engrailed


Effects of Mutation or Deletion (part 1/2)

In the trunk of the Drosophila embryo, the segment polarity genes are initially activated by the pair-rule genes; later, the segment polarity genes maintain one another's expression through a complex network of cross-regulatory interactions. These interactions, which are critical to cell fate specification, are similar in each of the trunk segments. To determine whether segment polarity gene expression is established differently outside the trunk, the regulation of the genes hedgehog (hh), wingless (wg), and engrailed (en) was studied in each of the segments of the developing head. The cross-regulatory relationships among these genes, as well as their initial mode of activation in the anterior head are significantly different from those in the trunk. In addition, each head segment exhibits a unique network of segment polarity gene interactions. It is proposed that these segment-specific interactions evolved to specify the high degree of structural diversity required for head morphogenesis (Gallitano-Mendel, 1997).

The proposed interactions betweeh hh, wg and en are described below.

1. The intercalary segment. In this cephalic segment, hh expression is en-independent. In addition, ptc mutations cause the loss of wg rather than ectopic wg expression The dependence of wg, en, and hh expression on ptc indicates a unique role for segment polarity genes in the intercalary segment. Unlike wg action in the trunk and gnathal segments, wg restricts rather than maintains en and hh expression in this segment. Finally, en expression, as it occurs in the trunk, depends on hh function. However, this dependence cannot be mediated through wg, since wg does not maintain en expression in the intercalary segment.

2. The antennal segment. As in the trunk, hh antennal expression depends on en, while wg expression requires hh. The requirement for hh is presumably mediated through ptc, which represses wg in this segment. Unlike in the trunk, wg restricts the expression domains of both en and hh. As in the intercalary segment, regulation of en by hh is wg- independent.

3. The ocular segment. In this segment, hh is en-independent and wg expression does not require hh. Although the wg domain (the head blob) does not expand in ptc mutant embryos, noncontiguous ectopic wg expression appears in its vicinity. Unlike its action in the trunk and the other head segments, wg is required to initiate en expression in the ocular segment. However, hh expression still expands in wg mutant embryos (as in the intercalary and antennal segment). As in the intercalary and antennal segments, regulation of en by hh does not depend on wg.

It is concluded that cross-regulatory interactions among the segment polarity genes in the anterior head are very different from those in the posterior head and trunk segments. The mode of patterning of the anterior head (the acron and cephalic segments) is thought to be more ancient than that of the posterior head (the gnathal segments). This distinction appears to be reflected in the segmentation mechanism used by certain present day short germ insects and primitive arthropods. In these organisms, the early germ band includes only the acron, cephalic segments, and tail. Gnathal and trunk segments are generated later in embryogenesis by a progressive budding process (Gallitano-Mendel, 1997).

The somatic muscles, the heart, the fat body, the somatic part of the gonad and most of the visceral muscles are derived from a series of segmentally repeated primordia in the Drosophila mesoderm. This work describes the early development of the fat body and its relationship to the gonadal mesoderm, as well as the genetic control of the development of these tissues. The first sign of fat body development is the expression of serpent in segmentally repeated clusters within the trunk mesoderm in parasegments 4-9. Segmentation and dorsoventral patterning genes define three regions in each parasegment in which fat body precursors can develop. The primary and secondary dorsolateral fat body primordia are formed ventral to the visceral muscle primoridium in each parasegment. The ventral secondary cluster forms more ventrally in the posterior portion of each parasegment. Fat body progenitors in these regions are specified by different genetic pathways. Two dorsolateral regions require engrailed and hedgehog (within the even-skipped domain) for their development while the ventral secondary cluster is controlled by wingless. Ubiquitous mesodermal en expression leads to an expansion of the primary clusters into the sloppy-paired domain, resulting in a continuous band of serpent-expressing cells in parasegments 4-9. The observed effect of en on fat body development is seen not only on mesodermal overexpression but also when en is overexpressed in the ectoderm. Loss of wingless leads to an expansion of the dorsolateral fat body primordium. decapentaplegic and one or more unknown genes determine the dorsoventral extent of these regions. High levels of Dpp repress serpent, resulting in the formation of visceral musculature, an alternative cell fate (Reichmann, 1998).

In each of parasegments 10-12 one of these primary dorsolateral regions generates somatic gonadal precursors instead of fat body. The balance between fat body and somatic gonadal fate in these serially homologous cell clusters is controlled by at least five genes. A model is suggested in which tinman, engrailed and wingless are necessary to permit somatic gonadal develoment, while serpent counteracts the effects of these genes and promotes fat body development. In wg mutant embryos, all dorsolateral mesodermal cells, including those in parasegments 10-12, acquire fat body fate. This phenotype can be interpreted as the combined effects of two separate functions of wg: (1) wg is necessary to repress fat body development in the dorsolateral mesoderm underlying the wg domain in all parasegments; (2) wg is required in the primary cluster to permit somatic gonadal precursor instead of fat body development in parasegments 10-12. Loss of engrailed results in the absence of demonstrable somatic gonadal precursors, similar to the situation in tinman mutants. Ubiquitous mesodermal en expression leads to the formation of additional somatic gonadal precursor cells in parasegments 10-12. The homeotic gene abdominalA limits the region of serpent activity by interfering in a mutually repressive feed back loop between gonadal and fat body development. It is unlikely that abdA represses srp directly, since srp can be expressed in cells in which abdA is active. abdA might prevent srp from inhibition of a somatic gonadal precursor competence factor (Riechmann, 1998).

Removing en activity causes incomplete morphological transformation from posterior to anterior fate in the wing, and fails to produce an ectopic anterior-posterior organizer. Complete transformation can only be effected by simultaneously eliminating activity of en and its homolog invected. Invected functions principally to specify posterior cell fate. Thus establishment of the anterior-posterior organizer and control of compartment identity are genetically distinguishable, and invected may perform a discrete subset of functions previously ascribed to en (Simonds, 1995).

Removing engrailed and invected from posterior wing cells created two new compartments: an anterior compartment consisting of mutant cells and a posterior compartment that grows from neighboring cells. In some cases, these compartments formed a complete new wing resulting from a duplication of anterior and posterior compartments. Increasing engrailed activity also affects patterning. Engrailed both directs the posterior compartment pathway and creates the compartment border (Tabata, 1995).

Engrailed and Huckebein are essential for development of serotonin neurons in the Drosophila CNS. en and hkb coexpress uniquely in the serotonin neurons and in neuroblast 7-3 (NB7-3). In the grasshopper, the analogous serotonin neurons originate from the first ganglion mother cell produced from NB7-3. The corresponding NB7-3 in Drosophila can be identified by its time of birth, size, and relative position within each hemisegment. The serotonin neurons can be identified during late embryogenesis by the appearance of DOPA decarboxylase (DDC) immunoreactivity. The DDC enzyme catalyzes the last step in the biosynthesis of serotonin and dopamine and can be used as a marker for both cell types. In the ventral ganglion there are three anatomically distinguished types of DDC immunoreactive cells per segment, a pair of ventrolateral serotonin cells (VL), a single midline dopamine cell (M) and the dorsal lateral (DL) dopamine cells. en and hkb are coexpressed in the VL cells but not the DL or M cells. The high selectivity of coexpression of these two gene products suggests that their combined activities may be important for the development of NB7-3 progeny. Serotonin neuron differentiation is abnormal in en and hkb mutants. Although neither mutant shows a complete loss of DDC immunoreactive serotonin cells, the few escaper serotonin neurons may be due to low levels of functional hkb gene product in a hypomorphic allele. Since NB 7-3 appears normal in hkb mutants, the effect of hkb on development of the serotonin cell lineage must be at a later stage of development, either at division of the neuroblast or ganglion mother cells or on the identity of the GMC progeny (Lundell, 1996). For more information on serotonin and dopamine neurons see Islet and Zn finger homeodomain 1).

Endoreduplication cycles that lead to an increase of DNA ploidy and cell size occur in distinct spatial and temporal patterns during Drosophila development. Only little is known about the regulation of these modified cell cycles. Fore- and hind-gut development have been investigated and evidence is presented that the knirps and knirps-related genes are key components to spatially restrict endoreduplication domains. Lack and gain-of-function experiments show that knirps and knirps-related, which both encode nuclear orphan receptors, transcriptionally repress S-phase genes of the cell cycle required for DNA replication and that this down-regulation is crucial for gut morphogenesis. Furthermore, both genes are activated in overlapping expression domains in the fore- and hind-gut in response to Wingless and Hedgehog activities emanating from epithelial signaling centers that control the regionalization of the gut tube. These results provide a novel link between morphogen-dependent positional information and the spatio-temporal regulation of cell cycle activity in the gut (Fuß, 2001).

The kni and knrl expression domains in the developing foregut and hindgut partially overlap with the expression domains of wingless and hedgehog, which define signaling centers that control morphogenetic movements during the regionalization of the gut. To investigate whether kni/knrl expression and consequently also the restriction of the endoreduplication pattern in the gut is coordinated the Wg and Hh signaling cascades, expression studies in various lack and gain-of-function situations were performed. In hh mutants, kni expression is only mildly reduced in the developing fore- and hind-gut expression domains. In early wg mutants, kni fails to be expressed in the esophagus primordium and is strongly reduced in the developing small intestine and rectum. wg mutant embryos lack a foregut at later stages and have a strongly reduced hindgut. Ectopic expression of hh in all the hindgut cells using the UAS-Hh effector and the 14-3fkh driver line does not alter the kni or knrl expression domains in the hindgut, even when the Hh dose is increased by using effector lines with multiple UAS-Hh transgene insertions. However, if the same experiment is carried out in engrailed mutants, kni/knrl can be induced ectopically in all the hindgut cells. In wild-type embryos, engrailed is expressed in the dorsal part of the large intestine and exerts a repressing function on kni/knrl expression that apparently cannot be overcome by ectopic Hh activity. However, ectopic wg expression in all the hindgut cells using the UAS-Wg effector and the 14-3fkh driver line does result in ubiquitous induction of kni and knrl expression. engrailed expression in the hindgut of these embryos is repressed under these conditions. To investigate whether ectopic Wg expression in the hindgut interferes with DNA replication activity required for endoreduplication, BrdU incorporation was examined. BrdU incorporation is absent in the hindgut of such embryos. Consistent with this result, S-phase genes such as RNR2 are transcriptionally repressed upon ectopic Wg expression in all the hindgut cells using the 14-3fkh-Gal4 driver and UAS-Wg. As has been observed for ectopic kni/knrl expression in the hindgut, the size of the hindgut cells are reduced in these embryos (Fuß, 2001).

The hindgut of the Drosophila embryo is subdivided into three major domains, the small intestine, large intestine, and rectum, each of which is characterized by specific gene expression. The expression of wingless, hedgehog, decapentaplegic, and engrailed corresponds to the generation or growth of particular domains of the hindgut. wg, expressed in the prospective anal pads, is necessary for activation of hh in the adjacent prospective rectum. hh is expressed in the prospective rectum, which forms anterior to the anal pads, and is necessary for the expression of dpp at the posterior end of the adjacent large intestine. wg and hh are also necessary for the development of their own expression domains, anal pads, and rectum, respectively. dpp, in turn, causes the growth of the large intestine, promoting DNA replication. en defines the dorsal domain of the large intestine, repressing dpp in this domain. A one-cell-wide domain, which delineates the anterior and posterior borders of the large intestine and its internal border between the dorsal and ventral domains, is induced by the activity of en. A model is proposed for the gene regulatory pathways leading to the subdivision of the hindgut into domains (Takashima, 2001).

The term 'tissue compartments' can be used to indicate the domains of the gut. In this report, the term 'domain' is used in order to avoid confusion with the term 'developmental compartment', which has been defined by clonal analysis of the wing disc. To clarify the use of anatomical descriptions, the organization of the hindgut domains, as revealed by specific gene expression patterns is described. The most anterior domain of the hindgut, which is just posterior to the midgut, is the small intestine. The small intestine is followed by the large intestine, then the rectum. The large intestine is further subdivided into a ventral and a dorsal domain. A one-cell-wide domain, which was designated as h4, forms at the anterior and posterior borders of the large intestine, as well as at the border between the dorsal and ventral domains of the large intestine. The cells in these regions are designated collectively 'border cells'. Until the end of stage 12, the hindgut tube is situated on the midline of the body, and is left-right symmetric. During early stage 13, the hindgut rotates to the left, resulting in the original dorsal and ventral domains coming to face the left and right side of the body, respectively. The orifice of the rectum (the anal slit) is surrounded by the anal pads, the development of which is tightly linked to that of the hindgut (Takashima, 2001).

dpp is first expressed at early stage 11 as a narrow ring anterior to the prospective rectum. After early stage 12, a weak expression appears in the ventral domain of the large intestine, which partly overlaps the former dpp-positive domain. en, initially expressed throughout the hindgut primordium at stage 9, is soon restricted to the dorsal domain of the large intestine. The en-positive dorsal domain and the dpp-positive ventral domain do not overlap when examined by double staining for En protein and DPP mRNA. The expression of en continues throughout embryogenesis and larval stages (Takashima, 2001).

The border cells differentiate at the anterior and posterior border of the large intestine and at the border between the dorsal and ventral domains of the large intestine. The border cells are first detected at stage 12 by lacZ expression of some enhancer-trap strains, and after stage 14, the cells are distinguished by marked expression of Crb and dead ringer. By double staining for En and beta-galactosidase protein of border cell-specific enhancer trap lines, the border cells are found to abut the En-positive domain and to express no En protein, suggesting that dpp-positive cells abutting the en-positive domain differentiate into border cells. It is noteworthy that the spatial organization of en, hh, wg, and dpp domains is quite different from that of the segmented epidermis or the imaginal discs, suggesting that a different patterning mechanism is working in the hindgut (Takashima, 2001).

dpp is expressed in two overlapping regions of the large intestine; these regions appear to be regulated independently. dpp expression at the posterior end of the large intestine depends on hh activity in the adjacent rectum, whereas the weak expression of dpp in the ventral domain of the large intestine is not affected in the hh mutant. In the dorsal domain of the large intestine, where dpp is not expressed except in the posterior-most portion, en is expressed throughout development. Double staining for En protein and dpp mRNA reveal that the en-domain and the dpp-domain do not overlap. To analyze the regulatory relationship between dpp and en, dpp expression was examined in an en mutant, in which en and its paralog invected (inv) are deficient. Expression of dpp expands to the dorsal domain of the large intestine in the en mutant, but overall morphology of the hindgut is almost normal except for a slight overgrowth. Repression of dpp by en is also demonstrated by ectopic expression of en. When en is expressed throughout the hindgut with the GAL4-UAS system, dpp expression in the hindgut becomes very weak except in the posterior-most portion of the large intestine, where the hh signal from the adjacent rectum activates dpp expression (Takashima, 2001).

The border cells form a one-cell-wide domain that is composed of three portions: an anterior and a posterior ring, and bilateral strands that connect the two rings. The border cells strongly express Crb after stage 14. The border cells abut but do not overlap the En-positive domain. Differentiation of the border cells in en mutant embryos was examined by Crb immuno-staining or by use of border cell-specific enhancer-trap marker strains. Border cells do not differentiate in en mutants, suggesting that en activity is necessary for the differentiation of border cells. A single mutation of either en or inv does not affect the development of border cells, indicating functional redundancy of en and inv genes. When en is ectopically expressed throughout hindgut by byn-GAL4, border cells fail to form except at the posterior border of the large intestine. These results indicate that the interaction of en-positive and en-negative cells is required for the differentiation of border cells. The absence of border cells does not affect the gross morphology of the hindgut (Takashima, 2001).

The Drosophila hindgut develops three morphologically distinct regions along its anteroposterior axis: small intestine, large intestine and rectum. Single-cell rings of 'boundary cells' delimit the large intestine from the small intestine at the anterior, and the rectum at the posterior. The large intestine also forms distinct dorsal and ventral regions; these are separated by two single-cell rows of boundary cells. Boundary cells are distinguished by their elongated morphology, high level of both apical and cytoplasmic Crb protein, and gene expression program. During embryogenesis, the boundary cell rows arise at the juxtaposition of a domain of Engrailed- plus Invected-expressing cells with a domain of Delta (Dl)-expressing cells. Analysis of loss-of-function and ectopic expression phenotypes shows that the domain of Dl-expressing cells is defined by En/Inv repression. Further, Notch pathway signaling, specifically the juxtaposition of Dl-expressing and Dl-non-expressing cells, is required to specify the rows of boundary cells. This Notch-induced cell specification is distinguished by the fact that it does not appear to utilize the ligand Serrate and the modulator Fringe (Iwaki, 2002).

At its anterior, the hindgut joins the posterior midgut; at its posterior, it forms the anus. Along this AP axis, the hindgut of the mature embryo consists of three morphologically distinct domains: the wide, looping small intestine, the long and narrow large intestine, and the tapered rectum. Beginning at stage 13, these domains are demarcated at their junctions by rings of unusually high accumulation of the apical surface protein Crumbs (Crb). The ring at the small intestine/large intestine junction is designated the anterior boundary cell ring, and the ring at the large intestine/rectum junction is designated the posterior boundary cell ring (Iwaki, 2002).

Patterning of the hindgut in the DV axis is detected at stage 10 (germ band extension) when the hindgut develops an interiorly directed (dorsal) convexity. The side of the hindgut closest to the interior of the embryo is dorsal and expresses both En and Inv; that closest to the exterior is ventral and expresses dpp. By the completion of germ band retraction, the convexity at the anterior of the hindgut has shifted toward the left side of the embryo. Thus at the anterior of the hindgut, the initially dorsal, En- and Inv-expressing side comes to lie on the outer (left-facing) curve, while the initially ventral, Dpp-expressing side of the hindgut comes to lie on the inner (right-facing) curve; the DV relationship is retained at the posterior connection to the rectum. These initially DV patterned domains of the large intestine persist to the end of embryogenesis and into the larval stages; they are referred to as large intestine dorsal (li-d) and large intestine ventral (li-v). At each of the two boundaries between li-d and li-v, there is a single row of cells with high levels of Crb expression running the length of the large intestine, from the anterior boundary cell ring to the posterior boundary cell ring. These are designated the 'boundary cell rows'. In addition to their high level of Crb expression, the boundary cell rows and rings express the nuclear protein Dead ringer (Dri). Double antibody staining reveals that boundary cell rows at the border of the En/Inv-expressing li-d domain and the Dpp-expressing li-v domain express Dri in their nuclei and have strong Crb expression at their apical surfaces (Iwaki, 2002).

The boundary cell rows form at the junction of the li-d and li-v domains, which express different genes. To investigate whether the spatially restricted gene expression observed in these domains is essential for establishment of boundary cell rows, embryos homozygous for loss-of-function alleles of en, inv, dpp, dri, Dl, Ser, Notch, or fng were examined. The presence or absence of boundary cells was assessed by anti-Crb staining, since this delineates their characteristic morphology, and also detects one of their unique differentiated features (i.e. the cytoplasmic accumulation of Crb) (Iwaki, 2002).

In embryos lacking only en, the boundary cell rows and rings form normally. Similarly, many embryos lacking only inv form boundary cell rows and rings. In a significant number of inv embryos, however, gaps were observed in the posterior of the boundary cell rows. This is the only embryonic phenotype known for inv. When both en and inv are removed [in Df(enE) embryos], the phenotype is much more dramatic: boundary cell rows and rings are completely absent. Consistent with previous studies demonstrating a functional redundancy of en and inv, it is concluded that en and inv are required largely redundantly to establish the boundary cells. However, while inv can substitute completely for en, there is a requirement for inv that cannot be completely substituted by en. This is likely not due to a difference in protein structure, but rather to the fact that, in the hindgut, inv is expressed earlier and at a higher level than en. As their functions are so closely intertwined, the activities of en and inv, and the highly related proteins that they encode, are referred to as single entities: en/inv and En/Inv (Iwaki, 2002).

Since the experiments described in the preceding sections show that both spatially localized En/Inv and a boundary of Dl expression are required to establish the boundary cells, it was asked whether En/Inv might control the boundary of Dl expression. In Df(enE) embryos, Dl is not restricted to li-v, but rather is uniform in the hindgut circumference, indicating that en/inv is required to repress Dl. In the large intestine, uniform expression of En/Inv results in an absence of Dl expression. Expression of En/Inv in li-d is thus both necessary and sufficient to restrict Dl expression to li-d. While it represses Dl throughout the large intestine, ectopic En/Inv does not affect Dl expression in the rectum. Embryos with ectopic En/Inv not only express Dl at the anterior of the rectum, they also form the posterior boundary cell ring. Thus a boundary of Dl-expressing with Dl-non-expressing cells is required not only to establish the boundary cell rows but also likely to establish the posterior ring; the posterior ring also requires En/Inv activity, but this activity does not need to be localized (Iwaki, 2002).

Consistent with observations that En and Inv are repressors with the same targets, the data presented in this study demonstrate that Dl expression in the large intestine is restricted to the li-v domain by the repressive activity of En/Inv in li-d (Iwaki, 2002).

The data presented here support the following model. En/Inv is expressed in li-d and represses Dl in that domain; Dl expression is thereby restricted to the li-v domain. At the li-v/li-d transition, the Dl-expressing cells induce, by Notch signaling, a row of Dl-non-expressing cells to become a boundary cell row. Since En/Inv is not detected in differentiated boundary cells, Notch activation likely represses En/Inv expression. Notch activation also leads to Dri expression and an upregulation of Crb expression. While all of these transcriptional changes could be mediated by Su(H), they could also be further downstream (Iwaki, 2002).

In summary, three steps in the establishment of the Drosophila hindgut boundary cell rows are similar to steps characterized in other Notch dependent boundary-forming systems. (1) A homeodomain transcription factor (En/Inv in the case of the boundary cells) is expressed on one side of the forming boundary; (2) this transcription factor defines two domains, one which expresses Dl and one which does not; (3) Notch activation in the Dl-non-expressing cells that confront Dl-expressing cells leads to a unique cell fate (Iwaki, 2002).

Given the essential role of spatially restricted En/Inv expression in establishing the boundary cells, it is of interest to consider how En/Inv expression is restricted to the li-d domain. The activation of en expression in the large intestine at stage 10 requires the T-domain transcription factor brachyenteron (byn), which is expressed uniformly in the hindgut. Since dissection of the en regulatory region has identified fragments that drive reporter expression in all hindgut cells, en expression is likely restricted to li-d by a repressor that remains to be identified (Iwaki, 2002).

Boundary cells could be imagined to provide adhesive differences important for cell rearrangement; alternatively, their AP elongation might provide a mechanical force to drive hindgut elongation. In spite of these tempting scenarios, however, the normal appearance (overall size, diameter, and length) of Notch and Df(enE) hindguts, which completely lack both boundary cell rows and rings, demonstrates conclusively that the boundary cell rows and rings are not required to establish normal hindgut morphology (Iwaki, 2002).

A concerted action of Engrailed and Gooseberry-Neuro in neuroblast 6-4 is triggering the formation of embryonic posterior commissure bundles

One challenging question in neurogenesis concerns the identification of cues that trigger axonal growth and pathfinding to form stereotypic neuronal networks during the construction of a nervous system. This study shows that in Drosophila, Engrailed (En) and Gooseberry-Neuro (GsbN) act together as cofactors to build the posterior commissures (PCs), which shapes the ventral nerve cord. Indeed, these two proteins are acting together in axon growth and midline crossing, and that this concerted action occurs at early development, in neuroblasts. More precisely, their expressions in NB 6-4 are necessary and sufficient to trigger the formation of the PCs, demonstrating that segmentation genes such as En and GsbN play a crucial role in the determination of NB 6-4 in a way that will later influence growth and guidance of all the axons that form the PCs. Also, more specific function was demonstrated of GsbN in differentiated neurons, leading to fasciculations between axons, which might be required to obtain PC mature axon bundles (Colomb, 2008).

One of the most fascinating aspects of nervous system development is the establishment of stereotypic neuronal networks. An essential step in this process is the outgrowth and precise navigation of axons. Most CNS growth cones initially head straight towards the midline, and only after crossing, they change their behavior as they turn and follow specific longitudinal pathways. In Drosophila, the majority of axons cross the midline within either anterior or posterior commissures. The formation of commissures starts at stage 12 of embryonic development and involves dynamic, but reproducible interactions between: growth of the neurons, their fasciculation with other neurons to form the different bundles, apoptosis of neuronal cells, and migration of glial cells. In Drosophila, formation of posterior and anterior commissures are not believed to be related, and different cells and possibly different signals appear to be used for the guidance of the different commissures. Each neuron makes a choice as whether to cross the midline and, for those that do cross, whether to grow through the anterior or the posterior commissure, where axons are arranged in fascicles. One central issue is the identification of the intrinsic pathfinding abilities at the different steps of the neural development that are involved in the differential neuronal behavior. Whereas the process of construction of longitudinal tracts has been previously analyzed, as has the formation of ACs, little is known about the formation of the PCs (Colomb, 2008).

Obvious candidates for organizing the intrasegmental distribution of guidance cues along the antero-posterior axis are the segment polarity genes. Consistent with this assumption, embryos mutant for En/Inv and for Gsb/GsbN have severely reduced, and often missing, posterior commissures. Segment polarity genes occupy an intriguing position within the segmentation hierarchy. They are required in the epidermis to specify cell fates within each segment, and are also active both before and during the delamination of neuroblasts to generate the CNS. In particular, the specification of neuroblast identity within a given hemisegment depends upon interactions between segment polarity genes such as Engrailed and Invected with Gsb. Whereas gsb is expressed at early stage 6 and begins to be detectable when NBs start to delaminate (stage 9), GsbN is only detectable starting in stage 10 embryos and appears simultaneously to the disappearance of Gsb. En and GsbN are expressed in NBs of rows 6 and 7. Interestingly, NB 6-4 appears during the S3 wave of delamination at stage 10, just as GsbN expression begins. The axons that pioneer the first tracts will appear later, at stage 12, by which time Gsb expression is nearly completely switched off (Colomb, 2008).

This report has developed several lines of evidence for a concerted action of En and GsbN in neuroblasts. Indeed, it was shown that whereas heterozygous en/inv or gsb/gsbN deletions (respectively (Df enX31/+) and (Df gsbX62 /+)) show a normal architecture of the VNC, double heterozygotes (Df enX31/Df gsbX62) do not form PCs properly, resulting with high penetrance in loss of PCs. This result clearly indicates that En/Inv and Gsb/GsbN act together to form PCs. En has already been shown to have a major function in PC formation, comparatively to Inv. In view of the observation of physical interactions between En protein and GsbN protein, whether GsbN might be responsible for the absence of PCs in the transheterozygous (Df enX31/Df gsbX62) genetic background was analyzed. Using rescue experiments, it was indeed found that expression of GsbN was able to rescue the phenotype. This shows that, genetically, En and GsbN act together to build the posterior PC commissures, which are part of the VNC. There are several reasons to suspect that GsbN might act as a cofactor of En for PC formation. First, no evidence was found for a direct regulation of En on gsbN, since no En binding fragments were isolated within the GsbN locus by chromatin immunoprecipitation. This corroborates the observation that En does not bind the GsbN locus (60F region) on polytene chromosomes. Moreover, it was shown that missing PC phenotype resulting from En misexpression is not associated with a loss of gsbN function. In addition, En and GsbN proteins interact in vitro (as evidenced by GST-pull down and coIP experiments), in yeast (demonstrated using a two-hybrid assay), and in vivo in Drosophila, since they were found to bind common loci on polytene chromosomes. Together, these results support the notion that En and GsbN act as cofactors in the construction of the VNC. Interestingly, it was only possible to rescue the missing PC phenotype of transheterozygous (Df enX31/Df gsbX62) embryos when GsbN was restored from early stages in neuroblasts, but not in differentiated neurons. This shows that the formation of the PCs involves an early function of GsbN, which is consistent with a concerted action with En, since it has been shown that the early function of En is responsible for PC axon growth. It is known from previous studies that PCs are formed from neurons originating in rows 6 and 7 (which express both En and GsbN), as well as from neurons issued in other rows, such as row 5 (that only express GsbN). These observations strongly suggest that NBs expressing both the En and GsbN transcription factors might contain instructions for PC formation. In a first step, En/Inv and Gsb (not GsbN) were shown to be involved in NB specification. In particular expressions of En and Gsb were found to be necessary in the formation of NB 6-4. However, since gsbN is not expressed in the ventral neuroectoderm during the time of NB specification, it hence cannot play a role in neural specification at this level. Therefore, it is expected that the interaction between En and GsbN does not interfere directly in the formation and segregation of the NBs, but rather to happen after the NBs are formed. In particular, this study shows that En and GsbN are involved in the further determination of NB 6-4 to form posterior commissure. Indeed, En and GsbN functions in NB 6-4 not only influence NB 6-4 behavior, but also the behaviors of other neurons that construct the PCs, strongly suggesting that this concerted action of En and GsbN is involved in triggering formation of the PC bundles. One hypothesis is that they act together in a same complex to activate functions that are required for the development of the NBs and that will be necessary for further axon growth and pathfinding. Indeed, driving GsbN in NB 6-4 using different drivers such as eagle-Gal4 or collier-Gal4 was sufficient to rescue axonal growth and crossing of the midline of the PC formers in the transheterozygous (Df enX31/Df gsbX62) background (Colomb, 2008).

However, whereas axon growth and crossing of the midline seem to be rescued in both cases, separation between ACs and PCs were incomplete. One possible explanation for the fusion of the commissures, was provided by the analysis of the neuronal behavior of eagle-positive neurons. When GsbN is expressed in eagle-expressing NBs/neurons (corresponding to NB 6-4 and NB 7-3 progeny projecting through PCs, and to NB 2-4 and NB 3-3 progeny projecting through ACs), a rescue of axonal growth of PCs was observed, but it was also found that neurons projecting through ACs were fasciculating with the PCs. This suggests that the 'fuzzy' separation of ACs and PCs observed with the eagle-Gal4 driver probably resulted from abnormal axonal pathfinding in ACs. Therefore, formation of PC bundles requires at later stages, a specific function of GsbN in the neurons that is driving the fasciculation and guidance of axons forming PC commissures. This latter function of GsbN might correspond to a late function, since expression of GsbN with both early acting (in NBs, neurons, and glial cells) and late acting (in differentiated neuronal cells) Gal4 drivers was found to misroute axons that would have otherwise fasciculated to other axons at the midline. In this case too, all the axons seem to fasciculate, leading to a fuzzy separation of the commissures, sometimes collapsing at the midline. These observations support the idea that this late function of GsbN in the neurons is involved in their axonal pathfinding and in formation of fasciculations that are required to form the bundles, and that are a property of the follower neurons. En function on axonal pathfinding was found to occur early in the neuroblasts, but not in differentiated neurons. Expression of GsbN in differentiated neurons was also not able to trigger axonal growth and crossing of the midline of PC formers. Therefore, a two-step involvement of GsbN occurs in the formation of the PCs. In first, a concerted action of En and GsbN is necessary in NB 6-4 to trigger the axon growth of PC formers, whereas axonal guidance per se might rather result from independent role of En and GsbN, a specific action of GsbN on guidance occurring in differentiated neurons (Colomb, 2008).

Important questions relate to the behavior of NB 6-4 in different genetic contexts and the exact role of En and GsbN in this process. Since NB 6-4 generates both neurons and glial cells, one hypothesis is that they act together in the glial cells that are known to play a crucial role in axonal guidance. Several hypothesis could be drawn: (1) GsbN expression is needed to form NB 6-4 progeny. However, in transheterozygous (Df enX31/Df gsbX62) embryos, eg expressing neuronal cells were formed, but their axons were not growing. As well, it was found that glial cells were formed from NB 6-4, which does not favor this hypothesis. (2) GsbN acts directly on glial cell function. However whereas En is expressed in the glia, it was found that GsbN was not expressed in the glia, which also excludes this hypothesis. (3) En and GsbN are activating a function that will be expressed in NB 6-4 glial progeny and that is triggering axon growth and crossing of the midline, making these particular glial cells central in this process. However ectopic expression of GsbN in all the glia does not lead to abnormal architecture of the VNC, which does not favor for an indirect effect of GsbN in the glia. (4) Finally, functions activated by En and GsbN in NB 6-4 will be used in its neuronal progeny to 'show the way' of GsbN expressing neurons. The data rather favor a central role of NB 6-4 neuronal progeny in triggering the formation of the PCs. This of course does not exclude, as shown for longitudinal tracts, a crucial role between on one hand these particular neurons and the NB 6-4 glial progeny, followed by a crosstalk between these glial cells and the GsbN expressing neurons (Colomb, 2008). The molecular mechanisms involved in these processes will be particularly informative in arriving at an understanding of how neuronal axon trajectories are dictated to construct the VNC. The next challenge will be also to understand what cellular events and downstream functions are regulated in NBs by both En and GsbN to construct PC bundles, since their expression in NB 6-4 seems to be crucial to trigger the whole process of PC formation, and to understand what are the specific downstream functions regulated more specifically by GsbN to specify fasciculations between GsbN expressing neurons, a property associated to the followers (Colomb, 2008).

One way to address these questions would be to identify genes that are directly regulated by En and GsbN and that would therefore likely be misregulated in the transheterozygous (Df enX31/Df gsbX62) genetic background (Colomb, 2008).

Finally, the identification of direct targets of GsbN or of common direct targets of En and GsbN would allow a better understanding of the downstream functions involved in the specification and differentiation of the different neurons, which ultimately drive axon growth and axonal pathfinding (Colomb, 2008).

The observations that vertebrate homologs of En (En1 and En2) and Gsb/GsbN (Pax3 and Pax7), but also other Pax genes are required in neural fate specification, and that they are involved in axon growth, strongly suggests that the molecular mechanisms acting in Drosophila are relevant to and probably conserved in higher organisms (Colomb, 2008).

Continued: engrailed Effects of Mutation part 2/2 |


engrailed: Biological Overview | Evolutionary Homologs | Transcriptional regulation | Targets of activity | Protein Interactions | Developmental Biology | References

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

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