Different roles of TM5, TM6, and ECL3 in the oligomerization and function of human ABCG2

Abstract

ABCG2 is a member of the ATP-binding cassette transporter superfamily, and its overexpression causes multidrug resistance (MDR) in cancer chemotherapy. ABCG2 may also protect cancer stem cells by extruding cytotoxic materials. ABCG2 has previously been shown to exist as a high-order homo-oligomer consisting of possibly 8-12 subunits, and the oligomerization domain was mapped to the C-terminal domain, including TM5, ECL3, and TM6. In this study, we further investigate this domain in detail for the role of each segment in the oligomerization and drug transport function of ABCG2 using domain swapping and site-directed mutagenesis. We found that none of the three segments (TM5, TM6, and ECL3) is essential for the oligomerization activity of ABCG2 and that any one of these three segments in the full-length context is sufficient to support ABCG2 oligomerization. While TM5 plays an important role in the drug transport function of ABCG2, TM6 and ECL3 are replaceable. Thus, each segment in the TM5-ECL3-TM6 domain plays a distinctive role in the oligomerization and function of ABCG2.

Figure 1.

(A) Schematic presentation of ABCG2 topology. (B) Schematic presentation of HA-tagged full-length and Myc-tagged ABCG2 constructs. The solid circle and ovals represent HA and Myc tags, respectively. The numbered boxes and lines represent transmembrane (TM) segments and loops (L), respectively. Abbreviations: ECL, extracellular loop; NBD, nucleotide-binding domain; MSD, membrane-spanning domain. (C) Co-IP. HEK293 cells with stable expression of ABCG2^HA-WT were transiently transfected with ABCG2^Myc-TM5–6, ABCG2^Myc-TM1–2, or ABCG2^{Myc-TM1–2/L1*} followed by immunoprecipitation with the Myc antibody and Western blot analysis of the precipitates with HA and Myc antibodies. (D) Quantitative analysis of Co-IP. The expression and Co-IP level of constructs from three independent experiments as shown in panel C were quantified using ScnImage and calculation of the relative ratio of Co-IP to the expression level followed by normalization to that of the positive control construct ABCG2^Myc-TM5–6.

Figure 2.

(A) Schematic presentation of HA-tagged full-length and Myc-tagged ABCG2 constructs. The triplet dots represent cysteine mutations, while the arrowheads indicate mutation of the QXXS motif. Other symbols are the same as described for Figure 1. (B) Co-IP. HEK293 cells with stable expression of ABCG2^HA-WT were transiently transfected with ABCG2^Myc-TM1–4, ABCG2^Myc-WT, ABCG2^Myc-F-TM5*6*, ABCG2^{Myc-F-TM5*6*CL}, or ABCG2^{Myc-F-TM5*6*CLM} followed by immunoprecipitation with the HA or Myc antibody and Western blot analysis of the precipitates with both HA and Myc antibodies. (C and D) Quantitative analysis of Co-IP. The expression and Co-IP level of constructs from three independent experiments as shown in panel B were quantified using ScnImage and calculation of the relative ratio of Co-IP to the expression level followed by normalization to that of the wild-type positive control construct.

Figure 3.

(A) Schematic presentation of HA-tagged full-length and Myc-tagged ABCG2 constructs. The symbols are the same as described for Figure 1. (B) Co-IP. HEK293 cells with stable expression of ABCG2^HA-WT were transiently transfected with ABCG2^Myc-ECL3*, ABCG2^Myc-WT, or ABCG2^Myc-TM1–4 followed by immunoprecipitation with the HA or Myc antibody and Western blot analysis of the precipitates with both HA and Myc antibodies. (C and D) Quantitative analysis of Co-IP. The expression and Co-IP level of constructs from three independent experiments as shown in panel B were quantified using ScnImage and calculation of the relative ratio of Co-IP to the expression level followed by normalization to that of the wild-type positive control construct.

Figure 4.

(A) Schematic presentation of HA-tagged full-length and Myc-tagged ABCG2 constructs. The symbols are the same as described for Figure 1. (B) Co-IP. HEK293 cells with stable expression of ABCG2^HA-WT were transiently transfected with ABCG2^Myc-TM5*L3*, ABCG2^Myc-L3*TM6*, ABCG2^Myc-WT, or ABCG2^Myc-TM1–4 followed by immunoprecipitation with the HA or Myc antibody and Western blot analysis of the precipitates with both HA and Myc antibodies. (C and D) Quantitative analysis of Co-IP. The expression and Co-IP level of constructs from three independent experiments as shown in panel B were quantified using ScnImage and calculation of the relative ratio of Co-IP to the expression level followed by normalization to that of the wild-type positive control construct.

Figure 5.

Drug resistance function of wild-type and mutant ABCG2. HEK293 cells with stable expression of ABCG2^Myc-WT, ABCG2^Myc-ECL3*, ABCG2^Myc-TM5*6*, and vector-transfected control (Vec) were subjected to treatment with mitoxantrone (A) and doxorubicin (B) followed by the MTT assay for relative resistance factors (RRF) as described in Materials and Methods. (C—F) Analysis of RRF to mitoxantrone (C and E) and doxorubicin (D and F) of HEK293 cells with stable expression of ABCG2^Myc-TM5* and ABCG2^Myc-TM6* (C and D) or ABCG2^Myc-TM5*L3* and ABCG2^Myc-L3*TM6* (E and F) using the MTT assay as shown in panels A and B. **p < 0.01.

Figure 6.

Effect of mutations in ABCG2 on drug efflux function. HEK293 cells with stable expression of ABCG2^Myc-WT, ABCG2^Myc-TM5*6*, ABCG2^Myc-ECL3*, ABCG2^Myc-TM5*L3*, ABCG2^Myc-L3*TM6*, and vector-transfected control cells (Vec) were subjected to mitoxantrone (MTZ) accumulation analysis using flow cytometry as described in Materials and Methods. The relative accumulation of mitoxantrone was calculated after normalization to the expression level of ABCG2 and then to that of the vector-transfected control from three independent experiments. **p < 0.01.

Figure 7.

Chemical cross-linking of wild-type and mutant ABCG2. HEK293 cells with expression of ABCG2^Myc-WT, ABCG2^Myc-ECL3*, ABCG2^Myc-TM5*6*, ABCG2^Myc-TM5*, ABCG2^Myc-TM6*, ABCG2^Myc-L3*TM6*, and ABCG2^Myc-TM5*L3* were treated without or with DSS followed by isolation of plasma membranes and Western blot analysis of ABCG2 using the Myc antibody. The molecular weight of each cross-linked oligomer was estimated on the basis of linear regression of native protein markers used. The number of asterisks indicates the number of subunits cross-linked.

Figure 8.

Schematic model of ABCG2 interactions. Three different possible interaction sites contributed by TM5, TM6, and ECL3 are shown. The minimal stable tetrameric unit is shown in a box with dashed lines. The empty circles with numbers represent TM segments.