(566g) Development and Characterization of a Promoter Library for Probiotic E. coli Nissle Using a High-Throughput Barcoding Approach

Knock-in of genes or pathways with therapeutic potential to the gut microbiota using engineered probiotic strains has the potential to revolutionize human health. However, the dosage of therapeutics delivered in vivo must be precisely controlled both spatially and temporally. This degree of tunability requires detailed knowledge of transcriptional regulator activity in situ. To date, engineered organisms have been largely developed in the aerobic, homogeneous, resource-rich, noncompetitive environment of the test tube. The gut environment is largely the opposite, so it is unclear how well existing databases of transcriptional regulators will function in gut-resident probiotics. However, accurate determination of gene expression in the gut context requires a vertebrate host, quickly making the traditional approach of testing each promoter in separate animals infeasible and unethical. We have optimized a high-throughput approach for in situ measurement of promoter activity using DNA barcodes, enabling expression assays for entire promoter libraries to be performed in single animals using high-throughput sequencing. To demonstrate this approach, we curated 32 candidate promoters through analysis of metatranscriptome datasets and promoter libraries, and transferred these synthetic constructs into E. coli Nissle. We confirmed that our barcoding strategy had an undetectable effect on gene expression, but interestingly found that each barcode exhibited a significant (up to 5-fold) and reproducible bias during amplification and sequencing. In response, we developed a model of this bias enabling accurate recapitulation of known ratios of barcode-containing DNA in standard curve experiments. Using this validated method, we then analyzed promoter activity in E. coli Nissle grown in vitro and in situ under a variety of environmental conditions (oxygen content, diet, and body site). Collectively, these experiments reveal exciting dynamics in synthetic constructs during passage through the mammalian gut and inform the design of probiotic strains engineered to improve human health. More broadly, these experiments also reveal PCR bias as a significant confounder of barcoded abundance measurements and suggest strategies to overcome its effects.