Fig 1.
Schematic diagram of the membrane yeast two-hybrid system (MYTH) to screen for interaction partners of the ABC-transporter BSEP.
The MYTH system is based on a split-ubiquitin approach. The bait, BSEP, is fused to the C-terminus of ubiquitin (Cub) and a transcription factor (LexA-VP16). The preys are soluble or membrane-associated liver proteins, introduced via a cDNA library. They are fused to the N-terminus of ubiquitin (NubG). Upon interaction of BSEP and the liver protein the ubiquitin moieties, which have a low affinity for each other due to a mutation in the N-terminus, come into close proximity. The reassembled ubiquitin is recognized by endogenous ubiquitin specific proteases (UBP). The transcription factor is cleaved off and activates the reporter genes (HIS3, ADE2, lacZ).
Fig 2.
Functional control assay for the membrane yeast two-hybrid with BSEP and self-activation control.
The yeast strain NMY51 was transformed with the BSEP bait construct and control plasmids coding either for a nonsense peptide with the NubI (wild-type ubiquitin) or the NubG (mutated ubiquitin) -tag. Due to its affinity to its C-terminal half the NubI moiety activates the system regardless of bait interaction. Both controls show equal transformation efficiency on selective media (SD-LW) for the plasmids (upper panels). The positive control shows yeast growth on selective medium (SD-LWH) due to the affinity of the wild-type ubiquitin moieties. This confirms expression of the BSEP fusion protein. In contrast, the negative control shows no reporter gene activation based on unspecific interaction of BSEP with the NubG-nonsense peptide. Additionally, no self-activation of the MYTH system with BSEP was detected in a library-scale transformation of the empty pPR3-N library vector.
Table 1.
Protein and gene names of identified interaction partners of BSEP derived from the MYTH assay.
Fig 3.
co-IP / MS/MS identifies BSEP interaction partners.
(A) Immunoprecipitation of BSEP coupled to complex mass spectrometry reveals new interaction partners of BSEP in human liver. Crude canalicular membrane preparations were solubilized in either digitonin or Triton X-100, and immunoprecipitated samples were subjected to MS/MS. For each detergent, protein frequencies from two co-IPs with BSEP antibody are plotted against the respective negative control with naïve mouse IgG. Interaction partners of interest are labeled. For the sake of clarity, proteins found exclusively in either the BSEP or control co-IP are depicted with an MS-score of one instead of zero on the other axis. (B) Immunoblot analysis of co-IPs shown inA.
Fig 4.
BSEP interacts with radixin and the bile acyl-CoA synthetase.
(A) MYTH bait dependency test of radixin1-318 and BACS640-690 against BSEP or a non-interacting control bait. (B) Pull-down of BSEP with radixin1-318, BACS77-690 and AP-2 μ1. Protein interaction partners were immobilized as bait on Strep-Tactin Sepharose and purified BSEP was added. BSEP and the interaction partners were detected by immunoblot analysis with monoclonal antibodies against BSEP or the Strep-tag, respectively. Strep-Tactin Sepharose without bait protein served as negative control. (C) Tag-less BSEP can be pulled down with radixin1-318, but not with full-length, non-activated radixin. Protein interaction partners were immobilized on Strep-Tactin Sepharose. Purified, tag-less BSEP was added to the beads and the complexes were eluted after washing. BSEP and the interacting proteins were detected by immunoblot analysis with monoclonal antibodies against BSEP or the Strep-tag, respectively.
Fig 5.
Schematic overview of the subcellular localization of identified BSEP interaction partners.
The ten novel interaction partners of BSEP are shown in a cellular context. Depicted are the nucleus (N), endoplasmic reticulum (ER), Golgi apparatus (Golgi) and cytoskeleton at the apical membrane (Cyts).
Table 2.
Proteins identified as interaction partners of BSEP in MYTH and co-IP / MS/MS screens.
Fig 6.
Bait dependency test with BSEP interaction partners in the early secretory pathway.
To confirm the interaction with prey proteins from the initial MYTH screen, reporter gene activation was tested individually for each prey against the bait, BSEP, or a non-interacting control bait, large T antigen. The figure shows one representative result for the positively tested preys, which are denoted by their respective gene name.
Fig 7.
Venn diagram of potential BSEP interaction partners obtained with MYTH and co-IP / MS/MS.
The numbers represent the potential BSEP interaction partners identified by MYTH and co-IP / MS/MS and the overlapping set of proteins found in both screens.