Search

Encoderless combinatorial chemistry requires high-throughput product identification without the use of chemical or other tags. We developed a novel nano-layered substrate plate and combined it with a microarraying robot, matrix-assisted laser desorption/ionization (MALDI) mass spectrometry, and custom software to produce a high-throughput small molecule identification system. To optimize system performance, we spotted 5 different chemical entities, spanning a m/z range of 195 to 1338, in 20,304 spots for a total of 101,520 molecules. The initial spot identification rate was 99.85% (20,273 spots), and after a proofreading algorithm was added, 100% of 20,304 spots and 101,520 molecules were identified. An internal recalibration algorithm also significantly improved mass accuracy to as low as 45 ppm. Using this optimized system, 47 different chemical entities, spanning a m/z range of 138 to 1,592, were spotted over 5,076 spots and could be identified with 100% accuracy. Our study lays the foundation for improved encoderless combinatorial chemistry.

The discovery of small molecule ligands targeted to the surface of live pathogenic bacteria would enable an entirely new class of antibiotics. We report the development and validation of a microarray-based high-throughput screening platform for bacteria that exploits 300 μm diameter chemical spots in a 1 in. × 3 in. nanolayered glass slide format. Using 24 model compounds and 4 different bacterial strains, we optimized the screening technology, including fluorophore-based optical deconvolution for automated scoring of affinity and cyan–magenta–yellow–key (CMYK) color-coding for scoring of both affinity and specificity. The latter provides a lossless, one-dimensional view of multidimensional data. By linking in silico analysis with cell binding affinity and specificity, we could also begin to identify the physicochemical factors that affect ligand performance. The technology we describe could form the foundation for developing new classes of antibiotics.