The Interactive Fly

Genes involved in tissue and organ development

Central Nervous System (CNS) - Ventral Midline

Ventral midline gene expression
Formation of midline precursors (MPs) and MP neurons in Drosophila from Multiple Notch signaling events control Drosophila CNS midline neurogenesis, gliogenesis and neuronal identity
MidExDb: gene expression data of Drosophila CNS midline cells
The midline and its embryonic origin
Discussion of genes affecting the midline
Glia and axonogenesis
Gene expression profiling of the developing Drosophila CNS midline cells
Midline progenitors revealed by dye tracing studies
Chromatin profiling of Drosophila CNS subpopulations identifies active transcriptional enhancers

Genes active in the ventral midline

Glia and axonogenesis

Separate sites in The Interactive Fly link to the genes involved in glia morphogenesis and axonogenesis.

The midline of the central nervous system

The midline mesectoderm (or ventral midline) is a cell population extending along the ventral surface of the embryo. It boasts a specific embryological origin and developmental fate, and plays a pivotal role in the development of the central nervous system.

Embryonic origin of the midline

The cellular blastoderm consists of three ventral cell populations. The most ventral is presumptive mesoderm. Upon gastrulation, presumptive mesoderm invaginates to become the interior of the gastrula. The invagination of the mesoderm brings together mesectodermal cells that formerly lay on either side of the mesoderm. These cells, the second ventral cell population, comprise the presumptive mesectoderm. This tissue is fated to become the midline. The third ventral cell population is the presumptive neurogenic ectoderm. Thus the presumptive mesectodermal cells occupy a ventrolateral strip, one cell wide, separating the presumptive mesoderm from the neurogenic region.

At the end of gastrulation each presumptive segment is only four cells wide, distributed along the anterior-posterior axis. After invagination, the eight midline cells, four from each side, become positioned side by side, in two rows of four cells. During germ band elongation these two rows intermingle, forming a single row of eight ventral midline cells. Subsequently each cell divides once. The two daughter cells reside side by side adjacent to the midline, forming two rows of eight cells. This is the last division for some of these cells, but others divide again producing more progeny. Each of the progeny cells, however, adopts a specific fate, and will proceed to fulfill a very specific role, identical for each segment, in the structuring and function of the central nervous system. During stage 11 the mesectoderm loses contact with the outer surface of the gastrula and becomes positioned in the dorsal area of the developing CNS (Klambt, 1991).

Mesectoderm cells give rise to neurons as well as glia cells. They give rise to midline structures of the CNS, including pioneering neurons of the intersegmental connectives. But mesectoderm cells differ from the adjacent lateral neurectoderm in two important aspects. (1) Mesectoderm cells never form typical neuroblasts. Instead, they remain integrated as epithelial cells in the surface ectoderm until quite late in development (stage 12; 8-9 hours), as opposed to neuroblasts, which segregate from the ectoderm during stages 9 to 11 (before 6 hours). Mesectoderm cells undergo a single division in the horizontal plane (i.e., parallel to the surface), whereas neuroblasts divide in a stem cell mode with a perpendicularly oriented mitotic spindle. (2) Mesectoderm cells invaginate from the surface ectoderm as a coherent group and uniformly give rise to neurons or glia cells, but not epidermis, whereas neuroblasts forming in the lateral neurectoderm are surrounded by cells that stay at the surface and later form the epidermis (Dumstrei, 1998 and references).

One of the main tasks of the midline cells is to provide cues for incoming axons for the neural cells of the neurectoderm. These cues help guide axons along and across the midline. Longitudinal axon fibers form the intersegmental connective. Transverse fibers form two commissures in each segment, a posterior commissure, and an anterior commissure. During commissure formation, the specific midline cells actually migrate, switching positions, so that each has new neighbors, different from the original ones. When the two commissures initially form, they are in close proximity, but soon separation takes place correlating with the migration of a specific glial cell (Klambt, 1991).

A model is presented for the formation of axon commissures. First, the growth cones that pioneer both commissures appear to be attracted toward the midline from a distance, suggesting a possible role for chemotropism. Second, the growth cones that pioneer the posterior commissure are guided toward medial precursor neurons (MP1) and ventral unpaired medial neurons (VUMs) neural derivatives of one and two respectively of the eight midline precursors. The commissure appears to form in two steps: initially, the growth cones extend toward the VUM cells, and once reaching them, they then turn anteriorly over the MP1 neurons to cross the midline anterior to the VUM cell nuclei. Third, the growth cones that pioneer the anterior commissure are guided toward the anterior midline glia (MGA) and the VUM growth cones. Thus different midline cell types, and possibly different signals, appear to be involved in the pioneering of the two commissures. Fourth, the last step of axon commissure formation requires the separation of the two commissures. This event involves the posterior migration of the middle midline glia (MGM) over the top of the MGA glia and along the VUM growth cones between the two commissures. This model predicts that disruptions of any one of these proposed functions should lead to a predictable phenotype (Klambt, 1991).

Midline progenitors revealed by dye tracing studies

At the blastoderm stage, the mesectodermal or midline progenitor cells are arranged as one row on either side separating the anlage of the mesoderm from the neurogenic ectoderm. During gastrulation, the mesoderm invaginates (forming the ventral furrow) and the two rows of mesectodermal cells become juxtaposed along the ventral midline. At this stage, these cells can be unambiguously identified in the living embryo due to their position and their characteristic shape. To disclose the dynamics of their early behavior until they delaminate from the ectoderm, the midline cells were continuously monitored in vivo using time-lapse videomicroscopy (Bossing, 1994).

About 25 minutes (at 24°C) after the onset of gastrulation, (a.o.g.) the peripheral diameter of the midline cells significantly increases. About 10 minutes later, they alternately rotate clockwise and counterclockwise (at least five times) for 3-4 minutes. These rotations are followed by their first postblastodermal division. This behavior can also be observed for other dividing cells in the neurogenic ectoderm. Orientation of the mitotic spindle is in parallel to the epithelial layer but seems to be random within this plane. About 40 minutes a.o.g., the midline cells complete their mitosis and their daughter cells occupy the medial 2-3 rows of the ectoderm. In the course of further germ band elongation, they become stretched along the longitudinal axis forming a single row on either side in which sibling cells remain next to each other. At the end of stage 9, their basal part becomes progressively shifted interiorly, but they still maintain a prominent cytoplasmic extension to the periphery, which progressively stretches along the longitudinal axis until completion of germ band elongation (late stage 10). Finally, during the second half of stage 11, all midline cells completely delaminate from the ectoderm. As a rule, enlargement of cell diameters, division and delamination begin near the cephalic furrow and progress posteriorly (Bossing, 1994).

Further divisions among midline cells are only performed by the MNB and by progenitors that exceptionally give rise to larger compound clones. In these cases, the second division takes place between stages 11 and 13 and the third division between stages 14 and 16. Taken together, all midline progenitors divide while still in the ectoderm, and delamination is not completed before the end of stage 11. The sequence of early dynamic processes, including enlargement of cell diameters, rotatory movements, division and segregation, follows a precisely regulated temporal and spatial pattern (Bossing, 1994).

Thoracic midline progenitors numbers 8-15 on either side (as counted from the cephalic furrow) were individually labelled about 35 times each. In total 537 clones were analyzed. These fall into five different classes. Clones of class 1, the VUM neuron class, consist of cells known as ventral unpaired median (VUM) neurons. 215 clones of VUM neurons (40% of all clones) were obtained. In 169 cases (79% of all VUM clones), the lineage consisted of only two cells, one motoneuron and one interneuron. VUM clones that contained four cells (n=32; 15% of all VUM clones) or six cells (n=14; 6% of all VUM clones) were obtained. At stage 17 the closely associated cell bodies of the VUM neurons are located medially in the ventral and posterior cortex region of the neuromere. Their fibers run tightly fasciculated dorsally towards the neuropile where they separate to form a motoneuronal projection bifurcating in the dorsal part of the anterior commissure and an interneuronal projection bifurcating in the ventral part of the posterior commissure. First, outgrowth of axonal processes is detected at stage 13. In stage 16, the VUM neurons appear to be shifted from a position close to the posterior commissure towards their final position in the most ventral cortex region (Bossing, 1994).

In most instances (n=121), two types of 2-cell VUM clones could be distinguished differing with respect to their motoneuronal projection: In one type (n=76; 63%), the fiber on either side leaves the CNS through the posterior root of the anterior fascicle (intersegmental nerve) and ends in a dorsolateral to dorsal region. In the other type (n=45; 37%), the fiber leaves the CNS through the posterior fascicle (segmental nerve) and ends in a ventral-to-ventrolateral region of the embryo. The course of the interneuronal projection shows no significant variations. In 28 cases of the 4-cell VUM clones, the projection pattern was examined in some detail. In most instances (n=22; 79%), two bifurcated motoneuronal as well as two bifurcated interneuronal projections were detected. The interneuronal projections leave the posterior commissure tightly fasciculated and do not separate until close to their termination sites in the connectives. The motoneuronal projections in most cases (n=13) leave the CNS separately through the segmental and intersegmental nerve to innervate muscles in the ventral-to-ventrolateral and dorsolateral-to-dorsal region, respectively (Bossing, 1994).

Seventy-eight clones (14.5% of all clones) consisted of clones of class 2. These are glial cells previously described as midline glia. In most cases, the clones were composed of 2 cells; in three cases, 3-4 cells and in five cases 1-2 cells were found. Since cells of each clone were closely attached to each other and their nuclei were sometimes obscured, the number of glial cells per clone could not be determined precisely. In 54 instances, the position of the cells were determined. At stage 17, they typically (n=39) enwrap the two commissures with one nucleus in a more ventral and the second nucleus in a more dorsal position between the commissures. In 9 cases the two cells were located dorsoposteriorly to the posterior commissure and in 4 cases dorsoanteriorly to the anterior commissure (Bossing, 1994).

Seventy-nine clones (14.7% of all clones) consisted of clones of class 3, the MP1 neurons, a pair of interneurons described as midline precursor 1 (MP1) neurons. At stage 17, the MP1 neurons reside slightly anterior and ventral to the posterior commissure in the corner formed by the connectives and the posterior commissure. Their ipsilateral projection bifurcates in an anterior and posterior branch, which runs within the medial sector of the connective. Since the branches at this stage still seemed to be growing, their length varied among the preparations. The posterior branch may span up to three neuromeres whereas the anterior branch may span up to two neuromeres. At stage 13, the MP1 neurons are located dorsally in close neighbourhood to each other. At this stage, outgrowth of fibers could be detected. Bifurcation of the fibers occurs at the end of stage 14. During the condensation of the nerve cord (stage 16/17), the cell bodies of the MP1 become shifted more laterally to their final position (Bossing, 1994).

Clones of class 4, MNB clones, were obtained in 71 instances (13.2% of all clones). The clones comprised 5-8 neurons, most of which consisted of 6 cells (n=35; 49%). Their cell bodies formed a dense medial cluster ventrally in the posterior cortex of the neuromere. They sent a fascicle dorsally towards the posterior commissure where it bent to run anteriorly along the midline to the anterior commissure. At the anterior edge of the anterior commissure, it split into short bilaterally projecting fibers. Between the commissures, one fiber left the main bundle and bifurcated at the dorsoposterior rim of the anterior commissure to project on either side to the lateral area of the connective. In late embryos (end of stage 17), this fiber may have reached the border of the ventral nerve cord by projecting through the intersegmental nerve. In one case, it had even left the CNS to end near a ventrolateral muscle, suggesting that it represents a late differentiating motoneuronal projection. In a number of late stage 17 embryos, a further projection was observed that medially entered the posterior commissure and split into short bilaterally projecting fibers similar to the anterior projection in the anterior commissure. No differentiation of fibers was detected before stage 16. At stage 17, in addition to three of the VUM neurons, a cluster of about 6 small cells stained positive with the anti-Engrailed antibody. Since the median neuroblast (MNB) and its progeny are known to express Engrailed and because these cells are located in the same position as the cluster of cells described here, it is thought that this type of clone derives from the MNB (Bossing, 1994).

Clones of class 5, the unpaired median interneurons (UMI) visualized in 75 clones (14% of all clones) consist of two interneurons that have not been described before. UMI cell bodies occupy a position dorsal to the VUM and MNB cells. One of them projected medially towards the anterior region of the anterior commissure. Here it bifurcated and both branches entered the connective where they turned anteriorly. The fiber of the second UMI neuron split in the posterior commissure to project bilaterally to the lateral part of the connective. Here it bifurcated again on either side into an anterior and a posterior branch. An additional small bilateral branch was formed in the anterior commissure (Bossing, 1994).

It is important to note that the system shows significant variability: (1) the number of midline progenitors (sim-lacZ expressing cells) generally varies between 6 and 8 cells, with an average of 7.5 cells per segment; (2) in a significant number of cases (12%) the composition of clones differs from the typical pattern in that they comprise additional cells either of the same type (4-cell and 6-cell VUM clones; 8.5% of all clones) or of different types (e.g. MP1/UMI; 3.5% of all clones). It is tempting to speculate that the generation of larger compound clones in the midline is a function of the number of progenitors available in a particular segment. In this way the variability in progenitor cell numbers may be compensated by variabilities in clone sizes to equip segments with an adequate final population of midline cells. Nevertheless, (3) in the fully differentiated embryo, the segmental number of sim expressing midline glia was found to vary between 2 and 4 cells. Furthermore, variability of cell numbers in the late embryo has also been reported for the lineage of the neuroblast NB1-1. Thus, the widespread assumption of invariant lineages in the CNS of Drosophila has to be reconsidered. Knowing the midline lineages and their variabilities, it is now possible to experimentally approach the mechanisms leading to their specification (Bossing, 1994).

Genes affecting the midline

Three groups of mutants elucidate the nature of the cell interactions that structure the CNS and the specific role of the midline. Group 1 mutants, consisting of single minded and slit, result in the complete absence of commissures and the collapse of the longitudinal tracts into a single, fused tract at the midline. single minded mutation results in the death of midline precursors, and slit mutation results in the failure of midline cell differentiation. Group 2 mutants, consisting of orthodenticle alone, results in a lack of the posterior commissure for each segment. In otd mutants, growth cones show no affinity for the midline and instead grow anteriorly and posteriorly on their own side. otd strongly influences the differentiation of several specific midline cells. Group 3 mutants, consisting of spitz, star and rhomboid, lead to a fusion of the two commissures. In all three mutants the glial cell lineage is affected (Klambt, 1991).

Subsequent analysis defines four steps involved in commissure development. single minded, jaywalker, Egf receptor and slit are involved in the first step in midline formation, the formation of the anlage of the CNS midline. Next the segment polarity genes hedgehog, engrailed, patched and wingless are involved in the specification of midline cell number. It is possible that midline and ectodermal pattern formations occur at the same time. In addition to the segment polarity genes other signaling mechanisms appear important. Notch, for example, is required to specify the different midline lineages. The third step in commissure formation consists of the formation of commissures. Once the midline cells have been specified, they guide commissural growth cones toward and across the midline. Here, the Netrins, frazzled, commissureless, weniger, schizo, roundabout and karussel play an essential role. The fourth step in commissure development involves the separation of the commissures. Contrary to midline specification and initial commissure formation this process occurs relatively late during embryogenesis and thus a maternal contribution is not likely to rescue a mutant phenotype. In addition, the separation of commissures requires not only the differentiation of the midline glial cells but also the differentiation of the midline neurons as well as interactions of these two cell types for normal migration to occur. This might explain the large number of genes identified (Hummel, 1999).

The analysis of mutations reveal two major phenotypic classes, the pointed and the tramtrack groups. pointed and tramtrack mediate different aspects of glial development. In pointed mutants no glial differentiation occurs, whereas ectopic pointed expression results in ectopic glial differentiation. tramtrack, in contrast, does not interfere with actual glial cell differentiation but appears to be required for the repression of neuronal differentiation in these cells. The pointed group consists of pointed itself, rhomboid, kastchen, klotzchen, kette, schmalspur, mochte gern, spitz, Star, cabrio and kubel. Mutations in eight genes lead to an axon phenotype initially described for tramtrack. In tramtrack-type mutation (tramtrack, shroud, disembodied, spook, shade, shadow, phantom, and rippchen) commissures appear fused, but in contrast to pointed group mutations, connectives are not affected (Hummell, 1999).

The very specific lineages of neural precursors, the specific migratory paths of cells, and specific interactions between cells in the developing central nervous system, all point to the special nature of Drosophila development, quite reminiscent of the unique specification of cell lineage in C. elegans. The identity of each cell is defined in a specific manner by the activation of specific genes for specific lineages. Each cell expresses a different combination of cell surface adhesion molecules, cell surface receptors, and ligands. These molecules, each of which mediate the interactions that take place between cells, as well as the growth of axons, ensure the correct function of the nervous system; the ultimate result they bring about is the accurate wiring of a nervous system for correct function (Hummell, 1999).

Gene expression profiling of the developing Drosophila CNS midline cells

The Drosophila CNS midline cells constitute a specialized set of interneurons, motorneurons, and glia. The utility of the CNS midline cells as a neurogenomic system to study CNS development derives from the ability to easily identify CNS midline-expressed genes. For this study, a variety of sources were used to identify 281 putative midline-expressed genes, including enhancer trap lines, microarray data, published accounts, and the Berkeley Drosophila Genome Project (BDGP) gene expression data. For each gene, expression was analyzed at all stages of embryonic CNS development and expression patterns were categorized with regard to specific midline cell types. Of the 281 candidates, 224 midline-expressed genes were identified; these include transcription factors, signaling proteins, and transposable elements. Fifty-eight genes are expressed in mesectodermal precursor cells, 138 in midline primordium cells, and 143 in mature midline cells -- 50 in midline glia, 106 in midline neurons. Additionally, twenty-seven genes expressed in glial and mesodermal cells associate with the midline cells. This work provides the basis for future research that will generate a complete cellular and molecular map of CNS midline development, thus allowing for detailed genetic and molecular studies of neuronal and glial development and function (Kearney, 2004).

To create a database of midline gene expression profiles, embryos at each developmental stage pertinent to embryonic CNS development (stages 5-17) were examined. These stages include (1) mesectoderm anlage in statu nascendi (ISN: stages 5-6); (2) mesectoderm anlage (stages 7-8); (3) midline primordium (stages 9-12), and (4) mature midline cells (stages 13-17). The annotation of these developmental time periods corresponds to the terminology proposed by the BDGP gene expression group. It is common for genes to be expressed in multiple stages and in multiple midline and lateral CNS cell types. Using alkaline phosphatase (AP) staining gene expression can be identified in subsets of midline cells. The anterior and posterior midline glia occupy opposite ends of the midline segment while midline neurons can be recognized as medial neurons (MP1 and MNB progeny) or ventral neurons (MP3 and VUM) based on their position in the segment. Midline accessory cells also occupy characteristic positions with respect to the mature midline cells. Definitive identification of expression in individual cell types requires confocal analysis, and will, largely, be the subject of subsequent work (Kearney, 2004).

Mesectoderm ISN expression (stages 5-6): The mesectoderm ISN occurs in blastoderm and gastrula stage embryos (stage 5-6) when mesectodermal cells cannot be recognized by morphology but only by gene expression. Thirty-seven genes expressed in the mesectoderm ISN were identified. These include transcription factors (14) and proteins implicated in cell signaling (12), consistent with early roles for these genes in dictating midline cell fates. These were subdivided into two categories: genes expressed in all mesectoderm ISN cells (18) and genes expressed in a subset of mesectoderm ISN cells (19). Not included are gap genes and others with either widespread or narrowly restricted expression patterns that are unlikely to play a role in the development of mesectoderm ISN cells or their progeny. Of the 18 genes expressed in all mesectoderm ISN cells, some are expressed exclusively in the mesectoderm, such as sim, HLHmβ, Sema-1b, and CG9598, whereas others, including brk and Tom, are expressed in the mesectoderm and neuroectoderm. There is a significant body of work describing the regulation of mesectoderm ISN gene expression by the embryonic dorsal-ventral (D/V) patterning cascade and the Notch signaling pathway, thus providing many tools to understand the regulation of these mesectoderm ISN-expressed genes (Kearney, 2004).

Genes that are expressed in subsets of mesectoderm ISN cells are expressed as either repeating stripes that bisect the mesectoderm ISN or small numbers of mesectoderm ISN cells. These genes include the well-known segmentation genes, such as dpn, wg, en, gsb-n, and slp2 as well as additional genes such as Sema-5c, Mes2, and Ect3. Although their functions in midline cells are unclear, many of these genes are likely involved in defining different classes of midline cell types, similar to their roles in intrasegmental patterning of the epidermis and lateral CNS (Kearney, 2004).

Mesectoderm anlage expression (stages 7-8): The mesectoderm anlage stage occurs after gastrulation (stages 7-8) when the two populations of mesectoderm ISN cells meet at the ventral midline and undergo a synchronous cell division producing 16 midline cells per segment. There are 22 genes expressed in all mesectoderm anlage cells, including sim, Wnt8, and edl, and 19 expressed in subsets, including Mes2, wg, slp2, esg, and Ect3. More than half of these genes (22/41) are also expressed in the mesectoderm ISN in similar patterns. Only 10 genes expressed in all mesectoderm anlage cells and six in subsets of cells initiate transcription during this stage. Consistent with roles in regulating the division and fate of midline cells, this newly expressed group includes genes involved in cell division (CycE, stg), transcriptional control (esg, Hgtx), and cell signaling (btl, edl, tkv, tsl) (Kearney, 2004).

Midline primordium expression (stages 9-12): During the midline primordium stage, midline precursor cells divide, change shape, migrate, and differentiate. There are 138 genes expressed in the midline primordium. All but six of the genes expressed in the mesectoderm anlage continue expression in the midline primordium. However, eight genes, including rho and vvl, refine their expression from all mesectoderm anlage cells to a subset of midline primordium cells. In contrast, four genes, including CenB1A and Stat92E, expand their expression from subsets of mesectoderm anlage to all midline primordium cells. Of note, 67% (92/138) of the midline primordium-expressed genes initiate their expression during this stage. Several genes (37) are expressed in all midline primordium cells including cdi, CenB1A, Ect3, mfas, rho, sog, Tl, and Wnt8. These genes mimic the expression pattern of sim and are likely direct transcriptional targets of Sim::Tgo heterodimers. Most genes at this stage are expressed in cellular subsets (101) and are noteworthy since they reveal the subdivision of midline progenitors into distinct cell types. It is likely that these genes are regulated by Sim::Tgo in combination with other transcription factors. Of note, a large number of transcription factors are expressed in subsets of midline primordium cells. Many of the midline primordium-expressed genes are also expressed in mature midline neurons (41) and will be useful in determining developmental relationships between midline primordium and mature midline cells (Kearney, 2004).

The 138 genes expressed in midline primordium cells include characterized and uncharacterized genes. Transcription factors compose the largest class. There are 40 putative transcription factors, including 15 zinc finger proteins, 9 homeobox proteins, and 7 bHLH proteins. Interestingly, zinc finger genes such as cas and Kr play prominent roles in generating ganglion mother cell (GMC) diversity in the lateral CNS and may play a similar role in generating the diversity of MNB progeny. Another set of related zinc finger proteins that function together in development of the lateral CNS neurons are esg, sna, and wor. All three are expressed in the midline primordium and have different expression profiles that likely overlap in some midline cells. An additional set of genes important in lateral CNS development and expressed in the midline primordium are the proneural genes sc and l(1)sc. Similar to their role in lateral CNS neuroblast formation, these genes may be important for the formation, differentiation, and division of the MNB and its progeny (Kearney, 2004).

Another 41 genes expressed in the midline primordium are membrane or intracellular signaling proteins or membrane receptors involved in cell adhesion. These genes include argos, btl, glec, Tl, tsl, and wrapper, some of which have been shown to be important for midline cell migration, apoptosis, midline-controlled axonogenesis, and glial-neuron interactions (Kearney, 2004).

Mature midline cells (stages 13-16): midline glial expression: The midline glia are specified as two separate populations of cells, anterior and posterior, that are reduced by apoptosis to generate the ~3 mature midline glia in each segment. Fifty genes expressed in midline glia were identified. Most (22) are expressed in both anterior and posterior midline glia, 11 are expressed prominently in anterior midline glia, and 4 are expressed prominently in posterior midline glia. An additional 13 genes are expressed in midline glia, but it is unclear whether they are expressed in all or a subset of midline glia. Many of the genes expressed in midline glia are also expressed in the midline primordium (35/50), potentially providing markers for glial precursors in the primordium stage. Several genes (10) are expressed in both midline glia and other glial cells, indicating that midline and lateral glia can share glial-expressed genes. Additionally, 13 genes are expressed in both midline glia and midline neurons. Genes expressed in midline glial cells are varied in nature, including transcription factors, signaling proteins for the spitz, netrin, and slit pathways, intracellular signaling proteins, and genes whose functions have not been tested experimentally. There are also genes, including CG8291, CG31116, CG31634, and Mdr65, that resemble ion channels and transport proteins that may be involved in glial function (Kearney, 2004).

Mature midline cells (stages 13-16): midline neuronal expression: Midline neurons consist of a number of different interneurons and motorneurons from distinct midline lineages. One hundred and six genes expressed in mature midline neurons have been identified: 52 in subsets of midline neurons, 29 in all or most midline neurons, and 25 whose expression is too complex or obscured to make an accurate assignment. Approximately half of these genes (51) initiate expression during the midline primordium stage, suggesting that many neuronal cell fates are likely determined prior to the mature midline stages. Genes expressed in the mature midline neurons can be categorized by their appearance in the medial or ventral regions of the CNS. The medial group includes the two MP1 neurons and the 5-8 MNB progeny, while the ventral group includes the MP3 and VUM neurons. Of the 52 genes identified in subsets of midline neurons, 16 are expressed in medial neurons and 36 in ventral neurons . Among genes expressed in midline neurons, notable are genes associated with neural function. These include neurotransmitter receptors (Glu-RI, 5-HT1a, 5-HT7, Nmdar1, NPFR1, NPFR76F) and proteins involved in neurotransmitter synthesis (Gad1, ple). In addition, this class contains many functionally uncharacterized genes. With their well-defined complement of gene expression, midline neurons will be a useful system for studying the molecular genetics of neural function (Kearney, 2004).

Mature midline (stages 13-16): midline accessory cell expression: A number of cells that lie along the CNS midline are derived from the lateral CNS or mesoderm. These include the MM-CBG, channel glia, and DM cells. Some of the genes that are expressed in the MM-CBG and channel glia have been described in a genome-wide search for glial-expressed genes. While a systematic analysis of accessory cell gene expression was not a major goal of this screen, a number of genes described from various sources as 'midline' are expressed in accessory cells, and for this reason they have been included (Kearney, 2004).

The MM-CBG consist of two to four glial cells per segment (two in abdominal and four in thoracic segments) that lie adjacent to the VUM neurons. Their function is unknown; however, due to their close proximity, MM-CBG may act as support cells for the VUM neurons. These cells originate from lateral CNS neuroblast 6-4 and then migrate to their final position adjacent to the midline. MM-CBG can be recognized based on their cell morphology, location, and co-localization with Mz840-Gal4, an enhancer trap line expressed in MM-CBG. Twelve genes expressed in MM-CBG were identified. Two genes, CG1124 and nrv2, were also expressed in midline glia. The genes expressed in MM-CBG are characteristic of glia, and include five that encode transporters (CG6070, CG10960, Eaat1, nrv1, nrv2). Studying the MM-CBG and VUM neurons will potentially provide insight into glial-neuronal interactions (Kearney, 2004).

The channel glia consist of six cells, two located dorsally, two medially, and two ventrally, that line the channel that lies along the midline between the posterior commissure and anterior commissure of the next ganglion. Like MM-CBG, the channel glia are derived from non-mesectodermal cells, including lateral CNS neuroblast 7-4, and migrate towards the midline. Channel glia can be identified based on their glial-like morphology, their location along the midline channel, and expression of previously identified markers, Mz820-Gal4 and en. Six genes were found to be expressed in channel glia. In addition to repo (a transcription factor shown to be expressed in many glia including channel glia) the functions of other channel glia-expressed genes are diverse and include a G protein-coupled receptor (CG4322), an esterase (Gli), a metallopeptidase (CG6225), and a transporter (CG3168) (Kearney, 2004).

The two mature DM cells in each segment are mesodermal derivatives that lie above the CNS. DM cells are associated with the midline channel and, thus, reside between the posterior commissure and the anterior commissure of the next ganglion. The DM cells extend a process to the body wall muscle attachment sites. CNS motorneurons and neurosecretory cells that form the transverse nerve use these processes to navigate toward their muscle synaptic targets. The formation of mature DM cells is dependent on an unknown signal from the CNS midline cells. The paired DM cells are easily identified based on their number, position atop the CNS, and characteristic morphology. These cells are noted for their expression of the buttonless gene and, like channel glia, these cells express numerous genes encoding basement membrane components. Ten genes were identified that were expressed in the DM cells, including the TE 412 transposable element. The genes expressed in DM cells are varied, including two transcription factors (zfh1 and CG12648), a cell adhesion protein (fas1), a protease (Mmp1), a basement membrane component (prc), a G protein-coupled receptor (CG4322), and a transporter (CG4726) (Kearney, 2004).

Chromatin profiling of Drosophila CNS subpopulations identifies active transcriptional enhancers

One of the key issues in studying transcriptional regulation during development is how to employ genome-wide assays that reveals sites of open chromatin and transcription factor binding to efficiently identify biologically relevant genes and enhancers. Analysis of Drosophila CNS midline cell development provides a useful system for studying transcriptional regulation at the genomic level due to a large, well-characterized set of midline-expressed genes and in vivo validated enhancers. In this study, Formaldehyde-Assisted Isolation of Regulatory Elements (FAIRE-seq) was performed on FACS-purified midline cells and the midline FAIRE data were compared with whole-embryo FAIRE data. It was found that regions of the genome with a strong midline FAIRE peak and weak whole-embryo FAIRE peak overlap with known midline enhancers and provide a useful predictive tool for enhancer identification. In a complementary analysis, a large dataset of fragments that drive midline expression in vivo was compared with the FAIRE data. Midline enhancer fragments with a midline FAIRE peak tend to be near midline-expressed genes, whereas midline enhancers without a midline FAIRE peak are often distant from midline-expressed genes and unlikely to drive midline transcription in vivo (Pearson, 2016).

genes active in ventral midline


Bossing, T and Technau, G. M. (1994). The fate of the CNS midline progenitors in Drosophila revealed by a new method for single cell labelling. Development 120: 1895-1906. PubMed ID: 7924995

Dumstrei, K., et al. (1998). EGFR signaling is required for the differentiation and maintenance of neural progenitors along the dorsal midline of the Drosophila embryonic head. Development 125(17): 3417-3426. PubMed ID: 9693145

Hummel T, Schimmelpfeng K, Klambt C. (1999). Commissure formation in the embryonic CNS of Drosophila. Identification of required gene functions. Dev. Biol. 209(2): 381-98. PubMed ID: 10328928

Kearney, J. B., et al. (2004). Gene expression profiling of the developing Drosophila CNS midline cells. Dev. Biol. 275: 473-492. PubMed ID: 15501232

Klambt, C., Jacobs, J. R. and Goodman, C. S. (1991). The midline of the Drosophila central nervous system: a model for the genetic analysis of cell fate, cell migration and growth cone guidance. Cell 64: 801-815. PubMed ID: 1997208

Pearson, J.C., McKay, D.J., Lieb, J.D. and Crews, S.T. (2016). Chromatin profiling of Drosophila CNS subpopulations identifies active transcriptional enhancers. Development 143: 3723-3732. PubMed ID: 27802137

Genes involved in organ development

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.