COOH-terminal cytoplasmic tails of chloride/bicarbonate anion exchangers (AE) bind cytosolic carbonic anhydrase II (CAII) to form a bicarbonate transport metabolon, a membrane protein complex that accelerates transmembrane bicarbonate flux. To determine whether interaction with CAII affects the downregulated in adenoma (DRA) chloride/bicarbonate exchanger, anion exchange activity of DRA-transfected HEK-293 cells was monitored by following changes in intracellular pH associated with bicarbonate transport. DRA-mediated bicarbonate transport activity of 18 ± 1 mM H+ equivalents/min was inhibited 53 ± 2% by 100 mM of the CAII inhibitor, acetazolamide, but was unaffected by the membrane-impermeant carbonic anhydrase inhibitor, 1-[5-sulfamoyl-1,3,4-thiadiazol-2-yl-(aminosulfonyl-4-phenyl)]-2,6-dimethyl-4-phenyl-pyridinium perchlorate. Compared with AE1, the COOH-terminal tail of DRA interacted weakly with CAII. Overexpression of a functionally inactive CAII mutant, V143Y, reduced AE1 transport activity by 61 ± 4% without effect on DRA transport activity (105 ± 7% transport activity relative to DRA alone). We conclude that cytosolic CAII is required for full DRA-mediated bicarbonate transport. However, DRA differs from other bicarbonate transport proteins because its transport activity is not stimulated by direct interaction with CAII.
- chloride/bicarbonate exchanger
- downregulated in adenoma
bicarbonate metabolismis essential in humans, because carbon dioxide is the metabolic end product of respiratory oxidation and CO2/HCO is the body's primary pH buffer system. The bicarbonate transport superfamily of genes (SLC4 and SLC26 gene families), responsible for transmembrane movement of membrane-impermeant HCO , comprises the Cl−/HCO anion exchanger (AE) family (1, 19, 20), the Na+/HCO cotransporter proteins (NBC) (3, 4), and the recently identified proteins pendrin (9, 32, 37) and downregulated in adenoma (DRA) (23, 27, 48).
Several lines of evidence have demonstrated an interaction between cytosolic carbonic anhydrase II (CAII) and the AE1, AE2, and AE3 anion exchanger isoforms. Binding of erythrocyte membranes to CAII increased CAII enzymatic activity (25), which suggests an interaction between these two proteins. CAII can be coimmunoprecipitated with solubilized AE1 and incubation with an extracellular lectin-caused agglutination of AE1 and a similar redistribution of CAII on the cytosolic surface of the erythrocyte membrane (45). A sensitive microtiter binding assay, using truncation and point mutation of the AE1 COOH terminus, led to the identification of the binding site of CAII in AE1 as LDADD (amino acids 886–890) (46) and the basic amino-terminal region of CAII as the binding site for AE1 (44).
The functional consequences of the AE/CAII interaction have been studied (42). Using HEK-293 cells transiently transfected with AE1 cDNA, we determined that inhibition of endogenous CAII activity with acetazolamide resulted in a decrease of AE1 transport activity. Mutation of the AE1, LDADD, and CAII-binding motif caused a loss of CAII binding and a corresponding 90% decrease of AE1 transport activity. Overexpression of the functionally inactive CAII mutant, V143Y (10), displaced wild-type CAII from all three members of the AE family and had a dominant-negative effect on anion transport, inhibiting transport by ∼50%. This demonstrated that binding of functional CAII to the COOH terminus of AE1, AE2, and AE3 proteins is required for maximal bicarbonate transport activity. The requirement of a physical interaction between CAII and AE for maximal bicarbonate transport provided the first direct evidence of a functional transport metabolon: a physically associated complex of proteins in a metabolic pathway (38, 39). The metabolon may accelerate the coupled production of bicarbonate and transport by minimization of the diffusional distance between CAII and the bicarbonate transporter. The interaction between CAII and a peptide corresponding to the LDADD motif has also been reported to stimulate CAII activity directly (34). Together, these effects will increase substrate concentration at the transport site, thereby stimulating bicarbonate transport.
Comparison of the amino acid sequences of the COOH-terminal tails of bicarbonate transport proteins shows that, with the exception of DRA (31), all of these proteins contain at least one consensus CAII-binding motif consisting of a hydrophobic residue followed by four amino acids, of which at least two are acidic residues (46). Formation of a transport metabolon with CAII may therefore also occur with other bicarbonate transport proteins. NBCs also physically interact with CAII, and this interaction, as in the case of the AE family, is necessary for their maximal transport activity [B. Alvarez, F. Loiselle, and J. Casey, unpublished observation (NBC1) and F. Loiselle and J. Casey, unpublished observation (NBC3)]. The absence of a potential CAII-binding site in the COOH-terminal region of DRA therefore raises the questions of whether DRA forms a complex with CAII and whether DRA requires the formation of such a complex to maximize the rate of bicarbonate transport. If DRA does not interact with CAII, it differs significantly from the other bicarbonate transporters, particularly in its mode of regulation.
Human DRA, cloned from a colon subtraction library, is expressed in the normal colon but not in most adenocarcinomas (31). The protein product of the DRA gene is a membrane glycoprotein predicted to span the membrane 10–14 times (5). The protein is related to the sulfate transporters DTSDT (13) and SAT-1 (2) and has been shown to transport sulfate when expressed in Xenopus oocytes (36). Although DRA has considerably less similarity to the Cl−/HCO anion exchange protein family, DRA also mediates Cl−/HCO exchange activity when expressed in cultured mammalian cells (12,23).
Cl−/HCO exchange function in the human colon and ileum has been attributed to DRA (23, 27), which works in concert with the Na+/H+ exchanger (NHE) to mediate NaCl absorption. Mutations in the DRA gene manifest as congenital chloride diarrhea (CLD), an autosomal recessive disorder of intestinal electrolyte absorption (14, 15). Studies of humans with CLD provide strong evidence for defects in Cl−/HCO exchange in the ileum and colon (16).
Cystic fibrosis (CF) is an autosomal recessive disease arising from inactivation or misprocessing of a cAMP-sensitive Cl−channel, known as the CF transmembrane conductance regulator (CFTR) (29). CF causes defective fluid and electrolyte secretion in secretory epithelia (26, 47), which impairs the respiratory, pancreatic, hepatobiliary, and genitourinary systems (29). Two recent papers have shown that the expression of DRA in both trachea epithelial cells and cultured pancreatic duct cells increases in the presence of active CFTR. This increase in expression of DRA is accompanied by an increase in Cl−/HCO transport activity (12,48). The authors concluded that the HCO secretion defect in patients with CF results in part from the downregulation of the Cl−/HCO exchange activity mediated by DRA.
In the present study, we investigated the interaction between DRA and CAII. Using HEK-293 cells transiently transfected with DRA cDNA, we determined that inhibition of CAII activity with acetazolamide resulted in a substantial decrease of DRA transport activity, indicating that DRA requires the presence of active CAII for maximal bicarbonate transport. A microtiter plate assay showed that the COOH-terminal tail of DRA binds CAII with a much lower affinity and capacity than AE1. Expression of functionally inactive V143Y CAII mutant (10) displaced wild-type CAII from AE1 and had a dominant-negative effect on anion transport (42). In contrast, overexpression of V143Y CAII had no effect on the rate of DRA-mediated bicarbonate transport. Taken together, these results indicate that although DRA activity requires the presence of CAII enzymatic activity in the cytosol, DRA and CAII do not form the physical complex required by AE and NBC. Therefore, DRA is thus unique among HCO transport proteins because it does not form a complex with CAII. This suggests that the regulation of DRA activity differs from NBCs and AEs.
MATERIALS AND METHODS
Rabbit anti-glutathione S-transferase (GST) antibody was from Santa Cruz Antibodies (Santa Cruz, CA). Biotinylated anti-rabbit IgG and peroxidase-labeled biotin/avidin, glutathione Sepharose, pGEX-6P-1 expression vector, and 5′ pGEX sequencing primer were from Amersham Pharmacia Biotech (Quebec, Canada). Sheep anti-human carbonic anhydrase II antibody was from Serotec (Raleigh, NC). Poly-l-lysine, nigericin, and o-phenylenediamine dihydrochloride were from Sigma Aldrich Canada (Oakville, Canada). Molecular Probes 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF-AM) was from Cedarlane Laboratories (Ontario, Canada). Glass coverslips were from Fisher Scientific Products (Nepean, Canada). Dulbecco's modified Eagle's medium (DMEM), and calf serum and fetal bovine serum were from GIBCO-BRL (Burlington, ON, Canada).
An expression construct for the human DRA protein was received as a generous gift from Dr. Manoocher Soleimani (12), and V143Y CAII mutant was from Dr. Carol Fierke (10). Expression constructs for CA and AE have been described previously (6,42). Plasmid DNA for transfections was prepared by using Qiagen columns (Qiagen, Mississauga, Canada).
Proteins were expressed by transient transfection of HEK-293 cells (11) by using the calcium phosphate method (30). Cells were grown at 37°C in an air/CO2(19:1) environment in DMEM supplemented with 5% (vol/vol) fetal bovine serum and 5% (vol/vol) calf serum.
Anion exchange activity assay.
Anion exchange activity was monitored by using a fluorescence assay described previously (41). Briefly, HEK-293 cells grown on poly-l-lysine-coated coverslips were transiently transfected as described earlier. Two days posttransfection, coverslips were rinsed in serum-free DMEM and incubated in 4 ml of serum-free medium containing 2 μM BCECF-AM (37°C, 15 min). Coverslips were then mounted in a fluorescence cuvette and perfused alternately with Ringer's buffer (5 mM glucose, 5 mM potassium gluconate, 1 mM calcium gluconate, 1 mM MgSO4, 2.5 mM NaH2PO4, 25 mM NaHCO3, and 10 mM HEPES, pH 7.4) containing either 140 mM NaCl or 140 mM sodium gluconate bubbled with air/CO2 (19:1). Fluorescence was monitored by using a Photon Technologies International RCR fluorimeter at excitation wavelengths 440 and 502.5 nm and emission wavelength 528.7 nm. After calibration using the high potassium/nigericin technique (43) at three pH values between 6.5 and 7.5, fluorescence ratios were converted to intracellular pH (pHi). Rates of change of pHi were determined by linear regression (Kaleidagraph software) of the initial HCO efflux/influx and converted to rates of H+ equivalent flux across the plasma membrane according to the equation: J = βtotal × ΔpHi (28), where, as determined previously, βtotal = 57.5 mM (41). In all cases, the transport activity of sham-transfected cells was subtracted from the total rate to ensure that these rates consist only of AE transport activity.
To assess the effect of the CA inhibitors acetazolamide and SDPPP, cells were subjected to a standard transport assay and then incubated with the appropriate inhibitor in Ringer's buffer for 10 min. Residual transport activity was then assessed in a standard assay with the appropriate inhibitor present in all buffers. Cells were also subjected to repeated anion exchange assays, in the absence of inhibitors, as a control for time-dependent changes in transport activity.
To determine whether any inhibitors had a direct effect on the transport activity of AE proteins, Cl−/NO exchange activity was assessed in the absence and presence of both inhibitors. Briefly, cells were loaded with SPQ (17), a chloride-sensitive fluorescent dye, by overnight incubation with 10 mM SPQ added to tissue culture medium (41). Cells were then perfused with Ringer's buffer containing either 140 mM KCl or 140 mM KNO3 and subjected to the same inhibition transport assay described above, with fluorescence monitored at excitation wavelength 350 nm and emission wavelength 438 nm. To correct for changes of fluorescence that occurred as a result of dye leakage from cells and cell loss from the coverslip, total changes in fluorescence were normalized as described previously (41).
GST-fusion protein construction and purification.
A GST-fusion protein consisting of the cDNA for GST fused to the 33 amino acid COOH-terminal tail of human AE1 was a generous gift from Dr. Reinhart Reithmeier (45). The GST-fusion protein, consisting of the COOH-terminal 42 amino acids of DRA, was constructed by using PCR to amplify the appropriate cDNA sequence. With the use of human DRA cDNA as template, the forward and reverse primers CGCGGATCCAAGAAAGATTACAGTACTTCAAAGTTTAATCC and CGCGGATCCGAATTTTGTTTCAACTGGCACCTCATATACCCACT, respectively, introduced a BamHI restriction site at both ends of the amplified product. The PCR product was digested withBamHI and cloned into the pGEX-6P-1 expression vector, digested in the same way. The construct GST-DRAct was verified by sequencing with a Beckman Instruments CEQ2000 DNA sequencer, and plasmid DNA was purified by using Qiagen columns. The GST-DRAct construct was transformed into Escherichia coli strain BL21-codon plus (Stratagene), and a single colony was used to inoculate 50 ml of Luria-Bertani (LB) medium. After overnight growth at 37°C with shaking, this culture was used to innoculate 1.2 l LB medium (5 ml/200 ml). The culture was grown at 37°C with shaking until the A600 was 0.6–1.0. Isopropylthiogalactoside (1 mM final) was added, and growth was allowed to continue for 2–6 h. The culture was then centrifuged at 10,000 g for 10 min, and bacterial pellets were resuspended in 4°C PBS (150 mM NaCl and 5 mM Na2HPO4, pH 7.5) containing protease inhibitors (Complete mini-protease inhibitor cocktail; Roche Applied Science, Quebec, Canada). Resuspended cells were disrupted by sonication [4 × 1 min, power level 9.5 with model W185 probe sonifier (heat systems; Ultrasonics, Plainview, NY)], and Triton X-100 was added to a final concentration of 1% (vol/vol) with slow stirring for 30 min. After centrifugation at 15,000 g for 10 min, the supernatant was added to 1.3 ml of glutathione Sepharose 4B (50% slurry equilibrated with PBS) and incubated at room temperature with gentle agitation for 1–2 h. The sample was centrifuged at 500g for 5 min, and the pellet was washed three times with PBS. The fusion protein was eluted by using 10 mM reduced glutathione in 50 mM Tris · HCl, pH 8.0.
The ability of CAII to bind the COOH-terminal tail of DRA was investigated by using a sensitive microtiter assay, described previously (42, 46). Briefly, 200 ng of purified human CAII (Sigma Aldrich) were chemically coupled to wells of a 96-well plate by using 1-cyclohexyl-3-(2-morpholino-ethyl)carbodimide metho-p-toluenesulfanate (Sigma Aldrich). Wells were washed with PBS and blocked with 2% (wt/vol) BSA in PBS. GST fusion proteins of the COOH-terminal tail of AE1 and DRA were purified as described above. After washing with antibody buffer [100 mM NaCl, 5 mM EDTA, 0.25% gelatin (wt/vol), 0.05% Triton X-100 (wt/vol), and 50 mM Tris, pH 7.5], plates were incubated with various concentrations (0–200 nM) of purified GST fusion proteins or GST alone. After being washed, bound fusion proteins were detected by sequential incubation with rabbit anti-GST antibody (Santa Cruz Antibodies), biotinylated anti-rabbit IgG (Amersham), and peroxidase-labeled biotin/avidin (Amersham). Plates were then incubated with the peroxidase substrate o-phenylenediamine dihydrochloride (Sigma), and product formation was detected at 450 nm in a Labsystems Mutiskan MCC microplate reader. Binding data were fitted by using Kaleidagraph software (Synergy Software).
Values are expressed as means ± SE. Statistical significance was determined by using a Student's paired t-test withP < 0.05 considered significant.
HCO transport activity of DRA.
To measure the anion exchange transport activity of the protein DRA, HEK-293 cells were transiently transfected with human DRA cDNA. The transfected cells were grown on coverslips and loaded with BCECF-AM, a pH-sensitive fluorescent dye. Cells were then perfused alternately with chloride-containing and chloride-free Ringer's buffer. In chloride-free Ringer's buffer, chloride leaves the cell via DRA in exchange for HCO , causing alkalinization of the cytosol. Conversely, in chloride-containing Ringer's buffer, chloride moves into the cell in exchange for the removal of HCO , resulting in cellular acidification. Changes in pHi are measured as changes in fluorescence of BCECF, and, after appropriate calibration, this experiment results in an indirect measurement of changes in pHi resulting from DRA-mediated HCO flux (Fig. 1). Rates of transport were obtained by linear regression of the initial changes of pHi after the change of Ringer's buffer. The anion exchange transport rate for cells transiently transfected with vector alone was subtracted from the total rate of transport obtained from the cells expressing DRA to give a transport rate for DRA alone. The light beam in the fluorimeter illuminates ∼1 × 104 cells so that each experiment represented the mean response of a large population of cells.
DRA transported bicarbonate with a rate of 0.3 ± 0.02 ΔpH/min. Previous determination of the intrinsic buffering capacity of these cells (21, 41) enables the rate of change in pHi to be converted to H+ flux. When applied to DRA, the ΔpHi/min rate represents a H+equivalent flux of 18 ± 1 mM/min. Thus, under similar transient transfection conditions, DRA has anion exchange activity about half of that found for human AE1 in transfected HEK-293 cells (40 ± 1 mM/min; Ref. 42).
Effects of CA inhibition on DRA Cl−/HCO exchange activity.
HEK-293 cells endogenously express CAII (42). However, immunoblots failed to detect any endogenously expressed extracellular anchored CAIV protein (data not shown). To assess the role of CA isoforms upon DRA activity, we performed DRA transport assays in the presence of CA inhibitors. Some CA inhibitors directly inhibit anion exchange activity of AE proteins, although acetazolamide does not (7, 8). To determine whether CA inhibitors affect DRA directly, transport needed to be assessed in an assay independent of CA activity. Because DRA accepts both Cl− and NO as substrates (23), we measured DRA Cl−/NO exchange activity, with the Cl−-sensitive fluorescent dye, SPQ (17). SDPPP and PTSP are membrane-impermeant, broad-spectrum CA inhibitors that inhibit CAIV with Ki of 0.4 nM and 8 nM, respectively (33). Cl−/NO exchange assays indicated that the presence of acetazolamide (100 μM) and SDPPP (40 nM) had no direct effect on DRA-mediated anion exchange (transport rates were 99 ± 6% and 102 ± 9% of control, respectively, after incubation with inhibitor). Interestingly, DRA Cl−/NO exchange activity was inhibited by 50 ± 3% in the presence of 50 nM PTSP. This indicates that DRA, like AE family Cl−/HCO exchangers, is susceptible to direct inhibition by some CA inhibitors.
Transiently transfected cells were grown on glass coverslips and subjected to Cl−/HCO exchange assays. After an initial perfusion with chloride-free and chloride-containing Ringer's buffer, cells were incubated for 10 min with 100 μM acetazolamide. Because acetazolamide does not covalently react with CA, all buffers used subsequently in the experiment also contained 100 μM acetazolamide. Figure 2 shows that treatment with 100 μM acetazolamide caused a 53 ± 2% reduction in DRA-mediated bicarbonate flux, indicating that DRA requires the presence of functional CAII for maximal transport activity. To control for time-dependent changes of DRA activity, DRA activity was measured before and after a 10-min incubation in Cl−-free Ringer's buffer without acetazolamide. DRA activity after the incubation was 102 ± 3% (n = 4) the rate before incubation, indicating no significant change of transport over this time frame.
Because SDPPP does not directly inhibit DRA transport activity, we used SDPPP to determine whether endogenous extracellular CA in HEK 293 cells contributed to DRA activity. Figure 2 shows that 40 nM SDPPP had no effect on DRA-mediated Cl−/HCO exchange activity (104 ± 5% of control), indicating that endogenous extracellular CA did not contribute to the rate of DRA-mediated transport activity in HEK 293 cells. In addition, the effect of acetazolamide on DRA Cl−/HCO exchange activity must be due to intracellular, not extracellular, CA activity.
Binding of CAII to DRA.
The amino acid sequences of the cytoplasmic COOH-terminal regions of HCO transport proteins all possess a potential CAII-binding site (a hydrophobic residue followed by a group of four amino acids, of which at least two must be negatively charged; Ref.46), except for DRA. Although the COOH-terminal tail of DRA does not contain a potential CAII-binding site, we cannot assume that it will not be able to bind to CAII. To investigate the ability of DRA COOH terminus to bind CAII, we constructed a GST fusion protein of the cytoplasmic COOH-terminal region of DRA and measured its capacity to bind CAII.
Figure 3 shows the results of the microtiter assay, which measured binding between GST fusion proteins (GST-AE1ct and GST-DRAct, respectively) and CAII immobilized on a 96-well plate. Both GST-AE1ct and GST-DRAct bound more CAII than did GST alone. However, AE1 bound CAII much more avidly than did DRA. Binding affinities of interactions were K d of 60 nM for DRA and 40 nM for AE1. The maximum binding capacity of DRA was only 25% of AE1 (Fig. 3). This result indicates that the COOH-terminal domain of DRA binds CAII to a much lower extent than AE1.
Effect of carbonic anhydrase on DRA Cl−/HCO exchange activity.
The above experiments suggest that, although DRA requires the presence of functional CAII in the cytosol for transport binding activity, little CAII will bind the DRA COOH-terminal tail. However, CAII may bind another region of DRA. To assess the effect of a DRA-CAII physical interaction upon transport activity, we employed a dominant-negative approach, previously used for AE proteins (42). HEK-293 cells transfected with V143Y CAII, a functionally inactive mutant (10), express the mutant CAII at levels about 20-fold higher than endogenous wild-type CAII (42). V143Y CAII retains its ability to bind to the AE HCO transport proteins (42) so that overexpression of V143Y CAII will compete with the endogenous wild-type CAII at any potential binding site in the cell. Thus, if DRA is activated by a physical interaction with CAII, as are other HCO transport proteins, overexpression of V143Y CAII will decrease the transport activity of DRA.
HEK-293 cells were cotransfected with CAII and DRA cDNAs. Figure4 shows that overexpression of either wild-type CAII protein or V143Y CAII had no effect on DRA transport activity. The lack of effect of overexpression of wild-type CAII suggests that the endogenous level of CAII is sufficient for full DRA transport activity, consistent with results for the AE Cl−/HCO exchangers (42). Because V143Y CAII did not alter DRA transport activity, DRA does not require a direct physical association with CAII for maximal transport activity.
We have found recently that the human AE1 Cl−/HCO anion exchanger also physically interacts with membrane-anchored CAIV, acting as the extracellular component of the bicarbonate transport metabolon (40). The loss of Cl−/HCO anion exchange activity of AE1 caused by V143Y CAII could be rescued by expression of CAIV. However, Fig. 4 shows that coexpression of CAIV did not have any effect on DRA activity. We conclude that, unlike AE1, DRA does not functionally interact with CAIV.
In this study, we have examined the physical and functional relationships between DRA and CAII. The lack of a CAII-binding site motif in the COOH-terminal tail of DRA thus far makes DRA a unique bicarbonate transport protein. To investigate whether CAII activity had any impact on the transport capability of DRA, we monitored DRA-mediated bicarbonate transport activity before and after incubation with 100 μM acetazolamide, a membrane-permeant CA inhibitor (7,8). Our data showed that inhibition of CAII by acetazolamide impaired DRA transport activity, which indicates that CAII activity has a substantial effect on DRA Cl−/HCO exchange activity. Acetazolamide had no direct effect on DRA-mediated Cl−/NO exchange activity. Therefore, the effect of acetazolamide on DRA-mediated Cl−/HCO exchange is an indirect effect due to inhibition of intracellular CAII. Thus DRA bicarbonate transport activity and CAII activity are functionally coupled. Because the membrane-impermeant CA inhibitor SDPPP had no effect on DRA-mediated Cl−/HCO exchange, any endogenously expressed external CA in HEK-293 cells does not affect DRA transport activity.
Binding assays showed that the ability of the DRA COOH-terminal tail to bind CAII is much less than the bicarbonate transporter, AE1. Because the binding assay showed some limited interaction between CAII and DRA, it is possible that interaction with CAII activates DRA. However, overexpression of the V143Y CAII mutant had no dominant-negative effect on DRA transport activity. This indicates that DRA bicarbonate transport activity is not activated by CAII binding at any site on DRA. We conclude that DRA may bind to CAII to some limited degree but that any physical interaction between the two proteins has no impact on the bicarbonate transport activity of DRA.
The bicarbonate transport metabolon is emerging as an important mechanism to accelerate and regulate bicarbonate transport activity. A transport metabolon is a physical complex of an enzyme and a transporter, which maximizes and may regulate substrate flux through the enzyme and across the membrane. The first example of a bicarbonate transport metabolon was provided by anion exchangers of the AE family, which form a functional and physical complex with cytosolic CAII (18, 25, 45) and with the extracellular enzyme, CAIV (40). NBC1 and NBC3 also form a metabolon with CAII [B. Alvarez, F. Loiselle, and J. Casey, unpublished observation (NBC1) and F. Loiselle and J. Casey, unpublished observation (NBC3)]. Comparison of the COOH-terminal tails of other bicarbonate transport proteins indicates that, with the exception of DRA, they all contain at least one potential CAII-binding site. The analysis performed in this report indicates that, consistent with the absence of a CAII-binding motif, binding of CAII or CAIV had no effect on DRA transport activity.
This is the first reported examination of the effect of carbonic anhydrase inhibitors on DRA transport activity. Because carbonic anhydrases and anion exchangers share HCO as a substrate, it is not surprising that both classes of protein are inhibited by compounds directed to the HCO binding site, as originally shown for AE1 (7, 8). The insensitivity of DRA Cl−/NO exchange activity to acetazolamide and SDPPP demonstrates that these CA inhibitors can be used to selectively inhibit CA in the presence of functional DRA. On the other hand, the sensitivity of DRA to the CA inhibitor PTSP shows that care needs to be taken in the selection of CA inhibitors in studies of DRA.
When DRA was first cloned from mouse (23), the Cl−/HCO transport activity of the protein was also characterized. The investigators monitored transport activity in the absence of HCO and found that DRA mediated a slow alkalinization upon removal of extracellular chloride. Cells did not recover from the alkalinization upon reintroduction of extracellular Cl−, which was attributed to a high specificity at the intracellular substrate site for HCO (23). The fact that DRA does not form a transport metabolon with CAII provides another explanation for the lack of recovery of pHi to a physiological level. In the absence of extracellular bicarbonate, readdition of extracellular chloride causes chloride to move into the cell in exchange for bicarbonate. We have shown that tethering CAII near the intracellular anion-binding site of AE1 increases the rate of HCO efflux. Because DRA does not bind CAII, the production of intracellular HCO occurs in the cytosol as opposed to at the site of anion exchange, thus slowing the efflux of HCO and thus decreasing the rate of recovery of pHi.
The data presented here show that DRA is unique among bicarbonate transporters because it does not interact with either CAII or CAIV. One reason for this may be that DRA does not function solely as a Cl−/HCO transport protein, consistent with reports that DRA transports sulfate, oxalate, and chloride (24, 36). Similarly, while investigating Cl−/HCO exchange in apical membrane vesicles from human proximal colon, it was noted that bromide, nitrate, and acetate inhibited uptake of 36Cl−, suggesting that the colonic Cl−/HCO exchange protein DRA might also accept these ions as substrates (22). In the present report, we also found that DRA functioned as a Cl−/NO exchanger. However, there is strong evidence that the physiological role of DRA in the mammalian colon is Cl−/HCO exchange (23).
The lack of interaction between DRA and carbonic anhydrases also suggests that DRA is regulated differently from other bicarbonate transporters. Direct interaction between carbonic anhydrase and a bicarbonate transporter accelerates the bicarbonate transport rate (42). Modulation of the interaction therefore presents a rapid way to alter the rate of bicarbonate transport, but DRA cannot use this mode of regulation. In the physiological context, cells may contain a membrane protein that binds CAII and localizes the enzyme near the DRA transport site. If such a protein exists, HEK-293 cells do not express it because dominant-negative V143Y CAII had no effect on DRA activity.
DRA is expressed on the apical membrane of pancreatic epithelia (12). In response to the release of secretin, the pancreas produces a HCO -rich alkaline fluid, which is secreted to the duodenum to neutralize the acidic chyme produced during digestion. Bicarbonate uptake across the basolateral membrane of pancreatic cells occurs via an NBC (35), whereas bicarbonate flux across the apical membrane may be mediated by DRA (12). CF patients frequently present with a greatly reduced bicarbonate secretion capacity (26). It has been shown recently that expression of DRA is increased in the presence of functional CFTR in both cultured pancreatic and tracheal cells (12, 48). The increase in DRA expression in pancreatic cells was associated with a twofold increase in Cl−/HCO transport activity, and the authors concluded that the decrease in HCO production in CF patients is in part due to downregulation of DRA expression and activity (12, 48). Our data are consistent with a model in which defective pancreatic bicarbonate secretion in CF results from the inability of DRA to bind directly to CAII. Failure to localize CAII close to the DRA bicarbonate transport site could reduce transport efficiency. It could be speculated that an intermediary protein may bring CAII to the vicinity of DRA for optimal DRA bicarbonate transport. This intermediary may be CFTR or some protein regulated by CFTR.
With the exception of DRA, the COOH-terminal tail of every bicarbonate transport protein contains a potential CAII-binding motif. Formation of a complex with CAII potentiates bicarbonate transport activity, and modulation of the interaction provides a cell with a potential regulatory mechanism to control bicarbonate flux (42). DRA is the only bicarbonate transport protein identified to date in which the CAII-binding site is absent. In this study, we investigated the physical and functional interaction between DRA and CAII. On the basis of our results, we conclude that although there is a need for the presence of CAII enzymatic activity in the cytosol, the weak physical interaction between the two proteins does not affect the functional activity of DRA. For full transport activity, DRA may require an intermediary protein to bind CAII and bring the enzyme close to DRA in the plasma membrane. This makes DRA unique among the bicarbonate transport protein superfamily examined to date.
We thank Dr. George Schwartz for CAIV cDNA.
This work was funded by an operating grant from the Heart and Stroke Foundation (HSF) of Canada. D. Sterling holds studentship trainee awards from the HSF and Alberta Heritage Foundation for Medical Research (AHFMR). N. Brown was supported by an AHFMR summer studentship. J. Casey is a Senior Scholar of AHFMR.
Address for reprint requests and other correspondence: J. R. Casey, Dept. of Physiology, Univ. of Alberta, Edmonton, Alberta, Canada T6G 2H7 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
July 24, 2002;10.1152/ajpcell.00115.2002
- Copyright © 2002 the American Physiological Society