Genome-wide analysis of transcription factor binding sites and their characteristic DNA structures

Abstract

Background: Transcription factors (TF) regulate gene expression by binding DNA regulatory regions. Transcription factor binding sites (TFBSs) are conserved not only in primary DNA sequences but also in DNA structures. However, the global relationship between TFs and their preferred DNA structures remains to be elucidated.

Results: In this paper, we have developed a computational method to generate a genome-wide landscape of TFs and their characteristic binding DNA structures in Saccharomyces cerevisiae. We revealed DNA structural features for different TFs. The structural conservation shows positional preference in TFBSs. Structural levels of DNA sequences are correlated with TF-DNA binding affinities.

Conclusions: We provided the genome-wide correspondences of TFs to DNA structures. Our findings will have implications in understanding TF regulatory mechanisms.

Figure 1

The landscape of TFs and their characteristic binding DNA structures. Rows represent TFs, and columns represent DNA structures. For each TF-structure pair, if structural conservation rate of its real TFBSs is significantly higher (P < 0.05, after Bonferroni correction for multiple testing) than those in 10,000 randomized experiments in which sequence conservation rates are the same as that of real TFBSs, it was colored red, otherwise it was colored black.

Figure 2

The refined landscape of TFs and their characteristic binding DNA structures. Using three criteria, we identified 27 pairs of TF-structure correspondences. TFBSs of these TFs are conserved in the corresponding DNA structures, independent of sequence conservation.

Figure 3

Structural conservation shows positional preferences in TFBSs. Real conservation rates of structures are shown for each position of TFBSs (black). Low levels correspond to high conservation rates. Average conservation rates of structures in 10,000 randomized experiments in which TFBSs are generated from real PWMs are also shown (red). Error bars were calculated by standard deviation. The names of TF-structure correspondences are also indicated. TFs show significantly higher conservation in their corresponding structures than those based on 10,000 randomized experiments in the following specific positions in TFBSs (P < 0.05, after Bonferroni correction for multiple testing): (A) The thirteenth and fourteenth positions; (B) The first and sixth positions; (C) The fourth position; (D) The first, second, third, fourth and fifth positions; (E) The second, third, fourth and fifth positions; (F) The first, second, third, fourth, fifth, seventh and eighth positions; (G) The third and fourth positions; (H) The third and fourth positions; (I) The third position; (J) The first, fourth, fifth, sixth, seventh, eighth and ninth positions; (K) The second, third, fourth and fifth positions.

Figure 4

Structural levels of DNA sequences are significantly correlated with TF-DNA binding affinities. (A) Shown is a scatter plot comparison between structural (duplex disrupt energy) levels of 8-mers and TF Dal80 binding affinities of 8-mers. To control for nonspecific protein-DNA binding, we restricted the analysis to the top 500 out of 65,536 8-mers with the highest Dal80 binding affinities. The Pearson correlation and the p-value of the scatter plot are indicated. The duplex disrupt energy property of DNA sequences facilitates Dal80 binding to DNA. (B) Same as (A), but for TF Swi4 and structure roll. To control for nonspecific protein-DNA binding, we restricted the analysis to the top 500 out of 65,536 8-mers with the highest Swi4 binding affinities. The roll property of DNA sequences inhibits Swi4 binding to DNA.