420198 Prediction of Peptide Solubility: An Experimental and Computational Framework

Wednesday, November 11, 2015: 4:55 PM
150G (Salt Palace Convention Center)
Chris A. Kieslich1,2, George Khoury3, James Smadbeck3, Fani Boukouvala1,2, Zoltan Szekely4, Patrick J. Sinko4 and Christodoulos A. Floudas1,2, (1)Texas A&M Energy Institute, Texas A&M University, College Station, TX, (2)Artie McFerrin Department of Chemical Engineering, Texas A&M University, College Station, TX, (3)Department of Chemical and Biological Engineering, Princeton University, Princeton, NJ, (4)Department of Pharmaceutics, Rutgers University, Piscataway, NJ

The design of peptides for therapeutic use has become an increasingly important and desirable approach for biomedical research. However, design of protein/peptide therapeutics is complicated by competing factors affecting the possible use of validated peptide agonists and antagonists, such as solubility, cell permeability, bioavailability, stability, and adverse side effects. As stark evidence for this fact, over 40% of drug candidates fail further pursuit because of adverse pharmacokinetics or bioavailability (1). For this reason, accurate methods for the prediction of peptide properties (e.g., solubility) based on amino acid composition alone, are highly desirable. In particular, amyloidogenic sequences are of significant concern for peptide solubility due to their role in the formation of fibrous aggregates.

In this study, we have utilized a combination of experimental and computational tools to investigate the factors leading to poor peptide solubility, with particular interest in developing predictive models of peptide solubility. We performed solid-phase peptide synthesis, using a Nautilus 2400 parallel synthesizer, to produce over 40 suspected amyloid peptides containing between 4 and 11 amino acids. The solubility of each peptide was measured using a modified version of the LYophilisation Solubility Assay (LYSA), as described in (2). SVM regression was performed based on predicted structure based features, including solvation free energy and solvent accessible surface area. Additionally, a novel docking algorithm, based on grey-box global optimization (3), was used to predict the structure of the optimal peptide aggregate. Finally, molecular dynamics simulations, as described previously (4), were also performed to evaluate the association affinity of each peptide.

  1. J. Wang, T. Hou, Recent Advances on Aqueous Solubility Prediction. Combinatorial Chemistry & High Throughput Screening 14, 328–338 (2011).

  2. D. Roy, F. Ducher, A. Laumain, J. Y. Legendre, Determination of the aqueous solubility of drugs using a convenient 96-well plate-based assay. Drug Development and Industrial Pharmacy 27, 107–109 (2001).

  3. F. Boukouvala, M.M.F. Hasan, C.A. Floudas, Global Optimization of General Constrained Grey-Box Models: New Method and its Application to Constrained PDEs. Journal of Global Optimization, Submitted (2015).

  4. J. Smadbeck et al., De Novo Design and Experimental Characterization of Ultrashort Self-Associating Peptides. Plos Computational Biology 10 (2014).

Extended Abstract: File Not Uploaded