Retention modeling of therapeutic peptides in sub-/supercritical fluid chromatography

Abstract

Chromatographic modeling software packages are valuable tools during method optimization steps. These are well established for reversed-phase applications utilizing retention time (RT) prediction to optimize separations and receive robust methods, which is of high interest for the analysis of pharmaceuticals. In contrast to liquid chromatography, the knowledge of RT prediction in supercritical fluid chromatography is limited to a manageable number of studies.

This study uses the software DryLab to predict the RTs of the peptides bacitracin (Bac), colistin, tyrothricin (Tyro), and insulin analogs. Gradient time, column temperature, and the ternary composition (terC) of carbon dioxide, methanol (MeOH), and acetonitrile (ACN) in the gradient elution are varied in a feasibility approach using a neutral (Viridis BEH) and an amino-derivatized aromatic (Torus 2-PIC) stationary phase. In the second part, chromatographic optimization is performed in silico through gradient adjustments to optimize the separation of the fingerprint of Bac. The final gradient method utilizes a Viridis BEH column (100 × 3.0 mm, 1.7 μm), carbon dioxide, and a modifier consisting of ACN/MeOH/water/methanesulfonic acid (60:40:2:0.1, v:v:v:v). In addition, changes in the retention order of Tyro compounds with the proportion of the terC in combination with a Torus Diol column are investigated.