INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

ONE STRATEGY ORGANISMS use to limit the toxic effects of xenobiotics is excretory transport. Energy-dependent transport of xenobiotics can be seen at multiple organizational levels. At the cellular level, plasma membrane transporters use the energy released from ATP splitting to drive potentially toxic chemicals out of the cell. These xenobiotic transporting ATPases are members of the ATP-binding cassette family of membrane proteins, which includes the multidrug resistance (MDR; Refs. 8, 10) transporters, also called P-glycoproteins, and the MDR-associated proteins (MRP; Ref. 5). Both types of transporter were discovered in tumor cells with a multidrug-resistant phenotype. It was subsequently shown that exposure to a surprisingly wide range of chemotherapeutic agents can result in overexpression of one or both transporters, and the resulting increase in xenobiotic export capacity is one mechanism underlying the phenomenon of multidrug resistance seen in certain tumors (1, 10, 11, 16). Such transporters are also expressed in normal tissues. For P-glycoprotein, the highest levels of expression are found in transporting epithelia, e.g., liver, intestine, and renal proximal tubule (7, 29). For MRP, liver and kidney express an epithelia-specific isoform (cMRP or MRP2; Refs. 4, 27). In epithelia, P-glycoprotein and MRP2 are inserted in the plasma membrane in a polar manner, being localized to the pole of the cell facing the compartment into which xenobiotics are being excreted, i.e., canalicular and luminal membranes. This puts these transporters in the proper location to mediate the pumping of xenobiotics into urine and bile. As with tumor cells in culture, P-glycoprotein levels in epithelia have been shown to increase in animals chronically dosed with substrate [cyclosporin A (CSA)], suggesting that rates of xenobiotic excretion would be similarly increased (15). Thus, at the cell, tissue, and organism levels, ATP-driven xenobiotic transporters can both protect against toxic chemicals and limit the usefulness of therapeutic drugs.

Recent evidence indicates that aquatic organisms possess one or more MDR-like transport mechanisms that allow them to restrict xenobiotic uptake and increase chances of survival in polluted environments (reviewed in Ref. 18). For example, Western blotting shows the presence of proteins immunologically related to P-glycoprotein in tissues from marine and freshwater invertebrates and vertebrates. Binding assays with membrane fractions show xenobiotic binding with the broad specificity characteristics of P-glycoprotein. Also, the ability of sponges, marine mussels, freshwater clams, and marine worms to limit accumulation of xenobiotics is reduced after exposure to P-glycoprotein modifiers, e.g., verapamil. Finally, several studies show induction of the multixenobiotic resistance phenotype in organisms exposed to polluted water (18). Together, these findings make a case for the presence of one or more xenobiotic excretion systems in these organisms. Such transport systems are present at some basal level in animals maintained in unpolluted environments and they appear to be upregulated on exposure to pollutants.

In contrast to mammals, where high levels of P-glycoprotein and MRP2 are found in the luminal membranes of the epithelial cells of the renal proximal tubule, intestine, and liver (4, 26, 29), we know little about the tissue distribution of xenobiotic transporters in aquatic animals. Limited data from teleost fish do indicate that P-glycoprotein and MRP2 are present in renal and hepatic tissue, where they could mediate the excretory transport of xenobiotics (12, 19, 27). However, aquatic organisms possess other excretory organs. One such organ is the rectal gland of marine elasmobranchs. This gland is a specialized, salt excretory organ composed of numerous blind-ended, branched tubules that empty into a central duct (9, 28). Rectal gland tubules contain only a single cell type, a columnar epithelial cell. The NaCl concentration of the fluid secreted by the shark rectal gland is higher than plasma, thus contributing to the ability of these hyperosmoregulators to volume and ion regulate. To date, NaCl and fluid excretion has been the only described function of the elasmobranch rectal gland. Here we use isolated rectal gland tubules and confocal microscopy to describe a second function of the shark rectal gland, xenobiotic excretion.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. Fluorescein (FL) and fluorescein methotrexate (FL-MTX) were purchased from Molecular Probes (Eugene, OR). Daunomycin, chlorodinitrobenzene (CDNB), leukotriene C₄ (LTC₄), probenecid, p-aminohippurate (PAH), tetraethylammonium (TEA) chloride, sulforhodamine 101 free acid, and verapamil were purchased from Sigma Chemical (St. Louis, MO). CSA and the fluorescent CSA derivative [N-(4-nitrobenzofurazan-7-yl)-D-Lys⁸]-cyclosporin (NBDL-CSA) were obtained from Dr. G. Fricker (27). All other chemicals were obtained from commercial sources at the highest purity available.

Animals. Male and female dogfish sharks (2-4 kg) were collected outside of Frenchman's Bay, ME, and maintained at the Mount Desert Island Biological Laboratory in large circular tanks with running seawater. After sharks were pithed, glands were removed and suspended in ice-cold elasmobranch saline (in mM: 270 NaCl, 4 KCl, 3 MgCl₂, 2.5 CaCl₂, 8 NaHCO₃, 1 KH₂PO₄, 0.5 Na₂SO₄, and 350 urea; pH adjusted to 7.8 by equilibration with 99% O₂-1% CO₂).

Tubule isolation. Tubules were isolated at 4°C using a modification of the procedure of Ecay and Valentich (6). The ends of each gland and the connective tissue were removed, and the gland was minced into 1-mm cubes using razor blades. The tissue was washed three times with 15 ml of elasmobranch saline to remove blood cells and digested in 25 ml elasmobranch saline containing 0.05 g of collagenase (Boehringer Mannheim). During digestion, the tubes were agitated and gassed with 99% O₂-1% CO₂. After 25 min, the digest was suctioned 10 times into a 25-ml pipette to break up the tubules. The mixture was allowed to settle for 30 s, then the supernatant was removed and put into a 50-ml centrifuge tube. The supernatant was centrifuged at 1,000 rpm for 45 s, and the resulting tubule pellet was resuspended in 10 ml of gassed elasmobranch saline. The tubules were stored at 4°C until used.

Confocal fluorescence microscopy. Rectal gland tubules were pipetted into a chamber (Bionique) containing 1 ml gassed elasmobranch saline, fluorescent substrates, and inhibitors. The chamber floor was a 4 × 4 cm glass coverslip to which the tubules adhered lightly and through which the tubules could be viewed by means of an inverted microscope; the chamber cover was a small petri dish. For measurements, the chamber was mounted on the stage of a Nikon Diaphot inverted microscope that was part of a Noran Confocal Microscope system. Tubules were viewed using a Nikon 20× Fluor objective (numerical aperature = 0.75). Illumination was provided by an Ar ion laser at 488 or 529 nm. The primary dichroic filter was a triple dichroic with cutoffs at 510 and 540 nm. Long-pass emission filters were 515 (488 nm illumination) and 550 (529 nm illumination) nm. Images were collected with a pinhole setting of 25 or 50 µm. Neutral density filters and reduced laser power were used to minimize photobleaching. With these settings and with photomultiplier gain adjusted so that the average pixel intensity in the lumens of control tubules was ~100, tissue autofluorescence was undetectable.

To obtain an image, dye-loaded tubules in the chamber were viewed under reduced, transmitted light illumination, and a single tubule with well-defined lumen and undamaged epithelium was selected. The plane of focus was adjusted to cut through the center of the tubular lumen. Then, in confocal fluorescence mode, 16-32 video frames were averaged (~0.5-1 s). The confocal image (512 × 512 × 8 bits) was viewed on a high-resolution monitor and saved to optical disk. Fluorescence intensities were measured from stored images using an Apple Power Macintosh 7100 computer and National Institutes of Health Image version 1.58 software as described previously (19, 20). Briefly, three to five adjacent cellular and luminal areas (at least 200-400 pixels each) were selected from each tubule. After background subtraction, the average pixel intensity for each area was calculated. The values used for that tubule were the means of all measured areas.

In the present experiments, shark rectal gland tubule fragments were incubated in medium containing a fluorescent xenobiotic, confocal fluorescence images were collected, and fluorescence distribution was analyzed to obtain an indication of dye distribution in the tissue. Data are presented as steady-state fluorescence intensity measurements made over the cellular and luminal regions of the tubules. Two caveats must be kept in mind when interpreting such measurements. First, the signal from a fluorescent probe is sensitive to environment, e.g., pH or solvent polarity. As a result, absolute calibration of dye concentration in a single region of a tissue or in a cell is difficult, and the constant relating probe fluorescence to concentration could vary from region to region. Second, the steady-state solute concentration in a tissue compartment is a function of all processes governing entry into and exit from that compartment. Changes in steady-state solute concentration indicate that one or more of those processes has been altered. Often, additional knowledge about the nature of the treatment causing the change in concentration can be used to identify the altered process.

Immunostaining. Rectal glands were removed and quick frozen in liquid nitrogen. A Leica Frigocut cryostat was used to prepare frozen sections of the gland. Sections were fixed in 20°C methanol and then processed for indirect immunofluorescence staining with primary and fluorophore-conjugated secondary antibodies. Primary antibodies used included a rabbit polyclonal anti-mammalian MRP2 (Ref. 4; a gift of Dr. Dietrich Keppler), and mouse monoclonal antibodies to fish cytokeratin (a gift of Dr. Jon Holy, University of Minnesota), and the shark Na-K-2Cl transporter (a gift of Dr. Bliss Forbush, Yale University School of Medicine). Stained sections were viewed on a Nikon Optiphot 2 epifluorescence microscope, and the images were recorded using Kodak TriX-400 35-mm film.

Statistics. Data are given as means ± SE. Means were considered to be statistically different when the probability value (P) was < 0.05 by use of the appropriate paired or unpaired t-test.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Xenobiotic transport in isolated rectal gland tubules. Light collagenase digestion of shark rectal gland fragments yielded a mixture of straight and branched tubular fragments of various lengths as well as clusters of tubule fragments that had not been separated by the collagenase (Fig. 1, A, C, E). Phase contrast microscopy showed that most of the fragments had an intact epithelium and a patent lumen. Moreover, a large percentage of the tubules in this preparation appeared to be closed at both ends, raising the possibility that once xenobiotics were transported into the lumen they would not leak out. In initial experiments, tubular fragments were exposed to fluorescent xenobiotics and cellular and luminal fluorescence was visualized using confocal microscopy. Based on previous studies with teleost renal proximal tubules, the following fluorescent substrates were tested: FL (Na-dependent organic anion system; Ref. 21), sulforhodamine 101 (Na-dependent and Na-independent organic anion systems; Ref. 19), FL-MTX (Na-independent organic anion system; Ref. 19), daunomycin (organic cation system and P-glycoprotein; Ref. 20), and NBDL-CSA (P-glycoprotein; Ref. 27). Note that in teleost renal proximal tubule some of these compounds are handled by more than one transport system.

View larger version (117K): [in this window] [in a new window]   Fig. 1.   Transmitted light (A, C, E) and corresponding confocal fluorescence (B, D, F) images of shark rectal gland tubule fragments incubated in medium with 5 µM sulforhodamine 101. Similar images were obtained for tubules incubated in medium with 1 µM fluorescein methotrexate (FL-MTX). Note that the tubule at the botttom of D appears to have an open lumen. Each white bar indicates 20 µm.

Three patterns of xenobiotic accumulation were found in rectal gland tubular fragments: 1) FL did not accumulate within the cellular or luminal compartments; 2) NBDL-CSA and daunomycin accumulated within the cells, but luminal fluorescence was always lower than cellular fluorescence; 3) sulforhodamine 101 and FL-MTX did not accumulate to high levels within rectal gland cells, but luminal fluorescence was substantially higher than cellular or medium fluorescence (Fig. 1, B, D, F). Thus rectal gland tubule fragments were able to selectively concentrate fluorescent xenobiotics in the luminal space. Of the five fluorescent substrates tested, only those handled by renal system for large organic anions showed concentrative secretion into the lumens of rectal gland tubules.

We used confocal microscopy and digital image analysis to characterize the mechanism driving sulforhodamine 101 transport from medium to tubular lumen. Sulforhodamine 101, a disulfonic acid, was chosen as test substrate rather than FL-MTX, based on resistance to photobleaching and insensitivity to changes in pH. Figure 2 shows the time course of 5 µM sulforhodamine 101 accumulation in rectal gland tubules. Cellular fluorescence intensity rose rapidly with time and reached a steady-state value that was about two-thirds of the level in the medium (in this experiment, medium fluorescence averaged 10 U). Luminal fluorescence also rose rapidly, but, at steady state, mean luminal fluorescence intensity was about eight times higher than cellular fluorescence and five times higher than in the medium. Addition of 0.1 mM KCN to the incubation medium reduced luminal fluorescence by over 70% (Fig. 2), but had little effect on cellular fluorescence.

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Time course of 5 µM sulforhodamine 101 uptake by shark rectal gland tubules. Tubules were incubated in medium with fluorescent substrate, without [control (Con)] and with 0.1 mM KCN. At the indicated times, images were acquired as described in MATERIALS AND METHODS. Each point represents the mean value for 7-10 tubules; variability is given as SE bars. Statistical comparisons: at all times, KCN significantly reduced luminal fluorescence (P < 0.01).

Figure 3 shows the effects of changing medium sulforhodamine 101 concentration on steady-state medium and cellular and luminal fluorescence. As one would expect, medium fluorescence increased linearly with sulforhodamine 101 concentration. Cellular fluorescence was also a linear function of medium sulforhodamine 101 concentration, but at each concentration cellular fluorescence was significantly lower than medium fluorescence (P < 0.05). Luminal fluorescence exceeded medium and cellular fluorescence (P < 0.01). Luminal fluorescence increased with medium sulforhodamine 101 concentration but, unlike medium and cellular fluorescence, those increases were less than proportional. That is, luminal fluorescence intensity with 5 and 10 µM sulforhodamine 101 was below the extrapolated line drawn through the origin and the 1 µM sulforhodamine 101 point (Fig. 3, dashed line). For 10 µM sulforhodamine 101, measured luminal fluorescence was 76 ± 5% of the value calculated from the extrapolated line (P < 0.01; ² test).

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Effect of increasing medium sulforhodamine 101 concentration on steady-state accumulation by rectal gland tubules. Tubules were incubated in medium with the indicated concentration of fluorescent substrate. After 30 min, images were acquired as described in MATERIALS AND METHODS. Each point represents the mean value for 11-14 tubules; variability is given as SE bars. Dashed line is the extrapolated line through the origin and the lowest sulforhodamine 101 concentration.

Having demonstrated that sulforhodamine 101 transport into rectal gland tubular lumens was concentrative, energy dependent, and saturable, we used chemicals known to be handled by specific transport systems as tools to characterize specificity limits of the mechanism involved. PAH and TEA, model substrates for the classical renal organic anion and organic cation transport systems, respectively (25), did not affect sulforhodamine 101 transport (Fig. 4). Additionally, verapamil, a substrate for both the renal organic cation transport system and P-glycoprotein did not affect sulforhodamine 101 transport either (20).

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Effects of transport inhibitors on sulforhodamine 101 and FL-MTX transport. Tubules were incubated in medium with 5 µM sulforhodamine 101 (A) or 1 µM FL-MTX (B) without (control) or with the indicated compounds. After 30 min, images were acquired as described in MATERIALS AND METHODS. Each point represents the mean value for 10-20 tubules; variability is given as SE bars. PAH, p-aminohippurate; TEA, tetraethylammonium; VERAP, verapamil. Statistical comparisons: for both sulforhodamine 101 and FL-MTX, leukotreine C4 (LTC4), and cyclosporin A (CSA) significantly reduced luminal fluorescence (P < 0.01).

In liver, secretion of large amphiphilic organic anions from hepatocyte to bile canaliculus is mediated by one or more multispecific organic anion transporting ATPases (cMOAT, also called MRP2; Refs. 17, 22, 24). MRP2 has also been localized to the luminal membrane of renal proximal tubule (26), and recent experiments suggest that it mediates the cell-to-lumen step in renal secretion of large organic anions (19). It is important to note that MRP2 is distinct from another xenobiotic transporting ATPase, P-glycoprotein, that is also present at high levels in the canalicular membrane of hepatocytes and in the luminal membrane of renal proximal tubule cells (15, 29). MRP2 is sensitive to inhibition by a variety of anionic compounds. Of these, cysteinyl leukotrienes have a particularly high affinity for the transporter, with LTC₄ exhibiting a Michaelis constant of 250 nM (13). Transport by MRP2 is also inhibited by CSA (inhibition constant 3 µM; Ref. 2) and by certain glutathione (GSH) conjugates (13, 14). To determine whether a transporter with similar specificity might be operating in rectal gland tubules, we measured the effects of LTC₄, CSA, and CDNB on sulforhodamine 101 and FL-MTX transport. For both fluorescent substrates, CSA and LTC₄ significantly reduced luminal fluorescence (Fig. 4); these compounds had no significant effects on cellular fluorescence.

For tubules incubated in medium with 5 µM sulforhodamine 101, luminal fluorescence was also reduced in a concentration-dependent manner by CDNB (Fig. 5). CDNB is an uncharged compound that enters cells by simple diffusion. Once in cells, CDNB is conjugated to GSH. The resulting DNB-GSH is both an anion and a high-affinity competitor for multispecific organic anion transporting ATPases such as cMOAT and MRP2 (13, 14, 22).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Inhibition of sulforhodamine 101 secretion by chlorodinitrobenzene (CDNB). Tubules were incubated in medium with 5 µM fluorescent substrate, without (control), and with the indicated concentration of CDNB. After 30 min, images were acquired as described in MATERIALS AND METHODS. Each point represents the mean value for 11-19 tubules; variability is given as SE bars. Statistical comparisons: CDNB significantly reduced luminal fluorescence (P < 0.05 for 5 µM CDNB; P < 0.01 for 10 and 25 µM).

Immunolocalization of MRP2 antibody. The above transport data are consistent with the luminal membrane of shark rectal gland cells, possessing a xenobiotic transport system with the substrate and inhibitor specificity characteristics of a multispecific organic anion transporter (MRP2). To determine whether such a protein was present in the luminal membrane, frozen sections of rectal gland were exposed to a polyclonal antibody to MRP2 (4, 26) and a fluorescent secondary antibody. Epifluorescence micrographs of this tissue showed that MRP2 immunostaining is predominantly associated with the apical region of the tubular epithelium (Fig. 6). This apical enrichment is clear in tubules cut in both cross section and longitudinally. Verification of this apical localization pattern for MRP2 staining comes from double-labeled sections in which the MRP2 staining clearly codistributes with the apical marker cytokeratin (Fig. 7). The MRP2 staining shown here was not a result of nonspecific labeling due to restricted antibody access, because no labeling was seen when rectal gland sections were stained with secondary antibodies alone and because an antibody against the basolateral shark Na-K-2Cl cotransporter labeled the basolateral membrane exclusively (not shown).

View larger version (85K): [in this window] [in a new window]   Fig. 6.   Immunolabeling of frozen rectal gland sections with anti-MRP2. The transporter shows a clear localization in the apical region of the tubular epithelium both in cross section (A) and longitudinal (B) profiles. Magnification ×465.

View larger version (64K): [in this window] [in a new window]   Fig. 7.   Double labeling of frozen rectal gland sections with anti-MRP2 and anticytokeratin. A: differential interference contrast image; B: anti-MRP2 staining; C: anticytokeratin staining. Magnification ×465.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Animals use a variety of adaptive strategies to survive environmental stress. For example, many marine animals possess extrarenal "salt glands" that help them to maintain osmotic and ion activity gradients between body fluids and the outside medium. In marine elasmobranchs, this function is subsumed by a rectal gland, which actively transports NaCl from blood to rectum. The osmoregulatory function of the elasmobranch rectal gland was first discovered ~40 years ago. Since then, work on the gland has focused on the physiology and cellular and molecular biology of ion transport (9, 28). To date, the only known function of the gland is salt and fluid excretion.

The results of the present confocal imaging study indicate that dogfish shark rectal gland serves a second function, also related to coping with environmental stress, xenobiotic excretion. This conclusion is based on the following observations. First, isolated rectal gland tubule fragments rapidly transported two fluorescent organic anions, sulforhodamine 101 and FL-MTX, from bath to lumen. In these tubules, sulforhodamine 101 and FL-MTX fluorescence in the lumen greatly exceeded that in the bath; with 5 µM sulforhodamine 101, the lumen-to-bath fluorescence ratio exceeded 5. Second, inhibitor studies showed that transport from bath to lumen was reduced when metabolism was inhibited by KCN or when tubules were exposed to certain xenobiotics, i.e., CSA, LTC₄, and CDNB. Third, transport from bath to lumen partially saturated when the medium sulforhodamine 101 concentration was raised from 1 to 10 µM. Together, these results indicate that rectal gland tubular fragments, like renal proximal tubules isolated from certain teleost fish (25), contain a luminal compartment that is sealed from the medium and that rectal gland tubules are capable of transport from medium to lumen of certain xenobiotics by a process that is uphill, energy dependent, specific, and saturable. In support of these findings for isolated rectal gland tubule fragments, our initial experiments with isolated, perfused rectal glands have shown that sulforhodamine 101 is transported from vascular perfusate to duct fluid by a process that is blocked by CDNB. That is, in four glands perfused with 1 µM sulforhodamine 101, the sulforhodamine 101 excretion rate was 2.8 ± 0.5 nmol · min¹ · g tissue¹; when 25 µM CDNB was added to the perfusate the rate was only 0.3 ± 0.1 nmol · min¹ · g tissue¹ (data from 3 glands; Masereeuw, Karnaky, and Miller, unpublished data).

Xenobiotic excretion across the rectal gland tubule epithelium involves transport across two membrane barriers. The present data do not yet allow us to define the mechanism responsible for the first step, cellular accumulation. Steady-state sulforhodamine 101 levels within the cells never exceeded those in the medium, and cellular accumulation of sulforhodamine 101 was not significantly reduced by any xenobiotic tested. Over the range of medium sulforhodamine 101 concentrations used, cellular fluorescence was a linear function of medium fluorescence. By these criteria, cellular accumulation of sulforhodamine 101 was not concentrative, inhibitable, or saturable. On the basis of these findings, one might conclude that cellular accumulation of sulforhodamine 101 was a result of simple diffusion. However, the situation may be more complicated. In control tubules, cellular levels of sulforhodamine 101 were lower than medium levels. Because we expect the electrical potential difference across the membrane to result in cell-to-medium concentration ratios less than unity for this organic anion, those data are consistent with diffusion across the basolateral membrane being one determinant of cellular levels. However, if the low steady-state cellular sulforhodamine 101 levels in controls were solely the result of the passive distribution of that organic anion across the basolateral membrane, we would have expected cellular levels to have increased when metabolism was inhibited by KCN and the cells were presumably depolarized. No such increase was seen. Additional experiments are needed to better understand the mechanisms that contribute to cellular accumulation of sulforhodamine 101 in these tubules.

In contrast to uptake at the basolateral membrane, sulforhodamine 101 transport across the second barrier, the luminal membrane, had all the hallmarks of an active, carrier-mediated process. At all medium sulforhodamine 101 concentrations tested (1-10 µM), steady-state luminal fluorescence was many times higher than medium or cellular fluorescence. Although luminal fluorescence did increase with increasing medium sulforhodamine 101, such increases were less than proportional, indicating partial saturation of a luminal carrier. In addition, luminal accumulation of sulforhodamine 101 was inhibited by the metabolic poison, KCN, and by other xenobiotics. By these criteria, transport from cell to lumen was concentrative, saturable, energy dependent, and specific.

Note that none of the compounds that inhibited sulforhodamine 101 transport from cell to lumen appreciably affects cellular sulforhodamine 101 accumulation (Fig. 5). This observation indicates that steady-state cellular levels of sulforhodamine 101 are set independent of events at the luminal membrane. A similar insensitivity of cellular xenobiotic accumulation to events at the luminal membrane was previously observed in studies of P-glycoprotein- and MRP2-mediated xenobiotic transport in isolated renal proximal tubules (19-21, 27).

Two lines of evidence implicate a shark form of MRP2 in sulforhodamine 101 and FL-MTX transport from rectal gland cell to lumen. First, transport from cell to lumen was reduced by CSA and LTC₄, both potent inhibitors of renal and hepatic MRP2 (2, 27). CDNB, a compound that is conjugated to glutathione within cells, inhibited sulforhodamine 101 secretion in rectal gland tubules and isolated perfused rectal glands. GSH-S conjugates are other potent inhibitors of MRP2 (13, 14). Competitors for the renal organic anion transport system (PAH), the renal organic cation transport system (TEA and verapamil), and P-glycoprotein (verapamil) were without effect. In addition, smaller fluorescent organic anions (fluorescein) and fluorescent substrates for the renal organic cation transport system (daunomycin) and for P-glycoprotein (daunomycin and NBDL-CSA) were not transported into rectal gland tubule lumens. Second, using an antibody to mammalian MRP2, a fluorescent secondary antibody, and frozen sections of rectal gland tissue, we found heavy and specific staining associated with the luminal membrane of the epithelial cells. On the basis of the observed inhibition profile and the clear cut immunolocalization of an MRP2 antibody at the luminal membrane, we conclude that a shark rectal gland form of MRP2 is responsible for the active secretion of sulforhodamine 101 and probably FL-MTX.

Perspectives

Second, the present results suggest that elasmobranch rectal gland could become a useful comparative model for the study of the function and regulation of MRP2 in epithelia. The tissue is comprised of a single cell type and, as with tissue from other poikilotherms, both intact glands and isolated tubules function for long periods when kept in a simple physiological saline solution (9). With the use of fluorescent and radiolabeled substrates, xenobiotic transport can be studied at the intact organ, isolated tubule, and cellular and plasma membrane levels. Importantly, the present results indicate that MRP2 is the only xenobiotic transporter expressed in rectal gland tubules. Finally, because this tissue is already a well-established model for the study of the cellular and molecular biology of chloride transport (9, 27), much is known about responses to hormones and the intracellular signaling systems involved.