Development of a New Model System to Study Long-Distance Interactions Supported by Architectural Proteins

Abstract

Chromatin architecture is critical for the temporal and tissue-specific activation of genes that determine eukaryotic development. The functional interaction between enhancers and promoters is controlled by insulators and tethering elements that support specific long-distance interactions. However, the mechanisms of the formation and maintenance of long-range interactions between genome regulatory elements remain poorly understood, primarily due to the lack of convenient model systems. Drosophila became the first model organism in which architectural proteins that determine the activity of insulators were described. In Drosophila, one of the best-studied DNA-binding architectural proteins, Su(Hw), forms a complex with Mod(mdg4)-67.2 and CP190 proteins. Using a combination of CRISPR/Cas9 genome editing and attP-dependent integration technologies, we created a model system in which the promoters and enhancers of two reporter genes are separated by 28 kb. In this case, enhancers effectively stimulate reporter gene promoters in cis and trans only in the presence of artificial Su(Hw) binding sites (SBS), in both constructs. The expression of the mutant Su(Hw) protein, which cannot interact with CP190, and the mutation inactivating Mod(mdg4)-67.2, lead to the complete loss or significant weakening of enhancer-promoter interactions, respectively. The results indicate that the new model system effectively identifies the role of individual subunits of architectural protein complexes in forming and maintaining specific long-distance interactions in the D. melanogaster model.

Keywords: CP190; Mod(mdg4); Su(Hw); architectural C2H2 proteins; enhancer–promoter communication; insulator; long-distance interactions.

Figure 1

Integrated constructs and control phenotypes. (A) Schematic representation of attP sites integrated into selected loci on homologous chromosomes and separated by 28 kb. The loxP sites are designated by horizontal arrows, and the pink box indicates the DsRed fluorescent marker. (B) Schematic representation of integrated constructs ES^×4y and WBS^×4w. Enhancers of eye (E), body (B), wings (W), and bristles (Br) are represented by ovals. The FRT (Flp recognition target) and loxP sites are indicated by blue and white horizontal arrows, respectively. Promoters of yellow (y) and white (w) genes are shown as squares. The direction of gene transcription is indicated by black horizontal arrows. The four grey boxes represent the four Su(Hw) binding site multiplicities (S^×4). White rectangles represent the attB integration sites. The yellow and white genes are indicated. (C) Photo showing the representative phenotypes of flies, with the corresponding constructs (marked below the photo) inserted in either the A or B genome locus. Numbers above the photo indicate the scores of white and yellow gene expression in the eyes (e), wings blades (w), and body cuticle (b). Scales of eye and abdominal pigmentation are presented in Figure S2A.

Figure 2

Trans interactions between enhancers and promoters of reporter genes in the model system. (A–E) Schematic representations of interacting transgenes. Thin red arrows indicate transcriptional activation by specific enhancers, and the blue arrow with a cross indicate the loss of the enhancer–promoter interaction. The Su(Hw) (blue ovals), CP190 (green circles), and Mod(mdg4)-67.2 (pink circles) proteins are designated. Brackets represent the deletion of the corresponding elements in the constructs. The photos show the representative phenotypes of the flies with a combination of corresponding constructs (marked below each photograph). Other designations are as in Figure 1.

Figure 3

Fly phenotypes (A–D) from the A-ES^×4y/B-WBS^×4w line with different mutant backgrounds. All designations are as in Figure 1. Sign “-” indicates that eye pigmentation was not scored.

Figure 4

ChIP–qPCR analysis of protein binding to the S^×4 site in the A-ES^×4y/B-WBS^×4w line with different mutant backgrounds. The Su(Hw) binding in the su(Hw)^Δ114/su(Hw)^Δ114 mutant background was detected by anti-FLAG antibodies because antibodies related to the N-terminal domain of Su(Hw) were not detect truncated Su(Hw). Tested mutant backgrounds are indicated by colored boxes, with designations as in Figure 3. IgG, immunoprecipitation with nonspecific IgG. The 62D and ras64B coding regions (Ras) were used as positive and negative controls, respectively. The percentage recovery of immunoprecipitated DNA (y-axis) was calculated relative to the amount of input DNA. Error bars indicate standard deviations of quadruplicate polymerase chain reaction (qPCR) measurements from two independent biological samples of chromatin. Asterisks indicate significance levels: * p < 0.05 and ** p < 0.01 (Student’s t-test).

Figure 5

Model system for testing long-distance cis interactions, (A,B). (C–G) Fly phenotypes from the A-ES×4y,B-WBS×4w/+ and A-EΔS×4y,B-WBS×4w/+ lines with different mutant backgrounds. All designations are as in Figure 1, Figure 2 and Figure 3.