No abstract

A rat genomic clone containing 4.5 kilobases of 5'-flanking DNA and the first exon of the type II beta regulatory subunit (RII beta) of cAMP-dependent protein kinase was isolated, restriction mapped, and sequenced. The proximal 400-basepair promoter region was GC rich, lacked TATA/CAAT box motifs, and initiated transcription at multiple sites. Bandshifting and DNase-I footprinting experiments using this region of the RII beta promoter detected several related specific DNA-protein complexes formed using crude and fractionated nuclear extracts from rat ovary, brain, adrenal gland, and liver. All binding in these experiments mapped to a domain within the same region found to confer cAMP inducibility to a chloramphenicol acetyltransferase (CAT) reporter gene when transfected into primary cultures of rat granulosa cells. Although GC boxes (putative SP1-binding sites) and activator protein-2 (AP-2) elements were present in this functional region, and although expression vectors containing AP-2 sites conferred high levels of cAMP regulation of the CAT gene in cultured ovarian cells, neither the GC boxes nor the AP-2 sites were protected by footprint analyses or required for band shift activity of nuclear extract protein. These known regulatory elements, therefore, may be involved in functional activity of the RII beta promoter, but additional cis-acting DNA and trans-acting factors (yet to be characterized) also appear to interact with the functional promoter of the RII beta gene and regulate the hormone-specific expression of the A-kinase subunit in ovarian and neuronal cells.

We explore a model for ‘quenching-like' repression by studying synthetic bacterial enhancers, each characterized by a different binding site architecture. To do so, we take a three-pronged approach: first, we compute the probability that a protein-bound dsDNA molecule will loop. Second, we use hundreds of synthetic enhancers to test the model's predictions in bacteria. Finally, we verify the mechanism bioinformatically in native genomes. Here we show that excluded volume effects generated by DNA-bound proteins can generate substantial quenching. Moreover, the type and extent of the regulatory effect depend strongly on the relative arrangement of the binding sites. The implications of these results are that enhancers should be insensitive to 10–11 bp insertions or deletions (INDELs) and sensitive to 5–6 bp INDELs. We test this prediction on 61 σ54-regulated qrr genes from the Vibrio genus and confirm the tolerance of these enhancers' sequences to the DNA's helical repeat.

We use an oligonucleotide library of >10,000 variants to identify an insulation mechanism encoded within a subset of σ promoters. Insulation manifests itself as reduced protein expression for a downstream gene that is expressed by transcriptional readthrough. It is strongly associated with the presence of short CT-rich motifs (3-5 bp), positioned within 25 bp upstream of the Shine-Dalgarno (SD) motif of the silenced gene. We provide evidence that insulation is triggered by binding of the ribosome binding site (RBS) to the upstream CT-rich motif. We also show that, in E. coli, insulator sequences are preferentially encoded within σ promoters, suggesting an important regulatory role for these sequences in natural contexts. Our findings imply that sequence-specific regulatory effects that are sparsely encoded by short motifs may not be easily detected by lower throughput studies. Such sequence-specific phenomena can be uncovered with a focused oligo library (OL) design that mitigates sequence-related variance, as exemplified herein.

No abstract