Neural development during embryonic and metamorphic development in insects

L. P. Tolbert

ARL Division of Neurobiology,
University of Arizona
P.O. Box 210077, Tucson, AZ
85721-0077, USA

INTRODUCTION

Why study the development of insect nervous systems? Beyond satisfying sheer curiosity, knowledge of insect neural development holds the hope of revealing novel and specific paths for biologically sensitive intervention to protect or to control specific populations of insects. Furthermore, for developmental biologists, insects offer a rich source of material. The development of the nervous system in insects follows different paths, depending on the life history of the species, yet many of the cellular and molecular mechanisms underlying neural development appear to be common across disparate insect species, and even common between insects and mammals (see Arendt & Nubler-Jung 1999). Individual species variations confer particular experimental advantages to the investigator using insects. For all of these reasons, recent years have seen a huge research effort to understand neural development in insects. Unable to review all of insect nervous system development in a one-hour presentation, I will provide a very selective review. I will focus on the growing understanding of the importance and nature of cellular interactions in insect neural development. The nervous systems of certain insects have been excellent systems in which to study these interactions. Drosophila melanogaster, a superb organism for genetic studies, has been used to great advantage to reveal the cellular and molecular bases for developmental influences in the peripheral and central nervous systems. Large holometabolous insects, such as the hawkmoth Manduca sexta, have different advantages; for instance, developing sensory and central neural structures are readily accessible throughout a major postembryonic wave of development (during the metamorphosis of the larva to the adult), when the animal is large and hardy, readily amenable to surgical manipulations. Manduca, besides being useful for cellular studies, also has been especially useful for studies of the molecular bases of hormone action, a special type of long-range cellular interaction, in neural development (Levine et al. 1995). Insect nervous systems develop along widely differing timetables, depending on the life stages of the species. In this review, I will provide background on embryonic development and metamorphosis of the nervous system, but will focus mostly on intercellular interactions that play key roles no matter what the timetable, no matter what the extent of postembryonic reorganization, of the developing nervous system. I will go into most depth on intercellular interactions during development of the antennal system in Manduca.

EMBRYONIC DEVELOPMENT

Investigation of the development of the nervous system during embryonic development in insects has a long history and has been reviewed many times. The central nervous system, or CNS, arises from the ventral ectoderm of the embryo (Bate 1976, Doe & Goodman 1985). Regularly spaced cells in the ventral ectoderm enlarge and bud off, into the interior of the embryo, in clusters that form the primordia of the ganglia and brain. These cells then undergo a stereotyped set of mitotic divisions to produce neurons; the lineage of many of the uniquely identified neurons in the grasshopper and in Drosophila, is known completely, and is invariant from individual to individual. Soon after neurons are born, they begin to extend axons along specific paths toward their targets (see Bastiani et al. 1985). Once they reach their targets, they form synapses, and the specific neural substrates for behavior are elaborated.

METAMORPHIC ADULT DEVELOPMENT

After hatching, the nervous system continues to change. In species that do not undergo metamorphosis, the changes are likely to be subtle, mostly in response to input, falling under the rubric of “experience-dependent plasticity.” In hemimetabolous insects, such as grasshoppers and crickets, each molt involves the production of new sensory neurons, which then must send axons into the CNS to carry their input. This continuous addition of new sensory neurons and the challenge of integrating them into the animal’s neural circuitry has been studied extensively in the cercal system of the cricket (Murphey 1985). In grasshoppers, Arbas (1983), among others, showed regressive as well as progressive changes. In holometabolous insects, the larval nervous system produced during embryogenesis is much more dramatically altered during metamorphosis to produce an adult nervous system that includes some components of the larval system and some entirely new components (Levine et al. 1995). The overall shape of the CNS is modified, under the control of the Broad complex genes in Drosophila (Restifo et al. 1995). Best understood in Manduca, some of the larval circuitry undergoes extensive reorganization, involving regression and subsequent new growth of dendrites and axons. This reorganization occurs under the control of the ecdysteroid hormones (Truman 1996, Weeks & Levine 1990). In addition to reorganization of larval structures, new cells and circuits are added to the adult nervous system to accommodate adult-specific functions. Imaginal disks give rise to new appendages, with new sensory neurons and with new muscles to be innervated by motor neurons in the CNS. New neurons are born in the CNS and either become incorporated into existing circuitry, or make new neural centers (e.g., Sorensen et al. 1991). The visual and antennal systems have been studied in most detail. I will focus on the antennal system, below.

INTERCELLULAR INTERACTIONS IMPORTANT IN NEURAL DEVELOPMENT

How do only certain ectodermal cells acquire a neuronal fate? How do new neurons acquire a specific identity? How are neural circuits formed and modified? As in the developing nervous systems of all species with complex nervous systems that have been studied, molecular details are still sparse, but interactions between cells are being understood to be critical for normal development in insect nervous systems. Here, I will discuss some of the major types of interactions, and provide examples of work addressing the underlying mechanisms for the intercellular influences. The experiments cited will come from studies of insects with many different life patterns, in part to make the point that mechanisms appear to be broadly similar. In the part of the ventral ectoderm made competent to give rise to nervous system by the action of “proneural” genes, lateral inhibition determines which cells will become nervous system and which will remain in the ectoderm. Interactions between the products of the “neurogenic” genes Notch and Delta have been shown to be involved in this lateral inhibition and are being studied in detail (Anderson & Jan 1997). Once cells have delaminated from the ectoderm, they acquire specific fates; that is to say, they become specified to give rise to specific sets of cells (see Jan & Jan 1994, Skeath 1999). That some of this specification is accomplished via cell-cell interactions was revealed elegantly by, who used laser ablation methods to kill neuronal precursors in grasshopper embryos, to see whether new cells would replace their progeny. Whether the intercellular interactions trigger “master” regulatory genes that control large sets of genes encoding cellular properties, or whether smaller sets of genes are turned on individually is not known. Recent studies, however, have identified genes that specify unique cell fates in Drosophila (Doe & Skeath 1996). After neurons have adopted a specific fate, they begin to express their cell-specific characteristics. Most noticeable among these is the trajectory of the axon and choice of synaptic target. Axon extension is very complex and involves recognition of and reaction to many dozens of molecules in the environment. Growing axons, tipped by actively exploring growth cones, may respond to molecules deposited by other cells into the extracellular space, and to molecules on the surfaces of other cells, as well as to soluble molecules released by other cells (Auld 1999, Garcia-Alonso 1999, Van Vactor & Lorenz 1999). Glial cells play special roles in the guidance of early axons (Klambt et al. 1999). Work in Drosophila (e.g., Harrelson & Goodman 1988) has led to a comprehensive understanding of families of cell-surface molecules that play roles across phyla. Development of synapses has been shown to involve two-way interactions between pre- and postsynaptic elements. The developing neuromuscular junction has been used to great advantage in this area (Keshishian et al. 1996), in large part because pre- and postsynaptic partners can be identified at the level of the individual cell.

THE ANTENNAL SYSTEM AS A MODEL SYSTEM FOR STUDIES OF NEURON-GLIA INTERACTIONS

The developing antennal, or olfactory, system offers excellent opportunities for the exploration of intercellular interactions during neural development. Studies in several laboratories have shown the antennal (olfactory) system of the moth Manduca sexta, in particular, to be an advantageous system for detailed study of the interactions among neurons and between neurons and glial cells that lead to the creation of olfactory synaptic glomeruli, which are common to both vertebrate and invertebrate species, and, more recently, of the neuron-glia interactions important in axon guidance. This system is useful for developmental studies for a number of reasons. Olfactory receptor neurons (ORNs) and their postsynaptic targets arise independently during postembryonic development, and are located at some physical distance from each other, allowing the two populations of neurons to be manipulated independently with ease (Schneiderman et al. 1986). The antennal (olfactory) nerve is long from the earliest stages, and is convenient for imaging. Within the CNS, neurons and glial cells are born at different times, so that interference with glial proliferation does not affect neuronal proliferation (Oland et al. 1988). Although the number of cells involved is generally much smaller, cellular organization of the primary olfactory center in Manduca is strikingly similar to that of vertebrate olfactory bulbs (Boeckh et al. 1990, Hildebrand & Shepherd 1997), in that it is organized into discrete glomeruli that contain the synapses of olfactory receptors and their target neurons.

CELLULAR ORGANIZATION OF THE MATURE ANTENNAL SYSTEM IN MANDUCA SEXTA:

The ORNs have their cell bodies in 2cm-long paired antennae, and their axons project via the antennal nerve to the brain. In addition to other ORNs, male antennae have ORNs specialized to detect the female’s sex pheromone (Kaissling et al. 1989). The 330,000 axons (Oland & Tolbert 1988) of the ORNs of each antenna terminate in about 64 glomeruli of the ipsilateral antennal lobe of the brain (Rospars & Hildebrand 1992). In the male, the axons of the ORNs responsive to female sex pheromone describe a separate dorsal "macroglomerular complex” (Christensen & Hildebrand 1987). The "ordinary" glomeruli present in both sexes are arrayed in roughly a single layer around a coarse central neuropil. Antennal-lobe neurons branch in the glomeruli, and one major class of neuron, the projection neurons, project an axon out of the antennal lobe to higher centers in the protocerebrum. All glomeruli are surrounded by glial-cell borders (Tolbert et al. 1983; Oland & Tolbert 1987; Rössler et al. 1998).

DEVELOPMENT OF THE ANTENNAL SYSTEM:

The ORNs of the antennae and their targets, the antennal lobes, arise during metamorphosis. The antennae arise from imaginal disks that evert and begin to develop at the onset of metamorphosis. ORNs are born during stages 1 and 2 of the 18 stages (each roughly one day long) of metamorphic adult development. (Sanes & Hildebrand 1976). Almost immediately they extend axons, which begin to reach the brain at stage 3. ORN axons continue to grow into the brain from ever more distal antennal segments until about stage 9. The antennal lobe develops essentially de novo during metamorphosis (Kent 1985) from 5 neuroblasts that divide throughout larval life to produce the three clusters of neurons of the adult antennal lobe (Sorensen et al. 1991). By stage 2 of metamorphic development, neurons of the antennal lobe are postmitotic; they extend neurites into a small neuropil which is ensheathed by a continuous border of glial cells. As the first ORN axons begin to arrive, they pierce the glial border and encircle the neuropil, just beneath the layer of glia, before terminating in a fringe (Oland & Tolbert 1987; Oland et al. 1990). From late stage 5 through stage 6, ORN terminals elaborate and segregate into nodular "protoglomeruli", which become enveloped by glial cell bodies and processes. The neurites of antennal-lobe neurons reach outward to overlap with the axon terminals (Oland et al. 1990, Malun et al. 1994), and synapses are formed in large numbers (Tolbert 1989).

INFLUENCE OF ORN AXONS ON DEVELOPMENT:

Hildebrand et al. (1979) observed that if antennal ORN axons are prevented from innervating the antennal lobe during development in Manduca, the resulting lobe lacks glomeruli. We found that, without ORN input, development is abnormal beginning as soon as antennal axons would normally have begun to reach the brain (Oland & Tolbert 1987). Glial cells remain restricted to a rim surrounding the neuropil, and that neuropil is composed of neurites of antennal-lobe neurons that branch diffusely, rather than in glomerular tufts (Oland et al. 1990). Even more intriguingly, Schneiderman et al. (1986) and, more recently, Rössler et al. (1999) used antennal transplantation to show that the male-specific ORN axons of male antennae have the ability to induce a macroglomerular complex in a genetically female host antennal lobe. The axons from the transplanted male antenna induce some female antennal-lobe neurons to send a neurite branch into a macroglomerular complex, and these neurons now respond to female sex pheromone, clearly indicating a strong influence of antennal axons on development of target neurons.

NEURON-GLIA INTERACTIONS IN GLOMERULUS DEVELOPMENT:

Taking advantage of the fact that glia proliferate later than neurons in the antennal system, we reduced the number of glial cells, while maintaining apparently normal numbers of both antennal and antennal-lobe neurons, to ask whether glial cells act as intermediaries in developmental interactions between ORN axons and antennal-lobe neurons (Oland et al. 1988, Oland & Tolbert 1988, Baumann et al. 1996). When glial numbers are reduced significantly, ORN axons begin to form protoglomeruli, but the protoglomeruli dissipate before the neurites of antennal-lobe neurons grow out to meet them; the resulting neuropil lacks glomeruli, and closely resembles that of lobes that develop in the absence of ORN axons. On the other hand, the few glial cells that remain undergo the changes in shape and position seen in glia in normal lobes, indicating that they are responding to ORN axon ingrowth. These experiments lend major support to the notion that, despite the innate ability of ORN axon terminals to describe "protoglomeruli" upon entering the antennal lobe (Oland et al. 1990), glial cells are required to stabilize the protoglomerular organization of the axon terminals while the neurites of antennal-lobe neurons grow out to meet the axons. What are the molecular substrates for this neuron-glia interaction? We (Krull et al. 1994a) found evidence for the existence of tenascin-like molecules on the surfaces of the glial cells in the antennal nerve in the antennal-lobe neuropil during ORN axon ingrowth and glomerulus formation. Tenascin in other systems can inhibit the growth of axons, so, in a second set of experiments (Krull et al. 1994b), we tested the responses of antennal-lobe neurons to purified mouse CNS tenascin as a substrate for neurite outgrowth in culture and found that many antennal-lobe neurons avoided growing on the purified tenascin. These findings taken together support the idea that tenascin-like molecules on glial cells might constrain the growth of the neurites of certain antennal-lobe neurons in the developing glomeruli, although further work in this area is needed.

NEURON-GLIA INTERACTIONS IN SORTING AND TARGETING OF ORN AXONS:

In ongoing studies of the molecular mechanisms by which ORN axons find the appropriate target glomeruli, we have discovered that glial cells play a key role in sending ORN axons to the correct target glomeruli. We have found that a subset of the ORN axons in Manduca strongly expresses molecules likely to be identical to, or closely related to, the Manduca form of fasciclin II during the period of glomerulus development. Fasciclin II, a member of the immunoglobulin superfamily (Harrelson & Goodman 1988), is a well documented cell adhesion molecule, known to play roles in guidance of axons in the developing grasshopper (e.g., Bastiani et al. 1987) and in Drosophila (Grenningloh et al. 1991). In situ hybridization and immunocytochemical experiments indicate that scattered ORNs in the antenna of Manduca express fas II (Higgins et al. 1998). Serial reconstructions reveal that fas II-expressing ORN axons terminate in a reproducible subset of glomeruli. The fas II-positive axons are scattered singly or in small bundles throughout the nerve until they reach a glia-rich zone, at the entrance to the antennal lobe, where axons abruptly change their trajectory and sort into glomerulus-specific, fas II-positive or -negative bundles (Rössler et al. 1999). This finding led us to ask whether ingrowing ORN axons must interact with the cluster of glial cells that they encounter at the entrance to the antennal lobe in order to be able to sort into glomerulus-specific fascicles. Using the same paradigm as used previously to reduce the number of glomerulus-associated glial cells, we reduced the number of glial cells in the glia-rich axon sorting zone, and found that fas II-positive axons no longer sorted into fascicles based on their expression of fas II or terminated in the appropriate part of the antennal lobe (Rössler et al. 1999). Thus, glial cells appear to be essential to the sorting of ORN axons into bundles destined to terminate in specific glomeruli, just as they are essential to the stabilization of forming glomeruli.

CONCLUSIONS

A great deal is known about the development of the insect nervous system. In recent years, the critical importance of intercellular interactions has been recognized, and the molecular underpinnings of these interactions are being elucidated. Because neural development across disparate species involves solving many of the same problems, many cellular and molecular mechanisms being characterized in insects are being found to be very similar to those underlying nervous system development in other invertebrates and invertebrates. Thus studies in other species can inform us about likely mechanisms in insects, and studies in insects will continue to play an important role in helping to elucidate development in less accessible mammalian systems.

BIBLIOGRAPHY

Anderson, DJ, & YN Jan (1997) in Molecular and Cellular Approaches to Neural Development, Oxford University Press, NY, p. 26

Arbas, EA (1983) J. Comp. Neurol. 216:369

Arendt, D, & K Nubler-Jung (1999) Development 126:2309

Auld, V (1999) Cell Mol. Life Sci. 55:1377

Bastiani, MJ, CQ Doe, SL Helfand, & CS Goodman (1985) Trends Neurosci. 8:257

Bastiani MJ, AL Harrelson, PM Snow, & CS Goodman (1987) Cell 48:745

Bate, M. (1976) J. Embryol. Exp. Morphol. 35:107

Baumann, PM, LA Oland, & LP Tolbert (1996) J. Comp. Neurol. 373:118

Boeckh J, P Distler, KD Ernst, M Hosl, & D Malun (1990) In Chemosensory Information Processing (ed. D. Schild), Springer-Verlag, Berlin, p. 20

Christensen, TC, & JG Hildebrand (1987) J. Comp. Physiol. A 160:553.

Doe, CQ, & CS Goodman (1985) Dev. Biol. 111:193

Doe, CQ, & JB Skeath (1996) Curr. Opin. Neurobiol. 6:18

Garcia-Alonso, LA (1999) Cell Mol. Life Sci. 55:1386

Grenningloh G, EJ Rehm, & CS Goodman (1991) Cell 67:45

Harrelson, AL & CS Goodman (1988) Science 242:700

Higgins, MR, PA Eckholdt, NJ Gibson, LA Oland, PF Copenhaver, LP Tolbert (1998) Soc. Neurosci. Abstr. 24:1142

Hildebrand JG, & GM Shepherd (1997) Annu. Rev. Neurosci. 20:595

Hildebrand JG, LM Hall, & BC Osmond (1979) Proc. Natl Acad. Sci. 76:499 Jan, Y.N. & L.Y. Jan (1994) Current Opinion in Neurobiology 4:8-13

Kaissling, K-E, JG Hildebrand, & JH Tumlinson (1989) Arch. Insect Biochem. Physiol. 10:273

Kent, KS (1985) PhD dissertation, Harvard University, Cambridge, MA Keshishian, H, K Broadie, A Chiba,& M Bate (1996) Annu. Rev. Neurosci. 19:545

Klambt, C, K Schimmelpfeng, & T Hummel (1999) Adv. Exp. Med. Biol. 468:23

Krull CE, DB Morton, A Faissner, M Schachner, & LP Tolbert (1994a) J. Neurobiol. 15:515

Krull CE, LA Oland, A Faissner, M Schachner, & LP Tolbert (1994b) J. Neurobiol. 25:989

Malun, D, LA Oland, & LP Tolbert (1994) J. Comp. Neurol. 350:1

Murphey, R.K. (1985) Trends in Neurosci. 8:120

Oland, LA, & LP Tolbert (1987) J. Comp. Neurol. 255:196

Oland, LA, & LP Tolbert (1996) J. Neurobiol. 30:92

Oland LA, & LP Tolbert (1998) J. Neurobiol. 36:41

Oland LA, WM Pott, MR Higgins, & LP Tolbert (1998) J. Comp. Neurol. 398:119

Oland LA, G Orr, & LP Tolbert (1990) J. Neurosci. 10:2096

Restifo, LL, PS Estes, & C Dello Russo (1995) Eur. J. Entomol. 92:169

Rospars, JP, & JG Hildebrand (1992) Cell Tissue Res. 270:205

Rössler, W, LP Tolbert, & JG Hildebrand (1998) J. Comp. Neurol. 396:415

Rössler, W, LA Oland, MR Higgins, JG Hildebrand, & LP Tolbert (1999) J. Neurosci. 19:9865

Sanes JR, & JG Hildebrand (1976) Dev. Biol. 51:282

Schneiderman, AM, JG Hildebrand, MM Brennan, & JH Tumlinson (1986) Nature 323:801

Skeath, JB (1999) Bioessays 21:922

Sorensen, KA, NT Davis, & JG Hildebrand (1990) Soc. Neurosci. Abstr. 16:1147

Tolbert, LP (1989) Synapse 3:83 Tolbert LP, SG Matsumoto, & JG Hildebrand. (1983) J. Neurosci. 3:1158

Truman, JW (1996) Dev. Neurosci. 18:87

Van Vactor, D, & LJ Lorenz (1999) Cell Mol. Life Sci. 55:1355

Weeks, JC, & RB Levine (1990) Annu. Rev. Neurosci. 13:183

Copyright: The copyrights of this work belong to the author. This document appears in the Plenary Lectures Leslie Tolbert ABSTRACT BOOK I – XXI-International Congress of Entomology, Brazil, August 20-26, 2000.

More Topics on the Wing

About Us