Abstract
Drugs and xenobiotics bind to plasma proteins with varying degrees of affinity, and the amount of binding has a direct effect on free drug concentration and subsequent pharmacokinetics. Multiple active and facilitative transport systems regulate the excretion of anionic compounds from the blood in excretory and barrier tissues. Assumptions are made about in vivo substrate affinity and route of elimination based on data from plasma protein-free in vitro assays, particularly following expression of cloned transporters. Ochratoxin A (OTA), a fungal mycotoxin, is a high-affinity substrate for several renal secretory organic anion transporters (OATs), and literature suggests that this elimination pathway is the route of entry leading to proximal tubule-targeted toxicity. However, OTA is known to bind to several plasma proteins with a high affinity, particularly serum albumin, which may impact elimination. In this study, we have systematically examined the handling of OTA and other organic anions, estrone sulfate (ES) and methotrexate (MTX), by OATs in the presence of serum albumin. Increasing concentrations of albumin markedly reduced uptake of OTA by both Xenopus laevis oocytes expressing OATs 1, 3, and 4 and organic anion-transporting polypeptide 1. For all transporters tested, virtually all mediated OTA uptake was eliminated by an albumin concentration equivalent to 10% of that present in the blood plasma. Thus, OTA uptake is dependent on the free substrate concentration and severely limited by binding to human serum albumin. MTX and ES uptake were likewise dependent on free concentration.
The proximal tubule is the primary site for renal secretory elimination of drugs, toxins, and their metabolites from the body. In addition, the proximal tubule exhibits a very high metabolic demand since it mediates not only secretion of drugs and xenobiotics but also the reabsorption of 60 to 80% of the solute and water filtered by the glomerulus (Schnellmann, 2001). Thus, the proximal tubule is particularly susceptible to injury by drugs and toxins.
Elimination of drugs, toxins, and their metabolites is handled by several facilitative and active epithelial transport proteins expressed in the proximal tubule. Since the advent of cloning and the extensive study of these transporters in vitro, the impact of plasma protein has generally been overlooked when predicting physiological roles of transporters in the renal elimination of drugs. Nevertheless, plasma proteins are abundant, and the potential for altered transporter efficacy is substantial. For example, serum albumin, a major transport/carrier plasma protein [4% (w/v)], makes up approximately 60% of total plasma protein (Peters, 1996; Gekle, 2005). Indeed, it has long been known that to predict the efficacy of a drug in vivo, it is critical to account for plasma protein binding because it will determine the availability of free drug, its half-life, and its subsequent renal elimination (Weiner et al., 1964).
Certainly, the impact of binding will vary depending upon the affinity of plasma proteins for the drug of interest. The organic anion, p-aminohippuric acid (PAH), provides an example of a substrate impacted minimally by plasma protein binding. It binds with low affinity to albumin [binding constant (Kb) ∼ 2.3 × 103 M-1; Yoshimura et al., 1998], and its clearance (∼700 ml/min) greatly exceeds the glomerular filtration rate (GFR; ∼125 ml/min), indicating potent active renal tubular secretion (Schuster, 1985). In fact, at low doses, plasma PAH is completely cleared in a single pass through the kidney. This property makes PAH clearance a measure of renal plasma flow, an important clinical indicator of renal function (Schuster, 1985).
At the other extreme is ochratoxin A (OTA), a naturally occurring mycotoxin produced by several species of fungi (Van der Merwe et al., 1965a,b). Albumin binds OTA with a Kb of 5 × 106 M-1, i.e., more than 1000 times greater than that of PAH (Il'ichev et al., 2002), and leads to a renal clearance of only 0.11 ml/min, far lower than the GFR (Studer-Rohr et al., 2000). Consistent with these measurements, the half-life of OTA in man is 35 days (Studer-Rohr et al., 2000), and the fraction of free OTA in blood is less than 0.2% of the total (Hagelberg et al., 1989). Moreover, albumin-deficient mice show enhanced excretion of OTA, again indicating that albumin binding of OTA prolongs its systemic half-life in vivo (Kumagai, 1985). Nevertheless, data from in vivo studies (Stein et al., 1985), in vitro assays examining both isolated renal tissue (Sokol et al., 1988; Bahnemann et al., 1997; Groves et al., 1998; Welborn et al., 1998) and cloned transporters (Cha et al., 2000), suggest that OTA may be transported by basolateral organic anion transporters. Indeed, in vitro expression studies indicate that OTA is a high-affinity substrate for the basolateral organic anion transporter (OAT) isoforms 1 and 3 (rOAT1 Km, 2.1 μM; rOAT3 Km, 0.7 μM) (Kusuhara et al., 1999; Tsuda et al., 1999). Likewise, cloned organic anion transporters of both OAT and organic anion-transporting polypeptide (OATP) families expressed in the apical membranes of the proximal tubule show significant affinity for OTA, with a Km of 23 μM for human OAT4 and of 17 μM for rat Oatp1 (Kontaxi et al., 1996; Babu et al., 2002). However, these in vitro transport assays and expression studies were all conducted under albumin-free conditions. Thus, the contributions of these cloned transporters to the in vivo handling of OTA remain unclear. In this study, we have systematically examined the ability of the cloned renal transporters to mediate OTA transport under conditions more closely approximating those seen in vivo, i.e., in the presence of human serum albumin. Two in vitro models were used, Xenopus laevis oocytes expressing hOAT1, hOAT3, hOAT4, and rOatp1, and a mammalian cell line stably expressing hOAT1. OTA transport under these conditions was compared with transport of two other organic anions that bind with lower affinity to albumin, estrone sulfate (ES; Kb, 1 × 105 M-1) and methotrexate (MTX, Kb, 2.8 × 103 M-1) (Zini, 1984; Pan et al., 1985).
Materials and Methods
Materials
[3H]OTA (15 Ci/mmol), [3H]MTX (20 Ci/mmol), and [3H]ES (50 Ci/mmol) were obtained from American Radiolabeled Chemicals (St. Louis, MO). Unlabeled OTA, ES, MTX, probenecid, and fatty acid free human serum albumin (HSA) were purchased from Sigma-Aldrich (St. Louis, MO). All other chemicals were obtained from commercial suppliers and were of the highest purity available.
Methods
cRNA Synthesis. cRNA was synthesized from linearized plasmids containing hOAT1 (pcDNA3.1), hOAT3 (pcDNA3.1), hOAT4 (pcDNA3.1), and rOatp1 (pSPORT1, kindly provided by Dr. Peter Meier, University of Zurich, Switzerland) with an in vitro transcription kit (T7 mMessage mMachine; Ambion, Austin, TX). Extinction coefficients were utilized to determine cRNA concentration.
Xenopus Oocyte Uptake Assay. Stage V and VI oocytes from adult female X. laevis (Xenopus I, Ann Arbor, MI) were isolated and maintained as described previously (Sweet et al., 1997). Individual oocytes were injected with either cRNA (20-25 ng) or water. Two to three days after cRNA injection, they were separated into groups of six to eight oocytes and incubated at 18 to 22°C for 60 min in oocyte Ringer's 2 (82.5 mM NaCl, 2.5 mM KCl, 1 mM Na2HPO4, 3 mM NaOH, 1 mM CaCl2, 1 mM MgCl2, 1 mM pyruvic acid, and 5 mM HEPES, pH 7.6) containing [3H]substrate (1 μCi/ml). Following uptake, the oocytes were rinsed (×3) in ice-cold oocyte Ringer's 2 and placed into individual scintillation vials containing 300 μl of 1 M NaOH. The solution was incubated at 65°C for 20 min to lyse the oocytes, and 300 μl of 1 M HCl was added to neutralize the solution. Scintillation fluid (4 ml; Ecolume; MP Biomedicals, Irvine, CA) was added and radioactivity measured with a Packard Tri-Carb 2900TR liquid scintillation counter with external quench correlation. OTA uptake was calculated in picomoles per oocyte per hour from disintegrations per minute per oocyte and medium specific activity. Uptake data were normalized against water-injected controls and expressed as a percentage of control uptake.
MDCK Type II Cell Uptake. MDCK type II cells stably transfected with hOAT1 (Aslamkhan et al., 2003) were cultured in a humidified atmosphere of 5% CO2/95% O2 at 37°C in Eagle's modified essential medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum, 1 mM sodium pyruvate, and 200 μg/ml G418 (Invitrogen) to maintain selection. In the MDCK cell clone used in these experiments, hOAT1 is expressed on both the apical and basolateral membranes, permitting cellular uptake to be studied in cells grown on solid support (Aslamkhan et al., 2003). Cells were seeded (∼1 × 105 cells/ml) on eight-well borosilicate slide chambers (Nalge Nunc International, Naperville, IL) or 24-well plates (Corning Glassworks, Corning, NY) and cultured for 2 days. Prior to treatment, medium was aspirated, and cells were washed (×3) with phosphate-buffered saline. Cells were incubated (37°C) for 1 h with Hanks' balanced salt solution supplemented with 10 mM HEPES containing 5 μM OTA (slide chambers) or 50 nM [3H]OTA (1 μCi/ml) (plates) in the presence or absence of 0.1% (w/v) HSA. Utilizing the natural fluorescence of OTA, accumulation of 5 μM OTA was imaged using a Zeiss 510 UV laser scanning confocal microscope (Carl Zeiss GmbH, Jena, Germany). Uptake was also assessed following incubation with [3H]OTA. The cells were rinsed three times in ice-cold 0.2 M MgCl2, and 1 ml of 1 M NaOH was then added to each well. The cells were lysed overnight on a shaking platform. Samples were removed for analysis of protein content by the Bradford assay prior to the addition of 1 M HCl to neutralize the solution. Before radioactivity was measured, 4 ml of Ecolume (MP Biomedicals, Irvine, CA) was added. Radioactivity was measured with a Packard Tri-Carb 2900TR liquid scintillation counter with external quench correlation. OTA uptake was calculated in picomoles per milligram of protein per hour from disintegrations per minute per well, protein content (milligrams), and medium-specific activity. Data are expressed as a percentage of control uptake.
Statistics
All data are expressed as means ± S.D.. Significant p values are indicated (*) and were calculated by paired, two-tailed Student's t test. Statistical analyses were performed using Prism (GraphPad Software Inc., San Diego, CA).
Results
Basolateral Secretion: Organic Anion Transport. The effect of albumin on the uptake of OTA by the basolateral organic anion transporter, OAT1, was initially investigated using oocytes transfected with hOAT1. The data in Fig. 1A (modified from Dai et al., 2004) show that the uptake of 50 nM [3H]OTA was probenecid (organic anion transport inhibitor) sensitive and significantly reduced in a concentration-dependent manner by low levels of HSA. With 0.005% (w/v) HSA, OTA uptake was reduced by more than 98%. To assess the possibility that the observed inhibition of OTA uptake was characteristic only of nanomolar OTA concentrations, the experiment was repeated using a concentration of OTA, 100-fold greater, i.e., 5 μM. Figure 1B shows uptake of 5 μM OTA (4.38 ± 0.97 pmol/oocyte/h) was probenecid sensitive and also inhibited by low concentrations of HSA. With 0.1% (w/v) HSA, more than 98% of OTA uptake was blocked.
Since OTA is natively fluorescent, it is possible to directly observe the effect of albumin on uptake by MDCK type II cells stably transfected with hOAT1 (MDCKhOAT1). Diffuse cellular staining of 5 μM OTA (Fig. 2A) was seen in the absence of albumin, indicating cellular accumulation. OTA uptake was not detectable in the presence of 1% (w/v) HSA (Fig. 2B). Likewise, uptake of 50 nM [3H]OTA by MCDKhOAT1 was also nearly abolished (>95% inhibition) by 0.1% (w/v) HSA (Fig. 2C).
The effect of HSA on the transport of OTA by the second basolateral organic anion transporter, hOAT3, was also investigated in Xenopus oocytes. Control uptake of 5 μM [3H]OTA (3.61 ± 1.91 pmol/oocyte/h) was probenecid sensitive (uptake was below level of detection) and significantly reduced in a concentration-dependent manner by low levels of albumin (Fig. 3), just as seen with hOAT1. About 90% of the hOAT3-mediated OTA uptake was inhibited by 0.1% (w/v) HSA.
Reabsorption: Organic Anion Transport. The oocyte expression system confirmed that hOAT4 mediates probenecid sensitive uptake of 5 μM [3H]OTA (Fig. 4, inset). The effect of increasing albumin concentrations (Fig. 4) on OTA uptake was then examined. Similar to the data for hOAT1 and hOAT3, OTA uptake (4.21 ± 0.98 pmol/oocyte/h) by hOAT4 was significantly inhibited in a concentration-dependent manner by very low levels of albumin. Approximately 60% of uptake was inhibited by 0.05% (w/v) HSA and more than 90% by 0.1% (w/v) HSA. An investigation of uptake via rOatp1 shows that OTA uptake (Fig. 5) was probenecid sensitive and inhibited by increasing concentrations of HSA. Significant inhibition (∼60%) of OTA uptake (5.77 ± 1.59 pmol/oocyte/h) was observed at 0.05% (w/v) HSA. At 0.1% (w/v) HSA, more than 95% of control uptake was inhibited.
Other Substrates: Organic Anion Transport. The influence of albumin binding on substrate transport via the OATs in the oocyte expression system was investigated further using both methotrexate and estrone sulfate as substrates. These organic anions bind to albumin with a lower affinity than OTA; OTA (Kb ∼ 5.2 × 106 M-1) > ES (Kb ∼ 1 × 105 M-1) > MTX (Kb ∼ 2.8 × 103 M-1). [3H]ES (5 μM) uptake (20.85 ± 6.69 pmol/oocyte/h) via hOAT3 was cis-inhibited by 500 μM ES (Fig. 6A, inset) and was inhibited by increasing concentrations of HSA. [3H]MTX (25 μM) uptake (3.80 ± 1.05 pmol/oocyte/h) by hOAT1 was probenecid sensitive (Fig. 6B); however, the uptake did not appear to be significantly affected by increasing concentrations of HSA.
Discussion
The renal handling of drugs and xenobiotics and their subsequent metabolism are determined by their bioavailability, in particular their free concentration in plasma, which is highly dependent on binding to plasma proteins. Thus, plasma binding can have a critical effect on the renal handling (accumulation, metabolism, and elimination) of substrates.
Although the renal toxicant, OTA, is heavily bound to the plasma protein, albumin (>99%; Hagelberg et al., 1989), it has nevertheless been shown to accumulate in the proximal segment of the nephron. The transport of OTA is thought due to secretion by OATs and reabsorption from the glomerular ultrafiltrate by nonionic diffusion and by H+-dependent peptide and organic anion transporters (Gekle, 2005). In a recent study investigating the contributions of rabbit OAT1 and 3 to the accumulation of OTA within the proximal tubule, it was stated that both transporters can mediate the cellular accumulation of OTA (Zhang et al., 2004). However, when looking at these cloned organic anion transporters, the direct effect of plasma protein binding on substrate transport was not assessed and, to date, has been poorly addressed in the literature.
In the present study, we examined the impact of albumin substrate binding on uptake of OTA and other organic anions by individual organic anion transporters, particularly the OATs. In both the oocyte expression system and mammalian cell system, uptake of OTA via basolateral OATs was highly sensitive to free, not total, OTA concentration. At albumin concentrations nearly 10-fold lower [0.5% (w/v)] than found in vivo [4% (w/v)], the uptake of 5 μM [3H]OTA by hOAT1-and hOAT3-expressing oocytes was all but eliminated (Figs. 1 and 3). In a MDCK cell line expressing hOAT1, intracellular OTA accumulation was observed by confocal microscopy. No accumulation was found when uptake was assayed in the presence of 1% (w/v) HSA (Fig. 2B). Similarly uptake of [3H]OTA by MDCKhOAT1 cells was both probenecid-sensitive and significantly inhibited by 0.1% (w/v) HSA (Fig. 2C). With chronic environmental exposure to OTA, likely in the pico-molar to nanomolar range, and plasma concentrations of albumin in vivo at ∼4% (w/v), it is unlikely that basolateral OATs play a significant contribution to OTA accumulation within the proximal tubule. These data support evidence from in vivo data that OTA is not actively secreted (renal clearance ∼ 0.11 ml/min; Studer-Rohr et al., 2000).
If OTA accumulation in the kidney following chronic exposure to nanomolar concentrations is not due to secretion from blood via basolateral OATs, perhaps it can be explained by reabsorptive OTA accumulation from the ultrafiltrate. Following glomerular filtration of albumin bound-OTA, the proximal tubule and other segments of the nephron are known to reabsorb OTA. The mechanism of reabsorption has yet to be fully elucidated, although OTA has a high affinity for several apically expressed organic anion transporters, including hOAT4 and the organic anion-transporting polypeptide transporter, rOatp1 (Kontaxi et al., 1996; Babu et al., 2002). Although albumin levels present in the tubular ultrafiltrate are much lower than in blood plasma [0.0022-0.0032% (w/v)], albumin does appear to be more concentrated (40-fold) at the surface of the apical membrane (Gekle, 2005). With environmental exposure to OTA, likely in the nanomolar range, albumin would still bind to the free OTA and reduce reabsorptive transport. In fact, in our initial investigations of nanomolar OTA uptake by OAT1, 0.005% (w/v) albumin inhibited more than 98% of OTA accumulation (Fig. 1A). Investigation of apical transporters in the oocyte expression system (hOAT4; Fig. 4, rat Oatp1; Fig. 5) showed increasing albumin concentrations effectively reduced uptake of 5 μM[3H]OTA, just as it prevented OTA uptake by hOAT1 and 3 (Figs. 1, 2, 3). OTA is so tightly bound to albumin (unless there is acute in vivo exposure to high concentrations) that the impact of organic anion transporters to proximal tubule accumulation would be minimal at best.
To confirm that these results were due to albumin binding and not limited to just OTA, we examined the renal transport of MTX and ES, OAT substrates that bind with a lower affinity to albumin (OTA > ES > MTX). Increasing albumin concentrations to near physiological levels in oocytes expressing hOAT1 had no significant effect on [3H]MTX accumulation (Fig. 6B), yet accumulation of [3H]ES by hOAT3-expressing oocytes was significantly inhibited by albumin concentrations >0.5% (w/v) (Fig. 6A). This correlates well with ES in vivo data that report a low renal clearance (2.7 ml/min), indicative of no active secretion (Wright et al., 1978). Also for MTX, our results confirm both the in vivo and in vitro data from the literature; renal clearance of MTX (∼111 ml/min/m2) exceeds GFR, indicating active secretion (Pauley et al., 2004), which has been linked to OAT-mediated transport (Uwai et al., 2004). Therefore, the degree of binding affinity for albumin to substrate will determine whether or not a compound is actively secreted or reabsorbed by organic anion transporters. The high binding affinity of OTA to albumin greatly reduces the free concentration (less than 0.2% in blood plasma; Hagelberg et al., 1989) available for transport by organic anion transporters.
Using albumin binding constants, the amount of free substrate available to cloned transporters can be calculated. Based on such theoretical calculations, OTA (Kb, 5 × 106 M-1; Il'ichev et al., 2002) would be completely bound at 0.05% (w/v) albumin and ES (Kb, 1 × 105 M-1; Pan et al., 1985) at 0.1% (w/v). However, in our systems, inhibition was only observed at moderately higher concentrations, i.e., when albumin concentrations were >0.1% (w/v) (OTA; Figs. 1, 2, 3, 4, 5) and >1% (w/v) (ES; Fig. 6A). These data are consistent with previous demonstration that binding constants determined at low albumin concentrations may overestimate the amount of bound drugs and drug metabolites actually observed at physiological concentrations (Weisiger et al., 2001). Indeed, our in vitro transport data seem to reflect the free concentration of OTA (0.2%) and ES (3%) observed under physiological rather than theoretical conditions (Hagelberg et al., 1989; Tan et al., 2001). In contrast, for the more weakly bound substrate MTX, the calculated free fraction under physiological conditions of 47% (Zini, 1984) was reflected in our in vitro system (Fig. 6B). For OTA, it would appear that unless the on/off rate of albumin binding is very high at the site of carrier-mediated transport or some mechanism exists to displace the substrate from albumin, chronic low-level OTA exposure, subsequent renal accumulation, and toxicity are unlikely mediated via this route.
Nevertheless, chronic exposure to OTA does lead to accumulation and subsequent toxicity in the proximal tubule. In our study, we noted that there is still an albumin-insensitive, albeit small component of OTA accumulation present in the hOAT1-expressing MDCK cells (Fig. 2C). This is also present in both water-injected oocytes and nontransfected mammalian cells (data not shown). Perhaps very low levels of unbound OTA can accumulate by diffusion due to the lipophilic nature of OTA. The kidney receives approximately 20 to 25% of the resting cardiac output, and following glomerular filtration, the proximal tubule is the first segment of the nephron exposed to potential drugs and toxins present in the plasma. By virtue of its extensive salt and fluid reabsorption and the resulting high metabolic rate and oxygen demand, the proximal tubule is highly susceptible to injury (Schnellmann, 2001). A high blood flow exposing the proximal tubule to OTA, coupled with the lipophilic nature of OTA, may provide enough potential for entry by diffusion alone.
In the past, protein binding and its impact on the elimination of drugs and xenobiotics was routinely considered in analysis of the renal handling of drugs. Just over 30 years ago, Weiner (1973) drew attention to this issue, discussing the importance of protein binding in determining the route of renal elimination for organic anions. Around that time, much attention was paid to the influence of substrate-plasma protein binding on renal clearance of organic anions (Ochwadt and Pitts, 1956; Block and Burrows, 1960; Weiner et al., 1964; Weiner, 1973; Sheikh, 1976; Tanner et al., 1983; Berk et al., 1987). However, since the advent of molecular analysis and cloning of several OATs in the 1990s, this important factor has been generally overlooked, and assumptions are made about in vivo substrate affinity and route of elimination based on data from plasma protein-free in vitro assays using cloned transporters. As illustrated by the data presented above for kidney and complementary data on hepatic and intestinal transport by OATPs and multidrug resistance-associated protein 2, caution must be exercised when extrapolating in vitro data to the in vivo situation (Jacquemin et al., 1994; Kanai et al., 1996; Berger et al., 2003; Cui and Walter, 2003). Indeed, the data presented here argue strongly that renal substrate uptake is highly dependent upon not only transporter affinity but also the free concentration of drug. Thus, even though the mechanism(s) of OTA accumulation within the proximal tubule remain to be fully elucidated, the data presented here emphasize the importance of looking beyond simple in vitro assessment of transport by cloned carriers. To better understand the clinical effects of drugs in vivo, one must carefully assess factors, such as binding to plasma proteins or physiological regulation of the transporter, that may determine the effectiveness of secretion/excretion across excretory or barrier epithelia.
Acknowledgments
We thank Laura Hall, Chutima Srimaroeng, and Ramsey Walden (National Institute of Environmental Health Sciences, Research Triangle Park, NC) for technical expertise.
Footnotes
-
This work was supported by the Intramural Research Program of the National Institutes of Health and National Institute of Environmental Health Sciences.
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
doi:10.1124/jpet.105.093070.
-
ABBREVIATIONS: PAH, p-aminohippuric acid; GFR, glomerular filtration rate; OTA, ochratoxin A; OAT, organic anion transporter; r, rat; OATP, organic anion-transporting polypeptide; h, human; ES, estrone sulfate; MTX, methotrexate; HSA, human serum albumin.
- Received July 22, 2005.
- Accepted September 28, 2005.
- The American Society for Pharmacology and Experimental Therapeutics