2022 年 45 巻 4 号 p. 382-393
The organic cation transporter 2 (OCT2) belongs to the SLC22 family, while the multidrug and toxin extrusion 1 and 2-K (MATE1/MATE2-K) belong to the SLC47 family, are localized to the basolateral and apical membrane of human renal proximal tubular epithelial cells, respectively. They are polyspecific transporters that enable the transit of structurally diversified drugs with overlapping selectivity across plasma membranes. OCT2 and MATE1/2-K are critically involved in renal secretion, pharmacokinetics (PK), and toxicity of cationic drugs. Drug–drug interactions (DDIs) at OCT2 and/or MATE1/2-K have been shown to result in clinical impacts on PK, therapeutic efficacy and are probably involved in the renal accumulation of drugs. Sites of OCT2 and MATE1/2-K expression and function play an essential role in the pharmacokinetics and toxicity of drugs, such as cisplatin. Thus, knowing the sites (basolateral vs. apical) of the interaction of two drugs at transporters is essential to understanding whether this interaction helps prevent or enhance drug-induced nephrotoxicity. In this work, an overview of OCT2 and MATE1/2-K is presented. Primary structure, membrane location, functional properties, and clinical impact of OCT2 and MATE1/2-K are presented. In addition, clinical aspects of DDIs in OCT2 and MATE1/2-K and their involvement in drug nephrotoxicity are compiled.
The kidney is a critical organ for drug excretion and contributes to the control of systemic sodium and water homeostasis. Renal elimination of drugs occurs in the nephron through three concurrent processes: passive glomerular filtration, active tubular secretion, and tubular reabsorption. Multiple renal transporters located on the basolateral and apical membranes of proximal tubular cells are involved in active tubular secretion. Renal drug transporters are critically engaged in translocating drugs across the plasm membrane and are required to uptake and efflux drug molecules in the kidney. It's estimated that about 40% of all prescribed medications are organic cation,1–3) and most of these drugs are hydrophilic. Therefore, transporters are required to enable the passage of drugs across the plasma membrane. Many of these are inhibitors and/or substrates of organic cation transporters (OCTs). Interaction of drugs with each particular OCTs, in addition to the site of OCTs, is essential to predict the role of cation transporters on therapeutic efficacies, pharmacokinetics (PK), and potential drug toxicities. Moreover, it is also required to know which individual OCTs transport drugs or whether they are non-transported inhibitors. In the last decade, the issue of drug–drug interactions (DDIs) has received a lot of attention. The reasons are that patients receiving multiple drugs are becoming increasingly common, often exceeding more than five drugs, especially among the elderly, and that increasing knowledge of transporter-mediated DDI’s is an essential modifier of the PK and pharmacodynamics (PD) of drugs,4,5) and drug targets provide new options for preclinical in vitro testing for potential DDIs. In humans, OCT2 and the multidrug and toxin extrusion proteins 1 and 2-K (MATE1/2-K) are major transporters for the secretion of cationic drugs. Drugs that inhibit OCT2 and MATE1/2-K may alter the PK of the victim drug and potentially change its effects/toxicity.5–7) In addition, inhibition of OCT2 or/and MATE1/2-K can cause altered drug accumulation in renal tubular cells and possibly altered renal toxicities and drug efficacy.6,8) This review focuses on renal OCTs, particularly OCT2 and MATE1/2-K, and their significance in DDIs and the critical role of their inhibitors in preventing or enhancing drug-induced nephrotoxicity. We first briefly summarize the primary structures, membrane locations, principal tissue distributions, functional properties, biomedical functions, and regulations of organic cation transporters. Thereafter, clinical aspects of renal OCT2 and MATE1/2-K-mediated DDIs may highlight these transporters' critical role in clinically DDIs that result in alterations of PK and/or therapeutic efficacy of victim drug as such metformin. In addition, the clinically relevant significances on drug-induced nephrotoxicity are demonstrated using some examples in clinical DDI and some biomarkers that can directly detect proximal tubular injury. Lastly, a concise summary of the field is offered, as well as current challenges.
More than 400 membrane transporter are divided into two distinct classes: solute carrier (SLC) and ATP-binding cassette (ABC) families. ABC transporters are efflux transporters that export drugs out of cells by utilizing ATP as driving energy. Typically, there are over 300 SLC transporters, which are related to drug transport, and primarily include OCTs, MATEs, organic cation and carnitine transporters (OCTNs), organic anion transporters (OATs), organic anion-transporting polypeptides (OATPs), peptide transporters (PEPTs).5,9,10) In renal proximal tubular epithelial cells, the secretion and reabsorption of organic cations are mediated by a diverse array of transporters are shown in Fig. 1. OCT2, which is mainly expressed on the basolateral side of renal proximal tubular epithelial cells, is critically involved in the renal influx of cationic compounds.5,7) In turn, MATE1/2-K are critically involved in the efflux of those cationic compounds into the urine in the apical side of renal proximal tubular epithelial cells that cooperate with OCT2 in the tubular secretion of various cation drugs.5,7,11,12) OCT3 has complemented plays a role in the uptake of organic cations in the basolateral membrane,2,13–16) whereas OCTN1 and OCTN2 have supplemented to release cations compounds across the apical membrane into the tubular lumen. In addition, OCTN1 and OCTN2 in the tubular epithelium luminal membranes are efficient transporters for zwitterions.17,18) OCT1, OCTN1, or OCTN2 may mediate reabsorption of cation compounds from the tubular lumen, which are located in the apical side,18–20) in participation with organic cation release across the basolateral side via cation exchange by OCT2 and/or OCT3.2)
Translocation of organic cations by OCT2 and MATE1/2-K in renal proximal tubule cells showed by red arrows. Broad blue arrows depict the obligatory exchange of protons. Thin blue arrows indicated the translocation of organic cations, Na+ or zwitterions that can occur to transport organic cations. (ZI) zwitterion (OC+) organic cation.
OCTs belong to the SLC22 family.16) The SLC22 family includes three subtypes: OCTs with three isoforms: OCT1, OCT2, and OCT3; electroneutral OCTs (OCTN1, OCTTN2, and OCTN3); and the carnitine/cation transporter OCT6.9,15,16) OCT2 is exhibiting 70% amino acid identity to OCT1, while OCT3 is roughly 50 and 43% are identical to OCT1 and OCT2, respectively.21) OCTN1 and OCTN2 have 77% identity, and only 31–37 of the amino acids of OCTN1 and OCTN2 are identity with OCT1-3.16,20,22) The OCT1-3 and OCTN1/2 contain between 543 and 557 amino acids.21,23) They are predicted to have 12 transmembrane domains (TMDs) with intracellular NH2 and COOH termini (Fig. 2A). A large extracellular loop between TMD1 and TMD2 and a large intracellular loop between TMD6 and TMD7. A large extracellular loop and a large intracellular loop are contained potential N-glycosylation sites and multiple putative phosphorylation sites, respectively.10,16,24)
MATEs belong to the SLC47 family. In 2005, the first clones of MATE1 was identified. Soon subsequently, MATE2, MATE2-K, and MATE2-B were recognized. MATE1 and MATE2-K are functional proteins containing 570 and 566 amino acids, respectively, and are functionally active. They have 13 predicted TMDs with their short C-terminus on the extracellular and a large intracellular loop between TDM12 and TDM13 (Fig. 2B), while MATE2-B includes 209 amino acids with no mediate transport was observed.2,12,25,26)
3.2. Membrane LocationPrevious reviews have highlighted the patterns of OCTs expression and the location of the plasma membrane.2,6,18,27) In brief, OCT2 has a less widespread distribution. OCT2 is abundantly expressed in the kidney and is localized to the basolateral side of the human renal proximal tubular, but no expression in the liver, heart, small intestine, and colon was detected2,7) (Fig. 1). In addition, OCT2 is expressed in cerebral neurons and microvessels of the brain.27) OCT1 is one of the abundantly expressed in the sinusoidal membrane of hepatocytes. Expression of OCT1 was also detected in apical membranes of enterocytes of the small intestine and renal human proximal tubules, but Oct1 was localized on the basolateral side in the rat kidney.7,28) In addition, OCT1 is expressed in ciliated cells of airway epithelia.2) OCT3 has been widely expressed in many organs and tissues, including the heart, kidney, liver, adipose tissue, skeletal muscle, placenta, and salivary glands.16,29) OCT3 was localized to the basolateral membrane of renal proximal tubular and to the sinusoidal membrane liver.10,15,16) OCTN1 is expressed in many tissues, including the kidney, small intestine, muscle cells, lymphocytes, and monocytes.2,9) OCTN2 is expressed ubiquitously, including expression in the liver, kidney, small intestine, lymphocytes, monocytes, heart, and brain vessels.2,6,9) OCTN1 and OCTN2 proteins are both localized to the apical membrane of the small intestine and proximal renal tubules. MATE1 is highly expressed in numerous organs and tissues, including the kidney, liver, adrenal gland, skeletal muscle.6,12) Noteworthily, MATE1 was observed in brain vessels.30) MATE1 was localized to the apical side of proximal renal tubules and to the canalicular membrane of hepatocytes.7,12) In individual tubular cells, coexpression of MATE1 in the apical side with OCT2 and OAT in the basolateral side of proximal tubules has been revealed.7) MATE2-K has a more restricted expression pattern, mainly expressed in the human kidney, where it was localized to the apical side of the human proximal renal tubules.6,11,31) In addition, MATE2-K was also identified in brain microvessels.2,30) In individual tubular cells, coexpression of MATE2-K with OCT2 was revealed.7)
3.3. Functional Properties and Biomedical Functions of Renal Organic Cation TransportersOCT1, OCT2, and OCT3 have similar transport properties. First, OCTs are capable of translocating several cations with a wide range of molecular structures9); hence, they are referred to as polyspecific transporters. Most compounds transported by OCT1, OCT2, and OCT3 have a molecular mass below 500 and their smallest diameter below 4 Å.15,16) They are exhibiting broadly overlapping substrate specificities. Second, OCTs transport organic cations electrically.16,32) Third, OCTs operate in sodium- and proton-independent manners; however, the affinity of specific substrates is inversely related to their degree of ionization. Fourth, OCTs transport cations across the plasma membrane in both directions.2,6,9) It was found that OCTs transport most organic cations that are positively charged at physiological pH (Fig. 1). In addition, some evidence was presented that OCTs may be transported noncharged (such as cimetidine) or anions compounds (such as probenecid).15) OCTN1 is polyspecific transporters that accept zwitterions such as L-ergothioneine and L-carnitine and organic cations like tetraethylammonium (TEA) and quinidine and noncharged compounds in partially different ways.2,6,18) Evidence has been provided that OCTN1 can operate in both directions to different substrates.9) Of note, OCTN1 exhibits a high affinity for the zwitterionic antioxidant ergothioneine, whose uptake is stimulated by extracellular sodium, whereas uptake of TEA by OCTN1 is stimulated by intracellular protons and ATP.17,18) OCTN2 is a polyspecific transporter that accepts organic cations, noncharged, and zwitterions compounds in partially diverse manners.2,18) Its exhibits a great affinity for L-carnitine and functions as a sodium/L-carnitine cotransporter and OCTN2-mediated uptake of TEA independently of sodium2,8) (Fig. 1).
MATE1 is polyspecific transporter of organic cations and noncharged and zwitterions compounds, showing overlapping specificity of cationic substrates with OCT2 or other OCTs. In addition, MATE1 can transport some anionic compounds (e.g., anionic estrone sulfate), which implies that they may also collaborate with OATs to facilitate the renal elimination of anionic compounds.31) MATE1 are electroneutral proton-cation antiporters bi-directional across the plasma membrane.6,12,33) Like MATE1, MATE2-K is a polyspecific transporter of various organic cations and noncharged and zwitterions compounds, exhibiting extensively overlapping substrate specificities with MATE1.2,34) Similar to MATE1, an opposed proton gradient stimulates MATE-K2— mediated organic cation transport.11) Evidence has been provided that the affinity of MATE2-K for protons is higher than that of MATE1.2,34) OCTs are critically involved in intestinal absorption, tissue and cellular distribution and metabolism, and renal excretion of organic cation drugs. In addition, OCTs in the brain play an important role in the uptake of cation drugs and removing monoamine neurotransmitters from brain interstitium into neurons and glial cells.27,35) In the present review, we will focus on OCT2 and MATE1/2-K that are involved in the renal secretion of drugs and on DDIs that inhibit these functions. In proximal renal tubules, numerous transporters are mainly involved in the secretion and reabsorption of organic cations. OCT2 is critically involved in the renal influx of many cationic compounds. OCT2 collaborates with MATE1/2-K on the apical side that mediates the efflux of many cations into the urine (Fig. 1). The MATE1/2-K cation-proton exchangers ought to be most effective. Besides, OCT3 in the basolateral membrane side could contribute to the uptake of organic cations,2,13–15) which plays a role in the distribution and secretion of metformin due to its broad expression in peripheral tissues.36) Of note, in vitro assessment of apparent Michaelis Menten constant (Km) and IC50 values may help predict the individual cationic transporters' transport efficiency and inhibitory potential. In vivo clinically relevant drug concentrations, total plasma concentrations, and drug binding to plasma proteins must be determined to anticipate the clinical relevance of individual cationic transporters. The testing approach proposed includes three consecutive stages. First, it is proposed to assess new molecular entities (NMEs) for interaction with OCT2 and MATE1/2-K via estimating whether they inhibit the modal cation uptake as such metformin, 1-methyl-4-phenylpyridimium (MPP+), TEA, lamivudine, and cisplatin.2,6,37) Second, it is suggested to evaluate whether inhibitory NMEs are also transported. Third, information is obtained from the clinical trial's first phase about clinically significant NME concentrations.6)
OCT2 and MATE1/2-K are critically involved in the effects of the PK and/or PD of drugs. Unfortunately, there are still few clinical studies that exhibit the clinical impact of individual OCT2 and MATE1/2-K for PK and/or PD of specific drugs. This assignment is challenging because organic transport can be expressed in various sites that might be involved in uptake or efflux into cells. Furthermore, other transporters mediate are often found in the same membrane location with overlapping cation selectivity or in opposite membrane sides of cells that contribute to absorption or excretion. Knowledge detailed about the function of OCT2 and/or MATE1/2-K in different organs and information of how substrates are transported and about Km and IC50 values are required to select drugs for clinical testing of the DDIs. Although many drugs transported by OCT2 and MATE1/2-K and inhibit these transporters by many drugs have been studied extensively,2) the potential two drugs interaction in vivo at these transporters remains uncertain. Inhibition of tubular secretion is the most common form of interaction in the kidney. Inhibition of renal transporters can cause a decrease in renal elimination clearance, resulting in increased systemic exposure of the drug and altered effects/toxicity. Additionally, renal DDIs may alter drug accumulation intracellularly, leading to drug-induced nephrotoxicity and renal injury.5) The clinical DDI examples afterward show such instances. Consequently, knowing the specific interaction site at the basolateral or apical membrane is crucial for predicting whether an inhibitor causes nephrotoxicity or is nephroprotective in vivo. In the next section, we highlight the roles of renal OCT2 and MATE1/2-K in clinically relevant DDIs. Examples of renal OCT2 and MATE1-2/K-mediated DDIs are highlighted in Table 1.
Perpetrator drug | Victim drug | Renal transporter | Effect on victim drug PK | Effect or toxicity for victim drugs | References |
---|---|---|---|---|---|
Cimetidine (400 mg BID) | Metformin | MATE1/2-K > OCT2 | AUC ↑ (1.5-fold) | ↑ adverse effects | 41) |
ClR ↓ (− 45%) | |||||
Pyrimethamine (50 mg QD) | Metformin | MATE1, MATE2-K | AUC ↑ (39%) | ↑ adverse effects | 45,46) |
Cmax ↑ (42%) | |||||
ClR ↓ (− 35%) | |||||
Vandetanib (300 mg QD) | Metformin | MATE1, MATE2K | Cmax ↑(50%) | ↑adverse effects | 47) |
ClR ↓ (− 50%) | |||||
Trimethoprim (200 mg BID) | Metformin | MATEs >> OCT2 | Cmax ↑(22–38%) | ↑adverse effects (↑ plasma lactate) | 50,51) |
AUC ↑ (30–37%) | |||||
ClR ↓ (− 27–33%) | |||||
Ranolazine (1000 mg) Ranolazine (500 mg) | Metformin | Cmax ↑ (1.53-fold) | 52) | ||
AUC ↑ (1.79-fold) | |||||
Cmax ↑ (1.22-fold) | |||||
AUC↑ (1.37-fold) | |||||
Tucatinib (300 mg BID) | Metformin | MATE1 and/or MATE2-K | Cmax ↑(15%) | 6,54) | |
ClR ↓ (− 40%) | |||||
Famotidine (200 mg QD) | Metformin | MATE1 >MATE2K >> OCT2 | CLR ↑(28%) | ↓ 9% of glucose AUC | 55) |
AUC, Cmax no change | |||||
Dolutegravir (50 mg BID) | Metformin | OCT2 selective | Cmax ↑ (2.1 fold) | ↑ adverse effects | 56) |
AUC ↑ (2.5-fold) | |||||
Bictegravir (50 mg QD) | Metformin | OCT2 selective | AUC↑(39%), CLR↓(− 31%) | No change | 59,60) |
Trimethoprim/Sulfamethoxazole (160/180 mg QD) | Lamivudine | OCT2, MATE1, MATE2-K | AUC ↑ (1.4-fold) | ↑ adverse effects | 61,63) |
ClR ↓ (− 35%) |
OCT2 and MATE1/2-K are major transporters in renal tubular cells, co-mediate the tubular excretion of small hydrophilic drugs. Many cationic drugs interactions involve OCT2 or MATE1/2-K inhibition.4,5,8) Currently, the International Transporter Consortium (ITC) and U.S. Food and Drug Administration (FDA) recommend metformin as an in vivo probe to evaluate the inhibitory potential of NMEs towards OCT2 and MATE1/2-K.37,38) Metformin (a biguanide derivative) is the first choice for treating type 2 diabetes. It is a hydrophilic base that is positively charged in the blood, so metformin can hardly diffuse through plasma membranes and needs to be transported to permeate the plasma membrane.2,39) Metformin pharmacokinetics is mainly controlled by drug transporters, including OCT1, OCT2, OCT3, OCTN1, MATE1, MATE2-K, and PMAT.2,40) Metformin is not metabolized, but majority of it is excreted in urine with the prototype drug.39)
The transporter combination OCT2 (uptake) and MATE1/2-K (efflux) mediates renal secretion of metformin, and the critical role of transporters for PK and/or PD of metformin have been reinforced by numerous clinical studies, some of which will be discussed in the following section. Figure 3 illustrates the clinical effects that are anticipated when metformin uptake and efflux in proximal renal tubular. Increase the systemic exposure and reduce the renal clearance of metformin when inhibition of the renal metformin pathway occurs. Besides DDIs, administration of OCT2 inhibitors may be helpful for the reduction of drug-induced kidney injury, such as cisplatin-induced nephrotoxicity.
Cimetidine, an antihistamine, is a clinically relevant inhibitor of renal organic cationic secretion. About 20% of plasma cimetidine is protein-bound with maximum plasma concentration (Cmax) unbound in plasma of about 8 µM at 400 mg oral dose.8) Co-administration of cimetidine with metformin resulted in 1.5 fold increased area under the plasma concentration curve (AUC), and a 45% decrease in renal clearance of metformin was observed33,41) (Table 1). This could be due to the down-regulate functions of the renal OCT2 and MATE1/2-K that are involved in the renal elimination of metformin. However, it was later observed that cimetidine could inhibit metformin transport mediated by MATE1/2-K rather than by OCT2 at clinically relevant cimetidine concentrations in systemic blood.2,42–44) Considering the estimated IC50 values for cimetidine in inhibiting metformin uptake (10 µM) by MATE1/2-K are lower than the Cmax unbound of cimetidine in the blood, and identified the IC50 values for OCT2 are at least two times as high2,42,43) (Table 2). In addition, the resulting Ki of cimetidine in uptake metformin mediated by OCT2 was 124 µM, whereas cimetidine was a more potent MATE1/2-K inhibitor with Ki values of 3.8 and 6.9 µM, respectively.42) These data also indicate that reduced renal clearance of metformin is possibly related to inhibition of MATE transporters.
IC50 (Ki) values for model substrate uptake inhibition, metformin (met), crizotinib (cri) or lamivudine (lam) [µM] | ||||||
---|---|---|---|---|---|---|
Perpetrator drug | Cmax un, (µM) | Victim drug | OCT2 | MATE1 | MATE2-K | References |
Cimetidine | 8 | Metformin | 15–146, 124 met | 1.1–170 | 2.1–370, 5.5 met,10, 6.9 met | 2,42,43) |
2.6 met,10, 3.8 met | ||||||
Pyrimethamine | 0.8 | Metformin | 1.6–23 | 0.04–1.2, 0.3 met,10 | 0.01–0.83, 0.2 met,10 | 2,43) |
Vandetanib | 0.4 | Metformin | 73, 8.8 met,1 | 1.2, 0.16 met,1 | 1.3, 0.30 met,1 | 6,48,49) |
Trimethoprim | 17 | Metformin Lamuvidine | 14–1318, 27 met, 13 lam,1 | 2.7– > 100, 4.1 met, 10, 6.3 met | 0.42–1.9, 0.42 met,10, 29 met, 0.66 lam,1 | 2,6,43,51,61) |
Ranolazine | 2 | Metformin | 19 met,2 | 6,52,53) | ||
Tucatinib | 1.4 | Metformin | 0.11, 15 met,10 | 0.09 0.34 met,10 | 0.14 met,10 | 54) |
Famotidine | 1 | Metformin | 36, 114, 66 met,10 | 0.3–6.7 | 3.1–36 3.1 met,10 2.5 met,10 | 2,43,55) |
0.91 met,10, 0.25 met,10 | ||||||
Dolutegravir | 0.14 | Metformin | 0.07, 1.9 | 4.7, 6.3 met, 10 | >100, 25 met,10 | 2,56,58) |
Entecavir | 0.013 | Crizotinib | 0.037 cri,1 | 93) | ||
Metformin | 16 | Trospium | 521–2370 | 47–667 | 89–6516 | 2,6) |
Verapamil | 0.06 | Metformin | 0.6–92, 10 met,16 | 28, 42 | 32, 38 | 2,3,6,106) |
IC50 values with superscripts indicated victim drugs and their micromolar concentrations used as substrates, IC50 values without superscripts indicated cationic model substrates that different concentrations. Cmax un estimated maximum plasma/serum concentration of unbound drug based on published data. Met, metformin; cri, crizotinib; lam, lamivudine IC50 and ki values for MATE1 and MATE2-K were determined for cis inhibition of cellular drug uptake.6)
Pyrimethamine is an antimalarial drug. Earlier studies reported that upon co-administration of pyrimethamine with metformin, the AUC and Cmax increased by 39 and 42%, respectively, and the renal clearance of metformin decreased by 35%45,46) (Table 1). In addition, this was accompanied by increased serum creatinine (SCr) of 21% and a decreased creatinine renal clearance (CLcr) of 45%45) (Table 3). Pyrimethamine is a high-affinity inhibitor of MATE1/2-K, due to that IC50 values obtained for MATE1/2-K inhibition by pyrimethamine were more than 100-fold lower than the IC50 values determined for OCT2 inhibition, and unbound pyrimethamine in systemic blood is more than 10-fold lower than the IC50 value for OCT2 inhibition2) (Table 2). These data indicate that the underlying mechanism of pyrimethamine and metformin interactions is primarily attributed to inhibitions of apical MATE1/2-K similar to cimetidine.
Perpetrator drugs (Dosing regimen) | % Increase of SCr | % Decrease of CLcr | References |
---|---|---|---|
Pyrimethamine (50 mg QD) | 21 | 45 | 45) |
Trimethoprim (200 mg BID) | 23 | 25 | 50) |
Vandetanib (300 mg QD) | 15 | NA | 48) |
Famotidine (200 mg QD) | 6 | 12 | 55) |
Dolutegravir (50 mg QD or BID) | 10–15 | 10 or 14 | 56,57) |
Bictegravir (50 mg QD) | NA | 39 | 59) |
NA, not available.
Vandetanib is an oral anti-cancer drug used to treat myeloid thyroid cancer. It is a potent inhibitor of MATE1/2-K (versus OCT2). Upon co-administration of metformin with vandetanib, the renal excretion of metformin is decreased by 50%, and the Cmax of metformin is increased by 50%47) (Table 1). The IC50 values for inhibition of OCT2, MATE1, and MATE2-K-mediated uptake of 1 µM metformin were 8.8 , 0.16 , and 0.3 µM, respectively,48) and estimated Cmax unbound of vandetanib is 0.4 µM in systemic blood49) (Table 2). These data indicate that vandetanib is a potent inhibitor of MATE1 and MATE2-K.
4.2.4. Trimethoprim and Metformin InteractionTrimethoprim is an antibiotic used for treating urinary tract infections. It reduces renal elimination of metformin by MATEs. During co-administration of trimethoprim and metformin, renal metformin clearance was reduced by 27–33% and an increase in AUC and Cmax of metformin in blood by 30–37 and 22–38%, respectively, accompanied by an increase in plasma lactate levels and decrease of CLcr50,51) (Tables 1, 3). Trimethoprim’s Ki values for inhibited metformin uptake by OCT2 and MATE1 were more than 50% higher than the estimated Cmax of unbound trimethoprim in systemic blood, whereas MATE2-K’s Ki value was 63% lower (Table 2), indicating an impact on MATE2-K.6,51) Therefore, the changes in renal secretion of metformin after co-administration of trimethoprim are probably involved in the inhibition of apical MATEs rather than basolateral OCT2.
4.2.5. Ranolazine and Metformin InteractionRanolazine is an antianginal drug used for the treatment of chronic angina. Co-administration of ranolazine (1000 mg) and metformin resulted in a 1.53 fold (50%) increase in Cmax of metformin, and when co-administration of ranolazine (500 mg) and metformin resulted in a 1.22 fold (23%) increase in Cmax of metformin, implicating a reduced renal metformin elimination52) (Table 1). The IC50 values for inhibition of OCT2 mediated uptake of 2 µM metformin was 19 µM, and it was estimated Cmax unbound of ranolazine is 2 µM in systemic blood52,53) (Table 2). Although it has not been demonstrated, the inhibition of MATE1/2-K may probably have contributed to the decrease in renal secretion.
4.2.6. Tucatinib and Metformin InteractionTucatinib is an antineoplastic drug used with trastuzumab and capecitabine to treat HER2-positive breast cancer. Upon co-administration of tucatinib with metformin, the renal clearance of metformin was decreased by about 40%, and the maximum metformin concentration was increased slightly (15%)54) (Table 1). In vitro, IC50 values for tucatinib inhibition of OCT2, MATE1, and MATE2-K metformin (10 µM) transport were respectively 14.7, 0.340, and 0.135 µM and 1.4 µM was estimated for Cmax unbound tucatinib in systemic blood54) (Table 2). Thus, the effect of tucatinib on decreased renal metformin secretion is probably due to the inhibition of metformin efflux mediated by renal MATEs in the apical membrane side.6)
4.2.7. Famotidine and Metformin InteractionFamotidine is an H2-receptor antagonist used to treat peptic ulcer and gastroesophageal reflux diseases. It is a high-affinity inhibitor of MATE1 but also inhibits MATE2 and OCT2. Unlike MATEs inhibitor that was discussed above. After co-administration of famotidine with metformin, renal clearance of metformin was increased with no change in plasma exposure of metformin55) (Table 1). Unexpectedly, famotidine improved the bioavailability of metformin and enhanced the glucose-lowering effects of metformin. Since the IC50 value determined for inhibition of MATE1-mediated metformin uptake by famotidine was 4-fold lower than the maximum concentration of famotidine in blood. In contrast, the IC50 values for inhibition of OCT1, OCT2, or MATE-2 K mediated metformin uptake by famotidine were at least 2.5 higher2,6,43,55) (Table 2). From this available data, it was speculated that MATE1 inhibition is critically involved. Despite the careful study, the underlying mechanisms behind the observed effects remain obscure, and intensive future research is required to clarify them.
4.2.8. Dolutegravir and Metformin InteractionDolutegravir is an antiretroviral agent for the treatment of human immunodeficiency virus (HIV)-1 infection. It is an inhibitor of OCT2 and MATE1/2-K. The co-administration of dolutegravir (50 mg q12h) and metformin increased the AUC and Cmax of metformin by 2.5-fold (145%) and 2.1-fold (110%), respectively, which is implicating a reduction of renal metformin secretion56) (Table 1), and this was accompanied by an increase in SCr but no change in the level of lactic acidosis56,57) (Table 3). In vitro, dolutegravir is a more potent inhibitor for OCT2 (IC50 = 1.9 µM) than for MATE1/2-K (IC50 = 6.3–25 µM). Dolutegravir probably increased plasma exposure to metformin by inhibiting OCT2 because the lowest IC50 value for OCT2 is two times lower than the unbound Cmax of dolutegravir in blood. In contrast, IC50 values for MATE1 and MATE2-K for uptake inhibition of metformin (10 µM) are more than 40 times higher than the unbound Cmax of dolutegravir6,56,58) (Table 2). Thus, the mechanism of dolutegravir and metformin interaction was mainly due to OCT2 inhibition, and it is suggested that other unidentified process(es).
4.2.9. Bictegravir and Metformin InteractionAmong selective OCT2 inhibitors is bictegravir, which is a potent anti-HIV drug. Upon co-administration of metformin with bictegravir, the plasma exposure of metformin increased by approx. 39% and reduced the renal excretion of metformin by approx. 31%. However, these changes in metformin exposure are not clinically relevant59) (Table 1). In vitro, the findings imply that bictegravir has a more potent inhibitor of OCT2 (IC50 = 0.42 µM) than MATE1 (IC50 = 8 µM) inhibition, similar to dolutegravir60) (Table 2).
4.2.10. Lamivudine and Trimethoprim InteractionLamivudine is an antiviral agent used to treat HIV infection and chronic hepatitis B (HBV). Its transported by OCT1-3 and MATE1/2-K.2,61) OCT1, OCT2, and OCT3 transport lamivudine with Km values (249–2140 µM).6,62) Lamivudine is absorbed rapidly with a bioavailability of over 80%, and the primary pathway of lamivudine excretion is via renal tubular secretion.63) During co-treatment of trimethoprim (combined with sulfisoxazole that no OCTs interact) with lamivudine, a 43% increase in AUC and 35% decrease in renal clearance of lamivudine was found when compared with lamivudine alone63) (Table 1). The findings imply that trimethoprim increases systemic exposure to lamivudine by reducing renal excretion. Considering the estimated Cmax unbound trimethoprim and IC50 values for OCT2 and MATE2-K inhibition of lamivudine uptake by trimethoprim and the IC50 values for trimethoprim inhibition of OCT1 and MATE1-mediated uptake of other substrates (Table 2). It has been speculated that drug–drug interaction at MATE2-K might be responsible for the reduced renal secretion of lamivudine.2,50,61)
4.3. Some Studies in Knockout Mice and Nonsynonymous Single-Nucleotide Polymorphisms (SNPs) on the Impact of OCT2 and MATE1/2-K on MetforminEarlier studies reported on the pharmacokinetics of metformin administered orally or intravenously in knockout (KO) mice (including Oct1/Oct2-KO and Mate1-KO mice) or individuals with nonsynonymous SNP have provided essential insights on the impact of OCT2 and MATE1 in metformin renal excretion. Studies with Mate1-KO mice proved that metformin secretion in proximal tubular epithelial cells contributes considerably to urinary metformin excretion. They demonstrate that MATE1 is critical for metformin transport across the apical membrane of renal proximal tubular cells and is critically involved in metformin transport across the biliary membrane of hepatocytes. After oral administration of metformin, an increased AUC and decreased renal excretion of metformin in Mate1-KO mice compared to wild-type (WT) mice was observed.64,65) In addition, the hepatic concentration of metformin was significantly higher in the Mate1-KO than in the WT mice.66) During applying metformin in Mate1-KO mice, increased liver concentration and impaired renal excretion of metformin led to lactic acidosis compared to WT mice. Seven days after metformin was administered with drinking water, there was significantly higher blood lactate and lower pH in Mate1-KO than in WT mice.66) Experiments with Oct1/Oct2-KO mice showed a high impaired renal excretion of organic cations, but in mice lacking Oct2, it had a little effect as such TEA.67) They provide evidence that Oct1 and Oct2 together are essential for renal secretion of organic cations and play a vital role in increased drug sensitivity and toxicity when inhibiting these transporters in mice.14,67) Some studies have found lower renal metformin clearance in individuals with SNP in the OCT2 gene, suggesting that OCT2 may play a role in the renal secretion of metformin.68,69) However, the available data about individuals with SNPs in MATE1 and MATE2-K genes are still uncertain and/or contradictory.6)
The kidney is a common site for drug-related toxicity due to its crucial role in eliminating therapeutic drugs through a combination of passive glomerular filtration and active tubular secretion and tubular reabsorption (passive or active) processes occurring in the nephron.5) Up to 25% of acute kidney injury (AKI) cases in the critically ill are related to drug-induced kidney toxicity.70) Therefore, during the past decade, there has been an increased focus on finding renal toxicity related to a drug. Many different mechanisms contribute to drug-induced nephrotoxicity. These include tubular cell toxicity, altered intraglomerular hemodynamics, inflammation, rhabdomyolysis, thrombotic microangiopathy, and crystal nephropathy.71) The following section focuses on clinical studies that provide nephrotoxicity related to renal cation drug transporters and biomarkers that are directly related to tubular toxicity. Drug transporters expressed on the proximal tubular cells’ basolateral and apical membrane sides are considered one of the critical physiological factors determining the extent of drug-induced nephrotoxicity. Many clinically relevant significant DDIs in the kidney have been reported and recognized to inhibit OCT2 and MATE1/2-K mediated.2,5) Inhibition of OCT2 and MATE1/2-K in basolateral and apical membrane respectively, both reduced renal secretion, the impact on intracellular accumulation and nephrotoxicity is entirely different for the drug. As highlighted in Fig. 4, inhibition of the OCT2 uptake transporter reduces drug accumulation inside renal tubular cells, playing an important role in nephron protection. In contrast, inhibition of MATE1 and/or MATE2-K (efflux) mediates reduced renal secretion and exacerbates the intracellular accumulation and nephrotoxicity of the drug. Therefore, knowing the specific interaction site at the basolateral or apical membrane is crucial for predicting whether an inhibitor causes nephrotoxicity or is nephroprotective in vivo. Occasionally, co-treatment of an inhibitor is used intentionally to change renal clearance or minimize the nephrotoxicity of another drug.
When the inhibition of OCT2 in the basolateral membrane occurs, renal secretion and intracellular accumulation decrease. In contrast, inhibition of MATE1 and/or MATE2-K in the apical membrane leads to decreased renal secretion, exacerbating intracellular drugs such as cisplatin.
Cisplatin, an anti-cancer agent, is widely used to treat various solid tumors, including testicular, cervical, ovarian, non-small cell lung, and bladder cancer.72,73) Although cisplatin is a highly effective anti-cancer drug, its association with nephrotoxicity mainly limits the clinical application.72–75) Eliminations of cisplatin occur in proximal renal tubules via a process mediated by OCTs such as OCT2. Cisplatin accumulation in renal proximal tubular epithelial cells causes nephrotoxicity,2,72,76) and OCT2 is a high-capacity transporter for cisplatin accumulation in renal cells.77,78)
Previous experiments with knockout (KO) mice provided proof of principle that Oct2 and Oct1 in the basolateral side and Mate1 in the apical side of proximal renal tubular epithelial cells in mice are critically involved in eliminations of cisplatin and have an effect on the severity of cisplatin-induced nephrotoxicity. The urine secretion and intracellular accumulation of cisplatin were reduced with no severe nephrotoxicity when removing Oct1 and Oct2 in mice compared to WT mice.77) Renal impairment did not occur in the Oct1/Oct2-KO mice, as shown by higher blood urea nitrogen (BUN) and urine glucose levels in WT mice.77,79) In contrast, cisplatin caused more severe nephrotoxicity combined with a shorter survival duration of mice in Mate1-KO mice compared to WT mice.80) And cisplatin concentrations in the blood and kidney were increased one hour after intraperitoneal injection, and more substantial elevations in BUN and SCr were observed three days after cisplatin administration in Mate1-KO mice.80) Furthermore, some studies have found that OCT2 is critically involved in cisplatin uptake and impacts nephrotoxicity in individuals with SNP in the OCT2 gene. When patients carrying a nonsynonymous SNP in OCT2 were treated with cisplatin, there was no decrease of renal creatinine clearance indicating nephrotoxicity, in contrast to control patients.76,77)
The transporter combination OCT2 (uptake) in the basolateral membrane side and MATE1/2-K (efflux) in the apical membrane side mediates renal secretion of cisplatin drug (Fig. 4). Hence, co-administration of cisplatin with drugs that inhibit OCT2 provides an appealing strategy for preventing cisplatin-induced nephrotoxicity. However, drugs inhibiting MATE1 and/or MATE2-K may exacerbate cisplatin’s intracellular accumulation, promoting cisplatin-induced nephrotoxicity.75,78) Unfortunately, conclusive clinical evidence is inadequate to demonstrate that principle. The argument of principle that OCT2 inhibitors are potent therapeutic agents to prevent cisplatin-induced nephrotoxicity was provided in a clinical trial that co-administrated verapamil or cimetidine with cisplatin in cancer patients.81) Although many studies indicate that cimetidine might be an approach to decreasing cisplatin-induced nephrotoxicity due to its potent inhibitor of OCT2, inconsistent results have been revealed concerning its effectiveness in nephroprotective agents78,81–84); it may be explained partly, if not wholly, by the evidence that cimetidine is a more effective inhibitor of MATE1 than OCT2 at clinically relevant concentrations.42) These data are consistent with clinical studies describing high doses of cimetidine might be an approach to reduce cisplatin-induced nephrotoxicity due to the down-regulated activity of OCT2.78,81) L-Tetrahydropalmatine might be a candidate because it has selective inhibition of OCT2 mediate and reduces cisplatin-induced renal injury in human primary renal tubular.85) Also, germacrone may be an ideal candidate for reducing cisplatin-induced nephrotoxicity.86) Another example to prove this principle, erlotinib may exhibit a role in renal protection from cisplatin due to potently inhibited OCT2 function and decreased uptake of ASP+ and TEA, which are considered substrates of OCT2.87) In a randomized phase II clinical trial study, the co-treatment of erlotinib with cisplatin resulted in reduced cisplatin-induced nephrotoxicity,88) probably related to the inhibiting function of OCT2. Like erlotinib, nilotinib is potently inhibited OCT2 function in vitro test.89) In addition, evidence has been provided that numerous OCT2 inhibitors such as carvedilol, imatinib, and tropisetron may protect against cisplatin-induced nephrotoxicity in animal studies.79,90–92) Of note, entecavir, dolutegravir, and trospium are promising possibilities because they have selective inhibition of OCT2-mediated transport of some studied substrates2,56,93) (Table 2). One example of enhancing cisplatin-induced nephrotoxicity by co-administration with selective MATEs inhibitor is ondansetron that widely used as an antiemetic agent. Co-administration of ondansetron with cisplatin increases the cisplatin nephrotoxicity as observed in preclinical and clinical studies.94,95) The data suggest that inhibition of renal MATEs transporters is probably accountable for enhancing cisplatin-induced nephrotoxicity. In addition to ondansetron, pyrimethamine and trimethoprim can cause increased cisplatin-induced nephrotoxicity in mice due to the selective inhibition of MATE transporters.80) Tucatinib, vandetanib, and rucaparib were also shown to increase cisplatin nephrotoxicity.48,54,96) For these drugs, MATE1 and MATE2-K have a lower IC50 value than OCT2 for inhibiting organic cations. Therefore, drug inhibitors must be closely considered when used as nephroprotective agents during cisplatin treatment since OCT2 and MATEs have opposing effects on the accumulation of cisplatin in the renal tissue and the toxicity that results. Finally, further intensive investigation in the field is required to understand the mechanism and clinical significance of the interaction between these transporters and nephrotoxicity.
5.3. Biomarkers and Nephrotoxicity (to Proximal Tubular)More and more attention is being paid to identifying sensitive biomarkers that can be used to more efficiently identify and predict nephrotoxicity during drug development and clinical research. Besides, SCr and BUN are imperfect markers of kidney function because they have been affected by various renal and nonrenal factors independent of kidney function. Therefore, identifying biomarkers is of great interest because they are mainly associated with injury to the proximal tubule or glomerulus sites. They represent the major toxicological sites of >90% of medicines.70) Here we discussed some of the biomarkers of proximal tubule injury in the following section. Protein-based urine biomarkers with lower molecular weight proteins that are easily filtered, such as cystatin C (CysC), Beta-2 microglobulin (B2M), interleukin18 (IL18), neutrophil gelatinase-associated lipocalin (NGAL), and osteopontin (OPN), are primarily used to assess kidney tubule epithelial cell injury.60,97) These filtered proteins are efficiently reabsorbed via many receptors, such as B2M and neonatal Fc receptor (FcRn) in the proximal tubules, and are catabolized and reabsorbed to different amino acids and peptides such as CysC under normal conditions. However, when tubule epithelial cells are injured, urine levels of these proteins rise because the injured tubules’ reabsorption ability is reduced.60) Most of the remaining urine biomarkers are large proteins like alpha-glutathione-S-transferase (α-GST), N-acetyl beta-D-glucosaminidase (NAG), clusterin (CLU), and kidney injury molecule-1 (KIM-1), which do not go through glomerular filtration. The epithelial cells of the kidney produce these proteins as biomarkers in response to injury; therefore, their presence in the urine indicates increased excretion and/or cytotoxic release by the tubular epithelial cells.60,98) For preclinical evaluation of drug-induced nephrotoxicity, the FDA and Euro-Mediterranean Economists Association (EMEA) have recognized and accepted CysC, KIM-1, B2M clusterin, and OPN as biomarkers,97,99,100) and these biomarkers have a great correlation with histopathology examination.60,98) The clinical application of urine biomarkers to evaluate drug-induced kidney injury is still in its early stages. Emerging evidence suggests that kidney injury biomarkers such as NGAL and IL-18 may help detect kidney damage.70,97,98) Evidence has been provided that the urinary NGAL detected acute kidney injury earlier than standard SCr when treating patients with amphotericin B, which commonly causes tubule injury.101)
Recently, evidence showed that urine levels of KIM-1 were consistently linked to acute kidney injury, whereas NGAL was not, in several clinical trials involving cisplatin.102,103) B2M is a commonly used urine biomarker in the clinic.97) Patients treated with cisplatin, cyclosporin A, and gentamicin, all of which are nephrotoxic agents, had higher levels of B2M in their urine.97) After cisplatin exposure in rats, increased levels of α-GST, clusterin, and KIM-1 were found in the urine.104) Experiments with rodent models exposed to proximal (by cisplatin and gentamicin) provided evidence for the essential roles of urinary α-GST and the other urine markers that are outperformed SCr in recognizing proximal tubule toxicity.98,105) Besides, clusterin helped identify injuries to both the proximal and collecting ducts105) and suggests it can be used as an early marker with a similar profile to KIM-1.98,104) These urine biomarkers should be employed as a panel rather than as a single biomarker to capture drug-induced kidney damage completely.99,100) Although these biomarkers show enormous promise and the potential to dramatically improve the helpfulness of preclinical trials and the safety of clinical trials, there is still much to learn about these novel markers of drug-induced kidney toxicity.
Over 26 years of intensive investigation on OCTs, multiple key roles of OCT2 and MATE1/2-K in physiology and pharmacology have been firmly established. These transporters have many functions due to the capability for polyspecific recognition and transit of cation compounds. Realizing the extensive effects of OCT2 and MATE1/2-K on PK, PD, and renal toxicity of drugs and the interplay of various drugs with individual transporters have been established. Unfortunately, comprehensive knowledge of the functions of OCT2 and MATE1/2-K, DDIs of their transporters, and the prediction of two drugs' interactions in vivo at one transporter at a specific site remains limited. There are remaining challenges to anticipate and recognize DDIs mediated by OCT2 and/or MATE1/2-K. The most significant challenge is locating the actual sites precisely (basolateral vs. apical membranes) of renal in vivo interaction of two drugs. Besides, previous research indicates substrate-dependent and time-dependent inhibitions,88,89) which further increases the difficulty of evaluating and predicting DDIs. Extensive future research is needed to understand the pivotal roles of physiology and pharmacology for OCT2 and MATE1/2-K in the kidney, such as nephrotoxicity related to these transporters, in order to gain a better understanding of the principle that drug interaction with OCT2 inhibitors may be beneficial in reducing nephrotoxicity. We strongly believe that the information discussed in this review will shed additional light on the critical function of drug transporters in modifying DDIs, their effectiveness and provide full certainty about drug safety in a real-world clinical setting.
The authors declare no conflict of interest.