Utility of Retention Prediction Model for Investigation of Peptide Separation Selectivity in Reversed-Phase Liquid Chromatography: Impact of Concentration of Trifluoroacetic Acid, Column Temperature, Gradient Slope and Type of Stationary Phase

On the utility of predictive chromatography to complement mass spectrometry based intact protein identification

The amino acid sequence determines the individual protein three-dimensional structure and its functioning in an organism. Therefore, "reading" a protein sequence and determining its changes due to mutations or post-translational modifications is one of the objectives of proteomic experiments. The commonly utilized approach is gradient high-performance liquid chromatography (HPLC) in combination with tandem mass spectrometry. While serving as a way to simplify the protein mixture, the liquid chromatography may be an additional analytical tool providing complementary information about the protein structure. Previous attempts to develop "predictive" HPLC for large biomacromolecules were limited by empirically derived equations based purely on the adsorption mechanisms of the retention and applicable to relatively small polypeptide molecules. A mechanism of the large biomacromolecule retention in reversed-phase gradient HPLC was described recently in thermodynamics terms by the analytical model of liquid chromatography at critical conditions (BioLCCC). In this work, we applied the BioLCCC model to predict retention of the intact proteins as well as their large proteolytic peptides separated under different HPLC conditions. The specific aim of these proof-of-principle studies was to demonstrate the feasibility of using "predictive" HPLC as a complementary tool to support the analysis of identified intact proteins in top-down, middle-down, and/or targeted selected reaction monitoring (SRM)-based proteomic experiments.

Retention behavior of peptides in hydrophilic-interaction chromatography

Selected hydrophilic interaction chromatography (HILIC) columns packed with bare silica, bridge-ethyl hybrid silica, or an amide sorbent chemistry were utilized for an investigation of chromatographic behavior and separation selectivity of tryptic peptides. Retention model was proposed allowing for retention prediction of peptides with correlation coefficient R(2)~0.92-0.97 for various columns. The values of optimized amino acid retention coefficients were compared to those obtained for reversed-phase liquid chromatography (Gilar et al., Anal. Chem. 2010, 82, 265-275) and used to elucidate the impact of different amino acid on peptide HILIC retention. In contrast to reversed-phase chromatography, where presence of Phe, Trp, Ile, and Leu amino acid residues in sequence strongly promoted, and presence of hydrophilic His, Lys and Arg residues strongly reduced peptide retention, the effects of these amino acid residues in HILIC were opposite (His, Lys and Arg promote, Phe, Trp, Ile and Leu demote peptide retention in HILIC). Retention coefficient optimized for pH experiments illustrated the impact of silanols on HILIC retention.

Simultaneous analysis of aspartame and its hydrolysis products of Coca-Cola Zero by on-line postcolumn derivation fluorescence detection and ultraviolet detection coupled two-dimensional high-performance liquid chromatography

An innovative two-dimensional high-performance liquid chromatography system was developed for the simultaneous analysis of aspartame and its hydrolysis products of Coca-Cola Zero. A C8 reversed-phase chromatographic column with ultraviolet detection was used as the first dimension for the determination of aspartame, and a ligand-exchange chromatographic column with on-line postcolumn derivation fluorescence detection was employed as the second dimension for the analysis of amino acid enantiomers. The fluorimetric derivative reagent of amino acid enantiomers was o-phthaldialdehyde. The hydrolysis of aspartame in Coca-Cola Zero was induced by electric-heating or microwave heating. Aspartame was quantified by the matrix matched external standard calibration curve with a linear concentration range of 0-50 μg mL(-1) (r(2)=0.9984). The limit of detection (LOD) and the limit of quantification (LOQ) were 1.3 μg mL(-1) and 4.3 μg mL(-1), respectively. The amino acid enantiomers was analyzed by the matrix matched internal standard calibration method (D-leucine as the internal standard) with a linear concentration range of 0-10 μg mL(-1) (r(2)=0.9988-0.9997). The LODs and LOQs for L- and D-aspartic acid and L- and D-phenylalanine were 0.16-0.17 μg mL(-1) and 0.52-0.55 μg mL(-1), respectively, that was 12-13 times more sensitive than ultraviolet detection. The overall analysis accuracy for aspartame and amino acid enantiomers was 90.2-99.2% and 90.4-96.2%, respectively. The overall analysis precision for aspartame and amino acid enantiomers was 0.1-1.7% and 0.5-6.7%, respectively. Generally, the extent of aspartame hydrolysis increases with the increase of electro-thermal temperature, microwave power, and the duration of hydrolysis time. D-aspartic acid and D-phenylalanine can be observed with the electro-thermal racemization at the hydrolysis temperature 120°C for 1 day and only D-aspartic acid can be observed at the hydrolysis temperature 90°C for 2 and 3 days. For the microwave induced hydrolysis, only L-aspartic acid was detected at the power 560 W for 1 min and 320 W for 3 min.