Fusion proteins are a class of proteins where two or more protein domains (or segments of sequences) are fused and integrated into one entity. In laboratories, fusion proteins are routinely constructed by researchers using recombinant DNA technology to label the protein of interest, facilitate affinity-based protein purification, improve protein solubility and stability, enhance consecutive enzyme reaction rates, and create innovative therapeutic proteins. The addition of extra modules confers new functions and often results in a fusion protein with distinct overall structure.
However, the three-dimensional structure of fusion proteins is not always available in the Protein Data Bank (PDB), which hinders the structure-function studies as well as rational engineering of these proteins. The lack of structural data is mostly due to the relatively large size of fusion proteins (except proteins fused with small tags) and the intrinsic structural heterogeneity at the interface or inter-domain linking region, which prevents the whole protein from being crystallized. Fortunately, in many cases, the structure of individual fusion partner is well known and this structural information can be used to direct computational modeling process. Our structure modeling procedures begin with selection of the two protein components to be fused, whose structures are either known or can be computationally modelled using properly selected templates. The connection of two fusion partners with linkers or domain insertion strategies are then employed to create an initial fusion model. After that, final models are generated by computationally sampling the conformations of regions that bridge the fusion partners, allowing the protein domains to move, rotate and interact to reach a stable conformation.
Profacgen takes advantage of computational modeling methods to help customers predict the three-dimensional structure of fusion proteins of interest. We have extensive experiences with computational protein modeling techniques, always ensuring that the fusion partners are physiochemical compatible and the predictions testable in laboratory experiments. The resultant models are all quality verified and can be used in downstream computational simulation, as well as for the design and engineering of novel fusion proteins with desired functionalities.
Homology modeling for fusion partners without known structures
Knowledge-based fusion strategy
Intensive sampling of flexible regions between fusion components
Refinement of final models to eliminate steric clashes
Inclusion of ligands, cofactors, and water in the model
Prediction of physical properties
Quality assessment by multiple criteria
Modeling protocol customizable based on experimental design
We provide the service in a fully customizable manner to meet the specific requirements from the customers. Please feel free to contact us for more details about our fusion protein modeling service.