INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

MAMMALIAN KIDNEYS have the ability to actively remove substances from circulating renal blood and transport them into the lumen of proximal tubules by a process called "renal secretion." In some cases the process is efficient such that 100% of the substance presented to the kidney is cleared from blood in a single passage, i.e., the renal plasma clearance equals the renal plasma flow. The major, known renal secretory systems are pleiotropic and have been placed into two distinct categories due to the nature of their substrates and inhibitors: the organic cation transporter (OCT) and the organic anion transporter (OAT) systems (18, 19). Classic substrates for the OCT are tetraethylammonium (TEA) and N¹-methylnicotinamide (NMN), whereas p-aminohippurate (PAH) and methotrexate are examples of substrates for the OAT. Classic inhibitors are cimetidine (CIM), cyanine 863, and quinidine for the OCT and probenecid for the OAT. Recently, three OCTs and two OATs have been cloned from rat kidney, proving the existence of two classes of carriers long ago predicted using pharmacological approaches. The OCTs have considerable amino acid sequence homology as well as some homology with the OATs (7, 16, 21, 23, 25, 30).

Some nucleosides are also secreted by human [deoxyadenosine (dAdo), 2-chlorodeoxyadenosine (CldAdo), azidothymidine (AZT)] (9, 10, 17) and mouse [dAdo, 2'-deoxytubercidin (dTub)] kidneys (10, 11). The secretion of dTub by mouse kidney was shown to involve an OCT because: 1) secretion of dTub and TEA, but not PAH, was inhibited in vivo by CIM (15); 2) dTub inhibited the active uptake of TEA, but not PAH, by mouse kidney slices (11); and 3) organic cations, but not organic anions, inhibited the active uptake of dTub by mouse kidney slices (11). Thus we hypothesize that an OCT exists that also recognizes some nucleosides, in particular dTub. To test this hypothesis and to obtain a pure source for the putative carrier, we tested several renal cell lines for their ability to transport TEA by a mechanism inhibitable by dTub or CIM. Among the cell lines examined, OK cells clearly and reproducibly demonstrated an efflux mechanism consistent with active secretion at the apical membrane. This report summarizes our findings regarding this transporter to date. Some of these data have been presented previously in abstract form (2, 3, 14).

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Cell lines. OK (opossum kidney proximal tubule), Caki-1 (human clear cell carcinoma, renal primary), LLC-PK₁ (pig kidney proximal tubule), and MDCK (dog kidney distal tubule) cells were purchased from the American Type Culture Collection (Rockville, MD). Renca (mouse renal carcinoma) cells were obtained from Dr. Isaiah J. Fidler of the Dept. of Cancer Biology, University of Texas M. D. Anderson Cancer Center (28). OK, LLC-PK₁, and MDCK cells were cultured at 37°C in DMEM with 10% (vol/vol) fetal bovine serum. Caki-1 and Renca cells were cultured in McCoy's 5a and Eagle's MEM, respectively.

Uptake and efflux measurement. The OK cells were cultured in 12-well dishes for 10 days. For uptake experiments, the monolayers were incubated in 600 µl of DMEM containing [¹⁴C]TEA (0.1 µCi/ml, 30 µM) at 37°C in the absence or presence of inhibitors (300 µM). After various incubation times, the media were removed by aspiration, and the monolayers were washed twice with PBS (pH 7.4) at 4°C. The cells were lysed with 500 µl of 1 N NaOH for 3 h. The lysates were removed, and the incubation wells were washed twice with 250 µl of 1 N HCl. The lysates and wash solutions were combined and transferred to 7-ml scintillation vials. After the addition of 5 ml liquid scintillation fluid (Research Product International, Mount Prospect, IL), radioactivity was determined in a scintillation spectrometer (model LS 3801; Beckman Instruments, Fullerton, CA).

For efflux experiments, the cell monolayers were preincubated in 600 µl DMEM containing [¹⁴C]TEA (0.1 µCi/ml, 30 µM) at 37°C for 3 h or overnight. The incubation media were then removed, and the monolayers were washed twice with PBS at 4°C. DMEM (500 µl) with or without inhibitors (300 µM) was then added to each well. After various incubation times, 500 µl of the media were transferred directly to scintillation vials for determination of radioactivity as described above. For measurements of the uptake and efflux of dTub, the cells were incubated in the same manner using [³H]dTub (0.1 µCi/ml, 1 µM). Uptake and efflux experiments using Caki-1, LLC-PK₁, Renca, and MDCK cells were performed after 4-7 days of incubation.

Efflux inhibitors. After incubating OK cells with [¹⁴C]TEA (0.1 µCi/ml, 30 µM) for 3 h, efflux was measured after a 20-min incubation in increasing concentrations of potential inhibitors with or without a maximally effective concentration of CIM (1 mM). The IC₅₀ is the concentration that inhibited the CIM-sensitive TEA efflux by 50% as determined by graphical analysis.

Metabolic and other dependence of TEA efflux. After loading OK cells with [¹⁴C]TEA (0.1 µCi/ml, 30 µM) by overnight incubation, the efflux was measured as described above under different conditions as follow. For effects of temperature, the efflux was measured at 25°C and 4°C over a 60-min interval. The effects of pH were determined by incubating the cells at 25°C for 10 min in DMEM at pH 6.5 (buffer by MES) and 7.4 and 8.1 (buffered by HEPES). The effects of metabolic inhibitors were determined by incubating the cells for 60 min at 25°C in the absence of glucose and in the presence of 10 mM deoxyglucose and 10 µM rotenone. ATP levels in the cells following 60-min incubation was determined by high-pressure liquid chromatographic analysis of neutralized, acid-soluble extracts (11).

Xenopus laevis oocytes and [¹⁴C]TEA transport measurements. The plasmid containing the rOCT2 cDNA (from rat kidney cortex) was a gift from Dr. John B. Pritchard (Laboratory of Pharmacology and Chemistry, National Institute of Environmental Health Sciences, Research Triangle Park, NC). The coding sequence for rOCT2 was located between the Not I and Sal I sites of a pSPORT1 vector (Life Technologies, Gaithersburg, MD) with the start site of RNA transcription positioned downstream from a T7 RNA polymerase promoter. To prepare a rOCT2 cRNA, the vector was made linear by digestion with Hind III, after which an in vitro transcription kit (T7 mMessage mMachine; Ambion, Austin, TX) was used to synthesize the 5'-capped cRNA. X. laevis oocytes were collected, defolliculated manually, and injected with 50 nl of water per oocyte with or without rOCT2 cRNA (0.4 ng/nl). Three days after injection, the oocytes were incubated in uptake solution [100 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 1 mM CaCl₂, 10 mM HEPES (pH 7.4)] containing 180 µM [¹⁴C]TEA and 1.8 mM of various compounds (except for CldAdo, 1 mM) for 90 min at 25°C. Where indicated, deoxycoformycin (DCF, 1 µM) was added to the uptake solution to inhibit adenosine deaminase (ADA). At the end of the uptake period, the oocytes were washed three times in ice-cold uptake solution, then lysed in 0.2% SDS with 0.2 N NaOH, and the radioactivity was determined by liquid scintillation counting as described above.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Uptake of TEA and dTub by renal cell lines in culture. In seeking an efflux carrier that may be expressed on the apical membrane of a renal tubular cell, we reasoned that inhibitors might increase the "apparent uptake" at or near steady state. Thus we measured the overnight uptake of a classic OCT substrate, [¹⁴C]TEA, in the presence and in the absence of CIM or dTub (Table 1). The uptake was enhanced by CIM in OK and Caki-1 cells, whereas the uptake was inhibited by CIM in the other cell lines. Enhanced uptake of TEA produced by dTub was significant in OK cells, whereas the trend was toward inhibition of uptake in the other cell lines examined. Thus OK cells represent a source of a putative TEA, and perhaps dTub, carrier that is sensitive to CIM, consistent with earlier data in mouse kidneys (11). [³H]dTub uptake was significantly enhanced by CIM, albeit biologically insignificant, in OK and LLC-PK₁ cells. The increased uptake was most apparent in Caki-1 cells, for which the uptake after a 24-h incubation in the presence of CIM was increased to 212% of control. This effect of CIM probably reflects inhibition of dTub efflux from the cells, since efflux of dTub was inhibited by CIM in this cell line (data not shown).

To further investigate the nature of TEA uptake by OK cells, we measured the uptake as a function of time in the presence and absence of CIM or dTub (Fig. 1). The uptake of TEA by OK cells is relatively rapid for this compound, which is generally excluded from cells in the absence of a carrier, and steady state is achieved after 30-min incubation. Although the uptake was inhibited by CIM at the earliest times examined (5 and 30 min), there was no inhibition by dTub at these times. Both compounds enhanced the uptake after 3-h incubation. These results suggest that both uptake and efflux are sensitive to CIM, whereas only the efflux is inhibited by dTub. To determine possible influences of these compounds on efflux per se, the following experiments were performed.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Effect of cimetidine (CIM) and 2'-deoxytubercidin (dTub) on [14C]tetraethylammonium ([14C]TEA) uptake by OK cell monolayers. OK cells were incubated in [14C]TEA (0.1 µCi/ml, 30 µM) in absence () or presence of 300 µM CIM () or dTub () at 37°C. Data shown are means ± SE obtained from 2 experiments, each performed in triplicate. [14C]inulin () (0.1 µCi/ml, 30 µM) was used to indicate restriction of uptake to the extracellular space.

Efflux of TEA by OK cells. Since the apparent enhanced uptake of TEA seen after 24-h incubation in the presence of CIM or dTub could reflect increased uptake, decreased efflux, or both, we tested effects of these compounds on egress of this substrate from the cells. The cells were loaded with [¹⁴C]TEA by overnight incubation with or without the presence of CIM or dTub. The efflux was then measured in the presence or absence of CIM or dTub (Fig. 2). Efflux from the cells is gradual and almost complete in controls within a 1-h incubation period. During this time, both CIM and dTub inhibit the efflux. These observations suggest the presence of a TEA efflux mechanism in OK cells that is sensitive to CIM and dTub. To determine whether dTub and CIM inhibit the same or different carriers, dose-response effects of dTub were determined in the absence or presence of a maximally effective concentration of CIM (1 mM, Fig. 3). As shown in Fig. 3, dTub did not add to the inhibition produced by CIM, consistent with inhibition at a common site.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Effect of CIM and dTub on efflux of [14C]TEA. After overnight incubation in [14C]TEA (0.1 µCi/ml, 30 µM) with or without 300 µM CIM or dTub, efflux from OK cells into media was measured in absence () or presence of 300 µM CIM () or dTub ().

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Inhibition of CIM-sensitive TEA efflux by dTub. OK cells were incubated in media containing [14C]TEA (0.1 µCi/ml, 30 µM) for 3 h. Efflux of TEA into media was measured after a 20-min incubation in increasing concentrations of dTub in presence or absence of 1 mM CIM; , dTub; , dTub plus 1 mM CIM.

The IC₅₀ values for inhibition of the CIM-sensitive TEA efflux produced by compounds tested to date are summarized in Table 2. None of these inhibitors added to a maximally effective (1 mM) concentration of CIM, as illustrated for dTub in Fig. 3. Classic OCT inhibitors are the most potent, suggesting that this is an OCT that is inhibited by some nucleosides such as dideoxycytidine (ddC), dTub, AZT, and others. Of particular interest are the affinities of decynium 22, CIM, and NMN for this carrier, i.e., higher than those reported for OCT1 and OCT2 (7, 16). The pK_a values for CIM and dTub are 6.8 and 5.5, respectively (11, 13). According to Henderson-Hasselbalch equation, the positive charged moiety of CIM should be ~17-fold that of dTub at pH 7.4. If indeed the positive-charged form of CIM and dTub are the species that inhibit the TEA efflux, then they may have similar affinities to the transporter. Interestingly, TEA is an extremely poor inhibitor of its own efflux. Some structural specificity for inhibition among nucleosides is also apparent in the data summarized in Table 2. Namely, base specificity is indicated by the differing potencies among dideoxyribonucleosides (ranging from 38 µM to more than 3,000 µM for cytosine, guanine, thymine, adenine, and hypoxanthine); and among the arabinosyl derivatives of guanine, cytosine, and adenine. Some sugar specificity is also suggested by the differing effects of the dideoxyribosyl and arabinosyl derivatives of cytosine. Surprisingly, the adenine derivatives known to be secreted by kidney (dAdo, CldAdo) are rather poor inhibitors of the efflux, compared with dTub, AZT, and some of the dideoxyribonucleosides. Nitrobenzylthioinosine, a potent inhibitor of the equilibrative, facilitated diffusion nucleoside transporter, also inhibited TEA efflux by the OK cells; however, the order of potency (25 µM) is suggestive of a carrier other than that known as the "es nucleoside transporter" that is ~1,000-fold more sensitive (1). It should be mentioned that the intracellular concentrations of these inhibitors may differ from that in the medium such that the actual affinities at the site of action may differ from the results shown in Table 2.

Further characterization of the TEA efflux by OK cells. The efflux is profoundly temperature-dependent as illustrated by the reduced rate of efflux at 4°C vs. 25°C (Table 3). Metabolic dependence of the efflux is also shown with ATP depletion by rotenone and deoxyglucose. This treatment reduced ATP levels to 5.1 ± 2.6% of the control value that averaged 1.4 µmol/10⁹ cells (data not shown). Qualitatively similar effects on ATP levels and TEA efflux were observed using sodium cyanide or iodoacetamide and deoxyglucose (data not shown). With the exception of iodoacetamide, these treatments did not reduce cell viability as assessed by trypan blue dye exclusion, i.e., the reduced TEA efflux produced by rotenone and deoxyglucose is not due to general cytotoxicity. There is only a minor influence of omission of sodium on the efflux, suggesting that the flux of TEA from inside to outside is not dependent upon sodium or that the substitution of choline for sodium produces a slight inhibition of efflux under these conditions. Finally, the efflux of TEA is clearly not an organic cation/proton exchange mechanism, since reduced pH failed to enhance the rate of efflux. In summary, the CIM-sensitive efflux of TEA by OK cells under the conditions studied is not coupled to proton exchange and is metabolically dependent and largely independent of extracellular sodium.

Nucleosides fail to inhibit the transport of TEA by a cloned OCT. Recently, two OCTs have been cloned from rat kidney: OCT1 (7) and OCT2 (16). These proteins have ~70% homology, and, although subject to validation, they probably function as basolateral carriers enhancing the uptake of TEA into the proximal tubule cells. As with the TEA efflux shown herein, classic OCT inhibitors inhibit the uptake induced by cRNAs from these cloned cDNAs in the X. laevis oocyte translation system (7, 16). The cRNA for the OCT2 enhances TEA uptake by frog oocytes markedly, and the enhanced uptake is inhibited by quinidine and verapamil (Fig. 4). However, none of the nucleosides or nucleoside analogs tested showed significant effect on the OCT2 cRNA-dependent uptake, including dTub, AZT, and ddC, which were all active toward TEA efflux by OK cells (Table 2). In short, this finding suggests that inhibition of TEA transport by nucleosides is not a phenomenon common to all OCTs. Additionally, in similar experiments, the cRNA for OCT2 had no effect on the uptake or efflux of [³H]dTub by frog oocytes (experiments not shown), indicating that it is not a carrier for this nucleoside that is secreted by mouse kidney (10, 15).

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Organic cation transporter OCT2 cRNA enhanced uptake of TEA by X. laevis oocytes. X. laevis oocytes (15-20 per group) were injected with 50 nl (20 ng) rOCT2 cRNA per oocyte. Control group was injected with 50 nl of water. Three days after injection, oocytes were incubated in 180 µM [14C]TEA with 1.8 mM various inhibitors [except for 2-chlorodeoxyadenosine (CldAdo), which was present at 1 mM] for 90 min. Oocytes were then washed and lysed as described in MATERIALS AND METHODS. Uptake of [14C]TEA was determined by liquid scintillation spectrometry. DCF, deoxycoformycin; AZT, azidothymidine; ddC, dideoxycytidine; AraC, adenine arabinoside. * Significantly different from control, P  0.05, as determined by unpaired Student's t-test.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

The mammalian kidney plays a major role in maintaining an extracellular fluid compatible with life. One mechanism by which the kidney accomplishes this task is via secretory systems that improve the elimination of toxins and drugs, transporting solutes from the blood and extracellular space across the tubular epithelia into the lumen of the nephrons. Barriers exist, therefore, on either side of the tubular epithelia such that solutes must enter the cell via the basolateral membrane, traffic inside the cell, and be transported into the lumen across the apical membrane. Characteristics of the putative carrier proteins involved at the basolateral and apical membrane are summarized in a recent classic textbook of pharmacology (8). With regard to organic cations, the active transport is thought to occur at the apical membrane; however, mediated movement of model compounds such as TEA occurs in membrane vesicles prepared from either the basolateral or apical membranes. Passage across apical membranes is primarily via a proton exchanger, being enhanced by increasing hydrogen ion concentrations. The gradient for protons, from lumen to intracellular fluid, is favorable in the forming urine to drive this exchange mechanism. This proton/organic cation exchanger has been demonstrated in experiments using various renal brush-border membranes, rabbit proximal tubules, and cultured OK cells (4, 29). In the latter case, the rapid uptake (30 s) of TEA by OK cells was studied and shown to be influenced by a proton gradient (29). The mediated movement of organic cations by basolateral membranes is not coupled to proton exchange. The cDNAs encoding some of these transporter proteins have been recently cloned from kidneys of rat, human, mouse, and rabbit as well as a pig kidney cell line (5, 6, 7, 12, 16, 22, 24, 27, 30). The OCTs are ~70% homologous to one another and are ~30% homologous to the rat OAT (21, 23, 25). Studies reported to date suggest that these carrier proteins reside on the basolateral membrane, i.e., they serve to facilitate the uptake of organic cations into the cell, with one exception. That is, OCT2 is said to be localized to the apical membranes of distal tubular cells (5, 6), an unexpected finding since OCT has traditionally been considered a proximal tubular function. However, as seen herein (Table 1), there is a CIM-sensitive carrier for TEA in the dog kidney distal tubular cell line, MDCK. It has been suggested that the distal tubular location may serve to reabsorb some cations such as dopamine that is synthesized in the proximal tubules (5). Recently, a human fetal liver proton/organic cation antiporter was cloned (hOCTN1, Ref. 26) which is also expressed in kidney. Metabolic dependence of TEA uptake was demonstrated in hOCTN1 transfected cells, and the amino acid sequence has a sugar transporter protein signature and a nucleotide binding site (26); however, the possible relevance of this TEA carrier to that reported herein is unknown.

Our interest in OCT arose from measurements of Ado and dAdo in the body fluids of a child with severe combined immunodeficiency disease (ADA deficiency) and in cancer patients treated with the ADA inhibitor, DCF (10). The renal handling of Ado and dAdo is paradoxically different, i.e., Ado undergoes a net reabsorption, whereas dAdo is secreted. The renal secretion of dAdo was confirmed to occur in mice (10), and to study the possible mechanism, we used the nonmetabolized and nontoxic dAdo analog, dTub (11). This compound was also secreted by mouse kidney, and the secretion was inhibited by CIM (15). Uptake of dTub into mouse kidney slices was energy dependent, and the uptake was inhibited by organic cations to a greater extent than by organic anions. Conversely, dTub inhibited the uptake of TEA but not PAH by the mouse kidney slices. Taken together, these criteria satisfy the requirements of a substrate for "classic" OCT. Measurement of dTub transport per se in cultured cells has been problematic, i.e., the compound has sufficient lipid solubility that passive diffusion frequently overshadows facilitated processes (unpublished observations). This problem should not be interpreted to indicate the situation that exists in the intact organ, i.e., it is difficult if not impossible to duplicate the whole organ condition where secretion of dTub definitely occurs. In an effort to identify a dTub transporter, therefore, we used the impermeable solute, TEA. As such, our goal was to identify a transporter in renal cells that was sensitive to dTub and CIM. We further assumed that the carrier most likely to be novel would be an efflux mechanism located on the apical membrane for reasons described above. OK cells appear to possess such a carrier based upon the preliminary characterization of TEA efflux reported herein. Namely, the efflux is sensitive to CIM and other OCT inhibitors, and it is also inhibited by dTub and other nucleosides (Fig. 3; Table 2). The observation that dAdo and CldAdo fail to inhibit the TEA efflux by OK cells (Table 2) suggests that this may not be the mechanism by which they are secreted. Renal secretion of CldAdo in pediatric cancer patients equaled the renal plasma flow, indicative of very efficient, active transport (9). The observation that CIM failed to inhibit the secretion of dAdo in mice under conditions in which it did inhibit that of TEA and dTub (15) also suggests that there may be other carriers for dAdo and CldAdo. However, failure of TEA to inhibit its own efflux by OK cells (Table 2) suggests that one should not interpret failure of putative substrates to inhibit too broadly. Clearly, much is yet to be learned regarding the various steps and carriers in the process of renal secretion, and proof of the mechanisms involved will come only after cloning and evaluation of all relevant carriers. To that end, this work identifies a TEA efflux transporter in OK cells that is consistent with an energy-dependent, nucleoside-sensitive process. Work in progress using OK cell membrane vesicles will define further the nature of this putative, novel carrier, e.g., ATP dependence, nucleoside sensitivity.