br Materials and Methods br Results

Materials and Methods
Results
Discussion PDGF-BB/PDGFR-ββ signaling is a druggable target whose inhibition mitigates liver fibrosis severity (Liu et al., 2011). Several strategies to block PDGF-BB/PDGFR-ββ signaling have been applied in scientific research on liver fibrosis. For instance, PDGF-B monoclonal antibody, adenoviral dominant-negative soluble PDGFR-β and HSCs-specific PDGFR-β small interference RNA have been developed to inhibit PDGF-BB/PDGFR-ββ signaling to ameliorate liver fibrosis in rodent disease models (Chen et al., 2008; Reichenbach et al., 2012; Yoshida et al., 2014). Currently approved medical therapies acting through PDGF-BB/PDGFR-ββ inhibition are based on multi-targeting tyrosine-kinase inhibitors (Ehnman and Östman, 2014; Grassot et al., 2006; Horbert et al., 2015; Westra et al., 2014). Although monoclonal GSK J4 Supplier possess a selective affinity for their target, they also have limitations (Miller et al., 2013; Mócsai et al., 2014). Small-molecule inhibitors offer unique advantages, but they are usually nonselective or less selective (Ehnman and Östman, 2014; Horbert et al., 2015; Mócsai et al., 2014). Thus, the identification or design of small-molecule inhibitors with increased selective affinity to PDGFR-β represents an active area of research. Notably, every tyrosine kinase receptor has a unique extracellular domain for its own ligand, which constitutes a distinct protein–protein interaction interface between a ligand and its tyrosine kinase receptor. If we could identify a small-molecule compound that targets the PDGF-B/PDGFR-β interaction interface, it may serve as a selective inhibitor of PDGFR-β. However, modulating protein–protein interactions with small organic compounds remains enormously challenging because the interface areas are typically large (generally, an average interface area is close to 1150–10,000Å2) and flat (Ivanov et al., 2013). Additionally, the interface areas are typically hydrophobic and often lack the deep grooves that provide an interface for small molecule docking (Ivanov et al., 2013). Utilizing the proangiogenic VEGF/VEGFR protein–protein interface (Gautier et al., 2011) and the oncogenic ERK/ERK protein–protein interface (Herrero et al., 2015), researchers have found drug-like small molecules using structure-based screening, which could specifically inhibit the formation of the VEGFR1/VEGF complex and the dimeric ERK/ERK complex (Gautier et al., 2011; Herrero et al., 2015). These successful reports inspired us to search for a more selective inhibitor of PDGFR-β based on the PDGF-B/PDGFR-β interaction interface. Fortunately, we found that destruxin A5, a natural cyclopeptide, had the ability to block PDGF-BB/PDGFR-ββ signaling without binding to the ATP-binding pocket of human PDGFR-β, as observed with tyrosine kinase inhibitors (Fig. 1). This interesting and puzzling phenomenon attracted our attention. Surprisingly, destruxin A5 was able to bind to the extracellular domain of human PDGFR-β but did not directly interact with PDGF-BB (Fig. 2, Supplementary Fig. 5 and Supplementary Fig. 7). Importantly, destruxin A5 was able to block the interaction between PDGF-B and PDGFR-β and inhibit the dimerization of PDGFR-β that is induced by PDGF-BB stimulation of HSCs (Fig. 2), which indicated that the initiation of PDGF-B/PDGFR-ββ signaling was blocked by destruxin A5. We were also eager to explore the mechanism by which destruxin A5 was able to block the PDGF-B/PDGFR-β interaction. Encouragingly, our results suggested that destruxin A5 binds to human PDGFR-β via interactions with Phe136 and Phe138 (Fig. 3). It should be emphasized that the PDGF-B/PDGFR-β interface is formed by the side chains of Phe136, Phe138, Tyr205, Tyr207 from PDGFR-β and Leu38, Trp40, Ile75, Ile77, Pro82, Phe84 from PDGF-B (Shim et al., 2010). Additionally, the PDGFR-β sequences were highly conserved among mouse, rat and human (Supplementary Fig. 17). The overall PDGFR-β sequence similarity of mouse or rat with human PDGFR-β sequence was 92% and between mouse and rat 98%. Importantly, the PDGFR-β sequences of mouse, rat and human all contain Phe136, Phe138, Tyr205 and Tyr207. These data demonstrate that destruxin A5 is a compound that targets the PDGF-B/PDGFR-β interaction interface to block PDGF-BB/PDGFR-ββ signaling in mouse, rat and human. In the current study, one of our most important finding was the selective inhibitory property of destruxin A5 for PDGF-BB/PDGFR-ββ signaling, as destruxin A5 elicited no inhibitory effect on the kinase activity of the tyrosine kinase and had no direct effects on the interactions with the extracellular domains of human PDGFR-α, VEGFR1, VEGFR2, FLT3, KIT, EGFR or IFN-γR1 (Fig. 4), Moreover, destruxin A5 only slightly inhibited the cell proliferation induced by PDGF-AA, VEGF, FLT3 ligand and KIT ligand, whereas the cell proliferation induced by PDGF-BB was dose-dependently suppressed by destruxin A5 (Supplementary Fig. 9 and Supplementary Fig. 10). Thus, our study confirms the selective inhibitory effect of destruxin A5 on PDGF-BB/PDGFR-ββ signaling. These findings may confer substantial benefits to liver fibrosis patients. Indeed, our in vitro studies validated the suppressive effect of destruxin A5 on PDGF-BB/PDGFR-ββ signaling-mediated activation, proliferation, migration, cell cycle progression and fibrosis-related protein expression in HSCs (Fig. 5). Apoptosis is an important mechanism to reduce the number of activated HSCs (Liu et al., 2011). Thus, we evaluated the effect of destruxin A5 on apoptosis and found that destruxin A5 significantly increased the level of apoptosis in activated HSCs (Fig. 5G and H). Moreover, the antifibrotic effects of destruxin A5 on liver fibrosis were verified in ex vivo and in vivo models, and the mechanism underlying the antifibrotic effect of destruxin A5 was shown to be due to its selective inhibitory role in PDGF-BB/PDGFR-ββ signaling (Figs. 6, 7 and Supplementary Fig. 16). Importantly, the unique therapeutic effect of destruxin A5 on liver fibrosis carries extremely important and potential clinical significance because p-PDGFR-β is wildly expressed in cirrhotic liver tissues from HBV-positive patients (Fig. 8).