Human pendrin expressed in Xenopus laevis oocytes mediates chloride/formate exchange

Daryl A. Scott, Lawrence P. Karniski

Abstract

Pendred syndrome, characterized by congenital sensorineural hearing loss and goiter, is one of the most common forms of syndromic deafness. The gene causing Pendred syndrome (PDS) encodes a protein designated pendrin, which is expressed in the thyroid, kidney, and fetal cochlea. Pendrin functions as an iodide and chloride transporter, but its role in the development of hearing loss and goiter is unknown. In this study, we examined the mechanism of pendrin-mediated anion transport inXenopus laevis oocytes. Unlabeled formate added to the uptake medium inhibited pendrin-mediated 36Cl uptake in X. laevis oocytes. In addition, the uptake of [14C]formate was stimulated in oocytes injected with PDS cRNA compared with water-injected controls. These results indicate that formate is a substrate for pendrin. Furthermore, chloride stimulated the efflux of [14C]formate and formate stimulated the efflux of 36Cl in oocytes expressing pendrin, results consistent with pendrin-mediated chloride/formate exchange. These data demonstrate that pendrin is functionally similar to the renal chloride/formate exchanger, which serves as an important mechanism of chloride transport in the proximal tubule. A similar process could participate in the development of ion gradients within the inner ear.

  • Pendred syndrome
  • pendrin
  • chloride
  • formate

pendred syndrome was first described by Vaughan Pendred in 1896 (15) and is characterized by congenital sensorineural hearing loss and goiter. Pendred syndrome is one of the most common forms of syndromic deafness and may account for as much as 10% of all hereditary deafness (2, 16). It has been estimated that 86% of patients with Pendred syndrome will have structural malformations of the cochlea identifiable by computerized tomography (16). The gene mutated in Pendred syndrome (PDS) encodes a transmembrane protein designated pendrin. PDS expression has been detected in the thyroid and kidney by Northern blot analysis and in a fetal cochlear cDNA library by PCR analysis (7). Recent observations thatPDS mutations may result in deafness in the absence of other syndromic features suggest that hearing loss associated withPDS mutations may be underestimated (12).

Despite significant amino acid sequence homology with a family of sulfate transport proteins, expression of pendrin in Xenopus laevis oocytes and Sf9 cells demonstrates that it does not transport sulfate but functions instead as a sodium-independent transporter of chloride and iodide (19). Defects in iodide transport could explain the thyroid abnormalities observed in Pendred syndrome; however, the mechanism resulting in hearing loss in patients with Pendred syndrome is unknown.

The following studies were performed to characterize the function of pendrin. We report that pendrin has transport properties similar to that of the renal chloride/formate exchanger, which participates in the reabsorption of filtered chloride in the proximal tubule (9, 18, 20). A defect in the chloride transport properties of pendrin in the inner ear could contribute to the hearing loss associated with Pendred syndrome.

METHODS

Isolation of Xenopus laevis oocytes. The PDS coding sequence, spanning nucleotides 207–2564 was amplified by PCR from a human thyroid gland Quick-Clone cDNA library (Clontech). The generation of the PCR product, its cloning into a modified pGEM vector, and synthesis of capped cRNA have been described previously (19). Capped sat-1 cRNA was synthesized using a previously described clone as a template (11). Lobes of oocytes were dissected from femaleX. laevis frogs that had been anesthetized with 0.2% 3-aminobenzoic acid ethyl ester (tricaine). Oocytes for expression were prepared by gentle shaking for 1 h in 10 ml of a 1 mg/ml collagenase solution prepared in normal frog Ringer (NFR), consisting of 115 mM NaCl, 2.5 mM KCl, 1.8 mM CaCl2, 5 mM sodium pyruvate, 0.1 mg/ml gentamicin, and 10 mM HEPES, pH 7.35 with NaOH. Oocytes were washed once in calcium-free NFR, followed by two final washes in NFR. Oocytes were injected with a 50-nl volume of either water as the control or capped PDS cRNA in water using a NANOJECT microinjector (Drummond Scientific). After injection, the oocytes were incubated at 16°C for ∼72 h to allow for protein expression.

Transport assays. For uptakes, oocytes were washed at room temperature in 1.0 ml chloride-free buffer [115 mM sodium gluconate, 2.0 mM potassium gluconate, 2.5 mM Ca(OH)2, 11.0 mM HEPES/Tris, pH 7.35] and then incubated in 600 μl of uptake solution containing the radioisotopes to be tested. The composition of the uptake solutions are described in the figure legends. After timed incubation, oocytes were washed three times with 2 ml of ice-cold chloride-free buffer to remove unincorporated isotope. Individual oocytes were solubilized in 0.2 ml of 10% SDS and added to 3 ml scintillation fluid for measurement of radioisotope uptake by scintillation spectroscopy.

For efflux measurements, oocytes that had been injected withPDS cRNA were first preloaded with radioisotope by incubating for 2 h in chloride-free buffer containing either [14C]formate or 36Cl. Before the initiation of efflux measurements, a group of 8–10 oocytes was washed three times with 2 ml of ice-cold chloride-free buffer, and the amount of isotope in each oocyte was quantitated by scintillation spectroscopy as described above. This zero time point represents radioisotope within the oocytes at the start of the efflux experiment. To initiate efflux, the remaining oocytes were incubated in 1.0 ml of efflux buffer, with the composition of the efflux buffers described in the legend for Fig. 3. The quantity of radioisotope remaining in each oocyte at various time points was measured and compared with the zero time point to calculate the percent efflux of isotope over time.

For determination of the apparent K m, the effects of increasing concentrations of substrate on the rate of transport were analyzed using nonlinear regression (GraphPad Prism, version 3.0). Statistical significance was calculated using the two-tailed Student'st-test.

Materials. H36Cl and [14C]sodium formate were from ICN. [14C]oxalic acid, and collagenase type IA were purchased from Sigma. Scintillation cocktail (3a70B) was purchased from Research Products International.

RESULTS

Pendrin has been identified as a chloride and iodide transport protein in both the X. laevis oocyte and baculovirus expression systems (19). To identify anions that might serve as substrates for pendrin and participate in anion exchange, we examined the ability of various anions to inhibit the uptake of 36Cl in oocytes expressing pendrin. As expected, 5 mM concentrations of unlabeled chloride and iodide significantly inhibited 36Cl uptake, consistent with their known roles as substrates for pendrin (Fig.1). The inorganic anions nitrate and bromide also inhibited the uptake of 36Cl, whereas sulfate did not. Of the organic anions tested, only formate was able to significantly inhibit 36Cl uptake. The order of potency for inhibition of pendrin-mediated 36Cl uptake was iodide > nitrate = formate > bromide > chloride.

Fig. 1.

Cis-inhibition of pendrin-mediated chloride uptake. The 1-h uptake of chloride was measured in oocytes injected with 0.35 ng Pendred syndrome gene (PDS) cRNA. The uptake solutions contained 113 mM sodium gluconate, 1.9 mM potassium gluconate, 2.3 mM Ca(OH)2, 5 mM [36Cl]tetramethylammonium chloride, 10 mM HEPES/Tris (pH 7.35) as the control or with gluconate replaced by 5 mM of various test anions. The results are expressed as percent of control uptake and represent means ± SE of 3–4 experiments for each anion tested. Each experiment represents the average flux measurement in 7–10 oocytes isolated from a single frog. Bars extending below the x-axis indicate stimulation of 36Cl uptake compared with controls.

The inhibition of pendrin-mediated chloride uptake by formate suggests that formate can serve as a substrate for pendrin. To examine whether formate can be transported via pendrin, we measured the rate of [14C]formate uptake in oocytes injected with either water or PDS cRNA. As shown in Fig.2 A, the rate of formate uptake in oocytes injected with PDS cRNA is enhanced compared with water-injected controls, and the rate of uptake is stimulated by increasing the amount of injected cRNA. In addition, the 2 h uptake of [14C]formate in oocytes expressing pendrin was inhibited 84% when 5 mM unlabeled chloride was added to the uptake medium (Fig. 2 B). Taken together, these results are most consistent with formate serving as a substrate for the pendrin anion transporter. Saturability of the rate of pendrin-mediated chloride and formate transport was observed with increasing concentrations of substrate (data not shown). The apparentK m values for chloride and formate are 2.5 ± 0.2 and 0.58 ± 0.2 mM, respectively.

Fig. 2.

Pendrin-mediated formate uptake. A: uptake of 10 μM [14C]formate in chloride-free uptake solution was determined at various time points in oocytes injected with either 50 nl water or PDS cRNA in water. Each value represents the mean ± SE of 3 experiments. B: 2-h uptake of 10 μM [14C]formate in chloride-free buffer (−Cl) or with the isosmotic replacement of sodium gluconate by 5 mM NaCl (+Cl) was measured in oocytes injected with either water or 0.35 ng PDS cRNA. Each value represents mean ± SE of 18–20 oocytes isolated from 2 different frogs. P < 0.01, significant difference between PDS (−Cl) vs. PDS (+Cl).

In the kidney, ∼60% of filtered chloride is reabsorbed in the proximal tubule, and a significant portion of this chloride transport is mediated by a chloride/formate exchanger located on the apical membrane of the proximal tubule cell (1, 20). To determine whether pendrin can also mediate chloride/formate exchange, we examined whether external formate can stimulate the rate of chloride efflux and whether external chloride can stimulate the rate of formate efflux.

To determine the effects of external formate on chloride efflux,36Cl was preloaded in oocytes expressing pendrin, and the rate of 36Cl efflux was measured during incubation in a solution containing either formate or acetate. Acetate is a short-chain fatty acid similar to formate, but it is not a substrate for pendrin, as indicated by its inability to inhibit pendrin-mediated36Cl uptake (Fig. 1). As shown in Fig.3, the rate of chloride efflux from oocytes expressing pendrin is 2.1-fold greater when formate is added to the efflux medium compared with acetate. We also examined the rate of [14C]formate efflux from oocytes expressing pendrin during incubation in a solution containing either chloride or gluconate. As illustrated in Fig. 3, the rate of formate efflux is 2.8-fold greater in the presence of external chloride compared with gluconate controls. There was no significant difference in the rate of36Cl efflux in water-injected oocytes between external acetate vs. external formate conditions or in the rate of [14C]formate efflux in control oocytes between external gluconate vs. chloride conditions (data not shown). The observations that external formate stimulates chloride efflux and external chloride stimulates formate efflux from oocytes expressing pendrin is consistent with pendrin-mediated chloride/formate exchange.

Fig. 3.

Pendrin-mediated chloride and formate efflux. Oocytes injected with 0.35 ng PDS cRNA were incubated for 2 h in either 1.7 mM [36Cl]tetramethylammonium chloride or 10 μM [14C]sodium formate. Rate of chloride efflux was determined by comparing the amount of 36Cl in oocytes before the initiation of efflux (zero time point) to the amount of36Cl remaining in oocytes incubated for 30 min in chloride-free buffer containing either 50 mM sodium acetate or 50 mM sodium formate. Results represent mean ± SE of 7 experiments. Rate of formate efflux was determined by comparing the amount of [14C]formate in oocytes at the zero time point to the amount of [14C]formate remaining in oocytes following incubation for 15 min in buffer containing either 50 mM sodium gluconate or 50 mM NaCl. Results represent mean ± SE of 5 experiments. Each experiment represents average flux measurement in 6–8 oocytes isolated from a single frog. P < 0.05, significant difference within each group (acetate vs. formate; gluconate vs. chloride).

Two different anion transporters that accept formate as a substrate have been described in rat and rabbit kidneys (10, 20) and are labeled the chloride(formate)/oxalate exchanger and the chloride/formate exchanger. Under the appropriate conditions, both transporters can mediate chloride/formate exchange; however, the renal chloride(formate)/oxalate exchanger accepts oxalate as a substrate, whereas the renal chloride/formate exchanger does not. Because of the similarity between pendrin and the two renal chloride/formate exchangers, we examined whether pendrin can transport oxalate in addition to chloride and formate. As shown in Fig.4, there is no significant difference in [14C]oxalate uptake between oocytes injected with pendrin cRNA compared with water-injected controls. In contrast, the expression of sat-1, which is known to be an oxalate transport protein (11), results in a significant stimulation of [14C]oxalate uptake. These results demonstrate that the expression of oxalate transport in oocytes can be detected under the test conditions and that the absence of pendrin-mediated oxalate transport indicates that oxalate is not a substrate for pendrin.

Fig. 4.

Oxalate uptake in PDS-injected oocytes. The 1-h uptake of 352 μM [14C]oxalate in chloride-free uptake buffer was measured in oocytes injected with either water, 0.5 ngPDS,or 0.5 ng sat-1 cRNA. Results are expressed as mean ± SE of 4 experiments, with each experiment representing average flux measurement in 6–8 oocytes isolated from a single frog. P< 0.01, significant difference between sat-1 injected oocytes and water-injected controls.

DISCUSSION

In this study we have demonstrated that pendrin does not accept acetate, lactate, succinate, and oxalate as substrates but is able to mediate chloride/formate exchange. Interestingly, the selectivity for formate among carboxylic acids is similar to the selectivity observed for the renal chloride/formate exchanger (10). Kinetic parameters for the renal chloride/formate exchanger have not been reported; however, nonlinear regression analysis of previously published data (10) on the effects of increasing concentrations of chloride on chloride/formate exchange in rabbit renal microvillus membrane vesicles yields an apparent K m for chloride of 2.8 mM. This is nearly identical to the value of 2.5 mM reported in this study for pendrin. Furthermore, the concentrations of the anion exchange inhibitors furosemide, DIDS, and probenecid needed to achieve 50% inhibition of the renal chloride/formate exchanger (10) are similar to those previously reported for pendrin (19). These data demonstrate that pendrin is functionally related to the renal chloride/formate exchanger.

In the proximal tubule of the kidney, intracellular formate drives the energetically uphill accumulation of chloride into the cell across the apical membrane. Chloride then exits across the basolateral membrane via one or more transport pathways (1). To maintain sufficient intracellular formate concentrations for the reabsorption of large quantities of filtered chloride, pathways for the recycling of formate across the proximal tubule apical membrane have been described (9, 17). Chloride/formate exchange, in parallel with the Na/H exchanger, provides a mechanism for NaCl and volume reabsorption in the proximal tubule of the kidney (18, 20) and may provide insight into the role of pendrin in the inner ear.

The composition of the endolymph of the inner ear is low in sodium but high in potassium and chloride relative to plasma and the perilymph. The generation of these ion gradients is accomplished by multiple ion transporters located in various cell types within the inner ear. Mutations in the genes encoding ion transporters expressed in the inner ear, including the KVLQT1 potassium channel (14), the secretory Na+-K+-2Cl cotransporter (6), and the H+-ATPase (8) each result in hearing loss, indicating that maintenance of the normal electrolyte composition of the inner ear is critical for normal hearing.

In Pendred syndrome, the abnormal morphology of the inner ear ranges from dilation of the vestibular aqueducts to Mondini malformations, where the apical turns of the cochlea form a common cavity. While the inner ear abnormalities associated with PDS mutations could be developmental in nature, a loss of chloride transport within the inner ear could lead to abnormal salt and water flux, with subsequent dilation of the vestibular aqueduct and loss of the normal architecture of the cochlea. The fact that the renal chloride/formate exchanger performs a significant role in proximal tubule volume reabsorption suggests that a similar process could participate in the maintenance of the electrolyte gradients and volume homeostasis within the inner ear.

The mechanism of the thyroid abnormalities observed in Pendred syndrome is unknown but is presumably related to the iodide transport properties of pendrin. Mutations that would lead to a loss of iodide transport across the apical membrane of the thyrocyte would explain the organification defects observed in individuals with Pendred syndrome, where iodide is taken up normally in thyrocytes across the basolateral membrane but is incompletely bound to thyroglobulin in the colloid (4,5, 13).

It is interesting to note that pendrin expression has been detected only in the thyroid, inner ear, and kidney (7). Although Pendred syndrome is defined by abnormalities in the thyroid and inner ear, renal abnormalities have not been described in patients with Pendred syndrome. Given the functional similarities between pendrin and the rabbit and rat renal chloride/formate exchangers, pendrin could provide the mechanism for chloride/formate exchange in the human proximal tubule. Although it is possible that chloride/formate exchange does not play a significant role in chloride reabsorption in the human kidney, clinically significant defects in chloride transport may be difficult to detect in Pendred syndrome due to the redundancy of chloride transport processes along the nephron. Alternatively, the renal chloride/formate exchanger could be an isoform of pendrin, or it may it may be a structurally unrelated protein, with transport properties similar to that of pendrin.

In conclusion, we have demonstrated that pendrin is functionally similar to the renal chloride/formate exchanger. These transport properties may provide an explanation for the role of pendrin in the inner ear, thyroid, and kidney where it is expressed.

Acknowledgments

This work was supported by the Office of Research and Development, Dept. of Veterans Affairs, by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-47881, and in part by the March of Dimes Birth Defects Foundation Grant FY99-507.

Footnotes

  • Address for reprint requests and other correspondence: L. P. Karniski, Dept. of Internal Medicine, Univ. of Iowa Hospitals, 200 Hawkins Dr., Iowa City, IA 52242 (E-mail: lawrence-karniski{at}uiowa.edu).

  • 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. §1734 solely to indicate this fact.

REFERENCES

View Abstract