Rigorous pattern-recognition methods for DNA sequences. Analysis of promoter sequences from Escherichia coli

The basic nature of the sequence features that define a promoter sequence for Escherichia coli RNA polymerase have been established by a variety of biochemical and genetic methods. We have developed rigorous analytical methods for finding unknown patterns that occur imperfectly in a set of several sequences, and have used them to examine a set of bacterial promoters. The algorithm easily discovers the "consensus" sequences for the -10 and -35 regions, which are essentially identical to the results of previous analyses, but requires no prior assumptions about the common patterns. By explicitly specifying the nature of the search for consensus sequences, we give a rigorous definition to this concept that should be widely applicable. We also have provided estimates for the statistical significance of common patterns discovered in sets of sequences. In addition to providing a rigorous basis for defining known consensus regions, we have found additional features in these promoters that may have functional significance. These added features were located on either side of the -35 region. The pattern 5', or upstream, from the -35 region was found using the standard alphabet (A, G, C and T), but the pattern between the -10 and the -35 regions was detectable only in a sub-alphabet. Recent results relating DNA sequence to helix conformation suggest that the former (upstream) pattern may have a functional significance. Possible roles in promoter function are discussed in this light, and an observation of altered promoter function involving the upstream region is reported that appears to support the suggestion of function in at least one case.     

Expectation maximization algorithm for identifying protein-binding sites with variable lengths from unaligned DNA fragments.

An Expectation Maximization algorithm for identification of DNA binding sites is presented. The approach predicts the location of binding regions while allowing variable length spacers within the sites. In addition to predicting the most likely spacer length for a set of DNA fragments, the method identifies individual sites that differ in spacer size. No alignment of DNA sequences is necessary. The method is illustrated by application to 231 Escherichia coli DNA fragments known to contain promoters with variable spacings between their consensus regions. Maximum-likelihood tests of the differences between the spacing classes indicate that the consensus regions of the spacing classes are not distinct. Further tests suggest that several positions within the spacing region may contribute to promoter specificity.     

High level heterologous expression in E. coli using mutant forms of the lac promoter.

Single base deletions in the lac promoter which reduced the 18bp spacing between the -35 and -10 homology regions to 17bp, increased the strength of the promoter. A single base substitution (T----G) in the -35 region to generate the consensus sequence TTG-ACA increased the strength further and no longer required a 17bp spacing. The mutated lac promoter was as powerful as a shorter form of the tac promoter which lacked two AT-rich regions upstream of the -35 region, and expressed the P69 surface antigen (pertactin) of Bordetella pertussis to 30-40% total cell protein and tetanus toxin fragment C to 16-20% total cell protein.     

DNA sequence recognition by a eukaryotic sequence-specific endonuclease, Endo.SceI, from Saccharomyces cerevisiae

A eukaryotic sequence-specific endonuclease, Endo.SceI, causes sequence-specific double-stranded scission of double-stranded DNA to produce cohesive ends with four bases protruding at the 3' termini. Unlike in the case of restriction enzymes, an asymmetric 26-base pair consensus sequence was found around the cleavage site for Endo.SceI instead of a common sequence. We analyzed the base pairs that interacted with Endo.SceI on the recognition of its cleavage sites. A region comprising -10 through +16 base pairs from the center of the cleavage site was shown to be essential and sufficient for the sequence-specific cutting with Endo.SceI by experiments involving synthesized DNAs. Methylation interference experiments indicate that bases in the region comprising the +7 through +14 base pairs is involved in close contact with Endo.SceI in its recognition of the cleavage site. This +7 through +14-base pair region overlaps the most stringently conserved sequence in the consensus sequence for the cleavage site, suggesting that this region constitutes the core for the recognition by Endo.SceI.     

Up-promoter mutations in the lpp gene of Escherichia coli.

The promoter of the gene for the major outer membrane lipoprotein, the most abundant protein in Escherichia coli, is considered to be one of the strongest promoters in E. coli. The nucleotide sequences of the -10 and the -35 regions of the lpp promoter were altered in a step-wise manner to conform to their respective consensus sequences by synthetic oligonucleotide-directed site-specific mutagenesis. The mutated promoters were then fused to the lacZ gene to measure promoter activity. The beta-galactosidase activity increased approximately 1.9 and 2.4 fold when the -10 region (AATACT) was altered to TATACT(P1) and TATAAT (consensus sequence; P2), respectively. Similarly, it increased approximately 1.2 and 4.2 fold, when the -35 region (TTCTCA) was altered to TTCACA(R1) and TTGACA (consensus sequence; R2), respectively. When the mutations at the -10 and -35 regions were combined, the overall improvement of the promoter activity for R2-P1 was 4.0 fold over that of the wild-type promoter, while it was only 2.5 fold for R2-P2. These results indicate that substantial improvement of the promoter activity can be achieved by changing either of the two key regions to their respective consensus sequences. However, the complete conformity to consensus sequences at both regions does not necessarily result in the highest activity. With use of the improved lpp promoter in an expression cloning vehicle pIN-III-ompA, staphylococcal nuclease A was produced at a level of approximately 47% of the total cellular protein.     

Functional Analysis and Randomization of the Nisin-Inducible Promoter for Tuning Gene Expression in Lactococcus lactis

Three putative promoter regions were identified preceding the nisZ gene in Lactococcus lactis HSM-22. To investigate their function in the control of nisZ biosynthesis, green fluorescence protein (GFP) was adopted as probe to determine activities of the three promoters. The results showed that PnisZ-0 containing two sets of the ?35 and ?10 regions exhibited the same maximum activity as promoter PnisZ-2 containing the putative promoter region near the start codon. However, the GFP expression level directed by PnisZ-0 was twofold higher than that found with PnisZ-2 under low-dose nisin, indicating that promoter PnisZ-1 distant from the start codon could be important in response to the inducer nisin. Then, Pnis-2 was randomized to develop functional promoters through the degenerate oligonucleotide approach in L. lactis. 35 inducible promoters and 14 constitutive promoters were obtained, covering 3–5 logs of expression levels in small increments of activity. Sequence analysis revealed that base changes in both consensus sequence and spacing sequence resulted in remarkable decrease of promoter activity, while the sequence outside ?35 and ?10 regions would influence the promoter function radically. The functional promoters were evaluated for the efficiency and stability to control β-galactosidase (Gal) expression in L. lactis. High correlation was obtained between the Gal activity and promoter strength, suggesting that promoters developed here have the potential for fine tuning gene expression in L. lactis.     

Sequence elements determining ampC promoter strength in E. coli

A number of spontaneous up-promoter mutations have been isolated in the ampC beta-lactamase gene of Escherichia coli. The mutants were analyzed by DNA sequencing, and the level of ampC gene expression was determined. Six mutants with a 21-fold increase in promoter strength compared with the wild-type were mutated in the -35 promoter region from TTGTCA to the consensus sequence TTGACA . The -10 region sequence TACAAT was mutated to the consensus sequence TATAAT in three mutants exhibiting an ampC promoter seven times stronger than the wild-type. We have previously described a 1-bp insertion mutant ( Jaurin et al., 1981) that changes the inter-region distance to the consensus 17 bp. Thus, all the up-mutations found in the ampC promoter represent corrections of the three major discrepancies between the ampC promoter and the consensus E. coli promoter. We conclude that the three consensus elements of E. coli promoters, the -35 and -10 regions and an optimal inter-region distance of 17 bp, are the main elements determining the promoter strength.