The blood-cerebrospinal fluid barrier is a major pathway of cerebral creatinine clearance: involvement of transporter-mediated process

Authors


Address correspondence and reprint requests to Professor Ken-ichi Hosoya, Graduate School of Medicine and Pharmaceutical Sciences, University of Toyama, Toyama 930-0194, Japan. E-mail: hosoyak@pha.u-toyama.ac.jp

Abstract

There is still incomplete evidence for the cerebral clearance of creatinine (CTN) which is an endogenous convulsant and accumulates in the brain and CSF of patients with renal failure. The purpose of this study was to clarify the transporter-mediated CTN efflux transport from the brain/CSF. In vivo data demonstrated that CTN after intracerebral administration was not significantly eliminated from the brain across the blood-brain barrier. In contrast, the elimination clearance of CTN from the CSF was 60-fold greater than that of inulin, reflecting CSF bulk flow. Even in renal failure model rats, the increasing ratio of the CTN concentration in the CSF was lower than that in the plasma, suggesting a significant role for the CSF-to-blood efflux process. The inhibitory effects of inhibitors and antisense oligonucleotides on CTN uptake by isolated choroid plexus indicated the involvement of rat organic cation transporter 3 (rOCT3) and creatine transporter (CRT) in CTN transport. rOCT3- and CRT-mediated low-affinity CTN transport with Km values of 47.7 and 52.0 mM, respectively. Our findings suggest that CTN is eliminated from the CSF across the blood-CSF barrier as a major pathway of cerebral CTN clearance and transporter-mediated processes are involved in the CTN transport in the choroid plexus.

Abbreviations used
BBB

blood-brain barrier

BCSFB

blood-cerebrospinal fluid barrier

CRT

creatine transporter

CSF

cerebrospinal fluid

CTN

creatinine

HEK293

human embryonic kidney cells

rOCT3-AS

antisense oligodeoxynucleotides against rat organic cation transporter 3

rOCT3-SCR

scrambled sequences of rOCT3-AS

TEA

tetraethylammonium

Creatinine (CTN) is a catabolic product of creatine which plays an essential role in the energy storage and transmission of phosphate-bound energy (Wyss and Kaddurah-Daouk 2000). It has been reported that intracisternal and intracerebroventricular injections of CTN lead to convulsions in animals (Jinnai et al. 1969; De Deyn et al. 1992), although CTN is present in the brain (329 μM in humans; Marescau et al. 1992) and the cerebrospinal fluid (CSF, 67 μM in humans; De Deyn et al. 2001). These reports suggest that increasing levels of CTN in the brain and CSF affect CNS function.

Renal failure leads to the accumulation of CTN in the brain and CSF as well as in the serum of patients (De Deyn et al. 1995), which could contribute to the neurological complications suffered by patients (De Deyn et al. 2001). However, despite the importance of understanding how CTN accumulates in the brain of patients, there is still incomplete evidence for the molecular mechanism(s) of CTN transport between the circulating blood and the brain. In particular, clarifying the cerebral clearance system of CTN from the brain and CSF will shed new insight into how the abnormal accumulation of CTN in the brain and CSF of patients with renal failure can be prevented.

As no specific saturable uptake exists for CTN in human red blood cells (Ku and Passow 1980), it has been postulated that CTN diffuses out of the brain. On the other hand, an active elimination of CTN from the CSF was suggested in a study using brain ventricular perfusion methods (Anderson and Heisey 1975; Bierer and Heisey 1975). Furthermore, Urakami et al. (2004, 2005) reported that human OCT2 mediates the tubular secretion of CTN in renal epithelial cells. These findings pose the question as to whether CTN undergoes efflux from the brain only by diffusion.

The brain possesses the blood-brain barrier (BBB) and the blood-cerebrospinal fluid barrier (BCSFB) which are formed by tight junctions of brain capillary endothelial cells and choroid plexus epithelial cells, respectively, to strictly regulate the exchange between the brain/CSF and the circulating blood. These barriers act as the brain detoxification system by expressing several kinds of brain/CSF-to-blood efflux transporters which facilitate elimination of xenobiotics and endogenous metabolites (Hosoya et al. 2004; Ohtsuki and Terasaki 2007). This prompted us to hypothesize that the BBB and/or BCSFB possess a transporter-mediated efflux transport system of CTN to remove it from the brain/CSF.

The purpose of the present study was to investigate the efflux transport of CTN from the brain/CSF across the BBB and/or BCSFB by means of the intracerebral micro-injection method, i.e. the Brain Efflux Index method (Kakee et al. 1996), intracerebroventricular administration method and the uptake by isolated choroid plexus, and to identify which transporters mediate CTN transport using expression systems of Xenopus laevis oocytes and human embryonic kidney cells (HEK293) and antisense oligonucleotides. The contribution of CTN efflux transport from the CSF under pathological conditions was evaluated using glycerol- or cisplatin-induced renal failure rat model.

Materials and methods

Animals

Adult male Wistar rats (260–280 g) were purchased from Japan SLC (Hamamatsu, Japan). Mature female Xenopus laevis were purchased from Kato-S-Science (Chiba, Japan) and maintained in a controlled environment. All experiments were approved by the Animal Care Committee, University of Toyama.

Reagents

[3H]Creatinine (50 Ci/mmol) and [14C]CTN (55 mCi/mmol) were obtained from American Radiolabeled Chemicals (St. Louis, MO, USA). [Carboxyl-14C]Inulin ([14C]inulin, 2.64 mCi/g) was purchased from ICN Biochemicals (Costa Mesa, CA, USA). [14C]Tetra-ethylammonium ([14C]TEA; 3.2 mCi/mmol) and [3H]water (18 mCi/mol) was obtained from Perkin-Elmer Life Sciences (Boston, MA, USA). [14C]Butanol (2 mCi/mmol) was purchased from GE Healthcare (Piscataway, NJ, USA). All other chemicals were commercial products of analytical grade.

In vivo CTN efflux and influx study

In vivo efflux experiments of [3H]CTN from the brain across the BBB were performed by the intracerebral micro-injection technique (Brain Efflux Index method) (Kakee et al. 1996). [3H]CTN efflux from the CSF after intracerebroventricular administration was studied using the method described previously (Kitazawa et al. 2000). The apparent blood-to-brain/CSF influx clearances of CTN (CLbrain, influx and CLCSF, influx) were determined by integration plot analysis as previously described (Ohtsuki et al. 2002). The details of the methods and the data analyses are included in Appendix S1.

Uptake study by freshly isolated rat choroid plexus

The CTN uptake by rat choroid plexus was examined using the centrifugal filtration method described previously (Suzuki et al. 1987). The rats were decapitated and the choroid plexus was isolated from the lateral ventricles and incubated at 37°C for 1 min in 300 μL extracellular fluid buffer containing (in mM): 122 NaCl, 25 NaHCO3, 3 KCl, 0.4 K2HPO4, 1.4 CaCl2, 1.2 MgCl2, 10 d-glucose and 10 HEPES. Incubation medium containing [3H]CTN or [14C]CTN in the presence or absence of inhibitors was added to initiate uptake. The tissue-to-medium concentration ratio was calculated using [14C]butanol or [3H]water as the cell volume of the choroid plexus. The final concentration of incubation medium was [3H]CTN (17–30 nM), [14C]CTN (35 μM), [14C]butanol (157 μM) and [3H]water (1.7 M), respectively. The radioactivity in the specimens was determined using a liquid scintillation spectrophotometer (LSC-5000; Aloka, Tokyo, Japan).

Induction of renal failure in rats

A rat model of glycerol- or cisplatin-induced renal failure was produced as reported previously in detail (Shulman et al. 1993; Okabe et al. 2002). Saline-treated rats were used as controls. Concentrations of CTN in the CSF and plasma were measured using a kit, Creatinine Test Wako (Wako, Osaka, Japan) according to the manufacturer’s instructions.

Uptake study using rat OCT3- and CRT-expressing Xenopus laevis oocytes

Using T7 RNA polymerase, capped cRNA was transcribed from NotI-linearized pGEM-HEN containing an open reading frame of rat organic cation transporter 3 (rOCT3) and rat creatine transporter (CRT) cDNAs as described elsewhere (Mori et al. 2003). Defolliculated oocytes were injected with 23 nL water or the capped cRNA (30–50 ng) and incubated at 18°C in freshly prepared standard oocyte saline solution containing 100 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 5 mM HEPES, 25 μg/mL gentamycin, 2.5 mM pyruvic acid and 1% bovine serum albumine, pH 7.5. The standard oocyte saline solution used to incubate the oocytes was replaced with fresh solution daily. Experiments were performed after incubation for 4 to 6 days. The uptake study was performed as described previously (Mori et al. 2003). The detailed methods of uptake study are included in Appendix S1.

Infusion of antisense oligonucleotides into the third ventricle of rats

The design of antisense oligodeoxynucleotides against rOCT3 (rOCT3-AS) and their infusion into the third ventricle of rats were performed as previously described (Nakayama et al. 2007). The detailed methods are included in Appendix S1.

Uptake study by HEK293 cells stably over-expressing rat CRT

The uptake study was performed as described previously (Ohtsuki et al. 2002). The detail methods of construction of an HEK293 cell line stably over-expressing rat CRT (CRT/HEK293 cells) and uptake study of [14C]CTN uptake by CRT/HEK293 cells are included in Appendix S1.

Kinetic analyses

The kinetic parameters for CTN uptake by Xenopus laevis oocytes expressing rOCT3 and CRT/HEK293 cells were obtained from the following equation:

image

where V is the uptake rate of CTN, C is the CTN concentration in the medium, Km is the Michaelis–Menten constant and Vmax is the maximum uptake rate. To obtain kinetic parameters, the equation was fitted using the iterative non-linear least squares regression analysis program, multi (Yamaoka et al. 1981).

Statistical analysis

Unless otherwise indicated, all data are presented as the mean ±SEM. The kinetic parameters are presented as the mean ± SD. An unpaired, two-tailed Student’s t-test was used to determine the significance of differences between two group means. anova followed by the modified Fisher’s least squares difference method was used to assess the statistical significance of differences among means of more than two groups.

Results

Limited elimination of CTN from rat brain parenchyma across the BBB

Figure 1a shows the time profile of the percentage of [3H]CTN remaining in the ipsilateral cerebrum after micro-injection into the parietal cortex area 2 region of rat brain. The percentage of [3H]CTN remaining in the brain at each time-point up to 120 min slightly decreased but was not significantly different from that at 2 min. The apparent elimination rate constant which was determined from the slope was 1.26 × 10−3 ± 0.67 × 10−3 per minute and the half-life was 550 min. This result suggests that the contribution of the BBB to CTN efflux transport out of the brain is very limited.

Figure 1.

In vivo brain/CSF-to-blood transport of creatinine (CTN) across the blood-brain barrier and blood-cerebrospinal fluid barrier. (a) Time-course of [3H]CTN remaining in the ipsilateral cerebrum following intracerebral micro-injection. [3H]CTN (0.1 μCi) dissolved in 0.5 μL extracellular fluid buffer was injected into the parietal cortex area 2 region of the brain. Each point represents the mean ± SEM (= 4–5). (b) CSF concentration versus time profiles of [3H]CTN (open circle) and [14C]inulin (open square) after intracerebroventricular administration. Each point represents the mean ± SEM (= 3–4). [14C]Inulin was co-administered as a reference for CSF turnover and passive diffusion into the brain parenchyma. The concentrations of CTN and inulin remaining in the cisternal CSF were determined at the designated times. The values are expressed as the percentage of the dose remaining per milliliter of CSF. BEI, Brain Efflux Index.

Elimination of CTN from rat CSF after intracerebroventricular administration

Figure 1b shows the residual CSF concentration of [3H]CTN and [14C]inulin after intracerebroventricular administration as a function of time. [3H]CTN was eliminated from the CSF with a greater rate constant of 0.113 ± 0.011 per minute than that of [14C]inulin (0.009 ± 0.017 per minute), a reference compound for CSF turnover and diffusion into the brain interstitional space through the ependymal space. The half-life of CTN elimination from the CSF was estimated to be 6.13 min. The elimination clearance of CTN from the CSF (86.1 μL/min) was approximately 60-fold greater than that of inulin (1.46 μL/min). The elimination clearance of inulin was close to the CSF bulk flow rate (2.9 μL/min) obtained by Suzuki et al. (1985).

Blood-to-brain and blood-to-CSF influx transport of CTN across the BBB and BCSFB

The in vivo blood-to-brain and blood-to-CSF influx transport of [14C]CTN across the BBB and BCSFB was evaluated by integration plot analysis after intravenous administration of [14C]CTN to rats (Fig. 2a and b). The apparent blood-to-brain influx clearance (CLbrain, influx) of [14C]CTN was determined to be 3.96 ± 0.53 μL/(min · g brain) (Fig. 2a). The apparent blood-to-CSF influx clearance (CLCSF, influx) of [14C]CTN was determined to be 4.45 ± 0.58 μL/(min · mL in CSF) (Fig. 2b). Assuming that the volume of CSF is 250 μL per rat (Cserr and Berman 1978) and the weight of rat brain is 1.6 g per rat, the CLCSF, influx of [14C]CTN was estimated to be 1.11 μL/min per rat [4.45 μL/(min · mL in CSF) × 0.25 mL CSF per rat] and 0.694 μL/(min · g brain) (1.11 μL/min per rat/1.6 g of brain per rat), respectively.

Figure 2.

In vivo blood-to-brain/CSF transport of creatinine (CTN) across the blood-brain barrier and blood-cerebrospinal fluid barrier. Integration plot analysis of the initial uptake of [14C]CTN by the brain (a) and CSF (b) after intravenous administration. [14C]CTN (3 μCi/head) was injected via the femoral vein. Each point represents the mean ± SEM (= 3–4). (c) The relationship between the increasing ratios of CTN concentrations in the plasma and CSF of rats with renal failure. Open circle: control rats, open square: glycerol-induced renal failure model rats, open triangle: cisplatin-induced renal failure model rats. Each point represents the mean ± SEM (= 4–6). AUC, area under curve.

The relationship between the increasing ratios of the CTN concentration in the plasma and the CSF of model rats with renal failure

As shown in Fig. 2c, the increasing ratio of the CTN concentration in the CSF of model rats with renal failure was smaller than that in the plasma of those rats. The average concentrations of CTN in the plasma and CSF of control rats were 59.2 ± 7.9 and 30.5 ± 7.0 μM, respectively. Even in rats with cisplatin-induced severe renal failure, the average concentrations of CTN in the plasma (243 ± 57.4 μM) increased by 4.1-fold, whereas the CSF concentration (77.0 ± 12.3 μM) increased by 2.5-fold, compared with that of control rats. This result suggests the greater transport rate of the CSF-to-blood CTN efflux process than that of the blood-to-CSF influx process and/or the limited permeability of the blood-to-CSF influx transport across the BCSFB.

Characteristics of the CTN uptake by isolated rat choroid plexus

To examine whether CTN was eliminated from the CSF via the BCSFB, the uptake study of [3H]CTN by the isolated choroid plexus was performed. [3H]CTN uptake by the isolated choroids plexus exhibited a time-dependent increase linearly for up to 2 min of incubation with an initial uptake rate of 0.887 ± 0.101 μL/(min · μL choroid plexus) (Fig. 3). To characterize the transporter(s) involved in [3H]CTN uptake by the isolated choroid plexus, the inhibitory effects of various compounds on the [3H]CTN uptake were investigated (Table 1). The uptake of [3H]CTN was significantly inhibited by 20, 50 and 80 mM CTN, 2 and 5 mM MPP+ (a substrate of rOCT1-3), 20 mM TEA (a substrate of rOCT1-3), 2 mM decynium22 (a potent cyanine-related inhibitor of rOCT1-3), 2 mM corticosterone (a selective and potent inhibitor of rOCT3) and 5 mM creatine (a substrate of CRT). MPP+ and creatine added together, each at a concentration of 5 mM, reduced the CTN uptake by 61%. In contrast, 10 mM choline (a substrate of rOCT1 and rOCT2 but not of rOCT3) exhibited only 5% inhibition. These results suggest the involvement of rOCT3 and CRT in the CTN uptake by the isolated choroid plexus.

Figure 3.

 Time-course of [3H]creatinine (CTN) uptake by isolated choroid plexus. Choroid plexus was incubated with [3H]CTN (30 nM) at 37°C. Each point represents the mean ± SEM (= 3).

Table 1.   Effect of various compounds on [3H]CTN uptake by the isolated choroid plexus
InhibitorsConcentration of inhibitors (mM)% of control
  1. *p < 0.01, **p < 0.05 significantly different from control.

  2. [3H]CTN (17 nM) uptake by the isolated choroid plexus was measured at 37°C for 2 min. Each value represents the mean ± SEM (= 4–8).

Control 100 ± 7
Creatinine (CTN)2076.5 ± 3.6*
CTN5072.8 ± 3.8*
CTN8072.3 ± 1.5*
MPP+277.0 ± 3.5*
MPP+565.6 ± 5.0*
Tetraethylammonium (TEA)2078.6 ± 6.1**
Decynium22257.7 ± 1.2*
Corticosterone285.2 ± 3.0**
Choline1095.5 ± 1.9
Creatine573.5 ± 3.8*
MPP+ and creatine 5 and 539.1 ± 2.0*

Contribution of rOCT3 and CRT to the CTN uptake by isolated choroid plexus

To further elucidate the in vivo contribution of rOCT3, we studied the inhibitory effects of antisense oligonucleotides against rOCT3 (rOCT3-AS) on the [14C]CTN uptake by the isolated choroid plexus (Table 2). [14C]CTN uptake by the isolated choroid plexus of rats which had rOCT3-AS infused into the third ventricles decreased by approximately 20% compared with that of untreated rats and rats treated with scrambled sequences of rOCT3-AS (rOCT3-SCR). There was no significant difference between untreated rats and rats treated with rOCT3-SCR, with regard to the CTN uptake. Five millimolar MPP+ additionally reduced the CTN uptake by the choroid plexus of rats treated with rOCT3-AS by 39% compared with that of untreated rats. MPP+ and creatine added together, each at a concentration of 5 mM, further reduced the CTN uptake by the choroid plexus of rats treated with rOCT3-AS by 62%. To distinguish the contribution of CRT from that of rOCT3, we investigated the Na+ and Cl dependence of the CTN uptake by the isolated choroid plexus. As shown in Table S1, the absence of Na+ or Cl reduced [14C]CTN uptake by 61% and 14%, respectively. Five and two millimolars of MPP+ and corticosterone, respectively, did not additionally inhibit the CTN uptake under Na+-free conditions, whereas 5 mM MPP+ inhibited the CTN uptake by 17% under Cl-free conditions.

Table 2.   Effect of antisense oligonucleotides against rOCT3 on [14C]CTN uptake by the isolated choroid plexus
Conditions/Inhibitors% of control
  1. rOCT3, rat organic cation transporter 3; rOCT3-SCR, scrambled sequences of rOCT3-AS; rOCT3-AS, antisense oligodeoxynucleotides against rat organic cation transporter 3; CTN, creatinine.

  2. *p < 0.01 significantly different from control.

  3. [14C]CTN (35 μM) uptake by the isolated choroid plexus of rats infused with antisense oligonucleotides against rOCT3 (rOCT3-AS) and their scrambled sequences (rOCT3-SCR) into the third ventricles was measured at 37°C for 2 min in the absence (control) or presence of MPP+ and creatine. Each value represents the mean ± SEM (= 4–12). Untreated rats were used as a control.

Control100 ± 4
rOCT3-SCR99.1 ± 6.6
rOCT3-AS80.4 ± 4.6*
rOCT3-AS + 5 mM MPP+61.2 ± 4.0*
rOCT3-AS + 5 mM MPP+ + 5 mM creatine38.2 ± 0.8*

CTN uptake mediated by rOCT3 and CRT

Although the inhibitory effects shown in Tables 1 and 2 and Table S1 suggest the involvement of rOCT3 and CRT in CTN uptake by the isolated choroid plexus, it remains unknown whether rOCT3 and CRT actually mediate CTN transport. In order to clarify the characteristics of rOCT3 and CRT-mediated CTN transport, a [14C]CTN uptake study was performed using rOCT3 or CRT-expressing oocytes and HEK293 cells.

[14C]Tetraethylammonium used as a positive control and [14C]CTN uptake by rOCT3-expressing oocytes was 8.4- and 7.3-fold greater than that of water-injected oocytes, respectively (Fig. S1). As shown in Fig. 4a, [14C]CTN was significantly taken up in a time-dependent manner by rOCT3-expressing oocytes compared with water-injected oocytes. The uptake of [14C]CTN exhibited saturable kinetics with a Km of 47.7 ± 10.7 mM and a Vmax of 118 ± 21 pmol/(min · oocyte) (Fig. 4b). The inhibitory effects of various compounds on [14C]CTN uptake mediated by rOCT3 were examined (Table 3). MPP+, TEA and methylguanidine at a concentration of 2 mM significantly inhibited rOCT3-mediated uptake by 78%, 52% and 53%, respectively. In contrast, guanidine at the concentration of 2 mM did not have an inhibitory effect. Furthermore, creatine (a substrate of CRT) and β-guanidinopropionic acid (an potent inhibitor of CRT) had no significant effect on the [14C]CTN uptake, suggesting that creatine and a potent CRT inhibitor do not affect rOCT3-mediated CTN transport.

Figure 4.

 Rat organic cation transporter 3 (rOCT3)-mediated creatinine (CTN) uptake by rOCT3-expressing Xenopus laevis oocytes. (a) Time-courses of [14C]CTN uptake (70 μM) by Xenopus laevis oocytes injected with water (closed circle) and rOCT3 cRNA (open circle). Each point represents the mean ± SEM (= 12–15). (b) Concentration dependence of rOCT3-mediated uptake of [14C]CTN. The uptake was measured at the indicated concentration for 1 h. The inset graph shows the Eadie–Scathard plot of the same data. Each point represents the mean ± SEM (= 10–15).

Table 3.   Effect of various compounds on [14C]CTN uptake by Xenopus laevis oocytes expressing rOCT3
Inhibitors% of control
  1. CTN, creatinine; rOCT3, rat organic cation transporter 3.

  2. *p < 0.01, **p < 0.05, significantly different from control.

  3. [14C]CTN (70 μM) uptake was measured at 20°C for 60 min in the absence (control) and presence of inhibitors (2 mM). rOCT3-mediated transport was obtained by subtracting the transport rate in water-injected oocytes from that in rOCT3-expressing oocytes. Each value represents the mean ± SEM (= 11–15).

Control100 ± 19
MPP+22.1 ± 2.8*
Tetraethylammonium (TEA)48.1 ± 8.3**
Methylguanidine 47.2 ± 18.9**
Guanidine137 ± 22
Creatine137 ± 24
β-Guanidinopropionate143 ± 21

As shown in Fig. S2, [14C]CTN uptake by CRT-expressing oocytes was 57-fold greater than that of water-injected oocytes. This uptake was inhibited 7% and 38% by unlabelled CTN at a concentration of 2 and 20 mM, respectively (Fig. S2). [14C]CTN was significantly taken up in a time-dependent manner by HEK293 cells stably expressing CRT compared with HEK293 cells without transfection (Fig. 5a). The uptake of [14C]CTN exhibited saturable kinetics with a Km of 52.0 ± 2.8 mM and a Vmax of 6.65 ± 0.26 μmol/(min · mg protein) (Fig. 5b). The absence of either Na+ or Cl reduced the [14C]CTN uptake by 98% and 89%, respectively (Table S2). Although 2 mM creatine significantly inhibited CRT-mediated [14C]CTN uptake by 90%, 2 mM CTN had no significant effect, while CTN at a concentration of 20 mM inhibited [14C]CTN uptake by 26% (Table S2).

Figure 5.

 Creatine transporter (CRT)-mediated creatinine (CTN) uptake by rat CRT-expressing human embryonic kidney (HEK293) cells. (a) Time courses of [14C]CTN uptake (9 μM) by HEK293 cell stably expressing rat CRT (open circle) and HEK293 cells without transfection (closed circle). Each point represents the mean ± SEM (= 4). (b) Concentration dependence of rat CRT-mediated uptake of [14C]CTN. The uptake was measured at the indicated concentration for 1 min at 37°C. The inset graph shows the Eadie–Scathard plot. Each point represents the mean ± SEM (= 8).

Expression of rOCT3 and CRT mRNAs in the choroid plexus

RT-PCR gave one amplified product at the expected size of 248 and 353 bp in the rat choroid plexus and brain, which was used as a positive control for rOCT3 and CRT mRNA expressions, respectively (Fig. S3). The nucleotide sequences of the products were identical with that of rat OCT3 (GenBank accession number: NM019230) and rat CRT (GenBank accession number: X66494).

Discussion

In the present study, we have demonstrated that CTN is eliminated from the CSF across the BCSFB as a major cerebral clearance pathway for CTN, and transporter-mediated processes are involved in the CTN transport in the choroid plexus. This study provides in vivo evidence that CTN after intracerebral administration was not significantly eliminated from the brain across the BBB (Fig. 1a). In contrast, the elimination clearance of CTN from the CSF was approximately 60-fold greater than that of inulin, reflecting bulk flow of the CSF (Fig. 1b). These results strongly suggest that CTN undergoes CSF-to-blood efflux transport via a carrier-mediated process at the BCSFB as a major route for its cerebral clearance. This is in close agreement with a previous report implying an active efflux transport of CTN from the CSF using brain ventricular perfusion methods (Anderson and Heisey 1975; Bierer and Heisey 1975).

To elucidate the contribution of the BCSFB to CTN elimination from the CSF, a [3H]CTN uptake study was performed using freshly isolated rat choroid plexus (Fig. 3). An elimination clearance of CTN via the BCSFB was estimated to be 5.32 μL/min per rat from the initial uptake rate of CTN by the isolated choroid plexus [0.887 μL/(min · μL choroid plexus) × 6 μL (the volume of total rat choroid plexuses per rat; Ogawa et al. 1994)]. The estimated value is 6.2% of the total CTN elimination clearance from the CSF in vivo (86.1 μL/min per rat). One possible explanation for the apparently low contribution of the BCSFB is that an additional active and rapid transport pathway of CTN, e.g. uptake by ependymal cells, besides its transport pathway across the BCSFB, is involved in the elimination of CTN from the CSF. Indeed, as shown in Fig. 1b, the residual CSF concentrations at 0 min of CTN and inulin after intracerebroventricular administration are significantly different. It has been also reported that ependymal cells express mouse OCT3 and CRT (Braissant et al. 2001; Vialou et al. 2004). However, the elimination clearance via the BCSFB (5.32 μL/min per rat) is still approximately fivefold greater than the apparent blood-to-CSF influx clearance of CTN (1.11 μL/min per rat, Fig. 2b), suggesting that CTN is asymmetrically transported across the BCSFB in the CSF-to-blood direction. In this regard, the cerebral clearance system of CTN at the BCSFB would play an important role in preventing the induction of brain convulsions resulting from the abnormal accumulation of CTN in the brain.

Focusing on CTN transport across the BBB, it appears that the blood-to-brain influx clearance (Fig. 2a) is much greater than the brain-to-blood efflux clearance (Fig. 1a). This notion may help us to understand previous reports indicating that the CTN concentration in the human brain (329 μM; Marescau et al. 1992) was four- to fivefold greater than that in the serum (80.8 μM in men and 65.3 μM in women; Marescau et al. 1997). However, the elimination clearance of CTN via the BCSFB, which can be normalized as 3.33 μL/(min · g brain) (5.32 μL/min per rat/1.6 g of brain per rat) assuming the weight of rat brain is 1.6 g per rat, is almost comparable with the total blood-to-brain and blood-to-CSF influx clearances [3.96 μL/(min · g brain) + 0.694 μL/(min · g brain)]. This suggests that the influx and efflux transport of CTN between the brain and the circulating blood are well balanced at least under normal conditions. On the other hand, CTN is non-enzymatically produced from creatine at a constant rate (approximately 1.7% of total body creatine per day) within cells (Wyss and Kaddurah-Daouk 2000). Because creatine exists at a high concentration of 9.0 mM in the rat brain (Marescau et al. 1992), the non-enzymatic production rate of CTN in the brain can be estimated to be 0.11 nmol/(min · g brain) (9.0 mM × 1 mL/g of rat brain × 1.7%/24 × 60 min). This production rate is almost comparable with and even greater than that of its blood-to-brain influx transport which is normalized as 0.066 nmol/(min · g brain) by the blood-to-brain influx clearance and the concentration of CTN in the serum (3.96 μL/(min · g brain) × 16.6 μM of CTN concentration in the rat serum (Marescau et al. 1992)]. In this regard, the greater CTN concentration in the brain than in the serum could be because of the constant production of CTN from creatine in the brain parenchymal cells in addition to its blood-to-brain influx transport.

The effects of several inhibitors on [3H]CTN uptake by the isolated choroid plexus were examined to identify the transporters involved (Table 1). Each concentration of inhibitor used, which is over 10-fold greater than the Km, Ki or IC50 values of each transporter, is high enough to almost completely inhibit the corresponding transporter. MPP+ and TEA are typical substrates of rOCTs [MPP+; Km = 9.6 μM (Busch et al. 1996), 9.4 μM (Gründemann et al. 1999) and 91 μM (Wu et al. 1998) for rOCT1, 2 and 3; TEA, Km = 38 μM, 45 μM (Urakami et al. 1998) and 2.5 mM (Kekuda et al. 1998) for rOCT1, 2 and 3, respectively]. Decynium22 is a potent cyanine-related inhibitor of OCT1-3 [IC50 = 0.36 μM for rOCT1 (Gründemann et al. 1994), IC50 = 0.58 μM for rOCT2 (Koepsell et al. 1999) and Ki = 16 nM for rOCT3 (Rüss et al. 1993)]. Our present results showed that 5 mM MPP+, 20 mM TEA and 2 mM decynium22 inhibited the [3H]CTN uptake by 34%, 21% and 42%, respectively, suggesting the involvement of OCTs in the CTN uptake by the choroid plexus. A previous report (Sweet et al. 2001) and the present study (Fig. S3) demonstrated that rOCT2 and rOCT3 mRNAs among OCTs are expressed in the choroid plexus. Sweet et al. (2001) also suggested the apical membrane localization of rOCT2 in the choroid plexus, whereas the localization of rOCT3 in the choroid plexus remains to be demonstrated. Two millimolar corticosterone, a selective and potent inhibitor of OCT3 (Hayer-Zillgen et al. 2002), inhibited the CTN uptake by 15%. On the other hand, choline, a substrate of rOCT1 (Km = 346 μM) and rOCT2 (Km = 441 μM) but not rOCT3 (Sweet et al. 2001), exhibited only 5% inhibition of the CTN uptake at a concentration of 10 mM. These results suggest a greater contribution of rOCT3 (15%) than that of rOCT2 (5% if any) to the CTN uptake by the choroid plexus. In support of this notion, the [14C]CTN uptake by the isolated choroid plexus of rats treated with the antisense oligonucleotide against rOCT3 (rOCT3-AS) was reduced by approximately 20% compared with that of untreated rats and rats treated with its scrambled sequences (Table 2). This inhibitory effect of rOCT3-AS on the CTN uptake (20%) is very close to that of corticosterone (15%, Table 1), confirming that the actual contribution of rOCT3 could be 15–20%. Therefore, treatment with rOCT3-AS could almost completely inhibit the rOCT3 function in the choroid plexus. However, the CTN uptake by the choroid plexus of rats treated with rOCT3-AS was additionally inhibited by 39% by 5 mM MPP+ (Table 2). This maximal inhibitory effect of rOCT3-AS and MPP+ is close to the inhibitory effect of 5 mM MPP+ alone on the CTN uptake by the choroid plexus (34%, Table 1). Furthermore, as shown in Table 1, the inhibitory effects of corticosterone (15%) and choline (5%) on the CTN uptake by the isolated choroid plexus are still smaller than that of MPP+ (34%) and decynium22 (42%). These results also raise the possibility that a MPP+ and/or decyanium22-sensitive unknown transport system is additionally involved in CTN transport in the choroid plexus.

Next, we focused on CRT as a candidate of a CTN transporter on the basis of the structural similarity of CTN to creatine. A previous report (Happe and Murrin 1995) and the present study (Fig. S3) demonstrated that CRT mRNA is expressed in the choroid plexus. We have obtained data showing that CRT is localized on the brush-border membrane of choroid plexus epithelial cells (Tachikawa et al., 2008). Creatine, a substrate of CRT (Km = 46 and 29 μM; Schloss et al. 1994; Saltarelli et al. 1996), inhibited the uptake by 27% at a concentration of 5 mM, suggesting that CRT makes a contribution (around 30%) to the CTN uptake by the choroid plexus. This is confirmed by our present result showing that the absence of Na+ or Cl reduced the CTN uptake by 61% and 14%, respectively (Table S1). However, the lack of agreement between the reduction ratio of the CTN uptake under Na+-free conditions (61%, Table S1) and the inhibitory effect of creatine (27%, Table 1) needs some explanation. One possible explanation is that OCTs-mediated CTN transport in the choroid plexus is reduced under the Na+-free conditions. Indeed, the reduction ratio of the CTN uptake under Na+-free conditions (61%, Table S1) is consistent with the inhibitory effect of MPP+ and creatine added together (61%, Table 1). MPP+ and corticosterone did not additionally inhibit the CTN uptake under Na+-free conditions (Table S1). Because it has been reported that OCTs mediate Na+- and Cl independently and the membrane potential-dependent transport (Wu et al. 1998), the physiological changes in ion-balance of the isolated choroid plexus under Na+-free conditions may affect the transport function of OCTs.

Taken together, approximately 50% of the total CTN uptake by the isolated choroid plexus could be explained by rOCT2-, rOCT3- and CRT-mediated processes. Indeed, MPP+ and creatine added together, each at a concentration of 5 mM, significantly inhibited the CTN uptake by 61% (Table 1). Treatment with rOCT3-AS, MPP+ and creatine also reduced the CTN uptake by 62% (Table 2). However, as endogenous compounds in the CSF may affect rOCT3- and CRT-mediated transport of CTN in the choroid plexus, the physiological conditions in the CSF should be considered to evaluate the actual contribution of rOCT3 and CRT. Creatine is present in human CSF at a normal concentration of 24–66 μM (Almeida et al. 2004). Considering that the Km values of CRT for creatine are reported to be 46 μM/29 μM (Schloss et al. 1994; Saltarelli et al. 1996), CRT on the brush-border membrane of the choroid plexus epithelial cells may be approximately 50% saturated by endogenous creatine in the CSF. Furthermore, the Km value of CRT for CTN (52 mM, Fig. 5b) is almost three orders of magnitude greater than the Km value of CRT for creatine. Therefore, it is unlikely that CRT plays a significant role in the elimination of CTN from the CSF under physiological conditions. In contrast, rOCT3-mediated CTN transport is not inhibited by creatine at a concentration of 2 mM (Table 3), suggesting that rOCT3-mediated CTN transport is involved, at least in part, in the elimination of CTN at the BCSFB. Furthermore, residual fraction of the CTN uptake (approximately 40%) by the choroid plexus may be explained by the non-saturable transport process that has been suggested by the previous report (Ku and Passow 1980) as even 80 mM CTN inhibited the CTN uptake by the choroid plexus by only 28% (Table 1).

In non-dialyzed patients with renal insufficiency, CTN levels in the brain and CSF increase from their normal values of 290 and 67.6 μM to 389–733 μM and 168–521 μM, respectively (De Deyn et al. 2001) in relation to the 2- to 10-fold elevation of the serum CTN concentration (Marescau et al. 1997). In the present study, we found that the increasing ratio of the CTN level in the CSF of model rats with renal failure was smaller than that in the plasma of those rats (Fig. 2c). Even in rats with cisplatin-induced severe renal failure, the average concentration of CTN in the plasma increased by 4.1-fold, whereas the CSF concentration increased by 2.5-fold, compared with that of control rats. Accordingly, it appears that the CTN efflux transport, at least in part via rOCT3, at the BCSFB play a crucial role in an efficient removal of CTN from the CSF even under pathological conditions. Indeed, the Km value of rOCT3-mediated CTN transport (47.7 mM; Fig. 4b) is still much higher than the CTN concentration in the CSF of patients with renal failure (168–521 μM; De Deyn et al. 2001), suggesting that rOCT3 functions as a CTN transporter without saturation. On the other hand, we also cannot exclude the possibility that the blood-to-CSF influx transport of CTN across the BCSFB is limited even if the serum CTN concentration is significantly increased. Therefore, it will be clinically beneficial, as far as the cerebral accumulation of CTN is concerned, to block the BBB influx transport of CTN in patients with renal insufficiency. The present study indicates that the blood-to-brain influx clearance of CTN [Figs 2a and 3, 0.96 μL/(min · g brain)] is approximately 14-fold greater than that of sucrose [0.29 μL/(min  · g brain)] used as a BBB non-permeable paracellular marker (Lightman et al. 1987). This evidence suggests the blood-to-brain transport of CTN via carrier-mediated transport rather than passive diffusion. CRT may be a candidate for the blood-to-brain CTN influx transport across the BBB because we have reported that CRT is localized on the luminal membrane of brain capillary endothelial cells and mediates creatine influx transport across the BBB (Ohtsuki et al. 2002; Tachikawa et al. 2004). However, considering that the Km value of CRT-mediated creatine transport at the mouse BBB is 16.2 μM, the function of CRT on the luminal membrane is suggested to be almost 95% saturated by endogenous creatine in the serum (272 μM in mice) (Ohtsuki et al. 2002). In this regard, the other mechanism(s) underlying the CTN influx transport across the BBB may be involved and this requires clarification in future studies.

In conclusion, the BCSFB is a major pathway of the cerebral CTN clearance and transporter-mediated processes, at least in part via rOCT3, are involved in CTN efflux transport from the CSF at the BCSFB under physiological conditions. The present findings provide a novel insight into how CTN is distributed in the brain and even accumulated in the brain of the patients with renal failure and help us to clarify the molecular mechanism of neurological complications in those patients.

Acknowledgements

We would like to thank Dr T. Abe for supplying the pGEM-HEN vector for protein expression in Xenopus laevis oocytes and Drs Y. Hashimoto and S. Ito for technical assistance. This study was supported, in part, by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports, and Culture, and by Research Fellowships of the Japan Society for the Promotion of Science for Young Scientists, and Nakatomi Foundation, Japan.

Ancillary