(384b) Providing a Molecular Basis to Modeling Protein Adsorption to Multimodal Carbohydrate-Based Ligands

Authors: 
Chundawat, S. - Presenter, Rutgers, State University of New Jersey
Brady, S. K., Vanderbilt University
Narayanan, V., Rutgers, The State University of New Jersey
Lang, M. J., Vanderbilt University
Recognition and binding of proteins to carbohydrate-based multimodal ligands plays a critical, but poorly understood, role in both fundamental cellular (e.g., cell glycocalyx and viral protein binding) and industrial processes (e.g., antibody drug bioseparations). Manufacturing costs of biopharmaceutical drugs can be drastically reduced through development of inexpensive multimodal (or mixed-mode) chromatographic bioprocesses for protein isolation and purification. However, unlike classical affinity or ion exchange based bioseparations that have been well studied, multimodal chromatographic (MMC) systems are still in their developmental stages [1]. Inexpensive and biocompatible MMC matrices/ligands derived from carbohydrates have tremendous potential for industrial bioseparations. However, we currently have a poor understanding of the molecular mechanisms driving protein binding to traditional carbohydrate (e.g., sepharose) and/or glycoprotein (e.g., concanavalin A) derived chromatographic matrices. This has been further stymied by the lack of suitable experimental tools and predictive adsorption models that explain the heterogeneous nature of protein-carbohydrate interactions. One of the challenges with traditional methods (e.g., solution depletion) used for monitoring protein adsorption is that bulk ensemble measurements provide limited insight into the underlying molecular mechanisms driving protein-ligand binding. Studying binding interactions of proteins and carbohydrate-based ligands at the single-molecule level is critical to explaining observed macroscale measurements, which will ultimately lead to the development of mechanistically-relevant protein macroscale adsorption models.

Here we study the adsorption of carbohydrate-binding proteins to two distinct model polysaccharide substrates with identical molecular composition but varying surface hydrophilicity using complementary bulk and single-molecule based adsorption techniques. We have modified a recently developed optical tweezers force spectroscopy technique [2] to characterize the single molecule protein binding-unbinding force dynamics to polysaccharide surfaces and probed the protein-ligand bond lifetimes at the single molecule level. Analysis of our records was used to identify trends in the force-binding lifetime datasets to find that the unbinding force and total binding lifetimes reflects the underlying heterogeneity of the available binding sites for carbohydrate-derived ligands. This distribution reflects the spectrum of varying affinity sites present on the polysaccharide surface and ultimately provides a molecular basis for using various multi-site Langmuir-type adsorption models to characterize protein adsorption to polysaccharide surfaces. Interestingly, simple regression analysis of bulk adsorption data using multi-site Langmuir-type models reveals that traditional protein bulk adsorption measurements are unable to provide data with sufficiently high resolution to discriminate between the low and high affinity protein binding sites for polysaccharides if the difference in the protein-carbohydrate ligand unbinding force is relatively small. In summary, we now have the necessary tools to systematically characterize protein adsorption to multimodal carbohydrate-derived ligands for optimizing industrial bioseparations using a first-principles approach.

References

[1] Pinto, I. F., Aires-Barros, M. R. & Azevedo, A. M. Multimodal chromatography: debottlenecking the downstream processing of monoclonal antibodies. Pharm. Bioprocess. 3, 263â??279 (2015).

[2] Brady, S. K., Sreelatha, S., Feng, Y., Chundawat, S. P. S. & Lang, M. J. Cellobiohydrolase 1 from Trichoderma reesei degrades cellulose in single cellobiose steps. Nature Communications. 6, 10149 (2015).