Gender-Specific Expression of ATP-binding Cassette (Abc) Transporters and Cytoprotective Genes in Mouse Choroid Plexus

Abstract

The choroid plexus (CP) and blood-brain barrier (BBB) control the movement of several drugs and endogenous compounds between the brain and systemic circulation. The multidrug resistance associated protein (Mrp) efflux transporters form part of these barriers. Several Mrp transporters are positively regulated by the transcription factor nuclear factor erythroid-2-related factor (Nrf2) in liver. The Mrps, Nrf2 and Nrf2-dependent genes are cytoprotective and our aim was to examine basal gender differences in expression of Mrp transporters, Nrf2 and Nrf2-dependent genes (Nqo1 and Ho-1) in the brain-barriers. Previous studies have shown higher expression of Mrp1, Mrp2 and Mrp4 in female mouse liver and kidney. We hypothesized that similar renal/hepatic gender-specific patterns are present in the brain-barrier epithelia interfaces. qPCR and immunoblot analyses showed that Mrp4, Ho-1 and Nqo1 expression was higher in female CP. Mrp1, Mrp2 and Nrf2 expression in the CP had no gender pattern. Female Mrp1, Mrp2 and Mrp4 mouse brain expressions in remaining brain areas, excluding CP, were higher than male. Functional analysis of Mrp4 in CP revealed active accumulation of the Mrp4 model substrate fluo-cAMP. WT female CP had 10-fold higher accumulation in the vascular spaces than males and 60% higher than Mrp4^−/− females. Probenecid blocked all transport. Methotrexate did as well except in Mrp4^−/− females where it had no effect, suggesting compensatory induction of transport occurred in Mrp4^−/−. Collectively, our findings indicate significant gender differences in expression of Mrp transporters and cytoprotective genes in the CP and BBB.

1.1 INTRODUCTION

Gender can be an important factor in drug and xenobiotic toxicity through differences in absorption, metabolism, distribution, and excretion. Many of the genes associated with these processes have gender-specific patterns in liver and kidney. For example, multidrug resistance-associated protein 4 (Mrp4/Abcc4) is a membrane transport ATPase expressed at much higher levels in female than in male murine kidney and liver (Cheng et al., 2008; Lu and Klaassen, 2008). Its transported substrates include chemotherapeutic drugs, antiretroviral agents, and anti-cancer agents, as well as several endogenous ligands, such as uric acid, conjugated bile acids, glucuronidated estrogens, and leukotrienes. Mrp4, together with several other efflux transporters, is not only associated with cellular resistance to the toxicity of short-term exposure to, for example, methotrexate (MTX) (Chen et al., 2002; Hooijberg et al., 1999), but it also facilitates transepithelial transport and excretion by liver and kidney (Chen et al., 2002; Hooijberg et al., 1999; Reichel et al., 2007). Thus, the effectiveness of certain drugs as well as the excretion rate of xenobiotics and endogenous metabolites may be significantly impacted by gender. Sex differences in the gene expression of these efflux transporters may influence both tissue specific and whole body substrate clearance and, thus, aid in risk assessment and drug dosing decisions.

Efflux transporters are a critical component of the active barrier functions of the blood-brain barrier (BBB) and blood-cerebrospinal fluid barrier (BCSFB) as they protect the brain from endogenous and exogenous compounds, including a variety of therapeutics, by exporting them from the cerebrospinal fluid (CSF) or brain parenchyma interstitial space (IS) to the blood for ultimate excretion. The choroid plexus (CP) epithelium (BCSFB) and the brain parenchyma capillary endothelium (BBB) exert significant control over what goes into and comes out of the central nervous system. The CP is a highly vascularized epithelium that receives as much blood as the rest of the brain combined and secretes about 80% of the CSF. Because it assists in the clearance of various soluble compounds such as organic anions (OA) from CSF it is referred to as the “kidney” of the brain. The BBB has an important role as well in controlling movement of substrates between the blood and the brain IS. These barriers are highly active, dynamic and selective; however, the gender-specific characterization of the ATP binding cassette (Abc) efflux transporters in the brain is still limited.

The Abc-type transporters act as gatekeepers by allowing or limiting the entrance of an enormous variety of substrates across cellular membranes. In CP epithelium, Mrp1, Mrp4 (Leggas et al., 2004) and Mrp5 (Abcc5) (Roberts et al., 2008) are located in the basolateral membranes – an orientation that allows active transport from cytosol to an interstitium contiguous with the CP’s fenestrated capillaries - whereas Mrp2 (Abcc2) is located in the apical membrane – transporting from cytosol to CSF. Mrps are also located in BBB endothelium of several species, including humans, rats and mice (Leggas et al., 2004; Rao et al., 1999; Roberts et al., 2008). P-glycoprotein (P-gp/Abcb1), breast cancer resistance protein (Bcrp/Abcg2), Mrp1, Mrp2, Mrp4 and Mrp5 are all localized to the BBB luminal membranes. In relatively lower amounts the Abc transporters are also expressed in astrocytes, microglia, neurons, and oligodendrocytes (Ballerini et al., 2002; Berezowski et al., 2004).

In addition to Mrps, several drug metabolizing enzymes are expressed in the CP and BBB including glutathione S-transferases and several isoforms of the cytochrome P450 family (Decleves et al., 2011; Miksys et al., 2000; Tyndale et al., 1999). There are transcription factors known to positively regulate the expression and activity of several cytoprotective genes during periods of oxidative stress. For example, the transcription factor Nrf2 is involved in many cellular functions including drug metabolism, oxidative stress mitigation and regulation of efflux transporter pathways. These genes include NAD(P)H quinone oxidoreductase 1 (Nqo1), heme-oxygenase 1 (Ho-1) and glutamate cysteine ligase catalytic subunit (Gclc), as well as certain efflux transporters such as Mrp4 (Aleksunes et al., 2008; Aleksunes et al., 2010). Biotransformation of foreign compounds via these phase I and II enzymes can result in metabolites which can be transported by Mrps. Consequently, metabolic enzymes and transporters can act together to eliminate harmful compounds from the brain to the blood for eventual body excretion. The extent to which metabolism contributes to the BCSFB and to the BBB remains to be established. Furthermore, the gender-specific patterns of detoxification enzyme profiles and the influence of Nrf2 on transporter expression have not been well characterized in CP.

There are numerous reports of sexual dimorphism of Abc-type transporters in kidney and liver of adult mice, rats and humans (Cheng et al., 2008; Lu and Klaassen, 2008), but data on sex differences in CP are extremely limited. Our primary goal was to test the hypothesis that Abc transporters and cytoprotective genes in adult mouse CP share kidney- and liver-like gender expression patterns. Here, gender specific expression patterns of several key cytoprotective genes and proteins in the CP and brain are presented together with functional data for Mrp4.

2. METHODS

2.1. Animal tissue collection

All animal studies were performed in accordance with institutional regulations for animal protection and were approved by the Institutional Animal Care and Use Committee of the University of Connecticut (Protocol A12-050). Age-matched male and female C57BL/6J mice (10–12 weeks of age) were purchased from Jackson Laboratory (Bar Harbor, ME, USA). Mrp4^−/− mice with a C57BL/6J background were kindly provided by Dr. John Schuetz from St. Jude’s Children’s Hospital, Memphis, TN. A colony of these mice is maintained at the University of Connecticut. All mice were housed in temperature, light, and humidity controlled conditions approved by the Association for Assessment and Accreditation of Laboratory Animal Care. All mice were provided standard laboratory chow (Harlan) and water ad libitum. Following euthanasia via decapitation, brains were removed and immersed in ice-cold PBS. The two lateral CPs of each animal were removed from the ventricles under a dissecting microscope. Tissues were then transferred into 0.5 ml tubes containing either RNA later (Qiagen, Valencia, CA, USA), neutral phosphate buffered 10 % formalin fixative or Laemmli sample buffer for future analyses.

2.2. Solutions and chemicals

8-(2-[Fluoresceinyl]aminoethylthio)adenosine-3', 5'-cyclic monophosphate (Fluo-cAMP) was purchased from Biolog Life Science Institute (Bremen, Germany). This fluorescent cAMP analogue is highly resistant to metabolic degradation by cAMP-specific phosphodiesterase. Probenecid and methotrexate (MTX) were purchased from Sigma (St. Louis, MO, USA). Information on primary and secondary antibodies for immunoblots is listed in Table 1. All other chemicals were obtained from Sigma.

Table 1.

Description of primary and secondary antibodies for immunoblots

Protein	Primary Antibody Conc. Tm/T	Primary Antibody Source	Molecular weight (kDa)	2° Antibody Conc. Tm/T	Primary and secondary blocking buffer solution
Mrp1	ab24102 1:500	Abcam	190	Anti-ratS 1: 2000 RT/ 1hr	5 % NFDM
Mrp2	ab3373 1:500 4°C/ ON	Abcam	190	Anti-mouseS 1:2000 RT/ 1hr	5 % NFDM
Mrp4	ab15602 1:500 4°C/ ON	Abcam	160	Anti-ratS 1:2000 RT/ 1hr	5 % NFDM
Mrp5	ab24107 1:500 4°C/ ON	Abcam	160	Anti-ratS 1:2000 RT/ 1hr	5 % NFDM
Nrf2	8882S 1:500 4°C/ ON	Cell Signaling	~100	Anti-ratS 1:2000 RT/ 1hr	5 % BSA
Gclc	- 1:10000 4°C/ ON	Washington	~75	Anti-rabbitS 1:20000 RT/ 1hr	5 % NFDM
Ho-1	SPA-895 1:5000 4°C/ ON	Stressgen Bioreagents	~33	Anti-rabbit 1:2000 RT/ 1hr	5 % NFDM
Nqo1	ab2346 1:1000 4°C/ ON	Abcam	~32	Anti-goatS 1:10000 RT/ 1hr	5 % NFDM
β-actin	ab8227 1:3500 4°C/ 1hr	Abcam	42	Anti-rabbitS 1:2000 RT/ 1hr	5 % NFDM

NFDM, non-fat dry milk

BSA, bovine serum albumin

^SAntibody purchase from Sigma

Con, Antibody concentration

Tm, incubation temperature

T, incubation time

ON, overnight

RT, room temperature

2.3. SDS-PAGE and immunoblotting

The two lateral CPs from each animal were suspended in 20 µL of Laemmli sample buffer (2.3 % SDS, 5 % β-mercaptoethanol, 0.5 % bromophenol blue, 62.5 mM Tris HCL, pH 6.8) and stored at −20° C. Equal amounts of sample were loaded on 10 % SDS-Polyacrylamide gels and transferred to polyvinylidene difluoride membranes (PVDF). After transfer, PVDF membranes were blocked at 4°C with 5 % nonfat dry milk (NFDM) in PBS containing 0.5 % Tween (PBS-T). Membranes were incubated overnight with appropriate primary antibodies, except for a loading-control that was incubated for one hour. Standard β-actin controls (anti-β actin; 1:500) were included to ensure equal loading and for normalization of the target antibody band intensity. A description of the individual antibody conditions (controls, primaries and secondaries) is provided in Table 1. All the primary and secondary antibodies where diluted in the same blocking buffer solution found in Table 1. A species-appropriate peroxidase labeled secondary antibody (Sigma-Aldrich) was diluted in blocking buffer and incubated with blots for 1 hour. Finally, protein bands were detected using a chemiluminescence (ECL) kit (Thermo Scientific) according to the manufacturer’s instructions prior to exposure to medical x-ray film (Fujifilm). Densitometry of detected bands was performed using ImageJ software (NIH) for quantitative analysis.

2.4. RNA isolation and Quantitative real time polymerase chain reaction (qRT-PCR)

Total RNA was prepared from ventricular male and female CPs using a Qiagen RNeasy mini kit (Qiagen, Valencia, CA, USA). A Thermo Scientific Nanodrop was used to measure RNA concentrations and purity by absorbance readings at 260 and 280 nm. For each sample, first-strand cDNA was synthesized from 0.5 µg total RNA using the Superscript III First-Strand Synthesis System (Invitrogen, Carlsbad, CA, USA) in a 20 µl reaction, according to the manufacturer’s instructions. cDNA reactions were conducted in a PCR Sprint thermocycler (Thermo Electron Corporation). To control for contaminants, no-amplification-controls (NACs) with no enzyme, and no-template-controls (NTCs) with no cDNA were included. qRT-PCR was performed on CFX96 touch™ Real-time PCR detection System (Biorad) using SsoFast™ EvaGreen Supermix and a 10 µl reaction volume. Cycling parameters for all reactions were as follows: denaturation at 95° C for 20 seconds; 40 cycles of denaturation at 95° C for 3 seconds; 30 seconds of annealing and extension at 60° C. Mouse-specific primer pairs were used for each gene (Table 2). Gene expression was quantified by the ΔΔCT method (Schmittgen and Livak, 2008), using β-Actin as our internal control gene.

Table 2.

Primer sequences for quantitative real-time-PCR

Gene name	Gene symbol	Primer forward and reverse sequences
ATP-binding cassette, sub-family C (MRP), member 1	Abcc1	F: 5’-TTCGGAAGGGAGAATCGGCTTCAA-3’
		R: 5’-TGTTCAGTGCACCTTGCTTGTTCG-3’
ATP-binding cassette, sub-family C (/MRP), member 2	Abcc2	F: 5’-TCC AGG ACC AAG AGA TTT GC-3’
		R: 5’-TCT GTG AGT GCA AGA GAC AGG T-3’
ATP-binding cassette, sub-family C (MRP), member 4	Abcc4	F: 5’-TAATGGAAGCAGACAAGGCCCAGA-3’
		R: 5’-AGAGGCCAGTGCAGATACATGGTT-3’
ATP-binding cassette, sub-family C (MRP), member 5	Abcc5	F: 5’-ACA GCC GCT ATG GAC ACA GAG ACA GA-3’
		R: 5’-AGG CGA AGT TTC AGC AGG ACA GGA TG-3’
Nuclear factor erythroid 2-related factor 2	Nrf2	F: 5’-TCT ATG TGC CTC CAA AGG-3’
		R: 5’-CTC AGC ATG ATG GAC TTG GA-3’
Glutamate--cysteine ligase catalytic subunit	Gclc	F: 5’-CTGCACATCTACCACGCAGT-3’
		R: 5’-TTCATGATCGAAGGACACCA-3’
Heme oxygenase 1	Ho-1	F: 5’-GAGCCTGAATCGAGCAGAAC-3’
		R: 5’-CCTTCAAGGCCTCAGACAAA-3’
NAD(P)H quinone dehydrogenase 1	Nqo1	F: 5’-TTT AGG GTC GTC TTG GCA AC-3’
		R: 5’-GTC TTC TCT GAA TGG GCC AG-3’
Actin, beta	Actb	R: 5’-GCA GAC AGC CAA GGA GCC CAA AGA CC-3’
		F: 5’-GCA ACG AGC GGT TCC G-3’

2.5. Transport

For transport studies, fresh lateral CPs were removed and transferred immediately into 12-well plates containing an artificial cerebrospinal fluid (aCSF; 103 mM NaCl, 25 mM NaHCO₃, 4.7 mM KCl, 1.2 mM KH₂PO₄, 1.2 mM MgSO₄ × 7H₂O, 10 mM glucose, 2.5 CaCl₂, at pH 7.4) with pre-dissolved gases (95 % O2, 5 % CO2). Then, CPs were pre-incubated with or without (control) modulators of organic anion transport for 30 minutes at room temperature. Probenecid was used as a standard well known broad spectrum inhibitor for organic anion transporters. MTX was used as a competitive inhibitor for Mrp4. All incubations were carried out in 12-well plates inside sealed Ziploc bags under 95 % O₂, 5 % CO₂ in the dark. After 30 minutes of modulator treatment, fluorescently labeled substrate was added, and the time to reach steady state fluorescence was determined. Following each measurement, the Ziploc bags were re-filled with 95 % O₂, 5 % CO₂, sealed and returned to darkness. Accumulation of fluo-cAMP, a known Mrp4 substrate, was measured in cells and in vascular/perivascular spaces with or without 50 µM MTX or 1 mM probenecid. Viability of the tissue was monitored by morphology and substrate accumulation over time. After pre-incubation and incubation, fluorescence intensity in the different compartments of CP was imaged by inverted confocal Leica SP8 DMIRB microscope.

2.6. Confocal imaging

After incubation, CPs were transferred to glass bottomed Teflon chambers and observed on an inverted confocal microscope under the 20 × objective. Under transmitted light, intact epithelia were selected as described by Valeska et al 2010 (Reichel et al., 2010). For each piece of CP, five areas of undamaged CP tissue were chosen, and confocal images were taken. A 543 nm HeNe laser was used along with appropriate dichroic and long-pass emission filters. All images collected had the same settings (offset, inset, pinhole, laser intensity, etc.). The laser intensity did not exceed 1 % of the maximum. Auto-fluorescence and photo bleaching were not detectable. Fluorescence intensities were measured using the Leica LAS AF 2D Analysis software. All data are reported as fluorescence intensities after background (bathing media fluorescence) subtraction. Confocal imaging method used as described previously (Breen et al., 2002; Breen et al., 2004; Miller, 2004; Reichel et al., 2010) and the glass bottomed Teflon chambers were a kind gift from David Miller at NIEHS.

2.7. Statistical/Data Analysis

All experiments were carried out with CP tissue isolated from 4–8 mice. Calculated intensity values were pooled and presented as mean values ± SEM. Experimental means for basal mRNA and protein expression in WT groups were evaluated using paired t-tests and significant differences were assumed if P < 0.05. Group means of WT and Mrp4^−/− mice were compared using one-way ANOVA and Dunnetts post hoc test. Differences were considered statistically significant when P < 0.05.

3. RESULTS

3.1. Mrp mRNA and protein expression in male and female mouse CP

Quantitative PCR showed significant gender differences in gene expression of Mrp4 and Mrp5 (Figure 1A). Females expressed higher mRNA levels of Mrp4 whereas males expressed higher mRNA levels of Mrp5. The almost 3-fold higher expression of Mrp4 in female CP is consistent with the reported expression in murine liver and kidney (Lu and Klaassen, 2008; Maher et al., 2006). Mrp1 and Mrp2 mRNA expression did not differ significantly in male and female CP.

Figure 1. Multidrug resistance-associated proteins (Mrp) 1, 2, 4 and 5 mRNA expression and immunoblot analyses in wild-type female and male mouse CP.

Gene expression or protein expression was normalized to housekeeping gene β-actin. (A) Relative mRNA expression of Mrp1, 2, 4, and 5. (B) Individual blots with corresponding β-actin together with a summary of relative protein expressions. The corresponding reference β -actin is found below each target protein, except for Mrp2 which was detected along with Mrp5. Each column represents an individual mouse lateral CP. The data are presented as means ± SE (n=4, qPCR; n =8, immunoblots). Asterisks (*) represent a statistical difference between female and male (p ≤ 0.05).

Similar to mRNA expression, Mrp4 protein expression was twice as high in females compared to males (Figure 1B). No significant differences between genders were observed for Mrp1 or Mrp5. Even though hepatic and renal Mrp1 mRNA and protein expression are reported to be higher in females (Cheng et al., 2008), we found no significant gender difference in the CP. The lack of a gender difference in Mrp5 protein expression in CP is consistent with its expression in the liver.

3.2. Nrf2, Ho-1, Nqo1 and Gclc mRNA and protein expression in male and female mouse CP

Figure 2A shows the mRNA expression of Nrf2 and several Nrf2 dependent genes in male and female mouse CP. Ho-1, Nqo1 and Gclc genes were selected because they are known to play a role in xenobiotic metabolism and oxidative stress, generating possible Mrp-transportable substrates. Together with Mrp4, they are positively regulated by Nrf2 activation in mouse liver (Rohrer et al., 2014). Neither Nrf2 nor Gclc gene expression differed significantly with CP gender (Figure 2A). However, basal mRNA expression of Ho-1 and Nqo1 were higher in the female CP with Nqo1 more than double that of males.

Figure 2. mRNA and protein expression of Nrf2 and Nrf2-regulated Ho-1, Nqo1 and Gclc in unstressed WT male and female mouse CP.

Gene or protein expression was normalized to housekeeping gene β-actin. (A) Relative mRNA expression of Nrf2, Ho-1, Nqo1 and Gclc. (B) Individual blots with corresponding β-actin together with a summary of relative protein expressions. The data are presented as means ± SE (n=4, qPCR; n =8, immunoblots). Asterisks (*) represent a statistical difference between female and male (p ≤ 0.05).

Consistent with the observed mRNA expression, Ho-1 and Nqo1 protein expression were significantly higher in female compared to male CP (Figure 2B). The data are consistent with the previously reported hepatic expression (Rohrer et al., 2014). Also in agreement with the mRNA data, no significant gender differences were observed for Gclc and Nrf2 protein expression in CP (Figure 2B).

3.3. Functional analyses and fluo-cAMP transport characteristics

Of the four Mrps with known localizations in the CP, only Mrp4 showed significant gender-related differences in protein expression. To determine whether this difference translated to functional differences in substrate handling capacity, the active accumulation of the model substrate fluo-cAMP was examined. Figure 3A–E shows an example of phase contrast (A and C) and fluorescence confocal images (B and D) of female mouse CP in the absence (B) and presence (D) of 1 mM probenecid, a relatively nonspecific organic anion transport inhibitor. Fluorescence intensity is greatest in the vascular/perivascular spaces (VS). These fenestrated vessels are just beneath the choroidal epithelium and show outlines of red cells which do not take up the fluo-cAMP. In the linear scan across the image shown in Figure 3E the CP epithelial cells seem to maintain fluo-cAMP below the substrate signal in the CSF. Maximum intensity (increase beyond CSF reading, X) is in the vascular/perivascular spaces, and uptake is totally blocked by probenecid (Figure 3C–D). Figure 3F shows time dependent fluo-cAMP accumulation into female and male mouse CP vascular/perivascular spaces from a concentration of 2 µM fluo-cAMP. Fluo-cAMP fluorescence greater than that of the bathing aCSF was detected in both female and male after 50 minutes and was at or near steady state between 60 and 110 minutes with males reaching steady state somewhat earlier than females. Figure 3G shows that 2 µM fluo-cAMP was well below saturation. All subsequent measurements of fluorescence intensity reported here were done following exposure to 2 µM fluo-cAMP for 90 minutes. Under control conditions fluorescence intensity in the vascular/perivascular spaces was always higher than in cells or bath, and cellular fluorescence was always below that of the surrounding bath (data not shown).

Figure 3. Time course and concentration-dependent fluorescence intensity of fluo-cAMP in female and male mouse CP.

Representative phase contrast (A and C) and confocal (B and D) images of freshly isolate female mouse CP incubated with 2µM fluo-cAMP at steady state. CP incubated with fluo-cAMP in the presence (D) or absence (B) of the inhibitor probenecid (1mM). Confocal and transmitted light images are isolated from different female mice. (E) Plot of maximum fluorescence intensity (X) in vascular/perivascular spaces of corresponding representative image. Background intensity was subtracted. (F) Time course of fluo-cAMP transport (2µM) shows that steady state was reached after 60 minutes in male and continued 110 minutes after. In female fluo-cAMP transport (2µM) shows that steady state was reached after 85 minutes and continue 110 minutes after. Female fluorescence accumulation in vascular/perivascular spaces exceeded male fluorescence. Cellular fluorescence was lower than the bath in all experiments. (G) Saturation of fluo-cAMP transport in control group (female mouse CP) was directly affected by concentration of fluo-cAMP. The 2µM fluo-cAMP shows fluorescence intensity below saturation but high enough to be detected and was used for all functional assays. Scale bar in images A–D represents 25 µm.

3.4. Effects of probenecid and MTX on fluo-cAMP transport

The fluorescence intensity of the vascular/perivascular spaces was much higher in female than in male mouse CP (Figure 4). Uptake by male CP was only about 10 % that of females. The accumulation in the vascular/perivascular spaces of the female was abolished by incubation with 1 mM probenecid (Figure 4). Uptake by male CP was not significantly different following treatment with probenecid. In the Mrp4^−/− female mice, CP accumulation averaged about 40 % that of WT female and was completely blocked by probenecid (Figure 4). Uptake by male Mrp4^−/− CP was lower than the Mrp4^−/− female and also was not significantly different than that following probenecid treatment.

Figure 4. Effect of standard organic anion inhibitor, probenecid, and competitive inhibitor, methotrexate (MTX) on fluo-cAMP transport in female and male mouse CP.

Quantification of fluorescence intensity of A) WT and Mrp4^−/− female and male mouse CP incubated with or without probenecid (1mM) and B) CP incubated with or without MTX (50 µM). WT females show higher accumulation of fluo-cAMP than WT males and higher than Mrp4^−/− females. When incubated with the standard inhibitor for organic anion transporters, probenecid, accumulation of fluorescence was abolished in all groups. MTX significantly reduced fluorescence intensity in wild type female and had no effect in either wild type or Mrp4^−/− male CP. Interestingly, the fluorescence accumulation in VS/PVS was not affected in Mrp4^−/− female CP by MTX. The fluorescence was measured in the vascular/perivascular spaces and background fluorescence was subtracted. Data are presented as the mean values of 4 animals per group and the variability is shown as ±S.E. Groups with different letters are statistically different from each other (p < 0.05). Groups with same letter are not statistically different. PVS=perivascular spaces, VS=vascular spaces, aCSF= artificial cerebrospinal fluid.

MTX is a known high-affinity substrate for Mrp4, but it is also transported by several other Mrps. Our results show that fluo-cAMP transport by WT female CP was completely blocked by 50 µM MTX (Figure 4, lower panel). Again, accumulation in the female Mrp4^−/− was only about 40 % that of wild type female; however, MTX had no effect on this component of transport indicating the possibility that a compensatory transport process may have been induced in knockout mice.

3.5. Expression of Mrp1, Mrp2 and Mrp4 in mouse BBB

The BCSFB and BBB act together as gatekeepers to control the movement of organic anions and protect the brain from harmful compounds. In whole brain, Abc transporters are located mainly in brain endothelial cells (Roberts et al., 2008). Mrp1, Mrp2 and Mrp4 are expressed at the luminal membrane, facing the blood. To compare the expression of BBB Mrps in female and male mouse, mRNA levels were analyzed after the removal of the CP. In female mouse brain Mrp1, Mrp2 and Mrp4 mRNA expressions were at least twice that of males (Figure 5).

Figure 5. qPCR and immunoblot analysis of multidrug resistance associated proteins 1, 2, and 4 in WT female and male mouse brain.

mRNA and protein expression were normalized to β-actin and to control (female mice). (A) Quantification of mRNA and (B) relative protein expression of the selected efflux transporters in brain. The data are presented as means ± SE (n=4, qPCR; n =4, immunoblots). Asterisks (*) indicate a statistical difference between female and male (p≤0.05).

Immunoblots showed that protein levels of Mrp1, Mrp2 and Mrp4 were consistent with the mRNA expression, and that in every case they were significantly higher in females compared to males (Figure 5). Thus, a role in preventing the accumulation of their overlapping and often toxic substrates may be gender specific at the BBB.

4. DISCUSSION

Notably, this work revealed that there is a gender difference in a major Mrp transporter in the CP consistent with a shared kidney- and liver-like sex expression pattern. Female mice expressed higher Mrp4 mRNA in CP, and this differential expression corresponded to both higher protein levels in female CP and gender-dependent functional differences (Summary: Figure 6). Mrp1 and Mrp2 showed no gender differences in CP, and though males had higher Mrp5 mRNA expression, protein levels were no different. In addition to the efflux transporters, the present study indicated gender differences in the CP gene expressions of the electrophile scavenging Nqo1 and the heme-degradating enzyme Ho-1. Again, female CP had higher basal levels of mRNA and protein expression compared to male. We observed no differences in the expression of Gclc, consistent with studies in other tissues (Rohrer et al., 2014). Though the hepatic basal expression of Nrf2 is reportedly higher in males than females (Rohrer et al., 2014), we found no gender-dependent differences in CP.

Figure 6. Summary of Abc type transporter mRNA/protein expression in blood brain interfaces.

Known localization of drug transporter expression at the choroid plexus (A–B) and the blood brain barrier (C–D). Localization of Mrp5 and 6 remains uncertain (Mrp5–6? Symbol) due to lack of immunohistochemically competent antibodies. Green cAMP symbols indicate that fluo-cAMP accumulation is highest in vascular/subepithelial spaces in females compared to males. CSF and TJ represents cerebrospinal fluid and tight junctions, respectively.

Leggas et al. (2004) showed that Mrp4 limits brain penetration of substrate drugs and is expressed in the CP and the BBB; however, they did not report gender differences. In their Mrp4^−/− mouse study, the Mrp4 substrate, topotecan, was accumulated to 10 times the level of that in the WT brain. Its location in the CP epithelial basolateral membranes facing the blood side and in the luminal membranes of the BBB endothelium make Mrp4 especially effective since it not only can prevent cytotoxicity by blocking the cellular entry of potentially toxic substrates, it is part of the transepithelial transport of substrates from CSF or IS into blood. These transporters thus limit CNS accumulation of many therapeutic drugs and endogenous metabolites (Chen et al., 2002; Leggas et al., 2004; Merrell et al., 2008; Zhang et al., 2004). Given the potential role for Mrp transport during treatment of CNS disease, it is important to determine the possibility of gender differences at these barriers. Our studies suggest that Mrp4 function may differentially affect CNS substrate clearance by female and male mouse CP.

Recent studies reported that in liver Mrp2 and Mrp4 expression can be altered in response to cellular stress differentially in male and female (Rohrer et al., 2014), but the implications are not yet understood. Oxidative stress is thought to contribute to the pathogenesis of neurodegenerative diseases, therefore enzymes and transporters with anti-oxidant properties can protect the brain from this fate. Nqo1 is an important enzyme in the detoxification pathways of reactive quinones and reduces oxidative stress (Hwang et al., 2015; Luo et al., 2015; Merrell et al., 2008). Ho-1 catalyzes the rate limiting step in heme catabolism and is also widely accepted as being cytoprotective against several stressors, including electrophiles through the production of bilirubin, a powerful antioxidant and Mrp4 substrate (Luo et al., 2015; Merrell et al., 2008). Gclc is the rate limiting enzyme in GSH biosynthesis, and GSH conjugates can be Mrp4 substrates. Our data demonstrate clear gender differences in the CP for the antioxidant genes Nqo1 and Ho-1 with females having higher basal levels of mRNA and protein expression compared to males. Consistent with studies in other tissues (Rohrer et al., 2014), we observed no basal gender differences in the expression of Gclc in CP, and although it has been reported that there is a somewhat surprisingly higher hepatic basal expression of Nrf2 in males compared to females (Rohrer et al., 2014), we saw no gender differences in Nrf2 expression in CP. This suggests that the potential for Nrf2 activation and GSH production are similar in male and female CP.

Given the gender differences reported here in the CP and the overlapping transporter substrate specificities, we explored the expression of Mrp1, Mrp2 and Mrp4 in the remainder of the brain and found evidence of gender specific patterns. Based on prior studies, we assumed the brain Abc transporters are located mainly in BBB endothelium (Roberts et al., 2008; Zhang et al., 2004). Our studies indicated that Mrp1, Mrp2 and Mrp4 are expressed in a gender specific fashion at both gene and protein levels in the BBB. Previously it was reported that there were no gender differences in the mRNA expression of these three transporters in the whole brain homogenates (Maher et al., 2005). However, we removed the CP and revealed differences at both mRNA and protein levels in remaining brain.

Wang et al. (2014) showed that sulforaphane activation of Nrf2, and its translocation to the nucleus in vivo or in vitro, increases Mrp2 and Ho-1 expression in rat and mouse brain capillaries. Considering the variety of therapeutic drugs that are handled by these efflux transporters, conditions leading to oxidative stress or chemical activation of Nrf2 could change drug or xenobiotic penetration into the brain, reducing their efficacy or toxicity. The levels of Nrf2 activation or translocation in females versus males at these CNS barriers remain to be explored. However, contrary to expectations and consistent with the basal liver Nrf2 expression in a previous report (Rohrer et al., 2014), in our study basal Nrf2 mRNA and protein expression was not higher in female CP compared to male CP. The same study revealed that the higher female liver basal expression of Nrf2 target genes, Nqo1 and Mrp3, did not correspond to the level of Nrf2 expression. Moreover, other nuclear transcription factors, including peroxisome proliferator–activated receptor a, aryl hydrocarbon receptor, and c-Jun N-terminal kinase (JNK), have been recognized as key mediators in the regulation of the expression and activity of many of these cytoprotective genes (Aleksunes and Klaassen, 2012; Moffit et al., 2006; Xu et al., 2005). The gender dependence of additional consequences of Nrf2 activation and translocation to the nucleus, and the role of other transcription factors, at these CNS barriers remains to be determined. Even though the expression of these Mrps has been studied extensively in both liver and kidney, the mechanisms behind the gender-specific CP and BBB expression patterns remains poorly understood.

Although, brain Mrp1, Mrp2 and Mrp4 expression may have an impact on gender-specific transport of organic anions in BBB, Mrp4 may play a more prominent role in female and male mouse CP. On a mass basis, CP is reported to have 4–5 times more Mrp4 than kidney (Leggas et al., 2004). Mrp4 basolateral membrane localization in CP suggests a role in the flux of important anions from the CSF to the blood and prevention of penetration of toxins to the brain. Studies by others have clearly shown that cAMP outflow is mediated by multi-specific membrane transporters belonging to the Abc family (Biondi et al., 2010). The present study further characterized the function of Mrp4 by examining the efflux of fluo-cAMP in female and male mouse CP. Previous work in rat CP by Valeska et al. (Reichel et al., 2010) showed that Mrp4 functions as an ATP dependent export pump for cAMP. Thus, based on inhibitor specificity and our Mrp4^−/− model, CFS-to-blood transport of fluo-cAMP was significantly reduced in the Mrp4^−/− female. This confirms a strong involvement of Mrp4 in the transport of this substrate. Even though it is likely that Mrp4 plays an important role in the transport of cAMP in mouse CP ex-vivo, there might be other factors in vivo that could contribute to cAMP transport. For example, depletion of intracellular GSH with DL buthionine-(S, R)-sulphoximine leads to decreased export of cAMP and corresponding increases in intracellular cAMP accumulation (Lai and Tan, 2002).

In our experiments, MTX did not affect the transport of fluo-cAMP in female Mrp4^−/− CP suggesting that there are compensatory changes in organic anion transport capacity when Mrp4 is not present. The Mrp4^−/− mice are healthy and have a normal life span (Leggas et al., 2004). Mrp1 through Mrp5 also confer resistance to various anticancer drugs including MTX (Chen et al., 2002; Russel et al., 2008); therefore, this type of compensation might be attributed to an unidentified organic anion transporter that is not sensitive to MTX.

Consistent with previous studies, fluo-cAMP proved to be a good substrate model to study Mrp4 function. Compared with other Mrps, Mrp4 has a higher affinity for cAMP (Chen et al., 2002). In-vivo cAMP may be involved in intercellular signaling and the physiological contribution of Mrp4 to the regulation of cyclic nucleotide signaling will need to be further examined.

Mrp gender differences in the CP and brain correlate with the gender differences found in liver and kidney. The wide distribution of these transporters in combination with the differences in tissue-specific expression patterns and substrate specificity make them of great interest with respect to different classes of therapeutics and xenobiotics. The regulation of Mrp expression and activity by endogenous metabolites, signaling compounds, including sex hormones, and pharmaceutical agents have been documented. In gonadectomized mice, 17β-estradiol replacement markedly stimulates the expression of Mrp3 in both male and female mouse kidney. 5α-dihydroxytestosterone treatment suppressed female-predominant Mrp4 expression (Maher et al., 2006). In addition, gender can profoundly influence relative toxicity. For example, male mice are more susceptible to APAP nephrototoxicity than female mice (Hoivik et al., 1995). CYP2E1 is involved in the bioactiviation of APAP in the kidney. Previous studies have shown that higher renal CYP2E1 expression in male kidney contributes to greater susceptibility to APAP induced renal toxicity compared to female mice (Hoivik et al., 1995; Mazer and Perrone, 2008). Our data on the CP gender differences may provide a further step toward better understanding the impact of drugs and other organic metabolites on the brain.

In summary, we show here that basal expressions of the Mrps and several antioxidant-relevant genes are higher in females compared to males. Although our experiments show gender differences in the expression of several of these genes, the mechanism remains unknown. The present results highlight significant gender differences in the transport of the Mrp4 substrate fluo-cAMP. The physiological contribution of Mrp4 to regulation of cyclic nucleotide signaling will need to be further examined. Our results provide some guidance in this matter and suggest that the movement of endogenous/ exogenous molecules between plasma and brain compartments may be gender dependent. Because Mrp4 is important in determining the disposition of anionic drugs, and as a consequence, in determining the pharmacological and/or adverse effects of substrate drugs, gender differences should be considered in the future.

Abbreviations

CP

choroid plexus

BBB

blood-brain barrier

CSF

cerebrospinal fluid

Mrp

multidrug resistance-associated protein

Abc

ATP binding cassette

MTX

methotrexate

Mrp4

multidrug resistance-associated protein 4

BCSFB

blood-cerebrospinal fluid barrier

OA

organic anions

IS

interstitial space

Nrf2

nuclear factor erythroid-2-related factor

Nqo1

NAD(P)H quinone oxidoreductase-1

Ho-1

heme oxygenase-1

Gclc

glutamate cysteine ligase catalytic subunit

APAP

acetaminophen

ANOVA

analysis of variance