BIOLOGICAL OVERVIEW

vnd/NK2 is the first neural gene expressed. It functions early in neural system development forming a prepattern, even before the proneural genes begin to function. It is required for the formation of a subset of segmental neuroblasts, and possibly as a neuroectodermal committment gene. vnd transcription is upstream of proneural achaete-scute complex. It is responsible for the induction of achaete and scute in a subset of neuroblasts. It also regulates genes of the Enhancer of split complex (Kramatschek, 1994).

It is clear that E(spl)-C gene expression is dependent on lateral inhibition and the Notch pathway acting through Suppressor of Hairless. The role of VND in the transcriptional activation of E(spl)-C genes is currently unclear. Perhaps VND activates proneural genes which in turn activate E(spl)-C genes. Physical interaction of VND with the E(spl)-C has not yet been verified.

The specificity of neuron function in regulating different segmentally repeated muscle groups has its origin in the establishment of a pattern of neuron differentiation in the neuroectoderm. This pattern is determined by a combination of pair rule and segment polarity genes, the former responsible for segmentation and the latter for subdivision of cell fates within segments. The appropriately named dorsoventral polarity genes establish gene expression along the DV axis. By stage eight of embryonic development, each ventral hemisegment (the left or right half of a segment) is subdivided into three longitudinal columns from the ventral midline: medial, intermediate and lateral. They are also divided in the anterior to posterior axis into rows A through D.

Expression of achaete-scute is dependent on VND/NK2 in two proneural clusters of the medial column, rows B and D, while VND/NK2 shows little or no effect on gene expression in rows A and C, and no effect on gene expression in the intermediate and lateral columns. Thus VND/NK2 is required along precise domains of both the anterior-posterior axis and the dorsal-ventral axis. Using Snail or Hunchback proteins as markers, it has been shown that medial neuroblasts from rows B and D are not formed in VND/NK1 mutants. Do other genes specify the remaining neuroblasts? To date these regulators have not been identified (Skeath, 1994).

It is suggested that muscle segment homeobox functions in a similar fashion to vnd/NK2 in neural patterning, and that these roles are conserved in insects and vertebrates. Interaction between DPP and Short gastrulation (vertebrate homologs BMPs and chordin) restrict msh (vertebrate homologs the Msx genes) expression to the lateral most column of proneural clusters (in vertebrates the lateral-most portions of the neural plate). In a similar fashion vnd is restricted to the medial column of proneural clusters. In vertebrates the vnd homolog Nkx-2 is restricted to the medial region of the neural plate. Both msh and vnd serve a similar conserved function, regulation of expression of the achaete-scute complex to particular neuroblasts. Likewise Msx and Nkx-2 regulate the expression of vertebrate achaete-scute homolog (ash) in the developing neural column (D'Alessio, 1996).

To assay CNS cell fates in embryos lacking vnd function (vnd embryos), markers were used that specifically label ventral, intermediate, or dorsal column neuroectoderm and neuroblasts. Three ventral column markers were used: Achaete, Odd-skipped (Odd), and Prospero. Achaete labels ventral and dorsal column neuroectoderm and neuroblasts of rows 3 and 7; Odd labels the ventral and dorsal column neuroblasts of row 1, whereas Prospero labels the ventral column neuroblast in row 3 (MP2). Two intermediate column markers were used: Ind and Huckebein. Intermediate neuroblast defective labels all intermediate column neuroectoderm and neuroblasts, whereas Huckebein labels the intermediate neuroectoderm and neuroblasts of row 3 (e.g., NB 4-2) and the ventral and intermediate neuroectoderm of rows 1 and 5. The dorsal marker Msh labels all dorsal column neuroectoderm and neuroblasts. Using these markers, an investigation was carried out of to determine whether vnd is necessary or sufficient to specify ventral column identity within the developing CNS. In vnd mutant embryos, Achaete, Odd, and Prospero are not detected in the ventral column neuroectoderm or neuroblasts. This phenotype is the result of a lack of gene expression in the neuroectoderm and in some neuroblasts, as well as a failure in neuroblast formation. For example, the ventral column NB 1-1 forms >80% of the time, yet it is never Odd+. However, absence of a Prospero+ MP2 is the result of a failure in MP2 formation. It is concluded that vnd is necessary to specify ventral column neuroectoderm and neuroblast identity, and to form specific ventral column neuroblasts (McDonald, 1998).

The ventral column in vnd embryos could be fully or partially transformed to a different columnar identity (intermediate or dorsal) or could assume a novel identity. To distinguish between these two possibilities, intermediate and dorsal column markers were examined in vnd embryos. In vnd embryos, intermediate column markers are ectopically expressed in the ventral column. Ind is detected in both ventral and intermediate column neuroectoderm and neuroblasts, although because some ventral neuroblasts do not form in vnd embryos, the intermediate column neuroblasts often shift to a more ventral position. The intermediate column marker Huckebein is also detected in both the ventral and intermediate columns of row 3 neuroectoderm and neuroblasts (13% in the ventral column). Expression of Msh in dorsal column neuroectoderm and neuroblasts is normal in vnd embryos. In addition, dorsal column expression of Achaete and Odd are unchanged. Taken together, these data show that vnd is necessary for the specification of ventral column identity and the repression of intermediate column identity within the CNS (McDonald, 1998).

During the course of this gene expression analysis, it was noticed that the ventral column neuroectoderm has a distinctive cellular morphology. The Vnd+ ventral column cells are frequently elongated along the DV axis to give them an asymmetry ratio of equal or >1.5 (long axis divided by short axis), whereas the Ind+ intermediate column cells are more often round, with an asymmetry ratio closer to 1.0. In vnd embryos most ventral column cells fail to assume an elongated morphology, instead showing a round morphology characteristic of intermediate column cells. Taken together, this molecular marker and cell morphology analyses show that vnd is required to establish ventral column gene expression profiles (perhaps by direct transcriptional activation and/or repression) as well as to induce a cell shape change characteristic of the ventral column neuroectoderm (McDonald, 1998).

To determine whether vnd is sufficient to specify ventral column fate, an hsp70-vnd transgene was used to ectopically express vnd in the intermediate and dorsal columns of neuroectoderm. In embryos carrying the hsp70-vnd transgene that are heat shocked to induce ubiquitous vnd expression (hs-vnd embryos), ventral markers are ectopically expressed in the intermediate and dorsal columns of the neuroectoderm and neuroblasts, and intermediate and dorsal markers are lost. However, the transformation of intermediate to ventral cell fate is more complete than that of dorsal to ventral cell fate transition. In hs-vnd embryos, the ventral column marker Achaete expands into the intermediate column, leading to Achaete expression extending continuously across ventral, intermediate, and dorsal columns of neuroectoderm and neuroblasts. Similarly, the ventral neuroblast markers Prospero and Odd also show ectopic expression in the intermediate column in hs-vnd embryos. Conversely, hs-vnd embryos show a loss of intermediate column marker expression. The intermediate column marker Ind is strongly repressed or completely abolished in hs-vnd embryos. The row 3 intermediate column marker Huckebein is also repressed in hs-vnd embryos; Huckebein row 5 expression appears unaffected, which is not surprising because both the ventral and intermediate columns express Huckebein in wild-type row 5 neuroectoderm. These results show that ectopic vnd results in a transformation of intermediate column to ventral column identity within both neuroectoderm and neuroblast cell types (McDonald, 1998).

The dorsal column is mis-specified in hs-vnd embryos, but is not fully transformed into ventral column identity. In hs-vnd embryos, the row 1 ventral column marker Huckebein is ectopically detected in the dorsal neuroectoderm, and the dorsal column marker Msh is partially repressed. However, the ventral column marker Prospero is ectopically expressed in the intermediate column but not in the dorsal column. Thus, ectopic vnd is sufficient to partially transform dorsal column to ventral column identity (McDonald, 1998).

To determine the extent to which Vnd controls ventral column neuroblast identity, a neuroblast cell lineage marker, Even-skipped (Eve), was used to assay the development of specific ventral and intermediate column neuroblasts. Eve labels the progeny of two ventral column neuroblasts (aCC/pCC neurons from NB 1-1; U/CQ neurons from NB 7-1) and the progeny of one intermediate column neuroblast (RP2/RP2sib neurons from NB 4-2). The pattern of Eve is a sensitive indicator for normal cell fates within these neuroblast cell lineages. In vnd embryos, the Eve+ aCC/pCC and U/CQ neurons, derived from ventral column neuroblasts, are never detected. NB 1-1 forms and produces Prospero+, Eve GMCs; thus loss of Eve from the aCC/pCC neurons is caused by an alteration in NB 1-1 identity or cell lineage. In contrast, the absence of the Eve+ U/CQ neurons is the result of failure of their parental NB 7-1 to form. In addition, vnd embryos show a partially penetrant duplication of the Eve+ RP2/RP2sib neurons derived from the intermediate column NB 4-2. This phenotype is the result of a transformation of a Huckebein ventral column neuroblast into a duplicate Huckebein+ intermediate column NB 4-2 (McDonald, 1998).

Is vnd sufficient to induce ventral neuroblast cell lineages in the intermediate or dorsal column neuroblasts? In hs-vnd embryos, an excess number of the ventral column Eve+ aCC/pCC and U/CQ neurons develop. Because hs-vnd produces intermediate to ventral transformations of neuroblast identity, it is likely that the excess aCC/pCC and U/CQ neurons develop from duplicated ventral column neuroblasts (NBs 1-1 and 7-1). However, the possibility cannot be ruled out that ectopic vnd triggers duplication of GMC identities or excess rounds of cell division. It is concluded that loss of vnd results in a transformation of ventral to intermediate column neuroblast identity that is maintained in the cell lineage of at least three neuroblasts (NBs 1-1, 7-1, and 4-2), and ectopic vnd results in the converse transformation of intermediate to ventral column neuroblast identity that is maintained in the cell lineage of at least two neuroblasts (NBs 1-1 and 7-1) (McDonald, 1998).

Defects in the formation of ventral and intermediate column neuroblasts occur in vnd embryos. There is a loss of ventral column neuroblasts, particularly MP2 and NB 7-1. In addition, in hs-vnd embryos there is precocious formation of neuroblasts in the intermediate column. In wild-type embryos, intermediate column neuroblasts form at mid-stage 9, but in hs-vnd embryos these neuroblasts form earlier, at early stage 9, about the time the adjacent ventral column neuroblasts are forming. These data provide additional evidence for a transformation of intermediate to ventral column identity (McDonald, 1998).

vnd also regulates dorsoventral patterning of the procephalic neuroectoderm. vnd, msh, and ind are each expressed in the procephalic ectoderm: Vnd in a ventral domain, Ind in three small clusters of cells at intermediate positions, and Msh in a dorsal domain. There are two differences in gene expression and regulation in the procephalic region compared with the thoracic and abdominal neuroectoderm: (1) Vnd and Msh share an extensive border, only interrupted by two small islands of Ind+ cells. In vnd embryos, Msh expands into the ventral domain of the procephalic neuroectoderm, showing that Vnd is required to repress msh expression in the head. Consistent with this result, misexpression of vnd leads to repression of msh. (2) The Ind+ anterior cell cluster 1 appears to coexpress Vnd; coexpression of Vnd and Ind is never observed in the thoracic and abdominal neuroectoderm. Surprisingly, vnd embryos show a loss of the Ind+ cluster 1, and misexpression of vnd does not affect Ind expression in cluster 1; thus, in this domain of the embryo, Vnd is required for the development of the Ind+ cluster 1. Because the Ind+ cells of cluster 1 are primarily restricted to neuroblasts, one possibility is that loss of vnd in the neuroectoderm leads to a failure of neuroblast formation and thus to a loss of Ind+ cells, rather than that Vnd directly activates ind transcription in this domain. The remaining two Ind+ cell clusters (2 and 3) are expressed and regulated in a manner consistent with the thoracic and abdominal neuroectoderm. Both Ind+ cell clusters 2 and 3 directly abut Vnd+ cells but do not express Vnd. In vnd embryos, the Ind+ cluster 3 expands ventrally into the domain normally expressing vnd, whereas Ind+ cluster 2 appears unaffected. Misexpression of vnd represses ind expression in clusters 2 and 3. Thus, vnd can both activate ind (cluster 1) or repress ind (clusters 2 and 3) depending on the position within the procephalic neuroectoderm (McDonald, 1998).

GENE STRUCTURE

Bases in 5' UTR - 341

Exons - two

Bases in 3' UTR - 394

PROTEIN STRUCTURE

Amino Acids 723

Structural Domains

VND has both a homeodomain and a PEST-domain, affording the protein a short half life due to rapid turnover through degradation. The region before the homeobox contains both an arginine rich region and an acidic region. The C terminal 30 amino acids are rich in histidine, proline, and glycine (Kim, 1989). Following the homeodomain is an NK2 box (Jimenez, 1995).

This study describes the NMR determination of the three-dimensional structure of a 77 amino acid residue protein, which consists of the 60 residue NK-2 homeodomain from Drosophila melanogaster and adjacent amino acid residues. The NK-2 homeodomain protein is part of a 723 amino acid residue protein that is expressed early in embryonic development in part of the central nervous system. NK-2 was characterized using both a natural abundance and a uniformly 15N enriched sample by two-dimensional and three-dimensional NMR experiments. The average root-mean-square deviation for 30 structures for residues 8 to 53 is 0.40 A for the backbone heavy-atoms and 0.72 A for the backbone and side-chain heavy-atoms. These structures were obtained from 986 NOE-derived upper and lower bound restraints. The three-dimensional structure contains three helices that consist of homeodomain amino acid residues 10 to 22, 28 to 38 and 42 to 52, as well as a turn between helix II and III, characteristic of homeodomains. Residues 53 to 60 of the DNA recognition helix are not fully ordered in the absence of DNA. In the free state, this segment adopts a flexible but helix-like structure between residues 53 and 56 and is disordered from residues 57 to 60, although the helix elongates by eight residues upon binding to DNA. Also discussed are the roles of variable residues 52, 54 and 56 in determining the structure and flexibility of the recognition helix, as well as the stability of the NK-2 homeodomain as manifested by its thermal denaturation (Tsao, 1995).

The Engrailed homeoprotein is a dominantly acting, so-called 'active' transcriptional repressor, both in cultured cells and in vivo. When retargeted via a homeodomain swap to the endogenous fushi tarazu gene (ftz), Engrailed actively represses ftz, resulting in a ftz mutant phenocopy. Functional regions of Engrailed have been mapped using this in vivo repression assay. In addition to a region containing an active repression domain identified in cell culture assays, there are two evolutionarily conserved regions that contribute to activity. The one that does not flank the HD is particularly crucial to repression activity in vivo. This domain is present not only in all engrailed-class homeoproteins but also in all known members of several other classes, including goosecoid, Nk1, Nk2 (vnd) and muscle segment homeobox. The repressive domain is located in the eh1 region, known as 'region three', found several hundred amino acids N-terminal to the homeodomain. The consensus sequence, arrived at by comparing Engrailed, Msh, Gsc, Nk1 and NK2 proteins from a variety of species, consists of a 23 amino acid homologous motif found in all these proteins. Thus Engrailed's active repression function in vivo is dependent on a highly conserved interaction that was established early in the evolution of the homeobox gene superfamily. Using rescue transgenes it has been shown that the widely conserved in vivo repression domain is required for the normal function of Engrailed in the embryo (Smith, 1996).

date revised: 20 January 99

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.