BIOLOGICAL OVERVIEW

Drosophila has four types of sensory elements: external sense organs, multiple dendritic neurons, chordotonal neurons, and photoreceptors. The latter two do not require the proneural genes of the achaete-scute complex (AS-C) in order to develop and function properly, but they do need atonal, a neurogenic gene that behaves in many ways like the genes of AS-C. atonal gets its name from the disruptive effects the gene's mutation has on chordotonal neuron differentiation. Mutants are completely difficient in this lateral sense organ (Jarman, 1995).

Early in embryonic development, atonal is expressed in all chordotonal organ progenitor cells, but its later expression is restricted to only a particular set of precursors, through a process of lateral inhibition. Notch does not seem to be required for this process in chordotonal organs, and the mechanism is not well understood. Mutation in atonal also disrupts eye development (Jarman, 1995).

Atonal protein is produced in all cells just anterior to the morphogenic furrow. Expression is downstream of hedgehog, an important gene involved in regulating the progression of the furrow. As the furrow progresses, most cells that express atonal in front of the furrow lose that capacity. atonal expression behind the furrow is confined to the R8 progenitors, whose fate atonal determines. atonal is produced within the context of the furrow, which is only rudimentary in atonal mutants. atonal is a neurogenic gene, functioning in the place of achaete-scute complex bHLH genes, both in the eye and in chordotonal organs.

The function of Atonal is best illustrated by its role in chordotonal organ development. A scolopidium, the basic unit of chordotonal organs, consists of four cells: a neuron with a single dendrite, the scolopale cell, cap cell and ligament cell. The scolopale cell (a glial cell) forms a sheath around the dendrite, while the cap cell and ligament cell mediate the attachment of the chordotonal organ to the body cell. Expression of atonal is restricted to a subset of atonal-requiring chordotonal precursors, called founder precursors. In atonal mutants, all chordotonal organs are absent except for one scolopidium of Ich5, the abdominal pentascolopidial organ (Ich5 consists of five scolopidia organized in a linear array). This one scolopidium formed in atonal mutants is atonal independent. The atonal independent precursor corresponds to the earliest chordotonal precursor (precursor C1) and corresponds to the P cell which gives rise to the anterior-most scolopidium of Ich5 (zur Lage, 1997).

EGF receptor signaling is required in neural recruitment during formation of Drosophila chordotonal sense organ clusters. A total of five neural precursors express atonal in abdominal segments, and this number is too few to explain the formation of the eight scolopidia in each abdominal segment. Of the five precursors, C-1, C2 and C3 contribute to Ich5, C4 migrates slightly anterodorsally and gives rise to v'ch1, the dorsal most scolopidium, which is solitary. C5 contributes to vchAB, a more ventral pair of scolopidia. The remaining precursors require Egf-R signaling for their selection. Signaling by the founder precursors is initiated by atonal activating (directly or indirectly) rhomboid expression in the founder cells. It should be noted that in some developmental processes, rhomboid appears to function in the signal-receiving cells, such as in the patterning of ovarian follicle cells. It is not believed that this is the case in chordotonal-precursor formation, because rho is expressed in precursors that do not require rhomboid function (C1-C5 are formed even in rhomboid mutants). Signaling by these founder precursors, presumably through the EGF receptor ligand Spitz, then provokes a response in the surrounding ectodermal cells, as shown by the activation of expression of the Egf-R target genes pointed and argos. The signal and response then leads to recruitment of some of the ectodermal cells to the chordotonal precursor cell fate. Egf-R hyperactivation by misexpression of rhomboid results in excessive chordotonal precursor recruitment. Argos functions in a feedback mechanism to prevent the excess recruitment of additional ectodermal cells. The increase in the number of scolopidia caused by Egf-R hyperactivation is confined to an enlargement of existing cluster sizes: no new chordotonal clusters are formed. A two step mechanism is postulated for the formation of clusters of chordotonal precursors. In the first step, precursors C1-C5 are selected as founder precursors by the conventional route of proneural gene expression and lateral inhibition. In a distinct second phase, these precursors then signal to adjacent ectodermal cells via the Egf-r pathway, inducing some of them to become chordotonal precursors (secondary or recruited precursors). This two-step process is strongly reminiscent of the way atonal acts in neurogenesis in compound eyes. Here, atonal expression is initially refined by lateral inhibition, until atonal is expressed in only the founding R8 precursor, which then recruits R1-R7 in a mechanism that does not require the activation of atonal in these cells (zur Lage, 1997).

The selection of Drosophila sense organ precursors (SOPs) for sensory bristles is a progressive process: each neural equivalence group is transiently defined by the expression of proneural genes (proneural cluster), and neural fate is refined to single cells by Notch-Delta lateral inhibitory signalling between the cells. Unlike sensory bristles, SOPs of chordotonal (stretch receptor) sense organs are tightly clustered. It has been shown that for one large adult chordotonal SOP array (the adult femoral chordotonal sense organ), clustering results from the progressive accumulation of a large number of SOPs from a persistent proneural cluster. This is achieved by a novel interplay of inductive epidermal growth factor- receptor (EGFR) and competitive Notch signals. EGFR acts in opposition to Notch signaling in two ways: it promotes continuous SOP recruitment despite lateral inhibition, and it attenuates the effect of lateral inhibition on the proneural cluster equivalence group, thus maintaining the persistent proneural cluster. SOP recruitment is reiterative because the inductive signal comes from previously recruited SOPs (zur Lage, 1999).

The adult femoral chordotonal sense organ arises from a group of some 70-80 SOPs. A developmental analysis of Ato expression has revealed that these SOPs accumulate over an extended period of time in the dorsal region of each leg imaginal disc during the third larval instar and early pupa. The continued expression of Ato implies a sustained requirement for proneural function throughout the process of SOP accumulation. Unusually, Ato is persistently expressed in a group of ectodermal cells identified as the proneural cluster (PNC). From this PNC, cells are funnelled inward into a cavity formed by the folding of the disc. This invagination later becomes visible as a distinctive 2-cell wide intrusion, which is referred to as the 'stalk'. Cells at the deepest end of the stalk undergo shape changes to form an amorphous inner SOP mass. Invaginating cells are characterised by upregulation of Ato expression, a characteristic of SOP commitment. Surprisingly, SOP markers (Ase protein and the A101 enhancer trap line) are not expressed in all the stalk SOPs. Instead, these markers are only apparent in older cells, particularly at the time when they become part of the inner mass (which is therefore referred to as mature SOPs). Despite this, entry into the stalk seems to mark SOP commitment, since both the stalk and the mature SOPs are absent in discs from ato mutant larvae. This apparent intermediate stage may not have a counterpart in external sense organ precursor formation, although there is some evidence for multiple steps between the uncommitted cell and the SOP (the so-called pre-sensory mother cell state). Initially, Ato remains activated in all invaginated SOPs. This extended period of proneural gene expression is unusual since AS-C proneural expression is typically switched off in SOPs shortly after commitment. Later, at approximately 6 hours before puparium formation (BPF), Ato expression is switched off synchronously in the mature SOPs, although expression remains in the stalk SOPs and the PNC. At this point there is very little overlap between Ato and Ase or A101 (zur Lage, 1999).

The process of chordotonal SOP formation described above is at odds in several respects with the well-known paradigm of SOP selection for sensory bristles. In the latter, the solitary SOP expresses Delta, which triggers expression in the PNC of genes of the E(spl)-C, thereby preventing further SOP commitment and forcing loss of AS-C expression and neural competence. In the case of the femoral chordotonal organ, newly committed cells from the PNC are in contact with previously committed SOPs in the stalk, but are apparently not receiving (or not responding to) lateral inhibition signals from these to prevent their commitment. Likewise, the presence of committed SOPs does not switch off ato expression in the PNC. Nevertheless, components of the N-Dl pathway are expressed in patterns consistent with lateral inhibition. The newly formed SOPs express Dl, suggesting that they send inhibitory signals, while the PNC expresses mgamma, a member of the E(spl)-C, suggesting that these cells are responding to the Notch-Delta signal. Indeed, mgamma is coexpressed with ato in the PNC throughout the development of the SOP cluster. Chordotonal SOP formation is shown to be sensitive to N inhibitory signaling. Strong activation of N signaling or its effectors can inhibit chordotonal SOP formation. Thus, N signaling has an important role to play: it acts to limit the process of SOP selection from the PNC. Some mechanism, however, must prevent N signaling from completely inhibiting multiple SOP formation (zur Lage, 1999).

The progressive accumulation of chordotonal SOPs suggests that a recruitment mechanism could explain the clustering of SOPs. The Drosophila Egfr signaling pathway is involved in a number of recruitment processes in development, and a role for Egfr signaling has been demonstrated in the induction of embryonic chordotonal precursors (zur Lage, 1997). Although there appear to be significant differences in the process of SOP formation in imaginal discs, as compared with the embryo, it was asked whether Egfr signaling is also involved in forming the femoral chordotonal cluster. To address this question, the pathway was conditionally disrupted by expressing a dominant negative form of Egfr protein. Expression of UAS-Egfr DN results in a dramatic loss of chordotonal SOPs in late third instar imaginal leg discs (as judged by Ase protein expression or the A101 enhancer trap line). This demonstrates that Egfr signaling is required for the process of femoral chordotonal SOP formation. In contrast, the appearance of bristle SOPs is unaffected, arguing against the possibility of a nonspecific effect on SOPs in general (zur Lage, 1999).

To determine whether Egfr signaling controls SOP number, expression of components of the Egfr pathway that determine the level of signaling was forced, thus resulting in hyperactivation of the pathway. pointed (pnt) is an effector gene that encodes a transcription factor and is activated in cells responding to Egfr signaling. Both rho and pnt are expressed during chordotonal SOP formation. Indeed, forced expression of rho or pnt increases chordotonal SOP formation. Egfr could promote SOP formation by stimulating the commitment of PNC cells or by stimulating proliferation of SOPs. Both functions would be consistent with known Egfr roles, but the current investigations favour the former. Analysis of Ato expression in leg discs in which rho has been misexpressed reveals a large invagination of cells and a smaller PNC. Shrinking of the PNC was confirmed by the reduced extent of mgamma expression. These observations are consistent with an increased rate of SOP commitment upon Egfr hyperactivation. Moreover, this effect is reminiscent of the effect of N loss of function on Ato expression, suggesting that Egfr signaling supplies the mechanism that interferes with lateral inhibition of SOP commitment (zur Lage, 1999).

Although it seems that cells of the PNC and stalk are held in a state of mitotic quiescence throughout the time that SOP fate decisions are being made, BrdU is incorporated in the older (mature) SOPs. The experiments so far have indicated that Egfr signaling affects SOP commitment from the PNC. To determine more precisely the spatial patterning of Egfr activity required for SOP clustering and N antagonism, the expression patterns of key components of the pathway were characterized. Localized expression of rho appears to play a central role in spatial restriction of Egfr activity in cases where Spi is the ligand; in these cases it appears to mark the cells that are a source of signaling. During development of the femoral chordotonal organ, rho is expressed in a very restricted pattern: RHO mRNA is only detected in the SOPs, becoming confined in the late third instar larva to the youngest SOPs at the top of the stalk. To identify the cells responding to rho-effected signaling, an antibody that detects the dual-phosphorylated (activated) form of the ERK MAP kinase (dp-ERK) was used. In leg imaginal discs, dp-ERK is detected in a confined area corresponding to the uppermost (youngest) stalk SOPs. Thus, like rho, dp-ERK is expressed in the newly formed stalk SOPs. Double labelling for RHO RNA and dp-ERK confirms this, but also suggests that the overlap in expression is not complete: dp-ERK is detected above the uppermost rho-expressing cells of the stalk, probably in one or a few cells of the proneural cluster as they funnel into the stalk. This suggests that Egfr promotes SOP commitment as a consequence of direct signaling from previous SOPs to overlying PNC cells. Since rho expression is itself activated upon SOP commitment, this process occurs cyclically: the newly recruited SOPs are in turn able to signal to further overlying PNC cells. That is, recruitment is reiterative. Egfr signaling via Spitz has been shown to help to maintain neural competence by attenuation of Notch directed lateral inhibition. The opposing forces of Notch and Egfr signaling are thought to be played out through direct Notch and Egfr signaling between the epidermal proneural cells, which bear Notch, and the SOP, which sends inhibitory signals through the Delta ligand, and stimulatory signals through the Spitz ligand (zur Lage, 1999).

Reiterative recruitment alone cannot entirely explain the accumulation of SOPs. Such an accumulation also relies on the persistence of the competent pool of PNC cells from which SOPs can be recruited. For AS-C PNCs, this does not occur, because the mutual inhibition required for continued competence is unstable and resolves quickly to a state of lateral inhibition once the SOP emerges from the PNC. This results in rapid shutdown of AS-C expression and hence competence within the PNC. It is possible that the members of E(spl)-C that are expressed in the PNC (notably mgamma and mdelta) are less aggressive inhibitors of proneural gene expression than the E(spl)-C members expressed in AS-C PNCs (m5 and m8). The results obtained in the femoral SOP suggest, however, that Egfr has a role to play in maintaining the PNC by partially attenuating lateral inhibition on a PNC-wide scale. Thus, the PNC is not completely shut off by inhibition from SOPs, but instead kept in check, allowing continued mutual inhibition and maintenance of competence but not allowing general SOP commitment. Since neither rho nor dp-ERK are detected in the PNC as a whole, this function of Egfr could be indirect and achieved through partial attenuation of Dl signaling from the stalk SOPs themselves. The trans- or auto-activation of EGFR signaling between the stalk SOPs (as suggested by the co-expression of dp-ERK and rho) might be an indicator of this function. It is also possible, however, that Egfr signaling is direct and that the dp-ERK antibody is not sensitive enough to detect expression in the PNC cells (zur Lage, 1999).

Specificity of Atonal and Scute bHLH factors: analysis of cognate E box binding sites and the influence of Senseless

The question of how proneural bHLH transcription factors recognize and regulate their target genes is still relatively poorly understood. It has been shown that Scute (Sc) and Atonal (Ato) target genes have different cognate E box motifs, suggesting that specific DNA interactions contribute to differences in their target gene specificity. This study shows that Sc and Ato proteins (in combination with Daughterless) can activate reporter gene expression via their cognate E boxes in a non-neuronal cell culture system, suggesting that the proteins have strong intrinsic abilities to recognize different E box motifs in the absence of specialized cofactors. Functional comparison of E boxes from several target genes and site-directed mutagenesis of E box motifs suggests that specificity and activity require further sequence elements flanking both sides of the previously identified E box motifs. Moreover, the proneural cofactor, Senseless, can augment the function of Sc and Ato on their cognate E boxes and therefore may contribute to proneural specificity (Powell, 2008).

The proneural proteins exhibit very precise specificity in activation of different neurogenesis programmes. It has been suggested that utilization of different E-box motifs as binding sites may partly underlie this specificity. This was based on the finding that E boxes from Sc- and Ato-specific target genes conform to different consensus motifs. This study found further support for observation. In a cell culture assay, artificial enhancers of concatemers of E_Ato or E_Sc sequences generally show specific activation by Ato or Sc proteins, respectively. Nevertheless, the results also show that E box activity and specificity depends on complex features of the DNA surrounding the proneural-specific motifs both in cell culture and in vivo. The task of predicting by sequence analysis how proneural proteins regulate targets remains formidable (Powell, 2008).

Transcription factor activity depends on a complex interplay of interactions with DNA and with other protein factors, including those bound to other sites within the enhancer. To concentrate on the role that proneural protein interaction with E-box binding sites plays in specificity, synthetic enhancers of concatemers of E-box-containing sequences were studied in a cell culture reporter gene assay. A previous study of Ato or Sc-specific enhancers relied on the analysis of expression patterns produced in transgenic flies carrying GFP reporter gene constructs. In that study, specific regulation by Sc or Ato was inferred indirectly from patterns of GFP expression. This study showed that much of this inferred specificity is also seen in a cell culture reporter gene assay, strongly supporting the conclusion that Ato and Sc directly use different E box motifs. Thus, in general, the specificity of E box response (ratio of response to Sc and Ato) could be predicted from matches to E_Sc or E_Ato motifs identified previously. In most cases, this specificity also corresponded to the specificity of the native enhancer from which the E box was taken. An interesting exception is sens-E1: while this E box is proposed to respond to both Ato and Sc in vivo, it responds slightly better to Sc than to Ato in culture, which is more consistent with its E_Sc motif. It will be important to determine what other enhancer features allow such an E box to function as a common target of Ato and Sc in vivo (Powell, 2008).

Importantly, E box specificity is achieved without the appropriate cellular and developmental context of neurogenesis: S2 cells are embryonic, non-neural cells of likely hematopoietic origin and are not expected to contain neural-specific factors. The results therefore indicate that proneural factors have intrinsic ability to use different E box motifs without the need for interactions with neural specific cofactors. The E_Sc and E_Ato motifs differ most notably in the bases immediately flanking the 5' end of the 6-bp core sequence (NG vs. AW). There is evidence from the crystal structure of the MyoD bHLH domain–DNA complex that protein contacts are made with bases in this position, suggesting that similar direct contacts may influence E box utilization by proneural proteins. The basic region amino acids making these contacts (R₁₁₀, R₁₁₇ and E₁₁₈) are conserved in the proneural proteins, but in Ato the arginines are separated by three amino acids (LAA, equivalent to MyoD KAA) that are absent in Sc. Thus despite the apparent conservation of DNA-contacting residues, one might predict strong differences in how the proneural proteins interact with the flanking nucleotides. SPR analysis shows Ato/Da to bind to ato-E1 and sc-E1 with similar affinity. Rather than affecting E box affinity, it is possible that subtle differences in binding contacts may cause conformational effects that affect the transactivation ability of the proneural protein (Powell, 2008).

The above results point to the importance of distinct Ato and Sc E box motifs for proneural specificity. Several findings, however, demonstrate that these motifs are heavily dependent on the wider DNA context. For instance, the E(spl)mγ-E2(C₄ > G), sens-E1 and sc-E1 binding sites show very large differences in activity in cell culture, even though they have identical perfect E_Sc motifs at their core (gCAGGTGt). The effect of DNA context is also seen in the general inability, in the cell culture assay, to swap the proneural specificities of sc-E1 and ato-E1 by mutating the immediate 5' flanking bases of the core E box. Such changes generally result in loss of E box activity rather than a clear change in specificity. These results indicate that the E_Sc and E_Ato motifs are generally not sufficient for activity or specificity in the cell culture assay and that the surrounding DNA context is important (even within the short 20-bp sequences used) (Powell, 2008).

Interestingly, in some circumstances specificity could be manipulated more successfully in vivo: (sc-E1 GG > AA)₆-GFP transgenic flies showed GFP expression consistent with strongly reduced activation by Sc and a gain of activation in some specific locations by Ato. However, this mutated motif did not respond to ectopically expressed Ato, perhaps suggesting that improved specificity in vivo results from cofactors expressed in locations of endogenous Ato expression and function (Powell, 2008).

Overall, the results above show that further sequences on both flanks of the E_Sc and E_Ato box motifs are also important for specificity and activity. One possibility is that the 20-bp DNA sequences used to construct the concatemers may include flanking sequences that interact with other protein factors to influence proneural specificity. Such adjacent sites have been identified for some mouse proneural E box binding sites. Moreover, in its native enhancer, ato-E1 is adjacent to an Ets-domain transcription factor binding site (although this site is mutated in the constructs used in this study). However, such cofactors would need to be expressed in S2 cells. Moreover, although the flanking sequences of the ato-E1 and sc-E1 sites are strongly conserved among Drosophila species, no obvious shared sequence motifs were found in the 5' and 3' flanks of known Drosophila E boxes that might be cofactor binding sites. Whilst there is a potential POU factor binding sequence 5' of the ato-E1 site, no members of the Drosophila POU family appear to be expressed during early neurogenesis. Alternatively, the further flanking bases may influence bHLH heterodimer interaction either through direct contacts or through indirect conformational effects. It is interesting that 3' bases appear important as these may be predicted to affect Da interaction. It is notable that the Da homologue, E2A, has different half-site preferences when bound to Twi or MyoD (Powell, 2008).

The specificity of E-box concatemer constructs is generally more complete in vivo than in the S2 luciferase assay -- notably proneural proteins can generally activate non-cognate E boxes to some extent in cell culture but not in vivo. One possibility is that the intrinsic specificity of proneural proteins must normally be enhanced by interaction with cofactors that are not present in S2 cells. In the cell culture assay, at high proneural levels it was found that Sens enhanced proneural activity in a general manner. None of the constructs tested contain Sens binding motifs, so it is likely that enhancement occurs in a DNA-binding independent manner via protein–protein interactions. At low proneural concentrations, however, the effect of Sens enhancement becomes selective. For many of the constructs tested, Sens only enhanced the activity of proneural proteins for concatemers consisting of their cognate E box. It is suggested that proneural–Sens interaction may enhance the specificity of proneural–E box interaction. Thus, this is an interesting case in which proneural specificity can be influenced by a common cofactor, rather than requiring interaction with different subtype-specific cofactors. It remains to be determined whether Sens would enhance specificity on native enhancers as well as concatemer constructs. Moreover, it seems unlikely that Sens is a specificity factor for all proneural target genes. However, the results are consistent with Sens acting as a specificity cofactor in certain contexts -- such as the proneural autoregulatory enhancers active in SOPs where there are high levels of Sens and proneural proteins present. Other non-DNA binding proneural protein interactors, such as Chip may have a similar effect in other contexts (Powell, 2008).

The effect of Sens could be explained by two models. First, interaction of a proneural protein with a specific E-box motif may give rise to a specific conformation which results in an increased affinity for Sens protein. Alternatively, the Sens–proneural protein interaction may alter the proneural bHLH domain conformation thereby increasing its affinity for its cognate binding site (i.e., an induced fit model). Indeed, variation in MyoD bHLH protein DNA sequence preferences have been previously observed to be the result of effects on basic region conformation arising because of binding partner differences or amino acid composition of the basic region. In this view, proneural specificity relies on a combination of cognate DNA sequence recognition and protein–protein interactions. Important future work will be the identification of the amino acid residues of Ato and Sc necessary for their interaction with Sens and the determination of whether these influence DNA recognition (Powell, 2008).

Patterning mechanisms diversify neuroepithelial domains in the Drosophila optic placode

The central nervous system develops from monolayered neuroepithelial sheets. In a first step patterning mechanisms subdivide the seemingly uniform epithelia into domains allowing an increase of neuronal diversity in a tightly controlled spatial and temporal manner. In Drosophila, neuroepithelial patterning of the embryonic optic placode gives rise to the larval eye primordium, consisting of two photoreceptor (PR) precursor types (primary and secondary), as well as the optic lobe primordium, which during larval and pupal stages develops into the prominent optic ganglia. This study characterize a genetic network that regulates the balance between larval eye and optic lobe precursors, as well as between primary and secondary PR precursors. In a first step the proneural factor Atonal (Ato) specifies larval eye precursors, while the orphan nuclear receptor Tailless (Tll) is crucial for the specification of optic lobe precursors. The Hedgehog and Notch signaling pathways act upstream of Ato and Tll to coordinate neural precursor specification in a timely manner. The correct spatial placement of the boundary between Ato and Tll in turn is required to control the precise number of primary and secondary PR precursors. In a second step, Notch signaling also controls a binary cell fate decision, thus, acts at the top of a cascade of transcription factor interactions to define photoreceptor subtype identity. This model serves as an example of how combinatorial action of cell extrinsic and cell intrinsic factors control neural tissue patterning (Mishra, 2018).

In the fruit fly Drosophila melanogaster, all parts of the visual system develop from an optic placode, which forms in the dorsolateral region of the embryonic head ectoderm. During embryogenesis, neuroepithelial cells of the optic placode are patterned to form two subdomains. The ventroposterior domain gives rise to the primordium of the larval eye and consists of two photoreceptor (PR) precursor types (primary and secondary precursors), whereas the dorsal domain harbors neuroepithelial precursors that generate the optic lobe of the adult visual system. The basic helix-loop-helix transcription factor Atonal (Ato) promotes PR precursor cell fate in the larval eye primordium. The orphan nuclear receptor Tailless (Tll) is confined to the optic lobe primordium and maintains non-PR cell fate. Hedgehog (Hh) and Notch (N) signaling are critical during the early phase of optic lobe patterning. The secreted Hh protein is required for the specification of various neuronal and non-neuronal cell types, while Notch acts as neurogenic factor preventing ectodermal cells from becoming neuronal precursors by a process termed lateral inhibition. In the optic placode Ato expression is promoted by Hh and the retinal determination genes sine oculis (so) and eyes absent (eya). Notch delimits the number of PR precursors and maintains a pool of non-PR precursors. Ato is initially expressed in all PR precursors in the placode and its expression gets progressively restricted to primary precursors. In a second step, primary precursors recruit secondary precursors via EGFR signaling: primary precursors express the EGFR ligand Spitz, which is required in secondary precursors to promote their survival. After this initial specification of primary and secondary PR precursors, the transcription factors Senseless (Sens), Spalt (Sal), Seven-up (Svp) and Orthodenticle (Otd) coordinate PR subtype specification. Sens and Spalt are expressed in primary PR precursors, while Svp contributes to the differentiation of secondary PR precursors. By the end of embryogenesis, primary PR precursors have fully differentiated into blue-tuned Rhodopsin5 PRs (Rh5), while secondary PR precursors have differentiated into green-tuned Rhodopsin6 PRs (Rh6). While the functional genetic interactions of transcription factors controlling PR subtype specification has been thoroughly studied, it remains unknown how the placode is initially patterned by the interplay of Hh and Notch signaling pathways. Similarly, the mechanisms of how ato and tll-expressing domains are set up to ensure the correct number of primary and secondary PR precursors as well as non-PR precursors of the optic lobe primordium remain unknown (Mishra, 2018).

This study describes the genetic mechanism of neuroepithelial patterning and acquisition of PR versus non-PR cell fate in the embryonic optic placode and provide the link to subsequent PR subtype identity specification. The non-overlapping expression patterns of ato and tll in the optic placode specifically mark domains giving rise to the larval eye precursors (marked by Ato) and the optic lobe primordium (marked by Tll). ato expression in the larval eye primordium is temporally dynamic and can be subdivided into an early ato expression domain, including all presumptive PR precursors and a late ato domain, restricted to presumptive primary PR precursors. The ato expression domain directly forms a boundary adjacent to tll expressing precursors of the optic lobe primordium. tll is both necessary and sufficient to delimit primary PR precursors by regulating ato expression. Hh signaling regulates the cell number in the optic placode and controls PR subtype specification in an ato- and sens-dependent manner. Finally, this study also shows that Notch has two temporally distinct roles in larval eye development. Initially, Notch represses ato expression by promoting tll expression and later, Notch controls the binary cell fate decision of primary versus secondary PR precursors by repressing sens expression. In summary, this study has identified a network of genetic interactions between cell-intrinsic and cell-extrinsic developmental cues patterning neuroepithelial cells of the optic placode and ensuring the timely specification of neuronal subtypes during development (Mishra, 2018).

Neurogenic placodes are transient structures that are formed by epithelial thickenings of the embryonic ectoderm and give rise to most neurons and other components of the sensory nervous system. In vertebrates, cranial placodes form essential components of the sensory organs and generate neuronal diversity in the peripheral nervous system. How neuronal diversity is generated varies from system to system, and different gene regulatory networks have been proposed for each particular type of neuron. Interestingly, some transcription factors, like Atonal, play an evolutionary conserved role during neurogenesis both in Drosophila and in vertebrates (Mishra, 2018).

Neuroepithelial patterning of the Drosophila optic placode exhibits unique segregation of larval eye and optic lobe precursors during embryogenesis. This study has identified genetic mechanisms that control early and late steps in specifying PR versus non-PR cell fate that ensure the expression of precursor cell fate determinants. During germband extension at stage 10, transcriptional regulators (so, eya, ato and tll) show complex and partially overlapping expression patterns in the optic placode. Their interactions with the Notch and Hh signaling pathways define distinct PR and non-PR domains of the larval eye and optic lobe primordium. Intriguingly, the results show a spatial organization of distinct precursor domains, supporting a new model of how the subdivision of precursor domains emerges. In agreement with previous studies initially the entire posterior ventral tip expresses Ato, defining the population of cells that give rise to PR precursors, while neuroepithelial precursors for the presumptive optic lobe are defined by Tll-expression in the anterior domain of the optic placode. Subsequently, Ato expression ceases in the ventral most cells and thus gets restricted to about four primary PR precursors that are located directly adjacent to the Tll expression domain. Hence, a few cell rows are between the primary PR precursors and the ventral most edge of the optic placode. This is in agreement with a recent observation on the transcriptional regulation of ato during larval eye formation. Thus, primary PR precursors are directly adjacent to the Tll-expressing cells while the Ato and Tll negative domain of secondary PR precursors is located at the posterior ventral most tip of the optic placode. Setting the Tll-Ato boundary is critical to define the number of putative secondary PR precursors, which can be recruited into the larval eye, probably via EGFR signaling. A model is proposed during which coordinated action of Hh, Notch and Tll restricts the initially broad expression of Ato to primary PR precursors (see Ato to primary PR precursors). Lack of Tll results in a de-repression of Ato and results in an increased number of primary PR precursors, which in turn recruit secondary PR precursors. Interestingly, while tll mutants show an increase in both primary and secondary PR precursors, the ratio between both subtypes is maintained. This notion further displays similarities of ommatidal formation in the adult eye-antennal imaginal disc, where Ato expressing R8-precursors recruit R1-R6. In the eye-antennal disc, specification of R8-precursors determines the total number of ommatida and therefore also the total number of PRs, the ratio of R8 to outer PRs however always remains the same. Thus, the initial specification of primary PR precursors defines the total number of PRs in the larval eye similarly to R8 PRs, and the ratio of founder versus recruited cells remains constant. Interestingly, the maintenance of primary versus secondary PR precursor ratio is also maintained in ptc mutants further supporting this model (Mishra, 2018).

During photoreceptor development in the eye-antennal imaginal disc hh is expressed in the posterior margin and is required for the initiation and progression of the morphogenetic furrow as well as the regulation of ato expression. During embryogenesis the loss of hh results in a complete loss of the larval eye, while increasing Hh signaling (by means of mutating ptc) generates supernumerary PRs in the larval eye. During early stages, an increase of Ato expression was found in ptc mutants suggesting that similarly to the eye-antennal disc Hh positively regulates ato expression. The observed increase of Ato-expressing cells is not due to a reduction of Tll but is likely due to increased cell proliferation in ptc mutants. Hh also controls proliferation during the formation of the Drosophila compound eye (Mishra, 2018).

During embryonic nervous system development Notch dependent lateral inhibition selects individual neuroectodermal cells to become neuroblasts. Notch represses neuroblast cell fate and promotes ectodermal cell fate. During compound eye development, Notch regulates Ato expression and acts through lateral inhibition to select Ato expressing R8 PR precursors. Similarly, during Drosophila larval eye development, Notch is required for regulating PR cell number by maintaining epithelial cell fate of the optic lobe primordium. Inhibiting Notch signaling leads to a complete transformation of the optic placode to PRs of the larval eye. In the absence of Notch signaling, Ato expression is expanded in the optic placode and as a result the total number of PRs is increased. Despite the increase of the overall PR-number the number of secondary PR precursors is significantly decreased or lost in the absence of Notch activity. In the compound eye Notch promotes R7 cell fate by repressing the R8-specific transcription factor Sens. It was also proposed that genetic interaction between Notch and Sens is required for sensory organ precursor (SOP) selection in the proneural field in a spatio-temporal manner. This study found that during PR subtype specification Notch represses Sens expression, thereby controlling the binary cell fate decision of primary versus secondary PR precursors. Therefore, in the absence of Notch signaling, Sens expression represses the secondary PR precursor fate. As a result, all PR precursors are transformed and acquire primary PR precursor identity. In conclusion, this study observed that Notch is essential for two aspects during optic placode patterning. First, Notch activity is critical for balancing neuroepithelial versus PR cell fate mediated through Tll-regulated Ato expression. Second, Notch regulates the binary cell fate decision of primary versus secondary PR precursor cell fate through the regulation of Sens expression (Mishra, 2018).

A temporal transcriptional switch governs stem cell division, neuronal numbers, and maintenance of differentiation

The importance of producing the correct numbers of neurons during development is illustrated by both evolutionary enhancement of cognitive capacities in larger brains, and developmental disorders of brain size. In humans, increased neuronal numbers during development is speculated to partly derive from a unique subtype of neural stem cells (NSCs) that undergo a phase of expansion through symmetric self-amplifying divisions before generating neurons. Symmetric amplification also appears to underlie adult neural stem maintenance in the mouse. However, the mechanisms regulating this behavior are unclear. This study reports the discovery of self-amplifying NSCs in Drosophila and shows that they arise by a spatiotemporal conversion of classical self-renewing NSCs. This conversion is regulated by a temporal transition in the expression of proneural transcription factors prior to cell division. A causal link was found between stem cell self-amplification and increased neuronal numbers. It was further shown that the temporal transcriptional switch controls both stem cell division and subsequent neuronal differentiation (Mora, 2018).

The development of functional organs relies on the coordinated production of cells of different identities with temporal, spatial, and numerical precision. In the brain, where information processing depends on the output of interconnected neuronal circuits, not only the ratios of different neuronal subtypes, but also absolute numbers are important for optimal function. The number of neurons in the adult brain is a direct consequence of a spatiotemporally coordinated sequence of divisions of neural stem cells (NSCs) during development. However, it remains unclear how NSCs alter their division patterns over time and whether these alterations are causal to the generation of the correct number of neurons. Less clear still is whether and how the temporal transitions in NSC division influence the differentiation of their progeny (Mora, 2018).

In both mammals and insects NSCs regulate neurogenesis through a series of self-renewing divisions. NSC division patterns can be broadly classified in five categories. In three of these, NSCs divide asymmetrically renewing themselves and giving rise to daughters that differ in their proliferation potentials: daughters that do not divide, daughters that divide once, and daughters that divide multiple times. In the other two, NSCs divide symmetrically. One type of symmetric division common to vertebrates and invertebrates signals the end of stemness through the generation of two daughter cells committed to differentiation. A second, much rarer type, expands the progenitor pool through the generation of two cells, which retain the expression of NSC markers and the ability to generate neurons. In mouse, self-renewal by symmetric division has recently been reported to be predominant during adult neurogenesis, in contrast to what is observed in embryonic stages where most NSCs divide asymmetrically. In the primate brain, embryonic self-amplifying divisions have been detected in the NSCs known as outer radial glia (oRG). Multiple lines of evidence support the hypothesis that oRGs' high abundance and proliferative capacity are critical for the vast increase of brain size in primates. However, the direct evidence for the impact of symmetric amplification of NSCs on neuronal numbers, the mechanisms that mediate the switch from self-renewal to self-amplification and then to neurogenesis, and the impact of such a switch on terminal differentiation remain unexplored (Mora, 2018).

The fruit fly Drosophila melanogaster has long been a powerful model system for the discovery of the genetic, cellular, and molecular underpinnings of the behavior of NSCs, as well as the generation and differentiation of their neuronal progeny. Drosophila NSCs are called neuroblasts (Nbs), and two major modes of neurogenesis have been described. The type I Nbs self-renew while giving rise to committed daughters called ganglion mother cells (GMCs) that in turn divide terminally to produce two neurons or glia. The type II Nbs also self-renew but produce intermediate progenitors that in turn undergo a limited number of self-renewing divisions giving rise to GMCs, which give rise to neurons. Thus, to date, all Nbs in the fly brain are thought to produce neurons by asymmetric self-renewal and no symmetrically dividing, self-amplifying, NSCs have been found (Mora, 2018).

The majority of the fly brain is dedicated to visual processing. The higher-order visual centers called the optic lobes (OLs) receive the visual input from the retina and are arranged in four neuropils called lamina, medulla, lobula, and lobula plate (LP); all four organized in retinotopic maps. OL neurons derive from two major proliferation zones, called the outer proliferation center and the inner proliferation center (IPC), containing actively dividing Nbs. The organization of the OLs is constrained by the characteristics of the compound eye, which is composed of 1~750 repetitive units of 8 photoreceptors covering the visual field and projecting to the OL in a retinotopic order. This integration of the retinal map requires a tight control of the diversity and stoichiometry of the neuronal populations. While temporal and spatial cues required to generate different types of neurons have been identified, the control of the production of large numbers of neurons is much less understood. One striking example is the motion detection neurons of the LP called T4/T5. For each of the 750 retinal units, the LP contains 8 different T4/T5 direction sensitive neurons (T4a, b, c, d, and T5a, b, c, d, respectively). Thus, the direction-selective T4/T5 lineage generates approximately 12,000 neurons, representing more than 10% of all neurons in the fly brain. How such a massive proportion of neurons is generated is entirely unknown (Mora, 2018).

Another highly conserved feature of neurogenesis is that it is regulated by a small and highly conserved set of transcription factors known as the proneural proteins. First described in Drosophila, basic loop-helix proneural factors regulate neurogenesis in insects as well as in mammals. There are three families of proneural proteins named after their founding members; the Atonal (ATH), Achaete-scute (AS), and Neurogenin families. Proneural proteins most conserved function is to provide progenitors with the neuronal fate. In addition, they have been found to promote asymmetric division, exit from cell cycle and initiation of differentiation. Whether proneural proteins can promote symmetric proliferation or if they can combine their proliferation and differentiation functions in the same neuronal linage is still unclear (Mora, 2018).

This study has identified the first symmetrically self-amplifying NSCs in Drosophila giving rise to the population of T4/T5 neurons. These Nbs are generated by a temporal conversion of asymmetrically dividing Nbs, which is accompanied by a temporal transition in proneural protein expression from the AS protein Asense (Ase) to the ATH protein Atonal (Ato). Furthermore, it was discovered that the switch from Ase to Ato is necessary and sufficient for the switch in stem cell division pattern and the generation of the correct number of neurons. Lastly, it was demonstrated that Atonal creates a quantitative change in target gene expression that is propagated throughout the lineage to ensure the commitment of T4/T5 neurons to terminal differentiation (Mora, 2018).

This is the first example of transient amplification by symmetric division of NSCs in a non-mammalian animal, namely the Drosophila fruit fly. These cells are termed type III Nbs to distinguish them from previously described Nb types 0, I, and II. Embryonic NSC symmetric expansion is common in mammals, especially in gyrencephalic mammals, where oRG are highly abundant. This includes ferrets, non-human primates, and humans, but not rodents. oRG is thought to be in part responsible for the brain size expansion that is observed in these species. Having simpler models that recapitulate at least some aspects of oRG biology could be particularly relevant to the study of fundamental questions surrounding the control of brain size. Understanding the symmetry of self-renewal is also relevant for the study of adult neurogenesis where symmetric division have recently been shown to be predominant. In this context, limited rounds of symmetric self-renewal and consuming symmetric differentiation division can explain how neurogenesis is sustained for extended periods of time. This work finds that Drosophila symmetrically amplifying Nbs expand the progenitor pool while at the same time scheduling the future terminal differentiation of their progeny. This study describes genetic and regulatory control mechanisms of these features and the consequences of interfering with such mechanisms for brain development (Mora, 2018).

The type III Drosophila Nbs described in this study are located in the visual system anlagen in a region known as the IPC, where they generate two different neuronal populations: the C2, C3, T2, and T3 neurons (C/T neurons) and the T4/T5 neurons, in that specific temporal order. IPC Nbs transit through two distinct types of proliferation: an earlier phase of type I asymmetric divisions to generate C/T neurons and a later phase of symmetric transit amplification. These two phases coincide with a change in neuronal fate and number. While late born T4/T5 constitute one of the largest lineages in the Drosophila brain, the early born C/T neurons are two times less abundant. Although it is difficult to know the exact number of symmetric divisions each upper-Nb undergoes, it is interesting to note that one symmetric amplification before the terminal production of GMCs would account for the doubling of the number of upper-Nbs compared with lower-Nbs that were observed, resulting in exactly four T4/T5 neurons per upper-Nb. A concurrent study (Pinto-Teixeira, 2018) proposes that this particular stoichiometry may be accounted for through a single terminal Nb division. While the current observations do not contradict the stoichiometry, the suggestion that there is no amplifying step prior to the terminal division is difficult to reconcile with the multiple lines of evidence presented in this study. Together with a study by Apitz (2018), these authors further show that the layer specificity of T4/T5 neurons relies on Dpp and that the T4 versus T5 fate is Notch dependent. The Apitz study further shows how the Dpp signal is maintained from NE progenitors to neurons through a temporal relay mechanism. Together, these studies open the door for understanding precisely how this very large and complex lineage combines numerical expansion, cell fate, and layer-specific targeting over a series of successive temporal developmental transitions (Zhang, 2018).

This study focuses on the temporal transition of proliferation properties and shows that they are regulated by the serial expression of two proneural proteins, Ase and Ato. Interestingly Ase and Ato had not been involved in C/T versus T4/T5 fate decision, suggesting that lineage size can be controlled independently of cell fate. Previous studies in the IPC have shown that the switch in neuronal fate depends on another temporal series of two factors called Tailless and Dichaete. It would be interesting to investigate the crosstalk between these two temporal series as a model to further understand how neuronal numbers and neuronal fate are integrated during development. The current findings provide one of the first examples of Nbs changing their proliferation properties to achieve lineage size proportions, where NSC amplification is causally linked to an increase in the number of neurons generated (Mora, 2018).

Drosophila, Ato acts as a transcriptional activator regulating the commitment of different subsets of epithelial cells to the neuronal fate. However, in the IPC Nbs, Ato plays a dual role. On the one hand, it promotes the amplification of progenitors that express it, and on the other hand it ensures the terminal differentiation of their neuronal progeny. Curiously, Atoh1, the mammalian homolog of Ato, has been described both as a tumor suppressor in colorectal cancer and as an oncogene in medulloblastoma, the most common malignant brain tumor in children. It is suspected that this context-dependent function may be related to the dual role of Ato in amplification and differentiation characterized in this study (Mora, 2018).

It is important to note that, in the IPC, Ato can robustly impose symmetric division when ectopically expressed. However, only a fraction of Nbs divisions are affected in its absence. This demonstrates that Ato is sufficient, but not always necessary, for symmetric division, and suggests the existence of an overlapping and independent mechanism controlling the process. Ato in this context likely acts to ensure robust transitions first to symmetric amplification and later to differentiation. It is proposed that the strong reduction of T4/T5 neuron numbers in ato mutant brains is due to an incomplete transition from asymmetric to symmetric division. However, the effect of other functions of Ato yet to be characterized, for example in ensuring neuronal survival, cannot be excluded (Mora, 2018).

The fact that Ato expression in Nbs controls the differentiation of T4/T5 neurons is demonstrated by the ectopic expression of Nbs markers and the global downregulation of differentiation genes in neurons of ato mutant animals. This resembles the de-differentiation phenotype previously found in Drosophila mutants of longitudinal lacking (lola). However, unlike lola, Ato itself is never expressed in neurons, not even transiently. It is proposed that a stable cellular memory of differentiation is initiated transcriptionally in stem cells and inherited through successive cell divisions to ensure terminal differentiation of neuronal progeny. What the mechanisms of such a memory are, how they are activated in stem cells, and how they relate to stem cell division mode are exciting questions for future investigation (Mora, 2018).

A recurring observation throughout this analyses is that quantitative, rather than all or nothing, changes in gene expression downstream of Ato control the temporal progression of developmental events. For example, premature Ato expression causes a relatively modest reduction in Ase expression, and yet suffices to induce symmetric division prematurely. Similarly, quantitative regulation of Brat levels is required for a dose-dependent maintenance of terminal differentiation in postmitotic neurons. Brat is a member of a family of evolutionarily conserved tumor suppressor proteins that regulate differentiation and growth. In type I and II Nbs, Brat is asymmetrically inherited to promote differentiation. In IPC Nbs, Brat is symmetrically inherited during the transient amplification but it does not prevent Nb gene expression. It is therefore proposed that it is the progressive accumulation through temporal quantitative regulation, rather than its expression per se, that schedules the onset and maintenance of differentiation (Mora, 2018).

How cell division and differentiation are coordinated to determine organ size is a fundamentally important but poorly understood process. In Drosophila, the intrinsic activity life time of given proneural transcription factor is both a developmental and evolutionary strategy for the control of cell number in the peripheral nervous system. During the development of mammalian telencephalon, the expression of Ascl1, the mammalian homolog of Drosophila Achaete-scute proteins such as Ase, oscillates in NSCs. These oscillations promote proliferation, while sustained expression of Ascl1 promotes neuronal differentiation. Finally, there is evidence that spatiotemporal transitions in cross-regulatory transcription factors control root meristem growth in plants. This study shows that a similar logic regulates brain size. These observations suggest that the differential, temporally restricted and quantitative regulation of transcription factors and their target genes may serve a universal role as molecular clocks underlying the coordinated temporal order of developmental events (Mora, 2018).

GENE STRUCTURE

date revised: 3 MAR 97 

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

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