Protein solubilization: a novel approach.

Area of Science:

• Biopharmaceutical Formulation
• Protein Chemistry
• Computational Biology

Background:

• Protein therapeutics require high solubility and minimal aggregation for effective formulation.
• Protein-protein interactions, quantified by the osmotic second virial coefficient (B-value), are key drivers of aggregation.
• Current formulation development methods face challenges in achieving optimal protein solubility and stability.

Purpose of the Study:

• To develop a high-throughput screening method for identifying additives that improve protein solubility.
• To utilize artificial neural networks (ANNs) for predicting optimal formulation conditions.
• To reduce aggregation and enhance solubility of protein therapeutics.

Main Methods:

• High-throughput measurement of osmotic second virial coefficient (B-value) to assess protein-protein interactions.
• Three-tiered screening approach: individual additive screening, incomplete factorial screening for combinations, and ANN model training.
• Validation of protein stability and ANN model accuracy throughout the process.

Main Results:

• Identification of individual additives that reduce protein interactions and improve solubility.
• Successful training of an ANN model capable of predicting B-values for over 4000 untested formulations.
• Demonstrated reduction in aggregation and increase in solubility for an IgG protein using the developed method.

Conclusions:

• The presented high-throughput B-value screening combined with ANN modeling offers an efficient strategy for protein formulation development.
• This approach accelerates the identification of optimal solution conditions, leading to enhanced protein solubility and stability.
• The method provides a robust platform for minimizing aggregation in therapeutic protein formulations.