Robustness testing of the high throughput HPLC-based analysis of plasma N-glycans.

Background: Plasma glycan analysis using high throughput HPLC-based 96 well platform became a standard procedure for analyzing a large pool of samples for studies comprising thousands of observed individuals. An analytical method which is used to obtain such a huge amount of data should be well characterized and all potentially critical steps should be known.

Methods: Robustness of the high throughput method was tested by Plackett Burman two level, 11-factor, 12 experiment screening design. It provides valuable information about the few most important factors on which further optimization should be focused. According to a long-term laboratory experience, eleven potentially critical factors were chosen for initial screening. Response variable was calculated as coefficient of variance between area % of each peak in each reaction and the area % obtained after performing the procedure according to the laboratory standard operating procedure.

Results: Six out of 16, by HPLC separated, glycan groups revealed significant changes according to changes in factor levels. As expected due to their structural and chemical differences, glycan groups did not display uniform response to 11 factors, but effect estimates for six significant glycan groups showed the same direction regarding high and low factor levels.

General Significance: Screening experiment provided quality data which resolved the questions about optimal conditions and robustness of the high throughput glycan analysis. Fraction factorial design used in this study enabled us to test a great deal of critical steps in time, labor and money saving manner. This article is part of a Special Issue entitled Glycoproteomics.