Raman spectroscopy detects phenotypic differences among Escherichia coli enriched for 1-butanol tolerance using a metagenomic DNA library

Advances in Raman spectroscopy are enabling more comprehensive measurement of microbial cell chemical composition. Advantages include results returned in near real‐time and minimal sample preparation. In this research, Raman spectroscopy is used to analyze E. coli with engineered solvent tolerance, which is a multi‐genic trait associated with complex and uncharacterized phenotypes that are of value to industrial microbiology. To generate solvent tolerant phenotypes, E. coli transformed with DNA libraries are serially enriched in the presence of 0.9% (v/v) and 1.1% (v/v) 1‐butanol. DNA libraries are created using degenerate oligonucleotide primed PCR (DOP‐PCR) from the genomic DNA of E. coli, Clostridium acetobutylicum ATCC 824, and the metagenome of a stream bank soil sample, which contained DNA from 72 different phyla. DOP‐PCR enabled high efficiency library cloning (with no DNA shearing or end‐polishing) and the inclusion un‐culturable organisms. Nine strains with improved tolerance are analyzed by Raman spectroscopy and vastly different solvent‐tolerant phenotypes are characterized. Common among these are improved membrane rigidity from increasing the fraction of unsaturated fatty acids at the expense of cyclopropane fatty acids. Raman spectroscopy offers the ability to monitor cell phenotype changes in near real‐time and is adaptable to high‐throughput screening, making it relevant to metabolic engineering. New methodology was developed that allowed the packaging and screening of all microbial DNA in an environmental soil sample. In this paper, the DNA was screened for genes that conferred tolerance to 1‐butanol, a potential biofuel that is highly toxic to fermenting cultures. Genes conferring greater tolerance were identified and Raman spectroscopy was used to determined how they changed the chemical composition of cells. Raman spectroscopy offers the ability to monitor cell phenotype changes in near real‐time and is adaptable to high‐throughput screening, making it relevant to metabolic engineering.