Biologic Encapsulation by Tuning their Organic Solubility with Amphiphilic Polymers

Project Summary Biologic encapsulation in polymers has long been utilized to control their release. While this approach may be well suited for hydrogels with an aqueous solvent, encapsulation within hydrophobic polymers is complicated by solvent incompatibility. To address this, the double emulsion technique provides a standard method to emulsify

proteins and polymers together and form encapsulated microparticles. However, this results in low encapsulation efficiency and burst release profiles due to non-specific surface adsorption of the protein to the polymer’s surface. Meanwhile, hydrophobic small molecule drugs are routinely encapsulated into hydrophobic polymers by co-

dissolving drug and polymer in the same organic solvent prior to particle formation. This allows for evenly dispersed drug in the polymer matrix and sustained release profiles. Ideally, bioactive proteins would also be co- dissolved in organic solvent such as dichloromethane to provide the same matching of polymer-protein miscibility

and even dispersion in the polymer matrix. Here, we aim to address this unmet need by wrapping bone morphogenic protein 2 (BMP-2) with stabilizing random heteropolymers to form amphiphilic polymer-protein hybrids. We hypothesize that precise tuning of the polymer chemistry will protect BMP-2 from denaturing while endowing it with solvent/polymer miscibility. To

facilitate this complex formulation process, we will use active machine learning on a robotic platform through an established Design-Build-Test-Learn workflow (Advanced Materials 2022). In Aim 1, we will implement multiple cycles of this workflow to continuously evolve a Gaussian process regressor to predict new generations of

copolymer designs with each cycle. Experimentally, the model will be trained on BMP-2 release data from microparticles. In Aim 2, lead BMP-2 polymer-protein hybrids will be used to form BMP-2 encapsulated microparticles, optimized for tunable release, and compared to microparticles prepared by double emulsion.

Also, fluorescently labelled BMP-2 will be encapsulated and visualized by confocal microscopy to study drug dispersion within the polymer matrix. Ultimately, this project will provide proof-of-concept for our proposed platform technology which may allow better methods for biologic encapsulation in hydrophobic drug release

systems.