Bacterial Glycoprofiling by Using Random Sequence Peptide Microarrays

Abstract

Current analytical methods have been slow in addressing the growing need for glyco-analysis. A new generation of more empirical high-throughput (HTP) tools is needed to aid the advance of this important field. To this end, we have developed a new HTP screening platform for identification of surface-immobilized peptides that specifically bind O-antigenic glycans of bacterial lipopolysaccharides (LPS). This method involves screening of random sequence peptide libraries in addressable high-density microarray format with the newly developed luminescent LPS-quantum dot micelles. Screening of LPS fractions from O111:B4 and O55:B5 serotypes of E. coli on a microarray consisting of 10,000 20-mer peptide features revealed minor differences, while comparison of LPS from E. coli O111:B4 and P. aeruginosa produced sets of highly specific peptides. Peptides strongly binding to the E. coli LPS were highly enriched in aromatic and cationic amino acids, and most of these inhibited growth of E. coli. Flow cytometry and isothermal titration calorimetry (ITC) experiments showed that some of these peptides bind LPS in-solution with a K(d) of 1.75 microM. Peptide selections against P. aeruginosa were largely composed of hydrogen-bond forming amino acids in accordance with dramatic compositional differences in O-antigenic glycans in E. coli and P. aeruginosa. While the main value of this approach lies in the ability to rapidly differentiate bacterial and possibly other complex glycans, the peptides discovered here can potentially be used off-array as antiendotoxic and antimicrobial lead compounds, and on-array/on-bead as diagnostic and affinity reagents.