Automated Assessment of the DNA-Binding Capacity of a Proteome by In Vitro Selection

Abstract

Methods for the high-throughput identification of transcription factor binding sites (TFBSs) are becoming increasingly relevant as new genomes are sequenced and explored. Although DNA:protein complexes can be purified from lysates and binding sites identified, this only provides a snapshot of the full DNA-binding potential of an organismal proteome. Instead, automated double-stranded DNA (dsDNA) selections can be used to identify target sites for all DNA-binding proteins in parallel. Double-stranded DNA selections were carried out against cell lysates from Escherichia coli. The resultant aptamer sequences showed enrichment for known TFBSs versus both the native pool and the E. coli genome, and underwent an overall change in sequence composition. Randomized versions of the selected pools did not show the same enrichment. Based on this automated method and accompanying bioinformatics algorithms, it may be possible to extract complex information about TFBSs from cellular lysates. It may further be possible to use these methods to model previously unknown DNA-binding sites in a variety of novel organisms. (JALA 2007;12:135–42)