The pattern of Cg expression in the developing visual ganglia was examined by in situ hybridization and by staining with an anti-Cg antiserum. Both analyses gave similar results. Cg is expressed most strongly in dorsal and ventral regions of the optic ganglia, and weakly or not at all in the midline region. cg transcript and protein are reduced in the vicinity of the wingless+ domains. The absence of Cg in the midline region is consistent with the lack of phenotypic effects of cg mutations in this region. Since ectopic wg+ activity is sufficient to induce the ectopic expression of wg target genes in the midline region, it is supposed that other factors are responsible for wg target gene regulation there. The reduction of cg expression found in the vicinity of wg+ target gene expression is consistent with the notion that wg+ induction of its optic lobe targets occurs via the downregulation of Cg expression (Song, 2000).
These observations along with the role of Cg as a negative regulator of wg target genes suggest that Cg expression might be regulated by wg+. Consistent with this notion, three consensus dTCF binding sites were identified within the first intron of the cg locus. To determine whether Cg expression is indeed under wg+ control, animals were generated carrying ectopic wg+ clones. The wg+ clones resulting from recombination between the repeated FRT sites were visualized by their failure to express the CD2 marker. The presence of ectopic wg+-expressing cells could also be inferred by the local nonautonomous induction of the target gene omb. The induction of omb by ectopic wg+ expression coincides with a reduction in Cg expression. This effect of wg+ is nonautonomous, as both the induction of Omb and the reduction of Cg expression have been found to extend beyond the boundary of marked wg+ clones. Cg expression thus appears to be under the nonautonomous control of wg+ activity (Song, 2000).
The dorsoventral axis of the Drosophila visual cortex is patterned by nonautonomous signals expressed at its dorsal and ventral margins. wingless (wg) expression at the margins induces decapentaplegic (dpp), optomotor blind (omb), and aristaless in adjacent domains. Combgap represses Wg target gene expression in the visual cortex. Wg signal reception downregulates combgap expression and derepresses target gene transcription. Combgap participates in a Hedgehog-controlled circuit in the developing wing and leg by regulating the expression of Cubitus interruptus. Combgap is thus a tissue-specific relay between Wingless and its target genes for the determination of cell fate in the visual cortex (Song, 2000).
At hatching, approximately 40 cortical cell precursors form a disc-shaped epithelium on the ventrolateral surface of each brain hemisphere. The epithelium is divided into lamina and medulla precursor zones, which can be distinguished by the expression of Cubitus interruptus (Ci) in the prospective lamina cortex. Cells in two domains at the prospective dorsal and ventral margins of the adult optic lobe begin to express wingless in the mid-first instar stage. dpp expression begins after the onset of wg expression and continues in domains immediately adjacent to the Wg-positive cells. Two additional dorsoventral-specific markers are optomotor blind (omb) and aristaless. Omb is expressed in dorsal and ventral domains that include both the Wg- and Dpp-positive cell populations. Omb-positive glia migrate from these domains toward the lamina midline. aristaless, as assayed by the al04352 enhancer trap insertion (al-lacZ), is expressed in a graded pattern with respect to distance from the Wg-positive cells. The expression of omb, dpp, and al-lacZ is induced by ectopic wg+ expression and absent under conditions of wg loss of function. These observations indicate that Wg is responsible for the expression of these three markers (Song, 2000).
In the wild-type brain, retinal axons project in a crescent-shaped array into the lamina target field. Domains of dpp-expressing cortical cell precursors lie at the ends of the crescent-shaped retinal axon array. In cgk11504 animals, ingrowing retinal axons form an irregular pattern of projections, with axons often straying outside the normal target field. When assayed by either introduction of the dpp-lacZ reporter construct BS3.0 or by staining with anti-Dpp antibody, dpp expression was found to extend, in the cg mutant, toward the midline beyond the normal positions of its dorsal and ventral domains. The domains of omb expression also extend beyond their usual boundaries toward the midline. Expression of the al-lacZ reporter does not diminish in a graded fashion with distance from the Wg domains in cg brains. With respect to all three markers, and on the basis of morphology, a region centered about the dorsoventral midline is relatively unaffected by cg loss of function. Similar results have been obtained with the stronger cgDelta10 deletion allele and with cg1/cgk11504 heterozygotes. Thus cg loss of function results in an extension of dorsal and ventral cell identities toward the midline, while a region centered about the midline remains relatively unaffected (Song, 2000).
A possible explanation for the cg phenotype would be that enhanced or ectopic wingless expression induces ectopic wg target gene expression. To address this issue, wg expression was examined by the use of a sensitive reporter of wg transcription, P{ry,lacZ}17en40, and by staining with an anti-Wg monoclonal antibody. Both assays reveal a normal pattern and level of Wg expression in both homozygous cgk11504 brains and cgk11504 somatic mosaic animals. The cg phenotype therefore does not appear to result from a change in the level or pattern of wg expression (Song, 2000).
The cell autonomy of combgap function was determined by generating somatic cgk11504 clones using the FLP, FRT method. Within cg clones outside of the midline region, Omb, Dpp, and al-lacZ are all expressed ectopically. Clones or portions of clones that fall within the midline region (30% of those examined) appeared phenotypically normal, consistent with the lack of a cg requirement for the midline region in homozygous animals. There are also position-specific effects observed within cg clones. For example, not all cells within a cg clone expressed the marker Dpp. The position-specific ectopic gene activation in cg clones might reflect the activity of other signals involved in cortical cell fate determination. cg thus behaves as an autonomous repressor of omb, dpp, and al-lacZ expression, except in the midline region where it is not required (Song, 2000).
combgap is required for pattern formation in the leg, wing, and eye but does not appear to control Wg target gene expression in any of these tissues. In the developing wing, Wg, expressed at the dorsoventral margin, induces a graded expression of Distal-less (Dll) that diminishes with distance from the Wg-expressing cells. Dll expression is not significantly changed within cgk11504 somatic clones. Wg also controls the induction of sensory neurons at the dorsoventral margin. Examination of cg homozygotes in which these neurons were marked by neuralized-lacZ reveals no difference with respect to wild type. The role of cg in wing, leg, and eye development appears to primarily involve regulation of the Hedgehog pathway transcription factor Cubitus interruptus (Ci). Ci expression is normally restricted to the anterior compartments of the wing and leg discs. Somatic cgk11504 clones in the anterior compartment have strongly reduced Ci expression, while clones in the posterior compartment, where Ci is normally absent, display a low level of Ci immunoreactivity. Ci expression in the posterior compartment is normally repressed by the posterior determinant Engrailed (En). While cg clones at the compartment border display loss of Engrailed expression, clones in more posterior positions display normal En expression. En expression at the compartment border is dependent on Hedgehog signaling. Somatic cg clones also display reduced expression of Dpp, another Hh target gene expressed at the compartment border. Similar observations have been made on the developing leg. These observations reveal that the primary role of Cg in these tissues is to regulate Ci expression. The effect of cg loss of function on Hedgehog signaling thus appears to be indirect (Song, 2000).
To place cg+ activity in the context of the Wg signal transduction cascade, cg homozygotes were examined in which Wg signaling was suppressed by the misexpression of the Drosophila axin gene homolog, Axn. It was supposed that expressing a UAS-Axn construct under the control of the omb-GAL4 (P{GawB}bimd65) driver would create a negative feedback loop in which omb expression would be suppressed. That is, wg+ activation of omb-GAL4 transcription would be countered by GAL4 driven Axn expression. This was indeed the case. Omb expression in omb-GAL4, UAS-Axn brains is greatly reduced in both dorsal and ventral domains and in the glia that migrate into the lamina field . The reduction of Omb expression varies among specimens; the complete absence of Omb expression in at least one domain is observed in ~30% of specimens. Dpp expression is also greatly reduced or undetectable in omb-GAL4, UAS-Axn brains. These effects on Omb and Dpp expression are associated with a defect in the projection pattern of the photoreceptor axons. Similar axon projection phenotypes have been observed with wg loss of function (Song, 2000).
The loss of cg function is epistatic to Axn misexpression. The developing visual ganglia of animals homozygous for cgk11504, harboring the omb-GAL4 and UAS-Axn transgenes, displays ectopic expression of the wg targets Dpp and Omb like that found in cgk11504 animals. Ectopic Axn does suppress the high level of Omb expression in the normal Omb domains in the cgk11504 background, consistent with the notion that high-level Omb expression remains wg+ dependent in cgk11504. A second set of experiments were performed utilizing the heat shock-inducible P{hsGAL4} driver to express the P{UAS-Axn} transgene. Multiple heat shocks during the larval period were applied to animals harboring the transgenes in either a wild-type or cg background. Similar results were obtained to those with the omb-GAL4 driver. These observations argue that cg functions downstream of Axn (Song, 2000).
Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.