Transcription factor-mediated regulation of the BCRP/ABCG2 efflux transporter: a review across tissues and species

Abstract

Introduction:

The breast cancer resistance protein (BCRP/ABCG2) is a member of the ATP-binding cassette superfamily of transporters. Using the energy garnered from the hydrolysis of ATP, BCRP actively removes drugs and endogenous molecules from the cell. With broad expression across the liver, kidney, brain, placenta, testes, and small intestines, BCRP can impact the pharmacokinetics and pharmacodynamics of xenobiotics.

Areas covered:

The purpose of this review is to summarize the transcriptional signaling pathways that regulate BCRP expression across various tissues and mammalian species. We will cover the endobiotic- and xenobiotic-activated transcription factors that regulate the expression and activity of BCRP. These include the estrogen receptor, progesterone receptor, peroxisome proliferator-activated receptor, constitutive androstane receptor, pregnane X receptor, nuclear factor e2-related factor 2, and aryl hydrocarbon receptor.

Expert opinion:

Key transcription factors regulate BCRP expression and function in response to hormones and xenobiotics. Understanding this regulation provides an opportunity to improve pharmacotherapeutic outcomes by enhancing the efficacy and reducing the toxicity of drugs that are substrates of this efflux transporter.

1. Background

The breast cancer resistance protein (human ABCG2/BCRP and rodent Abcg2/Bcrp), a 75 kDa membrane transporter, was first identified in a human breast cancer cell line, MCF-7/AdrVp, following the observation of an ATP-dependent mechanism that reduced the intracellular accumulation and toxicity of the chemotherapeutic drug, mitoxantrone [1]. A member of the human ABC transporter superfamily (G subfamily), BCRP utilizes ATP hydrolysis to actively transport substrates against their electrochemical gradient. The functional unit of the ABC transporters is characterized by two nucleotide binding domains (NBDs) and two transmembrane domains (TMDs) which together form the substrate translocation pathway. ATP hydrolysis in the cytoplasmic NBD region drives conformational changes in the TMD region that allow for substrate transport across the lipid bilayer. Unlike many other members of the ABC transporter superfamily of this transporter superfamily, the ABCG2 gene encodes for a half transporter with a single cytoplasmic NBD region followed by a single TMD region composed of six α-helices required for substrate recognition and transport. Recent crystal structure analysis has shown that BCRP exists in dimeric as well as higher order tetrameric states composed of two BCRP dimers [2,3]. It is thought that these higher order complexes are required in order for BCRP to be functional.

Human BCRP and rodent Bcrp isoforms transport a wide array of structurally- and chemically-unrelated substrates ranging from drugs to environmental contaminants as well as endogenous molecules [Reviewed in 4]. Commonly used chemotherapeutic drugs that are substrates of BCRP include mitoxantrone, methotrexate, doxorubicin, and topotecan [Reviewed in 4, and in 5]. Other BCRP substrates include the diabetes drug glyburide, antihypertensive prazosin, antibiotic nitrofurantoin, and anticancer tyrosine kinase inhibitors as well as porphyrins, bile acids, and estrones [Reviewed in 4, and in 5]. BCRP is most highly expressed in tissues regulating pharmacokinetics (colon, small intestine, liver, kidney) and blood-tissue barriers (brain, testis, placenta). Localized on the maternal surface of placental syncytiotrophoblasts as well as the luminal surface of hepatocyte canaliculi, proximal convoluted tubules, enterocytes and brain capillaries, BCRP serves a crucial protective role by limiting the cellular, and often tissue, accumulation of xenobiotics [Reviewed in 4,6, and in 7].

2. Pharmacological and physiological functions of BCRP

BCRP functions as an efflux transporter to limit the intracellular accumulation of various agents and toxicants. BCRP-mediated transport of drugs including topotecan, mitoxantrone, daunorubicin, and doxorubicin is often one of the factors responsible for poor oral bioavailability, resistance in tumor cells, and prevention of fetal exposure [8,Reviewed in 9,10]. Interestingly, the ABCG2 gene contains several polymorphisms and splicing variants that affect BCRP protein expression and function. The C421A variant results in a lysine for glutamine substitution in position 141 (Q141K) that markedly decreases BCRP protein function [10,11]. Such polymorphisms may have therapeutic implications for various drugs that are BCRP substrates. Patients heterozygous for the CA allele exhibit a 1.3-fold higher oral bioavailability of topotecan when compared to patients with the wild-type allele [12]. Similarly, intravenous administration of diflomotecan resulted in 3-fold higher plasma concentrations in patients heterozygous for the CA allele when compared to patients with the wild-type allele [13].

The ability of BCRP to actively efflux endogenous substrates and in turn, regulate cellular homeostasis has also been associated with pathophysiology. In the kidneys, BCRP localizes to the brush border membrane of proximal tubule cells where it participates in the renal elimination of uric acid, a by-product of purine metabolism. Compared to wild-type BCRP, expression of the common Q141K variant of the ABCG2 gene in Xenopus oocytes reduces uric acid transport by half [14]. Impaired renal excretion of uric acid can lead to the accumulation of urate in joints, which can precipitate attacks of gout [Reviewed in 15]. As a result, individuals carrying the loss-of-function BCRP variant are more likely to have a familial history of gout (odds ratio: 1.96), and earlier onset of gout (42 vs. 48 years). The variant is more common among gout patients (odds ratio: 3.26) compared to healthy individuals.[14,16–18].

In the placenta, mononuclear progenitor cells called cytotrophoblasts fuse together to form giant multinucleated syncytiotrophoblasts through the process of syncytialization. Syncytiotrophoblasts are critical for the production of pregnancy-related hormones, metabolism of xenobiotics, and expression of placental barrier transporters including BCRP. Notably, as primary trophoblasts differentiate into syncytiotrophoblasts, a concurrent increase in BCRP mRNA and protein expression occurs [19]. Likewise, forskolin-mediated stimulation of cell fusion in human BeWo choriocarcinoma cells, a model of first trimester trophoblasts, not only enhances syncytialization markers, but also induces BCRP mRNA and protein levels [20,21]. Interestingly, a genetic knockdown of BCRP in BeWo cells not only reduces transporter expression, but also decreases the expression of syncytialization markers [22]. Consistent with these results is the observation that BCRP mRNA levels as well as syncytialization markers are decreased in placentas obtained from gestational pathologies including preeclampsia and intrauterine growth restriction [23–25]. Taken together, these data suggest that BCRP may be important in the physiology and pathophysiology of the human placenta.

While the majority of studies have focused on the ability of BCRP to confer cellular protection by removing substrates such as drugs and toxicants, there is growing evidence that this transporter may possess an intrinsic role in preventing cellular stress from non-substrates as well. For example, silencing of BCRP in BeWo cells and primary trophoblasts significantly increased cytokine- and ceramide-induced apoptosis [22,25]. Similarly, subpopulations of cancer cell lines MCF-7 and H460, selected for resistance to the BCRP substrate mitoxantrone, exhibited a higher survival rate and decreased apoptosis in response to nutrient starvation as compared to their non-selected counterparts [26]. BCRP-expressing human embryonic stem cells were also more resistant to ultraviolet light-induced cell death, an effect that was abolished in the presence of a known BCRP inhibitor, KO143 [27]. Moreover, subpopulations of the bladder cancer line T24 and medulloblastoma cells that contained high levels of BCRP were also more resistant to radiation treatment [28,29]. Under basal conditions and following gamma radiation, Ding et al., 2016 observed that protein expression of autophagy markers, LC3-II and SQSTM1, was significantly higher in BCRP-expressing MCF7 and H460 cell lines [26]. Together, these data indicate that BCRP has an important cellular role in promoting survival, possibly by regulating cell stress and autophagy.

This review summarizes the tissue-specific regulation of BCRP across species using nuclear receptor and transcription factor signaling (Figure 1). Understanding how these signaling pathways influence BCRP expression and function can be used to improve therapeutic outcomes while reducing the toxicity of drugs that are substrates of this efflux transporter.

Figure 1. Transcription Factor-Mediated Activation of ABCG2 Gene.

The location of response elements for nuclear hormone receptors and transcription factors involved in regulating the transcription of ABCG2/BCRP are shown relative to the transcriptional start site.

3. Estrogen receptors (NR3A1/2)

Steroid hormones, such as 17ß-estradiol, influence development, differentiation, and growth in various cell types and tissues. Responses to 17ß-estradiol are largely mediated by two nuclear estrogen receptors, ERα and ERß, as well as the membrane-bound GPR30. The nuclear ER isoforms have distinct and overlapping patterns of expression and modulation of transcription. Ligand-bound estrogen receptors undergo conformational changes that enable dissociation from chaperone proteins in the cytoplasm. Upon dissociation, ligand-bound estrogen receptors translocate to the nucleus, dimerize, and bind to estrogen response elements (ERE) found in the promoters of target genes [30,Reviewed by 31,32]. Studies demonstrate that estrogens have the ability to regulate BCRP expression in human breast cancer cell lines, primary and immortalized cells of the placenta, as well as animal models (Table 1).

Table 1.

Estrogen Receptor Regulation of BCRP/Bcrp Expression and Function

Model	Species	Agonists	Genetic Targeting	mRNA	Protein	Function	References
MCF-7 Breast Cancer Cells	Human	Estrone (10 nM), Estradiol (3 nM), and Diethylstilbestrol (0.1 nM)	siRNA	↑↔	↑↓	↓; Topotecan	[35,36]
T47D Breast Cancer Cells	Human	17-ß estradiol (0.1–10,000 nM)		↑↔			[33,34]
BeWo Chorio-carcinoma Cells	Human	Estriol (10 nM-100 μM), Estradiol (10 nM-10 μM), Estrone (10 nM-10 μM)		↑↓↔	↑↓	↓; Mitoxantrone	[41–43,164]
Primary Trophoblasts	Human	Estradiol (100 nM)		↑	↑		[40]
PA-1 Ovarian Cancer Cells	Human	Estradiol (10 nM)		↑			[33]
Brain Capillaries	Rat	Propylpyrazoletriol (1 nM), Diarylpropionitrile (10 nM), Estradiol (1 nM-10 nM)			↓	↓↔; BODIPY FL prazosin	[49–51]
Brain Capillaries	Mouse	Estradiol (0.1 mg/kg)/(10 nM)	ERα and ERß Knockout Models		↓↔	↓↔; BODIPY FL prazosin	[49,51]

3.1. Human breast cancer cell lines

Although an ERE sequence has been identified in the promoter region of ABCG2 (between nucleotides −243 and −115), the influence of estrogen on BCRP transcription in various models has been conflicting [33]. Treatment of ER-positive human breast cancer T47D:A18 cells with 10 nM 17ß-estradiol for 24 h induced BCRP mRNA by 3-fold. Furthermore, mutations and deletions of the ERE in the ABCG2 promoter attenuated and abrogated, respectively, this induction suggesting a direct regulatory role for estrogen [33]. By comparison however, a T47D cell line transfected with an ABCG2 promoter luciferase construct revealed that 17ß-estradiol (0.1–10,000 nM) treatment for 24 h had no effect on luciferase activity [34]. The authors hypothesized that the lack of promoter activity could be attributed to a low endogenous level of ERα in these cells. Transfection of the T47D cell line with an ERα construct resulted in a significant induction of luciferase activity following 17ß-estradiol treatment, suggesting that the endogenous expression of nuclear receptors in model cell lines should be taken into careful consideration [34]. Likewise, treatment of human breast cancer MCF-7 cells with 17ß-estradiol (0.03–3 nM) for 72 h induced BCRP mRNA between 2.7- to 8-fold. This response was abolished by concomitant treatment with tamoxifen, a selective estrogen receptor modulator [35]. By comparison, work by a separate laboratory observed a time- and dose-dependent decrease in BCRP protein in MCF-7 cells treated with estrone (0.01–10 nM) and 17ß-estradiol (0.003–3nM), for up to 96 h. A similar decrease in BCRP protein expression was also observed with diethylstilbestrol (0.001–0.1 nM) treatment [36]. The ability of estradiol to inhibit BCRP expression was partially reversed by co-incubation with tamoxifen. Furthermore, genetic down-regulation of ERα in MCF-7 cells using siRNA prevented the repression of BCRP by estradiol, whereas, ERα-negative A549 cells exhibited no changes in BCRP protein expression following exposure to 17ß-estradiol [36]. Ultimately, these studies suggest that cell lineage, type of cell lines, duration of treatment, and the endogenous expression of ERα may impact estrogen receptor-mediated regulation of BCRP in breast cancer cell lines.

3.2. Placenta

During pregnancy, the plasma concentration of 17ß-estradiol increases steadily over 60-fold from preovulatory peaks to term gestation [Reviewed in 37]. 17ß-estradiol participates in the onset of parturition, production of progesterone, maturation of the fetal adrenal glands, angiogenesis of placental villi, and differentiation of trophoblasts [Reviewed in 37,38]. ERα protein is expressed in villus cytotrophoblasts, amniotic fibroblasts, and vascular pericytes, whereas ERß expression is confined to differentiated syncytiotrophoblasts [Reviewed in 39]. Similar to breast cancer cells, divergent findings have been observed in estrogen-mediated regulation of BCRP expression and function in in vitro primary and immortalized trophoblast models.

Primary trophoblasts isolated from human term placentas were treated with 100 nM 17ß-estradiol for 12 and 48 h. Estradiol induced BCRP mRNA (12 h) and protein (48 h) by 1.5-fold as compared to vehicle controls [40]. It is important to note that only modest changes in BCRP protein expression were observed at the 24 h time point. As in the case of breast cancer cells, the duration of treatment is an important aspect to consider when studying the effects of estrogen-mediated regulation of placental transporters in primary trophoblasts [40,41]. Treatment of human BeWo choriocarcinoma cells with 1–10 μM estriol for 48 h also induced BCRP mRNA and protein by 2-fold [42]. Transporter up-regulation was abolished by concomitant treatment with the ICI-182,780 estrogen receptor antagonist. The positive regulatory role of estrogens was also observed by Yasuda et al., 2006, where BeWo cells treated with different forms and concentrations of estrogen such as estrone (10 nM – 10 μM), estriol (10 nM – 10 μM), and 17ß-estradiol (10 nM – 10 μM) for 72 h exhibited a concentration-dependent induction of BCRP mRNA and protein [43].

Conversely, a study by Wang et al. noted that treatment of BeWo cells with 1 μM 17ß-estradiol for 72 h reduced BCRP mRNA and protein levels [41]. The same laboratory confirmed this effect in 2008, as treatment of BeWo cells with an even lower concentration of 5 nM 17ß-estradiol for 72 h reduced BCRP mRNA and protein by half [42]. It has also recently shown that treatment of BeWo cells with 1– 10 μM genistein, a soy isoflavone that acts as a phytoestrogen, results in the dose-dependent down-regulation of BCRP mRNA, protein, and function as measured by cellular accumulation of the substrate ³H-glyburide. Down-regulation of BCRP expression by genistein was attenuated by ICl 182,780 (1 μM) [44]. As gestation progresses, estrogen secretion from the placenta increases whereas BCRP protein expression is thought to decrease [43,45]. While only an association, it supports the ability of estrogens to reduce BCRP expression.

Secretion of other pregnancy hormones, such as progesterone, by the placental models can potentially explain disparities across studies. Concomitant treatment of BeWo cells with progesterone and estradiol together augmented the induction of BCRP mRNA and protein expression when compared to progesterone treatment alone [41]. Consequently, endogenous production and secretion of progesterone by primary trophoblasts and BeWo cells may influence data in studies examining the ability of estrogens to regulate BCRP in the placenta.

Subsequent studies have interrogated which ER receptor is responsible for regulating BCRP in the placenta. Unlike breast cancer cells, the transcriptional regulation of BCRP by estradiol in BeWo cells was primarily mediated through ERß. Reduction of ERα expression using siRNA did not alter the estradiol-mediated induction of BCRP expression, suggesting that ERß is the main regulatory ER isoform in BeWo cells [42]. Treatment of BeWo cells with 17ß-estradiol (1 μM) down-regulated ERß mRNA expression and had no significant effect on ERα mRNA, further pointing towards ERß as the predominant regulatory isoform [41]. In pregnant mice, the highest levels of placental Bcrp mRNA are on gestational day (GD) 15 [46]. As compared to GD10, the placental mRNA expression of ERα and ß is significantly lower and higher, respectfully, on GD15 [46]. This correlation further supports the likely involvement of ERß isoform in regulating BCRP expression in the placenta.

3.3. Brain capillaries

Due to the presence of aromatase, an enzyme responsible for the synthesis of 17ß-estradiol [47], and estrogen receptors in the cerebral microvasculature [48], the expression of BCRP in brain endothelial cells also makes this transporter a target of estrogen-mediated regulation. Unlike breast cancer and placenta models, isolated mouse and rat brain capillaries exhibit a consistent reduction of Bcrp mRNA, protein, and function in response to 17ß-estradiol treatment. Isolated rat brain capillary membranes exposed to 17ß-estradiol (1–10 nM) for 1 h and 6 h exhibited reduced Bcrp function, as measured by luminal accumulation of a fluorescent Bcrp substrate, BODIPY FL prazosin [49–52]. This reduction in activity was comparable to treatment with a known Bcrp inhibitor, fumitremorgin C [49,50]. Inhibition of Bcrp was attributed to ERß signaling as rat brain capillaries incubated with 10 nM diarylpropionitrile, a selective ERß agonist, demonstrated decreased Bcrp protein expression and function that mirrored responses to 17ß-estradiol [49,51]. By comparison, exposure of rat brain capillaries to an ERα-specific agonist (1 nM propyl pyrazole triol) or antagonist (100 nM methyl-piperidino-pyrazole) had no effect on Bcrp levels [49,51,52]. The authors speculated that initial estradiol signaling occurred through both ER isoforms leading to internalization of the Bcrp protein and thus a loss of function at early time points whereas the down-regulation of Bcrp protein expression was mediated directly through ERß activation and proteasomal degradation [49–51]. In comparison to breast cancer and placenta models, expression of Bcrp in rodent brain capillaries appears to be significantly more sensitive to regulation by estrogen as much lower concentrations and exposure times elicit transcriptional and functional responses.

4. Progesterone receptors (NR3C3)

Progesterone is a critical regulator of embryogenesis, menstruation, metabolism, and transcriptional regulation of membrane transporters. In humans, the progesterone receptor exists as two isoforms, PR-A and PR-B, which arise from alternate translation of a single precursor progesterone receptor mRNA [53]. While both PR-A and PR-B mediate gene transcription, each isoform acts as an independent transcription factor with divergent properties [Reviewed in 54,55]. Upon progesterone binding, the receptors undergo conformational changes that enable translocation, dimerization, and binding to progesterone response elements in the promoter regions of target genes to either induce or repress transcription. A progesterone response element has been identified in the promoter region of ABCG2, suggesting this hormone can regulate BCRP expression and activity [56].

4.1. Placenta

Incubation of BeWo cells with progesterone (1 – 10 μM) for up to 72 h induced BCRP mRNA and protein expression up to 2-fold (Table 2) [41,56]. Furthermore, progesterone decreased the intracellular accumulation of a classical BCRP substrate, mitoxantrone, consistent with enhanced BCRP activity [56]. Human BeWo choriocarcinoma cells express both PR isoforms [56]. Progesterone stimulation of PR-B-overexpressing BeWo cells preferentially stimulated BCRP protein expression when compared to PR-A-transfected cells, suggesting that PR-B is the major regulatory isoform [56]. As aforementioned, induction of BCRP was further enhanced with concomitant estradiol treatment and resulted in a 2-fold induction of PR-B mRNA expression, signifying additive roles in regulating BCRP levels [41]. Up-regulation of BCRP protein by progesterone was reversible through co-administration with the PR and glucocorticoid receptor antagonist, RU-486 [56]. It is important to note that progesterone-mediated regulation was concentration- and time-dependent in BeWo and primary human trophoblast cells. Exposure to progesterone at concentrations lower than 10 μM or for less than 24 h had no effect on BCRP expression [34,40,41].

Table 2.

Progesterone Receptor Regulation of BCRP/Bcrp Expression and Function

Model	Species	Agonists	mRNA	Protein	Function	References
T47D Breast Cancer Cells	Human	Progesterone (10 nM-100 μM)	↑			[34]
BeWo Choriocarcinoma Cells	Human	Progesterone (10 nM-100 μM)	↑↔	↑	↑; Mitoxantrone	[34,41]
Primary Trophoblasts	Human	Progesterone (100 nM)	↔	↔		[40]
Placenta	Mouse	Progesterone (16 mg/kg)	↔	↔		[165]

4.2. Human breast cancer cell lines

The ability of progesterone to regulate BCRP in human breast cancer cell lines is less consistent compared to regulation in placental cell lines. To examine the responsiveness of BCRP to progesterone, Yasuda et al., 2009 transfected the T47D cell line with an ABCG2 promoter-luciferase plasmid. Following exposure to a range of progesterone concentrations (1 nM – 10 μM) for 24 h, a ~2.6 fold induction in luciferase activity was demonstrated. Activation of luciferase activity could be reversed by concomitant exposure to 100 nM RU-486 [34]. Interestingly, following treatment with progesterone (0.0001 nM – 10 μM) for a longer time period (48 h), BCRP expression and activity were significantly down-regulated in both T47D and MCF7 cell lines [57,58]. As a result of progesterone treatment, T47D cells were approximately 5-fold more sensitive to mitoxantrone-induced cytotoxicity (LC₅₀ control: 1.36 μM; LC₅₀ progesterone: 0.26 μM) [57]. The discrepancy between progesterone-mediated changes in promoter activity and BCRP mRNA and protein expression may be explained by the ability of progesterone to also signal through posttranscriptional [59–62] and posttranslational mechanisms [63–66]. Likewise, BCRP is known to be highly regulated by posttranscriptional [67–70] and posttranslational modifications [71–73]. More studies are required to determine the direct effect of hormones such as progesterone on the posttranscriptional and posttranslational regulation of BCRP expression.

5. Peroxisome proliferator-activated receptors (NR1C1/2/3)

Peroxisome proliferator-activated receptors (PPAR) are nuclear hormone receptors with three predominant isoforms denoted as PPARα, PPARß, and PPARγ. Upon ligand binding, PPARs translocate to the nucleus, heterodimerize with retinoid x receptor alpha, and bind to PPAR response elements in the promoter regions of target genes and, in turn, modulate their expression. PPARs regulate lipid and glucose homeostasis, metabolic enzyme activity, and expression of drug transporters including BCRP [74–76].

5.1. PPARα

5.1.1. Brain

A number of studies have observed a consistent induction of BCRP mRNA, protein, and function in response to PPARα signaling (Table 3). Treatment of immortalized human cerebral microvascular endothelial hCMEC/D3 cells with PPARα ligands, GW7647 (20 nM) and clofibrate (100 μM), induced BCRP mRNA. Consistent with PPARα-mediated induction of BCRP mRNA, treatment of hCMEC/D3 cells with increasing GW7647 (1.25 nM – 20 nM) and clofibrate (200 nM – 125 μM) concentrations for 6–72 h revealed concentration- and time-dependent induction of BCRP protein and function [77]. Targeted knockdown of PPARα with siRNA in hCMEC/D3 cells reduced BCRP protein expression further confirming a mechanistic role for PPAR [77].

Table 3.

Peroxisome Proliferator-Activated Receptor Regulation of BCRP/Bcrp Expression and Function

Model	Species	Agonists	Genetic Targeting	mRNA	Protein	Function	References
hCMEC/D3 Blood Brain Barrier Cells	Human	Clofibrate, GW7647	siRNA	↑	↑	↑; Mitoxantrone	[77]
Dendritic cells	Human	Rosiglitazone, Troglitazone, GW7845		↑	↑	↑; Hoechst	[81]
CD-1 Brain Capillaries	Mouse	Clofibrate		↑	↑	↑; BODIPY FL prazosin	[78]
Liver	Mouse	Wy14643, GW7647, Clofibrate		↑	↑		[74,75]
Small Intestine	Mouse	Wy14643, GW7647		↑	↑		[74]

Induction of Bcrp mRNA and protein by clofibrate has also been observed in isolated brain capillaries from CD-1 mice exposed to 125 μM clofibrate for 6 h [78]. Interestingly, while the up-regulation of BCRP in human hCMEC/D3 cells was attenuated with known PPARα antagonists, MK886 (500 nM) and GW6471 (500 nM), the clofibrate-induced transport activity of Bcrp in CD-1 mouse brain capillaries was not, suggesting some potential species differences [77,78]. Treatment of isolated rat brain capillaries with other PPARα agonists including linoleic acid (10 μM), perfluorooctanoic acid (10 nM), and perfluorooctane sulfonate (10 nM) similarly increased Bcrp protein expression and transport activity. This induction was abrogated by the PPARα antagonist GW6471 [79]. Taken together, these data suggest that PPARα can regulate both rodent and human Bcrp/BCRP expression in the brain.

5.1.2. Liver, small intestine, and kidney

The ability of PPARα to regulate Bcrp expression extends beyond the brain to the liver, intestine, and kidney. Mice administered PPARα agonists, including clofibrate (500 mg/kg/d for 10 days i.p.), Wy14643 (0.1% enriched diet, ad libitum for 3 days), GW7647 (0.01% enriched diet, ad libitum for 3 days), or the environmental contaminant perfluorooctanoic acid (3 mg/kg/day for 7 days by oral gavage), demonstrated consistent up-regulation of hepatic Bcrp mRNA and protein up to 3-fold [74,75,80]. As expected, administration of these agonists to PPARα-null mice had no effect on the mRNA and protein expression of Bcrp in the liver. In fact, PPARα null mice exhibit significantly lower constitutive expression of hepatic Bcrp protein when compared to the wild-type mice, suggesting that PPARα also influences basal Bcrp regulation [74,75,80]. PPARα agonists also up-regulate Bcrp expression in the intestines and kidneys. Wild-type C57BL mice fed a diet containing 0.1% Wy14643 for 3 days exhibited a 1.5-fold induction in intestinal Bcrp mRNA when compared to mice on a control diet [74]. Similarly, exposure to perfluorooctanoic acid induced expression of renal Bcrp protein by 1.5-fold [80]. Taken together, these data support the ability of PPARα to up-regulate BCRP/Bcrp across a wide array of tissues in rodents and humans.

5.2. PPARγ

Ligand-activated PPARγ can also induce BCRP mRNA and protein expression in a number of cell types. Treatment of human dendritic cells and colorectal adenocarcinoma Caco-2 cells with PPARγ agonists, rosiglitazone (25 nM-10 μM), GW7845 (100 nM), and troglitazone (1000 nM), induced BCRP mRNA and protein [81,82]. As expected, co-treatment of dendritic cells with the PPARγ antagonist, GW9662, attenuated BCRP up-regulation [81]. Lin et al., 2016 has similarly shown that BCRP expression in the placenta is regulated in part by PPARγ. Treatment of human BeWo trophoblasts with the PPARγ agonist, rosiglitazone, resulted in the induction of BCRP mRNA, protein, and function. By comparison, treatment with a PPARγ antagonist T0070907 down-regulated expression of BCRP mRNA and protein [83]. From the available data, it appears that activation of PPARα and PPARγ can induce the transcription of BCRP expression across numerous primary and immortalized cell lines as well as in rodent tissues.

6. Constitutive androstane receptor (NR1I3)

The constitutive androstane receptor (CAR) regulates energy homeostasis, drug metabolism, and xenobiotic transport [84,Reviewed in 85,86,87]. Upon ligand binding, CAR undergoes translocation to the nucleus where it forms a heterodimer with the retinoid X receptor alpha and binds to cis elements in the promoter region of target genes to augment their transcription [Reviewed in 85,88,89]. As an orphan receptor, CAR is activated by a wide variety of xenobiotic chemicals often at micromolar concentrations. This differs from typical steroid hormone receptors that respond to nanomolar concentrations of endogenous ligands [Reviewed by 85]. CAR also differs from classical steroid hormone receptors as its activation can occur through a ligand-independent mechanism, as in the case of phenobarbital. Phenobarbital induces the nuclear translocation of CAR by competitively binding to the epidermal growth factor receptor and causing dephosphorylation of CAR [Reviewed in 85,88,89].

Several studies have pointed to the involvement of CAR in the transcriptional regulation of human efflux transporters including Bcrp (Table 4). Treatment of isolated rat brain capillaries with the CAR agonist, phenobarbital, increased Bcrp protein expression [90]. Accordingly, isolated rat and mouse brain capillaries treated with phenobarbital and 3,3′,5,5′-tetrachloro-1,4-bis(pyridyloxy) benzene (TCPOBOP) exhibited a significant enhancement of transport activity of Bcrp, as measured by the luminal accumulation of BODIPY-prazosin [90]. The induction of Bcrp by phenobarbital and TCPOBOP was attenuated using okadaic acid (10 nM), a protein phosphatase 2A inhibitor; together, these data support a regulatory role of CAR in Bcrp expression [90]. Isolated capillaries from CAR-null mice did not exhibit enhanced Bcrp function in response to TCPOBOP [90]. These data were further confirmed in vivo where mice and rats treated with TCPOBOP or phenobarbital, exhibited up to 2-fold induction of Bcrp protein expression in brain capillary and liver membranes [90].

Table 4.

Constitutive Androstane Receptor Regulation of BCRP/Bcrp Expression and Function¹

Model	Species	Agonists	mRNA	Protein	Function	References
Primary Hepatocytes	Human	Phenobarbital, CITCO	↑	↑		[88,91]
Liver	Mouse	TCPOBOP		↑		[90]
Brain Capillaries	Porcine	CITCO	↑	↑	↑; Hoechst	[89]
Brain Capillaries	Rat	Phenobarbital		↑	↑; BODIPY FL prazosin	[90]
Brain Capillaries	Mouse	TCPOBOP		↑	↑; BODIPY FL prazosin	[90]

¹Abbreviations: CITCO (6-(4-chlorophenyl)imidazo[2,1-b][1,3]thiazole-5-carbaldehyde O-(3,4-dichlorobenzyl) oxime), TCPOBOP (3,3′,5,5′-tetrachloro-1,4-bis(pyridyloxy) benzene)

While TCPOBOP induced Bcrp protein expression and transporter activity in mice and rats, it had no effect on expression or activity in isolated porcine brain capillary endothelial cells [89]. Alternatively, the same cells treated with a different CAR agonist, 6-(4-chlorophenyl)imidazo[2,1-b][1,3]thiazole-5-carbaldehyde O-(3,4-dichlorobenzyl) oxime (CITCO), exhibited significant and concentration-dependent increases in Bcrp mRNA, protein, and function [89]. Thus, selection of an appropriate ligand is important as receptor activation of CAR is species-specific. Similar results were observed in primary human hepatocytes infected with a human CAR expressing adenovirus. Treatment of these cells with 0.3 μM CITCO for 24 h increased BCRP mRNA expression by 1.7-fold [88]. Similarly, exposure of primary human hepatocytes (from three male and four female donors) to phenobarbital (3.2 mM) for 72 h up-regulated BCRP mRNA by over 3-fold [91]. Although not statistically significant, human HuH-7 hepatoma cells treated with phenobarbital (2 mM) up-regulated BCRP mRNA [92]. Together, these data indicate that CAR positively regulates BCRP transcription and expression across species and tissues.

7. Pregnane X receptor (NR1I2)

The pregnane X receptor (PXR), a member of the nuclear hormone receptor superfamily of transcription factors, regulates genes involved in xenobiotic and endobiotic oxidation, metabolism, transport, and conjugation [Reviewed in 93,94]. Similar to the mechanism of CAR, ligand binding initiates PXR translocation to the nucleus where it forms a heterodimer with retinoid X receptor alpha, binds to a xenobiotic responsive enhancer (XRE) element located upstream of the transcriptional start site and regulates transcription of target genes [Reviewed in 95].

Accumulating evidence suggests that PXR regulates BCRP expression in a species- and tissue-specific manner (Table 5). Isolated mouse, rat, and porcine brain capillaries treated with PXR ligands in vivo (pregnenolone 16α-carbonitrile (PCN; 50mg/kg/d; 2 days)) or in vitro (hyperforin (1 μM; 24h); rifampicin (5 μM; 24 h)) exhibited increased Bcrp expression and function [90,94]. PXR-mediated regulation of Bcrp was also observed in mouse TM4 Sertoli cells. Treatment of TM4 cells with dexamethasone (100 μM) and PCN (50 μM) for 24 h induced Bcrp mRNA and protein expression, which was abrogated by the PXR antagonist, ketoconazole (10 μM). Genetic knockdown of PXR in TM4 cells using siRNA also reduced Bcrp protein expression, further confirming the direct regulatory role of PXR in testis [96]. The PXR-mediated regulation of Bcrp expression was also evident in mouse placentas. Treatment of pregnant C57BL/6 mice with 50 mg/kg i.p. PCN from gestational days 13 to 17 resulted in higher placental Bcrp mRNA levels when compared to PXR-null and heterozygous controls [97].

Table 5.

Pregnane X Receptor Regulation of BCRP/Bcrp Expression and Function

Model	Species	Agonists	mRNA	Protein	Function	References
Primary Hepatocytes	Human	Rifampicin	↑			[91]
Liver	Human	Carbamazepine	↑			[166]
Liver	Mouse	Pregenolone-16 alpha-carbonitrile, 2-acetylaminofluorene	↑↔			[98,99]
hCMEC/D3 Blood Brain Barrier Cells	Human	Rifampicin	↔			[103]
Brain Capillaries	Porcine	Rifampicin	↑	↑	↑; Hoechst	[94]
Small Intestine	Mouse		↔			[99]

Still, discrepancies in the PXR-mediated regulation of hepatic Bcrp expression have been reported [98–100]. BALB/c mice treated with PCN (400 mg/kg i.p. for 4 days) had unaltered expression of liver Bcrp mRNA [99]. Similarly, C57BL/6 mice treated with the PXR agonist, PCN (50 mg/kg i.m.), and antagonist, RU486 (50mg/kg i.m.), for 3 days had no change in liver Bcrp mRNA [100]. Interestingly, treatment of C57BL/6 mice with 2-acetylaminofluorene (150 mg/kg and 300 mg/kg i.p.) for 7 days resulted in a 2–2.5 fold induction of liver Bcrp mRNA [98]. This induction was prevented in PXR-null mice, supporting the direct involvement of this nuclear receptor in the up-regulation of Bcrp [98]. Discrepancies in mouse liver studies may be due to the dose of the PXR agonist, the duration of treatment, and potentially the strain of mice used.

The ability of PXR to regulate BCRP expression was examined in various human cell lines. Hepatocytes, obtained from three male and four female donors, exhibited a ~2.7-fold increase in BCRP mRNA after treatment with the PXR ligand rifampicin (50 μM) for 72 h [91]. Similarly, HepG2 cells, a human liver cancer cell line, over-expressing PXR exhibited 7-fold higher BCRP mRNA expression when compared to parental cells [101]. Moreover, PXR-mediated regulation of BCRP expression was also observed in MCF and MDA-MD-231 breast cancer cells as treatment with SR12813 (0.3 μM), a PXR agonist, for 24 h resulted in ~3-fold induction of BCRP mRNA in both lines [102]. By comparison, human cerebral hCMEC/D3 endothelial cells exposed to rifampicin (25 μM) for 24 h had no observable change in BCRP mRNA expression [103]. The authors of the study noted that while PXR mRNA was detectable in hCMEC/D3 cells, the expression was low potentially explaining the absence of PXR-mediated regulation [103]. Future studies should also take into consideration the endogenous expression of nuclear receptors in models of interest as that may alter the extent of regulation observed.

8. Aryl hydrocarbon receptor (AHR)

The aryl hydrocarbon receptor (AHR) is a member of the basic helix-loop-helix/PER-AHR nuclear translocator superfamily of transcriptional factors. AHR regulates genes involved in cell proliferation, vascular and immune system functioning, phase I/II drug metabolism, and more recently, xenobiotic transport [91,104,105,Reviewed in 106]. AHR is bound by several chaperone proteins in the cytoplasm and remains in an inactive state [107]. Upon ligand binding, AHR dissociates from chaperones and translocates to the nucleus where it forms a heterodimer with the aryl hydrocarbon nuclear translocator and regulates the transcription of various target genes. AHR ligands include several toxicants such as dioxins and polychlorinated biphenyls [108,109].

AHR ligands positively regulate human intestinal expression and function of BCRP across several primary and immortalized cell models (Table 6). Treatment of human Caco-2 cells with benzo[a]pyrene (10 μM), 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD, 50 nM), benzo[k]fluoranthene (5 μM), and indolo[3,2-b]carbazole (2.5 μM) for 24 h induced BCRP mRNA expression between 3- and 6-fold [110]. By 72 h, a similar degree of induction was observed in BCRP protein [110]. Up-regulation of BCRP was abolished using the AHR antagonists, PD98059 (10 μM) and 3’-methoxy-4’-nitroflavone (10 μM) [110]. Significant induction of BCRP mRNA and protein expression has also been reported in C2bbe1 (subclone of Caco-2), LS180 (colorectal adenocarcinoma), LS174T (colorectal adenocarcinoma), MCF-7, HuH-7, HepaRG (human hepatoma cells), and HepG2 cells, as well as primary human colonocytes and hepatocytes in response to 10 nM TCDD treatment for 24 to 72 h [91,92,111–113]. This response was abrogated by the AHR antagonist, 3,4-dimethoxyflavone (10 μM) [111]. Moreover, siRNA knockdown of AHR in LS174T and MCF7 cells attenuated the up-regulation of BCRP mRNA by 3-methylcholanthrene (3MC; 1 μM) and TCDD (10 nM) [111,112]. Likewise, site-directed mutagenesis of various AHR response elements in the ABCG2 promoter region resulted in significant reduction of 3MC-mediated activation in LS1274T cells further pointing to the involvement of AHR in the transcriptional regulation of this transporter [112].

Table 6.

Aryl Hydrocarbon Receptor Regulation of BCRP/Bcrp Expression and Function¹

Model	Species	Agonists	Genetic Targeting	mRNA	Protein	Function	References
Primary Trophoblasts	Human	3MC		↔			[167]
Primary Hepatocytes	Human	TCDD		↑			[91,111]
Hepa1c1c7 Liver Hepatoma Cells	Mouse	TCDD		↔			[111]
Liver	Mouse	3MC		↔			[99]
Primary Colonocytes	Human	TCDD		↑			[111]
Caco-2 Colorectal Adenocarcinoma Cells	Human	TCDD, BP, indolo[3,2-b]carbazole, benzo[k]fluoranthene		↑	↑		[110]
C2bbe1 Colorectal Adenocarcinoma Cells	Human	TCDD, DBA, and 3MC	siRNA	↑	↑	↑; Mitoxantrone	[111]
LS180 Colorectal Adenocarcinoma Cells	Human	TCDD		↑			[111]
LS174T Colorectal Adenocarcinoma Cells	Human	TCDD		↑			[111,112]
CMT93 Rectal Carcinoma Cells	Mouse	TCDD		↔			[111]
Small Intestine	Mouse	TCDD (30 μg/kg on GD16)		↔			[111]
Small Intestine	Mouse	3MC		↔			[99]
hCMEC/D3 Blood Brain Barrier Cells	Human	TCDD		↑	↔		[103]
Brain Capillaries	Rat	TCDD			↑	↑; BODIPY FL prazosin	[115]
MCF-7 Breast Cancer Cells	Human	TCDD		↑			[111]
EMT 6 Breast Cancer Cells	Mouse	TCDD		↔			[111]

¹Abbreviations: 3MC (3-methylcholanthrene), TCDD (2,3,7,8-tetrachlorodibenzo-p-dioxin), BP (benzo[a]pyrene), DBA (dimethyl-benzo[a]pyrene), GD (gestational day)

The ability of AHR to regulate BCRP in the placenta has also been reported. Univariate regression analysis of BCRP and AHR mRNA expression across 137 racially and ethnically diverse term placentas revealed significant correlation (b= 0.29; P < 0.0001) between the transcription factor and transporter [11]. Pharmacological inhibition (3 μM CH223191) and genetic knockdown (shRNA) of AHR in primary human villous trophoblasts and immortalized placental in vitro models (BeWo, JEG3) significantly attenuated the 3MC-mediated induction of BCRP mRNA [114].

While AHR-mediated regulation of Bcrp was demonstrated in immortalized porcine brain microvascular endothelial cells as well as isolated rat spinal cord and brain capillaries, various in vivo and in vitro mouse studies observed no changes in Bcrp expression in response to prototypical AHR agonists [99,111,114–116]. Mouse mammary (EMT-6), hepatic (hepa1c1c7), and intestinal (CMT93) carcinoma cell lines exhibited no change in Bcrp mRNA expression in response to TCDD treatment [111]. Additionally, no significant changes in Bcrp mRNA levels were observed in the small intestine, liver, and colon of pregnant C57BL/6N mice treated with TCDD (30 μg/kg, i.p.) on gestational day 16 and evaluated 24 hr later [111]. The strain of mice does not seem to play a factor in this phenomenon as BALB/c mice treated with 3MC (30 mg/kg/d i.p. for 4 days) similarly had unchanged Bcrp levels in the liver and small intestine [99]. Comparison of the human ABCG2 and mouse Abcg2 genes revealed that although majority of the exons are conserved, most of the noncoding regions, such as the intron and 5’-flanking regions, are not [111]. From these studies, it can be concluded that AHR-mediated transcriptional regulation of BCRP appears to be species-specific.

9. Nuclear factor erythroid 2-related factor 2 (NFE2L2)

Nuclear Factor Erythroid 2-Related Factor 2 (NRF2), a basic region-leucine zipper type transcription factor, protects against inflammation, oxidative injury, apoptosis, and environmental toxins through activation of target genes involved in antioxidant and detoxification pathways [Reviewed in 117,118,119]. Under basal conditions, Kelch like ECH associated protein (KEAP1) negatively regulates this pathway by sequestering NRF2 in the cytoplasm and targeting it for proteasomal degradation. In response to oxidative stress or electrophilic attack, NRF2 dissociates from KEAP1, translocates to the nucleus where it heterodimerizes with small Maf proteins, and initiates transcription of antioxidant target genes [120,121].

Recently several studies have suggested that BCRP can be regulated by NRF2 (Table 7). Luciferase reporter assays and CHIP analysis revealed an antioxidant response element (ARE) located −431 to −420 bp upstream of the ABCG2 transcriptional start site, direct binding of NRF2 [121]. Treatment of primary human hepatocytes, HepG2, HuH-7, and NuLi cells (human epithelial cells) with NRF2 inducers, tert-butylhydroquinone (20 μM-200 μM; 24–48 h) and oltipraz (50 μM; 72 h), enhanced BCRP mRNA and protein expression in a concentration-dependent manner. Induction of BCRP was accompanied by increased expression of classical NRF2 target genes, NAD(P)H quinone oxidoreductase 1 and the glutamate-cysteine ligase modifier subunit [91,92,121,122]. Transfection with NRF2-specific shRNA also significantly lowered BCRP mRNA (up to ~5-fold) and protein expression and function in A548 cells [121]. Knockdown of NRF2 in DU145 cells (human prostate cancer model), SKOV3 cells (ovarian cancer model), HCT116 cells (human colorectal carcinoma), MDA-MB-231 (human breast carcinoma) and HepG2 cells also significantly reduced BCRP expression [121–125]. Consistent with these findings, transfection of NuLi and HF-2 cells with shRNAs targeted against KEAP1, the negative regulator of NRF2, up-regulated BCRP mRNA and protein expression [121,126]. In addition, univariate regression analysis of genes in 137 human placentas revealed a strong correlation between NRF2 and BCRP mRNA levels (b = 0.85; P < 0.0001) [11]. Taken together, NRF2 is a consistent inducer of human BCRP expression across numerous tissues and species.

Table 7.

Nuclear Factor Erythroid 2-Related Factor 2 Regulation of BCRP/Bcrp Expression and Function

Model	Species	Agonists	Genetic Targeting	mRNA	Protein	Function	References
Primary Hepatocytes	Human	Oltipraz		↑			[91]
HepG2 Liver Carcinoma Cells	Human	tert-Butylhydroquinone	siRNA	↑			[122]
Brain Capillaries	Rat	Sulforaphane			↑	↑; BODIPY-prazosin	[168]

10. Post -transcriptional and -translational regulation

Beyond transcriptional regulation, the expression, trafficking, and function of the BCRP transporter also relies on post-transcriptional and –translational mechanisms. A proximal miRNA response element has been identified in the 3’-untranslated region (3’-UTR) region of the ABCG2 gene [127]. Transfection of MCF-7 breast cancer cells with miRNA-328 or miRNA-519c expression plasmids resulted in a ~1.35-fold reduction of BCRP protein expression [127]. This reduction correlated with an accelerated ABCG2 mRNA degradation in the transfected cells suggesting the contribution of an mRNA-specific decay mechanism. Post-translational regulation of BCRP, including phosphorylation and glycosylation, has also been reported. With respect to N-glycosylation, a study has shown that mutation of a glycosylation site, Asn596, results in enhanced ubiquitination and subsequent decrease in BCRP protein expression [128]. The functionality of the Bcrp transporter has also been shown to be affected by N-glycosylation as maximal Bcrp activity correlated with its fully N-glycosylated isoform, as observed by excretion of fluorescently tagged methotrexate into a canalicular network of a sandwich-cultured rat hepatocyte model [129]. Trafficking of the BCRP transporter is also regulated by post-translational mechanisms as proper translocation to the plasma membrane depends on a phosphorylated threonine 362 [73]. Likewise, localization of BCRP protein to detergent-resistant lipid rafts within the plasma membrane is important for its function in the placenta [130].

11. Conclusion

Over the past decade, numerous studies have highlighted the involvement of hormones, nuclear receptors, and transcription factors in regulating drug transporter expression and function. As described in this review, BCRP expression can be regulated by steroid receptors (ER, PR), xenobiotic receptors (CAR, PXR, PPAR), and transcription factors (AHR, NRF2). The relative contribution of each receptor alone or combined to the basal and inducible expression of BCRP is not entirely clear and likely differs across tissues. Nonetheless, it is clear that multiple signaling pathways can up-regulate BCRP expression and function. Moving forward, additional research is needed to understand the impact of this transcriptional regulation on the overall pharmacokinetics, pharmacodynamics, and toxicity of BCRP substrates.

12. Expert opinion

Localized across several blood-tissue interfaces, BCRP actively limits the intracellular accumulation of numerous toxic xenobiotics by extruding them across cell membranes. During pregnancy, the BCRP transporter is localized to the placental syncytiotrophoblasts and limits fetal drug exposure by removing various drugs back into the maternal circulation. Correspondingly, the gestation diabetes drug and BCRP-substrate, glyburide, is prescribed during pregnancy due to its low fetal accumulation.

While this self-defense mechanism offers crucial protection under normal physiological circumstances, it is also a main component of multidrug resistance in cancer cells. Multidrug resistance in cancer therapy is often characterized by the overexpression of ABC efflux transporters and complicated by the broad substrate overlap between them. The pharmacokinetics of several chemotherapeutic drugs, including methotrexate, doxorubicin, and daunorubicin, have been shown to be altered due to their recognition and elimination by the BCRP, MDR1, and MRP transporters [12,131,132]. Additionally, ABC transporters share distinct overlap not only in their substrates but also in their transcriptional machinery. Similar to BCRP, the multidrug resistance protein 1 (ABCB1/MDR1/P-gp) and multidrug resistance-associated proteins (ABCC/MRPs) have been shown to be transcriptionally regulated by the estrogen receptor [40,133–138], progesterone receptor [40,135,136], peroxisome proliferator activated receptors [75,139–145], pregnane x receptor [141,145–147], aryl hydrocarbon receptor [145,146], nuclear factor erythroid 2-related factor 2 [143,145], and constitutive androstane receptor [145,146,148] across species and tissues. Extrapolation of data from models investigating the regulatory capacity of transcriptional factors requires careful characterization of other proteins and signaling pathways potentially affected. Unwarranted activation of transporters by ligands targeting a mutual regulatory transcription factor may result in confounding data with respect to changes in transporter functionality, especially when a test substrate recognized by multiple transporters is affected. This is of particular concern in drug resistance as simultaneous activation of multiple transporters via transcriptional pathways may render the translocation of substrate drugs not feasible.

Understanding the mechanisms that govern BCRP expression enables the identification of new molecular targets and greater expansion of precision medicine. Knockout mouse models targeting the Bcrp transporter demonstrated a significantly higher area under the curve of plasma concentration-time curve (AUC) for orally administered ciprofloxacin, afatinib, rucaparib, methotrexate, and vemurafenib, to name a few [149–153]. Similar effects were observed in individuals carrying common single nucleotide polymorphisms (SNPs) in the ABCG2 gene. SNPs, such as C421A/Q141K, result in decreased BCRP protein expression and function [11,44]. Accordingly, individuals carrying the C421A/Q141K SNP exhibit higher plasma AUC or C_max of BCRP substrate drugs such as sulfasalazine, gefitinib, rosuvastatin, and atorvastatin [Reviewed in 9]. With respect to transcriptional regulation, more studies are still needed to determine whether the transcriptional modifiers of BCRP discussed in this review alter the pharmacokinetic profiles of various BCRP substrates in vivo. While decreased BCRP expression and/or loss of its function typically correlates with enhanced plasma concentration of orally administered substrates, the transcriptional factors discussed regulate not only transporters but also other major xenobiotic disposition pathways including those involved in Phase I and II metabolism. In the 2020 US FDA guidance for metabolizing enzyme- and transporter-mediated drug interactions, no models or methods are recommended for the evaluation of transporter induction despite the known ‘cross-talk’ between regulators of metabolism and transport. It is clear that additional investigation and validation of in vitro models specifically for transporters is needed.

With respect to its critical role in drug disposition, BCRP has also been recognized by the U.S. Food and Drug Administration as a key transporter to consider in the evaluation of potential drug-drug interactions (DDIs) during drug development and prescribing [154]. Several DDIs between BCRP substrate drugs and BCRP inhibitors in humans have been identified. These include interactions between rosuvastatin and cyclosporine [155], sulfasalazine and curcumin [156], and atorvastatin and ritonavir [157], to name a few. It is also important to note that the expression and function of BCRP may be affected by pharmaceutical agents and toxicants alike, further complicating these interactions. Pharmaceutical agents such as gefitinib, topotecan, mitoxantrone, venlafaxine, promazine, etoposide, ifenprodil, among others, have been shown to affect the expression and/or function of the BCRP transporter [158–160]. Similarly environmental toxicants such as cadmium chloride, bisphenol A (BPA), genistein, and daidzein, to name a few, are also implicated in regulating BCRP expression and/or function [44,52,161–163]. Additional preclinical and clinical studies are warranted to determine whether drug interactions can be observed with drugs that induce or repress BCRP expression.

Review of current literature concerning the transcriptional regulation of BCRP revealed several aspects to consider when designing and evaluating studies across tissues and species. Many of the discrepancies between studies can be attributed to either 1. Inappropriate model selection or 2. A lack of an extensive model characterization. With respect to selection of an inappropriate model, several studies did not take into account the endogenous expression of target nuclear receptors or transcription factors. Accordingly, treatment with an agonist targeting a nuclear receptor that is not present in a given cell line or expressed at very low levels could result in a false negative and affect data interpretation across tissues and species. Another aspect to consider is the endogenous hormone secretion of a primary or immortalized cell line. Differences in the amount and type of hormones secreted between models may affect the observed transcriptional and translational changes in response to various agonists and become a confounding factor.

In terms of model characterization, disparate results observed by several studies can be attributed to the duration of agonist/antagonist treatment and/or their concentration. Future studies should consider including time courses and dose responses for each compound utilized as sensitivity and responsiveness may vary across different cell lines and tissues. Time courses are of particular importance when comparing changes in mRNA and protein expression. Moreover, genetic variability is another aspect to take into account especially when comparing results obtained from human primary cells and immortalized cell lines. Primary cells may contain single nucleotide polymorphisms spanning the promoter, exon, and intron regions of the target gene. With respect to the transcriptional regulation of BCRP expression, the promoter region of the BCRP gene should be screened for any potential polymorphisms that could affect transcriptional factor and/or nuclear receptor binding. Without this consideration, data obtained from primary cells may be significantly different from immortalized cell models.

Crosstalk between transcription factors and nuclear receptors has been identified in numerous species and tissues and hence adds another level of complexity. With numerous pharmaceuticals and toxicants affecting multiple signaling pathways, future studies need to examine the combined regulatory effect of multiple transcription factors of BCRP at the same time. Site directed mutagenesis targeting one or more of the response elements found in the promoter region of ABCG2 can identify the transcription factors that predominantly regulate BCRP expression. Ultimately, establishment of guidelines which cover each of the aforementioned concerns during the initial experimental design and characterization would aid in the standardization, translatability, and consistency of data regarding transcriptional regulation of transport proteins.

Article highlights:

• The breast cancer resistance protein (ABCG2/BCRP) is an ATP-dependent efflux transporter responsible for limiting the cellular accumulation of a wide variety of antibiotics, chemotherapeutics, anti-diabetics, anti-hypertensives, environmental contaminants, and endogenous molecules.

• BCRP maintains cellular homeostasis by actively eliminating endogenous intracellular substrates. Disruption of BCRP expression and/or function has been associated with pathophysiology that including gout.

• Drug transporter expression and function is regulated by nuclear hormone receptors and xenobiotic-activated transcription factors. Steroid receptors (ER, PR), xenobiotic receptors (CAR, PXR, PPAR), and transcription factors (AHR, NRF2) are all involved in the coordinated regulation of BCRP expression.

• While BCRP/Bcrp regulation by endobiotic- and xenobiotic-activated transcription factors is largely similar across mammalian species, there are notable exceptions that are important in translating studies from experimental models to humans.