PHYSIOLOGICAL REVIEWS   Vol. 79 No. 1 January 1999, pp. S145-S166
Copyright ©1999 by the American Physiological Society

CFTR Is a Conductance Regulator as well as a Chloride Channel

ERIK M. SCHWIEBERT, DALE J. BENOS, MARIE E. EGAN, M. JACKSON STUTTS, AND WILLIAM B. GUGGINO

Department of Physiology and Biophysics, Gregory Fleming James CF Research Center, University of Alabama at Birmingham, Birmingham, Alabama; Department of Pediatrics, Yale University School of Medicine, New Haven, Connecticut; Department of Medicine, UNC CF Research Group, University of North Carolina, Chapel Hill, North Carolina; and Department of Physiology and Division of Pediatrics, Department of Medicine, Johns Hopkins CF Research Group, Johns Hopkins University School of Medicine, Baltimore, Maryland

I. INTRODUCTION: CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR AS A MULTIFUNCTIONAL MOLECULE
II. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR-OUTWARDLY RECTIFYING CHLORIDE CHANNEL REGULATORY INTERACTION
    A. Importance of Cytoplasmic ATP
    B. Importance of ATP Release Mechanisms and Purinergic Receptors
    C. What Domains of CFTR Are Important for Regulation of ORCC?
    D. Integration of the CFTR-ORCC Regulatory Interaction Model: Identifying Targets for CF Pharmacotherapy
III. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR REGULATORY INTERACTIONS WITH OUTWARDLY RECTIFYING CHLORIDE CHANNELS IN PLANAR LIPID BILAYERS
    A. Experimental System
    B. CFTR and ORCC in Bilayers
    C. Importance of ATP and Direct Protein-Protein Coupling in CFTR-ORCC Interaction
IV. CAN CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR CONDUCT ADENOSINE 5'-TRIPHOSPHATE?
    A. Lessons From Airway Epithelial Cells
    B. Lessons From Other Epithelial Cell Preparations
    C. Lessons From Heterologous Mammalian Cells
    D. Lessons From the Oocyte Expression System
    E. Lessons From the Bilayer System
V. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR-EPITHELIAL SODIUM CHANNEL REGULATORY INTERACTION IN AIRWAY EPITHELIA AND HETEROLOGOUS CELLS
    A. Function of CFTR and ENaC in Airway Epithelial Ion Transport
    B. Heterologous Cell Models of Interaction Between CFTR and ENaC
    C. Potential Mechanisms of Interaction
VI. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR REGULATORY INTERACTIONS WITH EPITHELIAL SODIUM CHANNELS IN PLANAR LIPID BILAYERS
    A. CFTR and -rENaC in Bilayers
    B. Influence of Actin on CFTR-ENaC Interactions
VII. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR REGULATION OF RENAL OUTER MEDULLARY POTASSIUM CHANNEL SENSITIVITY TO GLIBENCLAMIDE: CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR AS A SULFONYLUREA RECEPTOR
    A. Sulfonylurea Receptors: Members of the ABC Transporter Family
    B. Sulfonylurea Sensitivity of Inwardly Rectifying K⁺ Channels
    C. CFTR Interactions with K⁺ Channels
    D. CFTR and ROMK
    E. Effects of Phosphorylation on CFTR-ROMK2 Interaction
    F. CFTR Domains Important to the CFTR-ROMK Interaction
VIII. SUMMARY AND CONCLUSIONS: CORRELATION OF REGULATORY FUNCTIONS WITH SEVERITY OF CYSTIC FIBROSIS LUNG DISEASE
IX. UPDATE: RECENT DEVELOPMENTS
REFERENCES

	ABSTRACT

Top References

Schwiebert, Erik M., Dale J. Benos, Marie E. Egan, M. Jackson Stutts, and William B. Guggino. CFTR Is a Conductance Regulator as well as a Chloride Channel. Physiol. Rev. 79, Suppl.: S145-S166, 1999.  Cystic fibrosis transmembrane conductance regulator (CFTR) is a member of the ATP-binding cassette (ABC) transporter gene family. Although CFTR has the structure of a transporter that transports substrates across the membrane in a nonconductive manner, CFTR also has the intrinsic ability to conduct Cl at much higher rates, a function unique to CFTR among this family of ABC transporters. Because Cl transport was shown to be lost in cystic fibrosis (CF) epithelia long before the cloning of the CF gene and CFTR, CFTR Cl channel function was considered to be paramount. Another equally valid perspective of CFTR, however, derives from its membership in a family of transporters that transports a multitude of different substances from chemotherapeutic drugs, to amino acids, to glutathione conjugates, to small peptides in a nonconductive manner. Moreover, at least two members of this ABC transporter family (mdr-1, SUR) can regulate other ion channels in the membrane. More simply, ABC transporters can regulate somehow the function of other cellular proteins or cellular functions. This review focuses on a plethora of studies showing that CFTR also regulates other ion channel proteins. It is the hope of the authors that the reader will take with him or her the message that CFTR is a conductance regulator as well as a Cl channel.

	I. INTRODUCTION: CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR AS A MULTIFUNCTIONAL MOLECULE

Top Next References

This review discusses published data, preliminary data, and hypotheses that support the view of the cystic fibrosis transmembrane conductance regulator (CFTR) as a bona fide "conductance regulator" in addition to itself playing a role as a "conductor" of Cl.

Before the identification and cloning of the cystic fibrosis (CF) gene and characterization of its protein product, CFTR, as a cAMP-regulated Cl channel (12, 39, 74, 103, 104, 125), a series of defects in cellular functions were described in CF epithelia, some of which could not be easily reconciled with "primary" mutations or abnormalities within a Cl channel protein (9, 10, 23, 25, 33, 40, 44, 61, 114, 120). These "secondary" defects included misregulation of a separate class of Cl channels, the outwardly rectifying Cl channels (ORCC), by cAMP-dependent protein kinase (PKA) and protein kinase C (PKC) in airway epithelial cells. Moreover, the ORCC Cl channel or a regulator of this Cl channel subtype was hypothesized originally to be the putative CF gene product before CFTR was identified (59, 81, 112). Another abnormality, apart from the defects in Cl channel cell biology proposed originally by Quinton (99) and identified by Boucher et al. (23), was hyperabsorption of Na⁺ across CF airway epithelia (23, 76) and hyperactivity of Na⁺ channels in isolated CF airway epithelial cells (33). This defect cannot be explained directly by mutations or abnormalities in a channel that conducts Cl. Moreover, misregulation of exocytosis and endocytosis by cAMP in CF cells (25) as well as lack of sufficient acidification of CF intracellular organelles has been documented (9, 10, 61). As a consequence, defective posttranslational modification and/or trafficking of any and all membrane glycoproteins may be abnormal in CF epithelial cells, leading possibly to "tertiary" complications within the CF cell.

These studies illustrate that CF disease may be caused by a complex combination of primary defects in CFTR Cl channel function (131); secondary defects in other CFTR-regulated ionic conductances such as ORCC (40, 44, 114), epithelial Na⁺ channels (ENaC) (23, 76), and renal outer medullary K⁺ channel (ROMK) or other inwardly rectifying K⁺ channels (89); or possible tertiary perturbations of protein modification or trafficking and defective protein processing. Finally, the concept of "modifier" CF genes has emerged to complicate interpretations of CF lung disease further and may also contribute to some secondary or tertiary defects in CF cells (106). Figure 1 illustrates the concept of a multifunctional CFTR, and Figure 2 introduces the concept of primary, secondary, and tertiary defects in a CF epithelium.

View larger version (32K): [in this window] [in a new window]   FIG. 1.   Multifunctional cystic fibrosis transmembrane conductance regulator (CFTR). Epithelial cell model summarizing multiple functions of CFTR: 1) Cl channel function, 2) facilitation of ATP release, 3) positive regulation of outwardly rectifying Cl channels (ORCC), 4) negative regulation of epithelial Na+ channels (ENaC), 5) regulation of vesicle trafficking, 6) regulation of intracellular compartment acidification and protein processing, and 7) modulation of ROMK sensitivity to sulfonylureas. Functions 2-4 and 7 are focused on in greater detail in additional model figures. ROMK, renal outer medullary potassium channel; TGN, trans-Golgi network; ER, endoplasmic reticulum.

View larger version (34K): [in this window] [in a new window]   FIG. 2.   Primary, secondary, and tertiary effects of CFTR dysfunction. Model of a segment of cystic fibrosis (CF) airway illustrating host of defects associated with CF. Impact of mutations with CFTR causing primary (1°) defects are subject of other reviews within this volume. Secondary (2°) and tertiary (3°) defects are also listed in this model figure and may result in a lack of CFTR function or lack of function of protein product of a modifier gene or genes that has yet to be found. Secondary and/or tertiary defect designations have been given to those defects where their relationship to CFTR function is less clear. CaCC, Ca2+-dependent Cl channels; ClC, voltage-dependent Cl channels; CNG, cyclic nucleotide-gated cation channels; AQP, aquaporins; PMN, polymorphonuclear cell or neutrophil; P.a., Pseudomonas aeruginosa; asialo-GM1, carbohydrate moieties with a lack of added sialic acid residues; APM, apical membrane domain; BLM, basolateral membrane domain.

Taken together, the multitude of functional defects in CF epithelial cells, illustrated in Figure 2, together with the characterization of CFTR as a Cl channel led to a key question in CF research: How does CFTR expression modify the regulation of separate populations of ion channels such as ORCC, ENaC, and ROMK and alter other cellular processes such as vesicle trafficking and Golgi acidification? Answers to this question have emerged from published and preliminary work summarized below. In actuality, the naming of CFTR as the cystic fibrosis transmembrane conductance regulator was done with much foresight (103, 104).

	II. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR-OUTWARDLY RECTIFYING CHLORIDE CHANNEL REGULATORY INTERACTION

Top Previous Next References

Early studies of epithelial Cl channels and CF showed that ORCC were present in CF epithelial cells but that these channels failed to respond to PKA and/or PKC (59, 81, 112). As such, an early hypothesis was that the CF gene product would be a regulator of ORCC (59, 81, 112). After identification of CFTR as a cAMP-regulated Cl channel distinct from ORCC, the validity of research concerning the ORCC and its defective regulation by protein kinases was questioned. Complementation studies, however, showed that reintroduction of wild-type CFTR into a CF cell corrected defective protein kinase regulation of ORCC (40). Moreover, patch-clamp recordings of airway epithelia derived from CF(/) mice demonstrated that ORCC were present but insensitive to PKA or PKC in cells devoid of CFTR (44). Taken together, these results suggested that kinase regulation of ORCC required CFTR.

To determine why CFTR expression was essential for PKA stimulation of ORCC, a convenient assay was needed to record simultaneously the activity of both CFTR and ORCC Cl channel populations. Previous whole cell recordings of normal airway epithelial cells showed that only CFTR Cl channel currents were stimulated in symmetrical Cl-containing solutions (bath and pipette solutions) with 1 mM Mg²⁺-ATP in the pipette (intracellular) solution (115). However, complete reconstitution of the physiological intracellular concentration of ATP with 5 mM Mg²⁺-ATP yielded cAMP-stimulated whole cell recordings in which both CFTR and ORCC Cl channel currents could be measured (115). The Cl channel blocking drug DIDS could be added subsequently to inhibit ORCC while not affecting CFTR channels (115). In excised or inside-out patch-clamp recordings, 1 mM Mg²⁺-ATP was added to the cytoplasmic face of excised membrane patches as a cofactor for PKA or PKC phosphorylation; at that time, however, it was not appreciated that intracellular ATP was important for CFTR modulation of ORCC activation by cAMP and PKA (40, 44, 59, 81, 112). Thus it was apparent that physiological concentrations of intracellular ATP were required for CFTR modulation of ORCC. The next challenge was to understand the role of intracellular Mg²⁺-ATP in CFTR-ORCC interactions.

A combination of single-channel and whole cell recordings and Ussing chamber short-circuit current recordings revealed that ATP had to cross the membrane from the cytoplasm to the extracellular space to exert its effects on ORCC and other Ca²⁺-dependent Cl conductances (60, 114). Addition of hexokinase or apyrase into the extracellular solution, agents which "scavenge" ATP or degrade ATP into its metabolites, prevented cAMP from stimulating ORCC without affecting cAMP stimulation of CFTR (114). In effect, ATP scavengers uncoupled CFTR from ORCC (114). This was a key observation, because it showed that cAMP-stimulated and CFTR-dependent release of ATP out of the cell was essential to CFTR regulatory interaction with ORCC. This finding has been confirmed by other laboratories using different experimental cell model systems (72, 114); moreover, these findings demonstrated that overexpression of CFTR not only correlated with a cAMP-regulated 9-pS Cl conductance but also a smaller 4- to 5-pS ATP conductance across the plasma membrane of a mammalian cell (114). This result served to integrate further this autocrine/paracrine ATP signaling between CFTR and ORCC. The simplest interpretation was that CFTR conducted the ATP itself; however, other possible mechanisms exist for how CFTR facilitated ATP conduction and/or release from cells (see sect. IV). Additional experiments independent of cAMP stimulation or CFTR expression revealed that extracellular UTP and ATP at nanomolar doses stimulated ORCC in both normal and CF airway epithelial cells (114). This result was consistent with the emerging model that the released ATP was regulating ORCC through a purinergic receptor (114). Also, the demonstrations that nanomolar and low micromolar doses of ATP were sufficient to stimulate ORCC was novel (60, 114). Moreover, this study corroborated previous evidence showing that a P_2u or P_2y2 G protein-coupled purinergic receptor stimulates Cl channels and Cl secretion across normal and CF airway epithelia (92, 93, 121, 122). Taken together, these results helped formulate a working model for a CFTR-ORCC regulatory interaction that required physiological concentrations of 5 mM Mg²⁺-ATP, functional CFTR, release of ATP as an agonist into the extracellular milieu, purinergic receptors that bind the ATP, and an ORCC that is regulated by ATP agonists (114). The different components of this model will be revisited below, and possible new players or stimuli that may also contribute to this model will be introduced. It is also very important to emphasize that there may be multiple mechanisms by which epithelial cells release their ATP (48), that there are multiple purinergic receptors expressed by a given epithelial cell (receptor expression may also differ between apical and basolateral domains) (1, 11, 42, 60), and that there may be multiple epithelial anion channels and, possibly, epithelial cation channels that are regulated by ATP via ATP receptors (45, 46, 60, 82, 93, 121, 122). There may also be multiple physiological roles for this CFTR-ATP autocrine signaling cascade. As we learn more information about the complexity of these and other systems, it is anticipated that novel targets for CF pharmacological therapy will emerge.

Severe truncation of the NH₂ terminus of CFTR by removal of amino acids 1 to 259 but with the Kozak consensus sequence surrounding methionine-265 intact resulted in the recording of a cAMP-activated 6-pS Cl channel (compared with the 9-pS wild-type Cl channel) and a cAMP-stimulated macroscopic Cl current that was 67% that of wild-type CFTR when expressed in Xenopus oocytes (96). When expressed in IB3-1 CF airway cells, this M265 construct functioned as a cAMP-stimulated Cl channel and conferred cAMP regulation on the population of endogenously expressed ORCC. In sharp contrast, a similar truncation to amino acid 259 but with methionine-265 altered to a valine (M265V) failed to function as a cAMP-stimulated Cl channel in both expression systems. However, despite this lack of Cl channel activity, M265V conferred cAMP regulation of ORCC. The results from these initial constructs suggested that the region of CFTR critical for Cl channel function lay in the amino acids immediately distal to methionine-265. This is consistent with the mutagenesis studies showing that alterations of residues in -helices 5 and 6 compromised single Cl channel conductance when compared with wild-type CFTR (32, 85, 117, 124). Similar to the results with M265V, insertion of "dual" arginine mutations (R334W with R347P) into CFTR eliminated Cl channel activity but conferred cAMP stimulation upon ORCC. These results also suggested that methionines alternative to methionine-1 could serve as "translation initiation" start sites for the CFTR protein (96).

Severe truncations of the COOH terminus also yielded fruitful and interesting results. Expression of a "half molecule" of CFTR (TNR-CFTR) in which amino acids 836 through 1480 were eliminated (TMD-1, NBD1, and R domains of CFTR intact) functioned in oocytes as a 9-pS single Cl channel similar to wild-type CFTR. At the macroscopic level, however, the Cl current from TNR-CFTR was only 10-20% that of wild-type current, suggesting that plasma membrane expression of TNR-CFTR was significantly less than wild-type CFTR. This result was confirmed in IB3-1 CF cells as well. Surprisingly, however, despite this low plasma membrane expression, TNR-CFTR was able to confer cAMP regulation upon ORCC in IB3-1 CF cells. Similar results were observed for a novel and naturally occurring splice variant in kidney medulla that creates a similar half molecule of CFTR (90). Thus further truncation was necessary to eliminate the portions of CFTR that were required for the regulatory interaction with ORCC. This goal was accomplished with a construct of CFTR in which only the TMD-1 domain of CFTR was left intact in the cDNA (amino acids 371-1480 removed). This construct functioned in part in oocytes and IB3-1 cells as a constitutively active Cl channel that did not require cAMP stimulation. Its activity, however, was potentiated by cAMP. This result was intriguing, because this domain of CFTR lacks consensus PKA phosphorylation sites (31). This result may be explained by the fact that cAMP could cause insertion of more TMD1-CFTR from subplasma membrane vesicles (25). Because net membrane conductance can be enhanced not only by activation of channels resident in the membrane but also by insertion of new channels into the membrane, this mutant may be useful in testing this aspect of CFTR cell biology. Most important for this work, however, TMD1-CFTR failed to confer cAMP stimulation of ORCC despite its profound intrinsic Cl channel activity.

Taken together, these preliminary results utilizing truncations of the CFTR cDNA from the NH₂ and COOH termini suggest that the domains critical for Cl channel function and ORCC regulatory coupling are distinct. One function could be eliminated with the other intact, and vice versa. The part of CFTR critical for Cl channel function lies at least in part in -helices 5 and 6 of TMD-1. In contrast, the domains of CFTR essential for interaction with ORCC are NBD1 and the R domain of CFTR. It is still not clear whether these domains are capable of transporting ATP themselves. NBD1 of CFTR does bind ATP, and there is evidence for ion channel activity intrinsic to NBD1 in bilayers (7) and mammalian cells (34) (see also below). Nevertheless, cAMP-stimulated ATP release and ORCC regulatory interaction are interconnected in CF airway epithelial cells. An alternative explanation is that NBD1 and the R domain interact with a separate ATP-specific channel or transporter or a population of ATP-filled vesicles that promotes efflux of the ATP.

With the use of insertion of mild versus severe lung disease-associated mutations, the NBD1 domain was implicated as a site important for ORCC regulation (43). A mild mutation within NBD1, A455E, caused a small reduction in Cl channel function but did not affect cAMP stimulation of ORCC. In contrast, G551D, a common and severe mutation in NBD1, caused a severe reduction in Cl channel activity and eliminated cAMP stimulation of ORCC. This result agrees with the results from truncation mutants described above and implicates NBD1 as an essential domain for regulatory interaction with ORCC. Again, the important question that is addressed below is whether NBD1 with the R domain is capable of transporting ATP itself or whether NBD1 together with the R domain interact with a separate molecule that conducts or transports ATP.

Figure 3 illustrates current hypotheses concerning the autocrine/paracrine model for CFTR-ORCC regulatory interaction. This model is an extension of the "working model" published in 1996 (114). The involvement of other regulatory molecules and, possibly, ATP-specific channels or transporters and/or ATP-filled vesicles is also shown.

View larger version (35K): [in this window] [in a new window]   FIG. 3.   Molecules important for CFTR-ORCC regulatory interaction. Illustration of hypotheses concerning additional cofactors, regulatory proteins, or stimuli that may be required to observe CFTR-ORCC regulatory interaction and CFTR-facilitated ATP release. CFTR could transport ATP itself, provided appropriate cofactors or regulators are present, or CFTR could modulate transport of ATP through a separate channel protein or could facilitate fusion of ATP-filled vesicles or both. PKA, cAMP-dependent protein kinase; PKC, protein kinase C; P2R, P2 purinergic receptor.

First, PKA plus ATP stimulation of CFTR is required for its Cl channel activity. Second, the mechanisms of ATP release and the role of CFTR and multidrug resistance protein (mdr) in facilitating ATP release are less clear. This issue is discussed in greater detail in section IV. Third, once released under basal or stimulated conditions, ATP is free as an agonist to regulate the ORCC either directly or via purinergic receptors. Initially, the P_2u receptor was implicated. A recent explosion in cloning of purinergic receptors reveals that at least three P_2y receptors, including the P_2u or P_2y2 receptor, bind UTP as well or better than ATP and could underlie the responses (1, 11, 42). Future studies are necessary to document the relative expression of different receptor isoforms. Perhaps ATP receptors in addition to P_2y(2y2) could also become targets for nucleotide-based therapies. Ideally, an ATP receptor that is shared by all exocrine epithelia affected by CF could be exploited as a therapeutic target for all CF. Finally, the signaling pathways whereby the ATP receptor regulates the ORCC are unknown. They may be membrane delimited and/or involve membrane-bound G proteins, phospholipases (phospholipases C, D, and A₂), phospholipids, and protein kinases, or they may be cytoplasmic and require cytosolic second messenger pathways like cyclic nucleotide pathways or PKC pathways (1, 11, 42).

The ORCC may not be the only epithelial Cl channel regulated by external nucleotides. The emerging family of ClC channels may be regulated in this manner; emerging preliminary work suggests that newer isoforms of this family (ClC-3, ClC-5) display outwardly rectifying current-voltage relationships (73, 109). Other ubiquitously expressed ClC isoforms, such as ClC-2, ClC-6, and ClC-7, may also be candidates. External nucleotides at micromolar concentrations trigger increases in intracellular Ca²⁺ (Ca²⁺_i) (46, 60, 108) and may, therefore, activate the recently identified CaCC family of Cl channels (37). "Maxi" Cl channels of high single-channel conductance (200-400 pS) are also expressed by a variety of exocrine epithelia and may be regulated by external nucleotides (116). One possibility is that different purinergic receptors may be coupled to different Cl or Na⁺ channels by utilizing different second messenger pathways in different epithelia to stimulate and/or inhibit different channel subtypes. This concept has precedence in the cardiac, neuronal, and brain ion channel literature (16, 17, 26, 35, 54, 79, 105).

	III. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR REGULATORY INTERACTIONS WITH OUTWARDLY RECTIFYING CHLORIDE CHANNELS IN PLANAR LIPID BILAYERS

Top Previous Next References

Planar lipid bilayers are a useful experimental technique to study regulatory interactions between membrane transport proteins. This system has been exploited to reconstitute and examine the interactions between CFTR and outwardly rectified anion channels (72) as well as the recently cloned ENaC (8, 15, 64, 65) (see sect. VI). The advantage of this system is that it provides a well-controlled model environment to study in a direct and physiologically meaningful way the functional properties of ion channels. With proper care, one can incorporate channels preferentially into the membrane with a specific and defined orientation, as well as control the number of channels present. Upon definition of the appropriate experimental conditions, the ease with which regulatory components can be added to the presumptive extracellular or intracellular side makes interpretation of these sorts of experiments unequivocal. This section describes recent studies utilizing this system to examine the regulatory relationships between CFTR and the ORCC.

Both CFTR and the ORCC were isolated and reconstituted simultaneously and functionally from vesicles derived from bovine tracheal epithelia (72). Remarkably, the intimate regulatory relationship between these two proteins is preserved throughout the purification procedure. Initial studies focused on the requirement of CFTR for PKA (with ATP as a cofactor) activation of the ORCC. Immunoprecipitation of CFTR from this preparation prevented PKA-induced activation of the ORCC, suggesting that the presence of CFTR is required for PKA plus ATP stimulation of the ORCC. Moreover, the hypothesis that CFTR was required to be in a "transport-capable" mode for PKA activation of the ORCC was tested. To this end, inhibitory anti-CFTR_505-511 antibodies were utilized to block CFTR transport function, or functional CFTR was replaced with a nonfunctional, mutated form, namely, G551D-CFTR (72). Only when CFTR was functional as a Cl channel could the ORCC be activated by "cytoplasmically" added PKA plus ATP.

As described above, there is also considerable debate about whether CFTR can function both as a halide and an ATP channel (2, 3, 51, 52, 60, 94, 97, 100, 102, 114). Guggino and co-workers (60, 114) proposed that the facilitation of ATP release or efflux by CFTR to the extracellular side of cells was a necessary prerequisite for PKA activation of the ORCC. The bilayer system was used subsequently to test if this same phenomenon could be observed. The addition of an ATP scavenging system [namely, hexokinase (0.5 units/ml) plus glucose (5 mM)] to the presumptive extracellular side of a bilayer containing both the ORCC channel and CFTR prevented PKA-mediated activation of the ORCC. In this experiment, no exogenous ATP was present in the "external" bathing compartment. Alternatively, in the absence of hexokinase, ATP could arrive at the extracellular surface only by passage through the bilayer membrane. Addition of ATP by itself to the extracellular side did not have any effect on the ORCC channel itself in the bilayer. If CFTR-facilitated ATP release was the only requirement for PKA-mediated activation of the ORCC, it should be possible to activate the ORCC by the addition of ATP to the extracellular solution and by application of PKA plus ATP to the cytoplasmic side of the system in the absence of CFTR. In these experiments, however, when ORCC was the only channel present in the bilayer membrane, the ORCC could not be activated by the addition of 100 µM ATP to the outside and PKA plus ATP to the inside or by the addition of PKA plus ATP to both sides of the membrane (72). Taken together, these results suggest that CFTR plays an additional required role in the regulation of the ORCC that may involve intramembrane or cytoplasmic interactions. This idea is supported further by the results of experiments in which the ORCC and the nonconducting G551D CFTR were incorporated simultaneously into the planar lipid bilayer. Under these conditions, the ORCC regains its PKA sensitivity but only in the presence of extracellular ATP. The next feature examined whether the NBD1 domain of CFTR could interact with the ORCC in the presence of extracellular ATP and be sufficient as a single domain of CFTR to mediate PKA activation of the ORCC. NBD1 alone was not able to replace full-length CFTR in this function (72). These results must be viewed with some caution because NBD1 cannot conduct ions (at least in our hands) and, as such, its presence in the bilayer membrane cannot be verified. Thus, on the basis of the findings with dysfunctional G551D-CFTR in the planar lipid bilayer, it was clear that the physical presence of CFTR was essential for PKA plus ATP activation of the ORCC. This result suggests that direct intramembrane or cytoplasmic protein-protein interaction may be involved in CFTR-ORCC "cross talk." Figure 4 is included to illustrate a working model of CFTR-ORCC regulatory interactions as deciphered by lipid bilayer studies.

View larger version (30K): [in this window] [in a new window]   FIG. 4.   CFTR-ORCC regulatory interaction in bilayer. Model of bilayer that summarizes observations concerning CFTR-ORCC interaction using bovine tracheal vesicle protein preparation. Evidence for indirect coupling via ATP export as well as direct protein-protein interaction has been accumulated. Asterisk next to ATP emphasizes a caveat to these bilayer studies that shows that purified CFTR in isolation fails to transport ATP. TMD, transmembrane domain; NBD, nucleotide-binding domain; R, R or regulatory domain.

	IV. CAN CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR CONDUCT ADENOSINE 5'-TRIPHOSPHATE?

Top Previous Next References

This question has stimulated vigorous debate in the CF community concerning the precise mechanism whereby CFTR facilitates the release of ATP out of cells. Several views exist concerning the role of CFTR and other ATP-binding cassette (ABC) transporters such as mdr in ATP release in particular and the physiological validity of ATP release from cells in general. Investigators in the laboratories of Cantiello and Ausiello (3, 97, 102) as well as those of Guidotti and Abraham (2, 3, 102) have published much data originally that support the simplest conclusion that CFTR and mdr conduct ATP. Investigators from other laboratories have also published reports (2, 3, 94, 97, 102, 114) or have generated preliminary data (see below) that show that cAMP-stimulated ATP release is dependent on CFTR, that expression of CFTR is associated with an ATP conductance, and that there are multiple interpretations of this ATP transport phenomenon. In these studies, the precise route of ATP through the membrane was not defined; however, ATP transport across the membrane occurred only in the presence of CFTR and mdr. There is an equally large number of laboratories that have published results showing that they could not demonstrate CFTR-dependent ATP conductance or release (51, 52, 80, 100, 101). Many laboratories and investigators have tackled this question and have debated over the various findings (2, 52, 101). One more recent study suggested that it was likely that these ATP transport phenomenon were observed because of mechanical artifacts during patch-clamp recordings or because of cell damage or lysis during ATP bioluminescence release assays (51). This section of the review summarizes these findings, outlines the current views surrounding this question, and attempts to integrate these findings and propose hypotheses concerning this issue to move this research forward.

Original studies by Guggino and co-workers (114) focused on normal, CF, and wild-type CFTR-complemented CF airway epithelial cells and established the CFTR-ORCC regulatory interaction model in which ATP linked CFTR to ORCC via an autocrine/paracrine mechanism involving ATP. From this work, a 4- to 5-pS ATP conductance was recorded in parallel with a 9-pS Cl conductance only from excised membrane patches of wild-type CFTR-complemented IB3-1 cells (specifically, clone S9) and stimulated with the catalytic subunit of PKA not from parental IB3-1 cells (114). This ATP conductance was discernible only in nominally Cl-free solutions in which Tris-Cl or NaCl was replaced completely with Tris-ATP or Na₂ATP (114). These conditions may be critical and may explain other results below.

Because of this work and because of the variability in findings between many laboratories, assays are being developed to study ATP release into the apical and basolateral medium from normal, CF, and wild-type CFTR-complemented epithelial cells grown as monolayers in air-fluid interface culture on collagen-coated permeable supports. Alternatively, heterologous cell cultures overexpressing CFTR or mdr either stably or transiently grown on collagen-coated 35-mm culture dishes can also be studied. An important aspect of these studies is that the cells are studied in real time inside the sealed chamber of a luminometer while in a serum-free medium designed to maintain cell viability during transfections for several hours. These studies are in progress, the assay is being perfected, and the data are preliminary. Cells are grown to confluence or as a monolayer tight to fluid, washed twice in phosphate-buffered saline, and placed in a sealed chamber of a luminometer in an OptiMEM serum-free medium containing luciferase-luciferin reagent (113). Any ATP molecule released from the cells will catalyze the luciferase-luciferin reaction and yield a photon of light collected by the luminometer in arbitrary light units (113). Non-CF epithelia and heterologous cells expressing or overexpressing CFTR and mdr have a significantly higher basal ATP release magnitude than their CF or null-expressing counterparts, observed without any stimulation by cAMP agonists, Ca²⁺ agonists, or other stimuli (Schwiebert, unpublished data). This difference in basal ATP release is a reproducible finding and has been observed in at least six pairs or panels of CFTR- or mdr-expressing cells versus their counterpart controls. Moreover, a hypothesis concerning the role of CFTR- and mdr-facilitated ATP release in cell volume regulation is being explored. Hypotonicity-induced ATP release is potentiated markedly by the expression of the ABC transporters, CFTR and mdr (Schwiebert, unpublished data). An involvement of CFTR and mdr in regulatory volume decrease during cell volume regulation has been proposed (47, 127, 128).

Grygorczyk and Hanrahan (51, 52) have used bioluminescence to address this issue as well. No differences were observed between CFTR-expressing cells and their counterparts in terms of basal release or cAMP-stimulated release. It was demonstrated, however, that mechanical perturbation (bending of coverslips held freely in a luminometer cuvette with forceps or injection of cell samples into the injection port of a luminometer) caused ATP release from cells. This was not surprising, since cell damage or cell lysis could occur; however, it was suggested that mechanical stimulation of cells may cause ATP release and may be important for epithelial cell biology (51). These methods are similar to those used by Cantiello and co-workers (97) and Abraham and co-workers (2, 3), but they are very different from others being employed as described above by the Schwiebert and Guggino laboratories and below with respect to ATP release from single Xenopus oocytes by the Engelhardt laboratory in which the integrity and viability of the cells is maintained by studying the cells in a serum-free medium or in their standard Ringer solution and in which the cells are placed inside the luminometer on the support on which they are grown normally and studied for a luminescence signal without perturbation.

Preparations of nonairway tissues such as sweat duct epithelia or Calu-3 submucosal gland serous cells failed to show CFTR-associated ATP single-channel or macroscopic conductance (100). No macroscopic ATP conductance was observed in isolated sweat duct preparations, whereas a robust Cl conductance was observed routinely. In single-channel or whole cell patch-clamp recordings of Calu-3 cells, no ATP single-channel or macroscopic conductance was observed, although Cl channels were observed routinely (100). One possible explanation for the differences between these studies and others may be the difference in the origin of the cells used. Airway surface epithelial cells were used exclusively for the studies by Guggino and co-workers, whereas cells derived from sweat gland/duct and airway submucosal gland were used by these investigators. These ATP release/transport mechanisms and their regulation may be very different depending on what epithelium is being studied.

Contradictory data concerning ATP conductance in heterologous mammalian cell systems including wild-type CFTR stably transfected Chinese hamster ovary (CHO) cells have been reported (53, 94). In a series of single-channel and whole cell experiments in which chemical gradients of Cl and ATP were manipulated on each side of the membrane, Cl conductance referable to CFTR was observed with a consistent lack of any measurable ATP conductance (53). In contrast, Pasyk and Foskett (94) were able to show that ATP and another adenine nucleotide, adenosine 3'-phosphate 5'-phosphosulfate, were conducted across the plasma membrane and the endoplasmic reticulum membrane only in CFTR-complemented CHO cells and not in parental CHO cells. As with the study of Guggino and co-workers (114), Pasyk and Foskett (94) were able to show ATP conductance at negative voltages and Cl conductance at positive voltages under bi-ionic conditions. However, these observations could not be confirmed by Wine and co-workers in Calu-3 cells (100) or Hanrahan and co-workers in these CHO cells (53).

This variability may be explained by the fact that additional regulators may need to be expressed in concert with CFTR and/or that additional cofactors may need to be present to transport the ATP and that these cofactors are regulated favorably by CFTR. These regulatory subunits or cofactors may be sensitive to factors within the serum added to the cell culture medium. Different order of solution changes, the concentration or ionic strength of anions in the solutions, the nature of the solution changes, or the osmotic strength of the solutions may also explain this variability.

Engelhardt and co-workers have developed a single Xenopus oocyte ATP bioluminescence release assay that studies the ability of CFTR to promote ATP release. Their preliminary results may shed light on many of the results discussed above. This cAMP-stimulated ATP release response in oocytes is variable and has many requirements (J. Engelhardt and Q. Jiang, unpublished data). First, only certain frogs yield oocytes that have cAMP-stimulated and CFTR-facilitated ATP release capability; however, if the frog yields oocytes in which the response is observed, it is observed consistently. Second, when the response is observed, it requires CFTR; uninjected or water-injected oocytes do not display the response. Wild-type and 259-M265V-CFTR promote cAMP-stimulated ATP release from oocytes, whereas TMD-1 CFTR fails to promote cAMP-stimulated ATP release. These truncated constructs of CFTR are described above (see sect. II). Finally, a strict protocol of solution changes must be followed to trigger this response in wild-type CFTR cRNA-injected oocytes. Oocytes must first be incubated in a 0 Cl solution for 5 min, followed by cAMP agonist stimulation in 0 Cl solution, with a subsequent change to a solution that contains >90 mM Cl with the continued presence of cAMP agonists. Upon the change to a Cl-containing solution under cAMP-stimulating conditions, a robust ATP release response is observed that is immediate and sustained. Taken together with results above, specifically with regard to the variability from frog to frog and from oocyte batch to oocyte batch, these results underscore the idea that additional regulators or cofactors must be expressed simultaneously with CFTR in the oocyte to observe ATP release (see Figs. 3 and 5).

View larger version (40K): [in this window] [in a new window]   FIG. 5.   How does ATP get released from cells in a CFTR-dependent manner? Three hypotheses are shown for how CFTR may facilitate ATP transport/release out of cell. For hypothesis A to occur based on an overview of data in literature, appropriate cofactors or regulatory proteins must be expressed in concert with CFTR. CFTR in isolation in bilayer cannot conduct ATP based on preliminary data of Ismailov and Benos and published results of Bear et al. (12) and of Gundersen and Kopito and co-workers (100). Hypotheses B and C suggest positive CFTR regulation of a separate and independent ATP transport mechanism and CFTR-facilitated fusion of ATP-filled vesicles, respectively.

With the use of immunopurified and functionally reconstituted CFTR from bovine tracheal epithelia and the nonconducting G551D-CFTR from transformed L cells, it has been shown that functional CFTR is required for the activation of the ORCC by PKA plus ATP unless ATP is supplied exogenously to the extracellular bathing solution (72). Cystic fibrosis transmembrane conductance regulator is, therefore, essential for the transport of ATP from the cytoplasm to the extracellular bathing solution where it can then interact with some external domain of the ORCC. It is unlikely, however, that CFTR itself expressed in isolation can function as an ATP channel. Results from other laboratories (K. Gunderson and R. R. Kopito; C. E. Bear) using different CFTR protein and bilayer preparations (80, 100) and preliminary data generated by Ismailov and Benos indicate that purified CFTR by itself cannot transport ATP in the planar lipid bilayer (I. I. Ismailov and D. J. Benos, unpublished data). Another possibility is that an ATP channel is present in some purified protein preparations and that CFTR facilitates the function of this ATP-conducting molecule. Indeed, preliminary work from Ismailov and Benos shows that CFTR reconstituted functionally within vesicles isolated from bovine tracheal epithelia, however, does result in appearance of an ATP-specific conductance (Ismailov and Benos, unpublished data).

The fact that immunopurified CFTR cannot conduct ATP (80, 100) but that CFTR within a vesicle that contains many other proteins including ORCC can conduct ATP suggests two possible explanations for this CFTR-dependent ATP conductance: 1) a separate channel protein conducts this ATP specifically or 2) CFTR requires additional "signals" to occur or "regulators" to be present to be "coaxed" into conducting this ATP itself. Some of these possible signals or regulators have been proposed in Figures 3 and 5.

In short, the precise molecular mechanisms that underlie CFTR-associated ATP release are becoming clearer but have not been defined fully. However, all of this published and preliminary work discussed above is of considerable value and has provided much information upon which to attempt to explain how CFTR and other ABC transporters facilitate ATP release. It is clear, however, that if CFTR can facilitate ATP release, CFTR must itself be modulated by regulators or stimuli in a concerted manner or CFTR itself must modulate an ATP release pathway or pathways that is expressed universally in epithelial cells, heterologous cells, and Xenopus oocytes.

	V. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR-EPITHELIAL SODIUM CHANNEL REGULATORY INTERACTION IN AIRWAY EPITHELIA AND HETEROLOGOUS CELLS

Top Previous Next References

Of equal or perhaps greater importance for CF airway pathophysiology, Na⁺ has been shown to be hyperabsorbed across CF airway epithelium when compared with its non-CF counterpart. It was postulated and subsequently proven by Stutts et al. (120) that CFTR dysfunction was responsible for this basal amiloride-sensitive hyperabsorption and for defects in cAMP regulation of amiloride-sensitive ENaC currents across CF airway epithelium. This section summarizes these results, provides new preliminary results, and hypothesizes about the possible mechanisms of CFTR-ENaC interactions. In contrast to the positive regulation of ORCC described in sections II and III, CFTR regulation of ENaC is negative or inhibitory in nature.

2. Abnormal regulation of Na⁺ absorption in CF airway epithelium

Two additional characteristics of CF airway epithelial ion transport were not directly interpretable in the context of CFTR's function as a cAMP-dependent Cl channel. First, the magnitude of basal Na⁺ absorption was two to three times greater than in normal airway epithelia, in studies carried out in vivo and in vitro in freshly excised nasal epithelium (

1. Decreased cAMP stimulation of ENaC in two coexpression systems

The precedent that CFTR regulates other ion channels was established by the positive cAMP regulation CFTR conveyed on the outward-rectifying Cl channel (

The molecular basis for negative regulation of ENaC by CFTR has not been identified. Several possibilities are depicted in Figure 6. First, although controversial, a role for CFTR in the export of intracellular ATP has been suggested (114). Adenosine 5'-triphosphate, released into the extracellular domain, could inhibit apical membrane Na⁺ channels, either through activation of phospholipase C and PKC linked to P_2y2 purinergic receptors or after metabolism of ATP by ectonucleotidases into adenosine and subsequent inhibition of adenylyl cyclase linked to A₁ adenosine receptors (84). Theoretically, CFTR could accomplish the transport of yet unidentified substrates that affect Na⁺ channels. Members of the related mdr transporters move phospholipids across membranes (13, 50, 129); thus phospholipid messengers may play a role in modulation of ENaC by CFTR. A second possibility is that CFTR interacts directly with one or more ENaC subunits, as suggested by the studies in planar lipid bilayers (see sect. VI). However, this possibility may be difficult to reconcile with the observation that the functional regulation of ENaC by CFTR seen in the airways is not seen in every tissue where CFTR and ENaC appear to be expressed in concert, such as the sweat duct epithelium. However, the recent correlative studies by Kunzelmann et al. (78) underscore the idea that direct-direct protein interaction may be likely. The third possibility is that CFTR and ENaC interact via elements of the membrane-associated cytoskeleton and that this indirect interaction is involved somehow in the ability of CFTR to interfere with the regulation of ENaC by PKA. Although the details of this mechanism have not been established fully, both CFTR (30) and ENaC (15) have been shown to be affected by interactions with cytoskeletal components, namely, actin. Variable expression of a key regulatory protein or proteins linking CFTR or ENaC to the cytoskeleton or to each other in different tissues could explain tissue-specific regulation of ENaC by CFTR.

View larger version (42K): [in this window] [in a new window]   FIG. 6.   Hypotheses for CFTR-ENaC regulatory interaction. Membrane model of CFTR-ENaC interaction in which autocrine/paracrine signaling facilitated by CFTR via ATP or another substrate, direct/indirect linkage in membrane between CFTR and ENaC, and CFTR modulation of insertion/retrieval of vesicles containing ENaC are illustrated. G , unknown heterotrimeric G protein.

	VI. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR REGULATORY INTERACTIONS WITH EPITHELIAL SODIUM CHANNELS IN PLANAR LIPID BILAYERS

Top Previous Next References

Although fluid and electrolyte secretion is inhibited in CF patients because of CFTR transport dysfunction, fluid absoprtion across the airway is enhanced because of a simultaneous increase in the rate of Na⁺ transport (24, 76). These transport events result in the accumulation of mucus, pulmonary congestion, and excessive dryness of the epithelial surface. The increased Na⁺ absorption observed in CF patients has been attributed to a two- to threefold greater activity of the amiloride-sensitive Na⁺ entry pathway when compared with epithelia derived from non-CF individuals (24, 76). Currently, amiloride inhalation is being pursued as a possible therapy for CF lung disease (14, 70, 75). As indicated earlier, the molecular pathogenesis of the aforementioned events lies in the ability of the CFTR molecule itself to inhibit tonically the activity of amiloride-sensitive Na⁺ channels. The results of Stutts et al. (120) support the hypothesis that the physical presence of functionally competent CFTR is necessary for the downregulation of ENaC channels.

Several other issues are important regarding the interaction of -rENaC and CFTR. First, the -rENaC subunits of ENaC are expressed in unequal proportions in airway epithelia (27). The -subunit is sufficient to produce functional amiloride-sensitive Na⁺ channels, although the presence of - and -subunits enhances significantly amiloride-sensitive macroscopic currents in oocytes as well as changes substantively the gating pattern of -rENaC (29, 64, 65). The importance of ENaC - and -subunits for channel function and their coupling to regulatory processes is unknown, although varying the ratio of -, -, and -subunits did not produce a variety of channel intermediates in bilayers (64, 65). Second, although the message for -rENaC was found in the lung (27), this channel exhibits a single-channel conductance, ion selectivity, gating kinetics, and amiloride-sensitivity profile similar to a low-conductance, highly selective amiloride-sensitive Na⁺ channel found in native Na⁺-reabsorbing epithelia other than in the airways (91). This is intriguing because the Na⁺ channels identified thus far in airway epithelia display quite variable biophysical characteristics (33, 71, 87, 134). Single-channel conductances vary from 4 to 28 pS, their Na⁺-to-K⁺ permselectivity ratios vary from 1 to >10, and the affinity for amiloride varies from an inhibitory constant of 0.9 to 8 µM. Moreover, the ethylisopropyl analog of amiloride (ethylisopropylamiloride) can inhibit pneumocyte Na⁺ channels better than amiloride itself (87).

Wild-type CFTR was found to decrease the P_o of -rENaC in planar lipid bilayers without affecting the unitary conductance of the ENaC channels (64). However, this reduction of P_o was only moderate (~20% reduction). Other preliminary experiments revealed that CFTR had no effect on the P_o of this channel formed by the -subunit alone in contrast to inhibition of all three subunits expressed together. These preliminary results suggest that the - and/or -subunit is important for CFTR to exert its regulatory influence on ENaC. However, this tentative conclusion is paradoxical in light of the recently published report of Kunzelmann et al. (78), who based on correlative yeast two-hybrid analysis and patch-clamp analysis concluded that CFTR interacts with the -subunit of ENaC via cytoplasmic domains.

The cytoskeletal protein actin interacts with -rENaC and affects dramatically the biophysical properties of the channel. Actin decreases the single-channel conductance, confers sensitivity of the channel to PKA, and changes the gating of the channel such that the mean open and closed times become more long lived (from millisecond to second time constants). Because of this effect of actin on ENaC single-channel properties, the hypothesis that CFTR may exert an enhanced regulatory influence on -rENaC in the presence of actin was tested. Moreover, we wanted to examine the influence of CFTR on the stimulatory effect of PKA on the activity of -rENaC. These results showed that immunopurified bovine tracheal epithelial CFTR coreconstituted into a bilayer with -rENaC lowered the single-channel P_o of -rENaC in the presence of actin by >60% (compared with only 20% in the absence of actin). In the presence of actin, PKA plus ATP or ATP alone activated ENaC both in the presence and in the absence of CFTR in a transient manner, with peak activation occurring ~40 min after PKA plus ATP addition. The presence of CFTR, however, attenuated greatly the activation of ENaC by PKA plus ATP. We also found that actin, but not CFTR, could interact directly with the channel-forming -subunit of -rENaC. As indicated above, CFTR did not affect any properties of the channels formed by the -subunit of rENaC alone but did affect the channels only if either or both the - and -subunits were present. Taken together, our results suggest that actin enhances the ability of CFTR to downregulate ENaC epithelial Na⁺ channels and that the interactions between -rENaC, CFTR, and actin involve different subunits of this cloned epithelial Na⁺ channel. Figure 7 illustrates a working model of CFTR-ENaC cross talk as elucidated by lipid bilayer recordings.

View larger version (34K): [in this window] [in a new window]   FIG. 7.   CFTR-ENaC regulatory interactions in bilayer. Membrane model summary of observations concerning CFTR-ENaC interaction in bilayer and impact of protein kinase A (PKA) and short actin filaments on this interaction.

	VII. CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR REGULATION OF RENAL OUTER MEDULLARY POTASSIUM CHANNEL SENSITIVITY TO GLIBENCLAMIDE: CYSTIC FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR AS A SULFONYLUREA RECEPTOR

Top Previous Next References

Not only may CFTR regulate other ion channels directly or indirectly via several protein-protein or signaling interactions, but CFTR may also be capable of conferring effects of agonists or antagonists on other proteins. These functions of CFTR derive from analogies to the functions of other ABC transporters such as mdr, MRP, SUR, HisP, the TAP, and yeast STE6 (55, 56). These transporters facilitate the transmembrane movement of a large variety of compounds such as anionic and cationic hydrophobic drugs, peptides, phospholipids, and, in some cases, large proteins (55, 56). In part, the concept that CFTR could be a regulator of many other cell functions derived from its membership in the ABC transporter family. Moreover, because CFTR regulates two transmembrane-spanning motif ion channels such as ENaC and ROMK (see sect. VII), it is possible that there is a common motif that CFTR recognizes and modulates in these channel subtypes. Because CFTR also regulates ORCC, it is possible that this channel may be an anion channel with a similar topological motif.

Sulfonylureas are a family of compounds that inhibit ATP-sensitive K⁺ channel current (I_KATP). The I_KATP channels are found in a number of tissues including the brain, heart, smooth muscle, kidney, and pancreatic -cells (41, 98). They are a subgroup of the inwardly rectifying K⁺ channels (IRK) channels. A defining characteristic of I_KATP channels is their inhibition by ATP or other nucleotides (41, 98). In native tissue, these channels are inhibited by sulfonylureas. It was believed originally that sulfonylureas were specific inhibitors of I_KATP channels. However, it is clear that the sulfonylurea glibenclamide can also inhibit CFTR channel activity (118), suggesting a link between this Cl channel and the family of ATP-sensitive K⁺ channels.

The I_KATP channels are believed to provide a link between the metabolic status of the cell and membrane potential. In the -cell of the pancreas, modulation of I_KATP channels controls insulin release. Clinically, sulfonylureas, such as glibenclamide and tolbutamide, are used to treat diabetes mellitus. They act by inhibiting I_KATP channels in pancreatic -cells, thereby leading to cell depolarization. The depolarization in turn activates L-type Ca²⁺ channels, and the resulting Ca²⁺ influx triggers exocytosis of insulin. It was shown that sulfonylurea sensitivity of the pancreatic I_KATP (Kir6.2) (62, 95) is conferred by SUR, a high-affinity sulfonylurea binding protein and a newly identified member of the ABC superfamily (4).

The binding protein SUR is a 140-kDa molecule that has pharmacological characteristics of a high-affinity receptor. Although SUR possesses a number of structural motifs that are characteristic of ABC transporters, such as nucleotide binding folds as well as transmembrane-spanning dom