Ex) Article Title, Author, Keywords
Ex) Article Title, Author, Keywords
Split Viewer
DTT 2023; 2(2): 111-123
Published online September 30, 2023
https://doi.org/10.58502/DTT.23.0021
Copyright © The Pharmaceutical Society of Korea.
Steffanus Pranoto Hallis1 
, Beam Ju Go2, Jun Min Yoo2, Geun Hee Cho2, Mi-Kyoung Kwak1,2
Correspondence to:Mi-Kyoung Kwak, mkwak@catholic.ac.kr
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
The development of therapy resistance in cancer is a complex mechanism with multifaceted and catastrophic consequences. BCRP (Breast Cancer Resistance Protein/ATP-Binding Cassette Subfamily G member 2; ABCG2), a drug efflux transporter, is recognized as a key regulator in obtaining multidrug resistance properties of cancer cells. NRF2 (Nuclear Factor Erythroid 2-Like 2; NFE2L2) plays an essential role in maintaining cellular redox homeostasis by regulating the expression of an array of antioxidant and detoxifying genes. In contrast to its protective roles in normal cells, a substantial body of evidence has demonstrated the negative aspects of NRF2 accumulation in cancer cells, including facilitated tumor growth, stemness, and therapy resistance. This review focuses on the linkage between BCRP and NRF2 in cancer and its consequences in acquiring therapy resistance. In particular, we highlight the role of the NRF2/BCRP axis in photodynamic therapy resistance and cancer stem cell maintenance, demonstrating that inhibition of this axis can be a beneficial strategy to overcome resistance to cancer therapy. A comprehensive understanding of the NRF2/BCRP regulating pathway will lead to the development of new therapeutic approaches and improve the prognosis of cancer patients.
KeywordsBCRP/ABCG2, NRF2/NFE2L2, multidrug resistance, chemoresistance, photodynamic therapy, cancer stem cells
The development of resistance in cancer cells remains a major obstacle to successful chemotherapy, contributing to treatment failure and disease progression. One key mechanism of chemoresistance is the overexpression of ATP-binding cassette (ABC) transporters, which are responsible for the efflux of cytotoxic anticancer agents from cancer cells (Robey et al. 2018). Among 48 human ABC transporters, which are classified into seven subfamilies, multidrug resistance protein 1 (MDR1 or P-glycoprotein/P-gp) encoded by ABC subfamily B member 1 (ABCB1) was initially described in Chinese hamster ovary (CHO) cells and later in sublines of human KB carcinoma cells bearing multidrug-resistance to either colchicine, vinblastine, or doxorubicin (Roninson et al. 1986). The multidrug resistance-associated protein 1 (MRP1, encoded by ABCC1) was identified in 1992 as the second member of multidrug ABC transporters from lung cancer cell line H69AR, a cell line model of doxorubicin, etoposide, and vincristine resistance (Cole et al. 1992). The third member of the multidrug ABC transporter was encoded by ABCG2 and concurrently reported by three groups. The first group identified ABCG2 in human MCF7 breast cancer cells, where it was named breast cancer resistance protein (BCRP) (Doyle et al. 1998). Additionally, ABCG2 was found in mitoxantrone-resistant S1-M1-80 human colon carcinoma cells (Miyake et al. 1999), as well as in placental tissue (Allikmets et al. 1998). The overexpression of multidrug ABC transporters in cancer has been linked to chemotherapy failure in clinical settings. As a result, numerous efforts have been made to overcome multidrug-resistant cancer through the development of transporter inhibitors (Gottesman et al. 2016).
Nuclear factor erythroid 2-like 2 (NFE2L2/NRF2) is a transcription factor that regulates the expression of gene arrays involved in protection against oxidative and electrophilic stress and maintenance of cellular homeostasis (Yamamoto et al. 2018). Despite its protective roles in normal tissues, constitutive activation of NRF2 confers cancer cells survival by multiple mechanisms: increasing drug elimination through phase II metabolizing enzymes, preventing intracellular drug accumulation, and suppressing apoptotic cell death (Bai et al. 2016; Choi and Kwak 2016). In particular, the antioxidant response element (ARE) has been identified as a binding site for NRF2 in the promoter region of MRP1 and BCRP, confirming the critical role of NRF2 signaling in the regulation of multidrug-resistant ABC transporters (Kurz et al. 2001; Singh et al. 2010; Ji et al. 2013). The review article focuses on providing an overview of BCRP and NRF2, and their association with therapy resistance in cancer. In particular, we highlight the role of the NRF2/BCRP axis in cancer stem cells (CSCs) and photodynamic therapy (PDT) resistance. We also suggest that inhibiting this axis can be an effective strategy to overcome resistance to cancer therapy.
BCRP, which is also known as mitoxantrone resistance-associated protein (MXR) or placenta-specific ABC transporter (ABCP), is a type of ABC transporter that belongs to the G subfamily of ABC transporters, also known as ABCG2. BCRP is responsible for translocating its substrates across cell membranes, against a concentration gradient, and this process is achieved through the binding and hydrolysis of ATP (Dean et al. 2001). BCRP is a 72-kDa protein consisting of 665 amino acids with six transmembrane segments and one extracellular loop. Unlike other subfamilies of ABC transporters, which contain two transmembrane domains (TMD) and two nucleotide (ATP)-binding domain (NBD), BCRP is a “half-transporter” with only one TMD and one NBD fused to a single polypeptide chain (Ozvegy et al. 2001; Taylor et al. 2017). Therefore, for BCRP to function properly, it needs to homodimerize, possibly through the formation of intra- and intermolecular disulfide bonds, or oligomerize (Bhatia et al. 2005; Henriksen et al. 2005).
BCRP has diverse physiological roles and is localized in various tissues and organs, including the placenta, mammary glands, gastrointestinal tract, kidney, blood-brain barrier, and hematopoietic stem cells (Robey et al. 2009). BCRP expression in the placenta is essential for protecting the fetus from topotecan, which has been administered in maternal animals. Treatment of pregnant mice with GF120918, a BCRP inhibitor, increased the relative fetal penetration of topotecan (Jonker et al. 2000). In Bcrp1(−/−) knockout mouse model, BCRP absence increased topotecan penetration to the brain by 1.5 folds, while the penetration increased by 12 folds in Mdr1a/b(−/−) Bcrp1(−/−) double knockout model, suggesting the critical role of these transporters in the blood-brain barrier (de Vries et al. 2007). Expression analysis of BCRP mRNA has revealed that its levels decrease along the gastrointestinal tract. The highest expression levels of BCRP mRNA were observed in the duodenum, while levels continuously decreased towards the rectum (Gutmann et al. 2005). The intestinal efflux of quercetin glucuronide, a flavonoid known substrate of BCRP, is impaired in Mrp2-deficient rats treated with BCRP inhibitors. This suggests that BCRP plays a role in facilitating the intestinal uptake of quercetin, potentially by transporting them across the intestinal epithelium. Interestingly, this impairment in quercetin glucuronide efflux was not observed in Mrp2-deficient rats that were not treated with BCRP inhibitors. This suggests that BCRP may compensate for the loss of Mrp2 function in the transport of quercetin glucuronide (Sesink et al. 2005). BCRP is also found to be localized in the apical membrane of human kidney proximal tubules, where it is expressed alongside other ABC transporters, contributing to the excretion of drugs and xenobiotics in the urine (Huls et al. 2008). BCRP has also been identified as a marker of hematopoietic stem cells, due to its high expression in primitive murine stem cells and sharp decrease with differentiation (Zhou et al. 2001).
The human ABCG2 gene is located on chromosome 4, specifically in the 4q21-q22 region. It spans over 66 kb and consists of 16 exons and 15 introns (Bailey-Dell et al. 2001). Transcriptional regulation studies of ABCG2 have identified several functional cis-elements within its promoter region that are recognized by several transcription factors, including estrogen (Ee et al. 2004), progesterone (Wang et al. 2008), peroxisome proliferator-activated receptor γ (Szatmari et al. 2006), hypoxia-inducible factor (HIF)-1α (Krishnamurthy et al. 2004), HIF-2α (He et al. 2019), and NRF2 (Singh et al. 2010) (Fig. 1). At the post-transcriptional modification level, alternative splicing of the ABCG2 gene has been reported to generate three splice variants of the untranslated exon 1 in the 5’ untranslated region (UTR) of the mRNA. These splice variants can serve as alternative promoters and result in the expression of isoforms of the protein with altered function and localization (Nakanishi et al. 2006). In addition, microRNAs (miRNAs) have been suggested to be involved in BCRP regulation. Rapid amplification of cDNA ends (RACE) analysis of the 3’ UTR of ABCG2 mRNA from drug-resistant S1 colon cancer cells identified several putative miRNA binding sites (To et al. 2008). MiR-328 has been reported to directly regulate ABCG2 in MCF-7 breast cancer cells and re-sensitize the cells to mitoxantrone (Pan et al. 2009). Additionally, expression of miR-520 h inhibited migration, invasion, and side populations of pancreatic cancer cells (Wang et al. 2010).
Figure 1.BCRP regulation and its association with chemoresistance. Overactivation of NRF2 in cancer increases BCRP (ABCG2) as one of its targets, as well as other targets in antioxidant defense, metabolism, and drug detoxification, thereby promoting cell survival and resistance to therapy. The transcriptional regulation of BCRP can be controlled by additional transcription factors, including HIF-1α, HIF-2α, PPARγ, and estrogen, which can bind to their enhancer elements in the promoter sequence of BCRP genes. At the post-transcriptional level, miR-328 and miR-520h can directly suppress BCRP transcription, while NRF2-regulated miR-206 targets BCRP via the c-MET/EGFR signaling pathway. sMAFs, small MAF proteins; BCRP, breast cancer resistance protein; PPARγ, peroxisome proliferator-activated receptor-γ; HIF, hypoxia-inducible factor; ARE, antioxidant response element; PPRE, peroxisome proliferator-response element; HRE, hypoxia response element; ERE, estrogen response element; EGFR, epidermal growth factor receptor; c-MET, mesenchymal-epithelial transition factor.
ABC transporters work together to mediate the efflux of various substrates, including anticancer agents such as anthracyclines, camptothecins, taxanes, and tyrosine kinase inhibitors (Robey et al. 2018). It is also well-established that BCRP plays a major role in multidrug resistance of cancer because of its distinct action. While MRP1 selectively transports products of phase II drug metabolism, BCRP transports anticancer drugs, such as irinotecan, methotrexate, mitoxantrone, and topotecan (Mao and Unadkat 2015). Histology data indicates that BCRP is uniformly expressed in all subsets of tumors, with frequent expression observed in adenocarcinomas of the gastrointestinal tract, endometrium, lung, and melanoma (Diestra et al. 2002). The Cancer Genome Atlas (TCGA) database analysis of RNA sequencing data showed that hepatocellular carcinoma and kidney cancer, both of which are frequently resistant to vincristine and doxorubicin, expressed the highest levels of BCRP compared to other cancer types (Robey et al. 2018).
In vitro studies using cancer cell lines from various solid tumors, including breast, colon, gastric, glioma, lung, and ovarian cancer, have identified high expression of BCRP in multidrug resistance models (Maliepaard et al. 1999; Ross et al. 1999; Martin et al. 2009). An earlier study using drug-selected cancer cell lines showed that mitoxantrone, doxorubicin, and verapamil increased the expression of BCRP mRNA in human breast carcinoma (MCF-7), colon carcinoma (HT29), gastric carcinoma (EPG85-257), fibrosarcoma (EPF86-079), and myeloma (8226) (Ross et al. 1999). The human IGROV1 ovarian cancer cell line, which has developed resistance to the chemotherapy drugs topotecan and mitoxantrone following continuous exposure, expresses high levels of BCRP (Maliepaard et al. 1999). In a panel of glioma cell lines and brain tumor stem cells, high expression levels of BCRP were linked to the overexpression of the Tie2 receptor, which is associated with tumor malignancy, and BCRP downregulation decreased Tie2 expression (Martin et al. 2009). In SN-38-resistant PC-6 human small-lung cancer cell lines, which had acquired resistance to the active metabolites of irinotecan and topotecan, there was a reduction in the intracellular accumulation of SN-38. This reduction was found to be associated with the overexpression of BCRP, and suppression of BCRP enhanced the sensitivity of these cells to SN-38 (Kawabata et al. 2001). A pharmacogenetic study with 299 renal carcinoma patients in the Japanese population found that differences in the BCRP genotype (the ABCG2 C421A [Q141K] polymorphism) were associated with reduced adverse effects of sunitinib, without affecting its efficacy (Low et al. 2016).
There has been consistent clinical evidence for the relationship between BCRP overexpression and drug resistance in hematologic malignancies. A gene profiling study of older acute myeloid leukemia (AML) patients found that 77% of 170 AML samples were chemoresistant and characterized by high expressions of BCRP and MDR1 (Wilson et al. 2006). A comparison between sensitive and resistant AML patients showed a significant difference in BCRP expression, and that this difference was linked to overall patients’ survival (Marzac et al. 2011). Moreover, high levels of BCRP were identified in a population of putative leukemic stem cells (CD34+CD38−) from AML patients. BCRP inhibition in these leukemic stem cells using Ko143 resulted in increased cellular levels of mitoxantrone accumulation (Raaijmakers et al. 2005). Thus, these preclinical and clinical datasets suggest that the inhibition of BCRP in multidrug-resistant AML may potentially lead to better treatment outcomes and patient survival.
Given the critical role of BCRP in the acquisition of chemoresistance, extensive efforts have been dedicated to discovering and developing BCRP inhibitors (Doyle and Ross 2003). The first BCRP inhibitor to be reported was Fumitremorgin C (FTC), a secondary metabolite from Aspergillus fumigatus. FTC has been shown to effectively reverse the multidrug resistance of colon carcinoma to mitoxantrone, doxorubicin, bisantrene, and topotecan through the specific inhibition of BCRP (Rabindran et al. 1998). Subsequent studies have confirmed the inhibitory effect of FTC on BCRP in various cancer cell lines, including SF295 human glioblastoma, KM12 colon cancer, and A549 non-small-cell lung cancer cell lines (Robey et al. 2001a). FTC has also been found to inhibit the transport of methotrexate in human embryonic kidney cells (Chen et al. 2003). At a concentration of 5 µM, FTC has been shown to effectively re-sensitize the multi-drug resistant MCF-7 FLV1000 cell line to mitoxantrone, SN-38, and topotecan by inhibiting the efflux of these drugs via BCRP (Robey et al. 2001b). Later, a tetracycline analog of FTC called Ko143 was identified as a more selective inhibitor of BCRP. In an experiment using mouse intestine, Ko143 selectively inhibited BCRP without affecting the function of other ABC transporters such as P-gp or MRP-1 at a concentration of 100 nM (Allen et al. 2002).
Elacridar (GF120918), which was originally developed as a P-gp inhibitor, is an acridone carboxamide and identified as a potent inhibitor of BCRP. Elacridar has been shown to effectively inhibit the efflux of mitoxantrone and topotecan (de Bruin et al. 1999). Additionally, flavonoid metabolites 6-prenylchrysin and tectochrysin have a specific inhibitory effect on BCRP. In a study conducted on BCRP-expressing cancer cells, the application of 0.5 µmol/L of 6-prenylchrysin or 1 µmol/L of tectochrysin was sufficient to sensitize the cells to mitoxantrone, with less toxicity compared to elacridar (Ahmed-Belkacem et al. 2005).
Several anti-retroviral drugs, including lopinavir, nelfinavir, and delavirdine have been shown to inhibit BCRP and increase the cellular accumulation of pheophorbide A (Pha), a photosensitizer substrate of BCRP, in canine kidney epithelial cell lines (Weiss et al. 2007). Additionally, statins such as atorvastatin and rosuvastatin were shown to inhibit BCRP in the biliary excretion system using Bcrp(−/−) mice, although some statins are substrates of BCRP (Hirano et al. 2005). Moreover, since estrogen can regulate BCRP expression and therefore, estrogen agonists and antagonists have been found to modulate BCRP activity. Diethylstilbestrol, an estrogen agonist, increased the accumulation of topotecan and reversed the chemoresistance of the BCRP-transduced K562 cell line. In contrast, tamoxifen, an estrogen antagonist, has been found to only enhance the uptake of topotecan (Sugimoto et al. 2003).
NRF2 is a cap‘n’collar (CNC) basic leucine zipper (bZIP) transcription factor that regulates cytoprotective genes and maintains cellular redox homeostasis. Although NRF2 is expressed in all cell types, its basal level is mainly low due to negative regulation by KEAP1-mediated proteasomal degradation (Yamamoto et al. 2018). The NRF2 protein is composed of six functional domains, known as NRF2-ECH Homology (Neh) 1-6, each of which plays a specific role in NRF2 activity. Neh1 primarily contains the CNC-bZIP domains that facilitate DNA binding and dimerization, while Neh3-5 serve as transactivation domains. In particular, Neh4 and Neh5 recruit the histone acetyl-transferase cAMP-responsive element binding protein (CBP) and mediator complex to facilitate the transcriptional activity of NRF2 (Yamamoto et al. 2018). Meanwhile, Neh2 and Neh6 primarily regulate NRF2 degradation. Neh2 contains DLG and ETGE motifs that serve as distinct and prominent binding sites for the KEAP1 protein, while Neh6 functions as a KEAP1-independent degron site of NRF2. Neh6 facilitates NRF2 degradation through the binding of β-TRCP (Tong et al. 2006; Chowdhry et al. 2013).
Under normal conditions, two KEAP1 molecules and one NRF2 molecule form a structure that accelerates the ubiquitination of NRF2. KEAP1 acts as an adaptor, forming cullin3 (CUL3)-based E3 ubiquitin ligases and RBX1 to create a functional E3 ubiquitin ligase (Kobayashi et al. 2004; McMahon et al. 2006). The KEAP1-CUL3 complex recognizes the DLG and ETGE motifs in the Neh2 domain of NRF2 and ubiquitinates it, ensuring a low level of NRF2 by subjecting it to proteasomal degradation. However, following exposure to electrophiles or reactive oxygen species (ROS), the cysteine residues of KEAP1 undergo modification, forming disulfide bonds or covalent linkages due to oxidation, which results in the alteration of the conformational structure of the KEAP1 (Sekhar et al. 2010). The disruption of the DLG-KEAP1 binding allows for the accumulation of newly synthesized NRF2, which can be translocated into the nucleus. Then, NRF2 can dimerize with small Maf proteins and activate transcription by recognizing and binding to the ARE in the promoter of its target genes.
The direct correlation between NRF2 and target genes through ARE was first reported by Itoh et al. (1997). They observed a marked decrease in the expression of phase II enzymes, including glutathione S-transferases (GSTs) and NAD(P)H quinone oxidoreductase 1 (NQO1), in butylated hydroxyanisole (BHA)-treated Nrf2-null mice. Since then, NRF2 has been identified as a master regulator of genes involved in phase I, II, and III metabolism. These include phase I enzymes such as Cyp2a5 and Cyp2b10 (Ashino et al. 2014), phase II enzymes such as GSTs (GSTA3, GSTA4, GSTM1, GSTM2, GSTM3, GSTM4, GSTM6), NQO1 (Hirotsu et al. 2012), and aldo-keto reductase family 1 member C1 (AKR1C1) (Jung et al. 2013), and phase III transporters, including MRPs (ABCC1, ABCC2, ABCC3, ABCC4) (Stöckel et al. 2000; Mahaffey et al. 2009; Xu et al. 2010; Ji et al. 2013) and BCRP (ABCG2) (Singh et al. 2010).
While NRF2 activity is crucial in normal unstressed conditions, persistent activation of NRF2 has been found to contribute to impairing the efficacy of chemotherapy in cancer (Rojo de la Vega et al. 2018). The hyperactivation of NRF2 in cancer can be resulted from somatic mutations or genomic alterations in NRF2 and KEAP1 genes or the expression of KEAP1-competing proteins such as p62/SQSTM1 (Kitamura and Motohashi 2018). In non-small-cell lung cancer patients and cell lines, genomic domain mutation and high percentage loss of heterozygosity at 19p13.2 of KEAP1 results in the specific abrogation of the NRF2 repressor system (Singh et al. 2006). Thus, NRF2 is accumulated and enhances the transcriptional activation of genes involved in xenobiotic detoxification and drug efflux pumps. Moreover, mutations in either the IVR or DGR region of KEAP1, or in the DLG or ETGE motifs of NRF2, have consequences for the constitutive expression of cytoprotective enzymes induced by NRF2 and can facilitate tumor malignancies (Ohta et al. 2008; Shibata et al. 2008). The genomic alteration of exon 2 deletions in NRF2 results in persistent NRF2 activation in lung cancers and head and neck carcinomas (Goldstein et al. 2016). KEAP1 gene methylation leads to reduced expression of KEAP1 and activation of NRF2 and its corresponding targets, such as NQO1 and AKR1C1, during the carcinogenesis and cancer progression (Hanada et al. 2012; Barbano et al. 2013). One well-established KEAP1-competing protein that has been identified as responsible for NRF2 accumulation in cancer cells is the autophagy adaptor protein p62/SQSTM1. Its phosphorylation at the S349 residue by the mammalian target of rapamycin complex 1 (mTORC1) increases the binding affinity to KEAP1, subsequently disrupting the DLG-mediated KEAP1-NRF2 interaction (Ichimura and Komatsu 2018). Pharmacological inhibition of phosphorylated p62/SQSTM1 and KEAP1 interaction by K67 resulted in the suppression of NRF2 activity, which inhibited growth and re-sensitized the tumors to anti-cancer agents (Saito et al. 2016).
The high expression of NRF2, which has been observed in multiple types of cancers, contributes to chemoresistance by inducing downstream genes involved in electrophiles/ROS detoxification and the efflux of cytotoxic anticancer drugs (Srivastava et al. 2022). A recent report demonstrated that the activation of NRF2 by FAM117B-mediated KEAP1 competitive binding in gastric cancer facilitated tumor growth and chemoresistance to 5-fluorouracil and oxaliplatin (Zhou et al. 2023). In pancreatic cancer, oncogenic KRAS upregulated NRF2-dependent metabolic rewiring of glutamine metabolism, which leads to gemcitabine resistance (Mukhopadhyay et al. 2020). Keap1-deficient KrasG12D mice model developed lung tumors with hyperactivation of NRF2 and subsequent activation of the pentose phosphate pathway (PPP) (Best et al. 2019). Similarly, in KrasG12D–induced lung cancer, constitutive KRAS activation increased NRF2 transcription and resulted in cisplatin resistance. In this context, treatment with the NRF2 inhibitor brusatol was found to enhance the anti-tumor efficacy of cisplatin in KrasG12D-induced lung cancer (Tao et al. 2014).
Several studies have shown a direct correlation between NRF2 and BCRP in cancer. An early report demonstrated that treatment with oltipraz, an activator of NRF2, increased the mRNA level of BCRP in cultured human hepatocytes (Jigorel et al. 2006). This observation was further confirmed by a study that used tert-butylhydroquinone (tBHQ), another potent activator of NRF2, in HepG2 cells (Adachi et al. 2007). In this study, treatment of HepG2 cells with tBHQ induced translocation of NRF2 to the nucleus, and concomitantly increased the mRNA level of BCRP, as well as other ABC transporters and NRF2 targets, such as the heavy and light chains of γ-glutamylcysteine synthase. The decrease and increase in expression of BCRP mRNA mediated by NRF2- and KEAP1-silencing, respectively, suggest the involvement of the NRF2/KEAP1 system in the transcriptional regulation of BCRP (Adachi et al. 2007).
An advanced study by Singh et al. identified the ARE region located in the 5’-flanking region of the BCRP gene at -431 to -420 bp, which is responsible for NRF2-mediated expression in lung cancer cells in vitro and in vivo (Singh et al. 2010) (Fig. 1). This observation confirms previous findings of ARE located in the promoter regions of other multidrug ABC transporters, such as MRP1-3 (Stöckel et al. 2000; Kurz et al. 2001; Mahaffey et al. 2009). Similarly, in normal kidney tubular cells, knockdown of KEAP1 was found to increase BCRP expression. However, unlike in the previous studies where NRF2 activation increased BCRP expression, pharmacological activation of NRF2 did not elevate BCRP mRNA in this case (Jeong et al. 2015). In addition to transcriptional regulation, NRF2 has been reported to regulate BCRP through a post-transcriptional mechanism as well. For example, in ovarian and renal carcinoma cell lines, silencing of NRF2 led to the activation of miR-206, which targeted c-MET/EGFR and downregulated BCRP expression, resulting in increased intracellular accumulation of the anticancer drug doxorubicin (Singh et al. 2013; Choi et al. 2017).
PDT is an emerging chemotherapy strategy that utilizes photosensitizing drugs to selectively kill cancer cells by generating cytotoxic ROS upon light activation (Agostinis et al. 2011). PDT has gained increasing attention as a minimally invasive and more target-specific alternative to conventional chemotherapy (Dolmans et al. 2003; Correia et al. 2021). PDT has been approved by the FDA for the treatment of several types of cancer, including bladder, esophageal, head and neck, lung, and skin cancer (Correia et al. 2021). The expression of BCRP in cancer cells can limit the effectiveness of PDT, and therefore there has been interest in investigating the relationship between BCRP and PDT. One study used imatinib mesylate, a tyrosine kinase inhibitor (TKI) known to reduce BCRP levels, and found that it enhanced the accumulation of various photosensitizers, including 2-(1-hexyloxethyl)-2-devinyl pyropheophorbide-a (Photochlor), protoporphyrin IX, and the benzoporphyrin derivative monoacid ring A (Verteporfin), in ABCG+ human and murine cell lines both in vitro and in vivo (Liu et al. 2007). Another study showed that BCRP-silencing using siRNA or inhibition of its activity using a chemical inhibitor, resulted in increased intracellular accumulation of chlorin e6 and improved PDT efficacy in cancer cells. This suggests that targeting BCRP may be a promising strategy to enhance the efficacy of PDT in cancer therapy (Roh et al. 2017). In the doxorubicin-resistant ovarian cancer A2780DR cells, the increase of BCRP-mediated drug resistance is associated with c-MET elevation. Treating the A2780DR cells with BCRP inhibitor Ko143 or c-MET inhibitor resulted in the accumulation of the photosensitizer Pba within cancer cells, which in turn sensitized the cells to PDT (Jung et al. 2015). Another study showed that A431/BCRP lung cancer cells, which overexpress BCRP, are more resistant to Photofrin-mediated PDT. However, inhibition of BCRP with FTC reversed this resistance and increased the efficacy of PDT (Usuda et al. 2010). In line with this, NRF2-knockdown in several types of cancer cells, including triple-negative breast cancer, renal carcinoma, and colon cancer, led to an increase in the effectiveness of PDT by lowering the levels of BCRP (Choi et al. 2014). Furthermore, a recent report has demonstrated that nanoparticle-mediated NRF2 silencing can increase the sensitivity of tumor cells to oxidative stress during PDT. These findings further support the importance of the NRF2/BCRP pathway in the response of cancer cells to PDT (Sun et al. 2023).
Ever since the initial discovery of a subpopulation of leukemic stem cells characterized by the surface markers CD34+ and CD38−, CSCs have been recognized as promising targets for cancer therapy (Lapidot et al. 1994). The unique characteristics of CSCs, including their ability of self-renewal and differentiation potential, have been shown to play a critical role in cancer malignancies such as metastasis, therapy resistance, and recurrence (Batlle and Clevers 2017). CSCs and normal stem cells share many common transcription factors and signaling pathways, including several pluripotent transcription factors such as octamer-binding transcription factor 4 (OCT4), krupple-like factor 4 (KLF4), MYC, homeobox protein Nanog (NANOG), and SRY (sex determining region Y)-box 2 (SOX2) (Yang et al. 2020). Furthermore, a subpopulation of cancer cells known as the side population has been identified as a chemo- and radio-therapy-resistant population, which is thought to be due to its high expression of BCRP. As a result, BCRP has become recognized as one of the CSC markers (Zhou et al. 2001).
Recent studies have suggested that NRF2 may contribute to the development or maintenance of CSC properties, including chemoresistance (Kahroba et al. 2019; Choi et al. 2021; Hallis et al. 2023a). In the colonosphere model of in vitro CSCs, increased levels of P-gp, BCRP, and concomitant NRF2 signaling were observed, which exhibited resistance to doxorubicin (Ryoo et al. 2016). Subsequently, NRF2 knockdown in the colonosphere model sensitized these cells to drug treatment by decreasing efflux transporter activities. Consistent with these findings, the CD133high population from HCT116 colon cancer cells exhibited high levels of NRF2 and BCRP, along with elevated CSC markers, indicating that they possess CSC phenotypes such as enhanced proliferation, migration, colony formation, chemoresistance, and sphere-forming capacity (Park et al. 2022). In this experimental setting, colonospheres derived from NRF2-silenced HCT116 expressed low levels of BCRP along with reduced KLF4 expression, which resulted in the suppression of sphere-forming capacity. A similar study involving NRF2 knockdown in pancreatic ductal adenocarcinoma cell lines BxPC-3 and CFPAC-1, which had high expression of NRF2, found that it decreased the expression of CSC transcription factors NANOG, OCT4, and CD133, as well as chemoresistance properties, by downregulating BCRP (Kim et al. 2020). Silencing of NRF2 in sorafenib-resistant Huh7 hepatocellular cell line also limited cancer stemness and migration, as well as expression of BCRP (Gao et al. 2021). In another drug resistance model, A2780DR ovarian cancer cells that exhibit doxorubicin resistance showed higher NRF2 signaling activity concomitantly with elevated expression of aldehyde dehydrogenase 1 (ALDH1) and BCRP, which supports to maintain CSC properties (Shim et al. 2009; Kim et al. 2018). Isolated CSCs from the side population of cervical cancer patients showed aberrant activation of NRF2, BCRP, and Bcl-2, which resulted in chemoresistance and apoptosis resistance (Jia et al. 2015). Meanwhile, our recent study demonstrated a positive association between NRF2 and HIF-2α-induced CSC phenotypes under chronic hypoxic conditions (Hallis et al. 2023b). As previous studies have reported that HIF-2α directly regulates BCRP and confers CSC properties in ovarian CSCs, leading to resistance to doxorubicin and promoting tumor growth, this may suggest the potential association of NRF2 with HIF-2α/BCRP in hypoxic tumor environment (He et al. 2019). Collectively, these studies indicate that NRF2 and BCRP are intricately intertwined in maintaining resistance to chemotherapeutic agents and photodynamic therapy, and the maintenance of CSC properties in cancer (Fig. 2).
Figure 2.Role of NRF2 and BCRP in CSC maintenance and PDT resistance. (A) NRF2 signaling is activated in CSCs models, contributing to CSC properties, including tumor initiation and growth, invasion, metastasis, and therapy resistance. NRF2 upregulates the antioxidants gene to maintain low ROS levels and increases ABC transporters such as BCRP to protect CSCs against external harmful chemicals, which eventually leads to therapy resistance. High levels of NRF2 are also associated with the upregulation of other transcription factors, including OCT4, NANOG, KLF4, MYC, and SOX2, which help to maintain CSC phenotypes. (B) Enhanced BCRP in cancers facilitates the efflux of photosensitizers such as Pba, photochlor, and photofirin, thus leading to PDT resistance. BCRP inhibition by chemical inhibitors increases intracellular accumulation of photosensitizers and consequent ROS generation, leading to cancer cell apoptosis. OCT4, octamer-binding transcription factor 4; KLF4, krupple-like factor 4; NANOG, homeobox protein nanog; SOX2, SRY (sex determining region Y)-box 2; FTC, fumitremorgin C.
In this review, we have highlighted the critical role of NRF2 and BCRP as potential targets to counteract the therapy resistance of cancers. Elevated expression of BCRP in most cancer cases results in a poor prognosis and is associated with multidrug resistance, thereby impeding the efficacy of cancer therapy (Robey et al. 2018). Many studies have demonstrated the potential effect of BCRP inhibitors in multiple models of resistant cancers, and current efforts are focusing on investigating BCRP inhibitors for clinical trials. Thus, the regulation of BCRP by transcription factors or other signaling pathways is of great interest. Although several transcription factors have been implicated, the identification of ARE sequences in the promoter region of BCRP has highlighted the direct regulation by NRF2 (Singh et al. 2010). This is important since NRF2 overexpression in cancer clinical settings is correlated with poor prognosis and chemoresistance (Solis et al. 2010). Moreover, recent insights suggest that the suppression of BCRP and its regulator NRF2 may have an impact on PDT resistance and the maintenance of CSC properties. Thus, ongoing efforts are focused on strategies aimed at suppressing NRF2 activity to increase the sensitivity of cancer patients to chemotherapy (Dinkova-Kostova and Copple 2023). In particular, compared to the inhibition of BCRP, the inhibition of NRF2 is expected to be a more effective strategy for cancer treatment in that it can suppress the expression of other target genes that contribute to cancer growth and development as well as control of BCRP-mediated anticancer drug resistance. Currently, extensive efforts are being made to develop NRF2 inhibitors. Several natural compounds, including luteolin, procyanidins, and wedelolactone, have demonstrated NRF2 inhibitory effects in various experimental settings involving NRF2-overactive cancers (Sharifi-Rad et al. 2023). Brusatol from Brucea javanica has shown potent anti-tumor and anti-metastatic activities by decreasing NRF2 protein levels, which is attributed to multiple mechanisms, including protein translation inhibition (He et al. 2023). In a high-throughput chemical library screening, halofuginone has been identified as an inhibitor of NRF2 protein synthesis (Tsuchida et al. 2017). ML385, derived from a high-throughput chemical library screening and subsequent chemical optimization, has been discovered to interfere with the interaction between NRF2 and sMAF by directly binding to the NRF2 protein. Consequently, ML385 demonstrated the ability to sensitize KEAP1 mutant lung cancer cells to chemotherapy (Singh et al. 2016). While NRF2 inhibitors hold strong potential for enhancing chemotherapy sensitivity and suppressing tumorigenesis in NRF2-high cancers, the dual functionality of NRF2 in normal cells presents a limitation for its clinical application. In this context, achieving cancer-selective inhibition of NRF2 becomes a crucial strategy. In a recent study employing a synthetic lethal strategy, mitomycin C has emerged as a selectively cytotoxic agent for NRF2-overactive cancer cells, primarily through enhanced bioactivation via NRF2-target genes, such as NQO1 (Baird and Yamamoto 2021). Geldanamycin-derived HSP90 inhibitors have demonstrated selective lethality towards NRF2-high cancer cells, and their mechanism of action has been attributed to the enhancement of NQO1-mediated bioactivation (Baird et al. 2020). In summary, although a strategy for NRF2 inhibitors that selectively act on NRF2-overexpressing cancers has not been developed, with the progression of more advanced studies on the NRF2-KEAP1 system, there is an optimistic expectation for the future discovery of highly selective and effective NRF2/BCRP inhibitors.
This work was supported by National Research Foundation of Korea (NRF) grants funded by the Korean government (MSIT) (2022R1A2C2011866).
The authors declare that they have no conflict of interest.
DTT 2023; 2(2): 111-123
Published online September 30, 2023 https://doi.org/10.58502/DTT.23.0021
Copyright © The Pharmaceutical Society of Korea.
Steffanus Pranoto Hallis1 , Beam Ju Go2, Jun Min Yoo2, Geun Hee Cho2, Mi-Kyoung Kwak1,2
1Department of Pharmacy, Graduate School of The Catholic University of Korea, Bucheon, Korea
2College of Pharmacy, The Catholic University of Korea, Bucheon, Korea
Correspondence to:Mi-Kyoung Kwak, mkwak@catholic.ac.kr
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
The development of therapy resistance in cancer is a complex mechanism with multifaceted and catastrophic consequences. BCRP (Breast Cancer Resistance Protein/ATP-Binding Cassette Subfamily G member 2; ABCG2), a drug efflux transporter, is recognized as a key regulator in obtaining multidrug resistance properties of cancer cells. NRF2 (Nuclear Factor Erythroid 2-Like 2; NFE2L2) plays an essential role in maintaining cellular redox homeostasis by regulating the expression of an array of antioxidant and detoxifying genes. In contrast to its protective roles in normal cells, a substantial body of evidence has demonstrated the negative aspects of NRF2 accumulation in cancer cells, including facilitated tumor growth, stemness, and therapy resistance. This review focuses on the linkage between BCRP and NRF2 in cancer and its consequences in acquiring therapy resistance. In particular, we highlight the role of the NRF2/BCRP axis in photodynamic therapy resistance and cancer stem cell maintenance, demonstrating that inhibition of this axis can be a beneficial strategy to overcome resistance to cancer therapy. A comprehensive understanding of the NRF2/BCRP regulating pathway will lead to the development of new therapeutic approaches and improve the prognosis of cancer patients.
Keywords: BCRP/ABCG2, NRF2/NFE2L2, multidrug resistance, chemoresistance, photodynamic therapy, cancer stem cells
The development of resistance in cancer cells remains a major obstacle to successful chemotherapy, contributing to treatment failure and disease progression. One key mechanism of chemoresistance is the overexpression of ATP-binding cassette (ABC) transporters, which are responsible for the efflux of cytotoxic anticancer agents from cancer cells (Robey et al. 2018). Among 48 human ABC transporters, which are classified into seven subfamilies, multidrug resistance protein 1 (MDR1 or P-glycoprotein/P-gp) encoded by ABC subfamily B member 1 (ABCB1) was initially described in Chinese hamster ovary (CHO) cells and later in sublines of human KB carcinoma cells bearing multidrug-resistance to either colchicine, vinblastine, or doxorubicin (Roninson et al. 1986). The multidrug resistance-associated protein 1 (MRP1, encoded by ABCC1) was identified in 1992 as the second member of multidrug ABC transporters from lung cancer cell line H69AR, a cell line model of doxorubicin, etoposide, and vincristine resistance (Cole et al. 1992). The third member of the multidrug ABC transporter was encoded by ABCG2 and concurrently reported by three groups. The first group identified ABCG2 in human MCF7 breast cancer cells, where it was named breast cancer resistance protein (BCRP) (Doyle et al. 1998). Additionally, ABCG2 was found in mitoxantrone-resistant S1-M1-80 human colon carcinoma cells (Miyake et al. 1999), as well as in placental tissue (Allikmets et al. 1998). The overexpression of multidrug ABC transporters in cancer has been linked to chemotherapy failure in clinical settings. As a result, numerous efforts have been made to overcome multidrug-resistant cancer through the development of transporter inhibitors (Gottesman et al. 2016).
Nuclear factor erythroid 2-like 2 (NFE2L2/NRF2) is a transcription factor that regulates the expression of gene arrays involved in protection against oxidative and electrophilic stress and maintenance of cellular homeostasis (Yamamoto et al. 2018). Despite its protective roles in normal tissues, constitutive activation of NRF2 confers cancer cells survival by multiple mechanisms: increasing drug elimination through phase II metabolizing enzymes, preventing intracellular drug accumulation, and suppressing apoptotic cell death (Bai et al. 2016; Choi and Kwak 2016). In particular, the antioxidant response element (ARE) has been identified as a binding site for NRF2 in the promoter region of MRP1 and BCRP, confirming the critical role of NRF2 signaling in the regulation of multidrug-resistant ABC transporters (Kurz et al. 2001; Singh et al. 2010; Ji et al. 2013). The review article focuses on providing an overview of BCRP and NRF2, and their association with therapy resistance in cancer. In particular, we highlight the role of the NRF2/BCRP axis in cancer stem cells (CSCs) and photodynamic therapy (PDT) resistance. We also suggest that inhibiting this axis can be an effective strategy to overcome resistance to cancer therapy.
BCRP, which is also known as mitoxantrone resistance-associated protein (MXR) or placenta-specific ABC transporter (ABCP), is a type of ABC transporter that belongs to the G subfamily of ABC transporters, also known as ABCG2. BCRP is responsible for translocating its substrates across cell membranes, against a concentration gradient, and this process is achieved through the binding and hydrolysis of ATP (Dean et al. 2001). BCRP is a 72-kDa protein consisting of 665 amino acids with six transmembrane segments and one extracellular loop. Unlike other subfamilies of ABC transporters, which contain two transmembrane domains (TMD) and two nucleotide (ATP)-binding domain (NBD), BCRP is a “half-transporter” with only one TMD and one NBD fused to a single polypeptide chain (Ozvegy et al. 2001; Taylor et al. 2017). Therefore, for BCRP to function properly, it needs to homodimerize, possibly through the formation of intra- and intermolecular disulfide bonds, or oligomerize (Bhatia et al. 2005; Henriksen et al. 2005).
BCRP has diverse physiological roles and is localized in various tissues and organs, including the placenta, mammary glands, gastrointestinal tract, kidney, blood-brain barrier, and hematopoietic stem cells (Robey et al. 2009). BCRP expression in the placenta is essential for protecting the fetus from topotecan, which has been administered in maternal animals. Treatment of pregnant mice with GF120918, a BCRP inhibitor, increased the relative fetal penetration of topotecan (Jonker et al. 2000). In Bcrp1(−/−) knockout mouse model, BCRP absence increased topotecan penetration to the brain by 1.5 folds, while the penetration increased by 12 folds in Mdr1a/b(−/−) Bcrp1(−/−) double knockout model, suggesting the critical role of these transporters in the blood-brain barrier (de Vries et al. 2007). Expression analysis of BCRP mRNA has revealed that its levels decrease along the gastrointestinal tract. The highest expression levels of BCRP mRNA were observed in the duodenum, while levels continuously decreased towards the rectum (Gutmann et al. 2005). The intestinal efflux of quercetin glucuronide, a flavonoid known substrate of BCRP, is impaired in Mrp2-deficient rats treated with BCRP inhibitors. This suggests that BCRP plays a role in facilitating the intestinal uptake of quercetin, potentially by transporting them across the intestinal epithelium. Interestingly, this impairment in quercetin glucuronide efflux was not observed in Mrp2-deficient rats that were not treated with BCRP inhibitors. This suggests that BCRP may compensate for the loss of Mrp2 function in the transport of quercetin glucuronide (Sesink et al. 2005). BCRP is also found to be localized in the apical membrane of human kidney proximal tubules, where it is expressed alongside other ABC transporters, contributing to the excretion of drugs and xenobiotics in the urine (Huls et al. 2008). BCRP has also been identified as a marker of hematopoietic stem cells, due to its high expression in primitive murine stem cells and sharp decrease with differentiation (Zhou et al. 2001).
The human ABCG2 gene is located on chromosome 4, specifically in the 4q21-q22 region. It spans over 66 kb and consists of 16 exons and 15 introns (Bailey-Dell et al. 2001). Transcriptional regulation studies of ABCG2 have identified several functional cis-elements within its promoter region that are recognized by several transcription factors, including estrogen (Ee et al. 2004), progesterone (Wang et al. 2008), peroxisome proliferator-activated receptor γ (Szatmari et al. 2006), hypoxia-inducible factor (HIF)-1α (Krishnamurthy et al. 2004), HIF-2α (He et al. 2019), and NRF2 (Singh et al. 2010) (Fig. 1). At the post-transcriptional modification level, alternative splicing of the ABCG2 gene has been reported to generate three splice variants of the untranslated exon 1 in the 5’ untranslated region (UTR) of the mRNA. These splice variants can serve as alternative promoters and result in the expression of isoforms of the protein with altered function and localization (Nakanishi et al. 2006). In addition, microRNAs (miRNAs) have been suggested to be involved in BCRP regulation. Rapid amplification of cDNA ends (RACE) analysis of the 3’ UTR of ABCG2 mRNA from drug-resistant S1 colon cancer cells identified several putative miRNA binding sites (To et al. 2008). MiR-328 has been reported to directly regulate ABCG2 in MCF-7 breast cancer cells and re-sensitize the cells to mitoxantrone (Pan et al. 2009). Additionally, expression of miR-520 h inhibited migration, invasion, and side populations of pancreatic cancer cells (Wang et al. 2010).
Figure 1. BCRP regulation and its association with chemoresistance. Overactivation of NRF2 in cancer increases BCRP (ABCG2) as one of its targets, as well as other targets in antioxidant defense, metabolism, and drug detoxification, thereby promoting cell survival and resistance to therapy. The transcriptional regulation of BCRP can be controlled by additional transcription factors, including HIF-1α, HIF-2α, PPARγ, and estrogen, which can bind to their enhancer elements in the promoter sequence of BCRP genes. At the post-transcriptional level, miR-328 and miR-520h can directly suppress BCRP transcription, while NRF2-regulated miR-206 targets BCRP via the c-MET/EGFR signaling pathway. sMAFs, small MAF proteins; BCRP, breast cancer resistance protein; PPARγ, peroxisome proliferator-activated receptor-γ; HIF, hypoxia-inducible factor; ARE, antioxidant response element; PPRE, peroxisome proliferator-response element; HRE, hypoxia response element; ERE, estrogen response element; EGFR, epidermal growth factor receptor; c-MET, mesenchymal-epithelial transition factor.
ABC transporters work together to mediate the efflux of various substrates, including anticancer agents such as anthracyclines, camptothecins, taxanes, and tyrosine kinase inhibitors (Robey et al. 2018). It is also well-established that BCRP plays a major role in multidrug resistance of cancer because of its distinct action. While MRP1 selectively transports products of phase II drug metabolism, BCRP transports anticancer drugs, such as irinotecan, methotrexate, mitoxantrone, and topotecan (Mao and Unadkat 2015). Histology data indicates that BCRP is uniformly expressed in all subsets of tumors, with frequent expression observed in adenocarcinomas of the gastrointestinal tract, endometrium, lung, and melanoma (Diestra et al. 2002). The Cancer Genome Atlas (TCGA) database analysis of RNA sequencing data showed that hepatocellular carcinoma and kidney cancer, both of which are frequently resistant to vincristine and doxorubicin, expressed the highest levels of BCRP compared to other cancer types (Robey et al. 2018).
In vitro studies using cancer cell lines from various solid tumors, including breast, colon, gastric, glioma, lung, and ovarian cancer, have identified high expression of BCRP in multidrug resistance models (Maliepaard et al. 1999; Ross et al. 1999; Martin et al. 2009). An earlier study using drug-selected cancer cell lines showed that mitoxantrone, doxorubicin, and verapamil increased the expression of BCRP mRNA in human breast carcinoma (MCF-7), colon carcinoma (HT29), gastric carcinoma (EPG85-257), fibrosarcoma (EPF86-079), and myeloma (8226) (Ross et al. 1999). The human IGROV1 ovarian cancer cell line, which has developed resistance to the chemotherapy drugs topotecan and mitoxantrone following continuous exposure, expresses high levels of BCRP (Maliepaard et al. 1999). In a panel of glioma cell lines and brain tumor stem cells, high expression levels of BCRP were linked to the overexpression of the Tie2 receptor, which is associated with tumor malignancy, and BCRP downregulation decreased Tie2 expression (Martin et al. 2009). In SN-38-resistant PC-6 human small-lung cancer cell lines, which had acquired resistance to the active metabolites of irinotecan and topotecan, there was a reduction in the intracellular accumulation of SN-38. This reduction was found to be associated with the overexpression of BCRP, and suppression of BCRP enhanced the sensitivity of these cells to SN-38 (Kawabata et al. 2001). A pharmacogenetic study with 299 renal carcinoma patients in the Japanese population found that differences in the BCRP genotype (the ABCG2 C421A [Q141K] polymorphism) were associated with reduced adverse effects of sunitinib, without affecting its efficacy (Low et al. 2016).
There has been consistent clinical evidence for the relationship between BCRP overexpression and drug resistance in hematologic malignancies. A gene profiling study of older acute myeloid leukemia (AML) patients found that 77% of 170 AML samples were chemoresistant and characterized by high expressions of BCRP and MDR1 (Wilson et al. 2006). A comparison between sensitive and resistant AML patients showed a significant difference in BCRP expression, and that this difference was linked to overall patients’ survival (Marzac et al. 2011). Moreover, high levels of BCRP were identified in a population of putative leukemic stem cells (CD34+CD38−) from AML patients. BCRP inhibition in these leukemic stem cells using Ko143 resulted in increased cellular levels of mitoxantrone accumulation (Raaijmakers et al. 2005). Thus, these preclinical and clinical datasets suggest that the inhibition of BCRP in multidrug-resistant AML may potentially lead to better treatment outcomes and patient survival.
Given the critical role of BCRP in the acquisition of chemoresistance, extensive efforts have been dedicated to discovering and developing BCRP inhibitors (Doyle and Ross 2003). The first BCRP inhibitor to be reported was Fumitremorgin C (FTC), a secondary metabolite from Aspergillus fumigatus. FTC has been shown to effectively reverse the multidrug resistance of colon carcinoma to mitoxantrone, doxorubicin, bisantrene, and topotecan through the specific inhibition of BCRP (Rabindran et al. 1998). Subsequent studies have confirmed the inhibitory effect of FTC on BCRP in various cancer cell lines, including SF295 human glioblastoma, KM12 colon cancer, and A549 non-small-cell lung cancer cell lines (Robey et al. 2001a). FTC has also been found to inhibit the transport of methotrexate in human embryonic kidney cells (Chen et al. 2003). At a concentration of 5 µM, FTC has been shown to effectively re-sensitize the multi-drug resistant MCF-7 FLV1000 cell line to mitoxantrone, SN-38, and topotecan by inhibiting the efflux of these drugs via BCRP (Robey et al. 2001b). Later, a tetracycline analog of FTC called Ko143 was identified as a more selective inhibitor of BCRP. In an experiment using mouse intestine, Ko143 selectively inhibited BCRP without affecting the function of other ABC transporters such as P-gp or MRP-1 at a concentration of 100 nM (Allen et al. 2002).
Elacridar (GF120918), which was originally developed as a P-gp inhibitor, is an acridone carboxamide and identified as a potent inhibitor of BCRP. Elacridar has been shown to effectively inhibit the efflux of mitoxantrone and topotecan (de Bruin et al. 1999). Additionally, flavonoid metabolites 6-prenylchrysin and tectochrysin have a specific inhibitory effect on BCRP. In a study conducted on BCRP-expressing cancer cells, the application of 0.5 µmol/L of 6-prenylchrysin or 1 µmol/L of tectochrysin was sufficient to sensitize the cells to mitoxantrone, with less toxicity compared to elacridar (Ahmed-Belkacem et al. 2005).
Several anti-retroviral drugs, including lopinavir, nelfinavir, and delavirdine have been shown to inhibit BCRP and increase the cellular accumulation of pheophorbide A (Pha), a photosensitizer substrate of BCRP, in canine kidney epithelial cell lines (Weiss et al. 2007). Additionally, statins such as atorvastatin and rosuvastatin were shown to inhibit BCRP in the biliary excretion system using Bcrp(−/−) mice, although some statins are substrates of BCRP (Hirano et al. 2005). Moreover, since estrogen can regulate BCRP expression and therefore, estrogen agonists and antagonists have been found to modulate BCRP activity. Diethylstilbestrol, an estrogen agonist, increased the accumulation of topotecan and reversed the chemoresistance of the BCRP-transduced K562 cell line. In contrast, tamoxifen, an estrogen antagonist, has been found to only enhance the uptake of topotecan (Sugimoto et al. 2003).
NRF2 is a cap‘n’collar (CNC) basic leucine zipper (bZIP) transcription factor that regulates cytoprotective genes and maintains cellular redox homeostasis. Although NRF2 is expressed in all cell types, its basal level is mainly low due to negative regulation by KEAP1-mediated proteasomal degradation (Yamamoto et al. 2018). The NRF2 protein is composed of six functional domains, known as NRF2-ECH Homology (Neh) 1-6, each of which plays a specific role in NRF2 activity. Neh1 primarily contains the CNC-bZIP domains that facilitate DNA binding and dimerization, while Neh3-5 serve as transactivation domains. In particular, Neh4 and Neh5 recruit the histone acetyl-transferase cAMP-responsive element binding protein (CBP) and mediator complex to facilitate the transcriptional activity of NRF2 (Yamamoto et al. 2018). Meanwhile, Neh2 and Neh6 primarily regulate NRF2 degradation. Neh2 contains DLG and ETGE motifs that serve as distinct and prominent binding sites for the KEAP1 protein, while Neh6 functions as a KEAP1-independent degron site of NRF2. Neh6 facilitates NRF2 degradation through the binding of β-TRCP (Tong et al. 2006; Chowdhry et al. 2013).
Under normal conditions, two KEAP1 molecules and one NRF2 molecule form a structure that accelerates the ubiquitination of NRF2. KEAP1 acts as an adaptor, forming cullin3 (CUL3)-based E3 ubiquitin ligases and RBX1 to create a functional E3 ubiquitin ligase (Kobayashi et al. 2004; McMahon et al. 2006). The KEAP1-CUL3 complex recognizes the DLG and ETGE motifs in the Neh2 domain of NRF2 and ubiquitinates it, ensuring a low level of NRF2 by subjecting it to proteasomal degradation. However, following exposure to electrophiles or reactive oxygen species (ROS), the cysteine residues of KEAP1 undergo modification, forming disulfide bonds or covalent linkages due to oxidation, which results in the alteration of the conformational structure of the KEAP1 (Sekhar et al. 2010). The disruption of the DLG-KEAP1 binding allows for the accumulation of newly synthesized NRF2, which can be translocated into the nucleus. Then, NRF2 can dimerize with small Maf proteins and activate transcription by recognizing and binding to the ARE in the promoter of its target genes.
The direct correlation between NRF2 and target genes through ARE was first reported by Itoh et al. (1997). They observed a marked decrease in the expression of phase II enzymes, including glutathione S-transferases (GSTs) and NAD(P)H quinone oxidoreductase 1 (NQO1), in butylated hydroxyanisole (BHA)-treated Nrf2-null mice. Since then, NRF2 has been identified as a master regulator of genes involved in phase I, II, and III metabolism. These include phase I enzymes such as Cyp2a5 and Cyp2b10 (Ashino et al. 2014), phase II enzymes such as GSTs (GSTA3, GSTA4, GSTM1, GSTM2, GSTM3, GSTM4, GSTM6), NQO1 (Hirotsu et al. 2012), and aldo-keto reductase family 1 member C1 (AKR1C1) (Jung et al. 2013), and phase III transporters, including MRPs (ABCC1, ABCC2, ABCC3, ABCC4) (Stöckel et al. 2000; Mahaffey et al. 2009; Xu et al. 2010; Ji et al. 2013) and BCRP (ABCG2) (Singh et al. 2010).
While NRF2 activity is crucial in normal unstressed conditions, persistent activation of NRF2 has been found to contribute to impairing the efficacy of chemotherapy in cancer (Rojo de la Vega et al. 2018). The hyperactivation of NRF2 in cancer can be resulted from somatic mutations or genomic alterations in NRF2 and KEAP1 genes or the expression of KEAP1-competing proteins such as p62/SQSTM1 (Kitamura and Motohashi 2018). In non-small-cell lung cancer patients and cell lines, genomic domain mutation and high percentage loss of heterozygosity at 19p13.2 of KEAP1 results in the specific abrogation of the NRF2 repressor system (Singh et al. 2006). Thus, NRF2 is accumulated and enhances the transcriptional activation of genes involved in xenobiotic detoxification and drug efflux pumps. Moreover, mutations in either the IVR or DGR region of KEAP1, or in the DLG or ETGE motifs of NRF2, have consequences for the constitutive expression of cytoprotective enzymes induced by NRF2 and can facilitate tumor malignancies (Ohta et al. 2008; Shibata et al. 2008). The genomic alteration of exon 2 deletions in NRF2 results in persistent NRF2 activation in lung cancers and head and neck carcinomas (Goldstein et al. 2016). KEAP1 gene methylation leads to reduced expression of KEAP1 and activation of NRF2 and its corresponding targets, such as NQO1 and AKR1C1, during the carcinogenesis and cancer progression (Hanada et al. 2012; Barbano et al. 2013). One well-established KEAP1-competing protein that has been identified as responsible for NRF2 accumulation in cancer cells is the autophagy adaptor protein p62/SQSTM1. Its phosphorylation at the S349 residue by the mammalian target of rapamycin complex 1 (mTORC1) increases the binding affinity to KEAP1, subsequently disrupting the DLG-mediated KEAP1-NRF2 interaction (Ichimura and Komatsu 2018). Pharmacological inhibition of phosphorylated p62/SQSTM1 and KEAP1 interaction by K67 resulted in the suppression of NRF2 activity, which inhibited growth and re-sensitized the tumors to anti-cancer agents (Saito et al. 2016).
The high expression of NRF2, which has been observed in multiple types of cancers, contributes to chemoresistance by inducing downstream genes involved in electrophiles/ROS detoxification and the efflux of cytotoxic anticancer drugs (Srivastava et al. 2022). A recent report demonstrated that the activation of NRF2 by FAM117B-mediated KEAP1 competitive binding in gastric cancer facilitated tumor growth and chemoresistance to 5-fluorouracil and oxaliplatin (Zhou et al. 2023). In pancreatic cancer, oncogenic KRAS upregulated NRF2-dependent metabolic rewiring of glutamine metabolism, which leads to gemcitabine resistance (Mukhopadhyay et al. 2020). Keap1-deficient KrasG12D mice model developed lung tumors with hyperactivation of NRF2 and subsequent activation of the pentose phosphate pathway (PPP) (Best et al. 2019). Similarly, in KrasG12D–induced lung cancer, constitutive KRAS activation increased NRF2 transcription and resulted in cisplatin resistance. In this context, treatment with the NRF2 inhibitor brusatol was found to enhance the anti-tumor efficacy of cisplatin in KrasG12D-induced lung cancer (Tao et al. 2014).
Several studies have shown a direct correlation between NRF2 and BCRP in cancer. An early report demonstrated that treatment with oltipraz, an activator of NRF2, increased the mRNA level of BCRP in cultured human hepatocytes (Jigorel et al. 2006). This observation was further confirmed by a study that used tert-butylhydroquinone (tBHQ), another potent activator of NRF2, in HepG2 cells (Adachi et al. 2007). In this study, treatment of HepG2 cells with tBHQ induced translocation of NRF2 to the nucleus, and concomitantly increased the mRNA level of BCRP, as well as other ABC transporters and NRF2 targets, such as the heavy and light chains of γ-glutamylcysteine synthase. The decrease and increase in expression of BCRP mRNA mediated by NRF2- and KEAP1-silencing, respectively, suggest the involvement of the NRF2/KEAP1 system in the transcriptional regulation of BCRP (Adachi et al. 2007).
An advanced study by Singh et al. identified the ARE region located in the 5’-flanking region of the BCRP gene at -431 to -420 bp, which is responsible for NRF2-mediated expression in lung cancer cells in vitro and in vivo (Singh et al. 2010) (Fig. 1). This observation confirms previous findings of ARE located in the promoter regions of other multidrug ABC transporters, such as MRP1-3 (Stöckel et al. 2000; Kurz et al. 2001; Mahaffey et al. 2009). Similarly, in normal kidney tubular cells, knockdown of KEAP1 was found to increase BCRP expression. However, unlike in the previous studies where NRF2 activation increased BCRP expression, pharmacological activation of NRF2 did not elevate BCRP mRNA in this case (Jeong et al. 2015). In addition to transcriptional regulation, NRF2 has been reported to regulate BCRP through a post-transcriptional mechanism as well. For example, in ovarian and renal carcinoma cell lines, silencing of NRF2 led to the activation of miR-206, which targeted c-MET/EGFR and downregulated BCRP expression, resulting in increased intracellular accumulation of the anticancer drug doxorubicin (Singh et al. 2013; Choi et al. 2017).
PDT is an emerging chemotherapy strategy that utilizes photosensitizing drugs to selectively kill cancer cells by generating cytotoxic ROS upon light activation (Agostinis et al. 2011). PDT has gained increasing attention as a minimally invasive and more target-specific alternative to conventional chemotherapy (Dolmans et al. 2003; Correia et al. 2021). PDT has been approved by the FDA for the treatment of several types of cancer, including bladder, esophageal, head and neck, lung, and skin cancer (Correia et al. 2021). The expression of BCRP in cancer cells can limit the effectiveness of PDT, and therefore there has been interest in investigating the relationship between BCRP and PDT. One study used imatinib mesylate, a tyrosine kinase inhibitor (TKI) known to reduce BCRP levels, and found that it enhanced the accumulation of various photosensitizers, including 2-(1-hexyloxethyl)-2-devinyl pyropheophorbide-a (Photochlor), protoporphyrin IX, and the benzoporphyrin derivative monoacid ring A (Verteporfin), in ABCG+ human and murine cell lines both in vitro and in vivo (Liu et al. 2007). Another study showed that BCRP-silencing using siRNA or inhibition of its activity using a chemical inhibitor, resulted in increased intracellular accumulation of chlorin e6 and improved PDT efficacy in cancer cells. This suggests that targeting BCRP may be a promising strategy to enhance the efficacy of PDT in cancer therapy (Roh et al. 2017). In the doxorubicin-resistant ovarian cancer A2780DR cells, the increase of BCRP-mediated drug resistance is associated with c-MET elevation. Treating the A2780DR cells with BCRP inhibitor Ko143 or c-MET inhibitor resulted in the accumulation of the photosensitizer Pba within cancer cells, which in turn sensitized the cells to PDT (Jung et al. 2015). Another study showed that A431/BCRP lung cancer cells, which overexpress BCRP, are more resistant to Photofrin-mediated PDT. However, inhibition of BCRP with FTC reversed this resistance and increased the efficacy of PDT (Usuda et al. 2010). In line with this, NRF2-knockdown in several types of cancer cells, including triple-negative breast cancer, renal carcinoma, and colon cancer, led to an increase in the effectiveness of PDT by lowering the levels of BCRP (Choi et al. 2014). Furthermore, a recent report has demonstrated that nanoparticle-mediated NRF2 silencing can increase the sensitivity of tumor cells to oxidative stress during PDT. These findings further support the importance of the NRF2/BCRP pathway in the response of cancer cells to PDT (Sun et al. 2023).
Ever since the initial discovery of a subpopulation of leukemic stem cells characterized by the surface markers CD34+ and CD38−, CSCs have been recognized as promising targets for cancer therapy (Lapidot et al. 1994). The unique characteristics of CSCs, including their ability of self-renewal and differentiation potential, have been shown to play a critical role in cancer malignancies such as metastasis, therapy resistance, and recurrence (Batlle and Clevers 2017). CSCs and normal stem cells share many common transcription factors and signaling pathways, including several pluripotent transcription factors such as octamer-binding transcription factor 4 (OCT4), krupple-like factor 4 (KLF4), MYC, homeobox protein Nanog (NANOG), and SRY (sex determining region Y)-box 2 (SOX2) (Yang et al. 2020). Furthermore, a subpopulation of cancer cells known as the side population has been identified as a chemo- and radio-therapy-resistant population, which is thought to be due to its high expression of BCRP. As a result, BCRP has become recognized as one of the CSC markers (Zhou et al. 2001).
Recent studies have suggested that NRF2 may contribute to the development or maintenance of CSC properties, including chemoresistance (Kahroba et al. 2019; Choi et al. 2021; Hallis et al. 2023a). In the colonosphere model of in vitro CSCs, increased levels of P-gp, BCRP, and concomitant NRF2 signaling were observed, which exhibited resistance to doxorubicin (Ryoo et al. 2016). Subsequently, NRF2 knockdown in the colonosphere model sensitized these cells to drug treatment by decreasing efflux transporter activities. Consistent with these findings, the CD133high population from HCT116 colon cancer cells exhibited high levels of NRF2 and BCRP, along with elevated CSC markers, indicating that they possess CSC phenotypes such as enhanced proliferation, migration, colony formation, chemoresistance, and sphere-forming capacity (Park et al. 2022). In this experimental setting, colonospheres derived from NRF2-silenced HCT116 expressed low levels of BCRP along with reduced KLF4 expression, which resulted in the suppression of sphere-forming capacity. A similar study involving NRF2 knockdown in pancreatic ductal adenocarcinoma cell lines BxPC-3 and CFPAC-1, which had high expression of NRF2, found that it decreased the expression of CSC transcription factors NANOG, OCT4, and CD133, as well as chemoresistance properties, by downregulating BCRP (Kim et al. 2020). Silencing of NRF2 in sorafenib-resistant Huh7 hepatocellular cell line also limited cancer stemness and migration, as well as expression of BCRP (Gao et al. 2021). In another drug resistance model, A2780DR ovarian cancer cells that exhibit doxorubicin resistance showed higher NRF2 signaling activity concomitantly with elevated expression of aldehyde dehydrogenase 1 (ALDH1) and BCRP, which supports to maintain CSC properties (Shim et al. 2009; Kim et al. 2018). Isolated CSCs from the side population of cervical cancer patients showed aberrant activation of NRF2, BCRP, and Bcl-2, which resulted in chemoresistance and apoptosis resistance (Jia et al. 2015). Meanwhile, our recent study demonstrated a positive association between NRF2 and HIF-2α-induced CSC phenotypes under chronic hypoxic conditions (Hallis et al. 2023b). As previous studies have reported that HIF-2α directly regulates BCRP and confers CSC properties in ovarian CSCs, leading to resistance to doxorubicin and promoting tumor growth, this may suggest the potential association of NRF2 with HIF-2α/BCRP in hypoxic tumor environment (He et al. 2019). Collectively, these studies indicate that NRF2 and BCRP are intricately intertwined in maintaining resistance to chemotherapeutic agents and photodynamic therapy, and the maintenance of CSC properties in cancer (Fig. 2).
Figure 2. Role of NRF2 and BCRP in CSC maintenance and PDT resistance. (A) NRF2 signaling is activated in CSCs models, contributing to CSC properties, including tumor initiation and growth, invasion, metastasis, and therapy resistance. NRF2 upregulates the antioxidants gene to maintain low ROS levels and increases ABC transporters such as BCRP to protect CSCs against external harmful chemicals, which eventually leads to therapy resistance. High levels of NRF2 are also associated with the upregulation of other transcription factors, including OCT4, NANOG, KLF4, MYC, and SOX2, which help to maintain CSC phenotypes. (B) Enhanced BCRP in cancers facilitates the efflux of photosensitizers such as Pba, photochlor, and photofirin, thus leading to PDT resistance. BCRP inhibition by chemical inhibitors increases intracellular accumulation of photosensitizers and consequent ROS generation, leading to cancer cell apoptosis. OCT4, octamer-binding transcription factor 4; KLF4, krupple-like factor 4; NANOG, homeobox protein nanog; SOX2, SRY (sex determining region Y)-box 2; FTC, fumitremorgin C.
In this review, we have highlighted the critical role of NRF2 and BCRP as potential targets to counteract the therapy resistance of cancers. Elevated expression of BCRP in most cancer cases results in a poor prognosis and is associated with multidrug resistance, thereby impeding the efficacy of cancer therapy (Robey et al. 2018). Many studies have demonstrated the potential effect of BCRP inhibitors in multiple models of resistant cancers, and current efforts are focusing on investigating BCRP inhibitors for clinical trials. Thus, the regulation of BCRP by transcription factors or other signaling pathways is of great interest. Although several transcription factors have been implicated, the identification of ARE sequences in the promoter region of BCRP has highlighted the direct regulation by NRF2 (Singh et al. 2010). This is important since NRF2 overexpression in cancer clinical settings is correlated with poor prognosis and chemoresistance (Solis et al. 2010). Moreover, recent insights suggest that the suppression of BCRP and its regulator NRF2 may have an impact on PDT resistance and the maintenance of CSC properties. Thus, ongoing efforts are focused on strategies aimed at suppressing NRF2 activity to increase the sensitivity of cancer patients to chemotherapy (Dinkova-Kostova and Copple 2023). In particular, compared to the inhibition of BCRP, the inhibition of NRF2 is expected to be a more effective strategy for cancer treatment in that it can suppress the expression of other target genes that contribute to cancer growth and development as well as control of BCRP-mediated anticancer drug resistance. Currently, extensive efforts are being made to develop NRF2 inhibitors. Several natural compounds, including luteolin, procyanidins, and wedelolactone, have demonstrated NRF2 inhibitory effects in various experimental settings involving NRF2-overactive cancers (Sharifi-Rad et al. 2023). Brusatol from Brucea javanica has shown potent anti-tumor and anti-metastatic activities by decreasing NRF2 protein levels, which is attributed to multiple mechanisms, including protein translation inhibition (He et al. 2023). In a high-throughput chemical library screening, halofuginone has been identified as an inhibitor of NRF2 protein synthesis (Tsuchida et al. 2017). ML385, derived from a high-throughput chemical library screening and subsequent chemical optimization, has been discovered to interfere with the interaction between NRF2 and sMAF by directly binding to the NRF2 protein. Consequently, ML385 demonstrated the ability to sensitize KEAP1 mutant lung cancer cells to chemotherapy (Singh et al. 2016). While NRF2 inhibitors hold strong potential for enhancing chemotherapy sensitivity and suppressing tumorigenesis in NRF2-high cancers, the dual functionality of NRF2 in normal cells presents a limitation for its clinical application. In this context, achieving cancer-selective inhibition of NRF2 becomes a crucial strategy. In a recent study employing a synthetic lethal strategy, mitomycin C has emerged as a selectively cytotoxic agent for NRF2-overactive cancer cells, primarily through enhanced bioactivation via NRF2-target genes, such as NQO1 (Baird and Yamamoto 2021). Geldanamycin-derived HSP90 inhibitors have demonstrated selective lethality towards NRF2-high cancer cells, and their mechanism of action has been attributed to the enhancement of NQO1-mediated bioactivation (Baird et al. 2020). In summary, although a strategy for NRF2 inhibitors that selectively act on NRF2-overexpressing cancers has not been developed, with the progression of more advanced studies on the NRF2-KEAP1 system, there is an optimistic expectation for the future discovery of highly selective and effective NRF2/BCRP inhibitors.
This work was supported by National Research Foundation of Korea (NRF) grants funded by the Korean government (MSIT) (2022R1A2C2011866).
The authors declare that they have no conflict of interest.
PDF
Standard view
Export citation
Share 
Download
Download
