eagle (eg) is expressed in neuroblasts and is involved in the fate determination of serotonergic neurons. Serotonin is an evolutionarily conserved neurotransmitter, found in both invertebrates and vertebrates, and involved in locomotor and behavioral roles. Serotonin is produced in descendents of neuroblast NB 7-3. NB7-3 expresses several genes including engrailed, huckebein, seven-up, pdm1 and eagle. Although eg is expressed in both lateral and medial NB 7-3 derived serotonin cells, the eg loss-of-function mutants often affect the development of only one serotonin cell from each pair. The two cells can be distinguished from one another by differential expression of zinc finger homeobox 2 (zfh-2). This dual domain transcription factor has been shown to bind to and activate the DDC gene (Lundell, 1992 and 1994). The differential expression of zfh-2 and of another gene, pdm-1, can be used to determine that the remaining serotonin positive single cell in eg mutants expresses markers characteristic of the more lateral serotonin cell. In a wild-type CNS, both zfh-2 and pdm1 are selectively expressed in the more lateral serotonin cell but not in the more medial cell. engrailed and eagle are expressed in both these serotonin cells. In eg mutants only the medial cell consistently fails to become a serotonin cell. Therefore, even though eg is normally expressed in both serotonin cells, the absence of Eg protein has a more dramatic effect on the fate of the more medial neuron. This important observation suggests that the lateral serotonin cell can maintain its fate in the absence of Eg (Lundell, 1998).
Analysis of gene expression in eg mutants shows that expression of zfh-2 and en is dependent on eg function but expression of pdm1 is independent of eg function. Loss of eg function appears to have no affect on the expression of pdm1. Clearly, the serotonin cell phenotype in eg mutants is not directly related to the expression of pdm1. Loss of eg function affects the expression of zfh-2 in the lateral serotonin neuron. Loss of eg function affects the expression of en in both serotonin neurons. Thus, eagle is necessary for the maintenance of both engrailed and zfh-2 expression in the serotonin neurons (Lundell, 1998).
The simplest explanation for the difference between the medial and lateral serotonin neurons is that the lateral cell contains a redundant mechanism that allows continued synthesis of serotonin in the absence of Eg protein. This redundant mechanism is not 100% efficient, since not all segments in an eg mutant CNS contain serotonin cells. Since zfh-2 but not pdm1, expression is affected in eg mutants, it is suggested that zfh-1 is a potential factor for this redundant pathway, which establishes eg-independent serotonin synthesis. In an eg-loss-of-function mutant, the loss of Ddc expression is always accompanied by the loss of en expression, but can occur independently in the two serotonin cells. In a hemisegment where both cells fail to express Ddc, neither cell shows en expression. In a hemisegment where only the lateral serotonin cells continues to express Ddc, this lateral cell shows en expression but the medial cell does not. It is concluded that the two serotonin cells have distinctive regulatory networks. In the medial cell, eagle is required for the serotonin fate, while in the lateral cell, engrailed and zhf-1 are required but eagle is not. It is shown that hypomorphic alleles of eagle can produce viable adults that have a dramatic reduction in the number of serotonin-producing neurons (Lundell, 1998).
The overlap between Zfh-2 and Wg throughout the larval stages suggests that zfh-2 may be activated by Wg signalling. In order to test this, the effect of ectopic Wg expression on zfh-2 expression was analyzed. dpp-GAL4 was used to drive the expression of a UAS-wg construct along the A/P boundary in all domains along the P/D axis. Under these conditions, Zfh-2 shows a broad expansion into the presumptive notum region but no ectopic expression in the wing pouch. This indicates that ectopic Wg can activate zfh-2 at a distance from its site of expression (Whitworth, 2003).
In the wild type proximal wing disc, Nub overlaps the Zfh-2 domain at the inner ring of Wg. This situation is recapitulated when ectopic Wg is driven by dpp-GAL4, since ectopic Nub is only detected in regions of high Wg expression. This indicates that ectopic Wg is inducing a response similar to the inner ring and suggests that the region expressing ectopic Zfh-2 is now differentiating as proximal wing (Whitworth, 2003).
To assess whether the cells expressing ectopic Zfh-2 have altered their fate, the expression of the notum marker Tsh was analyzed. Tsh is completely repressed throughout the region of ectopic Zfh-2, indicating that cells are no longer fated as notum. Since Wg is an important factor in the development of wing blade, the expression of a wing blade marker, Vg, was also examined. Vg shows no expansion into more proximal regions and is still restricted to the wing pouch (Whitworth, 2003).
These observations support the idea that Wg is able to direct the differentiation of proximal wing fate at the expense of notum. This can also be inferred from an examination of the phenotype of pharate adults of genotype dpp-GAL4/UAS-wg. An outgrowth of tissue is seen with characteristic proximal wing sclerites and a concomitant loss of macrochaete and scutellum normally associated with the notum (Whitworth, 2003).
This experiment to induce ectopic proximal wing was carried out in such a way that the ectopic Wg was expressed in a pattern that intersects the endogenous domain of Wg expression and results in a continuous region of Wg expression. The observed effects could therefore be interpreted as directed overgrowth of the endogenous proximal wing and not differentiation of proximal wing de novo. This is supported by observations that ectopic expression of Wg in the proximal wing anlagen causes disc overgrowth and consequently overgrows proximal wing tissue. In view of this, attempts were made to reproduce the effects of ectopic Wg in a manner that was discontinuous with the endogenous wg domain. To achieve this, clones of wg-expressing cells were induced that were contained entirely within the notal region, outside of the endogenous proximal wing. To prevent diffusion of ectopic Wg, a construct (UAS-Nrt-flu-wg) was used that directs the expression of a membrane-tethered form of Wg marked with a Flu epitope tag. The colocalization of Nrt-Wg bound to the cell surface and GAL4-expressing cells confirms two things: (1) that the only cells expressing GAL4 induce expression of the UAS construct; (2) that the Flu epitope marker is not detectable beyond the site of expression, indicating that the Flu/Wg hybrid molecule is membrane-bound and not detectably diffusible. When stained to reveal Zfh-2, it can be clearly see that Zfh-2 is induced at a distance of several cell diameters from the site of Nrt- Flu-Wg expression, producing a large zone of Zfh-2-expressing cells surrounded by an epithelial fold. This observation is surprising considering that Wg protein is believed to be tethered to the cell membrane. In the wing pouch, the same construct elicits a Wg signal response only in the expressing cell and its immediate neighbors. Although the nature of the long-range induction cannot be explained at present, this does confirm that ectopically expressed Wg is able to induce the expression of Zfh-2 and therefore drive differentiation of proximal wing fate (Whitworth, 2003).
Taken together, these results show Wg is sufficient to direct the differentiation of proximal wing fate. Furthermore, Wg can only induce ectopic Zfh-2, and thereby proximal wing fate, in the more proximal notum tissue and not in the more distal wing pouch (Whitworth, 2003).
To determine whether Wg is required for zfh-2 expression and how this changes through development, a number of methods were employed to remove Wg function at different developmental stages. The temperature-sensitive allele wgIL114 was used in trans to a wg-lacZ insertion line to create a conditional null mutant. When larvae were moved to the restrictive temperature just prior to L2, Zfh-2 expression was no longer detected in the wing primordium. This indicates that Wg function is required at least for initiation of zfh-2 expression in the L2 wing disc. Wg signal transduction can also be antagonized by the expression of a dominant negative TCF (DN-TCF), a component of the Wg signalling pathway. dpp-GAL4 was used to drive expression of DNTCF along the A/P boundary from early larval stages. zfh-2 fails to be activated in the presence of DN-TCF, even into L3. This further supports the findings that Wg signal transduction is absolutely required for initiation of zfh-2 expression during L2 (Whitworth, 2003).
However, when Wg signalling is removed later in L3, under all experimental conditions tested, no effect on zfh-2 expression is seen. Large wg null clones or the expression of a dominant negative form of wg during L3 shows no detectable reduction in Zfh-2 levels. Similarly, no loss of Zfh-2 is observed with clonal expression of DN-TCF. This shows that, after activation in the L2, Wg activity is no longer required during L3 for the maintenance of zfh-2 expression (Whitworth, 2003).
Taken together, these results show that the regulation of zfh-2 by Wg is temporally dynamic. Although Wg is required early to activate zfh-2, when both are extensively coexpressed, Wg appears not to be required later to maintain zfh-2 expression. This raises the possibility that, once activated, zfh-2 might regulate its own expression by an unknown mechanism. This interpretation would also mean that the downstream response to Wg signal is temporally dynamic, since it appears that one set of genes, e.g., those required to determine proximal wing fate, is activated early and later becomes independent of Wg, and then another set of genes is in turn activated, e.g., those delimiting the wing blade (Whitworth, 2003).
Ectopic expression of Wg can induce zfh-2 only in regions outside of the wing pouch. This suggests that some factor has a repressive effect on zfh-2 in the pouch that cannot be overcome by Wg activation. Genes fundamental to wing blade development may be responsible for this repression. Since Vg expression is restricted to the presumptive wing blade and is required for wing blade development, the effects of ectopic expression of vg on the proximal wing region were examined. Using dpp-GAL4 to direct expression of vg along the A/P boundary represses zfh-2 in the proximal wing region. Endogenous wg expression, monitored with the wg-lacZ reporter, also shows complete repression at the point of intersection. Conversely, in vg1 mutant discs, the Zfh-2 expression domain is expanded into the remnant of the wing pouch and shows a greater overlap with Nub expression than in the wild type. In vg1 discs, much of the wing pouch anlagen fails to develop, and this is accompanied by complete loss of Wg expression at the wing margin; however, the two rings of Wg delimiting the proximal wing are maintained. This suggests that derepression of the zfh-2 domain into the pouch region is not caused by ectopic Wg activity (Whitworth, 2003).
Since the loss of vg does not result in complete derepression of zfh-2, it suggests that another repressor must be acting with vg. Nub is also required for wing blade development. Hypomorphic nub alleles display a severely reduced wing phenotype and a transformation of distal structures into proximal ones. nub2 discs show a complete loss of the inner ring of Wg and an expansion of Wg expression at the wing margin. In nub2 mutant discs, Zfh-2 expression is expanded into the wing pouch, along the line of the wing margin. This indicates two things: (1) that Nub normally acts to repress zfh-2 expression, and thus proximal wing fate, within the wing pouch, and (2) that ectopic zfh-2 is induced where Wg is expressed. Therefore, in an environment of reduced Nub, it can be predicted that ectopic Wg would be able to induce ectopic Zfh-2. To test this, ectopic Wg was expressed in a nub mutant background. As in the nub2 background, Zfh-2 is ectopically induced in the wing pouch along the wing margin . In addition, Zfh-2 can now be detected in the wing pouch along the line of dpp-GAL4, where high levels of Wg are ectopically expressed. This demonstrates that, in an environment of reduced Nub, Zfh-2 expression can be induced wherever Wg is expressed and is no longer restricted from the pouch. It is noted that, whereas Wg expression is expanded at the wing margin in nub discs, where ectopic Wg is induced in a nub background, endogenous Wg is expressed normally at the wing margin; however, the reason for this is unknown (Whitworth, 2003).
In nub discs, vg expression is unaffected, but vg is upregulated by high levels of ectopic Wg. Thus, it appears that the increased levels of Vg are not sufficient to repress Zfh-2 in the absence of Nub when Wg is present at high levels. However, further from the source of ectopic Wg, Zfh-2 is not induced in the nub background, and presumably here, Vg alone can repress Zfh-2. Taken together, these data suggest that zfh-2 expression is regulated by a balance between activation by Wg and repression by a combination of Nub and Vg, acting together or independently. The loss of either Nub or Vg is enough to cause only a partial derepression of zfh-2 in the wing pouch, indicating that alone neither Nub nor Vg is sufficient to completely repress proximal wing fate. However, their combined action, as is the case in the wild type, is able to completely repress zfh-2 expression in the wing pouch. Thus, these factors act to restrict zfh-2 expression to the periphery of the wing disc, thereby defining the distal limit of the proximal wing primordium (Whitworth, 2003).
Recent work has indicated that the homeobox gene homothorax (hth) is required for the correct development of the proximal wing by both upregulating Wg expression in the proximal wing and limiting the area of wing blade differentiation. Since loss of Hth function in the proximal wing leads to a dramatic reduction in the level of Wg expression, attempts were made to determine whether Hth is also required for regulation of Zfh-2 expression. In hth- clones, neither the expression pattern nor the level of Zfh-2 is altered compared with neighboring wild type tissue. This is consistent with the observation that late removal of wg does not affect the expression of zfh-2. Similarly, ectopic expression of Hth shows no effect on zfh-2 expression. These data suggest that Hth does not play a role in establishing or regulating the determination of proximal wing fate, since no change in the expression of Zfh-2 was observed. Thus, it appears that the prime functions of Hth in the proximal wing are to maintain Wg expression and define the limits of the wing pouch (Whitworth, 2003).
A 40-bp upstream regulatory region of the DOPA decarboxylase gene (Ddc), that is important for cell-specific expression in the Drosophila CNS, has been investigated. This region contains two redundant elements which when simultaneously mutated result in lowered DDC expression in serotonin neurons. A protein binding site within one of these elements has been uncovered, and a factor has been cloned that binds to the site. This factor is the product of the zfh-2 gene, a complex homeodomain/zinc finger protein previously identified by binding to an opsin regulatory element. The in vivo profile of Zfh-2 in the larval CNS shows intriguing overlap with DDC in specific serotonin and dopamine neurons. Zfh-2 is related to a human transcription factor ATBF1. The multiple homeodomain and zinc finger motifs in these two proteins show a similar linear arrangement that implies coordinate action among the motifs. In addition, the homology defines a new homeodomain subtype (Lundell, 1992).
The central nervous system contains a wide variety of neuronal subclasses generated by neural progenitors. The achievement of a unique neural fate is the consequence of a sequence of early and increasingly restricted regulatory events, which culminates in the expression of a specific genetic combinatorial code that confers individual characteristics to the differentiated cell. How the earlier regulatory events influence post-mitotic cell fate decisions is beginning to be understood in the Drosophila NB 5-6 lineage. However, it remains unknown to what extent these events operate in other lineages. To better understand this issue, a very highly specific marker was used that identifies a small subset of abdominal cells expressing the Drosophila neuropeptide Capa: the ABCA neurons. The data support the birth of the ABCA neurons from NB 5-3 in a cas temporal window in the abdominal segments A2-A4. Moreover, it was shown that the ABCA neuron has an ABCA-sibling cell which dies by apoptosis. Surprisingly, both cells are also generated in the abdominal segments A5-A7, although they undergo apoptosis before expressing Capa. In addition, a targeted genetic screen was performed to identify players involved in ABCA specification. It was found that the ABCA fate requires zfh2, grain, Grunge and hedgehog genes. Finally, it was shown that the NB 5-3 generates other subtype of Capa-expressing cells (SECAs) in the third suboesophageal segment, which are born during a pdm/cas temporal window, and have different genetic requirements for their specification (Gabilondo, 2011).
The findings strongly suggest that the Capaergic abdominal ABCA neuron arises from NB 5-3. This conclusion is based on the expression in ABCA cells of gsb, wg and unpg, and the absence of the markers lbe(K) and hkb. However, even though gsb expression is known to be maintained specifically in the lineage of rows 5 and 6 NBs, whether expression of the other genetic markers used to identify NBs at stage 11 changes late in embryogenesis remains unanswered. Nonetheless, the specific combination of NB markers found in ABCA cells and their position in the hemineuromere are consistent with their birth from NB 5-3. Previous accounts showed that this NB gives rise to a lineage of 9–15 cells. Additionally, observations derived from studies in which PCD was blocked showed that NB 5-3 can potentially produce a large lineage (ranging from 19 to 27 cells), suggesting that it could generate 13 or 14 GMCs. The lack of a NB 5-3 specific-lineage marker prevented resolution of its complete lineage, and thus determining the birth order of the ABCA cell (Gabilondo, 2011).
Recent findings on the NB 5-6 and NB 5-5 demonstrate that cas and grh act together as critical temporal genes to specify peptidergic cell fates at the end of these lineages. cas mutants lack ABCA cells and Cas is expressed in these cells, while the normal pattern of ABCA cells is found in grh mutants, and Grh is not present in ABCA neurons. These data strongly support the birth of ABCA cells in a cas-only temporal window. This is different from the suboesophageal Capaergic SECA cells, which while also arising from NB 5-3, show a reduction in cell number in both pdm and cas mutants, demonstrating birth at a mixed pdm/cas temporal window. Previous studies in other lineages have shown that when a temporal gene is mis-expressed, all progeny cells posterior to that temporal window can be transformed to the specific fate born at that particular temporal window. However, cas mis-expression failed at inducing ectopic ABCA cells, suggesting that cas in necessary but not sufficient to specify the ABCA fate (Gabilondo, 2011).
Programmed cell death (PCD) is a basic process in normal development. The results suggest that the ABCA and its sibling are equivalent cells committed to achieve the ABCA fate. First, it was shown that the ABCA-sibling cell dies by apoptosis, but produces an ABCA-like Capaergic neuron if PCD is inhibited. Second, when PCD is blocked, NB 5-3 also produces a GMC generating two ABCA-like Capaergic cells in the A5–A7 segments. These data indicate that a segment-specific mechanism prevents death of the ABCA cells in A2–A4 neuromeres. Segment specific cell death has been previously reported for the NB 5-3 lineage, and detailed studies on segment-specific apoptosis of other lineages have shown that this process is under homeotic control. In addition, the results show a different timing in the PCD undergone by the ABCA sibling and the ABCA cells born in A5–A7. This interpretation is based on the differential effect of p35 expression when cas-Gal4 or elav-Gal4 drivers were used. Although elav-Gal4 is transiently express in NBs and GMCs, robust and maintained driver expression commences in differentiating neurons. In contrast, cas expression starts in the NB and is maintained in the GMC and neuronal progeny. Therefore, the finding that death of the ABCA-sibling cell can be prevented by directing p35 with cas-Gal4, but not with elav-Gal4, suggests that the death of the ABCA sibling occurs earlier in development than the death undergone by the ABCA cells in A5–A7 segments (Gabilondo, 2011).
In the ABLK/LK peptidergic fate (derived from the NB 5-5), activation of Notch (N) signaling in the peptidergic cell prevents its death, while its sibling, NOFF cell undergoes apoptosis. On the contrary, in the EW3/Crz peptidergic fate (derived from the NB 7-3), silencing of N signaling is essential for the neuron survival, and therefore for it proper specification. The current results are in accordance with the last scenario, in which the ABCA cell is NOFF, and it sibling, which undergoes apoptosis, is NON. Therefore, Notch signaling must be switch off for the proper specification of the ABCA neuron (Gabilondo, 2011).
To search for genes involved in specification of the ABCA neural fate, a reduced set of mutants was screened of genes that are expressed in the embryonic CNS at stage 11, a time at which distinctly defined sublineages are being generated from all active NBs. Even though this method will certainly overlook important genes, the results reveal that it is in fact a very effective way to find genes involved in specification of a particular neural fate. Indeed, the ratio of success has been very satisfactory: 33.3% of the genes analyzed display a significant phenotype. Moreover, the set of identified genes could be further expanded by, for example, searching in interactome databases, and performing the subsequent screen on those putative interactors (Gabilondo, 2011).
It is assumed that the specification of a concrete cellular fate requires the combination of several transcription factors, namely a genetic combinatorial code. Recently, a detailed combinatorial code has been reported for three neuropeptidergic fates: ap4/FMRFa, ap1/Nplp1 and ABLK/Lk. However, very little is known about the specification of the rest of the 30 peptidergical fates. This study has identified several genes involved in the specification of the ABCA fate, which fit into three categories. First, genes were found whose loss-of-function produces a relevant increase of the number of ABCA cells. Most remarkable are the klu and rn phenotypes, which consist of duplications of the ABCA cells. These phenotypes suggest that these two transcription factors repress the ABCA fate in other neural cells (or/and NBs/GMCs). Interestingly, the normal phenotype of nab mutants indicates that, contrary to its mode of action in the wing, Rn does not work with the transcription cofactor Nab in this context (Gabilondo, 2011).
Second, genes were found whose loss-of-function produces a significant decrease of the number of ABCA neurons. In this category, the zhf2, ftz and grain phenotypes stand out. The effects of mutations on ftz are in agreement with its early role in segmentation: ftz is a pair-rule segmentation gene that defines even-numbered parasegments in the early embryo, and absence of ABCA cells was found in the A3 segment in ftz mutants. However, zfh2 and grain seem to be part of the specific combinatorial code of the ABCA cells. The Drosophila GATA transcription factor Grain has been reported to be involved in the specification of other cell fates, such as the aCC motoneuron fate. Based on its expression, the zinc finger homeodomain protein zfh2 has been proposed to mediate specification of the serotoninergic fate, but this has not been further demonstrated. Interestingly, during wing formation, zfh2 is required for establishing proximo-distal domains in the wing disc, and it does so partly by repressing gene activation by Rn. The opposite phenotypes that was observed in rn and zfh2 mutants suggest that similar interactions occur during ABCA specification. Analyses aimed to test this hypothesis are currently being performed (Gabilondo, 2011).
Third, two genes were found whose loss-of-function abolishes the ABCA fate: Grunge and hh. Grunge encodes a member of the Atrophin family of transcriptional co-repressors that plays multiple roles during Drosophila development. Taken together, studies from C. elegans to mammals suggest that Atrophin proteins function as transcriptional co-repressors that shuttle between nucleus and cytoplasm to transduce extracellular signals, and that they are part of a complex gene regulatory network that governs cell fate in various developmental contexts. Similarly, Hh is an extracellular signaling molecule essential for the proper patterning and development of tissues in metazoan organisms. It is noteworthy that two genes implicated in extracellular signaling pathways, Grunge and hh, are absolutely required for ABCA fate. Further studies will be needed to identify at which step/s they exert their actions, and to unravel possible interactions between them and with other players of the combinatorial code for ABCA specification (Gabilondo, 2011).
Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.