal/bronchial epithelia (42). We used standard procedures for quantitative PCR and RT-PCR, Western blots, Ussing chamber analysis of electrolyte transport, fluorescence recovery after photobleaching, measurement of ASL pH, ASL antimicrobial activity, and adenovirus-mediated expression, and GFP quantification. The figures show means ± SEM. Unpaired Student’s t test, linear regression, ANOVA, and repeated measures ANOVA were performed when appropriate and as indicated in figure legends. P ≤ 0.05 was considered statistically significant.
SI Materials and Methods
Animals.
CFTR−/− (CF), CFTR+/− (heterozygote), and CFTR+/+ (WT) pigs were the product of CFTR+/− matings (28). CFTRΔF508/ΔF508 (CF), CFTR+/ΔF508 (heterozygote), and CFTR+/+ pigs were the product of CFTR+/ΔF508 matings (29). CFTR+/+, CFTR−/−, CFTR+/− newborn littermates and CFTRΔF508/ΔF508, CFTR+/ΔF508, CFTR+/+ newborn littermates were obtained from Exemplar Genetics, and airways were removed within 12–18 h of birth. We used male and female pigs. For euthanasia, pigs were sedated with i.v. propofol and then received i.v. euthasol.
Study Approval.
These studies were approved by the University of Iowa Animal Care and Use Committee.
Cultured Airway Epithelia.
Airway epithelial cells from the inferior turbinate were harvested, seeded onto collagen-coated, semipermeable membranes (0.33 cm2, 3470 polyester; Costar), and grown at an air–liquid interface as previously described (38, 42). Epithelia were generated with varying proportions of CF and WT cells. Epithelia were analyzed after they had differentiated and at least 14 d after seeding. We studied epithelia generated from inferior turbinate cells because they were readily available and the electrolyte transport characteristics are comparable between porcine nasal and tracheal/bronchial epithelia (49).
Western Blots.
Airway epithelial cells were scraped into 25 μL of lysis buffer (100 mM NaCl, 0.1 mM PMSF, 50 mM Tris⋅HCl, pH 7.4) and a mixture of protease inhibitors (7 µg/mL benzamidine-HCl, 1 µg/mL pepstatin A, 2 µg/mL aprotinin, 2 µg/mL leupeptin) plus 1% TX-100, rotated 15 min at 4 °C and centrifuged at 10,000 × g for 20 min to pellet cells. An equal volume of 2× sample buffer was added and incubated at 55 °C for 30 min with agitation. Samples were run on 7% Tris⋅HCl gels (Thermofisher) with high molecular mass standards (Life Technologies). Electrophoresed gels were transferred to PDF-FL (Millipore) overnight. Membranes were blocked in 0.01% casein buffer in PBS, immunostained with CFTR antibodies (1:1,000, antibody #596 and antibody #769; CFTR Antibody Distribution Program, Cystic Fibrosis Foundation Therapeutics) and secondary antibodies (Molecular Probes), and visualized and quantified on an Odyssey IR imager (LiCor).
Adenovirus-Mediated Expression.
An adenovirus 5 vector included a CMV promoter driving cDNA for CFTR linked to GFP at the N terminus prepared by the University of Iowa Viral Vector Core. Previous studies showed that CFTR localization and function were not altered by inclusion of the GFP tag (46). Various doses were applied to the basolateral surface of epithelia for 2 h. Experiments were conducted 3 d later.
GFP Quantification.
Epithelia were fixed with 4% paraformaldehyde, permeabilized with 0.2% Triton X-100, and stained with DAPI (Vectashield; Vector Laboratories). Z-sections through epithelia were imaged with an Olympus Fluoview FV1000 confocal microscope. GFP and DAPI (nuclei) positive cells were counted using ImageJ. The percentage of transduced cells was determined as follows: (number of GFP-positive cells)/(number of DAPI-positive cells).
Quantitative RT-PCR.
Primers were developed and validated using standard procedures. Total RNA was harvested using an RNeasy Lipid Tissue Mini Kit (Qiagen). cDNA was prepared using a High-Capacity cDNA Reverse Transcription Kit (Life Technologies) with random hexamers. Power SYBR Green PCR master mix (Life Technologies) was used for quantification. Fold-changes were determined using ΔΔCT values. The primers used were as follows: CFTR, CACCCAGCCATTTTTGGC; AGGAGCGATCCACACGAA; and RPL13a, GGCCCCTACCACTTCCG; ACTGCCTGGTACTTCCA.
PCR for CFTR.
Genomic DNA was harvested from cell cultures of the CFTRΔF508/ΔF508 and CFTR+/+ mixed in various ratios as per Qiagen Puregene Blood Core Kit A (catalog no. 158445). Then, 10 μg/μL DNA was amplified. Two primer sets were used. The forward primer for Neo amplified in CFTR exon 10 while the reverse primer was directed against the Neo cassette inserted in the intron between exon 10 and 11; this primer set detects ΔF508 CFTR. The CFTR primer set amplifies CFTR upstream of exon 10. PCR conditions were as follows: 2 min at 95 °C, 30 cycles of 95 °C for 20 s, 56 °C for 20 s, and 68 °C for 4 min, and then 68 °C for 5 min. PCR products were electrophoresed on a 1% agarose gel. Gels were imaged using UVP Bioimaging Systems Bioimager. Intensity was measured using ImageJ. To determine the percentage of normal CFTR (y axis of Fig. 1A), a ratio of the band intensity was calculated as follows: (CFTR − ΔF508 CFTR)/CFTR. These data are presented normalized with CF set at 0% and WT set at 100%. The primer sets used were as follows: Neo, AGAATTTCATTCTGCTCTCAGT; GAGGAAATTGCATCGCATTG; and CFTR, TTTCTCTTCTGCCTATTTCCC; AAGCCACAGAAGCATATGCAT.
Ussing Chamber Analysis of Electrolyte Transport.
Differentiated cultures of airway epithelia were mounted in modified Ussing chambers (Physiologic Instruments, Inc.). For measurements of transepithelial Cl− secretion, epithelia were bathed on both surfaces with a solution containing 135 mM NaCl, 2.4 mM K2HPO, 0.6 mM KH2PO4, 1.2 mM CaCl2, 1.2 mM MgCl2, 10 mM dextrose, and 5 mM Hepes (pH 7.4) at 37 °C and bubbled with air. For measurements of transepithelial HCO3− secretion, epithelia were bathed on both surfaces with a solution containing 115 mM Na isethionate (C2H5NaO4S), 25 mM NaHCO3, 3 mM Ca gluconate, 2.4 mM Mg gluconate, 2.4 mM K2HPO4, 0.6 mM KH2PO, and 10 mM dextrose and bubbled with 5% CO2 in air (pH 7.4). To calculate cAMP-induced changes in short-circuit current (Isc), we measured Isc in the presence of amiloride (100 µM) added apically and then after basolateral addition of 10 µM forskolin and 100 µM 3-isobutyl-2-methylxanthine (IBMX).
Fluorescence Recovery After Photobleaching and ASL Depth.
To assess ASL viscosity, we used methods similar to those we previously described (32). The apical surface was not washed for at least 2 wk before study. FITC-dextran (70 kDa; Sigma) was applied to the apical surface of epithelia as a dry powder 2 h before study. FRAP was assayed in a humidified chamber at 37 °C using a multiphoton confocal microscope (LSM 510 META; Zeiss). Fluorescence was bleached, and then images were acquired until maximal recovery was reached. Generally, at least six recovery curves from different locations in each culture were acquired and averaged to obtain data for one epithelia/donor. The time constant (τASL) was determined by regression analysis from fluorescence recovery curves. Data are expressed relative to the time constant of saline (τASL/τsaline).
For measurements of ASL depth, epithelia were covered with perfluorocarbon to prevent evaporation, and ASL depth was measured using a laser z-scanning confocal microscope (LSM 510 META; Zeiss) equipped with a 40× water-immersion objective. ASL depth was determined from z-image stacks. Generally, at least six sets of images from different locations (away from the meniscus) in each culture were acquired and averaged.
Assay of ASL Antimicrobial Activity.
To measure ASL antibacterial activity, we used Staphylococcus aureus-coated gold grids as previously described (33). Bacteria-coated gold grids were placed onto the apical surface of airway epithelia for 1 min. After removal, bacteria on the grids were assessed for viability using a Live/Dead BacLight Bacterial Viability assay (Invitrogen). Viability was determined in four to six fields to determine the percentages of dead bacteria.
Measurement of ASL pH.
ASL pH was measured using a ratiometric pH indicator, SNARF-conjugated dextran (Molecular Probes). SNARF powder was distributed onto the apical surface, 10 µM forskolin and 100 µM IBMX were added basolaterally, and ASL pH was measured 2 h later (32, 33). The basolateral solution was cell culture media containing 25 mM HCO3− with a 5% CO2 atmosphere. SNARF was excited at 488 nm, and emission was recorded at 580 nm and 640 nm using a Zeiss LSM 510 microscope. SNARF was dissolved in colorless pH standards to generate a standard curve to convert fluorescence ratios into pH using the same microscope settings.
Statistical Analysis.
The figures show means ± SEM. Unpaired Student’s t test, linear regression, ANOVA, and repeated measures ANOVA were performed when appropriate and as indicated in figure legends. A P ≤ 0.05 was considered statistically significant.
Acknowledgments
We thank the University of Iowa In Vitro Models and Cell Culture Core and the Viral Vector Core. This work was funded by National Institutes of Health Grants HL091842, HL51670, and HL11744; by the Cystic Fibrosis Foundation Research Development Program; and by the Roy J. Carver Charitable Trust. V.S.S. was supported by NIH Grants F30HL123239 and 5T32GM007337. M.J.W. is an Investigator of the Howard Hughes Medical Institute.
Footnotes
- ↵1To whom correspondence should be addressed. Email: michael-welsh{at}uiowa.edu.
Author contributions: V.S.S., S.E., X.X.T., P.H.K., C.P.P., L.S.O., and M.J.W. designed research; V.S.S., S.E., X.X.T., P.H.K., C.P.P., L.S.O., and M.J.W. performed research; V.S.S., S.E., X.X.T., P.H.K., C.P.P., L.S.O., and M.J.W. analyzed data; and V.S.S. and M.J.W. wrote the paper.
Reviewers: R.G.C., Weill Cornell Medical College; and J.M.W., University of Pennsylvania Perelman School of Medicine.
Conflict of interest statement: M.J.W. holds equity in Exemplar Genetics, which has licensed CF pigs from the University of Iowa.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1604905113/-/DCSupplemental.
Freely available online through the PNAS open access option.
References
Levels of CFTR expression and airway host defenses
Sign up for Article Alerts
Article Classifications
- Biological Sciences
- Medical Sciences
Jump to section
You May Also be Interested in
Adaptations in heart structure and function likely enabled endurance and survival in preindustrial humans.
Image courtesy of Pixabay/Skeeze.
Viscoelastic carrier fluids enhance retention of fire retardants on wildfire-prone vegetation.
Image courtesy of Jesse D. Acosta.
Water requirements may make desert bird declines more likely in a warming climate.
Image courtesy of Sean Peterson (photographer).
QnAs with NAS member and plant biologist Sheng Yang He
Image courtesy of Sheng Yang He.