Polyspecific Organic Cation Transporters and their Impact on Drug Intracellular Levels and Pharmacodynamics

Abstract

Most drugs are intended to act on molecular targets residing within a specific tissue or cell type. Therefore, the drug concentration within the target tissue or cells is most relevant to its pharmacological effect. Increasing evidences suggest that drug transporters not only play a significant role in governing systemic drug levels, but are also an important gate keeper for intra-tissue and intracellular drug concentrations. This review focuses on polyspecific organic cation transporters, which include the organic cation transporters 1-3 (OCT1-3), the multidrug and toxin extrusion proteins 1-2 (MATE1-2) and the plasma membrane monoamine transporter (PMAT). Following an overview of the tissue distribution, transport mechanisms, and functional characteristics of these transporters, we highlight the studies demonstrating the ability of locally expressed OCTs to impact intracellular drug concentrations and directly influence their pharmacological and toxicological activities. Specifically, OCT1-mediated metformin access to its site of action in the liver is impacted by genetic polymorphisms and chemical inhibition of OCT1. The impact of renal OCT2 and MATE1/2-K in cisplatin intrarenal accumulation and nephrotoxicity is reviewed. New data demonstrating the role of OCT3 in salivary drug accumulation and secretion is discussed. Whenever possible, the pharmacodynamic response and toxicological effects is presented and discussed in light of intra-tissue and intracellular drug exposure. Current challenges, knowledge gaps, and future research directions are discussed. Understanding the impact of transporters on intra-tissue and intracellular drug concentrations has important implications for rationale-based optimization of drug efficacy and safety.

1. Introduction

The ability of a drug molecule to move through cell membranes is a vital property affecting its pharmacokinetic and pharmacodynamics properties. Lipophilic drugs generally have high membrane permeability and their movement across cell membranes occurs primarily through passive diffusion, a non-mediated process discussed in great details elsewhere in this issue. Hydrophilic drugs, on the other hand, have low membrane permeability, and their efficient uptake into cells and tissues often involve facilitated mechanisms mediated by membrane transporters (also known as carriers). Different from passive diffusion where a drug molecule moves across membranes down its concentration gradient without energy input, carrier-mediated transport can be coupled to a cellular energy source to power uphill transport against the drug concentration gradient. Further, carrier-mediated drug transport is saturable, inhibitable, and highly dependent on the functional characteristics of the membrane transporters expressed in the specific tissues or cell types. In mammalian cells, there are two major types of membrane proteins involved in drug and solute transport: the solute carrier (SLC) and the ATP-binding cassette (ABC) transporters. The past two decades have witnessed an explosion of knowledge in our understandings of the basic biology and pharmacology of various SLC and ABC drug transporters. The in vivo roles of these transporters in drug disposition, efficacy, and toxicity are increasingly being appreciated. The clinical significance of transporters as a site of drug-drug interaction and a source for interindividual variability in drug response is also begining to be acknowledged [1–3].

Most drugs are intended to act on targets residing within a specific tissue or cells. While some drugs bind to external cell surface targets (e.g. G protein-coupled receptors), others act on intracellular enzymes and receptors residing inside the cell. Thus, it is the unbound drug concentration within the target tissue or cells that is directly responsible for eliciting its pharmacological effect. However, in the clinical setting, direct measurement of drug concentrations in target tissues and cells is difficult to achieve. Measurement of blood or plasma drug concentrations is thus commonly used to establish pharmacokinetic–pharmacodynamic relationships. For drugs that rapidly cross membranes by passive diffusion, plasma concentration is often a good surrogate for tissue concentration because the unbound drug concentration in tissue/cells is at equilibrium with its unbound concentration in plasma at steady state [4,5]. However, if a drug is transported by active uptake and/or efflux drug transporters, such a relationship may no longer exist. For drugs that are substrates of uptake transporters, tissue and/or intracellular drug concentrations can be much higher than drug concentrations in plasma. Conversely, for drugs that are substrates of efflux transporters, concentrations in tissues and cells may be substantially lower than predicted from plasma levels. Increasing evidences suggest that transporters expressed in specific tissues and cells can exert a great impact on local and intracellular drug concentrations, directly influencing their pharmacological and toxicological activities [4,5].

This review focuses on a special group of SLC drug transporters—the polyspecific organic cation transporters, which mediate cellular uptake and efflux of a broad spectrum of drugs, toxins, and endogenous compounds. We first briefly review the molecular and functional characteristics of major organic cation transporters with a special emphasis on their tissue distribution, cellular localization and transport mechanisms. We then highlight the impact of these transporters in controlling tissue and intracellular drug concentrations using literature examples where the roles of locally expressed organic cation transporters have been clearly demonstrated in several tissues (liver, kidney, salivary glands) in in vivo or clinical studies. The resulting consequence on pharmacodynamic response and toxicological effects of clinically used organic cation drugs is presented and discussed alongside. Lastly, the current challenges, knowledge gaps and future research directions in this field are briefly summarized and discussed.

2. Molecular and Functional Characteristics of Polyspecific Organic Cation Transporters

Organic cations are structurally diverse endogenous compounds (e.g. biogenic amines) and xenobiotics (e.g. drugs, environmental toxins) that carry a net positive charge at physiological pH. About 40% of the commonly prescribed drugs exist as organic cations at physiological pH [6]. Many organic cations are hydrophilic and rely on transporters to move across cell membranes. In humans and other mammals, there are a number of SLC transporters that appear to be evolved specifically to handle these structurally diverse organic cations. These polyspecific (or multispecific) organic cation transporters including the classic organic cation transporters 1-3 (OCT1-3) from the SLC22 family, the multidrug and toxin extrusion proteins 1-2 (MATE1-2) from the SLC47 family, and the plasma membrane monoamine transporter (PMAT) from the SLC29 family [7–12]. The molecular and functional characteristics of the major human polyspecific organic cation transporters are summarized below and in Table 1. A variety of clinically used drugs have been identified as the substrates of these transporters, and some selected drug substrate are listed in Table 1.

Table 1.

Molecular and Functional Characteristics of Major Human Polyspecific Organic Cation Transporters.

Transporters	Gene	Transport Mode	Tissue Distribution	Selected Drug Substrates	Reference
OCT1	SLC22A1	Electrogenic	Liver, small intestine, kidney, lung, brain, heart, skeletal muscle, placenta, mammary gland, adrenal gland, immune cells, adipose tissue	Acyclovir, atenolol, debrisoquine, furamidine, ganciclovir, lamivudine, lamotrigine, metformin, oxaliplatin, pentamidine, picoplatin, tropisetron, zalcitabine	[51,57,136–138]
OCT2	SLC22A2	Electrogenic	Kidney, brain, lung, small intestine, thymus, placenta	Amantadine, amiloride, atenolol, cimetidine, cisplatin, famotidine, ifosfamide, lamivudine, memantine, metformin, oxaliplatin, picoplatin	[51,57,136,139]
OCT3	SLC22A3	Electrogenic	Liver, skeletal muscle, heart, placenta, brain, kidney, small intestine, urinary bladder, cornea, mammary gland, lung	Cisplatin, etilefrine, lamivudine, lidocaine, metformin, oxaliplatin, pramipexole, quinidine	[57,97,136]
MATE1	SLC47A1	Electroneutral (H+/OC+ exchange)	Liver, kidney, skeletal muscle, adrenal gland	Acyclovir, atenolol, cimetidine, cisplatin, fexofenadine, guanidine, metformin, oxaliplatin, procainamide, topotecan	[22,51,109,140,141]
MATE2/2-K	SLC47A2	Electroneutral (H+/OC+ exchange)	Kidney	Atenolol, cimetidine, cisplatin, guanidine, metformin, oxaliplatin, procainamide, topotecan	[16,22,51,140]
PMAT	SLC29A4	Electrogenic pH sensitive	Brain, heart, small intestine, kidney, liver	Metformin, ritonavir	[11,12,25,46,47]

2.1. Molecular Features of OCTs, MATEs, and PMAT

The human OCTs are encoded by the SLC22 gene family and consist of three closely-related members: OCT1 (SLC22A1), OCT2 (SLC22A2) and OCT3 (SLC22A3). hOCT1 and hOCT2 are 70% identical in protein sequence, whereas hOCT3 shares 50% sequence identity with hOCT1 and hOCT2 [13]. The OCT proteins contain 542–556 amino acids with 12 predicted α-helical transmembrane domains (TMDs) [3]. The COOH- and NH2-terminal ends of the OCTs are intracellular. One large hydrophilic loop is localized to the extracellular side between TMD1 and TMD2 and contains several N-glycosylation sites. A large intracellular loop is localized between TMD6 and TMD7 with potential protein kinase C-dependent phosphorylation sites [14].

In excretory organs, OCTs frequently team up with the multidrug and toxin extrusion (MATE) proteins to mediate transepithelial transport of organic cations [15]. Encoded by the SLC47A gene family in humans, MATEs include two members: MATE1 (SLC47A1) and MATE2 (SLC47A2) [15]. Human MATE1 has only one isoform with 570 amino acids in length, while human MATE2 has three isoforms: the full length isoform hMATE2 (602 amino acids), hMATE2-K (566 amino acids) and hMATE2-B (220 amino acids) [9,16]. Both hMATE2 and hMATE2-K are functional, whereas hMATE2-B possesses no transport activity [16]. Human MATEs are predicted to have 13 TMDs with an extracellular carboxyl terminus and an intracellular amino terminus [15,17].

Beside OCTs and MATEs, a newer polyspecific organic cation transporter, the plasma membrane monoamine transporter (PMAT), was recently cloned and characterized by our laboratory [11,12]. By gene ontology, PMAT (SLC29A4) belongs to the equilibrative nucleoside transporter (SLC29) family. However, detailed functional characterization work demonstrated that PMAT functions as a polyspecific organic cation transporter that shares similar substrate specificity and functional characteristics to the OCTs [11,12,18,19]. PMAT is predicted to have 11 TMDs with an intracellular N- and an extracellular C-terminus [11].

2.2. Driving Forces of OCTs, MATEs, and PMAT

OCTs-mediated organic cations transport is independent of the sodium and chloride ions [20,21]. OCTs functions as electrogenic, facilitative transporters, and the transport direction is dependent on the electrochemical gradient of the transported organic cations [7,22]. In animal cells, the universally existing inside-negative membrane potential is used by the OCTs to drive cellular uptake of the organic cation substrate [23]. This allows the OCTs to accumulate a substrate with intracellular concentrations much higher (up to 10 fold) than its extracellular concentration [24]. Similar to the OCTs, PMAT also functions as an electrogenic transporter that utilizes the physiological inside-negative membrane potential as a driving force [25]. The transport activity of PMAT can be further stimulated by an acidic pH [25,26]. MATEs, however, are proton/organic cation exchangers [16]. They couple a transmembrane proton gradient to drive the transport of organic cations in the opposite direction [17], a process involving an electroneutral exchange of proton for a monovalent organic cation [27].

2.3. Tissue distribution and expression of OCTs, MATEs, and PMAT

Despite the similar structure and transport function, tissue distribution of OCT1-3 varies greatly. Oct1, the first OCT isoform identified from rat kidney, is highly expressed in the kidney, liver and small intestine in rodents [28,29]. In humans, however, OCT1 is mainly found in the liver and localized to the basolateral membrane of hepatocytes [30,31]. Besides, low expression of human OCT1 is also detected in other tissues including small intestine, colon, kidney, lung, brain, heart, skeletal muscle, peripheral leukocytes, adrenal gland, mammary gland, immune cells and adipose tissue [29,32,33]. Oct2 was isolated by homology screening from rat kidney and human OCT2 was also cloned later [32,34]. In humans, OCT2 is predominantly expressed on the basolateral membrane of renal proximal tubule cells in the kidney, and low expression has also been reported in brain, lung, small intestine, thymus, placenta and the inner ear [29,32]. Oct3 was independently cloned from rat brain and placenta [35,36], while human OCT3 was cloned from Caki-1 cells and originally identified as an extraneuronal monoamine transporter [37]. Different from hOCT1 and hOCT2, hOCT3 has a broader tissue distribution with relatively high expression in skeletal muscle, placenta, salivary glands, heart, brain, adrenal gland, trachea, small intestine, and uterus [29,35,38–40]. The cellular localization of OCT3 is also tissue-specific. For instance, it is expressed on the basolateral membrane of hepatocytes and placental epithelium [29,41], but in the lung, it is localized to the luminal membrane of bronchial epithelial cells [42]. In salivary glands, the OCT3 protein is localized at both basolateral and apical membranes of the secretory epithelial cells [43].

Human MATE1 was first cloned and characterized as an efflux transporter of organic cations [9]. hMATE1 is highly expressed in the liver, kidney, adrenal gland and skeletal muscle [9], and it is localized to the apical membrane of renal proximal tubule cells and hepatocytes [44]. In humans, MATE2 and MATE2-K are mainly expressed in the kidney, even though they are also detectable in various tissues [16,45]. In the kidney, MATE2 and MATE2-K are also restricted to the apical membrane of renal proximal tubule cells [45].

PMAT mRNA is most strongly expressed in the brain; but lower levels of expression are also found in heart, small intestine, kidney, and liver [11,46,47]. In the brain, PMAT transcripts are widely distributed in many brain regions and are particularly abundant in brain cortex, hippocampus, cerebellum and epithelial cells of the choroid plexus [11,48]. The PMAT protein is localized to the apical membranes of the enterocytes in the intestine and epithelial cells of the choroid plexus [19,49].

2.4. Models of Organic Cation Transport across Excretory Epithelia

In secretory organs such as the kidney and liver, the OCTs and MATEs form a functional alliance to mediate organic cation secretion from the body (Figure 1). For instance, renal secretion of organic cations consists of two steps. Circulating organic cations in blood are first transported into the renal proximal tubule cells by the basolateral OCT2 driven by negative membrane potential [50]. Once inside the tubular cells, organic cations are effluxed into urine by the MATE1 and MATE2-K on the apical membrane (Figure 1A) [51]. The physiological pH of urine is slightly acidic ( pH 6.0–6.8), which provides an inwardly-directed proton gradient that can efficiently drive MATE-mediated organic cation efflux [52,53]. A similar model has been proposed for organic cation transport in hepatocytes. Located on the basolateral (sinusoidal) membrane of hepatocytes, OCT1, the most abundant OCT isoform in human liver, plays a pivotal role in the uptake of organic cations from blood into the liver [54]. Once inside the cells, organic cations may be further secreted into the bile by MATE1 on the apical (canalicular) membrane (Figure 1B) [9]. However, the canalicular pH is 7.2 or higher [55]. Without an inwardly directed proton gradient, it is unclear if MATE1-mediated efflux of organic cations across canalicular membrane is likely to occur efficiently or rely on some other transporters at the canaliculi.

Figure 1.

Models of organic cation transport in liver hepatocytes (A) and kidney proximal tubule cells (B).

3. Impact of OCTs and MATEs on Intracellular Levels, Pharmacodynamics, and Toxicity

3.1. Impact of OCT1/Oct1 on Hepatic Drug Levels and Action

Located in the sinusoidal membrane of hepatocytes, OCT1 has been identified as a main organic cation transporter in the liver and is responsible for the uptake of basic compounds in hepatocytes [3,29,56,57]. Although OCT3 and MATE1 are also expressed in hepatocytes, their roles in hepatic drug disposition and elimination have not been well established as compared to OCT1 [29,56,58,59]. MATE1 appears to mediate some biliary excretion but its activity with in vivo probes is generally low and appears to be significantly less than in the kidney [59–62]. The apparent low activity of MATE1 was thought to be due to the lack of a significant pH gradient between hepatocytes and bile [63], which is necessary to drive MATE1-mediated organic cation efflux as a proton/organic cation antiporter [59]. The importance of OCT1 in influencing liver intracellular concentrations and pharmacodynamics can be best appreciated with studies on the antidiabetic drug, metformin, and antiviral drugs such as lamivudine [64–70].

Metformin is a first line oral antihyperglycemic agent used for the treatment of type II diabetes [71]. Carrier-mediated transport across cell membranes is particularly important for metformin because it is a hydrophilic (LogP=-1.43) drug with very low passive permeability [60]. In vitro and preclinical in vivo studies have demonstrated metformin is a substrate of hepatic OCT1 as well as OCT2, OCT3, MATEs and PMAT [19,43,62,70,72]. Metformin ameliorates hyperglycemia by decreasing hepatic glucose production, reducing gastrointestinal glucose absorption, and improving peripheral sensitivity to insulin [63]. In the liver, metformin acts on intracellular AMP-activated protein kinase (AMPK) to suppress glucose production [73,74]. As the liver is a major site of metformin action, OCT1, the major gate keeper for organic cation uptake into the liver, acts as an important determinant of intracellular levels and the pharmacodynamics of metformin. Indeed, the impact of Oct1/OCT1 on hepatic metformin concentrations and therapeutic response has been well demonstrated in a series of investigations including Oct1 knockout mouse studies as well as pharmacogenetics and drug-drug interaction studies in humans [64,65,72,75,76].

Shu et al. first demonstrated OCT1 regulates hepatic metformin levels and response in vivo [64]. They showed that metformin hepatic concentration was 4.2-fold greater in wildtype Oct1 (Oct1+/+) mice than in Oct1 knockout (Oct1−/−) mice while metformin concentrations in the plasma and other organs were similar between Oct1+/+ and Oct1−/− mice. When the pharmacodynamic action of metformin was measured, hepatic AMPK and ACC phosphorylation was reduced in Oct−/− mice. Additionally, metformin was unable to reduce the fasting glucose level in Oct1−/− mice in a hyperglycemic mouse model but did reduce fasting glucose levels more than 30% in Oct1+/+ mice [64]. The role of OCT1 in metformin pharmacokinetics and pharmacodynamics in humans was then investigated in a pharmacogenetics study in healthy volunteers carrying either the normal reference OCT1 allele or variant alleles (OCT1-R61C, -G401S, -G465R, and 420del) that showed reduced metformin transport activity in vitro. Metformin’s pharmacokinetics were determined and its glucose-lowering effects were measured with an oral glucose tolerance test (OGTT) in the two groups [64,77]. Before treatment with metformin, subjects with reference and variant alleles had similar baseline plasma glucose levels and area under the glucose concentration-time curve (AUCs) after OGTT. However, after metformin treatment, volunteers carrying the OCT1 polymorphisms had significantly higher plasma glucose levels than those carrying the reference allele in OGTT. The reduced pharmacological effect of metformin in individuals with reduced OCT1 function was unlikely to be due to a change in metformin systemic exposure because metformin plasma concentration-time curve (AUC) and maximal plasma concentration (C_max) was even increased in volunteers with the reduced OCT1 function polymorphisms [77]. These results are consistent with the data from the mice, further supporting that OCT1 is a critical determinant of liver concentrations of metformin, which then directly affects the therapeutic response. More recently, Cho et al. examined the impact of OCT1 on metformin hepatic concentrations and glucose-lowering action by increasing OCT1 expression using the pregnane X receptor (PXR) agonist rifampin in humans [75]. Their data showed OCT1 mRNA levels were increased 4.1-fold in peripheral blood cells as a surrogate for hepatic induction in vivo. Metformin’s OGTT glucose AUCs were reduced by more than 50% with treatment of rifampin. Rifampin also affected renal clearance and absorption with a net effect of slightly increased systemic exposure (13%). In spite of confounding factors, this study suggests metformin intracellular hepatic concentrations and activity can be increased with induction of OCT1.

Besides metformin, OCT1 may play a similar role for organic cation drugs with a hepatic site of action. A similar concept can be further extended to non-hepatic tissues where transporter-mediated drug uptake is needed for the drug molecule to reach its intracellular target. One of these areas is the treatment of viral hepatitis where drug access to the liver is imperative for interacting with replicating virus [78–80]. A number of anti-retrovirals have been identified as interacting with OCTs [81–83]. Lamivudine (3TC), a treatment for chronic hepatitis B virus (HBV) as well as HIV infection, and its OCT mediated cellular accumulation is one of the better studied examples [67,78,84,85]. In vitro studies showed that lamivudine was a substrate of OCT1-3 [67,68], and several genetic variants of OCT1 were shown to have reduced the uptake activity for lamivudine in vitro [69].

To date, few studies have examined the role of OCT1 in mediating lamivudine uptake into the liver and its impact on HBV treatment outcome. However, ex vivo experiments showed that lamivudine uptake into peripheral blood mononuclear cells (PBMC) was mediated by OCT1 and OCT2 [67,85]. Uptake of lamivudine into PBMCs is important for the treatment of HIV, and a similar effect may be anticipated for OCT1 in hepatic disposition of lamivudine [78]. In one study, the ex vivo uptake of lamivudine into CD4 cells correlated well (r>0.80) with the expression of OCT1 and OCT2 mRNA [67]. This study further demonstrated that OCT inhibitors could reduce the uptake into cells ex vivo. Others have shown that OCT1 polymorphisms can have a large impact on lamivudine uptake in vitro [69]. The role of these polymorphisms on the clinical efficacy of lamivudine has not yet been determined.

The impact of hepatic OCT1 on liver drug disposition and action highlighted the fact that pharmacodynamic responses do not always correlate with plasma drug concentration data (i.e. pharmacokinetics). Rather, drug concentrations in target tissues are more relevant to therapeutic activity. As discussed, a decreased pharmacodynamic response of metformin was observed in Oct1−/− mice without a corresponding change in plasma exposure [64]. In contrast, this decreased pharmacodynamic response in Oct−/− mice corroborated with a substantial reduction in metformin concentrations in the liver. A decreased response to metformin was observed in volunteers carrying OCT1 polymorphisms in spite of an increased metformin systemic exposure [64,77]. These studies clearly demonstrated the ability of locally expressed uptake transporters to impact intracellular drug concentrations at the site of action, and thus directly influencing the pharmacological activity of a medication. However, unlike plasma, in vivo drug concentrations in tissues are often difficult to obtain. Development of biomarkers or use of whole body imaging approaches may offer unique opportunities to probe the impact of uptake transporters on intracellular drug levels at a specific tissue of drug action or toxicity. In addition, genetically modified animals are particularly useful in proof-of-concept studies to dissect the tissue-specific role for uptake transporters in vivo.

3.2. Impact of OCT2/Oct2 on Renal Drug Accumulation and Nephrotoxicity

In human kidneys, OCT2 is the primary blood-facing organic cation uptake transporter [29,57,86,8]. The luminal-facing MATE1 and MATE2-K work in concert with OCT2 to mediate active renal secretion of basic drugs [51,56,87,88] (Figure 1). The roles of OCT2 and the MATE transporters in renal elimination of organic cations are well established [72,89–91]. Many cationic drugs, such as metformin and atenolol, are eliminated by active renal secretion by the OCT2/MATE pathway [66,90,92]. Changes in the activity of OCT2, MATE1, or MATE2-K can impact systemic levels of renally cleared drugs. Furthermore, an imbalance between OCT2-mediated uptake and MATE-mediated efflux may result in drug accumulation in proximal tubule cells, leading to drug-induced nephrotoxicity and kidney injury. This scenario can occur clinically and is thought to underlie the differential nephrotoxicity of platinum-based anticancer agents [93].

Cisplatin is a chemotherapeutic agent used in the treatment of lung, bladder, colon, testis, and brain cancer [94]. Nephrotoxicity, primarily in proximal tubules, is a major dose limiting toxicity of cisplatin [95]. In dogs, cisplatin accumulated between 4 to 8-fold in the kidney as compared with plasma concentrations [96]. Cisplatin is an excellent OCT2 substrate; however, it is a poor substrate of either MATE1 or MATE2-K [97–100]. The in vivo role of OCT2 in cisplatin-induced nephrotoxicity has been investigated in an elegant study by Filipski et al [101]. Unlike human kidneys which primarily express OCT2, both Oct1 and Oct2 are expressed in the rodent kidney [29]. Therefore, a mouse model, in which genes encoding both Oct1 and Oct2 were deleted, was used. It was shown that deletion of Oct1 and Oct2 resulted in significantly reduced cisplatin renal accumulation and impaired urinary excretion of cisplatin without an apparent influence on plasma levels. Further, the Oct1/Oct2-deficient mice were protected from severe cisplatin-induced renal tubular damage [101]. In cancer patients receiving cisplatin treatment, a nonsynonymous single-nucleotide polymorphism (rs316019) in the SLC22A2 gene was associated with reduced cisplatin-induced nephrotoxicity [101]. These studies established a critical role of OCT2 in the renal accumulation and nephrotoxicity of cisplatin. In addition, OCT2 is expressed in hair cells of the cochlea and OCT2-mediated cisplatin accumulation may, by the same token, underline cisplatin-induced ototoxicity [102,103].

The discovery of the critical role of OCT2 in cisplatin toxicity provided a rationale for using OCT2-selective inhibitors to mitigate the debilitating side effect of cisplatin. Indeed, rodent studies suggested a protective effect of Oct inhibitors against cisplatin-induced nephrotoxicity [102,104,105]. For example, co-administration of the OCT2 inhibitor imatinib reduced nephrotoxicity as demonstrated by kidney histology and renal biomarkers [105]. However, the complex interplay between OCT2 and MATE activity plays a crucial role in the nephrotoxicity of cisplatin [93]. Many OCT inhibitors also inhibits MATEs, which may increase intracellular cisplatin accumulation and toxicity. Selective inhibition of MATE transporters with pyrimethamine or ondansetron as well as genetic ablation of Mate1 was shown to increase the nephrotoxicity of cisplatin in mice [106,107]. Cimetidine is a selective inhibitor of MATE transporters at therapeutic doses due to its differential potencies for OCT2 (IC₅₀ ~100–150 μM) and MATE transporters (IC₅₀s~1–7 μM). Cimetidine can inhibit both OCT2 and MATEs in vivo at supratherapeutic doses [91]. A human study using high doses of cimetidine to inhibit OCT2 reduced cisplatin induced nephrotoxicity as measured by effective renal plasma flow and glomerular filtration rate [108]. A more recent high dose cimetidine study also demonstrated minimal changes in human pharmacokinetics or antitumor activity of cisplatin with cimetidine co-administration [104]. Nevertheless, the risk of using chemical inhibitors as a cisplatin nephroprotectant should be carefully addressed given the opposing effect of OCT2 and MATEs in cisplatin intrarenal accumulation and toxicity.

Newer platinum-based chemotherapeutic agent such as carboplatin, oxaliplatin, and nedaplatin have reduced risks of nephrotoxicity [93,97]. Differences in transport activity by OCT2 and MATEs account, at least in part, for their reduced toxicity [93,97,105]. Carboplatin and nedaplatin are not transported by OCT2; and therefore do not accumulate as highly in the kidney [97,109]. Interestingly, oxaliplatin is transported by OCT2 and does not have the same nephrotoxic effects as cisplatin [97,109,110]. Oxaliplatin is also a good substrate of the apical transporters MATE1 and MATE2-K, which efflux oxaliplatin out of tubular cells, thus minimizing the level of tissue accumulation and toxicity [97]. The dependence of oxaliplatin to be efficiently effluxed from proximal tubule cells by the MATE transporters, however, raises a potential concern that drugs selectively inhibiting MATE1 and MATE2-K may promote a nephrotoxic effect of oxaliplatin.

3.3. Impact of OCT3/Oct3 on Drug Accumulation and Secretion in Salivary Glands

The salivary glands play an important role in oral health, nutrient digestion, and immunity to microbial infection [111]. Dysfunction of the salivary glands can lead to xerostomia and dysgeusia [112]. Saliva drug concentrations have been used for therapeutic drug monitoring and well as illicit drug testing due to the ease of access to the diagnostic fluid and other potential benefits [113–116]. The impact of efflux transporters on salivary drug accumulation has been reported, and we recently demonstrated an impact of uptake transporters, particularly OCT3, in salivary gland drug accumulation, secretion and drug-induced taste disturbance [43,117–121].

Most drugs are assumed to be secreted into saliva by passive diffusion. This process is characterized by downhill, non-mediated diffusion of drug molecules across the membranes of the salivary gland secretory epithelial cells [122]. However, passive diffusion can neither explain salivary secretion of hydrophilic drugs nor account for high drug accumulation in salivary glands. We found that OCT3 is highly expressed in human parotid, submandibular, and sublingual salivary glands [43]. Other polyspecific organic cation transporters, including OCT1, 2, MATEs, and PMAT, are minimally expressed in human or rodent salivary glands. OCT3 protein is localized at both basolateral (blood-facing) and apical (saliva-facing) membranes of salivary gland acinar cells. OCT3 appears to mediate both epithelial uptake and efflux of organic cations in the secretory cells of salivary glands, where the OCT3-mediated drug uptake is likely to be facilitated by the inside negative membrane potential and the efflux is dependent on the electrochemical potential of the substrate (Figure 2A).

Figure 2.

(A) A proposed model of OCT3-mediated metformin transport in salivary gland epithelial cells. OCT3 on the basolateral membrane of epithelial cells mediates metformin uptake from the blood into the cells. Once metformin is highly concentrated inside the cells, OCT3 on the apical membrane can facilitate efflux of metformin into the saliva. (B) Impact of Oct3 on metformin exposure, defined as area under the concentration-time curve (AUC), in plasma, salivary glands and kidney. Data were compiled from [43].

Metformin, a widely used anti-diabetic drug, is known to induce taste disturbance. Patients on metformin therapy frequently complain about a lingering metallic taste in the mouth [123,124]. When dosed to humans (oral or IV), metformin is readily detectable in the saliva [125]. We found metformin is efficiently transported by human and mouse OCT3/Oct3. When dosed to wild-type mice, metformin was actively transported into the salivary glands and achieved very high level accumulation. The overall exposure in salivary gland tissue, defined as area-under-the-concentration (AUC), is as high as in the kidney, and much higher than plasma and the liver. In contrast, active uptake and accumulation of metformin in salivary glands were substantially attenuated in Oct3−/− mice, and its salivary exposure was reduced by more than 50% in Oct3−/− mice as compared with wildtype mice (Figure 2B). Our studies demonstrated a critical role of OCT3 in the salivary glands drug accumulation and secretion. The high levels of drug accumulation achieved in salivary tissue are alarming with respect to potential tissue-specific adverse effects.

The primary function of salivary glands is to secrete saliva, and about 0.75–1.5 liters of salivary fluid are secreted each day in healthy adults. Dysfunction of the salivary glands can lead to xerostomia (or dry mouth). Although xerostomia has many origins [126], excessive accumulation of foreign chemicals in salivary glands may lead to tissue toxicity and dysfunction of the salivary glands. More than 500 drugs list xerostomia as a side effect and 25% of elderly patients on polytherapy report dry mouth [127–129]. Dry mouth can have a major adverse impact on patients’ quality of life [127,130–132]. The more severe case of oral mucositis by cancer therapeutic agents can also be dose-limiting [129,131,133]. Drug-induced xerostomia and oral mucositis are severe side effects that may result from salivary gland drug accumulation and toxicity. Our in vivo studies revealed that Oct3-mediated active uptake can lead to very high drug accumulation in salivary glands. Oct3-mediated drug uptake and accumulation may thus intensify drug toxicity in salivary gland epithelial cells. As OCT3/Oct3 is a polyspecific transporter, it can also transport other circulating drugs or toxins into salivary glands. Therefore, it is possible that Oct3-mediated or other carrier-mediated drug accumulation may interfere with the normal secretory function of salivary glands, contributing to hyposalivation and xerostomia. In this regard, it is interesting to note that oxaliplatin, a known substrate of OCT3, can cause severe oral mucositis [97,134]. Several other cytotoxic cancer chemotherapeutics including irinotecan, vincristine, and melphalan have been reported to have increased sensitivity in cells expressing OCT3 [135]. Whether OCT3 plays a role in salivary gland accumulation and toxicity of these anticancer drugs has yet to be investigated.

4. Conclusions

In the past two decades, great progress has been made in molecular and functional characterization of drug transporters and understanding their roles in drug disposition and response. It is now becoming increasingly recognized that locally expressed transporters can exert a large impact on tissue and intracellular drug levels, directly influencing their pharmacological and toxicological activities. Drug concentrations in target tissues do not always correlate with plasma drug concentrations when transporters are involved. As exemplified with the OCTs in this review, a change in transporter activity either through drug-drug interactions or genetic polymorphisms, can substantially alter tissue and intracellular drug concentrations in target organs without significantly changing systemic drug exposure. This review highlights the importance of locally expressed organic cation transporters in controlling tissue drug concentrations and pharmacodynamics, similar results have been reported or anticipated for other drug uptake or efflux transporters [4,5]. As in vivo drug levels in tissues are often difficult to obtain, this presents a challenge in predicting in vivo drug effects that depend on the local drug concentration at the site of action. Thus it is critically important to understand the impact of drug transporters on the modulation of tissue and intracellular drug concentrations in vivo.

Understanding the impact of transporters on the modulation of tissue intracellular drug concentrations in vivo is still a challenging area due to significant difficulties and knowledge gaps in the field. First, drug distribution into tissues is a complex process that can be affected by many factors including tissue blood flow, passive permeability, tissue binding and sequestration, as well as expression of compensatory transport mechanisms. To definitively establish the impact of a specific transporter in vivo, these confounding factors should be considered. In this regard, genetically engineered animal models have been proven to be particularly useful, although species differences in transporter function and expression are still of concerns. As shown with the OCTs, pharmacogenetics and clinical drug-drug interactions studies are currently the major approaches to infer the in vivo roles of transporters in humans. However, the lack of specific drug probes and the difficulty to directly measure drug concentrations in target tissues and cells make it challenging to correlate pharmacodynamics effect with the tissue-specific roles of the transporter. This can be further complicated if the transporter is expressed at multiple tissues especially in a drug elimination organ (e.g. kidney, liver). For example, a reduction in OCT-mediated liver uptake can be compensated by an elevated drug concentration in the plasma due to a simultaneously impaired renal clearance, which may result in no change in drug exposure in the target tissue [72]. Accordingly, the roles of transporters in both systemic exposure and local tissue distribution need to be considered. Second, the ability for a drug transporter to accumulate a drug against its concentration gradient is ultimately determined by its energy coupling mechanism or its transport mode. While this has been defined for some drug transporters, it is still unclear for many others including the hepatic OATP transporters. Clarifying the cellular energy source and the coupling modes for these transporters will help to understand and predict the rate and extent of tissue drug uptake mediated by these transporters. Third, there are few well-established methods that can be commonly used to directly measure tissue and intracellular drug concentrations in vivo [4]. New experimental approaches and technologies are clearly needed in this regard to allow direct correlation of tissue and intracellular drug levels with pharmacological response. Finally, as stated earlier, it is the unbound drug concentration at the target site that is responsible for eliciting its pharmacological effect. After transporting into cells, drugs can be further sequestered or bind to intracellular organelles. Little is currently know regarding these processes. Understanding intracellular drug transport processes represents yet another frontier in our understanding of the intra-tissue and intracellular pharmacology of therapeutics.

Abbreviations

ABC

ATP-Binding Cassette

ACC

Acetyl-CoA Carboxylase

AMP

Adenosine Monophosphate

AMPK

AMP-Activated Protein Kinase

AUC

Area Under the Concentration Time Curve

C_max

Maximum Plasma Concentration

DDI

Drug-Drug Interaction

MATE

Multidrug and Toxin Extrusion

OAT

Organic Anion Transporter

OATP

Organic Anion-transporting Polypeptide

OCT

Organic Cation Transporter

OGGT

Oral Glucose Tolerance Test

PBMC

Peripheral Blood Mononuclear Cell

PMAT

Plasma Membrane Monoamine Transporter

PXR

Pregnane X Receptor

SLC

Solute Carrier

TMD

Transmembrane Domain