Primary Structure and Functional Expression of the Apical Organic Cation Transporter from Kidney Epithelial LLC-PK 1 Cells*

Renal secretion of organic cations involves at least two distinct transporters, located in the basolateral and apical membranes of proximal tubule cells. Whereas the basolateral transporter has recently been cloned, sequence information about the apical type was not yet available.Anorganic cation transporter, OCT2p, was cloned from LLC-PK 1 cells, a porcine cell line with properties of proximal tubular epithelial cells. OCT2p was heterolo-gously expressed and characterized in human embry- onic kidney 293 cells. OCT2p-mediated uptake of the prototypical organic cation [ 14 C]tetraethylammonium ([ 14 C]TEA) into 293 cells was saturable. There was a highly significant correlation between the K i values for the inhibition of apical [ 14 C]TEA uptake into LLC-PK 1 cells and 293 cells transfected with OCT2p ( r (cid:53) 0.995; p < 0.001; n (cid:53) 6). Although OCT2p is structurally related to OCT1r, the basolateral organic cation transporter from rat kidney, the transporters could be clearly discrimi-nated pharmacologically with corticosterone, decy- nium22, and O -methylisoprenaline. The findings at hand suggest that OCT2 corresponds to the apical type of organic cation transporter. Reverse transcriptase-po-lymerase chain reaction indicates that mRNA of OCT1r is limited to non-neuronal tissue, whereas OCT2r, the OCT2p homologue from rat, was found in both the kid- ney and central nervous regions known to be rich

Secretion of metabolites, xenobiotics, and drugs is an important physiological function of renal proximal tubules. Organic cations, like N 1 -methylnicotinamide, tetraethylammonium (TEA), 1 and 1-methyl-4-phenylpyridinium (MPP ϩ ), are secreted via transcellular transport, i.e. uptake from the blood across the basolateral membrane into the proximal tubular epithelial cells followed by extrusion across the brush-border membrane into the tubular fluid (1). Evidence from functional studies on membrane vesicles, isolated tubules, and tissue slices indicates that specific transport mechanisms for organic cations exist in both the luminal and basolateral membranes of proximal tubular cells.
Transport across the basolateral membrane is brought about by a facilitated, but passive, mechanism that is accelerated by the inside negative membrane potential. On the other hand, transport across the apical membrane is thought to result from electroneutral exchange of cellular organic cations with tubular protons or organic cations.
So far, little is known about the molecular structures of the transporters. However, the recent cloning by Grü ndemann et al. (2) of OCT1r, an organic cation transporter from rat kidney, has opened the field of organic cation transport for molecular studies. OCT1r has been assigned to the basolateral membrane of proximal tubular cells and is also functionally expressed in hepatocytes (3).
The polymerase chain reaction (PCR) with degenerate oligonucleotides derived from the OCT1r amino acid sequence was employed to search for OCT1-related transporters. Here we report the cloning and functional expression of OCT2p, a transporter from LLC-PK 1 cells, with characteristics of the apical type of organic cation transporter, as well as the tissue distribution of OCT2r, the presumptive OCT2p homologue from rat kidney.
RNA Purification, cDNA Library Construction, and Library Screening-Total RNA was extracted by the method of Chomczynski and Sacchi (5). The mRNA was selected twice by affinity chromatography on oligo(dT)-cellulose (6). cDNA was synthesized by a modification of the strategy of Gubler and Hoffman (7) using a NotI primer-linker and Superscript II reverse transcriptase (RT) (Life Technologies, Eggenstein, Germany). For cDNA library construction, following adapter ligation, adapter phosphorylation, and restriction of the primer-linker, size fractionated cDNA (2.8 -4.5 kb) was prepared by UV-protected agarose gel electrophoresis (8). The cDNA was ligated with pBluescript II SK(Ϫ) (Stratagene, Heidelberg, Germany) and electroporated into Escherichia coli DH10B as described (9). The library was screened with bacteria plated on nylon filters as described (10) using a 35 S-labeled single-stranded DNA probe (11).
Plasmids, PCR Amplification, and DNA Sequencing-If not noted otherwise, standard molecular biology techniques were employed (12). pcDNA3OCT2p contains the cDNA of OCT2p in the NotI and XhoI sites of pcDNA3 (Clontech, Palo Alto, CA). pcDNA3OCT2r contains the cDNA of OCT2r in the HindIII and XhoI sites.
Cell Culture, Transfection, and Uptake Measurements-LLC-PK 1 cells (ATCC CRL-1392) and 293 cells (ATCC CRL-1573) were cultured on 60-mm polystyrol dishes as described (3,15). After washing with serum-free medium, each culture dish was incubated at 37°C for 4 h with 2 ml of serum-free medium which contained 10 nmol of Tfx-50 reagent (Boehringer Ingelheim Bioproducts, Heidelberg, Germany) and 0.5 pmol of plasmid DNA. Subsequently, the transfection medium was replaced by standard culture medium. After 2 days, uptake was measured essentially as described elsewhere (3,15). Unless stated otherwise, 3 M [ 14 C]TEA was used. Inhibitors of uptake were present during both the preincubation and incubation periods. In the pH effect experiment, Dulbecco's phosphate-buffered saline, containing 5 mM glucose, pH adjusted with HCl, was used.
293 cells stably transfected with pcDNA3OCT2r were selected with G418 according to the protocol of the vendor (Boehringer Mannheim). Expression of OCT2r was verified by RT-PCR and functional characterization.
Calculations and Statistics-Time courses, saturation, and inhibition curves were analyzed as described (15). K i values are given as geometric mean with 95% confidence interval. Hill coefficients are given as arithmetic mean Ϯ S.E.
Sequences were analyzed with the GCG package, with default settings, as implemented in HUSAR (DKFZ, Heidelberg, Germany).

Pharmacological Profile of Apical TEA Uptake into LLC-PK 1
Cells-With LLC-PK 1 cells grown on plastic dishes for 6 days, the effects on initial rates of specific TEA uptake of various compounds were determined. Specific transport was defined as that fraction of total uptake that is sensitive to 10 M cya-nine863, which is a known inhibitor of renal secretion of organic cations (1). All tested compounds inhibited apical TEA uptake into LLC-PK 1 cells (Fig. 1). The K i values ranged from 2.5 nM for disprocynium24 to 0.58 mM for O-methylisoprenaline (Table I).
Cloning of an OCT1r Homologous Transporter from LLC-PK 1 Cells-Degenerate primers were derived from the amino acid sequence of OCT1r, and used in PCR on cDNA from LLC-PK 1 cells. A 212-base fragment, which was cloned into pUC19 and sequenced, revealed a substantial similarity (89%) to OCT1r. When the PCR fragment was used as a probe in Northern blot analysis of LLC-PK 1 mRNA, a single band with a length of approximately 3.6 kb was detected (data not shown).
Subsequently, an appropriately size-fractionated cDNA library was generated from LLC-PK 1 mRNA and screened by colony hybridization, with a probe derived from the PCR fragment. A clone with a length of 3.0 kb was isolated and sequenced.
Primary Structure of the Transporter-The amino acid sequence of the clone, generated by conceptual translation of the largest open reading frame of the cDNA (Fig. 2), is remarkably similar to the amino acid sequence of OCT1r (identity 67%, similarity 83%). The clone was named OCT2p (p denoting pig). From hydropathy analysis, OCT2p, with a length of 554 amino acids, is predicted to have a membrane topology much like OCT1r (Fig. 2). Within a framework of 12 transmembrane segments, a large extracellular loop with three potential Nglycosylation sites is formed between transmembrane segments 1 and 2, and potential intracellular phosphorylation sites are situated after the transmembrane segments 6 and 12.
Functional Characterization of OCT2p-For functional expression, the cDNA of OCT2p was inserted into the eucaryotic expression vector pcDNA3. With the resulting plasmid, pcDNA3OCT2p, 293 cells, a transformed cell line derived from human embryonic kidney (16), were transiently transfected. Control cells were transfected with pcDNA3. The cDNA of OCT2p induced expression of specific [ 3 H]MPP ϩ or [ 14 C]TEA transport activity (data not shown). Since TEA had a higher ratio of specific to nonspecific uptake, it was chosen to characterize OCT2p.
A detailed analysis of the time course of uptake of TEA into pcDNA3OCT2p-transfected cells (Fig. 3) revealed rate constants for inwardly (k in ) and outwardly (k out ) directed TEA fluxes of 0.88 Ϯ 0.07 l min Ϫ1 mg of protein Ϫ1 and 0.07 Ϯ 0.01 min Ϫ1 , respectively (n ϭ 18). The uptake at equilibrium (A max ) amounted to 38.0 Ϯ 3 pmol mg of protein Ϫ1 . Based on an intracellular water space of 6.7 l mg of protein Ϫ1 and a transfection efficiency of 25% (3), at equilibrium an 8-fold accumulation of TEA relative to medium can be estimated. An uptake period of 4 min was chosen for all subsequent experiments to approximate initial rates of transport.
Specific uptake of TEA into 293 cells transfected with pcDNA3OCT2p was saturable (Fig. 4). Since the Eadie-Hofstee plot is not compatible with a single uptake mechanism, the saturation curve was resolved into a high affinity component (K m ϭ 20 M, V max ϭ 7 pmol min Ϫ1 mg of protein Ϫ1 ), which was further characterized, and a low affinity component (K m ϭ 620 M, V max ϭ 106 pmol min Ϫ1 mg of protein Ϫ1 ), which was not analyzed in detail in the present study (n ϭ 18).
To compare OCT2p with the apical TEA uptake mechanism in LLC-PK 1 cells, the pharmacological profile of TEA uptake into pcDNA3OCT2p-transfected 293 cells had to be established. Hence, the effects on initial rates of specific TEA uptake of various compounds were determined. All compounds tested inhibited TEA uptake (Fig. 5). The K i values ranged from 51 nM for decynium22 to 0.88 mM for O-methylisoprenaline (Table I).
Comparison of the pharmacological profile of OCT2p transiently expressed in 293 cells to that of the apical TEA uptake mechanism present in LLC-PK 1 cells (Fig. 6) revealed a highly significant correlation of the K i values (r ϭ 0.995, n ϭ 6, p Ͻ 0.001). This indicates that identical transport mechanisms are expressed in LLC-PK 1 cells and in pcDNA3OCT2p-transfected 293 cells.
Inhibitory Effect of Corticosterone-Inhibition by corticosterone of apical TEA uptake into LLC-PK 1 cells was analyzed in more detail. Compared with controls (30-min preincubation, K i ϭ 0.20 (0.11, 0.37) M), corticosterone was equally inhibitory to TEA uptake after a preincubation period of only 1 min (Fig. 7; K i ϭ 0.25 (0.15, 0.41) M). Saturation analysis (Fig. 8) implies that corticosterone is a competitive inhibitor of TEA uptake, since in the presence of 0.3 M corticosterone, the apparent K m for TEA was increased from 48 (control) to 124 M, while the V max remained the same (243 versus 266 pmol min Ϫ1 mg of protein Ϫ1 ).
Tissue Distribution of OCT1r and OCT2r-The presumptive OCT2p homologue from rat kidney, OCT2r, was cloned and sequenced as described for OCT2p (see "Discussion"). The tissue distribution of OCT1r and OCT2r mRNA was analyzed by RT-PCR with specific oligonucleotide primers (Fig. 9). OCT1r mRNA was detected in liver, kidney, intestine, and veins, but not in the brain. In peripheral tissues, OCT2r mRNA is confined to the kidney. Interestingly, however, some regions of rat central nervous system contain OCT2r mRNA. The signal was most pronounced in substantia nigra, nucleus accumbens, and striatum.
Effect of pH on Transport Activity of OCT2r-The rate of total TEA uptake at pH 7.2 into 293 cells expressing OCT2r was 15.4 Ϯ 0.7 pmol min Ϫ1 mg Ϫ1 (control cells, stably transfected with pcDNA3: 0.44 Ϯ 0.03 pmol min Ϫ1 mg Ϫ1 ). By contrast, at pH 5.4 the rate of uptake was 9.8 Ϯ 0.5 pmol min Ϫ1 mg Ϫ1 (control: 0.31 Ϯ 0.05 pmol min Ϫ1 mg Ϫ1 ). Thus, an increase of the medium proton concentration significantly inhib-  its uptake of TEA (n ϭ 3; p Ͻ 0.01). A comparable proton dependence was observed with HEPES-buffered media at pH 8.5 versus 6.5 (data not shown). DISCUSSION LLC-PK 1 cells, a widely used established cell line from pig kidney (17), show many properties of proximal tubular epithelial cells (18). These polarized cells attach with their basolateral surface to culture dishes, while the apical membrane domain, which corresponds to the brush-border membrane, faces the culture medium. After several days in culture, they form monolayers, with microvilli and tight junctions (19). LLC-PK 1 cells, grown on permeable or impermeable support, have been used for selective exposure of apical or basolateral membranes to substrates and inhibitors (20 -25).
LLC-PK 1 cells are a model for the renal secretion of organic cations, since with monolayers grown on permeable filters, the transepithelial TEA flux in physiological direction markedly exceeds the flux in the reverse direction (23,24,26). Transepithelial transport of TEA from the basolateral side is potently inhibited by decynium22, applied to the apical side, whereas application to the basolateral side has no effect (15). Moreover, apical membrane vesicles from LLC-PK 1 cells contain, analogous to renal brush-border membrane vesicles from various species (27)(28)(29)(30)(31)(32), an organic cation uptake mechanism that markedly depends on pH (33). Together, these findings strongly suggest that LLC-PK 1 cells express the apical type of organic cation transporter. Essentially similar results have been obtained with OK cells (34,35).
In the present study, LLC-PK 1 mRNA was chosen as primary material for the molecular characterization of the apical type of organic cation transporter. A cDNA was isolated, which, upon heterologous expression in 293 cells, induces saturable uptake of TEA and MPP ϩ . The clone was named OCT2p, to distinguish it from OCT1, for two reasons. 1) We have cloned and sequenced another organic cation transporter from rat kidney, OCT2r. Pairwise sequence analysis indicates that OCT2p is closer related to OCT2r (identity 81%, similarity 90%) than to OCT1r (identity 67%, similarity 83%). Interestingly, the overall similarity of structures of OCT1 and OCT2 agrees well with the conclusion of Ullrich et Corticosterone does not qualify as an organic cation. Nevertheless, it is a potent inhibitor (K i ϭ 0.13 M) of the extraneuronal monoamine transport mechanism (uptake 2 ), which is distinct from, but probably related to, the renal organic cation transporters (38 -40). Inhibition of the apical organic cation transporter in LLC-PK 1 cells by corticosterone seems to involve a direct interaction with the transporter rather than a genomic effect. Inhibition was competitive and was effective after 1 min of incubation with the inhibitor, whereas genomic effects become apparent only after hours (41). Other steroids such as progesterone and testosterone also inhibit apical TEA uptake into LLC-PK 1 cells with high potency (data not shown), which does not fit in with any known intracellular steroid receptor.
The inferred K m of TEA uptake into pcDNA3OCT2ptransfected 293 cells (20 M) agrees well with those reported for apical TEA uptake into LLC-PK 1 (34 M (25), 27 M (15)), and OK cells (28 M (35)). More decisively, comparison of the K i values of structurally unrelated compounds for the inhibition of transiently expressed OCT2p to those of the apical TEA uptake in LLC-PK 1 cells revealed a highly significant correlation (Fig.  6). This is firm evidence that OCT2p is functionally expressed in LLC-PK 1 cells. Since the LLC-PK 1 cells were grown on plastic dishes for 6 days prior to measurements, uptake most probably was restricted to the apical membrane domain. Therefore, it must be concluded that OCT2p is identical with the apical organic cation transporter from LLC-PK 1 cells. Initial rates were determined with various concentrations of TEA at 37°C with an uptake period of 2 min in the absence (E) and presence (q) of 0.3 M corticosterone. Specific uptake was defined as that fraction of total uptake which is sensitive to 10 M cyanine863. Shown is mean Ϯ S.E. (n ϭ 4). Inset, corresponding Eadie-Hofstee plot.
FIG. 9. RT-PCR analysis of tissue distribution of OCT1r and OCT2r mRNA. PCR products were separated by agarose gel electrophoresis, followed by staining with ethidium bromide, as shown. The mRNA-derived signal is the difference of band intensities between samples generated in the presence (ϩ) or in the absence (Ϫ) of RT during reverse transcription. As control for the intactness of mRNA, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was detected with specific primers. In separate experiments it was verified that the primers used to detect OCT1r and OCT2r do not show any cross-reactivity.
OCT2r, the presumptive OCT2p homologue from rat kidney, was found, when expressed in Xenopus oocytes, to operate independently from medium pH (37). However, in contrast to these data, our clone of OCT2r, when expressed in 293 cells, was clearly inhibited by an increase in the medium proton concentration. This proton dependence is consistent with the reported characteristics of the apical organic cation transporter (25). However, it should be noted that detection of the proton dependence may depend strongly on the experimental conditions, since some attempts have succeeded (26,32,42), whereas others have failed (25,34) or produced small effects (35). Perhaps post-translational modifications or competing driving forces such as membrane potential or intracellular substrates account for variability. Thus, in general, identification of transporters would seem to be more reliable by comparison of pharmacological profiles rather than by analysis of driving forces.
The present RT-PCR analysis of tissue distribution of OCT1r mRNA confirms data from Northern analysis (2). In non-neuronal tissue, OCT2r mRNA was exclusively found in the kidney, which agrees well with previous data from Northern analysis and RT-PCR (37). In contrast to the data of Okuda et al. (37), however, OCT2r mRNA was clearly detectable in various brain regions (Fig. 9).
Is there a link between the presumptive expression of OCT2r in the kidney and in distinct brain regions? We speculate that this common denominator could be the monoamine transmitter dopamine. The brain regions with the strongest RT-PCR signals for OCT2r mRNA, i.e. substantia nigra, nucleus accumbens, and striatum, are known to contain high levels of dopamine (20 -22, 87-101, and 71-134 ng of dopamine ϫ mg of protein Ϫ1 , respectively), compared with e.g. frontal cortex (0.2-1.0 ng of dopamine ϫ mg of protein Ϫ1 ) (43). Moreover, the kidney is a major site for extraneuronal production of dopamine (44), and renal secretion of dopamine in vivo is sensitive to disprocynium24. 2 Finally, OCT2r, expressed in 293 cells, avidly transports radiolabeled dopamine (data not shown).
In conclusion, an organic cation transporter, OCT2p, was cloned from LLC-PK 1 cells, with characteristics of the apical TEA transporter. Although OCT2p is structurally related to the basolateral transporter OCT1r, the transporters can be clearly distinguished on a pharmacological basis. While expression of OCT1r is limited to peripheral tissues, the present study indicates that OCT2r is expressed both in the kidney and the central nervous system. Since with the isocyanines and pseudoisocyanines very potent and selective inhibitors are available (4,15), OCT2 could become a promising target for biomedical applications.