Heterodimer Formation of the Homodimeric ABC Transporter OpuA

Abstract

Many proteins have a multimeric structure and are composed of two or more identical subunits. While this can be advantageous for the host organism, it can be a challenge when targeting specific residues in biochemical analyses. In vitro splitting and re-dimerization to circumvent this problem is a tedious process that requires stable proteins. We present an in vivo approach to transform homodimeric proteins into apparent heterodimers, which then can be purified using two-step affinity-tag purification. This opens the door to both practical applications such as smFRET to probe the conformational dynamics of homooligomeric proteins and fundamental research into the mechanism of protein multimerization, which is largely unexplored for membrane proteins. We show that expression conditions are key for the formation of heterodimers and that the order of the differential purification and reconstitution of the protein into nanodiscs is important for a functional ABC-transporter complex.

Keywords: ABC-transporter; OpuA; affinity purification; homo- and heterodimeric complexes; mechanism of multimerization; membrane protein; nanodisc reconstitution.

Figure 1

Schematic representation of various OpuA constructs used in this study. The wild-type transporter OpuA is composed of two OpuAA subunits, each carrying a tandem cystathionine-β-synthase (CBS) domain (red) and the ATP-binding domain (orange), and two OpuABC subunits, each carrying a transmembrane domain (TMD) (green), including the scaffold domain (yellow) and the substrate-binding domain (SBD) (blue). (a) Homodimeric OpuA-H, the wild-type OpuA, with a His₆-tag (cyan circle) linked to the SBD; (b) Homodimeric OpuA-S, OpuA tagged with a StrepII-tag (pink hexagon) linked to the SBD; (c) Homodimeric OpuA-SS, OpuA containing a TwinStrepII-affinity tag (double pink hexagon) linked to the SBD; (d) Heterodimeric OpuA-HS, OpuA containing a His₆-tag in one SBD and a StrepII-tag in the other SBD; (e) Heterodimeric OpuA-HSS, OpuA composed of one SBD tagged with His₆-tag and another one with TwinStrepII-tag; (f) Schematic representation of OpuA-HSS in nanodiscs; lipids and MSP1D1 scaffolding protein are shown as grey discs.

Figure 2

Characterization of three differently tagged homodimeric OpuA constructs. (a) Schematic plasmid maps of the expression vectors. OpuAA, gene encoding the ATPase subunit and CBS domains of OpuA; opuABC, gene encoding the TMD and SBD of OpuA; pNisA, nisin-inducible promoter; H, His₆-tag; S, StrepII-tag; double S, TwinStreptII-tag; bent arrows and lollipop symbols represent the promoters and terminators, respectively. (b) SDS-PAGE analysis (12.5% polyacrylamide) of affinity purifications of the three homodimeric OpuA constructs (OpuA-H, OpuA-S, and OpuA-SS). The indicated proteins were purified from crude membrane extracts as explained in the text. The fractions tested were membrane vesicles (V), column flow through (FT), wash (W), and elution fractions (E). (c) Size exclusion chromatography profiles of homodimeric OpuA nanodiscs, using a Superdex 200 increase 10/300 GL column. The chromatograms were normalized to the highest peak. The peak fractions used for further analysis are indicated by the gray shading. (a) and (b) represent the peak fractions of aggregated and empty nanodiscs, respectively. (d) Typical peak fraction of nanodiscs analyzed by 12.5% SDS-PAGE, showing the presence of OpuAA, OpuABC, and the scaffold protein MSP1D1. (e) ATPase activity in the presence (black bars) and absence (white bars) of 62 μM substrate (glycine betaine). Error bars represent the standard deviation of independent triplicates.

Figure 3

Schematic of the purification of the heterodimeric OpuA. (a) L. lactis Opu401 strain carrying plasmids pNZopuAHis and pILopuASS was grown in glucose-M17 broth at 30 °C and the genes were expressed with 0.05% (v/v) nisin A* at 21 °C for 4 h. Three possible OpuA variants are formed in the cell: OpuA-H, OpuA-SS, and OpuA-HSS. (b) Solubilization of membrane vesicles were carried out as described in the Methods section. (c) Three different purification strategies were tested by varying the order of the different steps. The central bold lined square highlights the most efficient protocol.

Figure 4

Optimization of the heterodimer formation under different induction conditions. (a) L. lactis Opu401 strain, harboring plasmids pILopuAS and pNZopuAHis, was propagated at 30 °C in glucose-M17 broth as described in the text. When cultures reached an OD₆₀₀ of 0.5, they were induced at four different nisin A* concentrations: 0.05% (green square), 0.02% (red diamond), 0.01% (orange triangle), and 0.002% (blue circle). Then, cultures (50 mL) were incubated at two different temperatures, 21 or 30 °C, and induction times of 2, 4, and 8 h were tested. (b) Membrane vesicles were obtained, and proteins were purified with Ni²⁺-Sepharose resin. To check the presence of the heterodimeric OpuA variant, final elution fractions were analyzed by Western blot analysis, using monoclonal antibodies directed against the StrepII-tag. The Roman numerals indicate the different nisin concentrations: 0.05% (I), 0.02% (II), 0.01% (III), and 0.002% (IV).

Figure 5

SDS-PAGE (upper panel) and Western blot (two lower panels) analysis of the two-step affinity purification of OpuA. L. lactis Opu401 carrying plasmids pNZopuAHis plus pILopuAS was grown and induced under the following conditions (0.01 % nisin A*; 21 °C during induction; 4 h of induction). Membrane vesicles were obtained as described in the Methods section, and after solubilization of the membranes with 0.5 % (w/v) DDM, the lysate was subjected to two affinity purification steps: Ni²⁺-Sepharose followed by Strep-tactin (left panel) or vice versa (right panel). The following fractions were tested: vesicles (V), flow through (FT), wash (W), and elution fractions (E). Monoclonal antibodies directed against the His₆-tag and StrepII-tag were used, as indicated on the left side of the immunoblots.

Figure 6

Optimization of OpuA reconstitution in nanodiscs. Homodimeric OpuA-SS was purified and reconstituted in nanodiscs formed at different molar ratios and concentrations. (a) Size exclusion chromatography profile of nanodiscs formed at a OpuA/MSP1D1/lipids ratio of 1:20:2000 (dotted line) and 1:20:1000 ratio (solid line); in the latter case, we used a six-times higher concentration of OpuA. Star represents the peak fraction with active OpuA-SS nanodiscs. We verified the dimeric state of OpuA in the peak fraction from the intensity of OpuAA, OpuABC, and MSP1D1 bands on SDS-PAA gels. (b) ATPase activity of OpuA-SS reconstituted in nanodiscs formed at a ratio of 1:20:2000 (I) and 1:20:1000 (II). Black and white bars represent activity in the presence and absence of 62 μM glycine-betaine, respectively. Error bars represent the standard deviation of triplicates.

Figure 7

Purification and characterization of the heterodimeric OpuA-HSS. Membrane vesicles containing a mixture of OpuA-H, OpuA-SS, and OpuA-HSS were obtained as described in the Methods section and subjected to a series of purification steps: (i) Ni²⁺-sepharose purification; (ii) nanodisc reconstitution; (iii) size exclusion chromatography; and (iv) Strep-tactin purification. (a) Size exclusion chromatography profile of the OpuA-H and OpuA-HS nanodiscs. Star represents the peak fraction used for further studies. (b) Coomassie-stained 12.5% SDS-PAGE samples of the different stages of the purification process. The fractions tested were flow through (FT), wash (W), elution (E), and the peak fraction containing nanodiscs (N). Note that OpuABC-H and OpuABC-SS subunits can be distinguished by their different migration in 12.5% SDS-PAGE gels. (c) ATPase activity in the presence (black bar) and absence (white bar) of 62 μM glycine-betaine. Error bars represent the standard deviation of independent triplicates.