CFTR–SLC26 transporter interactions in epithelia

Abstract

Transport mechanisms that mediate the movements of anions must be coordinated tightly in order to respond appropriately to physiological stimuli. This process is of paramount importance in the function of diverse epithelial tissues of the body, such as, for example, the exocrine pancreatic duct and the airway epithelia. Disruption of any of the finely tuned components underlying the transport of anions such as Cl⁻, HCO₃ ⁻, SCN⁻, and I⁻ may contribute to a plethora of disease conditions. In many anion-secreting epithelia, the interactions between the cystic fibrosis transmembrane conductance regulator (CFTR) and solute carrier family 26 (SLC26) transporters determine the final exit of anions across the apical membrane and into the luminal compartment. The molecular identification of CFTR and many SLC26 members has enabled the acquisition of progressively more detailed structural information about these transport molecules. Studies employing a vast array of increasingly sophisticated approaches have culminated in a current working model which places these key players within an interactive complex, thereby setting the stage for future work.

Introduction

Epithelial tissues mediate the directional, or vectorial, movements of ions, solutes and water, all processes that rely on the polarized distribution of specialized transport proteins, operationally classified as either channels or transporters. Differential expression and localization of these key molecules determine whether an epithelium absorbs, secretes or transports bidirectionally. In addition to their polarized distribution, the intrinsic structural and functional integrity of transport proteins is essential for ensuring that a given epithelium produces the appropriate and sufficient fluxes.

Studies lending insights into such basic physiological processes often arise from the need to understand disease states. Disease mechanisms effectively provide object lessons that form a fundamental basis of molecular medicine. One well-studied example of this concept in action is cystic fibrosis (CF), an inherited disease that results from the mutation of the cystic fibrosis transmembrane conductance regulator gene (CFTR; Riordan et al. 1989). CFTR encodes a cAMP-regulated Cl⁻ channel that is essential for the function of diverse systems. Over 1,500 mutations in CFTR (http://www.genet.sickkids.on.ca/cftr/) result in either the complete loss of the gene product or, alternatively, defective CFTR which manifest distinct classes of functional impairment. These include defects in folding/stability/trafficking, regulation/gating and conduction and reduced synthesis levels, manifesting in a wide spectrum of disease severity (Choo-Kang and Zeitlin 2000). The availability of structural information allows the investigator to scrutinize, by deploying in silico modeling and simulations, the consequences of mutations of each class, thereby enabling insights into each disrupted process—folding, gating, conduction and synthesis—at atomic resolution.

Strategies utilizing correlations between CFTR genotype and CF phenotype provide a window into understanding molecular structure and function. In the case of CFTR and CF, this approach can furnish a means of combating other diseases in which wild-type CFTR function becomes disrupted by pathogens, as in the case of secretory diarrhea. In this case, cholera toxin stimulates Cl⁻ secretion via CFTR, suggesting the potential usefulness of CFTR inhibitors in the treatment for cholera. Detailed structures, together with functional models, provide critical tools that facilitate the design of potent inhibitory modulators, possibly also by aiding refinement of compounds identified by high-throughput screening (Ma et al. 2002; Muanprasat et al. 2004).

The collective knowledge of CFTR function in health and disease now includes important considerations of mechanisms involving intermolecular interactions. Thus, CFTR regulation of transporters residing apically in epithelial cells also bears physiological and clinical relevance. Solute carrier family 26 (SLC26) family members couple the exchange of Cl⁻ with the transport of a host of physiologically important anions, including HCO₃ ⁻, I⁻ and SCN⁻ . 1 CFTR (and other Cl⁻ channels) may furnish the Cl⁻ required for SLC26 transporter activity. Importantly, the disruption of genes encoding SLC26 transporters produces a spectrum of unique and important diseases which, in their own right, profoundly affect the human condition. Thus, elucidation of the molecular details of functional coupling between SLC26 transporters and CFTR likely will enable the identification of potential treatment strategies for conditions resulting from bona fide SLC26 disruption as well.

The present review aims to summarize current knowledge on how CFTR and SLC26 transporters function in a concerted manner in epithelial HCO₃ ⁻ secretion via intermolecular domain interactions. Substantial background information discussing this process is provided to lend the physiological context. Included is a brief description of post-synaptic density 95/discs large/zona occludens 1 (PDZ) domains, which are functional protein–protein interaction motifs found in many adaptor molecules. Such molecules act to scaffold both CFTR and SLC26, thereby promoting intermolecular interactions. The structural information available for the relevant and critical functional domains in both CFTR and SLC26 is reviewed. Although knowledge to date indicates that many challenges must be overcome, this also provides ample opportunities to decipher intermolecular interactions that are critical to human health and disease.

CFTR and SLC26 transporters critically determine anion secretion by diverse epithelia

The need for CFTR in HCO₃⁻ secretion

This important function is best illustrated by first considering transport in the pancreas. HCO₃ ⁻ buffers many epithelial secretions and nowhere is this more evident than in the ducts of the exocrine pancreas, where the concentration of HCO₃ ⁻ can achieve levels as high as 140 mM (Steward et al. 2005) and the alkalinity of secreted fluid can exceed pH 8.0. Such high pH and HCO₃ ⁻ concentrations are essential for mucin solubilization as well as controlled activation of zymogens released by the pancreatic acini (Palade 1975). In CF, impaired HCO₃ ⁻ secretion results in mucin precipitation and blockage of the ductal tree (Choi et al. 2001). Moreover, the resultant increased acidity promotes premature zymogen activation. The subsequent, progressive cycles of destruction lead ultimately to tissue fibrosis (Durie and Forstner 1989). But how does loss of a Cl⁻ channel, CFTR, lead to defective HCO₃ ⁻ secretion?

Parallel universes—CFTR mediates anion/HCO₃^- transport by diverse epithelia

The answer to this question applies not only to pancreas, but also to many other CFTR-expressing, anion/HCO₃ ⁻ -transporting epithelia. Of note are the salivary ducts (Shcheynikov et al. 2008), epithelia lining both the male (Carlin et al. 2006; Pierucci-Alves et al. 2011) and female (Muchekehu and Quinton 2010; Wang et al. 2003) reproductive tracts and the airway submucosal glands (Ballard and Inglis 2004). One can regard the complex mixture of mucins and anti-microbial peptides in airway submucosal gland secretions as analogous to the mucins and zymogens comprising pancreatic gland secretions. The two are, essentially, parallel universes. In these tissues, as in the pancreas, compromised CFTR function associates with impaired HCO₃ ⁻ secretion—with tissue-specific consequences.

How? Two possibilities immediately come to mind (Fig. 1). First, single channel studies of CFTR-overexpressing mammalian cell lines demonstrate permeation of other anions, including HCO₃ ⁻ (Linsdell et al. 1997; Poulsen et al. 1994). These early findings suggested that HCO₃ ⁻ secretion can result from direct efflux through CFTR; importantly, this has been measured in intact, native guinea pig pancreatic duct preparations (Ishiguro et al. 2009). Alternatively, HCO₃ ⁻ might exit into the lumen via distinct transport molecules that are coupled functionally to cAMP-activated (i.e. CFTR-mediated) Cl⁻ efflux (Garnett et al. 2011; Gray et al. 1989; Steward et al. 2005). In addition, both modes may co-exist in a given epithelium.

Fig. 1.

Possible means of transepithelial HCO₃ ⁻ secretion. Top Schema outlining cystic fibrosis transmembrane conductance regulator (CFTR)-dependent transport of HCO₃ ⁻ by either direct or indirect pathways. For simplicity, basolateral Cl⁻ and HCO₃ ⁻ uptake are shown with a single generic co-transport graphic element, but it should be noted that the Na⁺/K⁺/2Cl⁻and Na⁺/HCO₃ ⁻ cotransporters are distinct molecular entities. Not shown is basolateral K⁺ conductance that enables the recycling of K⁺ taken up by Na⁺/K⁺ ATPase and the Na⁺/K⁺/2Cl⁻ cotransporter. Left HCO₃ ⁻ enters the cell basolaterally by Na⁺-coupled cotransport and exits into the lumen directly through CFTR. Anions transported by solute carrier family 26 (SLC26) transporters are indicated by A ⁻. Other ions are indicated by standard nomenclature. Right Alternatively, Cl⁻ secretion mediated by CFTR provides a counterion for HCO₃ ⁻ exit via SLC26-mediated exchange pathways. Dotted arrow indicates functional interactions between CFTR and SLC26 transporters. Bottom In the absence of CFTR, both the direct (left) and the indirect (right) pathways are disabled

SLC26 transporters have diverse functional modes

Important candidates for CFTR regulation are members of the solute carrier family, SLC26, comprising 11 members having low amino acid sequence identity, cytoplasmically localized amino and carboxyl termini, and between 8 and 14 transmembrane domains. Mammalian SLC26 proteins bear homology to prokaryotic SO₄ ⁻ transporters (SulP). Collectively, they can be referred to as SLC26/SulP proteins. Many SLC26A/SulP family members transport anions (Dorwart et al. 2008b; Mount and Romero 2004). Of the mammalian members, SLC26A1 and -A2 transport SO₄ ⁻², whereas SLC26A3, -A4, -A6 exchange Cl⁻ for a wide range of anions, including HCO₃ ⁻. Interestingly, SLC26A3 and -A6 can also function as conductances (Shcheynikov et al. 2006). Recent mutagenesis studies based on predictions of in silico modeling localized a single glutamate residue critical to switching between their different functional modes: E367 and E357 in SLC26A3 and SLC26A6, respectively (Ohana et al. 2011). Initial expression studies showed that SLC26A7 and SLC26A9 behave as anion channels (Dorwart et al. 2007; Kim et al. 2005). However, SLC26A9 can function in anion exchanger mode and furthermore can transport Na⁺ (Chang et al. 2009). To date, the protein products of SLC26A8 and SLC26A11 have poorly understood transport capabilities, and SLC26A10 is a pseudogene. SLC26A5, which encodes a motor protein in the outer hair cells of the cochlea, also is capable of electrogenic SCN⁻ transport (Schaenzler and Fahlke 2011).

SLC26 transporters are relevant to human disease

Mutations in SLC26A2, SLC26A3 and SLC26A4 result in human diseases, namely diastrophic dysplasia (DTDST2) (Hastbacka et al., 1994; Superti-Furga et al. 1996), congenital chloride-losing diarrhea (CLD) (Holmberg 1986) and Pendred syndrome (Everett et al. 1997; Pendred 1896), respectively. Both CLD and Pendred syndrome highlight the roles of SLC26 transporters in epithelial function. The loss of SLC26A3 function causes CLD by abolishing the major pathway for colonic Cl⁻ reabsorption (Melvin et al. 1999). Pendred syndrome is associated with mutations in SLC26A4. It manifests primarily as profound sensorineural deafness which results from disruption of cochlear endolymphatic fluid HCO₃ ⁻ buffering, loss of the endocochlear potential and subsequent oxidative stress (Jabba et al. 2006; Singh and Wangemann 2008). Pendred syndrome also occasionally presents with euthyroid goiter, and indeed SLC26A4 normally localizes to the thyroid. This, and its ability to exchange I⁻ for Cl⁻, led to the notion that SLC26A4 comprises the long-sought-after apical I⁻ pathway in the thyroid. This conclusion is unsatisfactory on several levels (discussed in (Fong 2011; van den Hove et al. 2006)). Alternative conduits for I⁻ exit therefore must exist and, pertinent to the present discussion, CFTR is a candidate (Devuyst et al. 1997; Fong 2011; Li et al. 2010; Li et al. 2012). As for HCO₃ ⁻, CFTR may directly carry I⁻ or indirectly regulate its passage via SLC26A4 in the thyroid.

Domains in CFTR and SLC26 transporters interact to regulate their function

An early study provided evidence for the functional interaction between CFTR and select SLC26 transporters (SLC26A3, -A4 and -A6) (Ko et al. 2002). Using cell systems heterologously expressing these proteins, Ko et al. measured membrane potential, current and ΔpH, an obligatory consequence of SLC26 activity, in response to CFTR stimulation. Full responses were observed for cells expressing the wild type—but not mutants—of either the CFTR or SLC26 transporters. Later work from the same group (Ko et al. 2004) localized the relevant interacting regions to the R domain of CFTR and the carboxyl terminus [sulfate transporter and anti-sigma factor antagonist (STAS) domain] of SLC26 transporters (Aravind and Koonin 2000). Interestingly, disease-associated mutations in SLC26 transporters can localize to the STAS domain, as will be discussed.

CFTR and SLC26 transporters interact with scaffolding proteins

Specific carboxyl terminus residues in SLC26A3 and -A6 (-TKF and -TRL, respectively) interact with the PDZ domain protein, sodium–hydrogen exchange regulatory factor 1 (NHE-RF1) (Lamprecht et al. 2002; Lohi et al. 2003). PDZ domains mediate protein–protein interactions. Available structures show that the roughly 80–90 amino acids comprising PDZ domains assume an anti-parallel organization of six β strands (βA –βF), together with two intervening α helices (αA and αB). Figure 2a (top) summarizes the order of these secondary elements and some of the functionally pertinent residues, particularly those intervening between βA and βB and forming a carboxyl terminal-binding, connecting loop. A high-resolution structure (1.8 Å) of the PDZ-3 of postsynaptic density protein-95 (PSD-95) (Fig. 2a, bottom) shows the αB helix and the βB strand cradling its peptide binding partner. This interaction is promoted via main-chain interactions of the peptide with βB (Doyle et al. 1996). PDZ binding domains can be classified according to their preference for specific peptide sequences. Class I domains recognize a consensus motif containing T/S-X-ϕ, where X indicates any amino acid and ϕ denotes a hydrophobic residue (Sheng and Sala 2001; Songyang et al. 1997).

Fig. 2.

a. Depictions of post-synaptic density 95/discs large/zona occludens 1 (PDZ) domain organization and structure. Top Linear cartoon representation showing the predicted secondary structural elements above the relevant sequence (amino acid residues 309–393) from an archetypal PDZ domain, rat PSD-95 PDZ-3. Blue arrows β sheets, purple rectangles α helices. Residues shown in boldface contact the peptide binding partner. Note that the binding loop rests in the region intervening between βA and βB. Bottom Three-dimensional structure of the PDZ-3 of postsynaptic density protein-95 (PSD-95) showing the interacting peptide (orange arrow) wedged between the groove formed by the βB sheet and the αB helix; the carboxylate binding loop hovers above. This image is modified from Doyle et al. (1996), with permission from Elsevier. b Cartoon showing hypothesized apical macromolecular signaling complexes promoted by PDZ protein-facilitated interactions between the CFTR and SLC26 transporters. PDZ domain binding motif of SLC26A3, TKF, is shown in this figure; the counterpart for SLC26A6 is TRL. For CFTR, the motif is TRL as shown. This motif is lacking in SLC26A4, but the potential for regulation by sulfate transporter and anti-sigma factor antagonist (STAS) domain interactions with the R domain of CFTR is preserved

The carboxyl terminus tri-peptide of CFTR is identical to that of SLC26A6 and also binds PDZ proteins (Sheng and Sala 2001; Wang et al. 1998). Particularly important in CFTR function are sodium–hydrogen exchange regulatory factor 1, NHE-RF1 (or ezrin binding protein 50, EBP-50) and NHE-RF2 (exchanger type 3 kinase A regulatory protein, E3KARP) (Hall et al. 1998; Sun et al. 2000; Wang et al. 1998), both of which contain two PDZ interaction domains. Thus, PDZ proteins link CFTR and SLC26A3 (or -A6) together, behaving as associated adaptor molecules that thereby can promote interactions between the R and STAS domains (Ko et al. 2004) (Fig. 2b).

Although interactions with diverse PDZ proteins have been shown to regulate CFTR gating via phosphorylation, membrane trafficking, stability and turnover, an exhaustive treatment of these aspects is beyond the scope of the present discussion. Several excellent reviews highlight the scaffolding role of PDZ proteins in CFTR function (Li and Naren 2005; Seidler et al. 2009a, b). PDZ interactions also mediate associations between the lysophosphatidic acid receptor 2 (LPA2) and NHE-RF2. The binding of lysophosphatidic acid thereby can inhibit CFTR phosphorylation (Li et al. 2005; Zhang et al. 2011), as reviewed elegantly by Eckford and Bear (2011). Small-molecule disruption of the interactions of LPA2 with NHE-RF2 therefore enhances CFTR activity (Zhang et al. 2011).

It should be noted that PDZ associations regulating similar aspects of SLC26 transporter cell biology are not as extensively studied as those of CFTR. Such questions therefore represent potentially interesting and important areas for further investigations.

The R domain: a key structural element that is essential to the regulated function of CFTR

Cystic fibrosis transmembrane conductance regulator comprises 1,480 amino acids organized in functional domains typical of ATP-binding cassette (ABC) family proteins. Briefly, these include two membrane-spanning domains (MSD1, MSD2), each made up of six membrane-spanning α-helices forming the pore for Cl⁻ passage. Each is followed by one of two cytosolic, nucleotide binding folds, NBD1 and NBD2, respectively. Unique to CFTR is a regulatory R) domain, a region that intervenes between NBD1 and MSD2. The R domain contains multiple sites for phosphorylation by protein kinase A (PKA) as well as a number of other kinases, notably adenosine monophosphate kinase (AMPK) and protein kinase C (PKC). Inhibitory stretches are also contained in the R domain (Baldursson et al. 2001; Rich et al. 1993; Wilkinson et al. 1997; Xie et al. 2002). Phosphorylation of the R domain relieves inhibition and allows binding of two ATP molecules between the NBDs, vastly increasing the probability of channel opening (“gating”). ATP binding opens CFTR and permits Cl⁻ flux; the subsequent hydrolysis of one ATP molecule then closes the gate (Gadsby et al. 2006).

The CFTR R domain is disordered

As described previously, overexpression studies indicate functional interaction between CFTR’s phosphorylated R domain with the STAS domain of SLC26 transporters. What is known about the structure of the R domain? An important early study analyzing recombinant R domain protein provided a crucial clue (Ostedgaard et al. 2000): spectra obtained using circular dichroism (CD) characterized it as primarily random coil (disordered) having low helicity. The phosphorylated recombinant protein exhibited a slightly lower propensity to fold. Subsequent nuclear magnetic resonance (NMR) studies confirmed the intrinsic disorder of the R domain (Baker et al. 2007; Kanelis et al. 2011). Phosphorylation reduced the helical content, in agreement with the earlier CD measurements. Low residue relaxation rates confirmed its disordered nature and suggested transitional structural contacts and conformations. Computer simulations utilizing discrete molecular dynamics simulation (DMD) generated low-energy, R domain conformers with low helical content (Fig. 3), and subsequent comparisons of disorder patterns generated for different phosphorylated conformers yielded similar conclusions (Hegedus et al. 2008). Independent approaches thus have built a convincing case for disorder within the R domain. The notion that intrinsic disorder may prove advantageous is steadily gaining support. Notably, key, interactive “hub proteins” contain regions of high disorder which may underlie rapid and reversible interactions between multiple binding partners (Dosztanyi et al. 2006; Patil and Nakamura 2006; Radivojac et al. 2007).

Fig. 3.

Secondary structural propensity plot generated by molecular dynamics simulation (DMD) simulations. The plot shows population-weighted averages for residues within the R domain. The positive ordinate values (red; helix) correspond to α-helical probabilities, whereas negative values (black; strand) reflect likelihood for β-sheets. This image is modified from Hegedus et al. (2008), with permission from Elsevier

Implications for intermolecular interactions

The paucity of secondary structural elements in the R domain therefore presents unique challenges in evaluating how it might interact not only with other domains within CFTR (Chappe et al. 2005) but also with the STAS domain of SLC26 transporters. A given phosphorylation event will determine the structure that sets the stage for the subsequent phosphorylation event. In disease-associated R-domain mutants, this sequence may be disrupted. Taking into account nine potential PKA sites in the R domain itself, the maximum number of possible ordered sequences for phosphorylation is nine! or 362,880. The likelihood of following any given conformer sequence will be determined by folding and energetic barriers at any partially phosphorylated step, thereby limiting the actual number of possible conformers. Still, these likely will lend substantial flexibility to any intermolecular interactions as well. Disease-associated R domain mutants are expected to produce an altered sequence of intermediates which in turn would affect interactions with SLC26 transporters. It is worth noting that recent structural models of R domain regulation by the binding of the oppositely acting kinases, AMPK and PKA, highlight the functional importance of disorder-associated structural flexibility (Siwiak et al. 2012).

STAS domains: important, interactive regions in SLC26 transporters

Aravind and Koonin first identified STAS domains in comparisons of the carboxyl-termini of SLC26 transporters with bacterial anti-sigma factor antagonists—in particular, the sporulation regulatory factor, SpoIIAA (Aravind and Koonin 2000). The NMR solution structure of unphosphorylated SpoIIAA showed four β-sheet regions and four α-helices arranged into a binding fold, presumably mediating its interaction with the kinase SpoIIAB (Kovacs et al. 1998); alternatively, five α-helices were assigned in predictive alignments with SLC26 family members (Aravind and Koonin 2000). In secondary structural predictions, the loop between the third β-sheet and the second α-helix was conserved in SpoIIAA and mammalian STAS domains (Aravind and Koonin 2000). In SpoIIAA, the STAS domain has a mild NTPase activity, and this conserved loop may coordinate phosphate binding. Phosphorylation of a serine residue (S57) immediately proximal to the second α-helix acts as a switch for SpoIIAA.

Interestingly, a point mutation in SLC26A4, F667C, localizes to this region of the conserved loop and moreover associates with the sensorineural deafness and goiter characteristic of Pendred syndrome (Aravind and Koonin 2000; Everett et al. 1997). Furthermore, mutation G678V, located near the end of SLC26A2’s second α-helix, is associated with DTDST (Superti-Furga et al. 1996). These findings underscore the importance of the conserved loop to the overall regulatory and interactive functions of STAS domains in SLC26A transport proteins.

Information from non-mammalian STAS-type structures

Available structural information on bacterial and plant STAS domains can offer insights into understanding important aspects of mammalian SLC26 protein function (Babu et al. 2010; Compton et al. 2011; Price and Howitt 2011; Rouached et al. 2005; Sharma et al. 2011b; Shibagaki and Grossman 2004; Shibagaki and Grossman 2006). Recent studies of a bacterial SLC26 protein (Yersinia enterocolitica Slc26A2) yielded a low-resolution structure indicating dimeric transporter stoichiometry, with the cytosolically localized STAS domains projecting away from the associated transmembrane domains (Compton et al. 2011). The dimeric stoichiometry agrees with previous functional measurements of bacterial, zebrafish and mammalian SLC26 transport (Detro-Dassen et al. 2008).

Earlier studies entailing large STAS domain deletions of the Arabadopsis thaliana SO₄ ⁻ transporters, SULTR1.1 and SULTR1.2A, suggested a role in SO₄ ⁻ accumulation by mediating either bona fide substrate transport or, alternatively, by enabling SULTR membrane localization (Shibagaki and Grossman 2004). A three-dimensional homology model of the STAS domain of SULTR1.2 aligned well with the known SpoIIAA crystal and NMR structures (Kovacs et al. 1998; Rouached et al. 2005; Seavers et al. 2001). The homology model provided a navigation tool with which deletions and mutations within the STAS domain were targeted. Transport by STAS-region SULTR1.2 mutants was quantified by growth rescue of a yeast strain deficient in SO₄ ⁻ transport, defining regions critical to SO₄ ⁻ transport but dispensable for proper membrane localization (Rouached et al. 2005). Using an extensive random mutagenesis approach, regions necessary for trafficking as well as intermolecular binding were identified by a similar growth rescue assay (Shibagaki and Grossman 2006).

Understanding mammalian STAS domain structure elucidates disease mechanisms

The availability of bacterial structures thus can provide an important framework with which to generate testable hypotheses aimed at discerning STAS domain functions in mammalian SLC26 proteins, but their usefulness is also limited. A systematic and elegant analysis of four SLC26A3 STAS mutants (I675/6ins, G702Tins, ΔY526/7 and I544N) associated with CLD (Makela et al. 2002) demonstrated structural and functional disruption by either of two distinct mechanisms: (1) domain misfolding, with retention at the level of the endoplasmic reticulum or (2) impaired intramolecular associations leading to diminished membrane trafficking (Dorwart et al. 2008a). Interestingly, the two insertion mutants exert their effects via the first mechanism, whereas the deletion and point mutations traffic inadequately. In the case of ΔY526/7, growth at lowered temperature (30°C) restored membrane targeting and function.

An exciting breakthrough recently emerged with the attainment of a high-resolution (1.57 Å) structure for rat Slc26a5 (Pasqualetto et al. 2010) (PDB 3IIoA). The structure of the crystallized protein indicated six β-sheets and five α-helices (Fig. 4a). Notably, the amino and carboxyl terminal boundaries extend beyond those inferred from SpoIIAA, suggesting that hurdles encountered in earlier attempts to analyze SLC26A3 STAS domains (Dorwart et al., 2008a) might be overcome by extending these boundaries. The structure predicts a surface for homologous (e.g. with intracellular loops linking membrane-spanning SLC26 domains) as well as heterologous (e.g. the R domain of CFTR?) interactions. In addition to the amino-terminal proline, the interactive surface contains a highly conserved phenylalanine within the intervening loop (conserved loop) between β-strand 3 and α-helix 2 (Fig. 4b).

Fig. 4.

Information derived from the rat Slc26a5 high-resolution structure. a Representation of the structure of the rat Slc26a5 STAS domain; green connecting loops joining 5 yellow β-strands and 5 red α-helices. The three-dimensional (3D) image is reprinted from Pasqualetto et al. (2010), with permission from Elsevier. A linear map showing sequential organization of the secondary structural elements is shown beneath the 3D representation. C.L. Conserved loop. b Residues in human SLC26A3-A6 STAS domains that putatively interact with the cytosolic interface of the apical membrane were determined by alignment with the structure rat Slc26a5 (Pasqualetto et al. 2010). Residues are identical in numbering for both rat Slc26a5 and human SLC26A5. For the four human SLC26 transporters compared, note the conservation of the proline residue within the extended N-terminus region (N), as well as the phenylalanine at the end of the C.L. between β-sheet 3 and α-helix 2

Modeling CLD-associated STAS mutations in SLC26A3 to the rat Slc26a5 STAS coordinates, Pasqualetto and coworkers revisited and confirmed previously proposed disease mechanisms (Dorwart et al. 2008a; Pasqualetto et al. 2010). These studies established proof-of-principle that the structure of rat Slc26a5 STAS may be used as a template for examining other mammalian SLC26 STAS domains. Indeed, Sharma and colleagues recently reported a model for SLC26A4 based on the published Slc26a5 structure (Sharma et al. 2011a). Future work directed at modeling—and empirically testing—wild-type and other deleterious, disease-associated STAS domain interactions with the R domain of CFTR likely will benefit tremendously from the availability of the rat Slc26a STAS domain template structure.

Conclusion

The last decade has witnessed an explosion of multidisciplinary studies focused on the inter-woven roles of CFTR and SLC26 transporters in vectorial HCO₃ ⁻ transport, a process critical to the normal functioning of numerous organ systems. Several devastating human diseases result from impaired HCO₃ ⁻ transport, and the strides made in understanding the basic properties of CFTR and SLC26 proteins pave the way for therapies and cures for these conditions. As the result of melding the disciplines of biophysics, biochemistry, cell biology, physiology and molecular medicine, the key role of intermolecular domain interactions in epithelial HCO₃ ⁻ secretion is now recognized. These include direct interactions between CFTR and SLC26 transporters as well as those mediated by PDZ scaffolding proteins. Armed with the wealth of structural information now available for both CFTR and SLC26, the field is poised to accelerate toward a higher level of understanding these physiologically essential interactions. Multi-disciplinary approaches have facilitated—and will continue to facilitate—critical advances in this exciting field.