Regulation of the ATP‐binding cassette transporters ABCB1, ABCG2 and ABCC5 by nuclear receptors in porcine blood–brain barrier endothelial cells

Blood–brain barrier (BBB) ABCB1, ABCG2 and ABCC5 transporters influence central therapeutic drug distribution. Transporter expression is regulated by the NR3C1, NR1I3 and NR1I2 nuclear receptors, but their precise roles in brain are poorly understood. We investigated the effects of selective ligand‐based activation of NR3C1, NR1I3, NR1I2 and NR2B1 in porcine brain endothelial cells (PBECs).


| INTRODUCTION
The blood-brain barrier (BBB) is a complex and dynamic physical barrier composed of non-fenestrated endothelial cells, pericytes and astrocytes that restricts the passage of molecules and maintains brain homeostasis (Daneman & Prat, 2015).Brain endothelial cells express ATP-binding cassette efflux (ABC) transporters that are major contributors to the highly restrictive function of the BBB, limiting entry of substrates from the blood to the central nervous system (CNS) (Morris et al., 2017).Furthermore, ABC transporters may also contribute to the transport of compounds from the brain parenchyma to the systemic circulation, reducing CNS accumulation (Bauer et al., 2021).
Studies have identified ABCB1 (P-glycoprotein [P-GP]) and ABCG2 (breast cancer resistance protein [BCRP]) as the two most prominent transporters that are highly expressed in the luminal domain of BBB endothelial cells in many species.Their substrate specificity is broad and includes an array of xenobiotics, including therapeutic drugs and endogenous mediators (Chen et al., 2016;Qosa et al., 2015).
The multidrug resistance-associated protein 5 (MRP5; ABCC5), a member of the ABCC subfamily, is also highly expressed in human, porcine, rat and bovine brain endothelial cells (Warren et al., 2009).Compared with ABCB1 and ABCG2, the substrate profile of ABCC5 is more limited, though it is known to efflux organic anions (McAleer et al., 1999), folates (Wielinga et al., 2005), the cyclic nucleotides cAMP and cGMP (Jedlitschky et al., 2000), nucleoside monophosphate analogues and glutathione conjugates (Wijnholds et al., 2000) and antiviral and anticancer agents (Ritter et al., 2005).Although ABCC5 is highly expressed in porcine (Kubo et al., 2015;Shubbar & Penny, 2018;Warren et al., 2009), bovine (Zhang et al., 2000) and human (Warren et al., 2009) brain endothelial cells, it is relatively understudied and little is known of the regulatory mechanisms that govern expression of this transporter.
NR3C1, the glucocorticoid receptor (GR), has also been shown to influence the expression of 173 ABCB1 and ABCG2 in brain endothelial cells (Ghosh et al., 2018;Narang et al., 2008), but the regulatory role of NR3C1 has not yet been precisely defined.Studies carried out in rodent models have demonstrated that the glucocorticoid dexamethasone upregulates the expression of ABCB1 and ABCG2 in brain endothelial cells, with the authors postulating this may be due to the drug being a dual NR1I2/NR3C1 ligand (Narang et al., 2008).However, later studies reported that dexamethasone is not an NR1I2 ligand in higher species, including porcine (Kublbeck et al., 2015), and findings in porcine and human BBB models are contrasting.To date, the NR3C1-mediated effects on the activity and expression of ABCB1 and ABCG2 at the BBB have not been well characterised and the majority of the studies have focused on the role of NR3C1 in regulating expression of drug metabolising enzymes in human and rodent liver (Cooper et al., 2008;Huss & Kasper, 2000;Novotna & Dvorak, 2014;Pascussi et al., 2001).
In the present study, we sought to clarify the regulatory role of NR1I2, NR1I3 and NR3C1 on the activity and expression of ABCB1, ABCG2 and ABCC5 in the BBB, employing a well-established BBB model composed of primary porcine brain endothelial cells (PBECs).This model is physiologically relevant to human and possesses many of the key features of the in vivo BBB (Cantrill et al., 2012;Shubbar & Penny, 2018, 2020;Skinner et al., 2009;Torres-Vergara & Penny, 2018).

| Isolation and culture of porcine brain endothelial cells (PBECs)
Fresh brains from male and female Landrace cross Large White pigs, aged 22-24 weeks, weighing 105-110 kg, were obtained from a local

What is already known
• NR1I2 and NR1I3 up-regulate ABCB1 and ABCG2 transporter activity and protein expression in BBB.
• NR3C1 up-regulates ABCB1 and ABCG2 transporter activity and protein expression in BBB.

What does this study add
• NR3C1 agonist up-regulates ABCC5 protein expression.

What is the clinical significance
• Understanding regulation of BBB ABC transporters will facilitate drug delivery into the brain • It will also help with understanding the transport of endogenous mediators into the brain.
abattoir.The isolation of porcine brain microvessels was performed as described previously (Cantrill et al., 2012;Skinner et al., 2009).The isolated microvessels were seeded into rat type I collagen (125 μgÁml À1 )/human fibronectin (7.5 μgÁml À1 ) coated six-well plates and maintained in growth medium containing phenol red free low-glucose Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% (v/v) plasma-derived serum, 125 μgÁml À1 of heparin, 100 UÁml À1 of penicillin, 100 mgÁml À1 of streptomycin and 2 mM of glutamine in a humidified atmosphere at 37 C with 5% CO 2 for 24 h.Microvessels were then treated with puromycin dihydrochloride, 4 μgÁml À1 , for 48 h and subsequently maintained in 1:1 mixture of growth medium/astrocyte-conditioned medium for 6 days.PBEC lysates were then used for western blotting studies or subcultured into collagen/fibronectin-coated 96-well plates, 25,000 cells per well, for transporter activity assays and 25,000 cells per well in CELLview 10-well microscopy chambers for immunocytochemistry studies.

| Cell treatment with nuclear receptor ligands
PBECs were maintained in treatment medium composed of phenol red-free low-glucose Dulbecco's Modified Eagle's Medium supplemented with 1% (v/v) FBS and 2 mM of L-glutamine.After 1 h of equilibration, PBECs were incubated with selected concentrations of the human NR1I2 (hNR1I2) agonist rifampicin, L-sulforaphane (hNR1I2 antagonist), the human NR1I3 (hNR1I3) agonist CITCO, meclizine (NR1I3 inverse agonist), dexamethasone (NR3C1 agonist) and mifepristone (NR3C1 antagonist) at 37 C with 5% CO 2 for either 30 min or 24 h.All the compounds were made up in DMSO and the final concentration of DMSO for cell treatment never exceeded 0.001% (v/v).This concentration of solvent did not significantly affect ABCB1, ABCG2 and ABCC5 activity and expression.Compounds were used at non-cytotoxic concentrations, as determined in advance by the Neutral Red cell viability assay (data not shown).

| Measurement of ABCB1, ABCG2 and ABCC5 functional activity
PBECs in 96-well plates were washed twice with warm PBS and equilibrated with phenol red-free low-glucose Dulbecco's Modified Eagle's Medium supplemented with 2 mM of L-glutamine at 37 C with 5% CO 2 for 30 min.PBEC monolayers were then incubated with 0.5 μM of calcein-AM (to measure ABCB1 activity), 1 μM of Hoechst 33342 (to measure ABCG2 activity) or 4 μM of chloromethylfluorescein diacetate (CMFDA) [to measure ABCC5 activity] for 30 min.
PBECs were washed twice with ice-cold PBS and 100 μl of PBS was added to each well.
Intracellular fluorescence of calcein (excitation, 484 nm; emission, 530 nm), Hoechst 33342 (excitation, 370 nm; emission, 450 nm) and glutathione methylfluorescein (GS-MF) (excitation, 492 nm; emission, 516 nm) was measured immediately using a Hidex sense microplate reader (Hidex, Finland).Relative fluorescence units (RFU) were normalised to the protein content of each individual well and functional activity is expressed as % relative fluorescence units μg À1 protein compared with non-treated control.Protein content of cell lysates was determined using the Bradford reagent (Bio-Rad) using a standard curve of bovine serum albumin.

| Western blotting
PBECs were treated with the selected compounds and cell lysates were prepared using CelLytic M lysis buffer (Sigma-Aldrich) with 0.1% protease inhibitor cocktail (Sigma-Aldrich).Lysates were centrifuged at 15,000 g at 4 C for 15 min and the protein content of supernatants was determined using the Bradford reagent (Bio-Rad), using a standard curve of bovine serum albumin.
To detect ABCG2 and NR3C1 expression, cell lysates (15 μg of protein) were separated using an 8% sodium dodecyl sulfate (SDS)polyacrylamide gel, to detect NR1I2, NR2B1 (retinoid X receptor; RXR) and NR1I3 expression, those of 40 μg of protein were separated using a 10% SDS-polyacrylamide gel and for ABCB1 and ABCC5, those of 80 μg of protein were separated using an 8% SDS-polyacrylamide gel.
Electrophoresis was carried out at 120 V and proteins were subsequently electro-transferred onto PVDF membranes in 10 mM of N-cyclohexyl-3-aminopropanesulfonic acid (CAPS) buffer, pH 11.
For every experiment, the PVDF membrane blot was cut into half so that half of the blot was probed with antibody to detect protein of interest and the other half was probed with antibody to detect housekeeping protein as loading control.All the western blotting procedures and analysis were conducted in accordance with the recommendations detailed in the British Journal of Pharmacology (Alexander et al., 2018).

| Immunocytochemical analyses of NR1I2 and NR1I3
PBECs were grown in collagen/fibronectin-coated CELLview 10-well microscopy chambers and treated with nuclear receptor ligands as described above.After treatment, PBECs were fixed with 4% (w/v) formaldehyde for 10 min, the formaldehyde was quenched by incubation with 0.1 M of Tris buffer, pH 7.5, for 5 min and the cells were then permeabilised by incubation with 0.2% (v/v) Triton X-100 for 10 min.After washing with PBST (PBS and 0.1% [v/v] Tween 20), the cells were incubated with 10% (v/v) normal goat serum (Abcam) in PBST for 1 h.For detection of NR1I2 and NR1I3, fixed cells were incubated for 1 h at room temperature with either rabbit antipregnane X receptor polyclonal antibody or rabbit anti-constitutive androstane receptor polyclonal antibody (Abcam), 1:50 dilution in 5% (v/v) normal goat serum in PBST.All antibody concentrations used for immunocytochemistry were optimised within our laboratory.Cell monolayers were then washed with PBST and incubated with goat anti-rabbit IgG H&L (Alexa Fluor-488) pre-adsorbed secondary antibody (Abcam), 1:200 dilution in 5% (v/v) normal goat serum in PBST for 1 h in the dark at room temperature.After washing with PBST, the cells were incubated with 5 μgÁml À1 of DAPI (4 0 ,6-diamidino-2-phenylindole; Sigma-Aldrich) for 15 min at room temperature in the dark.Slides were mounted with Prolong Diamond antifade solution (Life Technologies, UK) and cured overnight at 4 C. Negative controls were performed in parallel, where the cells were treated with 5% (v/v) normal goat serum in PBST with no primary antibody.Samples were analysed using an IX83 inverted deconvolution microscope (Olympus, Japan) with a 100Â/1.35UPlanApo objective through fluorescein isothiocyanate (FITC) and DAPI filters.Images were captured with an R6 QImaging charged-coupled device (CCD) camera with a Z optical spacing of 0.2 μm controlled by the Metamorph v7.8.4.0 software (Olympus).Raw images were deconvolved using the Huygens Pro software (SVI) (Scientific Volume Imaging, Netherlands) and further processed with ImageJ software (NIH, USA).All the Immuno-related procedures used comply with the recommendations made by the British Journal of Pharmacology (Alexander et al., 2018).

| Data and statistical analysis
The data and statistical analysis comply with the recommendations of the British Journal of Pharmacology on experimental design and analysis in pharmacology (Curtis et al., 2022).Data analysis for all treatments with selected compounds was not blinded because these experiments were carried out by a single person and blinding would likely contribute to complications in data analysis.However, every effort was made to vary the location within the plate and the order in which compounds were added to 96-well plates, 6-well plates and CELLview 10-well microscopy chambers to minimise the potential confounds of evaporation or unequal exposure time.Experiments were carried out at n ≥ 5, as stated in the figure legends, where n = number of independent experiments.Post hoc tests were conducted when the F value achieved the necessary level (P < 0.01).Data are presented as mean ± SD.Differences were significant when P < 0.01.Based on the reanalysis of the data within GraphPad Prism by the ROUT method (with Q number set to the most rigorous setting), the data from one of the independent experiments met the criteria as an outlier.Therefore, the data from one independent experiment have been excluded from Figure 1c.Cell viability and transporter assay data were analysed by parametric one-way analysis of variance (ANOVA) followed by a Tukey's post hoc test, for comparing multiple groups, using GraphPad Prism software Version 7.00 (GraphPad Software, Inc., California, USA, RRID:SCR_002798).
The data for western blots and immunocytochemistry were normalised to the control group in order to control unwanted sources of variation and were thus analysed non-parametrically with the Kruskal-Wallis test with Dunn's post hoc analysis, using GraphPad Prism software Version 7.00.

| Nomenclature of targets and ligands
Key protein targets and ligands in this article are hyperlinked to corresponding entries in the IUPHAR/BPS Guide to PHARMACOLOGY http://www.guidetopharmacology.org and are permanently archived in the Concise Guide to PHARMACOLOGY 2021/22 (Alexander, Cidlowski, et al., 2021;Alexander, Kelly, et al., 2021).

| Effect of ligand-based activation of NR1I2 on ABCB1, ABCG2 and ABCC5 transporter activities and protein expression
Treatment of PBECs for 24 h with the NR1I2 agonist rifampicin, at a concentration of 10 μM, significantly decreased the intracellular ) and antagonist (L-sulforaphane [L-SFN]) on ABCB1, ABCG2 and ABCC5 activities in porcine brain endothelial cells.Effects of long-term exposure of RF and L-SFN on intracellular accumulation of calcein (a), Hoechst 33342 (b) and glutathione methylfluorescein (GS-MF) (c) in cells treated with 10 μM of RF and 5 μM of L-SFN for 24 h.Effects of short-term exposure of RF and L-SFN on intracellular accumulation of calcein (d), Hoechst 33342 (e) and GS-MF (f ) in cells treated with 10 μM of RF and 5 μM of L-SFN for 30 min.Data were analysed using parametric one-way analysis of variance followed by a Tukey's post hoc test and are presented as mean ± SD (n = 4 for (c) where one outlier was excluded and n = 5 for (a), (b) and (d)-(f)) with three replicates in each independent experiment.*P < 0.01, significantly different as indicated.accumulation of Hoechst 33342 but had no significant effect on calcein and GS-MF (Figure 1a-c), consistent with increased ABCG2 activity.
To determine whether the effects of rifampicin and L-sulforaphane treatments are related to direct, short-term interaction with the transporters or inhibition of ABCB1, ABCG2 and ABCC5 activity regulated via the NR1I2 pathway, PBECs were subjected to a short-term incubation, 30 min rather than 24 h, with either rifampicin or L-sulforaphane.Under these experimental conditions, exposure to 10 μM of rifampicin alone and 5 μM of L-sulforaphane alone did not significantly affect the intracellular accumulation of calcein, Hoechst 33342 and GS-MF (Figure 1d-f) when compared with control.
However, a 30-min co-treatment with rifampicin and L-sulforaphane significantly enhanced the accumulation of calcein by 2.3-fold and Hoechst 33342 and GS-MF by over 1.6-fold.Although this cotreatment had some effect at 30 min, the extent of the effects is substantially lower than is observed with a 24-h incubation.
The regulatory role of NR1I2 on the functional activity of ABCB1, ABCG2 and ABCC5 transporters observed in this study was confirmed by western blotting (Figure 2a-c).Densitometric analysis revealed that 24 h of treatment with rifampicin and L-sulforaphane did not significantly affect the expression of ABCB1, ABCG2 and ABCC5 (Figure 2); however, co-treatment with rifampicin and L-sulforaphane significantly decreased ABCB1 protein expression (Figure 2a,d).

| Effect of ligand-based activation of NR1I3 on ABCB1, ABCG2 and ABCC5 transporter activities and protein expression
Treatment of PBECs with the NR1I3 agonist CITCO, 5 μM for 24 h, significantly increased the intracellular accumulation of calcein, Hoechst 33342 and GS-MF by 6.8-fold, 1.8-fold and 2.4-fold, respectively (Figure 3a-c).
F I G U R E 2 Effects of 24-h treatment with NR1I2 agonist (rifampicin [RF]) and antagonist (L-sulforaphane [L-SFN]) on the protein expression of ABCB1 (a), ABCG2 (b) and ABCC5 (c) were detected by western blotting.Densitometric quantification of the relative protein expression levels of ABCB1 (d), ABCG2 (e) and ABCC5 (f ).Data were analysed using the non-parametric Kruskal-Wallis test followed by a Dunn's post hoc test and are presented as mean ± SD (n = 5).*P < 0.01, significantly different as indicated.
In order to determine whether the effects of CITCO and meclizine treatments are related to direct short-term interaction with the transporters, that is, inhibition of ABCB1, ABCG2 and ABCC5 activity by the NR1I3 agonist and inverse agonist, PBECs were treated for 30 min, rather than 24 h, with CITCO and meclizine.Under these experimental conditions, exposure to 5 μM of CITCO resulted in a significant 2.3-fold increase in the intracellular accumulation of calcein and over 1.4-fold increases in intracellular accumulation of Hoechst 33342 and GS-MF (Figure 3d-f).Treatment with meclizine, 10 μM, resulted in a significant 2.8-fold increase in intracellular accumulation of calcein and significant increases of over 1.3-fold and 1.4-fold in intracellular accumulation of Hoechst 33342 and GS-MF, respectively (Figure 3d-f).Interestingly, cotreatment with CITCO and meclizine significantly enhanced the accumulation of calcein by 10.5-fold and Hoechst 33342 and GS-MF by over 2.0-fold (Figure 3d-f).
To determine the role of NR1I3 in regulating ABCB1, ABCG2 and ABCC5 transporter expression, western blotting (Figure 4a-c) and densitometric analysis (Figure 4d-f) were carried out.Treatment of PBECs with CITCO or meclizine for 24 h did not significantly affect the ABCB1, ABCG2 and ABCC5 protein expression.Interestingly, with co-incubation with CITCO and meclizine, ABCB1, ABCG2 and ABCC5 protein expression was significantly reduced by 75.3%, 53.9% and 67.4%, respectively, compared with control.

| Effect of NR1I2 ligands on the expression of NR1I2 and NR2B1
To exert its cellular effects, NR1I2 dimerises with NR2B1 (retinoid x receptor).Thus far, no studies report the effect of any NR1I2 agonist ) and inverse agonist (meclizine [MC]) on ABCB1, ABCG2 and ABCC5 activity and expression in porcine brain endothelial cells.Effects of long-term exposure of CT and MC on intracellular accumulation of calcein (a), Hoechst 33342 (b) and glutathione methylfluorescein (GS-MF) (c) in cells treated with 5 μM of CT and 10 μM of MC for 24 h.Effects of short-term exposure of CT and MC on intracellular accumulation of calcein (d), Hoechst 33342 (e) and GS-MF (f) in cells treated with 5 μM of CT and 10 μM of MC for 30 min.Data were analysed using parametric one-way analysis of variance followed by a Tukey's post hoc test and are presented as mean ± SD (n = 5) with three replicates in each independent experiment.*P < 0.01, significantly different as indicated.
or antagonist on the expression of NR1I2 and NR2B1 at the protein level.Therefore, western blotting studies were carried out to determine the effect of the NR1I2 agonist rifampicin and antagonist L-sul- foraphane on NR1I2 and NR2B1 protein expression in PBECs.
Exposure to either rifampicin or L-sulforaphane, or co-treatment with both compounds, for 24 h, did not significantly affect the protein expression of NR1I2 (Figure 5a,c) and NR2B1 (Figure 5b,d) compared with vehicle-treated control cells.

| Effect of NR1I3 ligands on the expression of NR1I3 and NR2B1
As with NR1I2, NR1I3 dimerises with NR2B1 to exert its cellular effects.To date, no studies report the effect of any NR1I3 agonist or inverse agonist on the expression of NR1I3 and NR2B1 at the protein level.Hence, western blotting studies were carried out to investigate the effect of CITCO and meclizine on NR1I3 and NR2B1 protein expression in PBECs.Treatment with either CITCO or meclizine, or a co-incubation with both, for 24 h, did not significantly modify the expression of NR1I3 (Figure 6a,c) and NR2B1 (Figure 6b,d) compared with vehicle-treated control cells.

| Effect of NR1I3 ligands on nuclear translocation of NR1I3
Immunocytochemistry studies were carried out to examine the subcellular distribution of NR1I3.In control studies, NR1I3-associated fluorescence (green) was widely distributed in nuclear, perinuclear and cytoplasmic regions (Figure 8a).Exposure to the NR1I3 agonist CITCO or the NR1I3 inverse agonist meclizine, or co-treatment with both, did not alter this localisation pattern.No apparent non-specific binding of the Alexa Fluor ® 488-conjugated secondary antibody was observed.
In order to quantify the effects of CITCO and meclizine on the distribution of NR1I3-associated fluorescence, semi-quantitative immunofluorescence analysis was performed in randomly selected regions from the cytoplasm and nucleus using digital image analysis in approximately 100 cells (Figure 8b).The analyses confirmed that treatment with CITCO or meclizine, or both, did not influence the ratio of nuclear NR1I3 fluorescence intensity to cytosolic NR1I3 fluorescence intensity, compared with non-treated control cells.

| Effect of NR3C1 ligands on ABCB1, ABCG2 and ABCC5 activities and expression
To investigate whether NR3C1 is implicated in regulating ABC efflux transporter activity and expression in PBECs, the NR3C1 agonist dexamethasone and the selective antagonist mifepristone were used.
Treatment with dexamethasone significantly decreased intracellular accumulation of calcein, Hoechst 33342 and GS-MF by 39.3%, 20.1% and 28.5%, respectively (Figure 9a-c), reflecting increased transporter activity.However, co-treatment with dexamethasone and mifepristone counteracted the dexamethasone-mediated increase in ABCB1, it significantly attenuated dexamethasone-induced up-regulation, with expression levels reverting to similar levels observed in the control condition.

| Effect of NR3C1 ligands on the expression of NR1I2, NR1I3, NR2B1 and NR3C1
To determine the ability of NR3C1 agonist (dexamethasone) and
NR1I2 activation with rifampicin significantly up-regulated ABCG2 transporter activity after 24 h of exposure, consistent with Lemmen, Tozakidis and Galla (2013).Increases in ABCB1 and ABCC5 transporter activity and ABCG2, ABCB1 and ABCC5 protein expression were also observed, although the increases were not significant.This finding is somewhat consistent with those of McInerney et al. (2017), who reported that rifampicin treatment significantly increased ABCB1 expression in the hCMEC/D3 human brain endothelial cell line.Cotreatment with L-sulforaphane, a selective NR1I2 antagonist, completely blocked the rifampicin-mediated increase in ABCG2, ABCB1 and ABCC5 activity.The finding that co-treatment with rifampicin and sulforaphane reduced protein expression is unexpected and, at this stage, we are unable to explain the mechanism.Further studies would be required to identify the precise mechanism.
In this study, the increase in ABCG2 activity by rifampicin was not accompanied by an increase in protein expression.This finding is somewhat consistent with that of Lemmen, Tozakidis, Bele and Galla (2013) who reported that rifampicin mediated increases in ABCG2 and ABCB1 activity and protein expression.The apparent differences may be due to the fact that Lemmen, Tozakidis and Galla (2013) employed different methods for the isolation of PBECs and different culture media, which may cause the cells to respond differently to rifampicin treatment.
NR1I2 antagonist treatment significantly reduced ABCB1 endogenous activity.This finding is consistent with that of Zhou et al. activities, suggesting that together they exert a synergistic inhibitory effect.ABCB1, ABCG2 and ABCC5 possess broad substrate specificity and it is possible that the combination of L-sulforaphane and rifampicin binding, even for 30 min, significantly inhibits transporter activity.
The NR1I2 ligands rifampicin and L-sulforaphane did not significantly affect NR1I2 expression, suggesting that regulation of transporter expression is independent of changes in NR1I2 expression.To date, no study has reported the effect of rifampicin or L-sulforaphane on NR1I2 protein expression in BBB endothelial cells.The lack of effect of rifampicin on NR1I2 protein expression is consistent with its lack of effect on NR1I2 mRNA expression in human hepatocytes (Pascussi, Drocourt, et al., 2000) and LS180, Caco-2 and TC-7 cell lines (Pfrunder et al., 2003).Our findings, and those of Pascussi, Drocourt, et al. (2000) and Pfrunder et al. (2003), contrast those of Ott et al. (2009) and Lemmen, Tozakidis and Galla (2013) who reported that rifampicin increased NR1I2 mRNA expression in PBECs, although the latter two studies did not report the effect of rifampicin on NR1I2 protein expression.Furthermore, Ott et al. (2009) et al., 2005).Thus, the net effect of CITCO on NR1I3 activity is dependent on the ratio of coactivator and corepressor proteins, which may influence the activity state of NR1I3 (Jyrkkärinne et al., 2012;Lempiäinen et al., 2005;Makinen et al., 2003).It is possible that the PBEC system used in the current study contains a higher ratio of corepressor to coactivator proteins compared with Chan et al. (2011) and Lemmen, Tozakidis, Bele and Galla (2013), meaning CITCO acts as an inverse agonist.
We report for the first time that whilst the NR1I3 agonist CITCO decreases ABCC5 activity, it has no effect on protein expression in PBECs, although the NR1I3 ligands phenobarbital and polychlorinated biphenyl 99 significantly repressed ABCC5 mRNA expression in rat hepatocytes in vivo (Maher et al., 2006).In contrast, Merrell et al. (2008) reported phenobarbital induced ABCC5 mRNA expression in hepatocytes of female rats but observed no effect in male rats in vivo, whilst Ambroziak et al. (2010) reported phenobarbital had no effect on ABCC5 mRNA levels in an immortalised rat brain endothelial cell line.These apparent discrepancies may be explained by the fact that unlike CITCO, phenobarbital activates NR1I3 via an indirect, rather than a direct, mechanism (Mutoh et al., 2013).
In the current study, in PBECs, meclizine had a potent synergistic effect when combined with CITCO, down-regulating ABCB1, ABCG2 and ABCC5 activity and expression.This contrasts with the findings In the current study, treatment of PBECs with either CITCO or meclizine alone, or combined did not affect NR1I3 protein expression and nuclear translocation, consistent with the studies of Chan et al. (2011), who observed that NR1I3 is expressed throughout hCMEC/ D3 cells and that CITCO did not induce the nuclear translocation of NR1I3 (Chan et al., 2011).Furthermore, NR1I3 is constitutively expressed with nuclear localisation and activated even in the absence of ligand (Guo et al., 2007;Kanno et al., 2005;Kawamoto et al., 1999;Zelko et al., 2001).However, Lemmen, Tozakidis, Bele and Galla (2013) reported that NR1I3 is primarily expressed cytoplasmically and that on treatment with CITCO, it translocates to the nucleus.The difference in findings may be explained by the fact that whilst Lemmen, Tozakidis, Bele and Galla (2013) found CITCO to be an NR1I3 agonist, in the current study, CITCO is not an NR1I3 agonist, possibly due to differences in the relative proportions of NR1I3 coactivators and corepressors in the two studies.
fluorescence (green) was predominantly localised to the cytoplasm (Figure 7a).Exposure to the NR1I2 agonist rifampicin caused a redistribution of NR1I2-associated fluorescence, resulting in an increase in nuclear fluorescence intensity, suggesting translocation of NR1I2 from cytoplasm into nucleus.Furthermore, the rifampicin-mediated NR1I2 nuclear translocation was completely blocked by co-treatment with the NR1I2 antagonist L- sulforaphane.No apparent non-specific binding of the Alexa Fluor ® 488-conjugated secondary antibody was observed.In order to F I G U R E 4 Effects of 24-h treatment with NR1I3 agonist (CITCO [CT]) and inverse agonist (meclizine [MC]) on the protein expression of ABCB1 (a), ABCG2 (b) and ABCC5 (c) in porcine brain endothelial cells were detected by western blotting.Densitometric quantification of the relative protein expression levels of ABCB1 (d), ABCG2 (e) and ABCC5 (f).Data were analysed using the non-parametric Kruskal-Wallis test followed by a Dunn's post hoc test and are presented as mean ± SD (n = 5).*P < 0.01, significantly different as indicated.quantify the effects of NR1I2 agonist rifampicin and antagonist Lsulforaphane on NR1I2 translocation, semi-quantitative immunofluorescence analysis was performed in randomly selected regions of the cytoplasm and nucleus using digital image analysis in approximately 100 cells.Semi-quantitative analysis confirmed that rifampicin caused a significant 4.0-fold increase in the ratio of nuclear NR1I2 fluorescence intensity to cytosolic NR1I2 fluorescence intensity compared with non-treated control cells (Figure 7b).Coincubation with rifampicin and L-sulforaphane counteracted the rifampicin-mediated nuclear translocation of NR1I2, with the ratio of nuclear NR1I2 fluorescence intensity to cytosolic NR1I2 fluorescence intensity reverting to a similar level observed in the control condition.Treatment of PBECs with L-sulforaphane alone did not significantly alter the nuclear NR1I2 to cytosolic NR1I2 fluorescence ratio.
ABCG2 and ABCC5 transporter activities, reducing activities of the three transporters to levels observed in non-treated control cells.Treatment with mifepristone alone had no significant effect on the activities of all three transporters.To determine whether the NR3C1-mediated increases in ABCB1, ABCG2 and ABCC5 activities are associated with changes in transporter expression, western blotting and densitometric analysis were performed.Dexamethasone treatment significantly increased ABCB1 (Figure 9d,g), ABCG2 (Figure 9e,h) and ABCC5 (Figure 9f,i) expression.Whilst mifepristone alone had no significant effect on expression of all three transporters, F I G U R E 5 Effect of NR1I2 agonist (rifampicin [RF]) and antagonist (L-sulforaphane [L-SFN]) on NR1I2 and NR2B1 protein expression in porcine brain endothelial cells.Cells were treated with 10 μM of RF and 5 μM of L-SFN for 24 h.Representative western blot image of NR1I2 (a) and NR2B1 (b).Densitometric quantification of the relative protein expression levels of NR1I2 (c) and NR2B1 (d).Data were analysed using the nonparametric Kruskal-Wallis test followed by a Dunn's post hoc test and are presented as mean ± SD (n = 5).NS, not significantly different as indicated.
antagonist (mifepristone) to influence expression of the NR1I2, NR1I3, NR2B1 and NR3C1 receptors in PBECs, western blotting and densitometric analysis were performed.Exposure to dexamethasone or mifepristone or their co-treatment did not significantly affect the protein expression of NR1I2 (Figure 10a,e), NR1I3 (Figure 10b,f), NR2B1 (Figure 10c,g) and NR3C1 (Figure 10d,h), compared with control cells.

( 2016 )
who demonstrated that L-sulforaphane reduced ABCB1 mRNA expression.Although the mechanism of L-sulforaphane-mediated down-regulation of transporter expression is unknown in PBECs, the antagonist may block ABCB1 gene expression mediated by a non-NR1I2-dependent pathway.Short-term (30 min) treatment with rifampicin or L-sulforaphane alone had no effect on ABCB1, ABCG2 and ABCC5 activity, suggesting no direct inhibition.Interestingly, however, co-treatment of PBECs with agonist and antagonist significantly decreased all transporter F I G U R E 6 Effect of NR1I3 agonist (CITCO [CT]) and inverse agonist (meclizine [MC]) on NR1I3 and NR2B1 protein expression in porcine brain endothelial cells.Cells were treated with 5 μM of CT and 10 μM of MC for 24 h.Representative western blot image of NR1I3 (a) and NR2B1 (b).Densitometric quantification of the relative protein expression levels of NR1I3 (c) and NR2B1 (d).Data were analysed using the nonparametric Kruskal-Wallis test followed by a Dunn's post hoc test and are presented as mean ± SD (n = 5).NS, not significantly different as indicated.

F
I G U R E 7 Effect of NR1I2 agonist (rifampicin[RF]) and antagonist (L-sulforaphane [L-SFN]) on NR1I2 nuclear translocation.Cells were treated with 10 μM of RF and 5 μM of L-SFN for 24 h.NR1I2 protein was localised with anti-pregnane X receptor polyclonal antibody (1:20) and Alexa Fluor ® 188-conjugated secondary antibody (1:500).Nucleus was visualised with 4 0 ,6-diamidino-2-phenylindole (DAPI) staining.Control is untreated cells, stained with DAPI and both anti-pregnane X receptor polyclonal antibody and Alexa Fluor ® 188-conjugated secondary antibody and negative control is untreated cells, stained with only the Alexa Fluor ® 188-conjugated secondary antibody and DAPI.(a) Representative deconvolution microscopy image of immunocytochemical localisation of NR1I2 (green) and the nucleus (blue) in porcine brain endothelial cells.(b) Quantitative analysis of Alexa Fluor ® 188 fluorescence was performed in randomly selected regions of cytoplasm and nucleus.NR1I2 fluorescence ratio (nuclear to cytosolic) normalised to control was analysed using the non-parametric Kruskal-Wallis test followed by a Dunn's post hoc test.Data are presented as mean ± SD (n = 5, with 3 replicates in each independent experiment, with at least 30 ratios from 5 to 10 cells in each replicate).*P < 0.01, significantly different as indicated.Scale bar: 10 μm.
and   Lemmen, Tozakidis and Galla (2013)  employed different methods for F I G U R E 8 Effect of NR1I3 agonist (CITCO[CT]) and inverse agonist (meclizine[MC]) on NR1I3 nuclear translocation.Cells were treated with 5 μM of CT and 10 μM of MC for 24 h.NR1I3 protein was localised with anti-constitutive androstane receptor polyclonal antibody (1:20) and Alexa Fluor ® 188-conjugated secondary antibody (1:500).Nucleus was visualised with 4 0 ,6-diamidino-2-phenylindole (DAPI) staining.Control is untreated cells, stained with DAPI and both anticonstitutive androstane receptor polyclonal antibody and Alexa Fluor ® 188-conjugated secondary antibody and negative control is untreated cells, stained with only the Alexa Fluor ® 188-conjugated secondary antibody and DAPI.(a) Representative deconvolution microscopy image of immunocytochemical localisation of NR1I3 (green) and the nucleus (blue) in porcine brain endothelial cells.(b) Quantitative analysis of Alexa Fluor ® 188 fluorescence was performed in randomly selected regions of cytoplasm and nucleus.NR1I3 fluorescence ratio (nuclear to cytosolic) normalised to control was analysed using the non-parametric Kruskal-Wallis test followed by a Dunn's post hoc test.Data are presented as mean ± SD (n = 5, with 3 replicates in each independent experiment, with at least 30 ratios from 5 to 10 cells in each replicate).Scale bar: 10 μm.PBEC isolation and culture, which may account for the differences observed.Importantly, NR1I2 expression is controlled by complex transcriptional, post-transcriptional and post-translational mechanisms(Mackowiak & Wang, 2016), and differences in such mechanisms themselves may contribute to the differences observed.Here, we demonstrate rifampicin-dependent activation and significant nuclear accumulation of NR1I2 in PBECs, consistent withOtt et al. (2009) andLemmen, Tozakidis and Galla (2013)  who reported rifampicin-mediated activation of NR1I2 in PBECs.Even though we observed NR1I2 nuclear translocation, protein expression of all three ABC transporters was increased, but not significantly.Furthermore, the current study is the first to report that a NR1I2 antagonist (L-sulforaphane) blocks NR1I2 nuclear translocation and provides evidence that NR1I2 nuclear translocation does not necessarily lead to significant increases in ABC transporter protein expression.Increased ABCB1, ABCG2 and ABCC5 activity, without a significant increase in expression at the protein level, can be explained by the fact that the cell's trafficking machinery can significantly alter ABC transporter activity by rapid insertion and removal of transporters at the plasma membrane (Geisler & Hegedűs, 2020; Harris F I G U R E 9 Effects of 24-h treatment with NR3C1 agonist (dexamethasone [DX]) and antagonist (mifepristone [MF]) on ABCB1, ABCG2 and ABCC5 activity and expression in porcine brain endothelial cells.Cells were treated with 10 μM of DX and 1 μM of MF for 24 h.Effects of DX and MF on intracellular accumulation of calcein (a), Hoechst 33342 (b) and glutathione methylfluorescein (c).Data were analysed using one-way analysis of variance followed by a Tukey's post hoc test and are presented as mean ± SD (n = 5).*P < 0.01, significantly different from control.Effects of DX and MF on the protein expression of ABCB1 (d), ABCG2 (e) and ABCC5 (f) were detected by western blotting.Densitometric quantification of the relative protein expression levels of ABCB1 (g), ABCG2 (h) and ABCC5 (i).Data were analysed using the non-parametric Kruskal-Wallis test followed by a Dunn's post hoc test and are presented as mean ± SD (n = 5).*P < 0.01, significantly different as indicated.NS, not significantly different as indicated.et al., 2018; Roston et al., 2012), causing the regulation of ABC transporter activity independent of changes in protein expression.Ligand-based activation of NR1I3 with CITCO significantly decreased ABCB1 and ABCG2 functional activity but not protein expression, contrasting the findings of Lemmen, Tozakidis, Bele and Galla (2013) who reported that CITCO increases ABCB1 and ABCG2 expression in PBECs and Chan et al. (2011) who reported that CITCO increases ABCB1 expression in hCMEC/D3 cells.However, CITCO can act as an agonist or inverse agonist, recruiting either coactivators or corepressors (Jyrkkärinne et al., 2012; Lempiäinen F I G U R E 1 0 Effect of 24-h treatment with NR3C1 agonist (dexamethasone [DX]) and antagonist (mifepristone [MF]) on NR1I2, NR1I3, NR2B1 and NR3C1 protein expression in porcine brain endothelial cells.Cells were treated with 10 μM of DX and 1 μM of MF for 24 h.Representative western blot image of NR1I3 (a), NR1I3 (b), NR2B1 (c) and NR3C1 (d).Densitometric quantification of the relative protein expression levels of NR1I3 (e), NR1I3 (f), NR2B1 (g) and NR3C1 (h).Data were analysed using the non-parametric Kruskal-Wallis test followed by a Dunn's post hoc test and are presented as mean ± SD (n = 5).*P < 0.01, significantly different as indicated.

of
Chan et al. (2011) andLemmen, Tozakidis and Galla (2013)  who demonstrated that meclizine partially reduced the CITCO-mediated increase in ABCB1 and ABCG2 activity and expression in hCMEC/D3 cells and PBECs, respectively.Meclizine also significantly reduced the activity of ABCB1, ABCG2 and ABCC5.This can be attributed to the ability of meclizine to reduce endogenous NR1I3 activity, as demonstrated byHuang et al. (2004) andLi et al. (2009) in transfected HepG2 cells.The current study is the first to report the role of the NR1I3 ligands CITCO and meclizine in regulating ABCC5 protein expression in BBB endothelial cells.
ity and protein expression in BBB endothelial cells.As NR3C1 antagonist completely abolished the actions of dexamethasone, it is likely that the actions of the steroid are mediated exclusively by ligandbased activation of NR3C1.In confirmatory studies, the NR3C1 ligand hydrocortisone increased the activity of ABCB1, ABCG2 and ABCC5, whilst co-treatment with hydrocortisone and mifepristone completely blocked this effect, substantiating the role of NR3C1 in regulating ABC transporter activity.
),Wang et al. (2003) andPascussi et al. (2003) document a glucocorticoid response element in the NR1I3 gene promoter, although no glucocorticoid response element has been identified in the NR1I2 gene promoter.The most plausible explanation of increased ABCG2 and ABCC5 expression is that NR3C1 may directly activate transcription of ABC transporter-encoding genes independent of NR1I2 and NR1I3 by binding to glucocorticoid responsive elements within gene promoters.To date, no studies report glucocorticoid responsive elements in the promoter regions of the ABCG2 and ABCC5 genes.5 | CONCLUSIONNR1I2 and NR1I3 significantly influence ABCB1, ABCG2 and ABCC5 activity independent of protein expression.Our findings provide the first evidence that NR3C1 (glucocorticoid receptor) regulates ABC transporter activity and expression independent of NR1I2 and NR1I3 in porcine BBB endothelial cells, a model highly relevant to human.However, future studies will seek to establish the relevance of these in vitro findings in in vivo studies.These findings advance our understanding of the complex regulatory network of interactions between BBB NR3C1, NR1I2 and NR1I3 receptors and the ABC transporters, which influence the passage of therapeutic drugs, endogenous mediators and xenobiotics into the CNS.AUTHOR CONTRIBUTIONS Yu Siong Ho: Conceptualization (equal); data curation (equal); formal analysis (lead); investigation (lead); methodology (lead); project administration (lead); validation (equal); visualization (lead); writing-original draft (lead); writing-review and editing (equal).Pablo Torres-Vergara: Conceptualization (equal); formal analysis (supporting); investigation (supporting); methodology (supporting); validation (supporting); writing-review and editing (equal).Jeffrey Penny: Conceptualization (equal); data curation (equal); formal analysis (supporting); funding acquisition (lead); investigation (supporting); methodology (supporting); project administration (supporting); supervision (lead); validation (equal); visualization (supporting); writing-review and editing (equal).