ATP-binding cassette (ABC) efflux transporters, including multidrug resistance protein 1 (Mdr1), breast cancer resistance protein (Bcrp), and multidrug resistance-associated proteins (Mrps) extrude chemicals from the brain. Although ABC transporters are critical for blood-brain barrier integrity, less attention has been placed on the regulation of these proteins in brain parenchymal cells such as microglia. Prior studies demonstrate that inflammation after lipopolysaccharide (LPS) treatment alters transporter expression in the livers of mice. Here, we sought to determine the effects of inflammation on the expression and function of transporters in microglia. To test this, the expression and function of ABC efflux transport proteins were quantified in mouse BV-2 microglial cells in response to activation with LPS. Intracellular retention of fluorescent rhodamine 123, Hoechst 33342, and calcein acetoxymethyl ester was increased in LPS-treated microglia, suggesting that the functions of Mdr1, Bcrp, and Mrps were decreased, respectively. LPS reduced Mdr1, Bcrp, and Mrp4 mRNA and protein expression between 40 and 70%. Conversely, LPS increased expression of Mrp1 and Mrp5 mRNA and protein. Immunofluorescent staining confirmed reduced Bcrp and Mrp4 and elevated Mrp1 and Mrp5 protein in activated microglia. Pharmacological inhibition of nuclear factor κB (NF-κB) transcriptional signaling attenuated down-regulation of Mdr1a mRNA and potentiated up-regulation of Mrp5 mRNA in LPS-treated cells. Together, these data suggest that LPS stimulates microglia and impairs efflux of prototypical ABC transporter substrates by altering mRNA and protein expression, in part through NF-κB signaling. Decreased transporter efflux function in microglia may lead to the retention of toxic chemicals and aberrant cell-cell communication during neuroinflammation.
ATP-binding cassette (ABC) efflux transporters are a large group of transmembrane proteins responsible for the efflux of endogenous substances and xenobiotics. Although a significant amount of research has focused on the expression and regulation of these transporters in the liver and kidneys (reviewed in Klaassen and Aleksunes, 2010), research has more recently focused on the localization and function of these transporters in the brain (reviewed in Dallas et al., 2006). In the brain, efflux proteins are expressed on brain capillary endothelial cells and are critical in the blood-brain-barrier extrusion of toxic chemicals and drugs from the parenchyma (Warren et al., 2009; Elali and Hermann, 2011). Members from major classes of transporters [multidrug resistance protein 1 (MDR1/ABCB1), multidrug resistance-associated protein 1-5 (MRP1-5/ABCC1-5), and the breast cancer resistance protein (BCRP/ABCG2)] have been found in one or more cell types in the brain, including endothelial cells, neurons, astrocytes, and microglia (reviewed in Klaassen and Aleksunes, 2010).
Microglia are the resident immune cells of the central nervous system and have important responsibilities, including neuron support and nutrition as well as immune surveillance. In adult animals, microglia are generally present in a stationary, ramified state and continuously monitor the neural parenchyma through finger-like cytoplasmic projections. Microglia can be stimulated by infections, cerebral ischemia, traumatic brain injury, or neuronal/parenchymal damage to remove necrotic and apoptotic debris and coordinate immune responses (Rogers et al., 2007; Badoer, 2010; Cao et al., 2011a). Activated microglia progressively increase in number during the normal aging process and are also commonly present in neurodegenerative diseases. Activation of microglia is characterized by the production of proinflammatory cytokines as well as morphological changes, including retraction of cell processes, increased size, and an amoeboid shape (Kim and Joh, 2006). The degree of activation, localization, and types of cytokines produced varies by the type of insult. After resolution of an acute response, some microglia undergo apoptosis, whereas others return to the ramified state (Cao et al., 2011a). However, under certain circumstances, including neurodegenerative diseases, microglia can persist in an activated state (Kim and Joh, 2006; Yasuda et al., 2007; Long-Smith et al., 2009; Vázquez-Claverie et al., 2009). A role for these transporters in neurodegenerative diseases has been postulated because some ABC isoforms can efflux mediators of neuronal injury such as β-amyloid (Hartz et al., 2010). Likewise, the expression of ABC transporters can be modulated by inflammation. For example, the hepatic expression and function of Mdr1a/1b are reduced in rodent models of inflammation using exogenous administration of lipopolysaccharide (LPS) or proinflammatory cytokines (Ando et al., 2001; Hartmann et al., 2001).
With recent evidence suggesting that microglia play a prominent role in brain inflammation and neurodegenerative diseases, we hypothesized that ABC transporter function is disrupted in activated microglia, which could alter cell-cell communication and chemical sequestration in the brain. BV-2 cells are immortalized mouse microglia that produce a wide variety of cytokines and other substances, similar to in vivo inflammatory responses, making them an ideal tool for mechanistic studies (Henn et al., 2009; Cao et al., 2011b). One signaling pathway implicated in microglial activation is the transcription factor nuclear factor κB (NF-κB) (Cao et al., 2011b), which has also been shown to regulate expression of efflux transporters (Ronaldson et al., 2010; Yu et al., 2011). Therefore, the purpose of the present study was to 1) quantify ABC transporter function and expression in BV-2 microglia cells, 2) investigate whether transporter function and expression are altered in response to activation with LPS, and 3) determine whether inhibition of NF-κB signaling alters transporter mRNA regulation in activated microglia.
Materials and Methods
Unless otherwise specified, chemicals were obtained from Sigma-Aldrich (St. Louis, MO).
The immortalized mouse (C57BL/6) microglia BV-2 cell line was used for all experiments (Henn et al., 2009). Cells were grown in modified Eagle's medium (Mediatech, Herndon, VA) supplemented with 10% fetal bovine serum (Atlanta Biologicals, Norcross, GA), 5% sodium pyruvate (Mediatech), 5% nonessential amino acids (Mediatech), and 1% penicillin-streptomycin (Invitrogen, Carlsbad, CA). Cells were incubated at 37°C with 5% CO2. Cells were used in experiments when they reached 80 to 90% confluence. BV-2 cells were activated with LPS from Escherichia coli 026:B6 (Sigma-Aldrich) for up to 24 h. LPS was dissolved in sterile water at a concentration of 1 mg/ml (stock solution). An equal volume of sterile water was added to all control treatments.
Cell morphology was evaluated visually with phase contrast imaging using an inverted microscope (Zeiss Axio Observer; Carl Zeiss Inc., Thornwood, NY). Cell morphology was also characterized in real-time by using an xCELLigence RTCA DP analyzer (Roche Diagnostics, Pleasanton, CA). In brief, BV-2 cells were trypsinized, counted, and seeded evenly (5 × 103 cells) into each well of an E-Plate 16 (Roche Diagnostics). Impedance measurements were recorded every minute during the first 2 h of the attachment phase and then every 15 min overnight. The next morning, medium was removed and cells were treated with either fresh medium containing vehicle (sterile water) or LPS (100 ng/ml) (0-h time point). Impedance was measured every minute for 100 min and then every 15 min for 24 h. Impedance measurements were recorded as a combined cellular index of proliferation, viability, and morphology changes.
Quantitative Transporter Function.
Transporter function was quantitated on a Cellometer Vision cell counter (Nexcelom Bioscience LLC., Lawrence, MA) using fluorescent substrates. The technique was adapted from Robey et al. (2011). In brief, cells were grown in T75 flasks to 80% confluence, trypsinized, and seeded into round bottom wells of a 96-well plate. Cells were then centrifuged at 500g for 5 min at 5°C, resuspended, and loaded with fluorescent substrate with or without inhibitor for 30 min at 37°C and 5% CO2 (uptake period). Fluorescent substrate and inhibitor concentrations are provided in Table 1. All transporter substrates and inhibitors were initially dissolved in dimethyl sulfoxide (DMSO). Equal volumes of DMSO were added to all controls when appropriate (final concentration of DMSO was less than 0.5%). Cells were then washed, centrifuged, resuspended, and incubated in substrate-free medium in the presence and absence of inhibitor for 1 h (efflux period). Cells were then washed, resuspended in cold phosphate-buffered saline (PBS), and fluorescence was quantified by using the Nexcelom Vision cellometer. Twenty microliters of cell suspension were applied to the cell counting chamber, and each sample was analyzed by using the bright-field images for cell size and cell number. The intensity of fluorescence for each cell was subsequently analyzed by using filter cubes VB-450-302 (Hoechst 33342; excitation/emission: 375/450 nm), VB-535-402 (calcein, BODIPY, and NBD; excitation/emission: 475/535 nm), or VB-595-502 (Rhodamine 123; excitation/emission: 525/595 nm). The total number of cells analyzed for each sample ranged from 200 to 2000. Raw fluorescence intensity for each cell was normalized to cell size to account for any changes in cell size during microglial activation. The average fluorescence for each sample was determined, and the treatment average was based on four independent samples for each treatment.
RNA Isolation and Real-Time Quantitative PCR.
Cultured BV-2 cells (5 × 105 cells/well) were plated onto six-well tissue culture plates and incubated overnight until attachment. Quadruplicates were treated with vehicle or 1, 10, 100, or 500 ng/ml LPS in fresh medium. After 12 h of incubation, cell cultures were washed twice with PBS and lysed with buffer RLT (QIAGEN, Valencia, CA) containing 1% β-mercaptoethanol. Total RNA was isolated by using the RNeasy mini kit (QIAGEN). cDNA was generated with the First Strand SuperScript cDNA synthesis kit (Invitrogen). cDNA purity and concentration were assessed by using a Nanodrop 1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA). For qPCR, specific forward and reverse primers (Integrated DNA Technologies, Inc., Coralville, IA) for each gene were added to 1 μg of cDNA from each sample. Sybr Green (Applied Biosystems, Foster City, CA) was used for detection of amplified products. qPCR was performed in a 384-well plate format by using the ABI 7900HT PCR system (Applied Biosystems). Ct values were converted to ΔΔCt values by comparison to a reference gene, ribosomal protein 13a (Rpl13a). Primer sequences for all target mRNAs are provided in Supplemental Table 1.
Western Blot Analysis.
BV-2 cells (1 × 106 cells) were plated onto 60-mm cell culture dishes and allowed to attach overnight. Triplicates were treated with fresh media containing vehicle or 1, 10, 100, or 500 ng/ml LPS and incubated at 37°C for 24 h. Cells were then washed twice in PBS. Cells were collected, transferred to a microcentrifuge tube, and centrifuged at 1000g for 10 min. The supernatant was removed, and the pellet was resuspended in cell lysis buffer with protease inhibitors. Protein concentration was determined by using the bovine serum albumin protein kit (Thermo Fisher Scientific). Western blot analysis of ABC transporters was performed by using 50 μg of protein homogenate per well by SDS-polyacrylamide gel electrophoresis on NuPage 8% gels (Invitrogen). Proteins were transferred overnight to a polyvinylidene fluoride membrane. Membranes were blocked in 5% nonfat dry milk in PBS with 0.5% Tween 20. Primary antibodies were diluted in 2% nonfat dry milk and incubated with membranes for 3 h at the following concentrations: 1:1000 for Mdr1a/b (C219; Novus Biologicals, Inc., Littleton, CO), 1:2000 for Mrp1 (MRPr1; Enzo Life Sciences, Inc., Farmingdale, NY), 1:2000 for Mrp4 (M4I-80; Enzo Life Sciences, Inc.), 1:500 for Mrp5 (M5I-10; G. Scheffer, VU Medical Center, Amsterdam, The Netherlands), 1:5000 for Bcrp (BXP-53; Enzo Life Sciences, Inc.), and 1:5000 for Gapdh (Millipore Corporation, Billerica, MA). Primary antibodies were probed by using species-appropriate secondary antibodies (Sigma-Aldrich) and detected by using SuperSignal West Dura blotting reagents (Thermo Fisher Scientific). Detection and semiquantification of protein bands was performed with a FluorChem imager (Alpha Innotech, San Leandro, CA).
Indirect Immunofluorescent Staining.
BV-2 cells (5 × 105 cells) were seeded into chambers of a chamber slide (Thermo Fisher Scientific) and allowed to attach overnight. The next day, cells were washed twice in PBS. Chambers were either filled with medium alone or medium plus 100 ng/ml LPS and incubated for 24 h. Slides were fixed with 4% paraformaldehyde, washed twice in PBS and once in PBS with 0.1% Triton X-100, and then blocked with 5% goat serum in PBS with 0.1% Triton X-100. Slides were then incubated in primary antibody or 5% goat serum, washed, incubated again with the secondary antibody-conjugated to AlexaFluor 488 (Life Technologies, Carlsbad, CA), washed, and dried. Slides were coverslipped with Prolong Gold containing 4′,6-diamidino-2-phenylindole (Life Technologies). Images were acquired on a Zeiss Observer D1 microscope with an X-Cite series 120Q fluorescent illuminator and a Jenoptik camera with ProgRes CapturePro 2.8 software (Jenoptik, Easthampton, MA). Images were cropped, and brightness and contrast were adjusted equally in Adobe Photoshop CS2 (Adobe Systems, San Jose, CA). All sections were stained and imaged under uniform conditions for each antibody. Negative controls without primary antibody were included to ensure minimal nonspecific staining (data not shown).
Cultured BV-2 cells (5 × 105 cells/well) were preincubated with vehicle (DMSO) or the NF-κB inhibitor (5 μM) 3-(4-methylphenylsulfonyl)-2-propenenitrile (BAY 11-7082) (EMD Biosciences, San Diego, CA) for 30 min in triplicate. After preincubation, cells were treated with vehicle or 100 ng/ml LPS in the absence or presence of BAY 11-7082. After 12 h of incubation, cell culture medium was removed and frozen for quantification of tumor necrosis factor α (TNFα) protein. Cell cultures were washed and lysed. RNA isolation, cDNA synthesis, and qPCR were performed as described above.
TNFα Protein Quantification.
TNFα levels were quantified in cell culture medium by using a mouse enzyme-linked immunosorbent assay kit (Invitrogen) according to the manufacturer's instructions.
Prism version 5 software (GraphPad Software, Inc., San Diego, CA) was used for statistical analysis. Differences among groups were determined by either a two-tailed t test when only two groups were compared or a one-way analysis of variance with a Newman-Keul's post hoc test for multiple group comparisons. Differences were considered statistically significant at p < 0.05.
Basal Expression and Function of ABC Efflux Transporters in BV-2 Cells.
Resting BV-2 microglia express ABC transporter mRNA including Mrp1–5, Mdr1a, Mdr1b, and Bcrp. mRNA expression in BV-2 cells was compared with mouse liver after normalizing to the reference gene Rpl13a (Table 2). The expression of transporter mRNA differed between BV-2 microglia and mouse liver. After normalizing to Rpl13a, BV-2 microglia had negligible amounts of Mrp2 mRNA compared with mouse liver. BV-2 cells also expressed lower levels of Mrp3 and Bcrp mRNA. In contrast, microglia expressed high levels of Mrp1, Mrp4, Mrp5, Mdr1a, and Mdr1b mRNA compared with mouse liver. Basal transporter function was quantified in control BV-2 cells by using a fluorescent cell counter (Fig. 1) as described previously (Robey et al., 2011). A panel of fluorescent chemicals (rhodamine, calcein AM, and Hoechst 33342) and inhibitors against various ABC transporters were tested (Table 1). Inhibition of Mdr1 with 6-[(2S,4R,6E)-4-methyl-2-(methylamino)-3-oxo-6-octenoic acid]-7-l-valine-cyclosporin A (PSC833) significantly increased the accumulation of rhodamine 123 in BV-2 cells by 84%. Compared with control cells, 5-(3-(2-(7-chloroquinolin-2-yl)ethenyl)phenyl)-8-dimethylcarbamyl-4,6-dithiaoctanoic acid sodium salt (MK-571), a general inhibitor of Mrps, significantly enhanced calcein AM fluorescence by 108%. The Bcrp inhibitor (3S,6S,12aS)-1,2,3,4,6,7,12,12a-octahydro-9-methoxy-6-(2-methylpropyl)-1,4-dioxopyrazino[1′,2′:1,6]pyrido[3,4-b]indole-3-propanoic acid 1,1-dimethylethyl ester hydrate (KO143) significantly increased Hoechst 33342 levels by 51% (Fig. 1A). Images demonstrating fluorescent dye retention in BV-2 cells in the presence of prototypical inhibitors are shown in Fig. 1B.
Lipopolysaccharide Activation of BV-2 Microglia.
Altered BV-2 cell morphology in response to LPS was assessed quantitatively by changes in impedance with the use of real-time measurements (Fig. 2A) and qualitatively by light microscopy (Supplemental Fig. 1). Greater electrical impedances across interdigitating microelectrodes on the bottom of the culture plate reflect collective changes in cell morphology, viability, and proliferation. As cells occupy a larger surface area on the electrodes, the path that a small electrical current must travel through the cell membrane is longer, which results in increased electrical impedance. In the present study, the early attachment and longer growth phases of plated BV-2 cells were clearly observed with slow, gradual increases in impedance over 24 h (time = −24–0 h). Upon addition of 100 ng/ml LPS (time = 0 h), an immediate and dramatic change in the impedance was detected in LPS-treated cells compared with control cells. The peak of the impedance difference between control and LPS-treated cells occurred 4 h after the addition of LPS. Morphological changes were confirmed at 6 h with phase-contrast microscopy (Supplemental Fig. 1, A and B). BV-2 cells without exposure to LPS were 10 to 15 μm in diameter, uniformly round, with occasional thin cytoplasmic projections. After LPS activation, BV-2 cells were much larger in diameter (25–30 μm), polymorphic, with many exhibiting thick cytoplasmic projections. Compared with vehicle-treated cells, incubation with 100 ng/ml LPS for 24 h did not alter cell number or cell viability as determined by the alamar blue assay (data not shown). These data suggest that the quantitative differences in cell impedance were largely caused by the observed morphologic changes. By 20 h, cell impedance values were similar between control- and LPS-treated cells. In addition to morphologic changes, activated microglia produced proinflammatory cytokines including TNFα, interleukin (IL)-1β, and IL-6. mRNA levels of TNFα, IL-1β, and IL-6 were dose-dependently increased in BV-2 cells in response to LPS at 12 h (Fig. 2B). Stimulation of TNFα cytokine mRNA expression was achieved even at the lowest concentration of LPS (1 ng/ml).
Activation of Microglia Decreases Efflux Transporter Function and Increases Intracellular Retention of Substrates.
Activation of BV-2 microglia with 100 ng/ml LPS for 24 h enhanced intracellular accumulation of all fluorescent substrates tested (Fig. 3). Retention of rhodamine 123, an Mdr1 substrate, was increased approximately 25% in LPS-activated microglia (Fig. 3A). The distribution curve shows a larger percentage of cells had more fluorescence indicated by a right shift in the curve. Retention of calcein AM, an Mrp substrate, increased 110% in LPS-treated microglia. Most control BV-2 cells had minimal calcein AM fluorescence, suggesting efficient efflux during the 1-h incubation period. In contrast, the majority of activated microglia exhibited strong retention of calcein AM, indicated by a right shift in the distribution of cellular fluorescence. Hoechst 33342, a Bcrp substrate, also showed greater average fluorescence and retention in activated microglia, although at a much lower degree than the other efflux transporters. Two additional chemicals, including a pharmaceutical and an endogenous signaling molecule, were included as examples of relevant ABC efflux substrates. Efflux of the α-adrenergic antagonist prazosin and the proapoptotic molecule ceramide was also inhibited by LPS treatment (Fig. 3B). Mean fluorescence for prazosin and ceramide increased 39 and 90%, respectively, after microglial activation with LPS. Images demonstrating fluorescent dye retention in control- and LPS-treated BV-2 cells are shown in Supplemental Fig. 2. Cellular loading and uptake of all substrates tested was not altered by LPS treatment (Supplemental Fig. 3), demonstrating that changes in the cellular retention of substrates in response to LPS are caused by differences in chemical efflux rather than uptake.
Decreased Efflux Transporter Function Is Paralleled by Down-Regulation of Transporter mRNA.
mRNA expression of ABC transporters was quantified after microglial activation with LPS treatment (Fig. 4). Transporter mRNA expression of most efflux transporters decreased by 12 h after initiation of LPS treatment. Mrp2 and Mrp3 mRNA decreased 60% at the lowest concentration (1 ng/ml LPS) with no further declines observed at higher LPS concentrations. All other efflux transporters exhibited concentration-dependent changes with increasing LPS concentrations. In BV-2 cells treated with 500 ng/ml LPS levels of Mrp1 and Mrp5 mRNA rose by 150 and 300%, respectively, whereas Mdr1a, Mdr1b, and Bcrp mRNAs were reduced by 70, 75, and 55%, respectively.
Transporter Protein Expression Is Consistent with Altered Function and mRNA Levels.
Western blot analysis of transporter proteins showed high basal protein expression of Mdr1, Bcrp, Mrp1, and Mrp4 in cultured BV-2 microglia (Fig. 5). Mouse liver, kidney, and placenta homogenates were used as positive controls for each transporter (data not shown). Mrp5 basal protein expression was relatively low. Upon LPS stimulation of BV-2 microglia for 24 h, Mrp4, Mdr1, and Bcrp proteins were significantly decreased by up to 55, 42, and 42%, respectively. Mrp1 and Mrp5 proteins were elevated up to 140 and 210%, respectively, after LPS treatment. Protein expression for all transporters followed the same pattern as their respective mRNAs. It is noteworthy that Mrp2 and Mrp3 proteins were neither present in control cells nor induced by LPS (data not shown). Qualitative transporter expression was assessed in paraformaldehyde-fixed BV-2 cells by using specific transporter antibodies (Fig. 6). Treatment of BV-2 cells with 100 ng/ml LPS for 24 h intensified the diffuse and plasma membrane staining of Mrp1 and Mrp5 proteins and resulted in multiple intensely fluorescent foci. Decreased staining of Mrp4 and Bcrp proteins was observed in LPS-treated cells. There are no commercial antibodies capable of selectively staining Mdr1 protein in cells of mouse origin.
Inhibition of NF-κB Attenuates LPS-Induced Mdr1a Down-Regulation and Potentiates Mrp5 Up-Regulation.
LPS activation of Toll-like receptor-4 results in downstream NF-κB activation through phosphorylation of IκB, the inhibitory protein of NF-κB. NF-κB is then free to enter the nucleus where it increases the mRNA transcription of several genes, including TNFα, IL-1β, and IL-6. BAY 11-7082 inhibits the phosphorylation of IκB, resulting in the stabilization of the IκB/NF-κB complex and inhibition of NF-κB-mediated transcription (Min et al., 2011). Compared with LPS alone, cotreatment of BV-2 cells with LPS and BAY 11-7082 attenuated IL-1β and IL-6 cytokine mRNA induction by approximately 75% (Fig. 7, A and B). We were surprised to find that TNFα mRNA was increased further by cotreatment of cells with LPS and BAY 11-7082 (Fig. 7C). However, secreted TNFα protein in the cell culture media of BAY 11-7082-treated cells was decreased by approximately 75%, confirming inactivation of NF-κB signaling (Fig. 7D).
Inhibition of NF-κB with BAY 11-7082 attenuated the decrease in Mdr1a mRNA that resulted from LPS activation of microglia (Fig. 8). Treatment with BAY 11-7082 did not affect the LPS-induced down-regulation of Bcrp or Mdr1b mRNA. BAY 11-7082 also caused an increase in Mrp5 mRNA in control cells and potentiated the increase of Mrp5 in LPS-treated cells (Fig. 9). Although not statistically significant, Mrp1 mRNA showed a similar trend to Mrp5 by using BAY 11-7082. To further investigate the enhancement of Mrp5 mRNA by BAY 11-7082, the mRNA expression of the oxidative stress transcription factor nuclear factor e2-related factor 2 (Nrf2) and its target genes NAD(P)H quinone oxidoreductase 1 (Nqo1) and heme oxygenase-1 was quantified. This signaling pathway has been shown previously to regulate expression of some Mrps (Maher et al., 2007). It is noteworthy that BAY 11-7082 increased Nrf2 mRNA in LPS-treated cells and elevated mRNA levels of Nqo1 (Fig. 9) and heme oxygenase-1 (data not shown) in control- and LPS-treated cells.
The present study investigated the regulation of ABC efflux transporter function in activated microglia. Significant reductions in the cellular accumulation of fluorescent substrates after a 60-min efflux period demonstrated basal activity of Mrps, Mdr1, and Bcrp in BV-2 cells. On the other hand, LPS-treated microglia had greater intracellular fluorescence, which suggested that microglial activation decreased the function of efflux transporters. The retention of fluorescent substrates was similar to, and in some cases greater than, the retention observed with known transporter inhibitors. These data indicate that efficient efflux of chemicals in resting microglia is impaired during microglial activation and neuroinflammation.
BV-2 cells as a model for microglial activation.
In this study, we confirmed the presence of ABC efflux transporters in BV-2 cells and showed that they are similar to primary microglia and in vivo rodent studies. Several groups have shown that cultured rat primary microglia and the rat MLS-9 microglia cell line express functional Mrp1, Mrp4, Mrp5, and Mdr1 proteins but are deficient in Mrp2 protein (Lee et al., 2001; Ballerini et al., 2002; Hirrlinger et al., 2002; Dallas et al., 2003, 2004). Our initial studies comparing BV-2 cells with primary microglia from neonatal C57BL/6 mice have shown similar mRNA for most efflux transporters including Mrp1, Mrp2, and Mrp5. Mdr1a/1b mRNAs were up to 50% less abundant in primary microglia (data not shown). It is noteworthy that mRNA quantities were approximately 7-fold lower and 10-fold higher in primary microglia for Mrp4 and Bcrp, respectively. From a functional aspect, the MLS-9 cell line has been shown to lack Bcrp-mediated efflux of mitoxantrone (Lee et al., 2007). This is similar to our data demonstrating modest differences in Bcrp-mediated Hoechst 33342 retention in BV-2 cells under various experimental conditions. In addition, Dauchy et al. (2008) determined that MRP2 and MRP3 mRNA are not present in human cortex or isolated microvessels, and they concluded that BCRP and MDR1 represent the primary ABC transporters in human microvessels with relatively low amounts of MRP1, MRP4, and MRP5. Human cerebral cortex samples, on the other hand, had greater expression of MRP5, BCRP, and MRP1 mRNA, suggesting that other cell types, including microglia, express a unique set of efflux transporters (Dauchy et al., 2008; Warren et al., 2009).
Transporter Protein Response to Inflammation.
Although transporters function similarly to export substrates from cells, they are likely to be regulated by different mechanisms. This is supported by the fact that Mrp1 and Mrp5 mRNA and protein were up-regulated by LPS, whereas all other transporters were decreased. The reduced function of Mrp1 (calcein AM retention), despite the up-regulation of mRNA and protein, may be caused by altered membrane trafficking, loss of a transport cosubstrate, or a compensatory response. The multiple foci of Mrp1 staining in LPS-treated cells suggest that trafficking to the plasma membrane may be altered despite an increase in total protein. Alternatively, functional inhibition of Mrp1 may result from the depletion of glutathione, which is a transport cosubstrate. Resting microglia have high amounts of glutathione, which is rapidly oxidized in activated microglia that are producing reactive oxygen species (Hirrlinger et al., 2000; Rudd et al., 2011). Glutathione depletion could lead to decreased function of Mrp1. It is also conceivable that up-regulation of Mrp1 may be a compensatory response to impaired function.
Inflammation causes similar changes in transporter expression in the liver. Mrp1 and Mrp5 mRNA in livers of mice exposed to LPS are up-regulated, similar to activated BV-2 microglia (Lickteig et al., 2007). LPS also decreases the function, mRNA, and protein expression of Mdr1 in rodent livers (Piquette-Miller et al., 1998; Fernandez et al., 2004). Additional studies point to IL-6 and/or TNFα as key cytokines responsible for hepatic Mdr1 down-regulation in rodents during inflammation (Ando et al., 2001; Hartmann et al., 2001). Regulation of Mdr1 expression in the brain by LPS is less consistent; however, there seems to be a uniform decline in function in human cell culture models and rodent studies regardless of expression levels (Goralski et al., 2003; Fernandez et al., 2004; Hartz et al., 2006).
Initial studies conducted in primary microglia isolated from neonatal C57BL/6 mice demonstrate transporter mRNA changes after LPS activation that are similar to BV-2 cells (data not shown). Treatment of primary microglia with LPS increased Mrp1 mRNA and decreased Bcrp mRNA at 12 h. Additional dose and time-course responses of primary microglia are needed to further investigate the regulation of transporter expression as well as function during neuroinflammation.
Transcription Factor Signaling.
Few studies have investigated the signaling mechanisms underlying the regulation of Mdr1 and other transporters in response to LPS (Morgan et al., 2008; Teng and Piquette-Miller, 2008). In the present study, inhibition of the NF-κB transcription factor pathway, which regulates cytokine production, attenuated Mdr1a mRNA down-regulation. Evidence of impaired NF-κB signaling by BAY 11-7082 was observed in TNFα protein and IL-6 and IL-1ß mRNA. Somewhat paradoxically, TNFα mRNA was induced in cotreated cells, possibly through alternate pathways to compensate for lower protein levels.
In contrast to Mdr1, Mrp1 was not affected by the inhibition of NF-κB. In primary rat astrocytes, Ronaldson et al. (2010) found that expression of Mrp1 protein and transport increased after treatment with TNFα. This finding is in contrast to the functional decrease of Mrp1 in microglial cells. This discrepancy may be related to the inherently different functional responses of astrocytes and microglia to TNFα and LPS, respectively. From a mechanistic standpoint, inhibition of NF-κB did not prevent Mrp1 expression changes in astrocytes treated with TNFα (Ronaldson et al., 2010). Likewise, we found that inhibiting NF-κB in LPS-treated microglia did not interfere with Mrp1 induction, indicating that Mrp1 mRNA is not directly regulated by NF-κB.
Nrf2 regulates the expression of Mrp1 in cultured mouse fibroblasts (Hayashi et al., 2003). Likewise, pharmacological activation of Nrf2 increases Mrp5 mRNA in mouse liver (Maher et al., 2005). In the present study, mRNA expression of the Nrf2 target, Nqo1, was up-regulated by LPS and BAY 11-7082, supporting Nrf2 activation as a candidate transcriptional pathway for Mrp1 and Mrp5 up-regulation. Nrf2 activation is probably a result of NF-κB signaling, cytokine production, and oxidative stress associated with microglial activation. Nrf2 dampens LPS-induced microglial activation and lowers the production of proinflammatory cytokines in BV-2 cells (Koh et al., 2009, 2011; Lee et al., 2012). It is noteworthy that inhibition of NF-κB in microglia using BAY 11-7082 resulted in the up-regulation of Nqo1 rather than down-regulation as would be expected if there was direct cross-talk between NF-κB and Nrf2. One possible explanation is that BAY 11-7082 also acts independently of NF-κB to activate Nrf2. In fact, BAY 11-7082 has been shown to increase mRNA and protein of Nrf2 responsive genes in a Nrf2-dependent manner in human colorectal cells, likely through production of reactive oxygen species (Min et al., 2011).
Implications of Altered ABC Transport Function in Neuroinflammation.
Our study provides novel information regarding efflux processes in activated microglia that may alter signaling with other neuronal cells. Activation of microglia was initially believed to be a response to neurodegeneration. However, it is now widely accepted that microglia play a more active or even primary role in initiating neuronal injury (Kim and Joh, 2006; Rogers et al., 2007; Long-Smith et al., 2009; Cao et al., 2011a; Ros-Bernal et al., 2011; Yokoyama et al., 2011). Activated microglia have been detected in living and postmortem patients with Parkinson's disease (McGeer et al., 1988; Mirza et al., 2000; Gerhard et al., 2006) and in samples from humans and monkeys with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-induced Parkinson's disease (Langston et al., 1999; McGeer et al., 2003). Activated microglia are implicated in direct neuronal degeneration (Block et al., 2007). Amyloid β, lipid peroxides, prostaglandins, pesticides, and other chemicals have been found to either influence the expression of, or are themselves substrates for, efflux transporters (Sreeramulu et al., 2007; Oosterhuis et al., 2008; Kania et al., 2011; Rudd et al., 2011; Yang et al., 2011). The proapoptotic protein ceramide was also markedly retained in LPS-activated microglia in the present study. Collectively, these findings suggest the potential ability of transporters to regulate the accumulation of potentially detrimental compounds within the brain.
The active role of neuroinflammation and microglial activation in neurodegenerative disease has led to several studies examining the relationship of anti-inflammatory drug use and risk of neurodegeneration. Human epidemiologic studies have shown an inverse association between nonsteroidal anti-inflammatory drug (NSAID) use and the risk of developing Parkinson's disease (McGeer and McGeer, 1998; Chen et al., 2005). Likewise, the incidence of Alzheimer's disease may be delayed by NSAID use in asymptomatic patients (Breitner et al., 2011). Although the benefit of NSAIDs in these patient populations is related probably to the general suppression of inflammation, modulation of transporter expression in microglia may aid in restoring normal cell-cell communication. Studies examining drug transporters using NSAIDs have yielded conflicting results depending on the model system used and the type of NSAID. Jung et al. (2011) showed that aspirin induces MDR1 mRNA and activity in intestinal cells in vitro and rat intestine in vivo. In another study, NSAIDs increased MDR1 mRNA expression in Caco-2 cells but not efflux activity (Takara et al., 2009). Thus, the use of NSAIDs to attenuate neuroinflammation and microglial activation may have the added benefit of partially restoring normal transporter function.
We have demonstrated that the activation of microglia with LPS alters chemical transport by differentially regulating the expression of ABC transporters. Understanding how microglia communicate, react, and interact with the neuronal environment, particularly through the expression of transporters, may help in determining mechanisms of neurodegeneration and development of possible interventions. Future studies should aim to verify altered ABC transporter expression and function in animal models of neuroinflammation and neurological disease.
Participated in research design: Gibson, Richardson, and Aleksunes.
Conducted experiments: Gibson and Hossain.
Contributed new reagents or analytic tools: Richardson and Aleksunes.
Performed data analysis: Gibson.
Wrote or contributed to the writing of the manuscript: Gibson, Richardson, and Aleksunes.
We appreciate the use of the xCELLigence RTCA DP analyzer by Roche Diagnostics for real-time impedance measurements. We thank Dr. Bin Liu (University of Florida, Gainesville, FL) for BV-2 cells; Dr. George Scheffer (VU Medical Center, Amsterdam, The Netherlands) for antibodies; and Dr. Angela Slitt (University of Rhode Island, Providence, RI) for qPCR primers.
This work was supported in part by the National Institutes of Health National Institute of Diabetes and Digestive and Kidney Diseases [Grant DK080774]; the National Institutes of Health National Institute of Environmental Health Sciences [Grants ES020522, ES015991, ES007148, ES005022]; and the National Institutes of Health National Institute of Neurological Disorders and Stroke [Grant NS072097].
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- ATP binding cassette
- multidrug resistance protein
- multidrug resistance-associated protein
- breast cancer resistance protein
- nuclear factor κB
- inhibitor of κB
- nuclear factor-e2 related factor 2
- tumor necrosis factor-α
- NAD(P)H quinone oxidoreductase 1
- nonsteroidal anti-inflammatory drug
- acetoxymethyl ester
- phosphate-buffered saline
- ribosomal protein l13a
- glyceraldehyde-3-phosphate dehydrogenase
- quantitative polymerase chain reaction
- dimethyl sulfoxide
- BAY 11-7082
- 6-[(2S,4R,6E)-4-methyl-2-(methylamino)-3-oxo-6-octenoic acid]-7-l-valine-cyclosporin A
- 5-(3-(2-(7-chloroquinolin-2-yl)ethenyl)phenyl)-8-dimethylcarbamyl-4,6-dithiaoctanoic acid sodium salt hydrate
- (3S,6S,12aS)-1,2,3,4,6,7,12,12a-octahydro-9-methoxy-6-(2-methylpropyl)-1,4-dioxopyrazino[1′,2′:1,6]pyrido[3,4-b]indole-3-propanoic acid 1,1-dimethylethyl ester hydrate.
- Received May 20, 2012.
- Accepted August 30, 2012.
- Copyright © 2012 by The American Society for Pharmacology and Experimental Therapeutics