Mapping of single-stranded regions in duplex DNA at the sequence level: single-strand-specific cytosine methylation in RNA polymerase-promoter complexes.

A method based on the differential rate of cytosine methylation in single- and double-stranded nucleic acids by dimethyl sulfate [Peattie, D.A. & Gilbert, W. (1980) Proc. Natl. Acad. Sci. USA 77, 4679-4682] has been developed for probing unpaired cytosines in DNA and DNA-protein complexes at the sequence level. Application of the method to the complexes between Escherichia coli RNA polymerase (EC 2.7.7.6) and three related promoters, lac UV5, trp, and a hybrid promoter tac resulting from the fusion of the two, reveals distinct differences in the way RNA polymerase unpairs DNA in these promoters. No single-stranded region is detectable in the complex with the trp promoter. For the lac UV5 promoter, the cytosines at positions -6, -4, -2, and -1 are in an unpaired region. The same cytosines in the tac promoter, which is homologous in sequence to lac UV5 in this region, are also found to be single stranded. For the pair of promoters lac UV5 and tac, the cytosine methylation reaction has also been used to demonstrate the steep temperature dependence of opening of base pairs by RNA polymerase. One striking feature is that the midpoint of this transition for the tac promoter is 3 degrees C lower than the corresponding value for lac UV5, even though the sequence of the unpaired region in the two promoters is identical.     

Correlation of lac operator DNA imino proton exchange kinetics with its function.

The kinetics for imino hydrogen exchange, at individual base pairs in the DNA sequence corresponding to the lactose operon operator of Escherichia coli, has been examined by NMR saturation recovery measurements as a function of temperature. Three 17-base-pair subsections of the lac operator DNA were chemically synthesized for these studies. The results support our previous observations in the 36-base-pair complete lac operator DNA fragment that has been used in our previous NMR studies. The results indicate faster opening kinetics at a GTG/CAC that is also the site of operator mutations leading to the highest level of constitutive beta-galactosidase synthesis. The GTG/CAC sequence occurs frequently and often symmetrically in prokaryotic and eukaryotic DNA sites where one anticipates specific protein interaction for gene regulation or recombination.     

Retroregulation of the synthesis of ribosomal proteins L14 and L24 by feedback repressor S8 in Escherichia coli.

Previous studies on regulation of the spc operon containing genes for ribosomal proteins have shown that S8, encoded by the fifth gene of the operon in Escherichia coli, is a translational repressor and regulates the synthesis of the third gene product (L5) and distal gene products by acting at a site near the L5 mRNA translation initiation site. We have now shown that S8 also regulates the synthesis of the first and second gene products (L14 and L24) of the operon by acting at the same mRNA target site--that is, the site located distal to sites coding for L14 and L24--and that mRNA degradation is involved in this retroregulation. It was shown that single base substitutions in the target site, which abolish repression of the synthesis of L5 and L5-distal gene products by S8, also cause derepression of L14-L24 synthesis. Inhibition of L14-L24 synthesis by S8 was also shown by overproducing S8 in trans from a plasmid carrying the S8 gene under lac promoter/operator control. A strain carrying temperature-sensitive mutations in genes for polynucleotide phosphorylase and RNase II was found upon shift to nonpermissive temperature to show higher differential synthesis rates of L14-L24 (and L5) relative to those of several L5-distal spc operon gene products. We suggest that repression of distal ribosomal protein synthesis by S8 triggers nucleolytic cleavage of spc operon mRNA, followed by mRNA degradation by these 3'- to 5'- exonucleases, which is then responsible for inhibition of L14-L24 synthesis.     

5-Azido-2''-deoxyuridine 5''-triphosphate: a photoaffinity-labeling reagent and tool for the enzymatic synthesis of photoactive DNA.

We have synthesized the photoactive deoxyuridine nucleotide 5-azido-2'-deoxyuridine 5'-triphosphate (5-N3dUTP) and used it to synthesize light-sensitive DNA by enzymatic incorporation. In the absence of ultraviolet light, 5-N3dUTP is a substrate for Escherichia coli DNA polymerase I. In in vitro DNA synthesis reactions using bacteriophage M13 single-stranded DNA as the template and 5-N3dUTP in place of dTTP, a photoactive complementary strand was synthesized by DNA polymerase I. The complementary strand was not synthesized when the 5-N3dUTP was substituted for dCTP or when it was exposed to ultraviolet light prior to the addition of DNA polymerase I. Using a synthetic lac operator template of 26 bases and a 15-base primer, we generated a photoactive 26-base-pair lac operator by enzymatically incorporating 5-N3dUMP with DNA polymerase I. Crosslinking of this photoactive DNA fragment to lac repressor was totally dependent on the presence of UV light and was reduced 78% by 150 microM isopropyl beta-D-thiogalactoside. Under the same conditions no crosslinking to lac repressor was observed using a nonphotoactive 26-base-pair lac operator. Photoactivatable deoxyuridine analogs have potential application as reagents to crosslink DNA binding proteins to 5-azidouracil-containing DNA and as active-site-directed photoaffinity labelling reagents.     

Contacts between the lac repressor and the thymines in the lac operator.

We have identified important points of contact between the lac repressor and the lac operator by crosslinking the repressor to bromouracil-substituted operator. We substituted bromouracils for thymines in a 55-base-long restriction fragment containing the lac operator and labeled one or the other 5' end with 32P. Ultraviolet irradiation of this fragment produced single-strand breakds at the bromouracils. We examined breakage at each bromouracil in the sequence by denaturing the DNA and displaying the UV-generated fragments on a polyacrylamide gel. In the presence of lac repressor, UV radiation failed to break at specific sites. We attribute this to a competing reaction in which the DNA crosslinks to the repressor rather than breaking. These crosslinkable sites thus define positions at which the lac repressor protein lies close to the methyl group of a thymine in the major groove of DNA.     

Detection and Isolation of the Repressor Protein for the Tryptophan Operon of Escherichia coli

DNA from a transducing bacteriophage carrying a fusion of the tryptophan and lactose operons of E. coli (lambdadtrp-lac) has been used to direct cell-free synthesis of beta-galactosidase (EC 3.2.1.23). Whereas normal lac operon (lambdadlac) DNA requires adenosine-3':5'-cyclic monophosphate (cAMP) for beta-galactosidase synthesis, trp-lac DNA is unaffected by cAMP. This difference in cAMP dependence verifies the presence of a cAMP-requiring promoter in the lac operon that has been removed from the trp-lac DNA. Synthesis with trp-lac DNA is controlled by the protein product of the tryptophan repressor gene (trpR). Synthesis in extracts of trpR(-) (repressor-negative) cells is progressively reduced by increased additions of extract from trpR(+) cells. No trpR(-) product repression is seen when beta-galactosidase synthesis is programmed by normal lac DNA. This highly sensitive and specific assay has facilitated quantitation and partial purification of the trp repressor.     

Formylmethionyl-tRNA alters RNA polymerase specificity.

Escherichia coli fMet-tRNAfMet alters the pattern of promoter selection of E. coli RNA polymerase (RNA nucleotidyltransferase, nucleosidetriphosphate:RNA nucleotidyltransferase; EC 2.7.7.6), affecting RNA synthesis from the rRNA, suIII+tRNA, and lac promoters in different ways. The in vitro synthesis of the stable RNA species is selectively decreased, whereas that of lac RNA from both the wild-type and mutant UV5 promoters is selectively increased at high ionic strength. The functional effect of fMet-tRNAfMet resembles that of the nucleotide guanosine 3'-diphosphate 5'-diphosphate (ppGpp). This nucleotide competes with the binding of fMet-tRnafMet to RNA polymerase.     

Promoters largely determine the efficiency of repressor action.

Operator sequence and repressor protein regulate the activity of the lac promoter over a greater than 1000-fold range. Combinations of the lac operator with other promoter sequences, however, differ vastly in the level of repression. The data presented show that the extent of repression is determined largely by the rates of complex formation of the competing systems operator-repressor and promoter-RNA polymerase and by the rate at which RNA polymerase clears the promoter. Moreover, up to 70-fold differences in the level of repression were found when the operator was placed in different positions within the promoter sequence. A kinetic model is proposed that explains the observed effects and that allows predictions on promoters controlled by negatively acting elements.     

Possible ideal lac operator: Escherichia coli lac operator-like sequences from eukaryotic genomes lack the central G X C pair.

Five DNA fragments have been cloned from yeast, chicken, and mouse DNA that titrate lac repressor in an Escherichia coli lac+ I+Z+ wild-type strain when on a multi-copy plasmid. The five repressor-binding sequences have been identified by DNA sequence determinations and DNase cleavage-inhibition patterns. They share the 14-base-pair symmetrical consensus sequence 5' T-G-T-G-A-G-C:G-C-T-C-A-C-A 3' (the colon represents the center of symmetry), which is an inverted repeat of 7 base pairs of the left half of the E. coli lac operator. A similar perfect palindromic DNA fragment--an 11-base-pair inverted repeat of the left half of the lac operator--was synthesized. The cloned synthetic DNA 5' G-A-A-T-T-G-T-G-A-G-C:G-C-T-C-A-C-A-A-T-T-C 3' binds lac repressor 8-fold more tightly than does wild-type E. coli lac operator DNA.