UPSTREAM BOX TATA BOX ORDER IS THE MAJOR DETERMINANT OF THE DIRECTION OF TRANSCRIPTION

Mammalian gene promoters for transcription by RNA polymerase II are typically organized in the following order: upstream sequence motif(s)/TATA box/initiation site. Here we report studies in which the order, orientation and DNA sequences of these three elements are varied to determine how these affect polarity of transcription. We have constructed promoters with an 'octamer' upstream sequence ATTTGCAT (or its complement ATGCAAAT) in combination with several different TATA boxes and initiation (cap) sites, and tested these promoters in transfection experiments with cultured cells. TATA boxes derived from the adenovirus major late promoter (TATAAAA). immunoglobulin kappa light chain (TTATATA) and heavy chain (TAAATATA) promoter functioned equally well or even better when inverted. Only the beta-globin TATA box (CATAAAA) was poorly active when inverted. In addition, a symmetrical TATA box (TATATATA) derived from a casein gene was very active. Our results suggest that the asymmetry of most TATA boxes (consensus TATAAAA) is not a primary determinant of the polarity of transcription. We also found that the initiation (cap) site, which usually consists of an adenine embedded in a pyrimidine-rich region (PyPyCAPyPyPyPyPy), was permissive towards sequence alterations; even a randomly composed sequence worked well. However, an inverted, hence purine-rich, cap site reduced transcript levels to 1/7th, as did an oligo G sequence. Irrespective of the presence of a cap site, the configuration: 'TATA box/octamer' yielded a strong leftward, rather than rightward transcription. From this, we conclude that the polarity of transcription is primarily determined by the linear order of an upstream sequence relative to a TATA box, rather than by the individual orientations of either of these two elements.