Daniela Ham, Daniel Svozil, Bohdan Schneider


Centre for Biomolecules and Complex Molecular Systems, Institute of Organic Chemistry and Biochemistry, Academy of Sciences of Czech Republic, Fleming Square 2, 166 10 Prague 6, Czech Republic


Keywords: protein-DNA complexes, conformation, structural bioinformatics


Despite extensive effort to understand structural features of protein-DNA complexes governing specific DNA-protein interactions, neither general, nor simple rules have been found to date [1, 2]. Understanding of these principles would serve as a guideline for protein-DNA recognition in various families of regulatory and other DNA-binding proteins. We have recently reported conformational variability of DNA based on the analysis of torsion angles in almost eight thousand dinucleotides from well resolved DNA x-ray structures [3]. Although double helical A and B conformations represent majority of the population, we have defined many substates that represent A-to-B, B-to-A, BI-to-BII, and other unusual conformations. Such a detailed conformational characterization allowed us, in the current work, to ascribe different DNA conformational substates to four groups of transcription regulatory proteins, three groups of DNA-binding proteins and four groups of enzymes. The majority of examined DNA-protein complexes exhibit the BI and BII conformations as the most populated substates. These conformations are observed for binding with different protein motifs such as helix-turn-helix, helix-loop-helix, leucine zipper and zinc fingers. Within the BI and BII DNA conformations, no rules governing specific recognition of DNA sequences by individual protein structural motifs were observed. This conclusion confirms the results of the previous study [1] that the recognition via certain motifs does not depend on functional characteristics of protein-DNA complexes. In contrast to these findings, here we also show that other proteins in DNA-protein complexes bind a substantial number of A type or A-to-B type of DNA conformations. Analysis of these interactions suggests that different protein domains, such as TBP-like motif, are more specific for classification of protein-DNA complexes according to their function in transcription. Identification of amino acids involved in side chain/base interaction among regulatory and DNA binding proteins distributed into distinct families is also being performed to further explore the common and unique principles responsible for protein-DNA recognition.


This work was supported by grants LC512 from the Ministry of Education of the Czech Republic and NSF grant DBI 0110076 to the NDB.


1. P. Prabakaran et al., Structure, 14 (2006) 1355-1367

2. C.O. Pabo & L. Nekludova, J. Mol. Biol., 301 (2000) 597-624

3. B. Schneider & D. Svozil, Acta Cryst. (2006). A62, s133.