Dual transcriptional activities of SIX proteins define their roles in normal and ectopic eye development

Abstract

The SIX family of homeodomain-containing DNA-binding proteins play crucial roles in both Drosophila and vertebrate retinal specification. In flies, three such family members exist, but only two, Sine oculis (So) and Optix, are expressed and function within the eye. In vertebrates, the homologs of Optix (Six3 and Six6) and probably So (Six1 and Six2) are also required for proper eye formation. Depending upon the individual SIX protein and the specific developmental context, transcription of target genes can either be activated or repressed. These activities are thought to occur through physical interactions with the Eyes absent (Eya) co-activator and the Groucho (Gro) co-repressor, but the relative contribution that each complex makes to overall eye development is not well understood. Here, we attempt to address this issue by investigating the role that each complex plays in the induction of ectopic eyes in Drosophila. We fused the VP16 activation and Engrailed repressor domains to both So and Optix, and attempted to generate ectopic eyes with these chimeric proteins. Surprisingly, we find that So and Optix must initially function as transcriptional repressors to trigger the formation of ectopic eyes. Both factors appear to be required to repress the expression of non-retinal selector genes. We propose that during early phases of eye development, SIX proteins function, in part, to repress the transcription of non-retinal selector genes, thereby allowing induction of the retina to proceed. This model of repression-mediated induction of developmental programs could have implications beyond the eye and might be applicable to other systems.

Fig. 1.

Sine oculis and Optix have limited intrinsic ability to activate transcription. (A-C) Confocal images of third instar larval eye-antennal discs. cb49-GAL4 drives expression of the GFP reporter at low levels in all cells of the eye-antennal disc. Arrows in B and C mark the location of ectopic eyes. (D) Transcriptional activation of UAS-lacZ in yeast by Optix-FL, DSix4-FL, So-FL, So-ΔNT, So-ΔSD, So-ΔHD and So-ΔCT. Asterisks indicate that So-ΔSD and So-ΔCT failed to activate transcription of the UAS-lacZ reporter. (E) Transcriptional activation of UAS-HIS3 by So, Ey and Eya in yeast. Assay is carried out in the presence of increasing amounts of the 3AT histidine biosynthesis inhibitor to determine the relative strength of activation potential. (F) Schematic of So full-length and deletion constructs.

Fig. 2.

Transcriptional repressor forms of So and Optix initiate ectopic eye formation. (A,E) Schematic of So-VP16, So-EN, Optix-VP16 and Optix-EN constructs used in this study. (B,F) Immunoblot (B) and western blot (F) demonstrating that So-VP16 and Optix-VP16 proteins are made in S2 cells and Drosophila embryos. (C,D,G,H) Scanning electron microscopy and confocal images of ectopic eyes (arrows) formed by expression of So-EN and Optix-EN.

Fig. 3.

An activator form of So rescues all retinal defects associated with the so^[1] loss-of-function mutant allele. (A-F) Scanning electron microscopy images of adult Drosophila compound eyes and heads.

Fig. 4.

The antennal selector gene cut is a potential target of both Sine oculis and Optix. (A,B,D-M) Confocal images of third larval instar eye-antennal discs. Arrows in D,I indicate repression of ct expression; in G,L, they indicate de-repression of ey expression; in E,J,F,K, they indicate the formation of photoreceptors. (C) Model in which Ey and Cut mutually repress each other’s transcription.

Fig. 5.

Regulatory relationship between Eyeless, Sine oculis and Cut in the eye-antennal disc. (A-G) Confocal images of third instar eye-antennal discs. Brackets in A-D mark the zones of ct-positive cone cells. (H) Scanning electron microscopy of an adult head. Arrows indicate a partial antennal segment in the place of the eye.

Fig. 6.

Expression of activator (So-VP16) and repressor (So-EN) forms of Sine oculis differentially affect the developing eye. (A,C-H,J-N) Scanning electron microscopy images of adult compound eyes. Arrow in H indicates glazed portion of eye. (B,I) Light microscope images of adult compound eye retinal sections.

Fig. 7.

Formation of SIX-Eya and SIX-Gro transcriptional complexes. Immunoblots of biochemical interactions from Kc167 cells between the Drosophila SIX proteins, Eya and Gro. Production of individual proteins is shown in the nuclear lysate (NL) lanes. Successful isolation of individual proteins is shown in the immunoprecipitation (IP) lanes. Specificity of the pull-downs is shown in the mock IP lanes. Protein interactions are shown in the immunoblot (IB) lanes.

Fig. 8.

Model comparing normal and ectopic eye development. In this model, the first step in ectopic eye development is the repression of the antennal/head capsule specifying gene ct. In normal eye development, this is achieved through interactions between the SIX proteins So and Optix with co-repressors. In the second step, Ey activates the entire retinal determination network through So-Eya and Optix-Eya complexes. These two steps then lead to the promotion of eye development.