Transepithelial Cl− and HCO3− transport is critically important for the function of all epithelia and, when altered or ablated, leads to a number of diseases, including cystic fibrosis, congenital chloride diarrhea, deafness, and hypotension (78, 111, 119, 126). HCO3− is the biological buffer that maintains acid-base balance, thereby preventing metabolic and respiratory acidosis (48). HCO3− also buffers the pH of the mucosal layers that line all epithelia, protecting them from injury (2). Being a chaotropic ion, HCO3− is essential for solubilization of ions and macromolecules such as mucins and digestive enzymes in secreted fluids. Most epithelia have a Cl−/HCO3 exchange activity in the luminal membrane. The molecular nature of this activity remained a mystery for many years until the discovery of SLC26A3 and the realization that it is a member of a new family of Cl− and HCO3− transporters, the SLC26 family (73, 78). This review will highlight structural features, the functional diversity, and several regulatory aspects of the SLC26 transporters.
Structural Features of the SLC26 Family
The human SLC26 transporter family consists of 11 members, with SLC26A10 likely being a pseudogene. The relatedness of the family members within and among species is depicted in the evolutionary tree shown in FIGURE 1⇓. There is no functional information on the D. melanogaster, and C. elegans transporters, and SO4= transport activity was reported for only two of the A. thaliana Slc26 transporters (see below). The functional diversity of the SLC26 transporters makes it impossible to infer their function from sequence relatedness, which will have to be addressed experimentally. Interestingly, the A. thaliana transporters show closer relation to the mammalian transporters than the C. elegans transporters, raising the possibility that the C. elegans transporters may have other or additional functions to anion transport. We also note that SLC26A11 is more closely related to the Drosophila and Arabidopsis sequences than to the other mammalian SLC26 transporters, suggesting that SLC26A11 evolved before the other mammalian transporters and may retain functions that are more specific to the D. melanogaster transporters.
The first member of the human SLC26 transporters family identified and cloned was SLC26A1, also known as the sulfate anion transporter 1 (Sat-1) (9). SLC26A2 was discovered by positional cloning of the gene responsible for dystrophic dysplasia (DTDST) (33). SLC26A3 was found as a gene that is downregulated in human colon adenomas and adenocarcinomas and therefore named DRA (96). It was subsequently realized that SLC26A3 is the gene mutated in the human disease congenital chloride diarrhea (40). The gene responsible for Pendred syndrome, an autosomal-recessive disorder defined by hearing loss and goiter, was the fourth member of the SLC26 family found (25). SLC26A5 was identified by searching for the gene responsible for the electromotility of outer hair cells in the cochlea (134) and was subsequently linked to a form of non-syndromic hearing loss (61). SLC26A6 was discovered by database searches for novel members of the SLC26 transporters (62), and as the renal Cl−/oxalate and Cl−/formate exchanger (44, 50). SLC26A7-11 were all uncovered through database searches for new members of the SLC26 family (62, 63, 116, 117).
The SLC26 transporters are large proteins comprised of 700–1,000 amino acids, and the individual family members have 21–43% amino acid identity. The assembly pattern of a bacterial, a teleost, and two mammalian SLC26 transporters suggests that the SLC26 transporters exist as dimers, and the dimeric assembly affects their function (17). The NH2 terminal two-thirds of the protein are largely hydrophobic and encompass the transmembrane segments where the anion binding sites and pores are thought to be located. The COOH terminal third of the protein is predicted to contain a sulfate transporter and anti-sigma factor antagonist (STAS) domain (4). Immunolocalization suggests that the NH2 and COOH termini of SLC26A3, 5, and 6 are cytosolic, indicating an even number of transmembrane spans for these isoforms (64, 65, 80, 133). With the low sequence identity between the individual family members, it is difficult to accurately predict the exact number of SLC26 transporter transmembrane spans. Various membrane-spanning prediction algorithms predict anywhere between 8 and 14 transmembrane α-helices. However, we note that the bacterial Cl−/H+ exchanger has 16 membrane-embedded helices that could not be predicted by any algorithm (22, 23). A compounding problem is that several recently solved crystal structures of ion channels and transporters contain short membrane helices that do not span the entire membrane but rather break and turn in the lipid bilayer (30, 76). These broken helices expose backbone amid and carboxyl groups that often are part of ion and solute binding sites. Thus it would not be surprising if the solute binding site(s) of the SLC26 transporters contained one or more broken helices that may not be predicted by the different algorithms.
The SLC26 transporter COOH-terminal STAS domains were found to be homologous with the SpoIIAA anti-sigma factor antagonists from sporulating bacteria (4). This homology suggests that the SpoIIAA proteins and the SLC26 STAS domains have similar structures. The SpoIIAA structure has a central β-sheet that is flanked by two α-helices on one side and by two small α-helices at the end of the β-sheet (FIGURE 2A⇓). An important difference between the SpoIIAA protein and the human SLC26 transporter STAS domains is a large insertion or intervening sequence (IVS) between the first predicted α-helix and the third β-strand of the SLC26 STAS domains (4). Secondary structure prediction algorithms suggest that the IVS is likely to be disordered in solution.
The functional significance of this region is currently unknown. The function of the STAS domain is not well understood, whereas the function of SpoIIAA is well characterized. SpoIIAA is a transcriptional regulator that interacts with the protein kinase SpoIIAB, a SpoIIE phosphatase, and the RNA polymerase sigma factor F (αF) (24). SpoIIAB phosphorylates SpoIIAA on a conserved Ser57 at the “top face” of the domain (FIGURE 2A⇑), resulting in SpoIIAA dissociating from SpoIIAB. Like SpoIIAA, the SLC26 transporter STAS domains have been hypothesized to be protein-protein interaction domains. Indeed, the STAS domain of SLC26A8 can bind to MgcRacGAP, the STAS domains of human and mouse SLC26A3 and Slc26a6 interact with CFTR R-domain (52, 113), while the Slc26a6 STAS domain interacts with carbonic anhydrase II (3).
Additional information regarding the function of the STAS domains has come from genetic studies on the Arabidopsis thaliana SLC26 transporters (Sultr) (91, 103) and analysis of disease-causing mutations in the STAS domain of SLC26A3 (75), which suggests that the STAS domain is important for membrane targeting and for transporter function. The phosphorylatable residue analogous to Ser57 of B. sphaericus SpoIIAA is conserved in most SLC26 STAS domains (FIGURE 2B⇑). In Sultr1;2, this residue is Thr587, and, when mutated to Ser, Asp, Ala, or deleted, the SO4= transport activity of Sultr is markedly reduced (91). Furthermore, random mutagenesis of the Sultr1;2 STAS domain identified this potential phosphorylation site as well as other residues to be important for SO4= transport (103). Mapping these mutants onto the SpoIIAA structure indicates that most of the mutations are localized to the top face of the structure (103). This is the same face that forms the SpoIIAA-SpoIIAB dimer interface (70), suggesting that this part of the molecule may be involved in a protein-protein interaction. The C645S and C646S mutants in Sult1;2 also cause a loss of SO4= transport activity (91), suggesting that these residues are functionally important or that they are critical for STAS domain folding. Similarly, the linking region between the final transmembrane segment and the beginning of the STAS domain has been shown to be important for transporter function (102). Residues Y526/7 and I544 in SLC26A3 STAS domain, which are mutated in patients with congenital Cl−diarrhea, are involved in folding and exit of SLC26A3 from the ER (75).
The Transport Function of the SLC26 Transporters
Critical information for all transporters is the mode and stoichiometry of the transport, the identity of the ion binding sites and conduction pathways, and the ionic specificity/selectivity of the transport. We have grouped the SLC26 transporters based on the functional similarities known at the time of writing this review. Group 1 is comprised of selective sulfate transporters and includes SLC26A1 and SLC26A2; group 2 members are coupled Cl−/HCO3− exchangers and include SLC26A3, SLC26A4, and SLC26A6. Members of group 3 function as ion channels and include SLC26A7 and SLC26A9. The transport modes of SLC26A8 and SLC26A11 are not known, and SLC26A5 does not appear to function as anion transporter in mammals (95). The functional diversity of the SLC26 transporters is similar to that found within the ClC family, where the bacterial ClC-ec1 and mammalian ClC4 and ClC5 function as electrogenic Cl−/H+ exchangers, whereas other members of the family function as Cl− channels (135). Despite these diverse functions, the ClCs share a conserved molecular architecture (21). Similarly, the SLC26 transporters may also share a basic structure that is used to generate the functional diversity observed in this protein family.
SLC26A1 is expressed in the liver and kidney, and to a lesser extent in the pancreas and testis (71, 72, 87), and is targeted to the basolateral membrane of epithelial cells (47, 86, 88). SLC26A1 functions as a DIDS-sensitive SO4 = transporter that poorly transports Cl− (87). SLC26A1 also transports oxalate, and the extent of oxalate and SO4 = transport by SLC26A1 appears to be similar (4, 129). The mechanism of SO4= and oxalate transport by SLC26A1 is not known. Based on inhibitor sensitivity, it was suggested that SLC26A1 functions as SO4=/HCO3− exchanger (69). However, no direct measurement of SLC26A1-mediated HCO3− driven SO4= or SO4= driven HCO3− transport is available. SLC26A1-mediated SO4= uptake is stimulated by extracellular halides and acidic extracellular pH (129), suggesting that SLC26A1 is not likely to function as a SO4=/Cl− exchanger. Other potential modes of SO4= transport have not been examined. The physiological role for SLC26A1 is not well understood, but it has = homeostasis and been suggested to play a role in SO4 sulfation of proteoglycans in the liver (86). In the kidney, Slc26a1 is found in the basolateral membrane of the S1, S2, and S3 segments of the proximal tubule (47). Several functional characteristics of oxalate transport by SLC26A1 (79, 129) are similar to the characteristics of oxalate transport in the basolateral membrane of the proximal tubule as determined in isolated vesicles (55, 85, 89) and the intact kidney (10). This suggests that SLC26A1 participates in renal oxalate homeostasis.
SLC26A2 is SLC26A1’s closest paralog and appears to function as a SO4=/Cl− but not as a SO4=/HCO3− exchanger (94). SLC26A2 is ubiquitous and is highly expressed in rib cartilage and the small intestine (31, = needed for proteo-33, 94). SLC26A2 provides the SO4 glycan sulfation, which is critical for cartilage development and function (26, 35). Accordingly, mutations in the SLC26A2 gene have been linked to diastrophic dysplasia, which is characterized by short limbs and stature, and joint and bone malformations (33). Three other chondrodysplasias have also been linked to mutations in SLC26A2 (achondrogenesis type IB, atelosteogenesis type II, and autosomal recessive multiple epiphyseal dysplasia), highlighting the important role of SLC26A2 in development of cartilage and the skeletal system (33–35, 108–110). Indeed, deletion of Slc26a2 in mice is embryonic lethal, and partial loss of Slc26a2 function is sufficient to cause a diastrophic dysplasia-like phenotype (26).
The group 2 Cl−/HCO3− exchangers are found in the luminal membrane of secretory epithelia and mediate luminal Cl− absorption and HCO3− secretion. Slc26a3 functions as a coupled Cl−/HCO3− exchanger that can also mediate Cl−/OH− exchange (73, 101). Measurement of the effect of cell depolarization by high extracellular K+ led to the conclusion that mouse Slc26a3 (73) and human SLC26A3 (56) are electroneutral exchangers. However, depolarization by K+ may not accurately assay the electrogenicity of a transporter since electrogenic transporters markedly affect the membrane potential, in particular when the cells are exposed to large gradients of the transported ions, as is the case when Cl−/HCO3− exchange is measured by exposing the cell to Cl−-free media. The electrogenic nature of a transporter can be accurately determined by measuring the stoichiometry of the exchange while clamping the membrane potential at set voltages. Such a detailed analysis of Cl− and HCO3−transport by Slc26a3 expressed in Xenopus oocytes revealed that Slc26a3 is an electrogenic transporter with a 2Cl−/1HCO3− exchange stoichiometry (FIGURE 3A⇓) (51, 101). Surprisingly, Slc26a3 appears to function as both a coupled transporter and as a channel, depending on the ions being transported. A large Slc26a3-mediated current can be measured in the presence of NO3− or SCN− (101). A coupled transport and channel behavior has also been reported for the Na+-HCO3− co-transporter NBCn1 (13). The bacterial ClC-ec1 functions as a 2Cl−/H+ exchanger and mutating Glu148 or Tyr445, which line the ion conducting pathway, converts it to a Cl− channel (1). By analogy, it is possible that similar residues in Slc26a3 retard the binding and movement of Cl−through the transporter but not of NO3− and SNC−, allowing uncoupled conductive transport of these ions.
Consistent with the function of SLC26A3 as a Cl−/HCO3− exchanger, mutations in SLC26A3 have been linked to congenital chloride diarrhea (CLD) (40). CLD patients present clinically with watery acidic diarrhea containing high levels of Cl− and metabolic alkalosis (67). Mice lacking Slc26a3 display reduced luminal Cl−/HCO3− exchange activity in the colon epithelium that results in acidic diarrhea with high chloride content, volume depletion, and growth retardation (97).
The second member of group 2 is the ubiquitously expressed SLC26A6 (50, 62, 129). Slc26a6 functions as a coupled electrogenic Cl−/HCO3− exchanger with a stoichiometry of 1Cl−/2HCO3− (FIGURE 3A⇑) (51, 101, 129). On the other hand, Chernova et al. (12) failed to measure a Slc26a6-mediated change in the membrane potential. However, current and pHi measurements in these studies are somewhat problematic in that the current generated by oxalate, the pHi changes and the Cl− fluxes by the mouse and human transporters are in disagreement with each other (12), whereas two independent studies found that Slc26a6 is elecrogenic (51, 101, 129). Moreover, we reexamined the clone sent to us by Alper et al. and found that it mediates electrogenic Cl−/HCO3− exchange (Shcheynikov N, Muallem S, unpublished observation). The 1Cl−/2HCO3− stoichiometry of Slc26a6 is probably essential for the pancreatic duct to secrete a fluid containing 140 mM HCO3− (52, 106) and likely plays a role in secreting HCO3−-rich fluids by other epithelia like the salivary glands and the airway (14, 41). Another major function of Slc26a6 is Cl−/oxalate and Cl−/formate exchange, and thus it is also known as CFEX (50). The role of Slc26a6 in oxalate transport was revealed by the finding that deletion of Slc26a6 leads to calcium oxalate urolithiasis as a result of decreased intestinal oxalate secretion, which leads to increased net oxalate absorption and an increased plasma oxalate concentration. The increased plasma oxalate results in increased renal oxalate load and excretion (43). Slc26a6 also mediates the oxalate-dependent NaCl absorption in the proximal tubule (68, 124).
The physiological role of SLC26A6 is only partially understood, and most of the information comes from disruption of the Slc26a6 gene in mice. Slc26a6−/− mice show impaired intestinal and pancreatic duct Cl− absorption and HCO3− secretion (42, 99, 123), along with impaired pancreatic fluid secretion (123). Moreover, deletion of Slc26a6 in the pancreatic duct results in disregulation of CFTR (123), indicating a role of Slc26a6 in the overall epithelial fluid and electrolyte secretion. However, the pancreatic duct phenotype of two independently generated Slc26a6−/− mouse lines show significant differences. Although in both mouse lines deletion of Slc26a6 resulted in a marked increase in ductal Cl− and HCO3− permeability, in our mouse line deletion of Slc26a6 increased the spontaneous and impaired the stimulated fluid and HCO3− secretion (123). This is in contrast to the other mouse line, which showed high spontaneous basal fluid secretion in the wild-type duct and no further increase of the basal secretion or impairment of the stimulated secretion on deletion of Slc26a6 (42). RT-PCR analysis attributed the increased Cl− and HCO3− permeability to an increase in Slc26a3 mRNA expression (42), whereas our qRT-PCR analysis failed to detect a change in Slc26a3 mRNA expression (123) . Finally, knockdown of CFTR, but not of Slc26a3, by siRNA markedly reduced the increased spontaneous HCO3− permeability and fluid secretion in our Slc26a6−/− mice (123). The role of CFTR in the spontaneous and stimulated fluid and HCO3− secretion in the wild-type and Slc26a6−/− mouse line used by Ishiguro et al. was not investigated (42). Hence, although both Slc26a6−/− mouse lines point to the importance of Slc26a6 in pancreatic duct fluid and electrolyte secretion, the consequence of deletion of Slc26a6 appears quite different. The reason for the different phenotypes is not clear at present, although high basal fluid and HCO3− secretion (42) may mask the increased spontaneous secretion due to deletion of Slc26a6 (123). In addition, it is not obvious how the marked increase in Cl− and HCO3− permeability observed in the two mouse lines will have no apparent effect on ductal fluid and electrolyte secretion when Cl− and HCO3− permeability clearly play a central role in ductal secretion.
Another group 2 member is SLC26A4, which is expressed in the kidney, cochlea, thyroid follicular cells (25, 92, 93, 105, 125), and salivary gland (unpublished observation from our laboratory). Mutations in SLC26A4 have been linked to Pendred syndrome and non-syndromic hearing loss DFNB4 (11, 15, 25, 27, 59, 84, 115). In addition to hearing loss, Pendred syndrome patients often present with goiter as a result of impaired I− organification in the thyroid (27, 33). Accordingly, SLC26A4 functions as an I− transporter (98). However, impaired I− transport is not likely to account for the syndromic and nonsyndromic deafness caused by mutations in SLC26A4. Another function of SLC26A4 is Cl−/HCO3− exchange (105), and impaired Cl−/HCO3− exchange may explain the hearing phenotype of SLC26A4 mutations. Indeed, an important study recently showed that Slc26a4−/− mice have aberrant HCO3− secretion in the epithelial cells of the inner ear resulting in acidified endolymph (126). Another site of Slc26a4 expression is the distal convoluted tubule, connecting tubule and cortical collecting duct. Specifically, Slc26a4 is found in the apical plasma membrane of type B and non-A, non-B intercalated cells in the cortical collecting duct and participates in Cl− -dependent HCO3− secretion (118). Accordingly, Slc26a4 participates in control of vascular volume and arterial pH.
The transport mode of I−, Cl−, and HCO3− by SLC26A4 is poorly understood, and the available information is contradictory. Current measurements in HEK cells concluded that SLC26A4 functions as an I−and Cl− channel (132). A subsequent study could not corroborate SLC26A4 function as a Cl− or I− channel but reported that expression of SLC26A4 activates a small outward rectifying K+ current (19). Cl−/HCO3− exchange-like activity was reported with SLC26A4 stably expressed in HEK cells; however, the mode of transport was not investigated in any detail (105). Characterization of the transport mode by SLC26A4 awaits further studies.
The third group of SLC26 transporters includes SLC26A7 and SLC26A9, which function as selective Cl− channels (18, 49). SLC26A7 expression has been identified in high endothelial venules, kidney, stomach, nasal epithelium, and epididymal ducts (7, 54, 83, 117). SLC26A7 has been localized to the basolateral membrane of the gastric parietal cells, the renal outer medullary collecting duct and the thick ascending limb (8, 20, 82, 83, 131), the apical membrane of the proximal tubule (20), and in endosomes in MDCK cells when grown under isotonic conditions (131). The targeting motifs and regulatory mechanisms that control SLC26A7 trafficking to the basolateral or luminal membrane, or to endosomes, are not known.
Based on pHi measurement in Xenopus oocytes expressing SLC26A7, it was concluded that SLC26A7 functions as a Cl−/HCO3− exchanger (82, 83). However, current measurements in Xenopus oocytes and HEK cells expressing SLC26A7 revealed that SLC26A7 is a highly selective Cl− channel with minimal HCO3− permeability (FIGURE 3B⇑) (49). An interesting feature of SLC26A7 is its regulation by pHi, with acidic pHi increasing the selectivity of SLC26A7 to Cl− (49). The physiological role of SLC26A7 is not known at present, but it may function as a Cl− loading mechanism in parietal cells (53). A pHi-regulated Cl− channel may serve as a sensor of pHi in cells that secrete acid or base equivalents to tighten their plasma membrane Cl− selectivity and favor coupled on the expense of uncoupled Cl− transport.
SLC26A9 is expressed at high levels in the luminal membrane of ciliated airway bronchial and alveolar epithelial cells (63), as well as the gastric surface, and at low levels in other glandular cells (130). More recently, expression of SLC26A9 was reported to be widespread (104). Expression of Slc26a9 appears to be dynamic and is upregulated in the gastric mucosa of H. pylori-infected mice (39). Measurement of Cl−-dependent changes in pHi concluded that SLC26A9 functions as a Cl−/HCO3− exchanger that is inhibited by NH4+ (63, 131). However, we recently measured SLC26A9 activity expressed in Xenopus oocytes or HEK cells, and found that SLC26A9 is a Cl− channel with minimal HCO3− permeability (FIGURE 3C⇑) (18). A similar conclusion, suggesting that SLC26A9 is a Cl− channel, was reported elsewhere (104). The functional and pharmacological properties of SLC26A9 are quite different from the Cl−/HCO3− exchange activity and Cl− secretion properties of the gastric mucosa (16). This suggests that SLC26A9 is not likely the gastric epithelium HCO3− transporter.
SLC26A5, SLC26A8, and SLC26A11 do not currently fit into any of the three groups described above. The current knowledge of the transport properties of SLC26A8 and SLC26A11 is meager and not sufficient to confidently place them in one of the classified groups. SLC26A8, also know as testis anion transporter 1 (Tat1), is expressed in the testis and the brain and has been suggested to function as a SO4= and/or I− transporter (63, 113, 116). Due to its expression in the male germ line, SLC26A8 was suggested to play a role in male fertility (63, 113). This was questioned by a study that failed to find mutations in the SLC26A8 gene in a screen of sterile male patients (66). However, knockout of Slc26a8 in mice resulted in male infertility caused by abnormal flagellar differentiation and the loss of sperm cell motility (112). SLC26A11 expression is ubiquitous and mediates SO4− uptake when expressed in Sf9 insect cells (74, 116). The physiological role of SLC26A11 is not currently known.
SLC26A5, also known as Prestin, is expressed in the outer hair cells (OHC) of the cochlea in the mammalian ear (134). A splicing defect in SLC26A5 leads to a non-syndromic deafness in humans (61), and deletion of Slc26a5 in mice caused a lack of OHC electro-motility and hearing loss (60). Unlike other SLC26 transporters, mammalian SLC26A5 does not function as a transporter but rather functions as a molecular motor that alters the shape of the OHCs in response to changes in membrane potential (134), which is the basis for cochlear amplification (5). SLC26A5 uses bound Cl− and HCO3− ions as voltage sensors to detect the membrane potential and control cell shape change (81). Unexpectedly, a recent study found that the chicken and zebrafish SLC26A5 orthologs function as 1:1 SO4=/Cl− exchangers (95). This suggests that the SLC26A5 ion transport activity evolved into an ion binding ability as the mammalian cochlea evolved.
A number of regulatory mechanisms have been described for the SLC26 transporters, including transcriptional, protein trafficking, and posttranslational modifications. A number of physiological stimuli have been shown to regulate several SLC26 transporters by affecting their gene or protein expression. Thyroid hormone regulates the expression of SLC26A1, SLC26A4, and SLC26A5 (58, 92, 127), aldosterone increases expression of Slc26a4 (120), and vasopressin and a low-K+ diet increase expression of Slc26a7 in the cortical collecting duct (7). Physiological states such as inflammation and enteropathogenic E. coli infection reduce the expression of Slc26a3 (29), whereas H. pylori infections induce expression of Slc26a9 (39).
In many of these cases, the regulation alters expression of the SLC26 transporters in the plasma membrane by an unknown mechanism. However, recent findings raise the possibility that expression of the SLC26 transporters in the plasma membrane may be regulated by the With No lysine (K) kinases (WNKs). The WNKs are protein kinases that lack a highly conserved catalytic lysine in their kinase domain, which is replaced by a lysine in the second β strand of the kinase fold (77). Interest in the WNKs dramatically increased with the discovery that mutations in WNK1 and WNK4 are linked to pseudo-hypoaldosteronism type II (PHAII), which is characterized by excessive renal Na+, Cl−, and K+ retention (128). Several studies have revealed that the WNKs regulate expression of key Na+, K+, and Cl− transporters at the apical and basolateral membranes of selective renal tubular segments (46, 107). The WNKs determine expression of transporters at the plasma membrane by regulating intersectin-mediated endocytosis (36). When expressed in Xenopus oocytes, WNK4 inhibits the activity of Slc26a6 (45), presumably by affecting its expression at the plasma membrane. Recently, we showed that expression of SLC26A9 at the plasma membrane, and consequently SLC26A9 channel activity, is markedly reduced by WNK1, WNK3, and WNK4 (18). It would be interesting to determine whether plasma membrane expression of all SLC26 transporters is regulated by the WNKs and whether physiological stimuli are controlling the activity of the WNKs to determine plasma membrane expression of the SLC26 transporters.
Posttranslational modifications and macromolecular complex formation also regulate many of the SLC26 transporters. Based on loss of prostaglandin E2-stimulated HCO3− secretion in the intestine of Slc26a6−/− mice, it is possible that PGE2 stimulates Slc26a6 activity by an unknown mechanism (114). Stimulation of Slc26a6 by PGE2 is not likely to involve PKA because of the puzzling observation that forskolin-stimulated HCO3− secretion is normal in Slc26a6−/− intestine (114). PKC phosphorylation of Slc26a6 inhibits transport activity by disrupting carbonic anhydrase II binding and the formation of a HCO3− transport metabolon (3). PKC phosphorylation of Slc26a6 also reduces its plasma membrane localization by an unknown mechanism (3, 32). Since Slc26a6 is essential for pancreatic duct fluid and HCO3− secretion (123), inhibition of Slc26a6 activity by PKC may explain the inhibition by substance P of pancreatic duct fluid and HCO3− secretion that is mediated by substance P-mediated activation of PKC (37, 38).
Macromolecular complexes are formed by at least five of the SLC26 family members since they possess PDZ ligands at their COOH terminus, allowing them to bind to PDZ domain containing scaffolding proteins (63). SLC26A3 forms a complex with the scaffolding proteins E3KARP and CAP70 (57, 90), and Slc26a7 binds to E3KARP and NHERF (64). Interestingly, CFTR can bind to these same scaffolding proteins (121, 122) to form a macromolecular complex. In the complex, CFTR activates Slc26a3, SLC26A4, and Slc26a6 (51), and in turn Slc26a3 and Slc26a6 activate CFTR channel activity (52). Accordingly, CFTR-regulated fluid secretion by the pancreatic duct requires the regulatory interaction of Slc26a6 and CFTR. In the resting state, Slc26a6 inhibits the activity of CFTR, and in the stimulated state Slc26a6 enhances the Cl− channel activity of CFTR (123). The mutual activation of CFTR and the SLC26 transporters is dependant on the physical association of the STAS and R-domains, which is enhanced by PKA-mediated phosphorylation of the R domain (52). Importantly, the isolated Slc26a3-STAS domain is as effective as full-length Slc26a3 in activation of CFTR (52), and, when targeted to the plasma membrane, the isolated R domain activates Slc26a3 (100). Although the stoichiometry of the R-STAS domain interaction is not known at present, activation by the isolated domains indicates that the transport function of the respective transporters is not required for their reciprocal activation. However, under physiological conditions, CFTR is likely to recycle the Cl− absorbed by the SLC26 trans porters to sustain the HCO3− secretion (see discussion in Refs. 52, 106). In this respect, the nonphosphorylated R domain interacts with the first nucleotide binding domain (NBD1) of CFTR, and this interaction is weakened by phosphorylation of the R domain (6). The combined findings suggest the model in FIGURE 4⇓ by which interaction of the SLC26 transporters with CFTR regulates epithelial fluid and electrolyte secretion in the basal and stimulated states. In the resting state, the R domain is bound to NBD1, and this binding may be enhanced or stabilized by SLC26A6. In this state, CFTR is inhibited, resulting in inhibition of fluid and electrolyte secretion. Stimulation of PKA phosphorylates the R domain to reduce its interaction with NBD1 and increase its binding to the SLC26A6 STAS domain. This stabilizes the interaction between NBD1 and NBD2 to markedly activate both CFTR and SLC26A6 to stimulate fluid and electrolyte secretion (FIGURE 4⇓).
Although our knowledge of the SLC26 transporters has increased dramatically over the past few years, many important questions remain unanswered. One of the most pressing is what residues are important for the transport cycle and mediate ion permeation through this remarkable family of transporters and channels. What role the STAS domain might play in the SLC26 transport cycle needs to be investigated along with defining the residues that mediate the STAS and R-domain interaction. Lastly, and perhaps most importantly, the physiological role of the SLC26 transporters in health and disease states needs to be determined.
- © 2008 Int. Union Physiol. Sci./Am. Physiol. Soc.