1α,25-Dihydroxyvitamin D3-liganded vitamin D receptor increases expression and transport activity of P-glycoprotein in isolated rat brain capillaries and human and rat brain microvessel endothelial cells
Matthew R. Durk,
Department of Pharmaceutical Sciences, Leslie Dan Faculty of Pharmacy, University of Toronto, Toronto, Ontario, Canada
Address correspondence and reprint requests to Dr K. Sandy Pang, Leslie Dan Faculty of Pharmacy, University of Toronto, 144 College Street, Toronto, Ontario, Canada M5S 3M2. E-mail: firstname.lastname@example.org
Induction of the multidrug resistance protein 1 (MDR1)/P-glycoprotein (P-gp) by the vitamin D receptor (VDR) was investigated in isolated rat brain capillaries and rat (RBE4) and human (hCMEC/D3) brain microvessel endothelial cell lines. Incubation of isolated rat brain capillaries with 10 nM of the VDR ligand, 1α,25-dihydroxyvitamin D3 [1,25(OH)2D3] for 4 h increased P-gp protein expression fourfold. Incubation with 1,25(OH)2D3 for 4 or 24 h increased P-gp transport activity (specific luminal accumulation of NBD-CSA, the fluorescent P-gp substrate) by 25–30%. In RBE4 cells, Mdr1b mRNA was induced in a concentration-dependent manner by exposure to 1,25(OH)2D3. Concomitantly, P-gp protein expression increased 2.5-fold and was accompanied by a 20–35% reduction in cellular accumulation of the P-gp substrates, rhodamine 6G (R6G), and HiLyte Fluor 488-labeled human amyloid beta 1-42 (hAβ42). In hCMEC/D3 cells, a 3 day exposure to 100 nM 1,25(OH)2D3 increased MDR1 mRNA expression (40%) and P-gp protein (threefold); cellular accumulation of R6G and hAβ42 was reduced by 30%. Thus, VDR activation up-regulates Mdr1/MDR1 and P-gp protein in isolated rat brain capillaries and rodent and human brain microvascular endothelia, implicating a role for VDR in increasing the brain clearance of P-gp substrates, including hAβ42, a plaque-forming precursor in Alzheimer's disease.
low-density lipoprotein receptor-related protein 1
rat/human multidrug resistance protein 1
multidrug resistance-associated protein
pregnane X receptor
quantitative real-time polymerase chain reaction
receptor for advanced glycation endproducts
sodium dodecyl sulfate polyacrylamide gel electrophoresis
tris-buffered saline with 0.1% tween 20
vitamin D receptor
The blood–brain barrier (BBB) presents a major obstacle for drugs targeting the brain parenchyma. Lipophilic drugs should be able to permeate the BBB, but are often pumped out by efflux transporters such as P-glycoprotein (P-gp) and breast cancer resistance protein (BCRP), both expressed at the apical membrane of capillary endothelial cells (Lee et al. 2001). Recent studies show that these transporters are under regulation by nuclear receptors (Tirona and Kim 2005), notably the ligand-activated pregnane X receptor (PXR) (Bauer et al. 2004; Chan et al. 2011), the glucocorticoid receptor (GR) (Narang et al. 2008), the peroxisome proliferator activator receptor α (PPARα) (Hoque et al. 2012), and the constitutive androstane receptor (CAR) (Miller 2010; Wang et al. 2010; Chan et al. 2011). Among the transporters expressed at the BBB, P-gp, which is encoded by the multidrug resistance protein 1 (MDR1) gene in humans or Mdr1a and Mdr1b in rodents, has been most widely studied (Schinkel et al. 1994). P-gp is involved in the transport of a wide variety of substrates, ranging from chemotherapeutic agents to HIV protease inhibitors and immunosuppressants (Tsuji et al. 1993; Kim et al. 1998; Kim 2002), and presents a major challenge for central nervous system pharmacotherapy.
The vitamin D receptor (VDR) is a hormone nuclear receptor that primarily regulates Ca2+ homeostasis and bone resorption upon binding to its natural ligand, 1α,25-dihydroxyvitamin D3 [1,25(OH)2D3], formed via the sequential metabolism (DeLuca, 1976) of dietary vitamin D in the liver (Ponchon et al. 1969) and kidney (Gray et al. 1972). Interestingly, a response element for VDR was identified in the human MDR1 promoter (Saeki et al. 2008). VDR has been shown to play a role in the regulation of MDR1 and P-gp expression in Caco-2 cells (Fan et al. 2009), rat liver and kidney (Chow et al. 2010), and mouse kidney and brain in vivo (Chow et al. 2011). It was further shown that in mice receiving repeated intraperitoneal doses of 1,25(OH)2D3, brain clearance of the P-gp substrate, digoxin, is increased (Chow et al. 2011).
This study extends the findings of Chow et al. (2011). We found, in this study, that isolated rat brain capillaries incubated with 1,25(OH)2D3 exhibit increased expression of P-gp protein and increased P-gp-mediated [N-ε(4-nitrobenzofurazan-7-yl)-D-lys8]-cyclosporine A (NBD-CSA) transport activity. We also showed that exposure to 1,25(OH)2D3 up-regulates the MDR1 gene and P-gp protein expression in brain microvessel endothelial cell lines from rat and human. In both cell lines, exposure to 1,25(OH)2D3 reduced accumulation of the P-gp substrates, rhodamine 6G (R6G), and HiLyte Fluor 488-labeled hAβ42. The findings show that in rodent and human brain capillary endothelial cells, MDR1/P-gp expression is under the direct control of VDR. Our data on HiLyte Fluor 488-labeled hAβ42 accumulation further suggests a role for VDR in reducing the brain's burden of the neurotoxic protein in Alzheimer's disease (AD).
1,25(OH)2D3 was purchased from Sigma-Aldrich Canada (Mississauga, ON, Canada). All reagents for cDNA synthesis and quantitative real-time polymerase chain reaction (qPCR) were obtained from Applied Biosystems (Forster City, CA, USA). Anti-Pgp (C219) and anti-Gapdh (6C5) antibodies were purchased from Abcam (Cambridge, MA, USA). NBD-CSA was custom-synthesized by Novartis (Basel, Switzerland) (Bauer et al. 2004). R6G was purchased from Sigma-Aldrich Canada and HiLyte Fluor 488-labeled human Aβ42, from AnaSpec Inc. (Freemont, CA, USA). PSC833 (valspodar) was a kind gift from Novartis Pharma (Dorval, QC, Canada). All other reagents were obtained from Fisher Scientific (Mississauga, ON, Canada), Invitrogen (Burlington, ON, Canada) or Sigma-Aldrich Canada.
Capillaries were isolated from male Sprague-Dawley rats (270–300 g) as described previously (Bauer et al. 2004). Animal protocols were approved by the Institutional Animal Care and Use Committees of the National Institute of Environmental Health Sciences (NIEHS)/National Institutes of Health (NIH) in accordance with NIH guidelines. Briefly, rats (obtained from Taconic Farms, Germantown, NY, USA) were killed by CO2 inhalation and decapitated. Brains were removed immediately and the cortical gray matter was isolated, weighed, and homogenized gently in a volume 3x its weight with buffer A (103 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl2, 1.2 mM KH2PO4, and 1.2 mM MgSO4 in 15 mM HEPES). The mixture was centrifuged at low speed after the addition of Ficoll (final concentration 30%). The resulting pellet of capillaries was resuspended in buffer B [buffer A plus 25 mM NaHCO3, 10 mM glucose, 1 mM Na-pyruvate, and 0.5% (w/v) bovine serum albumin (BSA)], and filtered through a 200 μm nylon mesh. The filtrate containing the capillaries was passed over a glass bead column and washed with 500 mL buffer. Capillaries adhering to the beads were collected by gentle agitation, then centrifuged, and the pellet was resuspended in ice-cold, capillary phosphate-buffered saline (cPBS) (8 g/L NaCl, 0.2 g/L KCl, 1.15 g/L Na2HPO4, 0.2 g/L KH2PO4, 0.1 g/L CaCl2, 0.1 g/L MgCl2, 0.11 g/L sodium pyruvate, and 0.9 g/L glucose, pH 7.4, sterile filtered) gassed with 95% CO2 and 5% O2 and used immediately.
Capillaries (50 μL in cPBS) were incubated with 0.1% EtOH (vehicle control) and the inducers 1,25(OH)2D3 (1, 10 or 100 nM) for 4 or 24 h, pregnenolone-16α-carbonitrile (PCN) (5 μM) for 4 h or TNF-α (10 ng/mL) for 24 h at 22°C in the dark in a confocal chamber. At the end of the 4 h incubation, some of the prepared capillaries were added to lysis buffer and used for immunoblotting, while the remaining capillaries were used for the determination of P-gp function with the NBD-CSA assay, as described previously (Miller et al. 2000). For the 24 h incubation, capillaries were first incubated in cPBS (containing vehicle or 10 nM 1,25(OH)2D3) and continuously gassed with 95% CO2 and 5% O2 at 22°C in the enclosed teflon incubation chamber. The medium within the incubation chamber was refreshed at the rate of 0.016 mL/min cPBS, pre-gassed with 95% CO2 and 5% O2 and delivered by a peristaltic pump, while excess cPBS was removed continuously by a suction pump. Viability of the 24 h incubation was established by comparing the luminal accumulation of 2 μM NBD-CSA at 24 h (120 ± 5 fluorescence units) with the luminal accumulation in freshly isolated capillaries (119 ± 6 fluorescence units). The inhibitor, PSC833 (5 μM), was found to reduce accumulation of the P-gp substrate to the same extent in both preparations (40% and 35%, respectively, for freshly prepared vs. 24 h incubated capillaries).
Transport was measured as described previously (Miller et al. 2000). Following exposure to 1,25(OH)2D3, PCN, or TNF-α, capillaries were exposed to 2 μM NBD-CSA for another hour in the presence or absence of 5 μM PSC833 at room temperature. Confocal images of 10–15 capillaries were taken; luminal fluorescence was quantified using ImageJ software (NIH, rsbweb.nih.gov/ij/) as previously described (Miller et al. 2000).
The rat brain microvessel endothelial cell line (RBE4) was provided by Dr. F. Roux (Hôpital Fernand Widal, Paris, France); the human brain microvessel endothelial cell line (hCMEC/D3) was provided by Dr. O. Courad (Institut Cochin, Departement Biologie Cellulaire and Inserm, Paris, France). RBE4 cells were maintained in a 50 : 50 mixture of minimal essential medium-alpha and HAM-F10 media, supplemented with l-glutamine, geneticin (G418), gentamicin, fibroblast growth factor, and 10% fetal bovine serum (FBS) (Invitrogen), and grown on rat-tail collagen type I-coated flasks and plates. hCMEC/D3 cells were maintained in basal medium (EGM-2), supplemented with vascular endothelial growth factor, insulin-like growth factor 1, epidermal growth factor, fibroblast growth factor, hydrocortisone, ascorbate, gentamicin (Lonza, Portsmouth, NH, USA), and 5% FBS, and grown on rat-tail collagen type I-coated flasks and plates. Both RBE4 and hCMEC/D3 cell lines were maintained at 37°C under 5% CO2 atmosphere.
Our preliminary studies with vehicle-treated control cells (3.3 × 104 cells/cm2) established that a 5-day incubation period after seeding provided the most stable mRNA expression of MDR1 and Mdr1a/b in hCMEC/D3 and RBE4 cells, respectively (data not shown). Thus, a 1-day exposure to 0 to 100 nM 1,25(OH)2D3 at 4 days after seeding and a 3-day exposure of the cells, beginning at 2 days after seeding, were used. Cells, confluent on the 5th day following plating, were then harvested. We found no toxicity for 1 or 3 day exposure to 1,25(OH)2D3 concentrations ≤ 100 nM in RBE4 and hCMEC/D3 cells using an assay kit that measures leakage of lactate dehydrogenase (LDH) (BioVision, Mountain View, CA, USA), compared with controls treated with vehicle. Moreover, 1,25(OH)2D3 treatment for both the 1- and 3-day treatment did not increase cell permeability to [14C]sucrose nor change the total protein concentration (data not shown). Consequently, 100 nM 1,25(OH)2D3 was chosen as the highest concentration employed for study.
For isolation of total RNA, cells were lysed with 0.5 mL TRIzol and extracted according to the manufacturer's protocol (Invitrogen). Total RNA concentration was determined by UV absorbance and the RNA purity was verified from the absorbance ratio (≥ 1.8) at 260/280 nm and 260/230 nm. Then 1.5 μg RNA was converted to cDNA (High Capacity Reverse Transcriptase Kit; Applied Biosystems, Forster City, CA, USA), followed by qPCR using the SYBR Green detection system (Applied Biosystems 7500 Real-Time PCR System, Streetsville, ON, Canada). Information on the sequences of primers of genes of interest is summarized in Table 1, with primer specificity being verified by BLAST analysis (http://www.ncbi.nlm.nih.gov/BLAST/). For each target gene, the critical threshold cycle (CT) value was determined using the ABI Sequence Detection software version 1.4 (Life Technologies, Carlsbad, CA, USA), with CT values normalized to that of rat/human glyceraldehyde 3-phosphate dehydrogenase (Gapdh/GAPDH). The difference in CT values (∆CT) between the target gene and Gapdh/GAPDH was normalized to the corresponding ∆CT of the vehicle control (∆∆CT) and expressed as fold expression [2−(∆∆CT)] to assess the relative difference in mRNA for each gene.
Table 1. Primer sequences used for qPCR
Forward primer sequence (5′→3′)
Reverse primer sequence (5′→3′)
Cells, grown and treated on a 10-cm tissue culture dish (surface area of 58.2 cm2; BD Falcon, Becton Dickinson, Mississauga, ON, Canada), were scraped in 1.25 mL ice-cold PBS and collected by centrifugation at 500 g for 5 min at 4°C, then resuspended in CellLytic® Lysis Reagent (Sigma-Aldrich), which contained 1% protease inhibitor cocktail (Sigma-Aldrich). The mixture was sonicated three times on ice for 30 s each to yield the resultant lysate and centrifuged at 7500 g for 10 min at 4°C to remove cellular debris. After determination of protein concentration of the supernatant by the Lowry's method (Lowry et al. 1951), 20 μg was used for immunoblotting. The immunoblotting method for microvessel endothelial cells entailed denaturation of 20 μg protein by heating to 37°C for 15 min, followed by separation with sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on 10% acrylamide/bis-acrylamide gel and transference to nitrocellulose membranes (Amersham Biosciences, Piscataway, NJ, USA), as described previously (Chow et al. 2009).
The second immunoblotting method, intended for lysate samples obtained from isolated capillaries, has been described by Bauer et al. (2004). Briefly, lysis buffer was added to isolated capillaries and protein was quantified using the Bradford colorimetric assay; subsequently, 5 μg of protein sample was resolved by SDS-PAGE. For both methods, band intensity was quantified by scanning densitometry (NIH Image software; http://rsb.info.nih.gov/nih-image/). Protein bands of interest were normalized to that of GAPDH.
R6G/HiLyte Fluor 488-hAβ42 accumulation in brain microvessel endothelial cell lines
Following seeding (3.3 × 104 cells/cm2), cells were treated with 1,25(OH)2D3 or vehicle for 1 or 3 days. On the day of study (5 days after seeding), the culture medium was replaced by 0.5 mL of blank, transport buffer (0.01% BSA; Sigma-Aldrich), and 2.38 g/L HEPES in Hank's Balanced Salt Solution, pH 7.4, for 15 min at 37°C (Zastre et al. 2009). The study was initiated upon replacement of 300 μL transport buffer by transport buffer containing (i) 1 μM R6G and 0.1% dimethylsulfoxide (DMSO), or 1 μM R6G and 5 μM PSC833 in 0.1% DMSO or (ii) 0.5 μM HiLyte Fluor 488-hAβ42 and 0.1% DMSO, or 0.5 μM HiLyte Fluor 488-hAβ42 and 5 μM PSC833 in 0.1% DMSO. The plates were incubated at 37°C for 30 min for R6G (Zastre et al. 2009) and 60 min for HiLyte Fluor 488-hAβ42, a predetermined time found to achieve steady-state accumulation (data not shown).
Substrate accumulation was terminated by rapid removal of the transport buffer and rinsing of the wells three times with ice-cold PBS. To assess cellular accumulation of R6G, cells were lysed by adding 130 μL of ice-cold 1% Triton X-100 followed by a brief sonication provided by a hand-held cell disruptor (Caltech Scientific, Mississauga, ON, Canada), and washing of the well with an additional 100 μL of ice-cold 1% Triton X-100. Then 60 μL of the resultant lysate, in triplicate, was used for determination of the R6G concentration at the excitation wavelength of 530 nm and emission wavelength of 560 nm with a SpectraMax Gemini XS (Molecular Devices, Sunnyvale, CA, USA). To assess cellular accumulation of HiLyte Fluor 488-hAβ42, cells were lysed upon addition of 130 μL ice-cold 0.2% SDS and washed with 100 μL of 0.2% SDS. HiLyte Fluor 488-hAβ42 in lysate was determined at the wavelengths of 485 nm for excitation and 535 nm for emission. Quantification of R6G or HiLyte Fluor 488-hAβ42 in cell lysate was enabled by calibration curves that contained standards of known concentrations of R6G or HiLyte Fluor 488-hAβ42 in cell lysate, and processed in identical fashion. Results were normalized to the total protein content (pmol/mg protein), determined by the Bradford colorimetric method using BSA as the standard.
For assessment of P-gp protein in isolated capillaries after 1,25(OH)2D3 treatment, the non-parametric Mann–Whitney test was used to evaluate differences between treatment groups. For the NBD-CSA transport assay, difference in luminal fluorescence between treatment groups was assessed by one-way analysis of variance (anova). Differences between treatment groups in microvessel endothelial cells were evaluated using the Student's two-tailed t-test. A p-value of < 0.05 was considered to be statistically significant.
1,25(OH)2D3 exposure increases P-gp protein levels and transport function in isolated rat brain capillaries
Previous studies have shown that P-gp transport activity in isolated rodent brain capillaries could be assayed using NBD-CSA, a fluorescent P-gp substrate, with confocal microscopy to measure luminal accumulation of the substrate, in the presence or absence of PSC833, a P-gp inhibitor (Miller et al. 2000). In this assay, transporter activity is given as the difference in luminal fluorescence without and with PSC833. Exposing isolated rat brain capillaries to 10 or 100 nM 1,25(OH)2D3 for 4 or 24 h nearly doubled specific transporter activity (Fig. 1a). Additionally, exposure of capillaries to 5 μM PCN for 4 h or 10 ng/mL TNF-α for 24 h increased P-gp function to a similar extent as 1,25(OH)2D3. These effects were comparable to those found when transporter expression was increased through PXR activation by PCN (Bauer et al. 2006) and TNF-α (Bauer et al. 2007). Consistent with increased P-gp transport activity, 4 h exposure of brain capillaries to 10 nM 1,25(OH)2D3 quadrupled P-gp protein expression (Fig. 1b).
1,25(OH)2D3 exposure up-regulates Mdr1b mRNA and P-gp protein in RBE4 cells and in hCMEC/D3 cells
Exposing RBE4 cells for 1 or 3 days to 1,25(OH)2D3 increased the mRNA expression of VDR and Cyp24, known VDR targets (Jones and Tenenhouse 2002), in a concentration-dependent manner. 1,25(OH)2D3 exposure also increased mRNA expression of Mdr1b in a concentration-dependent manner, though Mdr1a expression was unchanged (Fig. 2a). Consistent with these data, P-gp protein was increased 2.5-fold following 3 days of 100 nM 1,25(OH)2D3 exposure (Fig. 2b). Similarly, in hCMEC/D3 cells, 1 and 3 day exposure to 10 and 100 nM 1,25(OH)2D3 increased Cyp24 mRNA expression; VDR expression was unaffected (Fig. 3a). In addition, 1 and 3 day exposure to 10–100 nM 1,25(OH)2D3 increased MDR1 mRNA (Fig. 3a). P-gp protein levels were increased threefold after 3 days exposure to 100 nM 1,25(OH)2D3, (Fig. 3b).
1,25(OH)2D3 exposure decreases R6G accumulation in RBE4 and hCMEC/D3 cells
For the fluorescent P-gp substrate, R6G (Zastre et al. 2009), PSC833 (5 μM), the P-gp inhibitor, doubled R6G accumulation in both RBE4 and hCMEC/D3 cells, observations that are consistent with inhibition of P-gp (Fig. 4). In both cell types, exposure to 100 nM 1,25(OH)2D3 tended to decrease R6G accumulation after 1 day and significantly decreased R6G accumulation after 3 days, by 35 and 25%, respectively, for RBE4 and hCMEC/D3 cells (Fig. 4). In cells treated with PSC833, 1,25(OH)2D3 exposure had no effect on R6G accumulation; P-gp inhibition canceled the increase in transport elicited by 1,25(OH)2D3 treatment.
1,25(OH)2D3 exposure decreases HiLyte Fluor 488-hAβ42 accumulation in RBE4 and hCMEC/D3 cells
Research within the last decade has implicated a role for P-gp (Lam et al. 2001) in the brain efflux of amyloid beta (Aβ), a neurotoxic protein and a key player in AD and in cerebral amyloid angiopathy. Aβ is formed in the brain by cleavage of the amyloid precursor protein (APP) by β- and γ-secretases, forming peptide segments of different lengths, with the 40 (Aβ40) and 42 (Aβ42) amino acid fragments being the most common (Hartmann et al. 1997). When allowed to accumulate in the brain, these peptides aggregate and form amyloid plaques, a hallmark of AD (Querfurth and LaFerla 2010). In particular, Aβ42, owing to its hydrophobicity, is considered more pathogenic and constitutes a major component of these plaques (Miller et al. 1993). It has been proposed that brain accumulation of Aβ in AD is a result of its decreased efflux and not increased synthesis from the brain, as summarized in the amyloid clearance hypothesis (Zlokovic et al. 2000). Brain accumulation of Aβ is modulated by the P-gp (Cirrito et al. 2005; Hartz et al. 2010), the low-density lipoprotein receptor-related protein 1 (LRP1), and the receptor for advanced glycation endproducts (RAGE) (Deane et al. 2004), balancing efflux (LRP1 and P-gp) and influx (RAGE), respectively. Upon analyses of these proteins by immunoblotting, neither RAGE nor LRP1 protein was found to be altered in hCMEC/D3 cells following 1,25(OH)2D3 treatment (data not shown).
In both RBE4 and hCMEC/D3 cells, PSC833 doubled accumulation of HiLyte Fluor 488-hAβ42, an observation consistent with P-gp-mediated hAβ42 efflux (Fig. 5). Exposing these cells to 100 nM 1,25(OH)2D3 for 3 days significantly decreased HiLyte Fluor 488-hAβ42 accumulation (p <0.05). In the presence of 5 μM PSC833 and 100 nM 1,25(OH)2D3, HiLyte Fluor 488-hAβ42 accumulation returned to the level seen with PSC833 alone (Fig. 5). These results indicate that increasing P-gp expression through VDR is an effective way to reduce hAβ42 accumulation.
Mdr1/MDR1 and its gene product, P-gp, are regulated by numerous ligand-activated nuclear receptors (Timsit and Negishi 2007). At the BBB, recent studies suggest that this is the case for P-gp. In the hCMEC/D3 cell line (derived from human brain endothelial cells), CAR and PXR ligands increase MDR1 expression (Chan et al. 2011). In intact rats in vivo and isolated rat and mouse brain capillaries, BBB P-gp expression and transport activity were found to be increased by ligands for PXR and CAR (Bauer et al. 2004; Wang et al. 2010). Dexamethasone, acting through the GR, also increases Mdr1 mRNA, P-gp protein and transport function in primary microvascular endothelial cells isolated from rats (Narang et al. 2008). In contrast to the promiscuity of these nuclear receptors, the interaction of 1,25(OH)2D3-liganded VDR with its brain target, MDR1, is fairly specific in that the binding of VDR to 1,25(OH)2D3 occurs at high affinity at the pM range (Coty 1980). Our present data confirmed that VDR activation occurs following 10–100 nM 1,25(OH)2D3 treatment and specifically up-regulates rodent/human Mdr1/MDR1 and P-gp only and not Mrp1/MRP1 or Bcrp/BCRP, transporters which are also present in the cell lines examined (Weksler et al. 2005). In addition, neither LRP1 nor RAGE protein expression was altered (data not shown).
There exists strong molecular evidence to suggest that the VDR activates the MDR1 gene in multiple tissues and among various animal species. Analysis of VDR response elements (VDREs) on the human MDR1 gene (Saeki et al. 2008) revealed that up-regulation of MDR1 by 1,25(OH)2D3 is likely a direct effect at the transcriptional level, as shown in Caco-2 cells (Schmiedlin-Ren et al. 1997). For the kidney, a VDR-rich organ, increased renal P-gp expression because of VDR activation would increase the renal clearance and lower the systemic concentrations of P-gp substrates. In vivo, 1,25(OH)2D3 treatment increased P-gp expression in rat (Chow et al. 2010) and mouse (Chow et al. 2011) kidney, affecting the renal clearance of the P-gp substrate, digoxin. Additionally, reduced accumulation of digoxin was observed in mouse brain (Chow et al. 2011). Results obtained in the brain-specific systems in this study support the idea that VDR activation can increase P-gp expression and transport activity at the BBB. We further verified the involvement of VDR by showing up-regulation of the VDR target gene, Cyp24/CYP24.
This role of VDR in increasing P-gp function at the BBB is of clinical importance because many transporter substrates are therapeutic drugs (Doran et al. 2005; Marquez and Bambeke 2011) and because P-gp appears to contribute to Aβ efflux from the brain (Cirrito et al. 2005; Hartz et al. 2010). Typically, vitamin D analogs are given to increase Ca2+ resorption in patients suffering from renal failure (Slatopolsky and Brown 1997) or as adjunct therapy to treat cancer because of the anti-proliferative effects which they exert (Bortman et al. 2002; Muindi et al. 2002). In addition to increasing P-gp function, 1,25(OH)2D3-liganded VDR activity further affects other transporters and enzymes, including the induction of human intestinal and/or liver CYP3A4 (Schmiedlin-Ren et al. 1997; Thummel et al. 2001; Khan et al. 2009), sulfotransferase 2A1 (SULT2A1) (Echchgadda et al. 2004), OATP1A2 (Eloranta et al. 2012), the folate transporter (Eloranta et al. 2009), and MRP4 (Fan et al. 2009; Maeng et al. 2012). The wide range of enzymes and transporters that are VDR targets suggests that drug–drug interactions may be more prevalent in patients in which VDR is targeted for therapy (Kota et al. 2011).
Finally, this study suggests that, through induction of BBB P-gp, activation of VDR could lead to increased clearance of hAβ from the brain. Increasing BBB P-gp expression through activation of PXR has already been shown to decrease brain hAβ levels in transgenic mice expressing mutant human APP (Hartz et al. 2010). Efflux of this toxic peptide appears to be mediated in part by LRP1 at the abluminal membrane of the brain capillary endothelial cells and by P-gp at the luminal membrane (Lam et al. 2001; Deane et al. 2004), though we failed to detect any change in LRP1 expression following 1,25(OH)2D3 treatment. Induction of P-gp may not be the only VDR-mediated mechanism of Aβ reduction. Ito et al. (2011) also observed higher increased cerebral clearance of Aβ40 following 1,25(OH)2D3 treatment, but had not attributed their findings to P-gp induction as the underlying mechanism. Rather, they attributed the observation to both the genomic and non-genomic actions of VDR, citing the confounding factor of lower LRP1 levels (51%) in cerebral vessels of mdr1a/b knockout mice (Ohtsuki et al. 2010); yet, it was later demonstrated that inhibition of LRP1 did not alter cerebral clearance of Aβ40 (Ito et al. 2010). Others have provided additional, plausible mechanisms in which 1,25(OH)2D3 treatment stimulates macrophages, thus promoting Aβ clearance from the brain as part of an immune response (Masoumi et al. 2009; Mizwicki et al. 2012). The present findings suggest that targeting VDR may present opportunities for the development of alternative therapies or preventative measures in response to the onset and progression of AD.
The authors have no conflict of interest to declare. This work was supported by the Canadian Institutes of Health Research (CIHR) and by the Intramural Research Program of the National Institute of Environmental Health Sciences, NIH. Matthew R. Durk is a recipient of a CIHR Strategic Training Grant in Biological Therapeutics (CIHR).