BOCEs are the second repeated mosaic element found on the E. coli chromosome (Bachellier S. et al., 1999). BOCEs are made of at least a boxC sequence (Bergler H. et al., 1992), associated with other short conserved motifs, called D and E.
BoxC sequences are 56 bp-long imperfect palindromes, with a pyrimidine-rich 5' end (the tail) and a purine-rich 3' end (the head). Below is shown the consensus of boxCs and their schematic drawing.
Like BIMEs, boxCs have been found only on chromosomal DNA. These sequences are non-coding, found only in intergenic regions. They also exist in Klebsiella and Enterobacter, but NOT in Salmonella. On the E. colichromosome, 23 extragenic regions contain either 1 or 2 boxCs, totalling 33 boxCs. Further examination of the nucleotide sequence of E. coli boxCs led us to separate them in two groups, called type 1 (in red above) and type 2 (in blue above) according to critical positions.
When two boxCs are found in the same region, they are always in inverted orientation, and in all but one case, there is one type 1 and one type 2. When the two boxCs are tail-to-tail, two conserved sequence motifs can be associated with them. A 11 bp-long sequence is found between the tails of a type 1 and a type 2 boxCs (or can be flanking the tail of any type of single boxC), and a 25 bp-long sequence is sometimes flanking the head of a type 1 boxC. The consensus derived from the alignment of the D and E sequences are given in the sketch below. We called this new mosaic element BOCE for boxC Composite Element. A similar organisation of boxCs and conserved motifs is found in Klebsiella and Enterobacter.
As was noticed for BIMEs, boxC sequences may be differently located in various strains of the same species (compare the araA-araD intergenic regions in E. coli K-12 and ECOR8 and the recA 3' end in E. coli K-12 and Shigella flexneri). The functional significance of boxCs and BOCEs remains unclear. It has been hypothesized that they could stabilise mRNA by forming very stable secondary structures, and/or be involved in regulation of transcription. They are potential protein binding sites, but there is no experimental data to confirm these hypotheses.
Back to the table of contents
Please send questions and comments to Sophie.Bachellier@pasteur.fr.