Update on drug-drug interaction at organic cation transporters: mechanisms, clinical impact, and proposal for advanced in vitro testing

ABSTRACT

Introduction: Organic cation transporters collectively called OCTs belong to three gene families (SLC22A1 OCT1, SLC22A2 OCT2, SLC22A3 OCT3, SLC22A4 OCTN1, SLC22A5 OCTN2, SLC29A4 PMAT, SLC47A1 MATE1, and SLC47A1 MATE2-K). OCTs transport structurally diverse drugs with overlapping selectivity. Some OCTs were shown to be critically involved in pharmacokinetics and therapeutic efficacy of cationic drugs. Drug-drug interactions at individual OCTs were shown to result in clinical effects. Procedures for in vitro testing of drugs for interaction with OCT1, OCT2, MATE1, and MATE2-K have been recommended.

Areas covered: An overview of functional properties, cation selectivity, location, and clinical impact of OCTs is provided. In addition, clinically relevant drug-drug interactions in OCTs are compiled. Because it was observed that the half maximal concentration of drugs to inhibit transport by OCTs (IC₅₀) is dependent on the transported cation and its concentration, an advanced protocol for in vitro testing of drugs for interaction with OCTs is proposed. In addition, it is suggested to include OCT3 and PMAT for in vitro testing.

Expert opinion: Research on clinical roles of OCTs should be reinforced including more transporters and drugs. An improvement of the in vitro testing protocol considering recent data is imperative for the benefit of patients.

KEYWORDS:

• Drug-drug interaction
• high affinity inhibition
• in vitro testing
• MATE1
• MATE2-K
• OCT1
• OCT2
• OCT3
• organic cation transport
• PMAT

1. Introduction

Transporters enable transit of drugs across plasma membranes and are required for absorption, cellular uptake, and excretion of hydrophilic drugs. About 40% of orally applied drugs are positively charged [1]. Most of these drugs are hydrophilic and exhibit no or only minor passive permeation across plasma membranes. Their translocation across plasma membranes is mediated by organic cation transporters (OCTs) that include OCT1 (SLC22A1), OCT2 (SLC22A2), OCT3 (SLC22A3), OCTN1 (SLC22A4), OCTN2 (SLC22A5), PMAT (SLC29A4), MATE1 (SLC47A1), and MATE2–K (SLC47A2) [2–4]. In recent years, the issue of drug-drug interaction has attracted increased attention. The reasons are that co-administration of several drugs is getting more and more popular, particularly for the elderly, and that an increasing knowledge of metabolizing enzymes, transporters, and drug targets provides new possibilities of preclinical in vitro testing for potential drug–drug interactions. In recent reviews various aspects of interaction of drugs at OCTs have been discussed focusing on effects on renal excretion and hepatic uptake of drugs [5–10]. In this review, the discussion is resumed on how in vitro testing of new molecular entities (NMEs) for interaction with OCTs should be performed [2,11]. Preclinical in vitro testing of OCT1, OCT2, MATE1, and MATE2-K for inhibition of organic cation transport by NMEs has been recommended by the Food and Drug Administration (FDA), European Medicines Agency (EMA), and Japan’s Pharmaceuticals and Medical Devices Agency (PMDA). The substrates and substrate concentrations for the recommended uptake measurements were not firmly defined. This article reviews data on functions and clinical relevance of OCTs published since 2015 and timely adapted protocols for in vitro testing are proposed.

In the first part, functional properties, main tissue distributions, membrane locations, and physiological functions of the OCTs are recalled. Thereafter, data is presented in which clinically relevant uptake of metformin, cisplatin, tropisetron, sumatriptan, and fenoterol by OCT1 and/or OCT2 were demonstrated by employing transporter polymorphisms with abolished or impaired transport function. Drug-drug interactions at OCT1, OCT2, MATE1, and/or MATE2-K are described that lead to changes of pharmacokinetics and/or therapeutic efficacy of metformin. In addition, effects of drug-drug interactions are reported that suggest clinically relevant drug transport by OCTs but do not identify the involved transporter(s) clearly. Reflecting on the roles of monoamine neurotransmitter uptake by OCT2, OCT3, and PMAT in brain and the interaction of psychotropic drugs with these transporters, anticipated drug-drug interactions in brain are discussed. Next, current molecular data on interaction of substrates and inhibitors with OCT1 and OCT2 is summarized. The data suggests that these transporters contain large clefts with different transport relevant binding sites and inhibitor high-affinity binding sites. Hypotheses are recalled on how OCT1 and OCT2 recognize and transport organic cations. These hypotheses may be expanded to the other OCTs. They provide explanations for the observation that the IC₅₀ values for inhibitors are dependent on the structure of the transported substrate and the concentration of the substrate. Finally, procedures for in vitro testing of the interaction of new molecular entities (NMEs) with OCTs are proposed that are adapted to current knowledge. A step-by-step approach including data on pharmacokinetics from the phase 1 clinical trial in a later phase is suggested. It is also suggested to include OCT3 and PMAT in routine in vitro testing.

2. Functional properties, locations, and biomedical functions of OCT1, OCT2, OCT3, OCTN1, OCTN2, MATE1, MATE2-K, and PMAT

2.1. Functional properties

OCT1, OCT2, and OCT3 are facilitative diffusion systems that translocate structurally diverse organic cations in both directions across the plasma membrane and are inhibited by additional compounds that are not transported [2,12]. The driving force for transport is provided by the concentration gradient of transported cations and the membrane potential. Equally, OCTN1, OCTN2, MATE1, MATE2–K, and PMAT are polyspecific transporters with overlapping substrate specificities [2–4,12]. OCTN1 and OCTN2 transport organic cations, zwitterions, and uncharged compounds in partially different ways [2,3]. Tetraethylammonium (TEA) uptake by OCTN1 is stimulated by intracellular protons and ATP [3,13], whereas uptake of zwitterions by OCTN1 may be stimulated or inhibited by extracellular sodium [14]. OCTN2 operates as Na⁺-carnitine cotransporter and mediates electrogenic sodium independent uptake of TEA [2]. MATE1 and MATE2-K are electroneutral proton-cation antiporters that mediate cellular cation efflux but can operate in both directions [2]. PMAT mediates electrogenic transport of organic cations that are stimulated by protons on the cis-side suggesting proton–cation cotransport [4].

2.2. Expression and membrane location

Tissue expression and plasma membrane location of OCTs have been described in detail in previous reviews [2–4,15]. In brief, OCT1, OCT3, OCTN1, OCTN2, MATE1, and PMAT are broadly expressed, whereas OCT2 and MATE2-K are distributed less widespread. The main sites of expression in humans and some additional sites where drug-drug interactions may occur are as follows. OCT1 is strongly expressed in the sinusoidal membrane of hepatocytes. In addition, OCT1 is expressed in brush-border membranes (BBMs) of small intestine and renal proximal tubules (Figure 1). Among others, expression of OCT1 was observed in brain micro vessels, adipocytes, lymphocytes, and monocytes. OCT2 is mainly expressed in kidney where it is located in the basolateral membrane (BLM) of proximal tubular cells (Figure 1). Expression of OCT2 was also detected in cerebral neurons and brain micro vessels. OCT3 is expressed in many organs including small intestine, liver, kidney, brain, heart, adipose tissue, skeletal muscle, lymphocytes, and monocytes. In small intestine, OCT3 was localized to the BBM of enterocytes, in liver it was localized to the sinusoidal membrane of hepatocytes, and in kidney to the BLM of proximal tubular cells (Figure 1). OCTN1 is expressed in small intestine, liver, and kidney, lymphocytes, and monocytes [3]. In small intestinal enterocytes and renal proximal tubules, OCTN1 was localized to the BBMs (Figure 1). OCTN2 is ubiquitously expressed including expression in small intestine, liver, kidney, lymphocytes, monocytes, and micro vessels of brain [2,3]. Similarly, MATE1 is ubiquitously expressed, amongst others in kidney, liver, skeletal muscle, and micro vessels of brain [2]. MATE1 was localized to the canalicular membrane of hepatocytes and to the BBM of renal proximal tubular cells (Figure 1). MATE2-K is mainly expressed in kidney where it was localized to the BBM of proximal tubular cells (Figure 1). In addition, MATE2–K was detected in small vessels of the brain [2]. Also PMAT is expressed in many organs including small intestine, liver, kidney, brain, and heart [4]. In small intestine, PMAT was observed in the BBM whereas in kidney, PMAT was localized to podocytes of the glomeruli (Figure 1). In brain, PMAT is expressed in neurons and glial cells, in the choroid plexus, and in micro vessels [15].

Figure 1. OCTs in small intestinal enterocytes, hepatocytes, and renal proximal tubular cells. Translocation of organic cations by OCTs during intestinal absorption, uptake into hepatocytes and renal secretion is indicated by red arrows. Obligatory exchange of protons is indicated in solid lined arrows. Translocation of organic cations, Na⁺ or zwitterions that can occur in exchange to transport of organic cations is depicted in dashed arrows. OC⁺ organic cation, ZI zwitterion

2.3. Accepted and proposed functions in small intestine, liver, kidney, and brain

OCTs are involved in absorption and/or excretion of organic cations in small intestine, uptake of organic cations into hepatocytes and their biliary excretion, and reabsorption or excretion of organic cations in kidney [2,16]. In addition, OCTs contribute to uptake of cationic drugs into brain and removal of monoamine neurotransmitters from the cerebral interstitial space into neurons and glial cells [15,17–20]. The present review will focus on OCTs that participate in small intestinal absorption, hepatic uptake, and renal secretion of drugs, and on drug-drug interactions that inhibit these functions. In addition, presumed roles of organic cations in the brain and potential drug-drug interactions in brain are discussed.

In the BBM of small intestinal enterocytes, many transporters have been identified that transport cations and may participate in uptake of cationic drugs. In addition to the choline transporter (CHT)1 (SLC5A7), the high-affinity Na⁺-cotransporters for norepinephrine (NET1, SLC6A2), dopamine (DAT1, SLC6A3), and serotonin (SERT, SLC6A4), the thiamine transporter (THTR)2 (SLC19A3), and the choline–like transporters (CLT)1-5 (CLC44A1-5), the polyspecific OCTs OCT1, OCT2, OCTN1, OCTN2 and PMAT depicted in Figure 1 are involved [2,4]. The substrate selectivities of these transporters overlap, showing distinct differences for individual substrates. Since the abundance of these transporters in the BBM has not been compared systematically and their turnover rates for individual drugs have not been determined, their functional relevance for uptake of individual drugs has not been resolved. Notably, there is a complete lack of knowledge concerning the transporters that are responsible for the second step of small intestinal cation absorption, the translocation of organic cations across the BLM.

Translocation of organic cations across the sinusoidal membrane of hepatocytes is mediated by OCT1 and OCT3 [2] (Figure 1). OCT1 is more abundantly expressed than OCT3 [21]. Compared to OCT1, OCT3 may be more relevant for hepatic uptake of drugs that have a higher turnover and/or lower K_m value for transport such as berberin, ipratropium, and trospium [2]. In hepatocytes, also the expression of NET, SERT, PMAT, and of the thiamine transporter (THTR) 1 (SLC19A2) has been observed [2]. Since these transporters are driven by the inside directed gradient of sodium or protons, they are probably located in the sinusoidal membrane. Export of organic cations from hepatocytes into the bile is performed by MATE1 and multidrug resistance protein (MDR)1 (ABCB1) in the biliary membrane [16]. MATE1 exchanges various organic cations against extracellular protons, whereas MDR1 is an ATP dependent pump that accepts some relatively large hydrophobic, so-called type 2 cations as substrates [2,16].

Renal secretion of organic cations mainly occurs in renal proximal tubules. OCT2 and OCT3 in the BLM mediate uptake of organic cations from interstitial space into renal proximal tubule cells [2]. In the BBM many transporters have been identified that have been shown or suggested to participate in export of organic cations into the tubular lumen [2] (Figure 1). The cation-proton exchangers MATE1 and MATE2-K should be most effective. However, also OCT1, OCTN1, and OCTN2 could contribute to cation secretion when they operate in exchange modes. OCT1, OCTN1, and OCTN2 may export intracellular cations in exchange for cations in the tubular lumen. OCTN1 may export cations in exchange for zwitterions that may be cotransported with sodium or protons, and OCTN2 may export cation in exchange for carnitine that is cotransported with Na⁺ [3,22]

OCT1, OCT2, OCT3 and PMAT are low-affinity transporters for the monoamine neurotransmitters norepinephrine (NE), dopamine (DA), and serotonin (5-hydroxytryptamine, 5-HT) [2,15]. Like OCTN2, these OCTs are expressed in the blood–brain barrier [2,15]. They are also expressed in neurons of various brain areas and/or in glial cells [18–20,23–25]. While a detailed localization of OCTs in brain was performed in rodents, only few localizations were confirmed in humans [15]. Our knowledge of physiological functions of OCT2 and OCT3, and the impact of OCT2 and OCT3 during application of psychotropic drugs is exclusively derived from experiments with rodents [2,15,17,20,25–27]. These experiments include behavioral tests in animal models of depression, stress, anxiety, and addiction. The localization of OCT1, OCT2, OCT3, and PMAT in endothelial cells of brain capillaries has been confirmed in humans [15,18]. It implies the involvement of these OCTs in cerebral uptake of drugs. As demonstrated in rodents, OCT2, OCT3, and PMAT in humans are supposed to be located at plasma membranes of neurons in various brain regions [15]. As demonstrated in rodents, OCT2 and OCT3 in humans are supposed to be located at nerve terminals close to the synaptic clefts. It is believed that they supplement removal of monoamine neurotransmitters by NET, SERT and/or DAT. Thereby, OCT2 and OCT3 are supposed to modulate activation of neuronal circuits that steer complex cerebral functions including regulation of cognitive functions, locomotion, mood control, stress, and/or anxiety [15]. As in rodents, these functions of OCT2 and OCT3 in humans are believed to gain specific relevance when psychiatric drugs inhibit the Na⁺ dependent high-affinity monoamine neurotransmitter transporters.

3. Clinically relevant drug transport by OCTs and drug-drug interactions at OCTs

3.1. General considerations

The clinical impact of individual OCTs for the pharmacokinetics and/or therapeutic efficacy of a specific drug has only been resolved in a few cases. Such an assignment is difficult because each OCT may be expressed in different locations that may be involved in absorption, excretion and/or uptake into target cells. In addition, other transporters with overlapping cation selectivity are often present in the same plasma membrane or in plasma membranes on opposite sides of epithelial cells that are involved in absorption or excretion. Moreover, membrane expression, Michaelis Menten constant values, and catalytic turnover numbers of transporters translocating identical drugs may be largely different. While plasma membrane expression and the Michaelis Menten constant values may be estimated on the basis of immunohistochemistry and in vitro uptake measurements no satisfactory method to determine in vivo turnover numbers is available. The critical roles of individual OCTs in absorption, excretion and/or efficacy of drugs have been suggested based on clinical studies. Unequivocal demonstration of clinical relevance of OCT1 and/or OCT2 for pharmacokinetics and/or therapeutic efficacy of metformin, cisplatin, tropisetron, sumatriptan, and fenoterol was provided by studies on transporters containing polymorphisms that induce impaired or abolished transport (Table 1).

Table 1. Functions of OCTs that have been shown or are strongly suggested to be critical for pharmacokinetics and/or therapeutic effects of drugs

Drug	OCTs shown (or supposed) to be critically involved in intestinal absorption, hepatic uptake, and/or renal excretion				References
	Small intestinal absorption	Uptake into hepatocytes	Renal secretion	Renal reabsorption
Metformin	(OCT1)	OCT1	OCT2, (MATE1), (MATE2–K)	OCT1, OCT2	[2,28-34]
Cisplatin			OCT2, (MATE1), (MATE2-K)		[2,35,36]
Tropisetron		OCT1			[37]
Sumatriptan		OCT1			[38]
Fenoterol		OCT1			[39]

3.2. Clinical relevance of OCTs for metformin treatment and related drug-drug interactions

3.2.1. Introductory remarks

Metformin is one of the most frequently prescribed drugs used for the first-line treatment of type 2 diabetes mellitus. Metformin enters hepatocytes where it increases glycogen synthesis and decreases gluconeogenesis [40]. Skeletal muscle and fat cells are additional target sites for the antidiabetic effect of metformin. Being positively charged in the blood, metformin cannot permeate the plasma membrane passively [2]. Transmembrane movement of metformin can be mediated by OCT1, OCT2, OCT3, OCTN1, PMAT, MATE1 and MATE2-K [2] (Table 2). After oral application, metformin is absorbed in small intestine involving OCTs in the BBM [2] (Figure 1). Metformin does not bind to plasma proteins and is not metabolized. It enters hepatocytes across the sinusoidal membrane via OCT1 and OCT3 (Figure 1, Table 2). Metformin could leave the hepatocytes across the biliary membrane via MATE1 and across the sinusoidal membrane via OCT1 and OCT3. Whereas biliary excretion of metformin is supposed to be absent or minimal [46], efflux of metformin from hepatocytes into sinusoids may be relevant. It may occur at low metformin concentration in the blood in exchange for cationic substrates. Metformin is excreted with urine. It is ultrafiltrated in the glomeruli and secreted in proximal tubules (Figure 1). Some reabsorption of metformin in the proximal tubule appears to be possible.

Table 2. K_m values of drugs transported by OCTs for which OCT-mediated uptake proved to be clinically relevant

Drug (Cmax unb, µM)	OCT Km [µM]								References
	OCT1	OCT2	OCT3	OCTN1	OCTN2	PMAT	MATE1	MATE2–K
Cisplatin (3.3)	t	11	n.t.d.	n.t.d.	n.t.d.		t	t	[2,35]
Crizotinb (0.1)	n.t.d.	1.2							[41]
Fenoterol (0.001)	1.8	t	20						[2,39]
Lamuvidine (6)	249, 1250	249, 1900	2140				t	t	[2,42]
Metformin (16)	1470, 2160	285–3170	1090, 2260	t		1320	227–780	1050, 1980	[2,43]
Sumatriptan (0.3)	55	t	t						[2,38]
Tropisetron (0.09)	t								[2,37]
Trospium (4)	15–106	0.6- 8	4.4				15	8.2	[2,44]
Varenicline (0.03)		370							[45]

C_{max unb} estimated maximal concentration of unbound drug in serum of systemic blood, t transported, n.t.d. no transport detected

The K_m values of MATE1 and MATE2-K were obtained by measuring cellular uptake. They may not represent the physiological relevant K_m values for cellular export in exchange to protons.

Figure 2 depicts clinical effects that are expected when metformin absorption in small intestine, metformin uptake into hepatocytes, or metformin secretion in renal proximal tubules is impaired. The concentration of metformin in the systemic blood may be decreased when small intestinal absorption is impaired and increased when renal metformin secretion is reduced. An impaired uptake of metformin into hepatocytes will diminish or abolish the therapeutic effect. In this situation, the concentration of metformin in systemic blood may be slightly increased. Renal clearance of metformin may be decreased when renal secretion of metformin is impaired and may be increased when renal reabsorption of metformin is inhibited.

Figure 2. Anticipated effects of drugs that inhibit metformin transport by OCT1, OCT2, MATE1 and/or/MATE2-K in small intestine, liver, and kidney on pharmacokinetics, therapeutic efficacy and/or renal clearance

3.2.2. Effects of polymorphisms in OCTs on metformin treatment

To determine the clinical impact of individual OCTs, single nucleotide variants (SNVs) in OCT1, OCT2, OCT3, OCTN1, PMAT, MATE1, and MATE2-K were investigated according to therapeutic efficacy, pharmacokinetics, and/or side effects of metformin treatment [2,28]. Ambiguous and/or contradictory data have been reported characterizing variants of OCT3, OCTN1, PMAT, MATE1, and MATE2-K. Clinically relevant roles of OCT1 for intestinal absorption, for uptake into hepatocytes, and for renal reabsorption have been demonstrated or suggested by studying the effects of reduced-function variants of OCT1. The involvement of OCT1 in intestinal absorption of metformin was suggested because the frequency of gastrointestinal side effects that are supposed to be caused by a high concentration of metformin in the small intestinal lumen was increased in patients with reduced-function variants of OCT1 [29]. This could promote a decrease in metformin in blood and decrease the therapeutic efficacy (Figure 2). Employing ¹¹C positron emission tomography, it was observed that metformin uptake into liver was decreased in individuals with reduced-function OCT1 variants [30]. Together with the observation that the metformin induced decrease in D-glucose in the oral glucose tolerance test was blunted in individuals with reduced function variants of OCT1 [31], it can be concluded that OCT1 is essential for hepatic uptake of metformin into hepatocytes which is required for a distinct antidiabetic effect. Impairment of metformin uptake into liver could also increase the concentration of metformin in the blood (Figure 2). The observation that the renal clearance was increased in individuals with reduced-function variants of OCT1 suggests that OCT1 in the BBM of renal proximal tubular cells is involved in renal metformin reabsorption (Figure 2) [32]. The multiple involvement of OCT1 in the pharmacokinetics of metformin may explain why an unchanged or increased blood peak concentration of metformin was observed in individuals with reduced–function OCT1 variants [32,47,48].

For individuals with SNVs in the OCT2 gene, a reduced renal metformin clearance was observed in some populations suggesting clinical relevance of OCT2 for renal metformin secretion [2,28,33,34]. Effects of drug-drug interactions described below also supported this. The renal clearance of metformin was also reduced in Korean individuals with SNVs in PMAT [28]. However, the mechanism for this effect remains enigmatic because the expression and the membrane location of PMAT in renal tubules have not been demonstrated [4].

3.2.3. Effects of drugs interacting with OCTs on metformin treatment

During treatment of type 2 diabetes mellitus with metformin, reduced therapeutic efficacy was observed when various drugs that interact with OCTs were co-administered (Tables 3 and 4). The drugs include verapamil and citalopram, suggesting drug-drug interactions at OCTs [2,10,64,65].

Table 3. In vitro determined IC₅₀ or K_i values of perpetrator drugs that inhibit transport of victim drugs by OCTs and affect their pharmacokinetic and/or therapeutic efficacy

	IC50 (Ki) values for uptake inhibition of model substrates, metformin (m), crizotinib (cri) or lamuvidine (la) [µM]
Perpetrator drug (Cmax unb, µM) Victim drug	OCT1	OCT2	OCT3	OCTN1	OCTN2	PMAT	MATE1	MATE2–K	References
Cimetidine (8) Metformin	95–2830, 104 ^m	15–146, 124 ^m	88, 240	435			1.1–170 2.6 ^m,10, 3.8 ^m	2.1–370, 5.5 ^m,10, 6.9 ^m	[2,49,50]
Citalopram (0.008) Metformin	3, 19	12	145			117			[2,51]
Dolutegravir (0.14) Metformin		0.07, 1.9	>100, >100 ^m,10			>100 ^m,10	4.7, 6.3 ^m,10	>100, 25 ^m,10	[2,52,53]
Entecavir (0.013) Crizotinib		0.037^cri,1							[41]
Famotidine (1) Metformin	28, >300, 19 ^m,10	36, 114, 66 ^m,10	6.7, 11				0.3–6.7 0.91 ^m,10, 0.25 ^m,10	3.1–36 3.1 ^m,10 2.5 ^m,10	[2,50,54]
Metformin (16) Trospium	1230, 9480	521–2370	2330				47–667	89–6516	[2]
Pefiticinib (0.44) Metformin	0.28 ^m,10	71 ^m,10					10 ^m,10	21 ^m,10	[55]
Pyrimethamine (0.8) Metformin	8.5, 14	1.6–23					0.04–1.2, 0.3 ^m,10	0.01–0.83 0.2 ^m,10	[2,50]
Ranolazine (2) Metformin		19 ^m,2							[56,57]
Trimethoprim (17) Metformin Lamuvidine	20–57	14-1318, 27 ^m, 13^la,1	12				2.7–>100, 4.1 ^m,10, 6.3 ^m	0.42–1.9 0.42 ^m,10, 29 ^m, 0.66^la,1	[2,42,50,58]
Tucatinib (1.4) Metformin		0.11, 15 ^m,10					0.09 0.34 ^m,10	0.14 ^m,10	[59]
Vandetanib (0.4) Metformin		73, 8.8 ^m,1					1.2, 0.16 ^m,1	1.3, 0.30 ^m,1	[60,61]
Verapamil (0.06) Metformin	1–33	0.6–92, 10 ^m,16	57	8-25	21, 51		28, 42	32, 38	[2,62,63]

Victim drugs and their micromolar concentrations used as substrates for determination of IC₅₀ values are indicated as superscripts. IC₅₀ values without superscripts were determined with different concentrations of cationic model substrates

C_{max unb} estimated maximal concentration of unbound drug in the systemic blood calculated from values reported in the literature, m metformin, cri crizotinib, la lamuvidine

The indicated IC₅₀ and K_i values for MATE1 and MATE2-K were determined for cis inhibition of cellular drug uptake. They may be different from the physiological values of trans inhibition or cis inhibition of cellular export of the drugs.

Table 4. Drug-drug interactions at OCTs with clinical impact on metformin treatment of type 2 diabetes mellitus

Perpetual drug	Observed (supposed) effects on metformin treatment					Supposed victim OCTs
	Intestinal absorption	Renal clearance	Hepatic uptake	Plasma concentration	Therapeutic effect
Verapramil	(no change)	no change	(decrease)	no change	decrease	OCT1
Pefiticinib	(decrease)		(decrease)	decrease	(decrease)	OCT1
Cimetidine		decrease		(increase)		MATE1 and/or MATE2-K
Pyrimethamine		decrease		(increase)		MATE1 and/or MATE2-K
Vandetanib		decrease		increase		MATE1 and/or MATE2-K
Trimethoprim		decrease		increase		MATE2-K
Ranolazine		(decrease)		increase
Tucatinib		decrease		increase		MATE1 and/or MATE2-K
Dolutegravir		(decrease)		increase		OCT2
Famotidine	(increase)	increase		no change	increase	MATE1

Although verapamil is able to inhibit various OCTs (Table 3), only the inhibition of OCT1 mediated metformin uptake into hepatocytes is supposed to be clinically relevant because the concentration of metformin in systemic blood and renal metformin secretion were unchanged (Table 4) [65]. In addition, the maximal concentration of unbound verapamil (C_{max unb}) in systemic blood is too low to inhibit OCT2 mediated metformin uptake in kidney (Table 3). In contrast, C_{max unb} of verapamil of 1.7 µM was estimated in the portal vein [66] that could be high enough to inhibit OCT1 mediated metformin uptake into hepatocytes (Table 3). Note that IC₅₀ values for inhibition of OCT1 mediated transport indicated in Table 3 may not be representative of the clinical situation because they were not determined with metformin as substrate.

The novel antirheumatic drug pefiticinib inhibited OCT1 mediated uptake of 10 µM metformin by OCT1 with an IC₅₀ value of 0.28 µM [55]. Since this value is 1.6-fold lower than the estimated C_{max unb} in systemic blood observed after a single oral dose and the respective IC₅₀ values for OCT2, MATE1 and MATE2–K are 23-161 times higher (Table 3), pefiticinib may inhibit OCT1 in small intestine, liver and/or kidney. Inhibition of OCT1 in small intestine is suggested because the C_max of metformin in blood after oral co-administration of pefiticinib and metformin was 17% decreased [55]. Because pefiticinib in the portal vein most probably inhibits metformin uptake into hepatocytes, it is expected that pefiticinib blunts the therapeutic efficacy of metformin (Table 4).

Upon co–administration of the antihistaminic cimetidine or the antimalarial drug pyrimethamine with metformin, a decrease in renal metformin clearance was observed [34,58,67,68]. This could be due to interactions with OCT2, MATE1 and/or MATE2-K that are supposed to be critically involved in renal metformin secretion. However, since the IC₅₀ values of cimetidine and pyrimethamine for uptake inhibition of 10 µM metformin by MATE1 and MATE2–K are 1.5-4 times lower than the estimated C_{max unb} of compounds in systemic blood, and the determined IC₅₀ values for OCT2 are at least twice as high (Table 3), the effective interaction probably occurs at MATE transporters in the BBM (Table 4).

Similarly, it is supposed that vandetanib used for treatment of myeloid thyroid cancer and the antibiotic trimethoprim reduces renal secretion of metformin by inhibiting MATE transporters. After co-administration of vandetanib and metformin, the renal metformin clearance was decreased by about 50% and C_max of metformin in systemic blood was increased by about 50% [69]. The IC₅₀ values for uptake inhibition of 1 µM metformin mediated by OCT2, MATE1, and MATE2-K were 8.2 µM, 0.16 µM, and 0.3 µM, respectively [60], and for C_{max unb} of vandetanib in systemic blood, a value of 0.4 µM was estimated [61] (Table 3). This suggests inhibition of MATE1 and MATE2-K.

Co-administration of trimethoprim with metformin induced a 32% decrease in renal metformin clearance and a 38% increase of C_max of metformin in blood after oral application [70]. The K_i values determined for trimethoprim inhibition of metformin uptake by OCT2 and MATE1 were more than 50% higher than the estimated maximal concentration of unbound trimethoprim in systemic blood, whereas the K_i value for MATE2-K was 63% lower suggesting an effect on MATE2-K [58] (Table 3).

Other examples of clinically relevant interactions with MATE transporters in the kidney are ranolazine that is employed for treatment of chronic angina pectoris, and the antineoplastic drug tucatinib. Co-administration of ranolazine with metformin caused an approximate 50% increase in the concentration of metformin in systemic blood implicating an impairment of renal metformin secretion (Table 4) [56]. Because for uptake inhibition of 2 $\mu \mathrm{M}$ metformin by OCT2 an IC₅₀ value of 19 µM was determined and 2 µM was estimated for C_{max unb} of ranolazine in systemic blood (Table 3) [56,57], the decrease in renal secretion may be mainly due to inhibition of MATE1 and MATE2-K that have not been characterized by inhibition of ranolazine. When tucatinib was administered in combination with metformin, the renal metformin clearance was reduced by about 40% and the maximal concentration of metformin in systemic blood was slightly increased by 15% (Table 4) [59]. For in vitro inhibition of metformin uptake by OCT2, MATE1, and MATE2-K with tucatinib, respective IC₅₀ values of 15 µM, 0.34 µM, and 0.14 µM were determined (Table 3) [59]. The uptake measurements were performed with 10 µM metformin. Combined with an estimated C_{max unb} of tucatinib of 1.4 µM in systemic blood it is probable that an inhibition of renal MATE transporters is responsible for the observed decrease in renal metformin secretion.

The antiretroviral drug dolutegravir probably inhibits renal metformin secretion by interaction with OCT2 in the BLM. Two daily applications of 50 mg dolutegravir during metformin treatment resulted in a 110% increase of C_max in systemic blood implying a reduction of renal metformin secretion (Table 4) [52]. An inhibition of OCT2 by dolutegravir is suggested because the lowest IC₅₀ value measured for uptake of model substrates by OCT2 is two times lower than the estimated C_{max unb} of dolutegravir in systemic blood (Table 3) [5,52,53]. In contrast, for uptake inhibition of 10 µM metformin by MATE1 and MATE2-K, IC₅₀ values were determined that are more than 40 times higher than C_{max unb} of dolutegravir. The large increase in metformin in blood observed after co-administration of dolutegravir suggests drug-drug interaction at an additional transporter.

C_max of metformin in systemic blood was unaffected after oral co-administration of the H2-receptor antagonist famotidine with metformin, although renal clearance was increased (Table 4) [54]. Unexpectedly, the bioavailability of metformin was increased and the ability of metformin to lower the blood glucose increase during the oral glucose tolerance test was reinforced. Since the IC₅₀ value for famotidine inhibition of MATE1 mediated metformin uptake was fourfold lower than the maximal concentration of famotidine in systemic blood, whereas the IC₅₀ values for famotidine inhibition of metformin uptake by OCT1, OCT2, or MATE-2 K were at least 2.5 higher, it was speculated that MATE1 inhibition is critically involved. The detailed mechanisms underlying the observed effects remain obscure.

3.3. Clinical relevance of OCTs for cisplatin treatment and related drug-drug interactions

3.3.1. Introductory remarks

Cisplatin is used for treatment of various solid tumors including non-small cell lung cancer and cancer of testis [2]. Therapeutic application is mainly limited by nephrotoxicity of cisplatin. Cisplatin is excreted in renal proximal tubules with OCTs and the nephrotoxicity is caused by accumulation of cisplatin in renal tubular epithelial cells. In vitro it was observed that cisplatin is transported by OCT1, OCT2, MATE1 and MATE2-K (Table 2) [2]. OCT2 is supposed to be most effective.

3.3.2. Data suggesting that OCT2 is critical for cisplatin nephrotoxicity

The critical impact of OCT2 located in the BLM of renal proximal tubular cells has been demonstrated in mice and humans. In proximal tubular cells of mice Oct1 and Oct2 are expressed in the BLM [22], whereas in humans OCT2 is expressed in the BLM while OCT1 is expressed in the BBM [2]. After removing Oct1 and Oct2 in mice, urinary secretion and renal accumulation of cisplatin were decreased and no severe nephrotoxicity was observed, in contrast to wildtype mice [35]. In patients carrying a nonsynonymous SNV in OCT2, no impairment of renal creatinine clearance indicating nephrotoxicity was observed after treatment with cisplatin, in contrast to control patients [35,36]. Since this mutation causes decreased transport of TEA, MPP, and propranolol [2] it may also impair the transport of cisplatin.

3.3.3. Interaction of drugs with cisplatin transport by OCTs

Assuming that secretion of cisplatin in renal proximal tubular cells is mainly mediated by OCT2 in the BLM and MATE1 and/or MATE2-K in the BBM (Figure 1), inhibition of OCT2 should reduce nephrotoxicity, whereas inhibition of MATE1 and MATE2-K should increase nephrotoxicity. Drugs that inhibit OCT2 as well as MATE transporters are supposed to have opposite effects that depend on their abilities to inhibit cisplatin uptake at clinically relevant concentrations. The proof of principle that drugs interacting with OCT2 can reduce cisplatin nephrotoxicity was provided in a small clinical study in which cimetidine or verapamil were co-administered to patients treated with cisplatin [71]. Because the IC₅₀ values for inhibition of cisplatin transport by cimetidine and verapramil via MATE transporters and OCT2 have not been determined, clinical situations cannot be predicted in which predominantly MATE transporters are inhibited, and nephrotoxicity is increased. It remains a challenge to identify a safe drug for renoprotection that selectively inhibits cisplatin uptake by OCT2. L-tetrahydropalmitine may be a candidate because it has been shown to inhibit organic cation transport by OCT2 and to reduce toxicity of cisplatin in primary cultured human tubular cells [72]. Promising candidates are also entecavir, dolutegravir, and trospium that inhibit OCT2 mediated transport of some tested substrates with very high affinity (Table 3) [2]. On the other hand, drugs like pyrimethamine, trimethoprim, tucatinib, vandetanib, and rucaparib carry a high risk of increasing cisplatin nephrotoxicity [73] (Table 3). For these drugs, lower IC₅₀ values were described for the inhibition of organic cations by MATE1 and/or MATE2-K compared to OCT2.

3.4. Clinical relevance of OCT1 mediated transport of drugs other than metformin indicated by studies on loss-of-function polymorphisms

3.4.1. Introduction

OCT1 is highly polymorphic in contrast to the other OCTs. Twenty-seven nonsynonymous SNVs in OCT1 have been investigated for their effects on expression, transport activity, and/or plasma membrane targeting [2]. Tzvetkov and coworkers defined groups of haplotype allele combinations of OCT1 that include groups with abolished transport activity [74]. The large number of loss-of-function polymorphisms together with their high prevalence in some ethnicities indicates that OCT1 mediated transport under physiological conditions can be substituted by co-expressed transporters. However, transport by OCT1 may be essential for pharmacokinetics and/or therapeutic efficacy of individual drugs. Studying effects of loss-of–function mutations, clinically relevant OCT1 mediated uptake into hepatocytes has been demonstrated for several drugs in addition to metformin: the bronchodilator fenoterol, the antimigraine drug sumatriptan, and the antiemetic drug tropisetron [37–39]. Moreover, data were reported suggesting that OCT1 is critically involved in hepatic uptake of morphine and anti-malaria drug proguanil [2]. In general, one can say that the pharmacokinetics of drugs that are transported by OCT1 and metabolized in liver may be changed when drugs that inhibit OCT1 are co-administered.

3.4.2. Transport of fenoterol by OCT1

Fenoterol is a beta₂ adrenoreceptor agonist that is employed for inhalation treatment of bronchial asthma and chronic obstructive pulmonary disease [39]. During fenoterol treatment, adverse cardiovascular and metabolic reactions may occur that correlate with systemic exposure. Fenoterol that enters the blood is taken up by hepatocytes, metabolized by sulfatation and glucuronidation, and mainly excreted as metabolite [2,39]. 99% of fenoterol in blood is positively charged and reliant on a transporter to cross plasma membranes. Evidence has been presented that OCT1, OCT2, and OCT3 accept fenoterol as a substrate (Table 2). For fenoterol transport by OCT1 and OCT3, K_m values of 1.8 µM and 20 µM were determined, respectively. A clinically relevant role of OCT1 mediated uptake into hepatocytes was demonstrated in a study comparing pharmacokinetics and side effects of fenoterol in healthy volunteers without and with heritable OCT1 deficiency [62]. After intravenous injection of fenoterol, the C_max of fenoterol in blood of individuals with OCT1 deficiency was twofold higher compared to individuals with functional OCT1 [39]. Consistently, side effects of fenoterol treatment – an increased heartbeat rate, a decrease in blood pressure, an increase in blood glucose, and a decrease in blood potassium concentration– reinforced. The data suggest that fenoterol exposure is increased because fenoterol uptake into hepatocytes is reduced and metabolism and excretion of fenoterol metabolites are slowed down.

3.4.3. Transport of sumatriptan by OCT1

The antimigraine drug sumatriptan has very low bioavailability due to poor small intestinal absorption and extensive first-pass metabolism involving monoamine oxidase A [38]. Sumatriptan is metabolized in several organs, however, metabolism in liver is most important. Being more than 95% positively charged in blood, sumatriptan has a very low passive plasma membrane permeability and transporters are required for intestinal absorption and uptake into hepatocytes. It has been demonstrated that sumatriptan is transported by OCT1, OCT2, and OCT3 (Table 2). For OCT1, a K_m value of 55 µM was determined [38]. After oral application of 50 mg sumatriptan to healthy volunteers with two loss-of-function OCT1 alleles compared to individuals possessing at least one functional OCT1 allele, the C_max of sumatriptan in blood was increased 60% [38]. This effect is probably due to a slowed down uptake of sumatriptan into hepatocytes resulting in a decreased metabolism. OCT1 in the BBM of small intestine is probably not critically involved in sumatriptan absorption because the sumatriptan concentration in the small intestinal lumen after oral application is supposed to by far exceed the K_m of OCT1.

3.4.4. Transport of tropisetron by OCT1

The 5-hydroxytryptamine 3 receptor antagonist tropisetron is an antiemetic drug that is administered during cytostatic therapies. Tropisetron is mainly metabolized in the liver by cytochrome P450 2D6 (CYP2D6) and mutants of CYP2D6 have been associated with the failure of antiemetic therapy [37]. Because about 97% of tropisetron is positively charged at pH 7.4 tropisetron is heavily dependent on transporters to cross plasma membranes. Tropisetron is transported by OCT1 and inhibits OCT1 mediated uptake of 1 µM (4-(4-(dimethylamino)styryl)-N-methylpyridinium (ASP) with an IC₅₀ value of 8.6 µM [37]. Three hours after an oral application of 5 mg tropisetron, patients with two nonfunctional OCT1 alleles showed an approximately threefold higher tropisetron concentration in the systemic blood compared to patients with one or two functional OCT1 alleles. The increased blood concentration was associated with an increased therapeutic effect that consisted of fewer episodes of vomiting. The data suggest that OCT1 mediated uptake of tropisetron into hepatocytes is critical for metabolism while OCT1 does not play a significant role in intestinal absorption.

4. Clinical relevance of drug-drug interactions at not yet identified OCTs

4.1. Introductory remarks

Drug-drug interactions may help to elucidate clinical relevance of OCT-mediated drug transport. Pharmacological functions of drugs transported by OCTs may be changed after co-administration of drugs that inhibit OCTs. So far, no inhibitory drugs have been identified that selectively inhibit drug transport by individual OCTs. In response to co-administration of drugs inhibiting OCTs with drugs that are transported by OCTs various pharmacological effects were observed, however, the effects did not allow a clear identification of the involved transporter. Below, the effects of co-administration of trimethoprim on pharmacokinetics of the antiviral drug lamuvidine and of co-administration of metformin on pharmacokinetics of the anticholinergic drug trospium are described. Recently, drug-drug interactions between the antiviral drug entecavir and the antineoplastic drug crizotinib that are both substrates of OCT2 have also been reported (Table 3) [41]. These data suggest clinical relevance of OCT2 for renal secretion of crizotinib.

4.2. Interaction between OCT mediated transport of lamuvidine and trimethoprim

The HIV-reverse transcriptase inhibitor lamuvidine is transported by OCT1, OCT2, OCT3, MATE1, and MATE2-K [2,42]. K_m values between 249 µM and 2140 µM were determined for lamuvidine transport by OCT1, OCT2, and OCT3 [75]. After oral administration, lamuvidine is rapidly absorbed, exhibiting a bioavailability of more than 80% [76]. The predominant route of elimination is urinary excretion that mainly occurs via secretion in proximal tubules [76]. Moore and coworkers investigated the effect of co-administration of the antibiotic trimethoprim (combined with sulfisoxazole that does not interact with OCTs) and lamuvidine to HIV-seropositive individuals on the pharmacokinetics of lamuvidine [76]. In patients receiving lamuvidine alone, C_max of lamuvidine in systemic blood was reached about two hours after oral application, after another two hours it was reduced by half. Upon co-administration of trimethoprim and lamuvidine, the area under the concentration time curve (AUC) of lamuvidine was increased by 43% and the renal clearance was decreased by 35%. The data suggest that trimethoprim increases lamuvidine exposition by decreasing urinary excretion. Considering the estimated C_{max unb} of trimethoprim, the IC₅₀ values for inhibition of lamuvidine uptake via OCT2 and MATE2-K by trimethoprim, and the IC₅₀ values for trimethoprim inhibition of OCT1 and MATE1-mediated uptake of other substrates (Table 3), it is probable that renal secretion of lamuvidine was decreased by drug-dug interaction at MATE2–K [2,42].

4.3. Interaction between OCT mediated transport of trospium and metformin

The anticholinergic drug trospium is used for treating overactive bladder syndrome. Trospium is positively charged and does not permeate the plasma membrane passively. Only about 10% of orally administered trospium is absorbed [44]. 70-80% of trospium entering the blood is eliminated with the urine while the remainder is excreted with the bile. In kidney, trospium is ultrafiltrated and secreted in proximal tubules. Trospium transport through OCT1, OCT2, OCT3, MATE1, and MATE2-K has been demonstrated and K_m values between 0.6 µM and 106 µM have been reported (Table 2). The lowest K_m values were determined for OCT1. Effects of metformin on pharmacokinetics of trospium were investigated [46]. Oral co-administration of metformin and trospium to healthy individuals led to a 30% reduction of C_max of trospium in the blood. This suggests a decrease in small intestinal trospium absorption due to inhibition of trospium uptake by OCT1, OCT3, OCTN1, and/or PMAT in the small intestinal BBM (Figure 1). It cannot be distinguished which transporter(s) is (are) involved because all these transporters accept metformin as substrate and may also transport trospium (Table 2). It has been shown that trospium is transported by OCT1 and OCT3, whereas OCTN1 and PMAT have not been tested for transport.

5. Anticipated dug-drug interactions at OCTs in brain

In rodents OCT2, OCT3 and PMAT were localized in cerebral neurons. They were shown to play important roles in complex cerebral functions and to influence treatment with psychoactive drugs [2,15,77,78]. Similar functions and impact for treatment with psychoactive drugs in humans are assumed. During treatment with inhibitors of the Na⁺ dependent high-affinity neurotransmitter transporters NET, SERT, and/or DAT, the OCTs are supposed to play a major role in the reuptake of neurotransmitters after neuronal activation. Drugs that inhibit neurotransmitter reuptake by NET, SERT, and/or DAT may also inhibit low-affinity neurotransmitter uptake by OCT2, OCT3, and PMAT or may be transported by these OCTs. Transport and inhibition of OCTs may be altered by drug-drug interactions. For example, OCT2, OCT3, and PMAT transport NE, DA, and 5-HT, and are inhibited by the antidepressants, citalopram, fluoxetine, amitriptyline, imipramine, and desipramine [2,4,15]. The dopamine receptor antagonist sulpiride used for treatment of depression and psychotic disorders is also transported by OCT2 and OCT3 [2,15]. Drug-drug interactions in OCT2 and OCT3 are anticipated in patients that are in therapy with psychoactive drugs when they receive additional drugs interacting with OCTs that enter the brain. Metformin, fenoterol, lamuvidine, and trospium may be examples. This issue as well as drug-drug interactions in OCTs located in blood–brain barrier deserve special attention in future clinical studies.

6. Data on substrate recognition by OCT1 and OCT2 that may be paradigmatic for the other OCTs

Trying to understand the molecular mechanisms underlying polyspecific recognition of cations with different molecular structures and their translocation by OCTs, extensive mutagenesis has been performed in rat Oct1 (rOct1) and rat Oct2 (rOct2) [2,79–83]. The generated mutants were functionally characterized by studying effects on transport after expression in oocytes of Xenopus laevis or in epithelial cells, effects on ligand induced fluorescence changes of fluorescence labeled rOct1 variants, and effects of ligand binding on rOct1 reconstituted in nano discs. The data were interpreted using homology models that were generated based on crystallized structures of transporters of the same transporter superfamily. In recent reviews these data were discussed in detail [2,79]. The obtained insights and hypotheses can be summarized as follows. Both rOct1 and rOct2 contain a large binding cleft that may be oriented to the extracellular or intracellular side of the plasma membrane. Binding of a cationic substrate to the inner part of the outside-oriented cleft induces structural changes. These changes include a state in which the substrate is occluded and ends up with an inwardly oriented cleft from which the substrate can be released into the intracellular compartment. The recognition of substrates and inhibitors with different molecular structures within the outwardly oriented cleft is enabled by the existence of different overlapping substrate-binding regions within the cleft. Ligand binding to regions within the inner part of the cleft may trigger translocation. Since more than one ligand can bind at the same time to binding regions within the cleft, different ligands may exhibit competitive replacement or short-distance allosteric interactions. The situation may be further complicated because two organic cationic substrates and small ions may be transported at the same time [82,84]. The interposition of the occluded state within the transport process excludes a substantial leak of inorganic ions during transport. In addition to ligand-binding sites with relatively low affinities in the micro- to millimolar range that are supposed to be located within the inner part of the binding cleft, high-affinity binding sites for organic cations with affinities in the nanomolar range have been identified [81]. The high-affinity binding sites are supposed to be located within the outer region of the outward-oriented cleft. At very low substrate concentrations, one organic cation is transported per transport cycle and the transport may be inhibited by binding of another cation to a high-affinity site [82]. This explains why high-affinity inhibition by organic cations was observed when transport was measured at very low substrate concentrations [75,85]. At substrate concentrations within the range of the respective K_m value, two substrates may be transported together as has been shown for MPP [79,82]. In this situation, the affinity of substrate binding that normally determines the K_m for translocation is probably not influenced by cation binding to the high-affinity site. In this case, only cations are supposed to interact that bind within the inner part of the outward-oriented cleft. Organic cation transport may be inhibited by competition, partial competition, and/or allosteric effects between inhibitory cation and transported cation [62,86,87]. In addition, allosteric effects between transported substrate and inhibitory cation may change the affinity of the inhibitory cation [62,88,89]. These complex and only partially understood interactions are why IC₅₀ values for rOct1 and rOct2 and the human transporters OCT1, OCT2, OCT3, OCTN1, OCTN2, MATE1, and MATE2–K were largely different when transport was measured on different substrates and when different substrate concentrations far below the respective K_m values were used for uptake measurements [2,22,62,88–91].

7. Proposal for advanced in vitro testing of NMEs for interactions with OCTs and for drug-drug interactions at OCTs

7.1. Introduction

Early in drug development, it is important to know, which plasma membrane transporters translocate and/or are inhibited by new molecular entities (NMEs) to anticipate oral bioavailability, pharmacokinetics, organ distribution, routes of excretion, adverse effects, and potential drug-drug interactions. Cloning of the OCTs and their expression in cultured cells makes it possible to study interactions of NMEs with individual transporters in vitro. The preclinical in vitro identification of targeted transporters helps to select NMEs for clinical drug development, to analyze toxicity studies in animals, and to plan phase 1 clinical trials. A more specific in vitro characterization carried out later and guided by knowledge of pharmacokinetics may help to identify clinically relevant drug-drug interactions.

Because no crystal structures of OCTs are available, reliable in silico prediction of drug interaction cannot be performed and pharmacophore modeling is of limited value. For example, by pharmacophore modeling of drugs for interaction with OCT1, OCT2, and MATE1, less than 85% of the drugs that inhibited the transporters in vitro were predicted [62,86,87,92,93]. Hence, in vitro testing is obligatory to identify drugs that interact with OCTs.

The following issues of in vitro testing will be addressed. First, it is proposed which OCT should be tested in vitro. Second, it is suggested which measurements should be performed with each transporter before a phase 1 clinical trial is started. This includes detailed proposals for experimental procedures and minimal requirements for test conditions. Third, it is proposed to perform additional in vitro measurements after data from the phase 1 clinical trial have become available. These data provide information about drug bioavailability after oral application, drug concentrations in blood, drug metabolism, and drug excretion. Finally, we emphasize the benefits to perform additional in vitro measurements to anticipate potential drug-drug interactions that are not in the current focus of attention such as drug-drug interactions in the brain.

7.2. Proposal for advanced in vitro testing procedures

During drug development, the FDA, EMA, and Japan’s PMDA recommended preclinical in vitro testing of NMEs for inhibition of uptake of model cations by OCT1, OCT2, MATE1, and MATE2-K in accordance with suggestions of the international transporter consortium [94,95]. In this review, in vitro testing is also suggested for OCT3 and PMAT for the following reasons. First, OCT3 and PMAT are supposed to have important pathophysiological functions. OCT3 has been localized in the sinusoidal membrane of hepatocytes where it is probably relevant for hepatic uptake of individual drugs. Expression of OCT3 and PMAT has been observed in the BBM of enterocytes where they are involved in absorption of cationic drugs, in the blood-brain barrier where they may be involved in drug uptake into brain, and in cerebral neurons where they are supposed to modulate removal of neurotransmitters from the interstitial space. Second, the current investigations of drug-drug interactions in OCTs have been focused on OCT1, OCT2, MATE1, and MATE2-K so that interactions in OCT3 and PMAT may have been missed. Third, preclinical in vitro testing can be easily performed with two more transporters using established experimental tools and simple protocols. Finally, it is important for the benefit of the patients to include transporters in in vitro screening that are most probably critically involved in pharmacokinetics and therapeutic efficacies of drugs rather than waiting for future coincidental detection of their clinical relevance for drug transport.

To determine with a great degree of security whether NMEs interact with OCT1, OCT2, OCT3, PMAT, MATE1, and/or MATE2-K, the first step should be to investigate whether transport of the positively charged model substrate MPP can be inhibited by the investigated NMEs. MPP is proposed because it is commercially available with very high specific radioactivity and reveals good signals for transport of all OCTs. However, since several cations inhibit OCT mediated transport of some drugs like metformin with higher affinities than MPP [2,62,88] it is recommended to test NMEs that do inhibit MPP uptake also for inhibition of metformin uptake. To ensure that inhibition via high-affinity binding sites can be detected, it is recommended to employ MPP and metformin concentrations for uptake measurements that are at least a hundred times smaller than the K_m values of the tested transporter (Table 5). To exceed the anticipated-free plasma concentrations of investigated NMEs and to get a hint on their affinity, it is suggested to employ NME concentrations of 1 µM, 10 µM, and 100 µM (Table 5). Recent data showed that for inhibition of metformin transport by OCT1, OCT2, MATE1, and MATE2-K with various drugs, lower IC₅₀ values were obtained when the cells were preincubated with perpetrator drugs [97]. For this reason, a 30 min preincubation with the NMEs is recommended.

Table 5. Proposal for preclinical in vitro testing of NMEs for inhibition of OCTs

Transporter	Experimental conditions for identification of inhibitors		Determination of IC50 values of NMEs for inhibition of drug transport
	Substrate (concentration, µM)	NME (concentrations, µM)	Drug substrate (concentration, µM)
OCT1	MPP (0.15) Metformin (1)	1, 10, 100	Metformin (3), tropisetron (0.02), sumatriptan (0.06), fenoterol (0.0002)
OCT2	MPP (0.03) Metformin (1)	1, 10, 100	Metformin (1.6), cisplatin (0.3), verapramil (0.06), entecavir (0.001), lamuvidine (0.6), trospium (0.4), cimetidine (0.4), varenicline (0.003)
MATE1	MPP (0.04) Metformin (1)	1, 10, 100	Cisplatin (0.3), entecavir (0.001), trospium (0.4), cimetidine (0.4), trimethoprim (1.7)
MATE2-K	MPP (0.03) Metformin (1)	1, 10, 100	Cisplatin (0.3), entecavir (0.001), trospium (0.4), cimetidine (0.4), trimethoprim (1.7)
OCT3	MPP (0.5) Metformin (1)	1, 10, 100	Berberin (0.005), trospium (0.4)
PMAT	MPP (0.3) Metformin (1)	1, 10, 100

It is proposed to test NMEs for inhibition of MPP transport by the OCTs. NMEs that do not inhibit MPP uptake should be also tested for inhibition of metformin transport. The proposed concentrations of MPP and metformin for identification of inhibitory NMEs represent at least one hundredth of the lowest K_m value reported for uptake of MPP or metformin by the respective transporter [2,96]. If NMEs inhibit transport of MPP and/or metformin their IC₅₀ values for inhibition of approved drugs with clinically relevant transport by OCTs should be determined. The drug concentrations proposed for testing represent about 20% (OCT1) or 10% (OCT2, MATE1, MATE2-K, OCT3) of the estimated C_{max unb} in the systemic blood. Smaller drug concentrations can be used whereas higher concentrations involve the risk to miss clinically relevant high affinity inhibition of the target drug.

For NMEs that inhibit MPP uptake by an OCT, IC₅₀ values should be determined for transport inhibition of drugs for which clinically relevant transport has been demonstrated or suggested. For these measurements, drug concentrations should be used that are in the lower concentration range of the respective drug in systemic blood. Together with the maximal plasma concentration of unbound NMEs in systemic blood (C_{max unb}) determined in phase 1 clinical trials, these IC₅₀ values allow for a more reliable decision concerning the potential clinical relevance of interaction of NMEs with co-administered drugs than using IC₅₀ values determined by inhibition of model cations. Based on IC₅₀ values determined for inhibition of model substrates, the EMA recommended a C_{max unb}/IC₅₀ of greater than 0.02 for OCT2, MATE1, and MATE2-K and a C_{max unb}/IC₅₀ of greater than 0.04 for OCT1. The FDA gave the same recommendation for MATE1 and MATE2-K but recommended a C_{max unb}/IC₅₀ of greater than 0.1 for OCT2. Based on IC₅₀ values determined for individual drugs a C_{max unb}/IC₅₀ of greater than 0.04 is recommended for OCT2, MATE1, MATE2-K, OCT3, and PMAT and a C_{max unb}/IC₅₀ of greater than 0.08 for OCT1.

If a significant inhibition of OCT-mediated transport of MPP by an NME is observed, it should be obligatory to determine whether the respective NME is transported. The reason is that OCT-mediated transport can be clinically relevant for absorption, organ distribution, and excretion and that these functions can be blunted or abolished by co-administered drugs. Notably, transport of an NME can be clinically relevant even if the free NME concentration in blood is much lower than the respective K_m value. For example, although the blood concentration of metformin in the portal vein after a therapeutic dose is about 30 times lower than the K_m value for metformin uptake by OCT1 metformin, transport by OCT1 in the liver is clinically relevant [2,98]. To come to an unambiguous decision on whether an NME is transported by an OCT, an NME concentration of 1 µM or below should be tested. The reason is that transport of more hydrophobic compounds by OCTs such as verapamil and dolutegravir may be hidden behind passive membrane permeation that may predominate at higher substrate concentrations, especially if the employed substrate concentration is much higher than the K_m value of the transporter.

To get a better idea of the potential clinical impact of OCT-mediated transport of an NME and the potential clinically relevant inhibition of OCT mediated NME uptake by co-administered drugs, additional in vitro studies are suggested using the information derived from the phase 1 clinical trial. Such information includes knowledge of oral availability, metabolism in liver, time course of NME plasma concentration after oral application, and renal excretion of unmodified NME. The proposed in vitro testing of NMEs is designed to anticipate drug-drug interactions in liver and kidney. Potential drug-drug interactions of orally administered drugs during absorption may not require advanced in vitro testing.

7.3. Advanced testing for interactions with OCT1

If NMEs inhibit OCT1 mediated transport of MPP, the IC₅₀ values of the NMEs for inhibition of OCT1 mediated uptake of metformin, tropisetron, sumatriptan, and fenoterol should be determined (Table 5). OCT1 mediated uptake of these drugs into hepatocytes has been shown to be clinically relevant (Table 1). Metformin induces its antidiabetic effect mainly in hepatocytes, whereas tropisetron, sumatriptan, and fenoterol are mainly metabolized in the hepatocytes. For uptake measurements, drug concentrations should be used that are in the low range of clinically relevant concentrations of unbound drugs in portal blood. Considering that the concentrations of orally administered drugs in portal blood are higher compared to those in systemic blood, it is suggested to use concentrations of 20% or less of the maximal concentration of unbound drugs in the systemic blood (C_{max unb}) (Table 5). Notably, K_m values reported for metformin, sumatriptan, and fenoterol are 100-fold, 183-fold, and 1800-fold higher than the respective C_{max unb} values indicating efficient transport at concentrations far below K_m (Table 2). The suggested drug concentrations allow the detection of inhibition via high-affinity binding sites. If the measurements were performed with substrate concentrations close to the respective K_m values, clinically relevant high-affinity inhibition of OCT1 may not be detected [75,79]. Using low drug concentrations, the measurements probably need to be performed with radioactively labeled drugs. For the inhibition experiments, NME concentrations between a hundredth of C_{max unb} and five times C_{max unb} measured in systemic blood are recommended. A 30 min preincubation period with the NMEs is suggested.

If preclinical in vitro testing reveals that an NME is transported by OCT1 and is metabolized in liver or exhibits its therapeutic effect in hepatocytes, additional in vitro tests should be performed to rule out the possibility that co-administered drugs impair its therapeutic efficacy. Drugs that interact with OCT1 [2] are recommended for inhibition of OCT1 mediated NME uptake. Various examples are presented in Table 6. Drugs that are expected to be most frequently co-administered should be tested preferentially. For the uptake measurements, an NME concentration should be used that is in the low range of the supposed concentration of unbound NME in the portal vein. An NME concentration amounting to 20% of C_{max unb} or less is suggested. Three drug concentrations may be used that cover the range of presumed concentrations of unbound drug in the portal vein, as estimated on basis of the respective C_{max unb} values. The IC₅₀ values should be determined for drugs that inhibit transported NMEs at a clinically relevant concentration.

Table 6. Proposal for in vitro testing whether NMEs that are transported by OCTs are inhibited by drugs in use

Transporter	Drugs that may be screened for inhibition of transported NMEs
OCT1	Acyclovir, amiloride, amisulpride, amitriptyline, apomorphine, berberine, chlorpromazine, cisplatin, cimetidine, citalopram, clomipramine, desipramine, diphenylhydramine, domperidone, doxepin, entecavir, famotidine, fluoxetine, haloperidol, imipramine, ketamine, lamuvidine, metformin, metoclopramide, odansetron, pefiticinib, promazine, pyrimethamine, ranitidine, rucaparib, scopolamine, sulpiride, sumatriptan, trimethoprim, trimipramine, tropisetron, trospium, verapramil
OCT2	Amiloride, amisulpride, amitriptyline, benztropine, buspirone, chlorpromazine, cimetidine, cisplatin, citaloproam, desipramine, dolutegravir, doxepin, entecavir, famotidine, fenoterol, fluoxetine, imipramine, lamuvidine, metformin, olanzapine, pefiticinib, procyclidine, propanolol, pyrimethamine, ranitidine, ranolazine, rucaparib, sulpiride, sumatriptan, trimethoprim, trimipramine, trospium, tucatinib, varenicline, verapramil
MATE1	Amiloride, cimetidine, metformin, nadolol, pentamidine, procainamide, propranol, ranitidine, sulpiride, trimethoprim, trospium
MATE2-K	Amiloride, cimetidine, metformin, nadolol, pentamidine, procainamide, propanol, ranitidine, sulpiride, trimethoprim, trospium, pentamidine
OCT3	Amisulpride, cimetidine, citalopram, desipramine, entecavir, famotidine, fenoterol, imipramine, ketamine, lamuvidine, metformin, pentamidine, phenytoin, procainamide, propranol, ranitidine, sertraline, sulpiride, trimethoprim, trospium, verapramil
PMAT	Amitriptyline, citalopram, desipramine, fluoxetine, imipramine, olanzapine, risperidone, sertraline, trimipramine

These tests should be performed after the C_{max unb} of the respective NME has been determined. Inhibition of OCT mediated transport of NMEs by the indicated drugs should be measured. For NME concentrations, 20% of C_{max unb} (OCT1) or 10% of C_{max unb} (OCT2, MATE1, MATE2-K, OCT3, PMAT) is suggested. The drug concentrations used for inhibition should cover the range of unbound drug concentrations in portal and systemic blood. It is suggested to use a tenth, a half, and two and a half times of C_{max unb}.

7.4. Advanced testing for interactions with OCT2

For NMEs that inhibit OCT2 mediated MPP uptake, IC₅₀ values should be determined for inhibition of OCT2-mediated uptake of drugs that are secreted in kidney. It is proposed to determine IC₅₀ values for uptake of metformin, cisplatin, entecavir, lamuvidine, trospium, cimetidine, and varenicline (Table 5). Drug concentrations amounting to 10% of C_{max unb} or less should be used for the measurements.

If NMEs are transported by OCT2 and predominantly eliminated by renal excretion, it should be investigated whether frequently applied drugs that also interact with OCT2 inhibit their transport. Table 6 shows drugs that may be tested.

7.5. Advanced testing for interactions with MATE1 and MATE2-K

A more detailed characterization is also suggested for NMEs that inhibit MATE1 and MATE2-K because they reduce the renal secretion of drugs and/or increase their nephrotoxicity. Determination of IC₅₀ values for inhibition of transport of metformin, cisplatin, entecavir, trospium, cimetidine, and trimethoprim is recommended (Table 5).

If NMEs are transported by MATE1 or MATE2-K and excreted to a considerable extent in kidney, it should be examined whether their transport can be decreased by drugs that enter proximal tubular cells via OCT2 and interact with MATE1 or MATE2-K. Table 6 shows drugs for testing.

7.6. Advanced testing for interactions with OCT3

If NMEs inhibit MPP uptake by OCT3, their IC₅₀ values for inhibition of transported drugs with potential clinical relevance should be tested. OCT3 is co-expressed with OCT1 in the sinusoidal membrane of hepatocytes and OCT2 in the basolateral of renal proximal tubular cells (Figure 1). Since the plasma membrane abundance of OCT1 or OCT2 is higher compared to OCT3 [2], OCT3 mediated uptake of drugs that are also transported by OCT1 and/or OCT2, may be clinically relevant if they have a lower K_m for OCT3 than for OCT1 and/or OCT2. For example, OCT3 has lower K_m values for berberin and trospium than OCT1 [2]. The IC₅₀ values for NME inhibition of OCT3-mediated uptake of berberin and trospium should be determined using low substrate concentrations as described above (Table 5).

NMEs that are transported by OCT3 may be clinically relevant if they exhibit therapeutic effects in hepatocytes and/or are metabolized in hepatocytes. The K_m values for OCT3 mediated uptake should be determined for these NMEs. Clinical relevance of hepatic uptake is expected when the NMEs are not transported by OCT1 or have much higher K_m values for transport by OCT1 versus OCT3. NMEs transported by OCT3 may be also clinically relevant if they are secreted in renal proximal tubules and are not transported or poorly transported by OCT2. Smaller transport of NMEs by OCT2 than by OCT3 is expected if the K_m values for OCT2 are much higher than OCT3.

It is recommended to measure inhibition of OCT3 mediated uptake of these NMEs by cationic drugs that interact with OCT3 [2]. Inhibition of NME uptake at clinically relevant drug concentrations would indicate drug-drug interactions with potential clinical relevance. Suggested drugs for testing are indicated in Table 6.

7.7. Advanced testing for interactions with PMAT

PMAT is expressed in human liver and kidney, however, the membrane locations have not been determined [4,96]. Advanced testing is recommended if clinical trials indicate that NMEs that are transported by PMAT exhibit their therapeutic effects in hepatocytes, are metabolized in hepatocytes, or are secreted into the urine. It should be determined, whether their transport is inhibited by clinically relevant concentrations of cationic drugs that inhibit PMAT (Table 6).

8. Call for reinforced research on functions of OCTs and related drug-drug interactions

8.1. Studies on OCTN1 and OCTN2

Since OCTN1 and OCTN2 are present in the BBM of enterocytes and renal proximal tubular cells (Figure 1), and OCTN2 is expressed in the blood-brain barrier, these transporters may also participate in small intestinal absorption, renal secretion, and renal reabsorption of drugs, and in drug uptake into brain. Various drugs are transported by OCTN1 and/or OCTN2 [2]. For example, the antiviral entecavir, the antibiotic ethambutol, and the antidepressant sulpiride are transported by OCNT1 and OCTN2, the antiepileptic gabapentin is transported by OCTN1, and the antineoplastic drug etoposide is transported by OCTN2 [2]. Because little data suggest clinically relevant drug transport by OCTN1 or OCTN2, it does not seem justified to stipulate routine in vitro screening of NMEs for interaction with OCTN1 and OCTN2 during drug development. However, one can foresee that the clinical relevance of these transporters will be detected during future investigations. Hence, it is strategically forward to also test NMEs that are excreted in urine for interaction with these transporters. It may be determined whether transport of 0.2 µM ergothioneine by OCNT1 and of 0.03 µM L-carnitine by OCTN2 can be inhibited by NME concentrations of a tenth, a half, and two and a half times their C_{max unb}. Inhibitory NMEs may be tested for transport to evaluate whether OCTN1 and OCTN2 can be involved in urinary excretion. These measurements may be performed after reconstitution of OCTN1 and OCTN2 into proteoliposomes allowing more defined experimental conditions than in cultured cells [3]. NMEs that are transported by OCTN1 and/or OCTN2 may be tested for inhibition by drugs that interact with the respective transporters [2]. If NMEs excreted in urine inhibit OCTN2 mediated uptake of L-carnitine, it should be determined whether the L-carnitine concentration in systemic blood is within the normal range to exclude the possibility of hereditary L-carnitine deficiency [2,99].

8.2. Studies on potential effects of NMEs interacting with OCT2, OCT3 and/or PMAT in brain

Drugs inhibiting SERT, NET, and/or DAT that are used for treatment of depression, anxiety, obsessive compulsive disorder, and attention deficit hyperactivity disorder may also inhibit OCTs. For example, the antidepressants citalopram, imipramine, and desipramine inhibit cation transport by OCT2, OCT3, and PMAT and the antidepressants fluoxetine and amitriptyline inhibit transport by OCT2 and PMAT [2,15]. Psychoactive drugs may also be transported by OCTs expressed in brain. They may contribute to drug uptake into the brain or to drug removal from interstitial space. For example, the antipsychotic dopamine receptor antagonists sulpiride and amisulpride, and the anesthetic drug ketamine are transported by OCT2 and OCT3. The transport functions may be blunted during co-administration of drugs such as metformin interacting with OCTs that enter the brain. This issue deserves future investigation. It includes an evaluation of which psychoactive drugs that have been identified as inhibitors of OCT2, OCT3, and/or PMAT are also transported [15].

Considering the cerebral functions of OCT2, OCT3 and PMAT, NMEs that interact with these transporters should be evaluated for potential cerebral functions and drug-drug interactions. As a first step, it should be clarified in animal experiments whether the NMEs enter the brain, and their intracerebral concentrations should be determined. For NMEs that enter the brain, IC₅₀ values for inhibition of uptake of psychoactive drugs by OCT2, OCT3, and/or PMAT should be determined. If IC₅₀ values in the relevant range of intracerebral NME concentrations observed in animal experiments are obtained, co-administration with the respective drug should be prohibited until clinical trials prove them harmless.

9. Conclusion

Advanced in vitro testing of NMEs during drug development is proposed based on increased knowledge about interaction of structurally different cations with OCTs and clinical significance of OCTs for pharmacokinetics and efficacy of drugs. It is proposed to adapt the testing protocol according to the advanced understanding and to include OCT3 and PMAT in addition to OCT1, OCT2, MATE1, and MATE2-K that are currently recommended by medical agencies. OCT3 and PMAT should be included because they participate with high probability in drug uptake into hepatocytes and/or are assumed to play a role in therapeutic effects of psychoactive drugs. The proposed testing protocol consists of three sequential stages. First, it is proposed to test all NMEs for interaction with OCTs by measuring whether they inhibit the uptake of a very low concentration of the model cation MPP. Second, it is proposed to determine whether inhibitory NMEs are also transported. Third, knowledge from the phase 1 clinical trial concerning clinically relevant NME concentrations and pharmacokinetics should be employed for further in vitro characterization of NMEs interacting with OCTs. It should be tested whether clinically relevant NME concentrations inhibit transport of clinically relevant drugs that are transported by the respective transporters. In addition, it should be determined whether drugs that may be co-administered inhibit NMEs that are transported by OCTs in relevant locations. Improved in vitro testing allows a more accurate prediction and verification of drug-drug interactions.

10. Expert opinion

In vitro testing of NMEs during drug development is of high medical relevance. It helps to select NMEs for future processing and to design meaningful investigations during animal experiments and clinical trials. The FDA, EMA, and/or Japanˋs PMDA have recommended in vitro testing of NMEs for interactions with OCT1, OCT2, MATE1, and MATE2-K. The selection of these transporters was apparently based on research providing examples for clinical relevance of these transporters or suggesting their clinical relevance during drug treatment. OCT1, OCT2, OCT3, MATE1, MATE2-K, and PMAT have overlapping, however distinctly different inhibitor and substrate specificities for various drugs. Importantly, different OCTs are often localized in identical plasma membranes. For example, OCT1, OCT3, and PMAT are present in the sinusoidal membrane of hepatocytes and OCT2 and OCT3 are present in the BLM of renal proximal tubular cells. Hence, it can be expected with high likelihood that all OCTs located in the same membrane can become clinically relevant for pharmacokinetics of individual drugs because they may transport partially different entities. Currently the clinical relevance of OCT3 and PMAT has not been demonstrated due to missing clinical data. Despite progress in investigating OCTs since 1994 when the first OCT was cloned [100], research concerning clinical functions of OCTs on pharmacokinetics of cationic drugs and on drug-drug interactions is in its infancy. Previous research on the roles of OCTs in pharmacokinetics focused mainly on a few frequently used drugs and the transporters OCT1, OCT2, MATE1, and MATE2-K, while selective inhibitors for individual OCTs were not identified. Moreover, many drugs that have been identified as inhibitors of OCTs have not been tested for transport [2]. The inclusion of OCT3 and PMAT in the in vitro screening is also emphasized because recent research strongly suggests that these transporters and OCT2 are involved in therapeutic effects of psychoactive drugs [15].

Recent data indicate that the IC₅₀ values for inhibition of OCT1, OCT2, and/or MATE1 by cationic drugs are not only highly dependent on the molecular structure of the substrate employed for uptake measurements but also on the substrate concentration. For example, it was demonstrated for OCT1 and OCT2 that much lower IC₅₀ values were obtained when the MPP concentrations used for uptake measurements were hundredfold below K_m versus about tenfold below K_m. The reason is that high-affinity binding sites can become inhibitory at very low substrate concentrations. Noteworthy, this situation is clinically relevant because the concentration of many cationic drugs in blood are orders of magnitude lower than their K_m values for transport, and clinically significant uptake has been observed at drug concentrations that are more than fifty times below the respective K_m values. For these reasons, it should be mandatory to use substrate concentration of at least a hundred times below K_m to test NMEs for inhibition of OCTs. For practical reasons it is recommended to perform the in vitro testing in consecutive steps. First, it should be determined whether NMEs interact with individual OCTs by measuring inhibition of MPP uptake using a very low MPP concentration. Inhibitory NMEs should be tested for transport because the potential clinical impact of transported NMEs is different from inhibitory ones. When the clinical phase 1 trial has been completed and data about pharmacokinetics of the NMEs are available more specific in vitro testing should be performed. When NMEs are metabolized in the liver, testing could be focused on potential drug-drug interactions during uptake into hepatocytes, whereas testing could be focused on renal secretion when NMEs are excreted in the urine. It should be determined whether NMEs interacting with OCTs in liver and/or kidney can inhibit clinically relevant transport of approved cationic drugs by the respective OCTs at clinically relevant concentrations. For NMEs that are transported by OCTs, it should be determined whether they are inhibited by cationic drugs interacting with the same transporters.

The use of very low concentrations of NMEs and of approved drugs for in vitro testing is considered inevitable. Unfortunately, this may prevent a rapid general acceptance because such measurements can likely only be performed with radioactively labeled compounds and may require highly specific labeling. Since there is no alternative to using very low substrate concentrations for transport measurements for in vitro studies, the additional technical effort must be made for the benefit of patients. It will also benefit the pharmaceutical companies in the end. Last but not least an advanced in vitro testing followed by more targeted clinical testing will increase the knowledge about pharmacokinetics of drugs and drug-drug interactions in the future. A byproduct of including more OCTs in routine in vitro testing using clinically relevant conditions will hopefully be the detection of compounds that inhibit specific OCTs in clinically relevant conditions. This would make it possible to decipher the roles of OCTs in brain functions and during treatment of patients with certain psychotropic drugs.

Article highlights

• Eight organic cation transporters (OCTs) named OCT1, OCT2, OCT3, OCTN1, OCTN2, MATE1, MATE2-K, and PMAT were cloned and characterized.

• The OCTs have an overlapping selectivity for cationic drugs and are partially co-expressed in plasma membranes of specific tissues.

• OCT1 and OCT3 are co-expressed in the sinusoidal membrane of hepatocytes where they are involved in drug uptake, whereas OCT2 and OCT3 are co-expressed in the basolateral membrane of renal proximal tubular cells where they are involved in drug secretion.

• The half maximal concentration for inhibition of OCT-mediated cation transport by cationic drugs is dependent on the structure and concentration of the transported cation.

• OCTs contain high-affinity binding sites. During drug transport at clinically relevant low drug concentrations binding of co-administered drugs to the high-affinity binding sites may promote inhibition of transport.

• An advanced protocol for in vitro testing of new drugs for interactions with OCTs is proposed in which substrate and concentration-dependent affinity of drug inhibition is considered. It is proposed to test OCT3 and PMAT in addition to the previously recommended transporters OCT1, OCT2, MATE1, and MATE2-K.

This box summarizes key points contained in the article.

Reviewer disclosures

Peer reviewers in this manuscript have no relevant financial or other relationships to disclose.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Funding

This paper was not funded.