Sie sind auf Seite 1von 50

2.

3 Functional Development of the Neuromusculature


D E Featherstone, University of Illinois at Chicago, Chicago, IL, USA K S Broadie, Vanderbilt University, Nashville, TN, USA
2005, Elsevier BV. All Rights Reserved.

2.3.1. Introduction and Overview 2.3.1.1. Types of Cells 2.3.1.2. Overview of Neuromuscular Connectivity 2.3.1.3. Overview of NMJ Neurotransmission 2.3.1.4. Modulation of NMJ Neurotransmission 2.3.2. Development of Motor Neuron Function 2.3.2.1. Motor Neuron Morphogenesis 2.3.2.2. Development of Motor Neuron Electrical Properties 2.3.2.3. Synaptogenesis in the CNS 2.3.3. Development of Glial Function 2.3.3.1. Central Glia and CNS Scaffolding 2.3.3.2. Peripheral Glia and Motor Neuron Pathfinding 2.3.3.3. Glial Wrapping of Peripheral Axons 2.3.4. Development of Muscle Function 2.3.4.1. Muscle Morphogenesis and Patterning 2.3.4.2. Development of Muscle Electrical Properties 2.3.4.3. Development of Muscle Contractile Properties and Patterned Movement 2.3.5. Development of NMJ Presynaptic Function 2.3.5.1. Functional Development of the Presynaptic Terminal 2.3.5.2. Maturation of Neurotransmitter Release 2.3.5.3. Maturation of Mature Synaptic Properties 2.3.5.4. Differentiation of Synaptic Vesicle Pools and Cycling 2.3.6. Development of NMJ Postsynaptic Function 2.3.6.1. The Muscle Membrane Prior to Innervation 2.3.6.2. Development of Postsynaptic Glutamate Receptors 2.3.6.3. Molecular Maturation of Postsynaptic Signaling 2.3.7. Postembryonic Maturation of the NMJ 2.3.7.1. Elaboration of Synaptic Architecture 2.3.7.2. Formation of the SSR 2.3.7.3. Developmental Feedback between Pre- and Postsynaptic Cells 2.3.8. Development of Activity-Dependent Synaptic Plasticity 2.3.8.1. Development of Activity-Dependent Functional Plasticity 2.3.9. Distant Signals: Hormonal Effects on NMJ Development and Function 2.3.9.1. Evidence for Hormonal Regulation of NMJ Development 2.3.9.2. Modulation: Short-Term Hormonal Regulation of NMJ Function 2.3.10. Unanswered Questions

85 86 92 93 95 96 96 101 102 103 103 104 104 106 106 109 110 111 111 112 113 113 114 114 114 117 118 118 119 119 120 120 120 120 121 121

2.3.1. Introduction and Overview


The neuromusculature drives movement. It is composed of three functionally intertwined cell types: (1) beautifully patterned striated body muscles; (2) architecturally elaborate motor neurons that precisely innervate these muscles; and (3) peripheral glia that maintain and modulate motor neuron transmission. For the last century or more, numerous insect species have been exploited to study the

basis of neuromuscular development and function. The primary attractions of insect systems include large individually identifiable neurons, simple neuromuscular architecture, and easy propagation and manipulation of insect species in the laboratory. A well-refined maps of insect neural circuits and detailed physiological insight into the bases of insect neuromuscular development and function are provided. Until approximately 20 years ago, most

86 Functional Development of the Neuromusculature

p0010

p0015

insect physiologists focused on species such as grasshoppers, crickets, cockroaches, or moths, which have particularly large, experimentally tractable cells. In the last two decades, increasingly, the vast majority of studies in insect physiology and neuromuscular development have shifted to the tiny fruit fly, Drosophila melanogaster. The reason, of course, is that Drosophila has emerged as a premier genetic model system. This amenability to genetic manipulation, coupled with the increasing molecular focus of physiology and development, has led to the fruit flys dominance as the insect model for studies in development, physiology, and behavior, despite the considerable disadvantage of small size compared to larger insect species. Although other insect species still dominate fields including ecology, toxicology, and endocrinology, modern insect development and physiology is, for the most part, now synonymous with Drosophila development and physiology. Thus, an accurate review of advances in insect neuromuscular development focuses on Drosophila, and this chapter will focus strongly on details of the Drosophila neuromusculature. The neuromusculature has been most intensively investigated in mammals, but, in general, neuromuscular function and development is very similar in insects and mammals. There are, however, some important differences. The most obvious difference is that in insects, muscle contraction is triggered by the excitatory neurotransmitter glutamate. The vertebrate neuromuscular junction (NMJ) uses acetylcholine. However, glutamate is the primary excitatory synapse type in the vertebrate central nervous system (CNS). Thus, in addition to being a general model of the insect NMJ, the Drosophila NMJ has also served as a leading genetic model of glutamatergic synapses. Drosophila studies have concentrated on glutamatergic type I NMJ synapses on embryonic and larval muscles, which drive muscle contraction. There has been much less focus on type II or type III synapses, which serve primarily modulatory functions. Therefore, unless specified otherwise, this chapter will discuss Drosophila type I NMJ synapses. The focus of this chapter is on the acquisition and maturation of physiological properties (e.g., functional development) of these NMJs in the embryo and larva. This chapter does not discuss the development that occurs in pupae during metamorphosis (see Chapters 2.2 2.4, and 2.5). This chapter begins with an overview of the components and function of the mature larval Drosophila neuromuscular system (see Section 2.3.1). The cell types of the neuromusculature, the physiological properties of these cells, and patterns

of neuromuscular connectivity are discussed (see Sections 2.3.1.1 and 2.3.1.2), and then summarize the functional properties of NMJ transmission (see Sections 2.3.1.3 and 2.3.1.4). The bulk of the chapter is then devoted to the functional development of this neuromuscular system. First, the origin and morphological development of the Drosophila neuromusculature is given, describing the specification of cellular identities and the maturation of neuromuscular patterning and connectivity (see Sections 2.3.22.3.4). This information is not intended to be exhaustive, but rather to provide references for the reader to accurately place functional maturation within the wider context of overall neuromuscular development. More comprehensive descriptions of neuromuscular structural and functional development can be obtained in Chapters 1.10 (neurogenesis), 2.1, and 2.2 (myogenesis), 2.4, and 2.5. Second, the functional development of the NMJ synapse is focused. The maturation of neurotransmitter release (see Section 2.3.5), induction of postsynaptic specializations and the appearance of postsynaptic receptor fields (see Sections 2.3.62.3.7), the interaction between pre- and postsynaptic cells (see Section 2.3.7), developmental changes in functional plasticity (see Section 2.3.8), and the effects of hormones on NMJ function (see Section 2.3.9) are discussed. The chapter is concluded by considering the most pressing unanswered questions and suggesting fruitful directions for investigation in the future.
2.3.1.1. Types of Cells

2.3.1.1.1. Motor neurons Compared to vertebrates, insects have a relatively simple neuromusculature composed of relatively a few cells (see Figures 1 and 2). Rather than a motor neuron pool (hundreds of cells) innervating a multicellular muscle (thousands of myotubes), as in mammals, insects are characterized by a single motor neuron innervating a single myotube. Moreover, as they are bilaterally symmetrical segmented animals, the basic unit of the insect neuromusculature is a single hemisegment, which is simply reiterated a number of times depending on the length of the animal. Each hemisegment is nearly identical, with a few segment specific modifications overlying a common cellular plan. Thus, in the Drosophila embryo/larva body, thoracic segments T1T3 are highly similar (although each is unique) and the abdominal segments A2A7 are identical (A1 and A8 have slight variations). Insect neurobiologists work with identified, named neurons. In most insect species, neurons are readily discernible via basic imaging techniques, and can be individually identified in living preparations based

Functional Development of the Neuromusculature

87

Figure 1 Diagrammatic representation of the Drosophila embryonic/larval neuromusculature, viewed in cross-section. One of each general type of neuron is shown: motor neuron (mn), interneuron (in), and sensory neuron (sn). The cell bodies of motor neurons and interneurons lie within the cortex of the ventral ganglion. The cell bodies of peripheral sensory cells are found near the cuticle in the periphery. Motor neurons send axons either contralaterally or (as shown) ipsilaterally along the ventral nerve and into the peripheral musculature, where the axons (yellow) synapse onto one (or sometimes more, depending on the neuron) appropriate muscle target(s) (red rectangles). Note that the most distal muscle targets are also most dorsal, since axons grow along the periphery from the ventral side of the animal. Simultaneously, sensory neurons project their axons from the periphery back toward the CNS along ventral nerves. Although the motor and sensory axons are shown traversing different nerves in this diagram for clarity, each nerve actually contains multiple motor axons going out and multiple sensory axons going in. Interneurons lie entirely within the ventral ganglion. Neuronneuron synapses occur within the neuropil (NP), the area of the ventral ganglion containing most axonal and dendritic tracts, but no cell bodies.

p0025

on cell body size, location, and axonal projection. For example, drosophilists commonly visualize a single raw prawn 3 (RP3) motor neuron innervating a specific pair of ventral, longitudinal muscle cells (muscles 6 and 7; Broadie et al., 1993). The application of cell-filling techniques and antibodies further increases the number of identified, named neurons. Due to their large size, clear axonal projections, and known function, motor neurons have received the most focused analysis and are best characterized. In the large grasshopper embryo, all identifiable motor neurons have been described. Motor neuron cell identities have been highly conserved among diverse insect species, allowing studies from different insect species to cross-fertilize and inform parallel studies in other species (Thomas et al., 1984). Knowledge gleaned from the grasshopper, in particular, provided the vital framework for subsequent studies in Drosophila. In the Drosophila embryo/larva, all motor neurons in the identical A2A7 abdominal segments have been named and mapped. A combination of transgenic markers and antibodies also assign unique molecular identities to many of these cells.

Motor neurons share two general organizational properties. First, the neuronal structure is simple, with either a unipolar structure or a weakly bipolar structure. Typically, motor neuron axonal projections are either ipsilateral or contralateral, with a limited dendritic tree branching from a single process near the soma. Second, motor neurons innervating muscles of similar position (dorsal versus ventral) and orientation are often clustered, with overlapping dendritic trees that converge in the same region of the neuropil. Even in cases where neurons innervating similar muscles have soma positions that are quite different, the cells share a common domain of dendritic arborization within the CNS (Landgraf et al., 1997). Each hemisegment has 30 muscles and each muscle is (with a few exceptions) innervated in a 1 : 1 arrangement by one of 30 major motor neurons. These motor neurons display distinctive, enlarged type I NMJ boutons that drive muscle contraction. In addition to these 30 major motor neurons, there are four motor neurons that each innervate a large field of muscles and have characteristically smaller NMJ boutons. These

88 Functional Development of the Neuromusculature

Figure 2 Microscopic views of the Drosophila embryonic neuromuscular anatomy. (a) Confocal image of an intact (whole mount) embryo viewed from above (anterior is left). Anti-FasII antibodies have been used to visualize the motor nerves (green). Note the segmentally repeated peripheral nerves extending from the ventral ganglion out toward the body wall neuromusculature. (b) An enlarged view of a section of the embryonic peripheral neuromusculature. As in (a), axons and growth cones (brown) have been visualized using antibodies that recognize the cell adhesion molecule FasII. In each hemisegment, a group of axons extends from the ventral ganglion, where they first encounter the ventral longitudinal muscles (two hemisegments of which are enclosed by the dashed green box). Some axons continue on toward more distal, dorsal muscles (the dorsal midline is near the bottom of the image). The embryo in (b) has been split dorsally and laid flat; in the intact animal, the axons extending toward the bottom of the image (and the muscles they innervate) would curve up out of the page and back toward the ventral ganglion, such that the bottom of the image (which shows the dorsal midline) would overlap the midline of the ventral ganglion. (c) A close-up of the ventral ganglion, showing a subset of neurons visualized using green fluorescent protein (GFP) expressed specifically in cholinergic neurons (GAL4-Cha;UAS-GFP). Note the two parallel areas devoid of cell bodies; these are the neuropile. (d) A more detailed image of approximately 2.5 neuromeres of ventral ganglion, showing axons (green) and cell bodies (red). The overall appearance of a ladder is formed by segmentally repeated lateral and commissural axon tracts. (e) A detailed view of a portion of two neighboring hemisegments showing NMJs (brown) on ventral longitudinal muscles 7 and 6. The edge of the ventral ganglion is visible in the upper left corner. (f) An electron micrograph cross-section of a stage 17 embryonic neuromuscular bouton. The arrow indicates a presynaptic T-bar, surrounded by synaptic vesicles.

Functional Development of the Neuromusculature

89

p0035

p9000

motor neurons are believed to mediate modulatory functions. Thus, in total, there are 34 identified motor neurons per abdominal hemisegment. Until recently, relatively little was known about the functional properties of Drosophila motor neurons in situ. Recent physiological assays have focused on a small, medial dorsal cluster of motor neurons with both ipsi- and contralateral axons terminating in type I glutamatergic muscle terminals (Hoang and Chiba, 2001; Rohrbough and Broadie, 2002). These cells have resting potentials of 50 to 60 mV and fire repetitive action potentials (AP) during a depolarizing current injection (Rohrbough and Broadie, 2002), consistent with the sharp bursts of action potentials that drive peristaltic movement in vivo (Broadie and Bate, 1993e; Cattaert and Birman, 2001). The majority of motor neurons exhibit adaptive AP firing patterns, characterized by a variable decrease, but not termination, in spike frequency that is likely specific to each cell type (Rohrbough et al., 2003). A few motor neurons exhibit tonic firing at constant frequency during prolonged depolarization. Similar adaptive and tonic AP firing characteristics have been previously described in cultured Drosophila neurons (Saito and Wu, 1991; ODowd, 1995; Zhao and Wu, 1997; Schmidt et al., 2000). In situ, the heterogeneity of AP firing properties among specific motor neurons is likely regulated by intrinsic variables such as different patterns of impinging synaptic input. A considerable amount is known about the voltage-gated ion channels that mediate electrical signaling in Drosophila motor neurons (Wu and Ganetzky, 1992; see also Chapters 2.6 and 5.1). The primary voltage-gated Na channel is encoded by the paralytic (para) gene, although at least one other voltage-gated Na channel is also present (Loughney et al., 1989). The other primary inward current is carried by Ca2 via several distinct classes of voltage-gated Ca2 channels. The gene cacophony (cac) encodes the L-type voltage-gated Ca2 channel in Drosophila (Kulkarni and Hall, 1987). Null mutations in the Dmca1A gene result in embryonic lethality; partial loss-of-function mutations reduce the L-type Ca2 current (Smith et al., 1996). Presynaptic N-type Ca2 channels regulate neurotransmitter release (Rieckhof et al., 2003). The hyperpolarizing outward K current can be divided into at least four components: two voltagegated K currents (IA, IK) and two calcium-gated K currents (ICF , ICS). The Shaker (Sh; encodes IA channel) and ether-a-gogo (eag; encodes a subunit of multiple K channels) genes were identified based on hyperactivity of mutants under anesthesia (Ganetzky and Wu, 1983; Ganetzky et al., 1999).

Double eag;Sh mutants show a highly elevated frequency of action potentials in motor neurons and heightened neurotransmission at the NMJ. In cell body recordings, the principal IA current is not carried by Sh channels but rather by channels encoded by the Shaker cognate l (Shal) gene, which exhibits different kinetics (Tsunoda and Salkoff, 1995a). Null Shaker mutants do not display altered IA currents in records from embryonic motor neuron cell bodies (Rohrbough et al., 2003). Thus, it has been suggested that Shal encoded IA dominates in neuronal cell bodies and/or axons, whereas Shaker encoded IA dominates in synaptic terminals (Tsunoda and Salkoff, 1995b). The neuronal IK current is carried by at least two channels; the larger component mediated by a slowly inactivating channel encoded by Shaker cognate b (Shab) and the smaller component mediated by a noninactivating channel encoded by Shaker cognate w (Shaw) (Solc and Aldrich, 1988; Tsunoda and Salkoff, 1995a). The ICF current is encoded by the slowpoke (slo) locus and slo mutants completely lack the Ca2 activated K current (Elkins et al., 1986; Elkins and Ganetzky, 1990). The channel or channels carrying the ICS current have not yet been identified (Chopra et al., 2000). Acetylcholine (ACh) and GABA are the primary excitatory and inhibitory neurotransmitters, respectively, in the insect CNS (see Chapters 5.4 and 5.9). This conclusion is based primarily on neurotransmitter and receptor expression studies, electrophysiology, and (in Drosophila) transgenic reporters specific to transmitters green fluorescent protein coupled to, for example, choline acetyltransferase expression (Schuster et al., 1993; Jonas et al., 1994; Aronstein et al., 1996; Yasuyama and Salvaterra, 1999). Glutamatergic cells and glutamate receptors (both NMDA and non-NMDA types; see also Section 2.3.1.3) are also expressed throughout the CNS (Littleton and Ganetzky, 2000), but less is known about their circuit function. Larval motor neurons in situ respond to iontophoretic application of ACh, GABA, and glutamate (Rohrbough and Broadie, 2002). As expected, ACh is strongly excitatory, generating sustained depolarizations and usually driving bursts of action potentials. ACh responses are mediated by nicotinic ACh receptors as evidenced by their reversible blockade by d-tubocurare (dtC; Rohrbough and Broadie, 2002). GABA and glutamate cause inhibitory responses mediated primarily by Cl channel currents. Glutamate-gated currents are particularly prolonged and act to strongly suppress AP firing. Both inhibitory transmitters display similar reversal potentials near normal resting potential ($55 mV), and

90 Functional Development of the Neuromusculature

are reversibly blocked by the Cl channel blocker picrotoxin (Rohrbough and Broadie, 2002). Thus, motor neurons in situ express functional receptors appropriate for excitatory cholinergic input and inhibitory GABAergic and glutamatergic input. Motor neurons display several distinct types of endogenous synaptic activity, ranging from fast, unitary events to sustained excitatory currents and potentials. Both endogenous and evoked events require external Ca2 and are reversibly blocked by dtC application, demonstrating cholinergic synaptic input. These cells display unitary spontaneous excitatory synaptic currents (sepscs; 550 pA) closely resembling fast cholinergic sepscs reported in Drosophila cultured neurons (Lee and ODowd, 1999; Lee and ODowd, 2000; Lee et al., 2003). Most striking, however, are endogenous events with much larger amplitudes and longer durations termed spontaneous rhythmic currents (SRCs) or potentials (SRPs) that appear to be a type of periodic endogenous excitatory motor output, supporting bursts of AP firing (Rohrbough and Broadie, 2002). Interestingly, blockade of NMDA-type glutamate receptors abolishes centrally generated endogenous rhythmic output at the NMJ (Cattaert and Birman, 2001). Thus, while cholinergic input provides the major excitatory synaptic drive to motor neurons, additional forms of central transmission, including glutamatergic input, likely shape the frequency and pattern of motor rhythms controlling coordinated movement. Compared to our knowledge of motor neurons residing in the comparatively simple and invariant central abdominal segments, relatively little is known about the physiology, identity, or connectivity of cells upstream of the motor neurons (Lohr et al., 2002; Landgraf et al., 2003; Taghert and Veenstra, 2003). There are an estimated 200 neurons per hemisegment neuromere. Apart from the few identified neurosecretory cells (Taghert and Veenstra, 2003), which also send projections into the periphery, the vast bulk of these cells are interneurons, both local and projection subtypes. Almost all of these interneurons remain unnamed and their functions remain uncharacterized. Interneurons functionally upstream of motor neurons have cell bodies scattered throughout the central and lateral CNS neuropil and numerous varicosities (presumed synaptic boutons) overlapping spatially with the dendritic arbors of motor neurons (Rohrbough et al., 2000; Rohrbough and Broadie, 2002). Paired synaptic recording and cell labeling experiments indicate that fast synaptic responses in motor neurons are efficiently evoked when stimulation is applied to the neuropil adjacent to the motor

neurons primary dendrites (Rohrbough and Broadie, 2002; Rohrbough et al., 2003), suggesting that cholinergic inputs are formed by ascending or descending longitudinal axons. 2.3.1.1.2. Peripheral glia As in other animals, glia are an abundant cell type in the insect CNS and peripheral nervous system (PNS). Just how abundant, is a matter of debate, as estimates range widely and little effort has been made to catalog glial subtypes. In the Drosophila CNS, glial cells adopt morphologies resembling vertebrate astrocytes, yet essentially nothing is known about their mature functions. In the periphery, defined glial cells are found where peripheral nerves exit the CNS, and at regular intervals along the peripheral nerves (Auld et al., 1995; Sepp et al., 2000). These glia associate tightly with both motor and sensory axons, and have extensive cytoplasmic processes that fully ensheath peripheral nerves. The number of insulating peripheral glial cells per nerve seems to be somewhat variable (810 per hemisegment in Drosophila) and the exact cell body positions are not well mapped. Therefore, there is no general nomenclature or numbering scheme for Drosophila peripheral glia, unlike the situation for both motor neurons and muscles. In addition to bona fide peripheral glia, there are other cell types closely associated with the peripheral nerves. For example, Drosophila also contains nerve associated cells that are characterized by persistent larval expression of the transcription factor Twist. These persistent Twist (PT) cells represent the mesodermal anlage of the adult muscle that divide during larval life to form the adult myoblasts (Bate et al., 1991; Broadie and Bate, 1991). In many animals, peripheral glia have been shown to be essential for action potential conduction, and thus neuromuscular transmission. Similarly, Drosophila peripheral glia tightly wrap around all exposed peripheral nerves, enclosing afferent and efferent axons in a common glial sheath that excludes the hemolymph (Auld et al., 1995). Adjacent glial processes are joined by a specialized intercellular junction called a pleated septate junction (pSJ), which acts as a diffusion barrier similar to the tight junctions of vertebrates (Schulte et al., 2003). Although less highly specialized than the myelin sheath of mammalian Schwann cells, this insect glial sheath seems to serve comparable functions. Mutation of any of three Drosophila genes encoding glial proteins (Gliatactin, Neurexin IV, and Axotactin) causes blockade of motor nerve action potential propagation and, therefore, paralysis (Auld et al., 1995; Baumgartner et al., 1996;

Functional Development of the Neuromusculature

91

Yuan and Ganetzky, 1999). The gliatactin and neurexin IV mutants manifest very subtle disruptions in the glial sheath surrounding the motor nerves, and the conduction blockade can be rescued by reducing [K] in the external bath, indicating that the glial sheath is essential for insulating the peripheral nerves from the high [K] of the insect hemolymph. In contrast, Axotactin appears to be part of a novel glianeuron signaling pathway required for proper electrical signaling in neuronal axons (Yuan and Ganetzky, 1999). The functional roles of the glia in neuronal signaling remain to be elucidated. Nuclei of glial sheath cells lie on the outside of the motor nerve just proximal to the site where nerves enter the body wall musculature. This nerve entry point on the muscle is the location where the glial sheath covering the motor nerve terminates; insect NMJs are not covered in any sort of capping glia, unlike vertebrate NMJs. 2.3.1.1.3. Muscle Drosophila muscle is similar to mammalian muscle: syncytial (multinucleate; arising from myoblast fusion) and beautifully striated with actinmyosin contractile units (sarcomeres). However, the insect muscle can be very different at the level of multicellular organization. For example, in the Drosophila embryo/larva, a given muscle is composed of a single syncytial cell, whereas in mammals, each muscle is composed of a large aggregate of syncytial muscle cells (myotubes). In the adult Drosophila thorax, however, bundles of muscle cells do become the basic contractile unit, paralleling the situation in mammals (see also Chapter 2.2). The insect muscle attaches directly to the epidermis via specialized tendon cells distributed so that individual muscles span longitudinal, transverse, or oblique domains. Each segment has a highly specific, invariant pattern of muscles, each of a certain size and orientation, typically arranged in a number of discrete layers (Bate, 1990; Chapter 2.2). Muscle cells vary greatly in size, depending on the number of myoblasts that fuse to give rise to the mature myotube. In the Drosophila embryo/larva, each abdominal hemisegment A2A7 contains exactly 30 muscles, arranged in three comparably sized layers ($10 muscles/layer; Bate, 1990; Chapter 2.2). Each Drosophila muscle has a distinguishing name, although two different naming schemes are routinely used (Bate, 1993). In Drosophila, larval muscles are geometrically simple (typically rectangular) and nearly isoelectric throughout; their passive electrical properties have been well characterized (Jan and Jan, 1976b; Wu and Ganetzky, 1992; Singh and Wu, 1999). Muscles are weakly electrically coupled within and between

p0070

segments (Ueda and Kidokoro, 1996). In contrast to vertebrate muscle, insect muscle lacks voltage-gated Na channels, and Ca2 influx is required for insect muscle contraction (Aidley, 1975; Jan and Jan, 1976b; Singh and Wu, 1999). In total, five prominent voltage-activated currents are present in mature larval muscle; the voltage-gated calcium current (ICa), two voltage-gated potassium currents (IA, IK), and two calcium-gated potassium currents (ICF, ICS) (Wu and Ganetzky, 1992; Singh and Wu, 1999). The voltage-gated Ca2 channels are present in two subtypes: one with the pharmacological and biophysical properties of vertebrate L-type Ca2 channels (high voltage activation; maximal at 10 to 0 mV threshold, inactivate slowly) and the other with the characteristics of the vertebrate T-type Ca2 channels (low voltage activation threshold) (Gielow et al., 1995). In Drosophila, two Ca2 channel a subunit genes, Dmca1A and Dmca1D, encode channels with six transmembrane domains, both similar to the vertebrate voltage-gated Ca2 channels (Zheng et al., 1995; Smith et al., 1996). Definitive genetic analyses of the function of these gene products in muscle has not yet been done. Of the two voltage-gated K currents, IA is rapidly activated and rapidly inactivated, whereas IK is slowly activated and shows little inactivation during sustained depolarization. IA is selectively eliminated in Sh mutants; the Drosophila Sh gene was the first cloned K channel a subunit, a six transmembrane voltage-gated channel (Kamb et al., 1987; Kamb et al., 1988). Thus, the molecular identity of the principal IA is different in neurons and muscle; Shaker IA is only detectable in a subpopulation of neurons, and even in these neurons does not constitute the major IA current (Baker and Salkoff, 1990; Tsunoda and Salkoff, 1995a). In contrast, the shaw and shab genes encoding IK channels in neurons are also responsible for the IK current present in muscle (Tsunoda and Salkoff, 1995a). Mutations in the Shab gene reduce IK without affecting IA. Of the two Ca2-activated K currents, ICF is rapidly activated downstream of Ca2 channel opening and inactivates, whereas ICS activates slowly and shows little inactivation. Mutations of the slowpoke (slo) gene eliminates ICF in muscle (Elkins et al., 1986; Wu and Ganetzky, 1992). The channel carrying ICS has not yet been identified (Chopra et al., 2000). Additional genes for K channel subunits have been identified in Drosophila muscle, including eag, a structural K channel subunit, and Hyperkinetic (Hk), an auxiliary, membrane associated b subunit (Yao and Wu, 1999). Mutations in eag alter all four K currents in the muscle, reducing all but eliminating none, and it has been

p9005

p9010

p9015

92 Functional Development of the Neuromusculature

p9020

suggested that eag acts as a modulatory subunit in several different K channels (Zhong and Wu, 1993). In contrast, mutations in Hk alter the kinetics of the IA current alone, without affecting any of the others (Yao and Wu, 1999; Wang et al., 2000). In addition to these voltage- and calcium-gated channels, a mechanically gated, stretch-activated large conductance K channel is present in Drosophila muscle (Zagotta et al., 1988; Gorczyca and Wu, 1991). The molecular identity and exact functional significance of these channels remains unknown. Synaptic transmission at the NMJ is mediated by ionotropic glutamate receptors (GluRs), which are l-glutamate gated cation channels expressed in all muscles (see Chapter 5.9). Electrophysiological evidence for other receptors exists (see later sections), but the molecular identity of these proteins remains unknown. The exact subunit composition of any Drosophila channel, either ligand or voltagegated, and the functional consequences of channel modulation also remain largely unknown.
2.3.1.2. Overview of Neuromuscular Connectivity

the most ventral, external muscle group. From the dorsal CNS exit extends the transverse nerve (TN) that runs along the anterior segmental border. The TN carries just a few motor axons together with the processes of neurosecretory cells (Taghert and Veenstra, 2003). 2.3.1.2.2. Morphological and functional classes of NMJ In Drosophila, motor neurons innervating larval muscle have been classified as types I, II, and III (see Chapters 2.1 and 2.2). Most motor neurons ($32/hemisegment) are type I. Type I neurons use glutamate as their primary neurotransmitter and are the principal mediators of muscle excitability and contraction. Type I terminals are usually restricted to one muscle, contain thick, variably branched arbors, and typically extend terminal branches for only a relatively short distance from the entry point. Type I NMJ varicosities, or boutons (Figure 2f), are deeply embedded in the muscle membrane, and are characterized presynaptically by dense pools of small, clear synaptic vesicles (SVs) clustered around T-bars marking glutamate release sites (active zones) and postsynaptically by an extensive field of elaborate muscle membrane foldings called the subsynaptic reticulum (SSR). The SSR is an expanded area of postsynaptic receptivity that contains the GluRs and dense concentrations of other proteins involved in postsynaptic glutamatergic transmission. Morphologically, type I neurons come in two subdivisions: (1) type Ib neurons possess big (25 mm) boutons that innervate muscles on a 1 : 1 basis and (2) type Is neurons possess small (13 mm) boutons that innervate several adjacent muscles. Type Is NMJ boutons have a greatly reduced SSR relative to type Ib. These two neurons drive physiologically different patterns of transmission, varying in excitation threshold, excitatory junctional current (EJC) amplitude, and functional plasticity properties (Kurdyak et al., 1994). It has been proposed that type Ib and Is NMJ terminals are analogous to the well-characterized crustacean tonic and phasic NMJs, respectively. There are only two type II motor neurons per hemisegment, which innervate a large number of adjacent muscles in the dorsal and ventral muscle fields. Compared to type I boutons, type II NMJ terminals are characterized by much longer, much thinner, branched arborizations studded with small (<1 mm) synaptic boutons. Type II terminals are glutamatergic and contain small, clear SVs, but they predominantly contain larger, elliptical dense core vesicles (DCV) holding a variety of neuromodulators (e.g., biogenic amine (octopamine) and small peptide (proctolin) transmitters; Taghert and

p0090

2.3.1.2.1. Patterns of motor innervation Many insects contain segmental ganglia. In Drosophila, however, segmental ganglia are fused into one ventral nerve cord (VNC), although segmental units (neuromeres) can be recognized by landmarks including segmental nerves. Each neuromere consists of a nearly identical set of cortical neuronal somata in characteristic positions, enclosing a soma-free zone (neuropil) dense with synaptic connections. The neuropil is scaffolded by a highly recognizable ladder of two lateral axon tracts (longitudinal connectives) connected across the midline by an anterior and posterior axon tract (transverse commissures). Synaptic contacts within the CNS are restricted to the neuropil excluding the neuronal soma. Most motor axons converge at lateral exits located on the side of each hemisegment neuromere. Two major axon tracts converge on these exits: (1) the anterior intersegmental nerve (ISN) tract and (2) the posterior segmental nerve (SN) tract. A small number of axons also exit the CNS through a dorsal exit point in each segment (see Figures 1 and 2). In the periphery, five major nerves extend from each lateral CNS exit point. First, the ISN carries the motor innervation to the most dorsal muscle set. Second, multiple SN branches innervate the lateral and ventral muscles: the A branch (SNa) innervates the lateral muscle group, the B branch (SNb) innervates the next most lateral/ventral muscle group, the C branch (SNc) innervates the internal ventral muscles, and the D branch (SNd) innervates

p0095

Functional Development of the Neuromusculature

93

p9025

Veenstra, 2003). Type II terminals do not appear to activate ionotropic receptors in the muscle, but likely work through G protein-coupled receptors, most of which have not been characterized. Hence, type II motor neurons are proposed to be modulatory regulators of NMJ synaptic and/or muscle function, although their exact roles are not at all well understood. Type III neuron terminals are similar to type II terminals, but contain larger DCVs harboring a different array of modulators (e.g., Insulin; Taghert and Veenstra, 2003). Type III neurons, like type II, are believed to mediate modulatory functions in the neuromusculature. At an ultrastructural level, the release sites of type III boutons are distinctive, and it is possible that these terminals are largely neurohemal, releasing peptide hormones to the entire body musculature. Types II and III NMJ terminals reside in shallow grooves in the muscle surface and are not associated with a postsynaptic SSR. Each identified muscle cell is innervated by a specific combination of types I, II, and III motor neurons. All muscles possess type Ib innervation, in a 1 : 1 relation, but different muscles are variably innervated by type Is and type II neurons. Type III innervation is usually restricted to the centrally located muscle 12. Each motor neuron contacts each muscle at a highly characteristic nerve entry point. The NMJ structure of each muscle is highly conserved from segment to segment and from animal to animal, so much so that any given NMJ is instantly recognizable to anyone familiar with the Drosophila neuromuscular anatomy. However, the terminals are not invariant and may differ in the number/length of the branches and the number of synaptic boutons. Indeed, the NMJ shows a high degree of structural plasticity that is sensitive to the relative mobility and density of the animals during postembryonic development (Sigrist et al., 2003). Type II terminals are particularly liable to structural modification as a consequence of environmental conditions. Several chemical classes of NMJ neurotransmission exist in different insect species. NMJs are commonly glutamatergic and excitatory, but some insects also have direct inhibitory inputs on the muscle mediated by GABA or glutamate (see also Chapters 5.4 and 5.9). Drosophila embryos and larvae have only excitatory glutamatergic NMJs with no inhibitory GABAergic or glutamatergic receptors (Featherstone et al., 2000). In addition to these classic, fast neurotransmitters, there are a number of other neuromodulators capable of mediating both ionotropic and metabotropic muscle responses (Taghert and Veenstra, 2003). These include

neuropeptides such as pituitary adenyl cyclase activating polypeptide (PACAP)-like peptide, which (probably indirectly) gates and modulates ionic currents (Zhong, 1996) and biogenic amines, such as octopamine and tyramine (Nishikawa and Kidokoro, 1999; Nagaya et al., 2002), which modulate the efficacy of synaptic transmission (presumably via G protein-coupled receptors). The number of these other transmitters, potentially quite large, and the scope of their physiological functions, remains largely unknown.
2.3.1.3. Overview of NMJ Neurotransmission

The presynaptic mechanisms of neurotransmitter release have been particularly highly conserved between invertebrate and vertebrate species. In Drosophila, presynaptic depolarizations are typically triggered by short bursts (1020 Hz) of action potentials, resulting in Ca2 influx through N-type voltage-gated Ca2 channels. Uniform, electronlucent 3040 nm SVs containing glutamate virtually fill the synaptic bouton, but can be functionally differentiated into three discrete pools including: (1) a reserve pool in the bouton interior removed from active cycling except under conditions of high demand; (2) a cycling pool clustered near release sites; and (3) a readily releasable pool (RRP) of vesicles morphologically docked at release sites, available for imminent release (Rodesch and Broadie, 2000). The SV release sites contain a distinctive electron-dense T-bar. In Drosophila and other dipterans, T-bars appear in cross-section as a bar-like density approximately 100200 nm wide, oriented parallel to the cell membrane and supported by a stem approximately 50 nm in height projecting into the cytosol (Lane, 1985; Jia et al., 1993; Prokop, 1999). The true shape of a T-bar, however, is apparently more similar to a table when considered in three dimensions: three or four electron-dense legs or prongs (Rheuben et al., 1999) support each bar; the density typically appears as a T because only one leg is typically visible in cross-section (Atwood et al., 1993). The appearance of the presynaptic density varies among insect species, suggesting that its function is not dependent on a particular shape. In Manduca, for example, the presynaptic density appears simply as a narrow 2 nm bar. In longitudinal sections, T-bars can extend for hundreds of nm (Lane, 1985). The T-bar is surrounded by clear SVs that are typically 3040 nm in Drosophila. Because the T-bar/presynaptic density is presumably the site of synaptic vesicle fusion (or near it), each T-bar traditionally identifies an active zone. However, the exact role of the T-bar (or presynaptic density in

p0115

94 Functional Development of the Neuromusculature

p0125

any organism) in synaptic vesicle cycling is not known. Omega-shaped membrane profiles indicative of exo/endocytosis are not always adjacent to T-bars (Lane, 1985). Omegas that are present occur some distance away and are coated, suggesting that they are endocytic. Verstreken et al. (2002) argued that basal release requires only partial (kiss and run) SV fusion and showed electron micrographs where the few existing vesicles that are interacting with the membrane are not clustered around T-bars. Haghigi et al. (2003) showed micrographs containing apparently functional active zones with clustered vesicles, but no T-bars. If vesicles mark sites of neurotransmitter release, then T-bar components are clearly not prerequisites for neurotransmitter release. The molecular composition of the T-bar is also unknown. In Drosophila, no molecules are yet known to be restricted to the active zone, and no mutants have been identified that eliminate the T-bar structure. Many proteins involved in SV trafficking, fusion, and endocytosis have been exhaustively assayed in Drosophila. Examples include: (1) synaptotagmin, the proposed Ca2 sensor for exocytosis (Littleton et al., 1993b; Broadie et al., 1994; DiAntonio and Schwarz, 1994); (2) multiple components of the soluble NSF attachment protein receptor (SNARE) complex, proposed to mediate SV fusion (Broadie et al., 1995; Schulze et al., 1995; Fergestad et al., 2001); (3) regulators of the SNARE complex such as Uncoordinated 13 (Unc-13) and Ras opposite (Rop)/ Unc-18 (Harrison et al., 1994; Schulze et al., 1994; Aravamudan et al., 1999; Aravamudan and Broadie, 2003); (4) regulators of Clathrin-mediated endocytosis such as a-Adaptin and Endophilin (GonzalezGaitan and Jackle, 1997; Fergestad et al., 1999; Fergestad and Broadie, 2001; Verstreken et al., 2002); and (5) mediators of SV endocytosis and recycling such as the GTPase Shibire/Dynamin (Koenig and Ikeda, 1989; van der Bliek and Meyerowitz, 1991). To a very high degree, the molecular machinery of neurotransmitter release and SV cycling is highly conserved between Drosophila and other species, including mammals. Many reviewers have considered the conserved nature of this machinery of neurotransmitter release in detail (Keshishian et al., 1996; Lloyd et al., 2000; Broadie and Richmond, 2002). Postsynaptic composition and function varies according to the nature of the neurotransmitter. In Drosophila, muscle contraction is generated by a burst of EJCs (510 Hz) at the glutamatergic NMJ. Current influx through glutamate receptors is primarily carried by Na and, to a lesser extent,

Ca2 (Jan and Jan, 1976a). The highly folded SSR is believed to greatly increase the density of voltagegated Ca2 channels near the NMJ. Insect muscles do not contain voltage-gated Na channels. Since the presynaptic arbor can cover a considerable length of muscle area, it is likely that the synaptic response is widespread in order to mediate nearsynchronous muscle contraction. In Drosophila, Ca2 based muscle action potentials can be recorded, but it is not generally thought that larval muscles depend on regenerative potentials in vivo (Broadie and Bate, 1993d). In larger insects, there is clear evidence of muscle action potentials. The nature of the postsynaptic glutamate receptor field at the Drosophila NMJ has been most thoroughly investigated. There are at least three functional classes of GluRs in Drosophila muscle including: (1) extrasynaptic ionotropic receptors; (2) synaptic ionotropic receptors; and (3) metabotropic receptors (Chang et al., 1994; Nishikawa and Kidokoro, 1995; Chang and Kidokoro, 1996; Parmentier et al., 1996; Ciani et al., 1997; Petersen et al., 1997; Zhang et al., 1999). Ionotropic GluRs are multimeric l-glutamategated cation pores that are each composed of four subunits (Madden, 2002). Three different subunits are known to be expressed in embryonic/larval NMJs: GluRIIA, GluRIIB, and GluRIII (Petersen et al., 1997; DiAntonio, personal communication). GluRIII is a required subunit, and GluRIIA and GluRIIB appear to compete for assembly with GluRIII. Thus, native receptors in Drosophila appear to be multimers of GluRIIA/GluRIII, or GluRIIB/GluRIII (DiAntonio, personal communication). The A and B subunits appear to have arisen through a fairly recent genetic duplication event, since the genes are adjacent in the genome, show a high level of sequence identity, and the genetic removal of either shows that neither alone is required for NMJ function and viability (Petersen et al., 1997). Removal of either GluRIIA or GluRIIB alters single channel current amplitude, however, suggesting that the A and B subunits confer different biophysical properties (Petersen et al., 1997). If both A and B subunits are eliminated, the mutant is embryonic lethal and apparently lacks all NMJ transmission. Thus, the presence of either GluRIIA or GluRIIB, but not both, along with GluRIII, is essential for function. Mammalian GluRs were originally classified pharmacologically, based on relative sensitivity to the agonists amino-3-hydroxy-S-methylisoxazole-4propionate (AMPA), kainate, and N-methyl-d-aspartate (NMDA). GluRIIA, IIB, and III amino acid

p9030

p0130

Functional Development of the Neuromusculature

95

sequences demonstrate that Drosophila NMJ GluRs have highest sequence similarity to mammalian kainate receptors. Insect muscle glutamate receptors, however, are activated by quisqualate (Usherwood, 1994), as are mammalian AMPA receptors. Furthermore, heterologously expressed GluRIIA receptors are relatively insensitive to both AMPA and kainate (Schuster et al., 1991). Thus, the ionotropic glutamate receptors at insect NMJs can safely be considered non-NMDA type, but further distinctions based on mammalian receptor pharmacology are probably meaningless. The only metabotropic glutamate receptor identified so far at the fly NMJ is encoded by the DmGluRA gene, which is expressed both in the CNS neuropil and at the NMJ (Parmentier et al., 1996; Mohrmann and Broadie, unpublished data). Null mutants in the DmGluRA gene are viable, and the receptor appears to play rather subtle functions in the regulation of NMJ structure and functional plasticity (Parmentier et al., 1996; Mohrmann and Broadie, unpublished data). Other than the actual GluRs, only a few other components of the postsynaptic density are known. These proteins are, almost exclusively, present on both sides of the synaptic cleft and include: (1) transmembrane receptors such as members of the position specific (PS) integrin family (3a and 1b subunit; Beumer et al., 1999, 2002) and homophilic neural cell adhesion molecules (NCAMs) Fasciclins (II and III) (Schuster et al., 1996a; Davis et al., 1997; Zito et al., 1997); (2) membrane associated PDZ-domain anchoring proteins such as Discs Large (Dlg, a PSD-95 homolog) and the associated Scribbler protein (Budnik et al., 1996; Roche et al., 2002); and (3) signaling complex molecules such as Ca2/Calmodulin-dependent kinase II (CamKII), Rho-type guanine nucleotide exchange factor, and protease activated kinase (PAK) (Griffith, 1997; Beumer et al., 2002; Roche et al., 2002; Kazama et al., 2003). Based on much longer lists compiled for other species, the majority of the components of the postsynaptic density in insects have yet to be elucidated.
2.3.1.4. Modulation of NMJ Neurotransmission
p0155

The Drosophila NMJ is plastic in both its structural and functional properties. The degree of electrical activity modulates the terminal size and architectural complexity of the NMJ. Hyperactivity through mutation of the K channel genes eag and Sh results in a striking increase in synaptic size and complexity (Budnik et al., 1990). This morphological change is

causally linked to a change in the expression of the homophilic NCAM Fasciclin II (Davis et al., 1996a). In addition, however, many other signaling molecules mutate to alter the degree of structural complexity including CamKII, Integrins, and many other proteins (Zhong and Shanley, 1995; Budnik, 1996; Griffith, 1997; Rohrbough et al., 1999, 2000, 2003). It is likely that only a subset of these proteins are mechanistically linked to activity-dependent plasticity, whereas others simply mediate synaptic growth. To date, only short-term (up to minutes) functional plasticity has been revealed at the Drosophila NMJ. It may be that longer forms of functional plasticity are not present at the insect NMJ, although practical limitations in recording configurations have simply prevented the undertaking of long-term assays. In addition, plastic features are only apparent at low [Ca2], which may not reflect the normal physiological condition (Broadie, 2000). A likely explanation for this restriction is that in higher [Ca2], synaptic depression caused by depletion of resources (e.g., SVs) tends to dominate over any manifestation of facilitation (Broadie et al., 1997; Broadie, 2000). Four forms of synaptic modulation are routinely assayed at the Drosophila NMJ. In order of duration, these are: (1) paired-pulse facilitation (PPF); (2) short-term facilitation (STF); (3) augmentation; and (4) post-tetanic potentiation (PTP). All four forms of plasticity result in elevated EJC sizes following specific patterns of stimulation. PPF is manifest on the millisecond time scale, when one stimulation is followed after a short interval by a second stimulation. Significant PPF begins at a $100 ms interval and increases rapidly at shorter intervals. STF occurs during short, high frequency stimulation trains on the second time scale. Significant Ca2 dependent facilitation usually begins with $2 Hz stimulation trains, and increases rapidly at higher frequencies. Augmentation is a similar form of maintained facilitation that occurs during prolonged, high frequency stimulation trains. Augmentation increases with the frequency of the stimulus train and will persist for the period of stimulus. The longest term modulation, PTP, follows a prolonged, high-frequency stimulus train (510 Hz for >30 s) but persists after a return to a basal frequency (0.5 Hz) of stimulation. PTP represents a prolonged, low-level (2550%) potentiation that lasts for the duration of the recording (>20 min). PTP at the Drosophila NMJ occurs via cAMP dependent signaling (Zhong and Wu, 1991, 1993; Renden and Broadie, 2003).

96 Functional Development of the Neuromusculature

2.3.2. Development of Motor Neuron Function


2.3.2.1. Motor Neuron Morphogenesis

2.3.2.1.1. Early neurogenesis In insects, neurons are produced by progenitor cells called neuroblasts that arise from a specialized neurogenic region called the ventral neuroectoderm (for relevant reviews, see also Chapters 1.101.12). Neuroblasts are stem cells that divide repeatedly to form a specific, limited lineage of progeny. In Drosophila, the neuroectoderm first becomes distinguishable soon

after gastrulation during the very early stages of germ-band extension (see Figure 3). Following gastrulation, a regularly spaced array of neuroblasts develop in the ventral domain of each embryonic hemisegment, interspersed among cells forming the ventral epidermis (epidermoblasts). These two sets of progenitors segregate based on a competitive interaction called lateral inhibition. The genes mediating this inhibition mechanism are called neurogenic genes, since they allow all progenitor cells to acquire a neural fate. The lateral inhibition hypothesis proposes that all cells within the

Figure 3 A timeline of Drosophila NMJ development. The time scale on the left represents hours after egg laying (AEL) at 25  C. Under these conditions, embryogenesis takes approximately 22 h. Embryonic/larval NMJ synaptogenesis occurs during the period delimited by the golden color.

Functional Development of the Neuromusculature

97

neuroectoderm acquire the competence to develop as neuronal progenitors (a competence group), but within each group stochastic interactions single out only a single cell to develop as a specific neuroblast. This neuroblast subsequently inhibits all the surrounding cells within the group, forcing them to assume a secondary epidermal fate. The selected neuroblasts segregate individually into the interior of the embryos (delaminate) prior to initiating rounds of stem cell division. In Drosophila, neuroblasts delaminate in three distinct phases, eventually forming a monolayer of neural stem cells between the ectoderm and the underlying mesoderm (30 neuroblasts/hemisegment). The surprisingly regular neuroblast array has allowed the production of spatial maps with individually defined, named neuroblasts, increasingly accurately defined based on the expression of specific molecular markers (Bossing et al., 1996; Schmid et al., 1999). A well-defined coordinate system of positional information (anteriorposterior, dorsalventral) controls the identity of neuroblasts. In Drosophila, each individual neuroblast can be specifically identified based on the combination of genes it expresses and, later, the characteristic lineage it gives rise to. Soon after leaving the ectodermal layer, neuroblasts become more spherical and begin to divide in a series of asymmetric divisions to generate ganglion mother cells (GMCs), which then divide symmetrically to produce two neurons. Most neuroblasts perform multiple rounds of division, with most mitotic activity ending at a period corresponding to the completion of germ-band retraction, although rare divisions may continue in late embryonic stages. In Drosophila, these same neuroblasts reappear, after a period of quiescence, in early postembryonic stages, to again begin producing neurons that will form the adult nervous system (Prokop and Technau, 1991). 2.3.2.1.2. Establishing motor neuron identity Cell ablation studies indicate that neuroblasts are initially quite labile in terms of fate. Elimination of a neuroblast can be compensated for by the reversal of an epidermal fate decision and the emergence of a new neuroblast. In Drosophila and grasshopper embryos, neuroblast ablation and transplantation experiments, together with genetic studies, have shown that GMC birth order is the primary determinant of identity (Doe and Goodman, 1985; Furst and Mahowald, 1985; Prokop and Technau, 1994; Schmid et al., 1999). Neuroblasts grown in singlecell isolated cultures produce defined neuronal lineages (Huff et al., 1989; Schmidt et al., 2000), suggesting that neuronal identity is largely derived

from intrinsic determinants. Single identified neuroblasts grown in culture give rise to progeny that express molecular markers and functional properties expected for that lineage (Luer and Technau, 1992; Schmidt et al., 2000), suggesting that at least some aspects of neuronal differentiation are cell autonomous. The identity of neurons arising from early neuroblast divisions is controlled, at least largely, by intrinsic fate determinants that are asymmetrically distributed to the daughter cells. A hierarchy of transcription factors, which first function during segmentation, later regulate temporal identity in the nervous system, including hunchback (hb), castor (cas), and Kruppel (Kr) (Gaul et al., 1987; Cui and Doe, 1992; Mellerick et al., 1992; Isshiki et al., 2001). These transcription factors are sequentially expressed in neuroblasts, with GMCs maintaining the expression profile present at birth. This transcription factor expression specifies temporal identity and thus defines neuronal cell fate (Isshiki et al., 2001). Although the neuronal type is dependent on the GMC, the two daughter cells also acquire different fates. Cell ablation studies in the grasshopper clearly show that cell interactions between the siblings determine their respective fates (Kuwada and Goodman, 1985). In a Drosophila hemisegment neuromere, the 30 named neuroblasts generate $200 neurons including a small number (34) of motor neurons, a large number of interneurons (>150), and a few neurosecretory cells (<10). These same neuroblasts generate a large number of glial cells. Comprehensive cell lineage studies have defined the clone of neurons and glia produced by each Drosophila neuroblast (Schmid et al., 1999). These studies have identified the founding neuroblasts for all the cell types of the embryonic CNS: motor neurons, local and projection interneurons, neurosecretory cells, and glia. An important observation is that neuroblasts at similar dorsoventral positions usually generate similar cell lineages, including the same motor neuron subtype. Lineage analysis and connectivity mapping have been completed on the 31 motor neurons identified in the Drosophila embryo (Landgraf et al., 1997). Most (26 of 31) motor neurons derive from just a few neuroblasts in the ventral neuroectoderm. Although neuroblasts give rise to more than one type of motor neuron, structurally similar motor neurons generally derive from a common neuroblast and innervate operationally related muscles. This suggests a developmental mechanism linking the progeny of a specific neuroblast with a specific subset of muscles, which likely mediates formation of nervemuscle partnerships. Despite this general principle, neurons deriving

98 Functional Development of the Neuromusculature

from different neuroblast lineages can also innervate operationally related muscles, so this rule is not invariable. 2.3.2.1.3. Axonogenesis and pathfinding Growth cones lead extending axonal processes toward muscle targets. The small Rho GTPases are vital for regulation of the actin cytoskeleton dynamics that mediate growth cone extension during axonal pathfinding. In Drosophila, Cdc42 is involved in overall neuron morphogenesis, Rac1 is involved in axonal outgrowth and steering, and RhoA is essential for dendritic but not axonal growth (Luo et al., 1994; Lee et al., 2000; Ng et al., 2002). Growth cones use a complex cocktail of secreted cues and cell surface markers on epithelial cells, mesodermal cells, glial cells, and other neurons in their guidance. In the CNS, glial cells provide one of the most important guidepost functions for pioneer neurons. In the periphery, motor axons are guided by a complex combination of attractive and repulsive cues to initially select the appropriate axon tracts, then to select the appropriate point to defasciculate from this tract, and finally to select the correct target muscle within a domain of overlapping muscles. Among the best characterized of the attractive signals are members of the NCAM Immunoglobulin (IgG) superfamily (Tessier-Lavigne and Goodman, 1996), especially the Fasciclins (Bastiani et al., 1987; Patel et al., 1987; Bieber et al., 1989). Among the best characterized repulsive signals, are both secreted and membrane bound forms of the Semaphorin family (Kolodkin et al., 1993; Winberg et al., 1998). Many axons extend selectively along the surface of pioneer axons (Bate, 1976), forming axon bundles or fascicles, a process called selective fasciculation that requires NCAMs (Hortsch and Goodman, 1991; Tessier-Lavigne and Goodman, 1996). In addition, growth cones are guided to targets via diffusible, secreted ligands. Growth cones can also be guided by contact-dependent and diffusible repulsive signals that establish inhibitory domains. Much of the knowledge of motor neuron guidepost cells and attractive growth substrates comes from work on the large grasshopper embryo (Hortsch and Goodman, 1991). Both glial cells and the epidermis have been shown to provide permissive, guiding substrates for pioneer axons (Bastiani et al., 1986; OConnor et al., 1990). In both grasshoppers and Drosophila, the anterior corner cell (aCC) motor neuron pioneers the ISN, which carries the axons innervating the dorsal muscle field (Thomas et al., 1984). Later-growing axons depend on fasciculation to these pioneer axonal tracts, coupled to appropriate turning responses at

p0195

distinctive choice points. These choice points are often associated with glia (Bastiani et al., 1986; Doe et al., 1986). The initial efforts to identify the guiding molecules depended on monoclonal antibody staining screens in both the grasshopper and Drosophila (Hortsch and Goodman, 1991). This approach identified the surface glycoproteins Fasciclin IIV and Neuroglian, which are each expressed on a distinctive, but overlapping set of growth cones, fascicles, and guidepost glial cells (Bastiani et al., 1987; Patel et al., 1987; Bieber et al., 1989; Kolodkin et al., 1992). Three of these proteins are members of the IgG superfamily related to vertebrate NCAM (Tessier-Lavigne and Goodman, 1996). In vitro aggregation studies show that these proteins function as homophilic receptors. The dynamic expression of these proteins on axons and guidepost glia is highly conserved across insect species, with similar patterns on identified motor neurons in both the grasshopper and Drosophila (Thomas et al., 1984). In the Drosophila embryo, motor neurons, and the muscles they target, often express the same transcription factor prior to any contact. This paired specification is believed to drive the expression of matched homophilic receptors that will guide motor neuron pathfinding and target selection. For example, motor neurons targeting the dorsal muscles express the homeobox transcription factor evenskipped (eve), and this expression is both necessary and sufficient to guide these axons via the ISN to dorsal muscles (Landgraf et al., 1999b). Similarly, motor neurons targeting the ventral muscle field express the Lim-1/Isl-1/Mec-3 (LIM) homeobox transcription factor islet, which is sufficient to specify ventral motor axon projection (Thor and Thomas, 1997). Thus, these two genes constitute a bimodal switch regulating motor axon pathfinding to either ventral or dorsal muscle targets. The Eve transcription factor regulates the expression of NCAMs that promote adhesion to the ISN and so direct motor axons into the dorsal muscle field (Landgraf et al., 1999b). Misexpression of eve in normally ventrally projecting motor neurons will reroute them to the dorsal muscle field, bypassing their normal ventral targets. Thus, regulation of selective fasciculation is sufficient to direct motor axons towards the correct muscle group in the periphery. What controls when motor axons leave the nerve fascicle and select specific muscle targets? One mechanism is to decrease the adhesion between motor axons, either by decreasing their NCAM expression or locally introducing repulsion. In Drosophila, the beaten path (beat) gene encodes

p0200

p0210

Functional Development of the Neuromusculature

99

a secreted IgG superfamily protein, expressed in motor axons coincident with the period of defasciculation. In beat mutants, motor axons fail to defasciculate and travel past their intended muscle (Fambrough and Goodman, 1996). Moreover, misexpression of the anti-adhesive Beat protein in the dorsal-projecting eve expressing cells will cause them to prematurely defasciculate into the ventral muscle field (Landgraf et al., 1999b). Thus, it appears that Beat decreases the attractiveness of motor axons for other motor axons and directs the motor axon nerve departure choice point. A complementary mechanism that triggers defasciculation is increased adhesion to an alternative substrate, specifically muscle progenitors and target muscles. In mutants lacking muscle targets, motor axons extend into peripheral nerves, but do not defasciculate at the proper choice points (Prokop et al., 1996). Thus, muscles, and in particular muscle progenitors in the process of muscle formation, act to trigger proper defasciculation of motor axons (Landgraf et al., 1999b). One molecule that increases the adhesiveness of muscle is the Drosophila sidestep (side) gene, which encodes a novel cell surface protein (Sink et al., 2001). The Side protein is expressed on muscles during motor axon defasciculation and side mutant axons ignore their appropriate exit points and continue to extend along the motor nerves (Sink et al., 2001). Ectopic Side expression results in continued contact by motor axon growth cones and inappropriate pathfinding decisions. Thus, Side functions as an attractant, presumably as a ligand for an unknown receptor on growth cones. The coordinate control of appropriate adhesion and repulsion guides Drosophila motor axons along their peripheral routes. 2.3.2.1.4. Dendrite formation in the CNS At the same time that motor axons are pathfinding towards muscle outputs in the periphery, the dendrites of the motor neurons are similarly developing highly individual morphologies within the CNS neuropil. Compared to our knowledge of axon guidance, much less is known about the mechanisms responsible for directed outgrowth of dendrites. Very recent work has begun to examine dendrite formation in the Drosophila embryo and examine the partitioning of the neuropil into functionally distinct regions of dendritic arborization. The architecture of the neuropil has been mapped based on modality-specific sensory projections (Schrader and Merritt, 2000) and the scaffolding revealed by Fasciclin II expressing axonal tracts (Landgraf et al., 2003). Another class of reliable neuropil landmark is the constant topography of neuropeptidergic

projections including Serotonin, Insulin-, FMRF-, Allatostatin-, Corazonin-, Substance-P-, and Leucokinin-1-positive neurons (Nassel et al., 1990; Gorczyca et al., 1993; Nichols et al., 1999; Landgraf et al., 2003; Taghert and Veenstra, 2003). Particularly powerful genetic mosaic strategies have been used to map the arrangement of output synapses on individual neurites in the neuropil (Lohr et al., 2002). These studies indicate three distinctive neuropil compartments: (1) compartments that contain primary transport pathways mostly lacking output synapses; (2) compartments that contain output presynaptic terminals; and (3) compartments that are postsynaptic. In Drosophila, as in other insects, the dorsal neuropil contains the arbors of efferent neurons that show little overlap with sensory projections, indicating that direct connection between sensory and motor neurons rarely, if ever, occur. Individual efferent neurons differentiate dendritic arbors in distinctive patterns (Landgraf et al., 1997). An important defining principle is that motor neurons that innervate muscles of similar position and orientation are often clustered and have overlapping dendritic trees. Guidance strategies in the CNS neuropil have only recently begun to be analyzed, therefore the state of our understanding is primitive compared to motor axon targeting in the periphery. A recent surprising finding is that projections in the neuropil are controlled, at least in part, by guidance cues provided from the CNS midline rather than the target cells with which they will form synapses. In Drosophila, the Slit protein secreted by midline glia acts as a repellent to position axon fascicles in the neuropil at specific distances from the midline. The precise distance depends on differential combinations of the axonal Slit receptors, called Roundabout (Robo) receptors (Simpson et al., 2000). The position of dendritic arbors also depends on expression of Robo receptors responding to the same midline-derived Slit repellent (Zlatic et al., 2003). Robo has likewise been shown to regulate dendritic patterning and synaptic connectivity in the giant fiber system of adult Drosophila (Godenschwege et al., 2002). A recent study on cell-autonomous guidance signals showed that both Frazzled and Robo receptors on dendrites respond to diffusible signals from the midline to select their field of arborization (Furrer et al., 2003). It is probable that a combination of such diffusible morphogen signals and contact-dependent target signals (as yet unknown) together control dendritic pathfinding and target recognition in the CNS. Motor neurons innervating similar muscles are often, but not necessarily, related by lineage or

100 Functional Development of the Neuromusculature

position. Interestingly, however, in instances where no such developmental linkage exists, the motor neurons still share a common domain of dendritic arborization within the neuropil (Landgraf et al., 1997). Thus, a general rule appears to be that muscles related by function (position and orientation) are innervated by motor neurons with a common domain of dendritic arborization, irrespective of whether the neuronal somata share a common precursor/position or not. A feature of these developing dendrites is that they express the same homophilic adhesion molecules as the axons, proteins known to mediate selective adhesion during fasciculation and target recognition. For example, motor neuron dendrites expressing the homophilic cell adhesion molecule connectin project together to the same region lateral to the anterior commissure (Nose et al., 1992, 1997; Landgraf et al., 2003). These parallels support the hypothesis that connectin-mediated adhesion may be another mechanism governing dendritic branching in the neuropil. 2.3.2.1.5. Target recognition and selection Identified motor neuron projections to specific target muscles were mapped first in the grasshopper, and later in Drosophila (Ball et al., 1985; Sink and Whitington, 1991). In the grasshopper, motor neuron growth cones come into contact with muscle pioneers (see below) and show highly selective affinities for their specific targets. In Drosophila, similar muscle progenitors are the source of cue(s) that trigger motor axon defasciculation, targeting axons to restricted muscle fields (Landgraf et al., 1999a). Subsequent to defasciculation, motor growth cones explore the forming muscle cells within the target domain, initially quite widely. But increasingly, filopodia are restricted to only a specific target myotube. When such a specific muscle target is ablated in the grasshopper, the motor axon growth cone that would have innervated it, instead continues its exploration to a more distal region (Ball et al., 1985). In Drosophila, a combination of surgical and genetic manipulations have allowed specific muscle targets to be either duplicated or deleted in order to test targeting specificity. When ventral longitudinal muscles 6/7 are removed, for example, their motor neuron, RP3, grows past the normal target area and eventually innervates incorrect muscles in a variable pattern (Sink and Whitington, 1991). The motor growth cone terminates and differentiates into a synapse in the absence of its normal target, but only after a period of delay and continued searching. Similar observations are seen when other specific muscles are ablated (Cash et al., 1992; Chiba et al., 1993). Perhaps more interestingly, if a muscle target

is duplicated, both muscles are innervated equally by the correct motor neuron. These observations show an extreme high level of specificity in motor axon target recognition (see also Chapters 2.1 and 2.2). Motor neuron axons are guided to their proper muscles by a hierarchy of cues, starting with a neuronal transcription factor code that underlies their delivery to particular regions of the muscle field, and continuing with muscle target-specific transcription factor codes that drive the expression of defining cell surface receptors (Thor et al., 1999; Odden et al., 2002; see also Chapters 2.1 and 2.2). Different muscles express unique combinations of cell surface glycoproteins that mark them and, in at least some cases, clearly delineate the future site of motor neuron innervation in a subdomain of the myotube surface. For example, the homophilic NCAM Fasciclin III is expressed both on the RP3 growth cone and muscles 6/7, transiently at the time of target recognition (Halpern et al., 1991), but is downregulated as synaptogenesis proceeds. In mutants where innervation is blocked or delayed, the myotubes still clearly express Fasciclin III in a restricted domain corresponding to the normal site of the NMJ (Broadie and Bate, 1993b). Likewise, the leucine-rich repeat (LRR) CAMs Connectin and Toll are also expressed on a small subset of muscles and the growth cones that innervate these muscles. Following target recognition, Toll is localized to synaptic sites on the muscle and downregulated (Nose et al., 1992; Broadie and Bate, 1993b). These examples suggest that each muscle simply expresses a specific code of surface CAMs that allow selective adhesion with growth cones carrying a complementary set of CAMs. However, the mechanism is not simple. For example, genetic ablation of Fasciclin III, which prefigures the muscles synaptic site prior to growth cone contact and is expressed by the growth cone during contact, does not prevent NMJ formation (Chiba et al., 1995). Rather, target recognition appears, at most, only slightly impaired, consistent with the vague idea that multiple redundant recognition cues direct the first stages of synaptogenesis. Consistently, ectopic expression of these recognition factors uniformly causes more severe perturbation of neuromuscular connectivity (Chiba et al., 1995). These embryonic connectivity studies have, in general, only detected a single axon innervating any muscle. However, most muscles are known to have multiple innervations in the mature larva (Keshishian et al., 1996). Much of this later stage innervation is derived from the widespread terminals of ventral unpaired median (VUM) neurons, the likely source of octopaminergic type II NMJs of

Functional Development of the Neuromusculature

101

larval muscles (Monastirioti et al., 1995; Taghert and Veenstra, 2003). Nothing is yet known about the mechanism driving this later stage of neuromuscular target recognition and synaptogenesis.
2.3.2.2. Development of Motor Neuron Electrical Properties

p9035

p0245

Most of what we know about the electrical properties of developing insect neurons comes from studies in acutely dissociated cell cultures. In Drosophila, neurons developing in vitro from cells harvested at gastrulation, as well as cleavage arrested neuroblasts that give rise to giant neurons in culture, have been extensively used for defining the development of excitability and the function of genes encoding ion channels and their regulators (Solc et al., 1987; ODowd and Aldrich, 1988; Saito and Wu, 1991; ODowd, 1995; Tsunoda and Salkoff, 1995a). However, after nearly two decades of Drosophila culture studies, spontaneous synaptic transmission between neurons in vitro has only recently been recorded electrophysiologically (Lee and ODowd, 1999, 2000), and evoked neuronal transmission has never been recorded. This progress in fruit flies contrasts sharply with early strides made using other species. For example, limited cell migration and axon outgrowth were observed in cultured mosquito larva ganglia in 1937 (Pfeiffer, 1937) and, since then, researchers using the cockroach and other species have had great success propagating neuromuscular tissues in culture, both as explants and individual cells. The morphology, ultrastructure, and electrophysiology of these preparations (including that of NMJs formed in vitro) are relatively well described (Beadle and Hicks, 1985; Thomas et al., 1987; Levine and Weeks, 1996). In contrast to this long history of in vitro studies, in vivo motor neuron electrophysiology in Drosophila has been limited to just a few recent studies, all concentrated on an accessible cluster of five motor neurons (aCC and RP14), and one interneuron (pCC), located in a superficial, dorsal position within the embryonic VNC (Rohrbough et al., 2003). In the embryo, these identified neurons are accessible to patch-clamp recording from the cell body following removal of the overlying neurolemma with protease (Baines and Bate, 1998; Baines et al., 1999). In developing motor neurons of the Drosophila VNC, the first voltage-gated currents appear during mid-embryogenesis (stage 16, 1314 h at 25  C; $65% embryonic development; Baines and Bate, 1998). During later embryogenesis (mid-stage 16 onwards; 70% development), motor neurons display a sequential appearance of voltage-gated ionic

currents, including several outward K currents, an inward Ca2 current, and an inward Na current (Baines and Bate, 1998; Baines et al., 2001). Northern analyses show that shaw and shab transcripts encoding the IK K channels are expressed simultaneously in developing motor neurons from stage 13 (9 h at 25  C; $45% development; Tsunoda and Salkoff, 1995a). Consistently, the first detectable current to appear in cell body recordings is the delayed outward K current (IK). Subsequently, the calcium current (ICa) is detectable at 15 h ($70% development), the sodium current (INa) at 16 h ($75% development), and the rapidly inactivating potassium current (IA) at 17 h ($80% development). The developmental sequence of current appearance is somewhat distinct from that seen in both Drosophila embryonic and pupal muscle (see below; also, Salkoff, 1985; Broadie and Bate, 1993d). Action potentials are first recorded at 17 h ($80% development), closely coincident with the appearance of IA. By 85% development, Drosophila motor neurons manifest the ability to fire repetitive action potentials (Baines et al., 2001). Drosophila motor neurons exhibit multiple ligand gated receptors, although only a small subset of these have been documented in situ. The first ligand gated current to emerge is an ACh-gated inward cation current (IACh) during mid-embryogenesis (stage 16, 13 h at 25  C; 62% development), at the same time that the first voltage-gated K current is detectable (Baines and Bate, 1998). Picrotoxin-sensitive GABA receptors are present in cultured embryonic motor neurons (Lee and ODowd, 1999; Lee et al., 2003), and are known to be present in in situ motor neurons during late embryogenesis (Rohrbough and Broadie, unpublished data), but the appearance and development of the GABA current has not been reported. Interestingly, the dorsal unpaired median (DUM) neurons of the grasshopper embryos respond to ACh (and GABA) very considerably earlier in embryogenesis (40% development), coincident with the initiation of axon outgrowth (Goodman and Spitzer, 1979). Thus, different species differ considerably in the timing of their neuronal electrical development, both in absolute terms and relative to the stage of neuronal morphological maturation. Inhibitory GluRs are also present in Drosophila motor neurons, and a Drosophila homolog of this receptor family, expressed in oocytes, mediates a glutamate-gated chloride current (Cully et al., 1996). The inhibitory glutamate-gated current is present in mature embryonic motor neurons during late embryogenesis (Rohrbough and Broadie, unpublished data), but its development has not been charted.

p0250

p9040

102 Functional Development of the Neuromusculature

p9045

Endogenous activity in Drosophila motor neurons is first detected at 15 h ($71% development) in the form of very small (peak amplitude, 5 pA) spontaneous events that probably result from the opening of a small number of ion channels. Over the next hour, larger events (peak amplitude, 25 pA) first appear that represent synaptic currents, the first indication of synaptic transmission. The timing of the onset of synaptic function in Drosophila (7075% development) is comparable to the appearance of transmission in DUM motor neurons in grasshoppers (75% development) but delayed relative to synaptic development between cercal afferents and central giant interneurons in cockroaches (55% development; Goodman and Spitzer, 1979; Blagburn et al., 1996). The peripheral neuromuscular junction in Drosophila first becomes functional just slightly early (6570% development; Broadie and Bate, 1993e), corresponding to the time that motor neurons first display their full complement of ion channels.
2.3.2.3. Synaptogenesis in the CNS

p0260

Despite its molecular and genetic advantages, the small size of Drosophila has greatly hampered studies in the CNS of this species. Progress in understanding fruit fly CNS function is all very recent, and will be reviewed here. However, there is an extensive body of literature describing synaptic transmission between other insect species neurons in vivo (Levine, 1984; Callec, 1985; Weeks et al., 1997). Most of this work focuses on mature synapses; relatively little is known about the development of CNS function. In Drosophila cultured embryonic neurons, pharmacological analyses reveal that the majority of fast excitatory synaptic currents are cholinergic and mediated by nicotinic AChRs (Lee and ODowd, 1999, 2000). Less common inhibitory synaptic currents, mediated by picrotoxin sensitive GABA gated chloride receptors, are observed in a subset of cultured neurons. In Drosophila motor neurons in situ, the appearance of electrical currents and ligand gated neurotransmitter channels is followed closely by the onset of endogenous synaptic activity (Baines and Bate, 1998). The inputs onto these motor neurons are predominantly excitatory, cholinergic interneurons (Baines et al., 1999, 2001; Rohrbough and Broadie, 2002). In situ, both Drosophila motor neurons (aCC and RP2) and interneurons (pCC) exhibit infrequent fast, spontaneous events after 16 h ($75% of development; Baines and Bate, 1998), and then subsequently sustain inward currents supporting APs after 1920 h (Baines et al., 1999, 2001). Both of these types of endogenous activity require external Ca2, and are reduced/eliminated by

mutations that block AP firing, deplete SV supplies, or specifically eliminate cholinergic transmission, thus indicating that they represent cholinergic synaptic input (Baines and Bate, 1998; Baines et al., 1999, 2001). By 1819 h at 25  C (8590% of development), synaptic input to motor neurons is present in the form of both fast AP mediated synaptic currents and periodic, sustained (up to 1 s duration) episodes of excitatory input occurring at least several times per minute (Baines et al., 1999, 2001). This sustained, rhythmic excitatory activity requires cholinergic synaptic input, since it is eliminated by conditionally blocking ACh synthesis in cholinergic interneurons (Baines et al., 2001). Sustained bouts of cholinergic input activate large excitatory potentials accompanied by bursts of AP firing in the motor neuron (Rohrbough and Broadie, 2002; Rohrbough et al., 2003). The output of this centrally generated motor activity is manifest at the NMJ as periodic bursts of high-frequency glutamatergic transmission to drive the peristaltic muscle contraction underlying coordinated locomotion (Broadie and Bate, 1993a; Baines et al., 2001; Cattaert and Birman, 2001). In Drosophila, motor neurons develop in the absence of CNS synaptic transmission without any apparent change in synaptic connectivity (Baines et al., 2001). Although it is difficult to be certain that central synapses do not form in inappropriate areas under blockade conditions, it seems probable that there is no requirement for synaptic activity in establishing central synapses. In the absence of synaptic transmission, however, motor neurons do develop slightly abnormal basal electrical properties, including greater voltage activated Na and K currents, which result in altered neuronal excitability (Baines et al., 2001). The increased currents appear to be due, at least in part, to upregulated expression of para (INa) and slowpoke (IK(Ca))(Rohrbough et al., 2003). When neurotransmission is blocked, the number of action potentials fired by a depolarizing voltage ramp (60 to 60 mV) is increased, and the amplitude of the initial action potential is significantly larger. The modulation of neuronal electrical properties is not permanent, but rather can be reversed by restoration of synaptic function. This finding suggests that synaptic transmission has at least some role in regulating the development of electrical properties. Somewhat surprisingly, eliminating synaptic transmission selectively in motor neurons, rather than globally as above, does alter synapse formation onto these cells (Baines et al., 1999). Ultrastructural analyses reveal fewer presynaptic input sites and a lower frequency of endogenous synaptic events onto synaptically silenced Drosophila motor neurons (Baines et al., 1999).

Functional Development of the Neuromusculature

103

These results suggest that synaptic output by the postsynaptic motor neuron regulates synaptogenesis by interneurons onto these cells. Why might panneuronal transmission blockade cause different effects than selective blockade only in a few motor neurons? Perhaps CNS competition between neurons is a decisive factor in regulating circuit formation in the Drosophila embryo. A similar mechanism in mammals is central to synapse assembly in systems ranging from retinotectal projections to motor terminals. In addition to its role in axon fasciculation (above), the homophilic NCAM Fasciclin II operates to stabilize synaptic connections and mediate activity dependent plasticity in the postembryonic NMJ (Schuster et al., 1996a, 1996b; Davis et al., 1997; Davis and Goodman, 1998). At the NMJ, Fasciclin II is expressed in pre- and postsynaptic cells from early stages of synaptogenesis, although it is not required for synapse formation (Schuster et al., 1996a, 1996b). In contrast, overexpression of Fasciclin II in embryonic muscle can initiate ectopic synapse formation, suggesting that postsynaptic Fasciclin II may be limiting during synaptogenesis (Davis et al., 1997; Davis and Goodman, 1998). Based on these findings, the role of Fasciclin II in CNS synaptogenesis was examined, and it was shown that Fasciclin II acts as an important determinant of central synapse regulation (Baines et al., 2002). As observed at the NMJ, synaptogenesis by identified neurons in the embryonic CNS does not require Fasciclin II, but the postembryonic elaboration of these synaptic connections is impaired in the absence of Fasciclin II. In addition, just like the NMJ, central synapses are highly sensitive to alterations in the balance of Fasciclin II between pre- and postsynaptic cells. Increasing Fasciclin II in either cell during embryogenesis disrupts synaptic connectivity, causing defects comparable to the targeted blockade of synaptic transmission (Baines et al., 2002). In contrast to the findings at the NMJ, however, and to work on the Fasciclin II homolog in the mammalian hippocampus, elevated Fasciclin II reduces the number of functional synapses in the Drosophila CNS. Thus, in flies, Fasciclin II is not required for synaptogenesis but can regulate the elaboration of synaptic contacts, especially during postembryonic development, in a synapse-type specific manner.

2.3.3. Development of Glial Function


2.3.3.1. Central Glia and CNS Scaffolding
p0275

Compared to neurons, relatively little has been revealed so far about the functional development of glial cells in insects, although their importance

for insect development has long been recognized (Anderson et al., 1980; Oland and Tolbert, 2003). Many classes of glial cells (glia) have been identified in insects. Two major classes include: (1) glia that surround or enclose neuronal cell bodies and (2) glia that first prefigure and later wrap around axon tracts (Jacobs and Goodman, 1989). In the CNS, a prominent set of glia exist at the midline. One subset of midline glia act as the source of signals that attract (Netrin) and repulse (Slit) motor axons from crossing the midline, and thus determine ipsilateral versus contralateral motor axon projections. The Slit signal working on related Roundabout receptors (Robo I, II, and III) also determines the lateral distance traveled prior to axonal turning into discrete fascicles. As discussed above, Slit signaling also plays a role in establishing dendritic domains relative to the CNS midline. A separate subset of midline glia mediate contact-dependent guidance and, later, insulation functions. In Drosophila, six midline glia per hemisegment are characterized by unique gene expression patterns, and mutation of these genes reveal critical axonal guidance functions (Klambt et al., 1991; Hidalgo, 2003). These six glia first prefigure the future path of the axons, just before axons are extended. These cells continue to divide and, in later stages, enwrap the longitudinal axonal connectives in the CNS. Many other glia are not associated with axons but rather reside in the cell body layers in direct contact with the neuronal soma. The role of these glial cells is not well established. The mesodermally derived belt glia extend circumferentially around the CNS, generating a tough enclosing sheath that acts as a bloodbrain barrier to chemically isolate the CNS from the rest of the body. Recent work in Drosophila has begun to systematically reveal genes specific to glial cell functional development. A genomic screen utilizing glial cells missing (gcm, a gene encoding a protein that determines whether cells assume glial or neuronal fates, Jones et al., 1995; see also below) mutants, has identified >80 new Drosophila genes with expression restricted to glia (Freeman et al., 2003). These genes show expression patterns varying from pan-glial, through peripheral/CNS specific, to select subpopulations of glia. This work shows that a midline-derived netrin signal orients glia cell migration via a repulsive netrin receptor (dUnc-5), and draper, one of the newly identified genes, mediates glial wrapping (Freeman et al., 2003; see also below). Importantly, many (>80%) of the identified genes have close homologs expressed in corresponding sets of mammalian glia, indicating conserved molecular bases of glial function.

p9050

104 Functional Development of the Neuromusculature

2.3.3.2. Peripheral Glia and Motor Neuron Pathfinding

Peripheral glia arise from the neural crest in vertebrates and the lateral edge of the CNS in insects. Peripheral glia in Drosophila are molecularly distinct from CNS glia. Both express unique proteins, such as gliotactin in the peripheral glia (Auld et al., 1995), and wrapper in the central glia (Noordermeer et al., 1998). In Drosophila, each hemisegment neuromere has two peripheral nerve roots, by which motor axons exit and sensory axons enter the CNS. Several classes of glial cells are associated with these peripheral nerves at different points. First, there are a small number of glial cells associated with the segmental (middle of neuromere) and intersegmental (neuromere boundary) nerve roots, where they diverge from the main CNS axonal tracts. These cells are present prior to axon extension, and ablation of this glial class in the grasshopper prevents formation of the peripheral nerve (Bastiani et al., 1986). Second, the so-called exit glia are located just outside the CNS, where the segmental and intersegmental neurons diverge into peripheral axonal pathways. In addition, there are several glial cells that associate with peripheral axonal pathways, and enwrap the axons once the paths have been established. Most peripheral glia are born in the CNS near the lateral exit points for the peripheral nerves. Lineage analyses in Drosophila have shown that approximately 810 (some variation) peripheral glia per hemisegment arise specifically from neuroblasts 13 and 25 (Schmid et al., 1999; Sepp et al., 2001). At least one further glial cell is born in the periphery, associated with the ISN, and remains in its birth place through later development (Hidalgo, 2003). During mid-embryogenesis, the centrally born glia migrate over great distances to their final peripheral locations. These cells migrate as a continuous chain of glia, with the pioneering glial cells actively exploring the environment using Actinbased filopodia (Sepp and Auld, 2003). The small GTPases RhoA and Rac1 mediate the Actin cytoskeletal rearrangements required for this migration (Sepp and Auld, 2003). In the CNS, the immature glia seem to act as intermediate guidepost cells for pioneer motor axons migrating into the periphery. Peripheral glia prefigure axon paths through the CNS/periphery transition zone through which axons migrate into and out of the CNS. When peripheral glia are ablated early, via targeted expression of the cell death genes grim and ced-3, motor axon trajectories are initially perturbed, although later largely corrected

(Sepp et al., 2001). This suggests that peripheral glia do act as early guideposts for motor axons, but that other cues in the periphery are sufficient to guide these axons to their correct target muscles. Sensory axons entering the CNS are also disrupted, but these pathfinding errors are not corrected (Sepp et al., 2001). Following this early guidance, motor axon growth cones extend past the immature glia and begin to pathfind into the developing muscle field. At this point, these motor axons appear to now serve as the guidance substrate for the migration of glial cells into the periphery (Sepp et al., 2000). For example, in Drosophila, the aCC neuron pioneers the ISN during stage 12 ($40% development). Once it has passed the 810 glia at the CNS exit point, the glia begin to follow the aCC axon into the periphery (Sepp et al., 2000). Importantly, the glial growth cones never appear to extend past the aCC pioneer growth cone, suggesting that the aCC is acting as the substrate for glial cell migration. During stages 1315 (4760% development), the peripheral glia first follow the outgrowing motor axons through the ventral field, and then the mixed motor/sensory axons, through the dorsal field (Sepp et al., 2000). In Drosophila, genetic mutations that impair glial differentiation result in profound defects in peripheral nerves. For example, mutations in the homeobox transcription factor gcm gene disrupt the specification of most glia and cause severe axonal defasciculation and pathfinding errors (Jones et al., 1995). However, these mutations affect all glial cells, not just the peripheral glia selectively, so the cause of these defects is uncertain. Ablation or genetic disruption of peripheral glia results in defasciculated axon tracts (Sepp et al., 2001), but it has yet to be genetically proven that peripheral glia mediate axonal pathfinding. Indeed, the developmental profile given above suggests the opposite: in peripheral axons, first motor axons ventrally and then sensory axons dorsally provide guidance for the peripheral glia to reach their mature destinations.
2.3.3.3. Glial Wrapping of Peripheral Axons

During migration from the CNS to their final peripheral destinations, glial cells already begin to extend processes that enwrap surrounding motor and sensory axons. Although considerable glial wrapping occurs during the last several hours of Drosophila embryogenesis, glia do not entirely wrap motor axons in the embryo, and the terminal ends of motor axons remain unsheathed at hatching (Auld et al., 1995; Sepp et al., 2000). This suggests that peripheral glia are not directly involved in the generation of neuromuscular contacts or the establishment

Functional Development of the Neuromusculature

105

of functional NMJs. During postembryonic development, peripheral glia continue to extend cytoplasmic process, until axonal wrapping is complete and glial processes extend completely to the NMJ (Sepp et al., 2000). The glial ensheathment surrounds all the axons on the nerve fascicle in a single bundle, insulating a common axonal compartment from the surrounding hemolymph. This process occurs gradually over the first and second instar stages, and yet must be remarkably rapid to keep up with the impressive rate of larval growth. All wrapping in larvae appears to be due to extension of existing glial processes, since further glial division does not appear to occur in postembryonic stages (Sepp et al., 2000). By the third instar, glial processes extend completely to the NMJ. In both type I and type II NMJs, glial processes are usually observed extending to the first bouton on the synapse, although never beyond this point (Sepp et al., 2000). Glial cells regulate neural activity by insulating neurons and controlling the axonal microenvironment. Peripheral glia help maintain the appropriate ionic environment to facilitate AP propagation in motor axons. In insects, for example, peripheral glia shield axons from the high K concentration in the hemolymph, which would be predicted to largely abolish neuronal excitability. Little is known about the molecular mechanisms underlying glial wrapping of axons. However, a Drosophila mutant named fray has revealed a novel class of serine/threonine kinase in peripheral glia required for axon ensheathment (Leiserson et al., 2000). The fray mutants do not ensheath peripheral axons properly, leading to axonal defasciculation and early larval death (Leiserson et al., 2000). In normal embryos, the tight glial barrier is maintained by septate junctions between glial cells, which serve the function of vertebrate tight junctions to maintain a selective-diffusion barrier. Tight junctions are absent in insects. Drosophila has two types of septate junction (SJ), smooth (sSJs) and pleated (pSJs) (Campos-Ortega and Hartenstein, 1997). There are five characterized SJ associated proteins in Drosophila, all with clear roles in SJ formation: Gliotactin (Gli), Neurexin IV (Nrx), Discs large (Dlg), Scribble (Scrib), and Coracle (Cora). Mutation of any of these genes disrupts SJ formation (Woods and Bryant, 1993; Auld et al., 1995; Baumgartner et al., 1996; Bilder et al., 2003). Gli is a cholinesterase-like molecule (noncatalytically active) that is a member of a class of adhesion proteins termed Electrotactins (Auld et al., 1995; Schulte et al., 2003). Nrx is a transmembrane protein, primarily characterized as a synapse associated protein (Baumgartner et al., 1996). Neurexins were

first identified as synaptic receptors for Latrotoxin, the component of black widow spider venom that triggers massive exocytosis, and are thought to be involved in presynaptic differentiation in vertebrates (Scheiffele et al., 2000). Dlg and Scrib are membrane associated PSD-95/Dlg/ZO-1 (PDZ) domain containing proteins that act as multiprotein complex scaffolds (Woods and Bryant, 1993; Bilder, 2001; Bilder et al., 2003). Cora is a Band 4.1 homolog with a four point One/Ezrin/Vadixin/Moesin proteinprotein interaction domain that associates Band 4.1 protein attaches the cell cytoskeleton to transmembrane signaling proteins. Gliotactin is expressed throughout the processes of peripheral glia in late embryos and larvae and is required for the formation of the peripheral blood nerve barrier (Auld et al., 1995). In gli mutants, peripheral glia develop normally at the light microscope level, but ultrastructurally they contain small gaps in the gliaglia cell interfaces, compromising the peripheral bloodnerve barrier. This can be shown experimentally by exposing a nerve to an electron-dense dye, which is completely excluded from the axonal compartment in normal embryos but penetrates the impaired glial sheath of the mutants (Auld et al., 1995). In gli mutants, the high [K] hemolymph also contacts motor axons, action potentials fail to propagate, and the embryos are paralyzed. Low [K] saline rescues both action potential propagation and movement (Auld et al., 1995). Furthermore, in gli mutants, Nrx, Dlg, and Cora expression is strongly perturbed. In contrast to the other SJ mutants, SJ are present in gli mutants, but are morphologically immature (Schulte et al., 2003). Neurexin IV (Nrx) localizes to septate junctions in peripheral and subperineural glial junctions (Baumgartner et al., 1996). As in gli mutants, nrx mutants maintain peripheral glial cells that look normal at the light microscope level. However, ultrastructural examination reveals that nrx mutants lack the glialglial SJs that are required to maintain the tight bloodnervebrain barrier. As a result, nrx mutants, like gli mutants, lack action potentials and are paralyzed (Baumgartner et al., 1996). Loss of Nrx causes the mislocalization of Cora, the Drosophila protein 4.1 homolog, and eliminates ultrastructurally defined SJs. Another member of the Neurexin protein superfamily is encoded by the Drosophila axotactin (axo) gene (Yuan and Ganetzky, 1999). Unlike the transmembrane proteins described above, Axo is secreted by glia and subsequently localized to axonal tracts. Mutations in axo are temperature sensitive; at 37  C, action potentials are blocked (Yuan and Ganetzky, 1999). However, the bloodnervebrain barrier is

106 Functional Development of the Neuromusculature

intact in axo mutants. This suggests that Axo is a component of a glialneuronal signaling pathway that acutely regulates the electrical properties of target axons. To summarize, these results demonstrate that insect glial cells are vital for the acute maintenance of neuronal signaling. Defined glial functions include: (1) glial wrapping to insulate motor axons from the ionic composition of the hemolymph and (2) glial-neuronal communication required to develop or maintain neuronal signaling properties.

2.3.4. Development of Muscle Function


2.3.4.1. Muscle Morphogenesis and Patterning

2.3.4.1.1. Early myogenesis Fate-mapping techniques place the primordial cells of mesodermal structures, prominently including the somatic muscles, in the midventral regions of the blastoderm embryo. These cells will become mesoderm, but they do not appear to possess any finer fate restriction at this stage. Thus, zygotic gene expression in the blastoderm defines the mesodermal anlage, but later information is required to specify precise developmental pathways, such as those leasing to development of fat body, heart, blood, somatic, and visceral muscle (see also Chapters 2.1, 2.2, 2.6, and 2.9). During gastrulation, presumptive mesodermal cells move into the interior of the embryo along the invaginating ventral furrow. This embryonically defined germ layer corresponds with a distinctive domain of gene expression (e.g., twist and snail). Twist (basic HLH transcription factor) and Snail (zinc finger transcription factor) are required for the formation of the mesoderm as mutants in both genes fail to form any mesodermal derivatives. Twist activates many mesoderm specific genes, whereas Snail suppresses the expression of nonmesodermal genes (Leptin, 1991; Casal and Leptin, 1996). Different domains of Twist expression (high and low) are apparent in the mesoderm following gastrulation, and the somatic (body wall) muscles are derived from only the high [Twist] domains. Twist drives the transcription of the Drosophila homolog of myocyte enhancer factor (dMef2), an essential regulator of the myogenic pathway from the mesoderm (Baylies and Bate, 1996; Artero et al., 1998). In extended germ band embryos, such as Drosophila, the patterning of the mesoderm takes place beginning at the period between gastrulation and germ band retraction. This period is characterized by migration, proliferation (cell division), and restructuring of the mesodermal progenitor cells. Invaginating cells appear to be committed to mesodermal fates at the time of gastrulation, but their

specific fates within the mesoderm are still labile at this time. Numerous waves of mitosis occur in the mesoderm (at least four rounds of division in Drosophila), with the final divisions continuing during late germ band retraction and soon after shortening is completed. The exact number of cell divisions seems to vary widely between different progenitor cells. Mesodermal cells begin to acquire specific differentiated fates just prior to or at the time of germ band retraction. Cell transplantation studies indicate that the fate of a mesodermal cell is dictated by its position within the invaginated mesoderm and by its final position relative to the embryonic ectoderm. Following migration, mesodermal cells form a thin sheet covering the inner surface of the developing ectoderm. Classic experiments performed in the 1940s to 1950s on beetle and lacewing embryos indicated that mesodermal fates are dependent on the adjacent ectoderm. These studies suggested that early mesodermal cells are equivalent until fate restrictions are dictated by the ectodermal cells they come in contact with. Genetic studies in Drosophila have revealed two important ectodermal signals: Decapentaplegic (Dpp), a transforming growth factor-b protein, and Wingless (Wg), a Wnt signaling protein. Dpp acts as an inductive signal to divide myoblasts into dorsal and ventral domains, whereas Wg appears to regulate Twist levels and is required for the specification of the majority of muscles (Staehling-Hampton et al., 1994; Baylies et al., 1995; Riechmann et al., 1997). The formation of somatic muscles is first detectable during germ band retraction. The somatic mesoderm gives rise to myoblasts that have the capacity to fuse with appropriate neighboring cells. Nascent syncytia, formed when myoblasts begin to fuse, first appear ventrally, immediately adjacent to the developing CNS, at the onset of germ band retraction (Bate, 1990). As germ band retraction continues, myoblasts fuse with syncytia in lateral and dorsal locations. Each developing syncytial myotube continues to enlarge via fusion with neighboring myoblasts, so that the syncytia rapidly become more prominent while the pool of unfused myoblasts dwindles. Myoblast fusion continues for several hours, with the size of individual myotubes dictated by the number of fusion events (Bate, 1990). For any given muscle, the number of nuclei is tightly controlled but is not invariant. While fusion continues, and following its completion, the enlarging myotube extends growth cone-like processes that reach over the surface of the epidermis and probe for attachment sites (Bate, 1990). By $60% embryonic development in Drosophila, muscle assembly

p0350

Functional Development of the Neuromusculature

107

appears complete: fusion events stop and muscles are attached to the epidermis to form the specific muscle field pattern. The end of muscle assembly coincides with the arrival of motor neuron growth cones exploring the surface of the myotubes in preparation of forming NMJs. Innervation plays no part in establishing or maintaining the muscle pattern (Broadie and Bate, 1993c). Myoblast fusion is a multiple-step process involving initial cell recognition/adhesion, formation of a prefusion complex, plaque formation, and, finally, plasma membrane fusion driving to myotube formation. This cellular process appears well conserved between Drosophila and vertebrates. Genetic analyses in Drosophila have identified a growing number of genes specifically required for myoblast fusion. The rolling stone (rost) gene, for example, encodes an adhesive transmembrane protein involved in the early stages of myoblast recognition/adhesion and development of the prefusion complex (Frasch and Leptin, 2000). The myoblast city (mbc) gene encodes a homolog of human DOCK180, and functions as an upstream regulator of the Rho family GTPase Rac1 (Rushton et al., 1995; Frasch and Leptin, 2000). The blown fuse (blow) gene encodes a cytoplasmic protein required for progression between intercellular recognition/ adhesion and formation of the prefusion complex (Doberstein et al., 1997). The titin gene encodes an enormous elastic protein known to regulate cytoskeletal scaffolds. During myoblast fusion, Titin organizes cytoskeletal elements required for formation of membrane plaques between prefused cells, and later controls organization of the contractile cytoskeleton (Zhang et al., 2000). 2.3.4.1.2. Establishing muscle identity Drosophila syncytial muscles form a precise pattern. Each muscle has a unique identity defined by its size (number of nuclei), shape (arrangement of epidermal attachment sites), and pattern of innervation (Bate et al., 1999). This identity is presumably determined by transcription factor cascades that trigger expression of slightly different sets of genes, such that each muscle cell displays a specific number of myoblast fusion events and a set of cell surface proteins. Depending on the cell surface proteins they express, myoblasts recognize the appropriate attachment sites on the epidermis and the appropriate innervating neurons extending from the CNS. Consistent with this hypothesis, a small number of spatially restricted transcription factors have been identified (Michelson et al., 1990; Riechmann et al., 1997; Carmena et al., 1998) that define small subsets of myotubes from the onset of myogenesis. In several

cases, the expression of these transcription factors is restricted to a single myoblast, but expression extends to subsequent nuclei following fusion. These observations suggest that unique cells may be specified based on position, provided an identity based on expression of one or more distinguishing transcription factors, and endowed with the power to recruit neighboring myoblasts through fusion and subsequent influence of gene expression (see also Chapters 2.1 and 2.2). The founder cell hypothesis for establishing muscle identity was proposed by Michael Bate in 1990 (Bate, 1990). In this model, the essential step in crystallizing the muscle pattern is a process that selects a unique founder cell for each muscle in every segment. In Drosophila, 30 founders are specified for the 30 muscles in an abdominal hemisegment. The founder cells have the ability to initiate fusion with the surrounding cloud of myoblasts and are, thus, termed fusion-competent cells. These cells are proposed to be naive or equivalent myoblasts that are subsequently entrained to a particular muscle identity only following fusion with a founder cell myoblast. Muscle founder cells express combinations of transcriptional regulators that confer a specific identity, such as Kruppel (Kr) and S59/ Slouch (Slou) (Ruiz-Gomez et al., 1997; Baylies and Michelson, 2001). The dumbfounded (duf/kire) gene is expressed only in founder cells, and its product acts as an attractant or adhesive required to initiate fusion with fusion-competent cells (Ruiz-Gomez et al., 2000). Ectopic expression of dumbfounded causes inappropriate cellular fusion, indicating that this protein alone confers the ability in founder cells to initiate fusion. The expression of genes such as sticks and stones (sns) and myoblasts incompetent (minc) is restricted to fusion-competent cells, suggesting they are a selected population of cells determined by distinctive regulatory mechanisms (Ruiz-Gomez et al., 2002). In the absence of these genes, myoblasts fail to fuse and remain as an arrested population of mesodermal cells (see also Chapters 2.1 and 2.2). Although derived primarily from observations in Drosophila, this founder cell hypothesis was influenced by earlier work in grasshopper embryos. Cell ablation studies in the grasshopper have shown that muscle progenitors, called muscle pioneers, are uniquely identifiable cells containing fate restriction to become specific muscles (Ho et al., 1983; Ball et al., 1985). In the grasshopper, muscle pioneers are distinctive large cells that extend growth cones to define the mature muscle placement, and subsequently fuse with neighboring myoblasts to form the syncytial myotubes. Ablation studies have shown

108 Functional Development of the Neuromusculature

that removal of a pioneer results in the loss of the muscle, despite the fact that the bulk of the fusioncompetent myoblasts remain in place (Ball and Goodman, 1985a, 1985b). In Drosophila myoblast city mutants that block fusion, founder cells act similarly to grasshopper pioneer cells in spanning the area of the future muscle and otherwise acquiring the mature muscle properties (Rushton et al., 1995; Prokop et al., 1996). Thus, the founder cell hypothesis proposes that unique cells, founder cells in Drosophila and muscle pioneers in the grasshopper, seed the muscle pattern and are irreplaceable for the specification of muscle identity. In Drosophila, the adult muscles appear seeded in a similar way. When twist expression declines rapidly following the onset of germ band retraction, it leaves behind a small subset of twist expressing cells as isolated single cells associated with the developing PNS (Bate et al., 1991). These single cells divide starting in early larval stages to form the progenitors for specific subsets of adult muscles (Broadie and Bate, 1991; see also Chapter 2.2). There are considerable parallels between the specification of muscle founder cells and neuroblast specification in the developing nervous system. As in the nervous system, the earliest indication of muscle specification is the appearance of competence domains expressing the proneural gene lethal of scute (l(1)sc) (Carmena et al., 1995, 1998). As in the nervous system, each muscle competence group gives rise to a single progenitor (the founder cell), which specifies a specific muscle in the mesoderm. Furthermore, as in the nervous system, the Notch signaling pathway blocks all other cells from adopting the founder cell fate (lateral inhibition mechanism). In neurogenic mutants, the number of founder cells in the mesoderm, defined by patterns of gene expression, is dramatically increased. Rather than discrete individual cells, clusters of presumptive founder cells are observed at similar positions in the developing mesoderm (Ruiz-Gomez et al., 1997). Thus, as in the nervous system, neurogenic genes appear to act in a lateral interference mechanism to select a specific muscle founder cell from a group of competent cells. 2.3.4.1.3. Development of muscle integrity Prior to recognition of the muscle target by the motor axon growth cone, Drosophila myoblast fusion is complete and the newly formed syncytial myotubes have finalized their attachments to the epidermis, generating the mature muscle pattern. Consistent with this timing, the arrest of motor axon outgrowth does not alter the development of the muscle pattern (Broadie and Bate, 1993c). However, at this

stage (1213 h at 25  C; $60% embryonic development), myotubes are extensively electrically and dyecoupled together into large multifiber networks, but lack individual integrity (Broadie and Bate, 1993e). Myotubes are interconnected to their neighbors via extensive cytoplasmic bridges, surrounded by thin, patchy basal laminae (Rheuben et al., 1999). Coupling occurs between myotubes of different positional classes (longitudinal, transverse, oblique) within a hemisegment, as well as, strangely, across segmental borders. It is not clear how cytosolic bridges span the segmental boundaries. The purpose of the extensive connectedness of developing myotubes is not known either, but it may aid in intercellular communication or help to maintain synchronous myogenesis within and between segments. During early stage 16 (1314 h at 25  C; 6265% development), Drosophila myotubes abruptly and synchronously uncouple (Broadie and Bate, 1993e). This uncoupling coincides with the contacts between myotubes and motor neuron partners. Uncoupling also appears to coincide with the onset of electrogenesis in the myotubes (Broadie and Bate, 1993d). However, because it is difficult to effectively voltage-clamp the coupled myotubes, it is formally possible that currents may be initiated at earlier stages. In the absence of innervation, myotubes uncouple at the correct time and to the correct extent, and the appearance and maturation of muscle electrical properties is not detectably perturbed (Broadie and Bate, 1993d). Therefore, uncoupling is not due to instruction from the motor neuron. After uncoupling, detectable dye transfer is restricted to <10% of dye injected myotubes, a figure that remains relatively constant for the remainder of embryonic and larval development (Ueda and Kidokoro, 1996). Although relatively strong electrical coupling between myotubes persists through maturity, tight voltage-clamp of individual myotubes is now readily accomplished. At maturity, the intersegmental coupling coefficients between muscles 6 and 7 remain high (range of 0.330.43), although coupling with other muscles within the segment is much lower or undetectable (Ueda and Kidokoro, 1996). Coupling of adjacent muscle 6 cells between segments (intrasegmental coupling) is still reliably detected (coupling coefficient of 0.16). The purpose of muscle electrical coupling is unknown, but it may be to aid the coordination of muscle contraction during locomotion, particularly between muscles acting like a single functional unit, such as the ventral longitudinal muscles 6 and 7. In the Drosophila neuromusculature, as in the neuromusculature of other insects (Walther, 1981), the myotubes appear to actively engage in guiding

Functional Development of the Neuromusculature

109

motor axons to the correct target area of innervation (Bate, 1990). Myotubes extend dynamic Actinbased processes towards the filopodia of the motor axon prior to contact. The resemblance of the two motile projections has led to the term myopodia to describe the myotube processes (Ritzenthaler et al., 2000). Following the completion of myoblast fusion, myopodia appear to extend randomly from the myotube, but as they contact and court the motor axon filopodia, they become progressively restricted to the developing site of innervation. This restriction to the innervation site is due to interaction with the neuronal growth cone, as myopodia clustering does not occur in the absence of motor neuron outgrowth (Ritzenthaler et al., 2000). It is unclear whether restriction of myopodia to the innervation site is due to instruction from the motor axon growth cone or reflects mutual interactions between the synaptic partners. However, genetic manipulation of the expression of the myotube proteins a-Integrin (PSII) and Toll, which regulate synaptic adhesiveness, reduces the frequency of myopodia clustering during innervation (Ritzenthaler and Chiba, 2003). It was, therefore, suggested that myopodia clustering is due to the adhesion between a subset of myopodia and filopodia of the innervating growth cone, which maintains these processes while the others are eliminated. Although these observations show that myopodia join filopodia in a dynamic dance during target recognition, there is as yet no indication of the role myopodia play in this process, or even whether they are required for neuromuscular synaptogenesis.
2.3.4.2. Development of Muscle Electrical Properties
p0400

The first intracellular recordings from insect muscle (and functional characterization of insect neuromuscular transmission) was published in 1953 by Castillo, Hoyle, and Machne (Castillo et al., 1953). Since then, insect muscles have been the subject of many detailed electrophysiological studies, and a large body of literature exists that describes these findings (Ashcroft, 1982; Aidley, 1985). Few of these studies have sought to understand development of neuromuscular transmission, however, or sought to trace changes in electrical properties over time (Saito and Kawai, 1987; Rose et al., 2001; Keyser et al., 2003). The relevant findings from Drosophila are reviewed below. As mentioned earlier (Section 2.3.1.1.3), five voltage-gated currents are present in the larval muscles; an inward Ca2 current (ICa), two outward K currents (IA, IK), and two outward Ca2 dependent K currents (ICF, ICS) (Singh and Wu, 1999). These

currents mature in a rapid sequential order and are first detectable at the time that the myotubes uncouple during early stage 16 ($62% embryonic development), when muscles can first be reliably voltage-clamped (Broadie and Bate, 1993a, 1993d). Apparent already when the myotubes uncouple, is a small inward, voltage-gated ICa. Initially, this is the only current apparent in voltage step recordings, indicating that it is the first voltage-gated current present in embryonic muscle (Broadie and Bate, 1993c, 1993d). A stretch-activated, outward K current is also recordable as soon as myotubes uncouple, and this persists throughout later development. Within 1 h of myotube uncoupling, at $14 h ($66% development), outward voltage-gated K currents appear in the myotubes. Two distinct classes of K current appear at the same time: (1) a delayed, noninactivating K current (IK) and (2) a rapid-onset, rapidly inactivating A current (IA) (Broadie and Bate, 1993d). Finally, much later in embryogenesis, during mid-late stage 17 ($18 h at 25  C; $85% development), calcium-dependent K currents appear. Again, two distinct classes of current are first detectable at or close to the same time: (1) a rapid, inactivating ICF and (2) a delayed noninactivating ICS (Broadie and Bate, 1993e). The developmental profile of these different myotube currents is specific for each class of current and very different between currents. ICa is already quite robust and easily detectable when the myotubes uncouple, and its amplitude increases gradually but continuously throughout embryogenesis; at hatching, the ICa current peak amplitude has approximately doubled (Broadie and Bate, 1993c, 1993d). In contrast, the amplitude of the K currents increases much more rapidly soon after their appearance. The IK current climbs quickly in the first 2 h following myotube uncoupling, then levels off to a relatively gradual increase for the rest of embryogenesis (Broadie and Bate, 1993d). The IA current climbs much more rapidly showing a dramatic increase in amplitude soon after myotube uncoupling at $18 h ($85% development). At this point, IA has peaked and shows no further increase through hatching (Broadie and Bate, 1993d). The late appearing calcium-gated K currents are small relative to the other currents, and display a gradual rise in amplitude during the last hours prior to hatching. As a result of these developmental profiles, the voltage-gated K currents dominate the myotube current profile throughout embryonic myogenesis, with the IA current having the largest amplitude and the IK current having the second largest. In whole cell recordings, the inward ICa current is masked by these outward K currents

110 Functional Development of the Neuromusculature

but, when the K currents are removed, the ICa is revealed to be robust and have the third greatest amplitude. The calcium-gated K currents contribute only marginally to the myotube current through the end of embryogenesis. Development of these currents in postembryonic muscles has not been reported. The timing and order of appearance of currents in myotubes differs from motor neurons. In general, electrogenesis in neurons is delayed, albeit relatively slightly, relative to myotubes. In both cases, ICa and IK are the first currents to appear, nearly synchronously, but the order of their very earliest detection is reversed in neurons relative to myotubes. In neurons, the next current to appear is INa, which is not present in insect muscles. Finally, IA appears markedly later in neurons than in muscles. In neurons, the appearance of IA coincides with the first action potentials (see above). The Ca2 dependent K currents have not been reported in embryonic neurons, either because they are not yet present or are not easily detectable. In embryonic muscles, the Ca2 dependent K currents are the last to appear (just prior to hatching), and remain very small and relatively hard to detect. The properties of currents in embryonic myotubes appear similar to currents in mature larval muscle. The notable exception is the embryonic IA current. During early myogenesis, prior to 14 h at 25  C (<65% development), the IA current has a midpoint of steady-state inactivation 40 mV more negative than in the mature embryo or larva (Broadie and Bate, 1993d). An IA current with this inactivation gate should be entirely inactivated at the normal embryonic myotube resting potential of $60 mV. Experimentally, a prolonged, negative voltage prepulse of $100 mV is required to completely activate the IA current in the early myotube (Broadie and Bate, 1993d). Therefore, the prominent IA current reported here is predicted to be completely nonfunctional under normal developmental conditions. As myogenesis proceeds, however, a developmental change can be observed, as the IA steady-state inactivation curve develops a biphasic character (Broadie and Bate, 1993d). This change suggests that a low-voltage inactivation IA channel is present early and replaced later by a high-voltage inactivation IA channel. The IA current at all developmental stages is completely eliminated in Shaker (ShKS133) mutants, suggesting that the Shaker IA channel may be differentially regulated during development relative to the mature muscle (Broadie and Bate, 1993d). The mechanism and significance of this developmental shift in IA channel properties remains unclear.

2.3.4.3. Development of Muscle Contractile Properties and Patterned Movement

The development of neuromuscular behavior in Drosophila has recently been quantitatively assayed using videomicroscopy to monitor peristaltic muscle contraction waves during the last quarter of embryogenesis, starting shortly after the completion of the muscle pattern and innervation through hatching (Suster and Bate, 2002). The muscle contractile machinery driving this movement closely resembles that of the vertebrate muscle: the basic contractile unit is the sarcomere, a Myosin thick filament surrounded by a regular array of Actin thin filaments (Zhang et al., 2000). Myosin and Actin are both expressed first in myoblasts during the process of fusion to form myotubes ($10 h at 25  C; 4050% development). If fusion is prevented, Myosin expression is clearly maintained in the unfused myoblasts (Rushton et al., 1995), indicating that the synthesis of contractile machinery is independent of myotube formation. Assembly of the contractile apparatus is a lengthy process, with the muscles first assembling recognizable sarcomeres at the end of stage 15 ($12.5 h at 25  C; $60% development). The first weak endogenous movements are observed at around 1415 h (6671% development), and electrical stimulation of the motor nerve first leads to a detectable muscle twitch during this same period (Broadie and Bate, 1993e). The embryo first begins to move vigorously during the period of 1618 h (7686% development), at which time clear body length peristaltic waves are first observed. Embryonic movements include vigorous, repeated forward and backward waves of peristalsis (Suster and Bate, 2002). When NMJ neurotransmission is blocked with transgenic toxin expression or in a range of mutants (syntaxin, synaptobrevin, dunc-13), peristalsis is eliminated, indicating that it is completely dependent on neural input (Broadie et al., 1995; Sweeney et al., 1995; Aravamudan et al., 1999). In contrast, local body contraction of one or a few segments persists, as do quite vigorous head movements, driven by the myogenic activity of the pharyngeal muscles. Thus, embryonic peristalsis, similar to other repetitive movements, depends on central pattern generating (CPG) circuits that functionally mature during the last quarter of embryogenesis (Suster and Bate, 2002). Interestingly, removal of sensory inputs during development has shown that these are not essential for the differentiation of the CPG or maturation of behaviorally normal movement (Suster and Bate, 2002). By 1819 h (8590% development), the embryos seem

p0425

p9055

Functional Development of the Neuromusculature

111

to have perfected larval locomotory movements. Within the egg shell, both forward and reverse locomotory movements are evident, involving coordinated waves of muscle contraction along all the body segments (Suster and Bate, 2002). Following this apparent entrainment period, the embryo becomes markedly inactive during the last few hours of embryogenesis. Quite abruptly, the hatching movement program is then engaged, initially involving vigorous head movements to puncture the egg shell and then forward peristaltic waves to drive the new larva free.

2.3.5. Development of NMJ Presynaptic Function


Once a motor neuron growth cone contacts an appropriate muscle target, the filopodia-ringed vacuolated sheet that characterizes the seeking axon terminus transforms itself into a branched presynaptic terminal with swellings (or boutons) specialized for neurotransmitter release (Rheuben et al., 1999). Functional presynaptic terminals normally differentiate after contact with postsynaptic muscle, but the muscle targets do not need to be appropriate, or even present, for formation of relatively normal presynaptic structures (Prokop et al., 1996). In the grasshopper Barrytettix, for example, presynaptic terminals normally persist in adulthood opposite inappropriate targets, such as glia and basal lamina, after previously innervated nymph muscles degenerate (Arbas and Tolbert, 1986). Thus, presynaptic development, once triggered by unknown signals, appears to proceed to completion without any target derived signal. Without a differentiated postsynaptic target, however, presynaptic specializations are oriented randomly in the presynaptic membrane rather than exclusively toward the postsynaptic muscle target (Prokop et al., 1996). For example, Drosophila twist mutants do not form myoblasts (see above) and, thus, motor terminals are completely deprived of target cells. Nevertheless, fully differentiated presynaptic terminals form, except that presynaptic densities, with associated clusters of SVs, are oriented randomly. Fully formed but improperly oriented presynaptic specializations are also seen in mef-2 mutants, which are capable of producing only undifferentiated myoblasts (see above; Prokop et al., 1996; Prokop, 1999). It is as if axons know when and approximately where to form a presynaptic terminal (presumably via regional cues during outgrowth) but, without interaction from a differentiated postsynaptic partner, they do not know exactly which way to point it.

As with other aspects of embryonic body wall development, differentiation of presynaptic nerve terminals occurs in a ventral to dorsal wave, as axons growing outward from the ventral ganglion around the periphery of the cylindrical embryo meet increasingly distant and more dorsal muscles (Broadie et al., 1993; see Figure 1). This wave of development occurs throughout the latter half of embryogenesis, but differentiation of individual terminals occurs much more rapidly. In Drosophila, where embryogenesis lasts $21 h at 25  C, individual terminals mature from seeking growth cones to functional presynaptic structures in a few (23) hours (Broadie et al., 1993; Yoshihara et al., 1997). By the end of embryogenesis, each muscle is functionally innervated, but new boutons continue to be added, grow and differentiate throughout larval development.
2.3.5.1. Functional Development of the Presynaptic Terminal

The basic processes of presynaptic function appear to be highly conserved across phyla. Thus, insect nerve terminals are structurally and functionally very similar to those described in other animals, including the extensively described terminals of frog, mouse, and rat. Development of presynaptic function is difficult to track because neurotransmitter release is typically measured using activation of postsynaptic receptors as an assay. Nevertheless, all evidence suggests that at least some functional synaptic vesicle release machinery assembles prior to muscle target contact (Fletcher et al., 1994). For example, experiments in Drosophila, using voltageclamped sniffer cells to detect neurotransmitter release from growing neuronal axons in culture, show that presynaptic terminals release neurotransmitters prior to contacting their postsynaptic target (Yao et al., 2000). Spontaneous and evoked activation of postsynaptic GluRs is also immediately detected after the nascent presynaptic terminal contacts target muscle, suggesting that presynaptic terminals are already functional before synaptogenesis per se (Broadie and Bate, 1993b, 1993e). A more direct assay for presynaptic function is the uptake and release of lipophilic fluorescent dyes such as FM1-43, which stains and destains functional terminals during synaptic vesicle endocytosis and fusion, respectively (Cochilla et al., 1999). Chick growth cones in culture stain with FM1-43 (Diefenbach et al., 1999), but this dye uptake and release may be due to extensive membrane turnover caused by neurite growth and reorganization rather than synaptic vesicle cycling. Drosophila embryonic motor terminals also cycle FM1-43 efficiently

p0440

p9060

112 Functional Development of the Neuromusculature

p0445

(Fergestad and Broadie, 2001), but their ability to do so prior to target contact has not been examined. Other hallmarks of presynaptic differentiation include immunocytochemical detection of vesicle proteins such as Synaptotagmin, Synaptobrevin, and Cysteine string protein (CSP) (Parfitt et al., 1995; Wu and Bellen, 1997), which appear as distinct puncta in differentiated presynaptic terminals (Featherstone et al., 2001). Functionally mature synaptic terminals also concentrate neurotransmitters in their cytosol, a necessary prerequisite for pumping it up a steep concentration gradient into SVs. Glutamate immunoreactivity, localization of synaptic proteins such as Synaptotagmin, Synaptobrevin and CSP, clustered SVs, and presynaptic dense bodies (T-bars) are all visible in growth cones and during the earliest stages of presynaptic terminal differentiation, again supporting the idea that the presynaptic specializations are functional before target contact (Rheuben and Kammer, 1981; Broadie and Bate, 1993e; Broadie, 1996). Spectrins are cytoskeletal membrane matrix proteins expressed ubiquitously in Drosophila, but concentrated at NMJs (Featherstone et al., 2001). Spectrins are also prominent cytoskeletal components of mammalian synapses (Phillips et al., 2001). Mutations that eliminate alpha and beta Spectrin in Drosophila show that both proteins are required for localization of most presynaptic proteins, including CSP, Synaptotagmin, Synapsin, and Syntaxin (Featherstone et al., 2001), but neither nerve terminal morphology nor synaptic vesicle number or distribution are affected. Because of the mislocalization of presynaptic proteins, synaptic transmission is reduced to approximately 25% of normal. These results suggest that synaptic proteins are organized (probably indirectly) by a Spectrin scaffold during synaptogenesis.
2.3.5.2. Maturation of Neurotransmitter Release

p0455

In the NMJ of the Drosophila ventral longitudinal muscle 6/7, evoked EJCs increase from about 2000 pA in late-stage embryos to approximately 400 000 pA in third instar larvae, a 200-fold increase (Broadie and Bate, 1993e; Rohrbough et al., 1999; Broadie, 2000). This increase is probably necessary in order to compensate for the large developmental drop in muscle input resistance (Lnenicka and Keshishian, 2000), which otherwise would lead to insufficient postsynaptic depolarization in larvae. This dramatic increase in synaptic efficacy could be accounted for by several factors: (1) an increase in the number of presynaptic boutons; (2) an increase in the number of active zones within each bouton; and/or (3) an increase in the number of

postsynaptic GluRs (Featherstone et al., 2002; Reiff et al., 2002). The embryonic muscle 6/7 NMJs contain approximately 10 boutons. In third instar larvae, there are 50100 boutons. Embryonic boutons typically contain 68 presynaptic active zones (Rheuben et al., 1999), while third instar larvae boutons possess two to three times that number (Atwood et al., 1993; Renger et al., 2000). Therefore, the number of synapses (active zones) at each NMJ increases 10- to 30-fold over the course of late embryonic and larval development. Each active zone also increases in size, from about 200 nm in embryos (Prokop et al., 1996; Prokop, 1999) to about 500 nm in third instar larvae (Atwood et al., 1993; Rheuben et al., 1999; Renger et al., 2000). Consistent with this presynaptic change seen ultrastructurally, postsynaptic glutamate receptor clusters imaged confocally also increase in size approximately threefold between the end of embryogenesis and the end of third instar stage (D.E. Featherstone, unpublished data). In other words, embryonic through larval development of each NMJ involves all three changes mentioned above, i.e., addition of boutons, addition of active zones within each bouton, and an increase in size of each active zone. Nevertheless, these preand postsynaptic morphological differences account for only about half of the increase in EJC size measured during larval development. Thus during larval development, individual active zones also undergo functional changes that increase neurotransmitter release. Such changes could be due to increased presynaptic calcium channel influx (Yoshihara et al., 2000; Reiff et al., 2002), facilitated SNARE-mediated vesicle fusion, perhaps due to increased calcium sensitivity (Littleton et al., 1993a, 1999), or more efficient vesicle cycling (Kuromi and Kidokoro, 1999). Neurotransmitter release is highly dependent on extracellular calcium, which invades the presynaptic terminal when voltage-gated calcium channels open during presynaptic depolarization. Comparison of different electrophysiological studies suggests that the calcium dependence of NMJ neurotransmitter release approximately doubles between late embryos and third instar larva (Broadie, 1999). Thus, more release is achieved due to increased sensitivity to calcium. In Drosophila, the terminal membrane area increases many fold during development (from about 3 mm2 in embryos to about 100 mm2 in third instar larvae), potentially increasing the number of routes for calcium influx. Unfortunately, developmental studies of calcium channel density or calcium influx in terminals have not been performed. One explanation for the increased calcium sensitivity could be a developmental shift in the

p0460

Functional Development of the Neuromusculature

113

expression of Synaptotagmin family members, which confer different calcium dependence to release (Littleton et al., 1999). Several Synaptotagmin genes are indeed expressed in Drosophila, but a developmental shift in their expression at the NMJ has not yet been reported. From embryonic synaptogenesis through larval stages, each active zone also becomes associated with an increasing number of SVs. In the embryonic NMJs, for example, T-bars are typically surrounded by 510 vesicles, with few or no vesicles in the surrounding cytosol. In third instar larvae, T-bars are often surrounded by 23 dozen vesicles, with hundreds more in the nearby cytosol. The increased availability of SVs during development probably results in the fusion of more vesicles per stimulus at each active zone, as well as in fatigue resistance due to a larger number of reserve vesicles. Thus, the efficacy of individual active zones also increases during development, although the molecular mechanisms underlying these presynaptic changes remain largely unexplored.
2.3.5.3. Maturation of Mature Synaptic Properties

Late-stage embryonic and larval NMJs also show increasing facility for activity-dependent synaptic plasticity, wherein synaptic properties change following specific patterns of stimulation (described in more detail in Sections 2.3.1.4 and 2.3.8). These changes may reflect maturation of intracellular signaling pathways and mechanisms for modulation (see discussion below).
2.3.5.4. Differentiation of Synaptic Vesicle Pools and Cycling

Early in synaptogenesis, when the motor nerve terminal first contacts the postsynaptic muscle (13.5 h; 60% embryonic development at 25  C), endogenous currents at the nascent NMJ can be observed, although evoked transmission is easily fatigued and variable (Broadie and Bate, 1993e). Before 15 h of embryogenesis (68% development), when NMJs have just formed, the synapse fatigues rapidly and completely even with relatively low frequency (<5 Hz) stimulation (Broadie and Bate, 1993e). Between 16 and 18 h (7282% development), NMJ transmission is more consistent, and patterned synaptic activity is observed such that higher stimulation frequencies (>5 Hz) are needed to induce fatigue, from which the terminal recovers within a few seconds (Broadie and Bate, 1993e). After 18 h, the NMJ maintains reliable transmission with little fatigue at stimulation frequencies up to about 10 Hz (Broadie and Bate, 1993e). Reliability and fatigue resistance increase throughout larval development. Thus, third instar Drosophila larvae are capable of maintaining transmission (albeit with marked depression) at very high stimulation frequencies (up to 50100 Hz) with almost no failures (Rohrbough et al., 1999; Featherstone, personal communications). These functional changes during development can be attributed again to several factors, including the addition of presynaptic boutons, the addition of active zones within those boutons, an increase in active zone size, an increase in functional efficacy of each active zone, and an enlargement of the postsynaptic receptor field.

Most (>70%) SVs in embryonic motor terminals cluster around the active zone; relatively few vesicles are distributed throughout the center of the bouton, despite the large amount of space available (Prokop et al., 1996; Prokop, 1999). In larval terminals, however, bouton centers are often filled with vesicles, and vesicles are stacked many rows deep around active zones (Atwood et al., 1993; Rheuben et al., 1999; Renger et al., 2000). Electron microscopy, in combination with electrophysiological and mutant analyses, has shown that the vesicles immediately adjacent to active zones are sufficient for basal neuromuscular transmission. Furthermore, ultrastructural and dye studies suggest that, typically, only a minority of presynaptic vesicles are usually involved in basal transmission. Locust terminals, for example, only show, on average, a 20% decrease in vesicle density after stimulation (Titmus, 1981). In Drosophila, high bath potassium, which depolarizes cells and triggers release, causes release of about half of the total number of SVs (Kuromi and Kidokoro, 1998). Only under very high frequency stimulation or, especially, in conditions where endocytosis is inhibited, are the majority of vesicles recruited (Kuromi and Kidokoro, 1998, 1999, 2000, 2002). These reluctantly recruited reserve vesicles are the morphologically central vesicles not associated with active zones. Under conditions of high vesicle demand, such as high-frequency stimulation or impaired endocytosis, these reserve vesicles supplement and/or replace vesicles from the primary pool. In embryos, the reserve pool is absent there are only vesicles immediately around T-bars, and the centers of boutons are devoid of SVs. Consistent with this, embryos display markedly lower fatigue resistance (Broadie and Bate, 1993e), because they have no reserve vesicle pool (Prokop, 1999; Fergestad and Broadie, 2001). Because embryos only contain vesicles around their T-bars, we can also infer that nascent vesicles, which all embryos have, are first targeted to active zones and, when these areas are sufficiently occupied, a reserve pool begins to develop. In mutant larvae that develop without

p0475

p0480

114 Functional Development of the Neuromusculature

Endocytosis (e.g., Drosophila Endophilin mutants), the cycling pool is greatly reduced and the reserve pool is completely absent (Verstreken et al., 2002), again arguing that vesicles are first targeted to active zones and then, as this area fills, to the reserve pool. Because Cytochalasin D, a selective disruptor of filamentous Actin, eliminates the reserve pool, it is thought that actin is responsible for incorporating or maintaining SVs in the reserve pool (Kuromi and Kidokoro, 1998). CyclicAMP signaling mutations also disrupt reserve vesicle recruitment (Kuromi and Kidokoro, 2000), demonstrating a need for this pathway in building and/or maintaining the reserve vesicle pool. In Drosophila, cAMP levels are increased by cytosolic calcium (Kuromi and Kidokoro, 2002), suggesting that calcium probably regulates vesicle pool size in insects as it does in mammals. Taken together, these observations suggest that the increased fatigue resistance seen in late larval NMJs, compared to embryonic NMJs, is due to gradual accumulation of a reserve vesicle pool, which forms after the primary vesicle pool.

2.3.6. Development of NMJ Postsynaptic Function


2.3.6.1. The Muscle Membrane Prior to Innervation

Prior to innervation, the Drosophila body wall muscle membranes express cell adhesion molecules (CAMs) such as FasIII, FasII, Connectin, and Toll, which appear to mediate motor axon guidance and/ or target recognition. FasII and Connectin are initially expressed throughout the muscle membrane, but are localized to NMJs during the initial stages of NMJ formation. FasIII, on the other hand, is initially loosely localized only near the normal site of muscle innervation, as if to mark the site, and then disappears after the start of NMJ development. Consistent with the idea that these CAMs mediate cellcell stickiness, overexpression of FasII results in ectopic synapses due to trapping of growth cones that normally explore several potential targets before selecting the proper postsynaptic cell. Similarly, misexpression of FasIII causes inappropriate synapses between axons that normally innervate FasIII-expressing muscles and the muscles that misexpress FasIII (Chiba et al., 1995; Suzuki et al., 2000). Toll, on the other hand, is repulsive to neurons that do not ordinarily innervate Toll-expressing muscles (Rose et al., 1997; Suzuki et al., 2000). Since each motor neuron innervates a unique muscle (or set of muscles), there is presumably a unique molecular cell adhesion code that allows matching between

appropriate pre- and postsynaptic partners (Chiba and Rose, 1998). The CAM interactions between undecided partners are apparently aided by mobile cellular protuberances on both the pre-(growth cone filopodia) and postsynaptic sides. Decades ago, it was observed that the embryonic insect muscle appears ruffled at the normal site of innervation (Walther, 1981). Close examination has revealed that the membrane ruffling is due to small muscle membrane projections reminiscent of the axonal growth cone filopodia. These projections, termed myopodia, are extended randomly from the muscle surface before innervation, but cluster at the site of innervation during contact by the appropriate motor neuron growth cone (Ritzenthaler and Chiba, 2003). Fast excitatory signaling at insect NMJs occurs via postsynaptic ionotropic GluRs (Usherwood, 1977). In Drosophila NMJs prior to innervation, functional receptors are scattered in small numbers throughout the postsynaptic muscle membrane, such that iontophoresis of glutamate opens only one or two channels regardless of application site (Broadie and Bate, 1993b). Discs Large (Dlg) is the Drosophila homolog of mammalian PSD-95, a MAGUK (membrane-associated guanylate kinase) protein that is prominent at glutamatergic synapses. Similar to GluRs, Dlg is diffusely expressed initially, but localized and upregulated at synapses after innervation (Thomas et al., 2000).
2.3.6.2. Development of Postsynaptic Glutamate Receptors

p0495

Once the proper axon and muscle contact each other, they form a functional, albeit weak, synapse within minutes (Broadie and Bate, 1993b, 1993e). Consistent with this, postsynaptic glutamate receptors are distributed at low levels throughout the muscle membrane prior to innervation. Under normal circumstances, receptors outside the area of muscle contact disappear within an hour, while the number of synaptic receptors simultaneously increases (Broadie and Bate, 1993b; Saitoe et al., 1997), presumably as preexisting receptors cluster at the site of innervation (however, the possibility that new synaptic receptors are inserted while extrasynaptic receptors are degraded, has not been excluded). In Drosophila, EJCs, measured from barely innervated ventral muscle 6 at 13.5 h (60% of embryonic development), increase from single channel size (about 12 pA at 60 mV) to approximately 700 pA by 15 h (68% development). After this initial phase of postsynaptic development, EJC size gradually triples (to about 2000 pA) by the end of embryogenesis (22 h) (Broadie and Bate, 1993b),

Functional Development of the Neuromusculature

115

as new receptors are synthesized and inserted. Glutamate receptor upregulation in Drosophila fails if presynaptic electrical activity is completely blocked using either Tetrodotoxin (TTX) or genetic ablation of voltage-gated sodium channels (Broadie and Bate, 1993a; Saitoe et al., 1997). Incomplete blockage or increased activity have relatively minor

effects on postsynaptic development (Broadie and Bate, 1993a) (see Figure 4). Without innervation, GluRs are neither localized nor upregulated (Broadie and Bate, 1993b). Furthermore, clustering and upregulation of receptors occurs at ectopic synapses. Thus, contact by the presynaptic nerve terminal is both necessary and

p0505

Figure 4 Confocal micrographs showing the development of Drosophila ventral longitudinal embryonic/larval NMJs, visualized using antibodies that recognize neuronal tissue (HRP, red) and glutamate receptor subunit GluRIIA (green). (ac) Motor axons from the intersegmental nerve (ISNb) branch b form NMJs on ventral longitudinal muscles 7, 6, 13, and 12, in very late-stage (22 h, 100% embryonic development) embryos (a), first instar larvae (b), and second instar larvae (c). (d) The 6/7 NMJ in a third instar larva. Scale bar: 10 mm in all panels. Note the slight change in scale bar length between images. Over development, NMJ arbors grow larger and branchier, and glutamate receptor cluster size increases about threefold in area. Note also the large number of extrasynaptic glutamate receptor clusters visible in third instar muscle (d). Panel (e) shows a three-dimensional isosurface rendering of the same NMJ presented in panel (d), rotated slightly to show the edge of the confocal section visible near the upper left, where the axons extend to meet the bottom side of the muscles (which is the side facing away from the interior of the larvae).

116 Functional Development of the Neuromusculature

sufficient for postsynaptic development. Similar to GluRs, Dlg is diffusely expressed initially, but localized and upregulated at synapses after innervation (Thomas et al., 2000). Like GluRs, Dlg is neither synaptically localized nor upregulated in the absence of innervation (D.E. Featherstone, unpublished data). The signal from nerve to muscle that triggers postsynaptic development, however, is unknown. In mammalian NMJs, the secreted proteoglycan Agrin triggers receptor clustering and upregulation of postsynaptic ACh receptors via a muscle specific receptor tyrosine kinase (MuSK) (Sanes and Lichtman, 1999). Glutamatergic synapses, including insect NMJs, however, appear to use a different signal for postsynaptic induction (Colledge and Froehner, 1998). No obvious Agrin homolog is encoded by the Drosophila genome, and several kinases similar to MuSK, such as Neurospecific receptor kinase (Nrk), Heartless (Htl), and RAR-related orphan receptor (Ror), appear to be expressed in neurons (where Ach receptors are expressed), but not muscle. The Wnt protein encoded by the wg gene has recently been shown to concentrate in developing larval NMJs, where it appears to be secreted by the presynaptic terminal. In wg mutants, or larvae with mutations in the TGF b type II receptor encoded by the gene wishful thinking (wit), NMJs remain underdeveloped (Packard et al., 2002). Thus, Wg signaling via Wit has been proposed to be an essential intercellular signal necessary for coordination of pre- and postsynaptic differentiation (Packard et al., 2003). Wnts are known to participate in many other cellcell interactions during development, and have been implicated in presynaptic differentiation in mammalian synapses (Salinas, 1999). However, although Wnt signaling appears to regulate larval NMJ development, it does not seem to be required for synaptogenesis, and it remains to be established whether Wnts are secreted by insect motor neuron growth cones prior to target contact (which would be required for the initial induction signal). The Spectrin cytoskeleton is required for proper localization of Dlg but not glutamate receptors (Featherstone et al., 2001), suggesting that although clustering and upregulation of both these postsynaptic proteins depends on a nerve-derived signal, the mechanisms by which they localize diverge thereafter. Postsynaptic glutamate receptor (GluR) mRNA is translated at NMJs and the translation initiation factor eIF4E as well as the poly(A)-binding protein PABP, which are required for translation, are localized in the postsynaptic muscle near NMJs. Their genetic disruption alters synaptic development and GluR function (Sigrist et al., 2000). Consistent with

local translation, Sigrist et al. (2000) also showed that GluR mRNA is localized near NMJs. This result, however, differs from earlier examinations which showed that GluR mRNA is widely distributed in embryonic and larval muscle fibers (Currie et al., 1995). Induction of GluR fields by the presynaptic nerve could trigger transcription of GluR mRNA, local translation of pre-existing mRNAs, or both. Preliminary evidence (Q. Sheng and D.E. Featherstone, unpublished data) suggests that innervation dependent upregulation of glutamate receptors during embryogenesis does not involve transcription, since mRNA levels remain unchanged while receptor protein accumulates. Once induced, postsynaptic GluRs in Drosophila form distinct clusters under presynaptic terminals (Featherstone et al., 2002). Between late embryogenesis and third instar larval stage, the GluR cluster size area increases threefold (from about 0.5 mm2 to about 1.5 mm2; D.E. Featherstone, unpublished data). This size increase matches almost perfectly the postsynaptic density diameter measured by electron microscopy (Rheuben et al., 1999) and the increases in miniature junction currents recorded electrophysiologically (which increase from about 150 pA in embryos to 500 pA in third instars) (Rohrbough et al., 1999; Featherstone and Broadie, 2000; Featherstone et al., 2001). Because changes in the receptor cluster area mirror the functional (mEJC size) alterations so closely, receptor density within each cluster probably does not change during development. The number of receptor clusters per bouton also does not increase dramatically during larval development, but the bouton number does; therefore, the number of clusters at each NMJ also increases. In addition to synaptic GluRs, extrajunctional GluRs have been found electrophysiologically and microscopically in the muscle membrane of all insects examined (Cull-Candy, 1976; Usherwood, 1994; Nishikawa and Kidokoro, 1995). There are two types of extrajunctional GluRs: D and H, which, when activated, depolarize and hyperpolarize the muscle membrane, respectively (Cull-Candy, 1976). The H receptor current appears to be carried by Avermectin-sensitive glutamate-gated chloride channels (Cully et al., 1996) and may be specific to adult muscle (Saito and Kawai, 1985). Excitatory D receptors are expressed during larval development, but the function of these nonsynaptic receptors is not understood. One hypothesis is that these extrajunctional D receptors represent synaptic receptors moving to/from the postsynaptic density. This is thought to be the case in embryonic muscle, where the disappearance of extrajunctional

p0520

Functional Development of the Neuromusculature

117

receptors immediately following innervation is coincident with the appearance of junctional receptors (Broadie and Bate, 1993a; Saitoe et al., 1997). As the postembryonic muscle matures, however, extrajunctional receptors return to prominence (in Drosophila, during second instar stage; see Figure 4). Extrajunctional receptors differ from junctional receptors in fundamental functional ways (Usherwood, 1994; Nishikawa and Kidokoro, 1995). For example, extrajunctional receptors have lower conductance, less rectification, and decreased calcium permeability compared to junctional ones (Usherwood, 1994; Nishikawa and Kidokoro, 1995). In the Drosophila muscle, this permeability difference is large enough to shift the receptor current reversal potential from near zero (for extrajunctional receptors) to positive (about 12 mV) values (Nishikawa and Kidokoro, 1995). Extrajunctional receptors also desensitize much more rapidly (in $15 ms; Heckmann and Dudel, 1997) compared to junctional receptors (Nishikawa and Kidokoro, 1995). If extrajunctional receptors represent a pool of synaptic receptors held in reserve or trafficked to/from the synapse, then the differences in functional properties must represent a form of regulation determined by differential localization. Although GABA-gated currents have also been demonstrated in some insect muscle (typically adult muscle), they are not present in the Drosophila embryonic/larval body wall muscle (Broadie, 1999; Featherstone and Broadie, 2000). Glutamic acid decarboxylase (GAD), the enzyme responsible for synthesizing the neurotransmitter GABA is, however, expressed in embryonic and larval motor nerve terminals, where it regulates cytosolic glutamate levels and (indirectly) postsynaptic GluR numbers (see Section 2.3.6.3). The presence of GAD protein in embryonic/larval motor neurons without apparent GABAergic transmission may also suggest that these motor neurons become GABAergic during metamorphosis. However, this possibility has not been investigated and no circumstantial evidence exists for it. Essentially nothing is known about the development of GABAergic transmission in insects, including whether insect GABA receptors are localized using mechanisms similar to those employed by vertebrates (Kardos, 1999; Sassoe-Pognetto and Fritschy, 2000; Fritschy and Brunig, 2003). The Drosophila protein Cinnamon (Cin), previously identified as a molybdenum cofactor, represents a fly homolog of Gephyrin, which is a mammalian protein important for glycine and GABA receptor field development (Kamdar et al., 1994; Wittle et al., 1999). Cinnamons role in insect synaptic development or inhibitory neurotransmission has not yet been explored.

2.3.6.3. Molecular Maturation of Postsynaptic Signaling

As mentioned earlier, Drosophila third instar larval muscles contain large aggregates of the translation initiation factors eIF4E and poly(A)-binding protein (PABP) in or near the SSR of NMJs (Sigrist et al., 2000). According to Sigrist et al. (2000), GluR mRNA is localized near these postsynaptic translation aggregates (but others claim that GluR mRNA is unlocalized; Currie et al., 1995). When synaptic activity or cAMP levels are upregulated using K channel (eag,sh) or cAMP-phosphodiesterase (dunce) mutants, respectively, the number of postsynaptic translation aggregates increases, the amount of postsynaptic glutamate receptors goes up (without a corresponding rise in mRNA), and EJC size increases (Sigrist et al., 2000). Increased larval locomotion also upregulates the number of postsynaptic translation aggregates and receptors (Sigrist et al., 2003). These activity-dependent changes can be rapidly reversed by using a temperature sensitive mutation in a voltage-gated sodium channel to induce paralysis and thus decrease NMJ activity (Sigrist et al., 2003). Taken together, these data suggest that Drosophila NMJ glutamate receptors are synthesized in or near the SSR by translation factors that are somehow regulated by synaptic activity. Postsynaptic synthesis probably allows faster synaptic changes; indeed, Sigrist et al. (2003) measured significant changes in synaptic strength after 2 h. Synthesis and insertion of new receptors is not the only mechanism by which Drosophila NMJs control glutamate receptor number. The number of NMJ postsynaptic glutamate receptors at any time after initial induction represents a balance between receptor addition and receptor downregulation. In Drosophila, the presynaptic neuron slows the addition of postsynaptic receptors by secreting a signal that triggers receptor downregulation. Surprisingly, this signal is glutamate itself, released via an as yet unknown nonvesicular mechanism. In a genetic screen for mutations that affect postsynaptic development, Featherstone et al. (2000) paradoxically discovered that GAD enzyme levels in glutamatergic terminals regulate the number of postsynaptic glutamate receptors. Since Drosophila embryonic/larval NMJs are not GABAergic (Jan and Jan, 1976a; Broadie and Bate, 1993e; Featherstone et al., 2000), it was hypothesized that GADs primary function in the NMJ was to degrade glutamate, and that glutamate caused downregulation of its postsynaptic receptors. By genetically manipulating the concentration of presynaptic cytosolic glutamate,

118 Functional Development of the Neuromusculature

Featherstone et al. (2002) showed that the number of postsynaptic glutamate receptors is inversely correlated with presynaptic glutamate levels, consistent with the hypothesis that extracellular glutamate released from presynaptic nerve terminals causes downregulation of postsynaptic glutamate receptors. Thus, the nerve terminal first induces the clustering and upregulation of glutamate receptors (see Section 2.3.6.2), and then gradually puts a brake on the process using glutamate. Interestingly, glutamate-triggered downregulation of glutamate receptors is not blocked by mutations that eliminate synaptic vesicle fusion (Broadie et al., 1995; Sweeney et al., 1995; Aravamudan et al., 1999; Featherstone et al., 2002), demonstrating that glutamate-triggered receptor downregulation is primarily mediated by nonvesicular glutamate. Even after excitatory glutamatergic transmission at insect NMJs was widely accepted, mammalian physiologists were still questioning whether glutamate could serve as a transmitter in the vertebrate CNS (Nicholls and Sihra, 1986). One of the main arguments against glutamate was the fact that extracellular glutamate levels in the vertebrate nervous system seemed too high. Extracellular glutamate also triggers downregulation of glutamate receptors in cultured mammalian neurons (Lissin et al., 1999), and nonvesicular neurotransmitter release through various antiporters, neurotransmitter pumps, gap junctions and other channels, has recently been demonstrated in mammalian cells (Attwell et al., 1993; Baker et al., 2002; Demarque et al., 2002; Duan et al., 2003; Ye et al., 2003). Regulation of glutamate receptors by nonvesicular glutamate may thus be a conserved mechanism by which presynaptic terminals balance presynaptic properties and postsynaptic sensitivity. However, the mechanisms by which nonvesicular glutamate is released and the methods that the muscle cell uses for receptor downregulation remain to be elucidated. It also remains unknown whether receptors containing specific subunits are preferentially downregulated, so that receptor properties are tuned depending on innervation.

it appears to be regulated by complex and incompletely described cascades of intracellular messengers.


2.3.7.1. Elaboration of Synaptic Architecture

2.3.7. Postembryonic Maturation of the NMJ


Although a functional NMJ is fully established (e.g., functional pre- and postsynaptic elements are in place) by the end of embryogenesis, the synapse continues to grow and elaborate dramatically during the larval stages (Walther, 1981). Almost none of this postembryonic development occurs without extensive input from the synaptic partner, and all of

Between hatching and late larval stages, the size and structural elaboration of NMJs increase dramatically. New synaptic boutons and axonal branches are added to make, in general, a larger, bushier presynaptic arbor. Time-lapse studies suggest that there is no specific growth zone for presynaptic elaboration (Zito et al., 1999). Rather, new terminals or branches are added, apparently randomly, either by splitting of pre-established boutons, the formation of new boutons on pre-established branches, or bouton formation at the end of terminal branches (Zito et al., 1999). NMJ growth and subsequent morphology depends strongly on synaptic activity. As described above (see Section 2.3.6.3), NMJ locomotory activity in larvae induces accumulation of postsynaptic translation aggregates, increased production of postsynaptic glutamate receptors, and increased EJC size within 2 h. On a longer time scale (1224 h), increased NMJ activity also induces dramatic synaptic growth (Budnik et al., 1990; Sigrist et al., 2000; Sigrist et al., 2003). How are NMJ activity and morphology linked? Increased activity, as well as genetic manipulation of cAMP levels, are associated with a large ($50%) reduction in pre- and postsynaptic FasII at the NMJ and increased NMJ size (Davis et al., 1996b; Schuster et al., 1996b; Cheung et al., 1999; Sigrist et al., 2000, 2003). Reductions in FasII are both necessary and sufficient for NMJ growth (Schuster et al., 1996b), but neither the FasII level changes nor the subsequent alterations in NMJ size affect synaptic function (Davis et al., 1996b). Activity-dependent changes in synaptic function associated with NMJ growth require cAMP signaling via the cAMP response element binding protein CREB (Davis et al., 1996b), in addition to the faster changes in glutamate receptor synthesis described above (see Section 2.3.6.3). Thus, NMJ activity induces cAMP levels and FasII downregulation, which upregulate synaptic function and growth, respectively. FasII levels may also be coupled to synaptic activity via presynaptic calcium released from intra- or extracellular stores, and triggered perhaps by phospholipase C activation, since changes in these processes also affect synaptic growth (Hannan and Zhong, 1999). Because FasII is a cell adhesion molecule, it was suggested that synaptic growth is regulated by a careful balance between adhesion (which makes the neuron grip the muscle) and repulsion (which allows the neuron to grow over the muscle surface). Consistent with this, integrins are also expressed at

Functional Development of the Neuromusculature

119

NMJs and required for normal morphology and function of Drosophila NMJs (Beumer et al., 1999, 2002; Rohrbough et al., 2000). Beta PS integrins appear to act through CamKII to regulate Dlg and FasII (Beumer et al., 2002). Dlg is itself regulated by CamKII and required for FasII localization and NMJ growth (Budnik et al., 1996; Thomas et al., 1997; Koh et al., 1999). In addition to a careful balance in cell adhesion, NMJ growth also depends on the microtubule cytoskeleton and control of its assembly (Hummel et al., 2000; Roos et al., 2000; Zhang et al., 2001; Eaton et al., 2002; Pennetta et al., 2002). Drosophila NMJ boutons contain microtubule loops oriented parallel to the postsynaptic muscle membrane (Roos et al., 2000). During formation of new boutons, these cytoskeletal loops are rearranged. The Drosophila Futsch protein is a MAP1B homolog required for the microtubule rearrangements and thus for new bouton formation (Roos et al., 2000). Futsch protein levels are controlled by the fragile X related protein (FXR), a mRNA binding protein and translational repressor (Zhang et al., 2001). In fxr mutants, Futsch levels are increased and NMJs are overgrown, similar to the synaptic overgrowth found in human victims of fragile X mental retardation (Zhang et al., 2001).
s0210

compositions around their nervous system that are dramatically different from the composition found elsewhere in the relatively open blood system (Treherne, 1985; see also Section 2.3.3). When the SSR is insufficiently developed due to mutations that affect levels of Dlg, the number of postsynaptic glutamate receptors is also decreased (Parnas et al., 2001). This change is probably due simply to a reduction in available postsynaptic membrane, since loss of Dlg does not affect glutamate receptor numbers or localization prior to formation of the SSR (D.E. Featherstone, unpublished data).
2.3.7.3. Developmental Feedback between Pre- and Postsynaptic Cells

2.3.7.2. Formation of the SSR

During Drosophila embryogenesis, there is no SSR at the NMJ. At about the time of hatching, postsynaptic muscle cells produce short arm-like membrane extensions around type I presynaptic terminals, which also begin to sink into a growing dip in the postsynaptic membrane. By late larval stages, the presynaptic type I terminal is typically buried within layers of elaborately infolded muscle membrane (Rheuben et al., 1999). The SSR forms in only a subset of NMJs. Specifically, it is most elaborate in type Ib, smaller in type Is, and undetectable in types II and III boutons (Lane, 1985; Rheuben et al., 1999). Presumably, type I terminals possess a specific secreted or transmembrane tag that allows muscles to recognize them as such and build an SSR. No candidates for this presynaptic tag have been proposed. On the postsynaptic side, however, it has been shown that Dlg is critical for SSR formation (Lahey et al., 1994; Budnik et al., 1996; Guan et al., 1996). The insect SSR, like the membrane folds at mammalian NMJs, is thought to allow a higher density of membrane proteins in the region of the presynaptic terminal. The SSR may also be important for control of the fluid microenvironment around the NMJ, since insects have been shown to maintain fluid

After initial synapse formation, the morphology and function of NMJs continue to change during larval development. Typically, postsynaptic muscle volume increases several tenfold, and presynaptic arbor area increases several fold. To maintain coordinated movement, it is important that muscle depolarization following nerve activity remain relatively invariant throughout this dramatic growth. To ensure this, presynaptic neurotransmitter release is developmentally regulated based on levels of postsynaptic depolarization. For example, Drosophila mutants with decreased muscle depolarization (due to manipulation of voltage-gated potassium channels or glutamate receptors in the muscle) develop larval presynaptic terminals with double the normal number of T-bars and more neurotransmitter release (Haghighi et al., 2003). This increase in neurotransmitter release increases transmission throughput and corrects the deficit in muscle depolarization. Because the apparent goal of these changes is to maintain the size of the postsynaptic depolarization, the term synaptic homeostasis has been used (Paradis et al., 2001). Although the retrograde (muscle to neuron) signal that mediates synaptic homeostasis remains unidentified, CamKII is required postsynaptically (presumably to sense depolarization levels), and a bone morphogenetic protein (BMP) type II receptor is required presynaptically (presumably to receive the retrograde signal) (Haghighi et al., 2003). Alterations in glutamate receptor number do not cause compensatory changes in embryonic NMJs, suggesting that homeostatic changes occur during larval development (Broadie and Bate, 1993a; Featherstone et al., 2000, 2002). Paradoxically, manipulations that increase postsynaptic glutamate receptor numbers in larvae (such as overexpression of the receptors themselves, overexpression of postsynaptic translation aggregate proteins, or increased NMJ activity; Sigrist et al., 2000, 2002, 2003) increase the density of

120 Functional Development of the Neuromusculature

presynaptic active zones and number of boutons (Sigrist et al., 2002). Synaptic homeostasis predicts that more receptors would cause a reduction of NMJ size, as the presynaptic cell tries to compensate for increased postsynaptic sensitivity.

2.3.8. Development of Activity-Dependent Synaptic Plasticity


2.3.8.1. Development of Activity-Dependent Functional Plasticity
p0595

sensor-1 (NCS-1), has been shown to be both necessary and sufficient for paired pulse facilitation in glutamatergic synapses (Sippy et al., 2003). Plasticity might improve during larval development as signaling pathways mature (e.g., proteins like Frequenin accumulate at the synapse), or the variable and easily fatigued nature of transmission at younger stages might just make activity-dependent plasticity more difficult to trigger and measure.

Two types of developmental plasticity have been recognized at the Drosophila NMJ: morphological plasticity and functional plasticity. Morphological plasticity involves changes in NMJ size (bouton number, terminal length, or branching), which may or may not alter the size of postsynaptic currents. Functional plasticity involves changes in the size of postsynaptic EJCs, which may or may not involve alterations in NMJ size. As described in previous sections, NMJ activity levels can have dramatic effects on both NMJ growth (morphological plasticity) and EJC size (functional plasticity). Because morphological changes typically take a relatively long time (many hours in Drosophila) to manifest, activity-dependent morphological plasticity is usually assayed in third instar NMJs following chronic genetic manipulations. Activity-dependent functional plasticity, however, can be observed following both chronic (days) and acute (milliseconds minutes) manipulations. The latter type of functional plasticity is considered short-term functional plasticity. As described in Section 2.3.1.4, several types of short-term functional plasticity can be observed in Drosophila NMJs: (1) paired-pulse facilitation (PPF); (2) short-term facilitation (STF); (3) augmentation; and (4) posttetanic potentiation (PTP). All forms of short-term plasticity first develop late in embryogenesis (Broadie et al., 1997), but are much more robust and reproducible in third instar larvae (Zhong et al., 1992; Rohrbough et al., 2000). Mutations that affect learning and memory also affect short-term plasticity in the fly NMJ, suggesting that learning and memory utilizes processes shared by very different synapses. Proteins shown to be required for both learning and memory and NMJ plasticity in Drosophila include integrins (Rohrbough et al., 2000), 14-3-3 proteins (Broadie et al., 1997), and molecules involved in cAMP signaling (Zhong et al., 1992). Frequenin is a novel Drosophila neuronal synaptic calcium binding protein that facilitates frequency dependent facilitation in NMJs (Pongs et al., 1993). The mammalian homolog of Frequenin, called Neuronal calcium

2.3.9. Distant Signals: Hormonal Effects on NMJ Development and Function


The embryonic/larval insect PNS and body wall musculature is, as in most invertebrates, bathed by a relatively open circulatory system. Thus, the NMJ is readily accessible to any of the many substances produced by neurosecretory cells throughout the animal. Any bioactive substance transported from relatively distant cells is a hormone, but since synaptic transmission is really a short-distance form of endocrine signaling specialized for speed and specificity, the difference between presynaptic effects and hormonal effects can degenerate into semantics. For example, in addition to motor inputs utilizing fast neurotransmitters (e.g., glutamate or GABA), the larval body wall musculature also contains terminals that secrete peptides and amines such as PACAP, Proctolin, Leucokinin, Insulin, and Octopamine (Lane, 1985; Hannan and Zhong, 1999). There are receptors for some of these substances on neighboring muscle, but active zones for these modulatory terminals are often oriented away from the muscle, toward the body cavity (Lane, 1985; Orchard and Loughton, 1985; Prokop, 1999). Thus, although the target(s) of these substances may include muscle cells immediately adjacent to the terminal, distant muscle cells and other tissues may also be affected (Orchard and Loughton, 1985). As a result, the peptides and amines could be considered both as transmitters and hormones. The literature covering insect endocrinology is extensive, since almost every aspect of postembryonic insect development is regulated by hormones. Here, just a very few examples of how hormones affect NMJ growth and function are considered.
2.3.9.1. Evidence for Hormonal Regulation of NMJ Development
p0600

Almost all aspects of insect postembryonic development depend strongly on hormones. Development of the NMJ is no exception. Ecdysone is a hormone predominantly associated with developmental

Functional Development of the Neuromusculature

121

changes during metamorphosis, which peaks in concentration during late larval stages just prior to pupation Chapters 2.4, 3.3, 3.5, and 3.10). In Drosophila mutants where ecdysone is eliminated, the strength of NMJ transmission in the early third instar larva is unchanged, but is increased slightly ($20%) in the late third instar larva (Li and Cooper, 2001). Consistent with this, late (but not early) third instar ecdysoneless mutants also show NMJs with longer branches and more synaptic boutons (Li and Cooper, 2001). Thus, ecdysone appears to slow larval NMJ growth prior to pupation. Calcium activated protein for secretion (CAPS) is a highly conserved protein that is required for fusion of dense cored vesicles containing neuropeptides and appears to be necessary for release of all peptide hormones. Genetic elimination of CAPS (and thus all peptide hormone secretion) in Drosophila results in reduced movement and embryonic/early larval lethality. Mutant embryonic NMJs have transmission strength reduced to approximately one-half the normal, with no obvious changes in synaptic morphology. Selective expression of CAPS only in presynaptic neurons does not correct the NMJ defects, demonstrating that the reduction in transmission is entirely due to loss of one or more (unidentified) peptide hormones originating in other cells (Renden et al., 2001). It is possible that, were CAPS mutants to live beyond the early first instar stage, more dramatic effects on NMJ development and function might be observed due to loss of peptidergic output. The specific neuropeptides(s) responsible for deficient NMJ development in dCAPS mutants remains unidentified.
2.3.9.2. Modulation: Short-Term Hormonal Regulation of NMJ Function

FMRFamide-related peptides increase larval muscle contraction. Like FLRFamide, the target of FMRFamide is the NMJ. Direct application of 0.1 mM FMRFamide in Drosophila increases postsynaptic currents to approximately 150% of normal (Hewes et al., 1998). Presumably, any hormone that constitutively affects NMJ function will eventually trigger activity-dependent developmental changes such as homeostatic upregulation of transmitter release or alterations in synapse morphology.

2.3.10. Unanswered Questions


The purpose of this chapter is twofold: (1) to summarize the current state of knowledge about functional neuromuscular development in insects (especially Drosophila); and (2) to highlight some of the many remaining havens of ignorance concerning this topic. Our focus here has been largely restricted to Drosophila, as it is almost exclusively in fruit flies that the genetic and molecular bases of neuromuscular development have begun to be dissected. Despite the rapid and remarkable progress made recently using flies, many questions remain regarding insect neuromuscular development and functional maturation of synapses. The following list of unanswered questions is, of course, not exhaustive. Rather, it highlights areas that, in our opinion, represent the highest priorities for investigation. First, there is a general deficit in truly developmental studies. Drosophila mutant screens designed to study NMJ proteins usually try to isolate either conditional (i.e., temperature sensitive) mutants, or morphological mutants. Conditional mutant screens intentionally seek to avoid developmental artifacts. Morphological mutants infer developmental problems, but the progression of the phenotype (which is typically measured in third instar larvae just prior to pupation) during NMJ development is almost never explored. Thus, we are typically left knowing that something at some time went wrong, but we do not know exactly what or when or where. This is inadequate. Drosophila neurobiologists have been baffled by the difficulty of perturbing motor neuronmuscle connectivity with traditional loss-of-function mutants. Mutations in most genes with accepted roles in axon outgrowth and target recognition typically display relatively low penetrance (e.g., the phenotypes are spotty). Low penetrance is typically blamed on redundancy, where each pair of pre- and postsynaptic targets recognizes each other using multiple cues. Redundancy seems to be required by the fact that motor neuron growth cones follow

Many hormones have dramatic effects on behavior and several studies have demonstrated shortterm modulation of NMJ function in larvae. This modulation can result in increases or decreases in NMJ transmission. For example, the mysosuppressing FLRFamide has been identified in the nervous systems of many insect species, including moths, cockroaches, locusts, flesh flies, and fruit flies (Nachman et al., 1996). Myosuppressins cause potent inhibition of muscle contraction. Direct application of FLRF or analogs to the mealworm NMJ results in reduced synaptic transmission, suggesting that one of the normal targets of myosuppressins is the NMJ (Nachman et al., 1996). FLRFamide analogs also affect other insect synapses (Nachman et al., 1996). FMRFamides are another class of neuropeptides found in many insect species.

122 Functional Development of the Neuromusculature

amazingly invariant routes, after which they unerringly innervate a specific section of a particular target and form a distinctly shaped presynaptic arbor. Wild-type axons are able to innervate their targets with almost perfect reliability, even despite developmental delays. Without redundancy, it is hard to believe that reliable innervation results from a long series of chancy guidance and target selection choices involving transiently expressed cues. For example, even if the error rate at any particular decision point were only 0.01, almost one quarter of the NMJs would show an aberration after only two dozen decisions. But most developmental biologists also accept the theory that connectivity is determined through the use of combinatorial codes, with growth cones and their postsynaptic targets each displaying a unique set of molecules that allow them to recognize each other. The presence of unique combinatorial codes and redundancy do not seem compatible, however. The already staggering number of unique codes required for connectivity in the nervous system becomes even more unmanageable if redundancy is involved. Accordingly, we believe that developmental biologists are still missing something fundamental in our concepts of pathfinding and target selection. Following pathfinding and target recognition, we still know almost nothing about the signals and mechanisms involved in the conversion of a growth cone into a synapse. Although a few candidate molecules have been implicated in this conversion process (e.g., the neural tetraspanin Late Bloomer (Lbm; Kopczynski et al., 1996), we have no clear models on how such molecules function. The few early reports (see, for example Fradkin, 2002) have mostly not been followed up. Thus, the interface between neuromuscular target recognition and the initiation of synaptogenesis seems unduly neglected, despite the detailed characterization of the process at the level of cell morphology (Yoshihara et al., 1997). What role does the postsynaptic cell play in signaling this transition? What is the nature of the intercellular signals? How do such signals regulate the synthesis and localization of the diverse new proteins needed for synaptic function? Once preand postsynaptic cells recognize their partner as appropriate, exactly how do they initially build a synapse? Studies on mammalian neurons indicate that active zones are preassembled in large PiccoloBassoon transport vesicles (PTVs; Piccolo and Bassoon are protein components of these vesicles with undefined function) (Shapira et al., 2003). Active zones are inserted preformed when one of these vesicles fuses with the cytoplasmic membrane. Since proper orientation of active zones requires a

properly differentiated postsynaptic partner, these PTVs must somehow be targeted to the appropriate location using transsynaptic cues from the postsynaptic cell, or the active zones must quickly migrate to the correct location after random insertion, and then anchored opposite the postsynaptic cell. Clearly homologous genes encoding the PTV proteins Piccolo and Bassoon have not been identified in the Drosophila genome, however, and PTVs have not been shown to exist in insects. On the postsynaptic side of the NMJ, there is an extremely short list of molecular players. Both biochemical and genetic approaches need to be employed to gain perspective on the number of molecules composing the postsynaptic density (PSD). One obvious priority is the identification of the junctional and extrajunctional GluR channel subunits. Two decades ago, the study of glutamate receptors was essentially synonymous with the study of insect (or at least arthropod) glutamate receptors. Since then, however, it is embarrassing how little progress has been made with regard to insect GluRs in the face of the great strides accomplished by mammalian glutamatergic synapse aficionados. Without knowing the insect receptor subunits, we cannot determine how each is synthesized, how they assemble, and how the receptors are localized. We also need the complete list of subunits in order to ask whether there is a developmental shift in receptor composition, comparable to the situation with AChRs at the vertebrate NMJ. We know even less about the classes, localization, or function of other types of receptors, such as metabotropic GluRs, peptide and nonpeptide receptors, which control a range of downstream signaling cascades. There are also large gaps in our lists of presynaptic proteins. For example, no molecules have yet been identified as restricted components of the active zone. The molecular nature of the T-bar remains unknown. What is the function of the T-bar? Is it required for targeted SV exocytosis or any other aspect of presynaptic function? Similarly, are there distinctive active zones of other presynaptic functions (e.g., SV endocytosis)? Multiple Ca2 and K channels play crucial roles in the development of efficacy and plasticity at the NMJ on both sides of the cleft. How do these channels get specified and targeted to the correct membrane domain? Do PDZ domain proteins adequately explain how channels become correctly associated with other membrane and membrane-associated proteins to generate localized assemblies required for efficient synaptic transmission? If so, what is the full complement of PDZ domain proteins in

p0640

Functional Development of the Neuromusculature

123

p0650

the pre- and postsynaptic cells, and how do they interact to compartmentalize functional domains? Although the presynaptic terminal has been defined in terms of active and periactive zones, the true functional nature of this division remains unknown. Is the periactive zone a functional compartment or is it merely the zone defined as being outside of the active zones? The questions regarding the nature of such functional domains and the ways in which they are established during synaptogenesis are particularly important. In Drosophila there is now plenty of evidence for the existence of multiple transsynaptic signals traveling in both anterograde and retrograde directions. In the embryo, for example, an anterograde signal establishes the GluR field by inducing both GluR clustering and de novo GluR synthesis (Broadie and Bate, 1993b), while a retrograde signal dictates the placement of presynaptic active zones opposite their muscle targets (Prokop et al., 1996). In the larva, modification of postsynaptic GluR function leads to a compensating change in presynaptic neurotransmitter release (Davis et al., 1998; DiAntonio et al., 1999; Haghighi et al., 2003). What are the signals and receptor mechanisms in these diverse signaling situations? Identification of the transsynaptic signals mediating synaptogenesis and matching presynaptic input to postsynaptic responsiveness is a major challenge. Is there a functional analog of Agrin at glutamatergic synapses? Is even the concept of a secreted signaling mechanism correct? Alternate transmembrane signaling mechanisms (e.g., Neuroligin-Neurexin signaling) have been compellingly proposed for the induction of presynaptic differentiation, but, so far, genetic support for such models has not been forthcoming and these mechanisms remain unexplored in insects. The primary focus of this chapter has been the maturation of the type I glutamatergic NMJ in Drosophila. Essentially nothing is known about the functional development of other classes of synapse. How, when, and why do the different types of NMJ bouton become morphologically, functionally, and molecularly distinct? Are different mechanisms involved in modulatory synapse formation, compared to fast excitatory (glutamatergic) synapses? In Drosophila we have a single candidate protein, dCAPS (Renden et al., 2001), which appears to be required for all peptidergic transmission. Are there other molecules specifically required for DCV formation and SV fusion? How are these molecules targeted to specific active zones? The investigation of modulatory synapse formation and function represents a major growth area.

In recent years, the first attempts have been made to examine CNS function and functional development in Drosophila. These advances are crucial to any true, integrated understanding of neural circuit formation and function utilizing the Drosophila genetic system. Very preliminary work to date has focused on three tasks. First, the functional properties of mature neurons and the nature of CNS synaptic communication have begun to be described (Baines and Bate, 1998; Baines et al., 1999; Rohrbough and Broadie, 2002; Baines, 2003; Rohrbough et al., 2003). A crucial need is to define circuits in which synaptic transmission can be driven, recorded, and modulated. Second, the wiring diagram of connectivity in the CNS has begun to be elucidated (Godenschwege et al., 2002; Lohr et al., 2002; Marin et al., 2002; Komiyama et al., 2003; Landgraf et al., 2003; Rohrbough et al., 2003; Scott et al., 2003). This work involves the definition of the functional compartments in the CNS, the mapping of the location and nature of the synaptic connections and, ultimately, the definition of neural circuits. The third, least advanced area is charting the normal development of function within CNS circuits (Baines and Bate, 1998; Allen et al., 1999; Baines et al., 1999). As with the NMJ, a firm grasp of the normal developmental processes is a basic prerequisite to any attempt to unravel mutant phenotypes. These early studies need to be supported and encouraged. The Drosophila NMJ is inherently important as a model for understanding the insect NMJ, as well as a more general model of the glutamatergic synapse. But with regard to understanding the development of general synaptic properties, the NMJ is but a foot in the door. It is time for Drosophila neurobiologists to grapple with the complexity of functional development and plasticity within the CNS.

Acknowledgments
The authors would like to thank Matthias Landgraf, Jeffrey Rohrbough, Kale Bodily, and Warren Davis for the images shown in Figure 2. We also appreciate the input from many other scientists who shared unpublished results, insights, and discussion.

References
Aidley, D.J., 1975. Excitation-contraction coupling and mechanical properties. In: Usherwood, P.N.R. (Ed.), Insect Muscle. Academic Press, New York, pp. 337356. Allen, M.J., Shan, X., Caruccio, P., Froggett, S.J., Moffat, K.G., et al., 1999. Targeted expression of truncated glued disrupts giant fiber synapse formation in Drosophila. J. Neurosci. 19, 93749384.

124 Functional Development of the Neuromusculature

Anderson, H., Edwards, J.S., Palka, J., 1980. Developmental neurobiology of invertebrates. Annu. Rev. Neurosci. 3, 97139. Aravamudan, B., Broadie, K., 2003. Synaptic Drosophila UNC-13 is regulated by antagonistic G-protein pathways via a proteasome-dependent degradation mechanism. J. Neurobiol. 54, 417438. Aravamudan, B., Fergestad, T., Davis, W.S., Rodesch, C.K., Broadie, K., 1999. Drosophila UNC-13 is essential for synaptic transmission. Nat. Neurosci. 2, 965971. Arbas, E.A., Tolbert, L.P., 1986. Presynaptic terminals persist following degeneration of flight muscle during development of a flightless grasshopper. J. Neurobiol. 17, 627636. Aronstein, K., Auld, V., french-Constant, R., 1996. Distribution of two GABA receptor-like subunits in the Drosophila CNS. Invert. Neurosci. 2, 115120. Artero, R., Prokop, A., Paricio, N., Begemann, G., Pueyo, I., et al., 1998. The muscleblind gene participates in the organization of Z-bands and epidermal attachments of Drosophila muscles and is regulated by Dmef2. Devel. Biol. 195, 131143. Ashcroft, F.M., 1982. Electrical properties of insect muscle. In: Podesta, R.B. (Ed.), Membrane Physiology of Invertebrates. Dekker, New York, pp. 567604. Attwell, D., Barbour, B., Szatkowski, M., 1993. Nonvesicular release of neurotransmitter. Neuron 11, 401407. Atwood, H.L., Govind, C.K., Wu, C.F., 1993. Differential ultrastructure of synaptic terminals on ventral longitudinal abdominal muscles in Drosophila larvae. J. Neurobiol. 24, 10081024. Auld, V.J., Fetter, R.D., Broadie, K., Goodman, C.S., 1995. Gliotactin, a novel transmembrane protein on peripheral glia, is required to form the blood-nerve barrier in Drosophila. Cell 81, 757767. Baines, R.A., 2003. Postsynaptic protein kinase A reduces neuronal excitability in response to increased synaptic excitation in the Drosophila CNS. J. Neurosci. 23, 86648672. Baines, R.A., Bate, M., 1998. Electrophysiological development of central neurons in the Drosophila embryo. J. Neurosci. 18, 46734683. Baines, R.A., Robinson, S.G., Fujioka, M., Jaynes, J.B., Bate, M., 1999. Postsynaptic expression of tetanus toxin light chain blocks synaptogenesis in Drosophila. Curr. Biol. 9, 12671270. Baines, R.A., Seugnet, L., Thompson, A., Salvaterra, P.M., Bate, M., 2002. Regulation of synaptic connectivity: levels of Fasciclin II influence synaptic growth in the Drosophila CNS. J. Neurosci. 22, 65876595. Baines, R.A., Uhler, J.P., Thompson, A., Sweeney, S.T., Bate, M., 2001. Altered electrical properties in Drosophila neurons developing without synaptic transmission. J. Neurosci. 21, 15231531. Baker, D.A., Xi, Z.X., Shen, H., Swanson, C.J., Kalivas, P.W., 2002. The origin and neuronal function of in vivo nonsynaptic glutamate. J. Neurosci. 22, 91349141.

Baker, K., Salkoff, L., 1990. The Drosophila Shaker gene codes for a distinctive K current in a subset of neurons. Neuron 4, 129140. Ball, E.E., Goodman, C.S., 1985a. Muscle development in the grasshopper embryo. II. Syncytial origin of the extensor tibiae muscle pioneers. Devel. Biol. 111, 399416. Ball, E.E., Goodman, C.S., 1985b. Muscle development in the grasshopper embryo. III. Sequential origin of the flexor tibiae muscle pioneers. Devel. Biol. 111, 417424. Ball, E.E., Ho, R.K., Goodman, C.S., 1985. Muscle development in the grasshopper embryo. I. Muscles, nerves, and apodemes in the metathoracic leg. Devel. Biol. 111, 383398. Bastiani, M.J., du Lac, S., Goodman, C.S., 1986. Guidance of neuronal growth cones in the grasshopper embryo. I. Recognition of a specific axonal pathway by the pCC neuron. J. Neurosci. 6, 35183531. Bastiani, M.J., Harrelson, A.L., Snow, P.M., Goodman, C.S., 1987. Expression of fasciclin I and II glycoproteins on subsets of axon pathways during neuronal development in the grasshopper. Cell 48, 745755. Bate, C.M., 1976. Pioneer neurones in an insect embryo. Nature 260, 5456. Bate, M., 1990. The embryonic development of larval muscles in Drosophila. Development 110, 791804. Bate, M., Landgraf, M., Ruiz Gomez Bate, M., 1999. Development of larval body wall muscles. Int. Rev. Neurobiol. 43, 2544. Bate, M., Rushton, E., Currie, D.A., 1991. Cells with persistent twist expression are the embryonic precursors of adult muscles in Drosophila. Development 113, 7989. Bate, M, Arias, A.M. (Eds.) 1993. The Development of Drosophila melanogaster. Cold Spring Harbor Laboratory Press, New York. Baumgartner, S., Littleton, J.T., Broadie, K., Bhat, M.A., Harbecke, R., et al., 1996. A Drosophila neurexin is required for septate junction and bloodnerve barrier formation and function. Cell 87, 10591068. Baylies, M.K., Bate, M., 1996. twist: a myogenic switch in Drosophila. Science 272, 14491450. Baylies, M.K., Martinez Arias, A., Bate, M., 1995. wingless is required for the formation of a subset of muscle founder cells during Drosophila embryogenesis. Development 121, 38293837. Baylies, M.K., Michelson, A.M., 2001. Invertebrate myogenesis: looking back to the future of muscle development. Curr. Opin. Genet. Devel. 11, 431439. Beadle, D.J., Hicks, D., 1985. Insect nerve culture. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Nervous System: Structure and Function. pp. 181212. Beumer, K., Matthies, H.J., Bradshaw, A., Broadie, K., 2002. Integrins regulate DLG/FAS2 via a CaM kinase II-dependent pathway to mediate synapse elaboration and stabilization during postembryonic development. Development 129, 33813391. Beumer, K.J., Rohrbough, J., Prokop, A., Broadie, K., 1999. A role for PS integrins in morphological growth

bib0150

Functional Development of the Neuromusculature

125

and synaptic function at the postembryonic neuromuscular junction of Drosophila. Development 126, 58335846. Bieber, A.J., Snow, P.M., Hortsch, M., Patel, N.H., Jacobs, J.R., et al., 1989. Drosophila neuroglian: a member of the immunoglobulin superfamily with extensive homology to the vertebrate neural adhesion molecule L1. Cell 59, 447460. Bilder, D., 2001. PDZ proteins and polarity: functions from the fly. Trends Genet. 17, 511519. Bilder, D., Schober, M., Perrimon, N., 2003. Integrated activity of PDZ protein complexes regulates epithelial polarity. Nat. Cell Biol. 5, 5358. Blagburn, J.M., Sosa, M.A., Blanco, R.E., 1996. Specificity of identified central synapses in the embryonic cockroach: appropriate connections form before the onset of spontaneous afferent activity. J. Comp. Neurol. 373, 511528. Bossing, T., Udolph, G., Doe, C.Q., Technau, G.M., 1996. The embryonic central nervous system lineages of Drosophila melanogaster. I. Neuroblast lineages derived from the ventral half of the neuroectoderm. Devel. Biol. 179, 4164. Broadie, K.S., 1996. Regulation of the synaptic vesicle cycle in Drosophila. Biochem. Soc. Trans. 24, 639645. Broadie, K.S., 1999. Development of electrical properties and synaptic transmission at the embryonic neuromuscular junction. Int. Rev. Neurobiol. 43, 4567. Broadie, K.S., 2000. Electrophysiological approaches to the neuromusculature. In: Sullivan, W., Ashburner, M., Hawley, R.S. (Eds.), Drosophila Protocols. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 273295. Broadie, K., Bate, M., 1991. Development of adult muscles in Drosophila: ablation of identified muscle precursor cells. Development 113, 103118. Broadie, K., Bate, M., 1993a. Activity-dependent development of the neuromuscular synapse during Drosophila embryogenesis. Neuron 11, 607619. Broadie, K., Bate, M., 1993b. Innervation directs receptor synthesis and localization in Drosophila embryo synaptogenesis. Nature 361, 350353. Broadie, K., Bate, M., 1993c. Muscle development is independent of innervation during Drosophila embryogenesis. Development 119, 533543. Broadie, K.S., Bate, M., 1993d. Development of larval muscle properties in the embryonic myotubes of Drosophila melanogaster. J. Neurosci. 13, 167180. Broadie, K.S., Bate, M., 1993e. Development of the embryonic neuromuscular synapse of Drosophila melanogaster. J. Neurosci. 13, 144166. Broadie, K., Bellen, H.J., DiAntonio, A., Littleton, J.T., Schwarz, T.L., 1994. Absence of synaptotagmin disrupts excitation-secretion coupling during synaptic transmission. Proc. Natl Acad. Sci. USA 91, 1072710731. Broadie, K., Prokop, A., Bellen, H.J., OKane, C.J., Schulze, K.L., et al., 1995. Syntaxin and synaptobrevin function downstream of vesicle docking in Drosophila. Neuron 15, 663673.

Broadie, K.S., Richmond, J.E., 2002. Establishing and sculpting the synapse in Drosophila and C. elegans. Curr. Opin. Neurobiol. 12, 491498. Broadie, K., Rushton, E., Skoulakis, E.M., Davis, R.L., 1997. Leonardo, a Drosophila 14-3-3 protein involved in learning, regulates presynaptic function. Neuron 19, 391402. Broadie, K., Sink, H., Van Vactor, D., Fambrough, D., Whitington, P.M., et al., 1993. From growth cone to synapse: the life history of the RP3 motor neuron. Devel. Suppl. 227238. Budnik, V., 1996. Synapse maturation and structural plasticity at Drosophila neuromuscular junctions. Curr. Opin. Neurobiol. 6, 858867. Budnik, V., Koh, Y.H., Guan, B., Hartmann, B., Hough, C., et al., 1996. Regulation of synapse structure and function by the Drosophila tumor suppressor gene dlg. Neuron 17, 627640. Budnik, V., Zhong, Y., Wu, C.F., 1990. Morphological plasticity of motor axons in Drosophila mutants with altered excitability. J. Neurosci. 10, 37543768. Callec, J.J., 1985. Synaptic transmission in the central nervous system. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Nervous System: Structure and Function. Pergamon, New York, pp. 139180. Campos-Ortega, J.A., Hartenstein, V., 1997. The Embryonic Development of Drosophila melanogaster. Springer, Berlin. Carmena, A., Bate, M., Jimenez, F., 1995. Lethal of scute, a proneural gene, participates in the specification of muscle progenitors during Drosophila embryogenesis. Genes Devel. 9, 23732383. Carmena, A., Gisselbrecht, S., Harrison, J., Jimenez, F., Michelson, A.M., 1998. Combinatorial signaling codes for the progressive determination of cell fates in the Drosophila embryonic mesoderm. Genes Devel. 12, 39103922. Casal, J., Leptin, M., 1996. Identification of novel genes in Drosophila reveals the complex regulation of early gene activity in the mesoderm. Proc. Natl Acad. Sci. USA 93, 1032710332. Cash, S., Chiba, A., Keshishian, H., 1992. Alternate neuromuscular target selection following the loss of single muscle fibers in Drosophila. J. Neurosci. 12, 20512064. Castillo, J., Hoyle, G., Machne, X., 1953. Neuromuscular transmission in a locust. J. Physiol. 121, 539547. Cattaert, D., Birman, S., 2001. Blockade of the central generator of locomotor rhythm by noncompetitive NMDA receptor antagonists in Drosophila larvae. J. Neurobiol. 48, 5873. Chang, H., Ciani, S., Kidokoro, Y., 1994. Ion permeation properties of the glutamate receptor channel in cultured embryonic Drosophila myotubes. J. Physiol. 476, 116. Chang, H., Kidokoro, Y., 1996. Kinetic properties of glutamate receptor channels in cultured embryonic Drosophila myotubes. Jpn. J. Physiol. 46, 249264. Cheung, U.S., Shayan, A.J., Boulianne, G.L., Atwood, H.L., 1999. Drosophila larval neuromuscular

126 Functional Development of the Neuromusculature

junctions responses to reduction of cAMP in the nervous system. J. Neurobiol. 40, 113. Chiba, A., Hing, H., Cash, S., Keshishian, H., 1993. Growth cone choices of Drosophila motoneurons in response to muscle fiber mismatch. J. Neurosci. 13, 714732. Chiba, A., Rose, D., 1998. Painting the target: how local molecular cues define synaptic relationships. Bioessays 20, 941948. Chiba, A., Snow, P., Keshishian, H., Hotta, Y., 1995. Fasciclin III as a synaptic target recognition molecule in Drosophila. Nature 374, 166168. Chopra, M., Gu, G.G., Singh, S., 2000. Mutations affecting the delayed rectifier potassium current in Drosophila. J. Neurogenet. 14, 107123. Ciani, S., Nishikawa, K., Kidokoro, Y., 1997. Permeation of organic cations and ammonium through the glutamate receptor channel in Drosophila larval muscle. Jpn. J. Physiol. 47, 189198. Cochilla, A.J., Angleson, J.K., Betz, W.J., 1999. Monitoring secretory membrane with FM1-43 fluorescence. Annu. Rev. Neurosci. 22, 110. Colledge, M., Froehner, S.C., 1998. Signals mediating ion channel clustering at the neuromuscular junction. Curr. Opin. Neurobiol. 8, 357363. Cui, X., Doe, C.Q., 1992. ming is expressed in neuroblast sublineages and regulates gene expression in the Drosophila central nervous system. Development 116, 943952. Cull-Candy, S.G., 1976. Two types of extrajunctional l-glutamate receptors in locust muscle fibres. J. Physiol. 255, 449464. Cully, D.F., Paress, P.S., Liu, K.K., Schaeffer, J.M., Arena, J.P., 1996. Identification of a Drosophila melanogaster glutamate-gated chloride channel sensitive to the antiparasitic agent avermectin. J. Biol. Chem. 271, 2018720191. Currie, D.A., Truman, J.W., Burden, S.J., 1995. Drosophila glutamate receptor RNA expression in embryonic and larval muscle fibers. Devel. Dyn. 203, 311316. Davis, G.W., DiAntonio, A., et al., 1998. Postsynaptic PKA controls quantal size and reveals a retrograde signal that regulates presynaptic transmitter release in Drosophila. Neuron 20(2), 305315. Davis, G.W., Goodman, C.S., 1998. Synapse-specific control of synaptic efficacy at the terminals of a single neuron. Nature 392, 8286. Davis, G.W., Schuster, C.M., Goodman, C.S., 1996a. Genetic dissection of structural and functional components of synaptic plasticity. III. CREB is necessary for presynaptic functional plasticity. Neuron 17, 669679. Davis, G.W., Schuster, C.M., Goodman, C.S., 1996b. Genetic dissection of structural and functional components of synaptic plasticity. III. CREB is necessary for presynaptic functional plasticity [see comments]. Neuron 17, 669679. Davis, G.W., Schuster, C.M., Goodman, C.S., 1997. Genetic analysis of the mechanisms controlling

target selection: target-derived Fasciclin II regulates the pattern of synapse formation. Neuron 19, 561573. Demarque, M., Represa, A., Becq, H., Khalilov, I., Ben-Ari, Y., et al., 2002. Paracrine intercellular communication by a Ca2- and SNARE-independent release of GABA and glutamate prior to synapse formation. Neuron 36, 10511061. DiAntonio, A., Patersen, S.A., et al., 1999. Glutamate receptor expression regulates quantal size and quantal content at the Drosophila neuromuscular junction. J. Neurosci. 19(8), 30233032. DiAntonio, A., Schwarz, T.L., 1994. The effect on synaptic physiology of synaptotagmin mutations in Drosophila. Neuron 12, 909920. Diefenbach, T.J., Guthrie, P.B., Stier, H., Billips, B., Kater, S.B., 1999. Membrane recycling in the neuronal growth cone revealed by FM1-43 labeling. J. Neurosci. 19(21), 94369444. Doberstein, S.K., Fetter, R.D., Mehta, A.Y., Goodman, C.S., 1997. Genetic analysis of myoblast fusion: blown fuse is required for progression beyond the prefusion complex. J. Cell Biol. 136, 12491261. Doe, C.Q., Bastiani, M.J., Goodman, C.S., 1986. Guidance of neuronal growth cones in the grasshopper embryo. IV. Temporal delay experiments. J. Neurosci. 6, 35523563. Doe, C.Q., Goodman, C.S., 1985. Early events in insect neurogenesis. II. The role of cell interactions and cell lineage in the determination of neuronal precursor cells. Devel. Biol. 111, 206219. Duan, S., Anderson, C.M., Keung, E.C., Chen, Y., Swanson, R.A., 2003. P2X7 receptor-mediated release of excitatory amino acids from astrocytes. J. Neurosci. 23, 13201328. Eaton, B.A., Fetter, R.D., Davis, G.W., 2002. Dynactin is necessary for synapse stabilization. Neuron 34, 729741. Elkins, T., Ganetzky, B., 1990. Conduction in the giant nerve fiber pathway in temperature-sensitive paralytic mutants of Drosophila. J. Neurogenet. 6, 207219. Elkins, T., Ganetzky, B., Wu, C.F., 1986. A Drosophila mutation that eliminates a calcium-dependent potassium current. Proc. Natl Acad. Sci. USA 83, 84158419. Fambrough, D., Goodman, C.S., 1996. The Drosophila beaten path gene encodes a novel secreted protein that regulates defasciculation at motor axon choice points. Cell 87, 10491058. Featherstone, D.E., Broadie, K., 2000. Surprises from Drosophila: genetic mechanisms of synaptic development and plasticity. Brain Res. Bull. 53, 501511. Featherstone, D.E., Davis, W.S., Dubreuil, R.R., Broadie, K., 2001. Drosophila alpha- and beta-spectrin mutations disrupt presynaptic neurotransmitter release. J. Neurosci. 21, 42154224. Featherstone, D.E., Rushton, E., Broadie, K., 2002. Developmental regulation of glutamate receptor field size by nonvesicular glutamate release. Nat. Neurosci. 5, 141146.

bib9000

Functional Development of the Neuromusculature

127

Featherstone, D.E., Rushton, E.M., Hilderbrand-Chae, M., Phillips, A.M., Jackson, F.R., et al., 2000. Presynaptic glutamic acid decarboxylase is required for induction of the postsynaptic receptor field at a glutamatergic synapse. Neuron 27, 7184. Fehon, R.G., Dawson, I.A., Artavanis-Tsakonas, S., 1994. A Drosophila homologue of membrane-skeleton protein 4.1 is associated with septate junctions and is encoded by the coracle gene. Development 120, 545557. Fergestad, T., Broadie, K., 2001. Interaction of stoned and synaptotagmin in synaptic vesicle endocytosis. J. Neurosci. 21, 12181227. Fergestad, T., Davis, W.S., Broadie, K., 1999. The stoned proteins regulate synaptic vesicle recycling in the presynaptic terminal. J. Neurosci. 19, 58475860. Fergestad, T., Wu, M.N., Schulze, K.L., Lloyd, T.E., Bellen, H.J., et al., 2001. Targeted mutations in the syntaxis H3 domain specifically disrupt SNARE complex function is synaptic transmission. J. Neurosci. 21(23), 91429150. Fletcher, T.L., De Camilli, P., Banker, G., 1994. Synaptogenesis in hippocampal cultures: evidence indicating that axons and dendrites become competent to form synapses at different stages of neuronal development. J. Neurosci. 14, 66956706. Fradkin, L.G., Kamphorst, J.T., et al., 2002. Genomewide analysis of the Drosophila tetraspanins reveals a subset with similar function in the formation of theembryonic sysnapse. Proc. Natl Acad. Sci. USA 99(21), 1366313668. Frasch, M., Leptin, M., 2000. Mergers and acquisitions: unequal partnerships in Drosophila myoblast fusion. Cell 102, 127129. Freeman, M.R., Delrow, J., Kim, J., Johnson, E., Doe, C.Q., 2003. Unwrapping glial biology: Gcm target genes regulating glial development, diversification, and function. Neuron 38, 567580. Fritschy, J.M., Brunig, I., 2003. Formation and plasticity of GABAergic synapses: physiological mechanisms and pathophysiological implications. Pharmacol. Ther. 98, 299323. Furrer, M.P., Kim, S., Wolf, B., Chiba, A., 2003. Robo and Frazzled/DCC mediate dendritic guidance at the CNS midline. Nat. Neurosci. 6, 223230. Furst, A., Mahowald, A.P., 1985. Differentiation of primary embryonic neuroblasts in purified neural cell cultures from Drosophila. Devel. Biol. 109, 184192. Ganetzky, B., Robertson, G.A., Wilson, G.F., Trudeau, M.C., Titus, S.A., 1999. The eag family of K channels in Drosophila and mammals. Ann. N. Y. Acad. Sci. 868, 356369. Ganetzky, B., Wu, C.F., 1983. Neurogenetic analysis of potassium currents in Drosophila: synergistic effects on neuromuscular transmission in double mutants. J. Neurogenet. 1, 1728. Gaul, U., Seifert, E., Schuh, R., Jackle, H., 1987. Analysis of Kruppel protein distribution during early Drosophila development reveals posttranscriptional regulation. Cell 50, 539547.

Gielow, M.L., Gu, G.G., Singh, S., 1995. Resolution and pharmacological analysis of the voltage-dependent calcium channels of Drosophila larval muscles. J. Neurosci. 15, 60856093. Godenschwege, T.A., Simpson, J.H., Shan, X., Bashaw, G.J., Goodman, C.S., et al., 2002. Ectopic expression in the giant fiber system of Drosophila reveals distinct roles for roundabout (Robo), Robo2, and Robo3 in dendritic guidance and synaptic connectivity. J. Neurosci. 22, 31173129. Gonzalez-Gaitan, M., Jackle, H., 1997. Role of Drosophila alpha-adaptin in presynaptic vesicle recycling. Cell 88, 767776. Goodman, C.S., Spitzer, N.C., 1979. Embryonic development of identified neurones: differentiation from neuroblast to neurone. Nature 280, 208214. Gorczyca, M., Augart, C., Budnik, V., 1993. Insulin-like receptor and insulin-like peptide are localized at neuromuscular junctions in Drosophila. J. Neurosci. 13, 36923704. Gorczyca, M.G., Wu, C.F., 1991. Single-channel K currents in Drosophila muscle and their pharmacological block. J. Membr. Biol. 121, 237248. Griffith, L.C., 1997. Drosophila melanogaster as a model system for the study of the function of calcium/calmodulin-dependent protein kinase II in synaptic plasticity. Invert. Neurosci. 3, 93102. Guan, B., Hartmann, B., Kho, Y.H., Gorczyca, M., Budnik, V., 1996. The Drosophila tumor suppressor gene, dlg, is involved in structural plasticity at a glutamatergic synapse. Curr. Biol. 6, 695706. Haghighi, A.P., McCabe, B.D., Fetter, R.D., Palmer, J.E., Hom, S., et al., 2003. Retrograde control of synaptic transmission by postsynaptic CaMKII at the Drosophila neuromuscular junction. Neuron 39, 255267. Halpern, M.E., Chiba, A., Johansen, J., Keshishian, H., 1991. Growth cone behavior underlying the development of stereotypic synaptic connections in Drosophila embryos. J. Neurosci. 11, 32273238. Hannan, F., Zhong, Y., 1999. Second messenger systems underlying plasticity at the neuromuscular junction. Int. Rev. Neurobiol. 43, 119138. Harrison, S.D., Broadie, K., van de Goor, J., Rubin, G.M., 1994. Mutations in the Drosophila Rop gene suggest a function in general secretion and synaptic transmission. Neuron 13, 555566. Heckmann, M., Dudel, J., 1997. Desensitization and resensitization kinetics of glutamate receptor channels from Drosophila larval muscle. Biophys. J. 72, 21602169. Hewes, R.S., Snowdeal, E.C., III, Saitoe, M., Taghert, P.H., 1998. Functional redundancy of FMRFamiderelated peptides at the Drosophila larval neuromuscular junction. J. Neurosci. 18, 71387151. Hidalgo, A., 2003. Neuron-glia interactions during axon guidance in Drosophila. Biochem. Soc. Trans. 31, 5055. Ho, R., Ball, E.E., Goodman, C.S., 1983. Muscle pioneers: large mesodermal cells that erect a scaffold for developing muscles and motoneurones in grasshopper embryos. Nature 301, 6669.

128 Functional Development of the Neuromusculature

Hoang, B., Chiba, A., 2001. Single-cell analysis of Drosophila larval neuromuscular synapses. Devel. Biol. 229, 5570. Hortsch, M., Goodman, C.S., 1991. Cell and substrate adhesion molecules in Drosophila. Annu. Rev. Cell Biol. 7, 505557. Huff, R., Furst, A., Mahowald, A.P., 1989. Drosophila embryonic neuroblasts in culture: autonomous differentiation of specific neurotransmitters. Devel. Biol. 134, 146157. Hummel, T., Krukkert, K., Roos, J., Davis, G., Klambt, C., 2000. Drosophila Futsch/22C10 is a MAP1B-like protein required for dendritic and axonal development. Neuron 26, 357370. Isshiki, T., Pearson, B., Holbrook, S., Doe, C.Q., 2001. Drosophila neuroblasts sequentially express transcription factors which specify the temporal identity of their neuronal progeny. Cell 106, 511521. Jacobs, J.R., Goodman, C.S., 1989. Embryonic development of axon pathways in the Drosophila CNS. I. A glial scaffold appears before the first growth cones. J. Neurosci. 9, 24022411. Jan, L.Y., Jan, Y.N., 1976a. l-Glutamate as an excitatory transmitter at the Drosophila larval neuromuscular junction. J. Physiol. (Lond.) 262, 215236. Jan, L.Y., Jan, Y.N., 1976b. Properties of the larval neuromuscular junction in Drosophila melanogaster. J. Physiol. (Lond.) 262, 189214. Jia, X.X., Gorczyca, M., Budnik, V., 1993. Ultrastructure of neuromuscular junctions in Drosophila: comparison of wild type and mutants with increased excitability. J. Neurobiol. 24, 10251044. Jonas, P.E., Phannavong, B., Schuster, R., Schroder, C., Gundelfinger, E.D., 1994. Expression of the ligandbinding nicotinic acetylcholine receptor subunit D alpha 2 in the Drosophila central nervous system. J. Neurobiol. 25, 14941508. Jones, B.W., Fetter, R.D., Tear, G., Goodman, C.S., 1995. Glial cells missing: a genetic switch that controls glial versus neuronal fate. Cell 82, 10131023. Kamb, A., Iverson, L.E., Tanouye, M.A., 1987. Molecular characterization of Shaker, a Drosophila gene that encodes a potassium channel. Cell 50, 405413. Kamb, A., Tseng-Crank, J., Tanouye, M.A., 1988. Multiple products of the Drosophila Shaker gene may contribute to potassium channel diversity. Neuron 1, 421430. Kamdar, K.P., Shelton, M.E., Finnerty, V., 1994. The Drosophila molybdenum cofactor gene cinnamon is homologous to three Escherichia coli cofactor proteins and to the rat protein gephyrin. Genetics 137, 791801. Kardos, J., 1999. Recent advances in GABA research. Neurochem. Int. 34, 353358. Kazama, H., Morimoto-Tanifuji, T., Nose, A., 2003. Postsynaptic activation of calcium/calmodulin-dependent protein kinase II promotes coordinated pre- and postsynaptic maturation of Drosophila neuromuscular junctions. Neuroscience 117, 615625. Keshishian, H., Broadie, K., Chiba, A., Bate, M., 1996. The drosophila neuromuscular junction: a model

system for studying synaptic development and function. Annu. Rev. Neurosci. 19, 545575. Keyser, M.R., Anson, B.D., Titus, S.A., Ganetzky, B., Witten, J.L., 2003. Molecular characterization, functional expression, and developmental profile of an ether a-go-go K channel in the tobacco hornworm Manduca sexta. J. Neurobiol. 55, 7385. Klambt, C., Jacobs, J.R., 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, 801815. Koenig, J.H., Ikeda, K., 1989. Disappearance and reformation of synaptic vesicle membrane upon transmitter release observed under reversible blockage of membrane retrieval. J. Neurosci. 9, 38443860. Koh, Y.H., Popova, E., Thomas, U., Griffith, L.C., Budnik, V., 1999. Regulation of DLG localization at synapses by CaMKII-dependent phosphorylation. Cell 98, 353363. Kolodkin, A.L., Matthes, D.J., Goodman, C.S., 1993. The semaphorin genes encode a family of transmembrane and secreted growth cone guidance molecules. Cell 75, 13891399. Kolodkin, A.L., Matthes, D.J., OConnor, T.P., Patel, N.H., Admon, A., et al., 1992. Fasciclin IV: sequence, expression, and function during growth cone guidance in the grasshopper embryo. Neuron 9, 831845. Komiyama, T., Johnson, W.A., Luo, L., Jefferis, G.S., 2003. From lineage to wiring specificity. POU domain transcription factors control precise connections of Drosophila olfactory projection neurons. Cell 112, 157167. Kopczynski, C.C., Davis, G.W., et al., 1996. A neural tetraspanin, encoded by late bloomer, that facilitates synapse formation. Science 271(5257), 18671870. Kulkarni, S.J., Hall, J.C., 1987. Behavioral and cytogenetic analysis of the cacophony courtship song mutant and interacting genetic variants in Drosophila melanogaster. Genetics 115, 461475. Kurdyak, P., Atwood, H.L., Stewart, B.A., Wu, C.F., 1994. Differential physiology and morphology of motor axons to ventral longitudinal muscles in larval Drosophila. J. Comp. Neurol. 350, 463472. Kuromi, H., Kidokoro, Y., 1998. Two distinct pools of synaptic vesicles in single presynaptic boutons in a temperature-sensitive Drosophila mutant, shibire. Neuron 20, 917925. Kuromi, H., Kidokoro, Y., 1999. The optically determined size of exo/endo cycling vesicle pool correlates with the quantal content at the neuromuscular junction of Drosophila larvae. J. Neurosci. 19, 15571565. Kuromi, H., Kidokoro, Y., 2000. Tetanic stimulation recruits vesicles from reserve pool via a cAMP-mediated process in Drosophila synapses. Neuron 27, 133143. Kuromi, H., Kidokoro, Y., 2002. Selective replenishment of two vesicle pools depends on the source of Ca2 at the Drosophila synapse. Neuron 35, 333343. Kuwada, J.Y., Goodman, C.S., 1985. Neuronal determination during embryonic development of the grasshopper nervous system. Devel. Biol. 110, 114126.

Functional Development of the Neuromusculature

129

Lahey, T., Gorczyca, M., Jia, X.X., Budnik, V., 1994. The Drosophila tumor suppressor gene dlg is required for normal synaptic bouton structure. Neuron 13, 823835. Landgraf, M., Baylies, M., Bate, M., 1999a. Muscle founder cells regulate defasciculation and targeting of motor axons in the Drosophila embryo. Curr. Biol. 9, 589592. Landgraf, M., Bossing, T., Technau, G.M., Bate, M., 1997. The origin, location, and projections of the embryonic abdominal motorneurons of Drosophila. J. Neurosci. 17, 96429655. Landgraf, M., Roy, S., Prokop, A., VijayRaghavan, K., Bate, M., 1999b. even-skipped determines the dorsal growth of motor axons in Drosophila. Neuron 22, 4352. Landgraf, M., Sanchez-Soriano, N., Technau, G.M., Urban, J., Prokop, A., 2003. Charting the Drosophila neuropile: a strategy for the standardised characterisation of genetically amenable neurites. Devel. Biol. 260, 207225. Lane, N.J., 1985. Structure and components of the nervous system. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Nervous System: Structure and Motor Function. Pergamon, Oxford, pp. 149. Lee, D., ODowd, D.K., 1999. Fast excitatory synaptic transmission mediated by nicotinic acetylcholine receptors in Drosophila neurons. J. Neurosci. 19, 53115321. Lee, D., ODowd, D.K., 2000a. cAMP-dependent plasticity at excitatory cholinergic synapses in Drosophila neurons: alterations in the memory mutant dunce. J. Neurosci. 20, 21042111. Lee, D., ODowd, D.K., 2000b. cAMP-dependent plasticity at excitatory cholinergic synapses in Drosophila neurons: alterations in the memory mutant dunce. J. Neurosci. 20, 21042111. Lee, D., Su, H., ODowd, D.K., 2003. GABA receptors containing Rdl subunits mediate fast inhibitory synaptic transmission in Drosophila neurons. J. Neurosci. 23, 46254634. Lee, T., Winter, C., Marticke, S.S., Lee, A., Luo, L., 2000. Essential roles of Drosophila RhoA in the regulation of neuroblast proliferation and dendritic but not axonal morphogenesis. Neuron 25, 307316. Leiserson, W.M., Harkins, E.W., Keshishian, H., 2000. Fray, a Drosophila serine/threonine kinase homologous to mammalian PASK, is required for axonal ensheathment. Neuron 28, 793806. Leptin, M., 1991. twist and snail as positive and negative regulators during Drosophila mesoderm development. Genes Dev. 5, 15681576. Levine, R.B., 1984. Changes in neuronal circuits during insect metamorphosis. J. Exp. Biol. 112, 2744. Levine, R.B., Weeks, J.C., 1996. Cell culture approaches to understanding the actions of steroid hormones on the insect nervous system. Devel. Neurosci. 18, 7386. Li, H., Cooper, R.L., 2001. Effects of the ecdysoneless mutant on synaptic efficacy and structure at the neuromuscular junction in Drosophila larvae during normal and prolonged development. Neuroscience 106, 193200.

Lissin, D.V., Carroll, R.C., Nicoll, R.A., Malenka, R.C., von Zastrow, M., 1999. Rapid, activation-induced redistribution of ionotropic glutamate receptors in cultured hippocampal neurons. [Published erratum appears in J. Neurosci. 1999; 19 (8), 3275.] J. Neurosci. 19, 12631272. Littleton, J.T., Bellen, H.J., Perin, M.S., 1993a. Expression of synaptotagmin in Drosophila reveals transport and localization of synaptic vesicles to the synapse. Development 118, 10771088. Littleton, J.T., Ganetzky, B., 2000. Ion channels and synaptic organization: analysis of the Drosophila genome. Neuron 26, 3543. Littleton, J.T., Serano, T.L., Rubin, G.M., Ganetzky, B., Chapman, E.R., 1999. Synaptic function modulated by changes in the ratio of synaptotagmin I and IV. Nature 400, 757760. Littleton, J.T., Stern, M., Schulze, K., Perin, M., Bellen, H.J., 1993b. Mutational analysis of Drosophila synaptotagmin demonstrates its essential role in Ca(2)-activated neurotransmitter release [see comments]. Cell 74, 11251134. Lloyd, T.E., Verstreken, P., Ostrin, E.J., Phillippi, A., Lichtarge, O., Bellen, H.J., 2000. A genome-wide search for synaptic vesicle cycle proteins in Drosophila. Neuron 26, 4550. Lnenicka, G.A., Keshishian, H., 2000. Identified motor terminals in Drosophila larvae show distinct differences in morphology and physiology. J. Neurobiol. 43, 186197. Lohr, R., Godenschwege, T., Buchner, E., Prokop, A., 2002. Compartmentalization of central neurons in Drosophila: a new strategy of mosaic analysis reveals localization of presynaptic sites to specific segments of neurites. J. Neurosci. 22, 1035710367. Loughney, K., Kreber, R., Ganetzky, B., 1989. Molecular analysis of the para locus, a sodium channel gene in Drosophila. Cell 58, 11431154. Luer, K., Technau, G.M., 1992. Primary culture of single ectodermal precursors of Drosophila reveals a dorsoventral prepattern of intrinsic neurogenic and epidermogenic capabilities at the early gastrula stage. Development 116, 377385. Luo, L., Liao, Y.J., Jan, L.Y., Jan, Y.N., 1994. Distinct morphogenetic functions of similar small GTPases: Drosophila Drac1 is involved in axonal outgrowth and myoblast fusion. Genes Devel. 8, 17871802. Madden, D.R., 2002. The structure and function of glutamate receptor ion channels. Nat. Rev. Neurosci. 3, 91101. Marin, E.C., Jefferis, G.S., Komiyama, T., Zhu, H., Luo, L., 2002. Representation of the glomerular olfactory map in the Drosophila brain. Cell 109, 243255. Mellerick, D.M., Kassis, J.A., Zhang, S.D., Odenwald, W.F., 1992. castor encodes a novel zinc finger protein required for the development of a subset of CNS neurons in Drosophila. Neuron 9, 789803. Michelson, A.M., Abmayr, S.M., Bate, M., Arias, A.M., Maniatis, T., 1990. Expression of a MyoD family

130 Functional Development of the Neuromusculature

member prefigures muscle pattern in Drosophila embryos. Genes Devel. 4, 20862097. Monastirioti, M., Gorczyca, M., Rapus, J., Eckert, M., White, K., et al., 1995. Octopamine immunoreactivity in the fruit fly Drosophila melanogaster. J. Comp. Neurol. 356, 275287. Nachman, R.J., Olender, E.H., Roberts, V.A., Holman, G.M., Yamamoto, D., 1996. A nonpeptidal peptidomimetic agonist of the insect FLRFamide myosuppressin family. Peptides 17, 313320. Nagaya, Y., Kutsukake, M., Chigusa, S.I., Komatsu, A., 2002. A trace amine, tyramine, functions as a neuromodulator in Drosophila melanogaster. Neurosci. Lett. 329, 324328. Nassel, D.R., Lundquist, T., Hoog, A., Grimelius, L., 1990. Substance P-like immunoreactive neurons in the nervous system of Drosophila. Brain Res. 507, 225233. Ng, J., Nardine, T., Harms, M., Tzu, J., Goldstein, A., et al., 2002. Rac GTPases control axon growth, guidance and branching. Nature 416, 442447. Nicholls, D.G., Sihra, T.S., 1986. Synaptosomes possess an exocytotic pool of glutamate. Nature 321, 772773. Nichols, R., McCormick, J.B., Lim, I.A., 1999. Structure, function, and expression of Drosophila melanogaster FMRFamide-related peptides. Ann. N.Y. Acad. Sci. 897, 264272. Nishikawa, K., Kidokoro, Y., 1995. Junctional and extrajunctional glutamate receptor channels in Drosophila embryos and larvae. J. Neurosci. 15, 79057915. Nishikawa, K., Kidokoro, Y., 1999. Octopamine inhibits synaptic transmission at the larval neuromuscular junction in Drosophila melanogaster. Brain Res. 837, 6774. Noordermeer, J.N., Kopczynski, C.C., Fetter, R.D., Bland, K.S., Chen, W.Y., et al., 1998. Wrapper, a novel member of the Ig superfamily, is expressed by midline glia and is required for them to ensheath commissural axons in Drosophila. Neuron 21, 9911001. Nose, A., Mahajan, V.B., Goodman, C.S., 1992. Connectin: a homophilic cell adhesion molecule expressed on a subset of muscles and the motoneurons that innervate them in Drosophila. Cell 70, 553567. Nose, A., Umeda, T., Takeichi, M., 1997. Neuromuscular target recognition by a homophilic interaction of connectin cell adhesion molecules in Drosophila. Development 124, 14331441. OConnor, T.P., Duerr, J.S., Bentley, D., 1990. Pioneer growth cone steering decisions mediated by single filopodial contacts in situ. J. Neurosci. 10, 39353946. Odden, J.P., Holbrook, S., Doe, C.Q., 2002. Drosophila HB9 is expressed in a subset of motoneurons and interneurons, where it regulates gene expression and axon pathfinding. J. Neurosci. 22, 91439149. ODowd, D.K., 1995. Voltage-gated currents and firing properties of embryonic Drosophila neurons grown in a chemically defined medium. J. Neurobiol. 27, 113126. ODowd, D.K., Aldrich, R.W., 1988. Voltage-clamp analysis of sodium channels in wild-type and mutant Drosophila neurons. J. Neurosci. 8, 36333643.

Oland, L.A., Tolbert, L.P., 2003. Key interactions between neurons and glial cells during neural development in insects. Annu. Rev. Entomol. 48, 89110. Orchard, I., Loughton, B.G., 1985. Neurosecretion. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Endocrinology I. Pergamon, New York, pp. 61108. Packard, M., Koo, E.S., Gorczyca, M., Sharpe, J., Cumberledge, S., et al., 2002. The Drosophila Wnt, wingless, provides an essential signal for pre- and postsynaptic differentiation. Cell 111, 319330. Packard, M., Mathew, D., Budnik, V., 2003. Wnts and TGF beta in synaptogenesis: old friends signalling at new places. Nat. Rev. Neurosci. 4, 113120. Paradis, S., Sweeney, S.T., Davis, G.W., 2001. Homeostatic control of presynaptic release is triggered by postsynaptic membrane depolarization. Neuron 30, 737749. Parfitt, K., Reist, N., Li, J., Burgess, R., Deitcher, D., et al., 1995. Drosophila genetics and the functions of synaptic proteins. Cold Spring Harb. Symp. Quant. Biol. 60, 371377. Parmentier, M.L., Pin, J.P., Bockaert, J., Grau, Y., 1996. Cloning and functional expression of a Drosophila metabotropic glutamate receptor expressed in the embryonic CNS. J. Neurosci. 16, 66876694. Parnas, D., Haghighi, A.P., Fetter, R.D., Kim, S.W., Goodman, C.S., 2001. Regulation of postsynaptic structure and protein localization by the Rho-type guanine nucleotide exchange factor dPix. Neuron 32, 415424. Patel, N.H., Snow, P.M., Goodman, C.S., 1987. Characterization and cloning of fasciclin III: a glycoprotein expressed on a subset of neurons and axon pathways in Drosophila. Cell 48, 975988. Pennetta, G., Hiesinger, P., Fabian-Fine, R., Meinertzhagen, I., Bellen, H., 2002. Drosophila VAP-33A directs bouton formation at neuromuscular junctions in a dosage-dependent manner. Neuron 35, 291306. Petersen, S.A., Fetter, R.D., Noordermeer, J.N., Goodman, C.S., DiAntonio, A., 1997. Genetic analysis of glutamate receptors in Drosophila reveals a retrograde signal regulating presynaptic transmitter release. Neuron 19, 12371248. Pfeiffer, H.H., 1937. Versuche nuit ganglionkulturen aus corethra larven. Archiv. Exp. Zellforschung 20, 225229. Phillips, G.R., Huang, J.K., Wang, Y., Tanaka, H., Shapiro, L., et al., 2001. The presynaptic particle web: ultrastructure, composition, dissolution, and reconstitution. Neuron 32, 6377. Pongs, O., Lindemeier, J., Zhu, X.R., Theil, T., Engelkamp, D., et al., 1993. Frequenin a novel calcium-binding protein that modulates synaptic efficacy in the Drosophila nervous system. Neuron 11, 1528. Prokop, A., 1999. Integrating bits and pieces: synapse structure and formation in Drosophila embryos. Cell Tissue Res. 297, 169186. Prokop, A., Landgraf, M., Rushton, E., Broadie, K., Bate, M., 1996. Presynaptic development at the Drosophila neuromuscular junction: assembly and localization of presynaptic active zones. Neuron 17, 617626.

Functional Development of the Neuromusculature

131

Prokop, A., Technau, G.M., 1991. The origin of postembryonic neuroblasts in the ventral nerve cord of Drosophila melanogaster. Development 111, 7988. Prokop, A., Technau, G.M., 1994. Early tagma-specific commitment of Drosophila CNS progenitor NB1-1. Development 120, 25672578. Reiff, D.F., Thiel, P.R., Schuster, C.M., 2002. Differential regulation of active zone density during long-term strengthening of Drosophila neuromuscular junctions. J. Neurosci. 22, 93999409. Renden, R., Berwin, B., Chin, C.T., Kreber, R., Ganetzky, B., et al., 2001. Drosophila CAPS is an essential gene required for the regulation of dense core vesicle release and synaptic vesicle fusion. Neuron 31, 421437. Renden, R.B., Broadie, K., 2003. Mutation and activation of Galpha s similarly alters pre- and postsynaptic mechanisms modulating neurotransmission. J. Neurophysiol. 89, 26202638. Renger, J.J., Ueda, A., Atwood, H.L., Govind, C.K., Wu, C.F., 2000. Role of cAMP cascade in synaptic stability and plasticity: ultrastructural and physiological analyses of individual synaptic boutons in Drosophila memory mutants. J. Neurosci. 20, 39803992. Rheuben, M.B., Kammer, A.E., 1981. Membrane structures and physiology of an immature synapse. J. Neurocytol. 10, 557575. Rheuben, M.B., Yoshihara, M., Kidokoro, Y., 1999. Ultrastructural correlates of neuromuscular junction development. Int. Rev. Neurobiol. 43, 6992. Riechmann, V., Irion, U., Wilson, R., Grosskortenhaus, R., Leptin, M., 1997. Control of cell fates and segmentation in the Drosophila mesoderm. Development 124, 29152922. Rieckhof, G.E., Yoshihara, M., Guan, Z., Littleton, J.T., 2003. Presynaptic N-type calcium channels regulate synaptic growth. J. Biol. Chem. 278(42), 4109941108. Ritzenthaler, S., Chiba, A., 2003. Myopodia (postsynaptic filopodia) participate in synaptic target recognition. J. Neurobiol. 55, 3140. Ritzenthaler, S., Suzuki, E., Chiba, A., 2000. Postsynaptic filopodia in muscle cells interact with innervating motoneuron axons. Nat. Neurosci. 3, 10121017. Roche, J.P., Packard, M.C., Moeckel-Cole, S., Budnik, V., 2002. Regulation of synaptic plasticity and synaptic vesicle dynamics by the PDZ protein Scribble. J. Neurosci. 22, 64716479. Rodesch, C.K., Broadie, K., 2000. Genetic studies in Drosophila: vesicle pools and cytoskeleton-based regulation of synaptic transmission. Neuroreport 11, R4553. Rohrbough, J., Broadie, K., 2002. Electrophysiological analysis of synaptic transmission in central neurons of Drosophila larvae. J. Neurophysiol. 88(2), 847860. Rohrbough, J., Grotewiel, M.S., Davis, R.L., Broadie, K., 2000. Integrin-mediated regulation of synaptic morphology, transmission, and plasticity. J. Neurosci. 20, 68686878. Rohrbough, J., ODowd, D.K., Baines, R.A., Broadie, K., 2003. Cellular bases of behavioral plasticity: establishing

and modifying synaptic circuits in the Drosophila genetic system. J. Neurobiol. 54, 254271. Rohrbough, J., Pinto, S., Mihalek, R.M., Tully, T., Broadie, K., 1999. latheo, a Drosophila gene involved in learning, regulates functional synaptic plasticity. Neuron 23, 5570. Roos, J., Hummel, T., Ng, N., Klambt, C., Davis, G.W., 2000. Drosophila Futsch regulates synaptic microtubule organization and is necessary for synaptic growth. Neuron 26, 371382. Rose, D., Zhu, X., Kose, H., Hoang, B., Cho, J., et al., 1997. Toll, a muscle cell surface molecule, locally inhibits synaptic initiation of the RP3 motoneuron growth cone in Drosophila. Development 124, 15611571. Rose, U., Ferber, M., Hustert, R., 2001. Maturation of muscle properties and its hormonal control in an adult insect. J. Exp. Biol. 204, 35313545. Ruiz-Gomez, M., Coutts, N., Price, A., Taylor, M.V., Bate, M., 2000. Drosophila dumbfounded: a myoblast attractant essential for fusion. Cell 102, 189198. Ruiz-Gomez, M., Coutts, N., Suster, M.L., Landgraf, M., Bate, M., 2002. myoblasts incompetent encodes a zinc finger transcription factor required to specify fusioncompetent myoblasts in Drosophila. Development 129, 133141. Ruiz-Gomez, M., Romani, S., Hartmann, C., Jackle, H., Bate, M., 1997. Specific muscle identities are regulated by Kruppel during Drosophila embryogenesis. Development 124, 34073414. Rushton, E., Drysdale, R., Abmayr, S.M., Michelson, A.M., Bate, M., 1995. Mutations in a novel gene, myoblast city, provide evidence in support of the founder cell hypothesis for Drosophila muscle development. Development 121, 19791988. Saito, M., Kawai, N., 1985. Developmental changes in the glutamate receptor at the insect neuromuscular synapse. Brain Res. 350, 97102. Saito, M., Kawai, N., 1987. Patch clamp study of single glutamate channels during development in insect muscle. Devel. Biol. 121, 9096. Saito, M., Wu, C.F., 1991. Expression of ion channels and mutational effects in giant Drosophila neurons differentiated from cell division-arrested embryonic neuroblasts. J. Neurosci. 11, 21352150. Saitoe, M., Tanaka, S., Takata, K., Kidokoro, Y., 1997. Neural activity affects distribution of glutamate receptors during neuromuscular junction formation in Drosophila embryos. Devel. Biol. 184, 4860. Salinas, P.C., 1999. Wnt factors in axonal remodelling and synaptogenesis. Biochem. Soc. Symp. 65, 101109. Salkoff, L., 1985. Development of ion channels in the flight muscles of Drosophila. J. Physiol. (Paris) 80, 275282. Sanes, J.R., Lichtman, J.W., 1999. Development of the vertebrate neuromuscular junction. Annu. Rev. Neurosci. 22, 389442. Sassoe-Pognetto, M., Fritschy, J.M., 2000. Mini-review: gephyrin, a major postsynaptic protein of GABAergic synapses. Eur. J. Neurosci. 12, 22052210.

132 Functional Development of the Neuromusculature

Scheiffele, P., Fan, J., Choih, J., Fetter, R., Serafini, T., 2000. Neuroligin expressed in nonneuronal cells triggers presynaptic development in contacting axons. Cell 101, 657669. Schmid, A., Chiba, A., Doe, C.Q., 1999. Clonal analysis of Drosophila embryonic neuroblasts: neural cell types, axon projections and muscle targets. Development 126, 46534689. Schmidt, H., Luer, K., Hevers, W., Technau, G.M., 2000. Ionic currents of Drosophila embryonic neurons derived from selectively cultured CNS midline precursors. J. Neurobiol. 44, 392413. Schrader, S., Merritt, D.J., 2000. Central projections of Drosophila sensory neurons in the transition from embryo to larva. J. Comp. Neurol. 435, 3444. Schulte, J., Tepass, U., Auld, V.J., 2003. Gliotactin, a novel marker of tricellular junctions, is necessary for septate junction development in Drosophila. J. Cell Biol. 161, 9911000. Schulze, K.L., Broadie, K., Perin, M.S., Bellen, H.J., 1995. Genetic and electrophysiological studies of Drosophila syntaxin-1A demonstrate its role in nonneuronal secretion and neurotransmission. Cell 80, 311320. Schulze, K.L., Littleton, J.T., Salzberg, A., Halachmi, N., Stern, M., et al., 1994. rop, a Drosophila homolog of yeast Sec1 and vertebrate n-Sec1/Munc-18 proteins, is a negative regulator of neurotransmitter release in vivo. Neuron 13, 10991108. Schuster, C.M., Davis, G.W., Fetter, R.D., Goodman, C.S., 1996a. Genetic dissection of structural and functional components of synaptic plasticity. I. Fasciclin II controls synaptic stabilization and growth [see comments]. Neuron 17, 641654. Schuster, C.M., Davis, G.W., Fetter, R.D., Goodman, C.S., 1996b. Genetic dissection of structural and functional components of synaptic plasticity. II. Fasciclin II controls presynaptic structural plasticity [see comments]. Neuron 17, 655667. Schuster, C.M., Ultsch, A., Schloss, P., Cox, J.A., Schmitt, B., et al., 1991. Molecular cloning of an invertebrate glutamate receptor subunit expressed in Drosophila muscle. Science 254, 112114. Schuster, C.M., Ultsch, A., Schmitt, B., Betz, H., 1993. Molecular analysis of Drosophila glutamate receptors. EXS 63, 234240. Scott, E.K., Reuter, J.E., Luo, L., 2003. Dendritic development of Drosophila high order visual system neurons is independent of sensory experience. BMC Neurosci. 4, 14. Sepp, K.J., Auld, V.J., 2003. Reciprocal interactions between neurons and glia are required for Drosophila peripheral nervous system development. J. Neurosci. 23, 82218230. Sepp, K.J., Schulte, J., Auld, V.J., 2000. Developmental dynamics of peripheral glia in Drosophila melanogaster. Glia 30, 122133. Sepp, K.J., Schulte, J., Auld, V.J., 2001. Peripheral glia direct axon guidance across the CNS/PNS transition zone. Devel. Biol. 238, 4763.

Shapira, M., Zhai, R.G., Dresbach, T., Bresler, T., Torres, V.I., et al., 2003. Unitary assembly of presynaptic active zones from PiccoloBassoon transport vesicles. Neuron 38, 237252. Sigrist, S.J., Reiff, D.F., Thiel, P.R., Steinert, J.R., Schuster, C.M., 2003. Experience-dependent strengthening of Drosophila neuromuscular junctions. J. Neurosci. 23, 65466556. Sigrist, S.J., Thiel, P.R., Reiff, D.F., Lachance, P.E., Lasko, P., et al., 2000. Postsynaptic translation affects the efficacy and morphology of neuromuscular junctions. Nature 405, 10621065. Sigrist, S.J., Thiel, P.R., Reiff, D.F., Schuster, C.M., 2002. The postsynaptic glutamate receptor subunit DGluRIIA mediates long-term plasticity in Drosophila. J. Neurosci. 22, 73627372. Simpson, J.H., Bland, K.S., Fetter, R.D., Goodman, C.S., 2000. Short-range and long-range guidance by Slit and its Robo receptors: a combinatorial code of Robo receptors controls lateral position. Cell 103, 10191032. Singh, S., Wu, C.F., 1999. Ionic currents in larval muscles of Drosophila. Int. Rev. Neurobiol. 43, 191220. Sink, H., Rehm, E.J., Richstone, L., Bulls, Y.M., Goodman, C.S., 2001. sidestep encodes a target-derived attractant essential for motor axon guidance in Drosophila. Cell 105, 5767. Sink, H., Whitington, P.M., 1991. Pathfinding in the central nervous system and periphery by identified embryonic Drosophila motor axons. Development 112, 307316. Sippy, T., Cruz-Martin, A., Jeromin, A., Schweizer, F.E., 2003. Acute changes in short-term plasticity at synapses with elevated levels of neuronal calcium sensor-1. Nat. Neurosci. 6, 10311038. Smith, L.A., Wang, X., Peixoto, A.A., Neumann, E.K., Hall, L.M., et al., 1996. A Drosophila calcium channel alpha1 subunit gene maps to a genetic locus associated with behavioral and visual defects. J. Neurosci. 16(24), 78687879. Solc, C.K., Aldrich, R.W., 1988. Voltage-gated potassium channels in larval CNS neurons of Drosophila. J. Neurosci. 8, 25562570. Solc, C.K., Zagotta, W.N., Aldrich, R.W., 1987. Singlechannel and genetic analyses reveal two distinct A-type potassium channels in Drosophila. Science 236, 10941098. Staehling-Hampton, K., Hoffmann, F.M., Baylies, M.K., Rushton, E., Bate, M., 1994. dpp induces mesodermal gene expression in Drosophila. Nature 372, 783786. Suster, M.L., Bate, M., 2002. Embryonic assembly of a central pattern generator without sensory input. Nature 416, 174178. Suzuki, E., Rose, D., Chiba, A., 2000. The ultrastructural interactions of identified pre- and postsynaptic cells during synaptic target recognition in Drosophila embryos. J. Neurobiol. 42, 448459. Sweeney, S.T., Broadie, K., Keane, J., Niemann, H., OKane, C.J., 1995. Targeted expression of tetanus toxin light chain in Drosophila specifically eliminates

Functional Development of the Neuromusculature

133

synaptic transmission and causes behavioral defects. Neuron 14, 341351. Taghert, P.H., Veenstra, J.A., 2003. Drosophila neuropeptide signaling. Adv. Genet. 49, 165. Tessier-Lavigne, M., Goodman, C.S., 1996. The molecular biology of axon guidance. Science 274, 11231133. Thomas, J.B., Bastiani, M.J., Bate, M., Goodman, C.S., 1984. From grasshopper to Drosophila: a common plan for neuronal development. Nature 310, 203207. Thomas, U., Ebitsch, S., Gorczyca, M., Koh, Y.H., Hough, C.D., et al., 2000. Synaptic targeting and localization of discs-large is a stepwise process controlled by different domains of the protein. Curr. Biol. 10, 11081117. Thomas, U., Kim, E., Kuhlendahl, S., Koh, Y.H., Gundelfinger, E.D., et al., 1997. Synaptic clustering of the cell adhesion molecule fasciclin II by discs-large and its role in the regulation of presynaptic structure. Neuron 19, 787799. Thomas, W.E., Jordan, F.L., Townsel, J.G., 1987. The status of the study of invertebrate neurons in tissue culture phylum Arthropoda. Comp. Biochem. Physiol. A 87, 215222. Thor, S., Andersson, S.G., Tomlinson, A., Thomas, J.B., 1999. A LIM-homeodomain combinatorial code for motor-neuron pathway selection. Nature 397, 7680. Thor, S., Thomas, J.B., 1997. The Drosophila islet gene governs axon pathfinding and neurotransmitter identity. Neuron 18, 397409. Titmus, M.J., 1981. Ultrastructure of identified fast excitatory, slow excitatory and inhibitory neuromuscular junctions in the locust. J. Neurocytol. 10, 363385. Treherne, J.E., 1985. Blood-brain barrier. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Nervous System: Structure and Motor Function. Pergamon, New York, pp. 115138. Tsunoda, S., Salkoff, L., 1995a. Genetic analysis of Drosophila neurons: Shal, Shaw, and Shab encode most embryonic potassium currents. J. Neurosci. 15, 17411754. Tsunoda, S., Salkoff, L., 1995b. The major delayed rectifier in both Drosophila neurons and muscle is encoded by Shab. J. Neurosci. 15, 52095221. Ueda, A., Kidokoro, Y., 1996. Longitudinal body wall muscles are electrically coupled across the segmental boundary in the third instar larva of Drosophila melanogaster. Invert. Neurosci. 1, 315322. Usherwood, P.N., 1977. Glutamatergic synapses in invertebrates [proceedings]. Biochem. Soc. Trans. 5, 845849. Usherwood, P.N.R., 1994. Insect glutamate receptors. In: Evans, P.D. (Ed.), Advances in Insect Physiology. Academic Press, New York, pp. 309341. van der Bliek, A.M., Meyerowitz, E.M., 1991. Dynaminlike protein encoded by the Drosophila shibire gene associated with vesicular traffic. Nature 351, 411414. Verstreken, P., Kjaerulff, O., Lloyd, T.E., Atkinson, R., Zhou, Y., et al., 2002. Endophilin mutations block clathrin-mediated endocytosis but not neurotransmitter release. Cell 109, 101112.

Walther, C., 1981. Synaptic terminals from an identified motoneuron in locust muscle: comparison between first instar larva and adult. Neurosci. Lett. 27, 237242. Wang, J.W., Humphreys, J.M., Phillips, J.P., Hilliker, A.J., Wu, C.F., 2000. A novel leg-shaking Drosophila mutant defective in a voltage-gated K() current and hypersensitive to reactive oxygen species. J. Neurosci. 20, 59585964. Weeks, J.C., Jacobs, G.A., Pierce, J.T., Sandstrom, D.J., Streichert, L.C., et al., 1997. Neural mechanisms of behavioral plasticity: metamorphosis and learning in Manduca sexta. Brain Behav. Evol. 50 (Suppl. 1), 6980. Winberg, M.L., Mitchell, K.J., Goodman, C.S., 1998. Genetic analysis of the mechanisms controlling target selection: complementary and combinatorial functions of netrins, semaphorins, and IgCAMs. Cell 93, 581591. Wittle, A.E., Kamdar, K.P., Finnerty, V., 1999. The Drosophila cinnamon gene is functionally homologous to Arabidopsis cnx1 and has a similar expression pattern to the mammalian gephyrin gene. Mol. Gen. Genet. 261, 672680. Woods, D.F., Bryant, P.J., 1993. ZO-1, DlgA and PSD-95/ SAP90: homologous proteins in tight, septate and synaptic cell junctions. Mech. Devel. 44, 8589. Wu, C.F., Ganetzky, B., 1992. Neurogenetic studies of ion channels in Drosophila. Ion Channels 3, 261314. Wu, M.N., Bellen, H.J., 1997. Genetic dissection of synaptic transmission in Drosophila. Curr. Opin. Neurobiol. 7, 624630. Yao, W.D., Rusch, J., Poo, M., Wu, C.F., 2000. Spontaneous acetylcholine secretion from developing growth cones of Drosophila central neurons in culture: effects of cAMP-pathway mutations. J. Neurosci. 20, 26262637. Yao, W.D., Wu, C.F., 1999. Auxiliary hyperkinetic beta subunit of K channels: regulation of firing properties and K currents in Drosophila neurons. J. Neurophysiol. 81, 24722484. Yasuyama, K., Salvaterra, P.M., 1999. Localization of choline acetyltransferase-expressing neurons in Drosophila nervous system. Microsc. Res. Tech. 45, 6579. Ye, Z.C., Wyeth, M.S., Baltan-Tekkok, S., Ransom, B.R., 2003. Functional hemichannels in astrocytes: a novel mechanism of glutamate release. J. Neurosci. 23, 35883596. Yoshihara, M., Rheuben, M.B., Kidokoro, Y., 1997. Transition from growth cone to functional motor nerve terminal in Drosophila embryos. J. Neurosci. 17, 84088426. Yoshihara, M., Suzuki, K., Kidokoro, Y., 2000. Two independent pathways mediated by cAMP and protein kinase A enhance spontaneous transmitter release at Drosophila neuromuscular junctions. J. Neurosci. 20, 83158322. Yuan, L.L., Ganetzky, B., 1999. A glial-neuronal signaling pathway revealed by mutations in a neurexin-related protein. Science 283, 13431345.

134 Functional Development of the Neuromusculature

Zagotta, W.N., Brainard, M.S., Aldrich, R.W., 1988. Single-channel analysis of four distinct classes of potassium channels in Drosophila muscle. J. Neurosci. 8, 47654779. Zhang, D., Kuromi, H., Kidokoro, Y., 1999. Activation of metabotropic glutamate receptors enhances synaptic transmission at the Drosophila neuromuscular junction. Neuropharmacology 38, 645657. Zhang, Y.Q., Bailey, A.M., Matthies, H.J., Renden, R.B., Smith, M.A., et al., 2001. Drosophila fragile X-related gene regulates the MAP1B homolog Futsch to control synaptic structure and function. Cell 107, 591603. Zhang, Y., Featherstone, D., Davis, W., Rushton, E., Broadie, K., 2000. Drosophila D-titin is required for myoblast fusion and skeletal muscle striation. J. Cell Sci. 113, 31033115. Zhao, M.L., Wu, C.F., 1997. Alterations in frequency coding and activity dependence of excitability in cultured neurons of Drosophila memory mutants. J. Neurosci. 17, 21872199. Zheng, W., Feng, G., Ren, D., Eberl, D.F., Hannan, F., et al., 1995. Cloning and characterization of a calcium channel alpha 1 subunit from Drosophila melanogaster with similarity to the rat brain type D isoform. J. Neurosci. 15, 11321143. Zhong, Y., 1996. Genetic dissection of signal transduction mechanisms underlying PACAP-like neuropeptide transmission in Drosophila: synergy of cAMP and

Ras/Raf pathways. Ann. N. Y. Acad. Sci. 805, 6779, discussion 7980. Zhong, Y., Budnik, V., Wu, C.F., 1992. Synaptic plasticity in Drosophila memory and hyperexcitable mutants: role of cAMP cascade. J. Neurosci. 12, 644651. Zhong, Y., Shanley, J., 1995. Altered nerve terminal arborization and synaptic transmission in Drosophila mutants of cell adhesion molecule fasciclin I. J. Neurosci. 15, 66796687. Zhong, Y., Wu, C.F., 1991. Altered synaptic plasticity in Drosophila memory mutants with a defective cyclic AMP cascade. Science 251, 198201. Zhong, Y., Wu, C.F., 1993. Differential modulation of potassium currents by cAMP and its long-term and short-term effects: dunce and rutabaga mutants of Drosophila. J. Neurogenet. 9, 1527. Zito, K., Fetter, R.D., Goodman, C.S., Isacoff, E.Y., 1997. Synaptic clustering of Fascilin II and Shaker: essential targeting sequences and role of Dlg. Neuron 19, 10071016. Zito, K., Parnas, D., Fetter, R.D., Isacoff, E.Y., Goodman, C.S., 1999. Watching a synapse grow: noninvasive confocal imaging of synaptic growth in Drosophila. Neuron 22, 719729. Zlatic, M., Landgraf, M., Bate, M., 2003. Genetic specification of axonal arbors: atonal regulates robo3 to position terminal branches in the Drosophila nervous system. Neuron 37, 4151.

Das könnte Ihnen auch gefallen