| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Physiology, College of Medicine, University of Arizona, Tucson, Arizona
ABSTRACT I. INTRODUCTION II. RENAL ORGANIC CATION TRANSPORT A. Overview of the Physiological Characteristics of Renal OC Transport B. Molecular Characteristics of Renal OC Transporters 1. OCT family of OC transporters A) OCT1. B) OCT2. C) OCT3. D) OCTN1. E) OCTN2. F) OCTN3. 2. MDR family of OC transporters A) MOLECULAR ASPECTS OF MDR1. B) LOCATION OF MDR1 EXPRESSION IN RENAL TISSUES. C) SPECIFICITY OF MDR1. C. Physiological Integration of Renal OC Transporters 1. Cellular organization and mechanisms of renal OC transport A) BASOLATERAL/CONTRALUMINAL TRANSPORT MECHANISM. B) INFLUENCE OF INORGANIC IONS ON BASOLATERAL OC TRANSPORT. 2. Specificity and multiplicity of basolateral OC transport A) SPECIFICITY OF BASOLATERAL OC TRANSPORTERS. B) MULTIPLICITY OF BASOLATERAL OC TRANSPORTERS. C) MULTIPLE TRANSPORTERS IN NATIVE RENAL CELLS. D) COMPARISON OF CLONED TRANSPORTERS TO TRANSPORT ACTIVITY IN NATIVE RENAL CELLS. 3. Regulation of basolateral OC transport 4. Apical/brush border/luminal OC transport: mechanism 5. Specificity of luminal OC transport 6. Role of MDR1 in renal OC secretion 7. Intracellular OC sequestration III. RENAL ORGANIC ANION TRANSPORT A. Overview of the Physiological Characteristics of Renal OA Transport 1. Peritubular OA transport 2. Luminal OA transport B. Molecular Characteristics of Renal OA Transporters 1. OCT family of OA transporters A) OAT1. B) OAT2-4/URAT1. 2. OATP family of OA transporters A) OATP1 (SLC21A1; 2.A.60.1.1). B) OAT-K1/OAT-K2. 3. NPT family of Pi/OA transporters 4. MRP family of OA transporters A) MRP1 (ABCC1; 3.A.1.208.1). B) MRP2 (ABCC2; 3.A.1.208.2). C) MRP3 (ABCC3; 3.A.1.208.9). D) MRP4 (ABCC4; 3.A.1.208.7). E) MRP5 (ABCC5). F) MRP6 (ABCC6; 3.A.1.208.10). C. Physiological Integration of Renal OA Transporters 1. Mechanisms, selectivity, and functional organization of renal OA transport A) BASOLATERAL/PERITUBULAR TRANSPORT. B) SPECIFICITY AND MULTIPLICITY OF BASOLATERAL OA TRANSPORT. B) MECHANISM OF LUMINAL PAH (I.E., TYPE I) OA TRANSPORT. C) SPECIFICITY AND MULTIPLICITY OF LUMINAL OA TRANSPORTER PATHWAYS. 2. Regulation of tubular OA transport A) KINASE-MEDIATED MODULATION OF RENAL BASOLATERAL OA TRANSPORT. B) KINASE-MEDIATED MODULATION OF RENAL LUMINAL OA TRANSPORT. C) STEROID-MEDIATED MODULATION OF RENAL OA TRANSPORT. 3. Intracellular OA sequestration IV. SPECULATIONS AND CONCLUSIONS NOTE ADDED IN PROOF: ACKNOWLEDGMENTS REFERENCES
 |
ABSTRACT |
|---|
 Top NextReferences |
|---|
|
I. INTRODUCTION |
|---|
|
TopPreviousNextReferences |
|---|
In this review we have attempted to compile the novel material on the molecular and cellular physiology of renal OC and OA transport that has accrued since the 1993 review of Pritchard and Miller (342). Readers interested in the material preceding 1993 are directed to that excellent discussion and to several earlier reviews that assessed the historical sweep of studies in this area (292, 355, 514, 515). Here, we limit ourselves to an examination of the several families of transport processes currently suspected of playing significant roles in the renal handling of organic cations and anions. Readers interested in more detail on selected issues should consult excellent recent reviews focused on specific subjects (38, 41, 89, 218–221, 227, 343, 426, 465, 496, 501, 547). We also note that, although there is considerable functional overlap of the OC and OA transport processes expressed in the kidney and the liver (and gastrointestinal tract), we have limited our focus to those processes believed to play major roles in the transport of these compounds in renal cells. Readers are directed to reviews on organic electrolyte transport expressed in the liver (12, 201, 231, 233, 273, 419, 501) and other organs (14, 157, 220, 224, 311, 398, 547).
The review is organized in the following fashion. The molecular and cellular physiology of organic cation secretion is presented first. This reflects the fact that the first of the associated processes to be cloned was a cation transporter (OCT1). The resulting family of transport processes, which includes, as it turns out, processes involved with the classical OA secretory pathway, is referred to as the "organic cation transporter family" (253, 361). For each class of substrate (OC or OA), we initially discuss the general physiology of renal secretion, particularly as it was understood in 1993. This is followed by an extended discussion of the molecular characteristics of individual members of the several families of transporters implicated in renal transport of these substrates. Then, in the light of increasing understanding of the molecular properties of these processes, we consider recent, novel observations on the cellular physiology of renal OC or OA transport, with a particular effort focused on highlighting where "molecular" information has helped clarify "cellular" information, and where physiological observations run counter to expectations arising from the results obtained studying cloned transporters. We close with a discussion of issues that we think require future study. Foremost among these is the need to integrate knowledge obtained on individual transporters into the cellular events that result in renal OC or OA secretion. These secretory processes involve the coordinated function of many separate processes working in the physiological context of intact cells and tissues. Finally, we offer some speculative views on evolutionary aspects of renal OC and OA transport.
|
II. RENAL ORGANIC CATION TRANSPORT |
|---|
|
TopPreviousNextReferences |
|---|
The proximal tubule is the primary site of renal OC secretion, as determined by stop-flow, micropuncture, and microperfusion studies (342, 355). The proximal tubule is also responsible for the reabsorption of a more limited number of cationic substrates (342). In general, substrates for the pathways involved in renal OC transport include a diverse array of primary, secondary, tertiary, or quaternary amines that have a net positive charge on the amine nitrogen at physiological pH. Although a number of endogenous OCs have been shown to be actively secreted by the proximal tubule [e.g., N1-methylnicotinamide (NMN), choline, epinephrine, and dopamine; see Ref. 22], it is generally accepted that the principal function of this process is clearing the body of xenobiotic agents (342), including a wide range of alkaloids and other positively charged, heterocyclic compounds of dietary origin; cationic drugs of therapeutic or recreational use; or other cationic toxins of environmental origin (e.g., nicotine). The secretory process is also a site of clinically significant interactions between OCs in humans. For example, therapeutic doses of cimetidine retard the renal elimination of procainamide (408, 409) and nicotine (17).
Until recently, models of the cellular basis of renal OC secretion typically depicted a single basolateral entry step and a single luminal exit step, a simple view that effectively explained existing physiological data. That view is now known to be an oversimplification of the suite of cellular events that underlies renal OC transport. In developing a model for the functional basis of this complexity, it is useful to consider the "type I" and "type II" classifications for different structural classes of organic cations developed to describe OC secretion in the liver (274). In general terms, type I OCs are comparatively small (generally <400 mol wt) monovalent compounds, such as tetraethylammonium (TEA), tributylmethylammonium (TBuMA), and procainamide ethobromide. Type II OCs, which are usually bulkier (generally >500 mol wt) and frequently polyvalent, include d-tubocurarine, vercuronium, and hexafluorenium.
Figure 1 shows a model for transcellular secretion of OCs by the proximal tubule that is consistent with observations obtained in studies with isolated renal plasma membranes and intact proximal tubules (342, 343) and that is supported by recent molecular data. OCs enter the cell from the blood across the peritubular membrane. For type I OCs, this entry step involves either an electrogenic uniport (facilitated diffusion), driven by the inside-negative electrical potential difference (PD) (407), or an electroneutral antiport (exchange) of OCs (82, 407) [it is likely that these two mechanisms represent alternative modes of action of the same transporter(s); Ref. 44]. The negative PD is sufficient to account for an accumulation of OCs within proximal cells to levels 10–15 times that in the blood. Studies by Ullrich and colleagues (468, 476) on the structural specificity of inhibition of peritubular OC transport in microperfused rat proximal tubules in vivo indicate a clear correlation between an increase in substrate hydrophobicity and an increase in inhibitory effectiveness, although it is also clear that steric factors influence the interaction of type I OCs with basolateral transporters (e.g., Refs. 15, 470). The marked hydrophobicity of many type II OCs probably results in a substantial diffusive flux across the peritubular membrane, providing an alternative, electrically conductive avenue for entry into proximal cells.
|
We can summarize the current, overall understanding of the cellular processes associated with secretion of organic cations as follows: type I OCs enter proximal cells across the peritubular membrane via electrogenic facilitated diffusion and leave cells across the luminal membrane via electroneutral exchange for H+. Type II OCs diffuse into proximal cells across the peritubular membrane and are exported into the tubule filtrate via the primary active MDR1 transporter. Importantly, considerable overlap appears to exist in the selectivity of these parallel transport pathways (502; see also Ref. 403). It is also evident, as discussed below, that OC transport across the basolateral and luminal membranes of renal proximal tubules involves the parallel activity of several distinct transport processes.
B. Molecular Characteristics of Renal OC Transporters
1. OCT family of OC transporters
Several alternative nomenclatures are currently in use for describing members of the OCT family of transporters, and we will commonly provide two of these when we introduce each member of the family. The first is the designation within the solute carrier superfamily (SLC), as defined by the Human Genome Nomenclature Committee, in which the OCTs (and the related OATs) are members of group SLC22A. The second reflects the classification system developed by Saier for transport proteins (the Transport Commission; Ref. 361), within which the organic cation transporters are listed as members of the major facilitator superfamily (MFS) and are classified as 2.A.1.19: 2 (electrochemical potential driven porters), A (uniporters, symporters, antiporters), 1 (major facilitator superfamily), 19 (organic cation transporter family). Schömig et al. (382) suggested the alternative name "amphiphilic solute facilitator" (ASF) family for this group of transporters within the MFS, focusing on the characteristics of substrate specificity which link all members of this group. Common structural features of MFS proteins, which are generally shared by the family 19 OCT transporters, include 12 putative transmembrane spanning domains (TMDs) and several highly conserved structural motifs, including a striking degree of conservation for a 13-residue sequence found between TMD2 and TMD3: G-[RKPATY]-L-[GAS]-[DN]-[RK]-[FY]-G-R-[RK]-[RKP]-[LIVGST]-[LIM]. Figure 2 shows the remarkable conservation of this "signature sequence" for 37 of the cation transporting members of the OCT family.
|
Selected amino acids are conserved in all, or virtually all, of the OCT family members cloned to date. These include 4 cysteines and 10 prolines, which suggest an importance for these residues in establishing the secondary structure of this group of proteins. A number of charged amino acids are also conserved, and their potential role in the function of these organic electrolyte transporters is discussed in upcoming sections.
A) OCT1. I) Structure. OCT1 (SLC22A1; 2.A.1.19.1) was cloned in 1994 by Koepsell and colleagues (142) from a cDNA library prepared from rat kidney.1Orthologs were subsequently cloned from three additional mammalian species [mouse (132), rabbit (451), and human (549)] displaying from 95 to 78% identity with rOCT1. In addition, related sequences have also been cloned from C. elegans (ceOCT1; 28.4% identical; Ref. 537) and Drosophila melanogaster (drOCT; 33.7% identical; Ref. 450). The mammalian isoforms vary in length from 554 to 556 amino acids.
By hydropathy analysis, OCT1 orthologs appear to have 11 or 12 TMDs. It is generally acknowledged that 12 TMDs are likely, in accord with the predicted secondary structure profiles of most MFS members (334), and the recent high-resolution analyses of the structures of the MFS members, LacY (lactose transporter of Escherichia coli; Ref. 1) and GlpT (glycerol-3-phosphate transporter of E. coli; Ref. 174). The secondary structure predicted from a 12 TMD configuration includes intracellular NH2 and COOH termini, a very long extracellular loop between TMD1 and TMD2 (which typically includes three to four N-linked glycosylation sites), and a comparatively long intracellular domain between TMD 6 and TMD 7 (Fig. 3). Conserved motifs within OCT1 across all species include the following (426): 13 cysteine residues, 25 proline residues, 3 N-linked glycosylation sites2 (N 71, 96, 112), 3 protein kinase C (PKC) consensus sites3 (Ser-285, Ser-291, and Ser/Thr-327), 3 protein kinase A (PKA) consensus sites (Thr-235, Thr-296, and Thr-347), 1 protein kinase G (PKG) consensus site (Thr-347), 2 casein kinase II (CKII) consensus sites (Ser-333 and Thr-524), and 1 Ca2+/calmodulin kinase II (CaMII) consensus site (Thr-347). Of these sites, 7 cysteine residues, 17 proline residues, 1 N-linked glycosylation site (N 72), 1 PKA site (Thr-347), 1 PKC site (Ser-285), 1 PKG site (Ser-347), and 2 CKII sites (Ser-333 and Thr-524) are conserved across OCT1, OCT2, and OCT3.
|
The human gene for OCT1 consists of seven exons and six introns (152). Several alternatively spliced variants of OCT1 have been described. The rOCT1A variant (548) arises from a 104-bp deletion between bp 451 and 556 of rOCT1. Although the resulting protein product lacks putative TMDs 1 and 2 and the large extracellular loop that separates those two TMDs, the gene product supports mediated transport of TEA (548). In the human, four alternatively spliced isoforms are present in human glioma cells (152): a long (full-length) form and three shorter forms. Only the long form (hOCT1G/L554) supports transport when expressed in HEK293 cells. The long form and one of the shorter forms (hOCT1G/L506) are present in human liver cDNA (152).
II) Tissue and cellular distribution and localization. OCT1 appears to be expressed in many tissues, although intriguing species differences have been noted. OCT1 is expressed in the liver of all species tested [rat, Northern (142); mouse, cDNA library (132); human, Northern (129, 549); rabbit, RT-PCR (451)]; however, expression in the kidney shows more variability among species. In rats, OCT1 is heavily expressed in the kidney (cortex) (402), with substantial expression in the liver (142). It is also well expressed in the colon, small intestine (142), skin, and spleen (402). In rat liver, immunocytochemistry shows that OCT1 expression is restricted to the basolateral (sinusoidal) aspect of hepatocytes (276). This distribution was confirmed in Western blots, which showed a strong 67-kDa band in isolated sinusoidal membranes and a weak reaction (perhaps due to sinusoidal contamination) with isolated biliary membranes (276). Western blots of isolated rat renal membrane vesicles also indicated that OCT1 is restricted to the basolateral domain of cortical cells (489). In a thorough immunocytochemical assessment of OCT distribution in rat kidney (195), OCT1 expression was restricted to the basolateral membrane of proximal tubule cells, with the greatest expression apparent in early (S1) segments of superficial, midcortical, and juxtamedullary nephrons. rOCT1 was also evident in proximal S2 segments, although not to the same extent as in S1 segments. Expression in late (S3) proximal tubule, although evident, was markedly less than that in the early (S1) proximal tubule. At the transition between outer stripe (S3) and inner stripe of the outer medulla, rOCT1 expression was clearly limited to the late S3 segment (and not in the thin descending limb of Henle's loop). Determination of rOCT1 mRNA in renal tubules by in situ hybridization revealed a distribution for the mRNA qualitatively similar to that for the protein shown by immunocytochemistry. However, the relative abundance of the mRNA in the different tubule segments differed from that of the protein. Whereas the highest protein content apparent was in the S1 segments, the highest concentration of mRNA appeared to be in the S2 and S3 segments (195). No rOCT1 expression (protein or mRNA) was evident in cells of glomeruli or distal tubules (see also Ref. 488).
Tissue distribution of rOCT1 during embryonic development appears to differ from that in adults. During early development (up to day E16), expression of rOCT1 mRNA is restricted to liver and brain (335). After E16, rOCT1 mRNA is expressed in the kidney, rising through birth (335) and thereafter (402). The increase in rat OCT1 mRNA from birth through adulthood (which is paralleled by increases in mRNA for OCT2 and OCT3; Ref. 402) coincides with increases in TEA uptake into mouse renal slices through this developmental period (93).
In rabbits, OCT1 is clearly expressed in the kidney (and small intestine), but it appears to be more strongly expressed in the liver (451). In humans, OCT1 expression appears primarily in the liver (129, 549). In one study (300), immunocytochemistry, Western analysis, and quantitative RT-PCR failed to demonstrate the expression of OCT1 in human renal tissue. However, the demonstration in some other studies [by Northern analysis (549) and by RT-PCR (129)] of modest expression of OCT1 in human renal tissue has led to the suggestion that this isoform may play a "housekeeping role" with respect to OC transport in human tissues other than the liver, where it may play a predominant role in secretion of type I OCs (129, 218).
III) Functional characteristics.
OCT1-mediated transport activity has been functionally expressed in several heterologous expressions systems: Xenopus oocytes (43, 44, 61, 88, 129, 130, 132, 142, 304, 307, 451, 502, 548, 549), HEK-293 cells (32, 43, 130, 256, 270, 276), HeLa cells (550–552), Madin-Darby canine kidney (MDCK) cells (488, 489), Chinese hamster ovary (CHO) (191), and COS-7 cells (553). All the orthologs tested support saturable transport of TEA, with apparent Michaelis constants ranging from 38 to 251 µM. This transport is sensitive to the trans-membrane electrical potential. Increasing the external concentration of K+ (which depolarizes the oocyte membrane by 
20–50 mV; Refs. 142, 549) reduces the rate of TEA uptake into Xenopus oocytes injected with cRNA for rOCT1 (142) and hOCT1 (549). The degree of inhibition can be small, as shown by the observation that in oocytes injected with mOCT1 cRNA, the reduction of TEA uptake associated with elevation of external K+ fails to be statistically significant (132).
Direct evidence that OCT1-mediated transport of TEA and other type I cations results in an inward current was obtained in studies using voltage-clamped oocytes injected with rOCT1 (44). The relationship between inward current and the external concentration of substrate shows saturation, with the apparent Michaelis constants varying as a function of the holding potential. In one study, for example, as the holding potential was changed from –90 to –10 mV, the apparent Kt (Michaelis constant; substrate concentration resulting in half-maximal transport) for TEA uptake increased from 14 to 49 µM (44). Interestingly, half-saturation constants derived from saturation of inward currents are typically severalfold lower than those measured using conventional radiotracer techniques. For example, as noted above, the apparent Kt for TEA uptake mediated by rOCT1 in oocytes is on the order of 100–250 µM (61, 142). In addition, the rate of TEA uptake in oocytes inferred from the inward current supported by rOCT1 is about four- to fivefold greater than the rate of radiolabeled TEA uptake measured in parallel experiments (44). This difference may reflect a degree of substrate-induced ion-leak current through rOCT1 (e.g., Ref. 509). It may also reflect the fact that transport of labeled substrates into oocytes (and other cells) that are not voltage clamped can be expected to depolarize the cells, thereby decreasing the driving forces that, in turn, influence (in this case, decrease) the maximal rate of transport (44).
Every OCT1 ortholog tested to date, including mOCT1, supports the saturable electrogenic transport of TEA and other small, type I OCs (e.g., choline, tetramethylammonium, NMN, dopamine) into oocytes (43, 44, 88, 130, 304). Initial studies of electrical currents suggested that rOCT1 also supports electrogenic transport of several large, mono- and divalent type II OCs, including quinine, quinidine, d-tubocurarine, and pancuronium (44). However, subsequent experiments (304) indicated that the electrogenic currents produced by exposure of rOCT1-expressing oocytes to large type II OCs actually reflect inhibition by these compounds of the rOCT1-mediated electrogenic efflux of endogenous type I OCs (e.g., choline) (304). Nevertheless, additional evidence supports the conclusion that the protonated form of quinidine, at least, can serve as a transported substrate of rOCT1 (502).
In addition to operating as an electrogenic uniporter, OCT1 can also mediate OC/OC exchange (44, 88, 304, 550–552). Preloading Xenopus oocytes with unlabeled TEA, for example, stimulates the uptake of [3H]MPP by human, rabbit, mouse, and rat OCT1 (88). The symmetry of this type of trans-effect is apparent in observations of accelerated efflux of preloaded [3H]MPP from rOCT1-expressing oocytes in the presence of inwardly directed gradients of unlabeled TEA (44) or MPP (304). Human OCT1 also supports trans-stimulation of both influx and efflux (of TEA), but quantitative differences in the extent of these stimulated fluxes produced by some substrates (e.g., TBuMA) have led to the suggestion of asymmetrical binding properties on the extracellular versus intracellular face of the transporter (550).
Trans-inhibition of OCT1-mediated transport has also been noted (44, 88, 304, 550–552), usually when cells are exposed to compounds that are also observed to be high-affinity cis-inhibitors of OCT1-mediated activity. For example, although quinine and quinidine are (at best) transported poorly by rOCT1 (304, 502), these substrates cis-inhibit MPP uptake and trans-inhibit MPP efflux in Xenopus oocytes (304). Similar results are seen with the type II OCs d-tubocuarine and cyanine 863 (304), suggesting that when loaded with bulky, hydrophobic substrates, OCT1 undergoes the conformational changes associated with translocation much more slowly (if at all) than the unloaded transporter. At least two alternative explanations can be offered. First, following the conformational change(s) associated with translocation, hydrophobic substrates may dissociate slowly from the binding site, thereby reducing the "physiological turnover" of the substrate-transporter complex (i.e., the sum total of events associated with net substrate transport: binding of substrate to the transporter at the cis-aspect of the membrane, "translocation" of the substrate-transporter complex, dissociation of substrate from the transporter at the trans-face of the membrane, and a second translocation event that returns the transporter to a cis-facing conformation). Second, hydrophobic cations may diffuse across the membrane and exert what amounts to a cis-inhibition of efflux at the cytoplasmic face of the transport protein (also see Ref. 34).
Because of the importance of proton gradients in energizing the active flux of OCs across the luminal membrane of proximal tubules, via OC/H+ exchange, considerable attention has been paid to the effect of pH on the activity of all the cloned OC transporters to determine if one of these could function as such an exchanger. When rOCT1 is expressed in Xenopus oocytes, TEA uptake is unchanged over an external pH range of 6.5–8.5 (142). Moreover, when rOCT1 is stably expressed in MDCK cells, TEA uptake across the basolateral membrane shows a fivefold increase as external pH is increased from 5.4 to 8.4 (489), whereas rOCT1-mediated efflux of TEA from these cells is not influenced at all over this same range of extracellular pH. These data indicate that mediated exchange of TEA for H+ is an unlikely operational mode for rOCT1 (489).
The other cloned orthologs of OCT1 show a varied, and variable, response to changes in extracellular pH. At one extreme, the comparatively distantly related OCT1 ortholog from C. elegans, when expressed in HRPE cells, shows an almost complete elimination of mediated TEA transport as pH is decreased from 8.5 to 5.5 (537). Mouse OCT1 expressed in Xenopus oocytes shows a more modest (40%) decrease in TEA uptake over the range of 8.5 to 6.5 (132), and human OCT1, also expressed in oocytes, shows little (549) or no (129) response to extracellular pH. The failure of extracellular pH or transmembrane pH gradients to systematically influence OCT1 activity, combined with the previously discussed electrogenecity of OCT1-mediated OC transport, supports the conclusion that OCT1 operates as an electrogenic uniporter that can also support electroneutral OC/OC exchange.
IV) Substrate structural specificity. The issue of substrate selectivity of OCT1 has been examined in three ways: 1) directly, through measurement of transmembrane flux of labeled compounds (or transport-induced current); 2) indirectly, either through determination of the extent of inhibition of transport of a model substrate (e.g., TEA) produced by coexposure to a test agent, or as noted above, by gauging the stimulatory (or inhibitory) effect on transport of the model substrate following imposition of a trans-gradient of the test agent; and 3) by means of introducing mutations into the transporter sequence to gauge the influence of physicochemical alterations in protein structure on interaction with transported substrates. The latter approach, which has provided valuable insights into the mechanisms of transport protein activity (190), has been used sparingly, to date, in studies of OCT activity. Gorboulev et al. (130), reasoning that the binding of cationic substrates likely involves interaction with anionic residues on the surface(s) of OCTs, considered the effect on rOCT1 of mutating each of the six acidic residues found in all orthologs of OCT1, OCT2, and OCT3 (but not in the OCTNs or in the OATs). They observed a 15-fold decrease in the Kt for MPP uptake into oocytes injected with mRNA coding for a mutant rOCT1 in which Asp-475 was converted to Glu-475. In addition, the IC50 values for inhibition of TEA uptake into oocytes expressing the mutant transporter, measured for several n-tetraalkylammonium (n-TAA) compounds, all decreased, but the ratio of the IC50 measured for the wild-type versus mutant transporter increased with alkyl chain length. They concluded that rOCT1 contains a large cation-binding pocket with several interaction domains that may be responsible for high-affinity binding of structurally different cations and that Asp-475 is located close to one of these domains. A study by Chen et al. (60) employing chimeras of rOCT1 and rOCT2 implicated TMDs 2–7 as the locus of differences in the interaction of these homologs with selected nucleosides. Future studies assessing the influence of other regions of and selected residues in the sequence of OCT1 (and the other members of the family of OCT transporters) can be expected to add to the understanding of the molecular basis of substrate- transporter interaction.
The results of studies employing the first two approaches outlined above support the general conclusion, as discussed below, that OCT1 is a polyspecific carrier that transports a diverse array of cationic compounds. However, detailed conclusions of the type required to advance our understanding of the role this process plays in, for example, renal excretion of selected cationic drugs, are frequently complicated by confusion concerning the extent to which results have been influenced by choice of "model system," including both the specific ortholog (e.g., human versus rat) and the heterologous cell system in which the cloned transporter is expressed (e.g., Xenopus oocyte versus cultured mammalian cell).
Measurements of substrate transport, using either radiochemical or electrical methods, show that OCT1 accepts a broad array of type I organic cations, i.e., monovalent OCs with a molecular weight typically less than 400. This includes, in addition to what have become prototypical substrates for renal OC transporters, i.e., TEA, MPP, choline, and NMN, a variety of monoamines (e.g., dopamine, serotonin, and epinephrine) and nucleosides [e.g., 2-deoxytubercidin, cytosine arabinoside, and azidothymidine (AZT)]. Although there are clear similarities in the structural specificity displayed by the several OCT1 orthologs tested to date, there are also significant species differences in substrate selectivity. For example, electrophysiological measurements show that rat, mouse, human, and rabbit OCT1 orthologs all support mediated uptake of the n-TAA compounds, TMA and TEA (88). However, whereas hOCT1 and rbOCT1 also support uptake of tetrapropylammonium (TPrA) and tetrabutylammonium (TBuA), the two rodent orthologs do not (88).
The selectivity of OCT1-mediated transport has also been probed in studies measuring the inhibitory effectiveness of a vast array of cationic compounds (44, 142, 256, 318, 489, 549, 553). It is, however, difficult to discern a pattern in the molecular determinants associated with interaction of (putative) substrates with OCT1. Nevertheless, studies of the effect of increasing alkyl chain length on the inhibitory interaction of n-TAA ions with hOCT1 have revealed a strong correlation between increasing chain length (increasing hydrophobicity) and decreasing IC50 values for inhibition of TEA transport into HeLa cells transiently expressing hOCT1 (15, 550). This observation coincides closely with reports of a strong correlation between increasing hydrophobicity and increasing inhibitory interaction of test agents with contraluminal OC transport in intact rat (468) and rabbit (135) proximal tubules and with OC/H+ exchange in rabbit renal BBMV (536), at least for structurally related compounds (e.g., Ref. 534).
Three-dimensional variations in substrate structure produce steric constraints on substrate binding, although for OCT-mediated transport there has been little attention paid to this issue. In a study employing a computational approach to the development of a putative pharmacophore of substrate binding to hOCT1, Bednarczyk et al. (15) observed a systematic effect of planar hydrophobic mass on binding to the receptor: 4-phenylpyridinium compounds were more effective inhibitors of TEA transport than were 3-phenylpyridinium compounds, which were, in turn, more effective than quinolinium compounds. The deduced pharmacophore consisted of a single cationic recognition site and three hydrophobic features. Inhibitors that interacted most effectively with hOCT1 contained all these features, and weaker inhibitors generally displayed fewer of these sterically defined features. It was telling, however, that despite a strong correlation between predicted and measured IC50 values (r = 0.86), a number of outliers were noted, as in the IC50 values for a "test set" of eight compounds. Algorithms used to develop a pharmacophore typically seek a single, best-fit structure for interaction with a receptor that is assumed to possess a marked degree of structural specificity. However, OC transporters do not display the narrow specificity of the typical receptor, which generally accepts a "best" structure to the exclusion of most others. OC transporters, in contrast, apparently because of the protective role they play, must accept a broad array of substrate structures, including compounds to which the host organism may never have been exposed (e.g., dietary toxins or synthetic drugs). Consequently, one might predict a selective advantage arising from a transport process that can interact effectively with a diverse array of environmental chemicals, thus making it desirable for OC transporters to accept chemical structures that fit a generalized format, rather than one represented by a classical pharmacophore. QSAR analysis, which depends less on steric elements and more on physicochemical properties of the members of the training set, resulted in a model of the basis of substrate binding to hOCT1 (r = 0.95) that more adequately described the members of the test set (15).
Before proceeding further with our discussion of substrate specificity, however, we need to highlight the following issue. There is sufficient variability in reported values for the kinetic interaction of substrates and inhibitors of OCT1 (and other cloned transporters) to render suspect most conclusions concerning the quantitative basis of substrate-transporter interaction. There are at least three sources of this variability in results that may obscure the true structural selectivity of OCT1-mediated transport. These include differences in 1) species, 2) experimental technique or methodology, and 3) expression system. The potential influence of each of these sources of variability must be considered.
The possible influence of species differences in both quantitative and qualitative aspects of substrate specificity of OCT1 (and all other transporters, as well) is obvious. There is ample evidence that changes in a single amino acid residue can exert profound differences in transporter selectivity (e.g., Ref. 512). As noted earlier, substitution of a glutamate for an aspartate at amino acid 475 in rOCT1 decreases the Kt for TEA transport by 15-fold (130). The fact that the orthologs for OCT1 cloned to date generally differ by at least 15% from each other with respect to amino acid identity (despite conservation of many specific residues at what are likely to be key sites for transporter function; Refs. 41, 130, 426), makes it unlikely that any one of these proteins will be quantitatively identical to any other with respect to functional characteristics. Dresser et al. (88) examined this directly by comparing the kinetics of n-TAA interaction with the human, rabbit, rat, and mouse orthologs of OCT1 (expressed in Xenopus oocytes). Significant differences in the inhibition of OCT1-mediated MPP uptake were noted for all the orthologs, with the greatest differences noted between human OCT1 and the two rodent transporters; rabbit OCT1 showed intermediate properties. In addition, whereas both hOCT1 and rbOCT1 supported mediated transport of TMA, TEA, TPrA and tetrabutylammonium (TBA), the rodent orthologs transported only TMA and TEA (although, interestingly, TPrA and TBA were more potent inhibitors of MPP transport mediated by rOCT1 and mOCT1, than by hOCT1 or rbOCT1). Thus the kinetic and selectivity characteristics noted in one species cannot be assumed to hold for any other species, a fact that can have important implications in efforts to extrapolate from animal transport models to humans.
The degree of functional variation among OCT1 orthologs (i.e., species differences) is difficult to assess given the extent of variation in the literature regarding the kinetic characteristics of individual orthologs. As noted above, this variation may reflect differences in technique between laboratories. Values for the Kt or inhibition constant (Ki) for MPP interaction with rOCT1 expressed in oocytes ranges from 3 µM (130) to 60 µM (318); for NMN, the values range from 125 µM (130) to >2 mM (318). The basis of such variation is not clear, but it could reflect differences in the level of expression, methods of analysis, or other aspects related to the technique of measuring transport. Unfortunately, there are no clear "benchmark" values for the kinetic constants of OCT1-mediated transport in oocytes (or any other expression system); the database is not yet large enough.
The other potential source of variability that has not received sufficient attention is the influence of the expression system on the quantitative characteristics of expressed transport. Again, the database is small and examination of the kinetic constants for different OCT1 orthologs expressed in different systems reveals sufficient variability that systematic influences of the expression system used are not clear (Fig. 4). However, a pair of studies by Inui and colleagues (318, 489) suggests that systematic differences between expression systems may exist, at least with respect to the activity of OCT transporters. These studies examined a battery of compounds as inhibitors of the rat orthologs of OCT1 and OCT2 when transiently expressed in Xenopus oocytes (318) or stably expressed in MDCK cells (489). For every test compound, the apparent Ki (for inhibition of TEA transport) was lower in the mammalian cell line than in the oocytes. The extent of the difference was not, however, "predictable." For example, whereas the Ki for TEA self-inhibition decreased about 3-fold when the transporter was expressed in MDCK cells rather than oocytes, the Ki for cimetidine inhibition of TEA uptake decreased more than 50-fold. Indeed, whereas in oocytes the Ki for inhibition of radiolabeled TEA transport by cimetidine was greater than the Ki for inhibition by unlabeled TEA (329 versus 129 µM; Ref. 130), in MDCK cells it was much less (5.7 versus 38 µM). Interestingly, the pattern of these differences was quite similar for rOCT1 and rOCT2. Are such differences noted for all substrates or for all OCT transporters? There are not sufficient data to draw a conclusion. Neither is it known if systematic differences exist between transporters expressed in different mammalian cell lines (e.g., epithelial cells versus nonpolarized cells).
|
B) OCT2. I) Structure. OCT2 (SLC22A2; 2.A.1.19.5) was isolated originally from a rat renal library (317). However, orthologs have now been cloned from human (129), mouse (296), pig (141), and rabbit (553). The initial rOCT2 clone had an open reading frame that coded for a 593-amino acid protein (317). The human, mouse, pig, and rabbit orthologs, in contrast, are 555, 553, 554, and 554 amino acids in length, respectively. Gründemann et al. (144) isolated a clone of rOCT2 with a 555-amino acid open reading frame and suggested that the last 38 amino acids of the original clone were erroneously included due to difficulties associated with sequencing a poly(A) region of the 3'-end of the cDNA. The sequences of the several OCT2 orthologs are 68–70% identical to that of hOCT1 and 82–92% identical to one another. Conserved motifs within OCT2 across all species include 13 cysteine residues, 24 proline residues, 1 N-linked glycosylation site (N 72), 2 PKA consensus sites (Thr-348, Thr-489), 1 PKC consensus site (Ser-286), 1 PKG consensus site (Thr-348), and 3 CKII consensus sites (Ser-334, Ser-472, Thr-525).
Human kidney expresses at least one splice variant of OCT2. Designated hOCT2-A, it is characterized by the insertion of a 1,169-bp sequence arising from the intron found between exons 7 and 8 of hOCT2 (486). This insertion adds an in-frame stop codon resulting in a truncated protein of 483 amino acids that is missing the last three putative TMDs (i.e., 10, 11, 12). Despite the absence of the last three TMDs, hOCT2-A retains the capacity to transport TEA and cimetidine, although guanidine transport is lost. The affinity of hOCT2-A for cationic substrates differs markedly from that of OCT2. In side-by-side experiments of expression in HEK 293 cells, the apparent Kt for TEA was about sevenfold lower in hOCT2-A (63 versus 431 µM). The inhibitory profiles of the truncated and parent transporters also suggest that, for most compounds (including cimetidine, procainamide, and nicotine), hOCT2-A has a higher affinity for substrate than does hOCT2. However, for a few inhibitors (e.g., norepinephrine and dopamine), hOCT2 displays a higher apparent affinity than does hOCT2-A. The physiological role of two closely related OCT2s expressed in the kidney is not evident, although possible differences in regional distribution or functional regulation have been suggested (486).
II) Tissue and cellular distribution and localization. Northern blots show expression of rOCT2 in the kidney, and in neuronal tissue, but not in the liver (42, 317); hOCT2 and mOCT2 show a similar tissue distribution (129, 296). This restricted distribution was confirmed by RT-PCR: rat OCT2 is expressed in rat kidney cortex and medulla, but not in rat liver (489). RT-PCR of adult mouse tissue gives a similar distribution (296). Intriguingly, RT-PCR on isolated tubule segments from rat kidney showed expression of OCT2 not only in superficial and juxtamedullary proximal straight and convoluted tubules, but also in medullary thick ascending limbs, distal convoluted tubules, and cortical collecting ducts (488). OCT2 mRNA levels in male rat kidneys are approximately four times greater than those in female rat kidneys when determined using branched DNA signal amplification (402). This correlates with greater TEA transport into renal slices from male than from female rat kidneys (487). In rats, in situ hybridization revealed that OCT2 mRNA is expressed throughout the proximal tubules within the cortex and outer stripe of the outer medulla, with the heaviest hybridization in cells of the outer stripe (143, 195).
Like OCT1 expression, OCT2 expression in the kidney appears to be restricted to the basolateral membrane of proximal tubule cells. Immunocytochemistry showed that rOCT2 is restricted to the basolateral membrane of mid to late proximal tubules (S2 and S3 segments) (195, 415). Similarly, in the human kidney, OCT2 expression is restricted to the basolateral membrane of proximal tubule cells (300), although it is not clear if the levels of expression differ along the length of the tubule. Functional mapping of OCT1 and OCT2 transport activity in single isolated rabbit renal proximal tubules (RPT) showed that TEA uptake in the S2 segment is dominated by OCT2-mediated transport (553). An early report of OCT2 expression in the apical membrane of human distal tubules (129) may have reflected complications associated with the quality of the tissue. Another line of evidence consistent with the basolateral localization of OCT2 in renal epithelia arises from observations obtained with a fusion protein construct consisting of rOCT2 plus green fluorescent protein (GFP) (423). When rOCT2/GFP was transiently transfected into polarized MDCK cells, fluorescence was localized to the basolateral membrane (423). Moreover, transient transfection of rOCT2/GFP into intact, isolated killifish renal proximal tubules resulted in expression of GFP fluorescence in the basal and lateral aspects of the tubule cells (423). It is noteworthy that in the porcine renal cell line, LLC-PK1, OCT2 appears to be expressed in the apical membrane. This localization is based on the comparison of the selectivity characteristics of pOCT2 expressed in oocytes with those observed for apical versus basolateral OC transport in the LLC-PK1 cells (91, 141).
III) Functional characteristics. OCT2 has been functionally expressed in Xenopus oocytes (5, 34, 42, 61, 129, 317, 318, 423, 425), HEK-293 cells (42, 143), NIH3T3 cells (333), MDCK cells (423, 425, 489), COS-7 cells (553), and CHO cells (13). The characteristics of all the OCT2 orthologs studied to date are qualitatively similar to those of OCT1. All OCT2 orthologs support mediated transport of TEA, with Michaelis constants ranging from 20 to 393 µM. Transport is electrogenic, with inwardly directed gradients of substrate resulting in the generation of inward currents (5, 34, 129, 425). These currents are saturable, with apparent Michaelis constants that are similar to those measured using radiolabeled substrates. Isolated giant patches of rOCT2-expressing Xenopus oocytes have been used to study the electrogenic operation of this process in some detail (34). The patch configuration permits measurements of currents that correspond to the efflux mode of transporter operation. Small, type I substrates, including TEA and choline, produce saturable efflux currents with Kt values for efflux (160 µM and 2 mM for TEA and choline, respectively) that are similar to those measured for the mediated uptake of these compounds. This result is consistent with the assumptions that only membrane potential and concentration gradients, i.e., the electrochemical potential, of an organic cation serve as driving forces for rOCT2-mediated transport and that this transport shows little rectification.
Larger, type II cations, including quinine and TBuA, typically do not produce OCT2-mediated currents (5, 34), but they do frequently inhibit OCT2-mediated transport activity, indicating some type of interaction with OCT2. The nature of this interaction is, however, difficult to predict. For example, the weak base quinine exerts a noncompetitive inhibition of TEA uptake in Xenopus oocytes expressing rOCT2 (5), although the results of inhibition experiments performed at different pH values suggest that this effect reflects nonionic diffusion of the uncharged substrate into the oocyte and a subsequent competitive interaction at the cytoplasmic face of the transporter (5). In contrast, the quaternary ammonium cation TBuA, which is not transported by rOCT2, is a competitive inhibitor of TEA transport by rOCT2. This inhibition presumably reflects TEA's interaction at the extracellular face of the transporter. Decynium-22, cyanine-863, and tetrapentylammonium, all of which have fixed cationic charges and are not transported by rOCT2 (5), are presumably restricted to accessing the extracellular face of the transporter. However, each exerts a noncompetitive inhibition of transport activity, suggesting the presence on the transporter of an allosteric site that may interact with hydrophobic cations.
The OCT2 binding site appears to have different affinities for (at least) selected compounds depending on whether it faces the outward (extracellular) or inward (cytoplasmic) face of the membrane. When rOCT2 was expressed in Xenopus oocytes, the nontransported compounds TBuA and corticosterone blocked both inward and outward currents generated by TEA and choline, as measured in whole cells (inward currents) and excised giant patches (outward currents) (507). However, the apparent affinity of the OCT2 binding site for TBuA was four times higher when the transporter was oriented to face the extracellular face of the membrane, whereas the affinity for corticosterone was 20-fold lower. The data suggest that the substrate-binding site of rOCT2 is like a pocket containing overlapping binding domains for ligands, and these binding domains may undergo separate structural changes.
OCT2, like OCT1, can support OC/OC exchange. Inwardly directed gradients of unlabeled MPP, TEA, and amantidine, for example, accelerate efflux of labeled MPP from hOCT2-expressing oocytes (42).
The influence of proton gradients on OCT2-mediated transport has received considerable attention because of early functional evidence that this transporter resides in the apical, rather than basolateral, membrane of selected (cultured) cells and, in addition, may act as an OC/H+ exchanger (141, 144). The initial studies indicating that pOCT2 may be located in the apical membrane (mentioned above) used decynium-22 as a means to discriminate between apical and basolateral transport of OCs in LLC-PK1 cells (141). Decynium-22 inhibits TEA uptake in oocytes mediated by the pOCT2 ortholog cloned from LLC-PK1 with a Ki of 5.1 nM. This Ki corresponds closely to the measured Ki of 6.7 nM for inhibition of apical TEA transport in confluent monolayers of LLC-PK1 (381) (basolateral TEA uptake in LLC-PK1 cells is not blocked by 30 nM decynium-22; Ref. 381), thus supporting the contention that OCT2 is expressed in the apical membrane of these cells (141). This conclusion has been supported by independent observations based on the selectivity profile of OC transport in LLC-PK1 cells (91).
This functional information on localization has been evaluated in the light of evidence that 1) LLC-PK1 cells can support net transepithelial transport of TEA (364) and cimetidine (16) and 2) apical membrane vesicles isolated from LLC-PK1 cells can support mediated OC/H+ (176). The apical localization of pOCT2 in LLC-PK1 cells would be consistent with the hypothesis that OCT2 is an OC/H+ exchanger involved in active OC secretion in LLC-PK1 cells. This hypothesis was extended from porcine to rodent kidneys on the basis of the selectivity profiles of the rat orthologs of OCT1, OCT2, and OCT3 (EMT) that were interpreted as supporting an apical, rather than basolateral, location of rOCT2 (144). Weakening this last conclusion and calling into question the general interpretation of studies relying on comparisons of substrate selectivity to determine subcellular location of individual transport processes is the clear demonstration by the use of immunocytochemistry and OCT/GFP constructs that rOCT2 is restricted to the basolateral membrane of native proximal tubule cells (see above) (195, 415).
In addition, OCT2 fails to show convincingly the physiological characteristics of the apical OC/H+ exchanger: 1) changes in extracellular H+ concentration exert a modest effect on OCT2-mediated transport (rat, Ref. 317; pig, Ref. 141); 2) trans-gradients of H+ do not stimulate rOCT2-mediated transport (425); and 3) OCT2-mediated transport (unlike OC/H+ exchange, Ref. 533) is electrogenic, as assessed by the observation of transport-induced currents in rOCT2-expressing oocytes (5, 34). In addition, the K+-induced collapse of the apical membrane potential inhibits apical metoprolol uptake in LLC-PK1 cells (91). It has been suggested that the discrepancy between the behavior of OC/H+ exchange as measured in membrane vesicles and the transport activity of OCT2 in heterologous expression systems reflects differences in the experimental procedures involved in measurement of transport (e.g., different buffer systems; Ref. 144). However, inwardly directed H+ gradients do stimulate TEA efflux and outwardly directed H+ gradients do stimulate TEA uptake across the apical membrane of LLC-PK1 cells (364, 430). Nevertheless, all the data taken together suggest that, although OCT2 may be expressed in the apical membrane of some renal cells (or some cultured cells of renal origin), it is unlikely to work as a secondarily active secretory process and, instead, in those cells may be limited to mediating trans-apical fluxes in response to the prevailing electrochemical gradient of the substrate in question.
III) Substrate structural specificity.
OCT2 is a polyspecific carrier with selectivity characteristics that are very similar to those observed for OCT1. All OCT2 orthologs tested to date (rat, human, rabbit, and pig) support the mediated transport of TEA, with Michaelis constants ranging from 20 to 400 µM, with an average value of 150 µM. The observation that, in addition to expression in the kidney, hOCT2 is expressed in neuronal tissues (42) has led to the study of the interaction of monoamine transmitters with this process. The human ortholog of OCT2 transports (in order of increasing Kt) serotonin (80 µM), dopamine (390 µM), histamine (1.3 mM), and norepinephrine (1.9 mM) (42). Since these comparatively low affinity substrates are all relatively small, polar molecules, these findings are consistent with the general view that hydrophobicity is an important criterion in the binding of substrates and inhibitors to OCT transporters (e.g., Ref. 465). However, as in the case of the substrate selectivity of OCT1-mediated transport, the database is too small and the variability in apparent binding and transport constants that exists within and between different orthologs and expression systems is too large to permit any specific conclusions concerning the molecular determinants involved in the transport activity of OCT2.
Of particular importance to future studies of the relative role of the various OCTs in the kidney is finding one or more substrates/inhibitors that can effectively discriminate between the activities of different homologs when expressed in native tissue (where multiple homologs may be coexpressed in the same cell). The parallel selectivity of OCT2 and OCT1 for at least some OCs is evident in the side-by-side comparisons, in two different expression systems, referred to previously (191; see also Ref. 488). It is, however, also evident that some substrates show strikingly different affinities for these two processes. Koepsell and colleagues (5) compared the selectivity of the rat orthologs of OCT1 and OCT2 when expressed in Xenopus oocytes. Whereas several compounds interacted very similarly with both orthologs (e.g., TEA, MPP, NMN), other compounds interacted very differently. Most notably, mepiperphenidol and O-methylisoprenaline had IC50 values for inhibition of transport by rOCT1 that were 60 times lower than the values for inhibition of transport by OCT2. On the other hand, guanidine and corticosterone were 25–35 times more potent in their inhibition of rOCT2 than rOCT1. It is also noteworthy that selected nucleotides, including 2-deoxytubercidine, 2-chlorodeoxyadenosine, and cytosine arabinoside, are transported by rOCT1 but not by rOCT2 (60–62). Similarly, human OCT1 and OCT2 display markedly different relative affinities for n-TAA compounds (90). The rabbit orthologs of OCT1 and OCT2 also display markedly different affinities for selected substrates (553). Whereas the affinities of these transporters for TEA is approximately the same, the apparent affinity of OCT2 for cimetidine and the fluorescent cationic substrate [2-(4-nitro-2,1,3-benzoxadiazol-7-yl)aminoethyl]trimethylammonium (NBD-TMA) is 35–100 times higher than that of OCT1 (when expressed in COS-7 cells).
Exploiting the marked differences in apparent affinity that different OCT homologs have for selected substrates can provide a means to map the functional distribution of OC transport activity in renal cells. The IC50 of 13 µM for cimetidine's inhibition of peritubular TEA transport in isolated single S2 segments of rabbit proximal tubules is close to the IC50 of 6 µM for cimetidine inhibition of TEA transport by heterologously expressed rbOCT2, but differs radically from the IC50 of >500 µM for such inhibition of transport by rbOCT1 similarly expressed (553). Although such comparisons require extrapolation of kinetic values obtained with heterologous expression systems to the quantitative behavior of transporters expressed in native tubules (with the attendant caveats noted previously), these data argue that TEA transport in the S2 segment of rabbit proximal tubule is dominated by activity of OCT2 (see also Ref. 191). Extending the demonstration of functional distribution of transporter activity to different regions of the nephron remains a critical aspect of integrating molecular information into the context of cellular and organ-level physiology.
The recent discovery of altered transport function of hOCT1 and hOCT2 containing single nucleotide polymorphisms present in different ethnic populations (237, 238, 393) has underscored the importance of understanding structure-activity relationships for these processes. For example, 28 variable sites in the hOCT2 gene were discovered in a collection of 247 ethnically diverse DNA samples (Caucasian, African-American, Asian-American, Mexican-American, and Pacific Islander). Eight of these polymorphisms caused nonsynonymous amino acid changes, of which four were present in at least 1% of an ethnic population. These four displayed altered transporter function as seen by a threefold change in Kt values for MPP and TBA, changes that could result in differences in the pharmacokinetics of renal drug excretion between individuals expressing different variants of hOCT2. However, population-genetic analysis suggests that selection has acted against amino acid changes to hOCT2, which may reflect a necessary role of OCT2 in the renal elimination of endogenous amines or xenobiotics (238).
C) OCT3. I) Structure. OCT3 (SLC22A3; 2.A.1.19.6) was initially cloned from a rat placental cDNA library (199). However, orthologs have now been cloned from mouse (521) and human [originally referred to as the extraneuronal monoamine transporter (EMT) (145)]. OCT3 is a 551-amino acid peptide with 12 putative TMDs. The three orthologs of OCT3 display 47–48% sequence identity with hOCT1 and 48–49% sequence identity with hOCT2. Conserved motifs within OCT3 across all species include the following (426): 15 cysteine residues, 25 proline residues, 3 N-linked glycosylation sites (N 72, 99, 119), 2 PKA consensus sites (Thr-351, Thr-437), 2 PKC consensus sites (Ser-291, Thr-297), and 3 CKII consensus sites (Thr-327, Ser-337, Thr-528).
II) Tissue and cellular distribution and localization. Northern analysis indicates that rOCT3 is expressed most heavily in the placenta and intestine, but it is also found in the kidney, heart, brain, and lung (199). Using a branched DNA signal amplific
