atonal

DEVELOPMENTAL BIOLOGY

Embryonic

atonal is expressed in the proneural clusters and sense organ precursors that give rise to the embryonic and adult chordotonal organs (Jarman, 1993). Expression begins around stage 8 [Images] in clusters of cells from which a single precursor of the lateral chordotonal organ will arise.

The chordotonal (Ch) organ, an internal stretch receptor located in the subepidermal layer, is one of the major sensory organs in the peripheral nervous system of Drosophila. Clues as to Rhomboid's function are provided in an analysis of the role of Rhomboid in the determination of Ch organ precursor cells (COPs). The rhomboid gene and the activity of the Drosophila Epidermal growth factor receptor (Egf-R) signaling pathway are necessary to specifically induce three of the eight COPs in an embryonic abdominal hemisegment. The cell-lineage analysis of COPs indicates that each of the eight COPs originate from an individual undifferentiated ectodermal cell. The eight COPs in each abdominal hemisegment seem to be determined by a two-phase induction: first, five COPs are determined by the action of the proneural gene atonal and neurogenic genes. Subsequently, these five COPs start to express the rho gene, and rho activates the Efg-R-signaling pathway in neighboring cells and induces argos expression. Three of these argos-expressing cells differentiate into the three remaining COPs and they prevent neighboring cells from becoming extra COPs. In the five atonal dependent COPs, Egf-R signaling activity is required, but this signaling does not seem to involve the cell autonomous activity of Rho. In rho null mutants five chordotonal organs remain intact. However, rho expression is required to activate Egf-R in adjacent cells, and these three adjacent cells express the neuronal marker asense. Argos functions from the second wave of cells as a lateral inhibitor, restricting the number of recruited cells to the original three. As the rho-expressing first wave of COPs is adjacent to the three argos and asense expressing double postive COPs, Argos may function to prevent the continuance of Egf-R-signal activation in additional neighboring cells. A model is favored in which Rho protein is required for the activation of an Egf-R ligand, the Spitz transmembrane protein, by processing it into the functional soluble form. An alternative model, invalid at least in Ch organ determination but still valid for follicle cell determination in oogenesis, suggests that Rho protein is expressed in cells that require the activation of the Egf-R signaling pathway, and that Rho protein interacts with Egf-R protein directly or indirectly to amplify Egf-R signaling (Okabe, 1997).

The transcription factors encoding genes tailless (tll), atonal (ato), sine oculis (so), eyeless (ey) and eyes absent (eya), and Efgr signaling play a role in establishing the Drosophila embryonic visual system. The embryonic visual system consists of the optic lobe primordium, which, during later larval life, develops into the prominent optic lobe neuropiles, and the larval photoreceptor (Bolwig's organ). Both structures derive from a neurectodermal placode in the embryonic head. Expression of tll is normally confined to the optic lobe primordium, whereas ato appears in a subset of Bolwig's organ cells that are called Bolwig's organ founders. Phenotypic analysis of tll loss- and gain-of-function mutant embryos using specific markers for Bolwig's organ and the optic lobe, reveals that tll functions to drive cells to an optic lobe fate, as opposed to a Bolwig's organ fate. Similar experiments indicate that ato has the opposite effect, namely driving cells to a Bolwig's organ fate. Since tll and ato do not regulate one another, a model is proposed wherein tll expression restricts the ability of cells to respond to signaling arising from ato-expressing Bolwig's organ pioneers. The data further suggest that the Bolwig's organ founder cells produce Spitz (the Drosophila TGFalpha homolog) signal, which is passed to the neighboring secondary Bolwig's organ cells where it activates the Epidermal growth factor receptor signaling cascade and maintains the fate of these secondary cells. The regulators of tll expression in the embryonic visual system remain elusive, no evidence for regulation by the 'early eye genes' so, eya and ey, or by Egfr signaling is found (Daniel, 1999).

The Drosophila visual system comprises the adult compound eye, the larval eye (Bolwig's organ) and the optic lobe (a part of the brain). All of these components are recognizable as separate primordia during late stages of embryonic development. These components originate from a small, contiguous region in the dorsal head ectoderm. During the extended germband stage, the individual components of the visual system can be distinguished morphologically as well as by spatially localized expression of the homeobox gene so and the adhesion molecule Fas II. Initially centered as an unpaired, oval domain straddling the dorsal midline, the anlage of the visual system subsequently elongates in the transverse axis and narrows in the anteroposterior axis. By late gastrulation (stage 8), the anlage occupies two bilaterally symmetric stripes that are anterior and adjacent to the cephalic furrow. The domain of so expression at this stage contains two regions with a high expression level [ol_ex (the external fold of the optic lobe) and ol_in]. Only these two regions will ultimately give rise to the optic lobe and Bolwig's organ; the so-positive cells dorsal and posterior to these domains will either form part of the dorsal posterior head epidermis (dph) or undergo apoptotic cell death. During the extended germband stage, the anlage of the visual system expands further ventrally until, around stage 10, it reaches the equator (50% in the dorsoventral axis) of the embryo. Shortly thereafter, ol_in, the portion of the anlage that will give rise to most of the optic lobe and Bolwig's organ, reorganizes into a placode of high cylindrical epithelial cells that differ in shape from the surrounding more squamous cells of the head ectoderm. During stage 12, this placode starts to invaginate, forming a V-shaped structure with an anterior lip (ol_al) and a posterior lip (ol_pl). Bolwig's organ, which consists of a small cluster of sensory neurons, derives from the basal part of the posterior lip and can be recognized during stage 12 as a distinct, dome-shaped protrusion. During stage 13, invagination of the optic lobe separates it from the head ectoderm; only the cells of Bolwig's organ remain in the ectoderm. The ectodermal region ol_ex is also internalized and forms an external 'cover' of the optic lobe; many cells of this population undergo apoptosis (Daniel, 1999).

atonal is expressed in and required for the development of Bolwig's organ. ato is expressed in the head in several small cell clusters, one of which is a group of three to four cells that is part of the Bolwig's organ primordium. Expression of ato in this domain begins during stage 11 and continues until stage 12. Initially, a group of 6-8 cells faintly expresses ato. By stage 12, their number has decreased to 3 cells. During this period, ato-expressing cells can be seen as a small group of cells within the dome-shaped Bolwig's organ primordium. Loss of ato function results in the absence of Bolwig's organ. Thus, similar to what has been demonstrated for the compound eye, even though only a small subset of photoreceptors actually expresses ato, lack of ato function results in absence of all photoreceptors. Since Bolwig's organ is enlarged in a tll mutant background, it was asked whether tll inhibits ato expression. The number and pattern of ato-positive cells in tll mutants is found to be normal. These results suggest that tll functions in parallel with, or downstream of ato in the development of the Bolwig's organ/optic lobe primordium (Daniel, 1999).

Misexpression has been used to examine the abilities of various proneuronal genes to determine SOP cell fate in the embryonic PNS. To determine whether different types of proneural genes confer SOP cells with distinct neuronal specificities, a test was made of the abilities of three proneural genes (sc, ato, and amos) to induce the formation of ES, CH, and MD neurons. For MD neuron formation, the lacZ expression was tested in E7-2-36 embryos carrying proneural genes misexpressed by sca-GAL4. Misexpression of absent MD neurons and olfactory sensilla (amos) leads to a significant increase of MD neurons through the whole segment. In the ventral region, the number of MD neurons (vmd5 + vpda) increases significantly. Misexpression of sc or ato slightly induces the formation of MD neurons, but not as significantly as misexpression of amos. To compare the induction abilities of these three proneural genes in CH neuron formation, embryos were stained with MAb22C10 (see Futsch), and the lateral CH neurons were scored. In embryos with ato or amos misexpressed by sca-GAL4, a slight increase in the number of CH neurons in the lateral region was observed. Few ectopic CH neurons were induced in embryos that misexpressed sc. A test was performed to see if amos could rescue ato mutants in CH organ formation. In ato¹ mutants, misexpression of amos driven by hairy-GAL4 restores CH neuron formation in odd, but not even, segments. The extent of CH neurons rescued by amos is comparable to that of neurons rescued by ato, suggesting that amos can replace ato functionally (Huang, 2000).

Atonal is expressed in a subset of brain cells and regulates neurite arborization

Drosophila atonal (ato) is the proneural gene of the chordotonal organs (CHOs) in the peripheral nervous system (PNS) and the larval and adult photoreceptor organs. ato is expressed at multiple stages during the development of a lineage of central brain neurons that innervate the optic lobes and are required for eclosion. A novel fate mapping approach shows that ato is expressed in the embryonic precursors of these neurons and that its expression is reactivated in third instar larvae (L3). In contrast to its function in the PNS, ato does not act as a proneural gene in the embryonic brain. Instead, ato performs a novel function, regulating arborization during larval and pupal development by interacting with Notch (Hassan, 2000).

To examine the profile of ato expression in the Drosophila central brain, in situ hybridization and immunohistochemistry experiments were performed on embryos, third instar larval (L3), and adult brains. In stage 13 embryos, two small clusters of cells express the ATO RNA. These cells are located in the dorso–lateral region of the central brain, proximal to the developing optic lobes. This expression is very transient, and no ATO RNA is detectable in these cells after stage 13 or in similar positions in first (L1) or second (L2) instar larvae. Using a new antibody against the ATO protein it has been found that Ato protein has an expression pattern identical to that of ATO RNA. The ATO-expressing clusters are composed of 3–5 cells in each brain hemisphere. The sizes of the cells suggests that at least one is a neuroblast, whereas the others are likely to be ganglion mother cells. During early larval development, the ATO RNA is detectable in the inner proliferation zone of the optic lobes, but no expression is detected in the central brain in L1 and L2. In L3 ATO, RNA is once again detectable in the central brain in two clusters comprised of 20–30 cells in the dorso–lateral region proximal to the optic lobe. In addition, the inner proliferation zone of the optic lobe continues to express ATO RNA. Immunohistochemistry on frozen adult head sections shows that Ato is expressed symmetrically in one dorsal cluster (DC) of cells adjacent to the lobula in each brain hemisphere, as well as in two small groups of ventral cells. In summary, Ato is transiently expressed in stage 13 embryos in the brain. Expression is reinitiated in a cluster of neurons in the L3 brain and observed in three clusters of neurons of the adult brain (Hassan, 2000).

A better understanding of the origin of the Ato-expressing neurons, and of the function of ato in the brain, has resulted from the identification of an enhancer fragment that recapitulates ato expression. A 5.6 kb regulatory fragment directly upstream of the ato open reading frame (ORF) is sufficient for ato expression in the embryonic PNS as well as in the L3 leg, antennal, and wing discs. This fragment also directs lacZ expression to the larval optic lobes and central brain. However, a 3.6 kb fragment directly upstream of the ato ORF directs expression exclusively to the central brain clusters in L3. No expression is observed with the 3.6 kb enhancer in imaginal discs or other areas of the CNS, but it directs lacZ expression in the embryonic CHO precursors (Hassan, 2000).

To further characterize the cells that express ato, the 3.6 kb enhancer was cloned upstream of Gal4, and two independent transgenic lines were generated (ato-Gal4 14a and ato-Gal4 10). The expression of both lines was examined throughout fly development to determine whether they mimic ato expression. ato-Gal4 is initially expressed in the sensory organ precursors of the embryonic chordotonal organs (CHOs). At stage 13, 3–5 cells in each brain hemisphere initiate ato-Gal4 expression. While lacZ expression lasts longer than does ATO RNA, it is nonetheless transient and is no longer detectable in stage 15 embryos. However, lacZ expression is reinitiated at L3 in a cluster of 25–30 neurons in each brain lobe. Each cluster of neurons sends a bundle of axons ventrally and then contralaterally toward the opposite optic lobes. At the junction between the central brain and the developing lobula, the axons derived from the contralateral cluster branch out of the commissure, into the developing optic lobes. These data indicate the 3.6 kb fragment recapitulates the expression of ato in embryos and larvae (Hassan, 2000).

Since Ato is expressed in adult brains, ato-Gal4/UAS-lacZ adult brains were stained with anti-beta-galactosidase (anti-beta-gal) and the data were analyzed with confocal microscopy. In adult flies, ato-Gal4 is expressed in a DC and in two ventral clusters (VCs: VLC and VBC) of neurons. Some axons of the DC project ipsilaterally over the lobula. However, most axons of the DC form a bundle that is a component of the dorsal commissure and project contralaterally toward the lobula complex and the medulla. These neurites fan out over the lobula complex and the inner chiasm. Ten to twelve tracts cross the outer chiasm, toward the medulla, in a ladder-like pattern. Over the medulla, the fibers branch and appear to contact one another to form a 'grid-like' lattice. No fibers cross the lamina (Hassan, 2000).

The ventral brain cluster (VBC) appears to be located in the central brain, while the ventral lobular cluster (VLC) is located in the lobula. The axons of the VBC project along the brain–lobula border toward the DC, while the VLC forms an extensive network of fibers in the ventral lobula. To determine when the expression of ato is initiated in the VCs, pupal brains were examined at 50% and 75% pupal development. All clusters express ato-Gal4 at both stages, suggesting that ato expression in the DC never ceases after L3 and that VLC and VBC initiate ato expression de novo during pupal development. To confirm that the arborization patterns observed with cytoplasmic lacZ reflect axonal projections, the axonal marker UAS-tau.lacZ driven by ato-Gal4 was used. tau.lacZ expression mimics that of cytoplasmic lacZ, revealing the arborizations in the lobula and medulla. Thus, it appears that the arborizations of the DC and VC neurons in the optic lobes are composed of axons, implying that the ato-expressing neurons output onto the optic lobes rather than receive input from them. The dendritic projections of these clusters were examined using UAS-Nod.lacZ, which has been shown to be localized to dendrites. The dendrites of the clusters are very short and connect to each other, forming dendrodendritic connections. These data indicate that the ato-expressing neurons are multipolar, unlike most neurons of the insect brain (Hassan, 2000).

The above data raise several questions. (1) Are the clusters of the adult brain derived from ato-expressing embryonic precursors? (2) What is the role of ato in embryonic precursors? (3) Is the expression of ato required in L3 and adults? (4) What is the function of the ato-expressing clusters of neurons? To answer these questions, advantage was taken of the viability of ato mutants, allowing examination of the presence and morphology of these neurons at various stages. The ato brain enhancer was used to (1) mark these neurons and (2) express a variety of gene products. To address the relationship between the ato-expressing precursor cells in the embryonic brain and the ato-expressing neurons in the L3 and adult brain, the progeny of the embryonic precursors expressing ato were fate mapped using a novel approach based on the UAS-Gal4 system, which introduces an amplification step between the ato-Gal4 and the UAS-lacZ steps via a UAS-Gal4 construct. The rationale of the methodology is as follows: the ato-Gal4 construct is used to activate the expression of the Gal4 protein in a precursor-specific manner. Gal4 binds to the UAS sites of the UAS-lacZ and UAS-Gal4 constructs. Extra Gal4 protein from the UAS-Gal4 construct binds to the UAS sites upstream of lacZ and autoactivates Gal4 production. Hence, the signal becomes self-propagating and is substantially amplified. Upon cell division, daughter cells inherit sufficient Gal4 protein to reactivate the cascade (Hassan, 2000).

This strategy was applied to the ato brain enhancer. Normally, expression of ato-Gal4 is no longer detectable by stage 15 and is absent throughout early larval development (L1 and L2). In contrast, when the UAS-Gal4 construct is introduced, the lineage can be easily followed with anti-beta-gal antibodies. At stage 15, the two brain clusters consist of about 8 cells each. In addition, specific PNS organs are strongly positive for beta-gal. These organs all correspond to the embryonic CHOs, in agreement with the expression of ato-Gal4 in the CHO precursor cells. In L1 brains, there are ~10 cells in each cluster, which consist of two small subclusters. In late L2, each cluster comprises about 16 cells, and a thin commissure can already be seen. In L3, a single cluster of 25–30 cells can be observed in each brain lobe, the same number observed without the UAS-Gal4 construct. Had the embryonic precursors given rise to a different set of neurons, two or more clusters of neurons would have been identified in L3. No other clusters have been observed in L1–L3 brains. It is therefore concluded that the DC is derived from the ato-expressing embryonic brain cells. Studies using other promoter/enhancer Gal4 constructs indicate that this fate mapping method may be generally applicable (Hassan, 2000).

What might the function(s) of ato be in the central brain? To address this question, the consequences of the loss and gain of ato function in the DC in L3 and adult brains was examined using the ato-Gal4 line as a marker and a driver. Surprisingly, all three clusters (DC, VBC, VLC) are present in brains homozygous for the ato¹ allele as well as in brains transheterozygous for ato¹ and a deficiency that uncovers the ato region. However, several defects are obvious in ato mutant brains. While heterozygous control L3 brains show a normal DC, mutant L3 brains show severe defects in DC position and organization. In addition, probably as a result of the loose morphology of the cluster, the descending axon bundles are defasiculated. These defects were observed with about 15% penetrance and suggest a weak or partially redundant differentiation requirement for ato in the precursors of the DC lineage. Note, however, that the DC axons form a commissural tract, strongly suggesting that their basic identity as commissural neurons is not affected by the ato mutation (Hassan, 2000).

To investigate if ato is required in the postmitotic DC neurons, as the reinitiation of its expression suggests, the morphology of the axonal projections of the DC neurons of adult ato mutant brains was examined. The DC forms a stereotypical axonal pattern, making it simple to detect aberrations that may be caused by the ato mutation. In adult brains, in addition to the aberrant positioning of the cluster seen in L3, the arborization pattern of the DC over the lobula is severely impaired in ato deficiency flies. Most axons enter the lobula either ventrally or dorsally and show very limited branching, failing to form a proper 'fan-shaped' pattern. Since ato mutant brains have a severely reduced medulla, the medullar part of the pattern could not be examined. While these data suggest a role for ato in axon arborization, alternative interpretations are possible: (1) the observed axonal defects may be caused by the significant loss of optic lobe structures associated with loss of ato function (no lamina or medulla, reduced lobula); (2) it is possible that the defects are a reflection of the improper differentiation of the DC cells rather than a reflection of a specific role for ato in arborization (Hassan, 2000).

To rule out the first possibility, the lobular innervation pattern of the DC was examined in sine oculis (so) mutant brains. so mutants lack eyes and a lamina, and exhibit a severely reduced medulla and a reduced lobula. In gross morphology and size, so brains and ato brains are indistinguishable. Since so mutants show variability in phenotypic penetrance, only flies that completely lack eyes were used. The axonal projections of the DC lobular pattern appears essentially wild-type in so mutant brains, indicating that the axonal defects in the ato mutant brains are due to the ato mutation. To demonstrate that this is indeed the case, ato was expressed using the ato-Gal4 driver in ato¹/Df(3R)p13 flies. This resulted in a rescue of the lobular innervation pattern. This indicates that the role of ato in DC axon arborization is specific and cell autonomous (Hassan, 2000).

Finally, if ato does play a role in the development of the axonal pattern of the DC neurons, its overexpression in these neurons may result in an aberrant axonal pattern without affecting the number or position of the DC neurons. The ato-Gal4 line was used to overexpress ato in the DC and the cells were marked simultaneously using UAS-ato and UAS-lacZ constructs. Consistent with a role for ato in neurite arborization, excessive branching is found in the lobula and the medulla. Importantly, the organization of the cluster, its size, and its position are unaffected in L3 brains and adult brains. In addition, the number of axons crossing the optic chiasma, though not their morphology, is wild-type. These data strongly suggest that the early differentiation of the cluster is unaffected when ato is overexpressed. It is concluded that the axonal defects observed in ato loss- and gain-of-function experiments reflect a specific requirement for ato in the control of the arborization pattern of the DC neurons (Hassan, 2000).

To understand the mechanism by which ato functions in axonal development, the role of Notch was examined in the development of the DC axonal pattern. Two observations make Notch a logical candidate. (1) ato and Notch interact in an antagonistic fashion during CHO development. For example, in the leg femoral CHO, gain of Notch function reduces the number of precursors selected from the ato-expressing proneural cluster. (2) Notch has been shown to be required for axon guidance, perhaps mediating axon–substrate interactions, in both the PNS and CNS. If ato and Notch act antagonistically in arborization, it is expected that Notch activity levels would be relatively low within the DC, where its function is repressed by ato, and relatively high in the substrate cells, where ato does not antagonize it. Thus, after ato expression is reinitiated in the DC neurons, a differential in Notch activity levels may occur between the arborizing axon and the substrate cells. Perturbations of this imbalance, either by raising Notch activity levels in the DC or by reducing them in the surrounding cells, may result in defective arborization patterns. This model allows for three specific predictions: (1) the generalized loss of Notch function may result in excessive arborization of the DC neurons, whereas the DC-specific loss of function would cause no significant defects; (2) the gain of Notch in the DC neurons is expected to inhibit arborization; (3) if the activation of ato in the DC neurons serves to antagonize Notch activity, then it is expected that the gain of Notch function will be epistatic to the gain of ato function (Hassan, 2000).

To determine the requirements for Notch in DC axon development, two Notch alleles were examined: a temperature-sensitive allele (N^ts) and a viable hypomorphic allele (facet notchoid [N^nd3]). In the first set of experiments, N^ts;ato-Gal4,UAS-lacZ larvae were raised in a cycling incubator delivering a 30 min, 34°C heat shock every 8 hr from late L1 through wandering L3. L3 brains were examined for DC defects. Reducing Notch activity during larval development has no effects on the number, morphology, or position of the DC neurons or on the formation of the commissure. In contrast, defects in axon branching out of the commissure into the optic lobe were observed. Specifically, excessive branching and defasiculation of the axon bundles entering the optic lobe were observed. Importantly, these defects were not rescued by a wild-type copy of Notch driven by the ato enhancer in the DC (N^ts; UAS-N⁺;ato-Gal4,UAS-lacZ), suggesting that the requirement for Notch in DC axon arborization is nonautonomous in contrast to the requirement for ato. Larvae reared under the cycling heat shock paradigm or at a consistent 28°C temperature between either L1 and adult or L3 and adult do not produce homozygous N^ts flies. Therefore, to examine the adult DC innervation pattern in a background of reduced Notch activity, N^nd3;ato-Gal4,UAS-lacZ male flies were used. These flies show excessive branching in the medulla, resulting in an aberrant innervation pattern similar to that caused by ato overexpression in the DC neurons. The disruption is milder due to the fact that N^nd3 is a weak allele of Notch (Hassan, 2000).

Do the defects observed in Notch mutants reflect an independent function for Notch in arborization? It is possible that the conditions used in this study result in mild neurogenic phenotypes generating extra DC target cells. This, in turn, would cause the DC neurons to arborize excessively to innervate new targets. To rule out this possibility, area density (number of cells per unit area) and optic lobe cortex volume (volume occupied by optic lobe cell bodies) analyses were performed on N^ts L3 brains and N^nd3 adult brains. By both criteria, there are no significant differences between wild-type and mutant brains. Therefore, these data support a model in which ato generates the branching pattern by antagonizing Notch activity in the DC (Hassan, 2000).

To show that Notch function is not required within the DC neurons themselves, a dominant-negative form of Notch (UAS-N^EC) was expressed using ato-Gal4. N^EC has no effect on the formation of the larval or adult axonal patterns. In contrast, N^EC expression in imaginal discs results in strong loss of Notch function phenotypes and pupal lethality, demonstrating that the construct is active. This shows that while Notch function is required specifically for arborization of the DC neurons, its requirement is nonautonomous. The prediction that ato represses Notch activity in the DC cells implies that gain of Notch function within the DC would result in inhibition of axonal branching, a phenotype similar to that of loss of ato function. The membrane-bound, wild-type form of Notch (N⁺) is known to be required for the rescue of axonal defects associated with the loss of Notch function. However, in all cases in which ato and Notch appear to interact in the PNS, it is the nuclear form of Notch that is thought to be involved. To evaluate the effects of the gain of Notch function, both forms of Notch were overexpressed in the DC neurons: the membrane-bound N⁺ and the nuclear form, N^intra. Overexpression of N⁺ has no effect on the axonal pattern, whereas overexpression of N^intra results, in comparison with controls, in a severe inhibition of axonal branching over the lobula and a complete failure of innervation of the medulla. These data suggest that the nuclear form of Notch, but not the membrane-bound form, affects the arborization pattern of the DC axons. Finally, if ato suppresses Notch signaling within the DC, gain of Notch function should be epistatic to the gain of ato function, placing Notch genetically downstream of ato. Therefore, the combined overexpression of ato and N^intra should result in the same phenotype as the overexpression of N^intra alone. Brains in which both ato and N^intra are overexpressed using ato-Gal4 have a phenotype identical to that of N^intra overexpression alone. Taken together, the data presented above support the hypothesis that ato acts to suppress Notch signaling within the DC and that this suppression is essential for the generation of the proper arborization pattern of the DC axons (Hassan, 2000).

What is the function of the ato-expressing neurons? To address this question, the cell death gene reaper and the cytotoxin Ricin were overexpressed in these cells using the ato-Gal4 lines 14a and 10. This leads to the loss of all ato-expressing cells in the brain. At 22°C, only 18% of the flies that express the Ricin in the ato neurons eclose, when compared with their control siblings, although almost all flies develop to the pharate adult stage. At 28°C, none of the flies carrying both constructs hatched. The same observations were made with reaper expression using ato-Gal4. The Ricin and reaper-expressing escaper flies show a 2–3 day delay in eclosion, as compared with control siblings. After hatching, however, they were viable and fertile, and displayed no obvious behavioral defects. In addition, it was noted that the overexpression of N^intra in the ato neurons also results in very few adult escapers (10% of expected progeny eclosed, and most die as pharate adults). When pharate adults who fail to hatch after a 5 day delay in the reaper, Ricin, and N^intra experiments are dissected out of their pupal cases while still alive, they fail to move and usually die within 24 hr. These data indicate that the ato-expressing cells and their proper arborization are important for proper eclosion. However, it is impossible at this point to distinguish between the requirements for the DC, the VBC, the VLC, or any combination thereof (Hassan, 2000).

Larval

atonal is expressed just anterior to the morphogenetic furrow in the developing eye antennal disc. As the furrow progresses, expression becomes confined to R8 (Jarman, 1995). Induction of R8 requires atonal, but induction of R1-7 does not (Jarman, 1994). In the leg disc a large patch of atonal forms in the presumptive femur region corresponding to the large femoral chordotonal organ. There is also expression in the wing disc (Jarman, 1993).

Pupal phase

The mechanisms underlying the development of the olfactory sense organs on the third segment of the antenna of Drosophila suggest the employment of a novel developmental strategy. Specification of the founder or precursor cell is not governed by the genes of the achaete-scute complex. Instead another basic helix-loop-helix encoding gene, atonal, is essential for determination of only a subset of the sensilla types ‚ the sensilla coeloconica. Therefore, the existence of additional proneural genes for the selection of sensilla trichoidea and sensilla basiconica is predicted. The choice of a founder cell from the presumed proneural domain is regulated by Notch activity. Soon after delamination of the founder cell, two to three additional neighboring cells also take on a sensory fate and these cells together form a presensillum cluster. The selection of neighbors does not occur when endocytosis is blocked using a temperature sensitive allele of shibire, suggesting that cell-cell communication is required for this step. The cells of the cluster divide once before the occurence terminal differentiation that is influenced by Notch activity. The final cell number within each sensillum is controlled by programmed cell death (Reddy, 1997). Another gene, lozenge, has been implicated in the process of determination of founder cells of basiconic sensilla (Ray, 1995).

It is increasingly apparent that proneural bHLH factors not only confer neural competence for SOP formation but also endow SOPs with neural subtype information. While expression and loss-of-function analyses show that amos, ato, and the ASC are required for different subtypes of sense organs, a direct test of the functional specificity of proneural proteins comes from comparison of their capabilities in misexpression assays. Misexpression of all proneural genes results in ectopic SOP formation, but the subtype of the sense organ depends on both the gene misexpressed and the developmental context of misexpression. In the context of antennal development, misexpression of amos or ato, but not an ASC gene, drives ectopic olfactory sensillum formation. This suggests that the shared features of the amos and ato bHLH domains provide functional information important for olfactory sensillum determination. One of the specific functions ascribed to ato in chordotonal SOP formation is the inhibition of cut, a subtype selector gene for external sense organs. Olfactory sensilla also do not require cut, and so inhibition of cut expression might be an important conserved function of ato and amos in olfactory SOP formation. Another shared function may be the activation of cell recruitment, since amos- and ato-dependent sense organs (olfactory sensilla, chordotonal organs, and ommatidia) all require some form of local cell recruitment in their development. The recruitment of cells to an ommatidium and the recruitment of SOPs into chordotonal clusters both require Egfr signaling, perhaps under the direct control of ato. It is not yet known whether Egfr is involved in the process of cell recruitment by olfactory SOPs, but it is clearly possible that ato and amos directly control the recruitment process. Outside the antenna, misexpressed ato drives ectopic chordotonal organ formation. This functional specificity has been found to reside in Ato's bHLH domain and largely in its basic region. Ato's basic region is completely conserved in Amos, apart from an R to K conservative substitution. Not surprisingly, therefore, amos can also mimic ato in directing chordotonal organ formation when misexpressed outside the antenna. Nevertheless, despite amos's ability to mimic ato and their common requirement in olfactory sensillum formation, these genes are not functionally interchangeable. Differences in their ability to determine olfactory sensillum subtypes are initially suggested by their differing loss-of-function phenotypes, despite their partly overlapping expression domains. Also, Gupta (1997) reported that ato misexpression in the antenna specifically results in increased sensillum coeloconica numbers, while amos increases all sensilla types in the experiments described here and can rescue the loss of sensilla basiconica in lz mutants. At the least, this shows that amos can supply some functional information that ato apparently cannot (i.e., basiconic/trichoid fate). Most significantly, only amos misexpression can drive ectopic sensilla of olfactory-like morphology outside the antenna. Thus, unlike ato, amos function in olfactory sensillum formation appears to be partially able to overcome the constraints of developmental context. Presumably, one or more target genes must be differentially regulated by amos and ato (Goulding, 2000a and references therein).

Differential developmental functions of amos and atonal

amos is a new candidate Drosophila proneural gene related to atonal. Having isolated the first specific amos loss-of-function mutations, it has been shown definitively that amos is required to specify the precursors of two classes of olfactory sensilla. Unlike other known proneural mutations, a novel characteristic of amos loss of function is the appearance of ectopic sensory bristles in addition to loss of olfactory sensilla, owing to the inappropriate function of scute. This supports a model of inhibitory interactions between proneural genes, whereby ato-like genes (amos and ato) must suppress sensory bristle fate as well as promote alternative sense organ subtypes (zur Lage, 2003).

Three mutant alleles of amos were generated in an EMS screen. amos¹ is predicted to result in a protein truncation that removes the second half of bHLH helix 2 and the C-terminal region thereafter. amos² is a missense mutation that changes a Ser to an Asn in helix 1 of the bHLH domain. This position is not part of the bHLH core consensus sequence and is not predicted to directly affect DNA binding or dimerization. Moreover, Asn is found in this position in the ato bHLH domain, and so the effect of this mutation would be predicted to be mild. amos³ contains a 230 bp deletion within the ORF, which also causes a frame-shift that brings a spurious downstream stop codon in frame. This allele gives a predicted peptide of 74 amino acids, of which only the first 30 are shared with amos. It therefore lacks the entire bHLH domain and is likely to be a null (zur Lage, 2003).

amos¹ mutant embryos lack two dorsal sensory neurons per segment, including the dorsal bipolar dendritic neuron. Nevertheless, all amos alleles are adult viable as homozygotes and hemizygotes. The antennae of mutant adult flies were mounted and examined by light microscopy in order to quantify the number and type of olfactory sensilla. Compared with wild-type, amos mutant antennae carried dramatically reduced numbers of sensilla and, as a consequence, the third segment is significantly smaller. In particular, sensilla basiconica and trichodea were completely absent in the probable genetic nulls, amos¹ and amos³, whereas sensilla coeloconica appear unaffected. These phenotypes support the assertion that amos is the proneural gene for sensilla basiconica and trichodea, whereas ato is the proneural gene for sensilla coeloconica. However, mutant antennae exhibit a further unexpected phenotype. In wild type, the third segment bears only olfactory sensilla; in amos mutant flies, this segment bears a number of ectopic external sensory bristles and other abnormally structured sensilla. These bristles do not have bracts (unlike bristles on the leg), and so this phenotype does not represent a transformation of antenna to leg. These phenotypes are highly unusual; a characteristic of all other loss-of-function proneural gene mutations is that they cause the loss of sense organ subsets without the concomitant appearance of new or abnormal sensory structures. Therefore, the amos null phenotype is unique for a Drosophila proneural gene (zur Lage, 2003).

Given its subtle molecular basis, the putative hypomorph amos² has a surprisingly strong phenotype: it has no sensilla trichodea, and sensilla basiconica are reduced very substantially. There are also fewer ectopic bristles than in the null alleles, but there are many sensilla of unusual morphology. In the case of this allele, these seem to represent intermediates between sensilla basiconica/trichodea and external sense organs (zur Lage, 2003).

Late pupal antennae were stained with a sensory neuron marker, MAb22C10, to visualize olfactory receptor neurons (ORNs). Olfactory sensilla are innervated by multiple sensory neurons, which can be seen as groups in the wild-type antenna. amos mutant antennae have many fewer neuronal groups, corresponding in number to the sensilla coeloconica and the bristles. There are instances of sensilla innervated by a single neuron, which appear to correspond to the ectopic bristles. In wild-type flies, ORN axons form three olfactory nerves leading to the antennal lobe of the brain. In amos mutant antennae, all three antennal nerves are still present, although consisting of fewer axons (comprising the axons of ato-dependent ORNs). Although thinner, the fascicles appear normal in structure and location. Thus, in contrast to ato, mutations of amos do not cause defects in routing or fasciculation of the olfactory nerves. This supports the conclusion that the ato-dependent sensory lineage provides the information for fasciculation of these nerves (zur Lage, 2003).

Olfactory precursors arise in the pupal antennal imaginal disc over an extended period of time. Given their high density, the appearance of olfactory precursors is complex and incompletely characterized. The evolution of this pattern was characterized by studying Senseless (Sens; a.k.a. Lyra) expression, which is a faithful indicator of proneural-derived sensory precursors and is probably a direct target of proneural proteins. Precursor formation occurs in three waves. Initially, Sens expression begins a few hours before puparium formation (BPF) in an outer semicircle of cells. A second wave begins at 0-4 hours after puparium formation (APF) to give a very characteristic pattern, including three semicircles of precursors. After this, a third wave appears over an extended period of time, with increasing numbers of cells appearing intercalated between the early precursors until no spatial pattern features can be observed (zur Lage, 2003).

Using a polyclonal antibody raised against the entire Amos protein, it was determined that amos expression begins at puparium formation in three distinct semicircles and then continues for the next 16 hours, with the semicircles becoming indistinct by around 8 hours APF. The characteristic early waves of SOPs arise between the Amos domains of expression and do not show overlap with Amos expression. However, the third wave of SOPs appears to arise from the Amos expression domains. These late SOPs co-express Amos, and their nuclei lie beneath the Amos expression domains, consistent with these cells being olfactory SOPs. Unusually, overlying amos proneural cluster expression is evidently not affected by lateral inhibition upon the appearance of the SOPs. ato is expressed much earlier than amos. All wave 1 and 2 SOPs appear to express Ato or to have arisen from Ato-expressing cells. Consistent with this, the entire early SOP pattern is missing in antennal discs from ato¹ mutant pupae. SOPs only begin to appear between 4 and 8 hours APF, corresponding to the third wave of precursors. These coincide very precisely with Amos expression, which itself appears unaffected (zur Lage, 2003).

In summary, there are three waves of olfactory precursor formation. The first and second waves are well defined, giving rise to the sensilla coeloconica of the sacculus and the antennal surface, respectively. These precursors express and require ato. The third wave of precursors is much more extensive and has little obvious pattern, giving rise to the more numerous sensilla basiconica and trichodea. Amos is expressed in a pattern entirely consistent with it being the proneural gene for the late wave precursors. Expression of Amos is complementary with that of Ato and is independent of Ato function. Thus, Ato- and Amos-expressing SOPs show a degree of spatial and temporal separation (zur Lage, 2003).

amos appears to be expressed in proneural domains and then in SOPs. For sc and ato, these two phases of expression are driven by separate enhancers, and SOP-specific enhancers have been identified. A 3.6 kb fragment upstream from amos was found to support GFP reporter gene expression in the pupal third antennal segment. Comparison with Amos and Sens expression shows that GFP coincides with the Amos but not Ato SOPs. This fragment therefore contains an amos SOP enhancer. Perduring GFP expression driven by the enhancer can be observed in large numbers of sensilla on the maturing pupal antenna. From their morphology, it is clear that the GFP-expressing subset are the sensilla trichodea and basiconica. This confirms that early SOPs form sensilla coeloconica whereas late SOPs produce sensilla trichodea and basiconica. Interestingly, these GFP-expressing sensilla differentiate late, because there is no overlap with the 22C10 marker until late in development. Thus, the timing of neuronal differentiation reflects the timing of SOP birth. These findings correlate with the differing effects of proneural genes on fasciculation as described above: the first-born ato-dependent cells organize the nerves, and the later amos-dependent ORNs follow passively (zur Lage, 2003).

Loss of SOPs is one of the defining characteristics of proneural gene mutations. SOP formation was examined in amos mutants relative to wild type by examining Sens expression. As expected from the expression analysis, the first two waves of SOP formation show little discernible difference in pattern between amos mutant antennal discs and wild-type discs. This is consistent with these early SOPs expressing and requiring ato, and they indeed express ato in a pattern indistinguishable from wild type. This shows that ato expression does not depend on amos function. After this, the later arising SOPs do not appear to form between rows of ato-dependent SOPs, corresponding to those cells shown to express amos. A few precursors do not express ato, and these may represent precursors of the ectopic bristles. This is supported by an analysis of Cut expression, which is the key molecular switch that must be activated to allow SOPs to take a bristle fate, and whose expression correlates with bristle SOPs. In the wild-type antenna, Cut is not expressed during olfactory SOP formation, although later it is expressed in differentiating cells of all olfactory sensilla. This expression normally appears after 16 hours and does not overlap with Amos. In amos mutant antennae, expression begins earlier than normal in a subset of SOPs that appear to correspond to the ones identified above (zur Lage, 2003).

By 16 hours APF, there is a large loss of Sens staining in amos mutants. The remaining cells tend to be in clusters as would be expected for the early ato-dependent sensilla, but otherwise the identity of these cells cannot be determined. Detection of Amos protein in amos¹ mutant antennal discs also shows that although the Amos domains are still present, the deeper Amos/Sens-expressing nuclei are absent. Thus, at least a large number of amos-associated SOPs are not formed in the amos mutant (zur Lage, 2003).

The processes and lineages by which olfactory SOPs lead to the differentiated cells of the olfactory sensillum are not entirely known. The limited information available comes from analysis of the early wave of SOPs, which represent the ato-dependent sensilla. After a SOP is selected there appears in its place a cluster of 2-3 cells expressing the A101 enhancer trap [the pre-sensillum cluster (PSC)]; this is apparently caused not by division of the SOP but perhaps by recruitment by the SOP, although the evidence for this is indirect. These PSC cells then divide to form the cells of the sensillum, including the outer support cells (hair and socket cells), inner support cells (sheath cells) and 1-4 neurons. For the early subset of SOPs, formation of the PSC occurs at a time in which amos is still expressed in the epithelial domains, and so amos could influence the development of these cells. Using A101 as a marker of the PSC cells, it was determined that amos is not expressed in recognisable PSCs at 8 or 16 hours APF. Moreover, there is also no apparent co-labelling of Amos and Pros (a marker of one of the PSC cells). This suggests either that early PSC cells do not derive from amos-expressing cells or that amos is switched off rapidly when cells join a PSC (zur Lage, 2003).

The situation appears different for the cells derived from amos-dependent SOPs. Surprisingly at 24 hours and beyond, the amos enhancer drives GFP expression in most or all cells of the differentiating sensilla basiconica and trichodea, including most or all of the neurons (recognised by Elav expression); the sheath cell (recognised by Pros expression), and the outer support cells (recognised by the higher expression of Cut). This suggests that the late PSC cells do derive from amos-expressing cells and that activation of an enhancer within the 3.6 kb regulatory fragment (possibly separate from the SOP enhancer) is part of their specification process, although amos expression itself may not be long lived in these cells (zur Lage, 2003).

amos mutant antennae have Cut-expressing SOPs, but, although cut expression decides SOP subtype fate, it does not specify ectodermal cells as SOPs de novo. To investigate the involvement of other proneural genes, it was first determined whether the bristles depended on ato, since ato is expressed in close proximity to the emerging bristle SOPs. Clones of amos¹ mutant tissue were induced in ato¹ mutant antennae. In such clones, all olfactory sensilla were absent, as expected, but ectopic bristles were still formed. Therefore the bristles do not depend on ato function (zur Lage, 2003).

Apart from giving rise to separate classes of olfactory precursor, there are interesting differences in the way that ato and amos are deployed in the antenna. Three waves of olfactory precursor formation were characterized. The first and second waves are well defined, giving rise to well-patterned sensilla coeloconica of the sacculus and the antennal surface, respectively. These precursors express and require ato. The third wave of precursors is much more extensive and has little obvious pattern; it gives rise to the much more numerous sensilla basiconica and trichodea. This wave expresses and requires amos. For the early waves, ato is expressed according to the established paradigm: it is expressed in small PNCs, each cluster giving rise to an individual precursor. The pattern of the PNCs is very precise and prefigures the characteristic pattern of precursors. amos expression is dramatically different. It is expressed in large ectodermal domains for an extended period of time. Densely packed precursors arise from this domain continuously without affecting the domain expression. This shows that singling out does not necessarily require shut down of proneural expression, and therefore has implications for how singling out occurs. In current models, it is assumed that PNC expression must be shut down to allow a SOP to assume its fate. The amos pattern better supports the idea that a mechanism of escaping from or becoming immune to lateral inhibition is more likely to be important generally. One prediction would be that amos and ato (and ac/sc) differ in their sensitivities to Notch-mediated lateral inhibition, a situation that has been noted for mammalian homologs (zur Lage, 2003).

Why are the proneural genes deployed so differently? One possibility is simply that there are very many more sensilla basiconica and trichodea than coeloconica. All the coeloconica precursors can be formed by ato action in a precise pattern in two defined waves. This would not be possible for the large number of basiconica and trichodea precursors, and so precursor selection has been modified for amos. Indeed, amos appears to be a particularly 'powerful' proneural gene when misexpressed. This may make amos a useful model of other neural systems in which large numbers of precursors must also be selected (zur Lage, 2003).

For most insects, the antenna is the major organ of sensory input. It is not only the site of olfaction, but also of thermoreception, hygroreception, vibration detection and proprioception, as well as of touch. Patterning the sensilla is therefore complex and three types of proneural gene are heavily involved to give different SOPs. It is clear that the study of antennal sensilla will provide a useful model for exploring the fate determining contribution of intrinsic bHLH protein specificity and extrinsic competence factors (zur Lage, 2003).

Programmed cell death and context dependent activation of the EGF pathway regulate gliogenesis in the Drosophila olfactory system: glial cells in the Atonal survive through activation of the EGF pathway

In the Drosophila antenna, sensory lineages selected by the basic helix-loop-helix transcription factor Atonal are gliogenic while those specified by the related protein Amos are not. What are the mechanisms that cause the two lineages to act differentially? Ectopic expression of the Baculovirus inhibitor of apoptosis protein (p35) rescues glial cells from the Amos-derived lineages, suggesting that precursors are removed by programmed cell death. In the wildtype, glial precursors express the extracellular-signal regulated kinase (phosphoERK) transiently, and antagonism of Epidermal growth factor pathway signaling compromises their development. It is suggested that all sensory lineages on the antenna are competent to produce glia but only those specified by Atonal respond to EGF signaling and survive. These results underscore the importance of developmental context of cell lineages in their responses to non-autonomous signaling in the choice between survival and death (Sen, 2004).

Several lines of investigation have ascertained that the first cells to divide in the sensory lineages are the secondary progenitors: PIIa, PIIb and PIIc. The numbers of sensory cells undergoing division at different times in the developing antenna were estimated by staining mitotic nuclei with antibodies against phosphorylated histone. A peak of cell division was observed between 16 and 24 h after puparium formation (APF). It has been considered that only in those sensory lineages specified by Ato, PIIb produces a glial cell and a tertiary progenitor, PIIIb, which in turn divides to form the sheath cell and a neuron. In Amos dependent lineages, PIIb is believed to directly give rise to a neuron and a sheath cell. The difference between the two lineages could be entirely dependent on the nature of the proneural genes activated; Amos, for example, could direct a non-gliogenic lineage. Alternatively, the two proneural genes could specify similar division patterns but the glial cell precursor in Amos-lineages could be removed by PCD, resulting in non-gliogenic lineages (Sen, 2004).

To test the latter possibility, cell death profiles were examined in developing pupal antennae using the terminal transferase assay (TUNEL) and attempts were made to correlate the timing of PCD with cell division profiles discussed above. The appearance of TUNEL-positive cells peaked between 22 and 24 h APF consistent with the occurrence of PCD immediately after division of secondary progenitors (Sen, 2004).

TUNEL reactions were performed on 22-24 h APF antennae from lz-Gal4; UAS-lacZnls and ato-Gal4; UAS-lacZnls animals. Double labeling with antibodies against ß-galactosidase marked sensory cells arising from the Lz and Ato lineages. Lz::lacZ overlaps the regions of the antennal disc where amos expression occurs and labels all the basiconic and trichoid sensilla in the mature (36 h APF) antenna. Hence for the purpose of this study, all cells in which lz-Gal4 expresses will be considered to belong to the Amos-dependent lineages. ato-Gal4 drives reporter activity in proneural domains of the disc and subsequently in all cells of the coeloconic sense organs (Sen, 2004).

Most of the apoptotic nuclei observed during olfactory sense organ development co-localized with Lz::LacZ suggesting that death occurred mainly within the 'Amos-dependent' sensory clusters. Only very few TUNEL-positive cells were detected in regions where ato-lacZ expressed and these did not co-localize with the reporter expression. If PCD is the mechanism used to remove glial precursors from Amos lineages, then their rescue would be expected to result in additional peripheral glia in the antenna (Sen, 2004).

The GAL4/UAS system was used to target ectopic expression of baculovirus inhibitor of apoptosis protein (p35) to different cell types within the developing antennal disc. distalless⁹⁸¹-Gal4 (henceforth called dll-Gal4), which drives expression in all cells of the antennal disc, resulted in the formation of >300 glial cells as compared to ~100 in the wildtype. Other sensory cells--neurons, sheath, socket and shaft cells--within sense organs were unaffected. Ectopic expression of p35 specifically in Ato lineages (ato::p35) did not alter glial number. This means that the `additional' glial cells rescued in dll::p35 must arise from lineages other than Ato. Mis-expression of p35 in Amos-dependent lineages using lz-Gal4, on the other hand, resulted in a significant increase in glial number. While other explanations are possible, it is believed that the somewhat lower number of glia obtained in lz::p35 as compared to dll::p35 could be accounted for by the strength of the P(Gal4) driver (Sen, 2004) (Sen, 2004).

In order to identify the cell within the Amos lineage that is fated to die, the cellular events during development of sense organs were re-examined. At approximately 12-14 h APF, most sensory cells are associated in clusters of secondary progenitors. Two cells in each cluster -- PIIb and PIIc -- express the homeodomain protein Prospero (Pros). pros-Gal4;UAS-GFP recapitulates Pros expression at this stage and marks PIIb and PIIc and their progeny in all olfactory lineages. In the wildtype, a Repo-positive cell was associated with only a few of the total sensory clusters, these were all located within the coeloconic domain of the antenna. Targeted expression of p35 using pros-Gal4 increased glial number indicating that cells which are the progeny of either PIIb or PIIc could be rescued from apoptosis. In the pros-Gal UAS-2XEGFP/UAS-p35 genotype, a glial cell was associated with most clusters at 18 h APF rather than in Ato lineages alone (Sen, 2004).

In order to directly visualize the cell undergoing apoptosis, 22-24 h APF antenna from the neu^A101 strain were stained with antibodies against ß-galactosidase to mark the sensory cells and with TUNEL. Sensory clusters located in basiconic and trichoid domains of the pupal antenna each had a single associated TUNEL positive cell. Since TUNEL reactivity data does not reflect the initiation of the death program, developing antennae were also stained at different time points with an antibody that recognized the activated caspase -- Drice. At 20 h APF, a single Drice-positive cell was found within each sensory cluster within the basiconic and trichoid domains of the pupal antenna. This cell also expressed low levels of Pros suggesting that it could arise from either PIIb or PIIc. This means that the PIIb/c in Amos lineages, like that in Ato, divides to give rise to a PIIIb and its sibling. The sibling in the former lineage was not previously detected because it expresses only low levels of Pros and soon dies. Since this cell is capable of expressing the glial-identity gene repo when rescued from death, it is denoted as a glial precursor (Sen, 2004).

How is apoptosis of a specific cell within the lineage regulated? In Drosophila three genes [reaper (rpr), grim and head involution defective (hid)] which all map under the Df(3L)H99 are necessary for the initiation of the death program. Heterozygotes of Df(3L)H99 show a small but significant increase in glial number over that of normal controls. hid-lacZ was used to follow expression during antennal development; reporter activity occurs at low levels ubiquitously including in glial cells. Levels of reporter expression indicate somewhat higher hid transcription in glia rescued by p35 mis-expression. The presence of Hid in the 'normal' glial precursors suggests a mechanism dependent on possible trophic factors to keep cells alive. In several other systems signaling, mainly through the EGFR pathway, results in an antagonism of Hid action and transcription. The sustained levels of hid transcription in the rescued glia, is not unexpected since inhibitors of apoptosis act by antagonizing a downstream event of caspase activation, rather than on Hid itself (Sen, 2004).

The peripheral olfactory glia ensheath the axonal fascicles as they project towards the brain. Antennae from animals deficient for ato (ato¹/Df(3R)p¹³), lack a large fraction of glia; the ~30 which remain, appear to arise elsewhere and migrate into the third segment of the antenna somewhat later during development (Sen, 2004).

Is death the default fate for all glial precursors and are there signals that keep Ato-glia alive? Several studies have provided compelling evidence for the role of receptor tyrosine kinase signaling in glial survival. In order to test this in the olfactory glia, 14-16 h APF pupal antennae were stained with antibodies against Pros and the phosphorylated form of ERK (pERK). The PIIb and PIIc cells are recognized by expression of Pros, which becomes asymmetrically localized in PIIb during mitosis. At this stage sensory cells do not express pERK. pERK was detected in the daughter of the first secondary progenitor to divide (probably PIIb). Staining with anti-Repo antibodies shows that this cell is glial. Clusters showing pERK expressing cells were observed only in regions of the pupal antenna from where the coeloconic sensilla originate and not in the Amos-lineage sensory clusters. pERK expression decays rapidly as Repo rises; only occasional cells can be detected that show immunoreactivity to both (Sen, 2004).

The presence of pERK in nascent glial precursors is evidence for receptor tyrosine kinase pathway signaling in these cells. Since EGFR activation is a well recognized signal for glial survival, the role of this pathway in glial cell development was tested. In order to demonstrate a role for EGF signaling during glial formation in the antenna, a hs-Gal4 strain and carefully timed heat-pulses were used to drive transient expression of various antagonists of the pathway. Ectopic expression of a dominant negative form of human Ras (DN-RasN17) between 14 and 15 h APF, severely reduced the numbers of antennal glia. Similarly, expression of the inhibitory ligand Argos or a dominant negative construct of Drosophila EGFR (DN-EGFR) compromises glial development. Since abrogation of signaling could, in principle, affect other developmental events, antenna from 36 h APF pupae of relevant genotypes were stained with support cell and neuronal markers to ascertain that these treatments did not affect sense organ development generally (Sen, 2004) (Sen, 2004).

Interference with EGFR activity affects glial number, suggesting that signaling is required for glial development and/or survival. Cells in the Amos-lineage also produce glial precursors, which do not express pERK. This implies that Amos-lineage glial precursors fail to experience EGF signaling and are fated to die. This hypothesis could, in principle, be tested by constitutively activating the pathway in Amos lineages. These experiments were not possible to carry out since activation of EGFR at the time when Amos-dependent secondary progenitors were undergoing division resulted in pupal lethality. The spatial and temporal expression of Vein-lacZ and Sprouty-lacZ (Sty-LacZ) supports a role in development of glia and requires further genetic analysis. Vein-LacZ was first detected at 18 h APF and expression was elevated at 20 h APF in a domain of the pupal antenna occupied mainly by coeloconic sensilla. The fact that this is the region of the antenna from where most glia originate, is interesting in the light of data from other systems that demonstrate that Vein acts as gliotrophin (Sen, 2004).

The spatial and temporal pattern of Vein and Sty expression is consistent with a role for these ligands in glial development. The region where Vein is expressed shows high pERK activity suggesting activation of the EFGR pathway in the coeloconic (Ato) domains of the developing antenna. Sty, however, is present in the basiconic and trichoid (Amos) domains of the antenna. At 10-14 h, when the secondary progenitors have not yet divided, Sty expressing cells lie adjacent to the PIIb and PIIc; these cells are labeled by Pros. This localization is consistent with a role for Sty in antagonizing EGF activity in the secondary progenitors (Sen, 2004) (Sen, 2004).

This work provides one more instance where PCD plays a crucial role in the selection of a specific population of cell types during development although the mechanisms employed still need to be elucidated. How does the development context of lineages of cells within a single epidermal field together with non-autonomous cues result in distinct consequences? The Drosophila antenna is a valuable system to address these issues because of the diversity of morphologically and molecularly distinct cell types located in a highly stereotyped architecture and the wealth of reagents available for study (Sen, 2004).

Effects of Mutation or Deletion

Chordotonal organs are eliminated in embryos carrying chromosomal deficiencies that remove atonal (Jarman, 1993). atonal is semi-lethal. Homozygotes are almost eyeless and also lack ocelli. There is a remnant of the morphogenetic furrow with extensive cell death posterior to the furrow. The optic lobe is reduced (Jarman, 1994).

In Drosophila, two sets of proneural genes, atonal and members of the achaete-scute complex are required for the formation of chordotonal (ch) organs. Overexpression of ato produces numerous ectopic ch organs along wing veins of the proximal quarter of the wing. About 64% of the ch organs form clusters ranging from two to five in a group. Overexpression of ATO also induces the formation of bristles. Overexpression of Scute does not increase the number of ch organs but does produce an abundance of ectopic bristles. Information that specifies ch organs resides in the bHLH domain of ATO: chimeras containing the basic domain of ATO and the HLH domain of Scute also induce ch organ formation, but to a lesser extent than those containing the bHLH domain of ATO. The basic domains of ATO and SC differ in seven residues. Mutations of these seven residues in the basic domain of ATO suggest that most or perhaps all of these residues are required for induction of ch organs. None of the seven residues is predicted to contact DNA directly by computer simulation using the structure of the myogenic factor MyoD as a model, implying that interaction of ATO with other cofactors is likely to be involved in neuronal type specification (Chien, 1996).

To test whether proneural genes not normally expressed in a specific precursor can function in the specification of neural precursor identity in the CNS, proneural proteins not normally expressed in the MP2 precursor cell were ectopically expressed in MP2 precursors. Unlike other CNS precursors, MP2 normally expresses achaete and scute only. MP2 is unique in several respects: in its localization of Prospero protein to the nucleus soon after delamination, in the subsequent expression of the Fushi tarazu protein, and in the expression of the 22C10 antigen shortly before the MP2 division. lethal of scute, atonal and asense, each not normally expressed in MP2, were tested for their ability to specifiy MP2 when ectopically expressed in the MP2 precursor. All proneural proteins are similarly able to promote the segregation of a neural precursor at the MP2 position but show distinct capabilities in its specification. achaete/scute mutants have CNS precursors in only 17% of the hemisegments, as determined with antibodies against Hunchback, a general neuroblast marker. A significant fraction of the remaining precursors do not express nuclear Prospero or Fushi tarazu, and none of the precursors express appreciable amounts of 22C10 antigen. Targeted expression of either achaete or scute is sufficient to sustain formation of normal MP2 precursors. Targeted expression of either l'sc, ase or ato promotes delamination of a neural precurosor at the MP2 position, however this putatitive MP2 precursor expresses FTZ only when its development is prompted by l'sc or ato; expression of 22C10 occurs only in the case of ato. Thus Ato seems to be an efficient replacement for Ac/Sc. Targeted Ato results in both progeny neurons projecting to the anterior, instead of the normal situation in which one progeny neuron projects to the anterior and the second projects to the posterior. Thus Ato causes the misspecification of the MP2 lineage. It is concluded that totally normal specification of the MP2 fate can only be attained by the proneural genes achaete or scute (Parras, 1996).

The antenna of the adult fruit fly, Drosophila melanogaster, is covered with three morphologically distinct types of olfactory sense organs. In addition, mechano- and hygro-sensitive receptors are also present on its surface. While much has been learned about the development of the peripheral nervous system in Drosophila, the mechanisms underlying the development of olfactory sensilla are just beginning to be unraveled. The antennal sense organs have several properties that make them distinct from other sense organs. While each sensillum type is arranged in a well-defined region of the antenna, the position of an individual sensillum is not fixed. The development of these sense organs appears to combine an initial step of cell recruitment, as in photoreceptors, followed by cell lineage mechanisms, as in the development of other external sense organs. The earliest step in development, the selection of a sensory organ precursor, involves the interaction of proneural and neurogenic genes. Efforts to identify the proneural gene for the antennal sense organs have not met with much success to date. The basic helix-loop-helix (bHLH) transcription factor encoded by atonal is here shown to be a proneural gene for one morphological type of olfactory sensilla on the antenna and for all the olfactory sensilla on the maxillary palp. Loss of function and overexpression experiments together reveal that ato is both necessary and sufficient to specify these sensilla. Immunohistochemical experiments show that Ato expresses in a dynamic pattern in the developing antennal disc. These results demonstrate that ato acts solely in the specification of antennal sensilla coeloconica. This along with the observation that the AS-C genes do not function in antenna suggests that other proneural genes must operate in the specification of sensilla basiconica and trichoidea. Overexpression of extramacrochaetae, a negative regulator of bHLH encoding genes, results in a significant reduction in the number of all three types of antennal sensilla. This suggests that the unidentified proneural gene(s) possibly encode bHLH factors (Gupta, 1997).

In mutants lacking both the Achaete-Scute Complex (ASC) and atonal, coding for proteins known to establish SOP cell fate, two to three neurons of the solo-MD type remain in the dorsal region of each abdominal hemisegment. The identity of these remaining neurons are controlled by absent MD neurons and olfactory sensilla (amos). Both ASC proteins and Atonal physically interact with the protein Daughterless (Da). Since the solo-MD neurons that exist in ASC;ato double mutants are eliminated in da mutants, this observation implies that the proneural gene for those solo-MD neurons also encodes a transcription factor of the bHLH family (Huang, 2000).

Amos is a novel bHLH protein and is expressed in proneural clusters. This expression is later restricted to SOP cells. The technique of double-stranded RNA interference (RNAi), a procedure that mimics the effect of mutation, was used to eliminate amos function in embryos. Loss of amos function eliminates MD neurons that remain in ASC;atonal mutants. These results suggested that amos is required for the formation of dbp and some dmd neurons, the neurons, which remain in ASC;ato double mutants. When misexpressed, amos induces the formation of all types of sensory neurons. These results provide strong evidenct that intrinsic differences among the proneural genes, specifically genes of the ASC, atonal and amos play a role in the induction of different types of sensory neurons (Huang, 2000).

During Drosophila eye development, the proneural gene atonal specifies founding R8 photoreceptors of individual ommatidia, evenly spaced relative to one another in a pattern that prefigures ommatidial organization in the mature compound eye. Beyond providing neural competence, however, it has remained unclear to what extent atonal controls specific R8 properties. Reduced Atonal function gives rise to R8 photoreceptors that are functionally compromised: both recruitment and axon pathfinding defects are evident. Conversely, prolonged Atonal expression in R8 photoreceptors induces defects in inductive recruitment as a consequence of hyperactive EGFR signaling. Surprisingly, such prolonged expression also results in R8 pattern formation defects in a process associated with both Hedgehog and Receptor Tyrosine Kinase signaling. These results strongly suggest that Atonal regulates signaling and other properties of R8 precursors (White, 2000).

The data indicate that in R8 ato modulates Hh signaling activity, initiates EGFR-dependent recruitment, and possibly regulates axon guidance. These findings support the concept that proneural bHLH transcription factors function not only to endow neural competence but also to control specific neural subtype properties. This has important implications not only for Drosophila neurogenesis, but also for the large number of vertebrate Ato and AS-C homologues that have been identified. Indeed, recent studies indicate that at least some of the vertebrate Ato homologues may also have subtype determining properties (White, 2000).

The hypothesis that ato determines characteristics within selected neural precursors was initially intimated from ato misexpression studies during chordotonal organ precursor selection. ato represses the homeobox gene cut, thereby diverting these precursors from an external sense organ fate and allowing them to adopt the alternative chordotonal organ fate. Clearly, such aspects of ato's subtype-determining properties must be contingent on the developmental context (chordotonal versus R8 precursor). One property, however, appears to be common to both developmental contexts: EGFR-mediated local recruitment. Both photoreceptors and chordotonal precursors are clustered as a consequence of EGFR-dependent recruitment. In the case of chordotonal recruitment, it has been proposed that ato directly controls the level of EGFR signaling. In the eye, although it is clear that R8 initiates recruitment, it was previously unclear whether this property is directly controlled by ato. The experiments described here suggest that it is. Prolonging ato activity in R8 results in excessive recruitment as a result of hyperactive Egfr signaling. Conversely, reducing ato activity specifically in selected R8 precursors results in ommatidia that contain fewer photoreceptors than wild type. It is postulated that these R8 precursors are unable to trigger recruitment efficiently. It is instructive to compare this with the Ellipse (Elp) mutation: superficially, the reduction in R8 selection observed in Elp mutants resembles that of ato mutation. In contrast to ato partial loss of function, however, the remaining ommatidia are of normal structure in Elp eyes, indicating that although fewer R8 precursors arise, they can function normally with respect to recruitment. Strikingly, whilst ato expression in selected R8 precursors is normal in Elp mutants, it is not apparent in the R8 precursors in ato partial loss of function, consistent with a role for Ato after R8 selection has occurred. In wild type, ato presumably triggers the initiation of signaling, since recruitment continues after ato is switched off. The appearance of ectopic R7 cells upon misexpression is, in this respect, an artefact of the delayed developmental timing of ato misexpression relative to wild type (White, 2000).

Expression of ato in R8 photoreceptors induces defects in pattern formation at multiple stages in the process of 'singling out' of R8 cells. Thus, ommatidial spacing is irregular as a result of aberrant spacing of the intermediate clusters, while twinning of R8 precursors is a consequence of a perturbation in fate refinement in the R8 equivalence group. Although ato is likely to be directly involved in these processes, the effect detected here is novel and unexpected, since ato has been misexpressed after these events have finished and so must primarily affect R8 behaviour subsequent to these processes. The effect of ato misexpression must therefore be communicated from the differentiating R8 cells to the MF. It is significant, therefore, that R8 patterning disruption in heightened ato appears to be associated with both Hh- and Raf-dependent signaling. In both cases, increased signaling is associated with enhanced R8 formation. The Raf-dependent signaling is therefore distinct from the role of ato and EGFR signaling in intermediate cluster spacing, which appears to be inhibitory via the activation of rough. Thus, a new ato- and Raf-dependent process has been detected that acts to promote R8 formation later on (White, 2000).

In the expectation that more bHLH genes are required in neurogenesis in Drosophila, new bHLH genes have been sought in PCR experiments using degenerate primers based on homology to atonal. In addition to ato itself, the PCR product contained potential bHLH-encoding sequences from two new closely related genes, which have been termed amos and cousin of atonal (cato). cato is expressed widely in the developing PNS after neural precursor selection but before terminal differentiation. Consistent with this pattern, cato appears to be required for proper sensory neuron morphology (Golding, 2000b).

It was asked whether cato would similarly be functionally indistinguishable from ato in this ectopic sense organ assay. Initial observations of Gal4^109-68/UAS-cato flies indicate that cato can indeed behave as an ato-like proneural gene in directing the formation of a mixture of ectopic external sense organs and chordotonal organs. Like ato, at low misexpression levels cato induces extra external sense organs, but at higher levels external sense organs fail to form. Concomitantly, ectopic chordotonal organs are now present within the scutellum. A more detailed analysis, however, showed several qualitative and quantitative differences between the UAS-cato and the UAS-ato phenotypes. One noteworthy difference is that only under conditions of very strong cato misexpression are ectopic chordotonal organs generated, and even then far fewer chordotonal organs are formed than for UAS-ato. Also, these sensilla tend to be isolated or in small, poorly formed clusters of chordotonal organs. Such clusters are quite different in appearance from the densely packed arrays produced in Gal4^109-68/UAS-ato flies. These results suggest that although cato can behave inappropriately as a chordotonal proneural gene, it cannot efficiently bypass the requirement for ato in chordotonal SOP formation, and thus the proteins are not functionally equivalent (Golding, 2000b).

A further difference with ato is that the few chordotonal organs formed by UAS-cato tend to be malformed. Indeed it is often difficult to identify ectopic chordotonal organs on morphology alone, their scolopale structures being poorly differentiated. Additionally, the scolopale cells often stain inappropriately with the neuronal marker 22C10. The differentiation of external sense organs is also affected, again most notably with the malformation of support cells (socket and hair). It seems likely that disrupting the normal pattern of cato expression (by misexpression) affects sense organ cellular differentiation. This effect is also seen after misexpression in the embryo using a hairy-Gal4 driver line. With this driver, misexpression of ato results in expansion of lateral chordotonal organ clusters (usually 6-10 neurons instead of 5) but the neurons are of normal morphology. In contrast, misexpression of cato results in less expansion of the chordotonal clusters (usually 5-7 neurons); instead, there are organizational and dendritic abnormalities in the chordotonal neurones that are reminiscent of those seen upon cato loss (Golding, 2000b).

During development of the Drosophila peripheral nervous system, different proneural genes encoding basic helix-loop-helix transcription factors are required for different sensory organs to form. atonal is the proneural gene required for chordotonal organs and R8 photoreceptors, whereas the achaete-scute complex contains proneural genes for external sensory organs such as the macrochaetae, large sensory bristles. Whereas ectopic ato expression induces chordotonal organ formation, ectopic scute expression produces external sensory organs but not chordotonal organs in the wing. Proneural genes thus appear to specify the sensory organ type. In the ommatidium, or unit eye, R8 is the first photoreceptor to form and appears to recruit other photoreceptors and support cells. In the atonal1 mutant, R8 photoreceptors fail to form, thereby resulting in the complete absence of ommatidia. Ectopic scute expression in the ato1 mutant induces the formation of ommatidia, which occasionally sprout ectopic macrochaetae. Remarkably, many scute-induced ommatidia lack R8 although they contain outer photoreceptors (Sun, 2000).

How might ectopic expression of scute induce R8-less ommatidia, given that R8 normally is required for the formation of other photoreceptors and support cells of the ommatidia? It is probably not because of the ability of scute to induce cut expression, because only cone cells in an ommatidium normally express cut and ectopic expression of scute in the eye causes only a small number of cells in the ommatidia to express Cut. Presumably, expression of genes that specify the eye primordia, such as the Pax gene eyeless, directs scute to act on a different set of downstream genes than those involved in the formation of sensory bristles. Although it is possible that scute induces latent R8 precursors that express none of the known R8 markers but can still recruit the R1-7 photoreceptors, the following possibility is favored. Once R8 is induced by ato during normal eye development, it recruits other photoreceptors by inducing the expression of yet unidentified gene(s) that encode basic helix-loop-helix protein. This protein would share significant sequence similarity with either Scute or both Scute and Atonal. Ectopic scute expression in ato1 mutants mimics the action of this gene(s) and induces the formation of outer photoreceptors, thereby bypassing the normal requirement of R8 founder photoreceptors. It will be very interesting to identify this hypothetical basic helix-loop-helix gene(s) and study its function in eye development (Sun, 2000).

To test the role of sensory feedback in song production. the courtship songs were studied of Drosophila males expressing auditory mutations. The courtship songs of atonal (ato), beethoven (btv) and touch-insensitive-larva-B (tilB) were compared to wild-type songs. These mutations have in common the fact that the chordotonal organs are disrupted. Since chordotonal organs subserve both hearing (in the antenna) and proprioception (from the wing), these two potential routes for sensory feedback are defective in the mutant flies. Six song characters were measured: pulse number within a train, inter-pulse interval, pulse duration, sine burst duration, the carrier frequency of the sine song and the relative amplitude of the sine song. Using multivariate analysis, significant differences were found between mutant and normal songs. In addition, many mutant flies exhibit an unusual wing position during singing. The results indicate that sensory feedback plays an important role in shaping the courtship song of Drosophila (Tauber, 2001).

Atonal and the Olfactory lobe

Patterning of the antennal lobe of adult Drosophila occurs through a complex interaction between sensory neurons, glia, and central neurons of larval and adult origin. Neurons from the olfactory sense organs are organized into distinct fascicles lined by glial cells. The glia originate from one of the three types of sensory lineages, specified by the proneural gene atonal. Gain-of-function as well as loss-of-function analysis validates a role for cells of the Atonal lineage in the ordered fasciculation of sensory neurons. Upon entry of the antennal nerve to central regions, sensory neurons at first remain closely associated with central glia that lie around the periphery of the lobe anlage. Coincident with the arrival of sensory neurons into the brain, glial precursors undergo mitosis and neural precursors expressing Dachshund appear around the lobe. Sensory neurons and glial cells project into the lobe at around the same time and are likely to coordinate the correct localization of different glomeruli. The influence of sensory neurons on the development of the olfactory lobe could serve to match and lock peripheral and central properties important for the generation of olfactory behavior (Jhaveri, 2000a).

Each antennal or maxillary palps sense organ contains between one and four neurons; a total of about 1200 sensory neurons terminate in 43 glomeruli in each antennal lobe of the brain. The organization of the Drosophila antennal lobe is very similar to that in other insects and is also comparable in design to that of vertebrates. Each sensory neuron arborizes in a single glomerulus and, in most cases, also sends a branch to the corresponding glomerulus on the contralateral side. Within each glomerulus, sensory neurons make contact with a number of local interneurons that connect within different glomeruli in both lobes. Output neurons are uniglomerular and connect to the lateral region of the protocerebrum; some interneurons also send collaterals to the calyx of the mushroom bodies (Jhaveri, 2000a).

Antennal lobe development in the moth Manduca sexta has been extensively studied. Afferent olfactory neurons have been shown to play a key role in the initiation of glomerular development. Glial cells present around the lobe prior to the entry of sensory neurons provide cues for sorting within the lobe. Hence the patterning of the antennal lobe involves interaction between sensory neurons, lobe interneurons, and glial cells. The nature of these interactive cues needs to be dissected using a genetically tractable system. In Drosophila, the adult antennal lobe develops by modification of its larval counterpart as well as by de novo neurogenesis. Olfactory neurons develop from cells of the antennal disc and are guided to the brain by persistent larval neurons (Jhaveri, 2000a).

An important role is played by the glia and/or neurons of the Ato lineage in the organization of sensory axons into fascicles before they enter the antennal nerve. Most of the peripheral glia arise from sensory precursors specified by Ato that migrate over the antenna, and, as seen by both loss-of-function and gain-of-function experiments, such precursors could play a key role in the fasciculation of sensory neurons. What regulates the migration of glial cells? How do glial cells organize neurons? Several studies point to possible regulatory mechanisms. Glial migration has been well studied in the developing Drosophila eye where glia act in guidance of the photoreceptor axons into the axon stalk. In the embryonic CNS, glial cells are essential for selective fasciculation of axons in longitudinal pathways. Evidence accumulating from a number of laboratories suggests an intimate interaction between neurons and glia that regulates survival and function of both types of cells. In the context of the developing antenna, it appears that the cells of the Ato lineage are involved in the bulk of peripheral gliogenesis. In strong lozenge (lz) hypomorphs in which all basiconica and most of the trichoid sense organs are absent, the correct numbers of glia are formed and these are located at correct positions on the antennal surface. However, in ato hypomorphs only about 30% of the antennal glia form. In these mutants none of the sensilla coeloconic form, suggesting that these cells arise from a lineage independent of Ato. Two possible origins for this population of cells are proposed. (1) They could arise from a Lz-independent trichoid lineage. In antenna from lz null alleles, about 45 of the 145 trichoid sensilla still form. These could define a distinct set of sensilla that, like the coeloconica, are also gliogenic. (2) The residual glial population could arise in the CNS and migrate peripherally. The appearance of this glial sub-population correlates with a time when the antennal nerve has connected to the brain. During eye development peripheral migration of glia does not require a connection to the CNS and can occur by chemotaxis. A clonal analysis is currently being carried out to decipher the origin of this Ato-independent subset of glial cells (Jhaveri, 2000a).

How do olfactory afferents identify the location of the olfactory lobe? The extent to which larval structures prefigure the organization and function of adult antennal glomeruli is not known. Like the adult, Drosophila larvae demonstrate well-defined olfactory behaviors and are able to detect and respond to a wide repertoire of chemical stimuli. Behavioral experiments have shown that olfactory experiences during the larval stage can influence odorant preferences in the adult. The olfactory system of the larva is relatively simple; there are 21 sensory neurons that project into the olfactory lobe, although the circuitry still awaits a detailed description. During metamorphosis most larval sensory structures are histolysed and replaced by adult structures which develop de novo from imaginal discs; the adult antenna develops from cells of the eye-antennal disc within the first 40 h of pupation (Jhaveri, 2000a).

Neural elements from the antennal disc are guided to the brain by the larval antennal nerve that escapes histolysis during pupation. The presence of glial cells and neurons is observed in the immature antennal lobe well before entry of the sensory neurons. While other explanations are possible, these cells could be larval in origin (Jhaveri, 2000a).

The GAL4 line SG18.1 was used to drive n-synaptobrevin GFP in the olfactory neurons, thus allowing the visualization of developing synapses. Synapses of the olfactory neurons remain on the periphery of the glomerulus creating a 'doughnut-shaped' appearance. Such preparations were double labeled with an antibody against the immunoglobulin superfamily member Fasciclin II (Fas II), which has been implicated in the process of fasciculation, growth cone guidance, and development of the synapses. A single glomerulus shows Fas II staining at 36 h after puparium formation (APFP). This glomerulus, presumably VA1, is the first to differentiate and express synaptic markers. By 60 h APF several glomeruli demonstrate Fas II staining although at differing intensities. Fas II has been used as a marker for glomerular formation in different mutant combinations. lz³ alleles lack all basiconic sense organs and show a decrease in the numbers of trichoidea sensilla. Despite a total reduction of about 60% of the sensory afferents, only one of the approximately 43 glomeruli fails to form. This glomerulus -- V -- receives input exclusively from the basiconic sense organs; its absence further strengthens the idea that sensory input glomeruli are innervated by projections from multiple sense organ types. Careful examination of serial sections of Fas II-stained antennal lobes from lz³ animals led the authors to conclude that all other glomeruli are present in their normal locations. Antennal lobes from ato mutants show a dramatic disorganization of all glomeruli. In most of the lobes examined, no glomeruli could be discerned at all, while in a few cases, only one or two glomeruli were apparent. This incompletely penetrant defect could be explained by the fact that the ato allele used has some residual gene function. The wide-spread effect on several glomeruli in these mutants is rather surprising since neurons from the coeloconic sensilla contribute only about 15% of the total sensory input. Further, the projections from these sense organs are known to terminate in glomeruli that also receive input from other sensillar types. This effect on glomerular development requires further investigation and could be explained by a role for Ato in neurite growth per se. It is noteworthy that Ato expression is detected in olfactory neurons at a time when sensory neurons reach the brain (Jhaveri, 2000a).

Studies on the moth Manduca provide strong evidence that sensory neuron ingrowth into the lobe plays an important role in the initiation of glomerular formation. Transplantation of a male antennal disc into a female leads to the formation of a macroglomerular structure that is known to be male specific. What are the mechanisms by which sensory neurons pattern the development of their central targets? While the inductive influence of sensory neurons upon lobe patterning has been extensively studied in the Drosophila visual system, there are likely to be some key differences between the mechanisms used in different situations. Incoming photoreceptor axons contact laminar neuronal precursors and induce them to divide, presumably due to transmission of the secreted molecule Hedgehog. Entry of the olfactory neurons into the brain does not elicit cell division among neuronal precursors. A large increase in the number of cells that express Dac, which marks neuronal precursors, is observed just prior to differentiation. The progenitors of these cells could not be identified, although they could arise from 'nests' of arrested postmitotic neurons that differentiate upon receiving 'inductive' cues during metamorphosis (Jhaveri, 2000a).

Mitosis in glial cell progenitors is observed coincident with the arrival of the sensory afferents into the lobe. The speculation that glial cells could act as intermediates between the afferents and the lobe interneurons was first made in the moth M. sexta. Here, glial cell changes are observed upon entry of the olfactory receptor neurons and ablation of glia results in a failure of glomerular formation (Jhaveri, 2000a).

The cellular organization of the olfactory glomeruli in both insects and vertebrates involves sensory axons, anaxonic lobe interneurons, and so-called output neurons that connect to the higher centers. In both cases, glomerular architecture is sculpted and maintained by glial cells. Activity mapping, calcium imaging, and voltage-sensitive dyes in the olfactory lobes of a variety of species suggests that the glomerulus is a unit of odor coding. Markers have been used to trace the development of the rat olfactory bulb and describe events remarkably similar to those observed in this study. This represents an amazing conservation of developmental programs in species separated by nearly 600 million years of evolutionary history. Whether this indicates the presence of a common ancestral olfactory lobe or convergent evolution requires further investigation. Such an analysis would require knowledge of molecular mechanisms underlying development of the olfactory system. The framework provided in this study allows a detailed genetic analysis of how sensory neurons, interneurons, and glia interact to form the structural units underlying olfactory coding (Jhaveri, 2000a).

The first centers for the processing of odor information lie in the olfactory lobe. Sensory neurons from the periphery synapse with interneurons in anatomically recognizable units, termed glomeruli, are seen in both insects and vertebrates. The mechanisms that underlie the formation of functional maps of the odor-world in the glomeruli within the olfactory lobe remains unclear. This study addresses the basis of sensory targeting in the Drosophila; one class of sensory neurons, those of the Atonal lineage, plays a crucial role in glomerular patterning. Atonal-dependent neurons pioneer the segregation of other classes of sensory neurons into distinct glomeruli (Jhaveri, 2002).

There are three morphologically distinct sense organs on the surface of the antenna: sensilla coeloconica, basiconica and trichoidea. Atonal (Ato) is necessary and sufficient for the formation of the coleoconic sensilla. A related proneural protein, Amos, is required for the specification of the other two types of sensilla. Expression of amos is regulated by the runt domain transcription factor Lozenge (Lz), which appears to regulate the choice between basiconic and trichoid sensilla in a dose-dependent manner. Neurons of the Ato lineage are the first to enter the nascent adult olfactory lobe and are necessary for the correct targeting of other sensory neurons into the lobe. In mutants where Ato neurons fail to form, the afferents of other sense organs at first remain stalled at the periphery and although they ultimately enter the lobe, they fail to target correctly. As a consequence, glomerular formation is disrupted. The projection of central glial cells within the lobe is affected and postsynaptic interneurons enter the lobe normally but fail to terminate in glomeruli. This suggests that an intimate interaction between sensory neurons, projection neurons and glia during development orchestrates glomerular patterning, with sensory neurons of one class pioneering the organization of all sensory inputs crucial for this process (Jhaveri, 2002).

Sensory progenitors selected by ato⁺ function give rise not only to neurons and support cells of the sensilla coleoconica but also to the majority of the antennal glia. Repo-positive glial cells migrate along the developing neurons and their processes ensheath the sensory axons into three distinct fascicles. In ato¹/Df(3R)p¹³ animals, peripheral patterning is affected; only one major fascicle is observed as the axons exit the antenna. There is also a striking reduction in the number of peripheral glia; only 35±1 glia could be counted compared with 100±6 in the wild type. The majority of the antennal glia therefore arise from the gliogenic Ato lineage, while a small population are Ato independent and could have a central origin (Jhaveri, 2002).

Since ato mutants affect both glial as well as sensory neurons, the primary cause of the fasciculation defect in the antenna is unclear. In order to test the possible causal role of glia in sensory fasciculation, development of these cells was specifically perturbed using the Gal4/UAS system. Mis-expression of a dominant negative form of Cdc42 (DN-Cdc42) using the glial-specific MZ317-Gal4 line, would be expected to block cell growth because of disruption of cytoskeletal polarity. A reduction in glial cell number was observed, presumably because of cell death, and residual cells were aggregated with no apparent processes. Sensory neurons show a range of defects in fasciculation, some of which closely resembled those of ato mutants (Jhaveri, 2002).

Results from the genetic manipulation of sensory neurons on the antenna can be interpreted in the context of the events that take place during normal development. The adult olfactory lobe in Drosophila is built upon a pre-existing larval structure by addition of interneurons that arise by proliferation of identified neuroblasts during larval life. The glomerular innervation of these interneurons is pre-specified by lineage and birth order, and they enter the pupal lobe well before sensory neurons are chosen from cells on the antennal disc. At this stage, however, there is no evidence of glomerular patterning and dendrites of the projection neurons have not terminated into their appropriate proto-glomeruli. The first olfactory neurons to enter the brain are those of the Ato-lineage, which remain at the periphery of the lobe for several hours after arrival. These afferents, together with those of the basiconic and trichoid sense organs, which arrive later, invade the lobe at around 30 hours APF. Glial processes are elaborated and dendritic arborization of projection neurons occurs after targeting of sensory afferents is initiated. Neurons of the Ato lineage are key players in these patterning events; in their absence, none of the other sensory afferents or the projection neurons or glia can target correctly to glomeruli\. It is proposed that neurons projecting from the Ato-derived sense organs play a functional role as pioneers (Jhaveri, 2002).

Drosophila larvae exhibit rather well developed olfactory behavior, although the level of discrimination is not as sophisticated as that in the adult. Exposure to chemical stimuli at the larval stage can influence olfactory behavior in the adult. It is tempting to speculate that the larval olfactory lobe provides a pre-pattern upon which adult structures are remodeled. There are about 21 larval olfactory neurons that project from the dorsal organ to glomerular-like terminals within the lobe. Before histolysis, the larval antennal nerve serves to guide axons of the adult olfactory neurons, which arise in the antennal disc, to the brain (Jhaveri, 2002).

Evidence in a variety of insect species has shown that in the adult, sensory inputs are instructive for glomerular development. The instructive role of afferents can be divided into two steps: initially, a crucial role for one class of sensory neurons and then a possible general role for sensory inputs. It is remarkable that the Ato-lineage neurons, which comprise only a small percentage of the total number (approximately 200 out of a total of 1200), exert a disproportionate effect on glomerular development. In addition, the presence of only a small fraction of neurons of this lineage (~50) observed in a weak hypomorphic (ato¹/ato²) allelic combination are sufficient to guide normal glomerular formation. In lz nulls, however, only the coeloconic and a few trichoid sensilla still remain, accounting for about 300 sensory afferents, i.e. about 900 neurons are absent. These afferents are capable of patterning the majority, if not all, glomeruli (Jhaveri, 2002).

Cobalt fills from the coeloconic-rich region of the third antennal segment provided evidence for projections, although weak, to a large number of glomeruli. This is supported by results from experiments in which the afferents from Ato-dependent neurons were traced in an Ato-Gal4; UAS-GFP strain. This would support a hypothesis that Ato-neurons project to most/all glomeruli and serve to guide neurons from other lineages to these glomeruli (Jhaveri, 2002).

While glomerular patterning fails in ato mutants, indicating that the Ato-lineage neurons are essential for this process, it could be argued that they are not sufficient to pattern glomeruli. In lz nulls, about 300 olfactory sensory neurons are seen, about half of which arise from Lz-independent trichoidea. Could these remnant sensilla trichoidea be involved in a pioneer role similar to those of the Ato-lineage? These results cannot exclude such a role for these neurons. However, there are several reasons, taken together, that suggest Ato-lineage neurons are sufficient for the process of glomerular patterning: (1) neurons of the Ato-lineage arrive at the olfactory lobe before other sensory afferents; (2) mis-expression of Ato in all lineages using Neuralized-Gal4 converts most neurons to coeloconica -- in this situation, glomerular patterning is normal. (3) Mutants in the proneural gene amos lack all basiconic and trichoid sensilla and, here too, glomerular formation takes place. Thus, with the current availability of reagents and methods, it appears that Ato-lineage neurons are sufficient for glomerular development (Jhaveri, 2002).

What is the role played by Ato-dependent neurons in setting up the functional map in the antennal lobe? Functional mapping experiments have shown that odor quality is represented as a spatial map of activity among the olfactory glomeruli. Neurons expressing a given olfactory receptor terminate within the same glomeruli in the Drosophila antennal lobe. A similar topographic mapping has been described in vertebrates: evidence from these systems suggests that odorant receptors themselves play a role in guiding neurons to their targets. The mechanisms that underlie such guidance phenomena are unclear and, in addition, raise the question of how receptor gene regulation is specifically regulated. In Drosophila, the onset of receptor expression relative to the time of lobe development makes it very unlikely that odorant receptors play a role in the primary events in receptor axon targeting. Moreover, ubiquitous expression of some of the olfactory receptors in all the sensory neurons throughout development does not affect glomerular patterning (Jhaveri, 2002).

Neurons of the Ato lineage have the capacity to target the lobe and project to most glomeruli. They may then serve as somewhat general guideposts for other sensory neurons that then distribute themselves to specific glomeruli in response to short-range signals. How does such a developmental model incorporate the invariant distribution of odorant receptors among sense organs on the antennal surface and their stereotypic projection to olfactory glomeruli? One possibility is that the Ato-dependent pioneers project specifically to targets of glomerular formation in the lobe primordium. These neurons attract sensory neurons from other lineages that are fated to express the same receptor gene. This, however, begs the question of how groups of pioneers acquire distinct properties to enable their specificity of projection and ability to select discrete populations of neurons. A somewhat similar phenomenon occurs during development of the zebrafish olfactory system where antigenically and clonally distinct pioneer neurons guide other neurons to specific targets within the lobe. These pioneers are not completely analogous to the Ato neurons, as they are transient and do not eventually express odorant receptors (Jhaveri, 2002).

The targeting of pioneer neurons and subsequent patterning of a functional map must involve combinatorial cues that could exist as a pre-pattern in the developing lobe. Identification of molecules that set up such a developmental field is now possible because of identification of a large number of putative cell surface molecules from Drosophila genome analysis. It is suggested that in response to these cues, Ato-dependent neurons not only project correctly but also play an active role in guiding other sensory neurons to defined glomeruli (Jhaveri, 2002).

It has been postulated that neurons from the input and output fields of the olfactory lobe are independently specified and that choice of glomerular targets of the projection neurons is predetermined early in development. In such a scenario, projection neurons should target and arborize correctly in the lobe, even in situations where sensory input is lacking. This is not the case: in ato^– mutants, the interneurons, though present within the lobe, fail to terminate within defined glomerular sites. It is proposed that despite the possible existence of autonomous developmental programs within the interneurons, cues from sensory inputs are still essential to trigger proper patterning (Jhaveri, 2002).

Induction and autoregulation of Bar during retinal neurogenesis

Neurogenesis in the Drosophila eye imaginal disc is controlled by interactions of positive and negative regulatory genes. The basic helix-loop-helix (bHLH) transcription factor Atonal (Ato) plays an essential proneural function in the morphogenetic furrow to induce the formation of R8 founder neurons. Bar homeodomain proteins are required for transcriptional repression of ato in the basal undifferentiated retinal precursor cells to prevent ectopic neurogenesis posterior to the furrow of the eye disc. Thus, precise regulation of Bar expression in the basal undifferentiated cells is crucial for neural patterning in the eye. Evidence is shown that Bar expression in the basal undifferentiated cells is regulated by at least three different pathways, depending on the developmental time and the position in the eye disc. (1) At the time of furrow initiation, Bar expression is induced independent of Ato by Hedgehog (Hh) signaling from the posterior margin of the disc. (2) During furrow progression, Bar expression is also induced by Ato-dependent EGFR (epidermal growth factor receptor) signaling from the migrating furrow. (3) Once initiated, Bar expression can be maintained by positive autoregulation. Therefore, it is proposed that the domain of Bar expression for Ato repression is established and maintained by a combination of non autonomous Hh/EGFR signaling pathways and autoregulation of Bar (Lim, 2004).

To identify activators of Bar expression in the basal undifferentiated cells, focus was placed on two different transcription factors, Lozenge (Lz) and Glass (Gl), as candidates. Both proteins are known to be required for normal Bar expression in R1/6 photoreceptor cells, but it has not been demonstrated whether they are also required for Bar expression in the basal undifferentiated cells. Lz is expressed in R1, 6 and 7 photoreceptor cells and is required for normal level of Bar expression in R1/6 cells. In the basal undifferentiated cells, Lz is co-expressed with Bar in a majority of Bar-expressing cells, except in a group of cells just posterior to the furrow. To test whether Lz is also required for Bar expression in the basal undifferentiated cells, Bar expression was examined in homozygous lz^r15 mutants and loss-of-function (LOF) clones of lz^r15, a null allele of lz. It was found that the expression level of Bar is strongly decreased but not completely eliminated in R1/6 photoreceptor cells within lz^r15 mutant clones. However, Bar expression in the basal undifferentiated cells is little changed compared with its expression level in adjacent wild-type cells. These results suggest that Lz is necessary to activate Bar expression in R1/6 cells, but not in the basal undifferentiated cells behind the furrow (Lim, 2004).

Next, the requirement for Gl was examined in undifferentiated cells. Gl is a zinc-finger protein expressed in all cells posterior to the furrow. Gl is not necessary for Bar expression in the basal undifferentiated cells although it is essential for Bar expression in R1/6 photoreceptor cells. Taken together, these results suggest that Bar expression requires other activators in the basal undifferentiated cells (Lim, 2004).

Based on the evidence presented in this study, a model is proposed for the regulation of Bar expression in the basal undifferentiated cells. Prior to photoreceptor differentiation at the time of furrow initiation, Bar expression in the basal undifferentiated cells near the posterior region of the disc is induced by secreted signaling factors from the posterior margin. Hh signaling from the posterior margin is required for the initial induction of Bar expression. During furrow progression, a narrow region of Bar expression immediately posterior to the furrow depends on Ato from the furrow. EGFR signaling may partially mediate Ato effects on Bar expression. Hh produced by photoreceptor cells generated behind the furrow may also be required in part for Bar expression near the furrow during furrow progression. Finally, Bar is autoregulated to maintain its expression. The properly expressed Bar proteins repress ato transcription in the basal undifferentiated cells, thereby preventing ectopic photoreceptor differentiation posterior to the furrow (Lim, 2004).

Hh expression is dynamic, depending on the time and the position in the developing eye disc. In the early third instar eye disc, Hh is expressed in the posterior margin and is required for the furrow initiation. During furrow progression, Hh is also produced in the differentiating photoreceptor cells generated posterior to the furrow and secreted anteriorly to promote furrow progression. During this process, Bar is specifically expressed in the basal undifferentiated cells posterior to the furrow and inhibits ectopic retinal neurogenesis by repressing proneural gene ato expression (Lim, 2004).

Hh signaling is required for Bar expression in the basal undifferentiated cells during initial eye development because Bar expression is strongly reduced or absent within smo LOF clones generated near the furrow or close to the posterior margin of the disc. Prior to the photoreceptor differentiation, Hh expressed in the posterior margin of the disc is responsible for Bar expression at specific distances from the posterior region of the eye disc proper. A graded expression of Bar near the posterior region in ato¹ mutant eye disc might be the effects of Hh secreted by the posterior margin (Lim, 2004).

During furrow progression, Hh signaling is required for Ato expression in the furrow, and Ato-mediated EGFR signaling is required for Bar activation. Therefore, it is possible that the loss of Bar expression near the furrow in smo LOF clones might be caused by indirect effects of reduced Ato expression rather than by direct effects of Hh signaling on Bar expression. Hh may partially contribute to Bar expression by activating normal levels of Ato expression in the furrow. Thus, the Hh-Ato-EGFR cascade activates Bar expression just posterior to the furrow. Alternatively, since Hh signaling may also affect furrow progression, it is possible that the loss of Bar expression near the furrow in smo LOF clones might be caused by indirect effects of slow furrow migration rather than by direct effects of Hh signaling on Bar expression (Lim, 2004).

The results suggest that Ato is required nonautonomously for the induction of Bar expression just posterior to the migrating furrow. Although Ato acts as an activator for Bar expression, expression of these proteins always show a juxtaposed complementary pattern along the furrow. This suggests that some mediator(s) is required for transducing Ato effects on Bar expression. EGFR activated by Ato in the furrow is required for Bar expression, suggesting that nonautonomous effects of Ato on Bar expression may be partially mediated by EGFR, as revealed by analysis of Egfr negative clones and temperature senstive mutants. Furthermore, EGFR is required for Bar expression not only in the eye disc but also in the antenna and leg discs in Drosophila, suggesting that EGFR signaling may be a common activator for Bar expression in different tissues or even in higher organisms (Lim, 2004).

Notch (N) signaling is also known to contribute to neuronal differentiation together with Hh and Dpp pathways. Thus, N signaling may play a role for Bar expression in the basal undifferentiated cells during furrow progression. Bar expression is strongly downregulated when N function is removed with a temperature-sensitive mutation (N^ts) or using the Enhancer-of-split [E(spl)] mutant clones in the eye disc. This suggests that N signaling may be required for Bar expression in the basal undifferentiated cells. However, it is equally possible that loss of Bar expression in the E(spl) LOF clones or in the N^ts eye disc may be an indirect secondary effect of the lack of the basal undifferentiated cells because nearly all cells in the basal region of the eye disc differentiate into photoreceptor cells without N function. Further analysis of Bar regulation at the molecular level will be helpful to identify direct regulators of Bar expression in the undifferentiated cells of the eye disc (Lim, 2004).

Epithelial cell adhesion in the developing Drosophila retina is regulated by Atonal and the EGF receptor pathway

In the Drosophila retina, photoreceptor differentiation is preceded by significant cell shape rearrangements within and immediately behind the morphogenetic furrow. Groups of cells become clustered into arcs and rosettes in the plane of the epithelium, from which the neurons subsequently emerge. These cell clusters also have differential adhesive properties: adherens junction components are upregulated relative to surrounding cells. Little is known about how these morphological changes are orchestrated and what their relevance is for subsequent neuronal differentiation. This study reports that the transcription factor Atonal and the canonical EGF receptor signalling cascade are both required for this clustering and for the accompanying changes in cellular adhesion. In the absence of either component, no arcs are formed behind the furrow, and all cells show low Armadillo and DE-cadherin levels, although in the case of EGFR pathway mutants, single, presumptive R8 cells with high levels of adherens junction components can be seen. Atonal regulates DE-cadherin transcriptionally, whereas the EGFR pathway, acting through the transcription factor Pointed, exerts its effects on adherens junctions indirectly, at a post-transcriptional level. These observations define a new function for EGFR signalling in eye development and illustrate a mechanism for the control of epithelial morphology by developmental signals (Brown, 2006).

The discovery that EGFR signalling regulates cellular morphology in the morphogenetic furrow adds to the multiple functions already ascribed to this pathway in the eye. The process of cluster formation is the earliest detectable stage of ommatidial development and is tightly coordinated with subsequent photoreceptor recruitment. However, the results presented in this study clearly demonstrate that clustering is a separable process from photoreceptor differentiation. Most directly, the fact that spitz mutant clones do not show defects in clustering, whereas they fail to differentiate any photoreceptors beyond the founding R8 cell, shows that the functions of the EGFR pathway in clustering and in recruitment are distinct. In addition, these roles are spatially and temporally separable. The initial source of the Spitz signal for photoreceptor recruitment is the R8 cell. However, at the time at which the first MAPK activation is seen, in the furrow, the R8 does not yet exist. atonal expressing cells in the furrow upregulate rhomboid levels, and presumably it is these cells that release the activating ligand to control clustering. Since spitz clones do not show aberrant clustering, it is believed that Keren may be the ligand required to activate the EGFR in this process (either alone or redundantly with Spitz), since it has been conjectured that Keren is also involved in the control of ommatidial spacing, cell survival behind the morphogenetic furrow and ommatidial rotation. Evidence is available supporting the role of Keren in several aspects of eye development (Brown, 2006).

There are clearly strong similarities between the function for the EGFR described in this study and its function in the control of ommatidial spacing. In both cases, the pathway is activated in the morphogenetic furrow, under the control of Atonal, and in both, it appears as though Keren, rather than Spitz, may be the principal activating ligand. Indeed, it seems likely that the same signalling event may be responsible for coordinating both these processes. In the case of R8 spacing, it has been hypothesised that activation of the EGFR pathway leads to the secretion of an as yet unidentified inhibitory molecule, which acts non-autonomously to repress atonal expression between proneural clusters. In clustering, the signalling pathway seems to be required for two related purposes: firstly, to regulate the cell shape changes that accompany rosette and arc formation, and secondly, to maintain high levels of AJ proteins in these cells. In contrast to R8 spacing, this function appears to be cell-autonomous—no significant rescue is seen of clustering close to the borders of mutant clones. It is proposed that the same EGFR signalling event that results in the expression and secretion of the spacing inhibitory factor also causes an autonomous change in the transcriptional program of cells and leads to their maintaining strong AJs and to undergoing the cell shape changes that are required for rosette and arc formation (Brown, 2006).

Although the results demonstrate that both Atonal and the EGFR signalling cascade act to control the adhesive and morphological changes of cells behind the morphogenetic furrow, they implicate two independent mechanisms by which they act. Atonal exerts a transcriptional effect upon DE-cadherin. On the contrary, shg-lacZ and arm-lacZ levels are unaffected in ras mutant clones, indicating a function for the pathway in controlling either translation of these components, or in regulating the subcellular distribution or stability of AJs. However, it should be noted that, while the effects upon adhesion proteins are post-transcriptional, the EGFR pathway is acting through its canonical pathway and via the transcription factor Pointed, which must therefore control the expression of some downstream factor (Brown, 2006).

The following course of events is proposed. As cells enter the morphogenetic furrow, they all upregulate their levels of Armadillo and DE-cadherin. This change in the adhesive properties of the cells is independent of either Atonal or the EGFR signalling pathway, and presumably occurs as a result of the earlier signals responsible for driving the progression of the furrow, such as Hedgehog or Dpp. Since shg-lacZ expression does not appear to be upregulated at this early stage, it is speculated that this increase in protein levels is the result of some post-transcriptional mechanism, for example by stabilisation of AJs, although this has not been investigated this further. Behind the furrow, cells not fated to differentiate as photoreceptors downregulate the levels of A