(Received for publication, November 18, 1996, and in revised form, January 16, 1997)
From the Department of Pharmacology, University of Heidelberg, Im Neuenheimer Feld 366, 69120 Heidelberg, Germany
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.
An organic cation transporter, OCT2p, was cloned from LLC-PK1 cells, a porcine cell line with properties of proximal tubular epithelial cells. OCT2p was heterologously expressed and characterized in human embryonic kidney 293 cells. OCT2p-mediated uptake of the prototypical organic cation [14C]tetraethylammonium ([14C]TEA) into 293 cells was saturable. There was a highly significant correlation between the Ki values for the inhibition of apical [14C]TEA uptake into LLC-PK1 cells and 293 cells transfected with OCT2p (r = 0.995; p < 0.001; n = 6). Although OCT2p is structurally related to OCT1r, the basolateral organic cation transporter from rat kidney, the transporters could be clearly discriminated pharmacologically with corticosterone, decynium22, and O-methylisoprenaline. The findings at hand suggest that OCT2 corresponds to the apical type of organic cation transporter. Reverse transcriptase-polymerase 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 kidney and central nervous regions known to be rich in the monoamine transmitter dopamine.
Secretion of metabolites, xenobiotics, and drugs is an important physiological function of renal proximal tubules. Organic cations, like N1-methylnicotinamide, tetraethylammonium (TEA),1 and 1-methyl-4-phenylpyridinium (MPP+), 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 Gründemann 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-PK1 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.
[14C]TEA
([1-14C]-tetraethylammonium bromide, 0.12 Bq/pmol)
was from NENTM Life Science Products (Bad Homburg, Germany).
Corticosterone, cyanine863
(1-ethyl-2-[(1,4-dimethyl-2-phenyl-6-pyrimidinylidene)methyl]quinolinium chloride), decynium22 (1,1-diethyl-2,2
-cyanine iodide), quinine hemisulfate, and tetraethylammonium bromide were from Sigma-Aldrich (Deisenhofen, Germany); O-methylisoprenaline hydrochloride
was from Boehringer (Ingelheim, Germany). Disprocynium24 was
synthesized as described previously (4).
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() (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
35S-labeled single-stranded DNA probe (11).
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.
For PCR on double-stranded cDNA from LLC-PK1 cells, the
following degenerate oligonucleotides, derived from the OCT1r amino acid sequence (2), were used (see Fig. 2): 5-GAR CTN TAY CCN AC-3
(forward primer, corresponding to nucleotide positions 1403-1416 of
the OCT1r cDNA), and 5
-TTN GTY TCN GGY AA-3
(reverse primer, reverse complementary to positions 1601-1614). The PCR reaction mix
(50 µl) contained 0.5 µM per primer, 0.2 mM
per dNTP, 2.3 mM MgCl2, and 2 units of
Taq DNA polymerase in the buffer provided (Boehringer
Ingelheim Bioproducts, Heidelberg, Germany). Thermocycling consisted of
38 cycles of 30 s at 94 °C, 30 s at 45 °C, and 1 min at
72 °C. The amplification product (0.2 kb) was cloned into pUC19 as
described (13) and sequenced by the dideoxy method (14).
RT-PCR
20 µg total RNA was incubated at 37 °C for 30 min with 2 units of RQ1-DNase (Boehringer Ingelheim Bioproducts,
Heidelberg, Germany) in 100 µl of 5 mM MgCl2,
50 mM triethanolamine-HCl, pH 7.5, to degrade any residual
DNA. The RNA was extracted with phenol-chloroform, precipitated with
ethanol and dissolved in water. For cDNA synthesis, 5 µg of the
RNA thus prepared was incubated at 45 °C for 1 h in a total
volume of 20 µl with 100 units of Superscript II RT (Life Technologies, Eggenstein, Germany) in 10 µM random
hexamers, 1.5 mM per dNTP, 7 mM
MgCl2, 75 mM KCl, 50 mM
triethanolamine-HCl, pH 8.7 (26 °C), 10 mM
dithiothreitol, and 0.75 units/µl RNase inhibitor from human
placenta. For paired negative controls, RT was omitted. Following heat
inactivation of the proteins (10 min at 95 °C), and addition of 5 µl of 0.5 mg/ml DNase-free RNase A (QIAGEN, Hagen, Germany) in 10 mM Tris-HCl, pH 7.5, the cDNA was incubated at 37 °C
for 30 min, to degrade unreacted mRNA. With 4 µl of this
preparation, PCR was performed as described above (60 °C annealing
temperature), with the following primers: 5-CCT GGG CTC CCT GGT TGT
GGG TTA-3
(forward primer OCT1r), 5
-AAT GAG GGG CAG GGC TTG CCA AA-3
(reverse primer OCT1r), 5
-CCG CTA TCC CTG GGC TGT GTC AAA-3
(forward
primer OCT2r), 5
-TGG CCC ACA GCT CCC TTG GGT ATT-3
(reverse primer
OCT2r), 5
-ACT GGC GTC TTC ACC ACC AT-3
(forward primer
glyceraldehyde-3-phosphate dehydrogenase), and 5
-TCC ACC ACC CTG TTG
CTG TA-3
(reverse primer glyceraldehyde-3-phosphate
dehydrogenase).
LLC-PK1 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 µM [14C]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 StatisticsTime courses, saturation, and inhibition curves were analyzed as described (15). Ki values are given as geometric mean with 95% confidence interval. Hill coefficients are given as arithmetic mean ± S.E.
Sequences were analyzed with the GCG package, with default settings, as implemented in HUSAR (DKFZ, Heidelberg, Germany).
With LLC-PK1 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 µM cyanine863, which is a known inhibitor of renal
secretion of organic cations (1). All tested compounds inhibited apical
TEA uptake into LLC-PK1 cells (Fig. 1). The
Ki values ranged from 2.5 nM for disprocynium24 to 0.58 mM for
O-methylisoprenaline (Table I).
|
Degenerate primers were derived from the amino acid sequence of OCT1r, and used in PCR on cDNA from LLC-PK1 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-PK1 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-PK1 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 TransporterThe 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 N-glycosylation 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 OCT2pFor 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 [3H]MPP+ or [14C]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 (kin) and outwardly
(kout) directed TEA fluxes of 0.88 ± 0.07 µl min1 mg of protein
1 and 0.07 ± 0.01 min
1, respectively (n = 18). The
uptake at equilibrium (Amax) amounted to
38.0 ± 3 pmol mg of protein
1. Based on an
intracellular water space of 6.7 µl mg of protein
1 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 (Km = 20 µM, Vmax = 7 pmol
min1 mg of protein
1), which was further
characterized, and a low affinity component (Km = 620 µM, Vmax = 106 pmol
min
1 mg of protein
1), which was not
analyzed in detail in the present study (n = 18).
To compare OCT2p with the apical TEA uptake mechanism in
LLC-PK1 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 Ki 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-PK1 cells (Fig. 6) revealed a
highly significant correlation of the Ki values
(r = 0.995, n = 6, p < 0.001). This indicates that identical transport mechanisms are
expressed in LLC-PK1 cells and in
pcDNA3OCT2p-transfected 293 cells.
Inhibitory Effect of Corticosterone
Inhibition by
corticosterone of apical TEA uptake into LLC-PK1 cells was
analyzed in more detail. Compared with controls (30-min preincubation,
Ki = 0.20 (0.11, 0.37) µM),
corticosterone was equally inhibitory to TEA uptake after a
preincubation period of only 1 min (Fig. 7;
Ki = 0.25 (0.15, 0.41) µM). Saturation analysis (Fig. 8) implies that corticosterone is a
competitive inhibitor of TEA uptake, since in the presence of 0.3 µM corticosterone, the apparent Km for
TEA was increased from 48 (control) to 124 µM, while the
Vmax remained the same (243 versus
266 pmol min1 mg of protein
1).
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 ± 0.7 pmol min1 mg
1 (control cells, stably
transfected with pcDNA3: 0.44 ± 0.03 pmol min
1
mg
1). By contrast, at pH 5.4 the rate of uptake was
9.8 ± 0.5 pmol min
1 mg
1 (control:
0.31 ± 0.05 pmol min
1 mg
1). Thus, an
increase of the medium proton concentration significantly inhibits
uptake of TEA (n = 3; p < 0.01). A
comparable proton dependence was observed with HEPES-buffered media at
pH 8.5 versus 6.5 (data not shown).
LLC-PK1 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-PK1 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-PK1 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-PK1 cells contain, analogous to renal brush-border membrane vesicles from various species (27-32), an organic cation uptake mechanism that markedly depends on pH (33). Together, these findings strongly suggest that LLC-PK1 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-PK1 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+. 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 al. (36) from their extensive in vivo characterization of organic cation transport in rat kidney, that the luminal and contraluminal organic cation transporters are similar, but not identical. Very recently, Okuda et al. (37) have published the OCT2r cDNA sequence and reported functional expression in Xenopus oocytes. Our amino acid sequence is identical to the published sequence except for positions 332 (Lys (per Okuda et al.) versus Asn) and 335 (Ile versus Phe).
2) Data from a previous study (3) allow a valid comparison of the affinities of OCT1r and OCT2p for various inhibitors. While the Ki values for cyanine863 (0.50 versus 0.67 µM for OCT2p and OCT1r, respectively) and quinine (5.5 versus 4.3 µM) are virtually identical, OCT2p has a clearly higher affinity for decynium22 (0.051 versus 0.50 µM), but lower affinity for O-methylisoprenaline (25 versus 880 µM). However, the most discriminative compound is corticosterone, which with a Ki of 0.67 µM inhibits OCT2p 100 times more potently than OCT1r (Ki = 72 µM). OCT1r as found in rat hepatocytes (3) or as expressed in Xenopus oocytes (2) also is rather resistant to corticosterone, with Ki values of 14 and >20 µM, respectively.
Corticosterone does not qualify as an organic cation. Nevertheless, it is a potent inhibitor (Ki = 0.13 µM) of the extraneuronal monoamine transport mechanism (uptake2), which is distinct from, but probably related to, the renal organic cation transporters (38-40). Inhibition of the apical organic cation transporter in LLC-PK1 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-PK1 cells with high potency (data not shown), which does not fit in with any known intracellular steroid receptor.
The inferred Km of TEA uptake into pcDNA3OCT2ptransfected 293 cells (20 µM) agrees well with those reported for apical TEA uptake into LLC-PK1 (34 µM (25), 27 µM (15)), and OK cells (28 µM (35)). More decisively, comparison of the Ki values of structurally unrelated compounds for the inhibition of transiently expressed OCT2p to those of the apical TEA uptake in LLC-PK1 cells revealed a highly significant correlation (Fig. 6). This is firm evidence that OCT2p is functionally expressed in LLC-PK1 cells. Since the LLC-PK1 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-PK1 cells.
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 × mg of protein1, respectively), compared
with e.g. frontal cortex (0.2-1.0 ng of dopamine × mg
of protein
1) (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-PK1 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.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) Y09400[GenBank].
We thank Anke Ripperger and Heike Grieshaber for skillful technical assistance.