The constitutive androstane receptor and pregnane X receptor in the brain

Abstract

Since their discovery, the orphan nuclear receptors constitutive androstane receptor (CAR;NR1I3) and pregnane X receptor (PXR;NR1I2) have been regarded as master regulators of drug disposition and detoxification mechanisms. They regulate the metabolism and transport of endogenous mediators and xenobiotics in organs including the liver, intestine and brain. However, with proposals of new physiological functions for NR1I3 and NR1I2, there is increasing interest in the role of these receptors in influencing brain function. This review will summarise key findings regarding the expression and function of NR1I3 and NR1I2 in the brain, hereby highlighting the need for further research in this field.

Abbreviations

3α,5α‐THP

3α,5α‐tetrahydroprogesterone

ABC

ATP‐binding cassette

AhR

aryl hydrobon receptor

BBB

blood–brain barrier

BCRP

breast cancer resistance protein

NR1I3/CAR

constitutive androstane receptor

CITCO

6‐(4‐chlorophenyl)‐imidazo[2,1‐b]thiazole‐5‐carbaldehyde

CPT1A

Carnitine palmitoyltransferase 1

CYP450

cytochrome P450

DBD

DNA‐binding domain

G6P‐ase

glucose‐6‐phosphatase

GADD45B

DNA‐damage inducible β protein

GST

GSH‐S‐transferase

HMGCS2

3‐hydroxymethylglutaryl CoA synthase 2

LBD

ligand‐binding domain

MCP

monocrotophos

MRP

multidrug resistance‐associated protein

NR

nuclear receptor

PEPCK

phosphoenolpyruvate carboxykinase

P‐GP

P‐glycoprotein

NR1I2/PXR

pregnane X receptor

RXR/NR2B

retinoid X receptor α, β, γ

SCD‐1

stearoyl‐CoA desaturase 1

TCPOBOP

(1,4‐bis‐[2‐(3,5‐dichloropyridyloxy)]benzene

UGT

uridine glucuronosyl transferase

1. INTRODUCTION

The nuclear receptor (NR) superfamily is a group of proteins that function as cellular transcription factors of target genes responsible for the expression of proteins critical for cell development, activity and survival (Sever & Glass, 2013). In order to exert their effects, NRs must be activated via binding with ligands that are chemically diverse and include endobiotics, such as hormones, vitamins and their metabolites, dietary components including nutrients and xenobiotics including drugs and endocrine disruptors (Krasowski, Ni, Hagey, & Ekins, 2011).

To date, the NR superfamily is composed of 48 members in human, 49 in mice and 47 in the rat (Sever & Glass, 2013; Zhang et al., 2004). NRs share a common structure, composed of an N‐terminal containing the ligand‐independent transcription activation function‐1 (AF‐1) domain, a highly conserved DNA‐binding domain (DBD), which allows direct binding with response elements, and a C‐terminal ligand‐binding domain (LBD) (Huang, Chandra, & Rastinejad, 2010).

Understanding the functional properties of NRs that have a known nominal ligand, as in the case with steroid and vitamin‐activated receptors, has increased significantly due to the development of synthetic ligands with agonistic or antagonistic properties. However, with the identification of NRs with no known nominal ligands over recent decades, an increasing number of studies have attempted to elucidate their physiological functions (Banerjee, Robbins, & Chen, 2015; Mullican, Dispirito, & Lazar, 2013; Oladimeji & Chen, 2018). These so‐called orphan NRs tend to possess much broader ligand selectivity, such ligands including numerous xenobiotics (Ihunnah, Jiang, & Xie, 2011). In this regard, signalling via the constitutive androstane receptor (CAR;https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=607) and pregnane X receptor (PXR;https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=606) are major mechanisms for cellular response to exposure to xenobiotics and endogenous molecules (Chai, Zeng, & Xie, 2013; Tolson & Wang, 2010). Drugs used for treatment of disease conditions with numerous aetiologies (e.g. epilepsy, inflammatory diseases and bacterial infections) and endogenous compounds, including hormones and their metabolites, are ligands for both receptors and therefore influence the expression of metabolising enzymes and membrane transporters involved in drug disposition (Cerveny et al., 2007; Jigorel, Le Vee, Boursier‐Neyret, Parmentier, & Fardel, 2006; Martin, Riley, Back, & Owen, 2008). Although the evidence provided by in vitro and in vivo studies appears to support a potential role of both NR1I3 and NR1I2 in influencing drug disposition, recent studies have proposed additional functions (Figure 1). For example, NR1I3 and NR1I2 can exert anti‐inflammatory effects by blocking the actions of https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=996 in a cell‐ or tissue‐dependent ways (Dou et al., 2013; Malekshah, Lage, Bahrami, Afshari, & Behravan, 2012; Mencarelli et al., 2011) and influence the metabolism of endogenous compounds, including bile acids and hormones, as well as vitamins and nutrients. Interestingly, NR1I3 and NR1I2 can also regulate cell proliferation and the activity of pro‐apoptotic pathways (Oladimeji, Cui, Zhang, & Chen, 2016).

FIGURE 1.

Known biological actions of NR1I3 (CAR) and NR1I2 (PXR). Since both nuclear receptors are ubiquitously expressed in organs and tissues of mammalians, they regulate the expression of proteins involved in metabolism and transport of nutrients, endogenous mediators and xenobiotics in a tissue‐dependent manner. Furthermore, NR1I3 and NR1I2 can also influence biological processes including inflammation and apoptosis through protein‐protein interactions. Recent research have demonstrated that in the brain, apart from their role in some of the above‐mentioned processes, NR1I3 and NR1I2 may participate in complex functions such as behaviour

The effects of NR1I2 and NR1I3 were initially studied in organs including the liver and small intestine (Figure 1). In subsequent years, expression of both receptors has been reported in endothelial cells of the blood–brain barrier (BBB), with several reports focusing on the effects of ligand‐based activation of NR1I2 and NR1I3 on the expression of transporters and enzymes (Dauchy et al., 2009; Miller, 2015; Qosa, Miller, Pasinelli, & Trotti, 2015). More recently, the study of NR1I2 and NR1I3 in the brain has been extended to other cell types, including neurons, with interesting findings that could define new biological roles.

The aim of this review is to summarise the key actions of NR1I3 and NR1I2 in the brain in light of what is known in other cells and tissues.

2. AN OVERVIEW OF NR1I2 (PXR) AND NR1I3 (CAR)

NR1I2 and NR1I3 possess many structural features present in classic NRs, including the DNA‐binding domain, an N‐terminal AF‐1 domain and a C‐terminal LB ligand‐binding domain D (Khorasanizadeh & Rastinejad, 2016; Mullican et al., 2013). NR1I2 and NR1I3 are 50‐kDa proteins highly expressed in liver and intestine (Giessmann et al., 2004; Nannelli et al., 2010). Further studies have demonstrated that their expression pattern is more ubiquitous than initially thought, as other organs including brain, kidneys and heart also express both of these receptors (Lamba et al., 2004; Marini et al., 2007; Miki, Suzuki, Tazawa, Blumberg, & Sasano, 2005; Nannelli et al., 2010; Nishimura, Naito, & Yokoi, 2004; Savkur et al., 2003).

Their cellular localisation is still a matter of controversy since, depending on the cell model employed, NR1I3 and NR1I2 can exhibit a nuclear or cytosolic localisation (Kanno, Suzuki, Nakahama, & Inouye, 2005; Kawamoto et al., 1999; Saradhi, Sengupta, Mukhopadhyay, & Tyagi, 2005; Squires, Sueyoshi, & Negishi, 2004; Yoshinari, Kobayashi, Moore, Kawamoto, & Negishi, 2003). For example, in primary cultures of rat hepatocytes, the ligand‐based activation of NR1I3 induces its nuclear translocation, while in several human‐derived immortalised cell lines, NR1I3 is primarily constitutively active and localised in the nucleus (Chan, Hoque, Cummins, & Bendayan, 2011; Kanno et al., 2005; Kobayashi, Sueyoshi, Inoue, Moore, & Negishi, 2003). Although ligand‐based activation of NR1I3 does not necessarily elicit nuclear translocation, it has been reported that use of 6‐(4‐chlorophenyl)‐imidazo[2,1‐b]thiazole‐5‐carbaldehyde (https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2758), a selective human NR1I3 (hNR1I3) ligand, results in increased expression of the ATP‐binding cassette (https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=136&familyType=TRANSPORTER) transporter P‐glycoprotein (P‐GP; https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=768) in the immortalised brain endothelial cell line hCMEC/d3 (Chan et al., 2011). The authors attributed this finding to increased nuclear NR1I3 activity, but as a recent study demonstrated that CITCO is a dual hNR1I3/human NR1I2 (hNR1I2) ligand (Lin et al., 2020), the above‐mentioned outcome might also be the result of NR1I2 activation.

As is the case with NR1I3, NR1I2 exhibits nuclear localisation in immortalised cell lines (Saradhi et al., 2005) and cytosolic localisation in primary cultures (Ott, Fricker, & Bauer, 2009; Squires et al., 2004) but is not constitutively active when it accumulates in the nucleus.

In the cytosol, NR1I3 and NR1I2 are bound to a chaperone protein complex composed of the https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=784 90 and the NR1I3 retention protein (Mackowiak & Wang, 2016; Wang, Ong, Chai, & Chen, 2012). Upon activation, NR1I2 and NR1I3 are released from this protein complex and translocate to the nucleus to form a heterodimer with the α, β, γ isoforms of the https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=92 (RXR;NR2B) (Koutsounas et al., 2015). The NR1I2/RXR and NR1I3/RXR heterodimers are permissive partners within the heterodimer, such that the activity of this interaction can be enhanced through further binding with an RXR ligand (Chen, Wang, & Wan, 2010; Wang et al., 2006). The heterodimers can bind to co‐activators or co‐repressors, prior to interacting with the respective response elements, subsequently initiating or repressing the transcription of target genes (Oladimeji et al., 2016).

Although NR1I2 and NR1I3 are activated by direct ligand binding, evidence suggests that both receptors can be activated via an indirect mechanism. Indirect activation of NR1I3 involves the activation of the ERK (https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1494/https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1495) and https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=621 signalling pathways (Mackowiak & Wang, 2016). The mechanism for indirect activation of NR1I2 remains elusive, but studies propose that the https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=570 signalling pathway may be involved (Dong, Lin, Wu, & Chen, 2010; Lin et al., 2008). Furthermore, the activities of NR1I3 and NR1I2 can be regulated by post‐translational modifications including phosphorylation, ubiquitination, acetylation and SUMOylation (Staudinger, Xu, Biswas, & Mani, 2011; Wang et al., 2012).

One important feature of NR1I2 and NR1I3 is that ligand‐mediated receptor activation is species dependent, due to differences in the amino acid sequence of the ligand‐binding domain (Gray, Peacock, & Squires, 2009; Gray, Pollock, Schook, & Squires, 2010). This issue makes comparisons of NR1I2 and NR1I3 activation between species difficult, although a limited number of selective ligands for rodent and hNR1I2 and hNR1I3 are available (Table 1) (Chan et al., 2011; Lemmen, Tozakidis, Bele, & Galla, 2013; Lemmen, Tozakidis, & Galla, 2013). To help address this issue, humanised rodent models have been developed (Bauer et al., 2006; Haines et al., 2018).

TABLE 1.

Ligands and biological targets of NR1I3 (CAR) and NR1I2 (PXR)

Nuclear receptor	NR1I3 (CAR)		(NR1I2) PXR		Reference
Species	Rodent	Human a	Rodent	Human a
Ligands
Selective agonists	TCPOBOP, phenobarbital	CITCO, phenobarbital	PCN	Rifampicin Hyperforin SR12813	(Jimenez, Quattrochi, Yockey, & Guzelian, 2000; Jones et al., 2000; Kliewer et al., 1998; Moore et al., 2000; Tzameli, Pissios, Schuetz, & Moore, 2000)
Antagonists	https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2760	CINPA‐1 https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2757	https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2568	Ketoconazole, l‐https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=6569, coumestrol, SPA70	(Cherian et al., 2018; Forman et al., 1998; Huang et al., 2007; Lemmen, Tozakidis, Bele, & Galla, 2013; Lin et al., 2017; Wang et al., 2008; Yeung, Sueyoshi, Negishi, & Chang, 2008)
Biological targets
Enzymes
CYP450 isoforms	Cyp1a1, Cyp1a2, Cyp2a4, Cyp2b10, Cyp2c29, Cyp2c37, Cyp2c55, Cyp3a11	CYP1A1, CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP3A4, CYP3A5	Cyp3a, Cyp3a11, Cyp2b9, Cyp2b10, Cyp2c55	CYP3A4, CYP3A23, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP1A	(Aleksunes & Klaassen, 2012; Chen, Ferguson, Negishi, & Goldstein, 2003; Ferguson, Chen, LeCluyse, Negishi, & Goldstein, 2005; Ferguson, LeCluyse, Negishi, & Goldstein, 2002; Goodwin, Hodgson, D'Costa, Robertson, & Liddle, 2002; Maglich et al., 2004; Xu, Wang, & Staudinger, 2009; Yoshinari, Yoda, Toriyabe, & Yamazoe, 2010; Zhang, Huang, Chua, Wei, & Moore, 2002)
UGTs	Ugt1a1, Ugt1a9, Ugt2b34, Ugt2b35, Ugt2b36	UGT1A1	Ugt1a1, Ugt1a5, Ugt1a9	UGT1A1, UGT1A6, UGT1A3, UGT1A4	(Aleksunes & Klaassen, 2012; Maglich et al., 2004; Sugatani et al., 2005)
Sulfotransferases	Sult1a1, Sult1d1, Sult1e1, Sult2a1, Sult2a2, Sult3a1, Sult5a1	SULT2A1	Sult1a1, Sult1e1, Sult2a1, Sult2a2, Sult5a1	SULT2A1	(Aleksunes & Klaassen, 2012; Alnouti & Klaassen, 2008; Chen, Zhang, Baker, & Chen, 2007; Echchgadda et al., 2007; Maglich et al., 2004)
GSH‐S‐transferases	Gsta1, Gsta3, Gsta4, Gstm1, Gstm2, Gstm3, Gstm4, Gstp, Gstt1, Gstt3		Gsta1, Gsta3, Gsta4, Gstm1, Gstm2, Gstm3, Gstm4, Gstm6, Gstp, Gstt1, MGst1	GSTA1, GSTA2, GSTM1, GSTP1‐1	(Aleksunes & Klaassen, 2012; Cui, Choudhuri, Knight, & Klaassen, 2010; Knight, Choudhuri, & Klaassen, 2008; Singh et al., 2012; Zhang et al., 2002)
Gluconeogenic enzymes	PEPCK, G6P‐ase		PEPCK, G6P‐ase	PEPCK, G6P‐ase	(Kodama, Koike, Negishi, & Yamamoto, 2004; Kodama, Moore, Yamamoto, & Negishi, 2007; Mackowiak et al., 2019)
Lipogenic enzymes	Scd‐1		Scd‐1	SCD‐1	(Nakamura, Moore, Negishi, & Sueyoshi, 2007; Zhang et al., 2013)
β‐oxidation‐related enzymes	Cpt1a		Cpt1a	CPT1A	(He et al., 2013; Moreau et al., 2009; Nakamura et al., 2007; Ueda et al., 2002)
Ketogenic enzymes	Hmgcs2		Hmgcs2		(Gao, He, Zhai, Wada, & Xie, 2009; He et al., 2013)
Transporters
ABC	P‐gp, Bcrp, Mrp2, Mrp4	P‐GP, BCRP, MRP2	P‐gp, Bcrp, Mrp2, Mrp4	P‐GP , BCRP, MRP2, ABCBA1, ABCG1	(Bauer, Hartz, Fricker, & Miller, 2004; Kast et al., 2002; Lemmen, Tozakidis, Bele, & Galla, 2013; Lemmen, Tozakidis, & Galla, 2013; Martin et al., 2008; Narang et al., 2008; Oswald et al., 2006; Ott et al., 2009; Petrick & Klaassen, 2007; Teng & Piquette‐Miller, 2005; Wang, Sykes, & Miller, 2010; Whyte‐Allman, Hoque, Jenabian, Routy, & Bendayan, 2017; Yamasaki, Kobayashi, & Chiba, 2018)
SLC	https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=165#876, https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=238#1219			https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=238#1219, https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=238#1219	(Aleksunes & Klaassen, 2012; Hassani‐Nezhad‐Gashti et al., 2018; Li et al., 2015; Meyer zu Schwabedissen, Tirona, Yip, Ho, & Kim, 2008)

^aStudies performed in higher mammalian models including porcine are included since ligands for human NR1I3 (CAR) and NR1I2 (PXR) also activate these orthologues.

2.1. Ligands and biological effects

Both NR1I3 and NR1I2 possess broad ligand specificity that confers the ability to sense structurally diverse endogenous and exogenous compounds. Exposure to such ligands, including therapeutic drugs, environmental contaminants and endocrine disruptors, elicits the activation of overlapping and distinct sets of genes that encode metabolising enzymes, including members of the https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=242 (CYP450) family, sulfotransferases, uridine glucuronosyl transferases (https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=988) and GSH‐S‐transferases (https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=270#1381) (Table 1), resulting in xenobiotic inactivation (Aleksunes & Klaassen, 2012; Wang et al., 2012). Furthermore, NR1I2 and NR1I3 are involved in the regulation of expression of membrane transporters belonging to the ABC and solute carrier families (Table 1), responsible for the disposition of xenobiotics and their metabolites (Klaassen & Aleksunes, 2010). Due to this body of evidence, NR1I3 and NR1I2 were mainly regarded as modulators of detoxification mechanisms and drug disposition. However, more extensive research has unveiled further biological roles attributed to NR1I3 and NR1I2 activation (Table 1), including regulation of nutrient metabolism, inflammation and apoptosis (Hakkola, Rysa, & Hukkanen, 2016; Hukkanen, Hakkola, & Rysa, 2014; Robbins, Bakke, Cherian, Wu, & Chen, 2016).

The role of NR1I3 and NR1I2 in hepatic glucose and lipid metabolism has been studied in rodent and human‐based models. These studies demonstrated that ligand‐based activation of NR1I3 and NR1I2 plays a significant role in controlling energy status, including regulation of gluconeogenesis and lipogenesis, in a model‐dependent manner (Hakkola et al., 2016). For example, the ligand‐based activation of NR1I2 is a negative regulator of gluconeogenesis in human‐derived liver cell lines and rodent models, with NR1I2 activation leading to down‐regulation of the expression of key enzymes including phosphoenolpyruvate carboxykinase (PEPCK) and glucose‐6‐phosphatase (G6P‐ase) (Kodama et al., 2004; Kodama et al., 2007). However, studies carried out in human primary hepatocytes demonstrated ligand activation of NR1I2 results in induction of gluconeogenesis (Gotoh & Negishi, 2015). The findings of studies investigating the effects of NR1I3 activation on the regulation of gluconeogenesis are more consistent across models, with NR1I3 activation leading to down‐regulation of gluconeogenesis (Kodama et al., 2004; Mackowiak et al., 2019).

In regard to lipid metabolism, NR1I3 is thought to be a negative regulator of lipogenesis by down‐regulating the expression of lipogenic enzymes including stearoyl‐CoA desaturase 1 (SCD‐1) (Dong et al., 2009; Gao et al., 2009; Marmugi et al., 2016). Furthermore, the ligand‐based activation of NR1I3 decreases the expression of carnitine palmitoyltransferase 1 (CPT1A), an enzyme involved in mitochondrial β‐oxidation of fatty acids and the ketogenic enzyme 3‐hydroxymethylglutaryl CoA synthase 2 (https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2432) (Gao et al., 2009). Conversely, NR1I2 exerts an inductive effect on lipogenesis through up‐regulation of SCD‐1 (Nakamura et al., 2007; Zhang et al., 2013), the fatty acid transporter https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=3104 (Zhou et al., 2008) and ttps://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=185#981 (Li et al., 2015), a transporter for citrate, which acts as a precursor for the synthesis of cholesterol and fatty acids. The ligand‐based activation of NR1I2 also decreases the expression of CPT1A and HMGCS2 (Bitter et al., 2015; He et al., 2013; Moreau et al., 2009; Nakamura et al., 2007).

In the regulation of inflammation, the ligand‐based activation of NR1I2 blocks https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=996‐mediated expression of pro‐inflammatory genes (Zhou et al., 2006) and this discovery has led to the development of new therapeutic approaches for management of intestinal‐associated inflammation (Calanni, Renzulli, Barbanti, & Viscomi, 2014). Furthermore, studies in human hepatocytes demonstrate the pro‐inflammatory cytokine https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4998 repressed the ligand‐based activation of both NR1I3 and NR1I2 (Pascussi et al., 2000).

Both NR1I3 and NR1I2 can also influence the survival of tumour cells by regulating additional cell processes including apoptosis and cell proliferation through other distinct mechanisms. In rodent, NR1I3 inhibits apoptosis by interacting with the DNA‐damage inducible β protein, GADD45B (Yamamoto, Moore, Flavell, Lu, & Negishi, 2010). The role of NR1I2 in cancer is more complex since the receptor can elicit, in a cell‐dependent manner, pro‐ or anti‐apoptotic effects by regulating the activity of the tumour suppressor protein p53 (Oladimeji et al., 2016). Furthermore, the interaction between NR1I2‐p53 and the biological outcome is dependent on the expression of the wild‐type (Robbins, Cherian, Wu, & Chen, 2016) or mutant (Ouyang et al., 2010) variant of p53.

3. NR1I3 (CAR) AND NR1I2 (PXR): EXPRESSION AND EFFECTS IN THE BRAIN

3.1. Expression profile in areas of the brain

Several early reports attempted to characterise the expression profile of NR1I3 and NR1I2 in different areas of the brain (Table 2). In human, levels of receptor expression are markedly lower (or even undetectable) when compared to liver and intestine (Lamba et al., 2004; Miki et al., 2005; Nishimura et al., 2004; Savkur et al., 2003). Further studies carried out in different regions of the brain from mammalians found expression of NR1I3 and/or NR1I2 in the cortex (Kajta et al., 2019; Marini et al., 2007; Nannelli et al., 2010), hippocampus (Litwa et al., 2016; Marini et al., 2007; Nannelli et al., 2010), midbrain (Frye, Koonce, & Walf, 2014d; Marini et al., 2007; Nannelli et al., 2010) and cerebellum (Marini et al., 2007; Nannelli et al., 2010; Xia et al., 2017) (Table 2). NR1I3 and NR1I2 expression have also consistently been reported in BBB endothelial cells isolated from cortex (Bauer et al., 2004; Chan et al., 2011; Lemmen, Tozakidis, Bele, & Galla, 2013; Ott et al., 2009).

TABLE 2.

Expression profile of NR1I3 (CAR) and NR1I2 (PXR) in areas of the brain

	NR1I3	NR1I2	Reference
Rodent	Cortex (mRNA, protein)	Cortex (mRNA, protein)	(Bauer et al., 2004; Chan et al., 2013; Frye, Koonce, Walf, & Rusconi, 2013; Kajta et al., 2019; Litwa et al., 2016; Narang et al., 2008; Wang et al., 2010)
	Hippocampus (mRNA, protein)	Hippocampus (mRNA, protein)
	Cortical capillaries (mRNA, protein)	Midbrain (mRNA, protein)
		Cortical capillaries (mRNA, protein)
Quail	Cerebrum (mRNA)	Cerebrum (mRNA)	(Du et al., 2017; Lin et al., 2018; Xia et al., 2017)
	Cerebellum (mRNA)	Cerebellum (mRNA)
Rabbit	Cortex (mRNA)	Cortex (mRNA)	(Marini et al., 2007)
	Hippocampus (mRNA)	Midbrain (mRNA)
	Hypothalamus (mRNA)	Cerebellum (mRNA)
	Midbrain (mRNA)
	Striatum (mRNA)
	Cerebellum (mRNA)
Porcine	Cortex (mRNA)	Cortex (mRNA)	(Lemmen, Tozakidis, Bele, & Galla, 2013; Lemmen, Tozakidis, & Galla, 2013; Nannelli et al., 2010; Ott et al., 2009)
	Cerebellum (mRNA)	Cerebellum (mRNA)
	Midbrain (mRNA)	Midbrain (mRNA)
	Hippocampus (mRNA)	Hippocampus (mRNA)
	Meninges (mRNA)	Meninges (mRNA)
	Cortical capillaries (mRNA, protein)	Cortical capillaries (mRNA, protein)
Human	Brain tissue a (mRNA)	Brain tissue a (mRNA)	(Chan et al., 2011; Lamba et al., 2004; Miki et al., 2005; Nishimura et al., 2004; Savkur et al., 2003)
	Caudate nucleus (mRNA)	Cerebrum (mRNA)
	Cortical capillaries (mRNA, protein)	Cortex (mRNA)
		Cerebellum (mRNA)
		Cortical capillaries (mRNA, protein)

^aRefers to no particular area of the brain.

As the brain is an organ whose divisions are both functional and anatomical, it is plausible to hypothesise that the regional expression of NR1I3 and NR1I2 may exert distinctive biological actions. Furthermore, as each cell subtype (i.e. neurons, glial cells and BBB endothelial cells) is phenotypically different, the set of genes activated by both NRs may vary. These premises have been demonstrated by studies conducted in animal models. For instance, ligand‐based activation of NR1I2 up‐regulated the mRNA levels of CYP enzymes in specific regions of the rodent (Yadav, Dhawan, Seth, Singh, & Parmar, 2006), porcine (Nannelli et al., 2010) and brain, while the total (Boussadia et al., 2016) or localised (Frye, Koonce, & Walf, 2014b) inhibition of NR1I2 expression in rodents influence brain functions including behaviour (Table 3).

TABLE 3.

Biological actions of NR1I3 (CAR) and NR1I2 (PXR) in neural cells

Species	Model	Ligand	Receptor	Level of expression	Biological effect	Reference
Mouse	Primary hippocampal neurons from Swiss CD‐1 mice embryos	Nonylphenol	NR1I3, NR1I2	mRNA, protein	Up‐regulation of NR1I3, Pxr and Rxrα Induction of apoptosis	(Litwa et al., 2016)
	Primary neocortical neurons from Swiss CD‐1 mice embryos	Triclocarban	NR1I3	mRNA, protein	Induction of apoptosis	(Kajta et al., 2019)
	NR1I3 (−/−) and Pxr (−/−) C57BL/6J mice		NR1I3, NR1I2		Impaired recognition memory and anxiety	(Boussadia et al., 2016; Boussadia et al., 2018)
Rat	Male albino Wistar rats	PCN, dexamethasone	NR1I2	mRNA	Up‐regulation of CYP3A in hypothalamus and hippocampus	(Yadav et al., 2006)
	Knocked down expression of Pxr in midbrain of Long‐Evans rats		NR1I2	mRNA	Reduced metabolism of progesterone to 3α,5α‐THP Impaired mating behaviour Reduced hippocampal BDNF levels	(Frye et al., 2013; Frye et al., 2014d; Frye, Koonce, & Walf, 2014a; Frye, Koonce, & Walf, 2014c)
Quail		Atrazine, di‐(2‐ethylhexyl)‐phthalate	NR1I3, NR1I2	mRNA	Up‐regulation of NR1I3, NR1I2 and AhR Induction of apoptosis Oxidative and mitochondrial damage Cerebral and cerebellar toxicity	(Du et al., 2017; Lin et al., 2018; Xia et al., 2017)
Porcine	4‐day treatment with rifampicin in White large × Landrace hybrid pigs	Rifampicin	NR1I2	mRNA	Up‐regulation of CYP3A22, CYP3A29 in cortex and hippocampus Up‐regulation of CYP2B22 in meninges	(Nannelli et al., 2010)
Human	Cord blood CD34+ stem cell‐derived differentiating neuronal cells	Rifampicin	NR1I2	Protein	Up‐regulation of CYP1A1, 2B6, 2E1, 3A4, AhR, NR1I3, NR1I2 and GSTP1‐1	(Singh et al., 2012)
	SHSY‐5Y cells	Monocrotophos	NR1I3, NR1I2	mRNA	Up‐regulation of NR1I3, NR1I2, CYP2C8 and 3A4 Induction of apoptosis Oxidative damage	(Tripathi et al., 2017)

3.2. NR1I3 (CAR) and NR1I2 (PXR): Comparison of activities between brain and liver

Since it has been reported that the brain expresses metabolising enzymes and transporters present in the liver (an organ commonly used as reference), it was thought that their expression could also be regulated by NRs such as NR1I3 and NR1I2. Although a number of studies have supported the above premise, they also demonstrate that the roles of NR1I3 and NR1I2 in the brain cannot always be widely extrapolated.

Several in vivo comparative studies conducted in animal models showed that the response to treatments with NR1I3 and/or NR1I2 ligands may vary between organs expressing both NRs (Marini et al., 2007; Messina, Nannelli, Fiorio, Longo, & Gervasi, 2009; Nannelli et al., 2009; Nannelli et al., 2010; Yamasaki et al., 2018). Furthermore, as discussed above, the effects of ligand‐based activation of NR1I3/NR1I2 differ between areas of the brain.

In rodent, administration of the NR1I3 ligand https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2804 to rats selectively up‐regulated the expression of CYP2B (https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1324 in cortex and midbrain (Schilter & Omiecinski, 1993; Upadhya, Chinta, Pai, Boyd, & Ravindranath, 2002). A later study showed that rats and mice treated with phenobarbital and the selective rodent NR1I3 ligand (1,4‐bis‐[2‐[3,5‐dichloropyridyloxy)]benzene (https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2756), respectively, exhibited increased protein expression of P‐gp https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=768) the multidrug resistance‐associated protein‐2 (MRP2;https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=780) and the breast cancer resistance protein (BCRP;https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=792) in brain capillaries (Wang et al., 2010). One common feature of the above studies is that the effects elicited by ligand‐based activation of NR1I3 were comparable to those observed in liver, which was used as a positive control. A similar approach was employed for the study of ligand‐based activation of NR1I2 in the rodent brain, where reports demonstrated that rats exposed to the selective rodent NR1I2 ligands pregnelolone‐16α‐carbonitrile (PCN) and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2768 showed increased expression of Cyp450 enzymes and ABC transporters in liver and specific areas of the brain. For example, PCN and dexamethasone up‐regulated the mRNA expression and activity Cyp3a1 (in midbrain, hypothalamus and hippocampus (Yadav et al., 2006), while P‐gp and Mrp2 were up‐regulated at the protein level in isolated brain capillaries, but not brain homogenates (Bauer et al., 2004; Bauer et al., 2008). A recent study showed that administration of PCN to mice up‐regulated P‐gp mRNA expression in liver, intestine, whole brain and brain cortex (Yamasaki et al., 2018). However, these changes were not observed at the protein level in both liver and whole brain. Interestingly, transgenic mice expressing hNR1I2 exhibited up‐regulated P‐gp protein levels in both liver and isolated brain capillaries upon treatment with rifampicin (Bauer et al., 2006).

In terms of higher animal models, treatment of rabbits with phenobarbital increased the hepatic mRNA expression of several CYP450 isoforms (CYP2A10, CYP2B4, CYP2B5 and CYP3A6), but these changes were not observed in areas of the brain including cortex, midbrain, striatum, hippocampus, hypothalamus and cerebellum (Marini et al., 2007). In pigs, the intraperitoneal administration of the hNR1I2 ligand https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2765 up‐regulated the mRNA expression of CYP450 isoforms (CYP3A22 CYP3A29, CYP3A46 and CYP2B22) and ABC transporters (P‐GP) in liver (Nannelli et al., 2010). However, at brain level, the mRNA expression of CYP3A22 and CYP3A29 was up‐regulated only in the cortex and hippocampus, while CYP2B22 was up‐regulated in meninges, with no significant increases in both microsomal and mitochondrial activities. Although transcripts of ABC transporters including P‐GP, MRP1 and MRP2 were found at different levels in all the brain regions studied, rifampicin failed to up‐regulate their expression. This outcome contrasts with those reported by studies conducted in rodent models, described above (Bauer et al., 2004; Bauer et al., 2006).

The study of relationships between NR1I3/NR1I2 activation and the expression of target genes in both liver and brain has also been extended to disease states. A recent report demonstrated that the induction of status epilepticus and spontaneous recurrent seizures in rats, through intra‐hippocampal injection of kainic acid, up‐regulated the expression of https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1330 and Cyp3a13 mRNA in liver and hippocampus of rats under spontaneous recurrent seizures, with no changes in the mRNA expression of NR1I3 and NR1I2 (Runtz et al., 2018). Although the authors did not demonstrate if NR1I3 and/or NR1I2 were involved in the up‐regulation of the Cyp450 isoforms, the fact that both brain and liver responded similarly to a common stimulus not related to direct ligand‐based activation of NR1I3 and NR1I2 with pharmacological agents is an interesting outcome that should be studied in greater depth.

Taken together, the above studies suggest that while certain areas of the brain respond in a similar manner to the liver upon ligand‐based activation of NR1I3 and NR1I2, this is not true of all brain regions. With a better understanding of the precise signalling pathways active in organs and tissues of the studied species, it would therefore be easier to predict the effects elicited by NR1I3 and NR1I2. In cells of the brain, there is evidence that both NR1I3 and NR1I2 translocate to the nucleus upon their ligand‐based activation (Kajta et al., 2019; Lemmen, Tozakidis, & Galla, 2013), but knowledge of any similarities or differences with liver in terms of post‐translational modifications, or how they latter interact with RXR and co‐activators/co‐repressors, is still scarce (Figure 2).

FIGURE 2.

Comparison of the downstream events of the NR1I3 (CAR) and NR1I2 (PXR) signalling pathways in liver and brain. Most of the research aiming to characterise the signalling pathways of NR1I3 and NR1I2 has been conducted in liver, whilst organs such as the brain are much less studied. This issue is particularly important as brain endothelial cells and neurons express both receptors and the response to the ligand‐based activation of NR1I3 and NR1I2 appears vary amongst areas of the brain. This outcome is consistent with the tissue‐dependent effects of

3.3. NR1I3 (CAR) and NR1I2 (PXR) in the BBB

The study of NR1I3 and NR1I2 function in BBB endothelial cells has been primarily focused on their effects on the expression of ABC transporters and metabolising enzymes involved in drug disposition. The effects of ligand‐based activation of NR1I3 and NR1I2 were first studied in rodent models (Bauer et al., 2004; Chan et al., 2013; Wang et al., 2010), with subsequent studies carried out in higher models (Lemmen, Tozakidis, Bele, & Galla, 2013; Ott et al., 2009) including human (Chan et al., 2011) (Table 4).

TABLE 4.

Effect of ligand‐based activation of NR1I3 (CAR) and NR1I2 (PXR) on expression of ABC transporters and metabolising enzymes in blood–brain barrier endothelial cells

Receptor	Species	Model	Ligand	Effect on expression and/or activity	Reference
NR1I3	Mouse	Isolated capillaries from wild‐type C3H mice	TCPOBOP	↑ P‐gp, Mrp2, Bcrp	(Wang et al., 2010)
		Isolated capillaries from NR1I3 (−/−) C3H mice		N/E
	Rat	Isolated capillaries from Sprague–Dawley rats	Phenobarbital	↑ P‐gp, Mrp2, Bcrp	(Wang et al., 2010)
	Porcine	Primary culture	CITCO	↑ P‐GP, BCRP	(Lemmen, Tozakidis, Bele, & Galla, 2013)
	Human	hCMEC/D3 cells	CITCO	↑ P‐GP	(Chan et al., 2011)
NR1I32	Mouse	Isolated capillaries from wild‐type CB6F1 mice	PCN	↑ P‐gp	(Bauer et al., 2006)
		Isolated capillaries from hNR1I2 humanised CB6F1 mice	Rifampicin Hyperforin	↑ P‐gp
		CD‐1 mice	Dexamethasone	↑ P‐gp (cortical capillaries), N/E Cyp3a11	(Chan et al., 2013)
	Rat	Isolated capillaries from Sprague–Dawley rats	PCN Dexamethasone	↑ P‐gp, Gstp	(Bauer et al., 2004; Bauer et al., 2008)
		Primary culture from Sprague–Dawley rats	Dexamethasone	↑ P‐GP, Mrp2, Bcrp	(Narang et al., 2008)
	Porcine	Primary culture	Rifampicin Hyperforin	↑ P‐GP, BCRP	(Ott et al., 2009); (Lemmen, Tozakidis, & Galla, 2013)
		White large × Landrace hybrid pigs	Rifampicin	N/E CYP3A22 CYP3A29, CYP3A46 and CYP2B22, P‐GP, MRP1, MRP2	(Nannelli et al., 2010)
	Human	hCMEC/D3 cells	Rifampicin SR12813	↑ P‐GP	(Chan et al., 2011)
		Primary culture		↑ CYP3A4, CYP2C9, CYP2E1	(Ghosh et al., 2017)

Note: ↑ denotes increased expression or activity, and ↓ decreased expression or activity.

Abbreviation: N/E, no effect.

Studies investigating NR1I3 and NR1I2 expression in rodent (mouse and rat) brain endothelial cells have provided consistent results in both in vitro and in vivo models. In mice, exposure of brain capillaries to PCN and dexamethasone up‐regulated the expression of PGP1(Bauer et al., 2006), while humanised mice for hNR1I2 were responsive to the ligands rifampicin and hyperforin by increasing both P‐gp expression and activity, findings that were reproduced in an in vivo model. The rodent NR1I2 ligand dexamethasone up‐regulated the expression of P‐gp in cortical brain capillaries but not in whole brain homogenates of mice (Chan et al., 2013). A later study that employed PCN in mice demonstrated a similar outcome, but using brain cortex homogenates instead of isolated brain capillaries (Yamasaki et al., 2018). The authors suggested that the PCN‐mediated up‐regulation of cortical P‐gp is a result of its effect on BBB endothelial cells, but the study did not provide findings supporting this statement.

Activation of NR1I2 has been reported to increase both P‐gp (Bauer et al., 2004), Mrp2 and the Gst isoform Gstp (Bauer et al., 2008) expression and activity in brain capillaries of rats exposed to PCN or dexamethasone, with similar findings observed in isolated brain capillaries. In primary cultures of rat brain endothelial cells, dexamethasone up‐regulated the expression of P‐gp, Mrp2 and the BCRP (Narang et al., 2008). Evidence of the expression of functional NR1I3 in rodent brain endothelial cells is provided by a study (Wang et al., 2010), which reported that in vivo and in vitro activation of NR1I3 with TCPOBOP and phenobarbital increased expression and activity of PGP1, MRP2 and the BCRP in mouse and rat brain capillaries, respectively.

Recent studies have proposed that alterations in NR1I2 and NR1I3 function could impair the expression of tight junction proteins at the BBB since deletion of NR1I2 and NR1I3 in rats leads to reduced expression of zonula occuldens‐1 protein in isolated brain microvessels, with no changes in claudin‐5 expression reported (Boussadia et al., 2016; Boussadia et al., 2018). Although this is the only evidence available to date at brain level, it is supported by a previous study that demonstrated the relationship between NR1I2 and expression of TJ proteins in the intestine, in which NR1I2 (−/−) mice exhibited higher intestinal permeability (Venkatesh et al., 2014).

There is convincing evidence of NR1I2 and NR1I3 expression in porcine brain endothelial cells. Although one study has reported that porcine NR1I2 (pNR1I2) is not expressed at the mRNA level in whole brain (Pollock, Rogatcheva, & Schook, 2007), a later report (Nannelli et al., 2010) found porcine NR1I3 and NR1I2 mRNA transcripts in porcine brain capillaries, albeit at lower levels compared to liver. As described in a previous section, the hNR1I2 ligand rifampicin in pigs did not affect capillary mRNA expression of CYP450 isoforms and ABC transporters such as P‐GP, MRP1 and MRP2 (Nannelli et al., 2010). However, in primary cultures of porcine brain endothelial cells, activation of pNR1I2 with the hNR1I2 ligands https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2764 and rifampicin increased expression of P‐GP at the mRNA and protein levels and the activity of P‐GP, although no such effect was achieved with the rodent NR1I2 ligand PCN (Ott et al., 2009). More recently, studies have confirmed the findings of Ott et al. (2009) and shown that activation of pNR1I2 up‐regulated the expression and activity of P‐GP and BCRP in porcine brain endothelial cells (Lemmen, Tozakidis, & Galla, 2013). Furthermore, studies have demonstrated that porcine brain endothelial cells exposed to the hNR1I3 ligand CITCO up‐regulated expression and activity of P‐GP and BCRP, but this up‐regulation was not achieved using TCPOBOP (Lemmen, Tozakidis, Bele, & Galla, 2013).

In human, the evidence is somewhat conflicting but tends to support the expression of functional NR1I3 and NR1I2 in the BBB. A quantitative proteomic analysis performed in freshly isolated human brain capillaries found that NR1I3 and NR1I2 are not expressed at the protein level (Shawahna et al., 2011). However, a later study demonstrated that NR1I2 is expressed in human brain endothelial cells, with a marked overexpression at both mRNA and protein levels in cultured endothelial cells isolated from epileptic brains (Ghosh et al., 2017). The authors suggested that overexpression of NR1I2 might be responsible for the increased expression of CYP enzymes, including https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1337, https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1326 and https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1330.

The hCMEC/D3 human brain endothelial cell line is widely used to study human BBB characteristics. Evidence of key orphan NR expression in these cells is mixed, with one study reporting NR1I3 and NR1I2 mRNA transcripts were not detectable (Dauchy et al., 2009). However, a later study confirmed expression of both receptors at the protein level, demonstrating that exposure of this cell line to CITCO (hNR1I3 ligand), rifampicin and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2763 (hNR1I2 ligands), increased P‐GP mRNA and protein expression (Chan et al., 2011). Furthermore, the same study showed that the NR1I3 and NR1I2‐mediated up‐regulation of P‐GP expression can be counteracted by employing antagonists of both NRs or down‐regulating their expression through use of siRNA.

3.4. NR1I3 (CAR) and NR1I2 (PXR): Roles in neurotoxicity

The mechanistic roles of NR1I2 and NR1I3 in the brain in neurotoxicity induced by xenobiotics, including chemotherapeutic drugs, pesticides and environmental contaminants, are of increasing interest (Table 3). One of the first pieces of evidence of a potential relationship between the ligand‐based activation of orphan NRs and neurotoxicity is the case of the NR1I3 ligand phenobarbital. For example, chronic exposure of rats to phenobarbital up‐regulated the expression of Cyp2b1 and enhanced the neurotoxic actions of the anticancer drug 9‐methoxy‐N²‐methtyellipticinium in reticular neurons (Upadhya et al., 2002). Although the authors did not experimentally demonstrate that NR1I3 is responsible for the outcome, the use of a well‐known NR1I3 ligand strongly suggests involvement of this NR.

Later studies focused their efforts in characterising the molecular mechanisms involved in the NR1I3 and NR1I2‐mediated neurotoxicity. A report showed that in isolated mouse embryonic hippocampal neurons maintained in culture, exposure to the environmental contaminant nonylphenol up‐regulated expression of NR1I3, NR1I2 and Rxrα (https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=610) at both the mRNA and protein levels (Litwa et al., 2016). These effects were accompanied by induction of apoptosis, due to increased https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1619 and lactate dehydrogenase activities. A later study showed that treatment of neocortical primary mouse neurons with the environmental contaminant triclocarban resulted in apoptosis mediated by NR1I3 and https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2951 (AhR) activation (Kajta et al., 2019). Furthermore, triclocarban inhibited SUMOylation, a post‐translational modification that could serve as a protective mechanism by reducing the activity of NR1I3 and AhR.

In cultures of human cord blood CD34⁺ stem cell‐derived differentiating neuronal cells, treatment with rifampicin up‐regulated the expression of NR1I3, NR1I2, CYP450 enzymes, including the 1A1, 2B6, 3A4 and 2E1 isoforms, and the GSTP1 isoform (Singh et al., 2012). The same study demonstrated that the pesticide and developmental toxicant monocrotophos (MCP) exerted similar effects on the above enzymes, which were enhanced by co‐treatment with MCP and rifampicin. This evidence showed that differentiating neural cells express inducible enzymes that respond to the ligand‐based activation of NR1I2, probably as a mechanism for inactivation of potentially neurotoxic xenobiotics.

Further studies with the human neuroblastoma SH‐SY5Y cell line confirmed the inductive effects elicited by MCP on the expression of NR1I3, NR1I2 and CYP3A4 and also reported enhanced induction of NR1I3, NR1I2 and CYP3A4 mRNA following co‐treatment with MCP and the anticancer drug https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=7154 (Tripathi et al., 2017). Furthermore, exposure of SH‐SY5Y cells to MCP increased the production of ROS, decreased the levels of reduced GSH and up‐regulated expression of pro‐apoptotic markers. Docking studies have demonstrated that MCP binds to both NR1I3 and NR1I2, suggesting that the up‐regulation of mitochondrial CYP450 enzymes, mediated by ligand‐based activation of both receptors, may be responsible for changes in levels of ROS and reduced GSH. However, no association between the pro‐apoptotic effects of MCP and the activation of NR1I3/NR1I2 is described.

Similar findings have been reported in a study carried out in quail that evaluated the actions of atrazine, a NR1I2 ligand (Abass et al., 2012) and widely employed neurotoxic herbicide (Lin et al., 2018). Chronic exposure to increasing doses of atrazine resulted in mitochondrial dysfunction, elevated ROS production, induction of apoptosis in the cerebrum and cerebellar toxicity (Lin et al., 2018; Xia et al., 2017). Furthermore, atrazine increased the cerebral and cerebellar expression of NR1I2, NR1I3 and AhR, suggesting that activation of these NRs is responsible for the up‐regulation of expression of CYP450 isoforms (Lin et al., 2018; Xia et al., 2017). The authors of the above studies proposed that the NR1I3, NR1I2 and AhR‐mediated changes in the expression of mitochondrial CYP450 enzymes are responsible for the mitochondrial dysfunction that leads to oxidative stress in the brain. This experimental approach was replicated for the assessment of the effects of the plasticizer di‐(2‐ethylhexyl)‐phthalate on cerebellar toxicity in quail, and the outcome was comparable to that observed for atrazine (Du et al., 2017).

The above findings are interesting since the ligand‐based activation of NR1I3 and NR1I2 partly mediates changes in expression and activities of key proteins involved in neurotoxicity. However, as most of the research attempting to characterise the cellular mechanisms has been conducted in cell‐based models with non‐selective ligands of NR1I3 and NR1I2, more studies are required to establish if the activity of both receptors is directly related to the neurotoxic effect of chemicals in an in vivo setting. This issue is important since the neurotoxic effects observed in animal models exposed to environmental contaminants and other chemicals might be the result of the activation of NRs other than NR1I3 and NR1I2.

3.5. NR1I3 (CAR) and NR1I2 (PXR): Roles in behaviour and other brain functions

The influence of NR1I3 and NR1I2 in the regulation of behaviour and other brain function, including memory, is receiving increased attention. Studies have proposed that the ligand‐based activation of NR1I2 may influence the metabolism of sex hormones, which themselves are ligands of this receptor, thereby altering levels of hormone metabolites involved in the regulation of behaviour.

Studies conducted in rat demonstrate that NR1I2 in the midbrain is required for formation of https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4108 (3α,5α‐THP), a metabolite of progesterone and neurosteroid that regulates motivated behaviours (e.g. mating) (Frye et al., 2013). The injection of a Nr1i2 antisense oligonucleotide into the ventral tegmental area of pro‐oestrous rats abrogated expression of NR1I2 and reduced 3α,5α‐THP levels, resulting in diminished sexual receptivity. A later study confirmed the above findings, also reporting that NR1I2 deletion reduced hippocampal levels of brain‐derived neurotropic factor, influencing behaviour and neural plasticity (Frye et al., 2014a).

More recently, studies report that deletion of NR1I3 and NR1I2 in rodents severely impairs recognition memory and increases development of anxiety‐like behaviours (Boussadia et al., 2016; Boussadia et al., 2018). Furthermore, both studies associated the absence of these NRs with alterations in the electroencephalographic signatures during sleep/awake periods. The authors suggested that these findings could be the result of alterations in the metabolism of endogenous mediators that regulate certain behaviours, thereby supporting the findings of previous studies.

4. CONCLUDING REMARKS

The current knowledge of the biological functions elicited by NR1I3 (CAR) and NR1I2 (PXR) in selected organs, such as the liver and intestine, has vastly improved over recent years; however, significantly less is known of the roles of these receptors in the brain. Constant exposure to endogenous and exogenous NR1I3 and NR1I2 ligands therefore necessitates an increased fundamental understanding of how the ligand‐based activation of these receptors, in healthy or pathological states, regulate brain biochemical functions, including proliferation and apoptosis, and psychological functions, such as behaviour. Despite an understanding of NR1I3 and NR1I2 functions in peripheral tissues, responses to ligand‐based activation of NR1I3 and NR1I2 are cell and tissue dependent, with differences attributed to downstream events (e.g. post‐translational modifications, co‐activation and co‐repression) within the NR1I3 and NR1I2 signalling pathways. The evidence from studies in cell lines and animal models have demonstrated that the effects of ligand‐based activation of NR1I3 and NR1I2 are cell dependent, so the approach to analyse their function in brain cells should be exploratory and circumscribed to the cell type used rather than based on assumptions generated from findings obtained on other models. In this regard, the development of selective ligands, comparative studies to assess responses in different tissues, characterisation of downstream events, studies of signalling pathways and the constitutive activity of NR1I3 in cells of the brain, and improved analyses of regional expression of NR1I3 and NR1I2 in the brain would prove useful in defining the roles of these receptors within the CNS.

4.1. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2019/20 (Alexander, Cidilowski et al., 2019; Alexander, Fabbro et al., 2019; Alexander, Kelly et al., 2019).

CONFLICT OF INTEREST

The authors declare no conflicts of interest.

Torres‐Vergara P, Ho YS, Espinoza F, Nualart F, Escudero C, Penny J. The constitutive androstane receptor and pregnane X receptor in the brain. Br J Pharmacol. 2020;177:2666–2682. 10.1111/bph.15055