Abstract:
In downstream processes for peptides, crystallization is still used as the state-of-the-art separation step for which the knowledge about the solubility of each single compound is mandatory. Since the determination of experimental temperature-dependent solubility data is time-consuming and expensive, modeling solubility based on physical properties such as melting properties is highly desired. Unfortunately, the direct determination of melting properties for biomolecules using conventional differential scanning calorimetry is not possible due to the decomposition of the peptides before their melting. In this work, fast scanning calorimetry (FSC) with heating rates up to 20,000 K s-1 was applied to measure the melting properties of 22 peptides with focus on isomeric dipeptides and tripeptides based on glycine, l-alanine, l-leucine, l-proline, and l-serine. The experimental determination of the aqueous solubility of these peptides was performed using the photometric method (UV/Vis spectrometer) and the gravimetric method of supersaturated solutions. Additionally, the pH value and the crystal structure of peptides were determined in order to ensure the neutral species in solution and to exclude crystal structure changes in the solid phase. The experimental FSC-measured melting properties were used as input data in the thermodynamic modeling framework PC-SAFT to model the peptide solubility in water. The PC-SAFT pure-component parameters of the peptides were determined following a weighted joint-parameter method introduced in this work. This approach allows determining the pure-component parameters of a peptide by joining the pure-component parameters of the parent amino acids. The binary interactions parameter between peptide and water was fitted to solubility-independent properties such as osmotic coefficients and mixture densities of aqueous peptide solutions. The modeled peptide solubility was in good agreement with the experimental solubility.
