Phylogentically, organic anion transporter (OAT)1 and OAT3 are closely related, whereas OAT2 is more distant. Experiments with human embryonic kidney-293 cells stably transfected with human OAT1, OAT2, or OAT3 were performed to compare selected transport properties. Common to OAT1, OAT2, and OAT3 is their ability to transport cGMP. OAT2 interacted with prostaglandins, and cGMP uptake was inhibited by PGE2 and PGF2α with IC50 values of 40.8 and 12.7 μM, respectively. OAT1 (IC50: 23.7 μM), OAT2 (IC50: 9.5 μM), and OAT3 (IC50: 1.6 μM) were potently inhibited by MK571, an established multidrug resistance protein inhibitor. OAT2-mediated cGMP uptake was not inhibited by short-chain monocarboxylates and, as opposed to OAT1 and OAT3, not by dicarboxylates. Consequently, OAT2 showed no cGMP/glutarate exchange. OAT1 and OAT3 exhibited a pH and a Cl− dependence with higher substrate uptake at acidic pH and lower substrate uptake in the absence of Cl−, respectively. Such pH and Cl− dependencies were not observed with OAT2. Depolarization of membrane potential by high K+ concentrations in the presence of the K+ ionophore valinomycin left cGMP uptake unaffected. In addition to cGMP, OAT2 transported urate and glutamate, but cGMP/glutamate exchange could not be demonstrated. These experiments suggest that OAT2-mediated cGMP uptake does not occur via exchange with monocarboxylates, dicarboxylates, and hydroxyl ions. The counter anion for electroneutral cGMP uptake remains to be identified.
- basolateral membrane
- organic anion transporters
- organic anion transporter 1
- organic anion transporter 2
- organic anion transporter 3
several members of the “solute carrier (SLC)22” family have been functionally characterized, comprising organic cation transporters (OCTs; OCT1–OCT3 and OCT6), organic zwitterion-cation transporters (OCTN; OCTN1 and OCTN2), and organic anion transporters (OATs; OAT1–OAT7, OAT10, and urate transporter 1) (summarized in Refs. 22 and 27). Common to all these transporters is their broad substrate specificity. Human OATs handle small, amphiphilic organic anions, uncharged molecules, and even some organic cations (for reviews, see Refs. 4, 27, 28, 31, and 40). OAT1, OAT2, and OAT3 show amino acid sequence identities between 38% and 51% (16, 38). The gene for human OAT2 (SLC22A7) is located on chromosome 6p21.2-1 (23) and is not paired with any other gene from the SLC22 family (8). In contrast, OAT1 and OAT3 are coded by tandem genes on chromosome 11q12.3 (30). OAT2 is highly expressed in the liver (36, 38), but its subcellular localization remains unclear. In the kidneys, OAT2 has been localized to the basolateral membrane of human kidney proximal tubule cells (7, 9), similar to OAT1 and OAT3 (26). Whereas OAT1 (24) and OAT3 (1, 37) work as organic anion/dicarboxylate exchangers, the driving force for OAT2 is unclear.
PGF2α (8, 19, 20, 25), estrone-3-sulfate (ES) (20, 21), cGMP (7, 8), uric acid (33), orotic acid (10, 29), and glutamate (29, 35) have been identified as substrates of human OAT2 expressed in Xenopus laevis oocytes (21), renal S2 cells (9, 19), and human embryonic kidney (HEK)-293 cells (7, 8, 10, 29, 35).
In the present report, we studied, in a comparative fashion, the influence of dicarboxylates, monocarboxylates, urate, and glutamate, the effects of extracellular pH, and interactions with prostaglandins on OAT1, OAT2, and OAT3 in stably transfected HEK-293 cells. Finally, we introduce MK571 as a high-affinity inhibitor for OAT1, OAT2, and OAT3.
MATERIALS AND METHODS
Reagents and chemicals.
Chemicals, including the constituents of mammalian Ringer solution (MRi), were from Applichem (Darmstadt, Germany). The antibiotics penicillin and streptomycin, the ionophore valinomycin, monocarboxylates and dicarboxylates, sodium, potassium, and calcium gluconate, tetraethylammonium chloride, and the compounds shown in Table 1 were purchased from Sigma-Aldrich (Taufkirchen, Germany) with the exception of MK571, which was from Tocris (Wiesbaden-Nordenstadt, Germany). [3H]cGMP, [3H]ES, l-[3H]glutamate, p-[3H]aminohippurate (PAH), and [14C]urate were from Perkin-Elmer (Rodgau, Germany).
Cell culture and transport experiments.
HEK-293 cells stably transfected with OAT1, OAT2, OAT3, or the empty pcDNA5-vector (12, 17) were grown in high-glucose DMEM (Life Technologies, Darmstadt, Germany) supplemented with 10% FCS (European Union approved origin, Life Technologies) and 1% penicillin-streptomycin. Cells were harvested and plated into 24-well polylysine-coated plastic dishes (Sarstedt, Nümbrecht, Gemany) at a density of 2 × 105 cells/well. After 72 h of incubation, cells were washed twice with 0.5 ml MRi containing (in mM) 130 NaCl, 4 KCl, 1 CaCl2, 1 MgSO4, 1 NaH2PO4, 20 HEPES, and 18 glucose at pH 7.4. Cells were incubated at 37°C in MRi that contained either 10 μM cGMP (10 nM [3H]cGMP plus 9.99 μM unlabeled cGMP), 20 nM or 100 nM l-[3H]glutamate, 0.5 μM [3H]PAH, 10 nM [3H]ES, or 10 μM [14C]urate in the absence or presence of monocarboxylates, dicarboxylates, glutamate, or prostaglandins at varying extracellular pH, Cl− substitution, Na+ substitution, and in the presence and absence of MK571. If test compounds were dissolved in DMSO (prostaglandins and MK571) or ethanol (valinomycin), controls also contained DMSO and ethanol to achieve comparable conditions. In up to 1% DMSO and 0.001% ethanol, no difference in uptake of the reference substrates (control) was observed. Uptake was terminated after 5 min by removal of the radiolabeled medium and immediate threefold washes with ice-cold MRi. Cells were dissolved in 0.5 ml of 1 N NaOH by gently shaking for 120 min, and the 3H or 14C content was determined by liquid scintillation counting (Tricarb 2900TR, Perkin-Elmer).
In cis-inhibition (competition) experiments, the uptake into vector-transfected HEK-293 cells was subtracted from the uptake into transporter-transfected cells to obtain the transporter-mediated uptake. For trans-stimulation experiments, cells were incubated at 37°C for 120 min in MRi containing 2 mM of the indicated monocarboxylates, dicarboxylates, PAH, cGMP, ES, or 2 and 5 mM glutamate or for 30 min in MRi containing PGE2 and MK571 at concentrations of 200 and 100 μM, respectively. Protein concentration was determined using the Bradford protein assay (3).
RNA extraction, reverse transcription, and real-time PCR assay.
Total RNA was extracted from vector-transfected HEK-293 cells using the RNeasy Mini Kit (Qiagen, Hilden, Germany). Two micrograms of each RNA preparation were reverse transcribed using a cDNA synthesis kit (Applied Biosystems, Forster City, CA). Real-time PCR was performed with a 4.5-μl aliquot of the total cDNA sample using TaqMan Gene Expression Assays (Life Technologies) for the following genes of interest: multidrug resistance protein (MRP)2 (Hs00166123_m1), MRP4 (Hs00988717_m1), OAT1 (Hs00537914_m1), OAT2 (Hs00198527_m1), and OAT3 (Hs00188599_m1). Human hypoxanthine phosphoribosyltransferase (HPRT)1 and GAPDH were used as an internal control for sample normalization. The conditions for real-time PCR were as follows: 2 min at 50°C followed by 10 min at 95°C and 40 amplification cycles (95°C for 15 s and 60°C for 60 s) using a Mx3000P real-time PCR cycler (Agilent Technologies, Waldbronn, Germany). Genes not detected within 40 amplification cycles were taken as being not expressed in HEK-293 cells. Relative levels of OAT1, OAT2, OAT3, MRP2, and MRP4 mRNAs were calculated using the following comparative threshold cycle (Ct) method: ΔCt = HRPT1 − gene of interest.
Statistics and calculations.
Data are provided as means ± SE. Statistical significance as obtained from unpaired Student's t-tests was set at P < 0.001. IC50 values were calculated using SigmaPlot software (Systat Software, Point Richmond, CA).
Interaction of urate and cGMP with OAT1, OAT2, and OAT3.
Previous reports have identified urate and cGMP as possible substrates for OAT1, OAT2, and OAT3 (7, 23, 35). In parallel setups, the impact of urate (each 0.5 mM) was tested on OAT1-, OAT2-, and OAT3-transfected HEK-293 cells (Fig. 1A). To proof for transporter-specific uptake, the established inhibitors of OAT1 (probenecid), OAT2 (indomethacin), and OAT3 (unlabeled ES), each at a concentration of 0.1 mM, were applied. Probenecid and urate inhibited PAH uptake in OAT1-transfected HEK-293 cells by 85.3 ± 1.6% and 26.6 ± 3.0%, respectively. In OAT2-transfected HEK-293 cells, indomethacin, but not urate, inhibited cGMP uptake by 95.1 ± 0.6%. The uptake of ES in OAT3-transfected HEK-293 cells was not altered significantly by urate. Probenecid- and indomethacin-sensitive urate uptake was observed only in OAT1- and OAT2-transfected HEK-293 cells but not in OAT3-transfected HEK-293 cells (Fig. 1B). Urate uptake was 1.90 ± 0.26, 2.58 ± 0.19, and 0.28 ± 0.10 pmol·mg−1·min−1 in OAT1-, OAT2-, and OAT3-transfected HEK-293 cells and decreased to 0.16 ± 0.45, 0.42 ± 0.29, and 0.17 ± 0.28 pmol·mg−1·min−1 in the presence of 100 μM probenecid (OAT1), 100 μM indomethacin (OAT2), and 100 μM probenecid (OAT3), respectively.
Recently, cGMP was identified as a reference substrate of OAT2 (7, 8). To test whether also OAT1 and OAT3 transport cGMP, the uptake of radiolabeled cGMP into OAT1-, OAT2-, and OAT3-transfected HEK-293 cells was measured (Fig. 2A). cGMP uptake was 38.3 ± 2.1, 63.7 ± 20.6, and 14.5 ± 1.3 pmol·mg−1·min−1 in OAT1-, OAT2-, and OAT3-transfected HEK-293 cells, respectively, and was decreased by the inhibitors probenecid (OAT1), indomethacin (OAT2), and unlabeled ES (OAT3) to 3.5 ± 0.1, 3.7 ± 0.7, and 3.3 ± 0.8 pmol·mg−1·min−1, respectively, indicating that, besides OAT2, OAT1 and OAT3 also transport cGMP.
To investigate whether OAT1-, OAT2, and OAT3-transfected HEK-293 cells can mediate PAH/cGMP, cGMP/cGMP, and ES/cGMP exchange, cells were incubated for 120 min without (control) or with 2 mM PAH, cGMP, or ES, respectively. Preincubation with cGMP or PAH significantly increased OAT1-mediated PAH uptake by 153.5 ± 11.9% and 173.7 ± 37.2%, whereas preincubation with ES decreased PAH uptake (Fig. 2B, OAT1). OAT2-mediated cGMP uptake was reduced after preincubation with cGMP and ES and unchanged with PAH (Fig. 2B, OAT2). The uptake of ES in OAT3-transfected cells was slightly but significantly higher after preincubation with cGMP and PAH but not with ES (Fig. 2B, OAT3).
To test whether OAT2 was able to transport, besides cGMP, also PAH or ES, the uptake of radiolabeled PAH and ES was determined in OAT2-transfected HEK-293 cells in the absence and presence of the OAT2-specific inhibitor indomethacin (Fig. 2C). Uptake of cGMP was 81.4 ± 8.3 pmol·mg−1·min−1 in the absence and 4.9 ± 1.0 pmol·mg−1·min−1 in the presence of 100 μM indomethacin. In contrast, the uptake of PAH and ES was very low and not affected by indomethacin. If OAT2 works as a uniporter, as recently suggested by Fork et al. (10) and Pfennig et al. (29), the uptake of cGMP should increase with depolarization of the cells. Raising the extracellular K+ concentration from 4 to 20 and 40 mM in the absence and presence of valinomycin, resulting in a calculated depolarization of up to 60 mV at 40 mM K+, did not significantly increase cGMP uptake. cGMP uptake was 63.6 ± 2.8 at 4 mM and changed to 69.8 ± 4.4 pmol·mg−1·min−1 at 40 mM in the absence of valinomycin and from 63.6 ± 3.6 to 72.2 ± 5.5 pmol·mg−1·min−1 in the presence of 100 nM valinomycin, respectively (Fig. 2D).
Interaction of OAT1, OAT2, and OAT3 with di- and monocarboxylates.
In a previous study (17), we showed inhibition of human OAT1 and OAT3 by dicarboxylates. Based on these data, glutarate (5 carbons), adipate (6 carbons), pimelate (7 carbons), and suberate (8 carbons) inhibited the uptake of PAH in OAT1-transfected cells by >85% (Fig. 3A, solid bars) and the uptake of ES in OAT3-transfected cells by 77.5 ± 4.3%, 66.8 ± 10.3%, 55.5 ± 22.1%, and 51.6 ± 18.8% (Fig. 3A, open bars), respectively. In contrast, these dicarboxylates did not affect cGMP uptake into OAT2-transfected cells (Fig. 3A, shaded bars) with the exception of suberate, which slightly inhibited uptake of cGMP by 8.5 ± 4.9%. In addition, the physiological substrate α-ketoglutarate (α-KG), which inhibited the uptake of PAH and uptake of ES by OAT1 and OAT3 with IC50 values of 4.7 ± 1.2 and 92.8 ± 33.6 μM, respectively (17), had no impact on cGMP uptake in OAT2-transfected HEK-293 cells: cGMP uptake in the absence and presence of 1 mM α-KG was 59.3 ± 0.7 and 59.0 ± 1.0 pmol·mg−1·min−1, respectively (three independent experiments; data not shown). Preincubation with either glutarate or α-KG failed to trans-stimulate OAT2-mediated uptake of cGMP: uptake was 60.3 ± 9.9 pmol·mg−1·min−1 in control cells, 63.0 ± 10.3 pmol·mg−1·min−1 in glutarate-preloaded cells, and 60.8 ± 10.2 pmol·mg−1·min−1 in α-KG-preloaded cells (Fig. 3B).
In Xenopus laevis oocytes expressing mouse Oat2, uptake of PGE2 was inhibited by the monocarboxylates propionate and butyrate (15). In the same study, propionate was identified as a substrate of mouse Oat2. With regard to human OAT2, butyrate and lactate did not affect the uptake of uric acid in OAT2-transfected HEK-293 cells (33). Using similar experimental conditions, none of the tested monocarboxylates, acetate, propionate, pyruvate, or butyrate, significantly affected OAT2-mediated cGMP uptake (Fig. 3C). Introduction of a hydroxyl group (lactate, 3-hydroxy butyrate, and glycolate) also failed to produce a significant inhibition of cGMP uptake. After preincubation with either 2 mM butyrate, 3-hydroxy butyrate, or lactate, cGMP uptake increased from 49.4 ± 7.5 to 58.5 ± 3.7, 65.6 ± 3.2, and 51.3 ± 3.7 pmol·mg−1·min−1, but this trans-stimulation was not statistically significant in three to four observations on different cell passages (data not shown).
Effect of extracellular pH and extracellular Cl− on OAT1, OAT2, and OAT3.
Within the range from pH 5.4 to 8.9, the uptake of PAH by OAT1 as well as the uptake of ES by OAT3 was higher at acidic than at alkaline extracellular pH with a steep increase between pH 7.4 and 6.4 (Fig. 4A). Uptake saturated at an extracellular pH < 5.9 and did not further decrease at pH > 7.9. Uptake at pH 6.4 exceeded the uptake at pH 7.4 by a factor of 1.8 (OAT1) and 3.8 (OAT3). The proton concentration at half-maximal substrate uptake for OAT1 and OAT3 was calculated to be 79 ± 12 and 117 ± 29 nM, respectively. In contrast, cGMP-mediated uptake into OAT2-transfected cells was only marginally affected by changes in extracellular pH (Fig. 4A). To investigate whether Cl− affects transport by OATs, all Cl− was substituted by gluconate, and the uptake of PAH in OAT1-transfected HEK-293 cells, uptake of cGMP in OAT2-transfected HEK-293 cells, and uptake of ES in OAT3-transfected HEK-293 cells were measured (Fig. 4B). Uptake of PAH during Cl− substitution by gluconate was reduced to values similar to those in the presence of probenecid (Fig. 4B, OAT1). Uptake of cGMP was not influenced by gluconate (Fig. 4B, OAT2), and uptake of ES was inhibited by 44.7 ± 8.2% during Cl− substitution by gluconate (Fig. 4B, OAT3).
Interaction of OAT1, OAT2, and OAT3 with glutamate.
Fork et al. (10) and Pfennig et al. (29) reported efflux of glutamate driven by extracellular orotate or benzoate and a three- to fourfold higher uptake of glutamate in OAT2-transfected over vector-transfected cells. In previous experiments (12), we demonstrated only a moderate interaction of OAT1 and no interaction of OAT3 with glutamate. OAT2 was not tested in these experiments. Under Na+-free conditions to block Na+-coupled glutamate uptake, a small probenecid-sensitive glutamate uptake was observed in OAT1-transfected cells (Fig. 5A, OAT1, shaded bars). Indomethacin-sensitive glutamate uptake by OAT2 was more pronounced (Fig. 5A, OAT2, shaded bars), whereas no ES-sensitive glutamate uptake was found in OAT3-transfected HEK-293 cells (Fig. 5A, OAT3, shaded bars).
Since OAT2-transfected HEK-293 cells accumulated glutamate two- to three-fold over vector-transfected cells, we investigated whether glutamate can trans-stimulate the uptake of cGMP by preloading OAT2- and vector-transfected cells with glutamate. In three independent cell preparations, cGMP uptake was 45.3 ± 4.4 pmol·mg−1·min−1 without glutamate and 49.7 ± 5.4 pmol·mg−1·min−1 with glutamate (Fig. 5B), and higher glutamate concentration did not improve the result (data not shown). Thus, there was no detectable cGMP/glutamate exchange. Simultaneous application of glutamate (1 mM) and cGMP significantly inhibited cGMP uptake by 35.2 ± 4.1% (data not shown).
Effect of prostaglandins and MK571.
In several studies (9, 13, 20, 21, 36), PGE2, PGI2, and PGF2α have been identified as substrates of OAT1, OAT2, and OAT3. PGE2, PGF2α, and PGI2 inhibited cGMP uptake by OAT2, respectively (Fig. 6A). Indomethacin reduced cGMP uptake to values observed in vector-transfected cells. For PGE2 and PGF2α, IC50 values of 40.8 ± 7.0 and 12.7 ± 0.8 μM were determined (Fig. 6B). The IC50 value for PGI2 approached values of >500 μM. Attempts to demonstrate an increased cGMP uptake by preincubation of OAT2-transfected HEK-293 cells with PGE2 and PGF2α for 30 min showed controversial results: whereas PGF2α (100 μM in 0.2% DMSO) slightly increased the uptake of cGMP from 50.5 ± 4.4 to 58.6 ± 10.1 pmol·mg−1·min−1, PGE2 (200 μM in 0.4% DMSO) decreased cGMP uptake from 62.4 ± 11.5 to 50.1 ± 6.0 pmol·mg−1·min−1. These changes in cGMP uptake were not statistically different (data not shown).
As reported by others (6) using Western blots and by us using real-time PCR experiments (Fig. 7A), HEK-293 cells possess endogenous MRP2 and MRP4. The amount of mRNA normalized to the housekeeping genes HPRT1 (Fig. 7A) or GAPDH (data not shown) was higher for MRP4 than for MRP2. The mRNA for OAT1, OAT2, and OAT3 was below the detection limit (>40 cycles).
MRP2 and MRP4 may extrude the prostaglandins accumulated by OAT2. Therefore, MK571 was used as an MRP inhibitor (for a review, see Ref. 18). However, MK571 turned out to be a potent inhibitor of OAT2. After a 30-min preincubation, cGMP uptake was 47.3 ± 9.6 pmol·mg−1·min−1 under control conditions and decreased to 3.1 ± 0.8 pmol·mg−1·min−1 after preincubation with 100 μM MK571. cGMP uptake after preincubation with PGE2 and with PGE2 plus MK571 was 35.7 ± 5.6 and 3.5 ± 0.2 pmol·mg−1·min−1, respectively (Fig. 7B). MK571 inhibited uptake of cGMP not only under preincubation but also under cis-inhibition conditions: 100 μM MK571 decreased the uptake of cGMP from 41.3 ± 3.1 to 4.2 ± 0.1 pmol·mg−1·min−1 (three subsequent cell passages; data not shown).
The impact of MK571 was also studied on PAH and ES uptake by OAT1 and OAT3. MK571 inhibited PAH uptake by OAT1 with an IC50 values of 23.7 ± 0.7, cGMP uptake by OAT2 with an IC50 value of 9.5 ± 0.7, and ES uptake by OAT3 with an IC50 value of 1.6 ± 0.2 μM, respectively (Fig. 7C).
Sharing 52% identical amino acids, OAT1 and OAT3 are close relatives and may have resulted from gene duplication on chromosome 11q12.3. With 37% identity to OAT3, OAT2 is more distantly related and is localized as a single gene on chromosome 6p21.1 (15, 23). In human kidneys, mRNA for OAT1 and OAT3 outweighed that for OAT2 (26), whereas, in the human liver, OAT2 dominated over OAT3 and OAT1 was not detectable at the mRNA level (2). OAT1 and OAT3 function as organic anion/α-KG exchangers. Whether also OAT2 can use the outwardly directed α-KG gradient for energizing organic anion uptake is not clear. In addition, there are two splice variants of human OAT2, with only the shorter one transporting cGMP (8).
In a search for a further physiological substrate, we tested the effects of urate on OAT1, OAT2, and OAT3. At a concentration of 0.5 mM, urate inhibited PAH uptake in OAT1-transfected HEK-293 cells and was itself transported by OAT1. These results are in line with those obtained by Ichida et al. (14), who identified urate as a low-affinity (KM: 943 μM), high-capacity (1,286 pmol·mg−1·min−1) substrate of OAT1. Although cGMP uptake by OAT2 was not affected by 0.5 mM urate, indomethacin-sensitive urate uptake was observed by us. The missing effect of urate on cGMP uptake is probably due to the low affinity of urate [KM: 1.168 mM (33)]. With respect to OAT3, there exist two reports on Xenopus oocytes investigating its interaction with urate (1, 5). Uptake of urate in OAT3-injected oocytes was only twice that of mocks (5), and for the inhibition of ES uptake by urate, an value IC50 of 275 μM was reported (1). In direct comparison with OAT1 and OAT2 (present study), neither inhibition of the uptake of ES nor transport of urate was observed, which may be explained by the low sensitivity (high IC50 value) and/or different expression systems.
Similarities between transport characteristics of human OAT1, OAT2, and OAT3.
In our experiments, besides OAT2, OAT1- and OAT3-expressing cells also took up significant amounts of cGMP, and transport was inhibited by probenecid and indomethacin, respectively. At 10 μM cGMP, uptake decreased in the following order: OAT2 > OAT1 > OAT3. Trans-stimulation by cGMP of PAH uptake by OAT1 and ES uptake by OAT3 indicates PAH/cGMP and ES/cGMP exchange as additional transport modes of OAT1 and OAT3, respectively. Surprisingly, intracellularly preloaded cGMP failed to trans-stimulate cGMP uptake by human OAT2, suggesting that cGMP self-exchange does not take place or is not detectable. Rat Oat2 has been reported to translocate cGMP (36) or to be unable to transport this cyclic nucleotide (10).
The physiological implication of cGMP transport by OATs remains uncertain. Uptake via OATs could prevent a loss of this intracellular second messenger and prolong its action. It has been discussed that increased renal cGMP levels have an antifibrotic benefit (34). OATs may also clear blood from cGMP by uptake into cells and subsequent release into the tubular lumen via MRP2 and MRP4.
Human OAT1, OAT2, and OAT3 transported PGE2 and PGF2α with high affinities (20). Here we show that, at 100 μM, PGF2α > PGE2 > PGI2 inhibited OAT2-mediated cGMP uptake. The IC50 values observed here for inhibition of cGMP uptake by PGE2 (40.8 μM) are considerably higher than the KM reported for PGE2 uptake [0.71 μM (20)], which could be due to competition between cGMP and PGE2. Our attempts to demonstrate cGMP/prostaglandin exchange failed, probably because of the presence of endogenous MRP2 and MRP4 in HEK-293 cells releasing preloaded prostaglandins. Blockade of MRPs by the cysteinyl leukotriene-based inhibitor MK571 (18) to prevent prostaglandin efflux failed to unmask cGMP/prostaglandin exchange, because MK571 turned out to be a potent inhibitor not only of OAT2 (IC50: 9.5 μM), but also of OAT1 (IC50: 23.7 μM) and OAT3 (IC50: 1.6 μM). Thus, OAT1, OAT2, and OAT3 interact with prostaglandins and are potently inhibited by MK571.
Actually, OAT2 was considered responsible for the long-known glutamate efflux from the liver (11). In the present study, we showed transport of glutamate by OAT2 but failed to show trans-stimulation of cGMP uptake by loading cells with glutamate. One reason for these conflicting results may high endogenous levels of intracellular glutamate that already saturated OAT2 at the intracellular binding site.
Differences between OAT2 and OAT1/OAT3.
Intracellular PAH and cGMP trans-stimulated OAT1-mediated PAH uptake and OAT3-mediated ES uptake but not OAT2-mediated cGMP transport. Radiolabeled PAH and ES were not taken up by OAT2, indicating that this transporter does not share PAH with OAT1 and ES with OAT3 as substrates.
OAT2-mediated cGMP uptake was not inhibited by a series of dicarboxylates in this study and in experiments performed by Sato et al. (33) using uric acid as an OAT2 substrate. In contrast, OAT1-mediated PAH uptake was strongly and OAT3-mediated ES uptake was considerably inhibited by glutarate, adipate, pimelate, and suberate (17). The interaction with dicarboxylates with a chain length of at least five carbons seems to be restricted to OAT1 and OAT3, i.e., to transporters performing organic anion/dicarboxylate exchange. Glutarate and α-KG failed to trans-stimulate cGMP uptake by OAT2, indicating that this transporter is apparently not able to perform organic anion/dicarboxylate exchange, as opposed to OAT1 and OAT3.
OAT1, OAT2, and OAT3 also differed with respect to their pH dependence. Whereas a decrease in pH from 8 to 6 increased transport activities of OAT1 and OAT3, cGMP uptake by OAT2 was hardly pH dependent, excluding H+-cGMP symport or cGMP/OH− exchange as possible modes of transport. Whether the strong pH dependence of OAT1 and OAT3 reflects symport with H+ or antiport against OH− or is merely based on a pH dependence of the transporter itself remains to be elucidated.
Substrate uptake by OAT1 and OAT3 are strongly Cl− dependent. For OAT1, the Cl− dependence was blunted by the mutation of the arginine at position 466 in the 11th transmembrane domain (32). The substitution of Cl− by gluconate did not change KM for the uptake of PAH in OAT1-transfected HEK-293 cells but caused a threefold decrease in the maximal uptake rate (39), whereas in OAT3-transfected HEK-293 cells, KM for ES decreased without an effect on the maximal uptake rate (39).
Transport mode of OAT2.
Having excluded cGMP/dicarboxylate and cGMP/OH− exchange, we tested for a possible interaction of OAT2 with monovalent anions. However, the tested monocarboxylic acids with two to four carbons neither inhibited cGMP uptake, when added to the bath, nor significantly trans-stimulated uptake after preloading into the cells, indicating a lack of interaction of OAT2 with short-chain monocarboxylates. Hence, cGMP/monocarboxylate exchange can be excluded as a transport mode of OAT2. An analogous conclusion can be drawn for a putative cGMP/Cl− exchange, because uptake of cGMP was not affected by Cl−.
The inability to demonstrate exchange of several structurally unrelated compounds with cGMP favors the possibility that cGMP uptake by OAT2 occurs as an uniporter. In this case, uptake of a net negative charge should render transport sensitive to membrane potential, i.e., depolarization should increase cGMP uptake. However, increasing extracellular K+ up to 10-fold in the presence of valinomycin did not result in a statistically significant increase in cGMP uptake. The independence of membrane potential would be rather compatible with an electroneutral exchange of cGMP for an intracellular organic anion, e.g., intracellular glutamate or an as-yet-unidentified compound.
Y. Hagos is a Professor of Physiology at the Center of Physiology and CEO of PortaCellTec. All other authors have no conflict of interest.
Author contributions: M.H., Y.H., and B.C.B. conception and design of research; M.H., Y.H., and W.K. performed experiments; M.H., Y.H., and B.C.B. analyzed data; M.H., Y.H., G.B., and B.C.B. interpreted results of experiments; M.H. and B.C.B. prepared figures; M.H., G.B., and B.C.B. drafted manuscript; M.H., G.B., and B.C.B. edited and revised manuscript; M.H., Y.H., G.B., and B.C.B. approved final version of manuscript.
The authors thank Sören Petzke and Andrea Paluschkiwitz for skillful technical assistance.
- Copyright © 2015 the American Physiological Society