Transcription terminators can be strong promoters

DNA sequences directing transcription initiation (promoters) and termination (terminators) are distinct and function through unrelated mechanisms. Class II terminators for bacteriophage T7 RNA polymerase are unusual in being short and lacking hairpin structures. Serendipitously, we find that two class II terminators, the natural bacteriophage T7 concatemer junction (CJ) and the artificial vesicular stomatitis virus (VSV) terminator, function as strong promoters for E. coli RNA polymerase in vivo. Mapping of transcription start sites, conservation analysis and mutagenesis revealed that the promoter function does not rely solely on the conserved class II terminator sequence. Instead, we identify a promoter motif with an extended -15TGn and a critical -12TA dinucleotide within the TATA box, different from canonical {sigma}70 promoters. Activities also correlated with Ts in the -35 box and AT richness of the adjacent core. Furthermore, analysis of published randomized-promoter libraries suggests that the activities of many members lacking canonical promoter boxes may be explained by the promoter motif identified with the class II terminator. The new principle of overlap between initiation and termination may be an unappreciated factor in streamlined bacterial genomes. Significance StatementThis study overturns a long-standing distinction between transcription initiation and termination signals. For decades, RNA synthesis has been thought to rely on separate, unrelated DNA signals governing promotion and termination. Unexpectedly, we discovered that two class II terminators for bacteriophage T7 RNA polymerase, the natural concatemer junction and an artificial vesicular stomatitis virus terminator, are strong promoters for Escherichia coli RNA polymerase. We identify a shared promoter motif that explains many previously-unexplained promoter activities. This finding reveals an unrecognized dual-function DNA element and suggests that compact terminator-promoter sequences may represent an efficient and previously overlooked strategy for genome streamlining.