The organisational principles of locomotor networks are less well understood than those of many sensory systems, where in-growing axon terminals form a central map of peripheral characteristics. Using the neuromuscular system of the Drosophila embryo as a model and retrograde tracing and genetic methods, principles underlying the organisation of the motor system have been uncovered. Dendritic arbors of motor neurons, rather than their cell bodies, are partitioned into domains to form a myotopic map, which represents centrally the distribution of body wall muscles peripherally. While muscles are segmental, the myotopic map is parasegmental in organisation. It forms by an active process of dendritic growth independent of the presence of target muscles, proper differentiation of glial cells, or (in its initial partitioning) competitive interactions between adjacent dendritic domains. The arrangement of motor neuron dendrites into a myotopic map represents a first layer of organisation in the motor system. This is likely to be mirrored, at least in part, by endings of higher-order neurons from central pattern-generating circuits, which converge onto the motor neuron dendrites. These findings will greatly simplify the task of understanding how a locomotor system is assembled. These results suggest that the cues that organise the myotopic map may be laid down early in development as the embryo subdivides into parasegmental units (Landgraf, 2003).

The analysis began by correlating the positions of motor neuron dendrites with the distribution of their muscle targets in the periphery. Motor neurons were retrogradely labelled in a pairwise fashion and the positions of their dendritic arbors were mapped. Because of an interest in the mechanisms that underlie the assembly of the motor system, focus was placed on stages when each motor neuron first establishes a characteristic domain of arborisation within the neuropile (early stage 17, 15h after egg-laying [AEL]) (Landgraf, 2003).

Motor axons project into the muscle field via two main nerves, the intersegmental (ISN) and the segmental nerve (SN). The transverse nerve (TN) runs along the segment border and has few motor axons. Choice of nerve root is one of several features that divide the motor neurons into two principal sets, the ISN and SN. (1) The cell bodies of SN motor neurons are located in the same segment as the muscles that they innervate, whereas ISN motor neuron somata are located in the segment next anterior (with the exception of the RP2 and two neuromodulatory efferent ventral unpaired median [VUM] neurons. (2) ISN motor neurons innervate internal muscles, which span a segment from anterior to posterior, whereas SN (and the TN) motor neurons innervate external muscles. External muscles are distinct from the internal set in several respects: (1) they are generally transverse; (2) unlike internal muscles, they require wingless (wg) signalling for their specification; (3) external (but not internal) muscles and their innervating motor neurons express the cell adhesion molecule (CAM) Connectin, with the single exception of muscle ventral transverse 1 (VT1) (Landgraf, 2003 and references therein).

In addition, ISN and SN motor neurons elaborate their dendrites in distinct regions of the neuropile. Dendrites of ISN motor neurons occupy a domain extending posteriorly from the posterior part of one neuromere into the anterior part of the next. SN motor neuron dendrites occupy a domain that lies between the domains of ISN motor neuron arbors (Landgraf, 2003).

Thus, the organisation of the body wall muscles into internal and external sets is reflected centrally in patterns of motor neuron arborisations. The innervating motor neurons project their axons through different nerves and elaborate their dendritic fields in distinct regions of the neuropile. Although dendritic arbors become progressively more elaborate and extensive over developmental time, their separate domains remain clearly recognisable and appear to be maintained at least until the motor system is fully functional (18 h AEL) (Landgraf, 2003).

Having established that there is a central representation of the muscle field, the organisation of the motor neuron dendrites was analyzed in greater detail. (1) The set of external muscles and their innervating (SN) motor neurons were examined. Muscles of similar anteroposterior positions, such as the ventral acute muscle (VA3) and the segment border muscle (SBM), are innervated by motor neurons whose dendritic arbors lie in a common region of the neuropile. Conversely, motor neurons supplying the anterior (lateral transverse 1-2 [LT1-LT2]) versus the posterior (SBM) muscles have dendritic arbors that are correspondingly separated in the anteroposterior axis of the CNS (Landgraf, 2003).

To put the idea of a regular map to the test, focus was placed on an unusual external motor neuron-muscle pair. Muscle VT1 is innervated by a TN rather than an SN motor neuron. However, VT1 lies at the same place in the anteroposterior axis as the SBM, although VT1 is ventral and the SBM more dorsal. The VT1 motor neuron dendritic field is found to overlaps with that of the SBM motor neuron. For the external set, it is concluded that differences in target muscle location in the anteroposterior axis are mapped centrally as regular differences in dendritic position, but dorsoventral distinctions are not (Landgraf, 2003).

It was next asked whether there is a similarly regular representation of the internal muscles in the developing CNS. While most external muscles are transverse and have unique anteroposterior locations, the internal muscles span the width of a segment so that positional distinctions between them are solely in the dorsoventral axis. It was found that the set of internal muscles is represented centrally by three dendritic domains. Motor neurons innervating ventral internal muscles elaborate their dendritic arbors in the anterior half of the ISN dendritic domain. Motor neurons with dorsolateral internal muscle targets (lateral longitudinal [LL] 1, dorsal acute [DA] 3, dorsal oblique 3-5 [DO3-DO5]) put their arbors into the posterior part of the ISN dendritic domain. Finally, dorsal muscles are represented by a motor neuron dendritic domain that lies between those representing ventral (anterior) and dorsolateral (posterior) internal muscle groups. Thus, the internal muscles are represented in the neuropile by three domains of dendritic arborisation that reflect their different dorsoventral locations in the periphery. Once again, it is concluded that there is a regular mapping of muscle position in the neuropile: in this case, it is positions in the dorsoventral axis peripherally that are represented centrally as differences in the anteroposterior locations of dendrites (Landgraf, 2003).

To test the idea that dendritic arbor positions relate to the distribution of muscles, an atypical motor neuron-muscle pair was examined. The RP2 motor neuron is reported to innervate dorsal muscle DA2, yet its dendrites span the domains that represent both dorsal and dorsolateral internal muscles. However, on careful analysis it was found that DA2 is, in fact, specifically innervated by a U neuron whose dendrites lie in the dorsal internal domain, whereas the RP2 axon forms endings generally on all dorsolateral and dorsal muscles by 19 h AEL. These seem to correspond to the type 1s boutons found in late larvae. Thus, the RP2 neuron puts its dendrites into a region of the neuropile that does indeed represent its targets, namely the dorsolateral and dorsal internal muscles (Landgraf, 2003).

Like the muscle field itself, the map of motor neuron dendrites is metamerically repeated. However, the boundaries of these two units are out of register with one another, since the dendrites of the motor neurons innervating internal muscles lie in the next anterior neuromere. The anterior border of the dendritic map, as defined by the extent of these anterior dendrites, coincides with the anterior margin of engrailed (en) expression. Thus, while the muscles are segmental in their organisation, the domains occupied by the dendrites of their innervating motor neurons are parasegmental (Landgraf, 2003).

To test whether genes that implement the parasegmental pattern in the epidermis are also required for the formation of the parasegmental organisation of the neuromuscular system, the formation of SN and ISN dendritic fields was studied in embryos singly mutant for the following segment polarity genes: en/invected (Df(en^E)), wg (wg^CX4), naked (nkd²), patched (ptc⁹), hedgehog (hh²¹), and gooseberry (Df2R(gsb)). Every one of the six different mutants that were analysed has partially aberrant patterns of neuroblasts (NBs). Nevertheless, SN and ISN motor neurons still form and can be identified by their characteristic axonal projections into the periphery. In addition, it was found that the fundamental separation between SN and ISN dendritic domains is present despite often severe perturbations in CNS structure. For example, in gsb mutant embryos, both nerve roots are frequently fused so that the SN and ISN share a common CNS exit point. Nevertheless, SN and ISN axons as well as their dendritic fields do not intermingle but remain separate. These results suggest that the subdivision of the neuropile into the principal ISN and SN dendritic domains is a robust feature of the system, which appears to be specified early in development, since the embryo subdivides into parasegmental units (Landgraf, 2003).

It was next asked what mechanisms underlie the formation of the myotopic map. Because ISN and SN motor neurons lie at different positions in the CNS and their axons grow out into the muscle field through different nerves, it is reasonable to suppose that at least the major subdivision of dendritic arborisations into internal and external domains could be a byproduct of the locations at which the motor neurons are generated and the paths taken by their growing axons. This ‘passive mapping' explanation can be excluded by considering a single motor neuron-muscle pair, namely dorsal transverse 1 (DT1) and its innervating motor neuron. DT1 is an external muscle (by position, orientation, wg dependence, and Connectin expression), yet its motor neuron is clustered with the internal muscle innervating set and its axon (uniquely for the external muscles) grows out through the ISN. Despite its packing within the ‘internal motor neuron' set, the DT1 motor neuron makes a long posterior projection through the internal muscle domain of the myotopic map to reach the external domain, where it arborises appropriately, reflecting the orientation and external nature of its target muscle. In contrast, motor neurons derived from the same NB as DT1 innervate neighboring internal muscles DO3-DO5 and put their dendrites in a more anterior region characteristic of the dorsolateral muscles. These findings strongly suggest that the mapping of the muscle field within the CNS is an active process of growth and arborisation that partitions dendrites into subdomains of the neuropile that are appropriate to their function, rather than a passive subdivision of available space by position of origin or axon trajectory (Landgraf, 2003).

Since dendritic arbors form after motor axons have reached their targets, the muscles could be instrumental in dictating the organisation of the central map. To test this idea, the UAS/GAL4 system was used to misexpress an activated form of Notch (Kidd et al. 1998) in the developing mesoderm, suppressing the formation of muscle founder cells while leaving other tissues intact. In such muscleless embryos, the main nerve trunks, SN and ISN, still form and project into the periphery. Retrograde labellings of these nerves show that SN and ISN motor neurons form relatively normal dendritic arbors that consistently conform to the characteristic separation of SN and ISN dendrites. Thus, the neuropile is partitioned into distinct fields of dendritic arborisation independently of the muscles. It is concluded that the mapping process is likely to be an autonomous property of the motor neurons and their neighboring cells (Landgraf, 2003).

It was next asked whether motor neuron dendritic fields could be patterned by the substrates on which they grow. In the Drosophila ventral nerve cord (VNC), motor neuron dendrites form in the dorsal-most region of the neuropile, sandwiched between longitudinal glia above and the underlying scaffold of axons. Glial cells can act as substrates for supporting and guiding axonal growth. To test whether they might also be required for the growth and spatial patterning of dendritic fields, dendritic arbors were analysed in glial cells missing (gcm) mutant embryos, which are defective in glial cell differentiation. Although the structure of the nervous system is disrupted in gcm mutant embryos and the dendritic arbors are abnormal, they continue to form in their characteristic locations and the fundamental distinction between the ISN and SN motor neuron dendritic fields is maintained. Remarkably, even the long posterior dendritic projection of the DT1 motor neuron forms and reaches its target region, the SN external muscle dendritic domain. These results suggest that the patterning of the neuropile into distinct motor neuron dendritic domains is a process that appears to be intrinsic to the motor neurons and their neighboring neurons, but does not require proper glial cell differentiation (Landgraf, 2003).

One likely explanation for the division of dendrites into separate domains is that there is a process of mutual exclusion between the arborisations of neighboring cells. Such a process of dendritic ‘tiling' has so far only been documented between particular classes of sensory neurons, but could also occur in the motor system. The idea of tiling was tested by considering two groups of motor neurons whose axons have a common trajectory, but whose dendritic fields form in adjacent territories. The DO3-DO5 and DT1 motor neurons project their dendrites posteriorly, and at their most-anterior point, these dendrites meet the axons and dendrites of the anterior corner cell (aCC) and U/CQ neurons. To show whether the aCC and U/CQ axons and/or dendrites inhibit the growth of DO3-DO5 and DT1 dendrites anteriorly, these neurons (as well as RP2 and the posterior corner cell [pCC] interneuron) were selectively ablated. Using anti-Even-skipped (Eve) staining as a marker for aCC, RP2, and U/CQs (there are an additional two medially located eve-expressing interneurons, pCC and friend of pCC [fpCC], it was found that these neurons can be selectively ablated before they form dendrites (at approximately 11 h AEL): on average, by 10.5 h AEL all but 0.6 and by 12 h AEL all but 0.06 of the seven medially located eve-expressing neurons have been ablated per half-neuromere. In no instance was a concomitant anterior expansion of the DO3-DO5 and DT1 motor neuron dendrites into the regions vacated by the aCC and U/CQ dendrites observed. It is concluded that, at least in this instance, the initial dendritic territory of one set of motor neurons (DO3-DO5 and DT1) is not defined by a process of tiling, in which they are excluded by neighboring (aCC and U/CQ) dendritic arbors. However, it is possible that the elaboration of motor neuron dendritic arbors during later developmental stages may involve interactions between neighboring dendritic territories, activity-dependent processes, or both (Landgraf, 2003).

Thus, in summary, these results suggest that the mechanisms that subdivide the neuropile into distinct dendritc domains are very robust and refractory to perturbations. They further suggest that the cues that organise the map may be laid down early in development as the embryo subdivides into parasegmental units (Landgraf, 2003).

Motoneurons are key points of convergence within motor networks, acting as the 'output channels' that directly control sets of muscles to maintain posture and generate movement. This study used genetic mosaic techniques to reveal the origins and architecture of the leg motoneurons of Drosophila. A small number of leg motoneurons are born in the embryo but most are generated during larval life. These postembryonic leg motoneurons are produced by five neuroblasts per hemineuromere, and each lineage generates stereotyped lineage-specific projection patterns. Two of these postembryonic neuroblasts generate solely motoneurons that are the bulk of the leg motoneurons. Within the largest lineage, lineage 15, distinct birth-order differences are seen in projection patterns. A comparison of the central projections of leg motoneurons and the muscles they innervate reveals a stereotyped architecture and the existence of a myotopic map. Timeline analysis of axonal outgrowth reveals that leg motoneurons reach their sites of terminal arborization in the leg at the time when their dendrites are elaborating their subtype-specific shapes. These findings provide a comprehensive description of the origin, development, and architecture of leg motoneurons that will aid future studies exploring the link between the assembly and organization of connectivity within the leg motor system of Drosophila (Brierley, 2012).

In insects that undergo a complete metamorphosis, like Drosophila, the ventral nerve cord is produced by two distinct phases of neurogenesis. The first wave occurs during embryonic development and produces the components required for the control of larval behavior. Some of the neurons generated at this time remodel and play a role in adult circuits. The bulk of neurons found in the adult fly are produced during the second, more prolonged neurogenic phase during larval and early pupal life. Within this process, this study determined whether the leg motoneurons are produced during the embryonic or postembryonic phases of neurogenesis (Brierley, 2012).

This work identified two distinct types of motoneuron clones in the third instar larva VNC, generated by embryonic heatshocks. One type has complex, highly branched dendrites with axons that exit the nerve cord and terminate on body wall muscles. This type of neuron in insects is uniquely identifiable and can have one of two different fates during metamorphosis; some remodel and take up a new adult-specific role, whereas others undergo programmed cell death. It was not possible to determine the identity, number, or fate of specific embryonic neurons using their larval morphology alone. However, it is known that in the beetle Tenebrio molitor and the moth Manduca sexta the larval leg motoneurons remodel to become adult leg motoneurons. The second type of motoneuron clones had a simple morphology in the third instar CNS reminiscent of single-cell postembryonic clones born during larval life. This second type of neuron is similar to the flight motoneuron MN5 and the persistent Broad positive neurons seen in the embryonic CNS. These neurons remain in an immature state throughout larval life, before completing their development during metamorphosis (Brierley, 2012).

To quantify how many types of leg motoneuron are born during the embryonic wave of neurogenesis, MARCM clones were induced in the embryo, and then single-cell motoneuron clones were identified in the adult ventral nervous system with axons in the leg. These data suggest that at least seven different leg motoneuron types are born in the embryo. Using this approach their origins or how many of each type there are could not be determined. These seven types could represent progeny from seven different neuroblasts. A previous study, Baek (2009), predicted that 13 lineages generate leg motoneurons in the embryo, but their data, like the current, cannot definitively answer this question. Although it is believed that these different neurons are bona fide types, it is possible that a single motoneuron could generate two very different terminal arborizations. This seems unlikely, as all the data points toward a high degree of morphological stereotypy in embryonic motoneurons. In future the availability of many more cell type-specific markers should enable identification each of these early born neurons (Brierley, 2012).

The data reveal that neurons born during larval life make the most significant contribution to the pool of leg motoneurons. These leg motoneurons are generated by five postembryonic Nbs. Of these, two lineages generate exclusively motoneurons: lineage 15, which contains on average 28 motoneurons, and lineage 24 which contains six motoneurons. This confirms the observations of Baek (2009), who also found these lineages. Three postembryonic lineages were found that contain one or two motoneurons along with a large number of interneurons (lineages 20, 21, and 22), whereas Baek (2009) only reported one. The motoneurons within these lineages are born soon after the onset of postembryonic neurogenesis, with the first ganglion mother cell (GMC) generating two siblings, a motoneuron and a local interneuron. Following this, every time a GMC divides the motoneuron sibling undergoes apoptosis, whereas the interneuron survives. Such hemilineage-based programs of cell death play a significant role in determining the type and number of network components in the thoracic nervous system of Drosophila (Brierley, 2012).

Most knowledge of the origins of Drosophila motoneurons comes from studies in the embryo. Clonal analysis in the embryo revealed that 17 of the 31 Nbs generate motoneurons and all are born early within these lineages, with most Nbs contributing one or two motoneurons, and at most six. If every embryonic born neuron is derived from a different Nb then the maximum number of lineages generating leg motoneurons would be 12, compared with six if all are derived from a single Nb. Baek (2009) concluded that 13 lineages contribute leg motoneurons (Brierley, 2012).

The general organization of Drosophila leg motoneurons within the CNS shows great similarity with that of the grasshopper Schistocerca americana, with the neurons being clustered into groups. Each of these eight groups are presumably derived from their own single Nb, with the primary neurites inserting into characteristic position in the neuropil (Brierley, 2012).

The success of holometabolous insects as a group is due largely to their ability to produce radically different body plans at larval and adult stages that allow them to exploit very different ecological niches. Some of the most striking adaptations within the Holometabola are seen in the articulated appendages, particularly the legs. How the developmental programs that control leg motoneuron connectivity have been modified is likely to be very interesting and may provide insights into evolution of neural networks. This census of Drosophila motoneurons is likely to help with comparative studies on leg motoneurons from other insect species (Brierley, 2012).

Regardless of the exact number of lineages that generate leg motoneurons, it is striking that just two postembryonic lineages contribute the bulk of the leg motoneurons (34 of the 47) and this raises the question of whether there is something fundamentally different about leg motoneuron specification compared with what was already know from studies in the embryo (Brierley, 2012).

One of the most notable differences between the larval and adult musculoskeletal system is that the muscles of adults are multifiber and often innervated by a number of isomorphic neurons. The dendrites of most leg motoneurons are located ipsilaterally and elaborate into the same neuromere in which they are born, unlike in the embryo where motoneurons can have extensive contralateral dendrites and are parasegmental in their nature. This difference in organization, i.e., segmental vs. parasegmental dendrites, could be an adaptation for larval locomotion, where the control of the next adjacent segment is critical (Brierley, 2012).

Although there is general agreement between many of the current findings and those of Baek (2009), there are some differences in detail, which may have important implications. Unlike Baek (2009), this study found that the largest leg motoneuron lineage, lineage 15, also innervates muscles in the body wall as well as intrinsic muscles in the femur and the tibia. Lineage 15 therefore has the most extensive coverage along the proximodistal axis of the leg and does not have a distal bias, as previously suggested (Baek, 2009). The extrinsic muscles in the body wall are extremely important, as they control the bodywall/coxal joint, which is in effect a universal joint allowing the leg a near 360
rotation. Although this study has presented a more complete picture of these motoneurons, more work is needed to identify the origins of the other motoneurons that innervate this complex group of muscles. Motoneurons within lineage 24 innervate muscles in the coxa, trochanter, and femur and control the movement of the femur and tibia. In this study no neuron from this lineage was seen innervating the tibia reductor muscle group, or indeed any other glutamatergic leg motoneurons terminating on two distinct muscle targets, as suggested by Baek (2009). The three lineages, lineages 20, 21, and 22, all innervate muscle groups in the coxa (Brierley, 2012).

What this lineage analysis highlighted is that within the CNS the dendrites of each of the postembryonic lineages each occupy distinct territories along the mediolateral, anteroposterior, and dorsoventral axes. This is important, as it is known from studies on the leg sensory system that the central afferent projections of different classes of sensory neuron occupy distinctive volumes within the dorsoventral axis depending on their modality (Brierley, 2012).

Lineage 15 has the most medially projecting dendrites located in both the anterior and posterior regions of the neuropil; the dendrites of motoneurons from lineage 24 motoneurons take up more lateral territories, whereas the central projections from lineages 20, 21, and 22 occupy the most lateral neuropil domains and span the anteroposterior axis. These lineage-specific patterns are reproducible, with no obvious variation in the muscles innervated or with a significant difference in the size or morphology of the axonal arborizations. This observation emphasizes that decoding lineage-specific programs of morphogenesis is likely to hold the key for understanding the development and organization of motoneurons within the leg network (Brierley, 2012).

To explore these motor lineages in more detail, individual neurons were visualized using the MARCM technique to determine how motoneuron birth date is correlated with aspects of morphology. As well as birth-dating, the single-cell clones allowed a close look at the relationship between muscle innervation and dendrite shape (Brierley, 2012).

In lineage 15, the sequential production was found of at least five distinct motoneuron subtypes during larval life. The first-born neuron innervates a muscle in the bodywall, the next subtype targets a muscle in the proximal femur, with the following subtype targeting a muscle in the proximal tibia. The next subtype innervates targets in the distal femur and then the distal tibia. Thus, there is no simple proximal to distal filling up of the leg, based on the birth-date of neurons; instead, neurons that innervate the most proximal target of a leg segment are born first. The central projections of these motoneuron subtypes were also very stereotyped, with the dendrites of early born cells spanning medial to lateral territories and late-born cells elaborating their dendrites in the lateral and ventral neuropil. Lineage 24 also shows a stereotyped birth-order based pattern of innervation along the proximodistal axis of the leg. It was found that lineage 24 generates three subtypes during larval life with both early and late-born neurons innervating the same muscle group located in the coxa and having dendrites that target lateral regions within the CNS. The second and third subtypes target the trochanter and the femur, respectively (Brierley, 2012).

It is interesting to speculate how a lineage like 15 may have evolved from an ancestral condition. The first motoneuron subtype innervates a body wall muscle and the next the long tendon muscle located in the femur. The long tendon muscle, also called the unguis retractor, attaches to the apodeme that controls the most distal element in the leg, the pretarsus. Could it be that these early born neuron subtypes are the most 'ancient' within the lineage, while the sequential addition of the later subtypes occurred as new leg segments were introduced? It would be intriguing to look at the homologous neurons in different outgroups (Brierley, 2012).

The long tendon muscle motoneurons are also unique among the glutamatergic leg motoneurons, as they are the only ones that elaborate dendrites in the contralateral hemineuromere. It is worthy of note that there are more long tendon muscle group (ltm) motoneurons than any other leg motoneuron. This is probably due to the need for the precise control of the pretarsal claw, which is fundamental to all locomotory and nonlocomotory behavior involving the leg. The difference in the birth-order of neuron types between the different lineages is also striking. Rigid birth-order-based rules that control the targeting of terminal processes have been described for other types of secondary neurons, including the antennal lobe projection neurons found in the fly's olfactory system. The sequential production of different neuron subtypes at distinct times during development is a common mechanism for generating the diversity of circuit components in many taxa, including vertebrates. In flies, there is strong evidence that individual Nbs express a sequence of progenitor transcription factors, such as Hunchback, Kruppel, Pdm, and Castor, which in turn regulate the postmitotic transcription factors to specify a distinct identity. The differences observed between neuronal birth-date and the dendritic and axonal arborizations in lineages 15 and 24 could be due to similar transient and sequential expression of temporally controlled transcription factors, like those observed in embryonic lineages or by other transcription factors such as Chinmo and Broad, which are deployed within postembryonic neuron subtypes. Although most studies in insects emphasize stereotyped lineage-specific specification a recent report describes how local interneuron populations within the Drosophila antennal lobe can have great morphological variability. It may be that particular neuronal classes, such as those that transfer information between one part of the nervous system and another, are more developmentally hard-wired than elements that perform mainly local processing (Brierley, 2012).

The data show that soma location is not an important descriptor of identity, but rather the location of their dendritic terminals. Taken as a whole, this work reveals that, although there is great stereotypy, there is no simple organizing principle that translates birthdate into projection pattern, i.e., that early born neurons innervating proximal leg segments and late born neurons targeting distal ones. Solving lineage-based codes within this system is likely to hold the key to understanding fundamental rules about how networks are assembled (Brierley, 2012).

Understanding how ordered patterns of synaptic connectivity are established between motoneurons and the rest of the motor network is a fundamental question in neurobiology. Landgraf (2003) revealed that the dendrites of motoneurons in the Drosophila embryo are organized to reflect the innervation of muscles in the periphery. They forwarded the idea that different territories within such a 'myotopic map' reflect patterns of connectivity with premotor elements and that such maps could be a general organizational principle of all motor systems. This study was directed to establishing whether leg motoneurons generate the same kind of myotopic map and thus explore the generality of this compelling idea (Brierley, 2012).

It was found that the dendrites of leg motoneurons occupy a large volume of the leg neuropil and showed a considerable degree of overlap, even though each occupies a slightly different volume. This is in marked contrast to the myotopic map seen in the embryo, where dendrites appear to generate exclusive, nonoverlapping territories: i.e., the dendrites of motoneurons that innervate internal muscles segregate into a different neuropilar domain from those innervating external muscles, alongside which the motoneurons of the internal muscles organize themselves into a map representing the dorsoventral axis of the body wall. This difference in organization may be due to differences in the skeleton and the biomechanics of the two systems. The adult leg is a complex multijointed appendage with many degrees of freedom, where interjoint coordination between and within legs is of paramount importance. In contrast, the fly larva locomotes using simple peristaltic waves of the abdominal wall and head turns. The organization of Drosophila leg motoneuron dendrites mirrors that of vertebrate somatic motoneurons in the spinal cord, where each motoneuron type has a dendritic arborization that covers a distinctive territory but at the same time has considerable overlap with the dendrites of other types (Brierley, 2012).

To step back from this, a systematic and unbiased analysis was performed of leg motoneuron dendrite position where the prothoracic neuropil was divided into sectors and the location of the arborizations of 13 different types was measured using single-cell MARCM clones. This approach allowed determination of the volume that each of the different motoneuron types samples within the neuropil and how the different types relate to each other. The dendrogram generated shows how closely related the different subtypes are. Any branch can be reversed around a node point, and the relatedness of different arborizations can be inferred using this. Using five neurons for each type helped provide a robust measure of the similarities/differences between the different motoneurons. It was found the 13 motoneuron types clustered into nine sets. Motoneurons of the same subtype tended to group together in most examples. Some motoneurons that innervate functionally related muscles also clustered together, e.g., the ltm1 and ltm2 muscles located in the femur and the tibia. It was also found that motoneurons that innervate the tibia levator and the tibia reductor muscle also formed a group: these are synergists (Brierley, 2012).

As a counterpoint to this, a number of motoneurons were seen that innervate antagonistic muscles cluster together, e.g., the trochanter levator and trochanter depressor muscles both located in the coxa being one set, and the tarsal levator and tarsal depressor muscles that control the tarsal segments being another. Baek found four sets that clustered together;they found the motoneurons that innervate the long tendon muscles in both the femur and tibia grouped together, as did antagonistic motoneurons that innervate the coxal segment. They also saw two different types of trochanter neurons cluster, which this study did not have data for. Baek suggested that different reductors clustered together but the neuron they proposed to innervate the tibia reductor does not. This study took the clustering data and remapped this onto the neuropil. It shows clearly the large overlap of the dendrites of many of the different motoneurons. What does this overlap mean functionally? First, it is important to emphasize that just because neurites occupy the same space, judged by light microscopy, it does not mean they make connections with each other or, as in this case, receive the same types of input. In the brachial spinal cord of the bullfrog Rana catesbeiana sensory axons from the triceps brachii muscle make connections with triceps motoneurons but do not innervate subscapularis and pectoralis motoneurons, which are in very close proximity. The monosynaptic connections between triceps brachii motoneurons and sensory neurons appears relatively late in development, after the dendrites have grown into a territory that contains an extant presynaptic terminal field (Brierley, 2012).

The specific connection occurs then as soon as the motoneuron arrives. Importantly, it says that if the terminals are not within a territory they cannot make connections with inputs there. The occurrence of pre- and postsynaptic elements in space is thus necessary but not sufficient for connectivity. Other examples in the vertebrate spinal cord show that there are considerable similarities in the morphology of somatic motoneuron dendrites within large parts of their arborization, but that key differences in specific regions can occur. A good example of this is seen in the dorsal dendrites in the lumbar motoneurons of the turtle Pseudemys scripta elegans, where such specialized differences in dendrite morphology might reflect a difference in synaptic input or the processing of input. The finding that the dendrites of Drosophila motoneurons that innervate antagonistic muscle pairs are similar is interesting (Brierley, 2012).

Are such dendritic organizations important for interpreting information from similar inputs? Future work exploring connectivity using physiological approaches should allow us to address whether this is important for function or for determining patterns of connectivity during development. Although there is no simple 'easy to read' map, what the data shows is that there are robust topological relationships between these dendritic arborizations (Brierley, 2012).

As described above, there exists a diversity of dendritic and axonal projection patterns within neural maps. A key question raised by this is how do neurons within such maps ensure that both the axonal and dendritic terminals execute appropriate programs of morphogenesis. We now know that dendrites, like axons, use conserved molecular cues and various transmembrane receptors to attain their distinct organizations. One possible mechanism for generating a diversity of dendrite shapes could be retrograde signaling from the target muscle. It was of interest to look at the relative timing of axon and dendrite outgrowth in this system to see if this could be possible. The data from lineage 15 reveals that two subtypes, which innervate different long tendon muscle sets, have nearly identical dendritic trees, but their axons target muscles in different segments of the leg. The timeline data shows that motoneuron axon outgrowth in the proximal leg occurs at the same time as dendritic elaboration in the CNS. This opens the possibility that retrograde signals may play a role in the development of neurons that innervate muscle targets in the leg (Brierley, 2012).

How modular are these programs? These events must be controlled at some level by transcription factor codes regulating blends of guidance receptors. Cell intrinsic temporal transcription factors can control a combinatorial code of postmitotic transcription factors. Feedback from the muscle field could also provide patterning information as seen in the vertebrate spinal cord, where motoneuron dendrite arborizations are controlled in part by the transcription factor Pea3, which is induced by retrograde signaling from target muscles. The respective timing of outgrowth of Drosophila leg motoneuron axons and dendrites opens this up as a possibility. Previously, laser ablation studies revealed that the dendrites of Drosophila flight motoneurons change their growth following the removal of their target muscle. It will be interesting to test this hypothesis experimentally by removing the muscle targets in the leg and quantifying the dendritic arborizations of known motoneurons (Brierley, 2012).

To move the body, the brain must precisely coordinate patterns of activity among diverse populations of motor neurons. This study used in vivo calcium imaging, electrophysiology, and behavior to understand how genetically-identified motor neurons control flexion of the fruit fly tibia. Leg motor neurons exhibit a coordinated gradient of anatomical, physiological, and functional properties. Large, fast motor neurons control high force, ballistic movements while small, slow motor neurons control low force, postural movements. Intermediate neurons fall between these two extremes. This hierarchical organization resembles the size principle, first proposed as a mechanism for establishing recruitment order among vertebrate motor neurons. Recordings in behaving flies confirmed that motor neurons are typically recruited in order from slow to fast. However, this study also found that fast, intermediate, and slow motor neurons receive distinct proprioceptive feedback signals, suggesting that the size principle is not the only mechanism that dictates motor neuron recruitment. Overall, this work reveals the functional organization of the fly leg motor system and establishes Drosophila as a tractable system for investigating neural mechanisms of limb motor control (Azevedo, 2020).

Dexterous motor behaviors require precise neural control of muscle contraction to coordinate force production and timing across dozens to hundreds of muscles. This coordination is mediated by populations of motor neurons, which translate commands from the central nervous system into dynamic patterns of muscle contraction. Although motor neurons are the final common output of the brain, the scale and complexity of many motor systems have made it challenging to understand how motor neuron populations collectively control muscles and thus generate behavior. For example, a human leg is innervated by over 150,000 motor neurons and a single calf muscle is innervated by over 600 motor neurons. How can the nervous system coordinate the activity of such large motor neuron populations to flexibly control the force, speed and precision of limb movements (Azevedo, 2020)?

One way to streamline motor control is to establish a hierarchy among neurons controlling a particular movement, such as flexion of a joint. This hierarchy allows premotor circuits to excite different numbers of motor neurons depending on the required force: motor neurons controlling slow or weak movements are recruited first, followed by motor neurons that control progressively stronger, faster movements. A recruitment order for vertebrate motor neurons innervating a single muscle was first postulated over 60 years ago. Subsequent work identified mechanisms associated with the recruitment order and synthesized these findings as the size principle, which states that small motor neurons, with lower spike thresholds, are recruited prior to larger neurons, which have higher spike thresholds. Evidence for the size principle has been provided by electrophysiological analysis of motor neurons in a number of species, from crayfish to humans. These studies have also described systematic relationships between motor neuron electrical excitability, recruitment order, conduction velocity, force production, and strength of sensory feedback and descending input (Azevedo, 2020).

A simplifying assumption of the size principle is that all motor neurons within a pool receive identical presynaptic input, such that recruitment order is entirely dictated by the gradient of motor neuron excitability. However, recordings in vertebrates have found that presynaptic inputs may vary within a motor pool and that recruitment order can change during specific movements. Are these violations of the size principle a fundamental feature of motor control circuits or rare exceptions to the rule? Answering this question would be greatly aided by a tractable system in which it was possible to genetically target identified motor neurons for recording, manipulation, and mapping of presynaptic inputs (Azevedo, 2020).

The leg of the fruit fly, Drosophila melanogaster, contains 14 muscles which are innervated by just 53 motor neurons. In spite of this tiny scale, the fly leg supports a variety of fast and flexible behaviors. During forward walking, the fly grips the substrate with the distal segment of its front leg (the tarsus) and flexes the femur-tibia joint, pulling the body forward. The femur-tibia joint of a walking fly flexes and extends 10-20 times per second, reaching swing speeds of several thousand degrees per second. Flies also use their legs to target other body parts during grooming, for social behaviors like aggression and courtship, and to initiate flight take-off and landing. These behaviors require a wide range of muscle force production across multiple timescales. However, little is currently known about the organization and function of leg motor control circuits in the fly. While a great deal of progress has been made on understanding the processing of sensory signals in the Drosophila brain, understanding of how this information is translated into behavior by the fly's ventral nerve cord (VNC) is lacking. Investigating motor control in Drosophila is important because the fly's compact nervous system and identified cell types make it a tractable system for comprehensive circuit analysis (Azevedo, 2020).

This study has investigated motor control of the Drosophila tibia. First the organization of tibia flexor motor units was mapped using calcium imaging from leg muscles in behaving flies. With electrophysiology, it was then discovered that motor neurons controlling tibia flexion lie along a gradient of anatomical and physiological properties that correlate with muscle force production. Slow motor neurons produce <0.1 μN per spike, while fast motor neurons produce ~10 μN per spike, approximately equal to the fly's weight. Recordings during spontaneous leg movements revealed a recruitment hierarchy: slow motor neurons typically fire first, followed by intermediate, then fast neurons. Interestingly, all tibia flexor motor neurons receive feedback from proprioceptors at the femur/tibia joint, but these sensory signals vary in amplitude, sign, and dynamics across the different motor neuron types. Optogenetic manipulation of each motor neuron type had unique and specific effects on the behavior of walking flies, consistent with their roles in controlling distinct force regimes (Azevedo, 2020).

Together, these data establish the organization and function of a key motor control module for the fly leg. Overall, this study found that motor neurons controlling the fly tibia exhibit many features consistent with the size principle. However, it was also observed that tibia motor neurons receive distinct proprioceptive feedback signals and that recruitment order is occasionally violated. Thus, in addition to the size principle, heterogeneous input from premotor circuits is likely to play an important role in coordinating neural activity within a motor pool (Azevedo, 2020).

This study discovered a gradient of properties that co-varies across the motor pool. At one end of the spectrum is the fast motor neuron, which has a larger diameter axon and dendrites, low input resistance, weak sensory input, and produces enough force with each spike to support the fly's entire body weight. At the other end, the slow motor neuron has fine axons and dendrites and high input resistance; each spike in the slow motor neuron produces <1% the force of a fast motor neuron spike. The slow motor neuron also has a high spontaneous firing rate that maintains resting muscle tension, and it receives strong proprioceptive feedback that continuously modulates its firing rate (Azevedo, 2020).

Leg motor neurons are recruited in a specific order: small, low gain motor neurons that generate weak forces are recruited first, and larger, more powerful motor neurons are recruited later. Consistent with this recruitment order, proprioceptive feedback has the greatest effect on the slow motor neuron: tibia movements of less than 1° significantly impact slow motor neuron firing. By comparison, proprioceptive feedback has a much smaller effect on the firing rate of intermediate and fast motor neurons. Fly behavior is also quite sensitive to aberrant optogenetic recruitment of motor neurons. For example, flies fail to initiate walking if unable to recruit slow motor neurons at the bottom of the recruitment hierarchy (Azevedo, 2020).

A relationship between force production and recruitment order is a common feature of vertebrate motor systems. This organization has been proposed to confer a number of computational and energetic advantages. Recruitment of additional motor units increases force nonlinearly, overcoming suppressive nonlinearities in spike rates and muscle force production. Furthermore, the relative increase in force does not decrease with successive recruitment, as it would if all motor units produced similar amounts of force. Thus, much like Weber's law describes the constant sensitivity to relative stimulus intensity, a recruitment hierarchy maximizes the resolution of motor unit force while also simplifying the dimensionality of the motor system (Azevedo, 2020).

A key assumption of the original size principle is that all motor neurons within a pool receive the same presynaptic input; the alternative is that different premotor neurons provide input to different subsets of motor neurons. The results of this study suggest the motor system controlling the fly tibia operates in a middle ground between these two extremes. Although tibia motor neurons generally follow a recruitment order in accordance with the size principle, it was found that the dynamics of proprioceptive feedback vary across motor neurons within the pool. A small but significant proportion of behavioral events were observed in which recruitment order was violated. These data suggest that although the tibia flexor motor neuron pool may share some presynaptic inputs, they are not identical. Thus, motor control of the fly tibia is more complex than a straightforward implementation of the size principle (Azevedo, 2020).

Similar exceptions to the strict dogma of the size principle have been observed in vertebrate species. For instance, the effect of proprioceptive feedback varies across a pool of motor neurons controlling the cat leg. During rapid escape behaviors in zebrafish, slow motor neurons can completely drop out of the population firing pattern. Muscles controlling human fingers can change recruitment order based on movement direction. These examples suggest that some behaviors exhibit more degrees of freedom than can be supported by recruitment of motor neurons based on their intrinsic excitability alone (Azevedo, 2020).

Investigations of the vertebrate spinal cord have also identified circuit motifs that support flexible control of motor neurons within a pool. For example, recordings in turtles and zebrafish have shown that motor neurons receive coincident excitation and inhibition, which could underlie selective de-recruitment. Zebrafish spinal cord premotor neurons provide patterned inhibition to speed-specific circuits and can flexibly switch between different motor neuron recruitment patterns (Azevedo, 2020).

This characterization of identified tibia motor neurons provides a handle to investigate similar premotor circuit motifs for flexible limb motor control in the fly. Little is currently known about the leg premotor circuitry in Drosophila. However, the hypotheses generated by this work should soon be testable using connectomic reconstruction of identified motor neurons and their presynaptic inputs in the Drosophila VNC (Azevedo, 2020).

Although programmed cell death (PCD) plays a crucial role throughout Drosophila CNS development, its pattern and incidence remain largely uninvestigated. This study provides a detailed analysis of the occurrence of PCD in the embryonic ventral nerve cord (VNC). The spatio-temporal pattern of PCD was traced and the appearance of, and total cell numbers in, thoracic and abdominal neuromeres of wild-type and PCD-deficient H99 mutant embryos were compared. Furthermore, the clonal origin and fate of superfluous cells in H99 mutants was examined by DiI labeling almost all neuroblasts, with special attention to segment-specific differences within the individually identified neuroblast lineages. These data reveal that although PCD-deficient mutants appear morphologically well-structured, there is significant hyperplasia in the VNC. The majority of neuroblast lineages comprise superfluous cells, and a specific set of these lineages shows segment-specific characteristics. The superfluous cells can be specified as neurons with extended wild-type-like or abnormal axonal projections, but not as glia. The lineage data also provide indications towards the identities of neuroblasts that normally die in the late embryo and of those that become postembryonic and resume proliferation in the larva. Using cell-specific markers it was possible to precisely identify some of the progeny cells, including the GW neuron, the U motoneurons and one of the RP motoneurons, all of which undergo segment-specific cell death. The data obtained in this analysis form the basis for further investigations into the mechanisms involved in the regulation of PCD and its role in segmental patterning in the embryonic CNS (Rogulja-Ortmann, 2007).

In this analysis of PCD distribution it was found that, macroscopically, the CNS of wt and PCD-deficient (H99) embryos do not show large differences. These observations indicate that the supernumerary cells do not disturb developmental events in the CNS of H99 embryos, such as cell migration and axonal pathfinding. The glial cells mostly find their appropriate positions accurately. The DiI-labeled NB lineages were, in the majority of cases, easily identifiable based on their shape, position and axonal pattern, despite the supernumerary cells. The FasII pattern showed that the axonal projections form and extend along their usual paths. In fact, the supernumerary cells themselves are capable of differentiating i.e. expressing marker genes and extending axons, as shown by clones of several NBs and by cell marker expression analysis in H99 (e.g. NB7-3) (Rogulja-Ortmann, 2007).

It has been shown that a large number of CNS cells undergo PCD during embryonic development. The distribution of activated Caspase-3-positive cells in wt embryos suggests that the death of some cells is under tight spatial and temporal control, as revealed by their regular, segmentally repeated occurrence. Other dying cells were rather randomly distributed, suggesting a certain amount of developmental plasticity. The overall counts of Caspase-3-positive cells give an estimate of the numbers of dying cells at a given time. They indicate that PCD becomes evident in the CNS at stage 11 and is most abundant in the late embryo (from stage 14). It is however difficult to estimate the total number of apoptotic cells throughout CNS development by anti-Caspase-3 labeling, because the cell corpses are removed fairly quickly. Therefore the total number of cells were counted per thoracic and abdominal hemineuromere in the late embryo. Comparison between stage 16 and stage 17 wt embryos indicates that 25-30 % of all cells are removed in both tagmata after stage 16, which in turn suggests that the total percentage of removed cells must be high, since PCD occurs at high levels already from stage 14 on. In comparison to the developing nervous system of C. elegans, where PCD removes about 10% of cells, and of mammals, where this number can be as high as 50-90%, PCD in the fly CNS appears to show an intermediate prevalence. This lends support to the hypothesis of an increasing contribution of PCD in shaping more advanced nervous systems during evolution (Rogulja-Ortmann, 2007).

Comparisons between wt and H99 reveal, as expected, a greater number of cells in both tagmata of H99 embryos (151% increase in the thorax and 162% in the abdomen at stage 17). These additional cells in H99 may reflect the total number of cells normally undergoing cell death until stage 17. However, there is a large variability in the total number of cells, especially within the H99 strain. In wt embryos, it seems to be more pronounced in the thorax and at stage 17, which might be a consequence of variable amounts of PCD occurring until this stage. The even higher variability within the H99 strain (both in thorax and abdomen) is likely to reflect variable numbers of additional cell divisions. The great majority of abdominal NBs are normally removed by PCD after they have generated their embryonic progeny, whereas in the thoracic neuromeres most of the NBs enter quiescence at the end of embryogenesis and continue dividing as postembryonic NBs in larval stages. Thus, there are few mitoses occurring in the wt CNS from stage 16 onwards. BrdU labeling experiments revealed a high number of BrdU-positive cells in some H99 embryos injected at early stage 17. It is assumed that these are progeny of mitotic NBs and/or GMCs that survive and continue dividing, generating cells that do not exist in wt. Clones obtained by DiI labeling in H99 confirm this conclusion. The finding that surviving cells divide already in the embryo complement results that showed that, in reaper mutants, NBs in the abdominal neuromeres survive and generate progeny in larval stages (Rogulja-Ortmann, 2007).

Among the DiI-labeled clones in H99 embryos, very few NB lineages were obtained which did not differ from their wt counterparts. The majority contained, as expected, supernumerary cells. In some cases axons projected by these cells could be identified, showing that they are specified as neurons. In fact, in three cases (NB4-2, NB5-3 and NB7-3), these additional cells were found to be specified as motoneurons. As additional axons within a fascicle were generally difficult to identify, it is possible that these are not the only lineages which make additional motoneurons in H99. Whether these cells are normally born and apoptose, or originate from additional divisions of surviving NBs or GMCs, cannot be determined from these experiments, but similar observations have been made for both cases. It is interesting that none of these cells, regardless of their origin, are specified as glia. No additional glia were observed in the NB clones in H99 embryos, and equal numbers of Repo-expressing glial cells were found in wt and H99. It is concluded that PCD occurs almost exclusively in neurons and/or undifferentiated cells, and that lateral glia are not produced in excess numbers in the embryo. Furthermore, because it is likely that NBs, which normally die, stay in a late temporal window in H99, one could speculate that NBs in this window normally do not give rise to glia. These results are not in agreement with the notion that LG are overproduced, and their numbers adjusted through axon contact. Occasional apoptotic LG have been observed and it is possible that the current method of counting does not allow a resolution fine enough to account for an occasional additional Repo-positive cell in H99 embryos. However, if LG were consistently overproduced, a higher number of glia in would be expected H99 embryos. It is assumed that LG cell death may reflect a small variability in the number of cells needed, and not a general mechanism for adjusting glial cell numbers (Rogulja-Ortmann, 2007).

Generally, no difference was found between Repo-expressing glia numbers in wt and H99. However, a small difference does become apparent when one separates the total cell counts into those in the CNS and those in the periphery: 25.67±0.45 cells/hs and 28.42±0.64 cells/hs for wt and H99, respectively, were counted in the CNS, whereas 8.50±0.28 cells/hs and 6.35±0.82 cells/hs for wt and H99, respectively, were found in the periphery. The reasons for this difference might be the greater width of the CNS in H99 embryos, and that the cues required for proper migration of the peripheral glia are disturbed by additional cells. Alternatively, the difference might be due to differentiation defects in these cells (Rogulja-Ortmann, 2007).

In addition to NB clones with too many cells and wild-type-like axon projections in H99, some lineages were obtained whose clones exhibited atypical projection patterns. These projections were found to belong both to motoneurons (e.g. in NB4-2) and interneurons (e.g. NB5-3, NB7-2 and NB-7-4). NB4-2 normally produces two motoneurons (RP2 and 4-2Mar) and 8-14 interneurons. In two out of three NB4-2 clones in H99 two additional motoneurons that project anteriorly were found, similar to RP2. One of the two clones was found in the thorax and had a normal cell number (16), whereas the other was abdominal and had too many cells (25). Thus, the two additional motoneurons are likely to be the progeny of divisions occurring in the wt, and not of an additional NB or GMC mitosis. The fact that the third NB4-2 clone (found in the abdomen and comprising 17 cells) did not show the same motoneuronal projections could be due to these cells not being differentiated at the time of fixation (clones of different ages were occasionally observed in the same embryo), or they may not have differentiated at all. It would be interesting to determine the target(s) of these additional motoneurons and thereby perhaps gain insight into physiological reasons for their death. However, such an experiment has to await tools that allow specifically labeling of the NB4-2 lineage, or these motoneurons, in the H99 mutant background (Rogulja-Ortmann, 2007).

The other three lineages (NB5-3, NB7-2 and NB7-4) all have atypical interneuronal projections. The cells which these atypical axons belong to may represent evolutionary remnants that are not needed in the Drosophila CNS. Alternatively, they might have a function earlier in development and be removed when this function is fulfilled. Such a role has been shown for the dMP2 and MP1 neurons, which are born in all segments and pioneer the longitudinal axon tracts. At the end of embryogenesis these neurons undergo PCD in all segments except A6 to A8, where their axons innervate the hindgut. It is known that some cells of the NB5-3 lineage express the transcription factor Lbe, and that H99 mutants show about three additional Lbe-positive neurons per hemisegment, which mostly likely belong to NB5-3. The DiI-labeling results complement this finding in that four or more additional neurons were also found in H99 clones. The supernumerary Lbe-positive neurons in H99 could possibly be the ones producing the atypical axonal projections (Rogulja-Ortmann, 2007).

In the wt embryo, only eight NB lineages show obvious tagma-specific differences in cell number and composition. Tagma-specific differences among serially homologous CNS lineages have been shown to be controlled by homeotic genes. Therefore, these lineages provide useful models for studying homeotic gene function on segment-specific PCD. In H99 embryos, further lineages were observed that were differently affected in the thorax and abdomen. How these tagma-specific differences arise in a PCD-deficient background is an interesting question. For example, NB4-3 shows a wild-type cell number in the thorax (8 and 12-13), but has too many cells in the abdomen (15, 15 and 22). There are a couple of plausible scenarios to explain this observation. (1) The development of the NB4-3 lineage, including the involvement of PCD, could actually differ in the thorax and abdomen of wt embryos, with the final cell number being similar by chance. The DiI-labeled clones allow determination of the final cell number, but do not reveal how this number is achieved. The difference would become obvious in an H99 mutant background, at least regarding the involvement of PCD. (2) This possibility does not exclude the first one, the thoracic NB4-3 could become a postembryonic NB (pNB) and the abdominal NB4-3 might undergo PCD after generating the embryonic lineage. In H99, the abdominal NB would be capable of undergoing a variable number of additional divisions to generate a variable number of progeny. This would easily explain larger discrepancies in cell number between individual clones in H99 (e.g. the abdominal NB4-3 clone with 22 cells), and is in agreement with occasional observations of H99 embryos with a very high CNS cell number per segment, and with the two observed classes of H99 embryos with high and low numbers of BrdU-positive cells (Rogulja-Ortmann, 2007).

NB6-2 is another lineage whose clones differ in the two tagmata of H99 embryos. In this case, the abdominal clones showed no difference to their wt counterparts, whereas the thoracic clones did (18 and 19 cells). Although no difference in cell number between thoracic and abdominal clones was reported for this lineage, a rather large count range (8-16 cells) was given, which would allow for a thorax-specific PCD of two to three postmitotic progeny. Alternatively, the thoracic NB6-2 might undergo cell death upon generating its progeny, which would make it the first identified apoptotic NB in the thorax. When PCD is prevented, this NB may undergo a few additional rounds of division. The data obtained in these experiments do not counter this notion, but the number of clones obtained in the thorax was not sufficient to draw a definite conclusion. As the abdominal NB6-2 lineage in H99 did not differ from the one in wt, its NB may be one of the few abdominal postembryonic NBs (Rogulja-Ortmann, 2007).

A specific set of NBs undergoes PCD in the late embryo, whereas surviving NBs resume proliferation in the larva as pNBs, after a period of mitotic quiescence. The identities of the individual NBs undergoing PCD versus those surviving as pNBs are still unknown. The sizes of NB lineages obtained in H99 embryos may provide hints for identifying candidate pNBs in the abdomen [12 NBs/hs in A1, four in A2 and three in A3 to A7, and NBs that undergo PCD in the thorax at the end of embryogenesis [seven NBs/hs in T1 to T3. In the abdomen, NB1-1a and NB6-2 are obvious candidates for pNBs, as they remained consistently unchanged in H99 embryos. Two other NBs, NB1-2 and NB3-2, are also potential abdominal pNBs as they mostly did not differ from their wt counterparts, and only occasionally contained one additional cell. On the other hand, clones which showed more than twice the cell number in H99 (NB2-1, NB5-4a and NB7-3) than in wt, strongly suggest that these NBs normally undergo PCD in the abdomen (but perform additional divisions in H99), because, even if one daughter cell of each GMC undergoes PCD, they still cannot account for all cells found in H99 clones (Rogulja-Ortmann, 2007).

Regarding thoracic NBs, it can only be speculated on account of low sample numbers. NBs which seem to become pNBs in the thorax, as they showed no difference between wt and H99 clones, are NB3-2, NB4-3 and NB4-4. Potential candidates for NBs which do not become pNBs, but undergo PCD in the thorax, are expected to consistently have a significant increase in cell number in H99. These are NB5-1 and NB5-5. In addition, lineages for which one clone was obtained in H99 but which also showed many more cells in the thorax than normal are NB2-2t, NB5-4t and NB7-3 (Rogulja-Ortmann, 2007).

In order to investigate the developmental signals and mechanisms involved in the regulation of PCD in the embryonic CNS, some of the apoptotic cells were identified which will be used as single-cell PCD models. These are the dHb9-positive RP neuron from NB3-1, Lbe-positive neurons from NB5-3, the Eg-positive GW neuron from NB7-3 and the Eve-positive U neurons from NB7-1. As not much is known about the dying RP motoneuron or the Lbe-positive neurons, the first goal will be to characterize each of these cells more closely, based on the combination of expressed molecular markers (Rogulja-Ortmann, 2007).

Some of the dying NB7-3 cells are already known to be undifferentiated daughter cells of the second and third GMC, which undergo PCD shortly after birth. Notch has been identified as the signal initiating PCD. The surviving daughters receive the asymmetrically distributed protein Numb, which counteracts the PCD-inducing Notch signal. The same had been shown in a sensory organ lineage of the embryonic peripheral nervous system, where cells produced in two subsequent divisions undergo Notch-dependent PCD. Both the PCD in the NB7-3 lineage and in the sensory organ lineage require the Hid, rpr and grim genes. It will be interesting to see whether the Notch-Numb interaction also plays a role in the segment-specific PCD of the differentiated GW motoneuron, or if another signal is used for the removal of this, and possibly other, differentiated cells (Rogulja-Ortmann, 2007).

The U motoneurons also show a segment-specific cell death pattern (they apoptose in A6 to A8), thus somewhat resembling the MP1 and dMP2 neurons. However, in contrast to MP1 and dMP2, the U neurons survive in the anterior segments and undergo PCD in the posterior ones. Whether homeotic genes play any role in the survival or death of these cells remains to be investigated (Rogulja-Ortmann, 2007).

What mechanisms control the sequential generation of neurons -- that is, how are unique fates acquired by the successive daughters of a neuroblast? After a neuroblast has been specified positionally, by the actions of segment polarity genes and a network of homeodomain genes, it delaminates from the ectoderm and begins to divide unequally into one large and one small daughter cell. The large cell (still called a neuroblast) continues to go on this way for a variable number of rounds. The small cell, called a ganglion mother cell (GMC), typically divides equally one more time to generate a pair of postmitotic neurons. Often these neurons form a stack on top of the neuroblast from which they originated. As postmitotic neurons in the insect CNS do not generally migrate, the position of a neuron in the CNS depends on whether it was generated early or late. In this way, a histogenetic order is built into the cellular cortex of the insect CNS, with early neurons deep and close to the neuropil and late neurons next to the surface of the brain (Kambadur, 1998; Harris, 2001).

This arrangement of cells according to relative birth date is also observed in laminated structures in the vertebrate CNS, the best example being the cerebral cortex. In the mammalian cortex, cells acquire their fates at the ventricular surface at the time they are born, and these postmitotic neurons cells then migrate to their specified laminar destinations (McConnell, 1995). In both the cortex and the retina, it is thought that progenitors are pluripotent and realize their particular fates by being exposed to an extracellular environment that changes with time (Harris, 1997). Whether intrinsically or extrinsically controlled, particular combinations of transcription factors are expressed in the neuroblasts of both the vertebrate retina and fly CNS over the course of development, and these factors appear to restrict the competence of neuroblasts to the fates that are appropriate (Harris, 2001 and references therein).

The first insights into this problem in the fly CNS were made in Odenwald's laboratory at the NIH. They showed that expression of the transcription factor genes hunchback (hb), pdm, and castor (cas) occur sequentially in the embryonic CNS of Drosophila (Kambadur, 1998). Furthermore, the CNS neuroblasts themselves sequentially express these three genes in a conserved order (Brody, 2000), and whichever of the three genes is expressed in the neuroblast when it divides continues to be expressed in the progeny. The transcription factor Grainyhead (Gh) appears to mark the NB after it has generated lineages marked by Hb, Pdm and Cas, and the Gh positive NB also generates a fourth lineage (Brody, 2000). Thus, the earliest generated neurons in the fly CNS tend to express hb, while later generated neurons express pdm, and still later generated neurons express cas followed by gh. By analogy to the spatial coordinate genes, these can be called 'temporal coordinate genes (Harris, 2001).

Isshiki (2001), working in Doe's laboratory, followed individual neuroblasts and their progeny. It appears that each neuroblast examined express four temporal coordinate genes -- hb, Kruppel (Kr), pdm, and cas, in that invariant order. By following the GMCs and their daughter neurons, Isshiki confirmed on a cellular level that each GMC maintains the expression profile of the temporal coordinate genes that its parent neuroblast displayed at the time the GMC was generated. The relevance of these temporal coordinates to neuronal fate was addressed with misexpression constructs and loss of function mutants in hb and Kr. These experiments led to respecification of GMCs and their progeny to earlier or later fates, as expected if these genes really are important to fate. Thus, these temporal coordinate genes play a similar role in fate specification along a histogenetic axis as the spatial coordinate genes play in the positional axes (Harris, 2001).

What, one might wonder, are the mechanisms that could generate these temporal transitions in the parent neuroblast? The first possibility is that global temporal cues, such as circulating hormones, or intracellular signaling such as Notch, Egf, Wingless, or Hedgehog, trigger these transitions. Since transitions in transcription factor expression take place in isolation, in cultured clones, it is concluded that once NBs initiate lineage development no additional signaling between NBs and the neuroectoderm and/or mesoderm is required to trigger the temporal progression of transcription factor expression during NB outgrowth (Brody, 2000; Harris, 2001).

Second, the cascade mechanism itself, in which each of these genes is responsible for turning on the next in the series, functions to regulate the transcription factor transitions. Overexpression of Hb activates Kr and represses Pdm and Cas; overexpression of Kr activates Pdm, represses Cas, but has no effect on Hb expression; Pdm positively regulates Cas expression; and Cas repressed Pdm expression (Kambdadur, 1998 and Isshiki, 2001). Although there are appropriate sequential regulatory interactions of this kind, mutations in any of the earlier genes only subtly affect the temporal expression of subsequently expressed genes. Thus, although these interactions refine sequential expression, there must be additional elements to temporal regulation. Isshiki (2001) has shown that individual NBs go through the same sequence of expression in their own sweet time, independent of the developmental stages at which they delaminate. A third possibility is therefore suggested, a more mysterious clock mechanism may also be responsible for generating the order. The clock in this case appears to be directly related to the cell cycle, since arresting cell division with the Cdc25 mutant, string, freezes the pattern in time (Cui, 1995; Weigmann, 1995; Harris, 2001 and references therein).

Group B Sox domain transcription factors play conserved roles in the specification and development of the nervous system in higher metazoans. However, comparatively little is known about how these transcription factors regulate gene expression, and the analysis of Sox gene function in vertebrates is confounded by functional compensation between three closely related family members. In Drosophila, only two group B Sox genes, Dichaete and SoxN, have been shown to function during embryonic CNS development, providing a simpler system for understanding the functions of this important class of regulators. Using a combination of transcriptional profiling and genome-wide binding analysis this study conservatively identified over 1000 high confidence direct Dichaete target genes in the Drosophila genome. Dichaete is shown to play key roles in CNS development, regulating aspects of the temporal transcription factor sequence that confer neuroblast identity. Dichaete also shows a complex interaction with Prospero in the pathway controlling the switch from stem cell self-renewal to neural differentiation. Dichaete potentially regulates many more genes in the Drosophila genome and was found to be associated with over 2000 mapped regulatory elements. This analysis suggests that Dichaete acts as a transcriptional hub, controlling multiple regulatory pathways during CNS development. These include a set of core CNS expressed genes that are also bound by the related Sox2 gene during mammalian CNS development. Furthermore, Dichaete was identified as one of the transcription factors involved in the neural stem cell transcriptional network, with evidence supporting the view that Dichaete is involved in controlling the temporal series of divisions regulating neuroblast identity (Aleksic, 2013).

The core Dichaete binding intervals identified in this study are enriched for Sox binding motifs but significant overrepresentation was also found of binding motifs for Vfl (Zelda), the GAGA-binding factor Trl, and the JAK-STAT pathway transcription factor Stat92E. All three of these factors have been identified as key elements in the regulatory programme that drives the onset of zygotic gene expression in the blastoderm embryo. Dichaete also plays a key role in early zygotic gene expression, regulating the correct expression of pair rule genes, and this study found overlapping Vfl/Dichaete binding at eve, h, and run stripe enhancers. While most of the work on Vfl has focused on understanding its function during the maternal to zygotic transition, the gene is expressed more widely after cellularisation, particularly in the CNS. Indeed recent work has shown a specific role for Vfl in the CNS midline, a tissue where Dichaete is known to be active, and this study found overlapping Vfl/Dichaete binding associated with slit and comm, known Dichaete midline targets. Post cellularisation functions for Trl and Stat92E are well established (Aleksic, 2013).

These three factors, particularly Vfl and Trl, have been strongly associated with enhancer activity driven by Highly Occupied Target (HOT) regions. HOT regions have been identified in large scale studies of the Drosophila, C. elegans and human genomes, and represent genomic sites where many functionally unrelated transcription factors bind, frequently in the absence of specific binding motifs. The finding that Dichaete binding locations are marked by overrepresentation of binding motifs for factors defining HOT regions, coupled with the widespread gene expression effects of Dichaete mutations, suggests that Dichaete may also play a role in regulatory interactions at HOT enhancers. It is notable that Dichaete, in common with all other characterised Sox proteins, is known to bend DNA upon binding. It is possible that Dichaete activity at HOT regions is mediated by this bending activity, helping to bring together complexes of other regulators. In this view, Dichaete would assist binding of factors at non-canonical target sites by favouring protein-protein interactions. In one of the bona fide Dichaete regulatory elements that have been studied in detail, the slit midline enhancer, Dichaete helps coordinate interactions between the POU factor Vvl and a Sim/Tango heterodimer (Aleksic, 2013).

During central nervous system (CNS) development, progenitors typically divide asymmetrically, renewing themselves while budding off daughter cells with more limited proliferative potential. Variation in daughter cell proliferation has a profound impact on CNS development and evolution, but the underlying mechanisms remain poorly understood. This study found that Drosophila embryonic neural progenitors (neuroblasts) undergo a programmed daughter proliferation mode switch, from generating daughters that divide once (type I) to generating neurons directly (type 0). This typeI> 0 switch is triggered by activation of Dacapo (mammalian p21CIP1/p27KIP1/p57Kip2) expression in neuroblasts. In the thoracic region, Dacapo expression is activated by the temporal cascade (castor) and the Hox gene Antennapedia. In addition, castor, Antennapedia, and the late temporal gene grainyhead act combinatorially to control the precise timing of neuroblast cell-cycle exit by repressing Cyclin E and E2f. This reveals a logical principle underlying progenitor and daughter cell proliferation control in the Drosophila CNS (Baumgardt, 2014).

Proliferation analysis of the developing Drosophila VNC reveals that most, if not all, lateral NBs initially divide in the type I proliferation mode, generating daughters that divide once. Three specific lineages, as well as many other NBs, subsequently switch to generating daughters that do not divide (type 0 mode). The full extent of the typeI>0 switch is currently difficult to precisely assess for several reasons. One such complicating issue pertains to possible developmental changes in daughter cell-cycle length over time. On this note, however, no obvious change was found in NB5-6T daughter divisions prior to the switch. In addition, if the accelerated decline in daughter division was indeed caused by a lengthening of the cell cycle rather than a typeI>0 switch, a long 'tail' of daughter divisions would be expected, perduring into St16-17. This is not the case; rather, daughter proliferation drops down to almost zero by St16-17. Similarly, no evidence was found for changes in NB cell-cycle length over time in the three specific NB lineages. Another complicating issue pertains to the fact that NBs differ in their time point of delamination, number of division rounds, and time point of switching, so that even if all NBs switched, only a fraction of the NBs would be in their type 0 window at the same time. However, these complications are more likely to lead to under- rather than overappreciation of the extent of the typeI>0 switch, and it is tempting to speculate that it may indeed involve the vast majority of NBs (Baumgardt, 2014).

Drosophila neuroblasts are an excellent model for investigating how neuronal diversity is generated. Most brain neuroblasts generate a series of ganglion mother cells (GMCs) that each make two neurons (type I lineage), but sixteen brain neuroblasts generate a series of intermediate neural progenitors (INPs) that each produce 4-6 GMCs and 8-12 neurons (type II lineage). Thus, type II lineages are similar to primate cortical lineages, and may serve as models for understanding cortical expansion. Yet the origin of type II neuroblasts remains mysterious: do they form in the embryo or larva? If they form in the embryo, do their progeny populate the adult central complex, as do the larval type II neuroblast progeny? This study presents molecular and clonal data showing that all type II neuroblasts form in the embryo, produce INPs, and express known temporal transcription factors. Embryonic type II neuroblasts and INPs undergo quiescence, and produce embryonic-born progeny that contribute to the adult central complex. These results provide a foundation for investigating the development of the central complex, and tools for characterizing early-born neurons in central complex function (Walsh, 2017).

It has been difficult to link embryonic neuroblasts to their larval counterparts in the brain and thoracic segments owing to the period of quiescence at the embryo-larval transition, and owing to dramatic morphological changes of the CNS that occur at late embryogenesis. Recent work has revealed the embryonic origin of some larval neuroblasts: the four mushroom body neuroblasts in the central brain and about 20 neuroblasts in thoracic segments. This study used molecular markers and clonal analysis to identify all eight known type II neuroblasts in each brain lobe and show they all form during embryogenesis, perhaps the last-born central brain neuroblasts. It was not possible to identify each neuroblast individually, however, owing to their tight clustering, movements of the brain lobes, and the lack of markers for specific type II neuroblasts (Walsh, 2017).

The single previously reported embryonic type II neuroblast formed from PntP1+ neuroectodermal cells with apical constrictions called a placode. This study did not investigate this neuroectodermal origin of type II neuroblasts in much detail, but multiple type II neuroblasts were seen developing from PntP1+ neuroectoderm. In the future, it would be interesting to determine whether all type II neuroblasts arise from PntP1+ neuroectoderm or from neuroectodermal placodes. Interestingly, one distinguishing molecular attribute of type II neuroblasts is PntP1, which is not detected in type I neuroblasts. Thus, a candidate for distinguishing type I/type II neuroblast identity is EGF signaling, which can be detected in the three head placodes and is required for PntP1 expression. Clearly, there are more PntP1+ neuroectodermal cells than there are type II neuroblasts, and expression of an EGF negative regulator such as Argos might be necessary to divert some of these neuroectodermal cells away from type II neuroblast specification. The earliest steps of type II neuroblast formation represent an interesting spatial patterning question for future studies (Walsh, 2017).

Now that the embryonic type II neuroblasts have been identified, it is worth considering whether there are differences between embryonic and larval type II neuroblasts or their INP progeny. To date, molecular markers do not reveal any differences between embryonic and larval type II neuroblasts, with the exception that embryonic neuroblasts transiently express the temporal transcription factor Pdm. Interestingly, type I embryonic neuroblasts require Cas to close the Pdm expression window, whereas this study found that cas mutants do not exhibit extension of the Pdm expression window in the earliest-born type II neuroblast or de novo expression of Pdm in the later-forming neuroblasts. Are there differences between embryonic and larval INPs? Larval INPs mature over a period of 6 h and then divide four to six times with a cell cycle of about 1 h. In contrast, embryonic INPs might have a more rapid maturation because Elav+ neurons were seen within 9D11+ INP lineages by stage 14, just 3 h after the first type II neuroblast forms. This study found that INPs undergo quiescence at the embryo-larval transition, as shown by the pools of INPs at stage 16 that do not stain for the mitotic marker pH3. The fate of these quiescent INPs -- whether they resume proliferation, differentiate or die -- remains to be determined (Walsh, 2017).

Neuroblasts in the embryonic ventral nerve cord use the temporal transcription factor cascade Hb>Krüppel>Pdm>Cas>Grh to generate neural diversity. This study shows that the type II neuroblasts are among the last neuroblasts to form in the embryonic brain, and that they sequentially express only the late temporal transcription factors Pdm (in the earliest-forming neuroblast) followed by Cas and grh (in all eight type II neuroblasts). It is unknown why most type II neuroblasts skip the early Hb>Krüppel>Pdm temporal transcription factors; perhaps it is due to their late time of formation, although several earlier-forming thoracic neuroblasts also skip Hb (NB3-3), Hb>Krüppel (NB5-5), or Hb>Krüppel>Pdm. This is another interesting spatial patterning question for the future. Furthermore, misexpression of the early factors (Hb and Krüppel) would be unlikely to affect the progeny produced by type II NBs during embryogenesis, as the competence window for Hb (i.e., the stage at which neuroblasts are responsive to Hb expression) closes with the loss of Dan/Danr expression in all neuroblasts at stage 12. Thus, most embryonic type II neuroblasts form after closing of the Hb competence window and would probably be unresponsive (Walsh, 2017).

Type I neuroblasts show persistent expression of the temporal transcription factors within neurons born during each window of expression (i.e. a Hb+ neuroblast divides to produce a Hb+ GMC which makes Hb+ neurons). In contrast, this study found that type II lineages do not show persistent Cas or grh expression in INPs born during each expression window, but do contain some Cas+ neurons. Both Cas and grh transcription factors can be seen in INPs immediately adjacent to the parental neuroblast, consistent with transient perdurance from the parental neuroblast, but they are typically lacking in INPs more distant. The function of Pdm, Cas and grh in embryonic type II neuroblasts awaits identification of specific markers for neural progeny born during each expression window (Walsh, 2017).

During larval neurogenesis, virtually all INPs sequentially express the temporal transcription factors Dichaete>Grh>Ey. In contrast, embryonic INPs express only Dichaete. These data, together with the short time frame of embryogenesis, suggest that INP quiescence occurs during the Dichaete window, preventing expression of the later Grh>Ey cascade. Interestingly, INPs in the posterior cluster (presumptive DL1 and DL2 type II neuroblast progeny) completely lack Dichaete; this is similar to the DL1 and DL2 larval lineages, which also do not express Dichaete. It is possible that DL1/DL2 neuroblasts make INPs that generate identical progeny (and thus do not require an INP temporal cascade), or perhaps these two neuroblasts use a novel temporal cascade in both embryonic and larval stages (Walsh, 2017).

Larval type II neuroblasts produce many intrinsic neurons of the adult central complex. This study shows that embryonic INPs also produce neurons that contribute to the adult central complex. The data show ~54 neurons (64 minus 10 due to 'leaky' expression) born from embryonic-born INPs survive to adulthood and innervate the central complex. It is likely that this is an underestimate, however, because (1) 9D11-gal4 expression is lacking from a few INPs in the embryonic brain and (2) the time to achieve sufficient FLP protein levels to achieve immortalization could miss the earliest born neurons. The variation in immortalization of the widefield ellipsoid body neuron might represent a neuron born early in the type II lineages, thus unlabeled in a subset of embryos. Additionally, some embryonic-born neurons might perform important functions in the larval/pupal stages but die prior to eclosion (Walsh, 2017).

How the highly stereotyped morphologies of individual neurons are genetically specified is not well understood. This study identified six transcription factors (TFs; Ems, Zfh1, Pb, Zfh2, Pros and Toy) expressed in a combinatorial manner in seven post-mitotic adult leg motor neurons (MNs) that are derived from a single neuroblast in Drosophila. Unlike TFs expressed in mitotically active neuroblasts, these TFs do not regulate each other's expression. Removing the activity of a single TF resulted in specific morphological defects, including muscle targeting and dendritic arborization, and in a highly specific walking defect in adult flies. In contrast, when the expression of multiple TFs was modified, nearly complete transformations in MN morphologies were generated. These results show that the morphological characteristics of a single neuron are dictated by a combinatorial code of morphology TFs (mTFs). mTFs function at a previously unidentified regulatory tier downstream of factors acting in the NB but independently of factors that act in terminally differentiated neurons (Enriquez, 2015).

Neurons are the most morphologically diverse cell types in the animal kingdom, providing animals with the means to sense their environment and move in response. In Drosophila, neurons are generated by neuroblasts (NBs), specialized stem cells dedicated to the generation of neurons and glia. As they divide, NBs express a temporal sequence of transcription factors (TFs) that contribute to the generation of neuronal diversity. For example, in the embryonic ventral nerve cord (VNC), most NBs express a sequence of five TFs (Hunchback, Krüppel, Pdm1/Pdm2, Castor, and Grainyhead), while in medulla NBs and intermediate neural progenitors of the Drosophila larval brain a different series of TFs have been described. In vertebrates, analogous strategies are probably used by neural stem cells, e.g., in the cerebral cortex and retina, suggesting that this regulatory logic is evolutionarily conserved. Nevertheless, although temporally expressed NB TFs play an important role in generating diversity, this strategy cannot be sufficient to explain the vast array of morphologically distinct neurons present in nervous systems. For example, in the Drosophila optic lobe there is estimated to be ~40,000 neurons, classified into ~70 morphologically distinct types, each making unique connections within the fly's visual circuitry neurons (Enriquez, 2015).

A second class of TFs has been proposed to specify subtypes of neurons. For example, in the vertebrate spinal cord, all motor neurons (MNs) express a common set of TFs at the progenitor stage (Olig2, Nkx6.1/6.2, and Pax6) and a different set of TFs after they become post-mitotic (Hb9, Islet1/2, and Lhx3). Hox6 at brachial and Hox10 at lumbar levels further distinguish MNs that target muscles in the limbs instead of body wall muscles. Subsequently, limb-targeting MNs are further refined into pools, where all MNs in a single pool target the same muscle. Each pool is molecularly defined by the expression of pool-specific TFs, including a unique combination of Hox TFs. In Drosophila embryos, subclasses of MNs are also specified by unique combinations of TFs: evenskipped (eve) and grain are expressed in six MNs that target dorsal body wall, and Hb9, Nkx6, Islet, Lim3, and Olig2 are required for ventral-targeting MNs. However, each neuronal subtype defined by these TFs includes multiple morphologically distinct neurons, leaving open the question of how individual neuronal morphologies are specified neurons (Enriquez, 2015).

A third class of TFs suggested to be important for neuronal identity is encoded by terminal selector genes. Initially defined in C. elegans, these factors maintain a neuron's terminally differentiated characteristics by, for example, regulating genes required for the production of a particular neurotransmitter or neuropeptide. Consequently, these TFs must be expressed throughout the lifetime of a terminally differentiated neuron. Notably, as with neurons that are from the same subtype, neurons that share terminal characteristics, and are therefore likely to share the same terminal selector TFs, can have distinct morphological identities. For example, in C. elegans two terminal selector TFs, Mec-3 and Unc-86, function together to maintain the expression of genes required for a mechanosensory fate in six morphologically distinct touch sensitive neurons neurons (Enriquez, 2015).

In contrast to the logic revealed by these three classes of TFs, very little is known about how individual neurons, each with their own stereotyped dendritic arbors and synaptic targets, obtain their specific morphological characteristics. This paper addresses this question by focusing on how individual MNs that target the adult legs of Drosophila obtain their morphological identities. The adult leg MNs of Drosophila offer several advantages for understanding the genetic specification of neuronal morphology. For one, all 11 NB lineages that generate the ~50 leg-targeting MNs in each hemisegment have been defined. More than two-thirds of these MNs are derived from only two lineages, Lin A (also called Lin 15) and Lin B (also called Lin 24), which produce 28 and 7 MNs, respectively, during the second and third larval stages. Second, each leg-targeting MN has been morphologically characterized-both dendrites and axons-at the single-cell level. In the adult VNC, the leg MN cell bodies in each thoracic hemisegment (T1, T2, and T3) are clustered together. Each MN extends a highly stereotyped array of dendrites into a dense neuropil within the VNC and a single axon into the ipsilateral leg, where it forms synapses onto one of 14 muscles in one of four leg segments: coxa (Co), trochanter (Tr), femur (Fe), and tibia (Ti). Not only does each MN target a specific region of a muscle, the pattern of dendritic arbors of each MN is also stereotyped and correlates with axon targeting. The tight correlation between axon targeting and dendritic morphology has been referred to as a myotopic map. The stereotyped morphology exhibited by each MN suggests that it is under precise genetic control that is essential to its function neurons (Enriquez, 2015).

This study demonstrates that individual post-mitotic MNs express a unique combination of TFs that endows them with their specific morphological properties. Focus was placed on Lin B, which generates seven MNs, and six TFs were identified that can account for most of the morphological diversity within this lineage. Interestingly, these TFs do not cross-regulate each other and are not required for other attributes of MN identity, such as their choice of neurotransmitter (glutamine) or whether their axons target muscles in the periphery, i.e., they remain terminally differentiated leg motor neurons. Consistent with the existence of a combinatorial code, when two or three, but not individual, TFs were simultaneously manipulated nearly complete transformations in morphology were observed. However, removing the function of a single TF, which is expressed in only three Lin B MNs, resulted in a highly specific walking defect that suggests a dedicated role for these neurons in fast walking. Together, these findings reveal the existence of a regulatory step downstream of temporal NB factors in which combinations of morphology TFs (mTFs) control individual neuron morphologies, while leaving other terminal characteristics of neuronal identity unaffected neurons (Enriquez, 2015).

Inherent in the concept of a combinatorial TF code is the idea that removing or ectopically expressing a single TF will only generate a transformation of fate when a different wild-type code is generated. Consistent with this notion, only when the expression of two or three mTFs were simultaneously manipulated was it possible to partially mimic a distinct mTF code and, as a result, transform the identity of one Lin B MN into another. In contrast, manipulating single TFs typically resulted in aberrant or neo-codes that are not observed in wild-type flies. For example, removing pb function from Lin B resulted in two MNs with a code (Ems+Zfh1) and MN morphology that are not observed in wild-type Lin A and Lin B lineages. Analogously, ectopic Pb expression in Lin A, which normally does not express this TF, generated aberrant codes and MN morphologies. This latter experiment was particularly informative because although Pb redirected a subset of Lin A dendrites to grow in an anterior region of the neuropil, it did not alter the ability of these dendrites to cross the midline. Thus, the dendrites of these MNs had characteristics of both Pb-expressing Lin B MNs (occupying an antero-ventral region) and Pb-non-expressing Lin A MNs (competence to cross the midline). Axon targeting of these MNs was also aberrant: although they still targeted leg muscles, Pb-expressing Lin A MNs frequently terminated in the coxa, which is not a normal characteristic of Pb-expressing Lin B MNs or of any Lin A MN. These observations suggest that the final morphological identity of a neuron is a consequence of multiple TFs executing functions that comprise a complete morphological signature. Some functions, such as the ability to occupy the antero-ventral region of the neuropil, can be directed by a single TF (e.g., Pb), while other functions, such as the ability to accurately target the distal femur, require multiple TFs (e.g., Pb+Ems). Further, because it was possible to generate MNs that have both Lin B and Lin A morphological characteristics, hte results argue against the idea that there are lineage-specific mTFs shared by all progeny derived from the same lineage. Instead, the data are more consistent with the idea that the final morphological identity of an MN depends on its mTF code neurons (Enriquez, 2015).

Drosophila NBs, and perhaps vertebrate neural stem cells, express a series of TFs that change over time and have therefore been referred to as temporal TFs. For Lin B, the sequence of these factors is unknown, in part because the Lin B NB is not easily identified in the second-instar larval VNC, the time at which it is generating MNs. Nevertheless, each MN derived from Lin B and Lin A has a stereotyped birth order, consistent with the idea that temporal TFs play an important role in directing the identities of MNs derived from these lineages and, therefore, the mTFs they express. For Lin B, this birth order is Co1->Tr1->Fe1->Tr2->Co2->Co3->Co4. Interestingly, according to the mTF code proposed in this study, each of these MNs differs by at most two mTFs in any successive step. For example, Tr1 has the code [Zfh1, Ems, Pb, Zfh2] while Fe1, the next MN to be born, has the code [Zfh1, Ems, Pb]. Thus, it is posited that the sequence of temporal TFs acting in the NB is responsible for directing each successive change in mTF expression in postmitotic MNs (e.g., in the Tr1->Fe1 step, repression of zfh2). Although a link between temporal TFs and TFs expressed in postmitotic neurons has been proposed in Drosophila, the role of these TFs in conferring neuron morphologies is not known. Further, there may be additional diversity-generating mechanisms in lineages that produce many more neurons than the seven MNs generated by Lin B. One additional source of diversity may come from NB identity TFs, which distinguish lineages based on their position. Such spatial information could in principle allow the same temporal TFs to regulate different sets of mTFs in different NB lineages. It is also likely that differences in the levels of some mTFs may contribute to neuronal identities. Consistent with this idea, the levels of Zfh2 and Pros differ in the Lin B MNs expressing these TFs, differences that are consistent in all three thoracic segments and between animals. Further, Zfh1 levels vary between Lin B MNs and its levels control the amount of terminal axon branching. Previous studies also demonstrated that TF levels are important for neuron morphology, including Antp in adult leg MNs derived from Lin A and Cut in the control of dendritic arborization complexity in multidendritic neurons. If the levels of mTFs are important, it may provide a partial explanation for why the transformations of morphological identity generated in this study with the MARCM technique, which cannot control levels, are typically only partially penetrant neurons (Enriquez, 2015).

Another distinction between temporal TFs and mTFs is that no evidence has been observed of cross-regulation between mTFs. In situations when mTFs were either removed (e.g., pb^-/-; ems^RNAi) or ectopically expressed (e.g., UAS-pb + UAS-ems) in postmitotic Lin B MARCM clones, the expression of the remaining mTFs was unchanged. In contrast, when an NB lineage is mutant for a temporal TF, the prior TF in the series typically continues to be expressed. These observations suggest that the choice of mTF expression is made in the NB and that once the postmitotic code is established, it is not further influenced by coexpressed mTFs neurons (Enriquez, 2015).

Spatial patterning specifies neural progenitor identity, with further diversity generated by temporal patterning within individual progenitor lineages. In vertebrates, these mechanisms generate 'cardinal classes' of neurons that share a transcription factor identity and common morphology. In Drosophila, two cardinal classes are Even-skipped (Eve)(+) motor neurons projecting to dorsal longitudinal muscles, and Nkx6(+) motor neurons projecting to ventral oblique muscles. Cross-repressive interactions prevent stable double-positive motor neurons. The Drosophila neuroblast 7-1 (NB7-1) lineage uses a temporal transcription factor cascade to generate five distinct Eve(+) motor neurons; the origin and development of Nkx6(+) motor neurons remains unclear. This study used a neuroblast specific Gal4 line, sparse labelling and molecular markers to identify an Nkx6(+) VO motor neuron produced by the NB7-1 lineage. Lineage analysis to birth-date the VO motor neuron to the Kr(+) Pdm(+) neuroblast temporal identity window. Gain- and loss-of-function strategies to test the role of Kr(+) Pdm(+) temporal identity and the Nkx6 transcription factor in specifying VO neuron identity. Lineage analysis identifies an Nkx6(+) neuron born from the Kr(+) Pdm(+) temporal identity window in the NB7-1 lineage, resulting in alternation of cardinal motor neuron subtypes within this lineage (Eve>Nkx6>Eve). Co-overexpression of Kr/Pdm generates ectopic VO motor neurons within the NB7-1 lineage - the first evidence that this TTF combination specifies neuronal identity. Moreover, the Kr/Pdm combination promotes Nkx6 expression, which itself is necessary and sufficient for motor neuron targeting to ventral oblique muscles, thereby revealing a molecular specification pathway from temporal patterning to cardinal transcription factor expression to motor neuron target selection. Thus this study shows that one neuroblast lineage generates interleaved cardinal motor neurons fates; that the Kr/Pdm TTFs form a novel temporal identity window that promotes expression of Nkx6; and that the Kr/Pdm > Nkx6 pathway is necessary and sufficient to promote VO motor neuron targeting to the correct ventral muscle group (Seroka, 2020).

Neural diversity from flies to mice arises from two major developmental mechanisms. First, neural progenitors acquire a unique and heritable spatial identity based on their position along the rostrocaudal or dorsoventral body axes. Second, temporal patterning based on neuronal birth-order results in individual progenitors producing a diverse array of neurons and glia. Temporal patterning is best characterized in Drosophila; neural progenitors (neuroblasts) located in the ventral nerve cord, central brain, and optic lobes all undergo temporal patterning, in which the neuroblast sequentially expresses a cascade of TTFs that specify distinct neuronal identities. Although all neuroblasts undergo temporal patterning, the TTFs are different in each region of the brain. Similar mechanisms are used in the mammalian cortex, retina, and spinal cord, although many TTFs remain to be identified (Seroka, 2020).

A major open question is how transient expression of TTFs like Kr and Pdm lead to long-lasting specification of molecular and morphological neuronal diversity. Good candidates for integrating spatial and temporal cues to consolidate motor neuron identity are homeodomain transcription factors expressed in post-mitotic motor neurons. In vertebrates, dorsoventral domains of the spinal cord are partitioned into 12 distinct cardinal classes of neurons - each characterized by development from a common progenitor domain, expression of unique homeodomain transcription factors with cross-repressive interactions to stabilize boundaries, and generating neurons with common morphology. This nomenclature is adapted to define Eve+ and Nkx6+ (Flybase: HGTX) motor neurons as two 'cardinal classes' of motor neurons: each class expresses a homeodomain transcription factor (Eve or Nkx6) with cross-repressive interactions, and each class consists of motor neurons with related neuronal morphology (Eve+ motor neurons project to dorsal and lateral longitudinal muscles; Nkx6+ motor neurons project to ventral muscle groups) (Seroka, 2020).

The Drosophila neuroblast 7-1 (NB7-1) is arguably the best characterized system for understanding TTF expression and function. Similar to most other ventral nerve cord neuroblasts, NB7-1 expresses the canonical TTF cascade Hb-Kr-Pdm-Cas with each TTF inherited by the GMCs born during an expression window, and transiently maintained in the two post-mitotic neurons produced by each GMC. The TTF cascade generates diversity among the five Eve+ U1-U5 motor neuron progeny of NB7-1: Hb specifies U1 and U2, Kr specifies U3, Pdm specifies U4, and Pdm/Cas together specify U5. Identifying TTF target genes, including transcription factors and cell surface molecules, will provide a comprehensive view of how developmental determinants direct neuronal morphology and synaptic partner choices (Seroka, 2020).

It has long been thought that the cardinal classes of motor neurons derive from distinct progenitors; Eve+ motor neurons derive from NB7-1, NB1-1, and NB4-2 whereas Hb9+ or Nkx6+ motor neurons derive from NB3-1 and others. However, DiI labeling of NB7-1 identified a potentially unknown motor neuron innervating ventral muscles, which is distinct from dorsal and lateral longitudinal muscles targeted by the Eve+ motor neurons. The observed ventral projection in this lineage could reflect transient exuberant outgrowths that are lost during larval life, or they could be due to an uncharacterized motor neuron that forms stable synapses with ventral muscles (Seroka, 2020).

This study shows that a newly discovered Kr/Pdm TTF window generates an Nkx6+ Eve-motor neuron, born between U3 and U4 in the NB7-1 lineage, that projects to ventral oblique (VO) muscles. It was also shown that overexpression of Kr/Pdm together, or Nkx6 alone, generates ectopic VO motor neurons based on molecular marker expression. Finally, this study demonstrates that Nkx6 is required for proper motor neuron axon targeting to ventral oblique muscles. These results establish a genetic pathway from TTFs (Kr/Pdm), to a cardinal motor neuron transcription factor (Nkx6) to motor axonal targeting. Also the unexpected discovery was made that a single progenitor can alternate production of different cardinal motor neuron classes (Seroka, 2020).

Kr/Pdm co-expression has been detected in several neuroblast lineages, but until now there has not been evidence that this TTF combination could specify neuronal identity. Previous work showed that the Kr/Pdm window generates a Kr/Pdm GMC, and this study shows that this GMC generates an Nkx6+ ventral-projecting motor neuron. It is unknown whether Kr/Pdm directly or indirectly activate nkx6 expression. The nkx6 gene lies in a 45kb region devoid of genes, and there are only a few, sparse predicted Kr or Pdm binding sites in this genomic expanse. How do Kr and Pdm together specify one fate (Nkx6+ VO motor neuron) whereas Kr or Pdm alone specify completely different fates (Eve+ dorsal motor neurons)? It is likely that Kr/Pdm together activate a different suite of target genes than either alone. For example, Kr/Pdm together may directly activate nkx6 expression, whereas neither alone has that potential. The emergence of single cell transcriptome and ChIP studies will help to reveal how the combination of Kr/Pdm TTFs generates different cell fate output compared to Kr or Pdm alone (Seroka, 2020).

The production of an Nkx6+ VO motor neuron in Kr/Pdm window interrupts the sequential production of Eve+ dorsal motor neurons in the NB7-1 lineage, resulting in an Eve > Nkx6 > Eve alternation of cardinal motor neuron production within the lineage. This is unusual, as in most cases neurons with similar morphology or function are produced together in a lineage. In mammals, progenitors generate neurons first, followed by glia; no examples are known of neuron>glia>neuron production from a single lineage. Similarly, Drosophila central brain neuroblast lineages produce the mushroom body γ neurons, then α'/ β' neurons, and lastly α/β neurons, with no evidence for alternating or interspersed fates. In the abdominal NB3-3 lineage, the early-born cells are in a mechanosensitive circuit, whereas the late-born cells are in a proprioceptive circuit. The only possible example of interleaved production of two morphological classes of neurons is in the Drosophila lateral antennal lobe neuroblast lineage, which alternate between uniglomerular and multiglomerular (AMMC) projection neurons. The use of clonal and temporal labeling tools will be needed to examine additional lineages to determine the prevalence of lineages producing temporally interleaved neuronal subtypes as in the NB7-1 lineage (Seroka, 2020).

Overexpression of Kr/Pdm or Nkx6 can induce only 2-3 ectopic VO motor neurons within the NB7-1 lineage. Clearly not all neurons in the lineage are competent to respond to these transcription factors. Early-born temporal identities specified by Hb and Kr (U1-U3) are unaffected by Kr/Pdm or Nkx6 overexpression, which is similar to previous data showing that early temporal fates are not affected by overexpression of later TTFs in multiple lineages. It remains a puzzle why the Kr+ U3 neuron does not switch to a VO fate upon overexpression of Kr/Pdm. There may need to be an equal level of Kr and Pdm to specify VO fate, although this would not explain why Kr/Pdm overexpression converts the Pdm+ U4 motor neuron to a VO fate. Alternatively, there may be an early chromatin landscape that blocks access to relevant Pdm target loci (Seroka, 2020).

It is noted that the assay of VO neuronal identity was done in newly-hatched larvae. Although motor circuits are functional at this time, larvae grow for five more days. These is no data on whether the ectopic VO motor neurons are functional or are maintained through the life of the larvae. This would be an important question for the future (Seroka, 2020).

Nkx6 and Eve have cross-repressive interactions (Broihier, 2004), but with some limitations: early-born Eve+ motor neurons are not affected by Nkx6 overexpression. Wild type animals even show sporadic expression of Nkx6 in the Eve+ U2 motor neuron, but in these neurons it has no effect on Eve expression, nor does it promote targeting to ventral oblique muscles. There appears to be a mechanism to block endogenous or overexpressed Nkx6 function in the early lineage of neuroblasts producing Eve+ motor neurons. The mechanism 'protecting' early-born Eve+ neurons from Nkx6 repression of Eve is unknown. Early lineages may lack an Nkx6 cofactor; Nkx6 could act indirectly via an intermediate transcription factor missing in early lineages; the early TTFs Hb or Kr may block Nkx6 function; or the eve locus could be in a subnuclear domain inaccessible to Nkx6 (Seroka, 2020).

Nkx6 promotes motor neuron specification in both Drosophila and vertebrates. In Drosophila, loss of Nkx6 reduces ventral projecting motor neuron numbers and increases the number of Eve+ neurons, while overexpression increases ventral projecting motor neuron numbers at the expense of Eve+ neurons. In vertebrates the Nkx6 family members Nkx6.1/Nkx6.2 appear to play a broader role in motor neuron specification. Nkx6.1/Nkx6.2 show early expression throughout the pMN domain; mice mutant for both Nkx6 family members lack most somatic motor neurons; and Nkx6.1 overexpression in chick or zebrafish can induce ectopic motor neurons. It would be interesting to investigate whether vertebrate Nkx6.1/Nkx6.2 are required to suppress a specific motor neuron identity, similar to the antagonistic relationship between Nkx6 and Eve in Drosophila (Seroka, 2020).

Neuroblasts in all regions of the Drosophila CNS (brain, ventral nerve cord, optic lobe) use TTF cascades to generate neuronal diversity, yet less is known about TTF target genes. It is likely that TTFs induce expression of suites of transcription factors that persist in neurons and confer their identity. Examples may include the 'morphology transcription factors' that specify adult leg motor neuron dendrite projections, but in this case it remains unknown whether these transcription factors control all other aspects of adult motor neuron identity. It is possible that 'morphology transcription factors' are one module downstream of a broader regulatory tier similar to the terminal selector genes in C. elegans (Seroka, 2020).

This study has identified a linear pathway from Kr/Pdm to Nkx6 which specifies VO motor neuron identity. TTFs could act by two non-mutually exclusive mechanisms: inducing a stable combinatorial codes of transcription factors that consolidate neuronal identity, or by altering the chromatin landscape to have a heritable, long lasting effect on motor neuron gene expression. The observation that Nkx6 is maintained in the VO neuron after fading of Kr/Pdm expression supports the former mechanism. Identification of Kr/Pdm or Nkx6 target genes would give a more comprehensive understanding of TTF specification of neuronal identity (Seroka, 2020).