A map of taste neuron projections in the Drosophila CNS

We provide a map of the projections of taste neurons in the CNS of Drosophila. Using a collection of 67 GAL4 drivers representing the entire repertoire of Gr taste receptors, we systematically map the projections of neurons expressing these drivers in the thoracico-abdominal ganglion and the suboesophageal ganglion (SOG). We define 9 categories of projections in the thoracico-abdominal ganglia and 10 categories in the SOG. The projection patterns are modular, and can be interpreted as combinations of discrete pattern elements. The elements can be interpreted in terms of the taste organ from which the projections originate, the structures from which they originate, and the quality of taste information that they represent. The extensive diversity in projection patterns provides an anatomical basis for functional diversity in responses elicited by different taste stimuli.     

Sensory neurons and peripheral pathways in Drosophila embryos

Summary The thoracic and abdominal segments of the Drosophila embryo contain 373 neurons innervating external sensory structures and 162 neurons innervating chordotonal organs. These neurons are arranged in ventral, lateral and dorsal clusters within each segment, in a highly invariant pattern. Two fascicles are formed in each segment as the sensory axons grow ventrally towards the CNS and meet motor axons growing dorsally from the CNS. In all but the last segment, the anterior fascicle is contributed by the dorsal and lateral neurons, while the posterior one is formed by the ventral neurons. Five distinct segmental patterns are described, corresponding to (1) the prothorax, (2) the other two thoracic segments, (3) the first seven abdominal segments, (4) the eighth and (5) the ninth (and possibly the tenth) abdominal segments.The publisher regrets that two companion papers unfortunately were published out of sequence. The present paper should have preceded the paper entitled The sense organs in the Drosophila larva and their relation to the embryonic pattern of sensory neurons, which appeared in Volume 195, Number 4 of the journal (pp 222–228)     

Calcium spikes in a leech nonspiking neuron

The NS neurons are nonspiking cells, present as pairs in each midbody ganglion of the leech nervous system, which display a very extensive arborization. They were shown to regulate the coactivation of motoneurons. Here we have investigated the electrophysiological properties of these neurons under the hypothesis that transmission along the extensive neurites requires the aid of voltage-dependent conductances. The results indicate that NS neurons respond to electrical stimulation with a spike-like event, which was not an all-or-none but rather a graded phenomenon that depended on the intensity and duration of the electrical stimulus. The spike-like response was activated at a membrane potential of approximately −50 mV; its amplitude was a logarithmic function of the extracellular Ca²⁺ concentration and was unaffected by a broad range of changes in the extracellular Na⁺ concentration; intracellular application of tetraethylammonium (TEA) caused a large increase in its amplitude and duration. These data indicate that NS neurons bear voltage-dependent low-threshold Ca²⁺ and TEA-sensitive K⁺ conductances that could contribute to shaping synaptic signals, or transmission along the extensive neuritic tree.     

Axon extension and retraction by leech neurons: severing early projections to peripheral targets prevents normal retraction of other projections

During leech embryogenesis, interactions between homologous neurons in neighboring segments lead to the selective retraction of longitudinal axonal projections by midbody AP and AE neurons, which maintain lateral axonal projections to the periphery. Results of experiments reported here show that disconnecting the lateral projections from the periphery rescues the projections normally fated to retract. We propose that these neurons normally progress through two states during early development, one in which they are insensitive to interactions with their homologs (state A) and a second in which they are sensitive (state B). Establishment of lateral connections with their targets triggers the switch from state A to state B; cutting these projections puts neurons back to state A.     

Tract formation and axon fasciculation of molecularly distinct peripheral neuron subpopulations during leech embryogenesis.

In leech, the central projections of peripheral sensory neurons segregate into specific axonal tracts, which are distinguished by differential expression of surface antigens recognized by the monoclonal antibodies Lan3-2 and Lan4-2. Lan3-2 recognizes an epitope expressed on axons that segregate into three distinct axon fascicles. In contrast, the Lan4-2-positive axons selectively project into only one of the Lan3-2-positive axon tracts. These observations provide evidence for a hierarchy of guidance cues mediating specific pathway formation in this system. Since the Lan3-2 antibody has been shown to perturb this process and since, as shown here, the Lan3-2 and Lan4-2 antigens are closely molecularly interrelated, these antibodies may help define molecules and epitopes mediating neuronal recognition and axonal guidance.     

Specific pathway selection by the early projections of individual peripheral sensory neurons in the embryonic medicinal leech

In leech, the central annulus of each midbody segment possesses seven pairs of sensilla, which are mixed clusters of primary peripheral sensory neurons that extend their axons into the CNS where they segregate into distinct fascicles. Pathway selection by individual afferent growth cones of sensillar neurons was examined by double labeling using intracellular dye-filling with anitobody labeling in early Hirudo medicinalis embryos. The monoclonal antibody Lan3–2 was used because sensillar neuronal tracts are specifically labeled by this antibody. Examining 68 individually filled neurons we found that sensillar neuron growth cones bifurcate within the CNS, that they project long filopodia capable to sampling the local environment, and that all of them appeared to choose a single particular CNS fascicle without apparent retraction or realignment of growth cones. Furthermore, each side of the bifurcating afferent growth cones always chose the same fascicle, implying a specific choice of a distinct labeled pathway. By dye-filling individual central neurons (P-cells), we show that there are centrally projecting axons present at the time sensillar afferents enter the ganglionic primordia and select a particular fascicle, and we confirm that at least the dorsal peripheral nerve is likely to be pioneered by central neurons, not by the peripheral afferent. In the sensillum studied here, we sound examples of sensory neurons extending axons into one of all the avilable fascicles. Thus, an individual embryonic sensillum possesses a heterogeneous population of afferents with respect to the central fascicle chosen. This is consistent with the idea that segregation into distinct axon fascicles may be based upon functional differences between individual afferent neurons. Our findings argue strongly in favor of specific pathway selection by afferents in this system and are consistent with previous suggestions that there exists a hierarchy of cues, including surface glycoconjugates that mediate navigation of the sensillar growth cones and the fasciculation of their axons. 1994 John Wiley & Sons, Inc.     

Morphological analysis of Drosophila larval peripheral sensory neuron dendrites and axons using genetic mosaics

Nervous system development requires the correct specification of neuron position and identity, followed by accurate neuron class-specific dendritic development and axonal wiring. Recently the dendritic arborization (DA) sensory neurons of the Drosophila larval peripheral nervous system (PNS) have become powerful genetic models in which to elucidate both general and class-specific mechanisms of neuron differentiation. There are four main DA neuron classes (I-IV)(1). They are named in order of increasing dendrite arbor complexity, and have class-specific differences in the genetic control of their differentiation(2-10). The DA sensory system is a practical model to investigate the molecular mechanisms behind the control of dendritic morphology(11-13) because: 1) it can take advantage of the powerful genetic tools available in the fruit fly, 2) the DA neuron dendrite arbor spreads out in only 2 dimensions beneath an optically clear larval cuticle making it easy to visualize with high resolution in vivo, 3) the class-specific diversity in dendritic morphology facilitates a comparative analysis to find key elements controlling the formation of simple vs. highly branched dendritic trees, and 4) dendritic arbor stereotypical shapes of different DA neurons facilitate morphometric statistical analyses. DA neuron activity modifies the output of a larval locomotion central pattern generator(14-16). The different DA neuron classes have distinct sensory modalities, and their activation elicits different behavioral responses(14,16-20). Furthermore different classes send axonal projections stereotypically into the Drosophila larval central nervous system in the ventral nerve cord (VNC)(21). These projections terminate with topographic representations of both DA neuron sensory modality and the position in the body wall of the dendritic field(7,22,23). Hence examination of DA axonal projections can be used to elucidate mechanisms underlying topographic mapping(7,22,23), as well as the wiring of a simple circuit modulating larval locomotion(14-17). We present here a practical guide to generate and analyze genetic mosaics(24) marking DA neurons via MARCM (Mosaic Analysis with a Repressible Cell Marker)(1,10,25) and Flp-out(22,26,27) techniques (summarized in Fig. 1).     

Modulation of the pattern of axonal projections of a leech motor neuron by ablation or transplantation of its target

The sixth segmental ganglion in the ventral nerve cord of the leech H. medicinalis contains a bilateral pair of rostral penile evertor motor neurons (RPEs) that in the adult innervate the male genitalia. During embryogenesis, the RPEs extend numerous extraganglionic projections. Only two of these innervate the target and are normally retained in the adult, while the others retract. Early, but not late, removal of the male genitalia results in the indefinite retention and continued growth of projections that would normally retract. Any of these projections can innervate targets transplanted to ectopic locations. We conclude that an RPE motor neuron requires a signal, provided by its interaction with the target organ during a critical period, in order to stop extending axons, stabilize those axons that contact the target, and retract those that do not.     

Calmodulin and son of sevenless dependent signaling pathways regulate midline crossing of axons in the Drosophila CNS

The establishment of axon trajectories is ultimately determined by the integration of intracellular signaling pathways. Here, a genetic approach in Drosophila has demonstrated that both Calmodulin and Son of sevenless signaling pathways are used to regulate which axons cross the midline. A loss in either signaling pathway leads to abnormal projection of axons across the midline and these increase with roundabout or slit mutations. When both Calmodulin and Son of sevenless are disrupted, the midline crossing of axons mimics that seen in roundabout mutants, although Roundabout remains expressed on crossing axons. Calmodulin and Son of sevenless also regulate axon crossing in a commissureless mutant. These data suggest that Calmodulin and Son of sevenless signaling pathways function to interpret midline repulsive cues which prevent axons crossing the midline.     

Localization of the myomodulin-like immunoreactivity in the leech CNS

The distribution of myomodulinlike immunoreactivity in the leech CNS was determined using an antiserum raised against Aplysia myomodulin. Segmental ganglia contained approximately 60 immunoreactive neurons. In addition, numerous fibers containing immunoreactive varicosities were found throughout the neuropil. Using a combination of Lucifer Yellow injections and immunocytochemistry, we identified neurons including the anterior Pagodas (AP), annulus erector (AE), motor neurons, Leydig, longitudinal muscle motoneurons (L), S cells, and coupling interneurons, all of which are active during the touch-elicited shortening reflex. FMRF-amide-like immunoreactivity in three of these cells (L, AP, and AE) was previously demonstrated. Specific staining for myomodulin was abolished by preadsorption of the antiserum with synthetic myomodulin, but not with FMRF-amide. These results suggest a potential role for myomodulin in both intrinsic and extrinsic modulation of the leech touch-elicited shortening reflex. Further, it is possible that several neurons mediating this reflex contain multiple neuromodulatory peptides. © 1996 John Wiley & Sons, Inc.     

Photosensory input pathways in the medicinal leech

Summary The medicinal leech,Hirudo medicinalis possesses two types of photosensory organs: five bilateral pairs of eyes embedded in two longitudinal rows in the dorsal surface of the head, and seven bilateral pairs of sensilla situated in both the dorsal and the ventral surface of each of the 21 body segments. The photoreceptor cells of each eye or sensillum project their axons centrally via a characteristic cephalic or segmental nerve which carries the photosensory input to the brain or to the segmental ganglion. In response to a pulse of light the photoreceptors produce a train of impulses whose frequency first rises to anearly peak and then declines to asteady state plateau at which it remains until the end of the pulse. The amplitude of the early peak response and the level of the steady state plateau rise linearly with the log of the light pulse intensity, but the dynamic range of the early peak response is much narrower than that of the plateau. Both ocular and sensillar photoreceptors adapt to the intensity of interpulse background illumination; the ocular receptors adapt so completely that their level of background activity is nearly independent of the background light intensity, whereas the ventral sensillar photoreceptors adapt incompletely, so that their background activity rises with the background light intensity. Ocular and sensillar photoreceptors make their maximal response to green light at a wavelength of about 540 nm. They are almost insensitive to red and violet light at both extremes of the visible spectrum. The photosensory response of a single eye is directionally selective, whereas that of a single sensillum has much less directional selectivity. Several higher order sensory neurons were identified in the segmental ganglion that receive photosensory input from the sensilla. One of these neurons has the sensillum in the ipsilateral dorso-medial body wall of the same segment as its receptive field and another neuron the bilateral set of ventral sensilla in the body wall of the next posterior segment.We are indebted to Frank S. Werblin for valuable advice and discussions. We thank Kenneth L. Carlock for designing and constructing much of the special electronic equipment used in this study. We also thank Alexander Petruncola for his helpful suggestions regarding the computational analysis of the experimental results and for writing the computer programs used in the processing of the data.This research was supported by Grant No. GB 31933X from the National Science Foundation, and NIH research grant No. GM 17866 and Training Grant No. GM 00829 from the Institute for General Medical Sciences.     

Leydig neuron activity modulates heartbeat in the medicinal leech

1. Leydig neurons fire spontaneously at low rates (less than 4 Hz), but their activity increases with mechanical stimulation or electrical stimulation of mechanosensory neurons. These conditions also cause acceleration of bursting in heart motor neurons. 2. The firing rate of Leydig cells was found to regulate heart rate in chains of isolated ganglia. When Leydig neurons were made to fire action potentials at relatively high frequencies (ca. 5-10 Hz), however, heart motor neurons ceased bursting and were either silenced or fired erratically. 3. Firing of Leydig neurons at high rates caused bilateral heart interneurons of ganglia 3 or 4 to fire tonically rather than in their normal alternating bursts Tonic firing of these heart interneurons accounts for the prolonged barrages of ipsps recorded in heart motor neurons and the disruption of their normal cyclic activity. 4. Preventing spontaneous activity of Leydig neurons with injected currents in isolated ganglia caused deceleration of the heartbeat rhythm but did not halt oscillation. 5. Electrical stimulation of peripheral nerve roots with Leydig neuron activity suppressed in isolated ganglia caused acceleration of heart rate.     

Developmental control of cell division in leech embryos

During embryogenesis, cell division must be spatially and temporally regulated with respect to other developmental processes. Leech embryos undergo a series of unequal and asynchronous cleavages to produce individually recognizable cells whose lineages, developmental fates and cell cycle properties have been characterized. Thus, leech embryos provide an opportunity to examine the regulation of cell division at the level of individual well-characterized cells within a community of different types of cells. Isolation of leech homologues of some of the highly conserved regulators of the cell division cycle, and characterization of their patterns of maternal and zygotic expression, indicate that the cell divisions of early leech embryos are regulated by cell type-specific mechanisms. These studies with leech embryos contribute to the emerging appreciation of the diverse mechanisms by which animals regulate cell division during early development.     

The role of matrix molecules in regeneration of leech CNS

Extracellular matrix (ECM) molecules extracted from the leech central nervous system (CNS) provide substrates that induce extensive growth of processes of identified leech nerve cells in culture. Two ECM molecules, laminin and tenascin, have been identified. The laminin-like molecule has been purified and shown to be a cross-shaped molecule similar to vertebrate laminin with subunits of 340, 220, 180, and 160 kD. Purified laminin as a substrate induces rapid outgrowth of Retzius (R) and Anterior Pagoda (AP) cells in culture. The tenascin molecule has been partially purified. In electronmicrographs, leech tenascin, like vertebrate tenascin, has six arms of equal size joined in a central globule. Highly enriched fractions of leech tenascin induce rapid and extensive outgrowth of Retzius and AP cells in culture. Substrate molecules not only induce outgrowth of processes but also affect the growth patterns of individual nerve cells. Neurites are straight with few branches in laminin, but curved with profuse branches on tenascin. During regeneration of the CNS in the animal, laminin appears at new sites associated with growth cones. The appearance of laminin correlates with the accumulation of microglial cells. Thus, ECM molecules with growth-promoting activity for leech nerve cells in vitro appear to be involved in inducing regeneration and allowing the neurites to reconnect with former targets. © 1992 John Wiley & Sons, Inc.     

Mapping motor neuron activity to overt behavior in the leech

As an initial step in constructing a quantitative biomechanical model of the medicinal leech (Hirudo medicinalis), we determined the passive properties of its body wall over the physiological range of dimensions. The major results of this study were:

1. The ellipsoidal cross section of resting leeches is maintained by tonic muscle activation as well as forces inherent in the structure of the body wall (i.e., residual stress).
2. The forces required for longitudinal and circumferential stretch to maximum physiological dimensions were similar in magnitude. Cutting out pieces of body wall did not affect the passive longitudinal or circumferential properties of body wall away from the edges of the cut.
3. The strain (i.e., the percentage change in dimension) of different body segments when subject to the same force was identical, despite differences in muscle crosssections.
4. Serotonin, a known modulator of tension in leech muscles, affected passive forces at all physiological muscle lengths. This suggests that the longitudinal muscle is responsible for at least part of the passive tension of the body wall.
5. We propose a simple viscoelastic model of the body wall. This model captures the dynamics of the passive responses of the leech body wall to imposed step changes in length. Using steady-state passive tensions predicted by the viscoelastic model we estimate the forces required to maintain the leech at any given length over the physiological range.

     

Developmental interdeterminacy in embryos of the leech Helobdella triserialis

In embryonic development of the leech Helobdella triserialis, each of the four paired positionally identifiable, ectodermal teloblasts (N, O, P, and Q) generates a bandlet of blast cell progeny that merges with ipsilateral bandlets into a germinal band. Left and right germinal bands coalesce into the germinal plate which gives rise to the segmental tissues of the leech and wherein the progeny of each teloblast generate a characteristic pattern of epidermal and neuronal cells. Experiments reported here show that the positionally identified O teloblast sometimes generates the P pattern and vice versa. The reversal of these teloblasts' generative identities was shown to correspond to the formation of chiasmata by their blast cell bandlets, so that the positions of their bandlets in the germinal band are reversed as well. Thus it is the position of the bandlet in the germinal band, rather than the position of the parent teloblast, which correlates with the fate of o and p blast cells. Moreover, two types of ablation experiments have shown that, in the absence of generative P teloblast progeny, those cells which would normally generate the O pattern take on a new fate and give rise to the P pattern in the nervous system, both at the gross pattern level in the segmental ganglia, and at the level of identified neurons in the peripheral nervous system. If related, these phenomena suggest that the O and P teloblasts, which derive from the symmetric cleavage of the OP proteloblasts, have a common developmental pluripotency. And in that case, the fates of their progeny are determined hierarchically on the basis of relative position in the nascent germinal band, with P-type fate being preferred.     

Centrifugation redistributes factors determining cleavage patterns in leech embryos

In the normal development of glossiphoniid leech embryos, cytoplasmic reorganization prior to the first cleavage generates visibly distinct domains of yolk-deficient cytoplasm, called teloplasm. During an ensuing series of stereotyped and unequal cell divisions, teloplasm is segregated primarily into cell CD of the two-cell stage and then into cell D of the four-cell and eight-cell stages. The subsequent fate of cell D is also unique in that it alone undergoes further cleavages which generate five bilateral pairs of embryonic stem cells, the mesodermal (M) and ectodermal (N, O/P, O/P, and Q) teloblasts. Here we report studies on the effects of centrifugation on cleavage pattern and protein composition of individual blastomeres of the leech Helobdella triserialis. Centrifugation partially stratifies the cytoplasm of each cell, generating a layer of clear cytoplasm in cell CD derived largely from teloplasm. After centrifuging embryos at the two-cell stage, clear cytoplasm present in cell CD and normally inherited by cell D is redistributed and can be inherited by both cells C and D at the second cleavage. The developmental fates of cells C and D in centrifuged embryos correlate with the amount of clear cytoplasm they receive. In particular, when clear cytoplasm has been distributed roughly equally between the two cells, both cell C and cell D undergo further cleavages resembling the pattern of divisions normally associated with cell D. Likewise, non-yolk-associated proteins, normally found in higher quantities in cell D than in cell C, appear evenly disbursed between the two cells under conditions which induce this fate change. These results are consistent with the idea that the fates of cells C and D are influenced by the distribution or cellular localization of cytoplasmic components.