Targeting bag-1S/C-raf interaction for therapeutic intervention in cancer

Abstract

In this context, this study aims to map the interaction surface of the complex formed by Bag-1 and C-Raf, which was accomplished through the use of both molecular and structural techniques. For this, the three dimensional structure and domain architecture of the small isoform of Bag-1 were first examined by Hydrogen-Deuterium Exchange Mass Spectrometry (HDX-MS), and the regions on Bag-1S that can accommodate small molecule binding were probed to assess its "druggability". To this end, Bag-1S was first purified from cell lysate using Ni-NTA affinity purification through the incorporated hexahistidine tag, and subsequently, the tag was cleaved with TEV protease. A subsequent Ni-NTA purification was carried out in a flow-through mode to collect Bag-1S separate from His-tagged TEV enzyme and impurities that contained neighboring histidines. The purified Bag-1S showed an apparent 33-kDa band in gel electrophoresis. Sample purity was estimated at over 90% using ImageJ analysis of SDS-PAGE gel. To monitor the deuteration level of the Bag-1S isoform, HDX-MS experiments with five time-points that ranged from 12 s to 24 h were carried out and revealed the identification of ~150 peptides of the Bag-1S with a sequence coverage of 98%. Using HDX-MS data, peptide-specific deuterium incorporation rates were projected onto the modeled structure of Bag-1S and deuterium uptake was analyzed on the Bag-1S full-length structure. BAG domain exhibited a more solvent-protected and stabilized structure compared to the UBL (ubiquitin-like) domain. While turn regions are more labile, the regions where the helical conditions exist remained unexchanged during the entire monitored time. Multiple interaction partners of the adapter protein Bag-1 engage specifically with the BAG domain. Interestingly, the interaction sites of these partners coincide with the regions that are most solvent-protected. The interaction site is supposed to be located in the solvent-protected region of the BAG domain, which is surrounded by charged and hydrophilic regions. This solvent-protected region in the BAG domain likely possesses an interaction region, revealing a potential "druggable" binding site. To further evaluate the binding stoichiometry of the Bag-1S with C-Raf, cross-linking assays were performed in the subsequent experiments. To accomplish this, C-Raf and Bag-1S proteins were affinity-purified, which was followed by the combination of purified proteins to form an in vitro complex. Covalent coupling of the formed complexes was then performed with a cross-linking agent, DSS (disuccinimidyl suberate). According to the results obtained after immunoblotting of cross-linked samples, Bag-1S and C-Raf formed a 2:2 stoichiometric complex, suggesting that Bag-1S might contribute to C-Raf activation by triggering its dimerization. After the Bag-1S/C-Raf interaction was affirmed and stoichiometrically tested, on-membrane in vitro binding experiments were conducted to selectively identify the interface of the complex. The purified C-Raf was immobilized on a PVDF membrane and incubated with purified Bag-1S in vitro. Bag-1S-bound peptides were recovered and analyzed by LC-MS/MS after the formed complex was subjected to limited tryptic digestion on the membrane. A 20-amino acid length peptide was identified as a plausible C-Raf interacting peptide in the BAG domain of Bag-1S. Further, an in silico docking study was also conducted using the protein structure of the kinase domain of C-Raf (PDB ID 5OMV) and the modeled full length protein structure of Bag-1. In some of the poses with the lowest docking energy score, K137, T140, Q144, K149, and L156 residues of Bag-1S were found to occupy the Bag-1/C-Raf binding site. This region coincides with the plausible "druggable" interaction site identified in HDX-MS and on-membrane in vitro binding experiments. Site-directed mutagenesis experiments were then carried out to confirm the identified binding interface and to evaluate if mutations in the determined peptide sequence affect the binding of Bag-1S/C-Raf or not. Upon mutagenesis, K149A and L156R substitutions significantly decreased the endogenous levels of p-C-Raf (S338) and p-MEK1/2 (217/221) in MCF-7 cells. Consistently, TAP-pull down experiments demonstrated that these substitutions impaired the interaction of Bag-1S with C-Raf, without affecting its HSP70 contact. They also led to a significant decrease in the survival of MCF-7 cells compared to wild-type Bag-1S. In addition, while these mutations did not affect the interaction of Bag-1S with its known direct interaction partners, Bcl-2 and HSP70, they resulted in the disruption of its interaction with the complexes involved in other regulatory cell survival pathways, including B-Raf, Beclin 1, and Akt. Subsequent in vitro binding experiments did not reveal a binary interaction of Bag-1S with either Beclin 1 or B-Raf, at least under our experimental conditions. Therefore, it has been hypothesized that the formation of a Bag-1/Beclin 1 or Bag-1/B-Raf complex might require the presence of C-Raf as a mediator. Further, Bag-1S interacting C-Raf region was identified by on-membrane in vitro binding experiment coupled with LC-MS/MS. Four different peptides derived from native Bag-1 and C-Raf sequences corresponding to the plausible interaction segments of the complex were designed and then synthesized by using solid-phase synthesis. The ability of the peptides to hamper the formation of a Bag-1S/C-Raf complex was tested in vitro. Of these peptides, Pep 3 that targets C-Raf binder region of Bag-1S significantly altered Bag-1S/C-Raf interaction. Pep 3 not only impeded the binary interaction of C-Raf with Bag-1S but also disrupted BAG-associated complexes of Bag-1 in TAP pull-down experiments. Inhibition of multiple Bag-1S interactions afforded by Pep 3 bolsters its potential to impair the prolonged survival of cancer cells. We therefore not only affirmed that this region on C-Raf is responsible for Bag-1 binding, but also discovered a novel peptide inhibitor targeting Bag-1S, which has the potential to be improved for cancer therapy.