Mutational analysis of the ompA promoter from Flavobacterium johnsoniae

Abstract

Sequences that mediate the initiation of transcription in Flavobacterium species are not well known. The majority of identified Flavobacterium promoter elements show homology to those of other members of the phylum Bacteroidetes, but not of proteobacteria, and they function poorly in Escherichia coli. In order to analyze the Flavobacterium promoter structure systematically, we investigated the -33 consensus element, -7 consensus element, and spacer length of the Flavobacterium ompA promoter by measuring the effects of site-directed mutations on promoter activity. The nonconserved sequences in the spacer region and in regions close to the consensus motifs were randomized in order to determine their importance for promoter activity. Most of the base substitutions in these regions caused large decreases in promoter activity. The optimal -33/-7 motifs (TTTG/TANNTTTG) were identical to Bacteroides fragilis sigma(ABfr) consensus -33/-7 promoter elements but lacked similarity to the E. coli sigma(70) promoter elements. The length of the spacer separating the -33 and -7 motifs of the ompA promoter also had a pronounced effect on promoter activity, with 19 bp being optimal. In addition to the consensus promoter elements and spacer length, the GC content of the core promoter sequences had a pronounced effect on Flavobacterium promoter activity. This information was used to conduct a scan of the Flavobacterium johnsoniae and B. fragilis genomes for putative promoters, resulting in 188 hits in B. fragilis and 109 hits in F. johnsoniae.

FIG. 1.

Mapping of the ompA promoter. (A) Alignment analysis of putative ompA promoters from various flavobacteria. The accession numbers of ompA genes for F. johnsoniae, Flavobacteriales bacterium HTCC2170, “Leeuwenhoekiella blandensis” MED217, Cellulophaga sp. strain MED134, Croceibacter atlanticus HTCC2559, and “Gramella forsetii” KT0803 are AAPM01000000, ZP_01105836, NZ_AANC01000009, NZ_AAMZ01000002, AAMP00000000, and NC_008571, respectively. The putative UP sequences are boxed. The −33 and −7 motifs are capitalized and in boldface. Identical nucleotides are shown with asterisks. (B) Organization of the wild-type ompA promoter sequence used for deletion assay. Only the 3′ end of the kbl gene (2-amino-3-ketobutyrate coenzyme A ligase) is shown. The sequences of ompA transcriptionally fused with gfp are in italics. The putative −7 and −33 motifs are capitalized and boxed. The capital A with a curved arrow is the TSP. The forward dot arrows below the sequence represent positions of the deletion N-terminal primers. (C) Quantitative analysis of gfpmut3 expression of the ompA promoter deletion vector series in F. johnsoniae including the wild type (W.T.) as a positive control (pFj29) and a negative control (N.C.) (pSCH144). Reactions were performed in triplicate, and standard deviations are marked by error bars. Results for each deletion promoter clone have been normalized to a promoter activity of 100% for the wild-type ompA clone (Fj29).

FIG. 2.

Effects of single-base-pair substitutions on the ompA promoter −33 region in F. johnsoniae. Promoter activities were determined as described in Materials and Methods. Reactions were performed in triplicate, and standard deviations are marked by error bars. Results for each deletion promoter clone were normalized to a promoter activity of 100% for the wild-type (W.T.) ompA clone (Fj29). The sequences and the numbers indicate the base pairs at the specific positions in the wild-type ompA promoter. N.C., negative control.

FIG. 3.

Effects of single-base-pair substitutions on the ompA promoter −7 region in F. johnsoniae. Promoter activities were determined as described in Materials and Methods. Reactions were performed in triplicate, and standard deviations are marked by error bars. Results for each deletion promoter clone were normalized to a promoter activity of 100% for the wild-type (W.T.) ompA clone (Fj29). The sequences and the numbers indicate the base pairs at the specific positions in the wild-type ompA promoter. N.C., negative control.

FIG. 4.

Correlation between promoter activities in F. hibernum and F. johnsoniae. The relative promoter strength in Flavobacterium was normalized to that of the clone carrying pSCH144 (defined as 1.0) and was plotted as a function of the relative promoter strength measured in F. johnsoniae.

FIG. 5.

Relative promoter strength as a function of the GC content of the core ompA promoter element in F. hibernum (A) and F. johnsoniae (B).

FIG. 6.

Distribution of the predicted promoters with 17 to 23 bp separating the −33 and −7 consensus motifs in F. johnsoniae and B. fragilis and alignment analysis of putative promoters with a 19-bp spacer. The noncoding regions located upstream of ORFs in the F. johnsoniae and B. fragilis genomes were scanned by PatScan or RSAT. Please note that we conducted a string pattern search as a primary study with the consensus promoter motifs which may not be necessary for other native promoters to be functional in members of the phylum Bacteroidetes. (A) Distribution of putative promoters with various spacer lengths (17 to 23 bp) in F. johnsoniae. The number of promoters in each group is indicated above the bar. (B) Alignment of the putative promoters with a 19-bp spacer in F. johnsoniae. (C) Distribution of putative promoters with various spacer lengths (17 to 23 bp) in B. fragilis. The number of promoters in each group is indicated above the bar. (D) Alignment of the putative promoters with a 19-bp spacer in B. fragilis.