Distance and helical phase dependence of synergistic transcription activation in cis-regulatory module

Abstract

Deciphering of the spatial and stereospecific constraints on synergistic transcription activation mediated between activators bound to cis-regulatory elements is important for understanding gene regulation and remains largely unknown. It has been commonly believed that two activators will activate transcription most effectively when they are bound on the same face of DNA double helix and within a boundary distance from the transcription initiation complex attached to the TATA box. In this work, we studied the spatial and stereospecific constraints on activation by multiple copies of bound model activators using a series of engineered relative distances and stereospecific orientations. We observed that multiple copies of the activators GAL4-VP16 and ZEBRA bound to engineered promoters activated transcription more effectively when bound on opposite faces of the DNA double helix. This phenomenon was not affected by the spatial relationship between the proximal activator and initiation complex. To explain these results, we proposed the novel concentration field model, which posits the effective concentration of bound activators, and therefore the transcription activation potential, is affected by their stereospecific positioning. These results could be used to understand synergistic transcription activation anew and to aid the development of predictive models for the identification of cis-regulatory elements.

Figure 1. Distance dependence of transcription activation by two GAL4-VP16 dimers.

GAL4 binding sites (gray box) with designed spacing from 0 to 48 bp, in steps of 2 bp, were placed 22 base pairs upstream of the TATA box of adenovirus early gene 4 (red box), followed by the coding sequence of luciferase gene. Normalized luciferase activity is plotted versus GAL4 binding site spacer length and shows local peak values at 8, 18, 28 and 36 bp. Each data point and error bar came from three parallel replicates. Each experiment was repeated twice.

Figure 2. Helical phase dependence of transcription activation by two GAL4-VP16 dimers.

A) Structural reconstruction of binding modes of two GAL4-VP16 dimers on the designed adenovirus promoter with two GAL4 binding sites. The GAL4-DBD dimers are shown in cartoon representation, from the experimental coordinates in PDB code 3COQ , bound to the promoter region. Two views are shown, longitudinal (left) and transervse (right). B) Overlay of structural reconstructions on the luciferase activity assay from Figure 1. C) Rationale for EMSA assay. Longitudinal views of structural reconstruction of two GAL4-VP16 dimers bound to promoter, left = dimers bound on the same face of DNA, right = dimers bound on opposite faces of DNA. D) EMSA. Lanes refer to experiments completed with templates increasing spacer length between GAL4 binding sites. E) Plot of logarithm(migration distance of saturated or free DNA probes in EMSA) versus GAL4 spacer length. The local minima of binding complex mobility were proven to be at separation distances of 8, 18, 28 bp and 38/40 determined according to the mobility of free probes as marker line. Note the inverted scale of log(migration distance).

Figure 3. Helical phase dependence was not affected by the spatial relationship between activators and transcription complex.

A) Structural reconstruction of promoter occupied by TFIIA/TBP complex and two GAL4-VP16 dimers. Four templates, 142, 145, 193 and 196, are shown. The distance between the proximal GAL4 binding site is 22 bp in templates 142 and 145, and 26 bp in templates 193 and 196. B) Luciferase activity assay using a series of transcription templates bearing two GAL4 binding sites (gray box) with increasing spacer length, with the proximal GAL4 binding site placed 26 base pairs upstream of the TATA box of adenovirus early gene 4 (red box). Each data point and error bar came from three parallel replicates. Each experiment was repeated twice.

Figure 4. Multiple evenly distributed activators around DNA double helix function more effectively.

A) Structural reconstructions of templates. All five templates, 190, 191, 192, 187 and 189, contain a proximal GAL4 binding site 22 bp upstream to the TATA box. Templates 190, 191 and 192 contain two additional GAL4 binding sites, designed to have GAL4-VP16 dimers bound on the same side of the DNA double helix (template 190), arranged regularly spaced around the DNA (template 191), or opposite from the proximal GAL4-VP16 (template 192). Templates 187 and 189 contain three additional GAL4 binding sites, with three arranged on the same face of the DNA as the transcription complex (template 187), or all four arranged opposite (template 189). B) Transcription activation assay with templates 142, 150, 190, 191, 192, 187 and 189. Each data point and error bar came from three parallel replicates. Each experiment was repeated twice.

Figure 5. Evaluation of concentration dependence.

Evenly distributed activators around the DNA double helix activate transcription more effectively at all activator concentrations. Transcription activation assays were completed with templates 142, 145 (A), 190 and 191 (B), with increasing concentration of GAL4-VP16 expression plasmid in the luciferase assay. Each data point and error bar came from three parallel replicates. Each experiment was repeated twice.

Figure 6. Distance and phase dependence of synergistic transcription activation by two ZEBRA transcription activators.

Two ZEBRA binding sites (ZIIIB, oblique-line box) with increasing spacers length (in steps of 2 bp) were placed 22 bp upstream to the TATA box (red box). A) Transcription activation luciferase assay using templates designed with increasing distance between ZIIIB sites. B) EMSA assay. Lanes show experiments completed with templates designed with increasing distance between ZIIIB sites. Below the main gel is zoom in of region of the gel showing two ZEBRA molecules bound. C) Transcription activation assays were completed with templates Z4Z (ZIIIBs were 4 bp spaced) and Z8Z (8 bp spaced), with increasing concentration of ZEBRA expression plasmid in the luciferase assay. Each data point and error bar came from three parallel replicates. Each experiment was repeated twice.

Figure 7. Concentration field model of transcription activation.

Transcription activator binds on the promoter and recruit transcription machinery components (TF) to the TATA box to form the transcription initiation complex. If multiple activators are bound on the same side of DNA, the proximal activator provides steric hindrance to the protein-protein interactions mediated by the distal activator, therefore decreasing the total recruitment of TFs to the TATA box. Conversely, an arrangement of activators on opposite faces of the DNA, activators are free to recruit TFs.