Functional regulation of P-glycoprotein at the blood-brain barrier in proton-coupled folate transporter (PCFT) mutant mice

Abstract

Folate deficiency has been associated with many adverse clinical manifestations. The blood-brain barrier (BBB), formed by brain capillary endothelial cells, protects the brain from exposure to neurotoxicants. The function of BBB is modulated by multiple ABC transporters, particularly P-glycoprotein. A proton-coupled folate transporter (PCFT)-deficient mouse has been previously described as a model for systemic folate deficiency. Herein, we demonstrate that exposing mouse brain capillaries to the antiepileptic drug, valproic acid (VPA; 5 μM), significantly increased P-glycoprotein transport function in the wild-type animals. A ligand to the aryl hydrocarbon receptor, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), produced a similar induction of P-glycoprotein, which tightened the BBB, thereby increasing the neuroprotection. However, VPA- or TCDD-induced P-glycoprotein transport was blocked in the PCFT-nullizygous mice, indicating that multiple neuroprotective mechanisms are compromised under folate-deficient conditions. Brain capillaries from S-folinic acid (SFA; 40 mg/kg)-treated PCFT-nullizygous mice exhibited increased P-glycoprotein transport following VPA exposure. This suggests that SFA supplementation restored the normal BBB function. In addition, we show that tight-junction proteins are disintegrated in the PCFT mutant mice. Taken together, these findings strongly suggest that folate deficiency disrupts the BBB function by targeting the transporter and tight junctions, which may contribute to the development of neurological disorders.—Wang, X., Cabrera, R. M., Li, Y., Miller, D. S., Finnell, R. H. Functional regulation of P-glycoprotein at the blood-brain barrier in proton-coupled folate transporter (PCFT) mutant mice.

Folic acid is an essential nutritional element involved in multiple biological processes and cell functions. Dietary intake is the major means of securing and maintaining stable intracellular folate homeostasis. Folate deficiency has been associated with many adverse clinical phenotypes, including an increased risk for selected congenital malformations, most significantly neural tube defects (NTDs) (1–3), as well as neurological disorders in pediatric and adult patients. Folic acid supplementation in the periconceptional period has been shown to have protective effects against many different birth defects (2, 3). In addition, it has been sporadically effective in treating cases of low cerebral spinal fluid (CSF) folate concentrations, which have been linked to several different neurological diseases of childhood (4–6), such as the cerebral folate deficiency (CFD) syndrome. While most patients with CFD-related syndromes respond to folinic acid therapy and show a short-term clinical improvement, the ability to sustain the clinical benefit of high-dose folate treatment has not been consistently achieved (7, 8). Thus, understanding folate transport, metabolism, and delivery to the central nervous system (CNS) may help develop new strategies to prevent, or provide effective treatment for many pediatric neurological disorders.

The proton-coupled folate transporter (PCFT) is one of the three folate transport systems that works with other folate transporters and mediates folate transport via folate receptor-mediated endocytosis. The PCFT transporter is found to be expressed in many tissues, including the intestine, liver, and choroid plexus, and it is highly active in a low-pH acidic environment, such as cytosol (9). The PCFT mutant mouse displays hereditary folate malabsorption and anemia secondary to the folate deficiency. Previous studies from our laboratory described a mouse model with an inactivated PCFT gene, which has proved to be a great tool to investigate the function of PCFT in folate transport and metabolism (10).

To reach the brain, folates must cross the blood-brain barrier (BBB), which is a tissue formed by brain capillary endothelium cells. The BBB is a selective modulator of solute and fluid exchange between blood and brain (11). This tissue regulates the neuronal extracellular environment and protects the central nervous system from neurotoxicants. Barrier function depends primarily on two elements: specific transport proteins expressed in endothelial cell plasma membranes in a polarized manner, and tight junctions that seal the spaces between cells. Transporters function to mediate brain uptake of essential nutrients and ions and remove endogenous substances, neurotoxic compounds, and waste products. Members of ATP-binding cassette (ABC) transporter superfamily, including P-glycoprotein (PGP), multidrug resistance-associated proteins (MRPs), and breast cancer resistance protein (BCRP), are all highly expressed in the BBB, where they function as ATP-driven xenobiotic efflux pumps (12). The BBB is formed by the end of the first trimester during human embryonic development. This barrier is fully developed well before birth, indicating that it may play a significant role in those pathogenetic mechanisms underlying folate-dependent birth defects and pediatric neurological disorders.

These transporters have been linked to cellular folate transport and metabolism in a number of studies (13–16). MRPs and BCRP can transport naturally occurring reduced folates or folate cofactors in several cell lines, including human kidney HEK293 cells (14). Recent evidence demonstrates the physiological role of the ABC transporter family in controlling cellular homeostasis of natural folates. In ABC transporter-overexpressing cells, cellular folate pools were reduced by 40% compared to untransfected cells (16). ABC transporter protein expression was significantly decreased in cells selected for growth under low folate concentrations in leukemia cells (13). These results indicate an important pathophysiological role of the ABC transporters in the regulation of cellular folate homeostasis. However, it is unclear whether abnormal folate levels would affect the BBB transporter function and expression.

Previous studies showed that up-regulation of PGP was mediated by several nuclear receptor-mediated xenobiotic compounds in rat brain capillaries, which functionally tightened the BBB, leading to increased neuroprotection (17–19). In the present study, we demonstrate that exposing isolated mouse brain capillaries to the antiepileptic drug Depakene [valproic acid (VPA)], and dioxin (the most potent inducer previously reported; ref. 18), induces PGP transport activity, resulting in increased protection again toxins. However, these neuroprotective mechanisms are compromised in PCFT-mutant mice. In addition, we show that tight-junction proteins, critical components of the BBB, are disintegrated in these PCFT-mutant mice. Therefore, these findings suggest folate deficiency disrupts the BBB function by targeting the transporter and tight junctions, which may be associated with an increased likelihood of developing neurological disorders.

MATERIALS AND METHODS

Materials

A mouse polyclonal PGP antibody was obtained from Covance (Princeton, NJ, USA), and antibodies against multidrug resistance-associated protein 2 (Mrp2) and BCRP were obtained from Alexis Biochemicals (San Diego, CA, USA). ZO1, claudin 1, and occludin antibodies were purchased from Abcam (Cambridge, MA, USA). The fluorescent substrate NBD-CSA was custom synthesized (20), and PSC833, a specific inhibitor of PGP, was kindly provided by Novartis (Basel, Switzerland). The mouse monoclonal β-actin antibody, Ficoll, VPA, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) stock solution, and all other chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA). All the reagents were of analytical grade or the best available pharmaceutical grade.

Animals

All experiments were performed in compliance with the Institutional Animal Care and Use Committee guidelines and approved by the Animal Care and Use Committee of the University of Texas at Austin. The mice were housed in temperature-controlled rooms under a 12-h light-dark cycle and were given ad libitum access to food and water. The PCFT-mutant mouse colony has been maintained in the Dell Pediatric Research Institute Vivarium for the past 2 yr. Heterozygous mice (PCFT^+/−) were interbred to produce homozygous (PCFT^−/−) mice. PCFT mice were genotyped by screening DNA from tail biopsy samples, as described previously (10). The wild-type and nullizygous PCFT mice (3–4 wk old) were used for brain capillary experiments. For in vivo experiments, PCFT-nullizygous mice treated with S-folinic acid (SFA; 40 mg/kg) were treated for 3–4 wk before capillary collection. Animals were euthanized by CO₂ inhalation followed by decapitation. Brain capillaries were isolated and immediately used for transport experiments; for Western blotting, capillary membranes were prepared and frozen for further analysis.

Capillary isolation

Detailed procedures for capillary isolation were described previously (21, 22). Briefly, white matter, meninges, midbrain, choroid plexus, blood vessels, and olfactory lobes were removed from the brains under a dissecting microscope, and the brain tissue was then homogenized. This homogenate was kept in cold PBS supplemented with glucose and pyruvate (2.7 mM KCl, 1.5 mM KH₂PO₄, 136.9 mM NaCl, 8.1 mM Na₂HPO₄, 1 mM CaCl₂, 0.5 mM MgCl₂, 5 mM d-glucose, and 1 mM sodium pyruvate) throughout the isolation procedure. An aliquot of 30% Ficoll was added to an equal volume of brain homogenate, and capillaries were separated from the parenchyma by centrifuging at 5800 g for 20 min. Capillary pellets were washed with 1% BSA in PBS and passed through a syringe column filled with glass beads. The capillaries bound to the glass beads were released by gentle agitation, then washed with PBS, and used immediately.

Immunohistochemistry

Brain capillaries were fixed in 3% paraformaldehyde/0.25% glutaraldehyde in PBS for 10 min, permeabilized with 0.5% Triton X-100 in PBS for 30 min, and washed with PBS. After blocking with 1% BSA in PBS for 30 min, capillaries were incubated with rabbit anti-PCFT antibody or anti-PGP antibody in PBS with 1% BSA at 37°C for 60 min, washed with PBS with 1% BSA, and then incubated with Alexa Fluor 488- or Alexa Fluor 555-conjugated secondary antibody (1:1000) in PBS with 1% BSA at 37°C for 60 min. After further washing in PBS with 1% BSA, tissue was transferred to chambers for confocal microscopy. Isotype controls were treated identically, except they were not exposed to anti-rabbit IgG primary antibody.

Transport assay

Confocal microscopy-based transport assays with isolated rat brain capillaries have been described previously (21). All experiments were carried out at room temperature in coverslip-bottomed imaging chambers filled with PBS. In general, brain capillaries were exposed to VPA both without and with additional inhibitors. Specifically, the isolated brain capillaries were pretreated with the inhibitor actinomycin D (1 μM) or cycloheximide (100 μg/ml) for 15 min before VPA was added for an additional 2 h. After that, fluorescent substrate NBD-CSA for PGP (21, 22) was added, and luminal substrate accumulation was assessed 1 h later. In each experiment, PSC833 as a specific PGP inhibitor was included in the incubation medium. To acquire images, the chamber containing the capillaries was mounted on the stage of a SP5X inverted confocal laser-scanning microscope and imaged through a ×40 oil-immersion objective (Leica Microsystems, Wetzlar, Germany) using a 488-nm laser line for NBD-CSA. Images were saved to disk and luminal fluorescence was quantitated by ImageJ software (U.S. National Institutes of Health, Bethesda, MD, USA) as before (22). Data shown are for a single experiment that is representative of 3–5 replicates.

Western blots

Membranes were isolated from the brain capillaries of PCFT wild-type and nullizygous mice, as described previously (19, 23). Membrane protein concentrations were assayed by the Bradford method. An aliquot of the membrane protein was mixed with 6× sample buffer, loaded onto 4–12% Bis-Tris gel, electrophoresed, and then transferred to an Immobilon-FL membrane (Millipore, Bedford, MA, USA). The membrane was blocked with Odyssey blocking buffer (Li-Cor Biosciences, Lincoln, NE, USA) at room temperature for 1 h, and then immunoblotted with antibodies against PGP. The membrane was stained with corresponding goat anti-mouse fluorescence dye IRDye 800 in PBS with 0.1% Tween 20 at room temperature for 45 min and then imaged using an Odyssey Infrared Imaging System (Li-Cor Biosciences). β-Actin (42 kDa; 1:2000) was used as a loading control.

Statistical analyses

Data are expressed as means ± se. Statistical analyses of differences between groups was by 1-way ANOVA (Newman-Keuls multiple-comparison test) using Prism 4.0 software (GraphPad, San Diego, CA, USA). Differences between two means were considered significant when P < 0.05.

RESULTS

Presence of PCFT in the BBB

The immunocytochemical results demonstrated the presence of PCFT in the brain capillaries of wild-type mouse brains (3–4 wk). The PCFT staining is visible in the luminal membrane (Fig. 1, arrow). The control experiment, in which there was a lack of primary antibody against PCFT or primary antibody and which was replaced with isotype control, showed no staining (Fig. 1A). PGP has been demonstrated to be highly expressed in the BBB of adult rats (21). By conducting double labeling immunocytochemistry, we demonstrated that both PCFT and PGP were located in capillary membranes (Fig. 1B).

Figure 1.

Localization of PCFT in the BBB. A) Mouse brain capillaries were freshly isolated from PCFT wild-type mouse brains and immunostained with anti-PCFT antibody. Confocal imaging showed the luminal location (arrow) of PCFT in the BBB. Negative control, no primary antibody added. Isotype control, rabbit IgG used as primary antibody. B) Freshly isolated mouse brain capillaries were dual-labeled with anti-PCFT (red) and anti-PGP (green) antibodies by immunocytochemistry. Image (xy section) is a single confocal slice from the center of Z-stack serial images. The xz sections show cross-sections of the capillaries reconstructed from the slices where the arrow points to. Confocal imaging showed PCFT and PGP were localized along the luminal membrane in the BBB. Scale bars = 5 μm.

VPA-induced PGP transport

A transport assay has been established to measure the transport activity of ATP-driven efflux transporters in freshly isolated brain capillaries using fluorescent substrates, confocal microscopy, and digital image analysis. We have previously shown these capillaries to be physiologically active (viable) for >8 h (21, 22). Transport activity is measured as the specific accumulation of a fluorescent substrate in capillary lumens (vascular space). This transport is a robust, specific, metabolism-dependent process. Previous studies have validated that NBD-CSA is a specific fluorescent substrate for PGP transport (21, 22). The quantification of NBD-CSA accumulation in the luminal space is an indicator of PGP-mediated transport activity at the BBB.

Figure 2 shows representative confocal images of mouse brain capillaries incubated to steady state (60 min) with the fluorescent substrate NBD-CSA. Fluorescence was concentrated in the lumen of the control tissue, and luminal fluorescence was greatly reduced when capillaries were exposed to 5 μM PSC833, a specific inhibitor of PGP (Fig. 2A). This concentration of PSC833 causes maximal inhibition of NBD-CSA accumulation in mouse brain capillaries. PGP-mediated transport activity (specific transport) is the difference in steady-state luminal fluorescence in the absence and presence of this specific inhibitor, i.e., accumulation above that measured in capillaries exposed to 5 μM PSC833 (Fig. 2A). Previous studies indicated that residual luminal fluorescence represents nonspecific accumulation, likely from passive diffusion plus nonspecific binding to cellular elements (21). Our initial experiments with xenobiotic-activated nuclear receptor ligands for constitutive androstane receptor (CAR), pregnane X receptor (PXR), and aryl hydrocarbon receptor (AhR), demonstrated concentration-dependent stimulation of PGP transport in mouse brain capillaries (17–19). Cumulative evidence showed that VPA induces PGP and confers pharmacological resistance to therapeutics (24, 25). In the present study, we showed that exposing mouse brain capillaries of wild-type mice to VPA for 3 h increased luminal accumulation of NBD-CSA. The maximum increase occurred at 5 μM, but it reached a plateau within 5–10 μM of VPA exposure (Fig. 2B, P<0.001). The increase in PGP transport activity caused by VPA exposure was abolished by actinomycin D and by cycloheximide, indicating dependence on transcription and translation. Neither actinomycin D nor cycloheximide by itself affected PGP transport (Fig. 3A, B). In wild-type animals, this VPA-induced up-regulation of PGP selectively tightens the BBB, decreasing drug permeability, while greatly increasing the capability of brain in restricting neurotoxic agents from getting into CNS.

Figure 2.

VPA-induced PGP transport in the BBB. A) Representative confocal images showing luminal accumulation of NBD-CSA, a fluorescent substrate for PGP, in freshly isolated mouse brain capillaries. Luminal fluorescence was reduced by 5 μM PSC 833, a specific inhibitor of PGP, and was increased following 3 h of exposure to 10 μM VPA. Scale bar = 5 μm. B) Exposing mouse brain capillaries to VPA increased luminal accumulation of NBD-CSA in a concentration-dependent manner. Values are expressed as means ± se for 8-12 capillaries from a single preparation (each containing pooled brain tissue from 3-5 mice). ***P < 0.001 vs. control.

Figure 3.

PGP transport blocked by transcription or translation inhibitors. Mouse brain capillaries were freshly isolated from PCFT wild-type mouse brains. Inhibiting transcription by 1 μM actinomycin D (ActD; A) or translation by 100 μg/ml cycloheximide (CHX; B) abolished the effect of 5 μM VPA on PGP transport activity. ***P < 0.001 vs. control.

Impaired PGP function in PCFT^−/− mice

The VPA-mediated PGP up-regulation may be associated with multiple cellular homeostasis. Whether folate status plays a role in ABC transporter function is still an unresolved question. Previously, it has been reported that folate concentrations affect PGP transport, and ABC transporters have been shown to transport naturally occurring reduced folates (14). To further confirm the role of VPA in the modulation of PGP transport activity, parallel experiments were conducted in brain capillaries isolated from PCFT-nullizygous mice. Figure 4A shows representative confocal images of brain capillaries freshly isolated from PCFT-nullizygous mice. Once isolated, the capillaries were treated with VPA, and the PGP inhibitor, PSC833. Contrary to what occurred in wild-type mice, VPA failed to induce PGP transport function in the PCFT-mutant mice. The luminal fluorescence signals were similar to control images. PSC833, as a negative control, still caused a significant reduction in PGP transport (Fig. 4A). The quantitation based on luminal fluorescence accumulation showed no significant differences between control and VPA-treated groups (Fig. 4A, right panel), indicating that PGP was impaired at the BBB under folate-deficient conditions.

Figure 4.

PGP transport induction abolished in nullizygous mice. A) Representative images and fluorescence quantitation show that luminal fluorescence accumulation was not altered in freshly isolated brain capillaries of PCFT-nullizygous mice following VPA exposure. Scale bar = 5 μm. B) TCDD (0.1 nM) induced PGP transport in freshly isolated brain capillaries of wild-type mice (left panel), but it was blocked in PCFT-nullizygous mice (right panel). ***P < 0.001 vs. control.

Results from transport assays using TCDD in brain capillaries isolated from wild-type (PCFT^+/+) mice demonstrated a similar induction in PGP transport as that in rat brain capillaries reported previously (18), while this stimulation was abolished in PCFT^−/− mice (Fig. 4B). Therefore, the TCDD-induced PGP transport is eliminated under folate-deficient conditions, similar to the aforementioned VPA-exposed capillary experiments (Fig. 4A). These results using PCFT-nullizygous mice confirm that PGP function is linked to folate status.

To determine whether in vivo folate supplementation restores PGP function, we treated PCFT-nullizygous mice with 40 mg/kg SFA by subcutaneous injection 2×/wk for 2 wk. The body weights of treated mice were significantly increased compared to those without supplemental folate treatment. At the end of SFA treatment, we isolated brain capillaries and treated them with VPA. In SFA-treated animals, exposing capillaries to VPA significantly increased PGP transport compared to the control group (Fig. 5). In vivo SFA-treated PCFT-nullizygous mice showed restored PGP transport function. Taken together, these data indicate PGP transport function is highly related to folate concentrations in the BBB. This is the first demonstration that folate homeostasis alters ABC transporter function.

Figure 5.

PGP transport function recovered in SFA-dosed PCFT-nullizygous mice. PCFT-nullizygous mice (4 wk old) were given 40 mg/kg SFA by subcutaneous injection 2×/wk for 2 wk. Brain capillaries were freshly isolated and treated with VPA (5 μM). Luminal NBD-CSA accumulation was analyzed for PGP transport function. SFA supplementation restored PGP transport function in the BBB. ***P < 0.001 vs. control.

Abnormally low expression of PGP in PCFT^−/− mice

Given that the VPA-mediated PGP transport was dependent on transcription and translation (Fig. 3), the PGP function in the PCFT mouse model may be decreased by a mechanism that involves a decrease in protein expression. This observation was confirmed in the protein expression assays using both wild-type and PCFT-nullizygous mice. Brain capillaries were extracted for Western blot studies of some key ABC transporter proteins, including PGP, Mrp2, and BCRP. The capillary membranes were subjected to Western blot analysis. ABC transporters were highly expressed on the cellular membranes in capillaries. Figure 6A shows a significant difference in PGP protein expression between wild-type and PCFT-nullizygous mice. In the wild-type mice, PGP showed abundant expression, while this expression was extremely low in the PCFT-nullizygous mice, with almost no immunoreactivity. The loading control β-actin was equally expressed when the same PVDF membrane was probed. Thus, it is likely that a significant reduction in PGP expression contributes to the functional changes of PGP in the BBB. Expression of other transporters, such as BCRP, showed a reduction, and an increase in Mrp2 was observed. The expression of tight-junction protein ZO1 significantly decreased in the nullizygous mice (Fig. 6A). Figure 6B shows an induction of PGP expression following exposure of brain capillaries isolated from wild-type mice to VPA (10 μM).

Figure 6.

Western blot of transporter proteins and tight-junction protein in wild type and PCFT-nullizygous mice. A) Antibodies against ABC transporters PGP, Mrp2, BCRP, and tight-junction protein ZO1 were used to probe the membranes for protein expression. Mouse brain capillary membranes were isolated from wild-type (WT) and PCFT-nullizygous (null) mice and analyzed by Western blots. β-Actin was used as a loading control. B) Mouse brain capillaries were isolated from wild-type mice and exposed to VPA (10 μM). Mouse brain capillary membranes were isolated for PGP expression. Representative lots, and quantified band intensities normalized to actin, involving measurement from 3 separate experiments are shown.

Disintegration of tight junctions at the BBB

Mouse brain capillaries were freshly isolated and immunostained with antibodies against tight-junction proteins: ZO1, claudin-1, and occludin. Representative confocal images from immunocytochemistry are shown in Fig. 7. Under the same parameters set up for image acquisition, the fluorescent signals were significantly reduced in PCFT-nullizygous mice as compared to the capillaries isolated from wild-type mice (Fig. 7). With respect to claudin-1 and occludin in control capillaries, fluorescent signals formed a clear line along the capillary membranes, but the line was discontinued in the capillaries of the PCFT-nullizygous mice, indicating the disruption of tight junctions in the BBB (Fig. 7). These data indicate that the neuroprotective mechanisms at the BBB mediated by PGP and tight junctions were compromised in PCFT-nullizygous mice. Thus, BBB permeability is likely increased under folate-deficient conditions, so that neurotoxicants can invade the neural system and cause many neurological diseases associated with birth defects.

Figure 7.

Expression of tight-junction proteins in wild-type and PCFT-nullizygous mice by immunocytochemistry. Brain capillaries were freshly isolated from mouse brains from wild-type or PCFT-nullizygous mice and immunostained with anti-ZO1, anti-claudin-1, or anti-occludin antibody. Tight junctional proteins were significantly impaired in the BBB of PCFT-nullizygous mice compared to control animals.

DISCUSSION

A PCFT-mutant mouse model was established following the creation of gene ablation and was preliminarily characterized with an emphasis on the hematological deficits of the mutant (10). The pathological phenotypes of the nullizygous mice closely mimic the symptoms of patients with hereditary folate malabsorption syndrome, and thus, they serve as a definitive model for this recently described clinical entity. In addition, they are the most accurate model for anemia secondary to folate deficiency, because PCFT-mediated intestinal folate uptake is disrupted in these animals. The animals showed significant reduced folate levels in plasma, as well as other organs in this model, e.g., liver and kidney (10). We have observed severe neurological complications accompanied by low concentrations of folates in PCFT-nullizygous mice, indicating that PCFT plays a critical role in delivering folates to brain. Thus, the PCFT mouse model provides a valuable tool for investigating the role of PCFT in folate transport, absorption, and metabolism. In the present study, we used PCFT-nullizygous mice to create folate-deprived conditions and examined the functional effect on the BBB.

The gene expression pattern of PCFT has been previously reported and appears to be widely expressed in various tissues, including liver, kidney, lung, bone, and heart, as well as in the brain and spinal cord (10). Using immunocytochemical staining and confocal microscopy, we have now demonstrated the presence of PCFT in isolated mouse brain capillary endothelial cells. By conducting 2-color confocal fluorescence microscopy, we confirmed the colocalization of PCFT with the major BBB transporter, PGP. As one of the ATP-driven efflux pumps, PGP is predominantly expressed in the luminal membrane of rat brain capillaries together with other ABC transporters, Mrp2 and BCRP (26). The colocalization of PCFT and PGP in mouse brain capillaries provides the physical and molecular basis for the potential functional interaction between these transporters in the BBB.

Studies using an isolated rat brain capillary preparation have identified and characterized several distinct signaling pathways that regulate the function and expression of PGP at the BBB. The modulators of BBB function typically involve xenobiotics, environmental pollutants, inflammatory factors, reactive oxygen species, HIV-1 viruses, and a host of other factors (26). Among them, the most significant factors that mediated up-regulation of PGP are ligand-activated nuclear receptors, including CAR, PXR, and AhR (17–19). As a consequence, the induction of PGP functionally tightens the BBB, resulting in an increase in expelling neurotoxicants out of the brain.

In the current study, we identified a new regulator of the BBB, the antiepileptic drug VPA, that caused a dose-dependent increase in PGP transport activity in isolated brain capillaries from wild-type mice. As nuclear receptor-mediated regulation, these results suggest a neuroprotective mechanism through which VPA tightens the BBB by inducing PGP transport function. This has potentially important implications in the understanding of the physiological role of ABC transporter proteins in protecting the CNS from cellular toxins. VPA-induced increases in PGP transport activity were abolished when capillaries were exposed to actinomycin D or cycloheximide, indicating a dependence on transcription and translation. These results demonstrate that the BBB is a VPA target tissue. We disclose a mechanism through which antiepileptic drugs alter the normal transport function of the barrier, possibly increasing neuroprotection, but significantly limiting delivery of therapeutic drugs to the brain. This explains the poor drug efficacy of conventional pharmacotherapy in most patients with epilepsy (27). The mechanism of VPA-mediated PGP regulation is consistent with other ligand-activated nuclear receptor-mediated regulation of PGP identified in the BBB. Previously, we showed that exposing rat brain capillaries to CAR ligands, phenobarbital, or phenobarbital-like compounds, induced transport function and protein expression of xenobiotic efflux transporters PGP, Mrp2, and BCRP. VPA, as a histone deacetylase (HDAC) inhibitor, can induce CAR target gene P-450 2B through an epigenetic mechanism by HDAC1 dissociation and the binding of steroid receptor coactivator-1 to CAR (28). Our findings certainly implicate VPA as a new modulator of PGP through direct activation of transcription; however, this does not exclude other possibilities, such as an indirect action through CAR-induced signaling, GSK3β-mediated regulation through β-catenin, or redistribution of PGP due to internalization or compartmentalization. Therefore, VPA, similar to other ligands that activated nuclear receptors, increased PGP transport function, leading to a more strengthened BBB associated with increased pharmacoresistance.

VPA-induced PGP transport was eliminated in PCFT-nullizygous mice, and these animals displayed impaired folate absorption and systemic folate deficiency. To further confirm this, we selected our previously characterized and most potent ligand, TCDD, and conducted the PGP transport assay. Among all the receptor-driven ligands that induce PGP transport function, TCDD (as the high-affinity ligand for AhR), elicits the most potent effect (18). TCDD is a widespread and persistent organic environmental pollutant and is known to more than double the increase in PGP transport in rat brain capillaries. Previous studies demonstrated that TCDD activated AhR at nanomolar levels to target BBB transporters, resulting in an increase in neuroprotection and decreased drug delivery to the brain (18, 29). In the current study, results from TCDD-treated mouse brain capillaries isolated from PCFT-nullizygous mice showed the abolishment of PGP transport induction. In light of previous reports, our study clearly suggests that multiple neuroprotective mechanisms at the BBB were compromised in PCFT-nullizygous mice. Concerning therapeutic interventions, one wonders about the extent to which targeting the BBB itself might improve symptoms in neurological abnormalities associated with folate deprivation.

The BBB is characterized by two important elements: ABC efflux transporters and tight junctions, both of which are critical components of the barrier and form the primary line of defense preventing neurotoxic compounds from entering the CNS. We have demonstrated impaired PGP transport function and abnormally low expression in the BBB of PCFT-nullizygous mice. The tight-junction proteins constitute the other target that can be used to restore the neuroprotective function of the brain. The data in this study demonstrate that BBB permeability is increased under folate-deficient conditions. Alterations in tight-junction protein expression have been reported in various neurological disorders, including multiple sclerosis, stroke, Alzheimer's disease, Parkinson's disease, and epilepsy (30). There are also a number of neurological diseases that are characterized biochemically by excessively low folates or 5-methyltetrahydrofolate (5-MTHF) in the CSF (8, 31, 32). This could be due to a primary genetic disorder that disrupts folate-related metabolic pathways or transport across the blood-CSF barrier. Low CSF folate levels result in seizures and major neurological impairment, a more common clinical sign in several different neurological diseases, including the CFD syndrome. The CFD syndrome refers to a broad spectrum of neurological diseases, including mitochondrial disorders, AGS syndrome (characterized by microcephaly, retardation, and dyskinesia), and hypomyelination with atrophy of the basal ganglia and cerebellum. In addition, there has been some suggestion that this lack of CSF folates may be a contributing factor to autism spectrum disorders, and is often observed among patients with Rett syndrome (characterized by progressive developmental regression after 18 mo) (7). Several folate-related transporters are involved in the processes of folate transport and folate metabolism in these diseases. Mutations in folate receptor α (FRα) were detected in children who were diagnosed as cerebral folate transport deficiency (33, 34). The role that PCFT is playing in CFD is not completely understood; however, it works closely in tandem with FRα by delivering folate from endosomes into the cytoplasm and facilitating FRα-mediated endocytosis (5, 35). Despite the lack of evidence in PGP transporting folates, our current study showed the function of PGP in the BBB is significantly affected by insufficient folate. Further functional characterization of PCFT is needed to address folate transport to the brain.

In most of the less severely affected CFD patients, some form of folinic acid supplementation significantly ameliorated the symptoms and reversed the depletion of 5-MTHF in the brain (34, 36). Our laboratory has tested various folate supplementation strategies to overcome the inherited folate-deficient anemia and increase the length of animal survival in PCFT-nullizygous mice. The results confirmed that folates are the essential nutrients deficient in the PCFT-nullizygous mice and that supplementation is sufficient to rescue these mice from premature death (unpublished observations). In the current study, we observed that SFA restored the function of PGP in the BBB, suggesting that this function may contribute to possible therapeutic strategies in dealing with complications of CNS as a result of folate restriction in neurological deficits.

A number of the ABC transporters are known to adversely affect folate conservation and antifolate excretion. These transporters include MRPs and BCRP, and several of these proteins are also able to transport folates with a low affinity, but at a high capacity (37, 38). Depending on their specific location on polarized cellular membranes, they function to either transport intestinal folates across the epithelia or to reabsorb folates into the intestine (39–41). However, there is no evidence that the expression is altered with folate deficiency in vivo. The present study provides the first demonstration of abnormal protein expression of major ABC transporters at the BBB in response to folate-deficient conditions. Particularly, the dramatic drop of PGP expression in PCFT-nullizygous mice implicates the potentially important correlation of this transporter with folate transport activities. The expression level of PGP may be involved in the regulation of intracellular folate levels through an unidentified mechanism.

In summary, the results of this study indicate that the widely prescribed antiepileptic medication VPA increases PGP transport activity in mouse brain capillaries, leading to increased tightness of the BBB, but this effect is blocked under folate-deficient conditions. Folate supplementation successfully restores the BBB function by increasing VPA-induced PGP transport activities. The multiple neuroprotective mechanisms mediated by ABC transporters and tight-junction proteins appear to be impaired in folate-deficient PCFT mice, providing new insight into understanding of the central role of the BBB in the etiology of neurological disorders.