Spacing mutatioas between the Escherichia coli PBAD RNA polymerase binding site and the araC (T) induction site

Mutations in the Escherichia coli promoter pgAD have been constructed which alter the spacing of the adjacent RNA polymerase and araC inducer protein binding sites. While deletion of a single base-pair or small insertions do not detectably affect araC protein binding to DNA and they do not alter the conserved sequence of the RNA polymerase binding site, stimulation of PBAD i£. vivo is greatly reduced. The experiments suggest that the distance or angle between the two proteins on the DNA is critical for promoter function. INTRODUCTION Transcription from the Escherichia coll L-arabinose operon promoter, PBAD> requires araC protein in the presence of arabinose (1-3). AraC protein represses its own synthesis from pc (4), and in the absence of arabinose, represses transcription from PBAD (2,3,5). Both pc and PBAD a r e stimulated by the cAMP receptor protein (CRP)-cAMP complex (2-4,6). Between the divergently oriented promoters, the DNA binding 9ites for these effectors occupy a 147 base pair noncodlng region in the order: RNA polymerase at prj, which overlaps araC protein repressor (0); cAMP receptor protein; araC protein inducer (I); and RNA polymerase at PgAD (7,8). Although the mechanism by which these proteins stimulate transcription Is not known, models for positive activation encompass two extremes. The proteins nay interact directly (9,10), or the effects may be Indirect, mediated through changes in the DHA structure over significant distances (11,12). If protein-protein contacts or very short-range DNA alterations are necessary for activation of RNA polymerase, the distance and rotational relationship on the DNA between araC protein and polymerase should be critical. On the other hand. If araC and/or CRP-cAMP proteins exclusively affect DNA structure over appreciable distances, promoter activity should not be greatly impaired by small changes In the relative positions of the © I RL Press Limited, Oxford, England. 1873 Nucleic Acids Research binding sites. In order to test the alternatives, DNA was inserted or deleted between araC inducer and pgAD polymerase binding sites. While the mutations do not detectably affect araC protein binding to DNA and they do not alter the conserved sequence of the RNA polymerase binding site, stimulation of Pg^D ^ vivo Is greatly reduced or eliminated. MATERIALS AND METHODS Plasmids and strains Plasmid pTD3 was constructed by Teresa Dunn (M.S. in preparation). It contains a 440 base-pair fragment with the arabinose operon control region flanked by Hindlll and EcoRI linkers. This piece has been insertd into the EcoRI-Hindlll large fragment of the promoter-cloning vehicle pKOl (13). The resulting construct places the galK gene of pKOl under control of PRAD a n d retains ampicillln resistance (Fig. 1). TMD29 is a galK derivative of C600 with a constitutive araC protein, araC (T. Dunn, unpublished). SH313 is galK, araB, thr, his, thl, dcm, Str (S. Hahn, unpublished). General methods Deoxynucleotlde triphosphates were obtained from PL Biochemicals, Y"-^P ATP was from ICN. All enzymes were from Bethesda Research Labs, New England Biolabs and PL Biochemicals. All media, plates, and standard procedures including plasmid ligation, transformation and CsCl gradient plasmid Isolation were as described in Schleif & Wensink (14). Base Insertions and deletions Samples of BamHI cleaved pTD3 DNA (3 yg) were Incubated with 0.5 unit DNA polymerase I large fragment in 30 ul of 50 mM Tris-KCl, pH 7.5, 10 mM MgC2, 7 mM B-mercaptoethanol and 200 uM dNTP's for 30 minutes at room temperature (Fig. 2). After phenol extraction, the samples were digested with 300 Vogt units/ml SI nuclease In 50 ul of 30 mM Na Acetate, pH 4.5, 200 mM NaCl, 5 mM ZnCl2 for 30 minutes at room temperature (15). After ligation and transformation Into TMD29 (AraC, GalK") cells were plated on MacConkey galactose amplcillin plates and the GalK phenotype scored. Several GalK"" and GalK" transformants from each of the three reactions were screened for loss of the BamHI restriction site and also tested for the presence of the 440 base-pair Hindlll-EcoRI fragment. Several of the mutants which had undergone large deletions or insertions were discarded.