De novo promoters emerge more readily from random DNA than from genomic DNA.

Abstract

Promoters are DNA sequences that initiate transcription. We studied the propensity of both random and genomic DNA to initiate transcription. To this end, we assayed the promoter activity of 17,129 random, synthetic DNA sequences and 91,866 <i>Escherichia coli</i> genomic sequences. Genomic DNA encodes ~1.3× more promoters than random DNA. This higher incidence of promoters also holds for intragenic regions, suggesting an underappreciated role for intragenic promoters. We also explored the propensity of nonpromoter sequences to become promoters. To this end, we chose 225 random and 60 genomic sequences without promoter activity, created over half a million DNA mutants from them, and assayed these mutants for novel promoter activity. Promoters emerge ~3× more readily from random DNA than from genomic DNA, because the genome contains fewer proto-binding sites for transcriptional activators. Our work shows that the genome has a smaller evolutionary potential to create new transcripts than random DNA.