Genome engineering-based analysis of Bearded family genes reveals both functional redundancy and a nonessential function in lateral inhibition in Drosophila

Abstract

Lateral inhibition mediated by Notch receptor signaling regulates the determination of sensory organ precursor cells (SOPs) in Drosophila. The selection of SOPs from proneural cluster cells appears to rely on a negative feedback loop linking activation of the Notch receptor to downregulation of its ligand Delta within each cell. The molecular basis of this regulatory feedback mechanism is not known. Here, we have tested the role of the Bearded (Brd) family genes in this process. The Drosophila genome encodes eight Brd family members that interact with the E3 ubiquitin ligase Neuralized (Neur) and act as inhibitors of Neur-mediated Delta signaling. Genome engineering technologies were used to create specific deletions of all eight Brd family genes. We find that the Brd family genes malpha, m4, and m6 encoded by the Enhancer of split Complex (E(spl)-C) are dispensable for Drosophila development and that deletion of the five Brd family genes encoded by the Brd Complex only reduces viability. However, deletion of all Brd family genes results in embryonic lethality. Additionally, the malpha, m4, and m6 genes act redundantly with the other five Brd family genes to spatially restrict Notch activation in stage 5 embryos. These data reveal that the Brd family genes have an essential but redundant activity. While the activity of all eight Brd genes appears to be dispensable for SOP determination, clone border studies indicate that both the relative activity levels of Neur and Brd family members influence competition for the SOP fate during lateral inhibition. We propose that inhibition of Neur-Delta interaction by Brd family members is part of the feedback loop that underlies lateral inhibition in Drosophila.

Figure 1.—

Genome engineering of the E(spl)-C and Brd-C. (A) Structure of the E(spl)-C locus, Df(3)E(spl)δ-6 deficiency, and BAC-based DpE(spl)δ-8 and DpE(spl)δ-8Δα46 duplications. The E(spl)-C encodes seven bHLH repressors (blue) and four Brd family genes (green). The P-elements used to generate Df(3)E(spl)δ-6 are shown (orange triangles). The mα, m4, and m6 genes are deleted in DpE(spl)δ-8Δα46. (B) Structure of the Brd-C, Df(3)Brd-C1 deficiency, DpBrd-C and DpCG13466 duplications. The P-elements used to generate Df(3)Brd-C1 are shown (orange triangles).

Figure 2.—

Brd family genes act redundantly. (A) Wild-type embryos express sim in a single row of cells at stage 5. (B) sim transcripts are not detected in neur^IF65 mutant embryos. (C) Deletion of the mα, m4, and m6 genes in Dp(3;2)E(spl)δ-8Δα46; Df(3)E(spl)δ-6 embryos does not affect sim expression. (D) Deletion of the Brd-C in Dp(3;2)E(spl)δ-8; Df(3)Brd-C1 Df(3)E(spl)δ-6 embryos leads to the ectopic expression of the sim gene in a few cells dorsal to the mesectoderm. (E) Deletion of all eight Brd family members in Dp(3;2)E(spl)δ-8Δα46; Df(3)Brd-C1 Df(3) E(spl)δ-6 embryos leads to strong ectopic expression of the sim gene in 3–5 cell rows. (F) The sim gene is weakly expressed in Dp(3;2)E(spl)δ-8Δα46; Df(3)Brd-C1 neur^IF65 Df(3)E(spl)δ-6 embryos, indicating that the phenotype seen upon loss of the Brd-C, mα, m4, and m6 genes is largely rescued by loss of zygotic neur activity. (G) Spatial regulation of Dl signaling by Brd family proteins. Mesodermal cells express neur but not Brd family genes. Brd family proteins inhibit Neur in nonmesodermal cells. As a result, Neur-dependent Dl signaling is restricted to the mesoderm. Notch is activated in cells in direct contact with the mesoderm, as shown by expression of the sim gene.

Figure 3.—

Brd family genes are not required for SOP determination. Sensory organ formation was analyzed in adult cuticle preparations (A–D) and dissected pupal nota (E). Controls include wild type (A) and Df(3)E(spl)δ-6 Dp(3;2)E(spl)δ-8 flies (B). Deletion of the mα, m4, and m6 genes in Dp(3;2)E(spl)δ-8Δα46; Df(3)E(spl)δ-6 flies did not affect bristle formation (C). Likewise, loss of Brd-C genes in Df(3)Brd-C1 clones marked by yellow (y+) did not perturb bristle development (D). Loss of Brd-C genes in Df(3)Brd-C1 clones marked by GFP (green) in Dp(3;2)E(spl)δ-8Δα46; Df(3)E(spl)δ-6 pupae did not detectably affect SOP specification (sensory cells marked by Sens in red; E).

Figure 4.—

The relative levels of neur and Brd family genes influence fate decisions. Competition for the adoption of the SOP fate was studied by scoring the genotype of SOPs located along the clone border separating cells that differ in the copy number of Brd family (A and B) and neur genes (C) (as numbered in A–C). Df(3)Brd-C1 mutant clones (A) and mα m4 m6 triple mutant clones (B) were marked by GFP whereas cells with one or two copies of either the Brd-C or the mα, m4, and m6 genes do not express GFP. In panel C, the three different genoptypes were identified: neur mutant cells were marked by strong GFP expression, cells with one copy of the wild-type neur gene were marked by weak GFP expression, and cells with two copies of the wild-type neur gene were GFP-negative. SOPs were identified using Sens (red). SOPs located along clone borders are indicated by arrows (GFP-negative SOPs) and arrowheads (GFP-positive SOPs). (D) Plots showing the percentage of GFP-negative SOPs (in red) and GFP-positive SOPs (in yellow) along clone borders (% values are indicated above each bar; n is the number of SOPs scored along clone borders; “gene dosage” indicates the number of wild-type copies). Genotypes 1–7 are described in the materials and methods section. Competition along Brd-C mutant clones was monitored in a wild-type background (2) as well as in a mα m4 m6 triple mutant background (3). Similar results were obtained. Control wild-type clones were studied for each corresponding chromosomal arm (1, 4, and 6). Significantly more GFP-positive SOPs were observed in 2 and 3 compared to 1, and in 5 compared to 4. In contrast, more GFP-negative SOPs were observed in 7 compared to 6 (χ² test, P < 0.01).

Figure 5.—

Model A negative feedback loop linking activation of Notch at the cell surface to downregulation of Dl within the same cell operates during lateral inhibition. We propose that inhibition of Neur–Dl interaction by Brd family members is part of this feedback loop, so that the transcriptional upregulation of Brd family genes by activated Notch results in the downregulation of Dl activity in non-SOP cells. Transcriptional regulation of Brd genes by Notch is indicated in black while post-transcriptional regulatory steps are in blue.