ECCB 2002 Poster sorted by: Author | Number

Next | Previous poster (in order of the view you have selected)
Title: Composing a promoter model for antibacterial response of epithelial cells
P150
Shelest, Ekaterina; Kel, Alexander; Gößling, Ellen; Wingender, Edgar

ksl@gbf.de
GBF German Research Centre for Biotechnology, Mascheroder Weg 1, D-38124 Braunschweig, Germany;BIOBASE GmbH, Halchtersche Str. 33, D-38304 Wolfenbüttel, Germany

Binding of a bacteria to a eukaryotic cell triggers a complex network of interactions in and between both cells. P.aeruginosa is a pathogen that causes acute and chronic lung infections by interacting with the pulmonary epithelial cells (DiMango et al., 1998; Smith et al., 2001). We use this example for understanding the ways of triggering the eukaryotic cell(s) response, leading us to understanding the details of the inflammatory process in general. After adhesion of P.aeruginosa to the epithelial cells, the response of these cells is triggered by at least two distinct agents: bacterial lipopolysaccharides (Zhang and Ghosh, 2001) or bacterial pilins or flaggelins (McNamara et al., 2001). Both pathways lead to NF-kappaB activation. NF-kappaB is well-known to participate in these kinds of reactions.In this work we searched for additional transcription factors that may cooperate with NF-kappaB and/or complement its effects. Identifying them would give a clue to realizing the mechanisms of their activation, in particular those connected with [Ca2+] alteration. The information about the participating transcription factors would enable us to construct a model for searching other, still not identified, target genes which are potentially involved in defensive mechanisms.
NF-kappaB often appears to be a part of composite elements (CE). Composite elements are combinations of two or more transcription factor (TF) binding sites which provide synergistic action of the TFs, qualitatively different from a purely additive effect. The most abundant CEs with NF-kappaB are those with C/EBP moiety. C/EBP is also interesting because of two further reasons: first, it is known to participate in immune response; second, it is one of the target factors in the pathways triggered by increase of [Ca2+]. To use the weight matrix approach we had first to re-evaluate the existing matrices for C/EBP binding sites. There are 8 matrices for C/EBP in TRANSFAC, but all of them exhibit rather weak consensi. We decided to make more precise matrices for C/EBP dividing the whole set into subgroups using a subtractive approach. We came up with a set of matrices, each of them representing a subset of C/EBP sites. To make the search for binding sites more comprehensive we combined these matrices in such a way that for a defined rate of false negatives (FN) a minimal rate of false positives (FP) is achieved, the overlap between individual matrices being minimized. FP is represented by fr, the frequency of matches per nucleotide of random sequences.The FP rate could be reduced by about two thirds compared to one of the previously used search patterns (CEBP_02).
Thorough analysis of the known CEs of C/EBP/NF-kappaB type led us to develop a model describing the relative orientation, distance and scoring of the constituents. We used this model to analyze our collection of human 5'-flanking sequences (about 13000 5'-upstream sequences) setting the parameters such to re-identify 80% of the true positives. Under these conditions, we identified about 200 genes as harboring at least one potential C/EBP/NF-kappaB composite element. Erasing all those that encode hypothetical products, we end up with a list of 135 genes which can be checked for plausibility.
Although C/EBP/NF-kappaB CE seems to be very important, it is not likely that it could be the only structure responsible for the cells' antibacterial reply. Using transformation of the nucleotide sequence into the sequence of matches of TFs and analysing it, we identified a new potential composite element: NF-kappaB/Sp1. Exhaustive literature search enabled us also to include CREB and Elk-1 as candidates into the promoter model. So, now the model consists of 5 TFs (NF-kappaB, C/EBP; CREB; Elk-1, Sp1), 3 of which can be joined into 2 composite elements: C/EBP/NF-kappaB and/or NF-kappaB/Sp1.
After defining the thresholds, relative distances, etc. for all components of the model and applying the model to the set of human promoter regions we can select potential target genes.
[1] DiMango E., Ratner A.J., Bryan R., Tabibi S., ansd Prince A. (1998) Activation of NF-kB by adherent Pseudomonas aeruginosa in normal and cystic fibrosis respuratory epithelial cells. J Clin Invest., 101,11,2598-2606.
[2] Goessling,E., Kel-Margoulis,O.V., Kel,A.E. and Wingender,E. (2001) MATCH - a tool for searching transcription factor binding sites in DNA sequences. Application for the analysis of human chromosomes. In: Proceedings of the German Conference on Bioinformatics GCB 2001, Wingender, E., Hofestädt, R., and Liebich, I. (eds.), Braunschweig, pp. 158-161.
[3] McNamara N., Khong A., McKemy D., Caterina M., BoyerJ., Julius D., and Basbaum C. (2001) ATP transduces signals from ASGM1, a glycolipid that functions as a bacterial receptor. Proc. Natl. Acad. Sci. USA, 98, Issue 16, 9086-9091.
[4] Smith, R.S., Fedyk, E.R., Springer, T.A., Mukaida, N., Iglewski, B.H., and Phipps, R.P.(2001) IL-8 production in human lung fibroblasts and epithelial cells activated by the Pseudomonas aeruginosa autoinducer N-3-oxodododecanoyl homoserine lactone is transcriptionally regulated by NF-kB and activator protein-2. The Journal of immunology, 167, 366-374.
[5] Wingender,E., Chen,X., Fricke,E., Geffers,R., Hehl,R., Liebich,I., Krull,M., Matys,V., Michael,H., Ohnhaeuser,R., Prueß,M., Schacherer,F., Thiele,S. and Urbach,S. (2001) The TRANSFAC system on gene expression regulation. Nucleic Acids Res. 29, 281-283.
[6] Zhang, G., Ghosh, S. (2001) Toll-like receptor?mediated NF-B activation: a phylogenetically conserved paradigm in innate immunity. J Clin Invest, 107, 1, 13-19.