Journal of Pharmacology And Experimental Therapeutics Fast Forward
First published on April 22, 2008; DOI: 10.1124/jpet.108.138115
0022-3565/08/3261-33-40$20.00
JPET 326:33-40, 2008
METABOLISM, TRANSPORT, AND PHARMACOGENOMICS
Effect of Cysteine Mutagenesis on the Function and Disulfide Bond Formation of Human ABCG2
Yang Liu,
Youyun Yang,
Jing Qi,
Hui Peng, and
Jian-Ting Zhang
Department of Pharmacology and Toxicology, IU Simon Cancer Center, Walther Oncology Center/Walther Cancer Institute, Indiana University School of Medicine, Indianapolis, Indiana
Received for publication
February 15, 2008
Accepted
April 21, 2008.
 |
Abstract
|
|---|
ABCG2 is a member of the ATP-binding cassette (ABC) transporter superfamily. Its overexpression causes multidrug resistance in cancer chemotherapy. Based on its apparent half size in sequence when compared with other traditional ABC transporters, ABCG2 has been thought to exist and function as a homodimer linked by intermolecular disulfide bonds. However, recent evidence suggests that ABCG2 may exist as a higher form of oligomers due to noncovalent interactions. In this study, we attempted to create a cysless mutant ABCG2 as a tool for further characterization of this molecule. However, we found that the cysless mutant ABCG2 is well expressed but not functional. Mapping of the cysteine residues showed that three cysteine residues (Cys284, Cys374, and Cys438) are required concurrently for the function of ABCG2 and potentially for intramolecular disulfide bond formation. We also found that the cysteine residues (Cys592, Cys603, and Cys608) in the third extracellular loop are involved in forming intermolecular disulfide bonds and that mutation of these residues does not affect the expression or drug transport activity of human ABCG2. Thus, we conclude that Cys284, Cys374, and Cys438, which may be involved in intramolecular disulfide bond formation, are concurrently required for ABCG2 function, whereas Cys592, Cys603, and Cys608, potentially involved in intermolecular disulfide bond formation, are not required.
Human ABCG2 is a member of the G subfamily of the ATP-binding cassette (ABC) transporter superfamily, which transports a wide variety of substrates (Xu et al., 2007b
). Overexpression of ABCG2 has been shown to cause increased drug efflux and resistance in model cancer cell lines (Doyle et al., 1998; Miyake et al., 1999; Allen and Schinkel, 2002). The role of ABCG2 in clinical resistance of human acute myeloid leukemia has also been established (Zhang, 2007). In addition, it has been suggested that ABCG2 plays an important protective role for hematopoietic and cancer stem cells (Zhang, 2007).
Human ABCG2 consists of 655 amino acids (GenBank accession no. Q9UNQ0) that make up one nucleotide-binding domain (NBD) at the amino terminal half and one transmembrane domain (TMD) at the carboxyl-terminal half of the protein (Fig. 1). Traditional full ABC transporters such as ABCB1 consist of two NBDs and two TMDs. Thus, ABCG2 has been thought to exist and work as a homodimer covalently linked by disulfide bonds (Kage et al., 2002; Litman et al., 2002; Ozvegy et al., 2002). However, it has been found recently that human ABCG2 exists in drug-resistant cells primarily as a higher form of oligomer containing 12 subunits with noncovalent interactions (Xu et al., 2004). Electron microscopy examination of purified human recombinant ABCG2 also revealed that it is a high form of homo-oligomer possibly with eight identical subunits (McDevitt et al., 2006). It seems that the noncovalent interactions among the subunits in the homo-oligomer are located in the domain, including TM5-loop-TM6 (Xu et al., 2007a).
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 1. Schematic topological and linear structure of ABCG2 with cysteine residues indicated. ABCG2 consists of six transmembrane segments (boxes) and 12 cysteine residues (solid balls) with both its amino and carboxyl termini located in cytoplasm. The cysteine residues indicated with asterisks are functionally important, whereas the underlined ones are important for putative intermolecular disulfide bond formation.
|
|
In a recent study, Henriksen et al. (2005) have demonstrated that one of the cysteine residues (Cys603) in the third extracellular loop between TM5 and TM6 is responsible for the formation of intermolecular disulfide bonds. Similar observations have been made by Wakabayashi et al. (2006). However, it has also been suggested that intermolecular disulfide bonds are oxidized during sample preparations (Xu et al., 2004; Bhatia et al., 2005). Furthermore, in the study by Henriksen et al. (2005), it was found that mutation of the cysteine residues (Cys592 and Cys608) in the third extracellular loop impaired trafficking and function of ABCG2. Wakabayashi et al. (2007) reported that mutation of these two residues decreased ABCG2 stability. However, Bhatia et al. (2005) observed no effect of similar mutations on the trafficking or function of the mutant form of human ABCG2. Furthermore, Kage et al. (2005) found that mutation of Cys603 had no effect on ABCG2 function. Thus, it is not yet clear whether formation of intermolecular disulfide bonds is responsible for ABCG2 dimerization and whether cysteine residues are important for its membrane targeting and function.
Single nucleotide polymorphisms have also been found to potentially affect human ABCG2 activity. Of 20 known nonsynonymous polymorphisms of human ABCG2, none involves a cysteine residue. It is interesting to note that two nonsynonymous mutations, R482G and R482T, resulted in the ability of ABCG2 to transport substrates, such as rhodamine 123, which cannot be transported by the wild-type isoform (Han and Zhang, 2004). In both mutation studies by Henriksen et al. (2005) and Bhatia et al. (2005), the mutant human ABCG2R482G, which was found naturally in the drug-selected cancer cells, was used.
Cysless mutant proteins have been powerful tools for studying ABC transporters such as ABCB1 (Loo and Clarke, 1995) and core ABCC1 (Lee and Altenberg, 2003). Because the cysteine residues of human ABCG2 involved in intermolecular disulfide bond formation may not be essential for ABCG2 function, we tested in this article the possibility of creating a cysless mutant human ABCG2R482G for use in future studies. We also used the insect expression system to eliminate the potential problem of folding in mammalian cells for mutant proteins as observed previously for human ABCC7 (Denning et al., 1992). The insect system has been shown to produce a high level of active proteins, and it is suitable for studies such as substrate transport and ATPase activities of wild-type and mutant ABC transporters (Bakos et al., 1997; Bakos et al., 2000; Ozvegy et al., 2002). We found that three cysteine residues (Cys284, Cys374, and Cys438) that may be involved in intramolecular disulfide bond formation are concurrently required for ABCG2 function, whereas three cysteine residues (Cys592, Cys603, and Cys608) potentially involved in intermolecular disulfide bond formation were not required.
|
Materials and Methods
|
|---|
Materials. Insect Spodoptera frugiperda SF9 cells, pVL1393 plasmid, and BaculoGold Transfection Kit were purchased from BD Biosciences (San Jose, CA). Insect cell culture media and oligonucleotides for site-specific mutagenesis were obtained from Invitrogen (Carlsbad, CA). The QuikChange multisite-directed mutagenesis kit was obtained from Stratagene (La Jolla, CA). Mitoxantrone, rhodamine 123, and Hoechst 33324 were obtained from Sigma-Aldrich (St. Louis, MO). Monoclonal antibody BXP-21 against ABCG2 was obtained from ID Labs Inc. (London, ON, Canada). Fumitremorgin C (FTC) was a gift from Susan Bates (National Cancer Institute, Bethesda, MD). All other chemicals were of molecular biology grade from Sigma-Aldrich or Thermo Fisher Scientific (Waltham, MA).
Engineering of ABCG2R482G cDNA in Baculovirus Vector and Multisite-Directed Mutagenesis. Cysteine wild-type full-length human ABCG2R482G cDNA was excised from pcDNA3-ABCG2 (Xu et al., 2004) using BamHI and XbaI and engineered into pVL1393. To increase expression efficiency, the 5'-untranslated region (UTR) of the cDNA was removed by polymerase chain reaction using the following primers: 5'-CGAGGATCCATGCACCATCACCATCACCATTCTTCCAGTAATGTCGAA-3' (forward) and 5'-GTCTAATCCAGTTGTAGG-3' (reverse). The polymerase chain reaction product lacking the 5'-UTR was digested with BamHI and SpeI and used to replace the corresponding regions of ABCG2R482G with the 5'-UTR sequence in pVL1393.
For multisite-directed mutagenesis, full-length ABCG2R482G cDNA was divided into two fragments by BamHI, SpeI, and XbaI, which were then cloned into a pCR-Blunt vector. The resulting constructs, containing the amino- and carboxyl-terminal half-encoding regions, were used as templates to perform site-directed mutagenesis to change cysteines to alanines using the QuikChange multisite-directed mutagenesis kit according to the manufacturer's instructions. The primers used for specific cysteine to alanine mutations are listed in Table 1. Each fragment containing single or multiple mutations was then used to replace the corresponding wild-type sequence in pVL1393. The mutations in the full-length ABCG2R482G cDNA were confirmed by sequencing.
Culture of Sf9 Cells and Infection. Recombinant baculovirus containing cysteine wild-type or mutant human ABCG2R482G cDNA were generated with the BaculoGold Transfection Kit according to the manufacturer's instruction. Baculovirus harboring ABCG2 were purified and titrated by plaque assay and endpoint dilution. The same multiplicity of infections were used for all constructs to obtain similar expression levels. Baculovirus containing XylE was used as a vector control. Uninfected Sf9 cells were cultured in suspension in insect culture medium at 28°C in spinner flasks. Virus-infected Sf9 cells were cultured on 100-mm dishes or six-well plates.
Western Blot and Immunofluorescence. Two days after viral infection, the Sf9 cells were harvested, washed once with phosphate-buffered saline, and lysed in TNN buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5% Nonidet P-40, 50 mM NaF, 1 mM sodium orthovanadate, 1 mM dithiothreitol (DTT), 0.1% SDS, and 2 mM phenylmethylsulfonyl fluoride). Next, 20 µg of proteins were separated using SDS-polyacrylamide gel electrophoresis (PAGE) in the absence or presence of 100 mM DTT followed by Western blot analysis probed using antibody BXP-21.
Confocal immunofluorescence imaging was conducted as described previously (Yang et al., 2002). In brief, Sf9 cells were cultured on cover glass in six-well plates followed by infection with baculovirus containing cysteine wild-type or mutant ABCG2R482G. The cells were then fixed with acetone/methanol (1:1) and stained with monoclonal antibody BXP-21 40 h after infection. The staining was visualized using fluorescein isothiocyanate-conjugated secondary antibody on a confocal microscope from Bio-Rad (Hercules, CA).
Flow Cytometry. Flow cytometry was used to determine the activity of human ABCG2 to transport substrates mitoxantrone, rhodamine 123, or Hoechst 33342 in Sf9 cells as described previously (Ozvegy et al., 2002). In brief, 40 h after infection with ABCG2-containing baculovirus, Sf9 cells were harvested and resuspended in HEPES-buffered RPMI 1640 medium (120 mM NaCl, 5 mM KCl, 400 µM MgCl2, 40 µM CaCl2, 10 mM HEPES, 10 mM NaHCO3, 10 mM glucose, and 5 mM Na2HPO4). The resuspended cells were preincubated in the presence or absence of 10 µM FTC for 5 min at 37°C followed by a 30-min incubation at 37°C in the presence of 20 µM mitoxantrone, 1 µM rhodamine 123, or 10 µM Hoechst 33342. The cells were then washed and resuspended in ice-cold HEPES-buffered RPMI 1640 medium containing 30 µg/ml propidium iodide followed by flow cytometry analysis. Dead cells were excluded based on propidium iodide staining. Relative activity = (Fmutant – Fvector)/(Fwt – Fvector), where F = peak fluorescence intensity of cells due to accumulation of fluorescent drugs.
|
Results
|
|---|
Characterization of the Cysless Mutant Human ABCG2R482G. It has been shown that mutation of the cysteine residues to alanine in the third extracellular loop (Cys592, Cys603, and Cys608) did not have any effect on ABCG2R482G trafficking and function (Bhatia et al., 2005). This observation forms a basis for testing the possibility of creating a tool, cysless mutant, for studying ABCG2. For this purpose, we mutated all 12 cysteines to alanines (cysless or CL) in ABCG2R482G (Fig. 1). We then expressed the cysless mutant ABCG2R482G along with the cysteine-wild-type ABCG2R482G in insect Sf9 cells using the baculovirus expression system. This expression system has previously been used successfully for functional expression of ABCG2 (Ozvegy et al., 2002). As shown in Fig. 2A, both the cysteine-wild-type and the cysless mutant ABCG2R482G are equally well expressed and targeted onto plasma membranes in Sf9 cells. However, the cysless ABCG2R482G has little activity in facilitating the efflux of mitoxantrone and rhodamine 123 when compared with the cysteine-wild-type ABCG2R482G (Fig. 2B). It seems that the efflux activity observed in Sf9 cells expressing the cysteine-wild-type ABCG2R482G is specific to the ectopic human ABCG2R482G, because the cells harboring the vector control do not have the efflux activity (Fig. 2C), and the drug efflux activity of the cysteine-wild-type ABCG2R482G was sensitive to the ABCG2 inhibitor FTC (Fig. 2C) as well as GF120918 and novobiocin (data not shown). These observations suggest that mutations of some of the 12 cysteine residues caused the loss of ABCG2R482G function.
View larger version (42K):
[in this window]
[in a new window]
|
Fig. 2. Expression and function of wild-type and cysless mutant ABCG2R482G in insect Sf9 cells. A, Western blot and confocal immunofluorescence analyses of Sf9 cells expressing XylE [vector (Vec.) control], wild-type (WT), or CL mutant ABCG2R482G. B, accumulation of mitoxantrone and rhodamine 123 in Sf9 cells expressing XylE, WT, or cysless mutant ABCG2R482G. C, effect of ABCG2 inhibitor FTC on mitoxantrone accumulation in Sf9 cells expressing XylE, WT, or cysless mutant ABCG2R482G. The baseline was the autofluorescence of cells in the absence of drug substrates.
|
|
Mapping Cysteine Residues That Are Functionally Important. To determine which of the 12 cysteine mutations is probably responsible for the loss of function in the cysless mutant ABCG2R482G, we used three constructs with mutations of cysteine residues in various domains of ABCG2R482G. These constructs were named C9-CL, I5-CL, and C4-CL as shown in Fig. 3A. Figure 3B shows that all of these three cysteine-mutant ABCG2s could be successfully expressed and targeted onto plasma membranes. This observation is expected because the CL mutant without any cysteine residues could be successfully expressed in Sf9 cells (Fig. 2). It is interesting to note that only the C4-CL mutant retained the majority of its mitoxantrone efflux activity, whereas both the C9-CL and I5-CL mutants lost most of these activities. These observations suggest that the functionally important cysteine residue(s) may include all or partially the internal five cysteine residues (Cys284, Cys374, Cys438, Cys491, and Cys544), which are commonly mutated in both the nonfunctional C9-CL and I5-CL but are intact in the functional C4-CL.
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 3. Expression and function of ABCG2R482G with cysteine mutations in three different domains. A, schematic linear structure of wild-type and mutant ABCG2R482G with mutated cysteines in three different domains. B, Western blot and confocal immunofluorescence analyses of Sf9 cells expressing the three mutant ABCG2R482G, C9-CL, I5-CL, and C4-CL. C, relative activity of mutant ABCG2R482G, CL, C9-CL, I5-CL, and C4-CL compared with wild-type ABCG2R482G as determined using flow cytometry for mitoxantrone accumulation in Sf9 cells. Data shown are from four independent experiments with S.D.
|
|
To verify whether Cys284, Cys374, Cys438, Cys491, and Cys544 residues are important for ABCG2R482G function, we engineered another construct (N3C4-CL; Fig. 4A) that maintains all of these five cysteine residues but has all of the remaining cysteines mutated to alanine. This mutant is well expressed (Fig. 4B) and retains 
80% of the mitoxantrone efflux activity of the cysteine-wild-type ABCG2R482G (Fig. 4C). Further mutation of N3C4-CL by changing Cys284, Cys374, and Cys438 to alanines (construct N6C4-CL; Fig. 4A) completely eliminated the mitoxantrone efflux activity of human ABCG2R482G (Fig. 4C), although they are also well expressed. Thus, it is possible that the functionally important cysteine residues are Cys284, Cys374, and Cys438.
View larger version (36K):
[in this window]
[in a new window]
|
Fig. 4. Mapping the functionally important cysteine residues. A, schematic linear structure of wild-type and mutant ABCG2R482G. B, Western blot analyses of Sf9 cells expressing XylE (vector; Vec.), wild-type, and mutant ABCG2R482G. C, relative activity of mutant ABCG2R482G, N3C4-CL, and N6C4-CL, compared with wild-type and CL ABCG2R482G as determined using flow cytometry for mitoxantrone accumulation in Sf9 cells expressing these ABCG2s. Data shown are from three independent experiments with S.D.
|
|
To map which of the three cysteine residues are functionally important, we engineered four more constructs by mutating all or partial of the three cysteines Cys284, Cys374, and Cys438 to alanines and generated constructs I3-CL, I2-CL, C284A, and C374A (see Fig. 5A). All of these cysteine-mutant ABCG2R482G constructs could be successfully expressed (Fig. 5B). However, only I3-CL with all three cysteines mutated to alanine lost the majority of its mitoxantrone efflux activity (Fig. 5C). Mutation of one or two of these residues (I2-CL, C284A, and C374A) did not significantly affect the ABCG2R482G activity to transport mitoxantrone. I3-CL with C284A, C374A, and C438A mutations also lost most of its ability to transport another substrate, Hoechst 33342 (Fig. 5D). These observations suggest that Cys284, Cys374, and Cys438 need to work together for the function of ABCG2R482G. In addition, these three cysteines do not seem to affect the substrate specificity.
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 5. Fine tuning the functionally important cysteine residues and substrates selectivity. A, schematic linear structure of wild-type and mutant ABCG2R482G. B, Western blot analyses of Sf9 cells expressing XylE (vector; Vec.), wild-type, and mutant ABCG2R482G. C, relative activity of mutant ABCG2R482G I2-CL, C284A, I3-CL, and C374A compared with wild-type and ABCG2R482G as determined using flow cytometry for mitoxantrone (MX) accumulation in Sf9 cells. Data shown are from three independent experiments with S.D. D, effect of C284A, C374A, and C438A mutations (construct I3-CL) on efflux of Hoechst 33342 (Hoechst) as determined using flow cytometry. Data shown are from four independent experiments with S.D.
|
|
Cysteine Residues That Are Involved in the Formation of Intra- and Intermolecular Disulfide Bonds. It has been thought previously that the formation of intermolecular disulfide bonds is important for ABCG2 homodimerization (Kage et al., 2002; Litman et al., 2002; Ozvegy et al., 2002). Because we have several ABCG2R482G constructs with mutations of various cysteine residues, it would be interesting to determine whether any of these mutations, especially the ones that eliminated ABCG2R482G activity, affect disulfide bond formation as observed on nonreducing SDS-PAGE. As shown in Fig. 6B, the cysteine-wild-type ABCG2R482G did show some dimeric molecules under the nonreducing condition (without DTT). However, the dimeric ABCG2R482G represents only a small fraction of the total ABCG2R482G detected. We also observed that the dimeric ABCG2R482G band is broad (probably with three different populations), probably due to the existence of different intramolecular disulfide bonds, which potentially cause generation of different shapes of molecules with different mobility on the nonreducing SDS-PAGE (Urbatsch et al., 2001) (see also below).
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 6. Reducing and nonreducing SDS-PAGE analyses of wild-type and cysteine-mutant ABCG2R482G. A, schematic linear structure of wild-type and mutant ABCG2R482G with cysteine mutations in three different domains. B, nonreducing and reducing SDS-PAGE of wild-type and cysteine-mutant ABCG2R482G CL, C9-CL, I5-CL, and C4-CL. C, nonreducing SDS-PAGE of cysteine-mutant ABCG2R482G constructs L3-CL, I2-CL, C284A, and I3-CL compared with wild-type and CL mutant.
|
|
The disulfide-bond-linked dimeric ABCG2R482G on the nonreducing SDS-PAGE is completely eliminated with mutations in the CL and C9-CL (Fig. 6B, lanes 2 and 3). The dimeric ABCG2R482G on nonreducing SDS-PAGE was observed with the I5-CL mutant (Fig. 6B, lane 4). Because 1) C9-CL has only three cysteines at the amino terminus (Cys43, Cys55, and Cys119) and does not form any dimers and 2) I5-CL has seven cysteines with three at the amino terminus, same as C9-CL (Cys43, Cys55, Cys119), four at the carboxyl terminus (Cys592, Cys603, Cys608, and Cys635), and could form dimers of fast mobility, we conclude that the formation of the intermolecular disulfide bond probably requires the last four cysteine residues (see Fig. 3A). In light of previous findings (Bhatia et al., 2005; Henriksen et al., 2005), we propose that the formation of the intermolecular disulfide bond will probably require the cysteine residues located in the third extracellular loop linking TM5 and TM6. To test this hypothesis, we engineered another construct that has all three cysteine residues in the third extracellular loop mutated to alanine (C592A, C603A, and C608A) to determine whether dimers linked by intermolecular disulfide bonds exist with this mutant. As shown in Fig. 6C, lane 3, this mutation (construct L3-CL; see Fig. 6A) essentially eliminated dimers on the nonreducing SDS-PAGE. Thus, it is possible that the cysteines in the third extracellular loop are responsible for the formation of intermolecular disulfide bonds.
However, it is interesting to note that the nonreducing SDS-PAGE profile of the dimeric I5-CL mutant is different from the cysteine-wild-type ABCG2R482G (Fig. 6B, compare lanes 1 and 4). Whereas the dimeric cysteine-wild-type ABCG2R482G band is broad, possibly with three different populations, the I5-CL seems to have only a distinct one-dimer band of fast mobility on the nonreducing SDS-PAGE (Fig. 6B, lane 4). This difference may be due to the mutation of other cysteine residues in the I5-CL that could potentially affect the formation of intramolecular disulfide bonds and, thus, the shape of the molecule and mobility on the nonreducing SDS-PAGE. To test this possibility, we studied other mutant I2-CL, C284A, and I3-CL using the nonreducing SDS-PAGE. As shown in Fig. 6C, the mutant I3-CL had a single dimeric ABCG2R482G band of fast mobility on the nonreducing SDS-PAGE (Fig. 6C, lane 8), similar to the I5-CL mutant (see Fig. 6B). However, the mutant C284A had an additional dimeric band of medium mobility, whereas the I2-CL has an additional dimeric band of slow mobility (Fig. 6C, lanes 6 and 7). The dimeric bands with slow and medium mobility may represent dimeric ABCG2R482G containing both inter- and intramolecular disulfide bonds. Because the mutant construct C284A has the wild-type Cys374 and Cys438 residues and both mutants I2-CL and I3-CL have these two cysteine residues mutated, it is possible that these two cysteines are involved in the formation of intramolecular disulfide bonds that result in the dimeric protein of medium mobility on nonreducing SDS-PAGE. Because the difference between I2-CL and I3-CL is that Cys284 is mutated in the later, but not the former, it is possible that the dimeric band of slow mobility observed with I2-CL may be due to the existence of Cys284. It is also noteworthy that this dimeric band of slow mobility was also observed, albeit at a low level, with the mutants C4-CL (Fig. 6B, lane 5) and L3-CL (Fig. 6C, lane 3), which contain wild-type Cys284, but not with C9-CL and I5-CL, which have a mutated Cys284, consistent with the conclusion that Cys284 may be responsible for the production of the dimeric protein of slow mobility on the nonreducing SDS-PAGE.
|
Discussion
|
|---|
In this study, we have created a cysless mutant of human ABCG2R482G and shown that mutating all 12 cysteine residues effectively eliminated ABCG2R482G activity. Further mapping studies showed that three cysteine residues (Cys284, Cys374, and Cys438) are essential for human ABCG2R482G function. However, mutating these residues, unless mutated concurrently, did not affect ABCG2R482G function. In a previous study, Kage et al. (2005) also found that single mutation of any of these cysteine residues did not affect ABCG2 transport activity. However, Kage et al. (2005) did not perform a study using mutations in combination. Based on these findings, we conclude that three cysteine residues (Cys284, Cys374, and Cys438) are required concurrently for ABCG2R482G activity.
Human ABCG2 has 12 cysteine residues with five in the amino terminal NBD and seven in the carboxyl-terminal TMD. The three functionally important cysteine residues (Cys284, Cys374, and Cys438) are located in the center of the molecule with Cys284 and Cys374 in the NBD and Cys438 in the TMD. Both Cys284 and Cys374 are not directly located in the Walker A (amino acid residues 80–88) or B (residues 206–211) motifs and, thus, may not be directly involved in affecting nucleotide binding. The observation that the single mutation of these two cysteine residues (C284A and C374A) did not substantially affect ABCG2R482G activity confirms the fact that their mutations did not severely influence the nucleotide-binding activity. The third residue C438 is located in TM2. Mutation of Cys438 to alanine along with Cys374 (construct I2-CL) did not affect ABCG2R482G activity. C438A mutation alone also did not affect ABCG2R482G function (Kage et al., 2005). Hence, these three cysteine residues need to work together to provide the protein with functionality.
The results shown in Fig. 6 also suggest that the Cys284, Cys374, and Cys438 residues may be involved in the formation of intramolecular disulfide bonds. Mutating Cys374 and Cys438 (construct I2-CL) effectively eliminated formation of the dimeric protein of medium mobility on nonreducing SDS-PAGE, whereas mutating Cys284 eliminated the dimeric protein of slow mobility. However, it is not yet known whether the intramolecular disulfide bond is important for ABCG2R482G function. Considering the fact that all of these three cysteine residues are functionally important, the formation of the potential intramolecular disulfide bonds by these residues may be important for the ABCG2R482G function. Clearly, further studies are needed to determine whether the disulfide bonds formed by these three residues play any role in ABCG2 function.
We also found that three cysteine residues in the third extracellular loop (Cys592, Cys603, and Cys608) are probably involved in the formation of intermolecular disulfide bonds, consistent with previous findings (Bhatia et al., 2005; Henriksen et al., 2005; Kage et al., 2005). However, we found no evidence that these three cysteine residues are required for the expression and drug efflux activity of human ABCG2R482G, in contrast to observations by Henriksen et al. (2005), but consistent with those by Bhatia et al. (2005) and Kage et al. (2005). The reason for the discrepancy among these studies with respect to the effect of mutations on ABCG2R482G function is not yet known. The only difference between these studies is the expression system used, which may or may not be the cause of the difference in function. Whereas HEK293 cells were used by Henriksen et al. (2005), HeLa, PA317, and Sf9 cells were used by Bhatia et al. (2005), Kage et al. (2005), and in our study, respectively.
However, it is noteworthy that the dimers linked by intermolecular disulfide bonds observed on the nonreducing SDS-PAGE represent only a very small fraction of ABCG2R482G expressed in this study. The vast majority of ABCG2R482G does not have intermolecular disulfide bonds. It also seems that the cysteine wild-type ABCG2R482G has several species of dimers on the nonreducing SDS-PAGE, suggesting the existence of several molecules with different intramolecular disulfides (see discussion above). It has previously been observed that the formation of intermolecular disulfide bonds may be due to oxidation during sample preparation (Xu et al., 2004). Based on these observations and the finding that elimination of intermolecular disulfide bonds by mutations does not affect ABCG2R482G activity, we conclude that these intermolecular disulfide bonds of ABCG2R482G may not exist in vivo and may not be necessary for the transport function of ABCG2R482G.
In summary, we have shown in this study that, unlike ABCB1 (Loo and Clarke, 1995) and core ABCC1 (Lee and Altenberg, 2003), a cysless mutant of human ABCG2 is not functional and, thus, cannot be used as a tool for future studies. Through this study, we have identified three functionally important cysteine residues (Cys284, Cys374, and Cys438) that may need to work together to sustain the functionality of human ABCG2R482G. Furthermore, the cysteine residues in the third extracellular loop are responsible for the minor population of dimeric ABCG2R482G molecules observed on nonreducing SDS-PAGE, and their mutations do not affect ABCG2 function.
|
Acknowledgements
|
|---|
We appreciate Jeff Russ for editorial proofreading.
|
Footnotes
|
|---|
This work was supported in part by National Institutes of Health Grants CA120221 and CA113384 and by Department of Defense Grant DAMD170010297. Y.Y. was supported in part by the National Research Service Award T32 HL07910 from the National Institutes of Health.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
doi:10.1124/jpet.108.138115.
ABBREVIATIONS: ABC, ATP-binding cassette; NBD, nucleotide-binding domain; TMD, transmembrane domain; FTC, fumitremorgin C; UTR, untranslated region; PAGE, polyacrylamide gel electrophoresis; DTT, dithiothreitol; CL, cysless; GF120918, N-(4-[2-(1,2,3,4-tetrahydro-6,7-dimethoxy-2-isoquinolinyl)ethyl]phenyl)-9,10-dihydro-5-methoxy-9-oxo-4-acridine carboxamide); WT, wild type.
Address correspondence to: Dr. Jian-Ting Zhang, Department of Pharmacology and Toxicology, Indiana University School of Medicine, 1044 W. Walnut St., R4-166, Indianapolis, IN 46202. E-mail: jianzhan{at}iupui.edu
|
References
|
|---|
Allen JD and Schinkel AH (2002) Multidrug resistance and pharmacological protection mediated by the breast cancer resistance protein (BCRP/ABCG2). Mol Cancer Ther 1: 427–434.[Free Full Text]
Bakos E, Evers R, Calenda G, Tusnady GE, Szakacs G, Varadi A, and Sarkadi B (2000) Characterization of the amino-terminal regions in the human multidrug resistance protein (MRP1). J Cell Sci 113: 4451–4461.[Abstract]
Bakos E, Klein I, Welker E, Szabo K, Muller M, Sarkadi B, and Varadi A (1997) Characterization of the human multidrug resistance protein containing mutations in the ATP-binding cassette signature region. Biochem J 323 (Pt 3): 777–783.[Medline]
Bhatia A, Schafer HJ, and Hrycyna CA (2005) Oligomerization of the human ABC transporter ABCG2: evaluation of the native protein and chimeric dimers. Biochemistry 44: 10893–10904.[CrossRef][Medline]
Denning GM, Anderson MP, Amara JF, Marshall J, Smith AE, and Welsh MJ (1992) Processing of mutant cystic fibrosis transmembrane conductance regulator is temperature-sensitive. Nature 358: 761–764.[CrossRef][Medline]
Doyle LA, Yang W, Abruzzo LV, Krogmann T, Gao Y, Rishi AK, and Ross DD (1998) A multidrug resistance transporter from human MCF-7 breast cancer cells. Proc Natl Acad Sci U S A 95: 15665–15670.[Abstract/Free Full Text]
Han B and Zhang JT (2004) Multidrug resistance in cancer chemotherapy and xenobiotic protection mediated by the half ATP-binding cassette transporter ABCG2. Curr Med Chem Anticancer Agents 4: 31–42.[CrossRef][Medline]
Henriksen U, Fog JU, Litman T, and Gether U (2005) Identification of intra- and intermolecular disulfide bridges in the multidrug resistance transporter ABCG2. J Biol Chem 280: 36926–36934.[Abstract/Free Full Text]
Kage K, Fujita T, and Sugimoto Y (2005) Role of Cys-603 in dimer/oligomer formation of the breast cancer resistance protein BCRP/ABCG2. Cancer Sci 96: 866–872.[CrossRef][Medline]
Kage K, Tsukahara S, Sugiyama T, Asada S, Ishikawa E, Tsuruo T, and Sugimoto Y (2002) Dominant-negative inhibition of breast cancer resistance protein as drug efflux pump through the inhibition of S-S dependent homodimerization. Int J Cancer 97: 626–630.[CrossRef][Medline]
Lee SH and Altenberg GA (2003) Transport of leukotriene C4 by a cysteine-less multidrug resistance protein 1 (MRP1). Biochem J 370: 357–360.[CrossRef][Medline]
Litman T, Jensen U, Hansen A, Covitz KM, Zhan Z, Fetsch P, Abati A, Hansen PR, Horn T, Skovsgaard T, et al. (2002) Use of peptide antibodies to probe for the mitoxantrone resistance-associated protein MXR/BCRP/ABCP/ABCG2. Biochim Biophys Acta 1565: 6–16.[Medline]
Loo TW and Clarke DM (1995) Membrane topology of a cysteine-less mutant of human P-glycoprotein. J Biol Chem 270: 843–848.[Abstract/Free Full Text]
McDevitt CA, Collins RF, Conway M, Modok S, Storm J, Kerr ID, Ford RC, and Callaghan R (2006) Purification and 3D structural analysis of oligomeric human multidrug transporter ABCG2. Structure 14: 1623–1632.[Medline]
Miyake K, Mickley L, Litman T, Zhan Z, Robey R, Cristensen B, Brangi M, Greenberger L, Dean M, Fojo T, et al. (1999) Molecular cloning of cDNAs which are highly overexpressed in mitoxantrone-resistant cells: demonstration of homology to ABC transport genes. Cancer Res 59: 8–13.[Abstract/Free Full Text]
Ozvegy C, Varadi A, and Sarkadi B (2002) Characterization of drug transport, ATP hydrolysis, and nucleotide trapping by the human ABCG2 multidrug transporter. Modulation of substrate specificity by a point mutation. J Biol Chem 277: 47980–47990.[Abstract/Free Full Text]
Urbatsch IL, Gimi K, Wilke-Mounts S, Lerner-Marmarosh N, Rousseau M-E, Gros P, and Senior AE (2001) Cysteines 431 and 1074 are responsible for inhibitory disulfide cross-linking between the two nucleotide-binding sites in human P-glycoprotein. J Biol Chem 276: 26980–26987.[Abstract/Free Full Text]
Wakabayashi K, Nakagawa H, Adachi T, Kii I, Kobatake E, Kudo A, and Ishikawa T (2006) Identification of cysteine residues critically involved in homodimer formation and protein expression of human ATP-binding cassette transporter ABCG2: a new approach using the flp recombinase system. J Exp Ther Oncol 5: 205–222.[Medline]
Wakabayashi K, Nakagawa H, Tamura A, Koshiba S, Hoshijima K, Komada M, and Ishikawa T (2007) Intramolecular disulfide bond is a critical check point determining degradative fates of ATP-binding cassette (ABC) transporter ABCG2 protein. J Biol Chem 282: 27841–27846.[Abstract/Free Full Text]
Xu J, Liu Y, Yang Y, Bates S, and Zhang JT (2004) Characterization of oligomeric human half-ABC transporter ATP-binding cassette G2. J Biol Chem 279: 19781–19789.[Abstract/Free Full Text]
Xu J, Peng H, Chen Q, Liu Y, Dong Z, and Zhang JT (2007a) Oligomerization domain of the multidrug resistance-associated transporter ABCG2 and its dominant inhibitory activity. Cancer Res 67: 4373–4381.[Abstract/Free Full Text]
Xu J, Peng H, and Zhang JT (2007b) Human multidrug transporter ABCG2, a target for sensitizing drug resistance in cancer chemotherapy. Curr Med Chem 14: 689–701.[CrossRef][Medline]
Yang Y, Chen Q, and Zhang JT (2002) Structural and functional consequences of mutating cysteine residues in the amino terminus of human multidrug resistance-associated protein 1. J Biol Chem 277: 44268–44277.[Abstract/Free Full Text]
Zhang JT (2007) Biochemistry and pharmacology of the human multidrug resistance gene product, ABCG2. Zhong Nan Da Xue Xue Bao Yi Xue Ban 32: 531–541.[Medline]
![[Feedback]](/icons/banner/feedbackACT.gif)
![[For Subscribers]](/icons/banner/subscriberACT.gif)
![[Archive]](/icons/banner/archiveACT.gif)
![[Search]](/icons/banner/searchACT.gif)
![[Contents]](/icons/banner/tocACT.gif)
Top
Abstract
Materials and Methods
