BIOLOGICAL OVERVIEW

It is possible to artificially direct eye morphogenesis to any appendage of the body. Genetically engineered sleight of hand has been used to express the eyeless gene in various tissues throughout the body of the fly in order to clarify an understanding of development.

Directed gene expression in such experiments is carried out using Enhancer Trap GAL4 lines. GAL4 is a gene activator. It turns genes on and is easily manipulated using genetic engineering techniques to modify the original genome. A genetic vector carrying the yeast GAL4 gene is randomly integrated into various sites along the genome. Some GAL4 vectors insert next to enhancers. These are promoter-like regions that regulate the induction of specific genes. The GAL4 vector comes under the control of enhancers expressed in various tissues of the fly. These modified genomes and the animals that carry them are termed enhancer trap GAL4 lines. Different enhancer trap GAL4 lines express GAL4 in different tissues depending on the activity of the enhancer into which the vector has inserted.

Other lines have been genetically engineered by introduction of a vector carrying the eyeless gene. The gene is regulated by a promoter region, termed an upstream activating sequence (UAS), activated by GAL4. By placing the eyeless gene under the regulation of GAL4, the gene can be activated wherever gene activator GAL4 is expressed. When these two lines are cross bred and the progeny examined, eyeless appears to be activated in the various tissues expressing the GAL4 transcription factor.

In this manner, by targeting expression of the eyeless complementary DNA in various imaginal disc primordia, ectopic eye structures have been induced on Drosophila wings, legs, and antennae. The ectopic eyes appear morphologically normal, consisting of groups of fully differentiated ommatidia with a complete set of photoreceptor cells. These experiments form the basis for considering eyeless to be the master control gene for eye morphogenesis (Halder, 1995).

Because homologous genes are present in vertebrates, ascidians (sea squirts), insects, cephalopods (squids and octopus), and nemerteans (worms), eyeless may function as a master control gene throughout the metazoa (Halder, 1995). eyeless is not the only homeodomain functioning early in eye morphgenesis. sine oculis is expressed in the optic primordium during stage 5 (the cellular blastoderm). optimotor blind, a brachyury homolog, is expressed in the optic placode from stage 12 onward (Cheyette, 1994).

The model for eyeless function suggests eyeless is required for the determination of cell fate in the eye (see Specification of the eye disc primordium and establishment of dorsal/ventral asymmetry). The embyonic and larval expression patterns of ey are consistent with the early requirement of this gene: it is expressed in the eye primordia in the embryo and before the morphogenetic furrow at the time of photoreceptor determination in the third instar larva (Halder, 1995). A subsequent study shows that ey is not only required at earlier developmental stages for the initiation of photoreceptor development, but also in later stages when terminal photoreceptor-specific genes are expressed. A novel form of ey is expressed in the adult stage. It contains both the first and the second exons. Exon 2 is in frame with exon 1 and its 5' boundary is 15 base pairs in front of the first methionine reported in the D1 transcript. Late ey expression begins in the late pupal stage and coincides with rhodopsin expression. The transition from the patterning of the eye disc to the expression of terminal photoreceptor-specific genes occurs during late pupal development. There is an initial decrease of ey expression in the first half of pupal life, but ey transcripts start to accumulate again starting at 2/3 of the way through pupal development (Sheng, 1997).

Eyeless directly regulates rhodopsin 1 (rh1) expression in photoreceptor cells. rh1 is expressed specifically in photoreceptor cells R1 to R6. eyeless is expressed in both larval and adult terminally differentiated photoreceptor cells. The homeodomain of Eyeless binds to a palindromic homeodomain binding site P3/RCS1 in the rh1 proximal promoter, which is essential for rh1 expression. These results suggest that Pax-6/Ey directly regulates rh1 gene expression by binding to the conserved P3RCS1 element in the promoter (Sheng, 1997).

eyeless is expressed both early and late in Bolwig's organ cells, which serve as the larval photoreceptor. The ey expression in Bolwig's organ occurs during embryonic development at the end of stage 12. Krüppel expression can also be detected in all 12 Bolwig's organ precursors, but whether ey and Krüppel are coexpressed in all precursors is unknown. ey is down-regulated and absence during most phases of Bolwig's organ development, which includes morphogenetic movement and axonal growth, elongation and projection (Sheng, 1997).

Drosophila contains a second Pax-6 gene, twin of eyeless (toy), due to a duplication during insect evolution. Toy is more similar to vertebrate Pax-6 proteins than Ey, with regard to overall sequence conservation, DNA-binding function, and early expression in the embryo. toy and ey share a similar expression pattern in the developing visual system, and targeted expression of Toy, like Ey, induces the formation of ectopic eyes. Genetic and biochemical evidence indicates, however, that Toy functions upstream of ey by directly regulating the eye-specific enhancer of ey. Toy is therefore required for initiation of ey expression in the embryo and acts through Ey to activate the eye developmental program (Czerny, 1999).

The predicted Toy protein with its 543 amino acids is considerably shorter (by 295 amino acids) than Ey and thus in terms of its size, much more resembles the known Pax-6 proteins of other animals. Ey and Toy share the same sequence identity (90%) in the homeodomain with vertebrate Pax-6 proteins, while Ey is more closely related in the paired domain to vertebrate Pax-6 proteins than Toy (95% vs. 91% sequence identity). In contrast to Ey, Toy shows significant sequence conservation in the C-terminal region with other Pax-6 proteins, Hence, the Drosophila Toy protein has more features (length, conserved sequence motifs) in common with Pax-6 proteins of other animal phyla than with its previously characterized Drosophila Pax-6 paralog, Eyeless. Two Pax-6 genes are present in D. melanogaster and in the closely related species D. virilis and in a more distantly related holometabolous insect, the silkmoth Bombyx mori. However, a single Pax-6 gene could be isolated from squid, sea urchin, mouse, and human, which represents three different animal phyla. The absence of a second Pax-6 gene in the springtail and grasshopper further suggests that the duplication of the Pax-6 gene occurred during insect evolution, in agreement with the observed conservation of intron positions in toy and ey (Czerny, 1999).

During embryogenesis, transcripts of the toy gene are first detected at the cellular blastoderm stage in the posterior procephalic region, including the optic lobe area. During subsequent development, the toy expression domain in the dorsolateral head ectoderm gives rise to the brain and to most, if not all, parts of the visual system, including the optic lobe, the larval eyes (Bolwig's organ), and the eye imaginal discs from which the adult compound eyes and the three ocelli develop. Throughout gastrulation, toy expression is confined to the head region anterior to the cephalic furrow. After germband retraction, toy expression is detected in the optic lobe primordia in a broad region of the brain, and in the anlagen of the eye-imaginal discs. At this stage, toy expression is also observed in a segmentally reiterated pattern in the ventral nerve cord. toy transcripts could not be detected in the differentiating photoreceptor cells of Bolwig's organ, which originates from the posterior procephalic head region. In third instar larvae, toy expression is observed in defined regions of the brain and in the eye-antennal imaginal discs, but not in the leg, wing, or haltere discs. In the eye disc, toy expression is restrict to the undifferentiated part that lies anterior to the morphogenetic furrow (Czerny, 1999).

In contrast to toy, ey is first expressed during late germband extension. ey transcripts are then detected in every segment of the developing ventral nerve cord, whereas toy expression is still absent in this structure. Later in embryogenesis, both genes are expressed in the ventral nervous system, although in different sets of cells. Moreover, ey is initially expressed only in a few cells of the developing brain at germband extension; when compared to toy, its expression remains more regionalized in both brain hemispheres during further development. However, the expression of toy and ey is very similar, if not identical, in the developing visual system, with the exception that during embryogenesis, toy is already expressed in the posterior procephalic region from where the optic primordia originate. Thus, ey and toy are coexpressed in the optic lobe and eye primordia of the late embryo as well as in the undifferentiated part of the eye imaginal discs of third instar larvae (Czerny, 1999)

In the absence of a characterized toy mutation, the epistatic relationship between toy and ey was determined by several different criteria demonstrating that Toy functions directly upstream of ey in the eye developmental pathway and that both proteins fulfill nonredundant functions in compound eye development: (1) toy is normally expressed in the developing eye discs of eyeless mutants. Consequently, toy cannot compensate for the loss of Ey activity in ey mutant flies, indicating that Toy and Ey fulfill nonredundant functions in eye development. (2) Targeted expression of Toy activates ey transcription at the ectopic expression site, whereas misexpression of Ey does not induce toy transcription in heterologous imaginal discs. Hence, toy acts upstream of ey but not vice versa. (3) toy requires ey function for activating the eye developmental pathway, since targeted misexpression of Toy is unable to induce the formation of ectopic eyes in eyeless mutant flies. However, Ey can induce ectopic eye development in the absence of toy, since toy expression is not activated in Ey-induced ectopic eyes. (4) Toy directly regulates ey gene by binding to Pax-6 sites, which are present in the eye-specific enhancer of ey and are essential for enhancer activation at the onset of eye development in the embryo. In conclusion, Toy and Ey have different functions and presumably regulate different sets of target genes during compound eye development. It is proposed that ey is a key target gene of Toy in the developing eye disc. Following activation by Toy, Ey regulates, in turn, downstream genes like sine oculis and eyes absent to further activate the eye development pathway (Czerny, 1999).

A role for eyeless in adult Drosophila brain development and function has been described. eyeless expression is detected in neurons, but not glial cells, of the mushroom bodies, the medullar cortex, the lateral horn, and the pars intercerebralis. Furthermore, severe defects in adult brain structures essential for vision, olfaction, and for the coordination of locomotion are provoked by two newly isolated mutations of eyeless that result in truncated proteins. Consistent with the morphological lesions, defective walking behavior has been observed for these eyeless mutants (Callaerts, 2001).

In the adult, strong immunopositive nuclear staining for Eyeless is observed in the Kenyon cells (mushroom body somata), in the medulla cortex, in neurons in the pars intercerebralis, and in neurons in the lateral horn. In situ hybridization with eyeless probes gives the same expression pattern. No eyeless expression is seen in glial cells. In the larval brain, strong expression is observed in the outer proliferation center of the optic lobes, in the mushroom body neuroblasts, and in Kenyon cell (mushroom body neuron) somata, as well as in other cells in the central brain (Callaerts, 2001).

Several candidate eyeless alleles were identified in a screen for dominant enhancers of an eye loss phenotype induced by ectopic expression of the homeotic Proboscipedia (HoxA2/B2) protein. The ey^JD and ey^DIDa mutations were identified as ey alleles through complementation analysis of previously characterized eyeless alleles. The majority of ey^JD homozygous mutant animals display severe eye defects, with eye sizes reduced to less than 25% the size of wild-type. ey^DIDa homozygotes have slightly weaker eye phenotypes. Homozygosity for ey^JD and ey^DIDa is associated with 85% and 13% lethality, respectively. Hemizygosity for ey^JD and ey^DIDa leads to an increase in lethality to 94% and 29%. Transheterozygous combinations of ey^JD and ey^DIDa with another eyeless allele display a reduced lethality, variable eye phenotypes, and severe mushroom body and central complex defects. Very severe mushroom body defects are also observed in homozygous ey^JD and ey^DIDa flies, and in transheterozygotes of ey^JD and ey^DIDa with Df(4)J2 (Callaerts, 2001).

The molecular nature of the defects underlying the new eyeless alleles is characterized by PCR amplification and sequencing. ey^JD and ey^DIDa have exonic mutations. ey^DIDa is caused by a microdeletion and frameshift in exon 9 encoding the C-terminal domain of the Ey protein. The predicted mutant protein is 682 amino acids long (compared to 838 in wild-type), with the last 29 amino acids of the mutant protein unrelated to the wild-type sequence. ey^JD is due to a point mutation in the homeobox (position 1382 in the cDNA), resulting in a stop codon, predicted to encode a protein truncated after the first helix of the homeodomain with a total length of 432 amino acids. The ey^JD allele leads to the production of a protein of the predicted, reduced size. Surprisingly, no Ey protein could be detected in the ey^DIDa samples. This lack of protein is not due to destabilization of the mRNA. These two new alleles, ey^DIDa and ey^JD, are the first identified in Drosophila that affect the Ey protein product. They may thus reveal previously undetected aspects of normal eyeless function (Callaerts, 2001).

The brains of homozygous ey^DIDa and ey^JD mutant flies are consistently smaller in size when compared to the heterozygous controls, which appeared normal. All homozygous individuals have clearly recognizable defects in the optic lobes, the central complex, and the mushroom bodies (Callaerts, 2001).

The optic lobe consists of four distinct neuropils; lamina, medulla (proximal and distal), lobula, and lobula plate. The lamina and medulla are connected via the first optic chiasm, and the medulla and lobula/lobula plate via the second optic chiasm. Several defects in the eyeless mutant optic lobes were detected. The lamina often appears smaller and flatter in mutants. The medulla is reduced in size and mispositioned, presumably due to incomplete rotation relative to the lamina. The medullar cortex is severely under-developed. The serpentine layer, which separates the distal and proximal part of the medulla, appears disorganized. The remaining two neuropils of the optic lobe, the lobula and lobula plate, are also reduced in size, the internal chiasm appears abnormal, and ectopic fiber bundles that appear to originate directly from the lamina can be observed in the lobula/lobula plate complex. Occasionally, a fusion of the optic lobe with the central brain was observed (Callaerts, 2001).

At the gross-morphological level, each of the two mirror-symmetrical mushroom bodies is a three-armed structure. Each mushroom body consists of about 2,500 Kenyon cells. Their cell bodies lie dorsocaudally in the brain cortex. Their dendrites constitute the calyx, and the axons form the peduncle and the different mushroom body lobes. In a frontal plane just anterior to the ellipsoid body of the central complex, the alpha and alpha' lobes point dorsally and the beta, beta', and gamma lobes point medially toward each other. The third arm of each mushroom body is formed by the peduncle extending obliquely from the dorsally and most posteriorly residing calyx to the comparatively anterior and ventral branching point of the lobes. The calyces are embedded in the dorsal cortex layer of the brain. The mushroom body lobes are mostly not recognizable in flies homozygous for either ey allele, though immunohistochemistry with anti-Fas II antibody reveals residual traces of the lobes. Peduncles are discernible in most brains, but are greatly reduced in diameter. The calyces of the mutant mushroom bodies are clearly diminished in size. Both ey alleles have similar qualitative effects on the mushroom bodies (Callaerts, 2001).

The central complex resides between the brain hemispheres just dorsal to the esophagus. It comprises four strongly connected neuropil regions. The anterior-most ellipsoid body is of torus shape, and it resides in the anterior concavity of the fan-shaped body. The paired noduli are located ventral to the ellipsoid body and dorsal to the esophagus. Finally, the protocerebral bridge is found posterior to the fan-shaped body at the border between cortex and protocerebral neuropil, and is flanked laterally by the calyces of the mushroom bodies. The central complex region was strongly disordered in all homozygous ey mutant flies examined, with the defects more pronounced in ey^JD flies. For both alleles, the ellipsoid body appears to be fused with the fan-shaped body. Both neuropil regions are not well separated from the surrounding protocerebral neuropil in ey^DIDa brains, and are almost fused with it in ey^JD brains. The protocerebral bridge is disintegrated into several chunks of neuropil. Despite this marked disorganization, the total volume of the central complex was nearly unchanged in the ey mutants. In addition to the described phenotypes, strongly perturbed Fas II-positive neuronal projections are observed in ellipsoid and fan-shaped bodies of all mutant brains analyzed (Callaerts, 2001).

The central complex has been identified as a higher center for the control of walking behavior. Walking behavior of the homozygous ey alleles was studied as one measure of brain function at two different levels of resolution: (1) In the object fixation task 'Buridan's paradigm', the test fly is seen as a point-like object while walking between two inaccessible landmarks, and information is gathered about its walking activity, speed, and orientation behavior; (2) on a stepping analyzer, the actions of the single legs are resolved and their coordination studied. Regardless of the experimental situation, homozygous flies of both alleles are extremely reluctant to walk. This was particularly surprising in Buridan's paradigm, a situation that prompts normal flies to walk spontaneously, sometimes for hours. For ey mutant flies, spontaneous walking activity is already very low at the beginning, then fades further during the 15 min tests, as is typical of flies with a defective central complex. The homozygous test flies show almost no variation in their mean step length, which stays about as low as during slow walking (i.e., long stepping periods). Their range of stepping frequencies is limited at the fast end to 13 steps per second, whereas control heterozygotes, like wild-type flies, carried out up to 16 steps/s with every leg. The coordination of swing phases is normal and resembles the usual alternating tripod gait. A plot of the relative differences in step length and swing phase duration between mutants and controls reveals a 15% smaller step length in ey mutants, a reduction which is not explained by a mean shortening of swing phases. The smaller step size therefore is explained by a lower swing speed of legs during swing phases of almost normal duration. This specific walking defect has been previously found in other central complex structural mutants with a disrupted protocerebral bridge (Callaerts, 2001).

GENE STRUCTURE

The transcription unit spans approximately 16 kb. The gene encodes two transcripts which differ with respect to their first exons. These are spliced to exon three, which is shared. The N-terminal paired domain is coded for by three exons, and is present in both proteins (Quiring, 1994).

cDNA clone length - 2832

Bases in 5' UTR - 98+

Exons - Eight exons encode each splice variant.

Bases in 3' UTR - 229

PROTEIN STRUCTURE

Amino Acids - Two coding variants are noted: cDNA E10 codes for a protein of 838 amino acids, and cDNA D1 codes for a protein of 857 amino acids.

Structural Domains

eyeless has an N-terminal paired domain and a central homeodomain. eyeless is homologous to the mouse Small eye (Pax-6) gene, and to the Aniridia gene in humans. These genes share extensive sequence identity, with 94% sequence identity in the paired domain and 90% identity in the homeodomain. The vertebrate genes are expressed similarly in the developing nervous system and in the eye during morphogenesis (Quering, 1994). In the human paired box, there is a 14 amino acid alternatively spliced extra exon, not present in the fly. One splice site in the homeodomain is shared with humans, but a second, present in the fly, is absent in humans (Quiring, 1994).

Pax6, a transcription factor containing the bipartite paired DNA-binding domain, has critical roles in development of the eye, nose, pancreas, and central nervous system. The 2.5 A structure of the human Pax6 paired domain with its optimal 26-bp site reveals extensive DNA contacts from the amino-terminal subdomain, the linker region, and the carboxy-terminal subdomain. The Pax6 structure not only confirms the docking arrangement of the amino-terminal subdomain as seen in cocrystals of the Drosophila Paired Pax protein, but also reveals some interesting differences in this region and helps explain the sequence specificity of paired domain-DNA recognition. In addition, this structure gives the first detailed information about how the paired linker region and carboxy-terminal subdomain contact DNA. The extended linker makes minor groove contacts over an 8-bp region, and the carboxy-terminal helix-turn-helix unit makes base contacts in the major groove. The structure and docking arrangement of the carboxy-terminal subdomain of Pax6 is remarkably similar to that of the amino-terminal subdomain, and there is an approximate twofold symmetry axis relating the polypeptide backbones of these two helix-turn-helix units. The structure of the Pax6 paired domain-DNA complex provides a framework for understanding paired domain-DNA interactions, for analyzing mutations that map in the linker and carboxy-terminal regions of the paired domain, and for modeling protein-protein interactions of the Pax family proteins (Xu, 1999).

date revised: 5 March 2001 

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.