Hotspots of transcription factor colocalization in the genome of Drosophila melanogaster

Abstract

Regulation of gene expression is a highly complex process that requires the concerted action of many proteins, including sequence-specific transcription factors, cofactors, and chromatin proteins. In higher eukaryotes, the interplay between these proteins and their interactions with the genome still is poorly understood. We systematically mapped the in vivo binding sites of seven transcription factors with diverse physiological functions, five cofactors, and two heterochromatin proteins at approximately 1-kb resolution in a 2.9 Mb region of the Drosophila melanogaster genome. Surprisingly, all tested transcription factors and cofactors show strongly overlapping localization patterns, and the genome contains many "hotspots" that are targeted by all of these proteins. Several control experiments show that the strong overlap is not an artifact of the techniques used. Colocalization hotspots are 1-5 kb in size, spaced on average by approximately 50 kb, and preferentially located in regions of active transcription. We provide evidence that protein-protein interactions play a role in the hotspot association of some transcription factors. Colocalization hotspots constitute a previously uncharacterized type of feature in the genome of Drosophila, and our results provide insights into the general targeting mechanisms of transcription regulators in a higher eukaryote.

Fig. 1.

Detailed view of DamID profiles generated by using genomic tiling arrays. (A) Predicted genes in a 0.5-Mb section of the Adh-cactus region (position 13,700,381 to 14,199,715 on chromosome 2L) of Drosophila. (B) DamID binding ratios of seven transcription factors and Pol II. (C) DamID binding ratios of Gal4p-DBD. Ratios that are statistically significant and at least 2-fold above background are highlighted by colors in B and C.

Fig. 2.

Complete view of DamID profiles of the entire 2.9-Mb Adh-cactus region. (A) DamID data for all 14 Drosophila proteins that were studied. Each probe on the array is represented by a vertical line, the color of which indicates the relative binding of the protein (ranging from white equals background level to black equals strongest binding). (B) Location of hotspots, i.e., probes that are part of cluster 1 in Fig. 4B. (C) Location of all predicted genes in the probed region. Genes that overlap with at least one hotspot are highlighted in orange, and their names are indicated.

Fig. 4.

Definition of hotspot regions and other types of chromatin by SOM analysis and K means clustering of DamID profiles of 14 proteins. (A) SOM analysis visualized for each protein in SOM component planes consisting of a matrix of 24 × 12 hexagonal nodes. Each node includes tiling array fragments with similar binding behavior across all 14 DamID profiles. Nodes that are close to each other in the SOM component plane are more alike than nodes that are farther away. The color scale represents the mean binding Z score in each node. (B Left) Further simplification of SOM data by adaptive K means clustering, resulting in the classification of the combined protein-binding patterns into eight types of chromatin, each marked in a different color. (B Right) The area of each node reflects the number of tiling array fragments included in that node. (C) DamID profile of GAL4-DBD and ChIP profile of Gaf visualized in the same SOM matrix as in A and B. (D) DamID profiles of Bcd DBD and Bcd^K50A visualized in the same SOM matrix as in A and B. Bcd^K50A binds strongly to the hotspots (“type 1,” outlined by a white line), whereas Bcd DBD does not. (E) SOM matrix visualization of (from left to right) DamID profile of the 18-kDa subunit of Pol II, ChIP profile of the CTD subunit of Pol II (data from ref. 27), mRNA expression levels (data from ref. 27), and histone H3.3 over H3 ratios (data from ref. 28). Note the enrichment of all of these transcription-related marks in the hotspot cluster (outlined by a white line).

Fig. 3.

Comparison of Gaf DamID to Gaf ChIP. Gaf DamID-binding ratios across the entire 2.9 Mb Adh-cactus region are plotted on the positive y axis (dark blue), and Gaf ChIP-binding ratios are plotted on the negative y axis (light blue).

Fig. 5.

Sequence signals enriched in the eight different chromatin types shown in Fig. 4B. Information from the TRANSFAC (30) and JASPAR (31) databases was used to construct a position-specific affinity matrix (31) representing the sequence specificity of each transcription factor. Each position-specific affinity matrix was used to predict a profile along the entire Adh-cactus region of the binding affinity relative to the optimal binding sequence. Difference between average for probed regions of a given chromatin type (see Fig. 4B) and all other regions, represented as a t value computed by using a paired t test. Note the increased predicted affinity of colocalization hotspots (type 1 regions) for Gaf, Jra, and Max.

Fig. 6.

Different developmental expression profiles of hotspot-containing genes (red) compared with hotspot-free genes (black) in the Adh-cactus region. Values indicate mean log₂ expression levels (arbitrary units; data taken from ref. 34). P values indicate the statistical significance of the difference for each developmental stage (Wilcoxon test, Bonferroni-corrected for multiple testing); NS, not significant (P > 0.01).