Molecular mechanism of Sclerostin-mediated Wnt-inhibition
Final Report Abstract
This project focused on the molecular mechanism by which Sclerostin antagonizes Wnt signaling and also included a detailed analysis how antibodies against Sclerostin can be developed that neutralize its activity in order to restore Wnt signaling. As antibodies of this type are of great value in novel osteoanabolic therapy strategies to fight diseases such as osteoporosis our data will certainly facilitate further development and improvement of such therapeutically valuable antibodies. Our results on the mechanism of the Sclerostin-mediated Wnt inhibition clearly show that Sclerostin can antagonize different Wnt factors (at least in vitro) by a mechanism that is certainly more complex than a simple competition mechanism suggested first by other groups. This is an important finding as for instance expression of Wnt3a is induced by Bone Morphogenetic Proteins, e.g. BMP2 to facilitate bone formation. The downregulation of BMP2-induced bone marker genes in the presence Sclerostin then led to the initial misinterpretation that Sclerostin acts as a direct BMP antagonist, and the predicted failure of Sclerostin to neutralize Wnt3a would be in line with this hypothesis. However, albeit inhibition of Wnt3a activity requires higher concentrations of Sclerostin, it still can fully abolish Wnt3a activity in a reporter gene analysis. During the project we also analyzed a large number of antibodies that were raised against Sclerostin first to support functional mapping of Sclerostin and secondly in order to potentially obtain antibodies that are capable to neutralize its ability to block Wnt signaling. While most antibodies were obtained when the project was funded in the consortium TALOS (EU FP7), the analysis was carried when the project was funded by the DFG. Not only could we functionally map a large number of antibodies, which at that time helped to characterize important functional sites in Sclerostin, but we were also able to identify a highly effective Sclerostin-neutralizing Fab, AbD09097, whose mode of action could be functionally and structurally characterized in detail. With this information at hand we could generate another Sclerostin-neutralizing Fab, AbD14310, via phage display using guided selection. We were able to also determine the structure of this Fab antibody and compare its Sclerostin-binding mechanism with that of AbD09097. Final analyses are currently in progress and the data will be published. For both neutralizing antibodies we have started projects (supported by Bachelor and Master thesis projects) to determine structures of these antibodies in complex to Sclerostin-derived peptides. As the Fab proteins contain sequences at their C-termini that somewhat mimic Sclerostin-derived peptides (see also E4 and E5), production of the Fab proteins in a different format is required. Therefore we started to produce own expression constructs, which will produce the Fab proteins in insoluble form. The Fab antibodies will then be obtained by in vitro refolding procedures, combined with chemically synthesized peptides and crystallized for structure analysis. For AbD09097 a complex bound to a short 9mer peptide could be produced and purified. First crystals of this complex could be obtained, however further optimization is required. A large effort (a Master thesis, two Bachelor thesis and several students doing advanced lab rotation courses) also went into the development of Sclerostin-mimicking peptides that might have the potential capacity to serve as novel compounds for the therapy of the hyperostosis diseases Sclerosteosis and van Buchem. While requiring several rounds of optimization to obtain a working panning process our phage-display selection procedure was finally successful providing a conformationally restrained 14mer peptide cyclized by a disulfide-bond, which was capable to effectively antagonize Wnt1 activity in our Wnt reporter gene assay and as monovalent peptide exhibited binding to LRP6 with low micromolar affinity. This peptide might well serve as a starting point to developed a peptide-based or non-peptidic Sclerostin mimic, but additional work is needed, i.e. the affinity to LRP6 should be improved to be in the double-digit nanomolar range to act as highly effective Sclerostin mimetic with IC50 values to be in the low nanomolar range. This is necessary to avoid unspecific (side) effects due to high concentrations of the compound in a cellular environment. While further improvements in the binding of the peptide to LRP5/6 might be obtained either from structure analysis of a complex of that peptide bound to LRP6, other rational approaches without the need for structure data can also be taken. For instance as already observed in phage ELISA, using the peptide in a bi- or multivalent form might easily enhance its binding affinity due to avidity effects. Another issue that remains to be solved is the usually short serum half times of peptides requiring their modification with PEG or carbohydrate moieties that attenuate cellular uptake and renal clearance. However, as these potential projects present a more applied topic they might receive funding from elsewhere.
Publications
- (2012) Selecting highly structure-specific antibodies using structured synthetic mimics of the cystine knot protein sclerostin. Protein Eng Des Sel 25, 251-259
Back JW, Frisch C, Van Pee K, Boschert V, van Vught R, Puijk W, Mueller TD, Knappik A, Timmerman P
(See online at https://doi.org/10.1093/protein/gzs012) - (2013) Anti-Sclerostin Antibody Inhibits Internalization of Sclerostin and Sclerostin-Mediated Antagonism of Wnt/LRP6 Signaling. PLoS One 8, e62295
van Dinther M, Zhang J, Weidauer SE, Boschert V, Muth EM, Knappik A, de Gorter DJ, van Kasteren PB, Frisch C, Mueller TD, Ten Dijke P
(See online at https://doi.org/10.1371/journal.pone.0062295) - (2013) Mutational analysis of sclerostin shows importance of the flexible loop and the cystineknot for wnt-signaling inhibition. PLoS One 8, e81710
Boschert V, van Dinther M, Weidauer S, van Pee K, Muth EM, Ten Dijke P, Mueller TD
(See online at https://doi.org/10.1371/journal.pone.0081710) - (2015) Crystallization and preliminary X-ray crystallographic analysis of the sclerostin-neutralizing Fab AbD09097. Acta Crystallogr F Struct Biol Commun 71, 388-392
Boschert V, Muth EM, Knappik A, Frisch C, Mueller TD
(See online at https://doi.org/10.1107/S2053230X1500360X) - (2016) The sclerostin-neutralizing antibody AbD09097 recognizes an epitope adjacent to sclerostin's binding site for the Wnt co-receptor LRP6. Open Biol 6
Boschert V, Frisch C, Back JW, van Pee K, Weidauer SE, Muth EM, Schmieder P, Beerbaum M, Knappik A, Timmerman P, Mueller TD
(See online at https://doi.org/10.1098/rsob.160120)
