Development of the 5-HT-like immunoreactive pedal plexus in the pond snail Lymnaea stagnalis appressa.
In the gastropod opisthobranch molluscs Tritonia diomedea and Pleurobranchaea californica, a plexus of axon processes that is immunoreactive to 5-HT (serotonin) antibodies lies over the pedal sole ciliary cells that produce ciliary locomotion. Because of the absence of peripheral neuronal cell bodies immunoreactive to 5-HT antibodies, this plexus was thought to originate from central serotonergic neurons (Moroz et al., 1997). In Lymnaea stagnalis, the European pond snail, a histochemical and ultrastructural study of pedal innervation (McKenzie et al., 1998) also found a neural plexus over pedal sole ciliary cells that was immunoreactive to 5-HT antibodies and that appeared to have serotonergic synapses to pedal sole ciliary cells and mucous glands. As in T. diomedea and P. califomica, these 5-HT-like immunoreactive (5-HT-LIR) processes were assumed to arise from central serotonergic neurons. Axons of the A-cluster neurons in the pedal ganglia of L. stagnalis have been traced into the foot and are believed to contribute to the 5-HT-LIR processes over the pedal sole ciliary cells (Syed et al., 1988; Syed and Winlow, 1989). Because 5-HT increases ciliary beating in molluscan species, control of these pedal cilia is a possible function of this 5-HT-LIR plexus.
A substantial amount of evidence suggests that 5-HT controls ciliary beating in molluscs; however, individual neurons that form the 5-HT-LIR plexus have not been identified. There are only two instances in gastropods where results show that serotonergic neuron activity increases ciliary beating. In T. diomedea, Audesirk (1978) identified a serotonergic neuron that increased pedal ciliary beating; and in Helisoma trivolis, Kuang and Goldberg (2001) showed that release of 5-HT by a serotonergic neuron drives the embryonic prototrochal cilia and the ciliary band on the pedal plate. Indirect evidence for neural control of ciliary cells by 5-HT has been shown by bath application of 5-HT on bivalve gill cilia (Aiello and Guideri, 1965), gastropod pedal sole cilia (Audesirk et al., 1979; Syed et al., 1988; Deliagina and Orlovsky, 1990; Willows et al., 1997), gastropod veliger ciliary cells (Koshtoyants et al., 1961; Mackie et al., 1976), and gastropod embryonic ciliary cells (Diefenbach et al., 1991; Uhler et al., 2000; Doran et al., 2004). Selective serotonin reuptake inhibitors that maintain a high 5-HT concentration after synaptic release of 5-HT also increase ciliary beating in gastropod embryos (Uhler et al., 2000).
In direct-developing freshwater snails such as Lymnaea, primordial cilia in prototrochal ciliary cells lateral to the mouth and in ciliary cells along the midline of the pedal plate are shown by scanning electron microscopy as the first cilia to appear (Morrill, 1982; Diefenbach et al., 1991, 1998; Koss et al., 2003). Later in development as metamorphosis begins in Lymnaea, clusters of ciliary cells on the anterolateral part of the foot, also shown by scanning electron microscopy, begin the formation of the adult ciliated pedal sole epithelium (Morrill, 1982). During development, these clusters of ciliary cells proliferate to cover the whole foot (Syed et al., 1988). In the ramshorn snail H. trivolvis, the areas covered by the primordial cilia can be recognized at 14% of the time until hatching (stage E14) by immunoreactivity to antibodies to 5-HT (Diefenbach et al., 1998). In this paper on Lymnaea stagnalis appressa Say (1821), a North American species similar to L. stagnalis, similar immunoreactivity during metamorphosis shows both the primordial ciliary cells and the initial pedal sole ciliary cells. Axon outgrowth from pedal ganglia neurons is not associated with the primordial ciliary cells, but it is directed toward developing pedal sole ciliary cells.
Two bilaterally symmetric neuron pairs described in this paper have their axons in the unpaired nerve that arises from the ventral pedal commissure. In prior work in L. stagnalis this small nerve has been called the medial columellar nerve (Janse. 1974; Slade et al., 1981). but in this nerve Janse (1974) found sensory responses from the foot and not from the column. My results, which are consistent with those of Marois and Croll (1992), show that this nerve descends to the ciliated pedal sole epithelium. It is renamed here the ventromediall pedal nerve (vpn). Axon anatomy suggests that the two pairs of axons seen in this nerve in the L. s. appressa embryo are homologs of axons from the identified neuron pairs PeD7 and PeVl in L. stagnalis (Kyriakides et al., 1989).
Marois and Croll (1992) described the time of appearance of some of the principal 5-HT-LIR neurons in L. stagnalis. In a related species, the North American L. s. appressa, this paper shows the axon destination of early pedal ganglion neurons that are 5-HT-LIR. Although other 5-HT-LIR neurons may contribute to the pedal plexus later in development, I show that pedal plexus formation begins soon after initiation of metamorphosis with the first immunoreactive pedal ganglion neurons that are seen. This resolves the previously unknown origin of this plexus to a specific group of large neurons. It suggests the question: Does this small number of potentially identifiable neurons affect pedal ciliary locomotion?
Materials and Methods
Individuals of Lymnaeu stagnalis appressa were collected locally from Egg Lake, San Juan Island, Washington, kept in the laboratory in pond water, and fed lettuce. Egg strings were collected from these snails, and the transparent capsules containing the eggs were separated from the egg string mucus. In Lymnaea, first cleavage of the egg occurs 3 to 4 h after egg strings are deposited. At first cleavage, the capsules were placed under constant illumination in filtered pond water at 25 [degrees]C. Shell and animal specimens of these snails are cataloged in the Smithsonian as USNM 883793.
The timing of metamorphosis to an adult-like form in L. s. appressa has not previously been given. Hatching occurs in L s. appressa in 11 days, which is the same as that described by Cumin (1972) for the European L. stagnalis when it is maintained at 25 [degrees]C. The developmental stages in L. s. appressa follow closely those described for L. stagnalis by Cumin (1972). From personal observations, growth of radula teeth in L. s. appressa embryos begins at 98 h [+ or -] 4.1 h SD (n = 6 egg strings), or at 37% of hatching time. A ring of eye pigment outlines the eye a few hours earlier, with about a day required for pigment to completely fill the eye. This, with growth of the shell and foot during this time, places metamorphosis to an adult-like form as occurring from about 36% to 48% of hatching time. In reports on laboratory-cultured L. stagnalis, hatching occurs in 8-9 days (Marois and Croll, 1992; Nagy and Elekes, 2000; Voronezhskaya and Elekes, 2003). In L. stagnalis individuals that hatch in 8 days, metamorphosis occurs at from 48% to 63% of hatching time (Nagy and Elekes, 2000). Although this percentage differs from that for metamorphosis in L. s. appressa, the time from first cleavage of the egg to metamorphosis is similar in the two species.
Staging here of an embryo is based on the time from first cleavage of the egg and is given as a percentage of the average hatching time of 1 1 days. Because of variability in development between different egg strings, as shown by the time of initial formation of radula teeth, this percentage does not indicate a precise stage of development for individual embryos. The actual stage of development in embryos with the same percentage of hatching time may differ by a few percent. The percentage of hatching time is indicated in each figure.
Embryos were removed from their capsules in bicarbonate-buffered saline and fixed in 0.1 mol [I.sub.-1] phosphate-buffered 4% formalin. Before fixation, embryos were either pinned through their lung opening or cut through the shell and column and pinned through the mantle. Saline solution was NaCl. 30 mmol [l.sup.-1]: KC1. 3 mmol [l.sup.-1]; Mg[Cl.sup.2], 2 mmol [l.sup.-1]; Ca[Cl.sub.2], 4 mmol [l.sup.-1]; NaHC[O.sub.3], 15 mmol [l.sub.-1]. Bicarbonate was added to the saline solution immediately before use. Fixed material was processed with antibodies at room temperature as previously described (Longley and Longley, 1986). Primary antibodies used at dilutions of 1:200 were rabbit antibodies to 5-HT from Sigma and anti-phospho-Histone H3 antibodies (06-570) from Upstate Biotechnology. These were followed by goat tetramethyl rhodamine isothiocyanate (TRITC) second antibodies (Jackson Immunoreseareh) at 1:200 dilution for fluorescence. Preadsorption of the 5-HT antibodies with 5-HT BSA conjugate from Immunostar at 10 [micro]g/ml gave negative results. Second antibodies atone also showed negative immunoreactivity. After application of antibodies, the shell and column in younger embryos or the mantle in older embryos was removed in the glycerol mounting medium, and the foot was positioned against the cover slip to give a ventral view of the nervous system and pedal axons. In some preparations the foot was cut sagittally and viewed laterally.
About 170 embryos that ranged in age from 34% to 73% of hatching time were processed for immunohistochemistry. These were examined with fluorescence microscopy, and preparations that could be clearly imaged as wholemounts were processed with a Biorad Radiance 2000 confocal microscope. Of these, confocal stacks were taken of 29 embryos between 37% and 42% of hatching time and of 23 embryos that were older than 42%. Single sections or stacks of confocal sections of the nervous system were projected in the z direction with ImageJ, ver. 1.37v, by using maximum intensity, sum, or standard deviation of the sections. Except where noted, figures that show projected sections of the nervous system are turned to give a dorsal view with anterior toward the top of the figure, and fluorescence is shown inverted to enhance the visual perception of" contrast.
The cerebral ganglia neurons in L. s. appressa are labeled c4, c5, c6, and c3, in the order of their appearance as in Marois and Croll (1992), and their time of appearance is given to provide a comparison to the developmental staging by Marois and Croll (1992). Because Marois and Croll (1992) see faster development (9 days in L. stagnalis as compared to 11 days in L. s. appressa), 5-HT-LIR cerebral neurons appear during the veliger stage in L. stagnalis. Although the first cerebral ganglion 5-HT-LIR neurons in L. s. appressa do not appear until after metamorphosis begins, the timing of this event as a percentage of hatching time is similar in the two species. The location and relative time of appearance of the first four pedal ganglion neurons to show 5-HT-LIR in L. s. appressa are the same as those seen in L. stagnalis by Marois and Croll (1992). The early immuno-reactive pedal ganglion neurons that are paired contralaterally and send axons to the ciliated pedal sole epithelium are labeled here as el through e7 in the order of their appearance.
Identification of ciliary cells
During metamorphosis (36% to 48% of time to hatching), the initial adult pedal sole ciliary cells and the primordial ciliary cells can be reliably seen either with antibodies to phosphorylated histone H3 (Fig. 1A, B, C) or with antibodies to 5-HT (Fig. 1D, E, F). In projections of confocal stacks taken perpendicular to the epithelium (Fig. 1A, B, E), the location of the immunoreactivity associated with the ciliary cells cannot be clearly seen, but a 90-degree rotation of a small segment through the center of one of these ciliary cells (Fig. IB) shows that the antibodies to phosphorylated histone H3 bind to an array of rods about 1.25 [micro]m in length (Fig. 1C). A similar result with antibodies to 5-HT is seen at 45% of hatching time in a 90-degree rotation of a confocal stack through pedal sole ciliary cells lateral to the midline (Fig. 1D). These immunoreactive organelles, which are much shorter than a typical pedal cilium length of 10 [micro]m. show the location of the ciliary cells.
In instances when primordial ciliary cell immunoreactivity to 5-HT antibodies was weak, a small bipolar neuron was occasionally seen in the primordial pedal plate cilia (Fig. IE). From the nomenclature of Koss et al. (2003) for Helisoma trivolvis, this neuron appears to lie between pedal plate ciliary cells pCl and pC2. A 90-degree clockwise rotation of this confocal stack shows that the perikaryon of this neuron is closer to the surface of the foot than the immunoreactive organelles associated with the ciliary cells (Fig. 1F). As seen during rotation of the confocal stack, the axon of this neuron lies at the level of the organelles and is obscured by their immunoreactivity after rotation of the confocal stack (supplemental video of stack rotation available at http://www.biobull.org/supplemental/). The perikaryon of this neuron extends more ventrally than the immunoreactive organelles, which indicates that these organelles lie within the ciliary cells. It is likely that this immunoreactivity is associated with ciliary basal bodies, but this has not been confirmed. With inverted fluorescence in the confocal stacks, there is a general background fluorescence not associated with the ciliary cell layer that accumulates in z projections to give a gray, stippled appearance. In - projections perpendicular to the foot that include the ciliary cells, their immunoreactivity is seen as a darker gray area superimposed on this background.
Early immunoreactive pedal ganglion neurons
5-HT-LIR cerebral and pedal ganglia neurons were not seen prior to the appearance of radula teeth, which occurs at 37% of hatching time. In an embryo at 39% of hatching time, cerebral ganglia axons from c4 neurons have reached the el neurons in the pedal ganglia (Fig. 2A). These el neurons lie immediately dorsal to the bilaterally symmetric clusters of pedal ciliary cells near the anterior edge of the foot and lateral to the embryonic pedal plate ciliary band (Fig. 2B). The el axon grows posteriorly and slightly laterally at the level of the pedal ganglion to a point posterior to the cluster of pedal sole epithelial ciliary cells. Here the el axon turns ventrally and anteriorly toward these cells. A second pedal ganglion neuron (e2, posterior to el) and the ventral commissure are faintly visible in this embryo.
[FIGURE 1 OMITTED]
In a slightly older embryo, this typical placement of the ciliary cells and outgrowth of the el axon can be seen on the right side of the foot, but on the left side, where the ciliary cell cluster is not under el, the el axon follows an atypical path (Fig. 2C). Its axon has a short posterior loop before it follows a path medially and anteriorly until its neurites branch directly over the pedal sole ciliary cells. In this embryo, el and its weakly immunoreactive axon that extends to the ventral commissure are also present. A third neuron (e3), which does not have an axon at this stage, is seen in the left pedal ganglion anterior to el. Filipodia-like neurites from the el neuron are also present and extend anteriorly along the path of the incoming c4 axon and to e3.
Axons of cerebral ganglia c4 neurons grow contralaterally through the cerebral commissure and then turn posteriorly toward the pedal ganglia. The time of appearance of immunoreactivity and outgrowth of axons in pedal and cerebral ganglia neurons were often different in left and right ganglia. In the cerebral and pedal ganglia, immunoreactivity to 5-HT antibodies in cerebral neuron c4 was usually seen first. Pedal ganglia immunoreactivity and outgrowth of el axons were frequently seen before c4 axons reached the pedal ganglia (Figs. 2C, 3).
[FIGURE 3 OMITTED]
Pedal 5-HT-LIR neurons in embryos at 38%-42% of hatching time showed various stages of growth. In two embryos from the same egg string at 42% of hatching time, axons from the first four pairs of pedal ganglion neurons can be identified before they are obscured by the arrival of the intensely immunoreactive cerebral c4 axons. The ventral pedal commissure and the ventromedian pedal nerve that branches from this commissure are present in these embryos before the c4 axons arrive (Fig. 3A, D). In the left pedal ganglia, outgrowth of the posteriorly and laterally directed el axon can be seen (Fig. 3A, D, indicated by arrows). Immunoreactivity within the ventromedian pedal nerve is from e2 and e3 axons. During development, the e3 axon is typically smaller and more intensely immunoreactive than the el axon. In magnified confocal substacks of the left pedal ganglion where the c4 axon has not grown past e3, the e3 axon extends posteriorly to follow the e2 axon into the ventral commissure (Fig. 3B, C). In a slightly more advanced stage of development where the c4 axon has not reached the left pedal ganglion (Fig. 3D, E), the e4 neuron is present between el and e2. The axon from e4 extends anteriorly and turns to follow the el axon laterally (Fig. 3F). The e3 axon is not visible in this left ganglion, but it is visible in an enlarged, contrast-enhanced view from a single confocal section of the right pedal ganglion (arrow, Fig. 3G). In this section, el and e4 can also be seen as separate neurons.
[FIGURE 4 OMITTED]
In an embryo from a different egg string at 43% of hatching time, six of the seven initially developing 5-HT-L1R neurons can be identified (Fig. 4). Axons from all of these neurons will contribute to the pedal plexus over the pedal sole epithelium. The first four neurons described above are identified by position and are distributed along the path in the pedal ganglia that is followed by the c4 and c5 cerebral ganglion axons (Fig. 4A, B). The c4 and c5 axons branch through the dorsal pedal commissure and pass through the ventral pedal commissure. The el and e4 axons extend laterally in the putative medial pedal nerve to the patch of ciliary cells that is expanding during growth of the foot. The relation of these ciliary cells to the foot and to the pedal ganglia can be seen in a lower magnification projection that includes the surface of the foot (Fig. 5). A fifth neuron (e5) with very weak immunoreactivity lies laterally to el. It can be recognized by its axon, which extends posteriorly in the putative inferior pedal nerve (Fig. 4C, D). No axon was seen from the sixth neuron (e6), which is faintly outlined anterior to the dorsal commissure (Fig. 4A, B).
Intermediate stages of development
By 47% of hatching time, the four cerebral neuron c4 and c5 axons in the cerebropedal connectives outline the pedal ganglia as they pass by the previously identified immuno reactive neurons (Fig. 6A). Within the ventral commissure the large, weakly immunoreactive el axons are visible, but at this stage the c4 and c5 axons obscure the e3 axons. In an enlarged view of this commissure, both the e2 and e3 axons can be seen where they enter the ventromedian pedal nerve (Fig. 6B). At a more advanced stage of development (66% of batching time), these four axons from the el and e3 pairs and the four axons from the c4 and c5 cerebral neuron pairs are the only 5-HT-LIR axons in the ventral commissure. At this time the e2 axons still have relatively weak immuno-reactivity to 5-HT antibodies (Fig. 6C). From examination of individual sections in confocal stacks, e3 and e6 are ventral to the cerebral ganglion axons, while the other four (including e5 not shown in Fig. 6A) are dorsal to these axons.
[FIGURE 5 OMITTED]
The patches of ciliary cells, which initially appear near the anterior edge of the foot, expand and move posteriorly as the foot grows. The el and e4 axons in the putative medial pedal nerve (mpn) project laterally as a closely apposed pair to the bilaterally symmetric clusters of these ciliary ceils (Fig. 6D, E). The close apposition of the el and e4 axons suggests these axons are contained in a nerve. Immunoreactivity of ciliary cells over which the el and e4 axons project are seen as a darker gray area (Fig. 6E). In a 90-degree rotation of this confocal stack cropped to a width of 18 [micro]m, the multiple branched neurites of el and e4 are shown to descend to lie directly over the ciliary cells (Fig. 6F). Immunoreactivity of individual organelles in the more ventral part of these ciliary cells is not resolved in this rotation, but can be seen in an area cropped to 1.16 [micro]m from the same confocal stack (Fig. ID). At this time only one axon is seen in the putative inferior pedal nerve and only one axon is seen in the putative superior pedal nerve (Fig. 6D). These axons apparently come from the e5 and e6 neurons, respectively. Axons growing out from the pedal ganglia toward the ciliated pedal sole epithelium are organized in what appear to be the adult superior (spn), medial (mpn), and inferior (ipn) pedal nerves (Fig. 6D). As development proceeds and the foot grows, the projection of immunoreactive axons in these nerves is consistent with extension of the superior pedal nerve toward the anterior part of the foot, the medial pedal nerve toward the lateral and posterior part of the foot, and the inferior pedal nerve toward the posterior part of the foot.
At 50% of hatching time, a third pair of cerebral neurons (c6) has sent axons to the pedal ganglia, and at 55% of hatching time a fourth cerebral neuron (c3) is also present. Axons of these neurons enter the pedal ganglia in a separate fascicle that is lateral in the pedal ganglia to the c4 and c5 axons (Fig. 7A). The c3 and c6 axons branch into the dorsal pedal commissure as do c4 and c5 axons, but they do not enter the ventral pedal commissure.
At 55% of hatching time, 5-HT-LIR axons from the pedal ganglia in the superior and medial pedal nerves branch over the ciliated pedal sole epithelium in the anterior part of the foot (Fig. 7A). At this age the embryo can adhere to the capsule with the anterior part of the foot, and pedal ciliary locomotion inside the capsule is seen. Axons in the ventro-median and inferior pedal nerves extend toward the posterior part of the foot, but they have only a few branches over the pedal sole epithelium where immunoreactive cilia are still mostly absent. At this time two axons can be seen in the superior pedal nerve (Fig. 7B). One axon turns anteriorly as it enters the pedal ganglion, apparently toward e6, while the other turns posteriorly toward e7, which is identified at this stage. It was not possible to determine whether el lies dorsally or ventrally to the cerebral ganglia axons. An additional neuron in each ganglion anterior to the dorsal commissure near e6 was also seen, but an axon from this neuron was not seen in the pedal nerves.
In embryos at 60% of hatching time e2 and e3 axons have two major branching patterns in the ventromedian pedal nerve (vpn). They extend posteriorly in the foot at the level of the pedal ganglia (Fig. 8A), and they also branch ven trally to the pedal sole epithelium in the central region of the foot (Fig. 8C). Axons with both weak (e2) and strong (e3) immunoreactivity can be resolved in the vpn (Fig. 8B). In the inferior pedal nerve (ipn), e5 parallels the posteriorly directed ventromedian pedal nerve and has branches laterally directed toward the edge of the foot (Fig. 8A). Neither the posteriorly directed axons in the ventromedian nerve nor the e5 axon in the inferior pedal nerve are in close contact at this time with the pedal sole epithelium in the posterior part of the foot. With growth of the foot, the initial anterior cluster of ciliary cells has moved more posteriorly and laterally. Axons from el and e4 in the medial pedal nerve have also moved with this expanding cluster of ciliary cells, while e6 and e7 axons in the superior pedal nerve turn more anteriorly (Fig. 8D).
[FIGURE 6 OMITTED]
The 5-HT-LIR plexus covers most of the pedal sole epithelium at 64% of hatching time. At this time the three lateral nerves can he identified at the level of the pedal ganglion (Fig. 8E). The e6 and e7 axons seen in the superior pedal nerve at 55% of hatching time can still be recognized by their branching pattern, but other immunoreactive axons are also present in this nerve. The e6 and e7 neurons develop at different times. Their axons leave the pedal ganglion separately, but after branching, they form a pair in close apposition (Fig. 8E, F), possibly because they are in a small nerve peripherally. At this time the el and e4 axons are still the only immunoreactive axons seen in the medial pedal nerve. These neurons develop at the same time, and their axons are in close apposition as they leave the pedal ganglion (Fig. 8G). In the inferior pedal nerve there is only one 5-HT-LIR axon present, apparently from e5 (Fig. 8E).
[FIGURE 7 OMITTED]
Pedal sole innervation by the ventromedian pedal nerve
In the adult Lymnaea, a layer of neuronal processes strongly immunoreactive to 5-HT antibodies lies directly over the pedal sole ciliary cells. At 73% of hatching time this 5-HT-LIR plexus has extended to lie over the posterior part of the foot (Fig. 9). At this time, axons from e2 and e.3 are major contributors to this plexus. When the foot is processed for immunoreactivity to 5-HT antibodies and cut in sagittal section, the pedal ganglion and a nerve that goes to this layer of neuronal processes can be seen (Fig. 10). By examining successive sections from the confocal micro scope, a third dimension can be seen that shows this nerve to be the ventromedian pedal nerve (ypn) that arises from the ventral pedal commissure. It still contains axons with weak (e2) and strong (e3) immunoreactivity as in earlier stages of development (Fig. 10A). This lateral projection of the right pedal ganglion near the commissure shows the dorsal commissure lying over the ventral commissure as in the adult nervous system. About 20 weakly immunoreactive neurons can be recognized near the commissure. Two neurons in this cluster near the commissure are strongly 5-HT-LIR (Fig. 10A), and a third neuron that lies more laterally in this cluster is also strongly immunoreactive (not shown).
In this sagittal view, the edge of the foot is folded against the cover slip so that its surface is not uniformly perpendicular to the field of view. This obscures the relation of the pedal sole ciliary cell layer to the fluorescent processes in the ventromedian pedal nerve (Fig. 10B). At higher magnification in a 1.0-imu]m-thick substack of this folded foot, the 5-HT-like immunoreactivity in these small processes from the ventromedian pedal nerve is shown to lie immediately over the pedal sole ciliary cells. Immunoreactivity in ciliary cells was not seen, but ciliary cells and cilia have sufficient contrast to be recognized (Fig. IOC).
The principal results of this study are that in Lymnaea stagnalis apressa (1) the initial appearance of pedal ganglion axons with 5-HT-like immunoreactivity is correlated with the appearance of adult ciliary cells on the pedal sole epithelium, (2) as these ciliary cells develop, processes from the first 5-HT-LIR pedal ganglion neuron that is seen turn to form a plexus immediately over these pedal sole ciliary cells, (3) axons from 5-HT-LIR pedal ganglion neurons spread over specific areas of the foot in the pedal plexus so that individual neurons are associated with specific areas of the ciliated pedal sole epithelium, and (4) two pairs of neurons seen in the embryo with axons in the pedal plexus can be identified in the adult nervous system by their axon projections in the ventromedian pedal nerve.
[FIGURE 8 OMITTED]
Location of identified neurons
Early in neurogenesis, 5-HT-LIR cerebral ganglion axons branch through both the dorsal and ventral commissures forming a closed loop uniting the left and right pedal ganglia (Fig. 11 A). Initially this loop tends to lie parallel to the foot surface (Marois and Croll, 1992). Later, beginning at about 60% of prehatch development in L. v. appressa as neurons are rapidly added to the ganglia, there is a rotation of the dorsal and ventral commissures so that the ventral commissure comes to lie under the dorsal commissure as in the adult (Fig. 10A). From this it can be inferred that neurons identified as lying ventral to the cerebral axons will be anterior to the commissures in the adult, and neurons dorsal to the cerebral axons will lie in the posterior part of the pedal ganglia.
[FIGURE 9 OMITTED]
[FIGURE 10 OMITTED]
The diagram of neuron locations and axon distributions shows e3 and e6 ventral to the cerebral ganglia axons, while the other neurons, except for e7, were found dorsal to these axons (Fig. 11 A). The location of e7 relative to the cerebral axons is uncertain. Of these neurons, only the e2 and e3 pairs can thus far be identified in the adult. With their axons in the ventromedian pedal nerve, they appear to be ho-mologs to the L. stagnalis PeVl and PeD7 neurons (Kyri-akides et al, 1989). PeD7 is anterior to the commissures, which is consistent with the location of e3 ventral to the cerebral axons; and the large PeVl neuron is in the posterior, ventral part of the pedal ganglia, which is consistent with the location of e2 above the cerebral axons. The other neurons identified here have not been related to adult neurons. Simple extrapolation by neuron location from the early embryo to the adult is difficult. Nevertheless, it may be possible to identify these neurons in the adult by their axon anatomy, immnunoreaetivity, and possible effect on ciliary beating.
Axons of these neurons in the putative pedal nerves (Fig. 11) are distributed to the pedal plexus. The innervation areas of these nerves has been described in L. stagnalis by recording sensory neuron responses of tactile stimuli to the foot (Janse, 1974). These results for sensory stimuli on the sole of the foot give areas that are larger than the areas covered by immunoreactive axons reported here; however, areas determined by morphological techniques as cited by Janse (1974) are consistent with results here. Axons from e6 and el (Fig. 11D) in the superior pedal nerve (spn) innervate the anterior part of the foot ipsilaterally; axons from el and e4 (Fig. 11C) in the medial pedal nerve (mpn) innervate the middle part of the foot ipsilaterally; and the axon from e5 (Fig. 1 IC) in the inferior pedal nerve (ipn) innervates the posterior part of the foot ipsilaterally. When recording sensory responses, Janse (1974) found no overlap of these nerves between the left and right side of the foot. The field of the ventromedian pedal nerve (vpn) in the foot has not previously been described. This small nerve, approximately 70 [micro]m in diameter in the adult, was thought to innervate the column instead of the foot. The axons of e2 and e3 (Fig. 11B) cover the posterior part of the foot and extend anteriorly at least to the area under the pedal ganglia. They overlap in the pedal plexus with the e5 axons and possibly with the el and e4 axons.
[FIGURE 11 OMITTED]
Intensity of immunoreactivity in pedal ganglion axons
Neither the neurotransmitter nor the mechanism involved in coupling neurons to gastropod pedal sole ciliary cells has been unequivocally determined. The plexus of neural processes over these cells in Tritonia diomedea, Pleurobran-chaea califomica, and L., stagnalis are suggested to beserotonergic on the basis of immunoreactivity to 5-HT antibodies (Moroz et ai, 1997; Syed et al., 1988; McKenzie et al, 1998) and also by the glyoxylic acid method in L. stagnalis (McKenzie et al., 1987). In T. diomedea, dopamine and the peptide TPep, in addition to 5-HT, have been shown to be possible neurotransmitters that increase the rate of pedal ciliary cell activity (Woodward and Willows, 2006). In the L. s. appressa embryo there is an immunore activity difference in 5-HT-LIR axons that innervate the pedal sole epithelial plexus. This difference is most apparent in the ventromedian pedal nerve, where the el axon has much weaker immunoreactivity than the e3 axon. In the e6 and e7 axons paired in the superior pedal nerve and the el and e4 axons paired in the medial pedal nerve, the e6 and e4 axons show weaker immunoreactivity. The unpaired e5 axon in the inferior pedal nerve also shows relatively weak immunoreactivity. These differences in immunoreactivity suggest that these neurons may have different functions or different combinations of neurotransmitters (Dyakonova et al., 1995), but these differences may also be caused by different rates of neurotransmitter synthesis.
Ciliary cell innervation
After metamorphosis, ciliary beating on the foot in L. s. appressa is first recognized by rotation of the embryo in the capsule fluid. At about 6 days of development (55% of hatching time), the embryo can adhere to the capsule with the anterior part of the foot and perform locomotion with pedal cilia on the inside of the capsule. In L. s. appressa, neither A-cluster neurons nor their peripheral axons are seen until about one day after the embryo begins ciliary locomotion while attached to the capsule. This behavior requires both control of ciliary beating and a particular type of mucous secretion. At 55% of hatching time the e6 and el axons, which may contribute to this behavior, are the only 5-HT-LIR axons seen in the superior pedal nerve that innervates the anterior part of the foot.
In L. stagnalis, A-cluster neurons are first seen at 75% of development (Marois and Croll, 1992). It has been suggested that A-cluster neurons are the ciliary motoneurons in L. stagnalis (Syed et al., 1988; Syed and Winlow, 1989), but the involvement of A-cluster neurons in control of pedal sole ciliary beating is not clear, since there are no results showing specific areas of innervation for these neurons or their effect on ciliary beating. Because they appear late in development, it is unlikely that A-cluster neurons control the ciliary activity at the time the embryo attaches to the capsule, although they may contribute to control of cilia later in development.
In the freshwater snail Planorbis corneas, Deliagina and Orlovsky (1990) found large pedal neurons that control ciliary beating on specific areas of the foot. A similar result has been obtained for a L. s. appressa pedal neuron with its axon in the inferior pedal nerve. This neuron, which was adjacent to the PeVl homolog in the adult ganglion, is located similarly to e5, which is adjacent to e2 in the embryo. In experiments with the isolated pedal ganglia and posterior part of the foot, it caused carbon grain movement on the posterior ipsilateral edge of the adult pedal sole in an area that is innervated by the e.5 axon (Longley and Peterman, 2000). Also in the adult L. s. appressa, an electrically coupled neuron pair (PeVl homologs), with the same axon anatomy as e2 in the vpn, causes carbon grain movement over the whole posterior part of the foot sole. Stimulation of the vpn with the nervous system removed also produces carbon grain movement on the posterior part of the foot similar to that caused by the PeVl homologs (R. Longley, unpubl. obs.). This suggests that other neurons in the el-e7 group may also affect pedal sole ciliary beating. Experiments in the adult ganglia with these 5-HT-LIR neurons that have axons in the pedal plexus can determine their role in the control of pedal cilia.
I thank the director of Friday Harbor Laboratories for use of the facilities during this work and the Center for Cell Dynamics for use of the confocal microscope. I thank Margaret Longley for assistance in editing.
Aiello, E., and G. Guideri. 1965. Distribution and function of the branchial nerve in the mussel. Biol. Bull. 129: 431-438.
Audesirk, G. 1978. Central neuronal control of cilia in Trilonia diame-dia. Nature 272: 541-543.
Audesirk, G., R. E. McCaman, and A. O. D. Willows. 1979. The role of serotonin in the control of pedal ciliary activity by identified neurons in Tritonia diomedea. Camp. Biochem. Physiol. 62C: 87-91.
Cumin, R. 1972. Normentafel zur Organogenese von Limnaeu stagnalis (Gastropoda, Pulmonata) mit besonderer Berucksichtigung der darmdruse. Rev. Suisse Zool. 79: 709-774.
Deliagina, T. G., and G. N. Orlovsky. 1990. Control of locomotion in the freshwater snail Planorbis corneas. II. Differential control of various zones of the ciliated epithelium. J. Exp. Biol. 152: 405-423.
Diefenbach, T. J., N. K. Koehncke, and J. I. Goldberg. 1991. Char acterization and development of rotational behavior in Helisoma embryos: role of endogenous serotonin. J. Neurobiol. 22: 922-934.
Diefenbach, T. J., R. Koss, and J. I. Goldberg. 1998. Early development of an identified serotonergic neuron in Helisoma trivolvis embryos: serotonin expression, de-expression, and uptake. J. Neurobiol. 34: 361-376.
Doran. S. A., R. Koss, C. H. Tran, K. J. Christopher, W. J. Gallin, and J. I. Goldberg. 2004. Effect of serotonin on ciliary beating and intracellular calcium concentration in identified populations of embryonic ciliary cells. J. Exp. Biol. 207: 1415-1429.
Dyakonova, V., M. Carlberg, D. Sakharov, and R. Elofsson. 1995. Anatomical basis for interactions of enkephalins with other transmitters in the CNS of a snail. J. Comp. Neurol. 361: 38-47.
Janse, C. 1974. A neurophysiological study of the peripheral tactile system of the pond snail Lynmaca stagnalis (L.). Neth. J. Zool. 24: 93-161.
Koshtoyants, Kh. S., G. A. Buznikov, and B. N. Manukhin. 1961. The possible role of 5-hydroxytryptamine in the motor activity of embryos of some marine gastropods. Comp. Biochem. Physiol. 3: 20 -26.
Koss, R., T. J. Diefenbach, S. Kuang, S. A. Doran, and J. I. Goldberg. 2003. Coordinated development of identified serotonergic neurons and their target ciliary cells in Helisoma trivolvis embryos. J. Comp. Neurol. 457: 313-325.
Kuang, S., and J. I. Goldberg. 2001. Laser ablation reveals regulation of ciliary activity by serotonergic neurons in molluscan embryos. J. Neurobiol. 47: 1-15.
Kyriakides, M., C. R. McCrohan, C. T. Slade, N. I. Syed, and W. Winlow. 1989. The morphology and electrophysiology of the neurones of the paired pedal ganglia of l.ymnaea stagnalis (L.) Comp. Biochem. Physiol. 93A: 861-876.
Longley, R. D., and A. J. Longley. 1986. Serotonin immunoreactivity of neurons in the gastropod Aplysia californica. J. Neurobiol. 17: 339-358.
Longley, R. D., and M. Peterman. 2000. Neural control of foot cilia in Lymnaea stagnalis appressa. Sac. Neuroxci. Abstr. 26: 985 (Abstract).
Mackie, G. O., C. L. Singla, and C. Thiriot-Quievreux. 1976. Nervous control of ciliary activity in gastropod larvae. Biol. Bull. 151: 182-199.
Marois, R., and R. P. Croll. 1992. Development of serotoninlike im munoreactivity in the embryonic nervous system of the snail Lymnaea stagnalis. J. Comp. Neurol. 322: 255-265.
McKenzie, J. D., N. Syed, J. Tripp, and W. Winlow. 1987. Are the pedal cilia in Lymnaea under neural control? Pp. 26-30 in Neurobiology: Molluscan Models, H. H. Boer, W. P. M. Geraerts, and J. Joose, eds. North Holland, Amsterdam.
McKenzie, J. D., M. Caunce, M. S. Hetherington, and W. Winlow. 1998. Serotonergic innervation of the foot of the pond snail Lymnaea stagnalis (L.) J. Neurocytol. 27: 459-470.
Moroz, L. L., L. C. Sudlow, J. Jing, and R. Gillette. 1997. Serotonin-immunoreactivity in peripheral tissues of the opisthobranch molluscs Pleurobranchaea californica and Tritonia diomedea. J. Comp. Neural. 382: 176-188.
Morrill, J. B. 1982. Development of the pulmonate gastropod Lymnaea. Pp. 309-483 in Developmemntal Biology of Freshwater Invertebrates, F. W. Harrison and R. R. Cowden, eds. Alan R. Liss, New York.
Nagy, T and K. Elekes. 2000. Embryogenesis of the central nervous system of the pond snail Lymnaea stagnalis L. An ultrastructural study. J. Neurocytol. 29: 43-60.
Slade, C. T., J. Mills, and W. Winlow. 1981. The neuronal organization of the paired pedal ganglia of Lymnnaea stagnalis (L.). Comp. Biochem. Physiol. 69A: 789-803.
Syed, N. I., and W. Winlow. 1989. Morphology and electrophysiology of neurons innervating the ciliated locomotor epithelium in Lymnaea stagnalis (L.). Comp. Biochem. Physiol. 93A: 633-644.
Syed, N. I., D. Harrison, and W. Winlow. 1988. Locomotion in Lymnaea--role of serotonergic motoneurones controlling the pedal cilia. Pp. 387-402 in Symposia Biolagica Hungarica, Vol. 36. J. Salanki and K. S.-Rozsa, eds. Akademiai Kiado, Budapest.
Uhler, G. C, P. T. Huminski, F. T. Les, and P. P. Pong. 2000. Cilia-driven rotational behavior in gastropod (Physc elliptica) in mbryos induced by serotonin and putative serotonin reuptake inhibitors (SSRIs). J, Exp. Zool. 286:414-421.
Voronezhskaya, E. E., and K. Elekes. 2003. Expression of FMRF-amide gene encoded peptides by identified neurons in embryos and juveniles of the pulmonate snail Lymnaea stagnalis 314: 297-313.
Woodward, O. M., and A. O. D. Willows. 2006. Dopamine modulation Of [Ca.sup.2+] dependent Cl~ current regulates ciliary beat freaquency controlling locomotion in Tritonia diomtdea. J. Exp. Biol 209: 2749-2764 2764.
Willows, A. O. D., G. A. Pavlova, and N. E. Phillips. 1997. Modulation of ciliary beat frequency by neuropeptides from identified molluscan neurons. J. Exp. Biol. 200: 1433-1439.
ROGER D. LONGLEY *
Friday Harbor Laboratories, 620 University Road, Friday Harbor, Washington 98250
Received 18 December 2007; accepted 30 April 2008.
* E-mail: email@example.com
Abbreviations: 5-HT-LIR, 5-HT-like immunoreactive; el-el, 5-HT-LIR neurons; ipn, inferior pedal nerve; mpn, medial pedal nerve; ppc, primordial pedal plate cilia; pc, prototrochal cilia: spn, superior pedal nerve; vpn, ventromedian pedal nerve.
|Printer friendly Cite/link Email Feedback|
|Author:||Longley, Roger D.|
|Publication:||The Biological Bulletin|
|Date:||Dec 1, 2008|
|Previous Article:||Upside-down gliding of Lymnaea.|
|Next Article:||Flow forces on seaweeds: field evidence for roles of wave impingement and organism inertia.|