Organic Anion and Cation Transporters in Renal Elimination of Drugs

Characteristics of Organic Anion Secretion in Proximal Tubules

Initial experiments revealed proximal tubules as the principle site of organic anion secretion in various species (for earlier literature see ). In most of these studies, p -aminohippurate (PAH) served as a model organic anion, and the substrate specificity of the so-called PAH transport system in the basolateral membrane was extensively tested by stop-flow competition experiments with rat kidneys in situ . This system preferred amphiphilic compounds carrying one negative charge or two negative charges separated 6–7 Å by a hydrophobic moiety and being able to form multiple hydrogen bridges. The “PAH transport system” turned out to interact with a multitude of chemically unrelated molecules including many drugs.

PAH uptake across the basolateral membrane occurred by an exchange against intracellular α-ketoglutarate. Thereby, PAH uptake is a tertiary active process, with the Na ⁺ ,K ⁺ -ATPase establishing the extracellular-to-intracellular Na ⁺ gradient (primary active process), the sodium-dicarboxylate cotransporter utilizing this Na ⁺ gradient for intracellular α-ketoglutarate enrichment (secondary active process), and the PAH/α-ketoglutarate exchanger using the intracellular-to-extracellular α-ketoglutarate gradient for PAH uptake against the inside negative membrane potential (tertiary active process; see also Figure 72.1 ). In addition, α-ketoglutarate absorbed from the primary urine or derived from metabolism can drive PAH uptake. The arrangement of three co-operating transporters is conserved among species including man. As opposed, earlier experiments on PAH release across the apical membrane revealed species differences.

With the advent of molecular cloning and heterologous expression techniques, several proximal tubular transporters for organic anions have been characterized. Figure 72.1A shows transporters that are involved in the secretion of anionic drugs and toxins, or are targets of drugs, in the human proximal tubule. These include OAT1, OAT2, OAT3, OATP4C1, and MRP6 in the basolateral cell membrane, and OAT4, OAT10, URAT1, NPT4, MRP2, and MRP4 in the apical membrane. In rodent proximal tubules, Oat1, Oat3, Oatp4c1, and Mrp3 were found in the basolateral membrane, and Oat2, Oat5, Urat1, Oat-K1/2, Oatp1a1, Npt1/Oat _v 1, Mrp2, and Mrp4 in the apical membrane ( Figure 72.1B ). Thereby, both cell sides contain antiporters (OATs/Oats, URAT1/Urat1, OATP4C1/Oatp4c1), ATP-driven systems (MRPs/Mrps) and a voltage-driven system (NPTs/Npts). Taken together, the picture of organic anion transport became more complicated than it was anticipated from earlier in vivo and in vitro functional studies.

The Organic Anion Transporter 1 (OAT1/Oat1; SLC22A6/Slc22a6 )

The Organic Anion Transporter 1 (humans, OAT1; other species, Oat1) was cloned from man, monkey, pig (p), rabbit (rb), rat (r), and mouse (m). OAT1/Oat1 is the member A6/a6 of family SLC22/Slc22 , that consist of at least 25 different genes. The gene coding for OAT1 was located to chromosome 11q12.3 paired with that of OAT3. OAT1 mRNA is expressed in kidneys, and to a small extent also in brain and other tissues. In rats, mRNA expression rose from day 10 after birth and reached a maximum in adult animals whereas, in mice, Oat1 mRNA did not rise before day 25 after birth. Importantly, gender differences were observed in adult rats and mice with higher Oat1 mRNA expression in male animals.

Immunohistochemical studies localized OAT1/Oat1 protein to the basolateral membrane of proximal tubules of human, monkey, rat and mouse kidneys. In humans, OAT1 was found along the complete proximal tubule whereas rOat1 was highest in the S2 segment. Immunohistochemistry in adult rats revealed more Oat1 protein in male than in female animals. Whether gender differences are present in humans is not known (for reviews on gender differences see ).

The expression of human OAT1 was under positive control of hepatocyte nuclear factors 1α (HNF-1α) and 4α (HNF-4α). An increased expression of rOat1 was found after treatment of kidney slices or proximal tubules with insulin and epidermal growth factor as well as after bile duct ligation in rats whereas, in several other circumstances such as fever, endotoxinemia, renal insufficiency, ischemia-reperfusion, and ureter obstruction, the abundance of rOat1 was—at least transiently—decreased (reviewed in ). A down-regulation of Oat1 expression most probably decreases the renal excretion of anionic drugs. In this regard, it is interesting that inhibition of COX-2 attenuated the endotoxin-induced down-regulation of rOat1.

Short-term down-regulation involves a protein kinase C (PKC) activity-dependent endocytosis of OAT protein. Thereby, the protein itself is not phosphorylated, but possibly ancillary proteins such as caveolin 2. The appearance of immunoreactive rOat1 in vesicles beneath the basolateral membrane after mercury treatment of rats suggests that endocytosis of Oat1 may occur also in vivo .

Several non-synonymous single nucleotide polymorphisms have been described for human OAT1 (summarized in ). Except for the amino acid change R50H which led to changes in the K _m for adefovir, cidofovir and tenofovir, other amino acid exchanges did not result in altered transport function, at least for the tested substrates.

Following heterologous expression, OAT1/Oat1 proved to be a p -aminohippurate (PAH)/α-ketoglutarate antiporter with in vitro affinities for PAH between 5 and 50 µM (range, 3.1 to 430 µM ). A prototypical, though not OAT1-specific inhibitor is probenecid with K _i or IC ₅₀ values between 1.4 and 18.6 µM. Numerous compounds were tested as possible substrates of OAT1/Oat1 either by checking their inhibition of the uptake of labeled PAH (or other labeled or fluorescent substrates), or by direct demonstration of transport of labeled test compounds (results compiled in ). Among endogenous substrates of OAT1/Oat1 are, e.g., medium-chain fatty acids, α-ketoglutarate, cAMP and cGMP, prostaglandins E ₂ and F _2α , urate, and acidic neurotransmitter metabolites. Drugs interacting with OAT1/Oat1 included β-lactam antibiotics, antiviral drugs, non-steroidal anti-inflammatory drugs, diuretics, ACE inhibitors, angiontensin II receptor 1-antagonists, and methotrexate. Furthermore, OAT1/Oat1 interacted with uremic toxins, environmental toxins, and carcinogens. Thus, OAT1/Oat1 covers a wide substrate spectrum, is involved in renal drug handling, and contributes to nephrotoxicity. The interaction of several compounds with the same transporter, OAT1/Oat1, can lead to drug-drug interaction during renal excretion, alter pharmacokinetics and cause potentially serious side effects (for examples see ).

For mouse Oat1, a quantitative structure-activity relationship was determined. For mono-anions, interaction with mOat1 increased (or K _i decreased) with increasing mass and hydrophobicity, e.g., with the chain length of fatty acids. For di-anions, steric and electrostatic factors became the main determinants of affinity. For transport velocity, other factors seemed important because no correlation was found between V _max and K _m or K _i . In general, QSAR on mOat1 was in agreement with earlier data obtained for PAH transport in rat kidney in situ . Mutational studies on OAT1/Oat1 were performed to find out residues important for transport function: amino acid residues surrounding the putative substrate binding site of human OAT1 include R466 (transmembrane helix TMH11), K382 (TMH8), Y353, Y354 (both in TMH7), and F374 (TMH8) (summarized in ). R466 is also important for the interaction of OAT1 with chloride that increases the V _max of the transporter, but has no effect on its affinity towards PAH.

An Oat1 knockout mouse was generated which did not show any gross abnormalities. The clearance of PAH, furosemide, bendroflumethiazide, and a number of endogenous hydroxyl-substituted short chain fatty acids was decreased. Thus, mOat1 contributes to the renal excretion of diuretics and of hitherto unknown endogenous metabolites.

The Organic Anion Transporter 2 (OAT2/Oat2; SLC22A7/Slc22a7 )

OAT2/Oat2 was cloned from man, rat, and mouse. The gene for human OAT2 is located on chromosome 6p21.1. In humans and male rats, the expression of OAT2/Oat2 mRNA was high in liver and low in kidneys. In female rats, Oat2 expression in kidneys exceeded its expression in liver. In male mice, message for Oat2 was expressed in kidneys, but hardly in liver whereas, in female mice, message was found both in liver and kidneys. Immunolocalization studies revealed human OAT2 at the basolateral membrane of proximal tubule cells. In rat and mouse kidneys, Oat2 appeared to be present in the apical membrane of proximal tubule cells and exhibited a higher expression in female animals. Thus there are gender and species differences in Oat2 expression.

In liver and kidneys, OAT2/Oat2 expression was under positive control of HNF-1α and HNF-4α. In diabetic rats and in cisplatin-treated mice, a decreased renal Oat2 expression was found. As regards single nucleotide polymorphisms, three non-synonymous SNPs were reported the functional consequence of which is unknown (see ).

Due to conflicting results in the literature, it is unclear whether OAT2/Oat2 interacts with and is driven by dicarboxylates. High affinities of human and rat OAT2/Oat2 were reported for prostaglandin E ₂ and prostaglandin F _2α . Other endogenous substrates of OAT2 are nucleobases, nucleosides and nucleotides such as adenine, adenosine, GMP, GDP, GTP, cGMP, and cAMP. Urate was transported by OAT2 with an apparent K _m of 1.17 mM, suggesting that OAT2 contributes to proximal tubular urate secretion. Among drugs, some diuretics (hydrochlorothiazide, furosemide), pravastatin, sartanes (losartan and telmisartan), cephalosporins, erythromycin, histamine receptor-2 blockers, NSAIDs, and uricosurics (probenecid, benzbromarone) interacted with OAT2. High affinities were reported for antineoplastic drugs (methotrexate, fluorouracil, taxol).

In humans, the basolateral localization of OAT2 would favour a role in organic anion (urate) and drug excretion whereas, in rodents, the apical location of Oat2 is hard to reconcile with secretion. Thus, more information is needed to fully appreciate the function of OAT2/Oat2.

The Organic Anion Transporting Polypeptide 4C1 (OATP4C1/Oatp4c1; SLCO4C1/Slco4c1 )

OATP4C1 was cloned from human and rat kidneys and in both species, OATP4C1/Oatp4c1 proteins were found in the basolateral membrane of proximal tubule cells. The gene for human OATP4C1 was located on chromosome 5q21.2. In humans, the expression was found in kidneys, liver, and weakly in lungs. In rats, predominant expression was found in the kidneys and a weaker one in the lungs. Gender differences were not apparent. Thus, OATP4C1 can be regarded as a kidney-specific OATP. Functional tests revealed the transport of the endogenous compounds cAMP, triiodothyronine, thyroxine, but not of cGMP, taurocholate, prostaglandin E ₂ , estradiol-17β-D-glucuronide, and PAH. Among drugs, digoxin, ouabain, and methotrexate were taken up by OATP4C1. It was assumed that the physiological role of OATP4C1/Oatp4c1 is to present thyroid hormones to proximal tubule cells. The extent to which this transporter contributes to the proximal tubular secretion of drugs remains to be determined.

The Multidrug Resistance-Associated Protein 6 (MRP6/Mrp6; ABCC6/Abcc6)

MRP6/Mrp6 is a member of the ATP binding cassette (ABC) proteins that is mainly expressed in liver and kidneys. The murine homologue, mMrp6, showed highest expression in the liver, but was also found in other organs including the skin. Within human and mouse kidneys, MRP6/Mrp6 was localized to the basolateral membrane of proximal tubule cells (see ). This ATP-driven transporter handled glutathione-conjugates, the endothelin receptor antagonist BQ-123, and conferred low level resistance to epipodophyllotoxins and anthracyclines; probenecid and indomethacin inhibited MRP6. Mutations in the second nucleotide binding domain were associated with the connective tissue disease, Pseudoxanthoma elasticum , but the exact causal relationship between MRP6 and this disorder is unknown. Given the localization in the basolateral membrane of proximal tubule cells and the function of an ATP-driven efflux transporter it is unlikely that MRP6 contributes to organic anion secretion.

MRP1 (ABCC1) was found in the basolateral membrane of Henle’s loops and collecting ducts, and MRP3 (ABCC3) in distal tubules. In rats, Mrp3 was localized also to the basolateral membrane of proximal tubules. Also these transporters are not likely involved in organic anion secretion but may protect renal tubule cells from cytotoxic compounds.

The Organic Anion Transporter 4 (OAT4, SLC22A11 )

This transporter (reviewed in ) occurs only in humans. The gene for OAT4 is located on chromosome 11q13, in close neighbourhood to that of URAT1. OAT4 expression was highest in the kidneys, lower in placenta, and virtually absent from all other tested organs. Immunohistochemical studies revealed the presence of OAT4 in the proximal tubules only where it is located at the apical membrane. Thereby, OAT4 interacts with the scaffolding proteins PDZK1 and NHERF1 through its three C-terminal amino acids. Factors influencing the expression of OAT4 are presently unknown.

Mutational analysis revealed that evolutionary conserved glycine residues at positions 241 and 400 (transmembrane helices 5 and 8, respectively) are functionally important. Replacement of these residues by long aliphatic amino acids abolished transport activity, whereas the glycine-to-alanine mutation (G241A, G400A) reduced transport rates and decreased affinities were observed. Glycine is thought to participate in helix-helix interactions that seem to be important for transport activity of OAT4.

Eight non-synonymous single nucleotide polymorphisms were found, one of which (R48X) is deleterious (reviewed in ). The impact of the other amino acid changes on the function of OAT4 is unknown.

Following heterologous expression (for results see ), OAT4 transported estrone-3-sulfate with high affinity ( K _t of 1 µM). Other endogenous compounds translocated by OAT4 were urate, the prostaglandins E ₂ and F _2α , and dehydroepiandrosterone sulfate; no inhibition was found with glucuronidated steroids, suggesting that OAT4 only handles sulfated compounds. Cholate, taurocholate, octanoate, and corticosterone inhibited OAT4. Intracellular glutarate trans -stimulated estrone-3-sulfate uptake, but did not cis- inhibit it when present in the extracellular medium, suggesting different affinities for this dicarboxylate at the inner and outer binding sites. Similar findings for PAH indicate that the internal binding site has a higher affinity to this organic anion than the external binding site. Labeled PAH was not taken up by OAT4, but its efflux was accelerated by estrone-3-sulfate in the medium. It was concluded that OAT may work asymmetrically, taking up estrone-3-sulfate or urate in exchange for intracellular (α-keto)glutarate (“influx mode”). Intracellular organic anions such as PAH are effluxed through OAT4, probably in exchange for external chloride (“efflux mode”).

As reviewed in, OAT4 interacted with β-lactam antibiotics, angiotensin II receptor-1 blockers, methotrexate, diuretics, NSAIDs, and was inhibited by probenecid. The secretion of diuretics such as torasemide can lead to an increased reabsorption of urate due to diuretic/urate antiport at OAT4, explaining the observation that therapy with diuretics can cause hyperuricemia. NPT4 is yet another site for an interaction between diuretics and urate (see later).

Taking together, OAT4 in the apical membrane of proximal tubule cells can contribute to the absorption of estrone sulfate and urate from the primary urine, and to the secretion of anionic drugs. Since OAT4 is not present in rodents it is not possible to learn more about its function from knockout animals.

The Organic Anion Transporter 5 (Oat5, Slc22a19 )

Oat5 was cloned from rat and mouse in which it was restricted to the kidney and was equally expressed in male and female animals. There is no human homologue to Slc22a19. The Oat5 protein was localized to the apical membrane in the late proximal tubule segments. The interaction of Oat5 with dicarboxylates is not yet settled, because conflicting results were reported on the ability of succinate to trans -stimulate uptake of estrone-3-sulfate. Endogenous substrates were dehydroepiandrosterone sulfate and estradiol sulfate, but not prostaglandins and urate. A few drugs were tested: furosemide, benzylpenicillin, diclofenac, and ibuprofen inhibited Oat5. As long as the driving force for Oat5 is not defined it is not possible to speculate on the role of this transporter in drug secretion or absorption.

The Organic Anion Transporter 10 (OAT10/Oat10; SLC22A13/Slc22a13 )

OAT10 (previously cloned as orphan transporter ORCTL3) is predominantly expressed in human kidneys and weakly in brain, heart and colon. The gene is located on chromosome 3p22.2. Western blots showed the rOat10 protein in the apical, but not in the basolateral membrane, and female animals showed a stronger expression of rOat10 than did males. Endogenous substrates of OAT10 were nicotinate, lactate, urate, succinate, and glutathione. Succinate served as counter anion for nicotinate and urate uptake from the lumen. Among the few drugs tested, furosemide, hydrochlorothiazide, sulfinpyrazone, and cyclosporine A inhibited OAT10. It needs to be established whether OAT10 is involved in cyclosporine A-induced nephropathy. The physiological role of OAT10 is probably the uptake of the vitamin nicotinate in small intestine and renal proximal tubule cells.

The Urate Anion Transporter 1 (URAT1/Urat1/Rst; SLC22A12/Slc22a12 )

URAT1/Urat1/Rst was cloned from human and mouse kidneys (for reviews see ). The gene for human URAT1 is located on chromosome 11q13.1 adjacent to that of OAT4. The highest though not exclusive expression was found in kidneys. Human and mouse URAT1/Urat1 were immunolocalized to the apical membrane of proximal tubule cells where they interact with the scaffolding protein PDZK1. Male mice showed a higher expression than female animals, indicating sex differences at least in this species.

Hepatocyte nuclear factor-1α and -1β induced the expression of human and mouse URAT1/Urat1 promoter, and mUrat1 expression was diminished in a HNF-1α knockout mouse. Furthermore, the mUrat1 promoter was found to be hypomethylated, suggesting a tissue-specific epigenetic control (reviewed in ).

URAT1 mutations have been linked to familial hypouricemia. Most of the hypouricemic patients in Japan and Korea carry a truncation mutation, W258X. Meanwhile, several more amino acid exchanges have been described (see for a compilation). Some mutations occur also with hyperuricemia and gout, but the molecular mechanism is unknown.

URAT1 expressed in oocytes displayed saturable uptake of urate. Extracellular chloride inhibited, and intracellular chloride trans -stimulated urate uptake, indicating that URAT1 can operate as urate/chloride exchanger. Preloading of oocytes with lactate trans -stimulated urate uptake, indicating urate/lactate exchange as another transport mode of URAT1. The endogenous compounds L- and D-lactate, acetoacetate, β-hdyroxybutyrate, succinate (but not α-ketoglutarate), and orotate inhibited urate uptake. Exogenous compounds inhibiting URAT1 were the uricosuric drugs probenecid, benzbromarone, losartan; the antiuricosuric drugs pyrazine carboxylate and pyrazine dicarboxylate; the loop diuretics furosemide and bumetanide; the NSAIDs phenylbutazone, salicylate and indomethacin; and the vitamin nicotinate. PAH and estrone sulfate did not affect URAT1.

Also mouse Urat1/Rst transported urate ( K _m 1.2 mM) and performed urate/chloride, urate/lactate and urate/pyrazinoate exchange. Uptake of labeled urate was inhibited, when the uricosurics probenecid and benzbromarone, the uremic toxins indoxyl sulfate, indole acetate, hippurate, and a number of acidic neurotransmitter metabolites were present in the medium.

In summary, URAT1/Urat1 is involved in urate absorption from the primary urine. Thereby, urate uptake is driven by exchange against lactate that is transported into the cells by a sodium-lactate symporter in the apical membrane, as has been convincingly demonstrated in a c/ebpδ knockout mouse. This mouse does not express the sodium-coupled lactate transporters Slc5a8 and Slc5a12 in the apical membrane of proximal tubules and excretes lactate and—at unchanged Urat1 expression—urate with the urine. In man, uricosurics inhibit URAT1, and loss-of-function mutations cause an increased urate clearance and a reduction in serum level of urate.

The Voltage-Sensitive Organic Anion Transporter 1 (Oat _v 1/Npt1; Slc17a1 )

Searching for the voltage-driven transporter for PAH and urate in urate-secreting species (rabbits, pigs), Oat _v 1 was identified by expression cloning from pig kidneys. The message for pOat _v 1 was restricted to liver and kidneys. pOat _v 1 is not a member of the SLC22 family, but showed highest identity to human and rodent NPT1/Npt1 ( SLC17A1/Slc17a1 ), previously cloned as sodium-dependent phosphate transporters from human, rabbit, and mouse kidneys and located to the apical membrane of the proximal tubule segments S1–S3 (for review see ). The human gene for NPT1 was found on chromosome 6p21.3-p23. The expression of mNpt1 was under control of heptatocyte nuclear factor-1α, and HNF-1α knockout mice had greatly diminished proximal tubular Npt1 expression. Fasting decreased and streptozotocin-induced diabetes increased renal and hepatic expression of NaPi-1/Npt1 in rats, respectively. Whether gender differences exist is not known.

Human NPT and mouse Npt1/NaPi-1 performed the uptake of radiolabeled organic anions including PAH, urate, estradiol-17β-D-glucuronide, benzylpenicillin, faropenem, and mevalonate. Benzylpenicillin, ampicillin, cephalexin, cefazolin, indomethacin, furosemide, benzoate, lactate, and probenecid inhibited uptake of labeled faropenem by mNpt1. Uptake of labeled PAH by porcine Oat _v 1 was inhibited by, e.g., estrone sulfate, penicillin G, salicylate, indomethacin, ibuprofen, diclofenac, furosemide, bumetanide, and probenecid. Organic anion transport was independent of sodium, but was inhibited by increasing chloride concentrations as well as by DIDS and NPPB.

It is likely that organic anion transport through NPT1/Npt1 is electrogenic. Busch et al. observed outward currents upon addition of benzylpenicillin, phenol red and probenecid in Xenopus laevis oocytes expressing rbNaPi-1 (Npt1 ). PAH uptake ( K _m 4.4 mM) by pOat _v 1/pNpt1 was increased at inside positive membrane potentials. More recently, mNpt1 was purified and inserted into proteoliposomes for functional characterization. Both, sodium-dependent phosphate transport and sodium-independent PAH and urate transport were observed, indicating that at least mNpt1 is able to carry out both functions. The affinity of mNpt1 for urate and PAH was similar ( K _m 1.1 mM) and uptake was accelerated by a proteoliposome-inside positive membrane potential. Chloride stimulated (whereas it inhibited in other studies) PAH uptake, but was not transported itself. Salicylate, aspirin and acetaminophen, but not other NSAIDs, inhibited PAH and urate transport; radiolabeled aspirin was translocated by reconstituted mNpt1.

The mutation of an arginine residue (R138A) conserved within the SLC17 family abolished transport. A lowered V _max was found with the T269I mutation that corresponds to a SNP in human NPT1 gene associated with gout. Taken these data together it is likely that NPT1/Npt1/Oat _v 1 is involved in the voltage-driven exit of organic anions including urate, PAH, β-lactam antibiotics, diuretics and some non-steroidal anti-inflammatory drugs.