Promoters were originally identified as consensus sequences upstream from transcriptional start sites. What additional evidence might support the assignment of these sequences as parts of promoters?