Elsevier

Bioorganic & Medicinal Chemistry

Volume 16, Issue 17, 1 September 2008, Pages 8224-8236
Bioorganic & Medicinal Chemistry

Structure–activity relationships of new inhibitors of breast cancer resistance protein (ABCG2)

https://doi.org/10.1016/j.bmc.2008.07.034Get rights and content

Abstract

At the end of the last century tariquidar (XR9576) was synthesized, pharmacologically investigated, and classified as a promising 3rd generation P-glycoprotein (P-gp) modulator. Following the discovery of BCRP in 1998 an increasing number of substances were studied in relation to their potency to interact with this transporter. Recently it has been shown that XR9576 inhibits both P-gp and BCRP transport function similarly to GF120918 (elacridar). This observation prompted us to investigate 5 XR compounds and 25 structurally related derivatives synthesized in our laboratory for their BCRP inhibitory effect. The biological activity data were determined by our new Hoechst 33342 assay that has been transferred from P-gp to BCRP overexpressing cells. 3D-QSAR models (CoMFA and CoMSIA) were generated and validated by the leave-many-out method and the scrambling stability test. The best models yielded an internal predictive squared correlation coefficient higher than 0.8 and contained steric, electrostatic, hydrophobic, and hydrogen bond donor fields. To our knowledge, this is the first 3D-QSAR analysis of BCRP inhibitors. Additionally the biological activity data determined in P-gp overexpressing cells on one side and BCRP overexpressing cells on the other side were compared to identify selective and non-selective inhibitors of P-gp and BCRP. The results may help to get a better insight which structural elements are necessary to direct the interaction of these compounds with P-gp and/or BCRP.

Introduction

Multidrug resistance is a major obstacle to successful cancer treatment by chemotherapy. One pivotal mechanism by which tumor cells can become resistant to cytotoxic drugs used in chemotherapy is the increased expression of certain ATP-binding cassette (ABC) transporters. These ABC transporters use the energy of ATP hydrolysis to transport a wide variety of substrates out of cells against a concentration gradient leading to a decreased intracellular concentration and in turn to failure of chemotherapy.

To date there have been 49 human ABC transporters identified and classified into seven subfamilies.1 In humans, members from three subfamilies are primarily associated with the phenomenon of multidrug resistance (MDR). These include a member of the B- (ABCB1, P-glycoprotein), the G- (ABCG2, BCRP) and several members of the C-subfamily (ABCC1-5, MRPs).2, 3 As members of these three different subfamilies can be simultaneously overexpressed in tumor cells, MDR emerges frequently as a multifactorial, complex problem.

The MDR-related ABC transporters are widely expressed in excreting organs and physiological barriers such as the intestine, liver, kidney, blood–brain barrier and placenta, where they play a physiological key role in the protection against toxic agents.4, 5, 6

With regard to their physiological role, MDR-ABC transporters most probably evolved as complex cellular defense systems for the recognition and the energy-dependent removal of toxic agents entering the living cells or organisms from their environment. Therefore, they can be considered as essential parts of an immune-like defense system.7 When expressed in tumor cells, ABC transporters such as P-gp, BCRP, and MRP1 cause multidrug resistance of such tumor cells. Therefore, it is in focus of research interest to overcome MDR-mediated resistance in cancer cells.

P-glycoprotein (P-gp) was discovered by Juliano and Ling in 19768 and is possibly the most intensely studied among the ABC transporters. In accordance with other ABC transporters, P-glycoprotein is a membrane-bound, energy-dependent drug transporter.9 P-gp consists of 1280 amino acids and has a molecular weight of approximately 170 kDa in the fully glycosylated form.10, 11, 12 The protein is folded into two homologous halves, each containing six transmembrane helices and one intracellular nucleotide binding domain (NBD). P-gp contains three glycosylation sites on an extracytoplasmatic domain13, which have no essential role for transport function and serve to anchor the protein in the membrane.10, 11

By actively effluxing substrates, P-gp reduces intracellular concentrations to subtherapeutic levels, thereby conferring resistance to a broad range of cytotoxic drugs. Its most striking property is the diversity in structure of substrates that can be transported, including a vast number of drugs, among them many cytotoxic anticancer drugs, such as anthracylines, Vinca alkaloids, taxanes, podophyllotoxin derivatives, and derivatives from Campthoteca acuminata (camptothecin, topotecan, irinotecan).14, 15 The consensus that has emerged is that P-gp-recognized substrates are amphipatic7 with a molecular mass of 200–1900 Da.10 In most cases anionic substrates show no interaction with P-gp.10

The breast cancer resistance protein (BCRP, ABCG2) was first identified as an overexpressed protein in the MCF-7/AdrVp cell line. This cell line was selected by continuous exposure to doxorubicin in combination with verapamil to avoid development of resistance due to expression of P-gp. Interestingly, MCF-7/AdrVp cells also did not express the multidrug resistance-associated protein MRP1.16, 17 BCRP was also discovered by investigating cell lines selected for resistance to mitoxantrone – a poor substrate for P-gp and MRP1.18

ABC transporters belonging to the subfamily G possess only one NBD and a single transmembrane domain (TMD), containing six membrane-spanning helices and thus are considered to be ‘half transporters’. BCRP is composed of 655 amino acids and has a molecular weight of 72 kDa. It forms a homodimer probably linked by a disulfide bond in the membrane.6 It migrates as a dimer without reducing agents on a SDS gel, but runs as a 72 kDa protein on a SDS gel under reducing conditions.18, 19 Further ABCG2 is extensively glycosylated on asparagine-596, which is located within the third extracellular loop of the polypeptide.20, 21, 22 In accordance with P-gp, this glycosylation has no effect on transport function, but anchors the protein in the membrane. While the majority of ABC transporters include an N-terminal TMD followed by a C-terminal NBD, the ABCG2 protein has a reverse structure with the NBD located N-terminal to the TMD.17 The structural differences serve to underline the importance of ABCG2 structural studies.

ABCG2 is not only expressed in cancer cells, but is also present in many normal tissues, for example, in the placenta, particularly in the syncytiotrophoblastic cells. The remarkable expression level of ABCG2 in the placenta led to the nomenclature ABCP, and suggests that ABCG2 is responsible for protecting the fetus from toxic xenobiotics.23, 24 Lower levels of BCRP were also found in the brain, in the colon, small intestine, breast tissue, testis, ovary, liver, and prostate. The tissue localization points to a physiologically protective role of BCRP.

In comparison to P-gp BCRP is able to efflux large molecules, both positively and negatively charged, with amphiphilic character. To the transported substrates belong sulfated hormone metabolites or glucuronidated methotrexate.25, 26, 27 Further substrates of BCRP include the anticancer drugs mitoxanthrone, campthotecin-based topoisomerase 1 inhibitors such as topotecan or irinotecan, tyrosin kinase inhibitors such as imatinib and gefitinib, the podophyllotoxins etoposid and teniposid and flavopiridol.28, 6, 29 A strategy to reverse BCRP-mediated multidrug resistance and to improve treatment outcome is the inhibition of this ABC protein with modulators showing high potency and efficacy. Selective inhibitors for BCRP include fumitremorgin C6 and novobiocin.30 More recently a non-toxic, synthetic analogue of fumitremorgin C, Ko143, has been identified as a potent and selective modulator of BCRP inhibiting BCRP at nanomolar concentrations.31 Further, natural products are known to interact with BCRP. Investigations of Ahmed-Belkacem et al. illustrated that the flavonoids 6-prenylchrysin and tectochrysin can be considered as BCRP-specific inhibitors. For these compounds no interaction with P-gp or MRP1 was detected whereas low concentrations of these substances led to a distinct interaction with BCRP comparable to that of GF120918 (elacridar).32 In another study, it was shown that the flavonoids narigenin and genistein possess higher inhibitory potencies towards BCRP than to MRP1 and P-gp.33 These findings underline the importance of further investigations of flavonoids towards BCRP due to their potential specific interaction with this ABC-transporter. Though different in structure and genetic background, P-gp and BCRP share several common substrates. P-gp inhibitors that have been reported to also interact with BCRP are for example GF120918 (elacridar),34 cyclosporin A,35 reserpin36 and tariquidar.37, 38 A dual specific inhibitor could be advantageous for reversing resistance to drugs that are substrates of both transporters.

In our laboratory, tariquidar (XR9576) served as a template to design new MDR modulators.39, 40 In recently published studies, we presented the pharmacological results of these compounds in relation to P-gp.39, 40 The biological activity data were obtained in the new Hoechst 33342 assay and the well established Calcein AM assay. 3D-QSAR models were generated by CoMFA and CoMSIA and yielded an internal predictive squared correlation coefficient higher than 0.8. The best models were based on electrostatic, steric, hydrogen bond acceptor, and hydrophobic fields. Considering that tariquidar inhibits not only P-gp function but also is a powerful BCRP modulator,29 we decided to determine the inhibitory potencies of five XR-derivatives and 25 newly synthesized structurally related compounds towards BCRP function. In this study we present their inhibitory activities determined in BCRP overexpressing tumor cells. The inhibitory potencies of these substances were determined in the new Hoechst 33342 assay described previously39 that was transferred to BCRP overexpressing cells. 3D-QSAR models using CoMFA and CoMSIA methodologies were generated and validated applying different techniques: the scrambling stability test, the leave-one-out, and leave-many-out cross-validation. The best models yielded internal predictive squared correlation coefficients of higher than 0.8. These models may serve as useful tools for the rational design of new BCRP inhibitors. To our knowledge this is the first 3D-QSAR study in relation to BCRP inhibitors. One additional, pivotal aim of this study was to compare the biological activity data of these compounds determined in P-gp overexpressing cells on one side and BCRP overexpressing cells on the other side. The intention of this comparison was to answer the question whether these substances inhibit P-gp and BCRP with identical potency, or not. The results may support a better insight which structural features direct an interaction with P-gp and/or BCRP.

Section snippets

Functional assays

For a selective investigation of BCRP it was important to illustrate that the MCF-7 MX cell line overexpresses the ABCG2 gene product BCRP only. As shown in different studies by several working groups the mitoxantrone-resistant MCF-7 MX cell line is supposed to express high levels of BCRP.41, 42, 43, 44 In this study, the overexpression of BCRP was proven applying the specific anti-BCRP monoclonal BXP-21 antibody and fluorescein-conjugated anti-mouse antibody with flow cytometry detection.45, 46

Discussion

The Hoechst 33342 assay for the determination of P-gp activity, presented in a previous study,39 was transferred in this study to the measurement of BCRP transport activity. The overexpression of BCRP in the cell line MCF-7 MX was confirmed in agreement with literature data.41, 42, 43, 44, 54 Further, the absence of P-gp was proven in the used cell line. This is an essential precondition to assure a selective investigation of BCRP because Hoechst 33342 is also known to be a P-gp substrate.

To

Chemicals

XR9576 (tariquidar) was a gift from Prof. Dr. Buschauer (University of Heidelberg, Germany). Ko143 was kindly provided by Dr. A. Schinkel (Amsterdam, The Netherlands). Doxorubicin was a gift from Medac (Hamburg, Germany), all other chemicals were purchased from Sigma Chemicals (Taufkirchen, Germany) unless otherwise stated.

The novel tetrahydroisoquinoline-ethyl-phenyl-amine-based MDR inhibitors, as well as the XR-derivatives, have been recently synthesized in our laboratory.66 Stock solutions

Acknowledgments

The authors thank Dr. A. H. Schinkel, The Netherlands Cancer Institute, Amsterdam, The Netherlands, for providing Ko14332 and Dr. Armin Buschauer, Institute of Pharmaceutical Chemistry, University of Regensburg, Germany, for tariquidar. This work was supported by DFG (Deutsche Forschungsgemeinschaft) grants GRK677.

References and notes (68)

  • B. Sarkadi et al.

    FEBS Lett.

    (2004)
  • A. Bodo et al.

    Toxicol. Lett.

    (2003)
  • A.H. Schinkel et al.

    Adv. Drug Deliv. Rev.

    (2003)
  • C.J. Chen et al.

    Cell

    (1986)
  • K.J. Bailey-Dell et al.

    Biochim. Biophys. Acta

    (2001)
  • F. Staud et al.

    Int. J. Biochem. Cell Biol.

    (2005)
  • C.J. Chen et al.

    Cell

    (1986)
  • M. Suzuki et al.

    J. Biol. Chem.

    (2003)
  • H. Müller et al.

    Bioorg. Med. Chem

    (2007)
  • H. Müller et al.

    Bioorg. Med. Chem.

    (2008)
  • M. Aihara et al.

    Blood

    (1991)
  • A.B. Shapiro et al.

    J. Biol. Chem.

    (1995)
  • I. Ifergan et al.

    J. Biol. Chem.

    (2004)
  • T. Fojo et al.

    Oncogene

    (2003)
  • P. Borst et al.

    J. Natl. Cancer Inst.

    (2000)
  • V.V. Rao et al.

    Proc. Natl. Acad. Sci. U.S.A.

    (1999)
  • L.A. Doyle et al.

    Oncogene

    (2003)
  • B. Sarkadi et al.

    Physiol. Rev.

    (2006)
  • R.L. Ling et al.

    Biochim. Biophys. Acta

    (1976)
  • V.P. Ling

    Am. J. Med.

    (1995)
  • U. Hoffmann et al.

    Drug Metab. Rev.

    (2004)
  • M. Wiese et al.

    Curr. Med. Chem.

    (2001)
  • S.V. Ambudkar et al.

    Annu. Rev. Pharmacol. Toxicol.

    (1999)
  • T. Litman et al.

    J. Cell Sci.

    (2000)
  • H. Kusuhara et al.

    Pflugers Arch.

    (2007)
  • L.A. Doyle et al.

    Proc. Natl. Acad. Sci. U.S.A.

    (1998)
  • R.W. Robey et al.

    Cancer Metastasis Rev.

    (2007)
  • N.K. Diop et al.

    Biochemistry

    (2005)
  • K. Mohrmann et al.

    Cancer Chemother. Pharmacol.

    (2005)
  • T. Takada et al.

    Pharm. Res.

    (2005)
  • H. Saito et al.

    J. Pharmacol. Exp. Ther.

    (2006)
  • E.L. Volk et al.

    Cancer Res.

    (2003)
  • S.E. Bates et al.

    J. Bioenerg. Biomembr.

    (2001)
  • R.W. Robey et al.

    Clin. Cancer Res.

    (2001)
  • Cited by (87)

    • Establishment of drug-resistant cell lines under the treatment with chemicals acting through different mechanisms

      2021, Chemico-Biological Interactions
      Citation Excerpt :

      Multiple drug resistance (MDR) is the main cause for a decrease in the effectiveness of anticancer therapy, the development of relapses and tumor metastases at different stages of the treatment, which leads to death in more than 90% cases of mature tumors [1,2].

    • Synthesis and biological evaluation of quinazoline derivatives – A SAR study of novel inhibitors of ABCG2

      2019, European Journal of Medicinal Chemistry
      Citation Excerpt :

      High Five insect cells as well as baculovirus were a kind gift from Dr. Christine Hrycyna (Purdue University, Indiana, USA). To investigate the inhibitory effect of the test compounds toward ABCG2 the Hoechst 33342 accumulation assay was performed according to literature with small modifications [23,35–46,50]. The assay was carried out with the corresponding compound dissolved in DMSO to a final concentration of 10 mM.

    View all citing articles on Scopus

    These authors contributed equally.

    View full text