Human autosomal recessive osteopetrosis is a severe disease characterized by an increase of bone density. The murine mutation osteosclerosis (oc) has a similar phenotype to the human disease, and the reduced in osteosclerosis transporter (Roct) and osteoclast-specific vacuolar proton ATPase (V-ATPase) subunit genes were identified as being located in the oc locus and related to the oc phenotype (Brady et al., 1999; Scimeca et al., 2000).

Expression of Roct protein is reduced in the kidney of mice homozygous for oc mutation (Brady et al., 1999), although the involvement of Roct in osteopetrosis is still unknown. V-ATPase is involved in the formation of an acidic extracellular compartment for solubilization of bone minerals by osteoclasts, and its 1.6-kilobase 5′ sequence, including the first ATG, is deleted in the oc mutant mouse (Scimeca et al., 2000). The loss of V-ATPase function is thought to be a direct cause of osteopetrosis in the oc mutant. Nevertheless, chronic renal failure leads to bone disease, including osteosclerosis (Adams, 2002), and recessive osteopetrosis in humans is associated with neural defects (Sobacchi et al., 2001). These symptoms may not be directly associated with bone overgrowth owing to the defect of V-ATPase function.

Roct cDNA encodes 537 amino acids, and its predicted amino acid sequence contains 12 putative transmembrane domains and shows high homology with those of organic anion transporters (OATs), especially rat OAT3 (rOAT3; 92.4%) and human OAT3 (hOAT3, SLC22A8; 78.0%). OAT3 is expressed in renal microtubules and at the blood-brain barrier (BBB) and is considered to play an important role in the excretion of uremic toxins and neurotransmitter metabolites (Cha et al., 2001; Deguchi et al., 2002; Ohtsuki et al., 2002; Mori et al., 2003). Furthermore, in humans, OAT3 and V-ATPase genes are located in the same genomic locus (ch11q11-13), like Roct and V-ATPase in mouse (Heaney et al., 1998; Cha et al., 2001). Therefore, if Roct functions as an OAT3 in the mouse kidney and BBB, impairment of its function would explain the relationship among osteopetrosis, renal failure, and neural defects. Indeed, Roct has been suggested to function at the kidney and choroid plexus based on a study of Roct knockout mice (Sweet et al., 2002). Regarding function, Roct cRNA-injected oocytes exhibited uptake of p-aminohippuric acid (PAH), estrone sulfate (E₁S), and taurocholate (Sweet et al., 2002), but otherwise, the transport properties and expression of Roct are still unclear.

The purposes of the present work were first, to clarify the transport properties of Roct using the Xenopus laevis oocyte expression system; and second, to investigate the expression and localization of Roct by means of Northern blot, RT-PCR, and immunohistochemical studies.

Materials and Methods

Animals. Adult male C57BL6 mice, weighing 23 to 27 g, were purchased from Nippon SLC (Hamamatsu, Japan). Mature female X. laevis were purchased from Hamamatsu Kyozai (Hamamatsu, Japan) and maintained in a controlled environment as described by Goldin (1992). All experiments were approved by the Animal Care Committee, Graduate School of Pharmaceutical Sciences, Tohoku University (Sendai, Japan).

Reagents. The pGEM-HEN vector was a kind gift from Dr. T. Abe (Tohoku University). para-[glycyl-2-³H]Aminohippuric acid ([³H]PAH, 3.25 Ci/mmol), [estradiol-6, 7-³H(N)]estradiol 17β-d-glucuronide ([³H]estradiol 17β-d-glucuronide, 45.0 Ci/mmol), [6, 7-³H(N)]estrone sulfate ammonium salt ([³H]E₁S, 43.1 Ci/mmol), and [³H(G)]digoxin ([³H]digoxin, 19.0 Ci/mmol) were purchased from PerkinElmer Life Sciences (Boston, MA). [phenyl-4(n)-³H]Benzylpenicillin ([³H]benzylpenicillin, 20.0 Ci/mmol) and [N-methyl-³H]cimetidine ([³H]cimetidine, 14.8 Ci/mmol) were purchased from Amersham Biosciences Inc. (Piscataway, NJ). [7-¹⁴C]Salicylic acid ([¹⁴C]salicylic acid, 55.5 mCi/mmol) and [2-¹⁴C]indomethacin ([¹⁴C]indomethacin, 20.0 mCi/mmol) were purchased from PerkinElmer Life Sciences. [3′,5′,7-³H(N)]Methotrexate, disodium salt ([³H]methotrexate, 23.0 Ci/mmol) was purchased from Moravek Biochemicals (Brea, CA). [1,5-¹⁴C]Glutarate (5.3 mCi/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, MO). [ring-³H]Indoxyl sulfate ([³H]indoxyl sulfate, 6.5 Ci/mmol) was synthesized and purified by PerkinElmer Life Sciences. [ring-³H]Homovanillic acid ([³H]HVA, 12.5 Ci/mmol) was synthesized and purified by IZOTOP (Budapest, Hungary). PAH, probenecid, salicylic acid, benzylpenicillin potassium salt, cimetidine, indomethacin, 5-hydroxyindole-3-acetic acid, 3′-azido-3′-deoxythymidine, prostaglandin E₂, taurocholate sodium salt, pravastatin sodium salt, calcitonin acetate, α-naphthyl β-d-glucuronide, and glutathione reduced form (GSH) were obtained from Wako Pure Chemicals (Osaka, Japan). Homovanillic acid, sulfobromophthalein, and uric acid were purchased from Nacalai Tesque (Kyoto, Japan). Indoxyl sulfate potassium salt, estrone-3-sulfate sodium salt, 1-hydroxyethane-1,1-diphosphonic acid (etidronic acid), parathyroid hormone, β-estradiol, tetraethylammonium, indoxyl β-d-glucuronide cyclohexylammonium salt, β-estradiol 3-sulfate sodium salt, β-estradiol 3-sulfate 17-(β-d-glucuronide) dipotassium salt, 1-naphthyl β-d-glucuronide sodium salt, l-glutathione oxidized disodium salt (GSSG), and N-acethyl-S-farnesyl-l-cysteine methyl ester were purchased from Sigma-Aldrich (St. Louis, MO). 3,5,3′-Triiodo-l-thyronine sodium salt and l-thyroxine sodium salt were purchased from Tokyo Kasei. Methotrexate was purchased from Calbiochem-Novabiochem (La Jolla, CA). All other chemicals were commercial products of analytical grade.

RT-PCR Analysis and Isolation of Roct cDNA. Total RNA was prepared from mouse tissues and brain capillary-rich fraction using TRIzol reagent (Invitrogen, Carlsbad, CA). Isolation of the brain capillary-rich fraction was performed as described previously (Hosoya et al., 2000). Total RNA (1 μg) was reverse-transcribed with ReverTra Ace (Toyobo, Osaka, Japan). The synthesized cDNA was used for subsequent polymerase chain reaction. Polymerase chain reaction primers for amplification of Roct cDNA were designed and synthesized based on the nucleotide sequence of Roct (GenBank accession no. XM_123355). There was a sense strand with a BamHI site (underline), 5′-CGGGATCCCGGGTTCATCTTGCCTGGTGCCATG-3′ (positions 63-85); antisense with an EcoRI site (underline), 5′-CGGAATTCCGAAAGAGGATTCTGTTGTTCTTAGCTA-3′ (positions 1694-1719). The amplification conditions were as follows: 94°C for 30 s, 60°C for 30 s, 74°C for 60 s, for 30 cycles. The PCR products were separated by electrophoresis on 1% agarose gels. The RT-PCR product amplified from kidney was subcloned into the EcoRV site of pBluescript SKII(+) (Stratagene, La Jolla, CA) using a BKL kit (Takara, Shiga, Japan). Both strands of the subcloned cDNA inserts were subjected to DNA sequencing (model 4200; Li-COR, Lincoln, NE), and cDNA clones with base sequences 100% identical to the coding region of Roct were isolated. Then, the Roct cDNA was cut out from pBluescript SKII(+)/Roct using BamHI and EcoRI and subcloned into pGEM-HEN for in vitro transcription.

Functional Expression of Roct in X. laevis Oocytes. Capped cRNA of Roct was transcribed from NotI-linearized pGEM-HEN containing Roct cDNA with T7 RNA polymerase as described previously (Deguchi et al., 2002). pGEM-HEN contains 3′ and 5′-untranslated regions of Xenopus β-globin, which stabilize RNA and enhance the protein synthesis efficiency in the oocyte expression system (Liman et al., 1992). Defolliculated oocytes were injected with 50 nl of water or capped Roct cRNA (10 ng) and incubated in Barth's solution [88 mM NaCl, 1 mM KCl, 0.33 mM Ca(NO₃)₂, 0.4 mM CaCl₂, 0.8 mM MgSO₄, 2.4 mM NaHCO₃, and 10 mM HEPES] containing 50 μg/ml gentamicin and 2.5 mM pyruvate, pH 7.4, at 20°C. After incubation for 3 days, uptake experiments were performed.

Uptake by Oocytes. The uptake experiment was initiated by replacing the ND96 solution (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, and 5 mM HEPES, pH 7.4) with one containing radiolabeled substrate and terminated by addition of ice-cold ND96 solution after a 1-h incubation at 20°C. In previous studies using OAT3-expressing oocytes, the uptake kinetics of [³H]E₁S, [¹⁴C]- or [³H]PAH, [³H]methotrexate, and [³H]cimetidine were determined at 1 h as well (Kusuhara et al., 1999; Cha et al., 2001), as was also the case in studies of [³H]estradiol 17β-d-glucuronide, [³H]indoxyl sulfate, [³H]homovanillic acid, [¹⁴C]glutarate, and [¹⁴C]salicylic acid uptake (Cha et al., 2001; Ohtsuki et al., 2002; Mori et al., 2003). The concentrations of radiolabeled substrates were different due to the differences in the specific radioactivity of substrates. Therefore, the transport activity was normalized by substrate concentration and indicated in the unit of microliters/(hour·oocyte). Na⁺-free ND96 solution was prepared by equimolar replacement of NaCl with choline chloride. In the case of Cl^--free ND96 solution, NaCl, KCl, CaCl₂, and MgCl₂ were replaced with equimolar sodium gluconate, potassium gluconate, calcium gluconate, and magnesium gluconate, respectively. Oocytes were washed four times with ice-cold ND96 solution and solubilized in 5% SDS, and the accumulated radioactivity was determined in a liquid scintillation counter (LS-6500; Beckman Coulter Inc., Fullerton, CA).

The kinetic parameters for the uptake of benzylpenicillin or E₁S by Roct were estimated from the following equation: V = V_max × C/(K_m + C), where V is the uptake rate of the benzylpenicillin or E₁S [pmol/(h · oocyte)], C is the benzylpenicillin or E₁S concentration in the medium (μM), K_m is Michaelis-Menten constant (μM), and V_max is the maximum uptake rate [pmol/(h · oocyte)]. To obtain the kinetic parameters, the equation was fitted to the Roct-specific transport rate, which was obtained by subtracting the transport rate in water-injected oocytes from that in Roct-expressing oocytes, by an iterative nonlinear least-squares method using the MULTI program (Yamaoka et al., 1981). The input data were weighted as the reciprocal of the observed values, and the damping Gauss-Newton algorithm was used for fitting. The fitted line was converted to the V/C versus V form (Eadie-Scatchard plot).

For the inhibition study, uptake of 500 nM [³H]benzylpencillin by Roct was measured in the presence or absence of unlabeled compounds in ND96 solution. The concentration of inhibitors was 1 mM in most cases. However, calcitonin was examined at 10 μM because this was the maximum concentration that could be prepared from one vial (200 μg), and triiodothyronine, thyroxine, and estradiol were examined at 0.1 mM due to their low solubility. Parathyroid hormone was tested at 0.1 μM, because it was reported to have an inhibitory effect on phosphate transport at this concentration (Ba et al., 2003), and indoxyl glucuronide, estradiol sulfate, estradiol glucuronide sulfate, naphthyl sulfate, naphthyl glucuronide, GSSG, GSH, and N-acetyl-S-farnesyl-cys were examined at 5 μM, because inhibitory effects of these conjugates were reported at this concentration in the case of OAT4 (Cha et al., 2000). The specific uptake was obtained by subtracting the uptake into water-injected oocytes from that into Roct-expressing oocytes.

For experiments involving coexpression of Roct and rat Na⁺-dicarboxylate cotransporter (rNaDC-1), rNaDC-1 cDNA was isolated from rat kidney by RT-PCR and inserted into pBluescript SKII(+). cRNA was synthesized from its cDNA using T3 RNA polymerase. rNaDC-1 cRNA injected oocytes exhibited significant uptake of [¹⁴C]glutarate compared with that by water-injected oocytes [0.108 ± 0.020 (μl/(h·oocyte) and 0.0409 ± 0.0220 (μl/(h·oocyte), respectively (P < 0.01)]. Roct and rNaDC-1 cRNA were coinjected at a ratio of 2:1. For trans-stimulation studies, oocytes were pretreated for 90 min with incubation in ND96 solution containing 100 μM glutarate or in Na⁺-free ND96 solution containing 100 μM glutarate and quickly washed four times with glutarate-free ice-cold ND96 solution before initiation of uptake.

Data Analysis. Unless otherwise indicated, all data are presented as the mean ± S.E.M. An unpaired, two-tailed Student's t test was used to determine the significance of differences between two group means. One-way analysis of variance followed by the modified Fisher's least-significant-difference method was used to assess the statistical significance of differences among means of more than two groups.

Northern Blot Analysis. Blots containing 2-μg aliquots of poly(A)⁺ RNA from various mouse tissues (mouse multiple tissue Northern blot; BD Biosciences Clontech, Palo Alto, CA) were hybridized with a cDNA fragment of Roct (corresponding to positions 557-1409) as a probe; this region encodes an amino acid sequence with 48 and 41% identity to the corresponding regions of mouse OAT1 and mouse OAT2, respectively. The filter was washed twice with 2× SSC (1× SSC = 0.15 M NaCl and 0.015 M sodium citrate) and 0.05% SDS for 30 min at room temperature, and then twice with 0.1× SSC and 0.1% SDS for 40 min at 50°C. Then, the filter was exposed to film (X-OMAT; Eastman Kodak, Rochester, NY).

Western Blot Analysis. The membrane protein fraction of mouse brain, kidney, and small intestine epithelium was prepared using the procedure described in a previous report (Hosoya et al., 2000). The protein (62.5 μg/lane) was subjected to SDS-polyacrylamide gel electrophoresis. The separated proteins were electrotransferred to nitrocellulose membrane (Toyo Roshi, Tokyo, Japan), and the membrane was treated with blocking solution (Block Ace; Dainihon Pharmaceutical, Osaka, Japan) for 1 h at room temperature and diluted affinity-purified anti-OAT3 antibody (1 μg/ml) for 16 h at 4°C. Anti-OAT3 antibody was raised against an oligopeptide (KTKQESEAEKAS) corresponding to amino acid residues 516-527 of rOAT3 (Mori et al., 2003). Membranes were washed five times with 0.1% Tween 20/phosphate-buffered saline and then incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG (1:5000; Kirkegaard and Perry Laboratories, Gaithersburg, MD) for 1 h at room temperature. Immunoreactivity was visualized with an enhanced chemiluminescence kit (Supersignal west pico chemiluminescent substrate; Pierce Chemical, Rockford, IL).

Immunostaining Analysis. Adult mice were anesthetized with an intramuscular injection of ketamine and xylazine before perfusion through the carotid artery with 2% paraformaldehyde in 0.1 M phosphate buffer. The cerebrum was excised and immersed in 0.5 M sucrose/0.1 M phosphate buffer solution. Sagittal sections (10 μm in thickness) were cut from the frozen cerebrum using a cryostat (CM1900; Leica, Heidelberg, Germany) and mounted on silanized slide glasses (DAKO, Carpinteria, CA). The sections were incubated with 10% goat serum (Histofine; Nichirei, Tokyo, Japan) for 1 h at room temperature. The primary antibody [1 μg/ml rabbit anti-Roct antibody or 3.1 μg/ml mouse anti-glial fibrillary acidic protein (GFAP) antibody; Sigma-Aldrich) was applied to the sections, which were incubated for two nights at 4°C. For single staining of Roct, sections were further incubated with Alexa 488-conjugated goat anti-rabbit IgG (1:200; Molecular Probes, Eugene, OR) as a secondary antibody. For double staining, Alexa 546-conjugated goat anti-mouse IgG (1:200; Molecular Probes) was used as a secondary antibody for anti-GFAP antibody. Sections were mounted using Vectashield mounting medium (Vector Laboratories, Burlingame, CA). The stained sections were viewed using a confocal laser microscope (TCS SP; Leica).

Results

Roct Transports Substrates of OAT3. Roct cDNA was isolated from mouse kidney. The oocytes injected with Roct cRNA exhibited 118-fold greater [³H]PAH uptake than water-injected oocytes (Table 1), indicating that Roct is a PAH transporter, like the OAT family. Roct cRNA-injected oocytes exhibited uptake of various OAT3 substrates, such as [³H]E₁S, [³H]benzylpenicillin, [³H]estradiol 17β-glucronide, [³H]indoxyl sulfate, and [³H]cimetidine. Furthermore, [¹⁴C]indomethacin, [³H]homovanillic acid, [¹⁴C]glutarate, [¹⁴C]salicylic acid, and [³H]methotrexate were taken up by the cRNA-injected oocytes. In contrast, [³H]digoxin, which is a substrate of organic anion transporting polypeptide 2 (oatp2; SLC21A5), but not of OAT family members, was not taken up by the cRNA-injected oocytes. These results indicate that Roct mediates transport of OAT3 substrates and other organic anions.

Transport Properties of Roct-Mediated Uptake of Benzylpenicillin and E₁S. The uptake of [³H]benzylpenicillin and [³H]E₁S by Roct cRNA-injected oocytes increased linearly for about 120 min (Fig. 1). Accordingly, analysis was carried out at 60 min in the following uptake study. As shown in Fig. 2, benzylpenicillin and E₁S uptake was concentration-dependent. The K_m and V_max values of benzylpenicillin uptake were obtained as 40.0 ± 7.8 μM and 42.8 ± 5.4 pmol/(h·oocyte), respectively (mean ± S.D.), and those of E₁S were found to be 5.54 ± 1.59 μM and 15.2 ± 1.8 pmol/(h·oocyte), respectively (mean ± S.D.). Under Na⁺-free and Cl^--free conditions, the [³H]benzylpenicillin uptake by Roct was 82.5 ± 9.0 and 69.0 ± 11% of the control, respectively, and the values are not significantly different from the control.

trans-Stimulation effect of glutarate on the uptake by Roct was examined by using oocytes coexpressing Roct and rat NaDC-1 (Fig. 3). Glutarate preloading increased the uptakes of benzylpenicillin and E₁S by 1.40- and 1.38-fold, respectively, compared with the uptake without glutarate, although these changes are not statistically significant. Na⁺-free conditions significantly reduced the stimulation of benzylpenicillin uptake by glutarate preloading (P < 0.05), and the stimulation of E₁S uptake was also reduced, although without statistical significance.

Inhibitory Effects of Various Compounds on Benzylpenicillin Uptake Mediated by Roct. To investigate the substrate selectivity of Roct, the cis-inhibitory effect of various compounds on Roct-mediated [³H]benzylpenicillin uptake was examined (Table 2). OAT3 substrates and inhibitors, such as probenecid, PAH, sulfobromophthalein, taurocholate, cimetidine, indomethacin, homovanillic acid, salicylic acid, methotrexate, 3′-azido-3′-deoxythymidine, 5-hydroxyindole-3-acetic acid, pravastatin, and uric acid exhibited distinct inhibitory potency. In contrast, tetraethylammonium, which is a substrate of organic cation transporter (OCT), did not significantly affect the uptake. Among the compounds involved in bone turnover, prostaglandin E₂ and triiodothyronine inhibited the uptake, whereas thyroxine, etidronic acid, calcitonin, parathyroid hormone, and estradiol had no significant effect.

The inhibitory effect of conjugated compounds was also examined (Table 3). Sulfate-conjugated estrone, indoxyl, and estradiol inhibited the uptake activity, and indoxyl glucuronide and estradiol glucuronide sulfate also exhibited a significant inhibitory effect. In contrast, naphthyl sulfate, naphtyl glucuronide, two forms of glutathione (GSSG and GSH), and N-acetyl-S-farnesyl-cys did not affect the uptake by Roct significantly.

Distribution of Roct mRNA in Mouse Tissues. Expression of Roct mRNA in mouse tissues was examined by Northern blot analysis (Fig. 4A). A strong 2.7-kilobase band was detected in the kidney. A similar, but faint, band was also detected in the brain, but not in the heart, spleen, lung, liver, skeletal muscle, or testis. The DNA probe used in this Northern blot hybridized to Roct with at least 100-fold greater affinity than to rOAT1 and rOAT2 in dot blot analysis, suggesting that the probe selectively detected Roct (data not shown). Furthermore, RT-PCR analysis showed that Roct mRNA was detected in the brain, brain capillary-rich fraction, kidney, eye, and choroid plexus at the expected size (1653 base pairs), whereas no such band was detected in the liver or small intestine (Fig. 4B).

Localization of Roct in Mouse Brain Capillaries. Cross-reactivity of anti-OAT3 polyclonal antibody was evaluated by Western blot analysis, because the antigen peptide for this antibody (KTKQESEAEKAS) is only one amino acid different from the corresponding peptide in Roct (KTKQEPEAEKAS). As shown in Fig. 5, a band was detected in crude membrane fraction of mouse kidney and brain at 50 kDa, which is the same size as reported in the rat kidney, brain, and brain capillary-rich fraction and in rOAT3-expressing LLC-PK1 cells (Hasegawa et al., 2002; Mori et al., 2003). This band was not detected in the crude membrane fraction of mouse small intestine. Therefore, this antibody was suggested to react selectively with Roct in mouse brain, including brain capillaries.

The result of the immunohistochemical analysis of Roct in mouse brain capillaries is shown in Fig. 6. Immunoreactivity of GFAP, an astrocyte marker, surrounds the brain blood vessels (Fig. 6, B, E, and H; red), and Roct immunoreactivity is observed inside the GFAP signals (Fig. 6, C, F, and I; green). Roct immunoreactivity was observed not only in relatively large blood vessels (Fig. 6, A and D) but also in capillaries (Fig. 6G). No significant fluorescence was observed in brain blood vessels with normal rabbit IgG (data not shown). Figure 6K shows the result of propidium iodide staining of nuclei (red) in capillary endothelial cells. Roct (green) is localized outside the nuclei (Fig. 6L). These results suggest that Roct is localized at the abluminal membrane of brain capillary endothelial cells.

Discussion

The Roct gene encodes a protein that has high amino acid sequence homology with rOAT3 (92.4%) and hOAT3 (78.0%). However, Roct also has 35 to 49% homology with OAT1 (SLC22A6), OAT2 (SLC22A7), OAT4 (SLC22A11), and OAT5 (SLC22A10), and it has not been investigated in detail whether Roct functions as mouse OAT3.

In this study, we have clarified the transport properties of Roct. Roct mediates transport of compounds that have been reported to be transported by rOAT3 and/or hOAT3 (Table 1) (Kusuhara et al., 1999; Cha et al., 2001; Deguchi et al., 2002; Hasegawa et al., 2003; Mori et al., 2003). Roct-expressing oocytes exhibited uptake of benzylpenicillin with K_m value of 40.0 μM (Fig. 2A) and in an Na⁺-independent manner, like that by rOAT3 (K_m = 85.1 μM) (Kusuhara et al., 1999; Hasegawa et al., 2003), and there was no significant effect under Cl^--free conditions. The K_m value of E₁S uptake by Roct was 5.54 μM (Fig. 2B), which is close to that of rOAT3 and hOAT3 (2.3 and 3.1 μM, respectively) (Kusuhara et al., 1999; Cha et al., 2001). rOAT3 has been reported to be a dicarboxylate exchanger (Sweet et al., 2003). The uptake of [³H]benzylpenicillin and [³H]E₁S seemed to be trans-stimulated by glutarate preloading with coexpression of rNaDC-1 (Fig. 3), suggesting that Roct is a dicarboxyl exchanger like rOAT3. [³H]Benzylpenicillin uptake by Roct was also inhibited by various compounds in a similar manner to that by rOAT3 and hOAT3 (Table 2) (Kusuhara et al., 1999; Cha et al., 2001). Although partial degradation of the radiolabeled compounds during uptake can not be ruled out, the properties exhibited by Roct-expressing oocytes were similar to those of rat and/or human OAT-expressing oocytes. Therefore, these transport properties suggest that Roct functions as mouse OAT3.

Compared with the transport properties of other organic ion transporters, Roct transported estradiol 17β-glucronide (Table 1), and this compound was transported by rOAT3 but not by rOAT1 (Sugiyama et al., 2001). The uptake by Roct was inhibited by indoxyl sulfate (Table 2), which does not inhibit prostaglandin F₂α uptake by rOAT2 (Enomoto et al., 2003). Digoxin is a substrate of rat oatp2 (Sugiyama et al., 2001), although Roct did not transport it (Table 1). The OCT family forms a part of the organic ion transporter superfamily (SLC22) with the OAT family, but an OCT substrate, tetraethylammonium, did not affect the uptake by Roct. These results are consistent with the idea that Roct acts as mouse OAT3.

Despite the similarity in the properties of rOAT3 and hOAT3, there are some differences in their transport properties and tissue expression. The transport properties of Roct resemble those of hOAT3 rather than rOAT3. [³H]E₁S uptake by rOAT3 was not inhibited by 1 mM indomethacin (Kusuhara et al., 1999). In contrast, the uptakes by hOAT3 and Roct were strongly inhibited by indomethacin (Table 2) (Cha et al., 2001), and Roct exhibited significant uptake of [³H]indomethacin (Table 1). Furthermore, the tissue distribution of Roct is similar to that of hOAT3. rOAT3 is expressed in liver (Kusuhara et al., 1999), whereas mRNAs of Roct and hOAT3 were not detected (Fig. 4) (Cha et al., 2001). Taking these results into consideration, we conclude that Roct functions as mouse OAT3, and its properties are similar to those of hOAT3 rather than rOAT3.

Roct is expressed in the kidney (Fig. 4), and the uptake of PAH, E₁S and taurocholate by renal slices is decreased in Roct-knockout mouse compared with the wild type (Sweet et al., 2002), suggesting that Roct plays a certain role in the kidney. Recently, Jutabha et al. (2003) identified OATv1 and proposed the renal excretion pathway, in which organic anions are taken up by the epithelial cells via OAT1 and OAT3 and exit from the cells via OATv1. The uptake activity by OATv1 was inhibited by various sulfate and glucuronide conjugates, whereas there was little information about the substrate specificity of OAT3/Roct as far as conjugates are concerned. The uptake by Roct was inhibited by sulfate and glucuronide conjugates (Table 3), and, among those, E₁S, indoxyl sulfate, indoxyl glucuronide, and estradiol sulfate inhibited both Roct and OATv1 transport, whereas glutathiones and a cysteine conjugate did not inhibit the uptake. These results suggest that OAT3/Roct accepts sulfate and glucuronide conjugates as a substrate, and is likely to be involved in renal excretion of those conjugates cooperating with OATv1.

Chronic renal failure is often associated with bone disease. One possible explanation is that OAT3/Roct is involved in the renal excretion of hormones related to bone turnover, and impairment of OAT3/Roct functions alters the plasma level of those hormones, resulting in a changing in bone turnover rate. In fact, E₁S, a metabolite of estrone, is a high-affinity substrate of Roct and OAT3 (Fig. 2B) (Kusuhara et al., 1999; Cha et al., 2001), and estrogens regulate bone formation and resorption. Moreover, the plasma level of estrone in post-menopausal women is correlated with bone mineral density (Takayanagi et al., 2002). However, the transport properties of OAT3/Roct have not been investigated in relation to bone turnover. In the present study, the inhibitory effect of compounds related to bone turnover was investigated (Table 2). Etidronic acid is a drug used to treat osteoporosis. Calcitonin and estradiol promote osteogenesis. Prostaglandin E₂, parathyroid hormone, and thyroid hormones influence the differentiation of osteoclasts (Katagiri and Takahashi, 2002). Among these compounds, prostaglandin E₂ and triiodothyronine inhibited the uptake (Table 2). Therefore, alteration in the transport of these hormones in the body is likely to influence bone turnover in renal failure patients and the oc mutant mouse.

Calcitonin and GSH increased the uptake by Roct by 52 and 40%, respectively (Tables 2 and 3). One possible explanation is that calcitonin and GSH increased the uptake by functioning as a driving force. The other is that the activity of Roct was promoted by modification, such as phosphorylation. Roct has eight putative protein kinase C phosphorylation sites, and calcitonin activates the protein kinase C pathway (Kajiya et al., 2003), whereas the effect of phosphorylation on OAT3/Roct is still unknown.

Our recent reports have shown that rOAT3 is expressed at the BBB and is involved in the efflux transport of uremic toxins and neurotransmitter metabolites from the brain (Ohtsuki et al., 2002; Mori et al., 2003). The present study shows that Roct is expressed at brain capillary endothelial cells (Figs. 4B and 6). Luminal and abluminal membranes are separated by only 0.1 to 0.2 μm, and it is difficult to distinguish between them, whereas they diverge to surround the endothelial nucleus and are separated sufficiently (approximately 1-2 μm) to be resolved under confocal microscopy (Stewart et al., 1996). Therefore, the signal between GFAP and endothelial nuclei suggests that Roct is localized at the abluminal membrane of capillary endothelial cells (Fig. 6). Further study by using an astrocyte foot process marker, such as aquaporin-4 (Simard et al., 2003), and analysis with electron microscopy are necessary to establish precisely the abluminal localization of Roct, because there are difficulties in differentiating endothelial abluminal membrane from astrocyte or pericyte membrane.

rOAT3 is expressed in choroid plexus and is involved in the cerebrospinal fluid (CSF)-to-blood transport at the blood-cerebrospinal fluid barrier (Nagata et al., 2002). Similarly, Roct is expressed in choroid plexus (Fig. 4B), and Roct knockout mouse showed a decreased accumulation of fluorescein in capillaries from CSF, compared with that of the wild type (Sweet et al., 2002). Roct mediates the transport of a uremic toxin, indoxyl sulfate, and a dopamine metabolite, homovanillic acid, and its activity was inhibited by a serotonin metabolite, 5-hydroxyindole-3-acetic acid (Tables 1 and 2). Therefore, Roct is suggested to be involved in the clearance system for uremic toxins and neurotransmitter metabolites from the brain and CSF in mouse.

OAT3/Roct and V-ATPase genes are located in the same genomic locus in both human and mouse (Heaney et al., 1998; Scimeca et al., 2000; Cha et al., 2001). In mouse, suppression of the V-ATPase gene would influence Roct expression, because deletion of only the V-ATPase gene has been reported in the oc mutant (Scimeca et al., 2000). Therefore, in humans, it is conceivable that mutation in the V-ATPase gene influences the expression of hOAT3, because the genomic locations of these genes are conserved between human (ch11q11-13) and mouse (Chr 19) (Hudson et al., 2001). A reduction of OAT3/Roct expression would change the levels of substrate compounds in the blood and brain and this change could, in part, be associated with phenotypes such as osteopetrosis and neural defect.

In conclusion, the present study shows that Roct functions as mouse OAT3, and we suggest that it transports not only sulfate conjugates, but also glucuronide conjugates. Therefore, Roct is likely to be involved in the brain-to-blood BBB efflux transport and renal excretion of various compounds, including uremic toxins, neurotransmitter metabolites, and conjugates. Furthermore, Roct interacts with compounds involved in bone turnover, such as E₁S, prostaglandin E₂, and triiodothyronine. These results provide important information to allow a better understanding of the physiological role of OAT3/Roct at the BBB, blood-cerebrospinal fluid barrier, and kidney.