Flow-Seq Evaluation of Translation Driven by a Set of Natural Escherichia coli 5'-UTR of Variable Length.

Flow-seq is a method that combines fluorescently activated cell sorting and next-generation sequencing to deduce a large amount of data about translation efficiency from a single experiment. Here, we constructed a library of fluorescent protein-based reporters preceded by a set of 648 natural 5'-untranslated regions (5'-UTRs) of Escherichia coli genes. Usually, Flow-seq libraries are constructed using uniform-length sequence elements, in contrast to natural situations, where functional elements are of heterogenous lengths. Here, we demonstrated that a 5'-UTR library of variable length could be created and analyzed with Flow-seq. In line with previous Flow-seq experiments with randomized 5'-UTRs, we observed the influence of an RNA secondary structure and Shine-Dalgarno sequences on translation efficiency; however, the variability of these parameters for natural 5'-UTRs in our library was smaller in comparison with randomized libraries. In line with this, we only observed a 30-fold difference in translation efficiency between the best and worst bins sorted with this factor. The results correlated with those obtained with ribosome profiling.

Isolation and sequencing of short cell-surface-bound DNA.

To isolate the DNA fragments bound to cell-surface proteins with high affinity, peptide-DNA complexes were obtained by mild treatment of human endothelial cells with trypsin. After binding of nucleoprotein complexes to BrCN-Sepharose and extensive washing, 7-11 nt DNA fragments were eluted with acidic buffer, labeled with (32)P, and isolated by PAGE. The fragments were ligated to double-stranded oligonucleotide adapters, amplified by PCR, and cloned in pBluescript II KS (+) vector. Sequences of 12 clones containing the insert of cell-surface-bound 7-11-mer DNA fragments were obtained.