Book of Abstracts: Albany 2005
Functional Interplay of Structure-specific and Sequence-specific Elements in the Bacterial Promoters
The identification of DNA-protein and protein-protein contacts in transcription complexes is currently a necessary stage of the description of each novel promoter. The experimental approaches used for this purpose allow affiliating the given promoter to the regulon of one of seven alternative σ subunits of E.coli RNA polymerase and determining the dependence of its activity on known protein and non-protein factors. The nucleotide sequence of the promoter DNA itself contains this information; therefore, if the transcription start point is known, the promoter type and the pattern of its protein regulation can be predicted with a high probability. An opposite task is much more difficult. The detection of the gene regulatory region usually does not permit identification of the transcriptional start point even if σ-specificity and the type of regulation are estimated. Several promoter-like signals are typically predicted, and the true promoter usully (>50%) does not gain the maximal score. This happens due to the extreme sequence variation within consensus elements and participation of other promoter-specific elements in the transcription complex formation, which are hardly taken into account. They include modules structure-specifically interacting with RNA polymerase α subunits as well as elements determining the overall architecture of the transcription complex by kinking or bending. Informational content in some regions upstream from the conservative elements is rather high, reflecting the presence of these additional elements. Their functional manifestation depends on the positioning in the promoter DNA, and it is generally assumed that they should be phased with the conservative elements recognized by σ.
Surprisingly we found, that informational content of the upstream sequences tends to decrease rather than increase if they are aligned in respect to the conservative elements. That means that structure-specific and sequence-specific elements in the promoter DNA are independent from each other and assumes probable linkage of additional elements with the transcriptional start points. Position-dependent scoring of these elements as additional promoter-specific determinants allowed correct detection of the start points for more than 80% of known promoters, which was much higher than in the case when only conservative elements were taken into account. The transcription start point selection by the RNA polymerase may, therefore, be dependent on the remote elements. Gene expression data were used to estimate the correlation between the values of thus estimated promoter scores and the transcription efficiency of corresponding genes in vivo. The set of analyzed genes contained both constitutive and regulated species. Thus the overall correlation was not high but the data obtained provide a possibility to select genes subjected to positive or negative regulation assuming a possibility to use our software as a classifying tool for yet uncharacterized genes.
Supported by RFBR grant 03-04-48339 and RFBR-Ministry of Industry and Science (Moscow Region) (grant 04-04-97280).
O. N. Ozoline1,*
1Institute of Cell Biophysics RAS