Neurogenesis in the chilopod <Emphasis Type="Italic">Lithobius forficatus</Emphasis> suggests more similarities to chelicerates than to insects

In a recent comparative study on neurogenesis in the diplopod Glomeris marginata we have shown that the millipede and the spider share several features that cannot be found in homologous form in insects and crustaceans. The most distinctive difference is that groups of neural precursors are singled out from the neuroectoderm of the spider and the diplopod, rather than individual cells (i.e. neuroblasts) as in insects or crustacean. This observation constitutes the first morphological indication for a close myriapod/chelicerate relationship that has otherwise only been suggested by molecular phylogenetic analysis. To see whether the pattern of neurogenesis described for the diplopod is representative for myriapods, we analysed neurogenesis in the basal chilopod Lithobius forficatus. We show here that groups of cells invaginate from the chilopod neuroectoderm at strikingly similar positions as the invaginating cell groups of the diplopod and the spider. Furthermore, the expression patterns of the proneural and neurogenic genes reveal more similarities to the chelicerate and the diplopod than to insects. Thus, chelicerates and myriapods share the developmental mechanism for neurogenesis, either because they are true sister groups, or because this reflects the ancestral state of neurogenesis in arthropods.Edited by P. Simpson     

Neurogenesis in myriapods and chelicerates and its importance for understanding arthropod relationships

Several alternative hypotheses on the relationships betweenthe major arthropod groups are still being discussed. We reexaminehere the chelicerate/myriapod relationship by comparing previouslypublished morphological data on neurogenesis in the euarthropodgroups and presenting data on an additional myriapod (Strigamiamaritima). Although there are differences in the formation ofneural precursors, most euarthropod species analyzed generateabout 30 single neural precursors (insects/crustaceans) or precursorgroups (chelicerates/myriapods) per hemisegment that are arrangedin a regular pattern. The genetic network involved in recruitmentand specification of neural precursors seems to be conservedamong euarthropods. Furthermore, we show here that neural precursoridentity seems to be achieved in a similar way. Besides theseconserved features we found 2 characters that distinguish insects/crustaceansfrom myriapods/chelicerates. First, in insects and crustaceansthe neuroectoderm gives rise to epidermal and neural cells,whereas in chelicerates and myriapods the central area of theneuroectoderm exclusively generates neural cells. Second, neuralcells arise by stem-cell-like divisions of neuroblasts in insectsand crustaceans, whereas groups of mainly postmitotic neuralprecursors are recruited for the neural fate in cheliceratesand myriapods. We discuss whether these characteristics representa sympleisiomorphy of myriapods and chelicerates that has beenlost in the more derived Pancrustacea or whether these characteristicsare a synapomorphy of myriapods and chelicerates, providingthe first morphological support for the Myriochelata group.     

Segment polarity gene expression in a myriapod reveals conserved and diverged aspects of early head patterning in arthropods

Arthropods show two kinds of developmental mode. In the so-called long germ developmental mode (as exemplified by the fly Drosophila), all segments are formed almost simultaneously from a preexisting field of cells. In contrast, in the so-called short germ developmental mode (as exemplified by the vast majority of arthropods), only the anterior segments are patterned similarly as in Drosophila, and posterior segments are added in a single or double segmental periodicity from a posterior segment addition zone (SAZ). The addition of segments from the SAZ is controlled by dynamic waves of gene activity. Recent studies on a spider have revealed that a similar dynamic process, involving expression of the segment polarity gene (SPG) hedgehog (hh), is involved in the formation of the anterior head segments. The present study shows that in the myriapod Glomeris marginata the early expression of hh is also in a broad anterior domain, but this domain corresponds only to the ocular and antennal segment. It does not, like in spiders, represent expression in the posterior adjacent segment. In contrast, the anterior hh pattern is conserved in Glomeris and insects. All investigated myriapod SPGs and associated factors are expressed with delay in the premandibular (tritocerebral) segment. This delay is exclusively found in insects and myriapods, but not in chelicerates, crustaceans and onychophorans. Therefore, it may represent a synapomorphy uniting insects and myriapods (Atelocerata hypothesis), contradicting the leading opinion that suggests a sister relationship of crustaceans and insects (Pancrustacea hypothesis). In Glomeris embryos, the SPG engrailed is first expressed in the mandibular segment. This feature is conserved in representatives of all arthropod classes suggesting that the mandibular segment may have a special function in anterior patterning.     

Neurogenesis in the spider: new insights from comparative analysis of morphological processes and gene expression patterns

While there is a detailed understanding of neurogenesis in insects and partially also in crustaceans, little is known about neurogenesis in chelicerates. In the spider Cupiennius salei Keyserling, 1877 (Chelicerata, Arachnida, Araneae) invaginating cell groups arise sequentially and in a stereotyped pattern comparable to the formation of neuroblasts in Drosophila melanogaster Meigen, 1830 (Insecta, Diptera, Cyclorrhapha, Drosophilidae). In addition, functional analysis revealed that in the spider homologues of the D. melanogaster proneural and neurogenic genes control the recruitment and singling out of neural precursors like in D. melanogaster. Although groups of cells, rather than individual cells, are singled out from the spider neuroectoderm which can thus not be homologized with the insect neuroblasts, similar genes seem to confer neural identity to the neural precursor cells of the spider. We show here that the pan-neural genes snail and the neural identity gene Krüppel are expressed in neural precursors in a heterogenous spatio-temporal pattern that is comparable to the pattern in D. melanogaster. Our data suggest that the early genetic network involved in recruitment and specification of neural precursors is conserved among insects and chelicerates.     

The morphological and molecular processes of onychophoran brain development show unique features that are neither comparable to insects nor to chelicerates

The phylogenetic position of onychophorans is still being debated; however, most phylogenies suggest that onychophorans are a sister group to the arthropods. Here we have analysed neurogenesis in the brain of the onychophoran Euperipatoides kanangrensis. We show that the development of the onychophoran brain is considerably different from arthropods. Neural precursors seem to be generated at random positions rather than in distinct spatio-temporal domains as has been shown in insects and chelicerates. The different mode of neural precursor formation is reflected in the homogenous expression of the proneural and neurogenic genes. Furthermore, the morphogenetic events that generate the three-dimensional structure of the onychophoran brain are significantly different from arthropods. Despite the different mode of neural precursor formation in insects and chelicerates (neuroblasts versus neural precursor groups), brain neurogenesis shares more similarities in these arthropods as compared to the onychophoran. Our data show that the developmental processes that generate the brain have considerably diverged in onychophorans and arthropods.     

The T-box genes H15 and optomotor-blind in the spiders Cupiennius salei, Tegenaria atrica and Achaearanea tepidariorum and the dorsoventral axis of arthropod appendages

SUMMARY Dorsoventral axis formation in the legs of the fly Drosophila melanogaster requires the T-box genes optomotor-blind ( omb ) and H15 . Evolutionary conservation of the patterning functions of these genes is unclear, because data on H15 expression in the spider Cupiennius salei did not support a general role of H15 in ventral fate specification. However, H15 has a paralogous gene, midline ( mid ) in Drosophila and H15 duplicates are also present in Cupiennius and the millipede Glomeris marginata . H15 therefore seems to have been subject to gene duplication opening the possibility that the previous account on Cupiennius has overlooked one or several paralogs. We have studied omb - and H15 -related genes in two additional spider species, Tegenaria atrica and Achearanea tepidariorum and show that in both species one of the H15 genes belongs to a third group of spider H15 genes that has an expression pattern very similar to the H15 pattern in Drosophila . The expression patterns of all omb -related genes are also very similar to the omb expression pattern in Drosophila . These data suggest that the dorsoventral patterning functions of omb and H15 are conserved in the arthropods and that the previous conclusions were based on an incomplete data set in Cupiennius . Our results emphasize the importance of a broad taxon sampling in comparative studies.     

Parasegmental organization of the spider embryo implies that the parasegment is an evolutionary conserved entity in arthropod embryogenesis

Spiders belong to the chelicerates, which is a basal arthropod group. To shed more light on the evolution of the segmentation process, orthologs of the Drosophila segment polarity genes engrailed, wingless/Wnt and cubitus interruptus have been recovered from the spider Cupiennius salei. The spider has two engrailed genes. The expression of Cs-engrailed-1 is reminiscent of engrailed expression in insects and crustaceans, suggesting that this gene is regulated in a similar way. This is different for the second spider engrailed gene, Cs-engrailed-2, which is expressed at the posterior cap of the embryo from which stripes split off, suggesting a different mode of regulation. Nevertheless, the Cs-engrailed-2 stripes eventually define the same border as the Cs-engrailed-1 stripes. The spider wingless/Wnt genes are expressed in different patterns from their orthologs in insects and crustaceans. The Cs-wingless gene is expressed in iterated stripes just anterior to the engrailed stripes, but is not expressed in the most ventral region of the germ band. However, Cs-Wnt5-1 appears to act in this ventral region. Cs-wingless and Cs-Wnt5-1 together seem to perform the role of insect wingless. Although there are differences, the wingless/Wnt-expressing cells and en-expressing cells seem to define an important boundary that is conserved among arthropods. This boundary may match the parasegmental compartment boundary and is even visible morphologically in the spider embryo. An additional piece of evidence for a parasegmental organization comes from the expression domains of the Hox genes that are confined to the boundaries, as molecularly defined by the engrailed and wingless/Wnt genes. Parasegments, therefore, are presumably important functional units and conserved entities in arthropod development and form an ancestral character of arthropods. The lack of by engrailed and wingless/Wnt-defined boundaries in other segmented phyla does not support a common origin of segmentation.     

The ‘Ventral Organs’ of Pycnogonida (Arthropoda) Are Neurogenic Niches of Late Embryonic and Post-Embryonic Nervous System Development

Early neurogenesis in arthropods has been in the focus of numerous studies, its cellular basis, spatio-temporal dynamics and underlying genetic network being by now comparably well characterized for representatives of chelicerates, myriapods, hexapods and crustaceans. By contrast, neurogenesis during late embryonic and/or post-embryonic development has received less attention, especially in myriapods and chelicerates. Here, we apply (i) immunolabeling, (ii) histology and (iii) scanning electron microscopy to study post-embryonic ventral nerve cord development in Pseudopallene sp., a representative of the sea spiders (Pycnogonida), the presumable sister group of the remaining chelicerates. During early post-embryonic development, large neural stem cells give rise to additional ganglion cell material in segmentally paired invaginations in the ventral ectoderm. These ectodermal cell regions – traditionally designated as ‘ventral organs’ – detach from the surface into the interior and persist as apical cell clusters on the ventral ganglion side. Each cluster is a post-embryonic neurogenic niche that features a tiny central cavity and initially still houses larger neural stem cells. The cluster stays connected to the underlying ganglionic somata cortex via an anterior and a posterior cell stream. Cell proliferation remains restricted to the cluster and streams, and migration of newly produced cells along the streams seems to account for increasing ganglion cell numbers in the cortex. The pycnogonid cluster-stream-systems show striking similarities to the life-long neurogenic system of decapod crustaceans, and due to their close vicinity to glomerulus-like neuropils, we consider their possible involvement in post-embryonic (perhaps even adult) replenishment of olfactory neurons – as in decapods. An instance of a potentially similar post-embryonic/adult neurogenic system in the arthropod outgroup Onychophora is discussed. Additionally, we document two transient posterior ganglia in the ventral nerve cord of Pseudopallene sp. and evaluate this finding in light of the often discussed reduction of a segmented ‘opisthosoma’ during pycnogonid evolution.     

Evolutionary changes in sensory precursor formation in arthropods: embryonic development of leg sensilla in the spider Cupiennius salei

We describe here for the first time the development of mechanosensory organs in a chelicerate, the spider Cupiennius salei. It has been shown previously that the number of external sense organs increases with each moult. While stage 1 larvae do not have any external sensory structures, stage 2 larvae show a stereotyped pattern of touch sensitive ‘tactile hairs’ on their legs. We show that these mechanosensory organs develop during embryogenesis. In contrast to insects, groups of sensory precursors are recruited from the leg epithelium, rather than single sensory organ progenitors. The groups increase by proliferation, and neural cells delaminate from the cluster, which migrate away to occupy a position proximal to the accessory cells of the sense organ. In addition, we describe the development of putative internal sense organs, which do not differentiate until larval stage 2. We show by RNA interference that, similar to Drosophila, proneural genes are responsible for the formation and subtype identity of sensory organs. Furthermore, we demonstrate an additional function for proneural genes in the coordinated invagination and migration of neural cells during sensory organ formation in the spider.     

Expression of the decapentaplegic ortholog in embryos of the onychophoran Euperipatoides rowelli

The gene decapentaplegic (dpp) and its homologs are essential for establishing the dorsoventral body axis in arthropods and vertebrates. However, the expression of dpp is not uniform among different arthropod groups. While this gene is expressed along the dorsal body region in insects, its expression occurs in a mesenchymal group of cells called cumulus in the early spider embryo. A cumulus-like structure has also been reported from centipedes, suggesting that it might be either an ancestral feature of arthropods or a derived feature (=synapomorphy) uniting the chelicerates and myriapods. To decide between these two alternatives, we analysed the expression patterns of a dpp ortholog in a representative of one of the closest arthropod relatives, the onychophoran Euperipatoides rowelli. Our data revealed unique expression patterns in the early mesoderm anlagen of the antennal segment and in the dorsal and ventral extra-embryonic tissue, suggesting a divergent role of dpp in these tissues in Onychophora. In contrast, the expression of dpp in the dorsal limb portions resembles that in arthropods, except that it occurs in the mesoderm rather than in the ectoderm of the onychophoran limbs. A careful inspection of embryos of E. rowelli revealed no cumulus-like accumulation of dpp expressing cells at any developmental stage, suggesting that this feature is either a derived feature of chelicerates or a synapomorphy uniting the chelicerates and myriapods.     

Expression of the decapentaplegic ortholog in embryos of the onychophoran Euperipatoides rowelli

The gene decapentaplegic (dpp) and its homologs are essential for establishing the dorsoventral body axis in arthropods and vertebrates. However, the expression of dpp is not uniform among different arthropod groups. While this gene is expressed along the dorsal body region in insects, its expression occurs in a mesenchymal group of cells called cumulus in the early spider embryo. A cumulus-like structure has also been reported from centipedes, suggesting that it might be either an ancestral feature of arthropods or a derived feature (=synapomorphy) uniting the chelicerates and myriapods. To decide between these two alternatives, we analysed the expression patterns of a dpp ortholog in a representative of one of the closest arthropod relatives, the onychophoran Euperipatoides rowelli. Our data revealed unique expression patterns in the early mesoderm anlagen of the antennal segment and in the dorsal and ventral extra-embryonic tissue, suggesting a divergent role of dpp in these tissues in Onychophora. In contrast, the expression of dpp in the dorsal limb portions resembles that in arthropods, except that it occurs in the mesoderm rather than in the ectoderm of the onychophoran limbs. A careful inspection of embryos of E. rowelli revealed no cumulus-like accumulation of dpp expressing cells at any developmental stage, suggesting that this feature is either a derived feature of chelicerates or a synapomorphy uniting the chelicerates and myriapods.     

Exploring the myriapod body plan: expression patterns of the ten Hox genes in a centipede

The diversity of the arthropod body plan has long been a fascinating subject of study. A flurry of recent research has analyzed Hox gene expression in various arthropod groups, with hopes of gaining insight into the mechanisms that underlie their evolution. The Hox genes have been analyzed in insects, crustaceans and chelicerates. However, the expression patterns of the Hox genes have not yet been comprehensively analyzed in a myriapod. We present the expression patterns of the ten Hox genes in a centipede, Lithobius atkinsoni, and compare our results to those from studies in other arthropods. We have three major findings. First, we find that Hox gene expression is remarkably dynamic across the arthropods. The expression patterns of the Hox genes in the centipede are in many cases intermediate between those of the chelicerates and those of the insects and crustaceans, consistent with the proposed intermediate phylogenetic position of the Myriapoda. Second, we found two 'extra' Hox genes in the centipede compared with those in DROSOPHILA: Based on its pattern of expression, Hox3 appears to have a typical Hox-like role in the centipede, suggesting that the novel functions of the Hox3 homologs zen and bicoid were adopted somewhere in the crustacean-insect clade. In the centipede, the expression of the gene fushi tarazu suggests that it has both a Hox-like role (as in the mite), as well as a role in segmentation (as in insects). This suggests that this dramatic change in function was achieved via a multifunctional intermediate, a condition maintained in the centipede. Last, we found that Hox expression correlates with tagmatic boundaries, consistent with the theory that changes in Hox genes had a major role in evolution of the arthropod body plan.     

Neurogenesis in the water flea Daphnia magna (Crustacea, Branchiopoda) suggests different mechanisms of neuroblast formation in insects and crustaceans

Within euarthropods, the morphological and molecular mechanisms of early nervous system development have been analysed in insects and several representatives of chelicerates and myriapods, while data on crustaceans are fragmentary. Neural stem cells (neuroblasts) generate the nervous system in insects and in higher crustaceans (malacostracans); in the remaining euarthropod groups, the chelicerates (e.g. spiders) and myriapods (e.g. millipedes), neuroblasts are missing. In the latter taxa, groups of neural precursors segregate from the neuroectoderm and directly differentiate into neurons and glial cells. In all euarthropod groups, achaete–scute homologues are required for neuroblast/neural precursor group formation. In the insects Drosophila melanogaster and Tribolium castaneum achaete–scute homologues are initially expressed in clusters of cells (proneural clusters) in the neuroepithelium but expression becomes restricted to the future neuroblast. Subsequently genes such as snail and prospero are expressed in the neuroblasts which are required for asymmetric division and differentiation. In contrast to insects, malacostracan neuroblasts do not segregate into the embryo but remain in the outer neuroepithelium, similar to vertebrate neural stem cells. It has been suggested that neuroblasts are present in another crustacean group, the branchiopods, and that they also remain in the neuroepithelium. This raises the questions how the molecular mechanisms of neuroblast selection have been modified during crustacean and insect evolution and if the segregation or the maintenance of neuroblasts in the neuroepithelium represents the ancestral state. Here we take advantage of the recently published Daphnia pulex (branchiopod) genome and identify genes in Daphnia magna that are known to be required for the selection and asymmetric division of neuroblasts in the fruit fly D. melanogaster. We unambiguously identify neuroblasts in D. magna by molecular marker gene expression and division pattern. We show for the first time that branchiopod neuroblasts divide in the same pattern as insect and malacostracan neuroblasts. Furthermore, in contrast to D. melanogaster, neuroblasts are not selected from proneural clusters in the branchiopod. Snail rather than ASH is the first gene to be expressed in the nascent neuroblasts suggesting that ASH is not required for the selection of neuroblasts as in D. melanogaster. The prolonged expression of ASH in D. magna furthermore suggests that it is involved in the maintenance of the neuroblasts in the neuroepithelium. Based on these and additional data from various representatives of arthropods we conclude that the selection of neural precursors from proneural clusters as well as the segregation of neural precursors represents the ancestral state of neurogenesis in arthropods. We discuss that the derived characters of malacostracans and branchiopods – the absence of neuroblast segregation and proneural clusters – might be used to support or reject the possible groupings of paraphyletic crustaceans.     

Conserved and novel functions for Netrin in the formation of the axonal scaffold and glial sheath cells in spiders

Netrins are well known for their function as long-range chemotropic guidance cues, in particular in the ventral midline of vertebrates and invertebrates. Over the past years, publications are accumulating that support an additional short-range function for Netrins in diverse developmental processes such as axonal pathfinding and cell adhesion. We describe here the formation of the axonal scaffold in the spiders Cupiennius salei and Achaearanea tepidariorum and show that axonal tract formation seems to follow the same sequence as in insects and crustaceans in both species. First, segmental neuropiles are established which then become connected by the longitudinal fascicles. Interestingly, the commissures are established at the same time as the longitudinal tracts despite the large gap between the corresponding hemi-neuromeres which results from the lateral movement of the germband halves during spider embryogenesis. We show that Netrin has a conserved function in the ventral midline in commissural axon guidance. This function is retained by an adaptation of the expression pattern to the specific morphology of the spider embryo. Furthermore, we demonstrate a novel function of netrin in the formation of glial sheath cells that has an impact on neural precursor differentiation. Loss of Netrin function leads to the absence of glial sheath cells which in turn results in premature segregation of neural precursors and overexpression of the early motor- and interneuronal marker islet. We suggest that Netrin is required in the differentiated sheath cells for establishing and maintaining the interaction between NPGs and sheath cells. This short-range adhesive interaction ensures that the neural precursors maintain their epithelial character and remain attached to the NPGs. Both the conserved and novel functions of Netrin seem to be required for the proper formation of the axonal scaffold.     

Duplicated Pax6 genes in Glomeris marginata (Myriapoda: Diplopoda), an arthropod with simple lateral eyes

Composite (facetted) eyes comprised by several units, termed ommatidia, are an ancestral feature in the arthropods. Some arthropods, however, do not possess composite eyes, obviously by secondary reduction. Reductions on the level of conserved eye developmental genes are one possibility to reduce the visual system. The genes of the Pax6 family have been shown to be key regulators of visual system development in a wide variety of animals. Reduction of Pax6 expression may therefore be expected in a species with reduced eyes. Here I have investigated the myriapod Glomeris marginata that displays very simple eyes. Glomeris, however, possesses two Pax6 genes that, based on their sequence, are similar to Drosophila eyeless (ey) and twin of eyeless (toy), respectively. Both genes are highly expressed in the optic lobes and the ventral nerve cord of developing embryos. Furthermore, homologs of other high-ranking eye developmental genes like hedgehog, decapentaplegic, dachshund, and homothorax are expressed in the optic lobes. This indicates that eye reduction in Glomeris is not realized at the level of the Pax6 genes or other genes on the upper levels of the eye development network. I suggest instead that the simple eyes of Glomeris are the product of changes at a much lower level in the network, probably at the level of genes directly regulating ommatidia development or ommatidia number and arrangement.     

Evolution of dorsal-ventral axis formation in arthropod appendages: H15 and optomotor-blind/bifid-type T-box genes in the millipede Glomeris marginata (Myriapoda: Diplopoda)

In Drosophila, the T-box genes optomotor-blind (omb) and H15 have been implicated in specifying the development of the dorso-ventral (DV) axis of the appendages. Results from the spider Cupiennius salei have suggested that this DV patterning system may be at least partially conserved. Here we extend the study of the DV patterning genes omb and H15 to a representative of the Myriapoda in order to add to the existing comparative data set and to gain further insight into the evolution of the DV patterning system in arthropod appendages. The omb gene of the millipede Glomeris marginata is expressed on the dorsal side of all appendages including trunk legs, maxillae, mandibles, and antennae. This is similar to what is known from Drosophila and Cupiennius and suggests that the role of omb in instructing dorsal fates is conserved in arthropods. Interestingly, the lobe-shaped portions of the mouthparts do not express omb, indicating that these are ventral components and thus may be homologous to the endites present in the corresponding appendages in insects. Concerning the H15 gene we were able to identify two paralogous genes in Glomeris. Both genes are expressed in the sensory organs of the maxilla and antenna, but only Gm-H15-1 is expressed along the ventral side of the trunk legs. The expression is more extensive than in Cupiennius, but less so than in Drosophila. In addition, no ventral expression domain is present in the maxilla, mandible, and antenna. Because of this, the role of H15 in the determination of ventral fate remains unclear.     

Recruitment of cell groups through Delta/Notch signalling during spider neurogenesis

Early neurogenesis in the spider is characterised by a stereotyped pattern of sequential recruitment of neural cells from the neuroectoderm, comparable with neuroblast formation in Drosophila: However, in contrast to Drosophila, where single cells delaminate from the neuroectoderm, groups of cells adopt the neural fate and invaginate into the spider embryo. This raises the question of whether Delta/Notch signalling is involved in this process, as this system normally leads to a singling out of individual cells through lateral inhibition. I have therefore cloned homologues of Delta and Notch from the spider Cupiennius salei and studied their expression and function. The genes are indeed expressed during the formation of neural cells in the ventral neuroectoderm. Loss of function of either gene leads to an upregulation of the proneural genes and an altered morphology of the neuroectoderm that is comparable with Delta and Notch mutant phenotypes in Drosophila: Thus, although Delta/Notch signalling appears to be used in the same way as in Drosophila, the lateral inhibition process produces clusters of invaginating cells, rather than single cells. Intriguingly, neuroectodermal cells that are not invaginating seem to become neural cells at a later stage, while the epidermal cells are derived from lateral regions that overgrow the neuroectoderm. In this respect, the neuroectodermal region of the spider is more similar to the neural plate of vertebrates, than to the neuroectoderm of Drosophila: