Review Article

Multi-Drug Resistance in Cancer - Role of ABC Transporter Proteins

Balázs Sarkadi¹, Gergely Szakás^1,2 and András Váradi³

1. National Medical Center, Institute of Hematology and Immunology, Budapest, Hungary
2. National Cancer Institute, NIH, Bethesda, MD, USA
3. Institute of Enzymology, Hungarian Academy of Sciences, Budapest, Hungary

Chemotherapy is the treatment of choice in many malignant diseases. A major form of resistance against a variety of the antineoplastic agents currently used involves the function of a group of membrane proteins that extrude cytotoxic molecules, thus keeping intracellular drug concentration below a cell-killing threshold. Multi-drug transporters belong to the superfamily of ATP Binding Cassette (ABC) proteins, present in organisms from bacteria to humans. The medical significance of ABC proteins exceeds their role in cancer therapy resistance; the transport function of several members was found to hinder the effective therapy of many other widespread diseases (e.g. malaria, AIDS), and inherited diseases were also linked to mutations in these genes. The transport activity of ABC proteins has an important effect in general pharmacology, that is, in modulating the absorption, distribution and excretion of numerous pharmacological compounds.

ABC transporter proteins are located in the plasma membrane of the cells or in the membrane of different cellular organelles, and mediate the translocation of various molecules across these barriers. These substrate molecules exhibit a wide variety of chemical structures. Some ABC proteins facilitate the transport of inorganic ions, whereas others pump various organic compounds, including lipids, bile acids, glutathione and glucuronide conjugates, or even short peptides. Most ABC proteins utilize the energy of ATP hydrolysis for this transport activity (active transporters), but some ABC transporters form specific membrane channels.

The typical structure of an ABC protein consists of membrane-embedded transmembrane domains (TMD) and ATP binding domains (ABC). Typically, the transmembrane regions anchor the protein to the membrane and form a pore through which the transport of a surprisingly large variety of substrates occurs. The cytoplasmic nucleotide binding domains provide the molecular compartment where the energy of ATP is released. It is not known how the energy is conveyed from the ABC domains to the site of the transport and the precise mechanism of transport also remains elusive.

In this mini-review we focus on the structure, function, diagnostics and possible therapeutic modulation of the human ABC transporters causing multi-drug resistance (MDR) in cancer. We describe their current nomenclature, phylogenetic relationship and basic membrane topology. We also summarize the anticancer drugs involved in this phenotype, the diagnostic tools developed for studying the expression and function of these proteins, and give an overview of the efforts to develop effective and selective modulating agents for combination chemotherapy. In some aspects of this review we refer the interested reader to more detailed current review articles.

(top of page)

Cancer Multi-Drug Resistance - The Players

Numerous clinical data revealed that the multi-drug resistance phenotype in tumors is associated with the overexpression of certain ABC transporters, termed MDR proteins. The P-glycoprotein (Pgp, MDR1, ABCB1)-mediated multi-drug resistance was the first discovered^1-3 and probably still is the most widely observed mechanism in clinical multi-drug resistance^4-7. Soon after the cloning and characterization of MDR1, it became evident that other efflux-pumps may also play a significant role in transport-associated drug resistance. There are two other ABC transporters, which have been definitively demonstrated to participate in the multi-drug resistance of tumors: the multi-drug resistance protein 1 (MRP1, ABCC1), and the mitoxantrone resistance protein (MXR/BCRP, ABCG2)^7-11. Furthermore, other human ABC proteins capable of actively transporting various compounds out of cells may also be players in selected cases of multi-drug resistance. These include ABCB4 (MDR3) and ABCB11 (sister Pgp or BSEP), two proteins residing predominantly in the liver with a function involved in the secretion of phosphatidyl choline and bile acids, respectively^12-14. MDR3 has been already shown to transport certain drugs as well¹⁵. In addition to MRP1, five homologues (MRP2-MRP6) have been recently cloned. Overexpression of MRP2 (an organic anion transporter which can also extrude hydrophobic compounds) was definitively shown to confer cancer MDR^9,16. MRP3, an organic conjugate transporter, and MRP5, a nucleoside transporter, are also candidate proteins for causing certain forms of drug resistance⁹. The variable contribution of all these transport proteins to the clinical multi-drug resistance has been widely investigated.

Basic mechanism of cancer MDR

The generally accepted mechanism of multi-drug resistance is that the MDR proteins actively expel the cytotoxic drugs from cells, maintaining the drug level below a cell-killing threshold. Drug extrusion mediated by these primary active transporters is driven by the energy of ATP hydrolysis. The most intriguing characteristic distinguishing the MDR proteins from other mammalian transporters is their wide substrate specificity. Unlike other selective (classical) transport proteins, multi-drug transporters recognize and handle a wide range of substrates. This wide substrate specificity explains the cross-resistance to several chemically unrelated compounds, the characteristic feature found in the multi-drug resistance phenotype^4-7.

Different tumors with MDR protein overexpression (e.g. hepatomas, lung or colon carcinomas) often show primary (or intrinsic) resistance to cancer chemotherapy. In addition, cancer chemotherapy itself might induce the overexpression of these proteins, so that the multi-drug resistant clones become less sensitive to chemotherapy (secondary drug resistance). Treatment failure due to multi-drug resistance is also found in connection with conditions other than cancer, including some autoimmune disorders and infectious diseases^17-19.

Nomenclature, basic structure and membrane topology of MDR proteins

The ABC superfamily is one of the largest families of proteins. The most recent annotation of the human genome sequence revealed 48 genes for ABC proteins. The ABC proteins were grouped into seven sub-classes, ranging from ABCA to ABCG [see: http://nutrigene.4t.com/humanabc.htm] ^20-24 based on genomic organization, order of domains and sequence homology. The phylogenetic tree of the ABC transporters involved in cancer MDR is presented in Figure 1. A thick line and a circle label the three definite players, while the close relatives, which may have a role in drug resistance are also indicated on this evolutionary diagram.

FIGURE 1: Phylogenetic tree of the MDR-related ABC transporters.

All ABC proteins contain at least three characteristic peptide sequences: the Walker A and B motifs and the so-called ABC-signature sequence. Whereas the Walker motifs are present in several classes of ATP binding proteins, the presence of the signature region is diagnostic for the ABC proteins. It is generally accepted that the minimum functional unit requirement for an ABC transporter is the presence of two transmembrane domains (TMD) and two ATP Binding Cassette (ABC) units. These may be present within one polypeptide chain ("full transporters"), or within a membrane-bound homo- or heterodimer of "half transporters"^{4-7, 23, 24}. There are no high-resolution structural data presently available for any mammalian ABC-transporter; therefore computer modeling and laborious biochemical experiments are necessary to elucidate the position and orientation of membrane spanning segments and other domains within the polypeptide chain. Figure 2 presents the most plausible membrane topology models for the key MDR-ABC transporters. As shown in Figure 2, Pgp-MDR1 (ABCB1) is a "full transporter" with six TM helices in both TMDs of the protein, each followed by an ABC domain. A similar membrane topology has been predicted for ABCB4 (MDR3), and ABCB11 (sister PgP) as well^20-24.

MRPs belong to the ABCC-subfamily, comprising eleven members in the human genome. Most of these proteins (ABCC1-6) have been identified as active, ATP-dependent membrane transporters for various drugs and organic anions^8,9,10,13. In contrast to these active transporters, the cystic fibrosis transmembrane conductance regulator, ABCC7 (CFTR) is a regulated chloride channel, while ABCC8 (SUR1) and ABCC9 (SUR2) are called sulfonylurea receptors and best described as intracellular ATP sensors, regulating the permeability of specific K⁺ channels. Nothing is currently known about the function of ABCC10 and ABCC11^7,9,23,24.

The predicted membrane topology of MRP1 is shown in Figure 2. According to our current notion, in addition to an MDR1-like core, MRP1 contains an additional N-terminal segment of about 280 amino acids. A major part of this region is membrane-embedded with five transmembrane helices (TMD0), while a small cytoplasmic loop of about 80 amino acids (L0) connects this area to the core region^25-28. Recent studies revealed that the TMD0 domain of ABCC1 does not play a crucial role in either the transport activity or the proper routing of the protein. However, the presence of the membrane-associated cytoplasmic L0 region (together with the core region) is necessary for both the transport activity and the proper intracellular routing of the protein. These studies indicate that the L0 region forms a distinct structural and functional domain, which interacts with the membrane and the core region of the MRP1 transporter²⁹.

FIGURE 2: Membrane topology models for the MDR-related ABC transporters.

The third ABC protein believed to play a role in clinical MDR, ABCG2 (MXR/BCRP) is a half transporter^{11, 30}, with a unique domain arrangement, where the ABC is located at the N-terminus (see Figure 2). This protein performs an active extrusion of hydrophobic, positively charged molecules from the cells in an N-glycosylated mature form, and - in contrast to many other ABC half-transporters - is probably localized in the plasma membrane. Recently, it has been shown that the human ABCG2 multi-drug resistance protein forms an active homodimer for its transport function^{31, 32}.

There is no high-resolution three-dimensional structure available for any of the mammalian ABC transporters, thus the structural background of the MDR molecular mechanism is currently unresolved. A low-resolution structure of the MDR1³³ indicates that the protein is embedded into the membrane as a cylinder with a large central pore, which is closed at the inner (cytoplasmic) face of the membrane. This structure also included an opening of this cylinder to the lipid phase.

The structure of a bacterial ABC-transporter, MsbA of E. coli, has recently been determined by X-ray crystallography³⁴. MsbA is a half-transporter with a TMD-ABC domain arrangement, organized as a homodimer. The structure reveals that each MsbA subunit contains a transmembrane domain with six transmembrane helices, an ABC-domain, and an "intracellular domain" which is composed of the three intracellular loops connecting the transmembrane segments to the ABC-domain. One of the most important conclusions of the MsbA structure is that the membrane-spanning segments of the polypeptide are indeed alpha helices. The organization and interactions of these peptide domains will probably be a valuable foundation towards elucidating the structures of mammalian multi-drug transporter ABC proteins.

(top of page)

Substrate Specificity and Cellular Distribution of MDR-ABC Transporter

The three major MDR proteins are highly promiscuous transporters; they share the ability of recognizing and translocating a large number of structurally diverse, mainly hydrophobic compounds. In addition to their overlapping substrate specificity, each transporter can handle unique compounds. Therefore, the suggestion that the list of substrates makes up half of the Sigma catalog may not be much of an exaggeration.

Pgp is a transporter for large hydrophobic, either uncharged or slightly positively charged compounds while the MRP family primarily transports hydrophobic anionic conjugates and extrudes hydrophobic uncharged drugs. The MRP1-related uncharged drug transport is quite an enigma, and is somehow linked to the transport or allosteric effect of cellular free reduced gluthatione⁹. The exact spectrum of the MXR (ABCG2) transported substrates has not yet been explored in detail, and these studies are complicated by the variable substrate-mutants of MXR observed in the most recent studies⁶².

In order to put the MDR substrates in their medical and pharmacological context, we present some of the key molecules in separate figures. In Figure 3A are cancer drugs, which are, unfortunately for the patients, also MDR substrates. Figure 3B shows the chemical MDR modulators used experimentally or in clinical trials, while Figure 3C compiles the best-known MDR substrates used for functional diagnosis of these proteins^{4-7, 35-38}.

FIGURE 3 (3A, 3B, 3C): Venn-diagram for selected compounds interacting with the key MDR-related ABC transporters^{7, 11, 76-83}.

Cellular and tissue distribution

The tissue distribution of the MDR-ABC proteins is as varied as their substrate specificity. MRP1 is almost ubiquitously expressed, while the expression of Pgp is more restricted to tissues involved in absorption and secretion^4-7. High level MDR1 expression has also been shown in certain pharmacological barriers of the body, such as the blood-brain barrier and the choroid plexus^39,40. It has been reported that MXR is highly expressed in the placenta, liver, and most interestingly, in various stem cells^{30, 41}. All multi-drug transporters are localized predominantly in the plasma membrane. In polarized cells, Pgp-MDR1 is localized in the apical (luminal) membrane surface (e.g. in the epithelial cells of the intestine and the proximal tubules of kidney, or in the biliary canalicular membrane of hepatocytes)^42-44. In contrast, MRP1 expression in polarized cells is restricted to the basolateral membrane. The expression of MRP2, MDR3, and of Sister Pgp (BSEP) is predominant in the canalicular membrane of hepatocytes, while MRP3 and MRP5 are expressed in the basolateral membranes of these cells (see Figure 4). MRP2 is also highly expressed in the apical membranes of kidney proximal tubules. In polarized cells MXR expression was reported to be mostly apical⁴⁵.

(top of page)

FIGURE 4: Multi-drug transporters in the human liver hepatocytes.

Molecular Mechanism of the Multi-Drug Pumps

Drug transport by MDR proteins requires the energy of ATP-hydrolysis, controlled by drug interaction, and closely coupled to the actual drug translocation. Interaction with the drug-substrate significantly enhances the basal ATPase activity of the multi-drug transporters, that is, the transported drug-substrates increase the rate of ATP cleavage^46-48. Several models have been published describing the kinetics and mechanism of this substrate-dependent ATP hydrolysis and the related transport phenomena^49-50. The schematic pictures of the proposed molecular mechanisms of the MDR1 and MRP1 proteins, as depicted in Figure 5, summarize some of the information available. Clearly our understanding of the details of these molecular events is limited.

The site(s) in multi-drug transporters interacting with the drug-substrates are probably encoded in the transmembrane domains. Detailed mutagenesis studies of MDR1 and photochemical labeling with reactive drug-derivatives revealed that transmembrane helices 5 and 6 (in the N-proximal transmembrane domain), helices 11 and 12 (in the C-proximal transmembrane domain), as well as the short cytoplasmic loops connecting these helices, are involved in the formation of an extended drug-binding site(s)⁵¹. There are strong indications that the hydrophobic substrates of MDR1 are recognized within the membrane bilayer or in its vicinity, and this type of recognition makes the MDR1 protein a highly effective pump, preventing the cellular entry of toxic compounds⁵². In the case of MRP1 a similar picture has emerged. Recent studies have explored some parts of the transmembrane domains involved in drug interactions⁵³.

FIGURE 5: Possible model for the molecular mechanism of multidrug transporters.

Based on the three-dimensional structures of bacterial ABC-units, the nucleotide binding sites appear as shallow, more or less open grooves, forming atypical active sites. The close interaction of the two ABC units' likely results in the formation of a fully competent catalytic site. The regions connecting the ABC units to the transmembrane domains have a key role in the transfer of conformational information within the protein, and the ABC signature region may have a special function in this regard⁵⁴.

The ATP-hydrolytic cycle of both MDR1 and MRP1 have been investigated in detail. As mentioned above, interaction with the transported drugs enhances the catalytic ATPase activity of MDR1. In all ABC transporters, ATP binding and hydrolysis occurs at the sites localized in the ABC domains. Thus, an allosteric control of the drugs on the ATPase activity requires intramolecular interaction between the drug binding and the catalytic regions of the protein, but the mechanism by which the transported drugs accelerate ATP-hydrolysis is presently unknown. It has been documented in detail that the interaction of the two ABC units is an essential requirement for the catalytic reaction^{55, 56}. Several lines of evidence indicate that both NBDs can bind ATP, and both catalytic sites are active and, at least in the case of MDR1, the two ABC domains enter alternately into the catalytic cycle^{57, 58}.

The transport and ATPase cycle of the MDR proteins is blocked by vanadate, a phosphate-mimicking inhibitory anion, which stabilizes a transition state intermediate of the ATPase cycle. An occluded nucleotide in the catalytic sites is locked within the ABCs in this interaction. This occluded ("trapped") nucleotide state can be visualized by using (a -³²P)-azido-ATP with covalent photoaffinity labeling, thus the individual steps of ATP-hydrolysis can be studied. Similar to their ATPase activity, the rate of the vanadate-dependent nucleotide occlusion in MDR-ABC proteins is greatly accelerated by the transported drug-substrates⁵⁵. This is consistent with a model where substrate recognition increases the rate of the transition-state formation, the rate-limiting step of the catalytic cycle. It has recently been shown, that in the case of MDR1 the MDR1*MgADP*Vi complex exhibits a dramatically reduced binding affinity for the transported drug substrate, as compared to the MDR1*MgATP complex⁵⁹. This observation suggests that the hydrolytic step triggers conformational changes, which reduce drug binding to the binding site (and presumably makes drug binding to another site favorable, from which the drug can be released to the extracellular space). The authors also demonstrated that hydrolysis of a second ATP-molecule is required for conformational changes to reset the MDR1 transporter to the high affinity drug-binding conformation. On the basis of these results the probable stoichiometry is two ATP cleaved per molecule of drug substrate transported.⁵⁹.

(top of page)

Diagnostics of Cancer MDR

The reliable detection and quantitative determination of the clinically relevant levels of the MDR proteins would be extremely important in the clinical treatment of various cancerous diseases. Combination chemotherapy treatment protocols could be adjusted and drug-resistance reversing agents could be applied accordingly.

A large body of laboratory methods have been devised to estimate the expression and/or surface appearance of P-glycoprotein by molecular biology or immunodetection methods. Quantitative PCR technologies are more accessible and applicable in this regard, and selective and high-affinity monoclonal antibodies have also been developed for each of the MDR-ABC transporters. Still, intra and interlaboratory differences are disturbingly large in such estimations, and no validated consensus protocols have been established as yet.^{7, 11, 17, 60, 61}

A different approach is the determination of MDR transporter function (e.g. by using fluorescent substrate compounds and a combination of selective inhibitors). This can be achieved by measuring the cellular uptake, efflux, or steady-state distribution of several fluorescent substrates such as; anthracyclines, Hoechst dyes, verapamil-derivatives, rhodamine 123 (RH 123), or fluorescent indicators (e.g. Fluo-3).¹⁹ Most of these assays have serious drawbacks in quantitation, because the fluorescence and cellular distribution of these compounds depends on pH, intracellular free calcium, or binding to various cell constituents. Currently, the most commonly used fluorescence assay method to discriminate between drug-resistant and -sensitive cells is based on the efflux of RH 123 by MDR1. However, there are many disadvantages to this method, and it is not applicable in several forms of multi-drug resistance: RH 123 is not a substrate for any MRPs, and only a weak substrate (and probably only in a mutant form) for MXR^{11, 62} (see Figure 3C).

The Calcein Assay is a functional MDR diagnostic method developed in our laboratory. In this method we follow the accumulation of the fluorescent free Calcein inside cells^{63, 64}. The non-fluorescent Calcein AM rapidly traverses the cell membrane and is converted to Calcein by non-specific esterases. In MDR1-expressing cells Calcein AM is extruded by the multi-drug transporter before its intracellular conversion to the non-MDR1 substrate, free calcein. However, fluorescent free calcein rapidly accumulates when an MDR pump interfering agent (e.g. verapamil) blocks this Calcein AM extrusion (see Figure 6).

FIGURE 6: The basic principle of the Calcein assay.

This assay possesses numerous advantages - it is simple, rapid, and Calcein has a bright green fluorescence, which is well suited to flow-cytometry or other basic fluorescence-detection equipments. Calcein fluorescence is practically insensitive to pH and intracellular ion concentrations, and does not show spectral changes upon accumulation in intracellular compartments or when bound to cellular components. The enzymatic enhancement of the dye trapping process makes this assay far more sensitive than other functional assays. Based on this phenomenon, the Calcein assay we developed is a functional and quantitative measurement, and is suitable for flow cytometry- clinical laboratory applications. The first clinical studies indicate a strong predictive value of the Calcein assay for chemotherapy response and survival in leukemic patients⁶⁵. Calcein AM and free Calcein are also substrates for MRP1, thus the same assay, combined with an MRP1 inhibitor, can also be applied for estimating the functional expression of MRP⁶⁶ (see Figure 4). However, the assay is not applicable for the functional detection of MXR because neither Calcein AM nor Calcein is a substrate of this transporter.

(top of page)

Circumvention of Cancer MDR

Prevention of clinical MDR should significantly improve therapeutic response in a large number of tumor patients. One way to achieve this goal would be to develop anticancer agents that do not interact with any of the multi-drug transporters. However, as cytotoxic drugs must penetrate the cell membrane, and the MDR proteins have an extremely wide recognition pattern, this seems to be a remote possibility. There are several suggested methods to prevent the expression or function of multi-drug transporters (see below), but pharmacological modulation seems to be the first choice at present.

MDR modifying agents

MDR modifying agents, which inhibit the function of the MDR proteins either competitively or non-competitively, are good candidates for such a pharmacological modulation. These compounds are expected to increase the cytotoxic action of MDR-related drugs by preventing the extrusion of anticancer drugs from the target cells.

The co-application of a non-toxic "MDR-modulating" compound with combination chemotherapy may significantly improve the cancer cure rate. Indeed, we have witnessed a growing interest in the use of such clinically active agents. Biochemical investigations led to the identification of several Pgp modulator compounds, as diverse in structure as the transported drugs. The first generation consisted of drugs that were already in clinical use. Calcium channel blockers (Verapamil, Diltiazem, Azidopine), quinine derivatives, calmodulin inhibitors (Trifluoroperazine, Chlorpromazine) and the immunosuppressive agent, Cyclosporin A, were all shown to interact with the MDR transporters in vitro and in vivo. (See figure 3B).

The second generation modulators consisted of derivatives of these first generation compounds, which had less pronounced effect on their original target, but retained their modulatory effects. Two prominent examples of this group may be R-verapamil and PSC-833, the latter being a cyclosporin analog without immunosuppressive effect. Because of the differences in the substrate-specificities and inhibitor-sensitivities of the different MDR-ABC proteins expressed in different tumor cells, proper therapeutic intervention requires an advanced diagnosis and targeted modulator agents.

The third generation of MDR modifiers is molecules specifically devised to interact with specific MDR transporters^{7,11,17,35-37} (a few examples are listed in Figure 3B). As these agents are mostly in the early development phase in various research laboratories, their clinical efficiency has yet to be proven. Small hydrophobic peptide derivatives (Reversins), interacting with P-gp/MDR1 with high affinity and selectivity, were developed in our laboratory. Reversins were shown to have a strong inhibitory effect on Pgp/MDR1-mediated drug efflux and the ability to eliminate this drug resistance without any toxic effect in the control cells (67).

Additional methods to prevent MDR in cancer

A distinct mode of MDR reversal includes the use of monoclonal antibodies. Several antibodies were reported to inhibit in vitro Pgp/MDR1 mediated drug efflux^68,69. Antibodies specific for intracellular epitopes are not suitable for in vivo inhibition because immunoglobulins cannot enter cells. On the other hand, antibodies that bind to extracellular epitopes of the transporter may provide a supplement to chemical agents for the reversal of MDR in cancer.

Another potential method to eliminating MDR is the use of macromolecular carriers. Conjugation of drugs to various drug carriers has a wide range of application (induction of immune response, production of epitope-specific antibodies, etc.) and anti-tumor agents can efficiently bind to such carriers. For methods involving inhibition of expression or functional removal of mRNA using antisense oligonucleotides or ribozymes targeted to various multi-drug resistance proteins we refer the reader to recent reviews found in references 7 and 70.

(top of page)

General Pharmacological Role and Medical Significance of the MDR Proteins

In addition to their contribution to multi-drug resistance in cancer, MDR proteins play an important physiological role in the protection of the body against xenobiotics (toxic chemicals) occurring in the environment. The active efflux of these toxic agents accomplishes this function.

The physiological function of MDR1 has been best studied in a knock out mouse model in which one or both of the MDR1 genes has been disrupted. Mice have two genes resembling MDR1: MDR1A and MDR1B. The expression of the two mouse proteins shares the same pattern as the single human MDR1. Although the MDR1A+B knockout mice were viable and fertile with no obvious physiological abnormalities, careful analysis showed that mice lacking both MDR1A and MDR1B transporters were hypersensitive to xenobiotic compounds. On the basis of its localization and ability to mediate the vectorial transport of a range of toxic molecules, we can conclude that MDR1 is involved in the protection of the body against xenobiotic compounds⁷¹.

This notion is strongly supported by the wide substrate specificity of these transporters and the fact that MRP1 also mediates the transport of partially detoxified compounds, such as glutathione and glucuronide conjugates. The role of MDR proteins in the protection against toxic agents is also supported by their tissue distribution. These transporters are expressed in important pharmacological barriers, such as the brush border membrane of intestinal cells, the biliary canalicular membrane of hepatocytes, and the luminal membrane in proximal tubules of the kidney. MDR proteins are also present in the endothelial cell of the brain capillaries, and in the epithelial cells in the choroid plexus, both contributing to the blood-brain barrier (see Figure 7).

FIGURE 7: Multi-drug transporters in the blood-brain barrier

An extremely important question in current pharmacological studies is whether certain drugs can cross these barriers. Since MDR transporters play a key role in these transport processes, information on the interaction between pharmaceuticals and MDR transporters is essential information in drug targeting. Therefore, testing the interaction between compounds and various ABC transporters may significantly increase the understanding of these phenomena.

In addition to their role in pharmacology, several important disease phenotypes are known to be associated with mutations in ABC transporter genes^{4, 7, 22-24}. Among these, the best-known example is cystic fibrosis, one of the most frequent inherited diseases, the primary cause of which is the mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. With respect to the MDR-ABC transporters, there is no known inherited disease connected to the malfunction of either Pgp-MDR1 or MRP1. The phenotype associated with mutations in the MRP2 gene is called the Dubin-Johnson syndrome¹³. Mutations in sPgp cause progressive familial intrahepatic cholestasis (PFIC) type 2, while mutations in the MDR3 gene result in PFIC type 3⁷². Mutations in the MRP6 (which may or may not be a multi-drug transporter) gene cause Pseudoxanthoma elasticum⁷³.

A recent medical application of the MDR proteins is based on a retrovirus-mediated gene transfer of their cDNA- allowing chemo-protection of specific target cells. Chemotherapy is often limited by the general toxicity of the applied drugs. When stem cells of patients with advanced cancer are transduced with a multi-drug resistance protein cDNA-containing retrovirus, the high-level expression of MDR protein in the bone marrow cells should selectively protect this tissue against unwanted toxicity^{7, 37}.

Beyond the physiological, pathological and pharmacological significance of human MDR proteins, an important medical issue arises from the finding that the phenomenon of multi-drug resistance also occurs in pathogenic bacteria (e.g. Mycobacteria) and parasites (Plasmodium and Leishmania), causing diseases like tuberculosis, malaria and leishmaniasis. The multi-drug resistance in these unicellular organisms is also associated with the drug transport activity of ABC transporters that are the non-mammalian homologues of human MDR proteins³⁹. Members of the ABC transporter family are also responsible for multi-xenobiotic resistance in a large variety of organisms living in chemically polluted environments. Thus, this transporter family may have a major role in determining the uptake, bioaccumulation and ultimate biological and health hazards of environmental toxins and man-made pollutants⁷⁵.

References

1. Juliano,R.L.and Ung,v., Biochim.Biophys.Acta, 455, 152-162(1976)
2. Chen,D. et aI., Cell,47, 381-389(1986).
3. Ueda,K. et al., Proc.Natl.Acad.Sci., 84, 3004-3008(1987)
4. Endicott,JA and Ling,v., Annu. Rev.Biochem., 58,137-171 (1989)
5. Higgins,C.E, Ann. Rev.CellBioi.,8, 67-113(1992)
6. Gottesman,MM. and Pastan,I., Annu. Rev.Biochem., 62, 385-427(1993)
7. Gottesman,M.M et aI., Nat. Rev.Cancer;2, 48-58 (2002)
8. Cole,S.P.c.et al., Science,258, 1650-1654(1992)
9. Borst,P.et aI.,J. Natl. Cancer.Inst.,92, 1295-1302(2000)
10. Deeley,R.G.and Cole,S.P.c., Sem.CancerBioI., 8, 193-204(1997)
11. Litman,l et aI., Cell.Mol. Life.Sci.,58, 931-959(2001)
12. Lecureur,V.et aI., Toxicol.,152,203-219 (2000)
13. Paulusma,c.c. et aI., Science,271,1126-1128 (1996)
14. Paulusma,c.c. et aI., Hepatology,25, 1539-1542(1997)
15. Smith,AJ. et aI., J. Bioi.Chem.,275, 23530-23539(2000)
16. Kool,M et aI., CancerRes.,57,3537-3547 (1997)
17. List,A.F.et al., Blood,98, 3212-3220 (2001)
18. Ambudkar,S.v.et al., Annu. Rev.Pharmacol.Toxicol., 39,361-398 (1999)
19. Sarkadi,B. et aI., Semin.CancerBioi.,8, 171-182(1997)
20. Kast,C. and Gros,P., J.Biol.Chem., 272, 26479-26487 (1997)
21. Kast,C. and Gros,P., Biochemistry, 37, 2305-2313(1998)
22. Dean,M et ai, GenomeRes.,11, 1156-1166(2001)
23. Klein,I. et aI., Biochim.Biophys.Acta, 1461, 237-262(1999)
24. Ueda,K. et aI., Sem.CancerBioi.,8, 151-159(1997)
25. Gao,M et aI., J.Bioi.Chem., 271,27782-27787(1996)
26. Bakos,E.et al., J. Bioi.Chem.,271, 12322-12326(1996)
27. Gao,M. et al., J. Bioi.Chem.,273, 10733-10740(1998)
28. Bakos,E.et al., J. Bioi.Chem.,273, 32167-32175(1998)
29. Bakos,E.et aI., J. Cell.Sci.,113,4451-4461(2000)
30. Bates,S.et aI.,1. Bioenerg.Biomembr.,33, 503-511(2001)
31. Ozvegy,C. et aI., Biochem.Biophys.Res.Commun., 285, 111-117(2001)
32. Kage,K. et aI., Int. J. Cancer.,97, 626-630(2002)
33. Rosenberg,M.Eet aI., J. Bioi. Chem., 272,10685-10694 (1997)
34. Chang,G. and Roth,C. B., Science,293, 1793-1800(2001)
35. Krishna,R.et aI., Eur.J. Pharm.Sci.,11,265-283 (2000)
36. Leslie,E.M.et aI., Toxicology 5, 3-23 (2001)
37. Ambudkar,S.v.et aI., Annu. Rev.Pharmacal.Toxico/., 39,361-398 (1999)
38. Lehne,G., Curr.Drug. Targets.,1,85-99 (2000)
39. Rao,v.v. et aI., Proc.Natl.Acad. Sci.,96, 3900-3905 (1999)
40. Bart,1.et aI., CancerTreat.Rev.,26, 449-462 (2000)
41. Zhou,S.et aI., Nat.Med., 9, 1028-1034(2001)
42. Thiebaut,Eet ai, Proc.Natl.Acad. Sci.,84, 7735-7738 (1987)
43. vanHelvoort,A et al., Cell,87, 507-517(1996)
44. Kipp,H. et al., Semin.LiverDis.,20, 339-351(2000)
45. Jonker,J.w. et aI., J. Natl.Cancer Inst.,92, 1651-1656(2000)
46. Sarkadi,B.et aI., J. Bioi. Chem.,267, 4854-4858(1992)
47. Ambudkar,S.v.et aI., Proc.Natl.Acad.Sci., 89, 8472-8476(1992)
48. Scarborough,GA, J. Bioenerg.Biomembr., 27,37-41 (1995)
49. Senior,AE. et al., Semin.Cancer.Bioi.,8, 143-150(1997)
50. vanVeen,H.W.,et aI., J. Mol. Microbiol.Biotechnol., 3,185-192 (2001)
51. Loo,IW. and Clarke,D.M., . Bioi.Chem., 276,36877-36880 (2001)
52. Higgins,C.Eand Gottesman,MM, Trends.Biochem.Sci., 17,18-21(1992)
53. Qian,Y.M.et al., J. Bioi. Chem.,276, 38636-38644(2001)
54. Szakacs,G. et aI., Biochem.J., 356, 71-75(2001)
55. Szab6,K. et al., J. Bioi.Chem.,273,10132-10138 (1998)
56. Hrycyna,C.A et aI., Biochemistry, 37, 13660-13673(1998)
57. Hrycyna,C.A et aI., J. Bioi.Chem.,273, 16631-16634(1998)
58. Sauna,Z.E.et aI., J. Bioi.Chem.,276, 11653-11661(2001)
59. Sauna,Z.E.et al., J. Bioenerg.Biomembr.,33, 481-491 (2001)
60. Szakacs,G. et aI., Patho/.Onco/.Res.,4, 251-257(1998)
61. Robert, J., Eur.J.Gin.Invest., 29,536-545(1999)
62. Honjo,Y. et aI., CancerRes.,61,6635-6639 (2001)
63. Homolya,L.et aI., J. Bioi.Chem.,268, 21493-21496(1993)
64. Homolya,L. et aI., Br.J. Cancer;73, 849-855(1996)
65. Karaszi,E.et aI., Br.J. Haematol.,112,308-314 (2001)
66. Hollo,Z. et aI., FEBSLett.,383, 99-104(1996)
67. Sharom,EJ.et aI., Biochem.Pharmacol.,58, 571-586(1999)
68. Pearson,J.w. et aI., J. Nat'/CancerInst., 83, 1386-1391(1991)
69. Mechetner,E.B.and Roninson,I.B., Proc.Natl.Acad.Sci., 89, 5824-5828(1992)
70. Pawlak,W. et aI.,CancerTreat.Rev.,5. 333-350(2000)
71. Schinkel,AH., Semin.Cancer.Bioi.,B. 161-170(1997)
72. Bull,L.N.et al., Nat. Genet.,18,219-224 (1998)
73. LeSaux,O. et aI., Nat. Genet.,25, 223-227(2000)
74. vanVeen,H.W.et aI., Biochim.Biophys.Acta, 1365,31-36(1998)
75. Kurelec,B., Environ.Health.Per.spect., 105,S4,855-860(1997)

(top of page)